Emacs Lisp

The GNU Emacs website is at https://www.gnu.org/software/emacs/.
For information on using Emacs, refer to the Emacs Manual.
To view this manual in other formats, click here.

This is the GNU Emacs Lisp Reference Manual corresponding to Emacs version 29.2.

Copyright © 1990–1996, 1998–2024 Free Software Foundation, Inc.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with the Invariant Sections being “GNU General Public License,” with the Front-Cover Texts being “A GNU Manual,” and with the Back-Cover Texts as in (a) below. A copy of the license is included in the section entitled “GNU Free Documentation License.”

(a) The FSF’s Back-Cover Text is: “You have the freedom to copy and modify this GNU manual. Buying copies from the FSF supports it in developing GNU and promoting software freedom.”

Table of Contents

Short Table of Contents


1 Introduction

Most of the GNU Emacs text editor is written in the programming language called Emacs Lisp. You can write new code in Emacs Lisp and install it as an extension to the editor. However, Emacs Lisp is more than a mere extension language; it is a full computer programming language in its own right. You can use it as you would any other programming language.

Because Emacs Lisp is designed for use in an editor, it has special features for scanning and parsing text as well as features for handling files, buffers, displays, subprocesses, and so on. Emacs Lisp is closely integrated with the editing facilities; thus, editing commands are functions that can also conveniently be called from Lisp programs, and parameters for customization are ordinary Lisp variables.

This manual attempts to be a full description of Emacs Lisp. For a beginner’s introduction to Emacs Lisp, see An Introduction to Emacs Lisp Programming, by Bob Chassell, also published by the Free Software Foundation. This manual presumes considerable familiarity with the use of Emacs for editing; see The GNU Emacs Manual for this basic information.

Generally speaking, the earlier chapters describe features of Emacs Lisp that have counterparts in many programming languages, and later chapters describe features that are peculiar to Emacs Lisp or relate specifically to editing.

This is the GNU Emacs Lisp Reference Manual, corresponding to Emacs version 29.2.


1.1 Caveats

This manual has gone through numerous drafts. It is nearly complete but not flawless. There are a few topics that are not covered, either because we consider them secondary (such as most of the individual modes) or because they are yet to be written. Because we are not able to deal with them completely, we have left out several parts intentionally.

The manual should be fully correct in what it does cover, and it is therefore open to criticism on anything it says—from specific examples and descriptive text, to the ordering of chapters and sections. If something is confusing, or you find that you have to look at the sources or experiment to learn something not covered in the manual, then perhaps the manual should be fixed. Please let us know.

As you use this manual, we ask that you send corrections as soon as you find them. If you think of a simple, real life example for a function or group of functions, please make an effort to write it up and send it in. Please reference any comments to the node name and function or variable name, as appropriate. Also state the number of the edition you are criticizing.

Please send comments and corrections using M-x report-emacs-bug. If you wish to contribute new code (or send a patch to fix a problem), use M-x submit-emacs-patch.


1.2 Lisp History

Lisp (LISt Processing language) was first developed in the late 1950s at the Massachusetts Institute of Technology for research in artificial intelligence. The great power of the Lisp language makes it ideal for other purposes as well, such as writing editing commands.

Dozens of Lisp implementations have been built over the years, each with its own idiosyncrasies. Many of them were inspired by Maclisp, which was written in the 1960s at MIT’s Project MAC. Eventually the implementers of the descendants of Maclisp came together and developed a standard for Lisp systems, called Common Lisp. In the meantime, Gerry Sussman and Guy Steele at MIT developed a simplified but very powerful dialect of Lisp, called Scheme.

GNU Emacs Lisp is largely inspired by Maclisp, and a little by Common Lisp. If you know Common Lisp, you will notice many similarities. However, many features of Common Lisp have been omitted or simplified in order to reduce the memory requirements of GNU Emacs. Sometimes the simplifications are so drastic that a Common Lisp user might be very confused. We will occasionally point out how GNU Emacs Lisp differs from Common Lisp. If you don’t know Common Lisp, don’t worry about it; this manual is self-contained.

A certain amount of Common Lisp emulation is available via the cl-lib library. See Overview in Common Lisp Extensions.

Emacs Lisp is not at all influenced by Scheme; but the GNU project has an implementation of Scheme, called Guile. We use it in all new GNU software that calls for extensibility.


1.3 Conventions

This section explains the notational conventions that are used in this manual. You may want to skip this section and refer back to it later.


1.3.1 Some Terms

Throughout this manual, the phrases “the Lisp reader” and “the Lisp printer” refer to those routines in Lisp that convert textual representations of Lisp objects into actual Lisp objects, and vice versa. See Printed Representation and Read Syntax, for more details. You, the person reading this manual, are thought of as the programmer and are addressed as “you”. The user is the person who uses Lisp programs, including those you write.

Examples of Lisp code are formatted like this: (list 1 2 3). Names that represent metasyntactic variables, or arguments to a function being described, are formatted like this: first-number.


1.3.2 nil and t

In Emacs Lisp, the symbol nil has three separate meanings: it is a symbol with the name ‘nil’; it is the logical truth value false; and it is the empty list—the list of zero elements. When used as a variable, nil always has the value nil.

As far as the Lisp reader is concerned, ‘()’ and ‘nil’ are identical: they stand for the same object, the symbol nil. The different ways of writing the symbol are intended entirely for human readers. After the Lisp reader has read either ‘()’ or ‘nil’, there is no way to determine which representation was actually written by the programmer.

In this manual, we write () when we wish to emphasize that it means the empty list, and we write nil when we wish to emphasize that it means the truth value false. That is a good convention to use in Lisp programs also.

(cons 'foo ())                ; Emphasize the empty list
(setq foo-flag nil)           ; Emphasize the truth value false

In contexts where a truth value is expected, any non-nil value is considered to be true. However, t is the preferred way to represent the truth value true. When you need to choose a value that represents true, and there is no other basis for choosing, use t. The symbol t always has the value t.

In Emacs Lisp, nil and t are special symbols that always evaluate to themselves. This is so that you do not need to quote them to use them as constants in a program. An attempt to change their values results in a setting-constant error. See Variables that Never Change.

Function: booleanp object

Return non-nil if object is one of the two canonical boolean values: t or nil.


1.3.3 Evaluation Notation

A Lisp expression that you can evaluate is called a form. Evaluating a form always produces a result, which is a Lisp object. In the examples in this manual, this is indicated with ‘’:

(car '(1 2))
     ⇒ 1

You can read this as “(car '(1 2)) evaluates to 1”.

When a form is a macro call, it expands into a new form for Lisp to evaluate. We show the result of the expansion with ‘’. We may or may not show the result of the evaluation of the expanded form.

(third '(a b c))
     → (car (cdr (cdr '(a b c))))
     ⇒ c

To help describe one form, we sometimes show another form that produces identical results. The exact equivalence of two forms is indicated with ‘’.

(make-sparse-keymap) ≡ (list 'keymap)

1.3.4 Printing Notation

Many of the examples in this manual print text when they are evaluated. If you execute example code in a Lisp Interaction buffer (such as the buffer *scratch*) by typing C-j after the closing parenthesis of the example, the printed text is inserted into the buffer. If you execute the example by other means (such as by evaluating the function eval-region), the printed text is displayed in the echo area.

Examples in this manual indicate printed text with ‘-|’, irrespective of where that text goes. The value returned by evaluating the form follows on a separate line with ‘’.

(progn (prin1 'foo) (princ "\n") (prin1 'bar))
     -| foo
     -| bar
     ⇒ bar

1.3.5 Error Messages

Some examples signal errors. This normally displays an error message in the echo area. We show the error message on a line starting with ‘error→’. Note that ‘error→’ itself does not appear in the echo area.

(+ 23 'x)
error→ Wrong type argument: number-or-marker-p, x

1.3.6 Buffer Text Notation

Some examples describe modifications to the contents of a buffer, by showing the before and after versions of the text. These examples show the contents of the buffer in question between two lines of dashes containing the buffer name. In addition, ‘’ indicates the location of point. (The symbol for point, of course, is not part of the text in the buffer; it indicates the place between two characters where point is currently located.)

---------- Buffer: foo ----------
This is the ∗contents of foo.
---------- Buffer: foo ----------

(insert "changed ")
     ⇒ nil
---------- Buffer: foo ----------
This is the changed ∗contents of foo.
---------- Buffer: foo ----------

1.3.7 Format of Descriptions

Functions, variables, macros, commands, user options, and special forms are described in this manual in a uniform format. The first line of a description contains the name of the item followed by its arguments, if any. The category—function, variable, or whatever—appears at the beginning of the line. The description follows on succeeding lines, sometimes with examples.


1.3.7.1 A Sample Function Description

In a function description, the name of the function being described appears first. It is followed on the same line by a list of argument names. These names are also used in the body of the description, to stand for the values of the arguments.

The appearance of the keyword &optional in the argument list indicates that the subsequent arguments may be omitted (omitted arguments default to nil). Do not write &optional when you call the function.

The keyword &rest (which must be followed by a single argument name) indicates that any number of arguments can follow. The single argument name following &rest receives, as its value, a list of all the remaining arguments passed to the function. Do not write &rest when you call the function.

Here is a description of an imaginary function foo:

Function: foo integer1 &optional integer2 &rest integers

The function foo subtracts integer1 from integer2, then adds all the rest of the arguments to the result. If integer2 is not supplied, then the number 19 is used by default.

(foo 1 5 3 9)
     ⇒ 16
(foo 5)
     ⇒ 14

More generally,

(foo w x y…)
≡
(+ (- x w) y…)

By convention, any argument whose name contains the name of a type (e.g., integer, integer1 or buffer) is expected to be of that type. A plural of a type (such as buffers) often means a list of objects of that type. An argument named object may be of any type. (For a list of Emacs object types, see Lisp Data Types.) An argument with any other sort of name (e.g., new-file) is specific to the function; if the function has a documentation string, the type of the argument should be described there (see Documentation).

See Lambda Expressions, for a more complete description of arguments modified by &optional and &rest.

Command, macro, and special form descriptions have the same format, but the word ‘Function’ is replaced by ‘Command’, ‘Macro’, or ‘Special Form’, respectively. Commands are simply functions that may be called interactively; macros process their arguments differently from functions (the arguments are not evaluated), but are presented the same way.

The descriptions of macros and special forms use a more complex notation to specify optional and repeated arguments, because they can break the argument list down into separate arguments in more complicated ways. ‘[optional-arg]’ means that optional-arg is optional and ‘repeated-args’ stands for zero or more arguments. Parentheses are used when several arguments are grouped into additional levels of list structure. Here is an example:

Special Form: count-loop (var [from to [inc]]) body…

This imaginary special form implements a loop that executes the body forms and then increments the variable var on each iteration. On the first iteration, the variable has the value from; on subsequent iterations, it is incremented by one (or by inc if that is given). The loop exits before executing body if var equals to. Here is an example:

(count-loop (i 0 10)
  (prin1 i) (princ " ")
  (prin1 (aref vector i))
  (terpri))

If from and to are omitted, var is bound to nil before the loop begins, and the loop exits if var is non-nil at the beginning of an iteration. Here is an example:

(count-loop (done)
  (if (pending)
      (fixit)
    (setq done t)))

In this special form, the arguments from and to are optional, but must both be present or both absent. If they are present, inc may optionally be specified as well. These arguments are grouped with the argument var into a list, to distinguish them from body, which includes all remaining elements of the form.


1.3.7.2 A Sample Variable Description

A variable is a name that can be bound (or set) to an object. The object to which a variable is bound is called a value; we say also that variable holds that value. Although nearly all variables can be set by the user, certain variables exist specifically so that users can change them; these are called user options. Ordinary variables and user options are described using a format like that for functions, except that there are no arguments.

Here is a description of the imaginary electric-future-map variable.

Variable: electric-future-map

The value of this variable is a full keymap used by Electric Command Future mode. The functions in this map allow you to edit commands you have not yet thought about executing.

User option descriptions have the same format, but ‘Variable’ is replaced by ‘User Option’.


1.4 Version Information

These facilities provide information about which version of Emacs is in use.

Command: emacs-version &optional here

This function returns a string describing the version of Emacs that is running. It is useful to include this string in bug reports.

(emacs-version)
  ⇒ "GNU Emacs 26.1 (build 1, x86_64-unknown-linux-gnu,
             GTK+ Version 3.16) of 2017-06-01"

If here is non-nil, it inserts the text in the buffer before point, and returns nil. When this function is called interactively, it prints the same information in the echo area, but giving a prefix argument makes here non-nil.

Variable: emacs-build-time

The value of this variable indicates the time at which Emacs was built. It uses the style of current-time (see Time of Day), or is nil if the information is not available.

emacs-build-time
     ⇒ (25194 55894 8547 617000)

(This timestamp is (1651169878008547617 . 1000000000) if current-time-list was nil when Emacs was built.)

Variable: emacs-version

The value of this variable is the version of Emacs being run. It is a string such as "26.1". A value with three numeric components, such as "26.0.91", indicates an unreleased test version. (Prior to Emacs 26.1, the string includes an extra final component with the integer that is now stored in emacs-build-number; e.g., "25.1.1".)

Variable: emacs-major-version

The major version number of Emacs, as an integer. For Emacs version 23.1, the value is 23.

Variable: emacs-minor-version

The minor version number of Emacs, as an integer. For Emacs version 23.1, the value is 1.

Variable: emacs-build-number

An integer that increments each time Emacs is built in the same directory (without cleaning). This is only of relevance when developing Emacs.

Variable: emacs-repository-version

A string that gives the repository revision from which Emacs was built. If Emacs was built outside revision control, the value is nil.

Variable: emacs-repository-branch

A string that gives the repository branch from which Emacs was built. In the most cases this is "master". If Emacs was built outside revision control, the value is nil.


1.5 Acknowledgments

This manual was originally written by Robert Krawitz, Bil Lewis, Dan LaLiberte, Richard M. Stallman and Chris Welty, the volunteers of the GNU manual group, in an effort extending over several years. Robert J. Chassell helped to review and edit the manual, with the support of the Defense Advanced Research Projects Agency, ARPA Order 6082, arranged by Warren A. Hunt, Jr. of Computational Logic, Inc. Additional sections have since been written by Miles Bader, Lars Brinkhoff, Chong Yidong, Kenichi Handa, Lute Kamstra, Juri Linkov, Glenn Morris, Thien-Thi Nguyen, Dan Nicolaescu, Martin Rudalics, Kim F. Storm, Luc Teirlinck, and Eli Zaretskii, and others.

Corrections were supplied by Drew Adams, Juanma Barranquero, Karl Berry, Jim Blandy, Bard Bloom, Stephane Boucher, David Boyes, Alan Carroll, Richard Davis, Lawrence R. Dodd, Peter Doornbosch, David A. Duff, Chris Eich, Beverly Erlebacher, David Eckelkamp, Ralf Fassel, Eirik Fuller, Stephen Gildea, Bob Glickstein, Eric Hanchrow, Jesper Harder, George Hartzell, Nathan Hess, Masayuki Ida, Dan Jacobson, Jak Kirman, Bob Knighten, Frederick M. Korz, Joe Lammens, Glenn M. Lewis, K. Richard Magill, Brian Marick, Roland McGrath, Stefan Monnier, Skip Montanaro, John Gardiner Myers, Thomas A. Peterson, Francesco Potortì, Friedrich Pukelsheim, Arnold D. Robbins, Raul Rockwell, Jason Rumney, Per Starbäck, Shinichirou Sugou, Kimmo Suominen, Edward Tharp, Bill Trost, Rickard Westman, Jean White, Eduard Wiebe, Matthew Wilding, Carl Witty, Dale Worley, Rusty Wright, and David D. Zuhn.

For a more complete list of contributors, please see the relevant change log entries in the Emacs source repository.


2 Lisp Data Types

A Lisp object is a piece of data used and manipulated by Lisp programs. For our purposes, a type or data type is a set of possible objects.

Every object belongs to at least one type. Objects of the same type have similar structures and may usually be used in the same contexts. Types can overlap, and objects can belong to two or more types. Consequently, we can ask whether an object belongs to a particular type, but not for the type of an object.

A few fundamental object types are built into Emacs. These, from which all other types are constructed, are called primitive types. Each object belongs to one and only one primitive type. These types include integer, float, cons, symbol, string, vector, hash-table, subr, byte-code function, and record, plus several special types, such as buffer, that are related to editing. (See Editing Types.)

Each primitive type has a corresponding Lisp function that checks whether an object is a member of that type.

Lisp is unlike many other languages in that its objects are self-typing: the primitive type of each object is implicit in the object itself. For example, if an object is a vector, nothing can treat it as a number; Lisp knows it is a vector, not a number.

In most languages, the programmer must declare the data type of each variable, and the type is known by the compiler but not represented in the data. Such type declarations do not exist in Emacs Lisp. A Lisp variable can have any type of value, and it remembers whatever value you store in it, type and all. (Actually, a small number of Emacs Lisp variables can only take on values of a certain type. See Variables with Restricted Values.)

This chapter describes the purpose, printed representation, and read syntax of each of the standard types in GNU Emacs Lisp. Details on how to use these types can be found in later chapters.


2.1 Printed Representation and Read Syntax

The printed representation of an object is the format of the output generated by the Lisp printer (the function prin1) for that object. Every data type has a unique printed representation. The read syntax of an object is the format of the input accepted by the Lisp reader (the function read) for that object. This is not necessarily unique; many kinds of object have more than one syntax. See Reading and Printing Lisp Objects.

In most cases, an object’s printed representation is also a read syntax for the object. However, some types have no read syntax, since it does not make sense to enter objects of these types as constants in a Lisp program. These objects are printed in hash notation, which consists of the characters ‘#<’, a descriptive string (typically the type name followed by the name of the object), and a closing ‘>’. (This is called “hash notation” because it begins with the ‘#’ character, known as “hash” or “number sign”). For example:

(current-buffer)
     ⇒ #<buffer objects.texi>

Hash notation cannot be read at all, so the Lisp reader signals the error invalid-read-syntax whenever it encounters ‘#<’.

We describe the read syntax and the printed representation of each Lisp data type where we describe that data type, in the following sections of this chapter. For example, see String Type, and its subsections for the read syntax and printed representation of strings; see Vector Type for the same information about vectors; etc.

In other languages, an expression is text; it has no other form. In Lisp, an expression is primarily a Lisp object and only secondarily the text that is the object’s read syntax. Often there is no need to emphasize this distinction, but you must keep it in the back of your mind, or you will occasionally be very confused.

When you evaluate an expression interactively, the Lisp interpreter first reads the textual representation of it, producing a Lisp object, and then evaluates that object (see Evaluation). However, evaluation and reading are separate activities. Reading returns the Lisp object represented by the text that is read; the object may or may not be evaluated later. See Input Functions, for a description of read, the basic function for reading objects.


2.2 Special Read Syntax

Emacs Lisp represents many special objects and constructs via special hash notations.

#<…>

Objects that have no read syntax are presented like this (see Printed Representation and Read Syntax).

##

The printed representation of an interned symbol whose name is an empty string (see Symbol Type).

#'

This is a shortcut for function, see Anonymous Functions.

#:

The printed representation of an uninterned symbol whose name is foo is ‘#:foo’ (see Symbol Type).

#N

When printing circular structures, this construct is used to represent where the structure loops back onto itself, and ‘N’ is the starting list count:

(let ((a (list 1)))
  (setcdr a a))
=> (1 . #0)
#N=
#N#

#N=’ gives the name to an object, and ‘#N#’ represents that object, so when reading back the object, they will be the same object instead of copies (see Read Syntax for Circular Objects).

#xN

N’ represented as a hexadecimal number (‘#x2a’).

#oN

N’ represented as an octal number (‘#o52’).

#bN

N’ represented as a binary number (‘#b101010’).

#(…)

String text properties (see Text Properties in Strings).

#^

A char table (see Char-Table Type).

#s(hash-table …)

A hash table (see Hash Table Type).

?C

A character (see Basic Char Syntax).

#$

The current file name in byte-compiled files (see Documentation Strings and Compilation). This is not meant to be used in Emacs Lisp source files.

#@N

Skip the next ‘N’ characters (see Comments). This is used in byte-compiled files, and is not meant to be used in Emacs Lisp source files.

#f

Indicates that the following form isn’t readable by the Emacs Lisp reader. This is only in text for display purposes (when that would look prettier than alternative ways of indicating an unreadable form) and will never appear in any Lisp file.


2.3 Comments

A comment is text that is written in a program only for the sake of humans that read the program, and that has no effect on the meaning of the program. In Lisp, an unescaped semicolon (‘;’) starts a comment if it is not within a string or character constant. The comment continues to the end of line. The Lisp reader discards comments; they do not become part of the Lisp objects which represent the program within the Lisp system.

The ‘#@count’ construct, which skips the next count characters, is useful for program-generated comments containing binary data. The Emacs Lisp byte compiler uses this in its output files (see Byte Compilation). It isn’t meant for source files, however.

See Tips on Writing Comments, for conventions for formatting comments.


2.4 Programming Types

There are two general categories of types in Emacs Lisp: those having to do with Lisp programming, and those having to do with editing. The former exist in many Lisp implementations, in one form or another. The latter are unique to Emacs Lisp.


2.4.1 Integer Type

Under the hood, there are two kinds of integers—small integers, called fixnums, and large integers, called bignums.

The range of values for a fixnum depends on the machine. The minimum range is −536,870,912 to 536,870,911 (30 bits; i.e., −2**29 to 2**29 − 1) but many machines provide a wider range.

Bignums can have arbitrary precision. Operations that overflow a fixnum will return a bignum instead.

All numbers can be compared with eql or =; fixnums can also be compared with eq. To test whether an integer is a fixnum or a bignum, you can compare it to most-negative-fixnum and most-positive-fixnum, or you can use the convenience predicates fixnump and bignump on any object.

The read syntax for integers is a sequence of (base ten) digits with an optional sign at the beginning and an optional period at the end. The printed representation produced by the Lisp interpreter never has a leading ‘+’ or a final ‘.’.

-1               ; The integer −1.
1                ; The integer 1.
1.               ; Also the integer 1.
+1               ; Also the integer 1.

See Numbers, for more information.


2.4.2 Floating-Point Type

Floating-point numbers are the computer equivalent of scientific notation; you can think of a floating-point number as a fraction together with a power of ten. The precise number of significant figures and the range of possible exponents is machine-specific; Emacs uses the C data type double to store the value, and internally this records a power of 2 rather than a power of 10.

The printed representation for floating-point numbers requires either a decimal point (with at least one digit following), an exponent, or both. For example, ‘1500.0’, ‘+15e2’, ‘15.0e+2’, ‘+1500000e-3’, and ‘.15e4’ are five ways of writing a floating-point number whose value is 1500. They are all equivalent.

See Numbers, for more information.


2.4.3 Character Type

A character in Emacs Lisp is nothing more than an integer. In other words, characters are represented by their character codes. For example, the character A is represented as the integer 65. That is also their usual printed representation; see Basic Char Syntax.

Individual characters are used occasionally in programs, but it is more common to work with strings, which are sequences composed of characters. See String Type.

Characters in strings and buffers are currently limited to the range of 0 to 4194303—twenty two bits (see Character Codes). Codes 0 through 127 are ASCII codes; the rest are non-ASCII (see Non-ASCII Characters). Characters that represent keyboard input have a much wider range, to encode modifier keys such as Control, Meta and Shift.

There are special functions for producing a human-readable textual description of a character for the sake of messages. See Describing Characters for Help Messages.


2.4.3.1 Basic Char Syntax

Since characters are really integers, the printed representation of a character is a decimal number. This is also a possible read syntax for a character, but writing characters that way in Lisp programs is not clear programming. You should always use the special read syntax formats that Emacs Lisp provides for characters. These syntax formats start with a question mark.

The usual read syntax for alphanumeric characters is a question mark followed by the character; thus, ‘?A’ for the character A, ‘?B’ for the character B, and ‘?a’ for the character a.

For example:

?Q ⇒ 81     ?q ⇒ 113

You can use the same syntax for punctuation characters. However, if the punctuation character has a special syntactic meaning in Lisp, you must quote it with a ‘\’. For example, ‘?\(’ is the way to write the open-paren character. Likewise, if the character is ‘\’, you must use a second ‘\’ to quote it: ‘?\\’.

You can express the characters control-g, backspace, tab, newline, vertical tab, formfeed, space, return, del, and escape as ‘?\a’, ‘?\b’, ‘?\t’, ‘?\n’, ‘?\v’, ‘?\f’, ‘?\s’, ‘?\r’, ‘?\d’, and ‘?\e’, respectively. (‘?\s’ followed by a dash has a different meaning—it applies the Super modifier to the following character.) Thus,

?\a ⇒ 7                 ; control-g, C-g
?\b ⇒ 8                 ; backspace, BS, C-h
?\t ⇒ 9                 ; tab, TAB, C-i
?\n ⇒ 10                ; newline, C-j
?\v ⇒ 11                ; vertical tab, C-k
?\f ⇒ 12                ; formfeed character, C-l
?\r ⇒ 13                ; carriage return, RET, C-m
?\e ⇒ 27                ; escape character, ESC, C-[
?\s ⇒ 32                ; space character, SPC
?\\ ⇒ 92                ; backslash character, \
?\d ⇒ 127               ; delete character, DEL

These sequences which start with backslash are also known as escape sequences, because backslash plays the role of an escape character; this has nothing to do with the character ESC. ‘\s’ is meant for use in character constants; in string constants, just write the space.

A backslash is allowed, and harmless, preceding any character without a special escape meaning; thus, ‘?\+’ is equivalent to ‘?+’. There is no reason to add a backslash before most characters. However, you must add a backslash before any of the characters ‘()[]\;"’, and you should add a backslash before any of the characters ‘|'`#.,’ to avoid confusing the Emacs commands for editing Lisp code. You should also add a backslash before Unicode characters which resemble the previously mentioned ASCII ones, to avoid confusing people reading your code. Emacs will highlight some non-escaped commonly confused characters such as ‘’ to encourage this. You can also add a backslash before whitespace characters such as space, tab, newline and formfeed. However, it is cleaner to use one of the easily readable escape sequences, such as ‘\t’ or ‘\s’, instead of an actual whitespace character such as a tab or a space. (If you do write backslash followed by a space, you should write an extra space after the character constant to separate it from the following text.)


2.4.3.2 General Escape Syntax

In addition to the specific escape sequences for special important control characters, Emacs provides several types of escape syntax that you can use to specify non-ASCII text characters.

  1. You can specify characters by their Unicode names, if any. ?\N{NAME} represents the Unicode character named NAME. Thus, ‘?\N{LATIN SMALL LETTER A WITH GRAVE}’ is equivalent to and denotes the Unicode character U+00E0. To simplify entering multi-line strings, you can replace spaces in the names by non-empty sequences of whitespace (e.g., newlines).
  2. You can specify characters by their Unicode values. ?\N{U+X} represents a character with Unicode code point X, where X is a hexadecimal number. Also, ?\uxxxx and ?\Uxxxxxxxx represent code points xxxx and xxxxxxxx, respectively, where each x is a single hexadecimal digit. For example, ?\N{U+E0}, ?\u00e0 and ?\U000000E0 are all equivalent to and to ‘?\N{LATIN SMALL LETTER A WITH GRAVE}’. The Unicode Standard defines code points only up to ‘U+10ffff’, so if you specify a code point higher than that, Emacs signals an error.
  3. You can specify characters by their hexadecimal character codes. A hexadecimal escape sequence consists of a backslash, ‘x’, and the hexadecimal character code. Thus, ‘?\x41’ is the character A, ‘?\x1’ is the character C-a, and ?\xe0 is the character à (a with grave accent). You can use one or more hex digits after ‘x’, so you can represent any character code in this way.
  4. You can specify characters by their character code in octal. An octal escape sequence consists of a backslash followed by up to three octal digits; thus, ‘?\101’ for the character A, ‘?\001’ for the character C-a, and ?\002 for the character C-b. Only characters up to octal code 777 can be specified this way.

These escape sequences may also be used in strings. See Non-ASCII Characters in Strings.


2.4.3.3 Control-Character Syntax

Control characters can be represented using yet another read syntax. This consists of a question mark followed by a backslash, caret, and the corresponding non-control character, in either upper or lower case. For example, both ‘?\^I’ and ‘?\^i’ are valid read syntax for the character C-i, the character whose value is 9.

Instead of the ‘^’, you can use ‘C-’; thus, ‘?\C-i’ is equivalent to ‘?\^I’ and to ‘?\^i’:

?\^I ⇒ 9     ?\C-I ⇒ 9

In strings and buffers, the only control characters allowed are those that exist in ASCII; but for keyboard input purposes, you can turn any character into a control character with ‘C-’. The character codes for these non-ASCII control characters include the 2**26 bit as well as the code for the corresponding non-control character. Not all text terminals can generate non-ASCII control characters, but it is straightforward to generate them using X and other window systems.

For historical reasons, Emacs treats the DEL character as the control equivalent of ?:

?\^? ⇒ 127     ?\C-? ⇒ 127

As a result, it is currently not possible to represent the character Control-?, which is a meaningful input character under X, using ‘\C-’. It is not easy to change this, as various Lisp files refer to DEL in this way.

For representing control characters to be found in files or strings, we recommend the ‘^’ syntax; for control characters in keyboard input, we prefer the ‘C-’ syntax. Which one you use does not affect the meaning of the program, but may guide the understanding of people who read it.


2.4.3.4 Meta-Character Syntax

A meta character is a character typed with the META modifier key. The integer that represents such a character has the 2**27 bit set. We use high bits for this and other modifiers to make possible a wide range of basic character codes.

In a string, the 2**7 bit attached to an ASCII character indicates a meta character; thus, the meta characters that can fit in a string have codes in the range from 128 to 255, and are the meta versions of the ordinary ASCII characters. See Putting Keyboard Events in Strings, for details about META-handling in strings.

The read syntax for meta characters uses ‘\M-’. For example, ‘?\M-A’ stands for M-A. You can use ‘\M-’ together with octal character codes (see below), with ‘\C-’, or with any other syntax for a character. Thus, you can write M-A as ‘?\M-A’, or as ‘?\M-\101’. Likewise, you can write C-M-b as ‘?\M-\C-b’, ‘?\C-\M-b’, or ‘?\M-\002’.


2.4.3.5 Other Character Modifier Bits

The case of a graphic character is indicated by its character code; for example, ASCII distinguishes between the characters ‘a’ and ‘A’. But ASCII has no way to represent whether a control character is upper case or lower case. Emacs uses the 2**25 bit to indicate that the shift key was used in typing a control character. This distinction is possible only on a graphical display such as a GUI display on X; text terminals do not report the distinction. The Lisp syntax for the shift bit is ‘\S-’; thus, ‘?\C-\S-o’ or ‘?\C-\S-O’ represents the shifted-control-o character.

The X Window System defines three other modifier bits that can be set in a character: hyper, super and alt. The syntaxes for these bits are ‘\H-’, ‘\s-’ and ‘\A-’. (Case is significant in these prefixes.) Thus, ‘?\H-\M-\A-x’ represents Alt-Hyper-Meta-x. (Note that ‘\s’ with no following ‘-’ represents the space character.) Numerically, the bit values are 2**22 for alt, 2**23 for super and 2**24 for hyper.


2.4.4 Symbol Type

A symbol in GNU Emacs Lisp is an object with a name. The symbol name serves as the printed representation of the symbol. In ordinary Lisp use, with one single obarray (see Creating and Interning Symbols), a symbol’s name is unique—no two symbols have the same name.

A symbol can serve as a variable, as a function name, or to hold a property list. Or it may serve only to be distinct from all other Lisp objects, so that its presence in a data structure may be recognized reliably. In a given context, usually only one of these uses is intended. But you can use one symbol in all of these ways, independently.

A symbol whose name starts with a colon (‘:’) is called a keyword symbol. These symbols automatically act as constants, and are normally used only by comparing an unknown symbol with a few specific alternatives. See Variables that Never Change.

A symbol name can contain any characters whatever. Most symbol names are written with letters, digits, and the punctuation characters ‘-+=*/’. Such names require no special punctuation; the characters of the name suffice as long as the name does not look like a number. (If it does, write a ‘\’ at the beginning of the name to force interpretation as a symbol.) The characters ‘_~!@$%^&:<>{}?’ are less often used but also require no special punctuation. Any other characters may be included in a symbol’s name by escaping them with a backslash. In contrast to its use in strings, however, a backslash in the name of a symbol simply quotes the single character that follows the backslash. For example, in a string, ‘\t’ represents a tab character; in the name of a symbol, however, ‘\t’ merely quotes the letter ‘t’. To have a symbol with a tab character in its name, you must actually use a tab (preceded with a backslash). But it’s rare to do such a thing.

Common Lisp note: In Common Lisp, lower case letters are always folded to upper case, unless they are explicitly escaped. In Emacs Lisp, upper case and lower case letters are distinct.

Here are several examples of symbol names. Note that the ‘+’ in the fourth example is escaped to prevent it from being read as a number. This is not necessary in the sixth example because the rest of the name makes it invalid as a number.

foo                 ; A symbol named ‘foo’.
FOO                 ; A symbol named ‘FOO’, different from ‘foo’.
1+                  ; A symbol named ‘1+
                    ;   (not ‘+1’, which is an integer).
\+1                 ; A symbol named ‘+1
                    ;   (not a very readable name).
\(*\ 1\ 2\)         ; A symbol named ‘(* 1 2)’ (a worse name).
+-*/_~!@$%^&=:<>{}  ; A symbol named ‘+-*/_~!@$%^&=:<>{}’.
                    ;   These characters need not be escaped.

As an exception to the rule that a symbol’s name serves as its printed representation, ‘##’ is the printed representation for an interned symbol whose name is an empty string. Furthermore, ‘#:foo’ is the printed representation for an uninterned symbol whose name is foo. (Normally, the Lisp reader interns all symbols; see Creating and Interning Symbols.)


2.4.5 Sequence Types

A sequence is a Lisp object that represents an ordered set of elements. There are two kinds of sequence in Emacs Lisp: lists and arrays.

Lists are the most commonly-used sequences. A list can hold elements of any type, and its length can be easily changed by adding or removing elements. See the next subsection for more about lists.

Arrays are fixed-length sequences. They are further subdivided into strings, vectors, char-tables and bool-vectors. Vectors can hold elements of any type, whereas string elements must be characters, and bool-vector elements must be t or nil. Char-tables are like vectors except that they are indexed by any valid character code. The characters in a string can have text properties like characters in a buffer (see Text Properties), but vectors do not support text properties, even when their elements happen to be characters.

Lists, strings and the other array types also share important similarities. For example, all have a length l, and all have elements which can be indexed from zero to l minus one. Several functions, called sequence functions, accept any kind of sequence. For example, the function length reports the length of any kind of sequence. See Sequences, Arrays, and Vectors.

It is generally impossible to read the same sequence twice, since sequences are always created anew upon reading. If you read the read syntax for a sequence twice, you get two sequences with equal contents. There is one exception: the empty list () always stands for the same object, nil.


2.4.6 Cons Cell and List Types

A cons cell is an object that consists of two slots, called the CAR slot and the CDR slot. Each slot can hold any Lisp object. We also say that the CAR of this cons cell is whatever object its CAR slot currently holds, and likewise for the CDR.

A list is a series of cons cells, linked together so that the CDR slot of each cons cell holds either the next cons cell or the empty list. The empty list is actually the symbol nil. See Lists, for details. Because most cons cells are used as part of lists, we refer to any structure made out of cons cells as a list structure.

A note to C programmers: a Lisp list thus works as a linked list built up of cons cells. Because pointers in Lisp are implicit, we do not distinguish between a cons cell slot holding a value versus pointing to the value.

Because cons cells are so central to Lisp, we also have a word for an object which is not a cons cell. These objects are called atoms.

The read syntax and printed representation for lists are identical, and consist of a left parenthesis, an arbitrary number of elements, and a right parenthesis. Here are examples of lists:

(A 2 "A")            ; A list of three elements.
()                   ; A list of no elements (the empty list).
nil                  ; A list of no elements (the empty list).
("A ()")             ; A list of one element: the string "A ()".
(A ())               ; A list of two elements: A and the empty list.
(A nil)              ; Equivalent to the previous.
((A B C))            ; A list of one element
                     ;   (which is a list of three elements).

Upon reading, each object inside the parentheses becomes an element of the list. That is, a cons cell is made for each element. The CAR slot of the cons cell holds the element, and its CDR slot refers to the next cons cell of the list, which holds the next element in the list. The CDR slot of the last cons cell is set to hold nil.

The names CAR and CDR derive from the history of Lisp. The original Lisp implementation ran on an IBM 704 computer which divided words into two parts, the address and the decrement; CAR was an instruction to extract the contents of the address part of a register, and CDR an instruction to extract the contents of the decrement. By contrast, cons cells are named for the function cons that creates them, which in turn was named for its purpose, the construction of cells.


2.4.6.1 Drawing Lists as Box Diagrams

A list can be illustrated by a diagram in which the cons cells are shown as pairs of boxes, like dominoes. (The Lisp reader cannot read such an illustration; unlike the textual notation, which can be understood by both humans and computers, the box illustrations can be understood only by humans.) This picture represents the three-element list (rose violet buttercup):

    --- ---      --- ---      --- ---
   |   |   |--> |   |   |--> |   |   |--> nil
    --- ---      --- ---      --- ---
     |            |            |
     |            |            |
      --> rose     --> violet   --> buttercup

In this diagram, each box represents a slot that can hold or refer to any Lisp object. Each pair of boxes represents a cons cell. Each arrow represents a reference to a Lisp object, either an atom or another cons cell.

In this example, the first box, which holds the CAR of the first cons cell, refers to or holds rose (a symbol). The second box, holding the CDR of the first cons cell, refers to the next pair of boxes, the second cons cell. The CAR of the second cons cell is violet, and its CDR is the third cons cell. The CDR of the third (and last) cons cell is nil.

Here is another diagram of the same list, (rose violet buttercup), sketched in a different manner:

 ---------------       ----------------       -------------------
| car   | cdr   |     | car    | cdr   |     | car       | cdr   |
| rose  |   o-------->| violet |   o-------->| buttercup |  nil  |
|       |       |     |        |       |     |           |       |
 ---------------       ----------------       -------------------

A list with no elements in it is the empty list; it is identical to the symbol nil. In other words, nil is both a symbol and a list.

Here is the list (A ()), or equivalently (A nil), depicted with boxes and arrows:

    --- ---      --- ---
   |   |   |--> |   |   |--> nil
    --- ---      --- ---
     |            |
     |            |
      --> A        --> nil

Here is a more complex illustration, showing the three-element list, ((pine needles) oak maple), the first element of which is a two-element list:

    --- ---      --- ---      --- ---
   |   |   |--> |   |   |--> |   |   |--> nil
    --- ---      --- ---      --- ---
     |            |            |
     |            |            |
     |             --> oak      --> maple
     |
     |     --- ---      --- ---
      --> |   |   |--> |   |   |--> nil
           --- ---      --- ---
            |            |
            |            |
             --> pine     --> needles

The same list represented in the second box notation looks like this:

 --------------       --------------       --------------
| car   | cdr  |     | car   | cdr  |     | car   | cdr  |
|   o   |   o------->| oak   |   o------->| maple |  nil |
|   |   |      |     |       |      |     |       |      |
 -- | ---------       --------------       --------------
    |
    |
    |        --------------       ----------------
    |       | car   | cdr  |     | car     | cdr  |
     ------>| pine  |   o------->| needles |  nil |
            |       |      |     |         |      |
             --------------       ----------------

2.4.6.2 Dotted Pair Notation

Dotted pair notation is a general syntax for cons cells that represents the CAR and CDR explicitly. In this syntax, (a . b) stands for a cons cell whose CAR is the object a and whose CDR is the object b. Dotted pair notation is more general than list syntax because the CDR does not have to be a list. However, it is more cumbersome in cases where list syntax would work. In dotted pair notation, the list ‘(1 2 3)’ is written as ‘(1 . (2 . (3 . nil)))’. For nil-terminated lists, you can use either notation, but list notation is usually clearer and more convenient. When printing a list, the dotted pair notation is only used if the CDR of a cons cell is not a list.

Here’s an example using boxes to illustrate dotted pair notation. This example shows the pair (rose . violet):

    --- ---
   |   |   |--> violet
    --- ---
     |
     |
      --> rose

You can combine dotted pair notation with list notation to represent conveniently a chain of cons cells with a non-nil final CDR. You write a dot after the last element of the list, followed by the CDR of the final cons cell. For example, (rose violet . buttercup) is equivalent to (rose . (violet . buttercup)). The object looks like this:

    --- ---      --- ---
   |   |   |--> |   |   |--> buttercup
    --- ---      --- ---
     |            |
     |            |
      --> rose     --> violet

The syntax (rose . violet . buttercup) is invalid because there is nothing that it could mean. If anything, it would say to put buttercup in the CDR of a cons cell whose CDR is already used for violet.

The list (rose violet) is equivalent to (rose . (violet)), and looks like this:

    --- ---      --- ---
   |   |   |--> |   |   |--> nil
    --- ---      --- ---
     |            |
     |            |
      --> rose     --> violet

Similarly, the three-element list (rose violet buttercup) is equivalent to (rose . (violet . (buttercup))). It looks like this:

    --- ---      --- ---      --- ---
   |   |   |--> |   |   |--> |   |   |--> nil
    --- ---      --- ---      --- ---
     |            |            |
     |            |            |
      --> rose     --> violet   --> buttercup

2.4.6.3 Association List Type

An association list or alist is a specially-constructed list whose elements are cons cells. In each element, the CAR is considered a key, and the CDR is considered an associated value. (In some cases, the associated value is stored in the CAR of the CDR.) Association lists are often used as stacks, since it is easy to add or remove associations at the front of the list.

For example,

(setq alist-of-colors
      '((rose . red) (lily . white) (buttercup . yellow)))

sets the variable alist-of-colors to an alist of three elements. In the first element, rose is the key and red is the value.

See Association Lists, for a further explanation of alists and for functions that work on alists. See Hash Tables, for another kind of lookup table, which is much faster for handling a large number of keys.


2.4.7 Array Type

An array is composed of an arbitrary number of slots for holding or referring to other Lisp objects, arranged in a contiguous block of memory. Accessing any element of an array takes approximately the same amount of time. In contrast, accessing an element of a list requires time proportional to the position of the element in the list. (Elements at the end of a list take longer to access than elements at the beginning of a list.)

Emacs defines four types of array: strings, vectors, bool-vectors, and char-tables.

A string is an array of characters and a vector is an array of arbitrary objects. A bool-vector can hold only t or nil. These kinds of array may have any length up to the largest fixnum, subject to system architecture limits and available memory. Char-tables are sparse arrays indexed by any valid character code; they can hold arbitrary objects.

The first element of an array has index zero, the second element has index 1, and so on. This is called zero-origin indexing. For example, an array of four elements has indices 0, 1, 2, and 3. The largest possible index value is one less than the length of the array. Once an array is created, its length is fixed.

All Emacs Lisp arrays are one-dimensional. (Most other programming languages support multidimensional arrays, but they are not essential; you can get the same effect with nested one-dimensional arrays.) Each type of array has its own read syntax; see the following sections for details.

The array type is a subset of the sequence type, and contains the string type, the vector type, the bool-vector type, and the char-table type.


2.4.8 String Type

A string is an array of characters. Strings are used for many purposes in Emacs, as can be expected in a text editor; for example, as the names of Lisp symbols, as messages for the user, and to represent text extracted from buffers. Strings in Lisp are constants: evaluation of a string returns the same string.

See Strings and Characters, for functions that operate on strings.


2.4.8.1 Syntax for Strings

The read syntax for a string is a double-quote, an arbitrary number of characters, and another double-quote, "like this". To include a double-quote in a string, precede it with a backslash; thus, "\"" is a string containing just one double-quote character. Likewise, you can include a backslash by preceding it with another backslash, like this: "this \\ is a single embedded backslash".

Since a string is an array of characters, you can specify the string characters using the read syntax of characters, but without the leading question mark. This is useful for including in string constants characters that don’t stand for themselves. Thus, control characters can be specified as escape sequences that start with a backslash; for example, "foo\r" yields ‘foo’ followed by the carriage return character. See Basic Char Syntax, for escape sequences of other control characters. Similarly, you can use the special read syntax for control characters (see Control-Character Syntax), as in "foo\^Ibar", which produces a tab character embedded within a string. You can also use the escape sequences for non-ASCII characters described in General Escape Syntax, as in "\N{LATIN SMALL LETTER A WITH GRAVE}" and "\u00e0" (however, see a caveat with non-ASCII characters in Non-ASCII Characters in Strings).

The newline character is not special in the read syntax for strings; if you write a new line between the double-quotes, it becomes a character in the string. But an escaped newline—one that is preceded by ‘\’—does not become part of the string; i.e., the Lisp reader ignores an escaped newline while reading a string. An escaped space ‘ is likewise ignored.

"It is useful to include newlines
in documentation strings,
but the newline is \
ignored if escaped."
     ⇒ "It is useful to include newlines
in documentation strings,
but the newline is ignored if escaped."

2.4.8.2 Non-ASCII Characters in Strings

There are two text representations for non-ASCII characters in Emacs strings: multibyte and unibyte (see Text Representations). Roughly speaking, unibyte strings store raw bytes, while multibyte strings store human-readable text. Each character in a unibyte string is a byte, i.e., its value is between 0 and 255. By contrast, each character in a multibyte string may have a value between 0 to 4194303 (see Character Type). In both cases, characters above 127 are non-ASCII.

You can include a non-ASCII character in a string constant by writing it literally. If the string constant is read from a multibyte source, such as a multibyte buffer or string, or a file that would be visited as multibyte, then Emacs reads each non-ASCII character as a multibyte character and automatically makes the string a multibyte string. If the string constant is read from a unibyte source, then Emacs reads the non-ASCII character as unibyte, and makes the string unibyte.

Instead of writing a character literally into a multibyte string, you can write it as its character code using an escape sequence. See General Escape Syntax, for details about escape sequences.

If you use any Unicode-style escape sequence ‘\uNNNN’ or ‘\U00NNNNNN’ in a string constant (even for an ASCII character), Emacs automatically assumes that it is multibyte.

You can also use hexadecimal escape sequences (‘\xn’) and octal escape sequences (‘\n’) in string constants. But beware: If a string constant contains hexadecimal or octal escape sequences, and these escape sequences all specify unibyte characters (i.e., less than 256), and there are no other literal non-ASCII characters or Unicode-style escape sequences in the string, then Emacs automatically assumes that it is a unibyte string. That is to say, it assumes that all non-ASCII characters occurring in the string are 8-bit raw bytes.

In hexadecimal and octal escape sequences, the escaped character code may contain a variable number of digits, so the first subsequent character which is not a valid hexadecimal or octal digit terminates the escape sequence. If the next character in a string could be interpreted as a hexadecimal or octal digit, write ‘ (backslash and space) to terminate the escape sequence. For example, ‘\xe0\  represents one character, ‘a’ with grave accent. ‘ in a string constant is just like backslash-newline; it does not contribute any character to the string, but it does terminate any preceding hex escape.


2.4.8.3 Nonprinting Characters in Strings

You can use the same backslash escape-sequences in a string constant as in character literals (but do not use the question mark that begins a character constant). For example, you can write a string containing the nonprinting characters tab and C-a, with commas and spaces between them, like this: "\t, \C-a". See Character Type, and its subsections for a description of the various kinds of read syntax for characters.

However, not all of the characters you can write with backslash escape-sequences are valid in strings. The only control characters that a string can hold are the ASCII control characters. Strings do not distinguish case in ASCII control characters.

Properly speaking, strings cannot hold meta characters; but when a string is to be used as a key sequence, there is a special convention that provides a way to represent meta versions of ASCII characters in a string. If you use the ‘\M-’ syntax to indicate a meta character in a string constant, this sets the 2**7 bit of the character in the string. If the string is used in define-key or lookup-key, this numeric code is translated into the equivalent meta character. See Character Type.

Strings cannot hold characters that have the hyper, super, or alt modifiers.


2.4.8.4 Text Properties in Strings

A string can hold properties for the characters it contains, in addition to the characters themselves. This enables programs that copy text between strings and buffers to copy the text’s properties with no special effort. See Text Properties, for an explanation of what text properties mean. Strings with text properties use a special read and print syntax:

#("characters" property-data...)

where property-data consists of zero or more elements, in groups of three as follows:

beg end plist

The elements beg and end are integers, and together specify a range of indices in the string; plist is the property list for that range. For example,

#("foo bar" 0 3 (face bold) 3 4 nil 4 7 (face italic))

represents a string whose textual contents are ‘foo bar’, in which the first three characters have a face property with value bold, and the last three have a face property with value italic. (The fourth character has no text properties, so its property list is nil. It is not actually necessary to mention ranges with nil as the property list, since any characters not mentioned in any range will default to having no properties.)


2.4.9 Vector Type

A vector is a one-dimensional array of elements of any type. It takes a constant amount of time to access any element of a vector. (In a list, the access time of an element is proportional to the distance of the element from the beginning of the list.)

The printed representation of a vector consists of a left square bracket, the elements, and a right square bracket. This is also the read syntax. Like numbers and strings, vectors are considered constants for evaluation.

[1 "two" (three)]      ; A vector of three elements.
     ⇒ [1 "two" (three)]

See Vectors, for functions that work with vectors.


2.4.10 Char-Table Type

A char-table is a one-dimensional array of elements of any type, indexed by character codes. Char-tables have certain extra features to make them more useful for many jobs that involve assigning information to character codes—for example, a char-table can have a parent to inherit from, a default value, and a small number of extra slots to use for special purposes. A char-table can also specify a single value for a whole character set.

The printed representation of a char-table is like a vector except that there is an extra ‘#^’ at the beginning.1

See Char-Tables, for special functions to operate on char-tables. Uses of char-tables include:


2.4.11 Bool-Vector Type

A bool-vector is a one-dimensional array whose elements must be t or nil.

The printed representation of a bool-vector is like a string, except that it begins with ‘#&’ followed by the length. The string constant that follows actually specifies the contents of the bool-vector as a bitmap—each character in the string contains 8 bits, which specify the next 8 elements of the bool-vector (1 stands for t, and 0 for nil). The least significant bits of the character correspond to the lowest indices in the bool-vector.

(make-bool-vector 3 t)
     ⇒ #&3"^G"
(make-bool-vector 3 nil)
     ⇒ #&3"^@"

These results make sense, because the binary code for ‘C-g’ is 111 and ‘C-@’ is the character with code 0.

If the length is not a multiple of 8, the printed representation shows extra elements, but these extras really make no difference. For instance, in the next example, the two bool-vectors are equal, because only the first 3 bits are used:

(equal #&3"\377" #&3"\007")
     ⇒ t

2.4.12 Hash Table Type

A hash table is a very fast kind of lookup table, somewhat like an alist in that it maps keys to corresponding values, but much faster. The printed representation of a hash table specifies its properties and contents, like this:

(make-hash-table)
     ⇒ #s(hash-table size 65 test eql rehash-size 1.5
                             rehash-threshold 0.8125 data ())

See Hash Tables, for more information about hash tables.


2.4.13 Function Type

Lisp functions are executable code, just like functions in other programming languages. In Lisp, unlike most languages, functions are also Lisp objects. A non-compiled function in Lisp is a lambda expression: that is, a list whose first element is the symbol lambda (see Lambda Expressions).

In most programming languages, it is impossible to have a function without a name. In Lisp, a function has no intrinsic name. A lambda expression can be called as a function even though it has no name; to emphasize this, we also call it an anonymous function (see Anonymous Functions). A named function in Lisp is just a symbol with a valid function in its function cell (see Defining Functions).

Most of the time, functions are called when their names are written in Lisp expressions in Lisp programs. However, you can construct or obtain a function object at run time and then call it with the primitive functions funcall and apply. See Calling Functions.


2.4.14 Macro Type

A Lisp macro is a user-defined construct that extends the Lisp language. It is represented as an object much like a function, but with different argument-passing semantics. A Lisp macro has the form of a list whose first element is the symbol macro and whose CDR is a Lisp function object, including the lambda symbol.

Lisp macro objects are usually defined with the built-in defmacro macro, but any list that begins with macro is a macro as far as Emacs is concerned. See Macros, for an explanation of how to write a macro.

Warning: Lisp macros and keyboard macros (see Keyboard Macros) are entirely different things. When we use the word “macro” without qualification, we mean a Lisp macro, not a keyboard macro.


2.4.15 Primitive Function Type

A primitive function is a function callable from Lisp but written in the C programming language. Primitive functions are also called subrs or built-in functions. (The word “subr” is derived from “subroutine”.) Most primitive functions evaluate all their arguments when they are called. A primitive function that does not evaluate all its arguments is called a special form (see Special Forms).

It does not matter to the caller of a function whether the function is primitive. However, this does matter if you try to redefine a primitive with a function written in Lisp. The reason is that the primitive function may be called directly from C code. Calls to the redefined function from Lisp will use the new definition, but calls from C code may still use the built-in definition. Therefore, we discourage redefinition of primitive functions.

The term function refers to all Emacs functions, whether written in Lisp or C. See Function Type, for information about the functions written in Lisp.

Primitive functions have no read syntax and print in hash notation with the name of the subroutine.

(symbol-function 'car)          ; Access the function cell
                                ;   of the symbol.
     ⇒ #<subr car>
(subrp (symbol-function 'car))  ; Is this a primitive function?
     ⇒ t                       ; Yes.

2.4.16 Byte-Code Function Type

Byte-code function objects are produced by byte-compiling Lisp code (see Byte Compilation). Internally, a byte-code function object is much like a vector; however, the evaluator handles this data type specially when it appears in a function call. See Byte-Code Function Objects.

The printed representation and read syntax for a byte-code function object is like that for a vector, with an additional ‘#’ before the opening ‘[’.


2.4.17 Record Type

A record is much like a vector. However, the first element is used to hold its type as returned by type-of. The purpose of records is to allow programmers to create objects with new types that are not built into Emacs.

See Records, for functions that work with records.


2.4.18 Type Descriptors

A type descriptor is a record which holds information about a type. Slot 1 in the record must be a symbol naming the type, and type-of relies on this to return the type of record objects. No other type descriptor slot is used by Emacs; they are free for use by Lisp extensions.

An example of a type descriptor is any instance of cl-structure-class.


2.4.19 Autoload Type

An autoload object is a list whose first element is the symbol autoload. It is stored as the function definition of a symbol, where it serves as a placeholder for the real definition. The autoload object says that the real definition is found in a file of Lisp code that should be loaded when necessary. It contains the name of the file, plus some other information about the real definition.

After the file has been loaded, the symbol should have a new function definition that is not an autoload object. The new definition is then called as if it had been there to begin with. From the user’s point of view, the function call works as expected, using the function definition in the loaded file.

An autoload object is usually created with the function autoload, which stores the object in the function cell of a symbol. See Autoload, for more details.


2.4.20 Finalizer Type

A finalizer object helps Lisp code clean up after objects that are no longer needed. A finalizer holds a Lisp function object. When a finalizer object becomes unreachable after a garbage collection pass, Emacs calls the finalizer’s associated function object. When deciding whether a finalizer is reachable, Emacs does not count references from finalizer objects themselves, allowing you to use finalizers without having to worry about accidentally capturing references to finalized objects themselves.

Errors in finalizers are printed to *Messages*. Emacs runs a given finalizer object’s associated function exactly once, even if that function fails.

Function: make-finalizer function

Make a finalizer that will run function. function will be called after garbage collection when the returned finalizer object becomes unreachable. If the finalizer object is reachable only through references from finalizer objects, it does not count as reachable for the purpose of deciding whether to run function. function will be run once per finalizer object.


2.5 Editing Types

The types in the previous section are used for general programming purposes, and most of them are common to most Lisp dialects. Emacs Lisp provides several additional data types for purposes connected with editing.


2.5.1 Buffer Type

A buffer is an object that holds text that can be edited (see Buffers). Most buffers hold the contents of a disk file (see Files) so they can be edited, but some are used for other purposes. Most buffers are also meant to be seen by the user, and therefore displayed, at some time, in a window (see Windows). But a buffer need not be displayed in any window. Each buffer has a designated position called point (see Positions); most editing commands act on the contents of the current buffer in the neighborhood of point. At any time, one buffer is the current buffer.

The contents of a buffer are much like a string, but buffers are not used like strings in Emacs Lisp, and the available operations are different. For example, you can insert text efficiently into an existing buffer, altering the buffer’s contents, whereas inserting text into a string requires concatenating substrings, and the result is an entirely new string object.

Many of the standard Emacs functions manipulate or test the characters in the current buffer; a whole chapter in this manual is devoted to describing these functions (see Text).

Several other data structures are associated with each buffer:

The local keymap and variable list contain entries that individually override global bindings or values. These are used to customize the behavior of programs in different buffers, without actually changing the programs.

A buffer may be indirect, which means it shares the text of another buffer, but presents it differently. See Indirect Buffers.

Buffers have no read syntax. They print in hash notation, showing the buffer name.

(current-buffer)
     ⇒ #<buffer objects.texi>

2.5.2 Marker Type

A marker denotes a position in a specific buffer. Markers therefore have two components: one for the buffer, and one for the position. Changes in the buffer’s text automatically relocate the position value as necessary to ensure that the marker always points between the same two characters in the buffer.

Markers have no read syntax. They print in hash notation, giving the current character position and the name of the buffer.

(point-marker)
     ⇒ #<marker at 10779 in objects.texi>

See Markers, for information on how to test, create, copy, and move markers.


2.5.3 Window Type

A window describes the portion of the screen that Emacs uses to display buffers. Every live window (see Basic Concepts of Emacs Windows) has one associated buffer, whose contents appear in that window. By contrast, a given buffer may appear in one window, no window, or several windows. Windows are grouped on the screen into frames; each window belongs to one and only one frame. See Frame Type.

Though many windows may exist simultaneously, at any time one window is designated the selected window (see Selecting Windows). This is the window where the cursor is (usually) displayed when Emacs is ready for a command. The selected window usually displays the current buffer (see The Current Buffer), but this is not necessarily the case.

Windows have no read syntax. They print in hash notation, giving the window number and the name of the buffer being displayed. The window numbers exist to identify windows uniquely, since the buffer displayed in any given window can change frequently.

(selected-window)
     ⇒ #<window 1 on objects.texi>

See Windows, for a description of the functions that work on windows.


2.5.4 Frame Type

A frame is a screen area that contains one or more Emacs windows; we also use the term “frame” to refer to the Lisp object that Emacs uses to refer to the screen area.

Frames have no read syntax. They print in hash notation, giving the frame’s title, plus its address in core (useful to identify the frame uniquely).

(selected-frame)
     ⇒ #<frame emacs@psilocin.gnu.org 0xdac80>

See Frames, for a description of the functions that work on frames.


2.5.5 Terminal Type

A terminal is a device capable of displaying one or more Emacs frames (see Frame Type).

Terminals have no read syntax. They print in hash notation giving the terminal’s ordinal number and its TTY device file name.

(get-device-terminal nil)
     ⇒ #<terminal 1 on /dev/tty>

2.5.6 Window Configuration Type

A window configuration stores information about the positions, sizes, and contents of the windows in a frame, so you can recreate the same arrangement of windows later.

Window configurations do not have a read syntax; their print syntax looks like ‘#<window-configuration>’. See Window Configurations, for a description of several functions related to window configurations.


2.5.7 Frame Configuration Type

A frame configuration stores information about the positions, sizes, and contents of the windows in all frames. It is not a primitive type—it is actually a list whose CAR is frame-configuration and whose CDR is an alist. Each alist element describes one frame, which appears as the CAR of that element.

See Frame Configurations, for a description of several functions related to frame configurations.


2.5.8 Process Type

The word process usually means a running program. Emacs itself runs in a process of this sort. However, in Emacs Lisp, a process is a Lisp object that designates a subprocess created by the Emacs process. Programs such as shells, GDB, ftp, and compilers, running in subprocesses of Emacs, extend the capabilities of Emacs. An Emacs subprocess takes textual input from Emacs and returns textual output to Emacs for further manipulation. Emacs can also send signals to the subprocess.

Process objects have no read syntax. They print in hash notation, giving the name of the process:

(process-list)
     ⇒ (#<process shell>)

See Processes, for information about functions that create, delete, return information about, send input or signals to, and receive output from processes.


2.5.9 Thread Type

A thread in Emacs represents a separate thread of Emacs Lisp execution. It runs its own Lisp program, has its own current buffer, and can have subprocesses locked to it, i.e. subprocesses whose output only this thread can accept. See Threads.

Thread objects have no read syntax. They print in hash notation, giving the name of the thread (if it has been given a name) or its address in core:

(all-threads)
    ⇒ (#<thread 0176fc40>)

2.5.10 Mutex Type

A mutex is an exclusive lock that threads can own and disown, in order to synchronize between them. See Mutexes.

Mutex objects have no read syntax. They print in hash notation, giving the name of the mutex (if it has been given a name) or its address in core:

(make-mutex "my-mutex")
    ⇒ #<mutex my-mutex>
(make-mutex)
    ⇒ #<mutex 01c7e4e0>

2.5.11 Condition Variable Type

A condition variable is a device for a more complex thread synchronization than the one supported by a mutex. A thread can wait on a condition variable, to be woken up when some other thread notifies the condition.

Condition variable objects have no read syntax. They print in hash notation, giving the name of the condition variable (if it has been given a name) or its address in core:

(make-condition-variable (make-mutex))
    ⇒ #<condvar 01c45ae8>

2.5.12 Stream Type

A stream is an object that can be used as a source or sink for characters—either to supply characters for input or to accept them as output. Many different types can be used this way: markers, buffers, strings, and functions. Most often, input streams (character sources) obtain characters from the keyboard, a buffer, or a file, and output streams (character sinks) send characters to a buffer, such as a *Help* buffer, or to the echo area.

The object nil, in addition to its other meanings, may be used as a stream. It stands for the value of the variable standard-input or standard-output. Also, the object t as a stream specifies input using the minibuffer (see Minibuffers) or output in the echo area (see The Echo Area).

Streams have no special printed representation or read syntax, and print as whatever primitive type they are.

See Reading and Printing Lisp Objects, for a description of functions related to streams, including parsing and printing functions.


2.5.13 Keymap Type

A keymap maps keys typed by the user to commands. This mapping controls how the user’s command input is executed. A keymap is actually a list whose CAR is the symbol keymap.

See Keymaps, for information about creating keymaps, handling prefix keys, local as well as global keymaps, and changing key bindings.


2.5.14 Overlay Type

An overlay specifies properties that apply to a part of a buffer. Each overlay applies to a specified range of the buffer, and contains a property list (a list whose elements are alternating property names and values). Overlay properties are used to present parts of the buffer temporarily in a different display style. Overlays have no read syntax, and print in hash notation, giving the buffer name and range of positions.

See Overlays, for information on how you can create and use overlays.


2.5.15 Font Type

A font specifies how to display text on a graphical terminal. There are actually three separate font types—font objects, font specs, and font entities—each of which has slightly different properties. None of them have a read syntax; their print syntax looks like ‘#<font-object>’, ‘#<font-spec>’, and ‘#<font-entity>’ respectively. See Low-Level Font Representation, for a description of these Lisp objects.


2.5.16 Xwidget Type

An xwidget is a special display element, such as a web browser, that can be embedded inside a buffer. Each window that displays an xwidget will also have an xwidget view, which on X-Windows corresponds to a single X window used to display the widget.

Neither of these objects are readable; their print syntaxes look like ‘#<xwidget>’ and ‘#<xwidget-view>’, respectively. See Embedded Native Widgets, for a more detailed description of xwidgets.


2.6 Read Syntax for Circular Objects

To represent shared or circular structures within a complex of Lisp objects, you can use the reader constructs ‘#n=’ and ‘#n#’.

Use #n= before an object to label it for later reference; subsequently, you can use #n# to refer the same object in another place. Here, n is some integer. For example, here is how to make a list in which the first element recurs as the third element:

(#1=(a) b #1#)

This differs from ordinary syntax such as this

((a) b (a))

which would result in a list whose first and third elements look alike but are not the same Lisp object. This shows the difference:

(prog1 nil
  (setq x '(#1=(a) b #1#)))
(eq (nth 0 x) (nth 2 x))
     ⇒ t
(setq x '((a) b (a)))
(eq (nth 0 x) (nth 2 x))
     ⇒ nil

You can also use the same syntax to make a circular structure, which appears as an element within itself. Here is an example:

#1=(a #1#)

This makes a list whose second element is the list itself. Here’s how you can see that it really works:

(prog1 nil
  (setq x '#1=(a #1#)))
(eq x (cadr x))
     ⇒ t

The Lisp printer can produce this syntax to record circular and shared structure in a Lisp object, if you bind the variable print-circle to a non-nil value. See Variables Affecting Output.


2.7 Type Predicates

The Emacs Lisp interpreter itself does not perform type checking on the actual arguments passed to functions when they are called. It could not do so, since function arguments in Lisp do not have declared data types, as they do in other programming languages. It is therefore up to the individual function to test whether each actual argument belongs to a type that the function can use.

All built-in functions do check the types of their actual arguments when appropriate, and signal a wrong-type-argument error if an argument is of the wrong type. For example, here is what happens if you pass an argument to + that it cannot handle:

(+ 2 'a)
     error→ Wrong type argument: number-or-marker-p, a

If you want your program to handle different types differently, you must do explicit type checking. The most common way to check the type of an object is to call a type predicate function. Emacs has a type predicate for each type, as well as some predicates for combinations of types.

A type predicate function takes one argument; it returns t if the argument belongs to the appropriate type, and nil otherwise. Following a general Lisp convention for predicate functions, most type predicates’ names end with ‘p’.

Here is an example which uses the predicates listp to check for a list and symbolp to check for a symbol.

(defun add-on (x)
  (cond ((symbolp x)
         ;; If X is a symbol, put it on LIST.
         (setq list (cons x list)))
        ((listp x)
         ;; If X is a list, add its elements to LIST.
         (setq list (append x list)))
        (t
         ;; We handle only symbols and lists.
         (error "Invalid argument %s in add-on" x))))

Here is a table of predefined type predicates, in alphabetical order, with references to further information.

atom

See atom.

arrayp

See arrayp.

bignump

See floatp.

bool-vector-p

See bool-vector-p.

booleanp

See booleanp.

bufferp

See bufferp.

byte-code-function-p

See byte-code-function-p.

compiled-function-p

See compiled-function-p.

case-table-p

See case-table-p.

char-or-string-p

See char-or-string-p.

char-table-p

See char-table-p.

commandp

See commandp.

condition-variable-p

See condition-variable-p.

consp

See consp.

custom-variable-p

See custom-variable-p.

fixnump

See floatp.

floatp

See floatp.

fontp

See Low-Level Font Representation.

frame-configuration-p

See frame-configuration-p.

frame-live-p

See frame-live-p.

framep

See framep.

functionp

See functionp.

hash-table-p

See hash-table-p.

integer-or-marker-p

See integer-or-marker-p.

integerp

See integerp.

keymapp

See keymapp.

keywordp

See Variables that Never Change.

listp

See listp.

markerp

See markerp.

mutexp

See mutexp.

nlistp

See nlistp.

number-or-marker-p

See number-or-marker-p.

numberp

See numberp.

overlayp

See overlayp.

processp

See processp.

recordp

See recordp.

sequencep

See sequencep.

string-or-null-p

See string-or-null-p.

stringp

See stringp.

subrp

See subrp.

symbolp

See symbolp.

syntax-table-p

See syntax-table-p.

threadp

See threadp.

vectorp

See vectorp.

wholenump

See wholenump.

window-configuration-p

See window-configuration-p.

window-live-p

See window-live-p.

windowp

See windowp.

The most general way to check the type of an object is to call the function type-of. Recall that each object belongs to one and only one primitive type; type-of tells you which one (see Lisp Data Types). But type-of knows nothing about non-primitive types. In most cases, it is more convenient to use type predicates than type-of.

Function: type-of object

This function returns a symbol naming the primitive type of object. The value is one of the symbols bool-vector, buffer, char-table, compiled-function, condition-variable, cons, finalizer, float, font-entity, font-object, font-spec, frame, hash-table, integer, marker, mutex, overlay, process, string, subr, symbol, thread, vector, window, or window-configuration. However, if object is a record, the type specified by its first slot is returned; Records.

(type-of 1)
     ⇒ integer
(type-of 'nil)
     ⇒ symbol
(type-of '())    ; () is nil.
     ⇒ symbol
(type-of '(x))
     ⇒ cons
(type-of (record 'foo))
     ⇒ foo

2.8 Equality Predicates

Here we describe functions that test for equality between two objects. Other functions test equality of contents between objects of specific types, e.g., strings. For these predicates, see the appropriate chapter describing the data type.

Function: eq object1 object2

This function returns t if object1 and object2 are the same object, and nil otherwise.

If object1 and object2 are symbols with the same name, they are normally the same object—but see Creating and Interning Symbols for exceptions. For other non-numeric types (e.g., lists, vectors, strings), two arguments with the same contents or elements are not necessarily eq to each other: they are eq only if they are the same object, meaning that a change in the contents of one will be reflected by the same change in the contents of the other.

If object1 and object2 are numbers with differing types or values, then they cannot be the same object and eq returns nil. If they are fixnums with the same value, then they are the same object and eq returns t. If they were computed separately but happen to have the same value and the same non-fixnum numeric type, then they might or might not be the same object, and eq returns t or nil depending on whether the Lisp interpreter created one object or two.

(eq 'foo 'foo)
     ⇒ t

(eq ?A ?A)
     ⇒ t

(eq 3.0 3.0)
     ⇒ t or nil
;; Equal floats may or may not be the same object.

(eq (make-string 3 ?A) (make-string 3 ?A))
     ⇒ nil

(eq "asdf" "asdf")
     ⇒ t or nil
;; Equal string constants or may not be the same object.

(eq '(1 (2 (3))) '(1 (2 (3))))
     ⇒ nil

(setq foo '(1 (2 (3))))
     ⇒ (1 (2 (3)))
(eq foo foo)
     ⇒ t
(eq foo '(1 (2 (3))))
     ⇒ nil

(eq [(1 2) 3] [(1 2) 3])
     ⇒ nil

(eq (point-marker) (point-marker))
     ⇒ nil

The make-symbol function returns an uninterned symbol, distinct from the symbol that is used if you write the name in a Lisp expression. Distinct symbols with the same name are not eq. See Creating and Interning Symbols.

(eq (make-symbol "foo") 'foo)
     ⇒ nil

The Emacs Lisp byte compiler may collapse identical literal objects, such as literal strings, into references to the same object, with the effect that the byte-compiled code will compare such objects as eq, while the interpreted version of the same code will not. Therefore, your code should never rely on objects with the same literal contents being either eq or not eq, it should instead use functions that compare object contents such as equal, described below. Similarly, your code should not modify literal objects (e.g., put text properties on literal strings), since doing that might affect other literal objects of the same contents, if the byte compiler collapses them.

Function: equal object1 object2

This function returns t if object1 and object2 have equal components, and nil otherwise. Whereas eq tests if its arguments are the same object, equal looks inside nonidentical arguments to see if their elements or contents are the same. So, if two objects are eq, they are equal, but the converse is not always true.

(equal 'foo 'foo)
     ⇒ t

(equal 456 456)
     ⇒ t

(equal "asdf" "asdf")
     ⇒ t
(eq "asdf" "asdf")
     ⇒ nil

(equal '(1 (2 (3))) '(1 (2 (3))))
     ⇒ t
(eq '(1 (2 (3))) '(1 (2 (3))))
     ⇒ nil

(equal [(1 2) 3] [(1 2) 3])
     ⇒ t
(eq [(1 2) 3] [(1 2) 3])
     ⇒ nil

(equal (point-marker) (point-marker))
     ⇒ t

(eq (point-marker) (point-marker))
     ⇒ nil

Comparison of strings is case-sensitive, but does not take account of text properties—it compares only the characters in the strings. See Text Properties. Use equal-including-properties to also compare text properties. For technical reasons, a unibyte string and a multibyte string are equal if and only if they contain the same sequence of character codes and all these codes are in the range 0 through 127 (ASCII).

(equal "asdf" "ASDF")
     ⇒ nil

The equal function recursively compares the contents of objects if they are integers, strings, markers, vectors, bool-vectors, byte-code function objects, char-tables, records, or font objects. Other objects are considered equal only if they are eq. For example, two distinct buffers are never considered equal, even if their textual contents are the same.

For equal, equality is defined recursively; for example, given two cons cells x and y, (equal x y) returns t if and only if both the expressions below return t:

(equal (car x) (car y))
(equal (cdr x) (cdr y))

Comparing circular lists may therefore cause deep recursion that leads to an error, and this may result in counterintuitive behavior such as (equal a b) returning t whereas (equal b a) signals an error.

Function: equal-including-properties object1 object2

This function behaves like equal in all cases but also requires that for two strings to be equal, they have the same text properties.

(equal "asdf" (propertize "asdf" 'asdf t))
     ⇒ t
(equal-including-properties "asdf"
                            (propertize "asdf" 'asdf t))
     ⇒ nil

2.9 Mutability

Some Lisp objects should never change. For example, the Lisp expression "aaa" yields a string, but you should not change its contents. And some objects cannot be changed; for example, although you can create a new number by calculating one, Lisp provides no operation to change the value of an existing number.

Other Lisp objects are mutable: it is safe to change their values via destructive operations involving side effects. For example, an existing marker can be changed by moving the marker to point to somewhere else.

Although numbers never change and all markers are mutable, some types have members some of which are mutable and others not. These types include conses, vectors, and strings. For example, although "cons" and (symbol-name 'cons) both yield strings that should not be changed, (copy-sequence "cons") and (make-string 3 ?a) both yield mutable strings that can be changed via later calls to aset.

A mutable object stops being mutable if it is part of an expression that is evaluated. For example:

(let* ((x (list 0.5))
       (y (eval (list 'quote x))))
  (setcar x 1.5) ;; The program should not do this.
  y)

Although the list (0.5) was mutable when it was created, it should not have been changed via setcar because it was given to eval. The reverse does not occur: an object that should not be changed never becomes mutable afterwards.

If a program attempts to change objects that should not be changed, the resulting behavior is undefined: the Lisp interpreter might signal an error, or it might crash or behave unpredictably in other ways.2

When similar constants occur as parts of a program, the Lisp interpreter might save time or space by reusing existing constants or their components. For example, (eq "abc" "abc") returns t if the interpreter creates only one instance of the string literal "abc", and returns nil if it creates two instances. Lisp programs should be written so that they work regardless of whether this optimization is in use.


3 Numbers

GNU Emacs supports two numeric data types: integers and floating-point numbers. Integers are whole numbers such as −3, 0, 7, 13, and 511. Floating-point numbers are numbers with fractional parts, such as −4.5, 0.0, and 2.71828. They can also be expressed in exponential notation: ‘1.5e2’ is the same as ‘150.0’; here, ‘e2’ stands for ten to the second power, and that is multiplied by 1.5. Integer computations are exact. Floating-point computations often involve rounding errors, as the numbers have a fixed amount of precision.


3.1 Integer Basics

The Lisp reader reads an integer as a nonempty sequence of decimal digits with optional initial sign and optional final period.

 1               ; The integer 1.
 1.              ; The integer 1.
+1               ; Also the integer 1.
-1               ; The integer −1.
 0               ; The integer 0.
-0               ; The integer 0.

The syntax for integers in bases other than 10 consists of ‘#’ followed by a radix indication followed by one or more digits. The radix indications are ‘b’ for binary, ‘o’ for octal, ‘x’ for hex, and ‘radixr’ for radix radix. Thus, ‘#binteger’ reads integer in binary, and ‘#radixrinteger’ reads integer in radix radix. Allowed values of radix run from 2 to 36, and allowed digits are the first radix characters taken from ‘0’–‘9’, ‘A’–‘Z’. Letter case is ignored and there is no initial sign or final period. For example:

#b101100 ⇒ 44
#o54 ⇒ 44
#x2c ⇒ 44
#24r1k ⇒ 44

To understand how various functions work on integers, especially the bitwise operators (see Bitwise Operations on Integers), it is often helpful to view the numbers in their binary form.

In binary, the decimal integer 5 looks like this:

…000101

(The ellipsis ‘’ stands for a conceptually infinite number of bits that match the leading bit; here, an infinite number of 0 bits. Later examples also use this ‘’ notation.)

The integer −1 looks like this:

…111111

−1 is represented as all ones. (This is called two’s complement notation.)

Subtracting 4 from −1 returns the negative integer −5. In binary, the decimal integer 4 is 100. Consequently, −5 looks like this:

…111011

Many of the functions described in this chapter accept markers for arguments in place of numbers. (See Markers.) Since the actual arguments to such functions may be either numbers or markers, we often give these arguments the name number-or-marker. When the argument value is a marker, its position value is used and its buffer is ignored.

In Emacs Lisp, text characters are represented by integers. Any integer between zero and the value of (max-char), inclusive, is considered to be valid as a character. See Character Codes.

Integers in Emacs Lisp are not limited to the machine word size. Under the hood, though, there are two kinds of integers: smaller ones, called fixnums, and larger ones, called bignums. Although Emacs Lisp code ordinarily should not depend on whether an integer is a fixnum or a bignum, older Emacs versions support only fixnums, some functions in Emacs still accept only fixnums, and older Emacs Lisp code may have trouble when given bignums. For example, while older Emacs Lisp code could safely compare integers for numeric equality with eq, the presence of bignums means that equality predicates like eql and = should now be used to compare integers.

The range of values for bignums is limited by the amount of main memory, by machine characteristics such as the size of the word used to represent a bignum’s exponent, and by the integer-width variable. These limits are typically much more generous than the limits for fixnums. A bignum is never numerically equal to a fixnum; Emacs always represents an integer in fixnum range as a fixnum, not a bignum.

The range of values for a fixnum depends on the machine. The minimum range is −536,870,912 to 536,870,911 (30 bits; i.e., −2**29 to 2**29 − 1), but many machines provide a wider range.

Variable: most-positive-fixnum

The value of this variable is the greatest “small” integer that Emacs Lisp can handle. Typical values are 2**29 − 1 on 32-bit and 2**61 − 1 on 64-bit platforms.

Variable: most-negative-fixnum

The value of this variable is the numerically least “small” integer that Emacs Lisp can handle. It is negative. Typical values are −2**29 on 32-bit and −2**61 on 64-bit platforms.

Variable: integer-width

The value of this variable is a nonnegative integer that controls whether Emacs signals a range error when a large integer would be calculated. Integers with absolute values less than 2**n, where n is this variable’s value, do not signal a range error. Attempts to create larger integers typically signal a range error, although there might be no signal if a larger integer can be created cheaply. Setting this variable to a large number can be costly if a computation creates huge integers.


3.2 Floating-Point Basics

Floating-point numbers are useful for representing numbers that are not integral. The range of floating-point numbers is the same as the range of the C data type double on the machine you are using. On almost all computers supported by Emacs, this is IEEE binary64 floating point format, which is standardized by IEEE Std 754-2019 and is discussed further in David Goldberg’s paper “What Every Computer Scientist Should Know About Floating-Point Arithmetic”. On modern platforms, floating-point operations follow the IEEE-754 standard closely; however, results are not always rounded correctly on some systems, notably 32-bit x86.

On some old computer systems, Emacs may not use IEEE floating-point. We know of one such system on which Emacs runs correctly, but does not follow IEEE-754: the VAX running NetBSD using GCC 10.4.0, where the VAX ‘D_Floating’ format is used instead. IBM System/370-derived mainframes and their XL/C compiler are also capable of utilizing a hexadecimal floating point format, but Emacs has not yet been built in such a configuration.

The read syntax for floating-point numbers requires either a decimal point, an exponent, or both. Optional signs (‘+’ or ‘-’) precede the number and its exponent. For example, ‘1500.0’, ‘+15e2’, ‘15.0e+2’, ‘+1500000e-3’, and ‘.15e4’ are five ways of writing a floating-point number whose value is 1500. They are all equivalent. Like Common Lisp, Emacs Lisp requires at least one digit after a decimal point in a floating-point number that does not have an exponent; ‘1500.’ is an integer, not a floating-point number.

Emacs Lisp treats -0.0 as numerically equal to ordinary zero with respect to numeric comparisons like =. This follows the IEEE floating-point standard, which says -0.0 and 0.0 are numerically equal even though other operations can distinguish them.

The IEEE floating-point standard supports positive infinity and negative infinity as floating-point values. It also provides for a class of values called NaN, or “not a number”; numerical functions return such values in cases where there is no correct answer. For example, (/ 0.0 0.0) returns a NaN. A NaN is never numerically equal to any value, not even to itself. NaNs carry a sign and a significand, and non-numeric functions treat two NaNs as equal when their signs and significands agree. Significands of NaNs are machine-dependent, as are the digits in their string representation.

NaNs are not available on systems which do not use IEEE floating-point arithmetic; if the read syntax for a NaN is used on a VAX, for example, the reader signals an error.

When NaNs and signed zeros are involved, non-numeric functions like eql, equal, sxhash-eql, sxhash-equal and gethash determine whether values are indistinguishable, not whether they are numerically equal. For example, when x and y are the same NaN, (equal x y) returns t whereas (= x y) uses numeric comparison and returns nil; conversely, (equal 0.0 -0.0) returns nil whereas (= 0.0 -0.0) returns t.

Here are read syntaxes for these special floating-point values:

infinity

1.0e+INF’ and ‘-1.0e+INF

not-a-number

0.0e+NaN’ and ‘-0.0e+NaN

The following functions are specialized for handling floating-point numbers:

Function: isnan x

This predicate returns t if its floating-point argument is a NaN, nil otherwise.

Function: frexp x

This function returns a cons cell (s . e), where s and e are respectively the significand and exponent of the floating-point number x.

If x is finite, then s is a floating-point number between 0.5 (inclusive) and 1.0 (exclusive), e is an integer, and x = s * 2**e. If x is zero or infinity, then s is the same as x. If x is a NaN, then s is also a NaN. If x is zero, then e is 0.

Function: ldexp s e

Given a numeric significand s and an integer exponent e, this function returns the floating point number s * 2**e.

Function: copysign x1 x2

This function copies the sign of x2 to the value of x1, and returns the result. x1 and x2 must be floating point.

Function: logb x

This function returns the binary exponent of x. More precisely, if x is finite and nonzero, the value is the logarithm base 2 of |x|, rounded down to an integer. If x is zero or infinite, the value is infinity; if x is a NaN, the value is a NaN.

(logb 10)
     ⇒ 3
(logb 10.0e20)
     ⇒ 69
(logb 0)
     ⇒ -1.0e+INF

3.3 Type Predicates for Numbers

The functions in this section test for numbers, or for a specific type of number. The functions integerp and floatp can take any type of Lisp object as argument (they would not be of much use otherwise), but the zerop predicate requires a number as its argument. See also integer-or-marker-p and number-or-marker-p, in Predicates on Markers.

Function: bignump object

This predicate tests whether its argument is a large integer, and returns t if so, nil otherwise. Unlike small integers, large integers can be = or eql even if they are not eq.

Function: fixnump object

This predicate tests whether its argument is a small integer, and returns t if so, nil otherwise. Small integers can be compared with eq.

Function: floatp object

This predicate tests whether its argument is floating point and returns t if so, nil otherwise.

Function: integerp object

This predicate tests whether its argument is an integer, and returns t if so, nil otherwise.

Function: numberp object

This predicate tests whether its argument is a number (either integer or floating point), and returns t if so, nil otherwise.

Function: natnump object

This predicate (whose name comes from the phrase “natural number”) tests to see whether its argument is a nonnegative integer, and returns t if so, nil otherwise. 0 is considered non-negative.

wholenump is a synonym for natnump.

Function: zerop number

This predicate tests whether its argument is zero, and returns t if so, nil otherwise. The argument must be a number.

(zerop x) is equivalent to (= x 0).


3.4 Comparison of Numbers

To test numbers for numerical equality, you should normally use = instead of non-numeric comparison predicates like eq, eql and equal. Distinct floating-point and large integer objects can be numerically equal. If you use eq to compare them, you test whether they are the same object; if you use eql or equal, you test whether their values are indistinguishable. In contrast, = uses numeric comparison, and sometimes returns t when a non-numeric comparison would return nil and vice versa. See Floating-Point Basics.

In Emacs Lisp, if two fixnums are numerically equal, they are the same Lisp object. That is, eq is equivalent to = on fixnums. It is sometimes convenient to use eq for comparing an unknown value with a fixnum, because eq does not report an error if the unknown value is not a number—it accepts arguments of any type. By contrast, = signals an error if the arguments are not numbers or markers. However, it is better programming practice to use = if you can, even for comparing integers.

Sometimes it is useful to compare numbers with eql or equal, which treat two numbers as equal if they have the same data type (both integers, or both floating point) and the same value. By contrast, = can treat an integer and a floating-point number as equal. See Equality Predicates.

There is another wrinkle: because floating-point arithmetic is not exact, it is often a bad idea to check for equality of floating-point values. Usually it is better to test for approximate equality. Here’s a function to do this:

(defvar fuzz-factor 1.0e-6)
(defun approx-equal (x y)
  (or (= x y)
      (< (/ (abs (- x y))
            (max (abs x) (abs y)))
         fuzz-factor)))
Function: = number-or-marker &rest number-or-markers

This function tests whether all its arguments are numerically equal, and returns t if so, nil otherwise.

Function: eql value1 value2

This function acts like eq except when both arguments are numbers. It compares numbers by type and numeric value, so that (eql 1.0 1) returns nil, but (eql 1.0 1.0) and (eql 1 1) both return t. This can be used to compare large integers as well as small ones. Floating-point values with the same sign, exponent and fraction are eql. This differs from numeric comparison: (eql 0.0 -0.0) returns nil and (eql 0.0e+NaN 0.0e+NaN) returns t, whereas = does the opposite.

Function: /= number-or-marker1 number-or-marker2

This function tests whether its arguments are numerically equal, and returns t if they are not, and nil if they are.

Function: < number-or-marker &rest number-or-markers

This function tests whether each argument is strictly less than the following argument. It returns t if so, nil otherwise.

Function: <= number-or-marker &rest number-or-markers

This function tests whether each argument is less than or equal to the following argument. It returns t if so, nil otherwise.

Function: > number-or-marker &rest number-or-markers

This function tests whether each argument is strictly greater than the following argument. It returns t if so, nil otherwise.

Function: >= number-or-marker &rest number-or-markers

This function tests whether each argument is greater than or equal to the following argument. It returns t if so, nil otherwise.

Function: max number-or-marker &rest numbers-or-markers

This function returns the largest of its arguments.

(max 20)
     ⇒ 20
(max 1 2.5)
     ⇒ 2.5
(max 1 3 2.5)
     ⇒ 3
Function: min number-or-marker &rest numbers-or-markers

This function returns the smallest of its arguments.

(min -4 1)
     ⇒ -4
Function: abs number

This function returns the absolute value of number.


3.5 Numeric Conversions

To convert an integer to floating point, use the function float.

Function: float number

This returns number converted to floating point. If number is already floating point, float returns it unchanged.

There are four functions to convert floating-point numbers to integers; they differ in how they round. All accept an argument number and an optional argument divisor. Both arguments may be integers or floating-point numbers. divisor may also be nil. If divisor is nil or omitted, these functions convert number to an integer, or return it unchanged if it already is an integer. If divisor is non-nil, they divide number by divisor and convert the result to an integer. If divisor is zero (whether integer or floating point), Emacs signals an arith-error error.

Function: truncate number &optional divisor

This returns number, converted to an integer by rounding towards zero.

(truncate 1.2)
     ⇒ 1
(truncate 1.7)
     ⇒ 1
(truncate -1.2)
     ⇒ -1
(truncate -1.7)
     ⇒ -1
Function: floor number &optional divisor

This returns number, converted to an integer by rounding downward (towards negative infinity).

If divisor is specified, this uses the kind of division operation that corresponds to mod, rounding downward.

(floor 1.2)
     ⇒ 1
(floor 1.7)
     ⇒ 1
(floor -1.2)
     ⇒ -2
(floor -1.7)
     ⇒ -2
(floor 5.99 3)
     ⇒ 1
Function: ceiling number &optional divisor

This returns number, converted to an integer by rounding upward (towards positive infinity).

(ceiling 1.2)
     ⇒ 2
(ceiling 1.7)
     ⇒ 2
(ceiling -1.2)
     ⇒ -1
(ceiling -1.7)
     ⇒ -1
Function: round number &optional divisor

This returns number, converted to an integer by rounding towards the nearest integer. Rounding a value equidistant between two integers returns the even integer.

(round 1.2)
     ⇒ 1
(round 1.7)
     ⇒ 2
(round -1.2)
     ⇒ -1
(round -1.7)
     ⇒ -2

3.6 Arithmetic Operations

Emacs Lisp provides the traditional four arithmetic operations (addition, subtraction, multiplication, and division), as well as remainder and modulus functions, and functions to add or subtract 1. Except for %, each of these functions accepts both integer and floating-point arguments, and returns a floating-point number if any argument is floating point.

Function: 1+ number-or-marker

This function returns number-or-marker plus 1. For example,

(setq foo 4)
     ⇒ 4
(1+ foo)
     ⇒ 5

This function is not analogous to the C operator ++—it does not increment a variable. It just computes a sum. Thus, if we continue,

foo
     ⇒ 4

If you want to increment the variable, you must use setq, like this:

(setq foo (1+ foo))
     ⇒ 5
Function: 1- number-or-marker

This function returns number-or-marker minus 1.

Function: + &rest numbers-or-markers

This function adds its arguments together. When given no arguments, + returns 0.

(+)
     ⇒ 0
(+ 1)
     ⇒ 1
(+ 1 2 3 4)
     ⇒ 10
Function: - &optional number-or-marker &rest more-numbers-or-markers

The - function serves two purposes: negation and subtraction. When - has a single argument, the value is the negative of the argument. When there are multiple arguments, - subtracts each of the more-numbers-or-markers from number-or-marker, cumulatively. If there are no arguments, the result is 0.

(- 10 1 2 3 4)
     ⇒ 0
(- 10)
     ⇒ -10
(-)
     ⇒ 0
Function: * &rest numbers-or-markers

This function multiplies its arguments together, and returns the product. When given no arguments, * returns 1.

(*)
     ⇒ 1
(* 1)
     ⇒ 1
(* 1 2 3 4)
     ⇒ 24
Function: / number &rest divisors

With one or more divisors, this function divides number by each divisor in divisors in turn, and returns the quotient. With no divisors, this function returns 1/number, i.e., the multiplicative inverse of number. Each argument may be a number or a marker.

If all the arguments are integers, the result is an integer, obtained by rounding the quotient towards zero after each division.

(/ 6 2)
     ⇒ 3
(/ 5 2)
     ⇒ 2
(/ 5.0 2)
     ⇒ 2.5
(/ 5 2.0)
     ⇒ 2.5
(/ 5.0 2.0)
     ⇒ 2.5
(/ 4.0)
     ⇒ 0.25
(/ 4)
     ⇒ 0
(/ 25 3 2)
     ⇒ 4
(/ -17 6)
     ⇒ -2

If you divide an integer by the integer 0, Emacs signals an arith-error error (see Errors). On systems using IEEE-754 floating-point, floating-point division of a nonzero number by zero yields either positive or negative infinity (see Floating-Point Basics); otherwise, an arith-error is signaled as usual.

Function: % dividend divisor

This function returns the integer remainder after division of dividend by divisor. The arguments must be integers or markers.

For any two integers dividend and divisor,

(+ (% dividend divisor)
   (* (/ dividend divisor) divisor))

always equals dividend if divisor is nonzero.

(% 9 4)
     ⇒ 1
(% -9 4)
     ⇒ -1
(% 9 -4)
     ⇒ 1
(% -9 -4)
     ⇒ -1
Function: mod dividend divisor

This function returns the value of dividend modulo divisor; in other words, the remainder after division of dividend by divisor, but with the same sign as divisor. The arguments must be numbers or markers.

Unlike %, mod permits floating-point arguments; it rounds the quotient downward (towards minus infinity) to an integer, and uses that quotient to compute the remainder.

If divisor is zero, mod signals an arith-error error if both arguments are integers, and returns a NaN otherwise.

(mod 9 4)
     ⇒ 1
(mod -9 4)
     ⇒ 3
(mod 9 -4)
     ⇒ -3
(mod -9 -4)
     ⇒ -1
(mod 5.5 2.5)
     ⇒ .5

For any two numbers dividend and divisor,

(+ (mod dividend divisor)
   (* (floor dividend divisor) divisor))

always equals dividend, subject to rounding error if either argument is floating point and to an arith-error if dividend is an integer and divisor is 0. For floor, see Numeric Conversions.


3.7 Rounding Operations

The functions ffloor, fceiling, fround, and ftruncate take a floating-point argument and return a floating-point result whose value is a nearby integer. ffloor returns the nearest integer below; fceiling, the nearest integer above; ftruncate, the nearest integer in the direction towards zero; fround, the nearest integer.

Function: ffloor float

This function rounds float to the next lower integral value, and returns that value as a floating-point number.

Function: fceiling float

This function rounds float to the next higher integral value, and returns that value as a floating-point number.

Function: ftruncate float

This function rounds float towards zero to an integral value, and returns that value as a floating-point number.

Function: fround float

This function rounds float to the nearest integral value, and returns that value as a floating-point number. Rounding a value equidistant between two integers returns the even integer.


3.8 Bitwise Operations on Integers

In a computer, an integer is represented as a binary number, a sequence of bits (digits which are either zero or one). Conceptually the bit sequence is infinite on the left, with the most-significant bits being all zeros or all ones. A bitwise operation acts on the individual bits of such a sequence. For example, shifting moves the whole sequence left or right one or more places, reproducing the same pattern moved over.

The bitwise operations in Emacs Lisp apply only to integers.

Function: ash integer1 count

ash (arithmetic shift) shifts the bits in integer1 to the left count places, or to the right if count is negative. Left shifts introduce zero bits on the right; right shifts discard the rightmost bits. Considered as an integer operation, ash multiplies integer1 by 2**count, and then converts the result to an integer by rounding downward, toward minus infinity.

Here are examples of ash, shifting a pattern of bits one place to the left and to the right. These examples show only the low-order bits of the binary pattern; leading bits all agree with the highest-order bit shown. As you can see, shifting left by one is equivalent to multiplying by two, whereas shifting right by one is equivalent to dividing by two and then rounding toward minus infinity.

(ash 7 1) ⇒ 14
;; Decimal 7 becomes decimal 14.
…000111
     ⇒
…001110

(ash 7 -1) ⇒ 3
…000111
     ⇒
…000011

(ash -7 1) ⇒ -14
…111001
     ⇒
…110010

(ash -7 -1) ⇒ -4
…111001
     ⇒
…111100

Here are examples of shifting left or right by two bits:

                  ;         binary values
(ash 5 2)         ;   5  =  …000101
     ⇒ 20         ;      =  …010100
(ash -5 2)        ;  -5  =  …111011
     ⇒ -20        ;      =  …101100
(ash 5 -2)
     ⇒ 1          ;      =  …000001
(ash -5 -2)
     ⇒ -2         ;      =  …111110
Function: lsh integer1 count

lsh, which is an abbreviation for logical shift, shifts the bits in integer1 to the left count places, or to the right if count is negative, bringing zeros into the vacated bits. If count is negative, then integer1 must be either a fixnum or a positive bignum, and lsh treats a negative fixnum as if it were unsigned by subtracting twice most-negative-fixnum before shifting, producing a nonnegative result. This quirky behavior dates back to when Emacs supported only fixnums; nowadays ash is a better choice.

As lsh behaves like ash except when integer1 and count1 are both negative, the following examples focus on these exceptional cases. These examples assume 30-bit fixnums.

                 ;      binary values
(ash -7 -1)      ; -7 = …111111111111111111111111111001
     ⇒ -4        ;    = …111111111111111111111111111100
(lsh -7 -1)
     ⇒ 536870908 ;    = …011111111111111111111111111100
(ash -5 -2)      ; -5 = …111111111111111111111111111011
     ⇒ -2        ;    = …111111111111111111111111111110
(lsh -5 -2)
     ⇒ 268435454 ;    = …001111111111111111111111111110
Function: logand &rest ints-or-markers

This function returns the bitwise AND of the arguments: the nth bit is 1 in the result if, and only if, the nth bit is 1 in all the arguments.

For example, using 4-bit binary numbers, the bitwise AND of 13 and 12 is 12: 1101 combined with 1100 produces 1100. In both the binary numbers, the leftmost two bits are both 1 so the leftmost two bits of the returned value are both 1. However, for the rightmost two bits, each is 0 in at least one of the arguments, so the rightmost two bits of the returned value are both 0.

Therefore,

(logand 13 12)
     ⇒ 12

If logand is not passed any argument, it returns a value of −1. This number is an identity element for logand because its binary representation consists entirely of ones. If logand is passed just one argument, it returns that argument.

                   ;        binary values

(logand 14 13)     ; 14  =  …001110
                   ; 13  =  …001101
     ⇒ 12         ; 12  =  …001100

(logand 14 13 4)   ; 14  =  …001110
                   ; 13  =  …001101
                   ;  4  =  …000100
     ⇒ 4          ;  4  =  …000100

(logand)
     ⇒ -1         ; -1  =  …111111
Function: logior &rest ints-or-markers

This function returns the bitwise inclusive OR of its arguments: the nth bit is 1 in the result if, and only if, the nth bit is 1 in at least one of the arguments. If there are no arguments, the result is 0, which is an identity element for this operation. If logior is passed just one argument, it returns that argument.

                   ;        binary values

(logior 12 5)      ; 12  =  …001100
                   ;  5  =  …000101
     ⇒ 13         ; 13  =  …001101

(logior 12 5 7)    ; 12  =  …001100
                   ;  5  =  …000101
                   ;  7  =  …000111
     ⇒ 15         ; 15  =  …001111
Function: logxor &rest ints-or-markers

This function returns the bitwise exclusive OR of its arguments: the nth bit is 1 in the result if, and only if, the nth bit is 1 in an odd number of the arguments. If there are no arguments, the result is 0, which is an identity element for this operation. If logxor is passed just one argument, it returns that argument.

                   ;        binary values

(logxor 12 5)      ; 12  =  …001100
                   ;  5  =  …000101
     ⇒ 9          ;  9  =  …001001

(logxor 12 5 7)    ; 12  =  …001100
                   ;  5  =  …000101
                   ;  7  =  …000111
     ⇒ 14         ; 14  =  …001110
Function: lognot integer

This function returns the bitwise complement of its argument: the nth bit is one in the result if, and only if, the nth bit is zero in integer, and vice-versa. The result equals −1 − integer.

(lognot 5)
     ⇒ -6
;;  5  =  …000101
;; becomes
;; -6  =  …111010
Function: logcount integer

This function returns the Hamming weight of integer: the number of ones in the binary representation of integer. If integer is negative, it returns the number of zero bits in its two’s complement binary representation. The result is always nonnegative.

(logcount 43)     ;  43 = …000101011
     ⇒ 4
(logcount -43)    ; -43 = …111010101
     ⇒ 3

3.9 Standard Mathematical Functions

These mathematical functions allow integers as well as floating-point numbers as arguments.

Function: sin arg
Function: cos arg
Function: tan arg

These are the basic trigonometric functions, with argument arg measured in radians.

Function: asin arg

The value of (asin arg) is a number between −pi/2 and pi/2 (inclusive) whose sine is arg. If arg is out of range (outside [−1, 1]), asin returns a NaN.

Function: acos arg

The value of (acos arg) is a number between 0 and pi (inclusive) whose cosine is arg. If arg is out of range (outside [−1, 1]), acos returns a NaN.

Function: atan y &optional x

The value of (atan y) is a number between −pi/2 and pi/2 (exclusive) whose tangent is y. If the optional second argument x is given, the value of (atan y x) is the angle in radians between the vector [x, y] and the X axis.

Function: exp arg

This is the exponential function; it returns e to the power arg.

Function: log arg &optional base

This function returns the logarithm of arg, with base base. If you don’t specify base, the natural base e is used. If arg or base is negative, log returns a NaN.

Function: expt x y

This function returns x raised to power y. If both arguments are integers and y is nonnegative, the result is an integer; in this case, overflow signals an error, so watch out. If x is a finite negative number and y is a finite non-integer, expt returns a NaN.

Function: sqrt arg

This returns the square root of arg. If arg is finite and less than zero, sqrt returns a NaN.

In addition, Emacs defines the following common mathematical constants:

Variable: float-e

The mathematical constant e (2.71828…).

Variable: float-pi

The mathematical constant pi (3.14159…).


3.10 Random Numbers

A deterministic computer program cannot generate true random numbers. For most purposes, pseudo-random numbers suffice. A series of pseudo-random numbers is generated in a deterministic fashion. The numbers are not truly random, but they have certain properties that mimic a random series. For example, all possible values occur equally often in a pseudo-random series.

Pseudo-random numbers are generated from a seed value. Starting from any given seed, the random function always generates the same sequence of numbers. By default, Emacs initializes the random seed at startup, in such a way that the sequence of values of random (with overwhelming likelihood) differs in each Emacs run. The random seed is typically initialized from system entropy; however, on obsolescent platforms lacking entropy pools, the seed is taken from less-random volatile data such as the current time.

Sometimes you want the random number sequence to be repeatable. For example, when debugging a program whose behavior depends on the random number sequence, it is helpful to get the same behavior in each program run. To make the sequence repeat, execute (random ""). This sets the seed to a constant value for your particular Emacs executable (though it may differ for other Emacs builds). You can use other strings to choose various seed values.

Function: random &optional limit

This function returns a pseudo-random integer. Repeated calls return a series of pseudo-random integers.

If limit is a positive integer, the value is chosen to be nonnegative and less than limit. Otherwise, the value might be any fixnum, i.e., any integer from most-negative-fixnum through most-positive-fixnum (see Integer Basics).

If limit is a string, it means to choose a new seed based on the string’s contents. This causes later calls to random to return a reproducible sequence of results.

If limit is t, it means to choose a new seed as if Emacs were restarting. This causes later calls to random to return an unpredictable sequence of results.

If you need a random nonce for cryptographic purposes, using random is typically not the best approach, for several reasons:

  • Although you can use (random t) to consult system entropy, doing so can adversely affect other parts of your program that benefit from reproducible results.
  • The system-dependent pseudo-random number generator (PRNG) used by random is not necessarily suitable for cryptography.
  • A call to (random t) does not give direct access to system entropy; the entropy is passed through the system-dependent PRNG, thus possibly biasing the results.
  • On typical platforms the random seed contains only 32 bits, which is typically narrower than an Emacs fixnum, and is not nearly enough for cryptographic purposes.
  • A (random t) call leaves information about the nonce scattered about Emacs’s internal state, increasing the size of the internal attack surface.
  • On obsolescent platforms lacking entropy pools, (random t) is seeded from a cryptographically weak source.

4 Strings and Characters

A string in Emacs Lisp is an array that contains an ordered sequence of characters. Strings are used as names of symbols, buffers, and files; to send messages to users; to hold text being copied between buffers; and for many other purposes. Because strings are so important, Emacs Lisp has many functions expressly for manipulating them. Emacs Lisp programs use strings more often than individual characters.

See Putting Keyboard Events in Strings, for special considerations for strings of keyboard character events.


4.1 String and Character Basics

A character is a Lisp object which represents a single character of text. In Emacs Lisp, characters are simply integers; whether an integer is a character or not is determined only by how it is used. See Character Codes, for details about character representation in Emacs.

A string is a fixed sequence of characters. It is a type of sequence called an array, meaning that its length is fixed and cannot be altered once it is created (see Sequences, Arrays, and Vectors). Unlike in C, Emacs Lisp strings are not terminated by a distinguished character code.

Since strings are arrays, and therefore sequences as well, you can operate on them with the general array and sequence functions documented in Sequences, Arrays, and Vectors. For example, you can access individual characters in a string using the function aref (see Functions that Operate on Arrays).

There are two text representations for non-ASCII characters in Emacs strings (and in buffers): unibyte and multibyte. For most Lisp programming, you don’t need to be concerned with these two representations. See Text Representations, for details.

Sometimes key sequences are represented as unibyte strings. When a unibyte string is a key sequence, string elements in the range 128 to 255 represent meta characters (which are large integers) rather than character codes in the range 128 to 255. Strings cannot hold characters that have the hyper, super or alt modifiers; they can hold ASCII control characters, but no other control characters. They do not distinguish case in ASCII control characters. If you want to store such characters in a sequence, such as a key sequence, you must use a vector instead of a string. See Character Type, for more information about keyboard input characters.

Strings are useful for holding regular expressions. You can also match regular expressions against strings with string-match (see Regular Expression Searching). The functions match-string (see Simple Match Data Access) and replace-match (see Replacing the Text that Matched) are useful for decomposing and modifying strings after matching regular expressions against them.

Like a buffer, a string can contain text properties for the characters in it, as well as the characters themselves. See Text Properties. All the Lisp primitives that copy text from strings to buffers or other strings also copy the properties of the characters being copied.

See Text, for information about functions that display strings or copy them into buffers. See Character Type, and String Type, for information about the syntax of characters and strings. See Non-ASCII Characters, for functions to convert between text representations and to encode and decode character codes. Also, note that length should not be used for computing the width of a string on display; use string-width (see Size of Displayed Text) instead.


4.2 Predicates for Strings

For more information about general sequence and array predicates, see Sequences, Arrays, and Vectors, and Arrays.

Function: stringp object

This function returns t if object is a string, nil otherwise.

Function: string-or-null-p object

This function returns t if object is a string or nil. It returns nil otherwise.

Function: char-or-string-p object

This function returns t if object is a string or a character (i.e., an integer), nil otherwise.


4.3 Creating Strings

The following functions create strings, either from scratch, or by putting strings together, or by taking them apart. (For functions that create strings based on the modified contents of other strings, like string-replace and replace-regexp-in-string, see Search and Replace.)

Function: make-string count character &optional multibyte

This function returns a string made up of count repetitions of character. If count is negative, an error is signaled.

(make-string 5 ?x)
     ⇒ "xxxxx"
(make-string 0 ?x)
     ⇒ ""

Normally, if character is an ASCII character, the result is a unibyte string. But if the optional argument multibyte is non-nil, the function will produce a multibyte string instead. This is useful when you later need to concatenate the result with non-ASCII strings or replace some of its characters with non-ASCII characters.

Other functions to compare with this one include make-vector (see Vectors) and make-list (see Building Cons Cells and Lists).

Function: string &rest characters

This returns a string containing the characters characters.

(string ?a ?b ?c)
     ⇒ "abc"
Function: substring string &optional start end

This function returns a new string which consists of those characters from string in the range from (and including) the character at the index start up to (but excluding) the character at the index end. The first character is at index zero. With one argument, this function just copies string.

(substring "abcdefg" 0 3)
     ⇒ "abc"

In the above example, the index for ‘a’ is 0, the index for ‘b’ is 1, and the index for ‘c’ is 2. The index 3—which is the fourth character in the string—marks the character position up to which the substring is copied. Thus, ‘abc’ is copied from the string "abcdefg".

A negative number counts from the end of the string, so that −1 signifies the index of the last character of the string. For example:

(substring "abcdefg" -3 -1)
     ⇒ "ef"

In this example, the index for ‘e’ is −3, the index for ‘f’ is −2, and the index for ‘g’ is −1. Therefore, ‘e’ and ‘f’ are included, and ‘g’ is excluded.

When nil is used for end, it stands for the length of the string. Thus,

(substring "abcdefg" -3 nil)
     ⇒ "efg"

Omitting the argument end is equivalent to specifying nil. It follows that (substring string 0) returns a copy of all of string.

(substring "abcdefg" 0)
     ⇒ "abcdefg"

But we recommend copy-sequence for this purpose (see Sequences).

If the characters copied from string have text properties, the properties are copied into the new string also. See Text Properties.

substring also accepts a vector for the first argument. For example:

(substring [a b (c) "d"] 1 3)
     ⇒ [b (c)]

A wrong-type-argument error is signaled if start is not an integer or if end is neither an integer nor nil. An args-out-of-range error is signaled if start indicates a character following end, or if either integer is out of range for string.

Contrast this function with buffer-substring (see Examining Buffer Contents), which returns a string containing a portion of the text in the current buffer. The beginning of a string is at index 0, but the beginning of a buffer is at index 1.

Function: substring-no-properties string &optional start end

This works like substring but discards all text properties from the value. Also, start may be omitted or nil, which is equivalent to 0. Thus, (substring-no-properties string) returns a copy of string, with all text properties removed.

Function: concat &rest sequences

This function returns a string consisting of the characters in the arguments passed to it (along with their text properties, if any). The arguments may be strings, lists of numbers, or vectors of numbers; they are not themselves changed. If concat receives no arguments, it returns an empty string.

(concat "abc" "-def")
     ⇒ "abc-def"
(concat "abc" (list 120 121) [122])
     ⇒ "abcxyz"
;; nil is an empty sequence.
(concat "abc" nil "-def")
     ⇒ "abc-def"
(concat "The " "quick brown " "fox.")
     ⇒ "The quick brown fox."
(concat)
     ⇒ ""

This function does not always allocate a new string. Callers are advised not rely on the result being a new string nor on it being eq to an existing string.

In particular, mutating the returned value may inadvertently change another string, alter a constant string in the program, or even raise an error. To obtain a string that you can safely mutate, use copy-sequence on the result.

For information about other concatenation functions, see the description of mapconcat in Mapping Functions, vconcat in Functions for Vectors, and append in Building Cons Cells and Lists. For concatenating individual command-line arguments into a string to be used as a shell command, see combine-and-quote-strings.

Function: split-string string &optional separators omit-nulls trim

This function splits string into substrings based on the regular expression separators (see Regular Expressions). Each match for separators defines a splitting point; the substrings between splitting points are made into a list, which is returned.

If separators is nil (or omitted), the default is the value of split-string-default-separators and the function behaves as if omit-nulls were t.

If omit-nulls is nil (or omitted), the result contains null strings whenever there are two consecutive matches for separators, or a match is adjacent to the beginning or end of string. If omit-nulls is t, these null strings are omitted from the result.

If the optional argument trim is non-nil, it should be a regular expression to match text to trim from the beginning and end of each substring. If trimming makes the substring empty, it is treated as null.

If you need to split a string into a list of individual command-line arguments suitable for call-process or start-process, see split-string-and-unquote.

Examples:

(split-string "  two words ")
     ⇒ ("two" "words")

The result is not ("" "two" "words" ""), which would rarely be useful. If you need such a result, use an explicit value for separators:

(split-string "  two words "
              split-string-default-separators)
     ⇒ ("" "two" "words" "")
(split-string "Soup is good food" "o")
     ⇒ ("S" "up is g" "" "d f" "" "d")
(split-string "Soup is good food" "o" t)
     ⇒ ("S" "up is g" "d f" "d")
(split-string "Soup is good food" "o+")
     ⇒ ("S" "up is g" "d f" "d")

Empty matches do count, except that split-string will not look for a final empty match when it already reached the end of the string using a non-empty match or when string is empty:

(split-string "aooob" "o*")
     ⇒ ("" "a" "" "b" "")
(split-string "ooaboo" "o*")
     ⇒ ("" "" "a" "b" "")
(split-string "" "")
     ⇒ ("")

However, when separators can match the empty string, omit-nulls is usually t, so that the subtleties in the three previous examples are rarely relevant:

(split-string "Soup is good food" "o*" t)
     ⇒ ("S" "u" "p" " " "i" "s" " " "g" "d" " " "f" "d")
(split-string "Nice doggy!" "" t)
     ⇒ ("N" "i" "c" "e" " " "d" "o" "g" "g" "y" "!")
(split-string "" "" t)
     ⇒ nil

Somewhat odd, but predictable, behavior can occur for certain “non-greedy” values of separators that can prefer empty matches over non-empty matches. Again, such values rarely occur in practice:

(split-string "ooo" "o*" t)
     ⇒ nil
(split-string "ooo" "\\|o+" t)
     ⇒ ("o" "o" "o")
Variable: split-string-default-separators

The default value of separators for split-string. Its usual value is "[ \f\t\n\r\v]+".

Function: string-clean-whitespace string

Clean up the whitespace in string by collapsing stretches of whitespace to a single space character, as well as removing all whitespace from the start and the end of string.

Function: string-trim-left string &optional regexp

Remove the leading text that matches regexp from string. regexp defaults to ‘[ \t\n\r]+’.

Function: string-trim-right string &optional regexp

Remove the trailing text that matches regexp from string. regexp defaults to ‘[ \t\n\r]+’.

Function: string-trim string &optional trim-left trim-right

Remove the leading text that matches trim-left and trailing text that matches trim-right from string. Both regexps default to ‘[ \t\n\r]+’.

Function: string-fill string length

Attempt to Word-wrap string so that no lines are longer than length. Filling is done on whitespace boundaries only. If there are individual words that are longer than length, these will not be shortened.

Function: string-limit string length &optional end coding-system

If string is shorter than length characters, string is returned as is. Otherwise, return a substring of string consisting of the first length characters. If the optional end parameter is given, return a string of the length last characters instead.

If coding-system is non-nil, string will be encoded before limiting, and the result will be a unibyte string that’s shorter than length bytes. If string contains characters that are encoded into several bytes (for instance, when using utf-8), the resulting unibyte string is never truncated in the middle of a character representation.

This function measures the string length in characters or bytes, and thus is generally inappropriate if you need to shorten strings for display purposes; use truncate-string-to-width or window-text-pixel-size or string-glyph-split instead (see Size of Displayed Text).

Function: string-lines string &optional omit-nulls keep-newlines

Split string into a list of strings on newline boundaries. If the optional argument omit-nulls is non-nil, remove empty lines from the results. If the optional argument keep-newlines is non-nil, don’t remove the trailing newlines from the result strings.

Function: string-pad string length &optional padding start

Pad string to be of the given length using padding as the padding character. padding defaults to the space character. If string is longer than length, no padding is done. If start is nil or omitted, the padding is appended to the characters of string, and if it’s non-nil, the padding is prepended to string’s characters.

Function: string-chop-newline string

Remove the final newline, if any, from string.


4.4 Modifying Strings

You can alter the contents of a mutable string via operations described in this section. See Mutability.

The most basic way to alter the contents of an existing string is with aset (see Functions that Operate on Arrays). (aset string idx char) stores char into string at character index idx. It will automatically convert a pure-ASCII string to a multibyte string (see Text Representations) if needed, but we recommend to always make sure string is multibyte (e.g., by using string-to-multibyte, see Converting Text Representations), if char is a non-ASCII character, not a raw byte.

A more powerful function is store-substring:

Function: store-substring string idx obj

This function alters part of the contents of the specified string, by storing obj starting at character index idx. The argument obj may be either a character (in which case the function behaves exactly as aset) or a (smaller) string. If obj is a multibyte string, we recommend to make sure string is also multibyte, even if it’s pure-ASCII.

Since it is impossible to change the number of characters in an existing string, it is an error if obj consists of more characters than would fit in string starting at character index idx.

To clear out a string that contained a password, use clear-string:

Function: clear-string string

This makes string a unibyte string and clears its contents to zeros. It may also change string’s length.


4.5 Comparison of Characters and Strings

Function: char-equal character1 character2

This function returns t if the arguments represent the same character, nil otherwise. This function ignores differences in case if case-fold-search is non-nil.

(char-equal ?x ?x)
     ⇒ t
(let ((case-fold-search nil))
  (char-equal ?x ?X))
     ⇒ nil
Function: string= string1 string2

This function returns t if the characters of the two strings match exactly. Symbols are also allowed as arguments, in which case the symbol names are used. Case is always significant, regardless of case-fold-search.

This function is equivalent to equal for comparing two strings (see Equality Predicates). In particular, the text properties of the two strings are ignored; use equal-including-properties if you need to distinguish between strings that differ only in their text properties. However, unlike equal, if either argument is not a string or symbol, string= signals an error.

(string= "abc" "abc")
     ⇒ t
(string= "abc" "ABC")
     ⇒ nil
(string= "ab" "ABC")
     ⇒ nil

A unibyte and a multibyte string are equal in the sense of string= if and only if they contain the same sequence of character codes all being in the range 0–127 (ASCII). See Text Representations.

Function: string-equal string1 string2

string-equal is another name for string=.

Function: string-equal-ignore-case string1 string2

string-equal-ignore-case compares strings ignoring case differences, like char-equal when case-fold-search is t.

Function: string-collate-equalp string1 string2 &optional locale ignore-case

This function returns t if string1 and string2 are equal with respect to the collation rules of the specified locale, which defaults to your current system locale. A collation rule is not only determined by the lexicographic order of the characters contained in string1 and string2, but also by further rules about relations between these characters. Usually, it is defined by the locale environment with which Emacs is running and by the Standard C library against which Emacs was linked3.

For example, characters with different code points but the same meaning, like different grave accent Unicode characters, might, in some locales, be considered as equal:

(string-collate-equalp (string ?\uFF40) (string ?\u1FEF))
     ⇒ t

The optional argument locale, a string, overrides the setting of your current locale identifier for collation. The value is system dependent; a locale "en_US.UTF-8" is applicable on POSIX systems, while it would be, e.g., "enu_USA.1252" on MS-Windows systems.

If ignore-case is non-nil, characters are compared case-insensitively, by converting them to lower-case. However, if the underlying system library doesn’t provide locale-specific collation rules, this function falls back to string-equal, in which case the ignore-case argument is ignored, and the comparison will always be case-sensitive.

To emulate Unicode-compliant collation on MS-Windows systems, bind w32-collate-ignore-punctuation to a non-nil value, since the codeset part of the locale cannot be "UTF-8" on MS-Windows.

If your system does not support a locale environment, this function behaves like string-equal.

Do not use this function to compare file names for equality, as filesystems generally don’t honor linguistic equivalence of strings that collation implements.

Function: string< string1 string2

This function compares two strings a character at a time. It scans both the strings at the same time to find the first pair of corresponding characters that do not match. If the lesser character of these two is the character from string1, then string1 is less, and this function returns t. If the lesser character is the one from string2, then string1 is greater, and this function returns nil. If the two strings match entirely, the value is nil.

Pairs of characters are compared according to their character codes. Keep in mind that lower case letters have higher numeric values in the ASCII character set than their upper case counterparts; digits and many punctuation characters have a lower numeric value than upper case letters. An ASCII character is less than any non-ASCII character; a unibyte non-ASCII character is always less than any multibyte non-ASCII character (see Text Representations).

(string< "abc" "abd")
     ⇒ t
(string< "abd" "abc")
     ⇒ nil
(string< "123" "abc")
     ⇒ t

When the strings have different lengths, and they match up to the length of string1, then the result is t. If they match up to the length of string2, the result is nil. A string of no characters is less than any other string.

(string< "" "abc")
     ⇒ t
(string< "ab" "abc")
     ⇒ t
(string< "abc" "")
     ⇒ nil
(string< "abc" "ab")
     ⇒ nil
(string< "" "")
     ⇒ nil

Symbols are also allowed as arguments, in which case their print names are compared.

Function: string-lessp string1 string2

string-lessp is another name for string<.

Function: string-greaterp string1 string2

This function returns the result of comparing string1 and string2 in the opposite order, i.e., it is equivalent to calling (string-lessp string2 string1).

Function: string-collate-lessp string1 string2 &optional locale ignore-case

This function returns t if string1 is less than string2 in collation order of the specified locale, which defaults to your current system locale. A collation order is not only determined by the lexicographic order of the characters contained in string1 and string2, but also by further rules about relations between these characters. Usually, it is defined by the locale environment with which Emacs is running, and by the Standard C library against which Emacs was linked.

For example, punctuation and whitespace characters might be ignored for sorting (see Sequences):

(sort (list "11" "12" "1 1" "1 2" "1.1" "1.2") 'string-collate-lessp)
     ⇒ ("11" "1 1" "1.1" "12" "1 2" "1.2")

This behavior is system-dependent; e.g., punctuation and whitespace are never ignored on Cygwin, regardless of locale.

The optional argument locale, a string, overrides the setting of your current locale identifier for collation. The value is system dependent; a locale "en_US.UTF-8" is applicable on POSIX systems, while it would be, e.g., "enu_USA.1252" on MS-Windows systems. The locale value of "POSIX" or "C" lets string-collate-lessp behave like string-lessp:

(sort (list "11" "12" "1 1" "1 2" "1.1" "1.2")
      (lambda (s1 s2) (string-collate-lessp s1 s2 "POSIX")))
     ⇒ ("1 1" "1 2" "1.1" "1.2" "11" "12")

If ignore-case is non-nil, characters are compared case-insensitively, by converting them to lower-case. However, if the underlying system library doesn’t provide locale-specific collation rules, this function falls back to string-lessp, in which case the ignore-case argument is ignored, and the comparison will always be case-sensitive.

To emulate Unicode-compliant collation on MS-Windows systems, bind w32-collate-ignore-punctuation to a non-nil value, since the codeset part of the locale cannot be "UTF-8" on MS-Windows.

If your system does not support a locale environment, this function behaves like string-lessp.

Function: string-version-lessp string1 string2

This function compares strings lexicographically, except it treats sequences of numerical characters as if they comprised a base-ten number, and then compares the numbers. So ‘foo2.png’ is “smaller” than ‘foo12.png’ according to this predicate, even if ‘12’ is lexicographically “smaller” than ‘2’.

Function: string-prefix-p string1 string2 &optional ignore-case

This function returns non-nil if string1 is a prefix of string2; i.e., if string2 starts with string1. If the optional argument ignore-case is non-nil, the comparison ignores case differences.

Function: string-suffix-p suffix string &optional ignore-case

This function returns non-nil if suffix is a suffix of string; i.e., if string ends with suffix. If the optional argument ignore-case is non-nil, the comparison ignores case differences.

Function: string-search needle haystack &optional start-pos

Return the position of the first instance of needle in haystack, both of which are strings. If start-pos is non-nil, start searching from that position in haystack. Return nil if no match was found. This function only considers the characters in the strings when doing the comparison; text properties are ignored. Matching is always case-sensitive.

Function: compare-strings string1 start1 end1 string2 start2 end2 &optional ignore-case

This function compares a specified part of string1 with a specified part of string2. The specified part of string1 runs from index start1 (inclusive) up to index end1 (exclusive); nil for start1 means the start of the string, while nil for end1 means the length of the string. Likewise, the specified part of string2 runs from index start2 up to index end2.

The strings are compared by the numeric values of their characters. For instance, str1 is considered less than str2 if its first differing character has a smaller numeric value. If ignore-case is non-nil, characters are converted to upper-case, using the current buffer’s case-table (see The Case Table), before comparing them. Unibyte strings are converted to multibyte for comparison (see Text Representations), so that a unibyte string and its conversion to multibyte are always regarded as equal.

If the specified portions of the two strings match, the value is t. Otherwise, the value is an integer which indicates how many leading characters agree, and which string is less. Its absolute value is one plus the number of characters that agree at the beginning of the two strings. The sign is negative if string1 (or its specified portion) is less.

Function: string-distance string1 string2 &optional bytecompare

This function returns the Levenshtein distance between the source string string1 and the target string string2. The Levenshtein distance is the number of single-character changes—deletions, insertions, or replacements—required to transform the source string into the target string; it is one possible definition of the edit distance between strings.

Letter-case of the strings is significant for the computed distance, but their text properties are ignored. If the optional argument bytecompare is non-nil, the function calculates the distance in terms of bytes instead of characters. The byte-wise comparison uses the internal Emacs representation of characters, so it will produce inaccurate results for multibyte strings that include raw bytes (see Text Representations); make the strings unibyte by encoding them (see Explicit Encoding and Decoding) if you need accurate results with raw bytes.

Function: assoc-string key alist &optional case-fold

This function works like assoc, except that key must be a string or symbol, and comparison is done using compare-strings. Symbols are converted to strings before testing. If case-fold is non-nil, key and the elements of alist are converted to upper-case before comparison. Unlike assoc, this function can also match elements of the alist that are strings or symbols rather than conses. In particular, alist can be a list of strings or symbols rather than an actual alist. See Association Lists.

See also the function compare-buffer-substrings in Comparing Text, for a way to compare text in buffers. The function string-match, which matches a regular expression against a string, can be used for a kind of string comparison; see Regular Expression Searching.


4.6 Conversion of Characters and Strings

This section describes functions for converting between characters, strings and integers. format (see Formatting Strings) and prin1-to-string (see Output Functions) can also convert Lisp objects into strings. read-from-string (see Input Functions) can convert a string representation of a Lisp object into an object. The functions string-to-multibyte and string-to-unibyte convert the text representation of a string (see Converting Text Representations).

See Documentation, for functions that produce textual descriptions of text characters and general input events (single-key-description and text-char-description). These are used primarily for making help messages.

Function: number-to-string number

This function returns a string consisting of the printed base-ten representation of number. The returned value starts with a minus sign if the argument is negative.

(number-to-string 256)
     ⇒ "256"
(number-to-string -23)
     ⇒ "-23"
(number-to-string -23.5)
     ⇒ "-23.5"

int-to-string is a semi-obsolete alias for this function.

See also the function format in Formatting Strings.

Function: string-to-number string &optional base

This function returns the numeric value of the characters in string. If base is non-nil, it must be an integer between 2 and 16 (inclusive), and integers are converted in that base. If base is nil, then base ten is used. Floating-point conversion only works in base ten; we have not implemented other radices for floating-point numbers, because that would be much more work and does not seem useful.

The parsing skips spaces and tabs at the beginning of string, then reads as much of string as it can interpret as a number in the given base. (On some systems it ignores other whitespace at the beginning, not just spaces and tabs.) If string cannot be interpreted as a number, this function returns 0.

(string-to-number "256")
     ⇒ 256
(string-to-number "25 is a perfect square.")
     ⇒ 25
(string-to-number "X256")
     ⇒ 0
(string-to-number "-4.5")
     ⇒ -4.5
(string-to-number "1e5")
     ⇒ 100000.0

string-to-int is an obsolete alias for this function.

Function: char-to-string character

This function returns a new string containing one character, character. This function is semi-obsolete because the function string is more general. See Creating Strings.

Function: string-to-char string

This function returns the first character in string. This is mostly identical to (aref string 0), except that it returns 0 if the string is empty. (The value is also 0 when the first character of string is the null character, ASCII code 0.) This function may be eliminated in the future if it does not seem useful enough to retain.

Here are some other functions that can convert to or from a string:

concat

This function converts a vector or a list into a string. See Creating Strings.

vconcat

This function converts a string into a vector. See Functions for Vectors.

append

This function converts a string into a list. See Building Cons Cells and Lists.

byte-to-string

This function converts a byte of character data into a unibyte string. See Converting Text Representations.


4.7 Formatting Strings

Formatting means constructing a string by substituting computed values at various places in a constant string. This constant string controls how the other values are printed, as well as where they appear; it is called a format string.

Formatting is often useful for computing messages to be displayed. In fact, the functions message and error provide the same formatting feature described here; they differ from format-message only in how they use the result of formatting.

Function: format string &rest objects

This function returns a string equal to string, replacing any format specifications with encodings of the corresponding objects. The arguments objects are the computed values to be formatted.

The characters in string, other than the format specifications, are copied directly into the output, including their text properties, if any. Any text properties of the format specifications are copied to the produced string representations of the argument objects.

The output string need not be newly-allocated. For example, if x is the string "foo", the expressions (eq x (format x)) and (eq x (format "%s" x)) might both yield t.

Function: format-message string &rest objects

This function acts like format, except it also converts any grave accents (`) and apostrophes (') in string as per the value of text-quoting-style.

Typically grave accent and apostrophe in the format translate to matching curved quotes, e.g., "Missing `%s'" might result in "Missing ‘foo’". See Text Quoting Style, for how to influence or inhibit this translation.

A format specification is a sequence of characters beginning with a ‘%’. Thus, if there is a ‘%d’ in string, the format function replaces it with the printed representation of one of the values to be formatted (one of the arguments objects). For example:

(format "The value of fill-column is %d." fill-column)
     ⇒ "The value of fill-column is 72."

Since format interprets ‘%’ characters as format specifications, you should never pass an arbitrary string as the first argument. This is particularly true when the string is generated by some Lisp code. Unless the string is known to never include any ‘%’ characters, pass "%s", described below, as the first argument, and the string as the second, like this:

  (format "%s" arbitrary-string)

Certain format specifications require values of particular types. If you supply a value that doesn’t fit the requirements, an error is signaled.

Here is a table of valid format specifications:

%s

Replace the specification with the printed representation of the object, made without quoting (that is, using princ, not prin1—see Output Functions). Thus, strings are represented by their contents alone, with no ‘"’ characters, and symbols appear without ‘\’ characters.

If the object is a string, its text properties are copied into the output. The text properties of the ‘%s’ itself are also copied, but those of the object take priority.

%S

Replace the specification with the printed representation of the object, made with quoting (that is, using prin1—see Output Functions). Thus, strings are enclosed in ‘"’ characters, and ‘\’ characters appear where necessary before special characters.

%o

Replace the specification with the base-eight representation of an integer. Negative integers are formatted in a platform-dependent way. The object can also be a floating-point number that is formatted as an integer, dropping any fraction.

%d

Replace the specification with the base-ten representation of a signed integer. The object can also be a floating-point number that is formatted as an integer, dropping any fraction.

%x
%X

Replace the specification with the base-sixteen representation of an integer. Negative integers are formatted in a platform-dependent way. ‘%x’ uses lower case and ‘%X’ uses upper case. The object can also be a floating-point number that is formatted as an integer, dropping any fraction.

%c

Replace the specification with the character which is the value given.

%e

Replace the specification with the exponential notation for a floating-point number.

%f

Replace the specification with the decimal-point notation for a floating-point number.

%g

Replace the specification with notation for a floating-point number, using either exponential notation or decimal-point notation. The exponential notation is used if the exponent would be less than −4 or greater than or equal to the precision (default: 6). By default, trailing zeros are removed from the fractional portion of the result and a decimal-point character appears only if it is followed by a digit.

%%

Replace the specification with a single ‘%’. This format specification is unusual in that its only form is plain ‘%%’ and that it does not use a value. For example, (format "%% %d" 30) returns "% 30".

Any other format character results in an ‘Invalid format operation’ error.

Here are several examples, which assume the typical text-quoting-style settings:

(format "The octal value of %d is %o,
         and the hex value is %x." 18 18 18)
     ⇒ "The octal value of 18 is 22,
         and the hex value is 12."

(format-message
 "The name of this buffer is ‘%s’." (buffer-name))
     ⇒ "The name of this buffer is ‘strings.texi’."

(format-message
 "The buffer object prints as `%s'." (current-buffer))
     ⇒ "The buffer object prints as ‘strings.texi’."

By default, format specifications correspond to successive values from objects. Thus, the first format specification in string uses the first such value, the second format specification uses the second such value, and so on. Any extra format specifications (those for which there are no corresponding values) cause an error. Any extra values to be formatted are ignored.

A format specification can have a field number, which is a decimal number immediately after the initial ‘%’, followed by a literal dollar sign ‘$’. It causes the format specification to convert the argument with the given number instead of the next argument. Field numbers start at 1. A format can contain either numbered or unnumbered format specifications but not both, except that ‘%%’ can be mixed with numbered specifications.

(format "%2$s, %3$s, %%, %1$s" "x" "y" "z")
     ⇒ "y, z, %, x"

After the ‘%’ and any field number, you can put certain flag characters.

The flag ‘+’ inserts a plus sign before a nonnegative number, so that it always has a sign. A space character as flag inserts a space before a nonnegative number. (Otherwise, nonnegative numbers start with the first digit.) These flags are useful for ensuring that nonnegative and negative numbers use the same number of columns. They are ignored except for ‘%d’, ‘%e’, ‘%f’, ‘%g’, and if both flags are used, ‘+’ takes precedence.

The flag ‘#’ specifies an alternate form which depends on the format in use. For ‘%o’, it ensures that the result begins with a ‘0’. For ‘%x’ and ‘%X’, it prefixes nonzero results with ‘0x’ or ‘0X’. For ‘%e’ and ‘%f’, the ‘#’ flag means include a decimal point even if the precision is zero. For ‘%g’, it always includes a decimal point, and also forces any trailing zeros after the decimal point to be left in place where they would otherwise be removed.

The flag ‘0’ ensures that the padding consists of ‘0’ characters instead of spaces. This flag is ignored for non-numerical specification characters like ‘%s’, ‘%S’ and ‘%c’. These specification characters accept the ‘0’ flag, but still pad with spaces.

The flag ‘-’ causes any padding inserted by the width, if specified, to be inserted on the right rather than the left. If both ‘-’ and ‘0’ are present, the ‘0’ flag is ignored.

(format "%06d is padded on the left with zeros" 123)
     ⇒ "000123 is padded on the left with zeros"

(format "'%-6d' is padded on the right" 123)
     ⇒ "'123   ' is padded on the right"

(format "The word '%-7s' actually has %d letters in it."
        "foo" (length "foo"))
     ⇒ "The word 'foo    ' actually has 3 letters in it."

A specification can have a width, which is a decimal number that appears after any field number and flags. If the printed representation of the object contains fewer characters than this width, format extends it with padding. Any padding introduced by the width normally consists of spaces inserted on the left:

(format "%5d is padded on the left with spaces" 123)
     ⇒ "  123 is padded on the left with spaces"

If the width is too small, format does not truncate the object’s printed representation. Thus, you can use a width to specify a minimum spacing between columns with no risk of losing information. In the following two examples, ‘%7s’ specifies a minimum width of 7. In the first case, the string inserted in place of ‘%7s’ has only 3 letters, and needs 4 blank spaces as padding. In the second case, the string "specification" is 13 letters wide but is not truncated.

(format "The word '%7s' has %d letters in it."
        "foo" (length "foo"))
     ⇒ "The word '    foo' has 3 letters in it."
(format "The word '%7s' has %d letters in it."
        "specification" (length "specification"))
     ⇒ "The word 'specification' has 13 letters in it."

All the specification characters allow an optional precision after the field number, flags and width, if present. The precision is a decimal-point ‘.’ followed by a digit-string. For the floating-point specifications (‘%e’ and ‘%f’), the precision specifies how many digits following the decimal point to show; if zero, the decimal-point itself is also omitted. For ‘%g’, the precision specifies how many significant digits to show (significant digits are the first digit before the decimal point and all the digits after it). If the precision of %g is zero or unspecified, it is treated as 1. For ‘%s’ and ‘%S’, the precision truncates the string to the given width, so ‘%.3s’ shows only the first three characters of the representation for object. For other specification characters, the effect of precision is what the local library functions of the printf family produce.

If you plan to use read later on the formatted string to retrieve a copy of the formatted value, use a specification that lets read reconstruct the value. To format numbers in this reversible way you can use ‘%s’ and ‘%S’, to format just integers you can also use ‘%d’, and to format just nonnegative integers you can also use ‘#x%x’ and ‘#o%o’. Other formats may be problematic; for example, ‘%d’ and ‘%g’ can mishandle NaNs and can lose precision and type, and ‘#x%x’ and ‘#o%o’ can mishandle negative integers. See Input Functions.

The functions described in this section accept a fixed set of specification characters. The next section describes a function format-spec which can accept custom specification characters, such as ‘%a’ or ‘%z’.


4.8 Custom Format Strings

Sometimes it is useful to allow users and Lisp programs alike to control how certain text is generated via custom format control strings. For example, a format string could control how to display someone’s forename, surname, and email address. Using the function format described in the previous section, the format string could be something like "%s %s <%s>". This approach quickly becomes impractical, however, as it can be unclear which specification character corresponds to which piece of information.

A more convenient format string for such cases would be something like "%f %l <%e>", where each specification character carries more semantic information and can easily be rearranged relative to other specification characters, making such format strings more easily customizable by the user.

The function format-spec described in this section performs a similar function to format, except it operates on format control strings that use arbitrary specification characters.

Function: format-spec template spec-alist &optional ignore-missing split

This function returns a string produced from the format string template according to conversions specified in spec-alist, which is an alist (see Association Lists) of the form (letter . replacement). Each specification %letter in template will be replaced by replacement when formatting the resulting string.

The characters in template, other than the format specifications, are copied directly into the output, including their text properties, if any. Any text properties of the format specifications are copied to their replacements.

Using an alist to specify conversions gives rise to some useful properties:

  • If spec-alist contains more unique letter keys than there are unique specification characters in template, the unused keys are simply ignored.
  • If spec-alist contains more than one association with the same letter, the closest one to the start of the list is used.
  • If template contains the same specification character more than once, then the same replacement found in spec-alist is used as a basis for all of that character’s substitutions.
  • The order of specifications in template need not correspond to the order of associations in spec-alist.

REPLACEMENT can also be a function taking no arguments, and returning a string to be used for the replacement. It will only be called when the corresponding LETTER is used in the TEMPLATE. This is useful, for example, to avoid prompting for input unless it is needed.

The optional argument ignore-missing indicates how to handle specification characters in template that are not found in spec-alist. If it is nil or omitted, the function signals an error; if it is ignore, those format specifications are left verbatim in the output, including their text properties, if any; if it is delete, those format specifications are removed from the output; any other non-nil value is handled like ignore, but any occurrences of ‘%%’ are also left verbatim in the output.

If the optional argument split is non-nil, instead of returning a single string, format-spec will split the result into a list of strings, based on where the substitutions were performed. For instance:

(format-spec "foo %b bar" '((?b . "zot")) nil t)
     ⇒ ("foo " "zot" " bar")

The syntax of format specifications accepted by format-spec is similar, but not identical, to that accepted by format. In both cases, a format specification is a sequence of characters beginning with ‘%’ and ending with an alphabetic letter such as ‘s’.

Unlike format, which assigns specific meanings to a fixed set of specification characters, format-spec accepts arbitrary specification characters and treats them all equally. For example:

(setq my-site-info
      (list (cons ?s system-name)
            (cons ?t (symbol-name system-type))
            (cons ?c system-configuration)
            (cons ?v emacs-version)
            (cons ?e invocation-name)
            (cons ?p (number-to-string (emacs-pid)))
            (cons ?a user-mail-address)
            (cons ?n user-full-name)))

(format-spec "%e %v (%c)" my-site-info)
     ⇒ "emacs 27.1 (x86_64-pc-linux-gnu)"

(format-spec "%n <%a>" my-site-info)
     ⇒ "Emacs Developers <emacs-devel@gnu.org>"

A format specification can include any number of the following flag characters immediately after the ‘%’ to modify aspects of the substitution.

0

This flag causes any padding specified by the width to consist of ‘0’ characters instead of spaces.

-

This flag causes any padding specified by the width to be inserted on the right rather than the left.

<

This flag causes the substitution to be truncated on the left to the given width and precision, if specified.

>

This flag causes the substitution to be truncated on the right to the given width, if specified.

^

This flag converts the substituted text to upper case (see Case Conversion in Lisp).

_ (underscore)

This flag converts the substituted text to lower case (see Case Conversion in Lisp).

The result of using contradictory flags (for instance, both upper and lower case) is undefined.

As is the case with format, a format specification can include a width, which is a decimal number that appears after any flags, and a precision, which is a decimal-point ‘.’ followed by a decimal number that appears after any flags and width.

If a substitution contains fewer characters than its specified width, it is padded on the left:

(format-spec "%8a is padded on the left with spaces"
             '((?a . "alpha")))
     ⇒ "   alpha is padded on the left with spaces"

If a substitution contains more characters than its specified precision, it is truncated on the right:

(format-spec "%.2a is truncated on the right"
             '((?a . "alpha")))
     ⇒ "al is truncated on the right"

Here is a more complicated example that combines several aforementioned features:

(setq my-battery-info
      (list (cons ?p "73")      ; Percentage
            (cons ?L "Battery") ; Status
            (cons ?t "2:23")    ; Remaining time
            (cons ?c "24330")   ; Capacity
            (cons ?r "10.6")))  ; Rate of discharge

(format-spec "%>^-3L : %3p%% (%05t left)" my-battery-info)
     ⇒ "BAT :  73% (02:23 left)"

(format-spec "%>^-3L : %3p%% (%05t left)"
             (cons (cons ?L "AC")
                   my-battery-info))
     ⇒ "AC  :  73% (02:23 left)"

As the examples in this section illustrate, format-spec is often used for selectively formatting an assortment of different pieces of information. This is useful in programs that provide user-customizable format strings, as the user can choose to format with a regular syntax and in any desired order only a subset of the information that the program makes available.


4.9 Case Conversion in Lisp

The character case functions change the case of single characters or of the contents of strings. The functions normally convert only alphabetic characters (the letters ‘A’ through ‘Z’ and ‘a’ through ‘z’, as well as non-ASCII letters); other characters are not altered. You can specify a different case conversion mapping by specifying a case table (see The Case Table).

These functions do not modify the strings that are passed to them as arguments.

The examples below use the characters ‘X’ and ‘x’ which have ASCII codes 88 and 120 respectively.

Function: downcase string-or-char

This function converts string-or-char, which should be either a character or a string, to lower case.

When string-or-char is a string, this function returns a new string in which each letter in the argument that is upper case is converted to lower case. When string-or-char is a character, this function returns the corresponding lower case character (an integer); if the original character is lower case, or is not a letter, the return value is equal to the original character.

(downcase "The cat in the hat")
     ⇒ "the cat in the hat"

(downcase ?X)
     ⇒ 120
Function: upcase string-or-char

This function converts string-or-char, which should be either a character or a string, to upper case.

When string-or-char is a string, this function returns a new string in which each letter in the argument that is lower case is converted to upper case. When string-or-char is a character, this function returns the corresponding upper case character (an integer); if the original character is upper case, or is not a letter, the return value is equal to the original character.

(upcase "The cat in the hat")
     ⇒ "THE CAT IN THE HAT"

(upcase ?x)
     ⇒ 88
Function: capitalize string-or-char

This function capitalizes strings or characters. If string-or-char is a string, the function returns a new string whose contents are a copy of string-or-char in which each word has been capitalized. This means that the first character of each word is converted to upper case, and the rest are converted to lower case.

The definition of a word is any sequence of consecutive characters that are assigned to the word constituent syntax class in the current syntax table (see Table of Syntax Classes).

When string-or-char is a character, this function does the same thing as upcase.

(capitalize "The cat in the hat")
     ⇒ "The Cat In The Hat"

(capitalize "THE 77TH-HATTED CAT")
     ⇒ "The 77th-Hatted Cat"

(capitalize ?x)
     ⇒ 88
Function: upcase-initials string-or-char

If string-or-char is a string, this function capitalizes the initials of the words in string-or-char, without altering any letters other than the initials. It returns a new string whose contents are a copy of string-or-char, in which each word has had its initial letter converted to upper case.

The definition of a word is any sequence of consecutive characters that are assigned to the word constituent syntax class in the current syntax table (see Table of Syntax Classes).

When the argument to upcase-initials is a character, upcase-initials has the same result as upcase.

(upcase-initials "The CAT in the hAt")
     ⇒ "The CAT In The HAt"

Note that case conversion is not a one-to-one mapping of codepoints and length of the result may differ from length of the argument. Furthermore, because passing a character forces return type to be a character, functions are unable to perform proper substitution and result may differ compared to treating a one-character string. For example:

(upcase "fi")  ; note: single character, ligature "fi"
     ⇒ "FI"
(upcase ?fi)
     ⇒ 64257  ; i.e. ?fi

To avoid this, a character must first be converted into a string, using string function, before being passed to one of the casing functions. Of course, no assumptions on the length of the result may be made.

Mapping for such special cases are taken from special-uppercase, special-lowercase and special-titlecase See Character Properties.

See Comparison of Characters and Strings, for functions that compare strings; some of them ignore case differences, or can optionally ignore case differences.


4.10 The Case Table

You can customize case conversion by installing a special case table. A case table specifies the mapping between upper case and lower case letters. It affects both the case conversion functions for Lisp objects (see the previous section) and those that apply to text in the buffer (see Case Changes). Each buffer has a case table; there is also a standard case table which is used to initialize the case table of new buffers.

A case table is a char-table (see Char-Tables) whose subtype is case-table. This char-table maps each character into the corresponding lower case character. It has three extra slots, which hold related tables:

upcase

The upcase table maps each character into the corresponding upper case character.

canonicalize

The canonicalize table maps all of a set of case-related characters into a particular member of that set.

equivalences

The equivalences table maps each one of a set of case-related characters into the next character in that set.

In simple cases, all you need to specify is the mapping to lower-case; the three related tables will be calculated automatically from that one.

For some languages, upper and lower case letters are not in one-to-one correspondence. There may be two different lower case letters with the same upper case equivalent. In these cases, you need to specify the maps for both lower case and upper case.

The extra table canonicalize maps each character to a canonical equivalent; any two characters that are related by case-conversion have the same canonical equivalent character. For example, since ‘a’ and ‘A’ are related by case-conversion, they should have the same canonical equivalent character (which should be either ‘a’ for both of them, or ‘A’ for both of them).

The extra table equivalences is a map that cyclically permutes each equivalence class (of characters with the same canonical equivalent). (For ordinary ASCII, this would map ‘a’ into ‘A’ and ‘A’ into ‘a’, and likewise for each set of equivalent characters.)

When constructing a case table, you can provide nil for canonicalize; then Emacs fills in this slot from the lower case and upper case mappings. You can also provide nil for equivalences; then Emacs fills in this slot from canonicalize. In a case table that is actually in use, those components are non-nil. Do not try to specify equivalences without also specifying canonicalize.

Here are the functions for working with case tables:

Function: case-table-p object

This predicate returns non-nil if object is a valid case table.

Function: set-standard-case-table table

This function makes table the standard case table, so that it will be used in any buffers created subsequently.

Function: standard-case-table

This returns the standard case table.

Function: current-case-table

This function returns the current buffer’s case table.

Function: set-case-table table

This sets the current buffer’s case table to table.

Macro: with-case-table table body…

The with-case-table macro saves the current case table, makes table the current case table, evaluates the body forms, and finally restores the case table. The return value is the value of the last form in body. The case table is restored even in case of an abnormal exit via throw or error (see Nonlocal Exits).

Some language environments modify the case conversions of ASCII characters; for example, in the Turkish language environment, the ASCII capital I is downcased into a Turkish dotless i (‘ı’). This can interfere with code that requires ordinary ASCII case conversion, such as implementations of ASCII-based network protocols. In that case, use the with-case-table macro with the variable ascii-case-table, which stores the unmodified case table for the ASCII character set.

Variable: ascii-case-table

The case table for the ASCII character set. This should not be modified by any language environment settings.

The following three functions are convenient subroutines for packages that define non-ASCII character sets. They modify the specified case table case-table; they also modify the standard syntax table. See Syntax Tables. Normally you would use these functions to change the standard case table.

Function: set-case-syntax-pair uc lc case-table

This function specifies a pair of corresponding letters, one upper case and one lower case.

Function: set-case-syntax-delims l r case-table

This function makes characters l and r a matching pair of case-invariant delimiters.

Function: set-case-syntax char syntax case-table

This function makes char case-invariant, with syntax syntax.

Command: describe-buffer-case-table

This command displays a description of the contents of the current buffer’s case table.


5 Lists

A list represents a sequence of zero or more elements (which may be any Lisp objects). The important difference between lists and vectors is that two or more lists can share part of their structure; in addition, you can insert or delete elements in a list without copying the whole list.


5.1 Lists and Cons Cells

Lists in Lisp are not a primitive data type; they are built up from cons cells (see Cons Cell and List Types). A cons cell is a data object that represents an ordered pair. That is, it has two slots, and each slot holds, or refers to, some Lisp object. One slot is known as the CAR, and the other is known as the CDR. (These names are traditional; see Cons Cell and List Types.) CDR is pronounced “could-er”.

We say that “the CAR of this cons cell is” whatever object its CAR slot currently holds, and likewise for the CDR.

A list is a series of cons cells chained together, so that each cell refers to the next one. There is one cons cell for each element of the list. By convention, the CARs of the cons cells hold the elements of the list, and the CDRs are used to chain the list (this asymmetry between CAR and CDR is entirely a matter of convention; at the level of cons cells, the CAR and CDR slots have similar properties). Hence, the CDR slot of each cons cell in a list refers to the following cons cell.

Also by convention, the CDR of the last cons cell in a list is nil. We call such a nil-terminated structure a proper list4. In Emacs Lisp, the symbol nil is both a symbol and a list with no elements. For convenience, the symbol nil is considered to have nil as its CDR (and also as its CAR).

Hence, the CDR of a proper list is always a proper list. The CDR of a nonempty proper list is a proper list containing all the elements except the first.

If the CDR of a list’s last cons cell is some value other than nil, we call the structure a dotted list, since its printed representation would use dotted pair notation (see Dotted Pair Notation). There is one other possibility: some cons cell’s CDR could point to one of the previous cons cells in the list. We call that structure a circular list.

For some purposes, it does not matter whether a list is proper, circular or dotted. If a program doesn’t look far enough down the list to see the CDR of the final cons cell, it won’t care. However, some functions that operate on lists demand proper lists and signal errors if given a dotted list. Most functions that try to find the end of a list enter infinite loops if given a circular list. You can use the function proper-list-p, described in the next section (see proper-list-p), to determine whether a list is a proper one.

Because most cons cells are used as part of lists, we refer to any structure made out of cons cells as a list structure.


5.2 Predicates on Lists

The following predicates test whether a Lisp object is an atom, whether it is a cons cell or is a list, or whether it is the distinguished object nil. (Many of these predicates can be defined in terms of the others, but they are used so often that it is worth having them.)

Function: consp object

This function returns t if object is a cons cell, nil otherwise. nil is not a cons cell, although it is a list.

Function: atom object

This function returns t if object is an atom, nil otherwise. All objects except cons cells are atoms. The symbol nil is an atom and is also a list; it is the only Lisp object that is both.

(atom object) ≡ (not (consp object))
Function: listp object

This function returns t if object is a cons cell or nil. Otherwise, it returns nil.

(listp '(1))
     ⇒ t
(listp '())
     ⇒ t
Function: nlistp object

This function is the opposite of listp: it returns t if object is not a list. Otherwise, it returns nil.

(listp object) ≡ (not (nlistp object))
Function: null object

This function returns t if object is nil, and returns nil otherwise. This function is identical to not, but as a matter of clarity we use null when object is considered a list and not when it is considered a truth value (see not in Constructs for Combining Conditions).

(null '(1))
     ⇒ nil
(null '())
     ⇒ t
Function: proper-list-p object

This function returns the length of object if it is a proper list, nil otherwise (see Lists and Cons Cells). In addition to satisfying listp, a proper list is neither circular nor dotted.

(proper-list-p '(a b c))
    ⇒ 3
(proper-list-p '(a b . c))
    ⇒ nil

5.3 Accessing Elements of Lists

Function: car cons-cell

This function returns the value referred to by the first slot of the cons cell cons-cell. In other words, it returns the CAR of cons-cell.

As a special case, if cons-cell is nil, this function returns nil. Therefore, any list is a valid argument. An error is signaled if the argument is not a cons cell or nil.

(car '(a b c))
     ⇒ a
(car '())
     ⇒ nil
Function: cdr cons-cell

This function returns the value referred to by the second slot of the cons cell cons-cell. In other words, it returns the CDR of cons-cell.

As a special case, if cons-cell is nil, this function returns nil; therefore, any list is a valid argument. An error is signaled if the argument is not a cons cell or nil.

(cdr '(a b c))
     ⇒ (b c)
(cdr '())
     ⇒ nil
Function: car-safe object

This function lets you take the CAR of a cons cell while avoiding errors for other data types. It returns the CAR of object if object is a cons cell, nil otherwise. This is in contrast to car, which signals an error if object is not a list.

(car-safe object)
≡
(let ((x object))
  (if (consp x)
      (car x)
    nil))
Function: cdr-safe object

This function lets you take the CDR of a cons cell while avoiding errors for other data types. It returns the CDR of object if object is a cons cell, nil otherwise. This is in contrast to cdr, which signals an error if object is not a list.

(cdr-safe object)
≡
(let ((x object))
  (if (consp x)
      (cdr x)
    nil))
Macro: pop listname

This macro provides a convenient way to examine the CAR of a list, and take it off the list, all at once. It operates on the list stored in listname. It removes the first element from the list, saves the CDR into listname, then returns the removed element.

In the simplest case, listname is an unquoted symbol naming a list; in that case, this macro is equivalent to (prog1 (car listname) (setq listname (cdr listname))).

x
     ⇒ (a b c)
(pop x)
     ⇒ a
x
     ⇒ (b c)

More generally, listname can be a generalized variable. In that case, this macro saves into listname using setf. See Generalized Variables.

For the push macro, which adds an element to a list, See Modifying List Variables.

Function: nth n list

This function returns the nth element of list. Elements are numbered starting with zero, so the CAR of list is element number zero. If the length of list is n or less, the value is nil.

(nth 2 '(1 2 3 4))
     ⇒ 3
(nth 10 '(1 2 3 4))
     ⇒ nil

(nth n x) ≡ (car (nthcdr n x))

The function elt is similar, but applies to any kind of sequence. For historical reasons, it takes its arguments in the opposite order. See Sequences.

Function: nthcdr n list

This function returns the nth CDR of list. In other words, it skips past the first n links of list and returns what follows.

If n is zero, nthcdr returns all of list. If the length of list is n or less, nthcdr returns nil.

(nthcdr 1 '(1 2 3 4))
     ⇒ (2 3 4)
(nthcdr 10 '(1 2 3 4))
     ⇒ nil
(nthcdr 0 '(1 2 3 4))
     ⇒ (1 2 3 4)
Function: take n list

This function returns the n first elements of list. Essentially, it returns the part of list that nthcdr skips.

take returns list if shorter than n elements; it returns nil if n is zero or negative.

(take 3 '(a b c d))
     ⇒ (a b c)
(take 10 '(a b c d))
     ⇒ (a b c d)
(take 0 '(a b c d))
     ⇒ nil
Function: ntake n list

This is a version of take that works by destructively modifying the list structure of the argument. That makes it faster, but the original value of list may be lost.

ntake returns list unmodified if shorter than n elements; it returns nil if n is zero or negative. Otherwise, it returns list truncated to its first n elements.

This means that it is usually a good idea to use the return value and not just rely on the truncation effect unless n is known to be positive.

Function: last list &optional n

This function returns the last link of list. The car of this link is the list’s last element. If list is null, nil is returned. If n is non-nil, the nth-to-last link is returned instead, or the whole of list if n is bigger than list’s length.

Function: safe-length list

This function returns the length of list, with no risk of either an error or an infinite loop. It generally returns the number of distinct cons cells in the list. However, for circular lists, the value is just an upper bound; it is often too large.

If list is not nil or a cons cell, safe-length returns 0.

The most common way to compute the length of a list, when you are not worried that it may be circular, is with length. See Sequences.

Function: caar cons-cell

This is the same as (car (car cons-cell)).

Function: cadr cons-cell

This is the same as (car (cdr cons-cell)) or (nth 1 cons-cell).

Function: cdar cons-cell

This is the same as (cdr (car cons-cell)).

Function: cddr cons-cell

This is the same as (cdr (cdr cons-cell)) or (nthcdr 2 cons-cell).

In addition to the above, 24 additional compositions of car and cdr are defined as cxxxr and cxxxxr, where each x is either a or d. cadr, caddr, and cadddr pick out the second, third or fourth elements of a list, respectively. cl-lib provides the same under the names cl-second, cl-third, and cl-fourth. See List Functions in Common Lisp Extensions.

Function: butlast x &optional n

This function returns the list x with the last element, or the last n elements, removed. If n is greater than zero it makes a copy of the list so as not to damage the original list. In general, (append (butlast x n) (last x n)) will return a list equal to x.

Function: nbutlast x &optional n

This is a version of butlast that works by destructively modifying the cdr of the appropriate element, rather than making a copy of the list.


5.4 Building Cons Cells and Lists

Many functions build lists, as lists reside at the very heart of Lisp. cons is the fundamental list-building function; however, it is interesting to note that list is used more times in the source code for Emacs than cons.

Function: cons object1 object2

This function is the most basic function for building new list structure. It creates a new cons cell, making object1 the CAR, and object2 the CDR. It then returns the new cons cell. The arguments object1 and object2 may be any Lisp objects, but most often object2 is a list.

(cons 1 '(2))
     ⇒ (1 2)
(cons 1 '())
     ⇒ (1)
(cons 1 2)
     ⇒ (1 . 2)

cons is often used to add a single element to the front of a list. This is called consing the element onto the list. 5 For example:

(setq list (cons newelt list))

Note that there is no conflict between the variable named list used in this example and the function named list described below; any symbol can serve both purposes.

Function: list &rest objects

This function creates a list with objects as its elements. The resulting list is always nil-terminated. If no objects are given, the empty list is returned.

(list 1 2 3 4 5)
     ⇒ (1 2 3 4 5)
(list 1 2 '(3 4 5) 'foo)
     ⇒ (1 2 (3 4 5) foo)
(list)
     ⇒ nil
Function: make-list length object

This function creates a list of length elements, in which each element is object. Compare make-list with make-string (see Creating Strings).

(make-list 3 'pigs)
     ⇒ (pigs pigs pigs)
(make-list 0 'pigs)
     ⇒ nil
(setq l (make-list 3 '(a b)))
     ⇒ ((a b) (a b) (a b))
(eq (car l) (cadr l))
     ⇒ t
Function: append &rest sequences

This function returns a list containing all the elements of sequences. The sequences may be lists, vectors, bool-vectors, or strings, but the last one should usually be a list. All arguments except the last one are copied, so none of the arguments is altered. (See nconc in Functions that Rearrange Lists, for a way to join lists with no copying.)

More generally, the final argument to append may be any Lisp object. The final argument is not copied or converted; it becomes the CDR of the last cons cell in the new list. If the final argument is itself a list, then its elements become in effect elements of the result list. If the final element is not a list, the result is a dotted list since its final CDR is not nil as required in a proper list (see Lists and Cons Cells).

Here is an example of using append:

(setq trees '(pine oak))
     ⇒ (pine oak)
(setq more-trees (append '(maple birch) trees))
     ⇒ (maple birch pine oak)

trees
     ⇒ (pine oak)
more-trees
     ⇒ (maple birch pine oak)
(eq trees (cdr (cdr more-trees)))
     ⇒ t

You can see how append works by looking at a box diagram. The variable trees is set to the list (pine oak) and then the variable more-trees is set to the list (maple birch pine oak). However, the variable trees continues to refer to the original list:

more-trees                trees
|                           |
|     --- ---      --- ---   -> --- ---      --- ---
 --> |   |   |--> |   |   |--> |   |   |--> |   |   |--> nil
      --- ---      --- ---      --- ---      --- ---
       |            |            |            |
       |            |            |            |
        --> maple    -->birch     --> pine     --> oak

An empty sequence contributes nothing to the value returned by append. As a consequence of this, a final nil argument forces a copy of the previous argument:

trees
     ⇒ (pine oak)
(setq wood (append trees nil))
     ⇒ (pine oak)
wood
     ⇒ (pine oak)
(eq wood trees)
     ⇒ nil

This once was the usual way to copy a list, before the function copy-sequence was invented. See Sequences, Arrays, and Vectors.

Here we show the use of vectors and strings as arguments to append:

(append [a b] "cd" nil)
     ⇒ (a b 99 100)

With the help of apply (see Calling Functions), we can append all the lists in a list of lists:

(apply 'append '((a b c) nil (x y z) nil))
     ⇒ (a b c x y z)

If no sequences are given, nil is returned:

(append)
     ⇒ nil

Here are some examples where the final argument is not a list:

(append '(x y) 'z)
     ⇒ (x y . z)
(append '(x y) [z])
     ⇒ (x y . [z])

The second example shows that when the final argument is a sequence but not a list, the sequence’s elements do not become elements of the resulting list. Instead, the sequence becomes the final CDR, like any other non-list final argument.

Function: copy-tree tree &optional vecp

This function returns a copy of the tree tree. If tree is a cons cell, this makes a new cons cell with the same CAR and CDR, then recursively copies the CAR and CDR in the same way.

Normally, when tree is anything other than a cons cell, copy-tree simply returns tree. However, if vecp is non-nil, it copies vectors too (and operates recursively on their elements).

Function: flatten-tree tree

This function returns a “flattened” copy of tree, that is, a list containing all the non-nil terminal nodes, or leaves, of the tree of cons cells rooted at tree. Leaves in the returned list are in the same order as in tree.

(flatten-tree '(1 (2 . 3) nil (4 5 (6)) 7))
    ⇒(1 2 3 4 5 6 7)
Function: ensure-list object

This function returns object as a list. If object is already a list, the function returns it; otherwise, the function returns a one-element list containing object.

This is usually useful if you have a variable that may or may not be a list, and you can then say, for instance:

(dolist (elem (ensure-list foo))
  (princ elem))
Function: number-sequence from &optional to separation

This function returns a list of numbers starting with from and incrementing by separation, and ending at or just before to. separation can be positive or negative and defaults to 1. If to is nil or numerically equal to from, the value is the one-element list (from). If to is less than from with a positive separation, or greater than from with a negative separation, the value is nil because those arguments specify an empty sequence.

If separation is 0 and to is neither nil nor numerically equal to from, number-sequence signals an error, since those arguments specify an infinite sequence.

All arguments are numbers. Floating-point arguments can be tricky, because floating-point arithmetic is inexact. For instance, depending on the machine, it may quite well happen that (number-sequence 0.4 0.6 0.2) returns the one element list (0.4), whereas (number-sequence 0.4 0.8 0.2) returns a list with three elements. The nth element of the list is computed by the exact formula (+ from (* n separation)). Thus, if one wants to make sure that to is included in the list, one can pass an expression of this exact type for to. Alternatively, one can replace to with a slightly larger value (or a slightly more negative value if separation is negative).

Some examples:

(number-sequence 4 9)
     ⇒ (4 5 6 7 8 9)
(number-sequence 9 4 -1)
     ⇒ (9 8 7 6 5 4)
(number-sequence 9 4 -2)
     ⇒ (9 7 5)
(number-sequence 8)
     ⇒ (8)
(number-sequence 8 5)
     ⇒ nil
(number-sequence 5 8 -1)
     ⇒ nil
(number-sequence 1.5 6 2)
     ⇒ (1.5 3.5 5.5)

5.5 Modifying List Variables

These functions, and one macro, provide convenient ways to modify a list which is stored in a variable.

Macro: push element listname

This macro creates a new list whose CAR is element and whose CDR is the list specified by listname, and saves that list in listname. In the simplest case, listname is an unquoted symbol naming a list, and this macro is equivalent to (setq listname (cons element listname)).

(setq l '(a b))
     ⇒ (a b)
(push 'c l)
     ⇒ (c a b)
l
     ⇒ (c a b)

More generally, listname can be a generalized variable. In that case, this macro does the equivalent of (setf listname (cons element listname)). See Generalized Variables.

For the pop macro, which removes the first element from a list, See Accessing Elements of Lists.

Two functions modify lists that are the values of variables.

Function: add-to-list symbol element &optional append compare-fn

This function sets the variable symbol by consing element onto the old value, if element is not already a member of that value. It returns the resulting list, whether updated or not. The value of symbol had better be a list already before the call. add-to-list uses compare-fn to compare element against existing list members; if compare-fn is nil, it uses equal.

Normally, if element is added, it is added to the front of symbol, but if the optional argument append is non-nil, it is added at the end.

The argument symbol is not implicitly quoted; add-to-list is an ordinary function, like set and unlike setq. Quote the argument yourself if that is what you want.

Do not use this function when symbol refers to a lexical variable.

Here’s a scenario showing how to use add-to-list:

(setq foo '(a b))
     ⇒ (a b)

(add-to-list 'foo 'c)     ;; Add c.
     ⇒ (c a b)

(add-to-list 'foo 'b)     ;; No effect.
     ⇒ (c a b)

foo                       ;; foo was changed.
     ⇒ (c a b)

An equivalent expression for (add-to-list 'var value) is this:

(if (member value var)
    var
  (setq var (cons value var)))
Function: add-to-ordered-list symbol element &optional order

This function sets the variable symbol by inserting element into the old value, which must be a list, at the position specified by order. If element is already a member of the list, its position in the list is adjusted according to order. Membership is tested using eq. This function returns the resulting list, whether updated or not.

The order is typically a number (integer or float), and the elements of the list are sorted in non-decreasing numerical order.

order may also be omitted or nil. Then the numeric order of element stays unchanged if it already has one; otherwise, element has no numeric order. Elements without a numeric list order are placed at the end of the list, in no particular order.

Any other value for order removes the numeric order of element if it already has one; otherwise, it is equivalent to nil.

The argument symbol is not implicitly quoted; add-to-ordered-list is an ordinary function, like set and unlike setq. Quote the argument yourself if necessary.

The ordering information is stored in a hash table on symbol’s list-order property. symbol cannot refer to a lexical variable.

Here’s a scenario showing how to use add-to-ordered-list:

(setq foo '())
     ⇒ nil

(add-to-ordered-list 'foo 'a 1)     ;; Add a.
     ⇒ (a)

(add-to-ordered-list 'foo 'c 3)     ;; Add c.
     ⇒ (a c)

(add-to-ordered-list 'foo 'b 2)     ;; Add b.
     ⇒ (a b c)

(add-to-ordered-list 'foo 'b 4)     ;; Move b.
     ⇒ (a c b)

(add-to-ordered-list 'foo 'd)       ;; Append d.
     ⇒ (a c b d)

(add-to-ordered-list 'foo 'e)       ;; Add e.
     ⇒ (a c b e d)

foo                       ;; foo was changed.
     ⇒ (a c b e d)

5.6 Modifying Existing List Structure

You can modify the CAR and CDR contents of a cons cell with the primitives setcar and setcdr. These are destructive operations because they change existing list structure. Destructive operations should be applied only to mutable lists, that is, lists constructed via cons, list or similar operations. Lists created by quoting are part of the program and should not be changed by destructive operations. See Mutability.

Common Lisp note: Common Lisp uses functions rplaca and rplacd to alter list structure; they change structure the same way as setcar and setcdr, but the Common Lisp functions return the cons cell while setcar and setcdr return the new CAR or CDR.


5.6.1 Altering List Elements with setcar

Changing the CAR of a cons cell is done with setcar. When used on a list, setcar replaces one element of a list with a different element.

Function: setcar cons object

This function stores object as the new CAR of cons, replacing its previous CAR. In other words, it changes the CAR slot of cons to refer to object. It returns the value object. For example:

(setq x (list 1 2))
     ⇒ (1 2)
(setcar x 4)
     ⇒ 4
x
     ⇒ (4 2)

When a cons cell is part of the shared structure of several lists, storing a new CAR into the cons changes one element of each of these lists. Here is an example:

;; Create two lists that are partly shared.
(setq x1 (list 'a 'b 'c))
     ⇒ (a b c)
(setq x2 (cons 'z (cdr x1)))
     ⇒ (z b c)

;; Replace the CAR of a shared link.
(setcar (cdr x1) 'foo)
     ⇒ foo
x1                           ; Both lists are changed.
     ⇒ (a foo c)
x2
     ⇒ (z foo c)

;; Replace the CAR of a link that is not shared.
(setcar x1 'baz)
     ⇒ baz
x1                           ; Only one list is changed.
     ⇒ (baz foo c)
x2
     ⇒ (z foo c)

Here is a graphical depiction of the shared structure of the two lists in the variables x1 and x2, showing why replacing b changes them both:

        --- ---        --- ---      --- ---
x1---> |   |   |----> |   |   |--> |   |   |--> nil
        --- ---        --- ---      --- ---
         |        -->   |            |
         |       |      |            |
          --> a  |       --> b        --> c
                 |
       --- ---   |
x2--> |   |   |--
       --- ---
        |
        |
         --> z

Here is an alternative form of box diagram, showing the same relationship:

x1:
 --------------       --------------       --------------
| car   | cdr  |     | car   | cdr  |     | car   | cdr  |
|   a   |   o------->|   b   |   o------->|   c   |  nil |
|       |      |  -->|       |      |     |       |      |
 --------------  |    --------------       --------------
                 |
x2:              |
 --------------  |
| car   | cdr  | |
|   z   |   o----
|       |      |
 --------------

5.6.2 Altering the CDR of a List

The lowest-level primitive for modifying a CDR is setcdr:

Function: setcdr cons object

This function stores object as the new CDR of cons, replacing its previous CDR. In other words, it changes the CDR slot of cons to refer to object. It returns the value object.

Here is an example of replacing the CDR of a list with a different list. All but the first element of the list are removed in favor of a different sequence of elements. The first element is unchanged, because it resides in the CAR of the list, and is not reached via the CDR.

(setq x (list 1 2 3))
     ⇒ (1 2 3)
(setcdr x '(4))
     ⇒ (4)
x
     ⇒ (1 4)

You can delete elements from the middle of a list by altering the CDRs of the cons cells in the list. For example, here we delete the second element, b, from the list (a b c), by changing the CDR of the first cons cell:

(setq x1 (list 'a 'b 'c))
     ⇒ (a b c)
(setcdr x1 (cdr (cdr x1)))
     ⇒ (c)
x1
     ⇒ (a c)

Here is the result in box notation:

                   --------------------
                  |                    |
 --------------   |   --------------   |    --------------
| car   | cdr  |  |  | car   | cdr  |   -->| car   | cdr  |
|   a   |   o-----   |   b   |   o-------->|   c   |  nil |
|       |      |     |       |      |      |       |      |
 --------------       --------------        --------------

The second cons cell, which previously held the element b, still exists and its CAR is still b, but it no longer forms part of this list.

It is equally easy to insert a new element by changing CDRs:

(setq x1 (list 'a 'b 'c))
     ⇒ (a b c)
(setcdr x1 (cons 'd (cdr x1)))
     ⇒ (d b c)
x1
     ⇒ (a d b c)

Here is this result in box notation:

 --------------        -------------       -------------
| car  | cdr   |      | car  | cdr  |     | car  | cdr  |
|   a  |   o   |   -->|   b  |   o------->|   c  |  nil |
|      |   |   |  |   |      |      |     |      |      |
 --------- | --   |    -------------       -------------
           |      |
     -----         --------
    |                      |
    |    ---------------   |
    |   | car   | cdr   |  |
     -->|   d   |   o------
        |       |       |
         ---------------

5.6.3 Functions that Rearrange Lists

Here are some functions that rearrange lists destructively by modifying the CDRs of their component cons cells. These functions are destructive because they chew up the original lists passed to them as arguments, relinking their cons cells to form a new list that is the returned value.

See delq, in Using Lists as Sets, for another function that modifies cons cells.

Function: nconc &rest lists

This function returns a list containing all the elements of lists. Unlike append (see Building Cons Cells and Lists), the lists are not copied. Instead, the last CDR of each of the lists is changed to refer to the following list. The last of the lists is not altered. For example:

(setq x (list 1 2 3))
     ⇒ (1 2 3)
(nconc x '(4 5))
     ⇒ (1 2 3 4 5)
x
     ⇒ (1 2 3 4 5)

Since the last argument of nconc is not itself modified, it is reasonable to use a constant list, such as '(4 5), as in the above example. For the same reason, the last argument need not be a list:

(setq x (list 1 2 3))
     ⇒ (1 2 3)
(nconc x 'z)
     ⇒ (1 2 3 . z)
x
     ⇒ (1 2 3 . z)

However, the other arguments (all but the last) should be mutable lists.

A common pitfall is to use a constant list as a non-last argument to nconc. If you do this, the resulting behavior is undefined (see Self-Evaluating Forms). It is possible that your program will change each time you run it! Here is what might happen (though this is not guaranteed to happen):

(defun add-foo (x)            ; We want this function to add
  (nconc '(foo) x))           ;   foo to the front of its arg.

(symbol-function 'add-foo)
     ⇒ (lambda (x) (nconc '(foo) x))

(setq xx (add-foo '(1 2)))    ; It seems to work.
     ⇒ (foo 1 2)
(setq xy (add-foo '(3 4)))    ; What happened?
     ⇒ (foo 1 2 3 4)
(eq xx xy)
     ⇒ t

(symbol-function 'add-foo)
     ⇒ (lambda (x) (nconc '(foo 1 2 3 4) x))

5.7 Using Lists as Sets

A list can represent an unordered mathematical set—simply consider a value an element of a set if it appears in the list, and ignore the order of the list. To form the union of two sets, use append (as long as you don’t mind having duplicate elements). You can remove equal duplicates using delete-dups or seq-uniq. Other useful functions for sets include memq and delq, and their equal versions, member and delete.

Common Lisp note: Common Lisp has functions union (which avoids duplicate elements) and intersection for set operations. In Emacs Lisp, variants of these facilities are provided by the cl-lib library. See Lists as Sets in Common Lisp Extensions.

Function: memq object list

This function tests to see whether object is a member of list. If it is, memq returns a list starting with the first occurrence of object. Otherwise, it returns nil. The letter ‘q’ in memq says that it uses eq to compare object against the elements of the list. For example:

(memq 'b '(a b c b a))
     ⇒ (b c b a)
(memq '(2) '((1) (2)))    ; The two (2)s need not be eq.Unspecified; might be nil or ((2)).
Function: delq object list

This function destructively removes all elements eq to object from list, and returns the resulting list. The letter ‘q’ in delq says that it uses eq to compare object against the elements of the list, like memq and remq.

Typically, when you invoke delq, you should use the return value by assigning it to the variable which held the original list. The reason for this is explained below.

The delq function deletes elements from the front of the list by simply advancing down the list, and returning a sublist that starts after those elements. For example:

(delq 'a '(a b c)) ≡ (cdr '(a b c))

When an element to be deleted appears in the middle of the list, removing it involves changing the CDRs (see Altering the CDR of a List).

(setq sample-list (list 'a 'b 'c '(4)))
     ⇒ (a b c (4))
(delq 'a sample-list)
     ⇒ (b c (4))
sample-list
     ⇒ (a b c (4))
(delq 'c sample-list)
     ⇒ (a b (4))
sample-list
     ⇒ (a b (4))

Note that (delq 'c sample-list) modifies sample-list to splice out the third element, but (delq 'a sample-list) does not splice anything—it just returns a shorter list. Don’t assume that a variable which formerly held the argument list now has fewer elements, or that it still holds the original list! Instead, save the result of delq and use that. Most often we store the result back into the variable that held the original list:

(setq flowers (delq 'rose flowers))

In the following example, the (list 4) that delq attempts to match and the (4) in the sample-list are equal but not eq:

(delq (list 4) sample-list)
     ⇒ (a c (4))

If you want to delete elements that are equal to a given value, use delete (see below).

Function: remq object list

This function returns a copy of list, with all elements removed which are eq to object. The letter ‘q’ in remq says that it uses eq to compare object against the elements of list.

(setq sample-list (list 'a 'b 'c 'a 'b 'c))
     ⇒ (a b c a b c)
(remq 'a sample-list)
     ⇒ (b c b c)
sample-list
     ⇒ (a b c a b c)
Function: memql object list

The function memql tests to see whether object is a member of list, comparing members with object using eql, so floating-point elements are compared by value. If object is a member, memql returns a list starting with its first occurrence in list. Otherwise, it returns nil.

Compare this with memq:

(memql 1.2 '(1.1 1.2 1.3))  ; 1.2 and 1.2 are eql.
     ⇒ (1.2 1.3)
(memq 1.2 '(1.1 1.2 1.3))  ; The two 1.2s need not be eq.Unspecified; might be nil or (1.2 1.3).

The following three functions are like memq, delq and remq, but use equal rather than eq to compare elements. See Equality Predicates.

Function: member object list

The function member tests to see whether object is a member of list, comparing members with object using equal. If object is a member, member returns a list starting with its first occurrence in list. Otherwise, it returns nil.

Compare this with memq:

(member '(2) '((1) (2)))  ; (2) and (2) are equal.
     ⇒ ((2))
(memq '(2) '((1) (2)))    ; The two (2)s need not be eq.Unspecified; might be nil or (2).
;; Two strings with the same contents are equal.
(member "foo" '("foo" "bar"))
     ⇒ ("foo" "bar")
Function: delete object sequence

This function removes all elements equal to object from sequence, and returns the resulting sequence.

If sequence is a list, delete is to delq as member is to memq: it uses equal to compare elements with object, like member; when it finds an element that matches, it cuts the element out just as delq would. As with delq, you should typically use the return value by assigning it to the variable which held the original list.

If sequence is a vector or string, delete returns a copy of sequence with all elements equal to object removed.

For example:

(setq l (list '(2) '(1) '(2)))
(delete '(2) l)
     ⇒ ((1))
l
     ⇒ ((2) (1))
;; If you want to change l reliably,
;; write (setq l (delete '(2) l)).
(setq l (list '(2) '(1) '(2)))
(delete '(1) l)
     ⇒ ((2) (2))
l
     ⇒ ((2) (2))
;; In this case, it makes no difference whether you set l,
;; but you should do so for the sake of the other case.
(delete '(2) [(2) (1) (2)])
     ⇒ [(1)]
Function: remove object sequence

This function is the non-destructive counterpart of delete. It returns a copy of sequence, a list, vector, or string, with elements equal to object removed. For example:

(remove '(2) '((2) (1) (2)))
     ⇒ ((1))
(remove '(2) [(2) (1) (2)])
     ⇒ [(1)]

Common Lisp note: The functions member, delete and remove in GNU Emacs Lisp are derived from Maclisp, not Common Lisp. The Common Lisp versions do not use equal to compare elements.

Function: member-ignore-case object list

This function is like member, except that object should be a string and that it ignores differences in letter-case and text representation: upper-case and lower-case letters are treated as equal, and unibyte strings are converted to multibyte prior to comparison.

Function: delete-dups list

This function destructively removes all equal duplicates from list, stores the result in list and returns it. Of several equal occurrences of an element in list, delete-dups keeps the first one. See seq-uniq for non-destructive operation (see Sequences).

See also the function add-to-list, in Modifying List Variables, for a way to add an element to a list stored in a variable and used as a set.


5.8 Association Lists

An association list, or alist for short, records a mapping from keys to values. It is a list of cons cells called associations: the CAR of each cons cell is the key, and the CDR is the associated value.6

Here is an example of an alist. The key pine is associated with the value cones; the key oak is associated with acorns; and the key maple is associated with seeds.

((pine . cones)
 (oak . acorns)
 (maple . seeds))

Both the values and the keys in an alist may be any Lisp objects. For example, in the following alist, the symbol a is associated with the number 1, and the string "b" is associated with the list (2 3), which is the CDR of the alist element:

((a . 1) ("b" 2 3))

Sometimes it is better to design an alist to store the associated value in the CAR of the CDR of the element. Here is an example of such an alist:

((rose red) (lily white) (buttercup yellow))

Here we regard red as the value associated with rose. One advantage of this kind of alist is that you can store other related information—even a list of other items—in the CDR of the CDR. One disadvantage is that you cannot use rassq (see below) to find the element containing a given value. When neither of these considerations is important, the choice is a matter of taste, as long as you are consistent about it for any given alist.

The same alist shown above could be regarded as having the associated value in the CDR of the element; the value associated with rose would be the list (red).

Association lists are often used to record information that you might otherwise keep on a stack, since new associations may be added easily to the front of the list. When searching an association list for an association with a given key, the first one found is returned, if there is more than one.

In Emacs Lisp, it is not an error if an element of an association list is not a cons cell. The alist search functions simply ignore such elements. Many other versions of Lisp signal errors in such cases.

Note that property lists are similar to association lists in several respects. A property list behaves like an association list in which each key can occur only once. See Property Lists, for a comparison of property lists and association lists.

Function: assoc key alist &optional testfn

This function returns the first association for key in alist, comparing key against the alist elements using testfn if it is a function, and equal otherwise (see Equality Predicates). If testfn is a function, it is called with two arguments: the CAR of an element from alist and key. The function returns nil if no association in alist has a CAR equal to key, as tested by testfn. For example:

(setq trees '((pine . cones) (oak . acorns) (maple . seeds)))
     ⇒ ((pine . cones) (oak . acorns) (maple . seeds))
(assoc 'oak trees)
     ⇒ (oak . acorns)
(cdr (assoc 'oak trees))
     ⇒ acorns
(assoc 'birch trees)
     ⇒ nil

Here is another example, in which the keys and values are not symbols:

(setq needles-per-cluster
      '((2 "Austrian Pine" "Red Pine")
        (3 "Pitch Pine")
        (5 "White Pine")))

(cdr (assoc 3 needles-per-cluster))
     ⇒ ("Pitch Pine")
(cdr (assoc 2 needles-per-cluster))
     ⇒ ("Austrian Pine" "Red Pine")

The function assoc-string is much like assoc except that it ignores certain differences between strings. See Comparison of Characters and Strings.

Function: rassoc value alist

This function returns the first association with value value in alist. It returns nil if no association in alist has a CDR equal to value.

rassoc is like assoc except that it compares the CDR of each alist association instead of the CAR. You can think of this as reverse assoc, finding the key for a given value.

Function: assq key alist

This function is like assoc in that it returns the first association for key in alist, but it makes the comparison using eq. assq returns nil if no association in alist has a CAR eq to key. This function is used more often than assoc, since eq is faster than equal and most alists use symbols as keys. See Equality Predicates.

(setq trees '((pine . cones) (oak . acorns) (maple . seeds)))
     ⇒ ((pine . cones) (oak . acorns) (maple . seeds))
(assq 'pine trees)
     ⇒ (pine . cones)

On the other hand, assq is not usually useful in alists where the keys may not be symbols:

(setq leaves
      '(("simple leaves" . oak)
        ("compound leaves" . horsechestnut)))

(assq "simple leaves" leaves)
     ⇒ Unspecified; might be nil or ("simple leaves" . oak).
(assoc "simple leaves" leaves)
     ⇒ ("simple leaves" . oak)
Function: alist-get key alist &optional default remove testfn

This function is similar to assq. It finds the first association (key . value) by comparing key with alist elements, and, if found, returns the value of that association. If no association is found, the function returns default. Comparison of key against alist elements uses the function specified by testfn, defaulting to eq.

This is a generalized variable (see Generalized Variables) that can be used to change a value with setf. When using it to set a value, optional argument remove non-nil means to remove key’s association from alist if the new value is eql to default.

Function: rassq value alist

This function returns the first association with value value in alist. It returns nil if no association in alist has a CDR eq to value.

rassq is like assq except that it compares the CDR of each alist association instead of the CAR. You can think of this as reverse assq, finding the key for a given value.

For example:

(setq trees '((pine . cones) (oak . acorns) (maple . seeds)))

(rassq 'acorns trees)
     ⇒ (oak . acorns)
(rassq 'spores trees)
     ⇒ nil

rassq cannot search for a value stored in the CAR of the CDR of an element:

(setq colors '((rose red) (lily white) (buttercup yellow)))

(rassq 'white colors)
     ⇒ nil

In this case, the CDR of the association (lily white) is not the symbol white, but rather the list (white). This becomes clearer if the association is written in dotted pair notation:

(lily white) ≡ (lily . (white))
Function: assoc-default key alist &optional test default

This function searches alist for a match for key. For each element of alist, it compares the element (if it is an atom) or the element’s CAR (if it is a cons) against key, by calling test with two arguments: the element or its CAR, and key. The arguments are passed in that order so that you can get useful results using string-match with an alist that contains regular expressions (see Regular Expression Searching). If test is omitted or nil, equal is used for comparison.

If an alist element matches key by this criterion, then assoc-default returns a value based on this element. If the element is a cons, then the value is the element’s CDR. Otherwise, the return value is default.

If no alist element matches key, assoc-default returns nil.

Function: copy-alist alist

This function returns a two-level deep copy of alist: it creates a new copy of each association, so that you can alter the associations of the new alist without changing the old one.

(setq needles-per-cluster
      '((2 . ("Austrian Pine" "Red Pine"))
        (3 . ("Pitch Pine"))
        (5 . ("White Pine"))))
⇒
((2 "Austrian Pine" "Red Pine")
 (3 "Pitch Pine")
 (5 "White Pine"))

(setq copy (copy-alist needles-per-cluster))
⇒
((2 "Austrian Pine" "Red Pine")
 (3 "Pitch Pine")
 (5 "White Pine"))

(eq needles-per-cluster copy)
     ⇒ nil
(equal needles-per-cluster copy)
     ⇒ t
(eq (car needles-per-cluster) (car copy))
     ⇒ nil
(cdr (car (cdr needles-per-cluster)))
     ⇒ ("Pitch Pine")
(eq (cdr (car (cdr needles-per-cluster)))
    (cdr (car (cdr copy))))
     ⇒ t

This example shows how copy-alist makes it possible to change the associations of one copy without affecting the other:

(setcdr (assq 3 copy) '("Martian Vacuum Pine"))
(cdr (assq 3 needles-per-cluster))
     ⇒ ("Pitch Pine")
Function: assq-delete-all key alist

This function deletes from alist all the elements whose CAR is eq to key, much as if you used delq to delete each such element one by one. It returns the shortened alist, and often modifies the original list structure of alist. For correct results, use the return value of assq-delete-all rather than looking at the saved value of alist.

(setq alist (list '(foo 1) '(bar 2) '(foo 3) '(lose 4)))
     ⇒ ((foo 1) (bar 2) (foo 3) (lose 4))
(assq-delete-all 'foo alist)
     ⇒ ((bar 2) (lose 4))
alist
     ⇒ ((foo 1) (bar 2) (lose 4))
Function: assoc-delete-all key alist &optional test

This function is like assq-delete-all except that it accepts an optional argument test, a predicate function to compare the keys in alist. If omitted or nil, test defaults to equal. As assq-delete-all, this function often modifies the original list structure of alist.

Function: rassq-delete-all value alist

This function deletes from alist all the elements whose CDR is eq to value. It returns the shortened alist, and often modifies the original list structure of alist. rassq-delete-all is like assq-delete-all except that it compares the CDR of each alist association instead of the CAR.

Macro: let-alist alist body

Creates a binding for each symbol used as keys the association list alist, prefixed with dot. This can be useful when accessing several items in the same association list, and it’s best understood through a simple example:

(setq colors '((rose . red) (lily . white) (buttercup . yellow)))
(let-alist colors
  (if (eq .rose 'red)
      .lily))
     ⇒ white

The body is inspected at compilation time, and only the symbols that appear in body with a ‘.’ as the first character in the symbol name will be bound. Finding the keys is done with assq, and the cdr of the return value of this assq is assigned as the value for the binding.

Nested association lists is supported:

(setq colors '((rose . red) (lily (belladonna . yellow) (brindisi . pink))))
(let-alist colors
  (if (eq .rose 'red)
      .lily.belladonna))
     ⇒ yellow

Nesting let-alist inside each other is allowed, but the code in the inner let-alist can’t access the variables bound by the outer let-alist.


5.9 Property Lists

A property list (plist for short) is a list of paired elements. Each of the pairs associates a property name (usually a symbol) with a property or value. Here is an example of a property list:

(pine cones numbers (1 2 3) color "blue")

This property list associates pine with cones, numbers with (1 2 3), and color with "blue". The property names and values can be any Lisp objects, but the names are usually symbols (as they are in this example).

Property lists are used in several contexts. For instance, the function put-text-property takes an argument which is a property list, specifying text properties and associated values which are to be applied to text in a string or buffer. See Text Properties.

Another prominent use of property lists is for storing symbol properties. Every symbol possesses a list of properties, used to record miscellaneous information about the symbol; these properties are stored in the form of a property list. See Symbol Properties.

Function: plistp object

This predicate function returns non-nil if object is a valid property list.


5.9.1 Property Lists and Association Lists

Association lists (see Association Lists) are very similar to property lists. In contrast to association lists, the order of the pairs in the property list is not significant, since the property names must be distinct.

Property lists are better than association lists for attaching information to various Lisp function names or variables. If your program keeps all such information in one association list, it will typically need to search that entire list each time it checks for an association for a particular Lisp function name or variable, which could be slow. By contrast, if you keep the same information in the property lists of the function names or variables themselves, each search will scan only the length of one property list, which is usually short. This is why the documentation for a variable is recorded in a property named variable-documentation. The byte compiler likewise uses properties to record those functions needing special treatment.

However, association lists have their own advantages. Depending on your application, it may be faster to add an association to the front of an association list than to update a property. All properties for a symbol are stored in the same property list, so there is a possibility of a conflict between different uses of a property name. (For this reason, it is a good idea to choose property names that are probably unique, such as by beginning the property name with the program’s usual name-prefix for variables and functions.) An association list may be used like a stack where associations are pushed on the front of the list and later discarded; this is not possible with a property list.


5.9.2 Property Lists Outside Symbols

The following functions can be used to manipulate property lists. They all default to comparing property names using eq.

Function: plist-get plist property &optional predicate

This returns the value of the property property stored in the property list plist. Comparisons are done with predicate, which defaults to eq. It accepts a malformed plist argument. If property is not found in the plist, it returns nil. For example,

(plist-get '(foo 4) 'foo)
     ⇒ 4
(plist-get '(foo 4 bad) 'foo)
     ⇒ 4
(plist-get '(foo 4 bad) 'bad)
     ⇒ nil
(plist-get '(foo 4 bad) 'bar)
     ⇒ nil
Function: plist-put plist property value &optional predicate

This stores value as the value of the property property in the property list plist. Comparisons are done with predicate, which defaults to eq. It may modify plist destructively, or it may construct a new list structure without altering the old. The function returns the modified property list, so you can store that back in the place where you got plist. For example,

(setq my-plist (list 'bar t 'foo 4))
     ⇒ (bar t foo 4)
(setq my-plist (plist-put my-plist 'foo 69))
     ⇒ (bar t foo 69)
(setq my-plist (plist-put my-plist 'quux '(a)))
     ⇒ (bar t foo 69 quux (a))
Function: lax-plist-get plist property

This obsolete function is like plist-get except that it compares properties using equal instead of eq.

Function: lax-plist-put plist property value

This obsolete function is like plist-put except that it compares properties using equal instead of eq.

Function: plist-member plist property &optional predicate

This returns non-nil if plist contains the given property. Comparisons are done with predicate, which defaults to eq. Unlike plist-get, this allows you to distinguish between a missing property and a property with the value nil. The value is actually the tail of plist whose car is property.


6 Sequences, Arrays, and Vectors

The sequence type is the union of two other Lisp types: lists and arrays. In other words, any list is a sequence, and any array is a sequence. The common property that all sequences have is that each is an ordered collection of elements.

An array is a fixed-length object with a slot for each of its elements. All the elements are accessible in constant time. The four types of arrays are strings, vectors, char-tables and bool-vectors.

A list is a sequence of elements, but it is not a single primitive object; it is made of cons cells, one cell per element. Finding the nth element requires looking through n cons cells, so elements farther from the beginning of the list take longer to access. But it is possible to add elements to the list, or remove elements.

The following diagram shows the relationship between these types:

          _____________________________________________
         |                                             |
         |          Sequence                           |
         |  ______   ________________________________  |
         | |      | |                                | |
         | | List | |             Array              | |
         | |      | |    ________       ________     | |
         | |______| |   |        |     |        |    | |
         |          |   | Vector |     | String |    | |
         |          |   |________|     |________|    | |
         |          |  ____________   _____________  | |
         |          | |            | |             | | |
         |          | | Char-table | | Bool-vector | | |
         |          | |____________| |_____________| | |
         |          |________________________________| |
         |_____________________________________________|

6.1 Sequences

This section describes functions that accept any kind of sequence.

Function: sequencep object

This function returns t if object is a list, vector, string, bool-vector, or char-table, nil otherwise. See also seqp below.

Function: length sequence

This function returns the number of elements in sequence. The function signals the wrong-type-argument error if the argument is not a sequence or is a dotted list; it signals the circular-list error if the argument is a circular list. For a char-table, the value returned is always one more than the maximum Emacs character code.

See Definition of safe-length, for the related function safe-length.

(length '(1 2 3))
    ⇒ 3
(length ())
    ⇒ 0
(length "foobar")
    ⇒ 6
(length [1 2 3])
    ⇒ 3
(length (make-bool-vector 5 nil))
    ⇒ 5

See also string-bytes, in Text Representations.

If you need to compute the width of a string on display, you should use string-width (see Size of Displayed Text), not length, since length only counts the number of characters, but does not account for the display width of each character.

Function: length< sequence length

Return non-nil if sequence is shorter than length. This may be more efficient than computing the length of sequence if sequence is a long list.

Function: length> sequence length

Return non-nil if sequence is longer than length.

Function: length= sequence length

Return non-nil if the length of sequence is equal to length.

Function: elt sequence index

This function returns the element of sequence indexed by index. Legitimate values of index are integers ranging from 0 up to one less than the length of sequence. If sequence is a list, out-of-range values behave as for nth. See Definition of nth. Otherwise, out-of-range values trigger an args-out-of-range error.

(elt [1 2 3 4] 2)
     ⇒ 3
(elt '(1 2 3 4) 2)
     ⇒ 3
;; We use string to show clearly which character elt returns.
(string (elt "1234" 2))
     ⇒ "3"
(elt [1 2 3 4] 4)
     error→ Args out of range: [1 2 3 4], 4
(elt [1 2 3 4] -1)
     error→ Args out of range: [1 2 3 4], -1

This function generalizes aref (see Functions that Operate on Arrays) and nth (see Definition of nth).

Function: copy-sequence seqr

This function returns a copy of seqr, which should be either a sequence or a record. The copy is the same type of object as the original, and it has the same elements in the same order. However, if seqr is empty, like a string or a vector of zero length, the value returned by this function might not be a copy, but an empty object of the same type and identical to seqr.

Storing a new element into the copy does not affect the original seqr, and vice versa. However, the elements of the copy are not copies; they are identical (eq) to the elements of the original. Therefore, changes made within these elements, as found via the copy, are also visible in the original.

If the argument is a string with text properties, the property list in the copy is itself a copy, not shared with the original’s property list. However, the actual values of the properties are shared. See Text Properties.

This function does not work for dotted lists. Trying to copy a circular list may cause an infinite loop.

See also append in Building Cons Cells and Lists, concat in Creating Strings, and vconcat in Functions for Vectors, for other ways to copy sequences.

(setq bar (list 1 2))
     ⇒ (1 2)
(setq x (vector 'foo bar))
     ⇒ [foo (1 2)]
(setq y (copy-sequence x))
     ⇒ [foo (1 2)]

(eq x y)
     ⇒ nil
(equal x y)
     ⇒ t
(eq (elt x 1) (elt y 1))
     ⇒ t

;; Replacing an element of one sequence.
(aset x 0 'quux)
x ⇒ [quux (1 2)]
y ⇒ [foo (1 2)]

;; Modifying the inside of a shared element.
(setcar (aref x 1) 69)
x ⇒ [quux (69 2)]
y ⇒ [foo (69 2)]
Function: reverse sequence

This function creates a new sequence whose elements are the elements of sequence, but in reverse order. The original argument sequence is not altered. Note that char-tables cannot be reversed.

(setq x '(1 2 3 4))
     ⇒ (1 2 3 4)
(reverse x)
     ⇒ (4 3 2 1)
x
     ⇒ (1 2 3 4)
(setq x [1 2 3 4])
     ⇒ [1 2 3 4]
(reverse x)
     ⇒ [4 3 2 1]
x
     ⇒ [1 2 3 4]
(setq x "xyzzy")
     ⇒ "xyzzy"
(reverse x)
     ⇒ "yzzyx"
x
     ⇒ "xyzzy"
Function: nreverse sequence

This function reverses the order of the elements of sequence. Unlike reverse the original sequence may be modified.

For example:

(setq x (list 'a 'b 'c))
     ⇒ (a b c)
x
     ⇒ (a b c)
(nreverse x)
     ⇒ (c b a)
;; The cons cell that was first is now last.
x
     ⇒ (a)

To avoid confusion, we usually store the result of nreverse back in the same variable which held the original list:

(setq x (nreverse x))

Here is the nreverse of our favorite example, (a b c), presented graphically:

Original list head:                       Reversed list:
 -------------        -------------        ------------
| car  | cdr  |      | car  | cdr  |      | car | cdr  |
|   a  |  nil |<--   |   b  |   o  |<--   |   c |   o  |
|      |      |   |  |      |   |  |   |  |     |   |  |
 -------------    |   --------- | -    |   -------- | -
                  |             |      |            |
                   -------------        ------------

For the vector, it is even simpler because you don’t need setq:

(setq x (copy-sequence [1 2 3 4]))
     ⇒ [1 2 3 4]
(nreverse x)
     ⇒ [4 3 2 1]
x
     ⇒ [4 3 2 1]

Note that unlike reverse, this function doesn’t work with strings. Although you can alter string data by using aset, it is strongly encouraged to treat strings as immutable even when they are mutable. See Mutability.

Function: sort sequence predicate

This function sorts sequence stably. Note that this function doesn’t work for all sequences; it may be used only for lists and vectors. If sequence is a list, it is modified destructively. This functions returns the sorted sequence and compares elements using predicate. A stable sort is one in which elements with equal sort keys maintain their relative order before and after the sort. Stability is important when successive sorts are used to order elements according to different criteria.

The argument predicate must be a function that accepts two arguments. It is called with two elements of sequence. To get an increasing order sort, the predicate should return non-nil if the first element is “less” than the second, or nil if not.

The comparison function predicate must give reliable results for any given pair of arguments, at least within a single call to sort. It must be antisymmetric; that is, if a is less than b, b must not be less than a. It must be transitive—that is, if a is less than b, and b is less than c, then a must be less than c. If you use a comparison function which does not meet these requirements, the result of sort is unpredictable.

The destructive aspect of sort for lists is that it reuses the cons cells forming sequence by changing their contents, possibly rearranging them in a different order. This means that the value of the input list is undefined after sorting; only the list returned by sort has a well-defined value. Example:

(setq nums (list 2 1 4 3 0))
(sort nums #'<)
     ⇒ (0 1 2 3 4)
     ; nums is unpredictable at this point

Most often we store the result back into the variable that held the original list:

(setq nums (sort nums #'<))

If you wish to make a sorted copy without destroying the original, copy it first and then sort:

(setq nums (list 2 1 4 3 0))
(sort (copy-sequence nums) #'<)
     ⇒ (0 1 2 3 4)
nums
     ⇒ (2 1 4 3 0)

For the better understanding of what stable sort is, consider the following vector example. After sorting, all items whose car is 8 are grouped at the beginning of vector, but their relative order is preserved. All items whose car is 9 are grouped at the end of vector, but their relative order is also preserved:

(setq
  vector
  (vector '(8 . "xxx") '(9 . "aaa") '(8 . "bbb") '(9 . "zzz")
          '(9 . "ppp") '(8 . "ttt") '(8 . "eee") '(9 . "fff")))
     ⇒ [(8 . "xxx") (9 . "aaa") (8 . "bbb") (9 . "zzz")
         (9 . "ppp") (8 . "ttt") (8 . "eee") (9 . "fff")]
(sort vector (lambda (x y) (< (car x) (car y))))
     ⇒ [(8 . "xxx") (8 . "bbb") (8 . "ttt") (8 . "eee")
         (9 . "aaa") (9 . "zzz") (9 . "ppp") (9 . "fff")]

See Sorting Text, for more functions that perform sorting. See documentation in Access to Documentation Strings, for a useful example of sort.

The seq.el library provides the following additional sequence manipulation macros and functions, prefixed with seq-.

All functions defined in this library are free of side-effects; i.e., they do not modify any sequence (list, vector, or string) that you pass as an argument. Unless otherwise stated, the result is a sequence of the same type as the input. For those functions that take a predicate, this should be a function of one argument.

The seq.el library can be extended to work with additional types of sequential data-structures. For that purpose, all functions are defined using cl-defgeneric. See Generic Functions, for more details about using cl-defgeneric for adding extensions.

Function: seq-elt sequence index

This function returns the element of sequence at the specified index, which is an integer whose valid value range is zero to one less than the length of sequence. For out-of-range values on built-in sequence types, seq-elt behaves like elt. For the details, see Definition of elt.

(seq-elt [1 2 3 4] 2)
⇒ 3

seq-elt returns places settable using setf (see The setf Macro).

(setq vec [1 2 3 4])
(setf (seq-elt vec 2) 5)
vec
⇒ [1 2 5 4]
Function: seq-length sequence

This function returns the number of elements in sequence. For built-in sequence types, seq-length behaves like length. See Definition of length.

Function: seqp object

This function returns non-nil if object is a sequence (a list or array), or any additional type of sequence defined via seq.el generic functions. This is an extensible variant of sequencep.

(seqp [1 2])
⇒ t
(seqp 2)
⇒ nil
Function: seq-drop sequence n

This function returns all but the first n (an integer) elements of sequence. If n is negative or zero, the result is sequence.

(seq-drop [1 2 3 4 5 6] 3)
⇒ [4 5 6]
(seq-drop "hello world" -4)
⇒ "hello world"
Function: seq-take sequence n

This function returns the first n (an integer) elements of sequence. If n is negative or zero, the result is nil.

(seq-take '(1 2 3 4) 3)
⇒ (1 2 3)
(seq-take [1 2 3 4] 0)
⇒ []
Function: seq-take-while predicate sequence

This function returns the members of sequence in order, stopping before the first one for which predicate returns nil.

(seq-take-while (lambda (elt) (> elt 0)) '(1 2 3 -1 -2))
⇒ (1 2 3)
(seq-take-while (lambda (elt) (> elt 0)) [-1 4 6])
⇒ []
Function: seq-drop-while predicate sequence

This function returns the members of sequence in order, starting from the first one for which predicate returns nil.

(seq-drop-while (lambda (elt) (> elt 0)) '(1 2 3 -1 -2))
⇒ (-1 -2)
(seq-drop-while (lambda (elt) (< elt 0)) [1 4 6])
⇒ [1 4 6]
Function: seq-split sequence length

This function returns a list consisting of sub-sequences of sequence of (at most) length length. (The final element may be shorter than length if the length of sequence isn’t a multiple of length.

(seq-split [0 1 2 3 4] 2)
⇒ ([0 1] [2 3] [4])
Function: seq-do function sequence

This function applies function to each element of sequence in turn (presumably for side effects), and returns sequence.

Function: seq-map function sequence

This function returns the result of applying function to each element of sequence. The returned value is a list.

(seq-map #'1+ '(2 4 6))
⇒ (3 5 7)
(seq-map #'symbol-name [foo bar])
⇒ ("foo" "bar")
Function: seq-map-indexed function sequence

This function returns the result of applying function to each element of sequence and its index within seq. The returned value is a list.

(seq-map-indexed (lambda (elt idx)
                   (list idx elt))
                 '(a b c))
⇒ ((0 a) (1 b) (2 c))
Function: seq-mapn function &rest sequences

This function returns the result of applying function to each element of sequences. The arity (see subr-arity) of function must match the number of sequences. Mapping stops at the end of the shortest sequence, and the returned value is a list.

(seq-mapn #'+ '(2 4 6) '(20 40 60))
⇒ (22 44 66)
(seq-mapn #'concat '("moskito" "bite") ["bee" "sting"])
⇒ ("moskitobee" "bitesting")
Function: seq-filter predicate sequence

This function returns a list of all the elements in sequence for which predicate returns non-nil.

(seq-filter (lambda (elt) (> elt 0)) [1 -1 3 -3 5])
⇒ (1 3 5)
(seq-filter (lambda (elt) (> elt 0)) '(-1 -3 -5))
⇒ nil
Function: seq-remove predicate sequence

This function returns a list of all the elements in sequence for which predicate returns nil.

(seq-remove (lambda (elt) (> elt 0)) [1 -1 3 -3 5])
⇒ (-1 -3)
(seq-remove (lambda (elt) (< elt 0)) '(-1 -3 -5))
⇒ nil
Function: seq-remove-at-position sequence n

This function returns a copy of sequence where the element at (zero-based) index n got removed. The result is a sequence of the same type as sequence.

(seq-remove-at-position [1 -1 3 -3 5] 0)
⇒ [-1 3 -3 5]
(seq-remove-at-position [1 -1 3 -3 5] 3)
⇒ [1 -1 3 5]
Function: seq-keep function sequence

This function returns a list of all non-nil results from calling function on the elements in sequence.

(seq-keep #'cl-digit-char-p '(?6 ?a ?7))
⇒ (6 7)
Function: seq-reduce function sequence initial-value

This function returns the result of calling function with initial-value and the first element of sequence, then calling function with that result and the second element of sequence, then with that result and the third element of sequence, etc. function should be a function of two arguments.

function is called with two arguments. initial-value (and then the accumulated value) is used as the first argument, and the elements in sequence are used for the second argument.

If sequence is empty, this returns initial-value without calling function.

(seq-reduce #'+ [1 2 3 4] 0)
⇒ 10
(seq-reduce #'+ '(1 2 3 4) 5)
⇒ 15
(seq-reduce #'+ '() 3)
⇒ 3
Function: seq-some predicate sequence

This function returns the first non-nil value returned by applying predicate to each element of sequence in turn.

(seq-some #'numberp ["abc" 1 nil])
⇒ t
(seq-some #'numberp ["abc" "def"])
⇒ nil
(seq-some #'null ["abc" 1 nil])
⇒ t
(seq-some #'1+ [2 4 6])
⇒ 3
Function: seq-find predicate sequence &optional default

This function returns the first element in sequence for which predicate returns non-nil. If no element matches predicate, the function returns default.

Note that this function has an ambiguity if the found element is identical to default, as in that case it cannot be known whether an element was found or not.

(seq-find #'numberp ["abc" 1 nil])
⇒ 1
(seq-find #'numberp ["abc" "def"])
⇒ nil
Function: seq-every-p predicate sequence

This function returns non-nil if applying predicate to every element of sequence returns non-nil.

(seq-every-p #'numberp [2 4 6])
⇒ t
(seq-every-p #'numberp [2 4 "6"])
⇒ nil
Function: seq-empty-p sequence

This function returns non-nil if sequence is empty.

(seq-empty-p "not empty")
⇒ nil
(seq-empty-p "")
⇒ t
Function: seq-count predicate sequence

This function returns the number of elements in sequence for which predicate returns non-nil.

(seq-count (lambda (elt) (> elt 0)) [-1 2 0 3 -2])
⇒ 2
Function: seq-sort function sequence

This function returns a copy of sequence that is sorted according to function, a function of two arguments that returns non-nil if the first argument should sort before the second.

Function: seq-sort-by function predicate sequence

This function is similar to seq-sort, but the elements of sequence are transformed by applying function on them before being sorted. function is a function of one argument.

(seq-sort-by #'seq-length #'> ["a" "ab" "abc"])
⇒ ["abc" "ab" "a"]
Function: seq-contains-p sequence elt &optional function

This function returns non-nil if at least one element in sequence is equal to elt. If the optional argument function is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-contains-p '(symbol1 symbol2) 'symbol1)
⇒ t
(seq-contains-p '(symbol1 symbol2) 'symbol3)
⇒ nil
Function: seq-set-equal-p sequence1 sequence2 &optional testfn

This function checks whether sequence1 and sequence2 contain the same elements, regardless of the order. If the optional argument testfn is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-set-equal-p '(a b c) '(c b a))
⇒ t
(seq-set-equal-p '(a b c) '(c b))
⇒ nil
(seq-set-equal-p '("a" "b" "c") '("c" "b" "a"))
⇒ t
(seq-set-equal-p '("a" "b" "c") '("c" "b" "a") #'eq)
⇒ nil
Function: seq-position sequence elt &optional function

This function returns the (zero-based) index of the first element in sequence that is equal to elt. If the optional argument function is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-position '(a b c) 'b)
⇒ 1
(seq-position '(a b c) 'd)
⇒ nil
Function: seq-positions sequence elt &optional testfn

This function returns a list of the (zero-based) indices of the elements in sequence for which testfn returns non-nil when passed the element and elt as arguments. testfn defaults to equal.

(seq-positions '(a b c a d) 'a)
⇒ (0 3)
(seq-positions '(a b c a d) 'z)
⇒ nil
(seq-positions '(11 5 7 12 9 15) 10 #'>=)
⇒ (0 3 5)
Function: seq-uniq sequence &optional function

This function returns a list of the elements of sequence with duplicates removed. If the optional argument function is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-uniq '(1 2 2 1 3))
⇒ (1 2 3)
(seq-uniq '(1 2 2.0 1.0) #'=)
⇒ (1 2)
Function: seq-subseq sequence start &optional end

This function returns a subset of sequence from start to end, both integers (end defaults to the last element). If start or end is negative, it counts from the end of sequence.

(seq-subseq '(1 2 3 4 5) 1)
⇒ (2 3 4 5)
(seq-subseq '[1 2 3 4 5] 1 3)
⇒ [2 3]
(seq-subseq '[1 2 3 4 5] -3 -1)
⇒ [3 4]
Function: seq-concatenate type &rest sequences

This function returns a sequence of type type made of the concatenation of sequences. type may be: vector, list or string.

(seq-concatenate 'list '(1 2) '(3 4) [5 6])
⇒ (1 2 3 4 5 6)
(seq-concatenate 'string "Hello " "world")
⇒ "Hello world"
Function: seq-mapcat function sequence &optional type

This function returns the result of applying seq-concatenate to the result of applying function to each element of sequence. The result is a sequence of type type, or a list if type is nil.

(seq-mapcat #'seq-reverse '((3 2 1) (6 5 4)))
⇒ (1 2 3 4 5 6)
Function: seq-partition sequence n

This function returns a list of the elements of sequence grouped into sub-sequences of length n. The last sequence may contain less elements than n. n must be an integer. If n is a negative integer or 0, the return value is nil.

(seq-partition '(0 1 2 3 4 5 6 7) 3)
⇒ ((0 1 2) (3 4 5) (6 7))
Function: seq-union sequence1 sequence2 &optional function

This function returns a list of the elements that appear either in sequence1 or sequence2. The elements of the returned list are all unique, in the sense that no two elements there will compare equal. If the optional argument function is non-nil, it should be a function of two arguments to use to compare elements, instead of the default equal.

(seq-union [1 2 3] [3 5])
⇒ (1 2 3 5)
Function: seq-intersection sequence1 sequence2 &optional function

This function returns a list of the elements that appear both in sequence1 and sequence2. If the optional argument function is non-nil, it is a function of two arguments to use to compare elements instead of the default equal.

(seq-intersection [2 3 4 5] [1 3 5 6 7])
⇒ (3 5)
Function: seq-difference sequence1 sequence2 &optional function

This function returns a list of the elements that appear in sequence1 but not in sequence2. If the optional argument function is non-nil, it is a function of two arguments to use to compare elements instead of the default equal.

(seq-difference '(2 3 4 5) [1 3 5 6 7])
⇒ (2 4)
Function: seq-group-by function sequence

This function separates the elements of sequence into an alist whose keys are the result of applying function to each element of sequence. Keys are compared using equal.

(seq-group-by #'integerp '(1 2.1 3 2 3.2))
⇒ ((t 1 3 2) (nil 2.1 3.2))
(seq-group-by #'car '((a 1) (b 2) (a 3) (c 4)))
⇒ ((b (b 2)) (a (a 1) (a 3)) (c (c 4)))
Function: seq-into sequence type

This function converts the sequence sequence into a sequence of type type. type can be one of the following symbols: vector, string or list.

(seq-into [1 2 3] 'list)
⇒ (1 2 3)
(seq-into nil 'vector)
⇒ []
(seq-into "hello" 'vector)
⇒ [104 101 108 108 111]
Function: seq-min sequence

This function returns the smallest element of sequence. The elements of sequence must be numbers or markers (see Markers).

(seq-min [3 1 2])
⇒ 1
(seq-min "Hello")
⇒ 72
Function: seq-max sequence

This function returns the largest element of sequence. The elements of sequence must be numbers or markers.

(seq-max [1 3 2])
⇒ 3
(seq-max "Hello")
⇒ 111
Macro: seq-doseq (var sequence) body…

This macro is like dolist (see dolist), except that sequence can be a list, vector or string. This is primarily useful for side-effects.

Macro: seq-let var-sequence val-sequence body…

This macro binds the variables defined in var-sequence to the values that are the corresponding elements of val-sequence. This is known as destructuring binding. The elements of var-sequence can themselves include sequences, allowing for nested destructuring.

The var-sequence sequence can also include the &rest marker followed by a variable name to be bound to the rest of val-sequence.

(seq-let [first second] [1 2 3 4]
  (list first second))
⇒ (1 2)
(seq-let (_ a _ b) '(1 2 3 4)
  (list a b))
⇒ (2 4)
(seq-let [a [b [c]]] [1 [2 [3]]]
  (list a b c))
⇒ (1 2 3)
(seq-let [a b &rest others] [1 2 3 4]
  others)
⇒ [3 4]

The pcase patterns provide an alternative facility for destructuring binding, see Destructuring with pcase Patterns.

Macro: seq-setq var-sequence val-sequence

This macro works similarly to seq-let, except that values are assigned to variables as if by setq instead of as in a let binding.

(let ((a nil)
      (b nil))
  (seq-setq (_ a _ b) '(1 2 3 4))
  (list a b))
⇒ (2 4)
Function: seq-random-elt sequence

This function returns an element of sequence taken at random.

(seq-random-elt [1 2 3 4])
⇒ 3
(seq-random-elt [1 2 3 4])
⇒ 2
(seq-random-elt [1 2 3 4])
⇒ 4
(seq-random-elt [1 2 3 4])
⇒ 2
(seq-random-elt [1 2 3 4])
⇒ 1

If sequence is empty, this function signals an error.


6.2 Arrays

An array object has slots that hold a number of other Lisp objects, called the elements of the array. Any element of an array may be accessed in constant time. In contrast, the time to access an element of a list is proportional to the position of that element in the list.

Emacs defines four types of array, all one-dimensional: strings (see String Type), vectors (see Vector Type), bool-vectors (see Bool-Vector Type), and char-tables (see Char-Table Type). Vectors and char-tables can hold elements of any type, but strings can only hold characters, and bool-vectors can only hold t and nil.

All four kinds of array share these characteristics:

  • The first element of an array has index zero, the second element has index 1, and so on. This is called zero-origin indexing. For example, an array of four elements has indices 0, 1, 2, and 3.
  • The length of the array is fixed once you create it; you cannot change the length of an existing array.
  • For purposes of evaluation, the array is a constant—i.e., it evaluates to itself.
  • The elements of an array may be referenced or changed with the functions aref and aset, respectively (see Functions that Operate on Arrays).

When you create an array, other than a char-table, you must specify its length. You cannot specify the length of a char-table, because that is determined by the range of character codes.

In principle, if you want an array of text characters, you could use either a string or a vector. In practice, we always choose strings for such applications, for four reasons:

  • They occupy one-fourth the space of a vector of the same elements.
  • Strings are printed in a way that shows the contents more clearly as text.
  • Strings can hold text properties. See Text Properties.
  • Many of the specialized editing and I/O facilities of Emacs accept only strings. For example, you cannot insert a vector of characters into a buffer the way you can insert a string. See Strings and Characters.

By contrast, for an array of keyboard input characters (such as a key sequence), a vector may be necessary, because many keyboard input characters are outside the range that will fit in a string. See Key Sequence Input.


6.3 Functions that Operate on Arrays

In this section, we describe the functions that accept all types of arrays.

Function: arrayp object

This function returns t if object is an array (i.e., a vector, a string, a bool-vector or a char-table).

(arrayp [a])
     ⇒ t
(arrayp "asdf")
     ⇒ t
(arrayp (syntax-table))    ;; A char-table.
     ⇒ t
Function: aref arr index

This function returns the indexth element of the array or record arr. The first element is at index zero.

(setq primes [2 3 5 7 11 13])
     ⇒ [2 3 5 7 11 13]
(aref primes 4)
     ⇒ 11
(aref "abcdefg" 1)
     ⇒ 98           ; b’ is ASCII code 98.

See also the function elt, in Sequences.

Function: aset array index object

This function sets the indexth element of array to be object. It returns object.

(setq w (vector 'foo 'bar 'baz))
     ⇒ [foo bar baz]
(aset w 0 'fu)
     ⇒ fu
w
     ⇒ [fu bar baz]

;; copy-sequence copies the string to be modified later.
(setq x (copy-sequence "asdfasfd"))
     ⇒ "asdfasfd"
(aset x 3 ?Z)
     ⇒ 90
x
     ⇒ "asdZasfd"

The array should be mutable. See Mutability.

If array is a string and object is not a character, a wrong-type-argument error results. The function converts a unibyte string to multibyte if necessary to insert a character.

Function: fillarray array object

This function fills the array array with object, so that each element of array is object. It returns array.

(setq a (copy-sequence [a b c d e f g]))
     ⇒ [a b c d e f g]
(fillarray a 0)
     ⇒ [0 0 0 0 0 0 0]
a
     ⇒ [0 0 0 0 0 0 0]
(setq s (copy-sequence "When in the course"))
     ⇒ "When in the course"
(fillarray s ?-)
     ⇒ "------------------"

If array is a string and object is not a character, a wrong-type-argument error results.

The general sequence functions copy-sequence and length are often useful for objects known to be arrays. See Sequences.


6.4 Vectors

A vector is a general-purpose array whose elements can be any Lisp objects. (By contrast, the elements of a string can only be characters. See Strings and Characters.) Vectors are used in Emacs for many purposes: as key sequences (see Key Sequences), as symbol-lookup tables (see Creating and Interning Symbols), as part of the representation of a byte-compiled function (see Byte Compilation), and more.

Like other arrays, vectors use zero-origin indexing: the first element has index 0.

Vectors are printed with square brackets surrounding the elements. Thus, a vector whose elements are the symbols a, b and a is printed as [a b a]. You can write vectors in the same way in Lisp input.

A vector, like a string or a number, is considered a constant for evaluation: the result of evaluating it is the same vector. This does not evaluate or even examine the elements of the vector. See Self-Evaluating Forms. Vectors written with square brackets should not be modified via aset or other destructive operations. See Mutability.

Here are examples illustrating these principles:

(setq avector [1 two '(three) "four" [five]])
     ⇒ [1 two '(three) "four" [five]]
(eval avector)
     ⇒ [1 two '(three) "four" [five]]
(eq avector (eval avector))
     ⇒ t

6.5 Functions for Vectors

Here are some functions that relate to vectors:

Function: vectorp object

This function returns t if object is a vector.

(vectorp [a])
     ⇒ t
(vectorp "asdf")
     ⇒ nil
Function: vector &rest objects

This function creates and returns a vector whose elements are the arguments, objects.

(vector 'foo 23 [bar baz] "rats")
     ⇒ [foo 23 [bar baz] "rats"]
(vector)
     ⇒ []
Function: make-vector length object

This function returns a new vector consisting of length elements, each initialized to object.

(setq sleepy (make-vector 9 'Z))
     ⇒ [Z Z Z Z Z Z Z Z Z]
Function: vconcat &rest sequences

This function returns a new vector containing all the elements of sequences. The arguments sequences may be proper lists, vectors, strings or bool-vectors. If no sequences are given, the empty vector is returned.

The value is either the empty vector, or is a newly constructed nonempty vector that is not eq to any existing vector.

(setq a (vconcat '(A B C) '(D E F)))
     ⇒ [A B C D E F]
(eq a (vconcat a))
     ⇒ nil
(vconcat)
     ⇒ []
(vconcat [A B C] "aa" '(foo (6 7)))
     ⇒ [A B C 97 97 foo (6 7)]

The vconcat function also allows byte-code function objects as arguments. This is a special feature to make it easy to access the entire contents of a byte-code function object. See Byte-Code Function Objects.

For other concatenation functions, see mapconcat in Mapping Functions, concat in Creating Strings, and append in Building Cons Cells and Lists.

The append function also provides a way to convert a vector into a list with the same elements:

(setq avector [1 two (quote (three)) "four" [five]])
     ⇒ [1 two '(three) "four" [five]]
(append avector nil)
     ⇒ (1 two '(three) "four" [five])

6.6 Char-Tables

A char-table is much like a vector, except that it is indexed by character codes. Any valid character code, without modifiers, can be used as an index in a char-table. You can access a char-table’s elements with aref and aset, as with any array. In addition, a char-table can have extra slots to hold additional data not associated with particular character codes. Like vectors, char-tables are constants when evaluated, and can hold elements of any type.

Each char-table has a subtype, a symbol, which serves two purposes:

  • The subtype provides an easy way to tell what the char-table is for. For instance, display tables are char-tables with display-table as the subtype, and syntax tables are char-tables with syntax-table as the subtype. The subtype can be queried using the function char-table-subtype, described below.
  • The subtype controls the number of extra slots in the char-table. This number is specified by the subtype’s char-table-extra-slots symbol property (see Symbol Properties), whose value should be an integer between 0 and 10. If the subtype has no such symbol property, the char-table has no extra slots.

A char-table can have a parent, which is another char-table. If it does, then whenever the char-table specifies nil for a particular character c, it inherits the value specified in the parent. In other words, (aref char-table c) returns the value from the parent of char-table if char-table itself specifies nil.

A char-table can also have a default value. If so, then (aref char-table c) returns the default value whenever the char-table does not specify any other non-nil value.

Function: make-char-table subtype &optional init

Return a newly-created char-table, with subtype subtype (a symbol). Each element is initialized to init, which defaults to nil. You cannot alter the subtype of a char-table after the char-table is created.

There is no argument to specify the length of the char-table, because all char-tables have room for any valid character code as an index.

If subtype has the char-table-extra-slots symbol property, that specifies the number of extra slots in the char-table. This should be an integer between 0 and 10; otherwise, make-char-table raises an error. If subtype has no char-table-extra-slots symbol property (see Property Lists), the char-table has no extra slots.

Function: char-table-p object

This function returns t if object is a char-table, and nil otherwise.

Function: char-table-subtype char-table

This function returns the subtype symbol of char-table.

There is no special function to access default values in a char-table. To do that, use char-table-range (see below).

Function: char-table-parent char-table

This function returns the parent of char-table. The parent is always either nil or another char-table.

Function: set-char-table-parent char-table new-parent

This function sets the parent of char-table to new-parent.

Function: char-table-extra-slot char-table n

This function returns the contents of extra slot n (zero based) of char-table. The number of extra slots in a char-table is determined by its subtype.

Function: set-char-table-extra-slot char-table n value

This function stores value in extra slot n (zero based) of char-table.

A char-table can specify an element value for a single character code; it can also specify a value for an entire character set.

Function: char-table-range char-table range

This returns the value specified in char-table for a range of characters range. Here are the possibilities for range:

nil

Refers to the default value.

char

Refers to the element for character char (supposing char is a valid character code).

(from . to)

A cons cell refers to all the characters in the inclusive range ‘[from..to]’. In this case, the function returns the value for the character specified by from.

Function: set-char-table-range char-table range value

This function sets the value in char-table for a range of characters range. Here are the possibilities for range:

nil

Refers to the default value.

t

Refers to the whole range of character codes.

char

Refers to the element for character char (supposing char is a valid character code).

(from . to)

A cons cell refers to all the characters in the inclusive range ‘[from..to]’.

Function: map-char-table function char-table

This function calls its argument function for each element of char-table that has a non-nil value. The call to function is with two arguments, a key and a value. The key is a possible range argument for char-table-range—either a valid character or a cons cell (from . to), specifying a range of characters that share the same value. The value is what (char-table-range char-table key) returns.

Overall, the key-value pairs passed to function describe all the values stored in char-table.

The return value is always nil; to make calls to map-char-table useful, function should have side effects. For example, here is how to examine the elements of the syntax table:

(let (accumulator)
   (map-char-table
    (lambda (key value)
      (setq accumulator
            (cons (list
                   (if (consp key)
                       (list (car key) (cdr key))
                     key)
                   value)
                  accumulator)))
    (syntax-table))
   accumulator)
⇒
(((2597602 4194303) (2)) ((2597523 2597601) (3))
 ... (65379 (5 . 65378)) (65378 (4 . 65379)) (65377 (1))
 ... (12 (0)) (11 (3)) (10 (12)) (9 (0)) ((0 8) (3)))

6.7 Bool-vectors

A bool-vector is much like a vector, except that it stores only the values t and nil. If you try to store any non-nil value into an element of the bool-vector, the effect is to store t there. As with all arrays, bool-vector indices start from 0, and the length cannot be changed once the bool-vector is created. Bool-vectors are constants when evaluated.

Several functions work specifically with bool-vectors; aside from that, you manipulate them with same functions used for other kinds of arrays.

Function: make-bool-vector length initial

Return a new bool-vector of length elements, each one initialized to initial.

Function: bool-vector &rest objects

This function creates and returns a bool-vector whose elements are the arguments, objects.

Function: bool-vector-p object

This returns t if object is a bool-vector, and nil otherwise.

There are also some bool-vector set operation functions, described below:

Function: bool-vector-exclusive-or a b &optional c

Return bitwise exclusive or of bool vectors a and b. If optional argument c is given, the result of this operation is stored into c. All arguments should be bool vectors of the same length.

Function: bool-vector-union a b &optional c

Return bitwise or of bool vectors a and b. If optional argument c is given, the result of this operation is stored into c. All arguments should be bool vectors of the same length.

Function: bool-vector-intersection a b &optional c

Return bitwise and of bool vectors a and b. If optional argument c is given, the result of this operation is stored into c. All arguments should be bool vectors of the same length.

Function: bool-vector-set-difference a b &optional c

Return set difference of bool vectors a and b. If optional argument c is given, the result of this operation is stored into c. All arguments should be bool vectors of the same length.

Function: bool-vector-not a &optional b

Return set complement of bool vector a. If optional argument b is given, the result of this operation is stored into b. All arguments should be bool vectors of the same length.

Function: bool-vector-subsetp a b

Return t if every t value in a is also t in b, nil otherwise. All arguments should be bool vectors of the same length.

Function: bool-vector-count-consecutive a b i

Return the number of consecutive elements in a equal b starting at i. a is a bool vector, b is t or nil, and i is an index into a.

Function: bool-vector-count-population a

Return the number of elements that are t in bool vector a.

The printed form represents up to 8 boolean values as a single character:

(bool-vector t nil t nil)
     ⇒ #&4"^E"
(bool-vector)
     ⇒ #&0""

You can use vconcat to print a bool-vector like other vectors:

(vconcat (bool-vector nil t nil t))
     ⇒ [nil t nil t]

Here is another example of creating, examining, and updating a bool-vector:

(setq bv (make-bool-vector 5 t))
     ⇒ #&5"^_"
(aref bv 1)
     ⇒ t
(aset bv 3 nil)
     ⇒ nil
bv
     ⇒ #&5"^W"

These results make sense because the binary codes for control-_ and control-W are 11111 and 10111, respectively.


6.8 Managing a Fixed-Size Ring of Objects

A ring is a fixed-size data structure that supports insertion, deletion, rotation, and modulo-indexed reference and traversal. An efficient ring data structure is implemented by the ring package. It provides the functions listed in this section.

Note that several rings in Emacs, like the kill ring and the mark ring, are actually implemented as simple lists, not using the ring package; thus the following functions won’t work on them.

Function: make-ring size

This returns a new ring capable of holding size objects. size should be an integer.

Function: ring-p object

This returns t if object is a ring, nil otherwise.

Function: ring-size ring

This returns the maximum capacity of the ring.

Function: ring-length ring

This returns the number of objects that ring currently contains. The value will never exceed that returned by ring-size.

Function: ring-elements ring

This returns a list of the objects in ring, in order, newest first.

Function: ring-copy ring

This returns a new ring which is a copy of ring. The new ring contains the same (eq) objects as ring.

Function: ring-empty-p ring

This returns t if ring is empty, nil otherwise.

The newest element in the ring always has index 0. Higher indices correspond to older elements. Indices are computed modulo the ring length. Index −1 corresponds to the oldest element, −2 to the next-oldest, and so forth.

Function: ring-ref ring index

This returns the object in ring found at index index. index may be negative or greater than the ring length. If ring is empty, ring-ref signals an error.

Function: ring-insert ring object

This inserts object into ring, making it the newest element, and returns object.

If the ring is full, insertion removes the oldest element to make room for the new element.

Function: ring-remove ring &optional index

Remove an object from ring, and return that object. The argument index specifies which item to remove; if it is nil, that means to remove the oldest item. If ring is empty, ring-remove signals an error.

Function: ring-insert-at-beginning ring object

This inserts object into ring, treating it as the oldest element. The return value is not significant.

If the ring is full, this function removes the newest element to make room for the inserted element.

Function: ring-resize ring size

Set the size of ring to size. If the new size is smaller, then the oldest items in the ring are discarded.

If you are careful not to exceed the ring size, you can use the ring as a first-in-first-out queue. For example:

(let ((fifo (make-ring 5)))
  (mapc (lambda (obj) (ring-insert fifo obj))
        '(0 one "two"))
  (list (ring-remove fifo) t
        (ring-remove fifo) t
        (ring-remove fifo)))
     ⇒ (0 t one t "two")

7 Records

The purpose of records is to allow programmers to create objects with new types that are not built into Emacs. They are used as the underlying representation of cl-defstruct and defclass instances.

Internally, a record object is much like a vector; its slots can be accessed using aref and it can be copied using copy-sequence. However, the first slot is used to hold its type as returned by type-of. Also, in the current implementation records can have at most 4096 slots, whereas vectors can be much larger. Like arrays, records use zero-origin indexing: the first slot has index 0.

The type slot should be a symbol or a type descriptor. If it’s a type descriptor, the symbol naming its type will be returned; Type Descriptors. Any other kind of object is returned as-is.

The printed representation of records is ‘#s’ followed by a list specifying the contents. The first list element must be the record type. The following elements are the record slots.

To avoid conflicts with other type names, Lisp programs that define new types of records should normally use the naming conventions of the package where these record types are introduced for the names of the types. Note that the names of the types which could possibly conflict might not be known at the time the package defining a record type is loaded; they could be loaded at some future point in time.

A record is considered a constant for evaluation: the result of evaluating it is the same record. This does not evaluate or even examine the slots. See Self-Evaluating Forms.


7.1 Record Functions

Function: recordp object

This function returns t if object is a record.

(recordp #s(a))
     ⇒ t
Function: record type &rest objects

This function creates and returns a record whose type is type and remaining slots are the rest of the arguments, objects.

(record 'foo 23 [bar baz] "rats")
     ⇒ #s(foo 23 [bar baz] "rats")
Function: make-record type length object

This function returns a new record with type type and length more slots, each initialized to object.

(setq sleepy (make-record 'foo 9 'Z))
     ⇒ #s(foo Z Z Z Z Z Z Z Z Z)

7.2 Backward Compatibility

Code compiled with older versions of cl-defstruct that doesn’t use records may run into problems when used in a new Emacs. To alleviate this, Emacs detects when an old cl-defstruct is used, and enables a mode in which type-of handles old struct objects as if they were records.

Function: cl-old-struct-compat-mode arg

If arg is positive, enable backward compatibility with old-style structs.


8 Hash Tables

A hash table is a very fast kind of lookup table, somewhat like an alist (see Association Lists) in that it maps keys to corresponding values. It differs from an alist in these ways:

Emacs Lisp provides a general-purpose hash table data type, along with a series of functions for operating on them. Hash tables have a special printed representation, which consists of ‘#s’ followed by a list specifying the hash table properties and contents. See Creating Hash Tables. (Hash notation, the initial ‘#’ character used in the printed representations of objects with no read representation, has nothing to do with hash tables. See Printed Representation and Read Syntax.)

Obarrays are also a kind of hash table, but they are a different type of object and are used only for recording interned symbols (see Creating and Interning Symbols).


8.1 Creating Hash Tables

The principal function for creating a hash table is make-hash-table.

Function: make-hash-table &rest keyword-args

This function creates a new hash table according to the specified arguments. The arguments should consist of alternating keywords (particular symbols recognized specially) and values corresponding to them.

Several keywords make sense in make-hash-table, but the only two that you really need to know about are :test and :weakness.

:test test

This specifies the method of key lookup for this hash table. The default is eql; eq and equal are other alternatives:

eql

Keys which are numbers are the same if they are equal, that is, if they are equal in value and either both are integers or both are floating point; otherwise, two distinct objects are never the same.

eq

Any two distinct Lisp objects are different as keys.

equal

Two Lisp objects are the same, as keys, if they are equal according to equal.

You can use define-hash-table-test (see Defining Hash Comparisons) to define additional possibilities for test.

:weakness weak

The weakness of a hash table specifies whether the presence of a key or value in the hash table preserves it from garbage collection.

The value, weak, must be one of nil, key, value, key-or-value, key-and-value, or t which is an alias for key-and-value. If weak is key then the hash table does not prevent its keys from being collected as garbage (if they are not referenced anywhere else); if a particular key does get collected, the corresponding association is removed from the hash table.

If weak is value, then the hash table does not prevent values from being collected as garbage (if they are not referenced anywhere else); if a particular value does get collected, the corresponding association is removed from the hash table.

If weak is key-and-value or t, both the key and the value must be live in order to preserve the association. Thus, the hash table does not protect either keys or values from garbage collection; if either one is collected as garbage, that removes the association.

If weak is key-or-value, either the key or the value can preserve the association. Thus, associations are removed from the hash table when both their key and value would be collected as garbage (if not for references from weak hash tables).

The default for weak is nil, so that all keys and values referenced in the hash table are preserved from garbage collection.

:size size

This specifies a hint for how many associations you plan to store in the hash table. If you know the approximate number, you can make things a little more efficient by specifying it this way. If you specify too small a size, the hash table will grow automatically when necessary, but doing that takes some extra time.

The default size is 65.

:rehash-size rehash-size

When you add an association to a hash table and the table is full, it grows automatically. This value specifies how to make the hash table larger, at that time.

If rehash-size is an integer, it should be positive, and the hash table grows by adding approximately that much to the nominal size. If rehash-size is floating point, it had better be greater than 1, and the hash table grows by multiplying the old size by approximately that number.

The default value is 1.5.

:rehash-threshold threshold

This specifies the criterion for when the hash table is full (so it should be made larger). The value, threshold, should be a positive floating-point number, no greater than 1. The hash table is full whenever the actual number of entries exceeds the nominal size multiplied by an approximation to this value. The default for threshold is 0.8125.

You can also create a hash table using the printed representation for hash tables. The Lisp reader can read this printed representation, provided each element in the specified hash table has a valid read syntax (see Printed Representation and Read Syntax). For instance, the following specifies a hash table containing the keys key1 and key2 (both symbols) associated with val1 (a symbol) and 300 (a number) respectively.

#s(hash-table size 30 data (key1 val1 key2 300))

Note, however, that when using this in Emacs Lisp code, it’s undefined whether this creates a new hash table or not. If you want to create a new hash table, you should always use make-hash-table (see Self-Evaluating Forms).

The printed representation for a hash table consists of ‘#s’ followed by a list beginning with ‘hash-table’. The rest of the list should consist of zero or more property-value pairs specifying the hash table’s properties and initial contents. The properties and values are read literally. Valid property names are size, test, weakness, rehash-size, rehash-threshold, and data. The data property should be a list of key-value pairs for the initial contents; the other properties have the same meanings as the matching make-hash-table keywords (:size, :test, etc.), described above.

Note that you cannot specify a hash table whose initial contents include objects that have no read syntax, such as buffers and frames. Such objects may be added to the hash table after it is created.


8.2 Hash Table Access

This section describes the functions for accessing and storing associations in a hash table. In general, any Lisp object can be used as a hash key, unless the comparison method imposes limits. Any Lisp object can also be used as the value.

Function: gethash key table &optional default

This function looks up key in table, and returns its associated value—or default, if key has no association in table.

Function: puthash key value table

This function enters an association for key in table, with value value. If key already has an association in table, value replaces the old associated value. This function always returns value.

Function: remhash key table

This function removes the association for key from table, if there is one. If key has no association, remhash does nothing.

Common Lisp note: In Common Lisp, remhash returns non-nil if it actually removed an association and nil otherwise. In Emacs Lisp, remhash always returns nil.

Function: clrhash table

This function removes all the associations from hash table table, so that it becomes empty. This is also called clearing the hash table. clrhash returns the empty table.

Function: maphash function table

This function calls function once for each of the associations in table. The function function should accept two arguments—a key listed in table, and its associated value. maphash returns nil.


8.3 Defining Hash Comparisons

You can define new methods of key lookup by means of define-hash-table-test. In order to use this feature, you need to understand how hash tables work, and what a hash code means.

You can think of a hash table conceptually as a large array of many slots, each capable of holding one association. To look up a key, gethash first computes an integer, the hash code, from the key. It can reduce this integer modulo the length of the array, to produce an index in the array. Then it looks in that slot, and if necessary in other nearby slots, to see if it has found the key being sought.

Thus, to define a new method of key lookup, you need to specify both a function to compute the hash code from a key, and a function to compare two keys directly. The two functions should be consistent with each other: that is, two keys’ hash codes should be the same if the keys compare as equal. Also, since the two functions can be called at any time (such as by the garbage collector), the functions should be free of side effects and should return quickly, and their behavior should depend on only on properties of the keys that do not change.

Function: define-hash-table-test name test-fn hash-fn

This function defines a new hash table test, named name.

After defining name in this way, you can use it as the test argument in make-hash-table. When you do that, the hash table will use test-fn to compare key values, and hash-fn to compute a hash code from a key value.

The function test-fn should accept two arguments, two keys, and return non-nil if they are considered the same.

The function hash-fn should accept one argument, a key, and return an integer that is the hash code of that key. For good results, the function should use the whole range of fixnums for hash codes, including negative fixnums.

The specified functions are stored in the property list of name under the property hash-table-test; the property value’s form is (test-fn hash-fn).

Function: sxhash-equal obj

This function returns a hash code for Lisp object obj. This is an integer that reflects the contents of obj and the other Lisp objects it points to.

If two objects obj1 and obj2 are equal, then (sxhash-equal obj1) and (sxhash-equal obj2) are the same integer.

If the two objects are not equal, the values returned by sxhash-equal are usually different, but not always. sxhash-equal is designed to be reasonably fast (since it’s used for indexing hash tables) so it won’t recurse deeply into nested structures. In addition; once in a rare while, by luck, you will encounter two distinct-looking simple objects that give the same result from sxhash-equal. So you can’t, in general, use sxhash-equal to check whether an object has changed.

Common Lisp note: In Common Lisp a similar function is called sxhash. Emacs provides this name as a compatibility alias for sxhash-equal.

Function: sxhash-eq obj

This function returns a hash code for Lisp object obj. Its result reflects identity of obj, but not its contents.

If two objects obj1 and obj2 are eq, then (sxhash-eq obj1) and (sxhash-eq obj2) are the same integer.

Function: sxhash-eql obj

This function returns a hash code for Lisp object obj suitable for eql comparison. I.e. it reflects identity of obj except for the case where the object is a bignum or a float number, in which case a hash code is generated for the value.

If two objects obj1 and obj2 are eql, then (sxhash-eql obj1) and (sxhash-eql obj2) are the same integer.

This example creates a hash table whose keys are strings that are compared case-insensitively.

(defun string-hash-ignore-case (a)
  (sxhash-equal (upcase a)))

(define-hash-table-test 'ignore-case
  'string-equal-ignore-case 'string-hash-ignore-case)

(make-hash-table :test 'ignore-case)

Here is how you could define a hash table test equivalent to the predefined test value equal. The keys can be any Lisp object, and equal-looking objects are considered the same key.

(define-hash-table-test 'contents-hash 'equal 'sxhash-equal)

(make-hash-table :test 'contents-hash)

Lisp programs should not rely on hash codes being preserved between Emacs sessions, as the implementation of the hash functions uses some details of the object storage that can change between sessions and between different architectures.


8.4 Other Hash Table Functions

Here are some other functions for working with hash tables.

Function: hash-table-p table

This returns non-nil if table is a hash table object.

Function: copy-hash-table table

This function creates and returns a copy of table. Only the table itself is copied—the keys and values are shared.

Function: hash-table-count table

This function returns the actual number of entries in table.

Function: hash-table-test table

This returns the test value that was given when table was created, to specify how to hash and compare keys. See make-hash-table (see Creating Hash Tables).

Function: hash-table-weakness table

This function returns the weak value that was specified for hash table table.

Function: hash-table-rehash-size table

This returns the rehash size of table.

Function: hash-table-rehash-threshold table

This returns the rehash threshold of table.

Function: hash-table-size table

This returns the current nominal size of table.


9 Symbols

A symbol is an object with a unique name. This chapter describes symbols, their components, their property lists, and how they are created and interned. Separate chapters describe the use of symbols as variables and as function names; see Variables, and Functions. For the precise read syntax for symbols, see Symbol Type.

You can test whether an arbitrary Lisp object is a symbol with symbolp:

Function: symbolp object

This function returns t if object is a symbol, nil otherwise.


9.1 Symbol Components

Each symbol has four components (or “cells”), each of which references another object:

Print name

The symbol’s name.

Value

The symbol’s current value as a variable.

Function

The symbol’s function definition. It can also hold a symbol, a keymap, or a keyboard macro.

Property list

The symbol’s property list.

The print name cell always holds a string, and cannot be changed. Each of the other three cells can be set to any Lisp object.

The print name cell holds the string that is the name of a symbol. Since symbols are represented textually by their names, it is important not to have two symbols with the same name. The Lisp reader ensures this: every time it reads a symbol, it looks for an existing symbol with the specified name before it creates a new one. To get a symbol’s name, use the function symbol-name (see Creating and Interning Symbols). However, although each symbol has only one unique print name, it is nevertheless possible to refer to that same symbol via different alias names called “shorthands” (see Shorthands).

The value cell holds a symbol’s value as a variable, which is what you get if the symbol itself is evaluated as a Lisp expression. See Variables, for details about how values are set and retrieved, including complications such as local bindings and scoping rules. Most symbols can have any Lisp object as a value, but certain special symbols have values that cannot be changed; these include nil and t, and any symbol whose name starts with ‘:’ (those are called keywords). See Variables that Never Change.

The function cell holds a symbol’s function definition. Often, we refer to “the function foo” when we really mean the function stored in the function cell of foo; we make the distinction explicit only when necessary. Typically, the function cell is used to hold a function (see Functions) or a macro (see Macros). However, it can also be used to hold a symbol (see Symbol Function Indirection), keyboard macro (see Keyboard Macros), keymap (see Keymaps), or autoload object (see Autoloading). To get the contents of a symbol’s function cell, use the function symbol-function (see Accessing Function Cell Contents).

The property list cell normally should hold a correctly formatted property list. To get a symbol’s property list, use the function symbol-plist. See Symbol Properties.

The function cell or the value cell may be void, which means that the cell does not reference any object. (This is not the same thing as holding the symbol void, nor the same as holding the symbol nil.) Examining a function or value cell that is void results in an error, such as ‘Symbol's value as variable is void’.

Because each symbol has separate value and function cells, variables names and function names do not conflict. For example, the symbol buffer-file-name has a value (the name of the file being visited in the current buffer) as well as a function definition (a primitive function that returns the name of the file):

buffer-file-name
     ⇒ "/gnu/elisp/symbols.texi"
(symbol-function 'buffer-file-name)
     ⇒ #<subr buffer-file-name>

9.2 Defining Symbols

A definition is a special kind of Lisp expression that announces your intention to use a symbol in a particular way. It typically specifies a value or meaning for the symbol for one kind of use, plus documentation for its meaning when used in this way. Thus, when you define a symbol as a variable, you can supply an initial value for the variable, plus documentation for the variable.

defvar and defconst are special forms that define a symbol as a global variable—a variable that can be accessed at any point in a Lisp program. See Variables, for details about variables. To define a customizable variable, use the defcustom macro, which also calls defvar as a subroutine (see Customization Settings).

In principle, you can assign a variable value to any symbol with setq, whether or not it has first been defined as a variable. However, you ought to write a variable definition for each global variable that you want to use; otherwise, your Lisp program may not act correctly if it is evaluated with lexical scoping enabled (see Scoping Rules for Variable Bindings).

defun defines a symbol as a function, creating a lambda expression and storing it in the function cell of the symbol. This lambda expression thus becomes the function definition of the symbol. (The term “function definition”, meaning the contents of the function cell, is derived from the idea that defun gives the symbol its definition as a function.) defsubst and defalias are two other ways of defining a function. See Functions.

defmacro defines a symbol as a macro. It creates a macro object and stores it in the function cell of the symbol. Note that a given symbol can be a macro or a function, but not both at once, because both macro and function definitions are kept in the function cell, and that cell can hold only one Lisp object at any given time. See Macros.

As previously noted, Emacs Lisp allows the same symbol to be defined both as a variable (e.g., with defvar) and as a function or macro (e.g., with defun). Such definitions do not conflict.

These definitions also act as guides for programming tools. For example, the C-h f and C-h v commands create help buffers containing links to the relevant variable, function, or macro definitions. See Name Help in The GNU Emacs Manual.


9.3 Creating and Interning Symbols

To understand how symbols are created in GNU Emacs Lisp, you must know how Lisp reads them. Lisp must ensure that it finds the same symbol every time it reads the same sequence of characters in the same context. Failure to do so would cause complete confusion.

When the Lisp reader encounters a name that references a symbol in the source code, it reads all the characters of that name. Then it looks up that name in a table called an obarray to find the symbol that the programmer meant. The technique used in this lookup is called “hashing”, an efficient method of looking something up by converting a sequence of characters to a number, known as a “hash code”. For example, instead of searching a telephone book cover to cover when looking up Jan Jones, you start with the J’s and go from there. That is a simple version of hashing. Each element of the obarray is a bucket which holds all the symbols with a given hash code; to look for a given name, it is sufficient to look through all the symbols in the bucket for that name’s hash code. (The same idea is used for general Emacs hash tables, but they are a different data type; see Hash Tables.)

When looking up names, the Lisp reader also considers “shorthands”. If the programmer supplied them, this allows the reader to find a symbol even if its name isn’t present in its full form in the source code. Of course, the reader needs to be aware of some pre-established context about such shorthands, much as one needs context to be to able to refer uniquely to Jan Jones by just the name “Jan”: it’s probably fine when amongst the Joneses, or when Jan has been mentioned recently, but very ambiguous in any other situation. See Shorthands.

If a symbol with the desired name is found, the reader uses that symbol. If the obarray does not contain a symbol with that name, the reader makes a new symbol and adds it to the obarray. Finding or adding a symbol with a certain name is called interning it, and the symbol is then called an interned symbol.

Interning ensures that each obarray has just one symbol with any particular name. Other like-named symbols may exist, but not in the same obarray. Thus, the reader gets the same symbols for the same names, as long as you keep reading with the same obarray.

Interning usually happens automatically in the reader, but sometimes other programs may want to do it. For example, after the M-x command obtains the command name as a string using the minibuffer, it then interns the string, to get the interned symbol with that name. As another example, a hypothetical telephone book program could intern the name of each looked up person’s name as a symbol, even if the obarray did not contain it, so that it could attach information to that new symbol, such as the last time someone looked it up.

No obarray contains all symbols; in fact, some symbols are not in any obarray. They are called uninterned symbols. An uninterned symbol has the same four cells as other symbols; however, the only way to gain access to it is by finding it in some other object or as the value of a variable. Uninterned symbols are sometimes useful in generating Lisp code, see below.

In Emacs Lisp, an obarray is actually a vector. Each element of the vector is a bucket; its value is either an interned symbol whose name hashes to that bucket, or 0 if the bucket is empty. Each interned symbol has an internal link (invisible to the user) to the next symbol in the bucket. Because these links are invisible, there is no way to find all the symbols in an obarray except using mapatoms (below). The order of symbols in a bucket is not significant.

In an empty obarray, every element is 0, so you can create an obarray with (make-vector length 0). This is the only valid way to create an obarray. Prime numbers as lengths tend to result in good hashing; lengths one less than a power of two are also good.

Do not try to put symbols in an obarray yourself. This does not work—only intern can enter a symbol in an obarray properly.

Common Lisp note: Unlike Common Lisp, Emacs Lisp does not provide for interning the same name in several different “packages”, thus creating multiple symbols with the same name but different packages. Emacs Lisp provides a different namespacing system called “shorthands” (see Shorthands).

Most of the functions below take a name and sometimes an obarray as arguments. A wrong-type-argument error is signaled if the name is not a string, or if the obarray is not a vector.

Function: symbol-name symbol

This function returns the string that is symbol’s name. For example:

(symbol-name 'foo)
     ⇒ "foo"

Warning: Never alter the string returned by that function. Doing that might make Emacs dysfunctional, and might even crash Emacs.

Creating an uninterned symbol is useful in generating Lisp code, because an uninterned symbol used as a variable in the code you generate cannot clash with any variables used in other Lisp programs.

Function: make-symbol name

This function returns a newly-allocated, uninterned symbol whose name is name (which must be a string). Its value and function definition are void, and its property list is nil. In the example below, the value of sym is not eq to foo because it is a distinct uninterned symbol whose name is also ‘foo’.

(setq sym (make-symbol "foo"))
     ⇒ foo
(eq sym 'foo)
     ⇒ nil
Function: gensym &optional prefix

This function returns a symbol using make-symbol, whose name is made by appending gensym-counter to prefix and incrementing that counter, guaranteeing that no two calls to this function will generate a symbol with the same name. The prefix defaults to "g".

To avoid problems when accidentally interning printed representation of generated code (see Printed Representation and Read Syntax), it is recommended to use gensym instead of make-symbol.

Function: intern name &optional obarray

This function returns the interned symbol whose name is name. If there is no such symbol in the obarray obarray, intern creates a new one, adds it to the obarray, and returns it. If obarray is omitted, the value of the global variable obarray is used.

(setq sym (intern "foo"))
     ⇒ foo
(eq sym 'foo)
     ⇒ t

(setq sym1 (intern "foo" other-obarray))
     ⇒ foo
(eq sym1 'foo)
     ⇒ nil

Common Lisp note: In Common Lisp, you can intern an existing symbol in an obarray. In Emacs Lisp, you cannot do this, because the argument to intern must be a string, not a symbol.

Function: intern-soft name &optional obarray

This function returns the symbol in obarray whose name is name, or nil if obarray has no symbol with that name. Therefore, you can use intern-soft to test whether a symbol with a given name is already interned. If obarray is omitted, the value of the global variable obarray is used.

The argument name may also be a symbol; in that case, the function returns name if name is interned in the specified obarray, and otherwise nil.

(intern-soft "frazzle")        ; No such symbol exists.
     ⇒ nil
(make-symbol "frazzle")        ; Create an uninterned one.
     ⇒ frazzle
(intern-soft "frazzle")        ; That one cannot be found.
     ⇒ nil
(setq sym (intern "frazzle"))  ; Create an interned one.
     ⇒ frazzle
(intern-soft "frazzle")        ; That one can be found!
     ⇒ frazzle
(eq sym 'frazzle)              ; And it is the same one.
     ⇒ t
Variable: obarray

This variable is the standard obarray for use by intern and read.

Function: mapatoms function &optional obarray

This function calls function once with each symbol in the obarray obarray. Then it returns nil. If obarray is omitted, it defaults to the value of obarray, the standard obarray for ordinary symbols.

(setq count 0)
     ⇒ 0
(defun count-syms (s)
  (setq count (1+ count)))
     ⇒ count-syms
(mapatoms 'count-syms)
     ⇒ nil
count
     ⇒ 1871

See documentation in Access to Documentation Strings, for another example using mapatoms.

Function: unintern symbol obarray

This function deletes symbol from the obarray obarray. If symbol is not actually in the obarray, unintern does nothing. If obarray is nil, the current obarray is used.

If you provide a string instead of a symbol as symbol, it stands for a symbol name. Then unintern deletes the symbol (if any) in the obarray which has that name. If there is no such symbol, unintern does nothing.

If unintern does delete a symbol, it returns t. Otherwise it returns nil.


9.4 Symbol Properties

A symbol may possess any number of symbol properties, which can be used to record miscellaneous information about the symbol. For example, when a symbol has a risky-local-variable property with a non-nil value, that means the variable which the symbol names is a risky file-local variable (see File Local Variables).

Each symbol’s properties and property values are stored in the symbol’s property list cell (see Symbol Components), in the form of a property list (see Property Lists).


9.4.1 Accessing Symbol Properties

The following functions can be used to access symbol properties.

Function: get symbol property

This function returns the value of the property named property in symbol’s property list. If there is no such property, it returns nil. Thus, there is no distinction between a value of nil and the absence of the property.

The name property is compared with the existing property names using eq, so any object is a legitimate property.

See put for an example.

Function: put symbol property value

This function puts value onto symbol’s property list under the property name property, replacing any previous property value. The put function returns value.

(put 'fly 'verb 'transitive)
     ⇒'transitive
(put 'fly 'noun '(a buzzing little bug))
     ⇒ (a buzzing little bug)
(get 'fly 'verb)
     ⇒ transitive
(symbol-plist 'fly)
     ⇒ (verb transitive noun (a buzzing little bug))
Function: symbol-plist symbol

This function returns the property list of symbol.

Function: setplist symbol plist

This function sets symbol’s property list to plist. Normally, plist should be a well-formed property list, but this is not enforced. The return value is plist.

(setplist 'foo '(a 1 b (2 3) c nil))
     ⇒ (a 1 b (2 3) c nil)
(symbol-plist 'foo)
     ⇒ (a 1 b (2 3) c nil)

For symbols in special obarrays, which are not used for ordinary purposes, it may make sense to use the property list cell in a nonstandard fashion; in fact, the abbrev mechanism does so (see Abbrevs and Abbrev Expansion).

You could define put in terms of setplist and plist-put, as follows:

(defun put (symbol prop value)
  (setplist symbol
            (plist-put (symbol-plist symbol) prop value)))
Function: function-get symbol property &optional autoload

This function is identical to get, except that if symbol is the name of a function alias, it looks in the property list of the symbol naming the actual function. See Defining Functions. If the optional argument autoload is non-nil, and symbol is auto-loaded, this function will try to autoload it, since autoloading might set property of symbol. If autoload is the symbol macro, only try autoloading if symbol is an auto-loaded macro.

Function: function-put function property value

This function sets property of function to value. function should be a symbol. This function is preferred to calling put for setting properties of a function, because it will allow us some day to implement remapping of old properties to new ones.


9.4.2 Standard Symbol Properties

Here, we list the symbol properties which are used for special purposes in Emacs. In the following table, whenever we say “the named function”, that means the function whose name is the relevant symbol; similarly for “the named variable” etc.

:advertised-binding

This property value specifies the preferred key binding, when showing documentation, for the named function. See Substituting Key Bindings in Documentation.

char-table-extra-slots

The value, if non-nil, specifies the number of extra slots in the named char-table type. See Char-Tables.

customized-face
face-defface-spec
saved-face
theme-face

These properties are used to record a face’s standard, saved, customized, and themed face specs. Do not set them directly; they are managed by defface and related functions. See Defining Faces.

customized-value
saved-value
standard-value
theme-value

These properties are used to record a customizable variable’s standard value, saved value, customized-but-unsaved value, and themed values. Do not set them directly; they are managed by defcustom and related functions. See Defining Customization Variables.

definition-name

This property is used to find the definition of a symbol in the source code, when it might be hard to find the definition by textual search of the source file. For example, a define-derived-mode (see Defining Derived Modes) might define a mode-specific function or a variable implicitly; or your Lisp program might generate a run-time call to defun to define a function (see Defining Functions). In these and similar cases, the definition-name property of the symbol should be another symbol whose definition can be found by textual search and whose code defines the original symbol. In the example with define-derived-mode, the value of this property of the functions and variables it defines should be the mode symbol. The Emacs Help commands such as C-h f (see Help in The GNU Emacs Manual) use this property to show the definition of a symbol via a button in the *Help* buffer where the symbol’s documentation is shown.

disabled

If the value is non-nil, the named function is disabled as a command. See Disabling Commands.

face-documentation

The value stores the documentation string of the named face. This is set automatically by defface. See Defining Faces.

history-length

The value, if non-nil, specifies the maximum minibuffer history length for the named history list variable. See Minibuffer History.

interactive-form

The value is an interactive form for the named function. Normally, you should not set this directly; use the interactive special form instead. See Interactive Call.

menu-enable

The value is an expression for determining whether the named menu item should be enabled in menus. See Simple Menu Items.

mode-class

If the value is special, the named major mode is special. See Major Mode Conventions.

permanent-local

If the value is non-nil, the named variable is a buffer-local variable whose value should not be reset when changing major modes. See Creating and Deleting Buffer-Local Bindings.

permanent-local-hook

If the value is non-nil, the named function should not be deleted from the local value of a hook variable when changing major modes. See Setting Hooks.

pure

If the value is non-nil, the named function is considered to be pure (see What Is a Function?). Calls with constant arguments can be evaluated at compile time. This may shift run time errors to compile time. Not to be confused with pure storage (see Pure Storage).

risky-local-variable

If the value is non-nil, the named variable is considered risky as a file-local variable. See File Local Variables.

safe-function

If the value is non-nil, the named function is considered generally safe for evaluation. See Determining whether a Function is Safe to Call.

safe-local-eval-function

If the value is non-nil, the named function is safe to call in file-local evaluation forms. See File Local Variables.

safe-local-variable

The value specifies a function for determining safe file-local values for the named variable. See File Local Variables. Since this value is consulted when loading files, the function should be efficient and should ideally not lead to loading any libraries to determine the safeness (e.g., it should not be an autoloaded function).

side-effect-free

A non-nil value indicates that the named function is free of side effects (see What Is a Function?), so the byte compiler may ignore a call whose value is unused. If the property’s value is error-free, the byte compiler may even delete such unused calls. In addition to byte compiler optimizations, this property is also used for determining function safety (see Determining whether a Function is Safe to Call).

undo-inhibit-region

If non-nil, the named function prevents the undo operation from being restricted to the active region, if undo is invoked immediately after the function. See Undo.

variable-documentation

If non-nil, this specifies the named variable’s documentation string. This is set automatically by defvar and related functions. See Defining Faces.


9.5 Shorthands

The symbol shorthands, sometimes known as “renamed symbols”, are symbolic forms found in Lisp source. They’re just like regular symbolic forms, except that when the Lisp reader encounters them, it produces symbols which have a different and usually longer print name (see Symbol Components).

It is useful to think of shorthands as abbreviating the full names of intended symbols. Despite this, do not confuse shorthands with the Abbrev system (see Abbrevs and Abbrev Expansion).

Shorthands make Emacs Lisp’s namespacing etiquette easier to work with. Since all symbols are stored in a single obarray (see Creating and Interning Symbols), programmers commonly prefix each symbol name with the name of the library where it originates. For example, the functions text-property-search-forward and text-property-search-backward both belong to the text-property-search.el library (see Loading). By properly prefixing symbol names, one effectively prevents clashes between similarly named symbols which belong to different libraries and thus do different things. However, this practice commonly originates very long symbols names, which are inconvenient to type and read after a while. Shorthands solve these issues in a clean way.

Variable: read-symbol-shorthands

This variable’s value is an alist whose elements have the form (shorthand-prefix . longhand-prefix). Each element instructs the Lisp reader to read every symbol form which starts with shorthand-prefix as if it started with longhand-prefix instead.

This variable may only be set in file-local variables (see Local Variables in Files in The GNU Emacs Manual).

Here’s an example of shorthands usage in a hypothetical string manipulating library some-nice-string-utils.el.

(defun some-nice-string-utils-split (separator s &optional omit-nulls)
  "A match-data saving variant of `split-string'."
  (save-match-data (split-string s separator omit-nulls)))

(defun some-nice-string-utils-lines (s)
  "Split string S at newline characters into a list of strings."
  (some-nice-string-utils-split "\\(\r\n\\|[\n\r]\\)" s))

As can be seen, it’s quite tedious to read or develop this code since the symbol names to type are so long. We can use shorthands to alleviate that.

(defun snu-split (separator s &optional omit-nulls)
  "A match-data saving variation on `split-string'."
  (save-match-data (split-string s separator omit-nulls)))

(defun snu-lines (s)
  "Split string S into a list of strings on newline characters."
  (snu-split "\\(\r\n\\|[\n\r]\\)" s))

;; Local Variables:
;; read-symbol-shorthands: (("snu-" . "some-nice-string-utils-"))
;; End:

Even though the two excerpts look different, they are quite identical after the Lisp reader processes them. Both will lead to the very same symbols being interned (see Creating and Interning Symbols). Thus loading or byte-compiling any of the two files has equivalent results. The shorthands snu-split and snu-lines used in the second version are not interned in the obarray. This is easily seen by moving point to the location where the shorthands are used and waiting for ElDoc (see Local Variables in Files in The GNU Emacs Manual) to hint at the true full name of the symbol under point in the echo area.

Since read-symbol-shorthands is a file-local variable, it is possible that multiple libraries depending on some-nice-string-utils-lines.el refer to the same symbols under different shorthands, or not using shorthands at all. In the next example, the my-tricks.el library refers to the symbol some-nice-string-utils-lines using the sns- prefix instead of snu-.

(defun t-reverse-lines (s) (string-join (reverse (sns-lines s)) "\n")

;; Local Variables:
;; read-symbol-shorthands: (("t-" . "my-tricks-")
;;                          ("sns-" . "some-nice-string-utils-"))
;; End:

9.5.1 Exceptions

There are two exceptions to rules governing Shorthand transformations:

  • Symbol forms comprised entirely of characters in the Emacs Lisp symbol constituent class (see Table of Syntax Classes) are not transformed. For example, it’s possible to use - or /= as shorthand prefixes, but that won’t shadow the arithmetic functions of those names.
  • Symbol forms whose names start with ‘#_’ are not transformed.

9.6 Symbols with Position

A symbol with position is a symbol, the bare symbol, together with an unsigned integer called the position. These objects are intended for use by the byte compiler, which records in them the position of each symbol occurrence and uses those positions in warning and error messages.

The printed representation of a symbol with position uses the hash notation outlined in Printed Representation and Read Syntax. It looks like ‘#<symbol foo at 12345>’. It has no read syntax. You can cause just the bare symbol to be printed by binding the variable print-symbols-bare to non-nil around the print operation. The byte compiler does this before writing its output to the compiled Lisp file.

For most purposes, when the flag variable symbols-with-pos-enabled is non-nil, symbols with positions behave just as bare symbols do. For example, ‘(eq #<symbol foo at 12345> foo)’ has a value t when that variable is set (but nil when it isn’t set). Most of the time in Emacs this variable is nil, but the byte compiler binds it to t when it runs.

Typically, symbols with position are created by the byte compiler calling the reader function read-positioning-symbols (see Input Functions). One can also be created with the function position-symbol.

Variable: symbols-with-pos-enabled

When this variable is non-nil, symbols with position behave like the contained bare symbol. Emacs runs a little more slowly in this case.

Variable: print-symbols-bare

When bound to non-nil, the Lisp printer prints only the bare symbol of a symbol with position, ignoring the position.

Function: symbol-with-pos-p symbol.

This function returns t if symbol is a symbol with position, nil otherwise.

Function: bare-symbol symbol

This function returns the bare symbol contained in symbol, or symbol itself if it is already a bare symbol. For any other type of object, it signals an error.

Function: symbol-with-pos-pos symbol

This function returns the position, a number, from a symbol with position. For any other type of object, it signals an error.

Function: position-symbol sym pos

Make a new symbol with position. sym is either a bare symbol or a symbol with position, and supplies the symbol part of the new object. pos is either an integer which becomes the number part of the new object, or a symbol with position whose position is used. Emacs signals an error if either argument is invalid.


10 Evaluation

The evaluation of expressions in Emacs Lisp is performed by the Lisp interpreter—a program that receives a Lisp object as input and computes its value as an expression. How it does this depends on the data type of the object, according to rules described in this chapter. The interpreter runs automatically to evaluate portions of your program, but can also be called explicitly via the Lisp primitive function eval.


10.1 Introduction to Evaluation

The Lisp interpreter, or evaluator, is the part of Emacs that computes the value of an expression that is given to it. When a function written in Lisp is called, the evaluator computes the value of the function by evaluating the expressions in the function body. Thus, running any Lisp program really means running the Lisp interpreter.

A Lisp object that is intended for evaluation is called a form or expression7. The fact that forms are data objects and not merely text is one of the fundamental differences between Lisp-like languages and typical programming languages. Any object can be evaluated, but in practice only numbers, symbols, lists and strings are evaluated very often.

In subsequent sections, we will describe the details of what evaluation means for each kind of form.

It is very common to read a Lisp form and then evaluate the form, but reading and evaluation are separate activities, and either can be performed alone. Reading per se does not evaluate anything; it converts the printed representation of a Lisp object to the object itself. It is up to the caller of read to specify whether this object is a form to be evaluated, or serves some entirely different purpose. See Input Functions.

Evaluation is a recursive process, and evaluating a form often involves evaluating parts within that form. For instance, when you evaluate a function call form such as (car x), Emacs first evaluates the argument (the subform x). After evaluating the argument, Emacs executes the function (car), and if the function is written in Lisp, execution works by evaluating the body of the function (in this example, however, car is not a Lisp function; it is a primitive function implemented in C). See Functions, for more information about functions and function calls.

Evaluation takes place in a context called the environment, which consists of the current values and bindings of all Lisp variables (see Variables).8 Whenever a form refers to a variable without creating a new binding for it, the variable evaluates to the value given by the current environment. Evaluating a form may also temporarily alter the environment by binding variables (see Local Variables).

Evaluating a form may also make changes that persist; these changes are called side effects. An example of a form that produces a side effect is (setq foo 1).

Do not confuse evaluation with command key interpretation. The editor command loop translates keyboard input into a command (an interactively callable function) using the active keymaps, and then uses call-interactively to execute that command. Executing the command usually involves evaluation, if the command is written in Lisp; however, this step is not considered a part of command key interpretation. See Command Loop.


10.2 Kinds of Forms

A Lisp object that is intended to be evaluated is called a form (or an expression). How Emacs evaluates a form depends on its data type. Emacs has three different kinds of form that are evaluated differently: symbols, lists, and all other types. This section describes all three kinds, one by one, starting with the other types, which are self-evaluating forms.


10.2.1 Self-Evaluating Forms

A self-evaluating form is any form that is not a list or symbol. Self-evaluating forms evaluate to themselves: the result of evaluation is the same object that was evaluated. Thus, the number 25 evaluates to 25, and the string "foo" evaluates to the string "foo". Likewise, evaluating a vector does not cause evaluation of the elements of the vector—it returns the same vector with its contents unchanged.

'123               ; A number, shown without evaluation.
     ⇒ 123
123                ; Evaluated as usual—result is the same.
     ⇒ 123
(eval '123)        ; Evaluated "by hand"—result is the same.
     ⇒ 123
(eval (eval '123)) ; Evaluating twice changes nothing.
     ⇒ 123

A self-evaluating form yields a value that becomes part of the program, and you should not try to modify it via setcar, aset or similar operations. The Lisp interpreter might unify the constants yielded by your program’s self-evaluating forms, so that these constants might share structure. See Mutability.

It is common to write numbers, characters, strings, and even vectors in Lisp code, taking advantage of the fact that they self-evaluate. However, it is quite unusual to do this for types that lack a read syntax, because there’s no way to write them textually. It is possible to construct Lisp expressions containing these types by means of a Lisp program. Here is an example:

;; Build an expression containing a buffer object.
(setq print-exp (list 'print (current-buffer)))
     ⇒ (print #<buffer eval.texi>)
;; Evaluate it.
(eval print-exp)
     -| #<buffer eval.texi>
     ⇒ #<buffer eval.texi>

10.2.2 Symbol Forms

When a symbol is evaluated, it is treated as a variable. The result is the variable’s value, if it has one. If the symbol has no value as a variable, the Lisp interpreter signals an error. For more information on the use of variables, see Variables.

In the following example, we set the value of a symbol with setq. Then we evaluate the symbol, and get back the value that setq stored.

(setq a 123)
     ⇒ 123
(eval 'a)
     ⇒ 123
a
     ⇒ 123

The symbols nil and t are treated specially, so that the value of nil is always nil, and the value of t is always t; you cannot set or bind them to any other values. Thus, these two symbols act like self-evaluating forms, even though eval treats them like any other symbol. A symbol whose name starts with ‘:’ also self-evaluates in the same way; likewise, its value ordinarily cannot be changed. See Variables that Never Change.


10.2.3 Classification of List Forms

A form that is a nonempty list is either a function call, a macro call, or a special form, according to its first element. These three kinds of forms are evaluated in different ways, described below. The remaining list elements constitute the arguments for the function, macro, or special form.

The first step in evaluating a nonempty list is to examine its first element. This element alone determines what kind of form the list is and how the rest of the list is to be processed. The first element is not evaluated, as it would be in some Lisp dialects such as Scheme.


10.2.4 Symbol Function Indirection

If the first element of the list is a symbol then evaluation examines the symbol’s function cell, and uses its contents instead of the original symbol. If the contents are another symbol, this process, called symbol function indirection, is repeated until it obtains a non-symbol. See Naming a Function, for more information about symbol function indirection.

One possible consequence of this process is an infinite loop, in the event that a symbol’s function cell refers to the same symbol. Otherwise, we eventually obtain a non-symbol, which ought to be a function or other suitable object.

More precisely, we should now have a Lisp function (a lambda expression), a byte-code function, a primitive function, a Lisp macro, a special form, or an autoload object. Each of these types is a case described in one of the following sections. If the object is not one of these types, Emacs signals an invalid-function error.

The following example illustrates the symbol indirection process. We use fset to set the function cell of a symbol and symbol-function to get the function cell contents (see Accessing Function Cell Contents). Specifically, we store the symbol car into the function cell of first, and the symbol first into the function cell of erste.

;; Build this function cell linkage:
;;   -------------       -----        -------        -------
;;  | #<subr car> | <-- | car |  <-- | first |  <-- | erste |
;;   -------------       -----        -------        -------
(symbol-function 'car)
     ⇒ #<subr car>
(fset 'first 'car)
     ⇒ car
(fset 'erste 'first)
     ⇒ first
(erste '(1 2 3))   ; Call the function referenced by erste.
     ⇒ 1

By contrast, the following example calls a function without any symbol function indirection, because the first element is an anonymous Lisp function, not a symbol.

((lambda (arg) (erste arg))
 '(1 2 3))
     ⇒ 1

Executing the function itself evaluates its body; this does involve symbol function indirection when calling erste.

This form is rarely used and is now deprecated. Instead, you should write it as:

(funcall (lambda (arg) (erste arg))
         '(1 2 3))

or just

(let ((arg '(1 2 3))) (erste arg))

The built-in function indirect-function provides an easy way to perform symbol function indirection explicitly.

Function: indirect-function function &optional noerror

This function returns the meaning of function as a function. If function is a symbol, then it finds function’s function definition and starts over with that value. If function is not a symbol, then it returns function itself.

This function returns nil if the final symbol is unbound. It signals a cyclic-function-indirection error if there is a loop in the chain of symbols.

The optional argument noerror is obsolete, kept for backward compatibility, and has no effect.

Here is how you could define indirect-function in Lisp:

(defun indirect-function (function)
  (if (and function
           (symbolp function))
      (indirect-function (symbol-function function))
    function))

10.2.5 Evaluation of Function Forms

If the first element of a list being evaluated is a Lisp function object, byte-code object or primitive function object, then that list is a function call. For example, here is a call to the function +:

(+ 1 x)

The first step in evaluating a function call is to evaluate the remaining elements of the list from left to right. The results are the actual argument values, one value for each list element. The next step is to call the function with this list of arguments, effectively using the function apply (see Calling Functions). If the function is written in Lisp, the arguments are used to bind the argument variables of the function (see Lambda Expressions); then the forms in the function body are evaluated in order, and the value of the last body form becomes the value of the function call.


10.2.6 Lisp Macro Evaluation

If the first element of a list being evaluated is a macro object, then the list is a macro call. When a macro call is evaluated, the elements of the rest of the list are not initially evaluated. Instead, these elements themselves are used as the arguments of the macro. The macro definition computes a replacement form, called the expansion of the macro, to be evaluated in place of the original form. The expansion may be any sort of form: a self-evaluating constant, a symbol, or a list. If the expansion is itself a macro call, this process of expansion repeats until some other sort of form results.

Ordinary evaluation of a macro call finishes by evaluating the expansion. However, the macro expansion is not necessarily evaluated right away, or at all, because other programs also expand macro calls, and they may or may not evaluate the expansions.

Normally, the argument expressions are not evaluated as part of computing the macro expansion, but instead appear as part of the expansion, so they are computed when the expansion is evaluated.

For example, given a macro defined as follows:

(defmacro cadr (x)
  (list 'car (list 'cdr x)))

an expression such as (cadr (assq 'handler list)) is a macro call, and its expansion is:

(car (cdr (assq 'handler list)))

Note that the argument (assq 'handler list) appears in the expansion.

See Macros, for a complete description of Emacs Lisp macros.


10.2.7 Special Forms

A special form is a primitive specially marked so that its arguments are not all evaluated. Most special forms define control structures or perform variable bindings—things which functions cannot do.

Each special form has its own rules for which arguments are evaluated and which are used without evaluation. Whether a particular argument is evaluated may depend on the results of evaluating other arguments.

If an expression’s first symbol is that of a special form, the expression should follow the rules of that special form; otherwise, Emacs’s behavior is not well-defined (though it will not crash). For example, ((lambda (x) x . 3) 4) contains a subexpression that begins with lambda but is not a well-formed lambda expression, so Emacs may signal an error, or may return 3 or 4 or nil, or may behave in other ways.

Function: special-form-p object

This predicate tests whether its argument is a special form, and returns t if so, nil otherwise.

Here is a list, in alphabetical order, of all of the special forms in Emacs Lisp with a reference to where each is described.

and

see Constructs for Combining Conditions

catch

see Explicit Nonlocal Exits: catch and throw

cond

see Conditionals

condition-case

see Writing Code to Handle Errors

defconst

see Defining Global Variables

defvar

see Defining Global Variables

function

see Anonymous Functions

if

see Conditionals

interactive

see Interactive Call

lambda

see Lambda Expressions

let
let*

see Local Variables

or

see Constructs for Combining Conditions

prog1
prog2
progn

see Sequencing

quote

see Quoting

save-current-buffer

see The Current Buffer

save-excursion

see Excursions

save-restriction

see Narrowing

setq

see Setting Variable Values

setq-default

see Creating and Deleting Buffer-Local Bindings

unwind-protect

see Nonlocal Exits

while

see Iteration

Common Lisp note: Here are some comparisons of special forms in GNU Emacs Lisp and Common Lisp. setq, if, and catch are special forms in both Emacs Lisp and Common Lisp. save-excursion is a special form in Emacs Lisp, but doesn’t exist in Common Lisp. throw is a special form in Common Lisp (because it must be able to throw multiple values), but it is a function in Emacs Lisp (which doesn’t have multiple values).


10.2.8 Autoloading

The autoload feature allows you to call a function or macro whose function definition has not yet been loaded into Emacs. It specifies which file contains the definition. When an autoload object appears as a symbol’s function definition, calling that symbol as a function automatically loads the specified file; then it calls the real definition loaded from that file. The way to arrange for an autoload object to appear as a symbol’s function definition is described in Autoload.


10.3 Quoting

The special form quote returns its single argument, as written, without evaluating it. This provides a way to include constant symbols and lists, which are not self-evaluating objects, in a program. (It is not necessary to quote self-evaluating objects such as numbers, strings, and vectors.)

Special Form: quote object

This special form returns object, without evaluating it. The returned value might be shared and should not be modified. See Self-Evaluating Forms.

Because quote is used so often in programs, Lisp provides a convenient read syntax for it. An apostrophe character (‘'’) followed by a Lisp object (in read syntax) expands to a list whose first element is quote, and whose second element is the object. Thus, the read syntax 'x is an abbreviation for (quote x).

Here are some examples of expressions that use quote:

(quote (+ 1 2))
     ⇒ (+ 1 2)
(quote foo)
     ⇒ foo
'foo
     ⇒ foo
''foo
     ⇒ 'foo
'(quote foo)
     ⇒ 'foo
['foo]
     ⇒ ['foo]

Although the expressions (list '+ 1 2) and '(+ 1 2) both yield lists equal to (+ 1 2), the former yields a freshly-minted mutable list whereas the latter yields a list built from conses that might be shared and should not be modified. See Self-Evaluating Forms.

Other quoting constructs include function (see Anonymous Functions), which causes an anonymous lambda expression written in Lisp to be compiled, and ‘`’ (see Backquote), which is used to quote only part of a list, while computing and substituting other parts.


10.4 Backquote

Backquote constructs allow you to quote a list, but selectively evaluate elements of that list. In the simplest case, it is identical to the special form quote (described in the previous section; see Quoting). For example, these two forms yield identical results:

`(a list of (+ 2 3) elements)
     ⇒ (a list of (+ 2 3) elements)
'(a list of (+ 2 3) elements)
     ⇒ (a list of (+ 2 3) elements)

The special marker ‘,’ inside of the argument to backquote indicates a value that isn’t constant. The Emacs Lisp evaluator evaluates the argument of ‘,’, and puts the value in the list structure:

`(a list of ,(+ 2 3) elements)
     ⇒ (a list of 5 elements)

Substitution with ‘,’ is allowed at deeper levels of the list structure also. For example:

`(1 2 (3 ,(+ 4 5)))
     ⇒ (1 2 (3 9))

You can also splice an evaluated value into the resulting list, using the special marker ‘,@’. The elements of the spliced list become elements at the same level as the other elements of the resulting list. The equivalent code without using ‘`’ is often unreadable. Here are some examples:

(setq some-list '(2 3))
     ⇒ (2 3)
(cons 1 (append some-list '(4) some-list))
     ⇒ (1 2 3 4 2 3)
`(1 ,@some-list 4 ,@some-list)
     ⇒ (1 2 3 4 2 3)

(setq list '(hack foo bar))
     ⇒ (hack foo bar)
(cons 'use
  (cons 'the
    (cons 'words (append (cdr list) '(as elements)))))
     ⇒ (use the words foo bar as elements)
`(use the words ,@(cdr list) as elements)
     ⇒ (use the words foo bar as elements)

If a subexpression of a backquote construct has no substitutions or splices, it acts like quote in that it yields conses, vectors and strings that might be shared and should not be modified. See Self-Evaluating Forms.


10.5 Eval

Most often, forms are evaluated automatically, by virtue of their occurrence in a program being run. On rare occasions, you may need to write code that evaluates a form that is computed at run time, such as after reading a form from text being edited or getting one from a property list. On these occasions, use the eval function. Often eval is not needed and something else should be used instead. For example, to get the value of a variable, while eval works, symbol-value is preferable; or rather than store expressions in a property list that then need to go through eval, it is better to store functions instead that are then passed to funcall.

The functions and variables described in this section evaluate forms, specify limits to the evaluation process, or record recently returned values. Loading a file also does evaluation (see Loading).

It is generally cleaner and more flexible to store a function in a data structure, and call it with funcall or apply, than to store an expression in the data structure and evaluate it. Using functions provides the ability to pass information to them as arguments.

Function: eval form &optional lexical

This is the basic function for evaluating an expression. It evaluates form in the current environment, and returns the result. The type of the form object determines how it is evaluated. See Kinds of Forms.

The argument lexical specifies the scoping rule for local variables (see Scoping Rules for Variable Bindings). If it is omitted or nil, that means to evaluate form using the default dynamic scoping rule. If it is t, that means to use the lexical scoping rule. The value of lexical can also be a non-empty alist specifying a particular lexical environment for lexical bindings; however, this feature is only useful for specialized purposes, such as in Emacs Lisp debuggers. See Lexical Binding.

Since eval is a function, the argument expression that appears in a call to eval is evaluated twice: once as preparation before eval is called, and again by the eval function itself. Here is an example:

(setq foo 'bar)
     ⇒ bar
(setq bar 'baz)
     ⇒ baz
;; Here eval receives argument foo
(eval 'foo)
     ⇒ bar
;; Here eval receives argument bar, which is the value of foo
(eval foo)
     ⇒ baz

The number of currently active calls to eval is limited to max-lisp-eval-depth (see below).

Command: eval-region start end &optional stream read-function

This function evaluates the forms in the current buffer in the region defined by the positions start and end. It reads forms from the region and calls eval on them until the end of the region is reached, or until an error is signaled and not handled.

By default, eval-region does not produce any output. However, if stream is non-nil, any output produced by output functions (see Output Functions), as well as the values that result from evaluating the expressions in the region are printed using stream. See Output Streams.

If read-function is non-nil, it should be a function, which is used instead of read to read expressions one by one. This function is called with one argument, the stream for reading input. You can also use the variable load-read-function (see How Programs Do Loading) to specify this function, but it is more robust to use the read-function argument.

eval-region does not move point. It always returns nil.

Command: eval-buffer &optional buffer-or-name stream filename unibyte print

This is similar to eval-region, but the arguments provide different optional features. eval-buffer operates on the entire accessible portion of buffer buffer-or-name (see Narrowing in The GNU Emacs Manual). buffer-or-name can be a buffer, a buffer name (a string), or nil (or omitted), which means to use the current buffer. stream is used as in eval-region, unless stream is nil and print non-nil. In that case, values that result from evaluating the expressions are still discarded, but the output of the output functions is printed in the echo area. filename is the file name to use for load-history (see Unloading), and defaults to buffer-file-name (see Buffer File Name). If unibyte is non-nil, read converts strings to unibyte whenever possible.

User Option: max-lisp-eval-depth

This variable defines the maximum depth allowed in calls to eval, apply, and funcall before an error is signaled (with error message "Lisp nesting exceeds max-lisp-eval-depth").

This limit, with the associated error when it is exceeded, is how Emacs Lisp avoids infinite recursion on an ill-defined function. If you increase the value of max-lisp-eval-depth too much, such code can cause stack overflow instead. On some systems, this overflow can be handled. In that case, normal Lisp evaluation is interrupted and control is transferred back to the top level command loop (top-level). Note that there is no way to enter Emacs Lisp debugger in this situation. See Entering the Debugger on an Error.

The depth limit counts internal uses of eval, apply, and funcall, such as for calling the functions mentioned in Lisp expressions, and recursive evaluation of function call arguments and function body forms, as well as explicit calls in Lisp code.

The default value of this variable is 1600. If you set it to a value less than 100, Lisp will reset it to 100 if the given value is reached. Entry to the Lisp debugger increases the value, if there is little room left, to make sure the debugger itself has room to execute.

Variable: values

The value of this variable is a list of the values returned by all the expressions that were read, evaluated, and printed from buffers (including the minibuffer) by the standard Emacs commands which do this. (Note that this does not include evaluation in *ielm* buffers, nor evaluation using C-j, C-x C-e, and similar evaluation commands in lisp-interaction-mode.)

This variable is obsolete, and will be removed in a future version, since it constantly enlarges the memory footprint of the Emacs process. For that reason, we recommend against using it.

The elements of values are ordered most recent first.

(setq x 1)
     ⇒ 1
(list 'A (1+ 2) auto-save-default)
     ⇒ (A 3 t)
values
     ⇒ ((A 3 t) 1 …)

This variable could be useful for referring back to values of forms recently evaluated. It is generally a bad idea to print the value of values itself, since this may be very long. Instead, examine particular elements, like this:

;; Refer to the most recent evaluation result.
(nth 0 values)
     ⇒ (A 3 t)
;; That put a new element on,
;;   so all elements move back one.
(nth 1 values)
     ⇒ (A 3 t)
;; This gets the element that was next-to-most-recent
;;   before this example.
(nth 3 values)
     ⇒ 1

10.6 Deferred and Lazy Evaluation

Sometimes it is useful to delay the evaluation of an expression, for example if you want to avoid performing a time-consuming calculation if it turns out that the result is not needed in the future of the program. The thunk library provides the following functions and macros to support such deferred evaluation:

Macro: thunk-delay forms…

Return a thunk for evaluating the forms. A thunk is a closure (see Closures) that inherits the lexical environment of the thunk-delay call. Using this macro requires lexical-binding.

Function: thunk-force thunk

Force thunk to perform the evaluation of the forms specified in the thunk-delay that created the thunk. The result of the evaluation of the last form is returned. The thunk also “remembers” that it has been forced: Any further calls of thunk-force with the same thunk will just return the same result without evaluating the forms again.

Macro: thunk-let (bindings…) forms…

This macro is analogous to let but creates “lazy” variable bindings. Any binding has the form (symbol value-form). Unlike let, the evaluation of any value-form is deferred until the binding of the according symbol is used for the first time when evaluating the forms. Any value-form is evaluated at most once. Using this macro requires lexical-binding.

Example:

(defun f (number)
  (thunk-let ((derived-number
              (progn (message "Calculating 1 plus 2 times %d" number)
                     (1+ (* 2 number)))))
    (if (> number 10)
        derived-number
      number)))

(f 5)
⇒ 5

(f 12)
-| Calculating 1 plus 2 times 12
⇒ 25

Because of the special nature of lazily bound variables, it is an error to set them (e.g. with setq).

Macro: thunk-let* (bindings…) forms…

This is like thunk-let but any expression in bindings is allowed to refer to preceding bindings in this thunk-let* form. Using this macro requires lexical-binding.

(thunk-let* ((x (prog2 (message "Calculating x...")
                    (+ 1 1)
                  (message "Finished calculating x")))
             (y (prog2 (message "Calculating y...")
                    (+ x 1)
                  (message "Finished calculating y")))
             (z (prog2 (message "Calculating z...")
                    (+ y 1)
                  (message "Finished calculating z")))
             (a (prog2 (message "Calculating a...")
                    (+ z 1)
                  (message "Finished calculating a"))))
  (* z x))

-| Calculating z...
-| Calculating y...
-| Calculating x...
-| Finished calculating x
-| Finished calculating y
-| Finished calculating z
⇒ 8

thunk-let and thunk-let* use thunks implicitly: their expansion creates helper symbols and binds them to thunks wrapping the binding expressions. All references to the original variables in the body forms are then replaced by an expression that calls thunk-force with the according helper variable as the argument. So, any code using thunk-let or thunk-let* could be rewritten to use thunks, but in many cases using these macros results in nicer code than using thunks explicitly.


11 Control Structures

A Lisp program consists of a set of expressions, or forms (see Kinds of Forms). We control the order of execution of these forms by enclosing them in control structures. Control structures are special forms which control when, whether, or how many times to execute the forms they contain.

The simplest order of execution is sequential execution: first form a, then form b, and so on. This is what happens when you write several forms in succession in the body of a function, or at top level in a file of Lisp code—the forms are executed in the order written. We call this textual order. For example, if a function body consists of two forms a and b, evaluation of the function evaluates first a and then b. The result of evaluating b becomes the value of the function.

Explicit control structures make possible an order of execution other than sequential.

Emacs Lisp provides several kinds of control structure, including other varieties of sequencing, conditionals, iteration, and (controlled) jumps—all discussed below. The built-in control structures are special forms since their subforms are not necessarily evaluated or not evaluated sequentially. You can use macros to define your own control structure constructs (see Macros).


11.1 Sequencing

Evaluating forms in the order they appear is the most common way control passes from one form to another. In some contexts, such as in a function body, this happens automatically. Elsewhere you must use a control structure construct to do this: progn, the simplest control construct of Lisp.

A progn special form looks like this:

(progn a b c …)

and it says to execute the forms a, b, c, and so on, in that order. These forms are called the body of the progn form. The value of the last form in the body becomes the value of the entire progn. (progn) returns nil.

In the early days of Lisp, progn was the only way to execute two or more forms in succession and use the value of the last of them. But programmers found they often needed to use a progn in the body of a function, where (at that time) only one form was allowed. So the body of a function was made into an implicit progn: several forms are allowed just as in the body of an actual progn. Many other control structures likewise contain an implicit progn. As a result, progn is not used as much as it was many years ago. It is needed now most often inside an unwind-protect, and, or, or in the then-part of an if.

Special Form: progn forms…

This special form evaluates all of the forms, in textual order, returning the result of the final form.

(progn (print "The first form")
       (print "The second form")
       (print "The third form"))
     -| "The first form"
     -| "The second form"
     -| "The third form"
⇒ "The third form"

Two other constructs likewise evaluate a series of forms but return different values:

Special Form: prog1 form1 forms…

This special form evaluates form1 and all of the forms, in textual order, returning the result of form1.

(prog1 (print "The first form")
       (print "The second form")
       (print "The third form"))
     -| "The first form"
     -| "The second form"
     -| "The third form"
⇒ "The first form"

Here is a way to remove the first element from a list in the variable x, then return the value of that former element:

(prog1 (car x) (setq x (cdr x)))
Special Form: prog2 form1 form2 forms…

This special form evaluates form1, form2, and all of the following forms, in textual order, returning the result of form2.

(prog2 (print "The first form")
       (print "The second form")
       (print "The third form"))
     -| "The first form"
     -| "The second form"
     -| "The third form"
⇒ "The second form"

11.2 Conditionals

Conditional control structures choose among alternatives. Emacs Lisp has five conditional forms: if, which is much the same as in other languages; when and unless, which are variants of if; cond, which is a generalized case statement; and pcase, which is a generalization of cond (see Pattern-Matching Conditional).

Special Form: if condition then-form else-forms…

if chooses between the then-form and the else-forms based on the value of condition. If the evaluated condition is non-nil, then-form is evaluated and the result returned. Otherwise, the else-forms are evaluated in textual order, and the value of the last one is returned. (The else part of if is an example of an implicit progn. See Sequencing.)

If condition has the value nil, and no else-forms are given, if returns nil.

if is a special form because the branch that is not selected is never evaluated—it is ignored. Thus, in this example, true is not printed because print is never called:

(if nil
    (print 'true)
  'very-false)
⇒ very-false
Macro: when condition then-forms…

This is a variant of if where there are no else-forms, and possibly several then-forms. In particular,

(when condition a b c)

is entirely equivalent to

(if condition (progn a b c) nil)
Macro: unless condition forms…

This is a variant of if where there is no then-form:

(unless condition a b c)

is entirely equivalent to

(if condition nil
   a b c)
Special Form: cond clause…

cond chooses among an arbitrary number of alternatives. Each clause in the cond must be a list. The CAR of this list is the condition; the remaining elements, if any, the body-forms. Thus, a clause looks like this:

(condition body-forms…)

cond tries the clauses in textual order, by evaluating the condition of each clause. If the value of condition is non-nil, the clause succeeds; then cond evaluates its body-forms, and returns the value of the last of body-forms. Any remaining clauses are ignored.

If the value of condition is nil, the clause fails, so the cond moves on to the following clause, trying its condition.

A clause may also look like this:

(condition)

Then, if condition is non-nil when tested, the cond form returns the value of condition.

If every condition evaluates to nil, so that every clause fails, cond returns nil.

The following example has four clauses, which test for the cases where the value of x is a number, string, buffer and symbol, respectively:

(cond ((numberp x) x)
      ((stringp x) x)
      ((bufferp x)
       (setq temporary-hack x) ; multiple body-forms
       (buffer-name x))        ; in one clause
      ((symbolp x) (symbol-value x)))

Often we want to execute the last clause whenever none of the previous clauses was successful. To do this, we use t as the condition of the last clause, like this: (t body-forms). The form t evaluates to t, which is never nil, so this clause never fails, provided the cond gets to it at all. For example:

(setq a 5)
(cond ((eq a 'hack) 'foo)
      (t "default"))
⇒ "default"

This cond expression returns foo if the value of a is hack, and returns the string "default" otherwise.

Any conditional construct can be expressed with cond or with if. Therefore, the choice between them is a matter of style. For example:

(if a b c)
≡
(cond (a b) (t c))

It can be convenient to bind variables in conjunction with using a conditional. It’s often the case that you compute a value, and then want to do something with that value if it’s non-nil. The straightforward way to do that is to just write, for instance:

(let ((result1 (do-computation)))
  (when result1
    (let ((result2 (do-more result1)))
      (when result2
        (do-something result2)))))

Since this is a very common pattern, Emacs provides a number of macros to make this easier and more readable. The above can be written the following way instead:

(when-let ((result1 (do-computation))
           (result2 (do-more result1)))
  (do-something result2))

There’s a number of variations on this theme, and they’re briefly described below.

Macro: if-let spec then-form else-forms...

Evaluate each binding in spec in turn, like in let* (see Local Variables, stopping if a binding value is nil. If all are non-nil, return the value of then-form, otherwise the last form in else-forms.

Macro: when-let spec then-forms...

Like if-let, but without else-forms.

Macro: while-let spec then-forms...

Like when-let, but repeat until a binding in spec is nil. The return value is always nil.


11.3 Constructs for Combining Conditions

This section describes constructs that are often used together with if and cond to express complicated conditions. The constructs and and or can also be used individually as kinds of multiple conditional constructs.

Function: not condition

This function tests for the falsehood of condition. It returns t if condition is nil, and nil otherwise. The function not is identical to null, and we recommend using the name null if you are testing for an empty list.

Special Form: and conditions…

The and special form tests whether all the conditions are true. It works by evaluating the conditions one by one in the order written.

If any of the conditions evaluates to nil, then the result of the and must be nil regardless of the remaining conditions; so and returns nil right away, ignoring the remaining conditions.

If all the conditions turn out non-nil, then the value of the last of them becomes the value of the and form. Just (and), with no conditions, returns t, appropriate because all the conditions turned out non-nil. (Think about it; which one did not?)

Here is an example. The first condition returns the integer 1, which is not nil. Similarly, the second condition returns the integer 2, which is not nil. The third condition is nil, so the remaining condition is never evaluated.

(and (print 1) (print 2) nil (print 3))
     -| 1
     -| 2
⇒ nil

Here is a more realistic example of using and:

(if (and (consp foo) (eq (car foo) 'x))
    (message "foo is a list starting with x"))

Note that (car foo) is not executed if (consp foo) returns nil, thus avoiding an error.

and expressions can also be written using either if or cond. Here’s how:

(and arg1 arg2 arg3)
≡
(if arg1 (if arg2 arg3))
≡
(cond (arg1 (cond (arg2 arg3))))
Special Form: or conditions…

The or special form tests whether at least one of the conditions is true. It works by evaluating all the conditions one by one in the order written.

If any of the conditions evaluates to a non-nil value, then the result of the or must be non-nil; so or returns right away, ignoring the remaining conditions. The value it returns is the non-nil value of the condition just evaluated.

If all the conditions turn out nil, then the or expression returns nil. Just (or), with no conditions, returns nil, appropriate because all the conditions turned out nil. (Think about it; which one did not?)

For example, this expression tests whether x is either nil or the integer zero:

(or (eq x nil) (eq x 0))

Like the and construct, or can be written in terms of cond. For example:

(or arg1 arg2 arg3)
≡
(cond (arg1)
      (arg2)
      (arg3))

You could almost write or in terms of if, but not quite:

(if arg1 arg1
  (if arg2 arg2
    arg3))

This is not completely equivalent because it can evaluate arg1 or arg2 twice. By contrast, (or arg1 arg2 arg3) never evaluates any argument more than once.

Function: xor condition1 condition2

This function returns the boolean exclusive-or of condition1 and condition2. That is, xor returns nil if either both arguments are nil, or both are non-nil. Otherwise, it returns the value of that argument which is non-nil.

Note that in contrast to or, both arguments are always evaluated.


11.4 Pattern-Matching Conditional

Aside from the four basic conditional forms, Emacs Lisp also has a pattern-matching conditional form, the pcase macro, a hybrid of cond and cl-case (see Conditionals in Common Lisp Extensions) that overcomes their limitations and introduces the pattern matching programming style. The limitations that pcase overcomes are:

  • The cond form chooses among alternatives by evaluating the predicate condition of each of its clauses (see Conditionals). The primary limitation is that variables let-bound in condition are not available to the clause’s body-forms.

    Another annoyance (more an inconvenience than a limitation) is that when a series of condition predicates implement equality tests, there is a lot of repeated code. (cl-case solves this inconvenience.)

  • The cl-case macro chooses among alternatives by evaluating the equality of its first argument against a set of specific values.

    Its limitations are two-fold:

    1. The equality tests use eql.
    2. The values must be known and written in advance.

    These render cl-case unsuitable for strings or compound data structures (e.g., lists or vectors). (cond doesn’t have these limitations, but it has others, see above.)

Conceptually, the pcase macro borrows the first-arg focus of cl-case and the clause-processing flow of cond, replacing condition with a generalization of the equality test which is a variant of pattern matching, and adding facilities so that you can concisely express a clause’s predicate, and arrange to share let-bindings between a clause’s predicate and body-forms.

The concise expression of a predicate is known as a pattern. When the predicate, called on the value of the first arg, returns non-nil, we say that “the pattern matches the value” (or sometimes “the value matches the pattern”).


11.4.1 The pcase macro

For background, See Pattern-Matching Conditional.

Macro: pcase expression &rest clauses

Each clause in clauses has the form: (pattern body-forms…).

Evaluate expression to determine its value, expval. Find the first clause in clauses whose pattern matches expval and pass control to that clause’s body-forms.

If there is a match, the value of pcase is the value of the last of body-forms in the successful clause. Otherwise, pcase evaluates to nil.

Each pattern has to be a pcase pattern, which can use either one of the core patterns defined below, or one of the patterns defined via pcase-defmacro (see Extending pcase).

The rest of this subsection describes different forms of core patterns, presents some examples, and concludes with important caveats on using the let-binding facility provided by some pattern forms. A core pattern can have the following forms:

_ (underscore)

Matches any expval. This is also known as don’t care or wildcard.

'val

Matches if expval equals val. The comparison is done as if by equal (see Equality Predicates).

keyword
integer
string

Matches if expval equals the literal object. This is a special case of 'val, above, possible because literal objects of these types are self-quoting.

symbol

Matches any expval, and additionally let-binds symbol to expval, such that this binding is available to body-forms (see Dynamic Binding).

If symbol is part of a sequencing pattern seqpat (e.g., by using and, below), the binding is also available to the portion of seqpat following the appearance of symbol. This usage has some caveats, see caveats.

Two symbols to avoid are t, which behaves like _ (above) and is deprecated, and nil, which signals an error. Likewise, it makes no sense to bind keyword symbols (see Variables that Never Change).

`qpat

A backquote-style pattern. See Backquote-Style Patterns, for the details.

(cl-type type)

Matches if expval is of type type, which is a type descriptor as accepted by cl-typep (see Type Predicates in Common Lisp Extensions). Examples:

(cl-type integer)
(cl-type (integer 0 10))
(pred function)

Matches if the predicate function returns non-nil when called on expval. The test can be negated with the syntax (pred (not function)). The predicate function can have one of the following forms:

function name (a symbol)

Call the named function with one argument, expval.

Example: integerp

lambda expression

Call the anonymous function with one argument, expval (see Lambda Expressions).

Example: (lambda (n) (= 42 n))

function call with n args

Call the function (the first element of the function call) with n arguments (the other elements) and an additional n+1-th argument that is expval.

Example: (= 42)
In this example, the function is =, n is one, and the actual function call becomes: (= 42 expval).

(app function pattern)

Matches if function called on expval returns a value that matches pattern. function can take one of the forms described for pred, above. Unlike pred, however, app tests the result against pattern, rather than against a boolean truth value.

(guard boolean-expression)

Matches if boolean-expression evaluates to non-nil.

(let pattern expr)

Evaluates expr to get exprval and matches if exprval matches pattern. (It is called let because pattern can bind symbols to values using symbol.)

A sequencing pattern (also known as seqpat) is a pattern that processes its sub-pattern arguments in sequence. There are two for pcase: and and or. They behave in a similar manner to the special forms that share their name (see Constructs for Combining Conditions), but instead of processing values, they process sub-patterns.

(and pattern1…)

Attempts to match pattern1…, in order, until one of them fails to match. In that case, and likewise fails to match, and the rest of the sub-patterns are not tested. If all sub-patterns match, and matches.

(or pattern1 pattern2…)

Attempts to match pattern1, pattern2, …, in order, until one of them succeeds. In that case, or likewise matches, and the rest of the sub-patterns are not tested.

To present a consistent environment (see Introduction to Evaluation) to body-forms (thus avoiding an evaluation error on match), the set of variables bound by the pattern is the union of the variables bound by each sub-pattern. If a variable is not bound by the sub-pattern that matched, then it is bound to nil.

(rx rx-expr…)

Matches strings against the regexp rx-expr…, using the rx regexp notation (see The rx Structured Regexp Notation), as if by string-match.

In addition to the usual rx syntax, rx-expr… can contain the following constructs:

(let ref rx-expr…)

Bind the symbol ref to a submatch that matches rx-expr.... ref is bound in body-forms to the string of the submatch or nil, but can also be used in backref.

(backref ref)

Like the standard backref construct, but ref can here also be a name introduced by a previous (let ref …) construct.

Example: Advantage Over cl-case

Here’s an example that highlights some advantages pcase has over cl-case (see Conditionals in Common Lisp Extensions).

(pcase (get-return-code x)
  ;; string
  ((and (pred stringp) msg)
   (message "%s" msg))
  ;; symbol
  ('success       (message "Done!"))
  ('would-block   (message "Sorry, can't do it now"))
  ('read-only     (message "The schmilblick is read-only"))
  ('access-denied (message "You do not have the needed rights"))
  ;; default
  (code           (message "Unknown return code %S" code)))

With cl-case, you would need to explicitly declare a local variable code to hold the return value of get-return-code. Also cl-case is difficult to use with strings because it uses eql for comparison.

Example: Using and

A common idiom is to write a pattern starting with and, with one or more symbol sub-patterns providing bindings to the sub-patterns that follow (as well as to the body forms). For example, the following pattern matches single-digit integers.

(and
  (pred integerp)
  n                     ; bind n to expval
  (guard (<= -9 n 9)))

First, pred matches if (integerp expval) evaluates to non-nil. Next, n is a symbol pattern that matches anything and binds n to expval. Lastly, guard matches if the boolean expression (<= -9 n 9) (note the reference to n) evaluates to non-nil. If all these sub-patterns match, and matches.

Example: Reformulation with pcase

Here is another example that shows how to reformulate a simple matching task from its traditional implementation (function grok/traditional) to one using pcase (function grok/pcase). The docstring for both these functions is: “If OBJ is a string of the form "key:NUMBER", return NUMBER (a string). Otherwise, return the list ("149" default).” First, the traditional implementation (see Regular Expressions):

(defun grok/traditional (obj)
  (if (and (stringp obj)
           (string-match "^key:\\([[:digit:]]+\\)$" obj))
      (match-string 1 obj)
    (list "149" 'default)))

(grok/traditional "key:0")   ⇒ "0"
(grok/traditional "key:149") ⇒ "149"
(grok/traditional 'monolith) ⇒ ("149" default)

The reformulation demonstrates symbol binding as well as or, and, pred, app and let.

(defun grok/pcase (obj)
  (pcase obj
    ((or                                     ; line 1
      (and                                   ; line 2
       (pred stringp)                        ; line 3
       (pred (string-match                   ; line 4
              "^key:\\([[:digit:]]+\\)$"))   ; line 5
       (app (match-string 1)                 ; line 6
            val))                            ; line 7
      (let val (list "149" 'default)))       ; line 8
     val)))                                  ; line 9

(grok/pcase "key:0")   ⇒ "0"
(grok/pcase "key:149") ⇒ "149"
(grok/pcase 'monolith) ⇒ ("149" default)

The bulk of grok/pcase is a single clause of a pcase form, the pattern on lines 1-8, the (single) body form on line 9. The pattern is or, which tries to match in turn its argument sub-patterns, first and (lines 2-7), then let (line 8), until one of them succeeds.

As in the previous example (see Example 1), and begins with a pred sub-pattern to ensure the following sub-patterns work with an object of the correct type (string, in this case). If (stringp expval) returns nil, pred fails, and thus and fails, too.

The next pred (lines 4-5) evaluates (string-match RX expval) and matches if the result is non-nil, which means that expval has the desired form: key:NUMBER. Again, failing this, pred fails and and, too.

Lastly (in this series of and sub-patterns), app evaluates (match-string 1 expval) (line 6) to get a temporary value tmp (i.e., the “NUMBER” substring) and tries to match tmp against pattern val (line 7). Since that is a symbol pattern, it matches unconditionally and additionally binds val to tmp.

Now that app has matched, all and sub-patterns have matched, and so and matches. Likewise, once and has matched, or matches and does not proceed to try sub-pattern let (line 8).

Let’s consider the situation where obj is not a string, or it is a string but has the wrong form. In this case, one of the pred (lines 3-5) fails to match, thus and (line 2) fails to match, thus or (line 1) proceeds to try sub-pattern let (line 8).

First, let evaluates (list "149" 'default) to get ("149" default), the exprval, and then tries to match exprval against pattern val. Since that is a symbol pattern, it matches unconditionally and additionally binds val to exprval. Now that let has matched, or matches.

Note how both and and let sub-patterns finish in the same way: by trying (always successfully) to match against the symbol pattern val, in the process binding val. Thus, or always matches and control always passes to the body form (line 9). Because that is the last body form in a successfully matched pcase clause, it is the value of pcase and likewise the return value of grok/pcase (see What Is a Function?).

Caveats for symbol in Sequencing Patterns

The preceding examples all use sequencing patterns which include the symbol sub-pattern in some way. Here are some important details about that usage.

  1. When symbol occurs more than once in seqpat, the second and subsequent occurrences do not expand to re-binding, but instead expand to an equality test using eq.

    The following example features a pcase form with two clauses and two seqpat, A and B. Both A and B first check that expval is a pair (using pred), and then bind symbols to the car and cdr of expval (using one app each).

    For A, because symbol st is mentioned twice, the second mention becomes an equality test using eq. On the other hand, B uses two separate symbols, s1 and s2, both of which become independent bindings.

    (defun grok (object)
      (pcase object
        ((and (pred consp)        ; seqpat A
              (app car st)        ; first mention: st
              (app cdr st))       ; second mention: st
         (list 'eq st))
    
        ((and (pred consp)        ; seqpat B
              (app car s1)        ; first mention: s1
              (app cdr s2))       ; first mention: s2
         (list 'not-eq s1 s2))))
    
    
    
    (let ((s "yow!"))
      (grok (cons s s)))      ⇒ (eq "yow!")
    (grok (cons "yo!" "yo!")) ⇒ (not-eq "yo!" "yo!")
    (grok '(4 2))             ⇒ (not-eq 4 (2))
    
  2. Side-effecting code referencing symbol is undefined. Avoid. For example, here are two similar functions. Both use and, symbol and guard:
    (defun square-double-digit-p/CLEAN (integer)
      (pcase (* integer integer)
        ((and n (guard (< 9 n 100))) (list 'yes n))
        (sorry (list 'no sorry))))
    
    (square-double-digit-p/CLEAN 9) ⇒ (yes 81)
    (square-double-digit-p/CLEAN 3) ⇒ (no 9)
    
    
    
    (defun square-double-digit-p/MAYBE (integer)
      (pcase (* integer integer)
        ((and n (guard (< 9 (incf n) 100))) (list 'yes n))
        (sorry (list 'no sorry))))
    
    (square-double-digit-p/MAYBE 9) ⇒ (yes 81)
    (square-double-digit-p/MAYBE 3) ⇒ (yes 9)  ; WRONG!
    

    The difference is in boolean-expression in guard: CLEAN references n simply and directly, while MAYBE references n with a side-effect, in the expression (incf n). When integer is 3, here’s what happens:

    • The first n binds it to expval, i.e., the result of evaluating (* 3 3), or 9.
    • boolean-expression is evaluated:
      start:   (< 9 (incf n)        100)
      becomes: (< 9 (setq n (1+ n)) 100)
      becomes: (< 9 (setq n (1+ 9)) 100)
      
      becomes: (< 9 (setq n 10)     100)
                                         ; side-effect here!
      becomes: (< 9       n         100) ; n now bound to 10
      becomes: (< 9      10         100)
      becomes: t
      
    • Because the result of the evaluation is non-nil, guard matches, and matches, and control passes to that clause’s body forms.

    Aside from the mathematical incorrectness of asserting that 9 is a double-digit integer, there is another problem with MAYBE. The body form references n once more, yet we do not see the updated value—10—at all. What happened to it?

    To sum up, it’s best to avoid side-effecting references to symbol patterns entirely, not only in boolean-expression (in guard), but also in expr (in let) and function (in pred and app).

  3. On match, the clause’s body forms can reference the set of symbols the pattern let-binds. When seqpat is and, this set is the union of all the symbols each of its sub-patterns let-binds. This makes sense because, for and to match, all the sub-patterns must match.

    When seqpat is or, things are different: or matches at the first sub-pattern that matches; the rest of the sub-patterns are ignored. It makes no sense for each sub-pattern to let-bind a different set of symbols because the body forms have no way to distinguish which sub-pattern matched and choose among the different sets. For example, the following is invalid:

    (require 'cl-lib)
    (pcase (read-number "Enter an integer: ")
      ((or (and (pred cl-evenp)
                e-num)      ; bind e-num to expval
           o-num)           ; bind o-num to expval
       (list e-num o-num)))
    
    
    
    Enter an integer: 42
    error→ Symbol’s value as variable is void: o-num
    
    Enter an integer: 149
    error→ Symbol’s value as variable is void: e-num
    

    Evaluating body form (list e-num o-num) signals error. To distinguish between sub-patterns, you can use another symbol, identical in name in all sub-patterns but differing in value. Reworking the above example:

    (require 'cl-lib)
    (pcase (read-number "Enter an integer: ")
      ((and num                                ; line 1
            (or (and (pred cl-evenp)           ; line 2
                     (let spin 'even))         ; line 3
                (let spin 'odd)))              ; line 4
       (list spin num)))                       ; line 5
    
    
    
    Enter an integer: 42
    ⇒ (even 42)
    
    Enter an integer: 149
    ⇒ (odd 149)
    

    Line 1 “factors out” the expval binding with and and symbol (in this case, num). On line 2, or begins in the same way as before, but instead of binding different symbols, uses let twice (lines 3-4) to bind the same symbol spin in both sub-patterns. The value of spin distinguishes the sub-patterns. The body form references both symbols (line 5).


11.4.2 Extending pcase

The pcase macro supports several kinds of patterns (see Pattern-Matching Conditional). You can add support for other kinds of patterns using the pcase-defmacro macro.

Macro: pcase-defmacro name args [doc] &rest body

Define a new kind of pattern for pcase, to be invoked as (name actual-args). The pcase macro expands this into a function call that evaluates body, whose job it is to rewrite the invoked pattern into some other pattern, in an environment where args are bound to actual-args.

Additionally, arrange to display doc along with the docstring of pcase. By convention, doc should use EXPVAL to stand for the result of evaluating expression (first arg to pcase).

Typically, body rewrites the invoked pattern to use more basic patterns. Although all patterns eventually reduce to core patterns, body need not use core patterns straight away. The following example defines two patterns, named less-than and integer-less-than.

(pcase-defmacro less-than (n)
  "Matches if EXPVAL is a number less than N."
  `(pred (> ,n)))

(pcase-defmacro integer-less-than (n)
  "Matches if EXPVAL is an integer less than N."
  `(and (pred integerp)
        (less-than ,n)))

Note that the docstrings mention args (in this case, only one: n) in the usual way, and also mention EXPVAL by convention. The first rewrite (i.e., body for less-than) uses one core pattern: pred. The second uses two core patterns: and and pred, as well as the newly-defined pattern less-than. Both use a single backquote construct (see Backquote).


11.4.3 Backquote-Style Patterns

This subsection describes backquote-style patterns, a set of builtin patterns that eases structural matching. For background, see Pattern-Matching Conditional.

Backquote-style patterns are a powerful set of pcase pattern extensions (created using pcase-defmacro) that make it easy to match expval against specifications of its structure.

For example, to match expval that must be a list of two elements whose first element is a specific string and the second element is any value, you can write a core pattern:

(and (pred listp)
     ls
     (guard (= 2 (length ls)))
     (guard (string= "first" (car ls)))
     (let second-elem (cadr ls)))

or you can write the equivalent backquote-style pattern:

`("first" ,second-elem)

The backquote-style pattern is more concise, resembles the structure of expval, and avoids binding ls.

A backquote-style pattern has the form `qpat where qpat can have the following forms:

(qpat1 . qpat2)

Matches if expval is a cons cell whose car matches qpat1 and whose cdr matches qpat2. This readily generalizes to lists as in (qpat1 qpat2 …).

[qpat1 qpat2qpatm]

Matches if expval is a vector of length m whose 0..(m-1)th elements match qpat1, qpat2qpatm, respectively.

symbol
keyword
number
string

Matches if the corresponding element of expval is equal to the specified literal object.

,pattern

Matches if the corresponding element of expval matches pattern. Note that pattern is any kind that pcase supports. (In the example above, second-elem is a symbol core pattern; it therefore matches anything, and let-binds second-elem.)

The corresponding element is the portion of expval that is in the same structural position as the structural position of qpat in the backquote-style pattern. (In the example above, the corresponding element of second-elem is the second element of expval.)

Here is an example of using pcase to implement a simple interpreter for a little expression language (note that this requires lexical binding for the lambda expression in the fn clause to properly capture body and arg (see Lexical Binding):

(defun evaluate (form env)
  (pcase form
    (`(add ,x ,y)       (+ (evaluate x env)
                           (evaluate y env)))
    (`(call ,fun ,arg)  (funcall (evaluate fun env)
                                 (evaluate arg env)))
    (`(fn ,arg ,body)   (lambda (val)
                          (evaluate body (cons (cons arg val)
                                               env))))
    ((pred numberp)     form)
    ((pred symbolp)     (cdr (assq form env)))
    (_                  (error "Syntax error: %S" form))))

The first three clauses use backquote-style patterns. `(add ,x ,y) is a pattern that checks that form is a three-element list starting with the literal symbol add, then extracts the second and third elements and binds them to symbols x and y, respectively. This is known as destructuring, see Destructuring with pcase Patterns. The clause body evaluates x and y and adds the results. Similarly, the call clause implements a function call, and the fn clause implements an anonymous function definition.

The remaining clauses use core patterns. (pred numberp) matches if form is a number. On match, the body evaluates it. (pred symbolp) matches if form is a symbol. On match, the body looks up the symbol in env and returns its association. Finally, _ is the catch-all pattern that matches anything, so it’s suitable for reporting syntax errors.

Here are some sample programs in this small language, including their evaluation results:

(evaluate '(add 1 2) nil)                 ⇒ 3
(evaluate '(add x y) '((x . 1) (y . 2)))  ⇒ 3
(evaluate '(call (fn x (add 1 x)) 2) nil) ⇒ 3
(evaluate '(sub 1 2) nil)                 ⇒ error

11.4.4 Destructuring with pcase Patterns

Pcase patterns not only express a condition on the form of the objects they can match, but they can also extract sub-fields of those objects. For example we can extract 2 elements from a list that is the value of the variable my-list with the following code:

  (pcase my-list
    (`(add ,x ,y)  (message "Contains %S and %S" x y)))

This will not only extract x and y but will additionally test that my-list is a list containing exactly 3 elements and whose first element is the symbol add. If any of those tests fail, pcase will immediately return nil without calling message.

Extraction of multiple values stored in an object is known as destructuring. Using pcase patterns allows to perform destructuring binding, which is similar to a local binding (see Local Variables), but gives values to multiple elements of a variable by extracting those values from an object of compatible structure.

The macros described in this section use pcase patterns to perform destructuring binding. The condition of the object to be of compatible structure means that the object must match the pattern, because only then the object’s subfields can be extracted. For example:

  (pcase-let ((`(add ,x ,y) my-list))
    (message "Contains %S and %S" x y))

does the same as the previous example, except that it directly tries to extract x and y from my-list without first verifying if my-list is a list which has the right number of elements and has add as its first element. The precise behavior when the object does not actually match the pattern is undefined, although the body will not be silently skipped: either an error is signaled or the body is run with some of the variables potentially bound to arbitrary values like nil.

The pcase patterns that are useful for destructuring bindings are generally those described in Backquote-Style Patterns, since they express a specification of the structure of objects that will match.

For an alternative facility for destructuring binding, see seq-let.

Macro: pcase-let bindings body…

Perform destructuring binding of variables according to bindings, and then evaluate body.

bindings is a list of bindings of the form (pattern exp), where exp is an expression to evaluate and pattern is a pcase pattern.

All exps are evaluated first, after which they are matched against their respective pattern, introducing new variable bindings that can then be used inside body. The variable bindings are produced by destructuring binding of elements of pattern to the values of the corresponding elements of the evaluated exp.

Here’s a trivial example:

(pcase-let ((`(,major ,minor)
	     (split-string "image/png" "/")))
  minor)
     ⇒ "png"
Macro: pcase-let* bindings body…

Perform destructuring binding of variables according to bindings, and then evaluate body.

bindings is a list of bindings of the form (pattern exp), where exp is an expression to evaluate and pattern is a pcase pattern. The variable bindings are produced by destructuring binding of elements of pattern to the values of the corresponding elements of the evaluated exp.

Unlike pcase-let, but similarly to let*, each exp is matched against its corresponding pattern before processing the next element of bindings, so the variable bindings introduced in each one of the bindings are available in the exps of the bindings that follow it, additionally to being available in body.

Macro: pcase-dolist (pattern list) body…

Execute body once for each element of list, on each iteration performing a destructuring binding of variables in pattern to the values of the corresponding subfields of the element of list. The bindings are performed as if by pcase-let. When pattern is a simple variable, this ends up being equivalent to dolist (see Iteration).

Macro: pcase-setq pattern value…

Assign values to variables in a setq form, destructuring each value according to its respective pattern.

Macro: pcase-lambda lambda-list &rest body

This is like lambda, but allows each argument to be a pattern. For instance, here’s a simple function that takes a cons cell as the argument:

(setq fun
      (pcase-lambda (`(,key . ,val))
        (vector key (* val 10))))
(funcall fun '(foo . 2))
    ⇒ [foo 20]

11.5 Iteration

Iteration means executing part of a program repetitively. For example, you might want to repeat some computation once for each element of a list, or once for each integer from 0 to n. You can do this in Emacs Lisp with the special form while:

Special Form: while condition forms…

while first evaluates condition. If the result is non-nil, it evaluates forms in textual order. Then it reevaluates condition, and if the result is non-nil, it evaluates forms again. This process repeats until condition evaluates to nil.

There is no limit on the number of iterations that may occur. The loop will continue until either condition evaluates to nil or until an error or throw jumps out of it (see Nonlocal Exits).

The value of a while form is always nil.

(setq num 0)
     ⇒ 0
(while (< num 4)
  (princ (format "Iteration %d." num))
  (setq num (1+ num)))
     -| Iteration 0.
     -| Iteration 1.
     -| Iteration 2.
     -| Iteration 3.
     ⇒ nil

To write a repeat-until loop, which will execute something on each iteration and then do the end-test, put the body followed by the end-test in a progn as the first argument of while, as shown here:

(while (progn
         (forward-line 1)
         (not (looking-at "^$"))))

This moves forward one line and continues moving by lines until it reaches an empty line. It is peculiar in that the while has no body, just the end test (which also does the real work of moving point).

The dolist and dotimes macros provide convenient ways to write two common kinds of loops.

Macro: dolist (var list [result]) body…

This construct executes body once for each element of list, binding the variable var locally to hold the current element. Then it returns the value of evaluating result, or nil if result is omitted. For example, here is how you could use dolist to define the reverse function:

(defun reverse (list)
  (let (value)
    (dolist (elt list value)
      (setq value (cons elt value)))))
Macro: dotimes (var count [result]) body…

This construct executes body once for each integer from 0 (inclusive) to count (exclusive), binding the variable var to the integer for the current iteration. Then it returns the value of evaluating result, or nil if result is omitted. Use of result is deprecated. Here is an example of using dotimes to do something 100 times:

(dotimes (i 100)
  (insert "I will not obey absurd orders\n"))

11.6 Generators

A generator is a function that produces a potentially-infinite stream of values. Each time the function produces a value, it suspends itself and waits for a caller to request the next value.

Macro: iter-defun name args [doc] [declare] [interactive] body…

iter-defun defines a generator function. A generator function has the same signature as a normal function, but works differently. Instead of executing body when called, a generator function returns an iterator object. That iterator runs body to generate values, emitting a value and pausing where iter-yield or iter-yield-from appears. When body returns normally, iter-next signals iter-end-of-sequence with body’s result as its condition data.

Any kind of Lisp code is valid inside body, but iter-yield and iter-yield-from cannot appear inside unwind-protect forms.

Macro: iter-lambda args [doc] [interactive] body…

iter-lambda produces an unnamed generator function that works just like a generator function produced with iter-defun.

Macro: iter-yield value

When it appears inside a generator function, iter-yield indicates that the current iterator should pause and return value from iter-next. iter-yield evaluates to the value parameter of next call to iter-next.

Macro: iter-yield-from iterator

iter-yield-from yields all the values that iterator produces and evaluates to the value that iterator’s generator function returns normally. While it has control, iterator receives values sent to the iterator using iter-next.

To use a generator function, first call it normally, producing a iterator object. An iterator is a specific instance of a generator. Then use iter-next to retrieve values from this iterator. When there are no more values to pull from an iterator, iter-next raises an iter-end-of-sequence condition with the iterator’s final value.

It’s important to note that generator function bodies only execute inside calls to iter-next. A call to a function defined with iter-defun produces an iterator; you must drive this iterator with iter-next for anything interesting to happen. Each call to a generator function produces a different iterator, each with its own state.

Function: iter-next iterator &optional value

Retrieve the next value from iterator. If there are no more values to be generated (because iterator’s generator function returned), iter-next signals the iter-end-of-sequence condition; the data value associated with this condition is the value with which iterator’s generator function returned.

value is sent into the iterator and becomes the value to which iter-yield evaluates. value is ignored for the first iter-next call to a given iterator, since at the start of iterator’s generator function, the generator function is not evaluating any iter-yield form.

Function: iter-close iterator

If iterator is suspended inside an unwind-protect’s bodyform and becomes unreachable, Emacs will eventually run unwind handlers after a garbage collection pass. (Note that iter-yield is illegal inside an unwind-protect’s unwindforms.) To ensure that these handlers are run before then, use iter-close.

Some convenience functions are provided to make working with iterators easier:

Macro: iter-do (var iterator) body …

Run body with var bound to each value that iterator produces.

The Common Lisp loop facility also contains features for working with iterators. See Loop Facility in Common Lisp Extensions.

The following piece of code demonstrates some important principles of working with iterators.

(require 'generator)
(iter-defun my-iter (x)
  (iter-yield (1+ (iter-yield (1+ x))))
   ;; Return normally
  -1)

(let* ((iter (my-iter 5))
       (iter2 (my-iter 0)))
  ;; Prints 6
  (print (iter-next iter))
  ;; Prints 9
  (print (iter-next iter 8))
  ;; Prints 1; iter and iter2 have distinct states
  (print (iter-next iter2 nil))

  ;; We expect the iter sequence to end now
  (condition-case x
      (iter-next iter)
    (iter-end-of-sequence
      ;; Prints -1, which my-iter returned normally
      (print (cdr x)))))

11.7 Nonlocal Exits

A nonlocal exit is a transfer of control from one point in a program to another remote point. Nonlocal exits can occur in Emacs Lisp as a result of errors; you can also use them under explicit control. Nonlocal exits unbind all variable bindings made by the constructs being exited.


11.7.1 Explicit Nonlocal Exits: catch and throw

Most control constructs affect only the flow of control within the construct itself. The function throw is the exception to this rule of normal program execution: it performs a nonlocal exit on request. (There are other exceptions, but they are for error handling only.) throw is used inside a catch, and jumps back to that catch. For example:

(defun foo-outer ()
  (catch 'foo
    (foo-inner)))

(defun foo-inner ()
  …
  (if x
      (throw 'foo t))
  …)

The throw form, if executed, transfers control straight back to the corresponding catch, which returns immediately. The code following the throw is not executed. The second argument of throw is used as the return value of the catch.

The function throw finds the matching catch based on the first argument: it searches for a catch whose first argument is eq to the one specified in the throw. If there is more than one applicable catch, the innermost one takes precedence. Thus, in the above example, the throw specifies foo, and the catch in foo-outer specifies the same symbol, so that catch is the applicable one (assuming there is no other matching catch in between).

Executing throw exits all Lisp constructs up to the matching catch, including function calls. When binding constructs such as let or function calls are exited in this way, the bindings are unbound, just as they are when these constructs exit normally (see Local Variables). Likewise, throw restores the buffer and position saved by save-excursion (see Excursions), and the narrowing status saved by save-restriction. It also runs any cleanups established with the unwind-protect special form when it exits that form (see Cleaning Up from Nonlocal Exits).

The throw need not appear lexically within the catch that it jumps to. It can equally well be called from another function called within the catch. As long as the throw takes place chronologically after entry to the catch, and chronologically before exit from it, it has access to that catch. This is why throw can be used in commands such as exit-recursive-edit that throw back to the editor command loop (see Recursive Editing).

Common Lisp note: Most other versions of Lisp, including Common Lisp, have several ways of transferring control nonsequentially: return, return-from, and go, for example. Emacs Lisp has only throw. The cl-lib library provides versions of some of these. See Blocks and Exits in Common Lisp Extensions.

Special Form: catch tag body…

catch establishes a return point for the throw function. The return point is distinguished from other such return points by tag, which may be any Lisp object except nil. The argument tag is evaluated normally before the return point is established.

With the return point in effect, catch evaluates the forms of the body in textual order. If the forms execute normally (without error or nonlocal exit) the value of the last body form is returned from the catch.

If a throw is executed during the execution of body, specifying the same value tag, the catch form exits immediately; the value it returns is whatever was specified as the second argument of throw.

Function: throw tag value

The purpose of throw is to return from a return point previously established with catch. The argument tag is used to choose among the various existing return points; it must be eq to the value specified in the catch. If multiple return points match tag, the innermost one is used.

The argument value is used as the value to return from that catch.

If no return point is in effect with tag tag, then a no-catch error is signaled with data (tag value).


11.7.2 Examples of catch and throw

One way to use catch and throw is to exit from a doubly nested loop. (In most languages, this would be done with a goto.) Here we compute (foo i j) for i and j varying from 0 to 9:

(defun search-foo ()
  (catch 'loop
    (let ((i 0))
      (while (< i 10)
        (let ((j 0))
          (while (< j 10)
            (if (foo i j)
                (throw 'loop (list i j)))
            (setq j (1+ j))))
        (setq i (1+ i))))))

If foo ever returns non-nil, we stop immediately and return a list of i and j. If foo always returns nil, the catch returns normally, and the value is nil, since that is the result of the while.

Here are two tricky examples, slightly different, showing two return points at once. First, two return points with the same tag, hack:

(defun catch2 (tag)
  (catch tag
    (throw 'hack 'yes)))
⇒ catch2

(catch 'hack
  (print (catch2 'hack))
  'no)
-| yes
⇒ no

Since both return points have tags that match the throw, it goes to the inner one, the one established in catch2. Therefore, catch2 returns normally with value yes, and this value is printed. Finally the second body form in the outer catch, which is 'no, is evaluated and returned from the outer catch.

Now let’s change the argument given to catch2:

(catch 'hack
  (print (catch2 'quux))
  'no)
⇒ yes

We still have two return points, but this time only the outer one has the tag hack; the inner one has the tag quux instead. Therefore, throw makes the outer catch return the value yes. The function print is never called, and the body-form 'no is never evaluated.


11.7.3 Errors

When Emacs Lisp attempts to evaluate a form that, for some reason, cannot be evaluated, it signals an error.

When an error is signaled, Emacs’s default reaction is to print an error message and terminate execution of the current command. This is the right thing to do in most cases, such as if you type C-f at the end of the buffer.

In complicated programs, simple termination may not be what you want. For example, the program may have made temporary changes in data structures, or created temporary buffers that should be deleted before the program is finished. In such cases, you would use unwind-protect to establish cleanup expressions to be evaluated in case of error. (See Cleaning Up from Nonlocal Exits.) Occasionally, you may wish the program to continue execution despite an error in a subroutine. In these cases, you would use condition-case to establish error handlers to recover control in case of error.

For reporting problems without terminating the execution of the current command, consider issuing a warning instead. See Reporting Warnings.

Resist the temptation to use error handling to transfer control from one part of the program to another; use catch and throw instead. See Explicit Nonlocal Exits: catch and throw.


11.7.3.1 How to Signal an Error

Signaling an error means beginning error processing. Error processing normally aborts all or part of the running program and returns to a point that is set up to handle the error (see How Emacs Processes Errors). Here we describe how to signal an error.

Most errors are signaled automatically within Lisp primitives which you call for other purposes, such as if you try to take the CAR of an integer or move forward a character at the end of the buffer. You can also signal errors explicitly with the functions error and signal.

Quitting, which happens when the user types C-g, is not considered an error, but it is handled almost like an error. See Quitting.

Every error specifies an error message, one way or another. The message should state what is wrong (“File does not exist”), not how things ought to be (“File must exist”). The convention in Emacs Lisp is that error messages should start with a capital letter, but should not end with any sort of punctuation.

Function: error format-string &rest args

This function signals an error with an error message constructed by applying format-message (see Formatting Strings) to format-string and args.

These examples show typical uses of error:

(error "That is an error -- try something else")
     error→ That is an error -- try something else

(error "Invalid name `%s'" "A%%B")
     error→ Invalid name ‘A%%B’

error works by calling signal with two arguments: the error symbol error, and a list containing the string returned by format-message.

Typically grave accent and apostrophe in the format translate to matching curved quotes, e.g., "Missing `%s'" might result in "Missing ‘foo’". See Text Quoting Style, for how to influence or inhibit this translation.

Warning: If you want to use your own string as an error message verbatim, don’t just write (error string). If string string contains ‘%’, ‘`’, or ‘'’ it may be reformatted, with undesirable results. Instead, use (error "%s" string).

When noninteractive is non-nil (see Batch Mode), this function kills Emacs if the signaled error has no handler.

Function: signal error-symbol data

This function signals an error named by error-symbol. The argument data is a list of additional Lisp objects relevant to the circumstances of the error.

The argument error-symbol must be an error symbol—a symbol defined with define-error. This is how Emacs Lisp classifies different sorts of errors. See Error Symbols and Condition Names, for a description of error symbols, error conditions and condition names.

If the error is not handled, the two arguments are used in printing the error message. Normally, this error message is provided by the error-message property of error-symbol. If data is non-nil, this is followed by a colon and a comma separated list of the unevaluated elements of data. For error, the error message is the CAR of data (that must be a string). Subcategories of file-error are handled specially.

The number and significance of the objects in data depends on error-symbol. For example, with a wrong-type-argument error, there should be two objects in the list: a predicate that describes the type that was expected, and the object that failed to fit that type.

Both error-symbol and data are available to any error handlers that handle the error: condition-case binds a local variable to a list of the form (error-symbol . data) (see Writing Code to Handle Errors).

The function signal never returns. If the error error-symbol has no handler, and noninteractive is non-nil (see Batch Mode), this function eventually kills Emacs.

(signal 'wrong-number-of-arguments '(x y))
     error→ Wrong number of arguments: x, y

(signal 'no-such-error '("My unknown error condition"))
     error→ peculiar error: "My unknown error condition"
Function: user-error format-string &rest args

This function behaves exactly like error, except that it uses the error symbol user-error rather than error. As the name suggests, this is intended to report errors on the part of the user, rather than errors in the code itself. For example, if you try to use the command Info-history-back (l) to move back beyond the start of your Info browsing history, Emacs signals a user-error. Such errors do not cause entry to the debugger, even when debug-on-error is non-nil. See Entering the Debugger on an Error.

Common Lisp note: Emacs Lisp has nothing like the Common Lisp concept of continuable errors.


11.7.3.2 How Emacs Processes Errors

When an error is signaled, signal searches for an active handler for the error. A handler is a sequence of Lisp expressions designated to be executed if an error happens in part of the Lisp program. If the error has an applicable handler, the handler is executed, and control resumes following the handler. The handler executes in the environment of the condition-case that established it; all functions called within that condition-case have already been exited, and the handler cannot return to them.

If there is no applicable handler for the error, it terminates the current command and returns control to the editor command loop. (The command loop has an implicit handler for all kinds of errors.) The command loop’s handler uses the error symbol and associated data to print an error message. You can use the variable command-error-function to control how this is done:

Variable: command-error-function

This variable, if non-nil, specifies a function to use to handle errors that return control to the Emacs command loop. The function should take three arguments: data, a list of the same form that condition-case would bind to its variable; context, a string describing the situation in which the error occurred, or (more often) nil; and caller, the Lisp function which called the primitive that signaled the error.

An error that has no explicit handler may call the Lisp debugger (see Invoking the Debugger). The debugger is enabled if the variable debug-on-error (see Entering the Debugger on an Error) is non-nil. Unlike error handlers, the debugger runs in the environment of the error, so that you can examine values of variables precisely as they were at the time of the error. In batch mode (see Batch Mode), the Emacs process then normally exits with a non-zero exit status.


11.7.3.3 Writing Code to Handle Errors

The usual effect of signaling an error is to terminate the command that is running and return immediately to the Emacs editor command loop. You can arrange to trap errors occurring in a part of your program by establishing an error handler, with the special form condition-case. A simple example looks like this:

(condition-case nil
    (delete-file filename)
  (error nil))

This deletes the file named filename, catching any error and returning nil if an error occurs. (You can use the macro ignore-errors for a simple case like this; see below.)

The condition-case construct is often used to trap errors that are predictable, such as failure to open a file in a call to insert-file-contents. It is also used to trap errors that are totally unpredictable, such as when the program evaluates an expression read from the user.

The second argument of condition-case is called the protected form. (In the example above, the protected form is a call to delete-file.) The error handlers go into effect when this form begins execution and are deactivated when this form returns. They remain in effect for all the intervening time. In particular, they are in effect during the execution of functions called by this form, in their subroutines, and so on. This is a good thing, since, strictly speaking, errors can be signaled only by Lisp primitives (including signal and error) called by the protected form, not by the protected form itself.

The arguments after the protected form are handlers. Each handler lists one or more condition names (which are symbols) to specify which errors it will handle. The error symbol specified when an error is signaled also defines a list of condition names. A handler applies to an error if they have any condition names in common. In the example above, there is one handler, and it specifies one condition name, error, which covers all errors.

The search for an applicable handler checks all the established handlers starting with the most recently established one. Thus, if two nested condition-case forms offer to handle the same error, the inner of the two gets to handle it.

If an error is handled by some condition-case form, this ordinarily prevents the debugger from being run, even if debug-on-error says this error should invoke the debugger.

If you want to be able to debug errors that are caught by a condition-case, set the variable debug-on-signal to a non-nil value. You can also specify that a particular handler should let the debugger run first, by writing debug among the conditions, like this:

(condition-case nil
    (delete-file filename)
  ((debug error) nil))

The effect of debug here is only to prevent condition-case from suppressing the call to the debugger. Any given error will invoke the debugger only if debug-on-error and the other usual filtering mechanisms say it should. See Entering the Debugger on an Error.

Macro: condition-case-unless-debug var protected-form handlers…

The macro condition-case-unless-debug provides another way to handle debugging of such forms. It behaves exactly like condition-case, unless the variable debug-on-error is non-nil, in which case it does not handle any errors at all.

Once Emacs decides that a certain handler handles the error, it returns control to that handler. To do so, Emacs unbinds all variable bindings made by binding constructs that are being exited, and executes the cleanups of all unwind-protect forms that are being exited. Once control arrives at the handler, the body of the handler executes normally.

After execution of the handler body, execution returns from the condition-case form. Because the protected form is exited completely before execution of the handler, the handler cannot resume execution at the point of the error, nor can it examine variable bindings that were made within the protected form. All it can do is clean up and proceed.

Error signaling and handling have some resemblance to throw and catch (see Explicit Nonlocal Exits: catch and throw), but they are entirely separate facilities. An error cannot be caught by a catch, and a throw cannot be handled by an error handler (though using throw when there is no suitable catch signals an error that can be handled).

Special Form: condition-case var protected-form handlers…

This special form establishes the error handlers handlers around the execution of protected-form. If protected-form executes without error, the value it returns becomes the value of the condition-case form (in the absence of a success handler; see below). In this case, the condition-case has no effect. The condition-case form makes a difference when an error occurs during protected-form.

Each of the handlers is a list of the form (conditions body…). Here conditions is an error condition name to be handled, or a list of condition names (which can include debug to allow the debugger to run before the handler). A condition name of t matches any condition. body is one or more Lisp expressions to be executed when this handler handles an error. Here are examples of handlers:

(error nil)

(arith-error (message "Division by zero"))

((arith-error file-error)
 (message
  "Either division by zero or failure to open a file"))

Each error that occurs has an error symbol that describes what kind of error it is, and which describes also a list of condition names (see Error Symbols and Condition Names). Emacs searches all the active condition-case forms for a handler that specifies one or more of these condition names; the innermost matching condition-case handles the error. Within this condition-case, the first applicable handler handles the error.

After executing the body of the handler, the condition-case returns normally, using the value of the last form in the handler body as the overall value.

The argument var is a variable. condition-case does not bind this variable when executing the protected-form, only when it handles an error. At that time, it binds var locally to an error description, which is a list giving the particulars of the error. The error description has the form (error-symbol . data). The handler can refer to this list to decide what to do. For example, if the error is for failure opening a file, the file name is the second element of data—the third element of the error description.

If var is nil, that means no variable is bound. Then the error symbol and associated data are not available to the handler.

As a special case, one of the handlers can be a list of the form (:success body…), where body is executed with var (if non-nil) bound to the return value of protected-form when that expression terminates without error.

Sometimes it is necessary to re-throw a signal caught by condition-case, for some outer-level handler to catch. Here’s how to do that:

  (signal (car err) (cdr err))

where err is the error description variable, the first argument to condition-case whose error condition you want to re-throw. See Definition of signal.

Function: error-message-string error-descriptor

This function returns the error message string for a given error descriptor. It is useful if you want to handle an error by printing the usual error message for that error. See Definition of signal.

Here is an example of using condition-case to handle the error that results from dividing by zero. The handler displays the error message (but without a beep), then returns a very large number.

(defun safe-divide (dividend divisor)
  (condition-case err
      ;; Protected form.
      (/ dividend divisor)
    ;; The handler.
    (arith-error                        ; Condition.
     ;; Display the usual message for this error.
     (message "%s" (error-message-string err))
     1000000)))
⇒ safe-divide

(safe-divide 5 0)
     -| Arithmetic error: (arith-error)
⇒ 1000000

The handler specifies condition name arith-error so that it will handle only division-by-zero errors. Other kinds of errors will not be handled (by this condition-case). Thus:

(safe-divide nil 3)
     error→ Wrong type argument: number-or-marker-p, nil

Here is a condition-case that catches all kinds of errors, including those from error:

(setq baz 34)
     ⇒ 34

(condition-case err
    (if (eq baz 35)
        t
      ;; This is a call to the function error.
      (error "Rats!  The variable %s was %s, not 35" 'baz baz))
  ;; This is the handler; it is not a form.
  (error (princ (format "The error was: %s" err))
         2))
-| The error was: (error "Rats!  The variable baz was 34, not 35")
⇒ 2
Macro: ignore-errors body…

This construct executes body, ignoring any errors that occur during its execution. If the execution is without error, ignore-errors returns the value of the last form in body; otherwise, it returns nil.

Here’s the example at the beginning of this subsection rewritten using ignore-errors:

  (ignore-errors
   (delete-file filename))
Macro: ignore-error condition body…

This macro is like ignore-errors, but will only ignore the specific error condition specified.

  (ignore-error end-of-file
    (read ""))

condition can also be a list of error conditions.

Macro: with-demoted-errors format body…

This macro is like a milder version of ignore-errors. Rather than suppressing errors altogether, it converts them into messages. It uses the string format to format the message. format should contain a single ‘%’-sequence; e.g., "Error: %S". Use with-demoted-errors around code that is not expected to signal errors, but should be robust if one does occur. Note that this macro uses condition-case-unless-debug rather than condition-case.


11.7.3.4 Error Symbols and Condition Names

When you signal an error, you specify an error symbol to specify the kind of error you have in mind. Each error has one and only one error symbol to categorize it. This is the finest classification of errors defined by the Emacs Lisp language.

These narrow classifications are grouped into a hierarchy of wider classes called error conditions, identified by condition names. The narrowest such classes belong to the error symbols themselves: each error symbol is also a condition name. There are also condition names for more extensive classes, up to the condition name error which takes in all kinds of errors (but not quit). Thus, each error has one or more condition names: error, the error symbol if that is distinct from error, and perhaps some intermediate classifications.

Function: define-error name message &optional parent

In order for a symbol to be an error symbol, it must be defined with define-error which takes a parent condition (defaults to error). This parent defines the conditions that this kind of error belongs to. The transitive set of parents always includes the error symbol itself, and the symbol error. Because quitting is not considered an error, the set of parents of quit is just (quit).

In addition to its parents, the error symbol has a message which is a string to be printed when that error is signaled but not handled. If that message is not valid, the error message ‘peculiar error’ is used. See Definition of signal.

Internally, the set of parents is stored in the error-conditions property of the error symbol and the message is stored in the error-message property of the error symbol.

Here is how we define a new error symbol, new-error:

(define-error 'new-error "A new error" 'my-own-errors)

This error has several condition names: new-error, the narrowest classification; my-own-errors, which we imagine is a wider classification; and all the conditions of my-own-errors which should include error, which is the widest of all.

The error string should start with a capital letter but it should not end with a period. This is for consistency with the rest of Emacs.

Naturally, Emacs will never signal new-error on its own; only an explicit call to signal (see Definition of signal) in your code can do this:

(signal 'new-error '(x y))
     error→ A new error: x, y

This error can be handled through any of its condition names. This example handles new-error and any other errors in the class my-own-errors:

(condition-case foo
    (bar nil t)
  (my-own-errors nil))

The significant way that errors are classified is by their condition names—the names used to match errors with handlers. An error symbol serves only as a convenient way to specify the intended error message and list of condition names. It would be cumbersome to give signal a list of condition names rather than one error symbol.

By contrast, using only error symbols without condition names would seriously decrease the power of condition-case. Condition names make it possible to categorize errors at various levels of generality when you write an error handler. Using error symbols alone would eliminate all but the narrowest level of classification.

See Standard Errors, for a list of the main error symbols and their conditions.


11.7.4 Cleaning Up from Nonlocal Exits

The unwind-protect construct is essential whenever you temporarily put a data structure in an inconsistent state; it permits you to make the data consistent again in the event of an error or throw. (Another more specific cleanup construct that is used only for changes in buffer contents is the atomic change group; Atomic Change Groups.)

Special Form: unwind-protect body-form cleanup-forms…

unwind-protect executes body-form with a guarantee that the cleanup-forms will be evaluated if control leaves body-form, no matter how that happens. body-form may complete normally, or execute a throw out of the unwind-protect, or cause an error; in all cases, the cleanup-forms will be evaluated.

If body-form finishes normally, unwind-protect returns the value of body-form, after it evaluates the cleanup-forms. If body-form does not finish, unwind-protect does not return any value in the normal sense.

Only body-form is protected by the unwind-protect. If any of the cleanup-forms themselves exits nonlocally (via a throw or an error), unwind-protect is not guaranteed to evaluate the rest of them. If the failure of one of the cleanup-forms has the potential to cause trouble, then protect it with another unwind-protect around that form.

For example, here we make an invisible buffer for temporary use, and make sure to kill it before finishing:

(let ((buffer (get-buffer-create " *temp*")))
  (with-current-buffer buffer
    (unwind-protect
        body-form
      (kill-buffer buffer))))

You might think that we could just as well write (kill-buffer (current-buffer)) and dispense with the variable buffer. However, the way shown above is safer, if body-form happens to get an error after switching to a different buffer! (Alternatively, you could write a save-current-buffer around body-form, to ensure that the temporary buffer becomes current again in time to kill it.)

Emacs includes a standard macro called with-temp-buffer which expands into more or less the code shown above (see Current Buffer). Several of the macros defined in this manual use unwind-protect in this way.

Here is an actual example derived from an FTP package. It creates a process (see Processes) to try to establish a connection to a remote machine. As the function ftp-login is highly susceptible to numerous problems that the writer of the function cannot anticipate, it is protected with a form that guarantees deletion of the process in the event of failure. Otherwise, Emacs might fill up with useless subprocesses.

(let ((win nil))
  (unwind-protect
      (progn
        (setq process (ftp-setup-buffer host file))
        (if (setq win (ftp-login process host user password))
            (message "Logged in")
          (error "Ftp login failed")))
    (or win (and process (delete-process process)))))

This example has a small bug: if the user types C-g to quit, and the quit happens immediately after the function ftp-setup-buffer returns but before the variable process is set, the process will not be killed. There is no easy way to fix this bug, but at least it is very unlikely.


12 Variables

A variable is a name used in a program to stand for a value. In Lisp, each variable is represented by a Lisp symbol (see Symbols). The variable name is simply the symbol’s name, and the variable’s value is stored in the symbol’s value cell9. See Symbol Components. In Emacs Lisp, the use of a symbol as a variable is independent of its use as a function name.

As previously noted in this manual, a Lisp program is represented primarily by Lisp objects, and only secondarily as text. The textual form of a Lisp program is given by the read syntax of the Lisp objects that constitute the program. Hence, the textual form of a variable in a Lisp program is written using the read syntax for the symbol representing the variable.


12.1 Global Variables

The simplest way to use a variable is globally. This means that the variable has just one value at a time, and this value is in effect (at least for the moment) throughout the Lisp system. The value remains in effect until you specify a new one. When a new value replaces the old one, no trace of the old value remains in the variable.

You specify a value for a symbol with setq. For example,

(setq x '(a b))

gives the variable x the value (a b). Note that setq is a special form (see Special Forms); it does not evaluate its first argument, the name of the variable, but it does evaluate the second argument, the new value.

Once the variable has a value, you can refer to it by using the symbol itself as an expression. Thus,

x ⇒ (a b)

assuming the setq form shown above has already been executed.

If you do set the same variable again, the new value replaces the old one:

x
     ⇒ (a b)
(setq x 4)
     ⇒ 4
x
     ⇒ 4

12.2 Variables that Never Change

In Emacs Lisp, certain symbols normally evaluate to themselves. These include nil and t, as well as any symbol whose name starts with ‘:’ (these are called keywords). These symbols cannot be rebound, nor can their values be changed. Any attempt to set or bind nil or t signals a setting-constant error. The same is true for a keyword (a symbol whose name starts with ‘:’), if it is interned in the standard obarray, except that setting such a symbol to itself is not an error.

nil ≡ 'nil
     ⇒ nil
(setq nil 500)
error→ Attempt to set constant symbol: nil
Function: keywordp object

function returns t if object is a symbol whose name starts with ‘:’, interned in the standard obarray, and returns nil otherwise.

These constants are fundamentally different from the constants defined using the defconst special form (see Defining Global Variables). A defconst form serves to inform human readers that you do not intend to change the value of a variable, but Emacs does not raise an error if you actually change it.

A small number of additional symbols are made read-only for various practical reasons. These include enable-multibyte-characters, most-positive-fixnum, most-negative-fixnum, and a few others. Any attempt to set or bind these also signals a setting-constant error.


12.3 Local Variables

Global variables have values that last until explicitly superseded with new values. Sometimes it is useful to give a variable a local value—a value that takes effect only within a certain part of a Lisp program. When a variable has a local value, we say that it is locally bound to that value, and that it is a local variable.

For example, when a function is called, its argument variables receive local values, which are the actual arguments supplied to the function call; these local bindings take effect within the body of the function. To take another example, the let special form explicitly establishes local bindings for specific variables, which take effect only within the body of the let form.

We also speak of the global binding, which is where (conceptually) the global value is kept.

Establishing a local binding saves away the variable’s previous value (or lack of one). We say that the previous value is shadowed. Both global and local values may be shadowed. If a local binding is in effect, using setq on the local variable stores the specified value in the local binding. When that local binding is no longer in effect, the previously shadowed value (or lack of one) comes back.

A variable can have more than one local binding at a time (e.g., if there are nested let forms that bind the variable). The current binding is the local binding that is actually in effect. It determines the value returned by evaluating the variable symbol, and it is the binding acted on by setq.

For most purposes, you can think of the current binding as the innermost local binding, or the global binding if there is no local binding. To be more precise, a rule called the scoping rule determines where in a program a local binding takes effect. The default scoping rule in Emacs Lisp is called dynamic scoping, which simply states that the current binding at any given point in the execution of a program is the most recently-created binding for that variable that still exists. For details about dynamic scoping, and an alternative scoping rule called lexical scoping, see Scoping Rules for Variable Bindings. Lately Emacs is moving towards using lexical binding in more and more places, with the goal of eventually making lexical binding the default. In particular, all Emacs Lisp source files and the *scratch* buffer use lexical scoping.

The special forms let and let* exist to create local bindings:

Special Form: let (bindings…) forms…

This special form sets up local bindings for a certain set of variables, as specified by bindings, and then evaluates all of the forms in textual order. Its return value is the value of the last form in forms. The local bindings set up by let will be in effect only within the body of forms.

Each of the bindings is either (i) a symbol, in which case that symbol is locally bound to nil; or (ii) a list of the form (symbol value-form), in which case symbol is locally bound to the result of evaluating value-form. If value-form is omitted, nil is used.

All of the value-forms in bindings are evaluated in the order they appear and before binding any of the symbols to them. Here is an example of this: z is bound to the old value of y, which is 2, not the new value of y, which is 1.

(setq y 2)
     ⇒ 2

(let ((y 1)
      (z y))
  (list y z))
     ⇒ (1 2)

On the other hand, the order of bindings is unspecified: in the following example, either 1 or 2 might be printed.

(let ((x 1)
      (x 2))
  (print x))

Therefore, avoid binding a variable more than once in a single let form.

Special Form: let* (bindings…) forms…

This special form is like let, but it binds each variable right after computing its local value, before computing the local value for the next variable. Therefore, an expression in bindings can refer to the preceding symbols bound in this let* form. Compare the following example with the example above for let.

(setq y 2)
     ⇒ 2

(let* ((y 1)
       (z y))    ; Use the just-established value of y.
  (list y z))
     ⇒ (1 1)

Basically, the let* binding of x and y in the previous example is equivalent to using nested let bindings:

(let ((y 1))
  (let ((z y))
    (list y z)))
Special Form: letrec (bindings…) forms…

This special form is like let*, but all the variables are bound before any of the local values are computed. The values are then assigned to the locally bound variables. This is only useful when lexical binding is in effect, and you want to create closures that refer to bindings that would otherwise not yet be in effect when using let*.

For instance, here’s a closure that removes itself from a hook after being run once:

(letrec ((hookfun (lambda ()
                    (message "Run once")
                    (remove-hook 'post-command-hook hookfun))))
  (add-hook 'post-command-hook hookfun))
Special Form: dlet (bindings…) forms…

This special form is like let, but it binds all variables dynamically. This is rarely useful—you usually want to bind normal variables lexically, and special variables (i.e., variables that are defined with defvar) dynamically, and this is what let does.

dlet can be useful when interfacing with old code that assumes that certain variables are dynamically bound (see Dynamic Binding), but it’s impractical to defvar these variables. dlet will temporarily make the bound variables special, execute the forms, and then make the variables non-special again.

Special Form: named-let name bindings &rest body

This special form is a looping construct inspired from the Scheme language. It is similar to let: It binds the variables in bindings, and then evaluates body. However, named-let also binds name to a local function whose formal arguments are the variables in bindings and whose body is body. This allows body to call itself recursively by calling name, where the arguments passed to name are used as the new values of the bound variables in the recursive invocation.

Example of a loop summing a list of numbers:

(named-let sum ((numbers '(1 2 3 4))
                (running-sum 0))
  (if numbers
      (sum (cdr numbers) (+ running-sum (car numbers)))
    running-sum))
⇒ 10

Recursive calls to name that occur in tail positions in body are guaranteed to be optimized as tail calls, which means that they will not consume any additional stack space no matter how deeply the recursion runs. Such recursive calls will effectively jump to the top of the loop with new values for the variables.

A function call is in the tail position if it’s the very last thing done so that the value returned by the call is the value of body itself, as is the case in the recursive call to sum above.

Warning: named-let works as expected only when lexical-binding is enabled. See Lexical Binding.

Here is a complete list of the other facilities that create local bindings:

Variables can also have buffer-local bindings (see Buffer-Local Variables); a few variables have terminal-local bindings (see Multiple Terminals). These kinds of bindings work somewhat like ordinary local bindings, but they are localized depending on where you are in Emacs.


12.4 When a Variable is Void

We say that a variable is void if its symbol has an unassigned value cell (see Symbol Components).

Under Emacs Lisp’s default dynamic scoping rule (see Scoping Rules for Variable Bindings), the value cell stores the variable’s current (local or global) value. Note that an unassigned value cell is not the same as having nil in the value cell. The symbol nil is a Lisp object and can be the value of a variable, just as any other object can be; but it is still a value. If a variable is void, trying to evaluate the variable signals a void-variable error, instead of returning a value.

Under the optional lexical scoping rule, the value cell only holds the variable’s global value—the value outside of any lexical binding construct. When a variable is lexically bound, the local value is determined by the lexical environment; hence, variables can have local values even if their symbols’ value cells are unassigned.

Function: makunbound symbol

This function empties out the value cell of symbol, making the variable void. It returns symbol.

If symbol has a dynamic local binding, makunbound voids the current binding, and this voidness lasts only as long as the local binding is in effect. Afterwards, the previously shadowed local or global binding is reexposed; then the variable will no longer be void, unless the reexposed binding is void too.

Here are some examples (assuming dynamic binding is in effect):

(setq x 1)               ; Put a value in the global binding.
     ⇒ 1
(let ((x 2))             ; Locally bind it.
  (makunbound 'x)        ; Void the local binding.
  x)
error→ Symbol's value as variable is void: x
x                        ; The global binding is unchanged.
     ⇒ 1

(let ((x 2))             ; Locally bind it.
  (let ((x 3))           ; And again.
    (makunbound 'x)      ; Void the innermost-local binding.
    x))                  ; And refer: it’s void.
error→ Symbol's value as variable is void: x

(let ((x 2))
  (let ((x 3))
    (makunbound 'x))     ; Void inner binding, then remove it.
  x)                     ; Now outer let binding is visible.
     ⇒ 2
Function: boundp variable

This function returns t if variable (a symbol) is not void, and nil if it is void.

Here are some examples (assuming dynamic binding is in effect):

(boundp 'abracadabra)          ; Starts out void.
     ⇒ nil
(let ((abracadabra 5))         ; Locally bind it.
  (boundp 'abracadabra))
     ⇒ t
(boundp 'abracadabra)          ; Still globally void.
     ⇒ nil
(setq abracadabra 5)           ; Make it globally nonvoid.
     ⇒ 5
(boundp 'abracadabra)
     ⇒ t

12.5 Defining Global Variables

A variable definition is a construct that announces your intention to use a symbol as a global variable. It uses the special forms defvar or defconst, which are documented below.

A variable definition serves three purposes. First, it informs people who read the code that the symbol is intended to be used a certain way (as a variable). Second, it informs the Lisp system of this, optionally supplying an initial value and a documentation string. Third, it provides information to programming tools such as etags, allowing them to find where the variable was defined.

The difference between defconst and defvar is mainly a matter of intent, serving to inform human readers of whether the value should ever change. Emacs Lisp does not actually prevent you from changing the value of a variable defined with defconst. One notable difference between the two forms is that defconst unconditionally initializes the variable, whereas defvar initializes it only if it is originally void.

To define a customizable variable, you should use defcustom (which calls defvar as a subroutine). See Defining Customization Variables.

Special Form: defvar symbol [value [doc-string]]

This special form defines symbol as a variable. Note that symbol is not evaluated; the symbol to be defined should appear explicitly in the defvar form. The variable is marked as special, meaning that it should always be dynamically bound (see Scoping Rules for Variable Bindings).

If value is specified, and symbol is void (i.e., it has no dynamically bound value; see When a Variable is Void), then value is evaluated and symbol is set to the result. But if symbol is not void, value is not evaluated, and symbol’s value is left unchanged. If value is omitted, the value of symbol is not changed in any case.

Note that specifying a value, even nil, marks the variable as special permanently. Whereas if value is omitted then the variable is only marked special locally (i.e. within the current lexical scope, or file if at the top-level). This can be useful for suppressing byte compilation warnings, see Compiler Errors.

If symbol has a buffer-local binding in the current buffer, defvar acts on the default value, which is buffer-independent, rather than the buffer-local binding. It sets the default value if the default value is void. See Buffer-Local Variables.

If symbol is already let bound (e.g., if the defvar form occurs in a let form), then defvar sets the toplevel default value, like set-default-toplevel-value. The let binding remains in effect until its binding construct exits. See Scoping Rules for Variable Bindings.

When you evaluate a top-level defvar form with C-M-x (eval-defun) or with C-x C-e (eval-last-sexp) in Emacs Lisp mode, a special feature of these two commands arranges to set the variable unconditionally, without testing whether its value is void.

If the doc-string argument is supplied, it specifies the documentation string for the variable (stored in the symbol’s variable-documentation property). See Documentation.

Here are some examples. This form defines foo but does not initialize it:

(defvar foo)
     ⇒ foo

This example initializes the value of bar to 23, and gives it a documentation string:

(defvar bar 23
  "The normal weight of a bar.")
     ⇒ bar

The defvar form returns symbol, but it is normally used at top level in a file where its value does not matter.

For a more elaborate example of using defvar without a value, see Local defvar example.

Special Form: defconst symbol value [doc-string]

This special form defines symbol as a value and initializes it. It informs a person reading your code that symbol has a standard global value, established here, that should not be changed by the user or by other programs. Note that symbol is not evaluated; the symbol to be defined must appear explicitly in the defconst.

The defconst form, like defvar, marks the variable as special, meaning that it should always be dynamically bound (see Scoping Rules for Variable Bindings). In addition, it marks the variable as risky (see File Local Variables).

defconst always evaluates value, and sets the value of symbol to the result. If symbol does have a buffer-local binding in the current buffer, defconst sets the default value, not the buffer-local value. (But you should not be making buffer-local bindings for a symbol that is defined with defconst.)

An example of the use of defconst is Emacs’s definition of float-pi—the mathematical constant pi, which ought not to be changed by anyone (attempts by the Indiana State Legislature notwithstanding). As the second form illustrates, however, defconst is only advisory.

(defconst float-pi 3.141592653589793 "The value of Pi.")
     ⇒ float-pi
(setq float-pi 3)
     ⇒ float-pi
float-pi
     ⇒ 3

Warning: If you use a defconst or defvar special form while the variable has a local binding (made with let, or a function argument), it sets the local binding rather than the global binding. This is not what you usually want. To prevent this, use these special forms at top level in a file, where normally no local binding is in effect, and make sure to load the file before making a local binding for the variable.


12.6 Tips for Defining Variables Robustly

When you define a variable whose value is a function, or a list of functions, use a name that ends in ‘-function’ or ‘-functions’, respectively.

There are several other variable name conventions; here is a complete list:

…-hook

The variable is a normal hook (see Hooks).

…-function

The value is a function.

…-functions

The value is a list of functions.

…-form

The value is a form (an expression).

…-forms

The value is a list of forms (expressions).

…-predicate

The value is a predicate—a function of one argument that returns non-nil for success and nil for failure.

…-flag

The value is significant only as to whether it is nil or not. Since such variables often end up acquiring more values over time, this convention is not strongly recommended.

…-program

The value is a program name.

…-command

The value is a whole shell command.

…-switches

The value specifies options for a command.

prefix--…

The variable is intended for internal use and is defined in the file prefix.el. (Emacs code contributed before 2018 may follow other conventions, which are being phased out.)

…-internal

The variable is intended for internal use and is defined in C code. (Emacs code contributed before 2018 may follow other conventions, which are being phased out.)

When you define a variable, always consider whether you should mark it as safe or risky; see File Local Variables.

When defining and initializing a variable that holds a complicated value (such as a syntax table for a major mode), it’s best to put the entire computation of the value into the defvar, like this:

(defvar my-major-mode-syntax-table
  (let ((table (make-syntax-table)))
    (modify-syntax-entry ?# "<" table)
    …
    table)
  docstring)

This method has several benefits. First, if the user quits while loading the file, the variable is either still uninitialized or initialized properly, never in-between. If it is still uninitialized, reloading the file will initialize it properly. Second, reloading the file once the variable is initialized will not alter it; that is important if the user has changed its value. Third, evaluating the defvar form with C-M-x will reinitialize the variable completely.


12.7 Accessing Variable Values

The usual way to reference a variable is to write the symbol which names it. See Symbol Forms.

Occasionally, you may want to reference a variable which is only determined at run time. In that case, you cannot specify the variable name in the text of the program. You can use the symbol-value function to extract the value.

Function: symbol-value symbol

This function returns the value stored in symbol’s value cell. This is where the variable’s current (dynamic) value is stored. If the variable has no local binding, this is simply its global value. If the variable is void, a void-variable error is signaled.

If the variable is lexically bound, the value reported by symbol-value is not necessarily the same as the variable’s lexical value, which is determined by the lexical environment rather than the symbol’s value cell. See Scoping Rules for Variable Bindings.

(setq abracadabra 5)
     ⇒ 5
(setq foo 9)
     ⇒ 9

;; Here the symbol abracadabra
;;   is the symbol whose value is examined.
(let ((abracadabra 'foo))
  (symbol-value 'abracadabra))
     ⇒ foo

;; Here, the value of abracadabra,
;;   which is foo,
;;   is the symbol whose value is examined.
(let ((abracadabra 'foo))
  (symbol-value abracadabra))
     ⇒ 9

(symbol-value 'abracadabra)
     ⇒ 5

12.8 Setting Variable Values

The usual way to change the value of a variable is with the special form setq. When you need to compute the choice of variable at run time, use the function set.

Special Form: setq [symbol form]…

This special form is the most common method of changing a variable’s value. Each symbol is given a new value, which is the result of evaluating the corresponding form. The current binding of the symbol is changed.

setq does not evaluate symbol; it sets the symbol that you write. We say that this argument is automatically quoted. The ‘q’ in setq stands for “quoted”.

The value of the setq form is the value of the last form.

(setq x (1+ 2))
     ⇒ 3
x                   ; x now has a global value.
     ⇒ 3
(let ((x 5))
  (setq x 6)        ; The local binding of x is set.
  x)
     ⇒ 6
x                   ; The global value is unchanged.
     ⇒ 3

Note that the first form is evaluated, then the first symbol is set, then the second form is evaluated, then the second symbol is set, and so on:

(setq x 10          ; Notice that x is set before
      y (1+ x))     ;   the value of y is computed.
     ⇒ 11
Function: set symbol value

This function puts value in the value cell of symbol. Since it is a function rather than a special form, the expression written for symbol is evaluated to obtain the symbol to set. The return value is value.

When dynamic variable binding is in effect (the default), set has the same effect as setq, apart from the fact that set evaluates its symbol argument whereas setq does not. But when a variable is lexically bound, set affects its dynamic value, whereas setq affects its current (lexical) value. See Scoping Rules for Variable Bindings.

(set one 1)
error→ Symbol's value as variable is void: one
(set 'one 1)
     ⇒ 1
(set 'two 'one)
     ⇒ one
(set two 2)         ; two evaluates to symbol one.
     ⇒ 2
one                 ; So it is one that was set.
     ⇒ 2
(let ((one 1))      ; This binding of one is set,
  (set 'one 3)      ;   not the global value.
  one)
     ⇒ 3
one
     ⇒ 2

If symbol is not actually a symbol, a wrong-type-argument error is signaled.

(set '(x y) 'z)
error→ Wrong type argument: symbolp, (x y)
Macro: setopt [symbol form]…

This is like setq (see above), but meant for user options. This macro uses the Customize machinery to set the variable(s) (see Defining Customization Variables). In particular, setopt will run the setter function associated with the variable. For instance, if you have:

(defcustom my-var 1
  "My var."
  :type 'number
  :set (lambda (var val)
         (set-default var val)
         (message "We set %s to %s" var val)))

then the following, in addition to setting my-var to ‘2’, will also issue a message:

(setopt my-var 2)

setopt also checks whether the value is valid for the user option. For instance, using setopt to set a user option defined with a number type to a string will signal an error.

Unlike defcustom and related customization commands, such as customize-variable, setopt is meant for non-interactive use, in particular in the user init file. For that reason, it doesn’t record the standard, saved, and user-set values, and doesn’t mark the variable as candidate for saving in the custom file.

The setopt macro can be used on regular, non-user option variables, but is much less efficient than setq. The main use case for this macro is setting user options in the user’s init file.


12.9 Running a function when a variable is changed.

It is sometimes useful to take some action when a variable changes its value. The variable watchpoint facility provides the means to do so. Some possible uses for this feature include keeping display in sync with variable settings, and invoking the debugger to track down unexpected changes to variables (see Entering the debugger when a variable is modified).

The following functions may be used to manipulate and query the watch functions for a variable.

Function: add-variable-watcher symbol watch-function

This function arranges for watch-function to be called whenever symbol is modified. Modifications through aliases (see Variable Aliases) will have the same effect.

watch-function will be called, just before changing the value of symbol, with 4 arguments: symbol, newval, operation, and where. symbol is the variable being changed. newval is the value it will be changed to. (The old value is available to watch-function as the value of symbol, since it was not yet changed to newval.) operation is a symbol representing the kind of change, one of: set, let, unlet, makunbound, or defvaralias. where is a buffer if the buffer-local value of the variable is being changed, nil otherwise.

Function: remove-variable-watcher symbol watch-function

This function removes watch-function from symbol’s list of watchers.

Function: get-variable-watchers symbol

This function returns the list of symbol’s active watcher functions.

12.9.1 Limitations

There are a couple of ways in which a variable could be modified (or at least appear to be modified) without triggering a watchpoint.

Since watchpoints are attached to symbols, modification to the objects contained within variables (e.g., by a list modification function see Modifying Existing List Structure) is not caught by this mechanism.

Additionally, C code can modify the value of variables directly, bypassing the watchpoint mechanism.

A minor limitation of this feature, again because it targets symbols, is that only variables of dynamic scope may be watched. This poses little difficulty, since modifications to lexical variables can be discovered easily by inspecting the code within the scope of the variable (unlike dynamic variables, which can be modified by any code at all, see Scoping Rules for Variable Bindings).


12.10 Scoping Rules for Variable Bindings

When you create a local binding for a variable, that binding takes effect only within a limited portion of the program (see Local Variables). This section describes exactly what this means.

Each local binding has a certain scope and extent. Scope refers to where in the textual source code the binding can be accessed. Extent refers to when, as the program is executing, the binding exists.

By default, the local bindings that Emacs creates are dynamic bindings. Such a binding has dynamic scope, meaning that any part of the program can potentially access the variable binding. It also has dynamic extent, meaning that the binding lasts only while the binding construct (such as the body of a let form) is being executed.

Emacs can optionally create lexical bindings. A lexical binding has lexical scope, meaning that any reference to the variable must be located textually within the binding construct10. It also has indefinite extent, meaning that under some circumstances the binding can live on even after the binding construct has finished executing, by means of special objects called closures.

The dynamic binding was (and still is) the default in Emacs for many years, but lately Emacs is moving towards using lexical binding in more and more places, with the goal of eventually making that the default.

The following subsections describe dynamic binding and lexical binding in greater detail, and how to enable lexical binding in Emacs Lisp programs.


12.10.1 Dynamic Binding

By default, the local variable bindings made by Emacs are dynamic bindings. When a variable is dynamically bound, its current binding at any point in the execution of the Lisp program is simply the most recently-created dynamic local binding for that symbol, or the global binding if there is no such local binding.

Dynamic bindings have dynamic scope and extent, as shown by the following example:

(defvar x -99)  ; x receives an initial value of −99.

(defun getx ()
  x)            ; x is used free in this function.

(let ((x 1))    ; x is dynamically bound.
  (getx))
     ⇒ 1

;; After the let form finishes, x reverts to its
;; previous value, which is −99.

(getx)
     ⇒ -99

The function getx refers to x. This is a free reference, in the sense that there is no binding for x within that defun construct itself. When we call getx from within a let form in which x is (dynamically) bound, it retrieves the local value (i.e., 1). But when we call getx outside the let form, it retrieves the global value (i.e., −99).

Here is another example, which illustrates setting a dynamically bound variable using setq:

(defvar x -99)      ; x receives an initial value of −99.

(defun addx ()
  (setq x (1+ x)))  ; Add 1 to x and return its new value.

(let ((x 1))
  (addx)
  (addx))
     ⇒ 3           ; The two addx calls add to x twice.

;; After the let form finishes, x reverts to its
;; previous value, which is −99.

(addx)
     ⇒ -98

Dynamic binding is implemented in Emacs Lisp in a simple way. Each symbol has a value cell, which specifies its current dynamic value (or absence of value). See Symbol Components. When a symbol is given a dynamic local binding, Emacs records the contents of the value cell (or absence thereof) in a stack, and stores the new local value in the value cell. When the binding construct finishes executing, Emacs pops the old value off the stack, and puts it in the value cell.

Note that when code using Dynamic Binding is native compiled the native compiler will not perform any Lisp specific optimization.


12.10.2 Proper Use of Dynamic Binding

Dynamic binding is a powerful feature, as it allows programs to refer to variables that are not defined within their local textual scope. However, if used without restraint, this can also make programs hard to understand. There are two clean ways to use this technique:

  • If a variable has no global definition, use it as a local variable only within a binding construct, such as the body of the let form where the variable was bound. If this convention is followed consistently throughout a program, the value of the variable will not affect, nor be affected by, any uses of the same variable symbol elsewhere in the program.
  • Otherwise, define the variable with defvar, defconst (see Defining Global Variables), or defcustom (see Defining Customization Variables). Usually, the definition should be at top-level in an Emacs Lisp file. As far as possible, it should include a documentation string which explains the meaning and purpose of the variable. You should also choose the variable’s name to avoid name conflicts (see Emacs Lisp Coding Conventions).

    Then you can bind the variable anywhere in a program, knowing reliably what the effect will be. Wherever you encounter the variable, it will be easy to refer back to the definition, e.g., via the C-h v command (provided the variable definition has been loaded into Emacs). See Name Help in The GNU Emacs Manual.

    For example, it is common to use local bindings for customizable variables like case-fold-search:

    (defun search-for-abc ()
      "Search for the string \"abc\", ignoring case differences."
      (let ((case-fold-search t))
        (re-search-forward "abc")))
    

12.10.3 Lexical Binding

Lexical binding was introduced to Emacs, as an optional feature, in version 24.1. We expect its importance to increase with time. Lexical binding opens up many more opportunities for optimization, so programs using it are likely to run faster in future Emacs versions. Lexical binding is also more compatible with concurrency, which was added to Emacs in version 26.1.

A lexically-bound variable has lexical scope, meaning that any reference to the variable must be located textually within the binding construct. Here is an example (see Using Lexical Binding, for how to actually enable lexical binding):

(let ((x 1))    ; x is lexically bound.
  (+ x 3))
     ⇒ 4

(defun getx ()
  x)            ; x is used free in this function.

(let ((x 1))    ; x is lexically bound.
  (getx))
error→ Symbol's value as variable is void: x

Here, the variable x has no global value. When it is lexically bound within a let form, it can be used in the textual confines of that let form. But it can not be used from within a getx function called from the let form, since the function definition of getx occurs outside the let form itself.

Here is how lexical binding works. Each binding construct defines a lexical environment, specifying the variables that are bound within the construct and their local values. When the Lisp evaluator wants the current value of a variable, it looks first in the lexical environment; if the variable is not specified in there, it looks in the symbol’s value cell, where the dynamic value is stored.

(Internally, the lexical environment is a list whose members are usually cons cells that are symbol-value pairs, but some of its members can be symbols rather than cons cells. A symbol in the list means the lexical environment declared that symbol’s variable as locally considered to be dynamically bound. This list can be passed as the second argument to the eval function, in order to specify a lexical environment in which to evaluate a form. See Eval. Most Emacs Lisp programs, however, should not interact directly with lexical environments in this way; only specialized programs like debuggers.)

Lexical bindings have indefinite extent. Even after a binding construct has finished executing, its lexical environment can be “kept around” in Lisp objects called closures. A closure is created when you define a named or anonymous function with lexical binding enabled. See Closures, for details.

When a closure is called as a function, any lexical variable references within its definition use the retained lexical environment. Here is an example:

(defvar my-ticker nil)   ; We will use this dynamically bound
                         ; variable to store a closure.

(let ((x 0))             ; x is lexically bound.
  (setq my-ticker (lambda ()
                    (setq x (1+ x)))))
    ⇒ (closure ((x . 0)) ()
          (setq x (1+ x)))

(funcall my-ticker)
    ⇒ 1

(funcall my-ticker)
    ⇒ 2

(funcall my-ticker)
    ⇒ 3

x                        ; Note that x has no global value.
error→ Symbol's value as variable is void: x

The let binding defines a lexical environment in which the variable x is locally bound to 0. Within this binding construct, we define a lambda expression which increments x by one and returns the incremented value. This lambda expression is automatically turned into a closure, in which the lexical environment lives on even after the let binding construct has exited. Each time we evaluate the closure, it increments x, using the binding of x in that lexical environment.

Note that unlike dynamic variables which are tied to the symbol object itself, the relationship between lexical variables and symbols is only present in the interpreter (or compiler). Therefore, functions which take a symbol argument (like symbol-value, boundp, and set) can only retrieve or modify a variable’s dynamic binding (i.e., the contents of its symbol’s value cell).


12.10.4 Using Lexical Binding

When loading an Emacs Lisp file or evaluating a Lisp buffer, lexical binding is enabled if the buffer-local variable lexical-binding is non-nil:

Variable: lexical-binding

If this buffer-local variable is non-nil, Emacs Lisp files and buffers are evaluated using lexical binding instead of dynamic binding. (However, special variables are still dynamically bound; see below.) If nil, dynamic binding is used for all local variables. This variable is typically set for a whole Emacs Lisp file, as a file local variable (see File Local Variables). Note that unlike other such variables, this one must be set in the first line of a file.

When evaluating Emacs Lisp code directly using an eval call, lexical binding is enabled if the lexical argument to eval is non-nil. See Eval.

Lexical binding is also enabled in Lisp Interaction and IELM mode, used in the *scratch* and *ielm* buffers, and also when evaluating expressions via M-: (eval-expression) and when processing the --eval command-line options of Emacs (see Action Arguments in The GNU Emacs Manual) and emacsclient (see emacsclient Options in The GNU Emacs Manual).

Even when lexical binding is enabled, certain variables will continue to be dynamically bound. These are called special variables. Every variable that has been defined with defvar, defcustom or defconst is a special variable (see Defining Global Variables). All other variables are subject to lexical binding.

Using defvar without a value, it is possible to bind a variable dynamically just in one file, or in just one part of a file while still binding it lexically elsewhere. For example:

(let (_)
  (defvar x)      ; Let-bindings of x will be dynamic within this let.
  (let ((x -99))  ; This is a dynamic binding of x.
    (defun get-dynamic-x ()
      x)))

(let ((x 'lexical)) ; This is a lexical binding of x.
  (defun get-lexical-x ()
    x))

(let (_)
  (defvar x)
  (let ((x 'dynamic))
    (list (get-lexical-x)
          (get-dynamic-x))))
    ⇒ (lexical dynamic)
Function: special-variable-p symbol

This function returns non-nil if symbol is a special variable (i.e., it has a defvar, defcustom, or defconst variable definition). Otherwise, the return value is nil.

Note that since this is a function, it can only return non-nil for variables which are permanently special, but not for those that are only special in the current lexical scope.

The use of a special variable as a formal argument in a function is not supported.


12.10.5 Converting to Lexical Binding

Converting an Emacs Lisp program to lexical binding is easy. First, add a file-local variable setting of lexical-binding to t in the header line of the Emacs Lisp source file (see File Local Variables). Second, check that every variable in the program which needs to be dynamically bound has a variable definition, so that it is not inadvertently bound lexically.

A simple way to find out which variables need a variable definition is to byte-compile the source file. See Byte Compilation. If a non-special variable is used outside of a let form, the byte-compiler will warn about reference or assignment to a free variable. If a non-special variable is bound but not used within a let form, the byte-compiler will warn about an unused lexical variable. The byte-compiler will also issue a warning if you use a special variable as a function argument.

A warning about a reference or an assignment to a free variable is usually a clear sign that that variable should be marked as dynamically scoped, so you need to add an appropriate defvar before the first use of that variable.

A warning about an unused variable may be a good hint that the variable was intended to be dynamically scoped (because it is actually used, but in another function), but it may also be an indication that the variable is simply really not used and could simply be removed. So you need to find out which case it is, and based on that, either add a defvar or remove the variable altogether. If removal is not possible or not desirable (typically because it is a formal argument and that we cannot or don’t want to change all the callers), you can also add a leading underscore to the variable’s name to indicate to the compiler that this is a variable known not to be used.)

Cross-file variable checking

Caution: This is an experimental feature that may change or disappear without prior notice.

The byte-compiler can also warn about lexical variables that are special in other Emacs Lisp files, often indicating a missing defvar declaration. This useful but somewhat specialized check requires three steps:

  1. Byte-compile all files whose special variable declarations may be of interest, with the environment variable EMACS_GENERATE_DYNVARS set to a nonempty string. These are typically all the files in the same package or related packages or Emacs subsystems. The process will generate a file whose name ends in .dynvars for each compiled Emacs Lisp file.
  2. Concatenate the .dynvars files into a single file.
  3. Byte-compile the files that need to be checked, this time with the environment variable EMACS_DYNVARS_FILE set to the name of the aggregated file created in step 2.

Here is an example illustrating how this could be done, assuming that a Unix shell and make are used for byte-compilation:

$ rm *.elc                                # force recompilation
$ EMACS_GENERATE_DYNVARS=1 make           # generate .dynvars
$ cat *.dynvars > ~/my-dynvars            # combine .dynvars
$ rm *.elc                                # force recompilation
$ EMACS_DYNVARS_FILE=~/my-dynvars make    # perform checks

12.11 Buffer-Local Variables

Global and local variable bindings are found in most programming languages in one form or another. Emacs, however, also supports additional, unusual kinds of variable binding, such as buffer-local bindings, which apply only in one buffer. Having different values for a variable in different buffers is an important customization method. (Variables can also have bindings that are local to each terminal. See Multiple Terminals.)


12.11.1 Introduction to Buffer-Local Variables

A buffer-local variable has a buffer-local binding associated with a particular buffer. The binding is in effect when that buffer is current; otherwise, it is not in effect. If you set the variable while a buffer-local binding is in effect, the new value goes in that binding, so its other bindings are unchanged. This means that the change is visible only in the buffer where you made it.

The variable’s ordinary binding, which is not associated with any specific buffer, is called the default binding. In most cases, this is the global binding.

A variable can have buffer-local bindings in some buffers but not in other buffers. The default binding is shared by all the buffers that don’t have their own bindings for the variable. (This includes all newly-created buffers.) If you set the variable in a buffer that does not have a buffer-local binding for it, this sets the default binding, so the new value is visible in all the buffers that see the default binding.

The most common use of buffer-local bindings is for major modes to change variables that control the behavior of commands. For example, C mode and Lisp mode both set the variable paragraph-start to specify that only blank lines separate paragraphs. They do this by making the variable buffer-local in the buffer that is being put into C mode or Lisp mode, and then setting it to the new value for that mode. See Major Modes.

The usual way to make a buffer-local binding is with make-local-variable, which is what major mode commands typically use. This affects just the current buffer; all other buffers (including those yet to be created) will continue to share the default value unless they are explicitly given their own buffer-local bindings.

A more powerful operation is to mark the variable as automatically buffer-local by calling make-variable-buffer-local. You can think of this as making the variable local in all buffers, even those yet to be created. More precisely, the effect is that setting the variable automatically makes the variable local to the current buffer if it is not already so. All buffers start out by sharing the default value of the variable as usual, but setting the variable creates a buffer-local binding for the current buffer. The new value is stored in the buffer-local binding, leaving the default binding untouched. This means that the default value cannot be changed with setq in any buffer; the only way to change it is with setq-default.

Warning: When a variable has buffer-local bindings in one or more buffers, let rebinds the binding that’s currently in effect. For instance, if the current buffer has a buffer-local value, let temporarily rebinds that. If no buffer-local bindings are in effect, let rebinds the default value. If inside the let you then change to a different current buffer in which a different binding is in effect, you won’t see the let binding any more. And if you exit the let while still in the other buffer, you won’t see the unbinding occur (though it will occur properly). Here is an example to illustrate:

(setq foo 'g)
(set-buffer "a")
(make-local-variable 'foo)
(setq foo 'a)
(let ((foo 'temp))
  ;; foo ⇒ 'temp  ; let binding in buffer ‘a
  (set-buffer "b")
  ;; foo ⇒ 'g     ; the global value since foo is not local in ‘b
  body…)
foo ⇒ 'g        ; exiting restored the local value in buffer ‘a’,
                 ; but we don’t see that in buffer ‘b
(set-buffer "a") ; verify the local value was restored
foo ⇒ 'a

Note that references to foo in body access the buffer-local binding of buffer ‘b’.

When a file specifies local variable values, these become buffer-local values when you visit the file. See File Variables in The GNU Emacs Manual.

A buffer-local variable cannot be made terminal-local (see Multiple Terminals).


12.11.2 Creating and Deleting Buffer-Local Bindings

Command: make-local-variable variable

This function creates a buffer-local binding in the current buffer for variable (a symbol). Other buffers are not affected. The value returned is variable.

The buffer-local value of variable starts out as the same value variable previously had. If variable was void, it remains void.

;; In buffer ‘b1’:
(setq foo 5)                ; Affects all buffers.
     ⇒ 5
(make-local-variable 'foo)  ; Now it is local in ‘b1’.
     ⇒ foo
foo                         ; That did not change
     ⇒ 5                   ;   the value.
(setq foo 6)                ; Change the value
     ⇒ 6                   ;   in ‘b1’.
foo
     ⇒ 6

;; In buffer ‘b2’, the value hasn’t changed.
(with-current-buffer "b2"
  foo)
     ⇒ 5

Making a variable buffer-local within a let-binding for that variable does not work reliably, unless the buffer in which you do this is not current either on entry to or exit from the let. This is because let does not distinguish between different kinds of bindings; it knows only which variable the binding was made for.

It is an error to make a constant or a read-only variable buffer-local. See Variables that Never Change.

If the variable is terminal-local (see Multiple Terminals), this function signals an error. Such variables cannot have buffer-local bindings as well.

Warning: do not use make-local-variable for a hook variable. The hook variables are automatically made buffer-local as needed if you use the local argument to add-hook or remove-hook.

Macro: setq-local &rest pairs

pairs is a list of variable and value pairs. This macro creates a buffer-local binding in the current buffer for each of the variables, and gives them a buffer-local value. It is equivalent to calling make-local-variable followed by setq for each of the variables. The variables should be unquoted symbols.

(setq-local var1 "value1"
            var2 "value2")
Command: make-variable-buffer-local variable

This function marks variable (a symbol) automatically buffer-local, so that any subsequent attempt to set it will make it local to the current buffer at the time. Unlike make-local-variable, with which it is often confused, this cannot be undone, and affects the behavior of the variable in all buffers.

A peculiar wrinkle of this feature is that binding the variable (with let or other binding constructs) does not create a buffer-local binding for it. Only setting the variable (with set or setq), while the variable does not have a let-style binding that was made in the current buffer, does so.

If variable does not have a default value, then calling this command will give it a default value of nil. If variable already has a default value, that value remains unchanged. Subsequently calling makunbound on variable will result in a void buffer-local value and leave the default value unaffected.

The value returned is variable.

It is an error to make a constant or a read-only variable buffer-local. See Variables that Never Change.

Warning: Don’t assume that you should use make-variable-buffer-local for user-option variables, simply because users might want to customize them differently in different buffers. Users can make any variable local, when they wish to. It is better to leave the choice to them.

The time to use make-variable-buffer-local is when it is crucial that no two buffers ever share the same binding. For example, when a variable is used for internal purposes in a Lisp program which depends on having separate values in separate buffers, then using make-variable-buffer-local can be the best solution.

Macro: defvar-local variable value &optional docstring

This macro defines variable as a variable with initial value value and docstring, and marks it as automatically buffer-local. It is equivalent to calling defvar followed by make-variable-buffer-local. variable should be an unquoted symbol.

Function: local-variable-p variable &optional buffer

This returns t if variable is buffer-local in buffer buffer (which defaults to the current buffer); otherwise, nil.

Function: local-variable-if-set-p variable &optional buffer

This returns t if variable either has a buffer-local value in buffer buffer, or is automatically buffer-local. Otherwise, it returns nil. If omitted or nil, buffer defaults to the current buffer.

Function: buffer-local-value variable buffer

This function returns the buffer-local binding of variable (a symbol) in buffer buffer. If variable does not have a buffer-local binding in buffer buffer, it returns the default value (see The Default Value of a Buffer-Local Variable) of variable instead.

Function: buffer-local-boundp variable buffer

This returns non-nil if there’s either a buffer-local binding of variable (a symbol) in buffer buffer, or variable has a global binding.

Function: buffer-local-variables &optional buffer

This function returns a list describing the buffer-local variables in buffer buffer. (If buffer is omitted, the current buffer is used.) Normally, each list element has the form (sym . val), where sym is a buffer-local variable (a symbol) and val is its buffer-local value. But when a variable’s buffer-local binding in buffer is void, its list element is just sym.

(make-local-variable 'foobar)
(makunbound 'foobar)
(make-local-variable 'bind-me)
(setq bind-me 69)
(setq lcl (buffer-local-variables))
    ;; First, built-in variables local in all buffers:
⇒ ((mark-active . nil)
    (buffer-undo-list . nil)
    (mode-name . "Fundamental")
    …
    ;; Next, non-built-in buffer-local variables.
    ;; This one is buffer-local and void:
    foobar
    ;; This one is buffer-local and nonvoid:
    (bind-me . 69))

Note that storing new values into the CDRs of cons cells in this list does not change the buffer-local values of the variables.

Command: kill-local-variable variable

This function deletes the buffer-local binding (if any) for variable (a symbol) in the current buffer. As a result, the default binding of variable becomes visible in this buffer. This typically results in a change in the value of variable, since the default value is usually different from the buffer-local value just eliminated.

If you kill the buffer-local binding of a variable that automatically becomes buffer-local when set, this makes the default value visible in the current buffer. However, if you set the variable again, that will once again create a buffer-local binding for it.

kill-local-variable returns variable.

This function is a command because it is sometimes useful to kill one buffer-local variable interactively, just as it is useful to create buffer-local variables interactively.

Function: kill-all-local-variables &optional kill-permanent

This function eliminates all the buffer-local variable bindings of the current buffer. As a result, the buffer will see the default values of most variables. By default, for variables marked as permanent and local hook functions that have a non-nil permanent-local-hook property (see Setting Hooks) won’t be killed, but if the optional kill-permanent argument is non-nil, even these variables will be killed.

This function also resets certain other information pertaining to the buffer: it sets the local keymap to nil, the syntax table to the value of (standard-syntax-table), the case table to (standard-case-table), and the abbrev table to the value of fundamental-mode-abbrev-table.

The very first thing this function does is run the normal hook change-major-mode-hook (see below).

Every major mode command begins by calling this function, which has the effect of switching to Fundamental mode and erasing most of the effects of the previous major mode. To ensure that this does its job, the variables that major modes set should not be marked permanent.

kill-all-local-variables returns nil.

Variable: change-major-mode-hook

The function kill-all-local-variables runs this normal hook before it does anything else. This gives major modes a way to arrange for something special to be done if the user switches to a different major mode. It is also useful for buffer-specific minor modes that should be forgotten if the user changes the major mode.

For best results, make this variable buffer-local, so that it will disappear after doing its job and will not interfere with the subsequent major mode. See Hooks.

A buffer-local variable is permanent if the variable name (a symbol) has a permanent-local property that is non-nil. Such variables are unaffected by kill-all-local-variables, and their local bindings are therefore not cleared by changing major modes. Permanent locals are appropriate for data pertaining to where the file came from or how to save it, rather than with how to edit the contents.


12.11.3 The Default Value of a Buffer-Local Variable

The global value of a variable with buffer-local bindings is also called the default value, because it is the value that is in effect whenever neither the current buffer nor the selected frame has its own binding for the variable.

The functions default-value and setq-default access and change a variable’s default value regardless of whether the current buffer has a buffer-local binding. For example, you could use setq-default to change the default setting of paragraph-start for most buffers; and this would work even when you are in a C or Lisp mode buffer that has a buffer-local value for this variable.

The special forms defvar and defconst also set the default value (if they set the variable at all), rather than any buffer-local value.

Function: default-value symbol

This function returns symbol’s default value. This is the value that is seen in buffers and frames that do not have their own values for this variable. If symbol is not buffer-local, this is equivalent to symbol-value (see Accessing Variable Values).

Function: default-boundp symbol

The function default-boundp tells you whether symbol’s default value is nonvoid. If (default-boundp 'foo) returns nil, then (default-value 'foo) would get an error.

default-boundp is to default-value as boundp is to symbol-value.

Special Form: setq-default [symbol form]…

This special form gives each symbol a new default value, which is the result of evaluating the corresponding form. It does not evaluate symbol, but does evaluate form. The value of the setq-default form is the value of the last form.

If a symbol is not buffer-local for the current buffer, and is not marked automatically buffer-local, setq-default has the same effect as setq. If symbol is buffer-local for the current buffer, then this changes the value that other buffers will see (as long as they don’t have a buffer-local value), but not the value that the current buffer sees.

;; In buffer ‘foo’:
(make-local-variable 'buffer-local)
     ⇒ buffer-local
(setq buffer-local 'value-in-foo)
     ⇒ value-in-foo
(setq-default buffer-local 'new-default)
     ⇒ new-default
buffer-local
     ⇒ value-in-foo
(default-value 'buffer-local)
     ⇒ new-default

;; In (the new) buffer ‘bar’:
buffer-local
     ⇒ new-default
(default-value 'buffer-local)
     ⇒ new-default
(setq buffer-local 'another-default)
     ⇒ another-default
(default-value 'buffer-local)
     ⇒ another-default

;; Back in buffer ‘foo’:
buffer-local
     ⇒ value-in-foo
(default-value 'buffer-local)
     ⇒ another-default
Function: set-default symbol value

This function is like setq-default, except that symbol is an ordinary evaluated argument.

(set-default (car '(a b c)) 23)
     ⇒ 23
(default-value 'a)
     ⇒ 23

A variable can be let-bound (see Local Variables) to a value. This makes its global value shadowed by the binding; default-value will then return the value from that binding, not the global value, and set-default will be prevented from setting the global value (it will change the let-bound value instead). The following two functions allow to reference the global value even if it’s shadowed by a let-binding.

Function: default-toplevel-value symbol

This function returns the top-level default value of symbol, which is its value outside of any let-binding.

(defvar variable 'global-value)
    ⇒ variable
(let ((variable 'let-binding))
  (default-value 'variable))
    ⇒ let-binding
(let ((variable 'let-binding))
  (default-toplevel-value 'variable))
    ⇒ global-value
Function: set-default-toplevel-value symbol value

This function sets the top-level default value of symbol to the specified value. This comes in handy when you want to set the global value of symbol regardless of whether your code runs in the context of symbol’s let-binding.


12.12 File Local Variables

A file can specify local variable values; Emacs uses these to create buffer-local bindings for those variables in the buffer visiting that file. See Local Variables in Files in The GNU Emacs Manual, for basic information about file-local variables. This section describes the functions and variables that affect how file-local variables are processed.

If a file-local variable could specify an arbitrary function or Lisp expression that would be called later, visiting a file could take over your Emacs. Emacs protects against this by automatically setting only those file-local variables whose specified values are known to be safe. Other file-local variables are set only if the user agrees.

For additional safety, read-circle is temporarily bound to nil when Emacs reads file-local variables (see Input Functions). This prevents the Lisp reader from recognizing circular and shared Lisp structures (see Read Syntax for Circular Objects).

User Option: enable-local-variables

This variable controls whether to process file-local variables. The possible values are:

t (the default)

Set the safe variables, and query (once) about any unsafe variables.

:safe

Set only the safe variables and do not query.

:all

Set all the variables and do not query.

nil

Don’t set any variables.

anything else

Query (once) about all the variables.

Variable: inhibit-local-variables-regexps

This is a list of regular expressions. If a file has a name matching an element of this list, then it is not scanned for any form of file-local variable. For examples of why you might want to use this, see How Emacs Chooses a Major Mode.

Variable: permanently-enabled-local-variables

Some local variable settings will, by default, be heeded even if enable-local-variables is nil. By default, this is only the case for the lexical-binding local variable setting, but this can be controlled by using this variable, which is a list of symbols.

Function: hack-local-variables &optional handle-mode

This function parses, and binds or evaluates as appropriate, any local variables specified by the contents of the current buffer. The variable enable-local-variables has its effect here. However, this function does not look for the ‘mode:’ local variable in the ‘-*- line. set-auto-mode does that, also taking enable-local-variables into account (see How Emacs Chooses a Major Mode).

This function works by walking the alist stored in file-local-variables-alist and applying each local variable in turn. It calls before-hack-local-variables-hook and hack-local-variables-hook before and after applying the variables, respectively. It only calls the before-hook if the alist is non-nil; it always calls the other hook. This function ignores a ‘mode’ element if it specifies the same major mode as the buffer already has.

If the optional argument handle-mode is t, then all this function does is return a symbol specifying the major mode, if the ‘-*- line or the local variables list specifies one, and nil otherwise. It does not set the mode or any other file-local variable. If handle-mode has any value other than nil or t, any settings of ‘mode’ in the ‘-*- line or the local variables list are ignored, and the other settings are applied. If handle-mode is nil, all the file local variables are set.

Variable: file-local-variables-alist

This buffer-local variable holds the alist of file-local variable settings. Each element of the alist is of the form (var . value), where var is a symbol of the local variable and value is its value. When Emacs visits a file, it first collects all the file-local variables into this alist, and then the hack-local-variables function applies them one by one.

Variable: before-hack-local-variables-hook

Emacs calls this hook immediately before applying file-local variables stored in file-local-variables-alist.

Variable: hack-local-variables-hook

Emacs calls this hook immediately after it finishes applying file-local variables stored in file-local-variables-alist.

You can specify safe values for a variable with a safe-local-variable property. The property has to be a function of one argument; any value is safe if the function returns non-nil given that value. Many commonly-encountered file variables have safe-local-variable properties; these include fill-column, fill-prefix, and indent-tabs-mode. For boolean-valued variables that are safe, use booleanp as the property value.

If you want to define safe-local-variable properties for variables defined in C source code, add the names and the properties of those variables to the list in the “Safe local variables” section of files.el.

When defining a user option using defcustom, you can set its safe-local-variable property by adding the arguments :safe function to defcustom (see Defining Customization Variables). However, a safety predicate defined using :safe will only be known once the package that contains the defcustom is loaded, which is often too late. As an alternative, you can use the autoload cookie (see Autoload) to assign the option its safety predicate, like this:

;;;###autoload (put 'var 'safe-local-variable 'pred)

The safe value definitions specified with autoload are copied into the package’s autoloads file (loaddefs.el for most packages bundled with Emacs), and are known to Emacs since the beginning of a session.

User Option: safe-local-variable-values

This variable provides another way to mark some variable values as safe. It is a list of cons cells (var . val), where var is a variable name and val is a value which is safe for that variable.

When Emacs asks the user whether or not to obey a set of file-local variable specifications, the user can choose to mark them as safe. Doing so adds those variable/value pairs to safe-local-variable-values, and saves it to the user’s custom file.

User Option: ignored-local-variable-values

If there are some values of particular local variables that you always want to ignore completely, you can use this variable. Its value has the same form as safe-local-variable-values; a file-local variable setting to the value that appears in the list will always be ignored when processing the local variables specified by the file. As with that variable, when Emacs queries the user about whether to obey file-local variables, the user can choose to ignore their particular values permanently, and that will alter this variable and save it to the user’s custom file. Variable-value pairs that appear in this variable take precedence over the same pairs in safe-local-variable-values.

Function: safe-local-variable-p sym val

This function returns non-nil if it is safe to give sym the value val, based on the above criteria.

Some variables are considered risky. If a variable is risky, it is never entered automatically into safe-local-variable-values; Emacs always queries before setting a risky variable, unless the user explicitly allows a value by customizing safe-local-variable-values directly.

Any variable whose name has a non-nil risky-local-variable property is considered risky. When you define a user option using defcustom, you can set its risky-local-variable property by adding the arguments :risky value to defcustom (see Defining Customization Variables). In addition, any variable whose name ends in any of ‘-command’, ‘-frame-alist’, ‘-function’, ‘-functions’, ‘-hook’, ‘-hooks’, ‘-form’, ‘-forms’, ‘-map’, ‘-map-alist’, ‘-mode-alist’, ‘-program’, or ‘-predicate’ is automatically considered risky. The variables ‘font-lock-keywords’, ‘font-lock-keywords’ followed by a digit, and ‘font-lock-syntactic-keywords’ are also considered risky.

Function: risky-local-variable-p sym

This function returns non-nil if sym is a risky variable, based on the above criteria.

Variable: ignored-local-variables

This variable holds a list of variables that should not be given local values by files. Any value specified for one of these variables is completely ignored.

The ‘Eval:’ “variable” is also a potential loophole, so Emacs normally asks for confirmation before handling it.

User Option: enable-local-eval

This variable controls processing of ‘Eval:’ in ‘-*-’ lines or local variables lists in files being visited. A value of t means process them unconditionally; nil means ignore them; anything else means ask the user what to do for each file. The default value is maybe.

User Option: safe-local-eval-forms

This variable holds a list of expressions that are safe to evaluate when found in the ‘Eval:’ “variable” in a file local variables list.

If the expression is a function call and the function has a safe-local-eval-function property, the property value determines whether the expression is safe to evaluate. The property value can be a predicate to call to test the expression, a list of such predicates (it’s safe if any predicate succeeds), or t (always safe provided the arguments are constant).

Text properties are also potential loopholes, since their values could include functions to call. So Emacs discards all text properties from string values specified for file-local variables.


12.13 Directory Local Variables

A directory can specify local variable values common to all files in that directory; Emacs uses these to create buffer-local bindings for those variables in buffers visiting any file in that directory. This is useful when the files in the directory belong to some project and therefore share the same local variables.

There are two different methods for specifying directory local variables: by putting them in a special file, or by defining a project class for that directory.

Constant: dir-locals-file

This constant is the name of the file where Emacs expects to find the directory-local variables. The name of the file is .dir-locals.el11. A file by that name in a directory causes Emacs to apply its settings to any file in that directory or any of its subdirectories (optionally, you can exclude subdirectories; see below). If some of the subdirectories have their own .dir-locals.el files, Emacs uses the settings from the deepest file it finds starting from the file’s directory and moving up the directory tree. This constant is also used to derive the name of a second dir-locals file .dir-locals-2.el. If this second dir-locals file is present, then that is loaded in addition to .dir-locals.el. This is useful when .dir-locals.el is under version control in a shared repository and cannot be used for personal customizations. The file specifies local variables as a specially formatted list; see Per-directory Local Variables in The GNU Emacs Manual, for more details.

Function: hack-dir-local-variables

This function reads the .dir-locals.el file and stores the directory-local variables in file-local-variables-alist that is local to the buffer visiting any file in the directory, without applying them. It also stores the directory-local settings in dir-locals-class-alist, where it defines a special class for the directory in which .dir-locals.el file was found. This function works by calling dir-locals-set-class-variables and dir-locals-set-directory-class, described below.

Function: hack-dir-local-variables-non-file-buffer

This function looks for directory-local variables, and immediately applies them in the current buffer. It is intended to be called in the mode commands for non-file buffers, such as Dired buffers, to let them obey directory-local variable settings. For non-file buffers, Emacs looks for directory-local variables in default-directory and its parent directories.

Function: dir-locals-set-class-variables class variables

This function defines a set of variable settings for the named class, which is a symbol. You can later assign the class to one or more directories, and Emacs will apply those variable settings to all files in those directories. The list in variables can be of one of the two forms: (major-mode . alist) or (directory . list). With the first form, if the file’s buffer turns on a mode that is derived from major-mode, then all the variables in the associated alist are applied; alist should be of the form (name . value). A special value nil for major-mode means the settings are applicable to any mode. In alist, you can use a special name: subdirs. If the associated value is nil, the alist is only applied to files in the relevant directory, not to those in any subdirectories.

With the second form of variables, if directory is the initial substring of the file’s directory, then list is applied recursively by following the above rules; list should be of one of the two forms accepted by this function in variables.

Function: dir-locals-set-directory-class directory class &optional mtime

This function assigns class to all the files in directory and its subdirectories. Thereafter, all the variable settings specified for class will be applied to any visited file in directory and its children. class must have been already defined by dir-locals-set-class-variables.

Emacs uses this function internally when it loads directory variables from a .dir-locals.el file. In that case, the optional argument mtime holds the file modification time (as returned by file-attributes). Emacs uses this time to check stored local variables are still valid. If you are assigning a class directly, not via a file, this argument should be nil.

Variable: dir-locals-class-alist

This alist holds the class symbols and the associated variable settings. It is updated by dir-locals-set-class-variables.

Variable: dir-locals-directory-cache

This alist holds directory names, their assigned class names, and modification times of the associated directory local variables file (if there is one). The function dir-locals-set-directory-class updates this list.

Variable: enable-dir-local-variables

If nil, directory-local variables are ignored. This variable may be useful for modes that want to ignore directory-locals while still respecting file-local variables (see File Local Variables).


12.14 Connection Local Variables

Connection-local variables provide a general mechanism for different variable settings in buffers with a remote connection (see Remote Files in The GNU Emacs Manual). They are bound and set depending on the remote connection a buffer is dedicated to.


12.14.1 Connection Local Profiles

Emacs uses connection-local profiles to store the variable settings to apply to particular connections. You can then associate these with remote connections by defining the criteria when they should apply, using connection-local-set-profiles.

Function: connection-local-set-profile-variables profile variables

This function defines a set of variable settings for the connection profile, which is a symbol. You can later assign the connection profile to one or more remote connections, and Emacs will apply those variable settings to all process buffers for those connections. The list in variables is an alist of the form (name . value). Example:

(connection-local-set-profile-variables
  'remote-bash
  '((shell-file-name . "/bin/bash")
    (shell-command-switch . "-c")
    (shell-interactive-switch . "-i")
    (shell-login-switch . "-l")))

(connection-local-set-profile-variables
  'remote-ksh
  '((shell-file-name . "/bin/ksh")
    (shell-command-switch . "-c")
    (shell-interactive-switch . "-i")
    (shell-login-switch . "-l")))

(connection-local-set-profile-variables
  'remote-null-device
  '((null-device . "/dev/null")))

If you want to append variable settings to an existing profile, you could use the function connection-local-get-profile-variables in order to retrieve the existing settings, like

(connection-local-set-profile-variables
  'remote-bash
  (append
   (connection-local-get-profile-variables 'remote-bash)
   '((shell-command-dont-erase-buffer . t))))
User Option: connection-local-profile-alist

This alist holds the connection profile symbols and the associated variable settings. It is updated by connection-local-set-profile-variables.

Function: connection-local-set-profiles criteria &rest profiles

This function assigns profiles, which are symbols, to all remote connections identified by criteria. criteria is a plist identifying a connection and the application using this connection. Property names might be :application, :protocol, :user and :machine. The property value of :application is a symbol, all other property values are strings. All properties are optional; if criteria is nil, it always applies. Example:

(connection-local-set-profiles
  '(:application tramp :protocol "ssh" :machine "localhost")
  'remote-bash 'remote-null-device)

(connection-local-set-profiles
  '(:application tramp :protocol "sudo"
    :user "root" :machine "localhost")
  'remote-ksh 'remote-null-device)

If criteria is nil, it applies for all remote connections. Therefore, the example above would be equivalent to

(connection-local-set-profiles
  '(:application tramp :protocol "ssh" :machine "localhost")
  'remote-bash)

(connection-local-set-profiles
  '(:application tramp :protocol "sudo"
    :user "root" :machine "localhost")
  'remote-ksh)

(connection-local-set-profiles
  nil 'remote-null-device)

Any connection profile of profiles must have been already defined by connection-local-set-profile-variables.

User Option: connection-local-criteria-alist

This alist contains connection criteria and their assigned profile names. The function connection-local-set-profiles updates this list.


12.14.2 Applying Connection Local Variables

When writing connection-aware code, you’ll need to collect, and possibly apply, any connection-local variables. There are several ways to do this, as described below.

Function: hack-connection-local-variables criteria

This function collects applicable connection-local variables associated with criteria in connection-local-variables-alist, without applying them. Example:

(hack-connection-local-variables
  '(:application tramp :protocol "ssh" :machine "localhost"))

connection-local-variables-alist
     ⇒ ((null-device . "/dev/null")
        (shell-login-switch . "-l")
        (shell-interactive-switch . "-i")
        (shell-command-switch . "-c")
        (shell-file-name . "/bin/bash"))
Function: hack-connection-local-variables-apply criteria

This function looks for connection-local variables according to criteria, and immediately applies them in the current buffer.

Macro: with-connection-local-application-variables application &rest body

Apply all connection-local variables for application, which are specified by default-directory.

After that, body is executed, and the connection-local variables are unwound. Example:

(connection-local-set-profile-variables
  'my-remote-perl
  '((perl-command-name . "/usr/local/bin/perl5")
    (perl-command-switch . "-e %s")))

(connection-local-set-profiles
  '(:application my-app :protocol "ssh" :machine "remotehost")
  'my-remote-perl)

(let ((default-directory "/ssh:remotehost:/working/dir/"))
  (with-connection-local-application-variables 'my-app
    do something useful))
Variable: connection-local-default-application

The default application, a symbol, to be applied in with-connection-local-variables. It defaults to tramp, but you can let-bind it to change the application temporarily (see Local Variables).

This variable must not be changed globally.

Macro: with-connection-local-variables &rest body

This is equivalent to with-connection-local-application-variables, but uses connection-local-default-application for the application.

Macro: setq-connection-local [symbol form]…

This macro sets each symbol connection-locally to the result of evaluating the corresponding form, using the connection-local profile specified in connection-local-profile-name-for-setq; if the profile name is nil, this macro will just set the variables normally, as with setq (see Setting Variable Values).

For example, you can use this macro in combination with with-connection-local-variables or with-connection-local-application-variables to lazily initialize connection-local settings:

(defvar my-app-variable nil)

(connection-local-set-profile-variables
 'my-app-connection-default-profile
 '((my-app-variable . nil)))

(connection-local-set-profiles
 '(:application my-app)
 'my-app-connection-default-profile)

(defun my-app-get-variable ()
  (with-connection-local-application-variables 'my-app
    (or my-app-variable
        (setq-connection-local my-app-variable
                               do something useful))))
Variable: connection-local-profile-name-for-setq

The connection-local profile name, a symbol, to use when setting variables via setq-connection-local. This is let-bound in the body of with-connection-local-variables, but you can also let-bind it yourself if you’d like to set variables on a different profile.

This variable must not be changed globally.

Variable: enable-connection-local-variables

If nil, connection-local variables are ignored. This variable shall be changed temporarily only in special modes.


12.15 Variable Aliases

It is sometimes useful to make two variables synonyms, so that both variables always have the same value, and changing either one also changes the other. Whenever you change the name of a variable—either because you realize its old name was not well chosen, or because its meaning has partly changed—it can be useful to keep the old name as an alias of the new one for compatibility. You can do this with defvaralias.

Function: defvaralias new-alias base-variable &optional docstring

This function defines the symbol new-alias as a variable alias for symbol base-variable. This means that retrieving the value of new-alias returns the value of base-variable, and changing the value of new-alias changes the value of base-variable. The two aliased variable names always share the same value and the same bindings.

If the docstring argument is non-nil, it specifies the documentation for new-alias; otherwise, the alias gets the same documentation as base-variable has, if any, unless base-variable is itself an alias, in which case new-alias gets the documentation of the variable at the end of the chain of aliases.

This function returns base-variable.

Variable aliases are convenient for replacing an old name for a variable with a new name. make-obsolete-variable declares that the old name is obsolete and therefore that it may be removed at some stage in the future.

Function: make-obsolete-variable obsolete-name current-name when &optional access-type

This function makes the byte compiler warn that the variable obsolete-name is obsolete. If current-name is a symbol, it is the variable’s new name; then the warning message says to use current-name instead of obsolete-name. If current-name is a string, this is the message and there is no replacement variable. when should be a string indicating when the variable was first made obsolete (usually a version number string).

The optional argument access-type, if non-nil, should specify the kind of access that will trigger obsolescence warnings; it can be either get or set.

You can make two variables synonyms and declare one obsolete at the same time using the macro define-obsolete-variable-alias.

Macro: define-obsolete-variable-alias obsolete-name current-name when &optional docstring

This macro marks the variable obsolete-name as obsolete and also makes it an alias for the variable current-name. It is equivalent to the following:

(defvaralias obsolete-name current-name docstring)
(make-obsolete-variable obsolete-name current-name when)

This macro evaluates all its parameters, and both obsolete-name and current-name should be symbols, so a typical usage would look like:

(define-obsolete-variable-alias 'foo-thing 'bar-thing "27.1")
Function: indirect-variable variable

This function returns the variable at the end of the chain of aliases of variable. If variable is not a symbol, or if variable is not defined as an alias, the function returns variable.

This function signals a cyclic-variable-indirection error if there is a loop in the chain of symbols.

(defvaralias 'foo 'bar)
(indirect-variable 'foo)
     ⇒ bar
(indirect-variable 'bar)
     ⇒ bar
(setq bar 2)
bar
     ⇒ 2
foo
     ⇒ 2
(setq foo 0)
bar
     ⇒ 0
foo
     ⇒ 0

12.16 Variables with Restricted Values

Ordinary Lisp variables can be assigned any value that is a valid Lisp object. However, certain Lisp variables are not defined in Lisp, but in C. Most of these variables are defined in the C code using DEFVAR_LISP. Like variables defined in Lisp, these can take on any value. However, some variables are defined using DEFVAR_INT or DEFVAR_BOOL. See Writing Emacs Primitives, in particular the description of functions of the type syms_of_filename, for a brief discussion of the C implementation.

Variables of type DEFVAR_BOOL can only take on the values nil or t. Attempting to assign them any other value will set them to t:

(let ((display-hourglass 5))
  display-hourglass)
     ⇒ t
Variable: byte-boolean-vars

This variable holds a list of all variables of type DEFVAR_BOOL.

Variables of type DEFVAR_INT can take on only integer values. Attempting to assign them any other value will result in an error:

(setq undo-limit 1000.0)
error→ Wrong type argument: integerp, 1000.0

12.17 Generalized Variables

A generalized variable or place form is one of the many places in Lisp memory where values can be stored using the setf macro (see The setf Macro). The simplest place form is a regular Lisp variable. But the CARs and CDRs of lists, elements of arrays, properties of symbols, and many other locations are also places where Lisp values get stored.

Generalized variables are analogous to lvalues in the C language, where ‘x = a[i]’ gets an element from an array and ‘a[i] = x’ stores an element using the same notation. Just as certain forms like a[i] can be lvalues in C, there is a set of forms that can be generalized variables in Lisp.


12.17.1 The setf Macro

The setf macro is the most basic way to operate on generalized variables. The setf form is like setq, except that it accepts arbitrary place forms in the first (left) argument of each pair rather than just symbols. For example, (setf (car a) b) sets the car of a to b, doing the same operation as (setcar a b), but without you having to use two separate functions for setting and accessing this type of place.

Macro: setf [place form]…

This macro evaluates form and stores its value in place, which must be a valid generalized variable form. If there are several place and form pairs, the assignments are done sequentially just as with setq. setf returns the value of the last form.

The following Lisp forms are the forms in Emacs that will work as generalized variables, and so may appear in the place argument of setf:

  • A symbol. In other words, (setf x y) is exactly equivalent to (setq x y), and setq itself is strictly speaking redundant given that setf exists. Most programmers will continue to prefer setq for setting simple variables, though, for stylistic and historical reasons. The macro (setf x y) actually expands to (setq x y), so there is no performance penalty for using it in compiled code.
  • A call to any of the following standard Lisp functions:
    aref      cddr      symbol-function
    car       elt       symbol-plist
    caar      get       symbol-value
    cadr      gethash
    cdr       nth
    cdar      nthcdr
    
  • A call to any of the following Emacs-specific functions:
    alist-get                     overlay-start
    default-value                 overlay-get
    face-background               process-buffer
    face-font                     process-filter
    face-foreground               process-get
    face-stipple                  process-sentinel
    face-underline-p              terminal-parameter
    file-modes                    window-buffer
    frame-parameter               window-dedicated-p
    frame-parameters              window-display-table
    get-register                  window-hscroll
    getenv                        window-parameter
    keymap-parent                 window-point
    match-data                    window-start
    overlay-end
    
  • A call of the form (substring subplace n [m]), where subplace is itself a valid generalized variable whose current value is a string, and where the value stored is also a string. The new string is spliced into the specified part of the destination string. For example:
    (setq a (list "hello" "world"))
         ⇒ ("hello" "world")
    (cadr a)
         ⇒ "world"
    (substring (cadr a) 2 4)
         ⇒ "rl"
    (setf (substring (cadr a) 2 4) "o")
         ⇒ "o"
    (cadr a)
         ⇒ "wood"
    a
         ⇒ ("hello" "wood")
    
  • The if and cond conditionals will work as generalized variables. For instance, this will set either the foo or the bar variable to zot:
    (setf (if (zerop (random 2))
    	  foo
    	bar)
          'zot)
    

setf signals an error if you pass a place form that it does not know how to handle.

Note that for nthcdr, the list argument of the function must itself be a valid place form. For example, (setf (nthcdr 0 foo) 7) will set foo itself to 7.

The macros push (see Modifying List Variables) and pop (see Accessing Elements of Lists) can manipulate generalized variables, not just lists. (pop place) removes and returns the first element of the list stored in place. It is analogous to (prog1 (car place) (setf place (cdr place))), except that it takes care to evaluate all subforms only once. (push x place) inserts x at the front of the list stored in place. It is analogous to (setf place (cons x place)), except for evaluation of the subforms. Note that push and pop on an nthcdr place can be used to insert or delete at any position in a list.

The cl-lib library defines various extensions for generalized variables, including additional setf places. See Generalized Variables in Common Lisp Extensions.


12.17.2 Defining new setf forms

This section describes how to define new forms that setf can operate on.

Macro: gv-define-simple-setter name setter &optional fix-return

This macro enables you to easily define setf methods for simple cases. name is the name of a function, macro, or special form. You can use this macro whenever name has a directly corresponding setter function that updates it, e.g., (gv-define-simple-setter car setcar).

This macro translates a call of the form

(setf (name args…) value)

into

(setter argsvalue)

Such a setf call is documented to return value. This is no problem with, e.g., car and setcar, because setcar returns the value that it set. If your setter function does not return value, use a non-nil value for the fix-return argument of gv-define-simple-setter. This expands into something equivalent to

(let ((temp value))
  (setter args… temp)
  temp)

so ensuring that it returns the correct result.

Macro: gv-define-setter name arglist &rest body

This macro allows for more complex setf expansions than the previous form. You may need to use this form, for example, if there is no simple setter function to call, or if there is one but it requires different arguments to the place form.

This macro expands the form (setf (name args…) value) by first binding the setf argument forms (value args…) according to arglist, and then executing body. body should return a Lisp form that does the assignment, and finally returns the value that was set. An example of using this macro is:

(gv-define-setter caar (val x) `(setcar (car ,x) ,val))
Macro: gv-define-expander name handler

For more control over the expansion, the gv-define-expander macro can be used. For instance, a settable substring could be implemented this way:

(gv-define-expander substring
  (lambda (do place from &optional to)
    (gv-letplace (getter setter) place
      (macroexp-let2* (from to)
        (funcall do `(substring ,getter ,from ,to)
                 (lambda (v)
                   (macroexp-let2* (v)
                     `(progn
                        ,(funcall setter `(cl--set-substring
                                           ,getter ,from ,to ,v))
                        ,v))))))))
Macro: gv-letplace (getter setter) place &rest body

The macro gv-letplace can be useful in defining macros that perform similarly to setf; for example, the incf macro of Common Lisp could be implemented this way:

(defmacro incf (place &optional n)
  (gv-letplace (getter setter) place
    (macroexp-let2* ((v (or n 1)))
      (funcall setter `(+ ,v ,getter)))))

getter will be bound to a copyable expression that returns the value of place. setter will be bound to a function that takes an expression v and returns a new expression that sets place to v. body should return a Emacs Lisp expression manipulating place via getter and setter.

Consult the source file gv.el for more details.

Function: make-obsolete-generalized-variable obsolete-name current-name when

This function makes the byte compiler warn that the generalized variable obsolete-name is obsolete. If current-name is a symbol, then the warning message says to use current-name instead of obsolete-name. If current-name is a string, this is the message. when should be a string indicating when the variable was first made obsolete (usually a version number string).

Common Lisp note: Common Lisp defines another way to specify the setf behavior of a function, namely setf functions, whose names are lists (setf name) rather than symbols. For example, (defun (setf foo) …) defines the function that is used when setf is applied to foo. Emacs does not support this. It is a compile-time error to use setf on a form that has not already had an appropriate expansion defined. In Common Lisp, this is not an error since the function (setf func) might be defined later.


12.18 Multisession Variables

When you set a variable to a value and then close Emacs and restart it, that value won’t be automatically restored. Users usually set normal variables in their startup files, or use Customize (see Customization Settings) to set user options permanently, and various packages have various files where they store the data (e.g., Gnus stores this in .newsrc.eld and the URL library stores cookies in ~/.emacs.d/url/cookies).

For things in between these two extremes (i.e., configuration which goes in the startup file, and massive application state that goes into separate files), Emacs provides a facility to replicate data between sessions called multisession variables. (This facility may not be available on all systems.) To give you an idea of how these are meant to be used, here’s a small example:

(define-multisession-variable foo 0)
(defun my-adder (num)
  (interactive "nAdd number: ")
  (setf (multisession-value foo)
        (+ (multisession-value foo) num))
  (message "The new number is: %s" (multisession-value foo)))

This defines the variable foo and binds it to a special multisession object which is initialized with the value ‘0’ (if the variable doesn’t already exist from a previous session). The my-adder command queries the user for a number, adds this to the old (possibly saved value), and then saves the new value.

This facility isn’t meant to be used for huge data structures, but should be performant for most values.

Macro: define-multisession-variable name initial-value &optional doc &rest args

This macro defines name as a multisession variable, and gives it the initial-value if this variable hasn’t been assigned a value earlier. doc is the doc string, and several keyword arguments can be used in args:

:package package-symbol

This keyword says that a multisession variable belongs to the package specified by package-symbol. The combination of package-symbol and name has to be unique. If package-symbol isn’t given, this will default to the first “segment” of the name symbol’s name, which is the part of its name up to and excluding the first ‘-’. For instance, if name is foo and package-symbol isn’t given, package-symbol will default to foo.

:synchronized bool

Multisession variables can be synchronized if bool is non-nil. This means that if there’re two concurrent Emacs instances running, and the other Emacs changes the multisession variable foo, the current Emacs instance will retrieve that modified data when accessing the value. If synchronized is nil or missing, this won’t happen, and the values in all Emacs sessions using the variable will be independent of each other.

:storage storage

Use the specified storage method. This can be either sqlite (in Emacs compiled with SQLite support) or files. If not given, this defaults to the value of the multisession-storage variable, described below.

Function: multisession-value variable

This function returns the current value of variable. If this variable hasn’t been accessed before in this Emacs session, or if it’s changed externally, it will be read in from external storage. If not, the current value in this session is returned as is. It is an error to call this function for a variable that is not a multisession variable.

Values retrieved via multisession-value may or may not be eq to each other, but they will always be equal.

This is a generalized variable (see Generalized Variables), so the way to update such a variable is to say, for instance:

(setf (multisession-value foo-bar) 'zot)

Only Emacs Lisp values that have a readable print syntax (see Printed Representation and Read Syntax) can be saved this way.

If the multisession variable is synchronized, setting it may update the value first. For instance:

(cl-incf (multisession-value foo-bar))

This first checks whether the value has changed in a different Emacs instance, retrieves that value, and then adds 1 to that value and stores it. But note that this is done without locking, so if many instances are updating the value at the same time, it’s unpredictable which instance “wins”.

Function: multisession-delete object

This function deletes object and its value from its persistent storage.

Function: make-multisession

You can also make persistent values that aren’t tied to a specific variable, but are tied to an explicit package and key.

(setq foo (make-multisession :package "mail"
                             :key "friends"))
(setf (multisession-value foo) 'everybody)

This supports the same keywords as define-multisession-variable, but also supports a :initial-value keyword, which specifies the default value.

User Option: multisession-storage

This variable controls how the multisession variables are stored. It value defaults to files, which means that the values are stored in a one-file-per-variable structure inside the directory specified by multisession-directory. If this value is sqlite instead, the values are stored in an SQLite database; this is only available if Emacs was built with SQLite support.

User Option: multisession-directory

The multisession variables are stored under this directory, which defaults to multisession/ subdirectory of the user-emacs-directory, which is typically ~/.emacs.d/multisession/.

Command: list-multisession-values

This command pops up a buffer listing all the multisession variables, and enters a special mode multisession-edit-mode which allows you to delete them and edit their values.


13 Functions

A Lisp program is composed mainly of Lisp functions. This chapter explains what functions are, how they accept arguments, and how to define them.


13.1 What Is a Function?

In a general sense, a function is a rule for carrying out a computation given input values called arguments. The result of the computation is called the value or return value of the function. The computation can also have side effects, such as lasting changes in the values of variables or the contents of data structures (see Definition of side effect). A pure function is a function which, in addition to having no side effects, always returns the same value for the same combination of arguments, regardless of external factors such as machine type or system state.

In most computer languages, every function has a name. But in Lisp, a function in the strictest sense has no name: it is an object which can optionally be associated with a symbol (e.g., car) that serves as the function name. See Naming a Function. When a function has been given a name, we usually also refer to that symbol as a “function” (e.g., we refer to “the function car”). In this manual, the distinction between a function name and the function object itself is usually unimportant, but we will take note wherever it is relevant.

Certain function-like objects, called special forms and macros, also accept arguments to carry out computations. However, as explained below, these are not considered functions in Emacs Lisp.

Here are important terms for functions and function-like objects:

lambda expression

A function (in the strict sense, i.e., a function object) which is written in Lisp. These are described in the following section. See Lambda Expressions.

primitive

A function which is callable from Lisp but is actually written in C. Primitives are also called built-in functions, or subrs. Examples include functions like car and append. In addition, all special forms (see below) are also considered primitives.

Usually, a function is implemented as a primitive because it is a fundamental part of Lisp (e.g., car), or because it provides a low-level interface to operating system services, or because it needs to run fast. Unlike functions defined in Lisp, primitives can be modified or added only by changing the C sources and recompiling Emacs. See Writing Emacs Primitives.

special form

A primitive that is like a function but does not evaluate all of its arguments in the usual way. It may evaluate only some of the arguments, or may evaluate them in an unusual order, or several times. Examples include if, and, and while. See Special Forms.

macro

A construct defined in Lisp, which differs from a function in that it translates a Lisp expression into another expression which is to be evaluated instead of the original expression. Macros enable Lisp programmers to do the sorts of things that special forms can do. See Macros.

command

An object which can be invoked via the command-execute primitive, usually due to the user typing in a key sequence bound to that command. See Interactive Call. A command is usually a function; if the function is written in Lisp, it is made into a command by an interactive form in the function definition (see Defining Commands). Commands that are functions can also be called from Lisp expressions, just like other functions.

Keyboard macros (strings and vectors) are commands also, even though they are not functions. See Keyboard Macros. We say that a symbol is a command if its function cell contains a command (see Symbol Components); such a named command can be invoked with M-x.

closure

A function object that is much like a lambda expression, except that it also encloses an environment of lexical variable bindings. See Closures.

byte-code function

A function that has been compiled by the byte compiler. See Byte-Code Function Type.

autoload object

A place-holder for a real function. If the autoload object is called, Emacs loads the file containing the definition of the real function, and then calls the real function. See Autoload.

You can use the function functionp to test if an object is a function:

Function: functionp object

This function returns t if object is any kind of function, i.e., can be passed to funcall. Note that functionp returns t for symbols that are function names, and returns nil for symbols that are macros or special forms.

If object is not a function, this function ordinarily returns nil. However, the representation of function objects is complicated, and for efficiency reasons in rare cases this function can return t even when object is not a function.

It is also possible to find out how many arguments an arbitrary function expects:

Function: func-arity function

This function provides information about the argument list of the specified function. The returned value is a cons cell of the form (min . max), where min is the minimum number of arguments, and max is either the maximum number of arguments, or the symbol many for functions with &rest arguments, or the symbol unevalled if function is a special form.

Note that this function might return inaccurate results in some situations, such as the following:

  • Functions defined using apply-partially (see apply-partially).
  • Functions that are advised using advice-add (see Advising Named Functions).
  • Functions that determine the argument list dynamically, as part of their code.

Unlike functionp, the next three functions do not treat a symbol as its function definition.

Function: subrp object

This function returns t if object is a built-in function (i.e., a Lisp primitive).

(subrp 'message)            ; message is a symbol,
     ⇒ nil                 ;   not a subr object.
(subrp (symbol-function 'message))
     ⇒ t
Function: byte-code-function-p object

This function returns t if object is a byte-code function. For example:

(byte-code-function-p (symbol-function 'next-line))
     ⇒ t
Function: compiled-function-p object

This function returns t if object is a function object that is not in the form of ELisp source code but something like machine code or byte code instead. More specifically it returns t if the function is built-in (a.k.a. “primitive”, see What Is a Function?), or byte-compiled (see Byte Compilation), or natively-compiled (see Compilation of Lisp to Native Code), or a function loaded from a dynamic module (see Emacs Dynamic Modules).

Function: subr-arity subr

This works like func-arity, but only for built-in functions and without symbol indirection. It signals an error for non-built-in functions. We recommend to use func-arity instead.


13.2 Lambda Expressions

A lambda expression is a function object written in Lisp. Here is an example:

(lambda (x)
  "Return the hyperbolic cosine of X."
  (* 0.5 (+ (exp x) (exp (- x)))))

In Emacs Lisp, such a list is a valid expression which evaluates to a function object.

A lambda expression, by itself, has no name; it is an anonymous function. Although lambda expressions can be used this way (see Anonymous Functions), they are more commonly associated with symbols to make named functions (see Naming a Function). Before going into these details, the following subsections describe the components of a lambda expression and what they do.


13.2.1 Components of a Lambda Expression

A lambda expression is a list that looks like this:

(lambda (arg-variables…)
  [documentation-string]
  [interactive-declaration]
  body-forms…)

The first element of a lambda expression is always the symbol lambda. This indicates that the list represents a function. The reason functions are defined to start with lambda is so that other lists, intended for other uses, will not accidentally be valid as functions.

The second element is a list of symbols—the argument variable names (see Features of Argument Lists). This is called the lambda list. When a Lisp function is called, the argument values are matched up against the variables in the lambda list, which are given local bindings with the values provided. See Local Variables.

The documentation string is a Lisp string object placed within the function definition to describe the function for the Emacs help facilities. See Documentation Strings of Functions.

The interactive declaration is a list of the form (interactive code-string). This declares how to provide arguments if the function is used interactively. Functions with this declaration are called commands; they can be called using M-x or bound to a key. Functions not intended to be called in this way should not have interactive declarations. See Defining Commands, for how to write an interactive declaration.

The rest of the elements are the body of the function: the Lisp code to do the work of the function (or, as a Lisp programmer would say, “a list of Lisp forms to evaluate”). The value returned by the function is the value returned by the last element of the body.


13.2.2 A Simple Lambda Expression Example

Consider the following example:

(lambda (a b c) (+ a b c))

We can call this function by passing it to funcall, like this:

(funcall (lambda (a b c) (+ a b c))
         1 2 3)

This call evaluates the body of the lambda expression with the variable a bound to 1, b bound to 2, and c bound to 3. Evaluation of the body adds these three numbers, producing the result 6; therefore, this call to the function returns the value 6.

Note that the arguments can be the results of other function calls, as in this example:

(funcall (lambda (a b c) (+ a b c))
         1 (* 2 3) (- 5 4))

This evaluates the arguments 1, (* 2 3), and (- 5 4) from left to right. Then it applies the lambda expression to the argument values 1, 6 and 1 to produce the value 8.

As these examples show, you can use a form with a lambda expression as its CAR to make local variables and give them values. In the old days of Lisp, this technique was the only way to bind and initialize local variables. But nowadays, it is clearer to use the special form let for this purpose (see Local Variables). Lambda expressions are mainly used as anonymous functions for passing as arguments to other functions (see Anonymous Functions), or stored as symbol function definitions to produce named functions (see Naming a Function).


13.2.3 Features of Argument Lists

Our simple sample function, (lambda (a b c) (+ a b c)), specifies three argument variables, so it must be called with three arguments: if you try to call it with only two arguments or four arguments, you get a wrong-number-of-arguments error (see Errors).

It is often convenient to write a function that allows certain arguments to be omitted. For example, the function substring accepts three arguments—a string, the start index and the end index—but the third argument defaults to the length of the string if you omit it. It is also convenient for certain functions to accept an indefinite number of arguments, as the functions list and + do.

To specify optional arguments that may be omitted when a function is called, simply include the keyword &optional before the optional arguments. To specify a list of zero or more extra arguments, include the keyword &rest before one final argument.

Thus, the complete syntax for an argument list is as follows:

(required-vars[&optional [optional-vars]]
 [&rest rest-var])

The square brackets indicate that the &optional and &rest clauses, and the variables that follow them, are optional.

A call to the function requires one actual argument for each of the required-vars. There may be actual arguments for zero or more of the optional-vars, and there cannot be any actual arguments beyond that unless the lambda list uses &rest. In that case, there may be any number of extra actual arguments.

If actual arguments for the optional and rest variables are omitted, then they always default to nil. There is no way for the function to distinguish between an explicit argument of nil and an omitted argument. However, the body of the function is free to consider nil an abbreviation for some other meaningful value. This is what substring does; nil as the third argument to substring means to use the length of the string supplied.

Common Lisp note: Common Lisp allows the function to specify what default value to use when an optional argument is omitted; Emacs Lisp always uses nil. Emacs Lisp does not support supplied-p variables that tell you whether an argument was explicitly passed.

For example, an argument list that looks like this:

(a b &optional c d &rest e)

binds a and b to the first two actual arguments, which are required. If one or two more arguments are provided, c and d are bound to them respectively; any arguments after the first four are collected into a list and e is bound to that list. Thus, if there are only two arguments, c, d and e are nil; if two or three arguments, d and e are nil; if four arguments or fewer, e is nil. Note that exactly five arguments with an explicit nil argument provided for e will cause that nil argument to be passed as a list with one element, (nil), as with any other single value for e.

There is no way to have required arguments following optional ones—it would not make sense. To see why this must be so, suppose that c in the example were optional and d were required. Suppose three actual arguments are given; which variable would the third argument be for? Would it be used for the c, or for d? One can argue for both possibilities. Similarly, it makes no sense to have any more arguments (either required or optional) after a &rest argument.

Here are some examples of argument lists and proper calls:

(funcall (lambda (n) (1+ n))        ; One required:
         1)                         ; requires exactly one argument.
     ⇒ 2
(funcall (lambda (n &optional n1)   ; One required and one optional:
           (if n1 (+ n n1) (1+ n))) ; 1 or 2 arguments.
         1 2)
     ⇒ 3
(funcall (lambda (n &rest ns)       ; One required and one rest:
           (+ n (apply '+ ns)))     ; 1 or more arguments.
         1 2 3 4 5)
     ⇒ 15

13.2.4 Documentation Strings of Functions

A lambda expression may optionally have a documentation string just after the lambda list. This string does not affect execution of the function; it is a kind of comment, but a systematized comment which actually appears inside the Lisp world and can be used by the Emacs help facilities. See Documentation, for how the documentation string is accessed.

It is a good idea to provide documentation strings for all the functions in your program, even those that are called only from within your program. Documentation strings are like comments, except that they are easier to access.

The first line of the documentation string should stand on its own, because apropos displays just this first line. It should consist of one or two complete sentences that summarize the function’s purpose.

The start of the documentation string is usually indented in the source file, but since these spaces come before the starting double-quote, they are not part of the string. Some people make a practice of indenting any additional lines of the string so that the text lines up in the program source. That is a mistake. The indentation of the following lines is inside the string; what looks nice in the source code will look ugly when displayed by the help commands.

You may wonder how the documentation string could be optional, since there are required components of the function that follow it (the body). Since evaluation of a string returns that string, without any side effects, it has no effect if it is not the last form in the body. Thus, in practice, there is no confusion between the first form of the body and the documentation string; if the only body form is a string then it serves both as the return value and as the documentation.

The last line of the documentation string can specify calling conventions different from the actual function arguments. Write text like this:

\(fn arglist)

following a blank line, at the beginning of the line, with no newline following it inside the documentation string. (The ‘\’ is used to avoid confusing the Emacs motion commands.) The calling convention specified in this way appears in help messages in place of the one derived from the actual arguments of the function.

This feature is particularly useful for macro definitions, since the arguments written in a macro definition often do not correspond to the way users think of the parts of the macro call.

Do not use this feature if you want to deprecate the calling convention and favor the one you advertise by the above specification. Instead, use the advertised-calling-convention declaration (see The declare Form) or set-advertised-calling-convention (see Declaring Functions Obsolete), because these two will cause the byte compiler emit a warning message when it compiles Lisp programs which use the deprecated calling convention.

The (fn) feature is typically used in the following situations:

  • To spell out arguments and their purposes in a macro or a function. Example:
    (defmacro lambda (&rest cdr)
      "…
    \(fn ARGS [DOCSTRING] [INTERACTIVE] BODY)"…)
    
  • To provide a more detailed description and names of arguments. Example:
    (defmacro macroexp--accumulate (var+list &rest body)
      "…
    \(fn (VAR LIST) BODY…)"
      (declare (indent 1))
      (let ((var (car var+list))
    	  (list (cadr var+list))
    …)))
    
  • To better explain the purpose of a defalias. Example:
    (defalias 'abbrev-get 'get
      "…
    \(fn ABBREV PROP)")
    

Documentation strings are usually static, but occasionally it can be necessary to generate them dynamically. In some cases you can do so by writing a macro which generates at compile time the code of the function, including the desired documentation string. But you can also generate the docstring dynamically by writing (:documentation form) instead of the documentation string. This will evaluate form at run-time when the function is defined and use it as the documentation string12. You can also compute the documentation string on the fly when it is requested, by setting the function-documentation property of the function’s symbol to a Lisp form that evaluates to a string.

For example:

(defun adder (x)
  (lambda (y)
    (:documentation (format "Add %S to the argument Y." x))
    (+ x y)))
(defalias 'adder5 (adder 5))
(documentation 'adder5)
    ⇒ "Add 5 to the argument Y."

(put 'adder5 'function-documentation
     '(concat (documentation (symbol-function 'adder5) 'raw)
              "  Consulted at " (format-time-string "%H:%M:%S")))
(documentation 'adder5)
    ⇒ "Add 5 to the argument Y.  Consulted at 15:52:13"
(documentation 'adder5)
    ⇒ "Add 5 to the argument Y.  Consulted at 15:52:18"

13.3 Naming a Function

A symbol can serve as the name of a function. This happens when the symbol’s function cell (see Symbol Components) contains a function object (e.g., a lambda expression). Then the symbol itself becomes a valid, callable function, equivalent to the function object in its function cell.

The contents of the function cell are also called the symbol’s function definition. The procedure of using a symbol’s function definition in place of the symbol is called symbol function indirection; see Symbol Function Indirection. If you have not given a symbol a function definition, its function cell is said to be void, and it cannot be used as a function.

In practice, nearly all functions have names, and are referred to by their names. You can create a named Lisp function by defining a lambda expression and putting it in a function cell (see Accessing Function Cell Contents). However, it is more common to use the defun macro, described in the next section. See Defining Functions.

We give functions names because it is convenient to refer to them by their names in Lisp expressions. Also, a named Lisp function can easily refer to itself—it can be recursive. Furthermore, primitives can only be referred to textually by their names, since primitive function objects (see Primitive Function Type) have no read syntax.

A function need not have a unique name. A given function object usually appears in the function cell of only one symbol, but this is just a convention. It is easy to store it in several symbols using fset; then each of the symbols is a valid name for the same function.

Note that a symbol used as a function name may also be used as a variable; these two uses of a symbol are independent and do not conflict. (This is not the case in some dialects of Lisp, like Scheme.)

By convention, if a function’s symbol consists of two names separated by ‘--’, the function is intended for internal use and the first part names the file defining the function. For example, a function named vc-git--rev-parse is an internal function defined in vc-git.el. Internal-use functions written in C have names ending in ‘-internal’, e.g., bury-buffer-internal. Emacs code contributed before 2018 may follow other internal-use naming conventions, which are being phased out.


13.4 Defining Functions

We usually give a name to a function when it is first created. This is called defining a function, and we usually do it with the defun macro. This section also describes other ways to define a function.

Macro: defun name args [doc] [declare] [interactive] body…

defun is the usual way to define new Lisp functions. It defines the symbol name as a function with argument list args (see Features of Argument Lists) and body forms given by body. Neither name nor args should be quoted.

doc, if present, should be a string specifying the function’s documentation string (see Documentation Strings of Functions). declare, if present, should be a declare form specifying function metadata (see The declare Form). interactive, if present, should be an interactive form specifying how the function is to be called interactively (see Interactive Call).

The return value of defun is undefined.

Here are some examples:

(defun foo () 5)
(foo)
     ⇒ 5

(defun bar (a &optional b &rest c)
    (list a b c))
(bar 1 2 3 4 5)
     ⇒ (1 2 (3 4 5))
(bar 1)
     ⇒ (1 nil nil)
(bar)
error→ Wrong number of arguments.

(defun capitalize-backwards ()
  "Upcase the last letter of the word at point."
  (interactive)
  (backward-word 1)
  (forward-word 1)
  (backward-char 1)
  (capitalize-word 1))

Most Emacs functions are part of the source code of Lisp programs, and are defined when the Emacs Lisp reader reads the program source before executing it. However, you can also define functions dynamically at run time, e.g., by generating defun calls when your program’s code is executed. If you do this, be aware that Emacs’s Help commands, such as C-h f, which present in the *Help* buffer a button to jump to the function’s definition, might be unable to find the source code because generating a function dynamically usually looks very different from the usual static calls to defun. You can make the job of finding the code which generates such functions easier by using the definition-name property, see Standard Symbol Properties.

Be careful not to redefine existing functions unintentionally. defun redefines even primitive functions such as car without any hesitation or notification. Emacs does not prevent you from doing this, because redefining a function is sometimes done deliberately, and there is no way to distinguish deliberate redefinition from unintentional redefinition.

Function: defalias name definition &optional doc

This function defines the symbol name as a function, with definition definition. definition can be any valid Lisp function or macro, or a special form (see Special Forms), or a keymap (see Keymaps), or a vector or string (a keyboard macro). The return value of defalias is undefined.

If doc is non-nil, it becomes the function documentation of name. Otherwise, any documentation provided by definition is used.

Internally, defalias normally uses fset to set the definition. If name has a defalias-fset-function property, however, the associated value is used as a function to call in place of fset.

The proper place to use defalias is where a specific function or macro name is being defined—especially where that name appears explicitly in the source file being loaded. This is because defalias records which file defined the function, just like defun (see Unloading).

By contrast, in programs that manipulate function definitions for other purposes, it is better to use fset, which does not keep such records. See Accessing Function Cell Contents.

Function: function-alias-p object &optional noerror

Checks whether object is a function alias. If it is, it returns a list of symbols representing the function alias chain, else nil. For instance, if a is an alias for b, and b is an alias for c:

(function-alias-p 'a)
    ⇒ (b c)

If there’s a loop in the definitions, an error will be signaled. If noerror is non-nil, the non-looping parts of the chain is returned instead.

You cannot create a new primitive function with defun or defalias, but you can use them to change the function definition of any symbol, even one such as car or x-popup-menu whose normal definition is a primitive. However, this is risky: for instance, it is next to impossible to redefine car without breaking Lisp completely. Redefining an obscure function such as x-popup-menu is less dangerous, but it still may not work as you expect. If there are calls to the primitive from C code, they call the primitive’s C definition directly, so changing the symbol’s definition will have no effect on them.

See also defsubst, which defines a function like defun and tells the Lisp compiler to perform inline expansion on it. See Inline Functions.

To undefine a function name, use fmakunbound. See Accessing Function Cell Contents.


13.5 Calling Functions

Defining functions is only half the battle. Functions don’t do anything until you call them, i.e., tell them to run. Calling a function is also known as invocation.

The most common way of invoking a function is by evaluating a list. For example, evaluating the list (concat "a" "b") calls the function concat with arguments "a" and "b". See Evaluation, for a description of evaluation.

When you write a list as an expression in your program, you specify which function to call, and how many arguments to give it, in the text of the program. Usually that’s just what you want. Occasionally you need to compute at run time which function to call. To do that, use the function funcall. When you also need to determine at run time how many arguments to pass, use apply.

Function: funcall function &rest arguments

funcall calls function with arguments, and returns whatever function returns.

Since funcall is a function, all of its arguments, including function, are evaluated before funcall is called. This means that you can use any expression to obtain the function to be called. It also means that funcall does not see the expressions you write for the arguments, only their values. These values are not evaluated a second time in the act of calling function; the operation of funcall is like the normal procedure for calling a function, once its arguments have already been evaluated.

The argument function must be either a Lisp function or a primitive function. Special forms and macros are not allowed, because they make sense only when given the unevaluated argument expressions. funcall cannot provide these because, as we saw above, it never knows them in the first place.

If you need to use funcall to call a command and make it behave as if invoked interactively, use funcall-interactively (see Interactive Call).

(setq f 'list)
     ⇒ list
(funcall f 'x 'y 'z)
     ⇒ (x y z)
(funcall f 'x 'y '(z))
     ⇒ (x y (z))
(funcall 'and t nil)
error→ Invalid function: #<subr and>

Compare these examples with the examples of apply.

Function: apply function &rest arguments

apply calls function with arguments, just like funcall but with one difference: the last of arguments is a list of objects, which are passed to function as separate arguments, rather than a single list. We say that apply spreads this list so that each individual element becomes an argument.

apply with a single argument is special: the first element of the argument, which must be a non-empty list, is called as a function with the remaining elements as individual arguments. Passing two or more arguments will be faster.

apply returns the result of calling function. As with funcall, function must either be a Lisp function or a primitive function; special forms and macros do not make sense in apply.

(setq f 'list)
     ⇒ list
(apply f 'x 'y 'z)
error→ Wrong type argument: listp, z
(apply '+ 1 2 '(3 4))
     ⇒ 10
(apply '+ '(1 2 3 4))
     ⇒ 10

(apply 'append '((a b c) nil (x y z) nil))
     ⇒ (a b c x y z)

(apply '(+ 3 4))
     ⇒ 7

For an interesting example of using apply, see Definition of mapcar.

Sometimes it is useful to fix some of the function’s arguments at certain values, and leave the rest of arguments for when the function is actually called. The act of fixing some of the function’s arguments is called partial application of the function13. The result is a new function that accepts the rest of arguments and calls the original function with all the arguments combined.

Here’s how to do partial application in Emacs Lisp:

Function: apply-partially func &rest args

This function returns a new function which, when called, will call func with the list of arguments composed from args and additional arguments specified at the time of the call. If func accepts n arguments, then a call to apply-partially with m <= n arguments will produce a new function of n - m arguments14.

Here’s how we could define the built-in function 1+, if it didn’t exist, using apply-partially and +, another built-in function15:

(defalias '1+ (apply-partially '+ 1)
  "Increment argument by one.")
(1+ 10)
     ⇒ 11

It is common for Lisp functions to accept functions as arguments or find them in data structures (especially in hook variables and property lists) and call them using funcall or apply. Functions that accept function arguments are often called functionals.

Sometimes, when you call a functional, it is useful to supply a no-op function as the argument. Here are three different kinds of no-op functions:

Function: identity argument

This function returns argument and has no side effects.

Function: ignore &rest arguments

This function ignores any arguments and returns nil.

Function: always &rest arguments

This function ignores any arguments and returns t.

Some functions are user-visible commands, which can be called interactively (usually by a key sequence). It is possible to invoke such a command exactly as though it was called interactively, by using the call-interactively function. See Interactive Call.


13.6 Mapping Functions

A mapping function applies a given function (not a special form or macro) to each element of a list or other collection. Emacs Lisp has several such functions; this section describes mapcar, mapc, mapconcat, and mapcan, which map over a list. See Definition of mapatoms, for the function mapatoms which maps over the symbols in an obarray. See Definition of maphash, for the function maphash which maps over key/value associations in a hash table.

These mapping functions do not allow char-tables because a char-table is a sparse array whose nominal range of indices is very large. To map over a char-table in a way that deals properly with its sparse nature, use the function map-char-table (see Char-Tables).

Function: mapcar function sequence

mapcar applies function to each element of sequence in turn, and returns a list of the results.

The argument sequence can be any kind of sequence except a char-table; that is, a list, a vector, a bool-vector, or a string. The result is always a list. The length of the result is the same as the length of sequence. For example:

(mapcar #'car '((a b) (c d) (e f)))
     ⇒ (a c e)
(mapcar #'1+ [1 2 3])
     ⇒ (2 3 4)
(mapcar #'string "abc")
     ⇒ ("a" "b" "c")

;; Call each function in my-hooks.
(mapcar 'funcall my-hooks)

(defun mapcar* (function &rest args)
  "Apply FUNCTION to successive cars of all ARGS.
Return the list of results."
  ;; If no list is exhausted,
  (if (not (memq nil args))
      ;; apply function to CARs.
      (cons (apply function (mapcar #'car args))
            (apply #'mapcar* function
                   ;; Recurse for rest of elements.
                   (mapcar #'cdr args)))))

(mapcar* #'cons '(a b c) '(1 2 3 4))
     ⇒ ((a . 1) (b . 2) (c . 3))
Function: mapcan function sequence

This function applies function to each element of sequence, like mapcar, but instead of collecting the results into a list, it returns a single list with all the elements of the results (which must be lists), by altering the results (using nconc; see Functions that Rearrange Lists). Like with mapcar, sequence can be of any type except a char-table.

;; Contrast this:
(mapcar #'list '(a b c d))
     ⇒ ((a) (b) (c) (d))
;; with this:
(mapcan #'list '(a b c d))
     ⇒ (a b c d)
Function: mapc function sequence

mapc is like mapcar except that function is used for side-effects only—the values it returns are ignored, not collected into a list. mapc always returns sequence.

Function: mapconcat function sequence &optional separator

mapconcat applies function to each element of sequence; the results, which must be sequences of characters (strings, vectors, or lists), are concatenated into a single string return value. Between each pair of result sequences, mapconcat inserts the characters from separator, which also must be a string, or a vector or list of characters; a nil value is treated as the empty string. See Sequences, Arrays, and Vectors.

The argument function must be a function that can take one argument and returns a sequence of characters: a string, a vector, or a list. The argument sequence can be any kind of sequence except a char-table; that is, a list, a vector, a bool-vector, or a string.

(mapconcat #'symbol-name
           '(The cat in the hat)
           " ")
     ⇒ "The cat in the hat"

(mapconcat (lambda (x) (format "%c" (1+ x)))
           "HAL-8000")
     ⇒ "IBM.9111"

13.7 Anonymous Functions

Although functions are usually defined with defun and given names at the same time, it is sometimes convenient to use an explicit lambda expression—an anonymous function. Anonymous functions are valid wherever function names are. They are often assigned as variable values, or as arguments to functions; for instance, you might pass one as the function argument to mapcar, which applies that function to each element of a list (see Mapping Functions). See describe-symbols example, for a realistic example of this.

When defining a lambda expression that is to be used as an anonymous function, you can in principle use any method to construct the list. But typically you should use the lambda macro, or the function special form, or the #' read syntax:

Macro: lambda args [doc] [interactive] body…

This macro returns an anonymous function with argument list args, documentation string doc (if any), interactive spec interactive (if any), and body forms given by body.

Under dynamic binding, this macro effectively makes lambda forms self-quoting: evaluating a form whose CAR is lambda yields the form itself:

(lambda (x) (* x x))
     ⇒ (lambda (x) (* x x))

Note that when evaluating under lexical binding the result is a closure object (see Closures).

The lambda form has one other effect: it tells the Emacs evaluator and byte-compiler that its argument is a function, by using function as a subroutine (see below).

Special Form: function function-object

This special form returns function-object without evaluating it. In this, it is similar to quote (see Quoting). But unlike quote, it also serves as a note to the Emacs evaluator and byte-compiler that function-object is intended to be used as a function. Assuming function-object is a valid lambda expression, this has two effects:

  • When the code is byte-compiled, function-object is compiled into a byte-code function object (see Byte Compilation).
  • When lexical binding is enabled, function-object is converted into a closure. See Closures.

When function-object is a symbol and the code is byte compiled, the byte-compiler will warn if that function is not defined or might not be known at run time.

The read syntax #' is a short-hand for using function. The following forms are all equivalent:

(lambda (x) (* x x))
(function (lambda (x) (* x x)))
#'(lambda (x) (* x x))

In the following example, we define a change-property function that takes a function as its third argument, followed by a double-property function that makes use of change-property by passing it an anonymous function:

(defun change-property (symbol prop function)
  (let ((value (get symbol prop)))
    (put symbol prop (funcall function value))))

(defun double-property (symbol prop)
  (change-property symbol prop (lambda (x) (* 2 x))))

Note that we do not quote the lambda form.

If you compile the above code, the anonymous function is also compiled. This would not happen if, say, you had constructed the anonymous function by quoting it as a list:

(defun double-property (symbol prop)
  (change-property symbol prop '(lambda (x) (* 2 x))))

In that case, the anonymous function is kept as a lambda expression in the compiled code. The byte-compiler cannot assume this list is a function, even though it looks like one, since it does not know that change-property intends to use it as a function.


13.8 Generic Functions

Functions defined using defun have a hard-coded set of assumptions about the types and expected values of their arguments. For example, a function that was designed to handle values of its argument that are either numbers or lists of numbers will fail or signal an error if called with a value of any other type, such as a vector or a string. This happens because the implementation of the function is not prepared to deal with types other than those assumed during the design.

By contrast, object-oriented programs use polymorphic functions: a set of specialized functions having the same name, each one of which was written for a certain specific set of argument types. Which of the functions is actually called is decided at run time based on the types of the actual arguments.

Emacs provides support for polymorphism. Like other Lisp environments, notably Common Lisp and its Common Lisp Object System (CLOS), this support is based on generic functions. The Emacs generic functions closely follow CLOS, including use of similar names, so if you have experience with CLOS, the rest of this section will sound very familiar.

A generic function specifies an abstract operation, by defining its name and list of arguments, but (usually) no implementation. The actual implementation for several specific classes of arguments is provided by methods, which should be defined separately. Each method that implements a generic function has the same name as the generic function, but the method’s definition indicates what kinds of arguments it can handle by specializing the arguments defined by the generic function. These argument specializers can be more or less specific; for example, a string type is more specific than a more general type, such as sequence.

Note that, unlike in message-based OO languages, such as C++ and Simula, methods that implement generic functions don’t belong to a class, they belong to the generic function they implement.

When a generic function is invoked, it selects the applicable methods by comparing the actual arguments passed by the caller with the argument specializers of each method. A method is applicable if the actual arguments of the call are compatible with the method’s specializers. If more than one method is applicable, they are combined using certain rules, described below, and the combination then handles the call.

Macro: cl-defgeneric name arguments [documentation] [options-and-methods…] &rest body

This macro defines a generic function with the specified name and arguments. If body is present, it provides the default implementation. If documentation is present (it should always be), it specifies the documentation string for the generic function, in the form (:documentation docstring). The optional options-and-methods can be one of the following forms:

(declare declarations)

A declare form, as described in The declare Form.

(:argument-precedence-order &rest args)

This form affects the sorting order for combining applicable methods. Normally, when two methods are compared during combination, method arguments are examined left to right, and the first method whose argument specializer is more specific will come before the other one. The order defined by this form overrides that, and the arguments are examined according to their order in this form, and not left to right.

(:method [qualifiers…] args &rest body)

This form defines a method like cl-defmethod does.

Macro: cl-defmethod name [extra] [qualifier] arguments [&context (expr spec)…] &rest [docstring] body

This macro defines a particular implementation for the generic function called name. The implementation code is given by body. If present, docstring is the documentation string for the method. The arguments list, which must be identical in all the methods that implement a generic function, and must match the argument list of that function, provides argument specializers of the form (arg spec), where arg is the argument name as specified in the cl-defgeneric call, and spec is one of the following specializer forms:

type

This specializer requires the argument to be of the given type, one of the types from the type hierarchy described below.

(eql object)

This specializer requires the argument be eql to the given object.

(head object)

The argument must be a cons cell whose car is eql to object.

struct-type

The argument must be an instance of a class named struct-type defined with cl-defstruct (see Structures in Common Lisp Extensions for GNU Emacs Lisp), or of one of its child classes.

Method definitions can make use of a new argument-list keyword, &context, which introduces extra specializers that test the environment at the time the method is run. This keyword should appear after the list of required arguments, but before any &rest or &optional keywords. The &context specializers look much like regular argument specializers—(expr spec)—except that expr is an expression to be evaluated in the current context, and the spec is a value to compare against. For example, &context (overwrite-mode (eql t)) will make the method applicable only when overwrite-mode is turned on. The &context keyword can be followed by any number of context specializers. Because the context specializers are not part of the generic function’s argument signature, they may be omitted in methods that don’t require them.

The type specializer, (arg type), can specify one of the system types in the following list. When a parent type is specified, an argument whose type is any of its more specific child types, as well as grand-children, grand-grand-children, etc. will also be compatible.

integer

Parent type: number.

number
null

Parent type: symbol

symbol
string

Parent type: array.

array

Parent type: sequence.

cons

Parent type: list.

list

Parent type: sequence.

marker
overlay
float

Parent type: number.

window-configuration
process
window
subr
compiled-function
buffer
char-table

Parent type: array.

bool-vector

Parent type: array.

vector

Parent type: array.

frame
hash-table
font-spec
font-entity
font-object

The optional extra element, expressed as ‘:extra string’, allows you to add more methods, distinguished by string, for the same specializers and qualifiers.

The optional qualifier allows combining several applicable methods. If it is not present, the defined method is a primary method, responsible for providing the primary implementation of the generic function for the specialized arguments. You can also define auxiliary methods, by using one of the following values as qualifier:

:before

This auxiliary method will run before the primary method. More accurately, all the :before methods will run before the primary, in the most-specific-first order.

:after

This auxiliary method will run after the primary method. More accurately, all such methods will run after the primary, in the most-specific-last order.

:around

This auxiliary method will run instead of the primary method. The most specific of such methods will be run before any other method. Such methods normally use cl-call-next-method, described below, to invoke the other auxiliary or primary methods.

Functions defined using cl-defmethod cannot be made interactive, i.e. commands (see Defining Commands), by adding the interactive form to them. If you need a polymorphic command, we recommend defining a normal command that calls a polymorphic function defined via cl-defgeneric and cl-defmethod.

Each time a generic function is called, it builds the effective method which will handle this invocation by combining the applicable methods defined for the function. The process of finding the applicable methods and producing the effective method is called dispatch. The applicable methods are those all of whose specializers are compatible with the actual arguments of the call. Since all of the arguments must be compatible with the specializers, they all determine whether a method is applicable. Methods that explicitly specialize more than one argument are called multiple-dispatch methods.

The applicable methods are sorted into the order in which they will be combined. The method whose left-most argument specializer is the most specific one will come first in the order. (Specifying :argument-precedence-order as part of cl-defmethod overrides that, as described above.) If the method body calls cl-call-next-method, the next most-specific method will run. If there are applicable :around methods, the most-specific of them will run first; it should call cl-call-next-method to run any of the less specific :around methods. Next, the :before methods run in the order of their specificity, followed by the primary method, and lastly the :after methods in the reverse order of their specificity.

Function: cl-call-next-method &rest args

When invoked from within the lexical body of a primary or an :around auxiliary method, call the next applicable method for the same generic function. Normally, it is called with no arguments, which means to call the next applicable method with the same arguments that the calling method was invoked. Otherwise, the specified arguments are used instead.

Function: cl-next-method-p

This function, when called from within the lexical body of a primary or an :around auxiliary method, returns non-nil if there is a next method to call.


13.9 Accessing Function Cell Contents

The function definition of a symbol is the object stored in the function cell of the symbol. The functions described here access, test, and set the function cell of symbols.

See also the function indirect-function. See Definition of indirect-function.

Function: symbol-function symbol

This returns the object in the function cell of symbol. It does not check that the returned object is a legitimate function.

If the function cell is void, the return value is nil. To distinguish between a function cell that is void and one set to nil, use fboundp (see below).

(defun bar (n) (+ n 2))
(symbol-function 'bar)
     ⇒ (lambda (n) (+ n 2))
(fset 'baz 'bar)
     ⇒ bar
(symbol-function 'baz)
     ⇒ bar

If you have never given a symbol any function definition, we say that that symbol’s function cell is void. In other words, the function cell does not have any Lisp object in it. If you try to call the symbol as a function, Emacs signals a void-function error.

Note that void is not the same as nil or the symbol void. The symbols nil and void are Lisp objects, and can be stored into a function cell just as any other object can be (and void can be a valid function if you define it with defun). A void function cell contains no object whatsoever.

You can test the voidness of a symbol’s function definition with fboundp. After you have given a symbol a function definition, you can make it void once more using fmakunbound.

Function: fboundp symbol

This function returns t if the symbol has an object in its function cell, nil otherwise. It does not check that the object is a legitimate function.

Function: fmakunbound symbol

This function makes symbol’s function cell void, so that a subsequent attempt to access this cell will cause a void-function error. It returns symbol. (See also makunbound, in When a Variable is Void.)

(defun foo (x) x)
(foo 1)
     ⇒1
(fmakunbound 'foo)
     ⇒ foo
(foo 1)
error→ Symbol's function definition is void: foo
Function: fset symbol definition

This function stores definition in the function cell of symbol. The result is definition. Normally definition should be a function or the name of a function, but this is not checked. The argument symbol is an ordinary evaluated argument.

The primary use of this function is as a subroutine by constructs that define or alter functions, like defun or advice-add (see Advising Emacs Lisp Functions). You can also use it to give a symbol a function definition that is not a function, e.g., a keyboard macro (see Keyboard Macros):

;; Define a named keyboard macro.
(fset 'kill-two-lines "\^u2\^k")
     ⇒ "\^u2\^k"

If you wish to use fset to make an alternate name for a function, consider using defalias instead. See Definition of defalias.


13.10 Closures

As explained in Scoping Rules for Variable Bindings, Emacs can optionally enable lexical binding of variables. When lexical binding is enabled, any named function that you create (e.g., with defun), as well as any anonymous function that you create using the lambda macro or the function special form or the #' syntax (see Anonymous Functions), is automatically converted into a closure.

A closure is a function that also carries a record of the lexical environment that existed when the function was defined. When it is invoked, any lexical variable references within its definition use the retained lexical environment. In all other respects, closures behave much like ordinary functions; in particular, they can be called in the same way as ordinary functions.

See Lexical Binding, for an example of using a closure.

Currently, an Emacs Lisp closure object is represented by a list with the symbol closure as the first element, a list representing the lexical environment as the second element, and the argument list and body forms as the remaining elements:

;; lexical binding is enabled.
(lambda (x) (* x x))
     ⇒ (closure (t) (x) (* x x))

However, the fact that the internal structure of a closure is exposed to the rest of the Lisp world is considered an internal implementation detail. For this reason, we recommend against directly examining or altering the structure of closure objects.


13.11 Open Closures

Traditionally, functions are opaque objects which offer no other functionality but to call them. (Emacs Lisp functions aren’t fully opaque since you can extract some info out of them such as their docstring, their arglist, or their interactive spec, but they are still mostly opaque.) This is usually what we want, but occasionally we need functions to expose a bit more information about themselves.

Open closures, or OClosures for short, are function objects which carry additional type information and expose some information about themselves in the form of slots which you can access via accessor functions.

OClosures are defined in two steps: first you use oclosure-define to define a new OClosure type by specifying the slots carried by the OClosures of this type, and then you use oclosure-lambda to create an OClosure object of a given type.

Let’s say we want to define keyboard macros, i.e. interactive functions which re-execute a sequence of key events (see Keyboard Macros). You could do it with a plain function as follows:

(defun kbd-macro (key-sequence)
  (lambda (&optional arg)
    (interactive "P")
    (execute-kbd-macro key-sequence arg)))

But with such a definition there is no easy way to extract the key-sequence from that function, for example to print it.

We can solve this problem using OClosures as follows. First we define the type of our keyboard macros (to which we decided to add a counter slot while at it):

(oclosure-define kbd-macro
  "Keyboard macro."
  keys (counter :mutable t))

After which we can rewrite our kbd-macro function:

(defun kbd-macro (key-sequence)
  (oclosure-lambda (kbd-macro (keys key-sequence) (counter 0))
      (&optional arg)
    (interactive "P")
    (execute-kbd-macro keys arg)
    (setq counter (1+ counter))))

As you can see, the keys and counter slots of the OClosure can be accessed as local variables from within the body of the OClosure. But we can now also access them from outside of the body of the OClosure, for example to describe a keyboard macro:

(defun describe-kbd-macro (km)
  (if (not (eq 'kbd-macro (oclosure-type km)))
      (message "Not a keyboard macro")
    (let ((keys    (kbd-macro--keys km))
          (counter (kbd-macro--counter km)))
      (message "Keys=%S, called %d times" keys counter))))

Where kbd-macro--keys and kbd-macro--counter are accessor functions generated by the oclosure-define macro for oclosures whose type is kbd-macro.

Macro: oclosure-define oname &optional docstring &rest slots

This macro defines a new OClosure type along with accessor functions for its slots. oname can be a symbol (the name of the new type), or a list of the form (oname . type-props), in which case type-props is a list of additional properties of this oclosure type. slots is a list of slot descriptions where each slot can be either a symbol (the name of the slot) or it can be of the form (slot-name . slot-props), where slot-props is a property list of the corresponding slot slot-name. The OClosure type’s properties specified by type-props can include the following:

(:predicate pred-name)

This requests creation of a predicate function named pred-name. This function will be used to recognize OClosures of the type oname. If this type property is not specified, oclosure-define will generate a default name for the predicate.

(:parent otype)

This makes type otype of OClosures be the parent of the type oname. The OClosures of type oname inherit the slots defined by their parent type.

(:copier copier-name copier-args)

This causes the definition of a functional update function, knows as the copier, which takes an OClosure of type oname as its first argument and returns a copy of it with the slots named in copier-args modified to contain the value passed in the corresponding argument in the actual call to copier-name.

For each slot in slots, the oclosure-define macro creates an accessor function named oname--slot-name; these can be used to access the values of the slots. The slot definitions in slots can specify the following properties of the slots:

:mutable val

By default, slots are immutable, but if you specify the :mutable property with a non-nil value, the slot can be mutated, for example with setf (see The setf Macro).

:type val-type

This specifies the type of the values expected to appear in the slot.

Macro: oclosure-lambda (type . slots) arglist &rest body

This macro creates an anonymous OClosure of type type, which should have been defined with oclosure-define. slots should be a list of elements of the form (slot-name expr). At run time, each expr is evaluated, in order, after which the OClosure is created with its slots initialized with the resulting values.

When called as a function (see Calling Functions), the OClosure created by this macro will accept arguments according to arglist and will execute the code in body. body can refer to the value of any of its slot directly as if it were a local variable that had been captured by static scoping.

Function: oclosure-type object

This function returns the OClosure type (a symbol) of object if it is an OClosure, and nil otherwise.

One other function related to OClosures is oclosure-interactive-form, which allows some types of OClosures to compute their interactive forms dynamically. See oclosure-interactive-form.


13.12 Advising Emacs Lisp Functions

When you need to modify a function defined in another library, or when you need to modify a hook like foo-function, a process filter, or basically any variable or object field which holds a function value, you can use the appropriate setter function, such as fset or defun for named functions, setq for hook variables, or set-process-filter for process filters, but those are often too blunt, completely throwing away the previous value.

The advice feature lets you add to the existing definition of a function, by advising the function. This is a cleaner method than redefining the whole function.

Emacs’s advice system provides two sets of primitives for that: the core set, for function values held in variables and object fields (with the corresponding primitives being add-function and remove-function) and another set layered on top of it for named functions (with the main primitives being advice-add and advice-remove).

As a trivial example, here’s how to add advice that’ll modify the return value of a function every time it’s called:

(defun my-double (x)
  (* x 2))
(defun my-increase (x)
  (+ x 1))
(advice-add 'my-double :filter-return #'my-increase)

After adding this advice, if you call my-double with ‘3’, the return value will be ‘7’. To remove this advice, say

(advice-remove 'my-double #'my-increase)

A more advanced example would be to trace the calls to the process filter of a process proc:

(defun my-tracing-function (proc string)
  (message "Proc %S received %S" proc string))

(add-function :before (process-filter proc) #'my-tracing-function)

This will cause the process’s output to be passed to my-tracing-function before being passed to the original process filter. my-tracing-function receives the same arguments as the original function. When you’re done with it, you can revert to the untraced behavior with:

(remove-function (process-filter proc) #'my-tracing-function)

Similarly, if you want to trace the execution of the function named display-buffer, you could use:

(defun his-tracing-function (orig-fun &rest args)
  (message "display-buffer called with args %S" args)
  (let ((res (apply orig-fun args)))
    (message "display-buffer returned %S" res)
    res))

(advice-add 'display-buffer :around #'his-tracing-function)

Here, his-tracing-function is called instead of the original function and receives the original function (additionally to that function’s arguments) as argument, so it can call it if and when it needs to. When you’re tired of seeing this output, you can revert to the untraced behavior with:

(advice-remove 'display-buffer #'his-tracing-function)

The arguments :before and :around used in the above examples specify how the two functions are composed, since there are many different ways to do it. The added function is also called a piece of advice.


13.12.1 Primitives to manipulate advices

Macro: add-function where place function &optional props

This macro is the handy way to add the advice function to the function stored in place (see Generalized Variables).

where determines how function is composed with the existing function, e.g., whether function should be called before, or after the original function. See Ways to compose advice, for the list of available ways to compose the two functions.

When modifying a variable (whose name will usually end with -function), you can choose whether function is used globally or only in the current buffer: if place is just a symbol, then function is added to the global value of place. Whereas if place is of the form (local symbol), where symbol is an expression which returns the variable name, then function will only be added in the current buffer. Finally, if you want to modify a lexical variable, you will have to use (var variable).

Every function added with add-function can be accompanied by an association list of properties props. Currently only two of those properties have a special meaning:

name

This gives a name to the advice, which remove-function can use to identify which function to remove. Typically used when function is an anonymous function.

depth

This specifies how to order the advice, should several pieces of advice be present. By default, the depth is 0. A depth of 100 indicates that this piece of advice should be kept as deep as possible, whereas a depth of −100 indicates that it should stay as the outermost piece. When two pieces of advice specify the same depth, the most recently added one will be outermost.

For :before advice, being outermost means that this advice will be run first, before any other advice, whereas being innermost means that it will run right before the original function, with no other advice run between itself and the original function. Similarly, for :after advice innermost means that it will run right after the original function, with no other advice run in between, whereas outermost means that it will be run right at the end after all other advice. An innermost :override piece of advice will only override the original function and other pieces of advice will apply to it, whereas an outermost :override piece of advice will override not only the original function but all other advice applied to it as well.

If function is not interactive, then the combined function will inherit the interactive spec, if any, of the original function. Else, the combined function will be interactive and will use the interactive spec of function. One exception: if the interactive spec of function is a function (i.e., a lambda expression or an fbound symbol rather than an expression or a string), then the interactive spec of the combined function will be a call to that function with the interactive spec of the original function as sole argument. To interpret the spec received as argument, use advice-eval-interactive-spec.

Note: The interactive spec of function will apply to the combined function and should hence obey the calling convention of the combined function rather than that of function. In many cases, it makes no difference since they are identical, but it does matter for :around, :filter-args, and :filter-return, where function receives different arguments than the original function stored in place.

Macro: remove-function place function

This macro removes function from the function stored in place. This only works if function was added to place using add-function.

function is compared with functions added to place using equal, to try and make it work also with lambda expressions. It is additionally compared also with the name property of the functions added to place, which can be more reliable than comparing lambda expressions using equal.

Function: advice-function-member-p advice function-def

Return non-nil if advice is already in function-def. Like for remove-function above, instead of advice being the actual function, it can also be the name of the piece of advice.

Function: advice-function-mapc f function-def

Call the function f for every piece of advice that was added to function-def. f is called with two arguments: the advice function and its properties.

Function: advice-eval-interactive-spec spec

Evaluate the interactive spec just like an interactive call to a function with such a spec would, and then return the corresponding list of arguments that was built. E.g., (advice-eval-interactive-spec "r\nP") will return a list of three elements, containing the boundaries of the region and the current prefix argument.

For instance, if you want to make the C-x m (compose-mail) command prompt for a ‘From:’ header, you could say something like this:

(defun my-compose-mail-advice (orig &rest args)
  "Read From: address interactively."
  (interactive
   (lambda (spec)
     (let* ((user-mail-address
             (completing-read "From: "
                              '("one.address@example.net"
                                "alternative.address@example.net")))
            (from (message-make-from user-full-name
                                     user-mail-address))
            (spec (advice-eval-interactive-spec spec)))
       ;; Put the From header into the OTHER-HEADERS argument.
       (push (cons 'From from) (nth 2 spec))
       spec)))
  (apply orig args))

(advice-add 'compose-mail :around #'my-compose-mail-advice)

13.12.2 Advising Named Functions

A common use of advice is for named functions and macros. You could just use add-function as in:

(add-function :around (symbol-function 'fun) #'his-tracing-function)

But you should use advice-add and advice-remove for that instead. This separate set of functions to manipulate pieces of advice applied to named functions, offers the following extra features compared to add-function: they know how to deal with macros and autoloaded functions, they let describe-function preserve the original docstring as well as document the added advice, and they let you add and remove advice before a function is even defined.

advice-add can be useful for altering the behavior of existing calls to an existing function without having to redefine the whole function. However, it can be a source of bugs, since existing callers to the function may assume the old behavior, and work incorrectly when the behavior is changed by advice. Advice can also cause confusion in debugging, if the person doing the debugging does not notice or remember that the function has been modified by advice.

For these reasons, advice should be reserved for the cases where you cannot modify a function’s behavior in any other way. If it is possible to do the same thing via a hook, that is preferable (see Hooks). If you simply want to change what a particular key does, it may be better to write a new command, and remap the old command’s key bindings to the new one (see Remapping Commands).

If you are writing code for release, for others to use, try to avoid including advice in it. If the function you want to advise has no hook to do the job, please talk with the Emacs developers about adding a suitable hook. Especially, Emacs’s own source files should not put advice on functions in Emacs. (There are currently a few exceptions to this convention, but we aim to correct them.) It is generally cleaner to create a new hook in foo, and make bar use the hook, than to have bar put advice in foo.

Special forms (see Special Forms) cannot be advised, however macros can be advised, in much the same way as functions. Of course, this will not affect code that has already been macro-expanded, so you need to make sure the advice is installed before the macro is expanded.

It is possible to advise a primitive (see What Is a Function?), but one should typically not do so, for two reasons. Firstly, some primitives are used by the advice mechanism, and advising them could cause an infinite recursion. Secondly, many primitives are called directly from C, and such calls ignore advice; hence, one ends up in a confusing situation where some calls (occurring from Lisp code) obey the advice and other calls (from C code) do not.

Macro: define-advice symbol (where lambda-list &optional name depth) &rest body

This macro defines a piece of advice and adds it to the function named symbol. The advice is an anonymous function if name is nil or a function named symbol@name. See advice-add for explanation of other arguments.

Function: advice-add symbol where function &optional props

Add the advice function to the named function symbol. where and props have the same meaning as for add-function (see Primitives to manipulate advices).

Function: advice-remove symbol function

Remove the advice function from the named function symbol. function can also be the name of a piece of advice.

Function: advice-member-p function symbol

Return non-nil if the advice function is already in the named function symbol. function can also be the name of a piece of advice.

Function: advice-mapc function symbol

Call function for every piece of advice that was added to the named function symbol. function is called with two arguments: the advice function and its properties.


13.12.3 Ways to compose advice

Here are the different possible values for the where argument of add-function and advice-add, specifying how the advice function and the original function should be composed.

:before

Call function before the old function. Both functions receive the same arguments, and the return value of the composition is the return value of the old function. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (apply function r) (apply oldfun r))

(add-function :before funvar function) is comparable for single-function hooks to (add-hook 'hookvar function) for normal hooks.

:after

Call function after the old function. Both functions receive the same arguments, and the return value of the composition is the return value of the old function. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (prog1 (apply oldfun r) (apply function r)))

(add-function :after funvar function) is comparable for single-function hooks to (add-hook 'hookvar function 'append) for normal hooks.

:override

This completely replaces the old function with the new one. The old function can of course be recovered if you later call remove-function.

:around

Call function instead of the old function, but provide the old function as an extra argument to function. This is the most flexible composition. For example, it lets you call the old function with different arguments, or many times, or within a let-binding, or you can sometimes delegate the work to the old function and sometimes override it completely. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (apply function oldfun r))
:before-while

Call function before the old function and don’t call the old function if function returns nil. Both functions receive the same arguments, and the return value of the composition is the return value of the old function. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (and (apply function r) (apply oldfun r)))

(add-function :before-while funvar function) is comparable for single-function hooks to (add-hook 'hookvar function) when hookvar is run via run-hook-with-args-until-failure.

:before-until

Call function before the old function and only call the old function if function returns nil. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (or (apply function r) (apply oldfun r)))

(add-function :before-until funvar function) is comparable for single-function hooks to (add-hook 'hookvar function) when hookvar is run via run-hook-with-args-until-success.

:after-while

Call function after the old function and only if the old function returned non-nil. Both functions receive the same arguments, and the return value of the composition is the return value of function. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (and (apply oldfun r) (apply function r)))

(add-function :after-while funvar function) is comparable for single-function hooks to (add-hook 'hookvar function 'append) when hookvar is run via run-hook-with-args-until-failure.

:after-until

Call function after the old function and only if the old function returned nil. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (or  (apply oldfun r) (apply function r)))

(add-function :after-until funvar function) is comparable for single-function hooks to (add-hook 'hookvar function 'append) when hookvar is run via run-hook-with-args-until-success.

:filter-args

Call function first and use the result (which should be a list) as the new arguments to pass to the old function. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (apply oldfun (funcall function r)))
:filter-return

Call the old function first and pass the result to function. More specifically, the composition of the two functions behaves like:

(lambda (&rest r) (funcall function (apply oldfun r)))

13.12.4 Adapting code using the old defadvice

A lot of code uses the old defadvice mechanism, which is largely made obsolete by the new advice-add, whose implementation and semantics is significantly simpler.

An old piece of advice such as:

(defadvice previous-line (before next-line-at-end
                                 (&optional arg try-vscroll))
  "Insert an empty line when moving up from the top line."
  (if (and next-line-add-newlines (= arg 1)
           (save-excursion (beginning-of-line) (bobp)))
      (progn
        (beginning-of-line)
        (newline))))

could be translated in the new advice mechanism into a plain function:

(defun previous-line--next-line-at-end (&optional arg try-vscroll)
  "Insert an empty line when moving up from the top line."
  (if (and next-line-add-newlines (= arg 1)
           (save-excursion (beginning-of-line) (bobp)))
      (progn
        (beginning-of-line)
        (newline))))

Obviously, this does not actually modify previous-line. For that the old advice needed:

(ad-activate 'previous-line)

whereas the new advice mechanism needs:

(advice-add 'previous-line :before #'previous-line--next-line-at-end)

Note that ad-activate had a global effect: it activated all pieces of advice enabled for that specified function. If you wanted to only activate or deactivate a particular piece, you needed to enable or disable it with ad-enable-advice and ad-disable-advice. The new mechanism does away with this distinction.

Around advice such as:

(defadvice foo (around foo-around)
  "Ignore case in `foo'."
  (let ((case-fold-search t))
    ad-do-it))
(ad-activate 'foo)

could translate into:

(defun foo--foo-around (orig-fun &rest args)
  "Ignore case in `foo'."
  (let ((case-fold-search t))
    (apply orig-fun args)))
(advice-add 'foo :around #'foo--foo-around)

Regarding the advice’s class, note that the new :before is not quite equivalent to the old before, because in the old advice you could modify the function’s arguments (e.g., with ad-set-arg), and that would affect the argument values seen by the original function, whereas in the new :before, modifying an argument via setq in the advice has no effect on the arguments seen by the original function. When porting before advice which relied on this behavior, you’ll need to turn it into new :around or :filter-args advice instead.

Similarly old after advice could modify the returned value by changing ad-return-value, whereas new :after advice cannot, so when porting such old after advice, you’ll need to turn it into new :around or :filter-return advice instead.


13.12.5 Advice and Byte Code

Not all functions can be reliably advised. The byte compiler may choose to replace a call to a function with a sequence of instructions that doesn’t call the function you were interested in altering.

This usually happens due to one of the three following mechanisms:

byte-compile properties

If a function’s symbol has a byte-compile property, that property will be used instead of the symbol’s function definition. See Byte-Compilation Functions.

byte-optimize properties

If a function’s symbol has a byte-optimize property, the byte compiler may rewrite the function arguments, or decide to use a different function altogether.

compiler-macro declare forms

A function can have a special compiler-macro declare form in its definition (see The declare Form) that defines an expander to call when compiling the function. The expander could then cause the produced byte-code not to call the original function.


13.13 Declaring Functions Obsolete

You can mark a named function as obsolete, meaning that it may be removed at some point in the future. This causes Emacs to warn that the function is obsolete whenever it byte-compiles code containing that function, and whenever it displays the documentation for that function. In all other respects, an obsolete function behaves like any other function.

The easiest way to mark a function as obsolete is to put a (declare (obsolete …)) form in the function’s defun definition. See The declare Form. Alternatively, you can use the make-obsolete function, described below.

A macro (see Macros) can also be marked obsolete with make-obsolete; this has the same effects as for a function. An alias for a function or macro can also be marked as obsolete; this makes the alias itself obsolete, not the function or macro which it resolves to.

Function: make-obsolete obsolete-name current-name when

This function marks obsolete-name as obsolete. obsolete-name should be a symbol naming a function or macro, or an alias for a function or macro.

If current-name is a symbol, the warning message says to use current-name instead of obsolete-name. current-name does not need to be an alias for obsolete-name; it can be a different function with similar functionality. current-name can also be a string, which serves as the warning message. The message should begin in lower case, and end with a period. It can also be nil, in which case the warning message provides no additional details.

The argument when should be a string indicating when the function was first made obsolete—for example, a date or a release number.

Macro: define-obsolete-function-alias obsolete-name current-name when &optional doc

This convenience macro marks the function obsolete-name obsolete and also defines it as an alias for the function current-name. It is equivalent to the following:

(defalias obsolete-name current-name doc)
(make-obsolete obsolete-name current-name when)

In addition, you can mark a particular calling convention for a function as obsolete:

Function: set-advertised-calling-convention function signature when

This function specifies the argument list signature as the correct way to call function. This causes the Emacs byte compiler to issue a warning whenever it comes across an Emacs Lisp program that calls function any other way (however, it will still allow the code to be byte compiled). when should be a string indicating when the variable was first made obsolete (usually a version number string).

For instance, in old versions of Emacs the sit-for function accepted three arguments, like this

  (sit-for seconds milliseconds nodisp)

However, calling sit-for this way is considered obsolete (see Waiting for Elapsed Time or Input). The old calling convention is deprecated like this:

(set-advertised-calling-convention
  'sit-for '(seconds &optional nodisp) "22.1")

The alternative to using this function is the advertised-calling-convention declare spec, see The declare Form.


13.14 Inline Functions

An inline function is a function that works just like an ordinary function, except for one thing: when you byte-compile a call to the function (see Byte Compilation), the function’s definition is expanded into the caller.

The simple way to define an inline function, is to write defsubst instead of defun. The rest of the definition looks just the same, but using defsubst says to make it inline for byte compilation.

Macro: defsubst name args [doc] [declare] [interactive] body…

This macro defines an inline function. Its syntax is exactly the same as defun (see Defining Functions).

Making a function inline often makes its function calls run faster. But it also has disadvantages. For one thing, it reduces flexibility; if you change the definition of the function, calls already inlined still use the old definition until you recompile them.

Another disadvantage is that making a large function inline can increase the size of compiled code both in files and in memory. Since the speed advantage of inline functions is greatest for small functions, you generally should not make large functions inline.

Also, inline functions do not behave well with respect to debugging, tracing, and advising (see Advising Emacs Lisp Functions). Since ease of debugging and the flexibility of redefining functions are important features of Emacs, you should not make a function inline, even if it’s small, unless its speed is really crucial, and you’ve timed the code to verify that using defun actually has performance problems.

After an inline function is defined, its inline expansion can be performed later on in the same file, just like macros.

It’s possible to use defmacro to define a macro to expand into the same code that an inline function would execute (see Macros). But the macro would be limited to direct use in expressions—a macro cannot be called with apply, mapcar and so on. Also, it takes some work to convert an ordinary function into a macro. To convert it into an inline function is easy; just replace defun with defsubst. Since each argument of an inline function is evaluated exactly once, you needn’t worry about how many times the body uses the arguments, as you do for macros.

Alternatively, you can define a function by providing the code which will inline it as a compiler macro (see The declare Form). The following macros make this possible.

Macro: define-inline name args [doc] [declare] body…

Define a function name by providing code that does its inlining, as a compiler macro. The function will accept the argument list args and will have the specified body.

If present, doc should be the function’s documentation string (see Documentation Strings of Functions); declare, if present, should be a declare form (see The declare Form) specifying the function’s metadata.

Functions defined via define-inline have several advantages with respect to macros defined by defsubst or defmacro:

  • They can be passed to mapcar (see Mapping Functions).
  • They are more efficient.
  • They can be used as place forms to store values (see Generalized Variables).
  • They behave in a more predictable way than cl-defsubst (see Argument Lists in Common Lisp Extensions for GNU Emacs Lisp).

Like defmacro, a function inlined with define-inline inherits the scoping rules, either dynamic or lexical, from the call site. See Scoping Rules for Variable Bindings.

The following macros should be used in the body of a function defined by define-inline.

Macro: inline-quote expression

Quote expression for define-inline. This is similar to the backquote (see Backquote), but quotes code and accepts only ,, not ,@.

Macro: inline-letevals (bindings…) body…

This provides a convenient way to ensure that the arguments to an inlined function are evaluated exactly once, as well as to create local variables.

It’s similar to let (see Local Variables): It sets up local variables as specified by bindings, and then evaluates body with those bindings in effect.

Each element of bindings should be either a symbol or a list of the form (var expr); the result is to evaluate expr and bind var to the result. However, when an element of bindings is just a symbol var, the result of evaluating var is re-bound to var (which is quite different from the way let works).

The tail of bindings can be either nil or a symbol which should hold a list of arguments, in which case each argument is evaluated, and the symbol is bound to the resulting list.

Macro: inline-const-p expression

Return non-nil if the value of expression is already known.

Macro: inline-const-val expression

Return the value of expression.

Macro: inline-error format &rest args

Signal an error, formatting args according to format.

Here’s an example of using define-inline:

(define-inline myaccessor (obj)
  (inline-letevals (obj)
    (inline-quote (if (foo-p ,obj) (aref (cdr ,obj) 3) (aref ,obj 2)))))

This is equivalent to

(defsubst myaccessor (obj)
  (if (foo-p obj) (aref (cdr obj) 3) (aref obj 2)))

13.15 The declare Form

declare is a special macro which can be used to add meta properties to a function or macro: for example, marking it as obsolete, or giving its forms a special TAB indentation convention in Emacs Lisp mode.

Macro: declare specs…

This macro ignores its arguments and evaluates to nil; it has no run-time effect. However, when a declare form occurs in the declare argument of a defun or defsubst function definition (see Defining Functions) or a defmacro macro definition (see Defining Macros), it appends the properties specified by specs to the function or macro. This work is specially performed by defun, defsubst, and defmacro.

Each element in specs should have the form (property args…), which should not be quoted. These have the following effects:

(advertised-calling-convention signature when)

This acts like a call to set-advertised-calling-convention (see Declaring Functions Obsolete); signature specifies the correct argument list for calling the function or macro, and when should be a string indicating when the old argument list was first made obsolete.

(debug edebug-form-spec)

This is valid for macros only. When stepping through the macro with Edebug, use edebug-form-spec. See Instrumenting Macro Calls.

(doc-string n)

This is used when defining a function or macro which itself will be used to define entities like functions, macros, or variables. It indicates that the nth argument, if any, should be considered as a documentation string.

(indent indent-spec)

Indent calls to this function or macro according to indent-spec. This is typically used for macros, though it works for functions too. See Indenting Macros.

(interactive-only value)

Set the function’s interactive-only property to value. See The interactive-only property.

(obsolete current-name when)

Mark the function or macro as obsolete, similar to a call to make-obsolete (see Declaring Functions Obsolete). current-name should be a symbol (in which case the warning message says to use that instead), a string (specifying the warning message), or nil (in which case the warning message gives no extra details). when should be a string indicating when the function or macro was first made obsolete.

(compiler-macro expander)

This can only be used for functions, and tells the compiler to use expander as an optimization function. When encountering a call to the function, of the form (function args…), the macro expander will call expander with that form as well as with args…, and expander can either return a new expression to use instead of the function call, or it can return just the form unchanged, to indicate that the function call should be left alone.

When expander is a lambda form it should be written with a single argument (i.e., be of the form (lambda (arg) body)) because the function’s formal arguments are automatically added to the lambda’s list of arguments for you.

(gv-expander expander)

Declare expander to be the function to handle calls to the macro (or function) as a generalized variable, similarly to gv-define-expander. expander can be a symbol or it can be of the form (lambda (arg) body) in which case that function will additionally have access to the macro (or function)’s arguments.

(gv-setter setter)

Declare setter to be the function to handle calls to the macro (or function) as a generalized variable. setter can be a symbol in which case it will be passed to gv-define-simple-setter, or it can be of the form (lambda (arg) body) in which case that function will additionally have access to the macro (or function)’s arguments and it will be passed to gv-define-setter.

(completion completion-predicate)

Declare completion-predicate as a function to determine whether to include a function’s symbol in the list of functions when asking for completions in M-x. This predicate function will only be called when read-extended-command-predicate is customized to command-completion-default-include-p; by default the value of read-extended-command-predicate is nil (see execute-extended-command). The predicate completion-predicate is called with two arguments: the function’s symbol and the current buffer.

(modes modes)

Specify that this command is meant to be applicable only to specified modes. See Specifying Modes For Commands.

(interactive-args arg ...)

Specify the arguments that should be stored for repeat-command. Each arg is on the form argument-name form.

(pure val)

If val is non-nil, this function is pure (see What Is a Function?). This is the same as the pure property of the function’s symbol (see Standard Symbol Properties).

(side-effect-free val)

If val is non-nil, this function is free of side effects, so the byte compiler can ignore calls whose value is ignored. This is the same as the side-effect-free property of the function’s symbol, see Standard Symbol Properties.

(speed n)

Specify the value of native-comp-speed in effect for native compilation of this function (see Native-Compilation Variables). This allows function-level control of the optimization level used for native code emitted for the function. In particular, if n is −1, native compilation of the function will emit bytecode instead of native code for the function.

no-font-lock-keyword

This is valid for macros only. Macros with this declaration are highlighted by font-lock (see Font Lock Mode) as normal functions, not specially as macros.


13.16 Telling the Compiler that a Function is Defined

Byte-compiling a file often produces warnings about functions that the compiler doesn’t know about (see Compiler Errors). Sometimes this indicates a real problem, but usually the functions in question are defined in other files which would be loaded if that code is run. For example, byte-compiling simple.el used to warn:

simple.el:8727:1:Warning: the function ‘shell-mode’ is not known to be
    defined.

In fact, shell-mode is used only in a function that executes (require 'shell) before calling shell-mode, so shell-mode will be defined properly at run-time. When you know that such a warning does not indicate a real problem, it is good to suppress the warning. That makes new warnings which might mean real problems more visible. You do that with declare-function.

All you need to do is add a declare-function statement before the first use of the function in question:

(declare-function shell-mode "shell" ())

This says that shell-mode is defined in shell.el (the ‘.el’ can be omitted). The compiler takes for granted that that file really defines the function, and does not check.

The optional third argument specifies the argument list of shell-mode. In this case, it takes no arguments (nil is different from not specifying a value). In other cases, this might be something like (file &optional overwrite). You don’t have to specify the argument list, but if you do the byte compiler can check that the calls match the declaration.

Macro: declare-function function file &optional arglist fileonly

Tell the byte compiler to assume that function is defined in the file file. The optional third argument arglist is either t, meaning the argument list is unspecified, or a list of formal parameters in the same style as defun (including the parentheses). An omitted arglist defaults to t, not nil; this is atypical behavior for omitted arguments, and it means that to supply a fourth but not third argument one must specify t for the third-argument placeholder instead of the usual nil. The optional fourth argument fileonly non-nil means check only that file exists, not that it actually defines function.

To verify that these functions really are declared where declare-function says they are, use check-declare-file to check all declare-function calls in one source file, or use check-declare-directory check all the files in and under a certain directory.

These commands find the file that ought to contain a function’s definition using locate-library; if that finds no file, they expand the definition file name relative to the directory of the file that contains the declare-function call.

You can also say that a function is a primitive by specifying a file name ending in ‘.c’ or ‘.m’. This is useful only when you call a primitive that is defined only on certain systems. Most primitives are always defined, so they will never give you a warning.

Sometimes a file will optionally use functions from an external package. If you prefix the filename in the declare-function statement with ‘ext:’, then it will be checked if it is found, otherwise skipped without error.

There are some function definitions that ‘check-declare’ does not understand (e.g., defstruct and some other macros). In such cases, you can pass a non-nil fileonly argument to declare-function, meaning to only check that the file exists, not that it actually defines the function. Note that to do this without having to specify an argument list, you should set the arglist argument to t (because nil means an empty argument list, as opposed to an unspecified one).


13.17 Determining whether a Function is Safe to Call

Some major modes, such as SES, call functions that are stored in user files. (See (ses)Simple Emacs Spreadsheet, for more information on SES.) User files sometimes have poor pedigrees—you can get a spreadsheet from someone you’ve just met, or you can get one through email from someone you’ve never met. So it is risky to call a function whose source code is stored in a user file until you have determined that it is safe.

Function: unsafep form &optional unsafep-vars

Returns nil if form is a safe Lisp expression, or returns a list that describes why it might be unsafe. The argument unsafep-vars is a list of symbols known to have temporary bindings at this point; it is mainly used for internal recursive calls. The current buffer is an implicit argument, which provides a list of buffer-local bindings.

Being quick and simple, unsafep does a very light analysis and rejects many Lisp expressions that are actually safe. There are no known cases where unsafep returns nil for an unsafe expression. However, a safe Lisp expression can return a string with a display property, containing an associated Lisp expression to be executed after the string is inserted into a buffer. This associated expression can be a virus. In order to be safe, you must delete properties from all strings calculated by user code before inserting them into buffers.


14 Macros

Macros enable you to define new control constructs and other language features. A macro is defined much like a function, but instead of telling how to compute a value, it tells how to compute another Lisp expression which will in turn compute the value. We call this expression the expansion of the macro.

Macros can do this because they operate on the unevaluated expressions for the arguments, not on the argument values as functions do. They can therefore construct an expansion containing these argument expressions or parts of them.

If you are using a macro to do something an ordinary function could do, just for the sake of speed, consider using an inline function instead. See Inline Functions.


14.1 A Simple Example of a Macro

Suppose we would like to define a Lisp construct to increment a variable value, much like the ++ operator in C. We would like to write (inc x) and have the effect of (setq x (1+ x)). Here’s a macro definition that does the job:

(defmacro inc (var)
   (list 'setq var (list '1+ var)))

When this is called with (inc x), the argument var is the symbol xnot the value of x, as it would be in a function. The body of the macro uses this to construct the expansion, which is (setq x (1+ x)). Once the macro definition returns this expansion, Lisp proceeds to evaluate it, thus incrementing x.

Function: macrop object

This predicate tests whether its argument is a macro, and returns t if so, nil otherwise.


14.2 Expansion of a Macro Call

A macro call looks just like a function call in that it is a list which starts with the name of the macro. The rest of the elements of the list are the arguments of the macro.

Evaluation of the macro call begins like evaluation of a function call except for one crucial difference: the macro arguments are the actual expressions appearing in the macro call. They are not evaluated before they are given to the macro definition. By contrast, the arguments of a function are results of evaluating the elements of the function call list.

Having obtained the arguments, Lisp invokes the macro definition just as a function is invoked. The argument variables of the macro are bound to the argument values from the macro call, or to a list of them in the case of a &rest argument. And the macro body executes and returns its value just as a function body does.

The second crucial difference between macros and functions is that the value returned by the macro body is an alternate Lisp expression, also known as the expansion of the macro. The Lisp interpreter proceeds to evaluate the expansion as soon as it comes back from the macro.

Since the expansion is evaluated in the normal manner, it may contain calls to other macros. It may even be a call to the same macro, though this is unusual.

Note that Emacs tries to expand macros when loading an uncompiled Lisp file. This is not always possible, but if it is, it speeds up subsequent execution. See How Programs Do Loading.

You can see the expansion of a given macro call by calling macroexpand.

Function: macroexpand form &optional environment

This function expands form, if it is a macro call. If the result is another macro call, it is expanded in turn, until something which is not a macro call results. That is the value returned by macroexpand. If form is not a macro call to begin with, it is returned as given.

Note that macroexpand does not look at the subexpressions of form (although some macro definitions may do so). Even if they are macro calls themselves, macroexpand does not expand them.

The function macroexpand does not expand calls to inline functions. Normally there is no need for that, since a call to an inline function is no harder to understand than a call to an ordinary function.

If environment is provided, it specifies an alist of macro definitions that shadow the currently defined macros. Byte compilation uses this feature.

(defmacro inc (var)
    (list 'setq var (list '1+ var)))

(macroexpand '(inc r))
     ⇒ (setq r (1+ r))

(defmacro inc2 (var1 var2)
    (list 'progn (list 'inc var1) (list 'inc var2)))

(macroexpand '(inc2 r s))
     ⇒ (progn (inc r) (inc s))  ; inc not expanded here.
Function: macroexpand-all form &optional environment

macroexpand-all expands macros like macroexpand, but will look for and expand all macros in form, not just at the top-level. If no macros are expanded, the return value is eq to form.

Repeating the example used for macroexpand above with macroexpand-all, we see that macroexpand-all does expand the embedded calls to inc:

(macroexpand-all '(inc2 r s))
     ⇒ (progn (setq r (1+ r)) (setq s (1+ s)))
Function: macroexpand-1 form &optional environment

This function expands macros like macroexpand, but it only performs one step of the expansion: if the result is another macro call, macroexpand-1 will not expand it.


14.3 Macros and Byte Compilation

You might ask why we take the trouble to compute an expansion for a macro and then evaluate the expansion. Why not have the macro body produce the desired results directly? The reason has to do with compilation.

When a macro call appears in a Lisp program being compiled, the Lisp compiler calls the macro definition just as the interpreter would, and receives an expansion. But instead of evaluating this expansion, it compiles the expansion as if it had appeared directly in the program. As a result, the compiled code produces the value and side effects intended for the macro, but executes at full compiled speed. This would not work if the macro body computed the value and side effects itself—they would be computed at compile time, which is not useful.

In order for compilation of macro calls to work, the macros must already be defined in Lisp when the calls to them are compiled. The compiler has a special feature to help you do this: if a file being compiled contains a defmacro form, the macro is defined temporarily for the rest of the compilation of that file.

Byte-compiling a file also executes any require calls at top-level in the file, so you can ensure that necessary macro definitions are available during compilation by requiring the files that define them (see Features). To avoid loading the macro definition files when someone runs the compiled program, write eval-when-compile around the require calls (see Evaluation During Compilation).


14.4 Defining Macros

A Lisp macro object is a list whose CAR is macro, and whose CDR is a function. Expansion of the macro works by applying the function (with apply) to the list of unevaluated arguments from the macro call.

It is possible to use an anonymous Lisp macro just like an anonymous function, but this is never done, because it does not make sense to pass an anonymous macro to functionals such as mapcar. In practice, all Lisp macros have names, and they are almost always defined with the defmacro macro.

Macro: defmacro name args [doc] [declare] body…

defmacro defines the symbol name (which should not be quoted) as a macro that looks like this:

(macro lambda args . body)

(Note that the CDR of this list is a lambda expression.) This macro object is stored in the function cell of name. The meaning of args is the same as in a function, and the keywords &rest and &optional may be used (see Features of Argument Lists). Neither name nor args should be quoted. The return value of defmacro is undefined.

doc, if present, should be a string specifying the macro’s documentation string. declare, if present, should be a declare form specifying metadata for the macro (see The declare Form). Note that macros cannot have interactive declarations, since they cannot be called interactively.

Macros often need to construct large list structures from a mixture of constants and nonconstant parts. To make this easier, use the ‘`’ syntax (see Backquote). For example:

(defmacro t-becomes-nil (variable)
  `(if (eq ,variable t)
       (setq ,variable nil)))

(t-becomes-nil foo)
     ≡ (if (eq foo t) (setq foo nil))

14.5 Common Problems Using Macros

Macro expansion can have counterintuitive consequences. This section describes some important consequences that can lead to trouble, and rules to follow to avoid trouble.


14.5.1 Wrong Time

The most common problem in writing macros is doing some of the real work prematurely—while expanding the macro, rather than in the expansion itself. For instance, one real package had this macro definition:

(defmacro my-set-buffer-multibyte (arg)
  (if (fboundp 'set-buffer-multibyte)
      (set-buffer-multibyte arg)))

With this erroneous macro definition, the program worked fine when interpreted but failed when compiled. This macro definition called set-buffer-multibyte during compilation, which was wrong, and then did nothing when the compiled package was run. The definition that the programmer really wanted was this:

(defmacro my-set-buffer-multibyte (arg)
  (if (fboundp 'set-buffer-multibyte)
      `(set-buffer-multibyte ,arg)))

This macro expands, if appropriate, into a call to set-buffer-multibyte that will be executed when the compiled program is actually run.


14.5.2 Evaluating Macro Arguments Repeatedly

When defining a macro you must pay attention to the number of times the arguments will be evaluated when the expansion is executed. The following macro (used to facilitate iteration) illustrates the problem. This macro allows us to write a for-loop construct.

(defmacro for (var from init to final do &rest body)
  "Execute a simple \"for\" loop.
For example, (for i from 1 to 10 do (print i))."
  (list 'let (list (list var init))
        (cons 'while
              (cons (list '<= var final)
                    (append body (list (list 'inc var)))))))

(for i from 1 to 3 do
   (setq square (* i i))
   (princ (format "\n%d %d" i square)))
→
(let ((i 1))
  (while (<= i 3)
    (setq square (* i i))
    (princ (format "\n%d %d" i square))
    (inc i)))

     -|1       1
     -|2       4
     -|3       9
⇒ nil

The arguments from, to, and do in this macro are syntactic sugar; they are entirely ignored. The idea is that you will write noise words (such as from, to, and do) in those positions in the macro call.

Here’s an equivalent definition simplified through use of backquote:

(defmacro for (var from init to final do &rest body)
  "Execute a simple \"for\" loop.
For example, (for i from 1 to 10 do (print i))."
  `(let ((,var ,init))
     (while (<= ,var ,final)
       ,@body
       (inc ,var))))

Both forms of this definition (with backquote and without) suffer from the defect that final is evaluated on every iteration. If final is a constant, this is not a problem. If it is a more complex form, say (long-complex-calculation x), this can slow down the execution significantly. If final has side effects, executing it more than once is probably incorrect.

A well-designed macro definition takes steps to avoid this problem by producing an expansion that evaluates the argument expressions exactly once unless repeated evaluation is part of the intended purpose of the macro. Here is a correct expansion for the for macro:

(let ((i 1)
      (max 3))
  (while (<= i max)
    (setq square (* i i))
    (princ (format "%d      %d" i square))
    (inc i)))

Here is a macro definition that creates this expansion:

(defmacro for (var from init to final do &rest body)
  "Execute a simple for loop: (for i from 1 to 10 do (print i))."
  `(let ((,var ,init)
         (max ,final))
     (while (<= ,var max)
       ,@body
       (inc ,var))))

Unfortunately, this fix introduces another problem, described in the following section.


14.5.3 Local Variables in Macro Expansions

In the previous section, the definition of for was fixed as follows to make the expansion evaluate the macro arguments the proper number of times:

(defmacro for (var from init to final do &rest body)
  "Execute a simple for loop: (for i from 1 to 10 do (print i))."
  `(let ((,var ,init)
         (max ,final))
     (while (<= ,var max)
       ,@body
       (inc ,var))))

The new definition of for has a new problem: it introduces a local variable named max which the user does not expect. This causes trouble in examples such as the following:

(let ((max 0))
  (for x from 0 to 10 do
    (let ((this (frob x)))
      (if (< max this)
          (setq max this)))))

The references to max inside the body of the for, which are supposed to refer to the user’s binding of max, really access the binding made by for.

The way to correct this is to use an uninterned symbol instead of max (see Creating and Interning Symbols). The uninterned symbol can be bound and referred to just like any other symbol, but since it is created by for, we know that it cannot already appear in the user’s program. Since it is not interned, there is no way the user can put it into the program later. It will never appear anywhere except where put by for. Here is a definition of for that works this way:

(defmacro for (var from init to final do &rest body)
  "Execute a simple for loop: (for i from 1 to 10 do (print i))."
  (let ((tempvar (make-symbol "max")))
    `(let ((,var ,init)
           (,tempvar ,final))
       (while (<= ,var ,tempvar)
         ,@body
         (inc ,var)))))

This creates an uninterned symbol named max and puts it in the expansion instead of the usual interned symbol max that appears in expressions ordinarily.


14.5.4 Evaluating Macro Arguments in Expansion

Another problem can happen if the macro definition itself evaluates any of the macro argument expressions, such as by calling eval (see Eval). You have to take into account that macro expansion may take place long before the code is executed, when the context of the caller (where the macro expansion will be evaluated) is not yet accessible.

Also, if your macro definition does not use lexical-binding, its formal arguments may hide the user’s variables of the same name. Inside the macro body, the macro argument binding is the most local binding of such variable, so any references inside the form being evaluated do refer to it. Here is an example:

(defmacro foo (a)
  (list 'setq (eval a) t))
(setq x 'b)
(foo x) → (setq b t)
     ⇒ t                  ; and b has been set.
;; but
(setq a 'c)
(foo a) → (setq a t)
     ⇒ t                  ; but this set a, not c.

It makes a difference whether the user’s variable is named a or x, because a conflicts with the macro argument variable a.

Also, the expansion of (foo x) above will return something different or signal an error when the code is compiled, since in that case (foo x) is expanded during compilation, whereas the execution of (setq x 'b) will only take place later when the code is executed.

To avoid these problems, don’t evaluate an argument expression while computing the macro expansion. Instead, substitute the expression into the macro expansion, so that its value will be computed as part of executing the expansion. This is how the other examples in this chapter work.


14.5.5 How Many Times is the Macro Expanded?

Occasionally problems result from the fact that a macro call is expanded each time it is evaluated in an interpreted function, but is expanded only once (during compilation) for a compiled function. If the macro definition has side effects, they will work differently depending on how many times the macro is expanded.

Therefore, you should avoid side effects in computation of the macro expansion, unless you really know what you are doing.

One special kind of side effect can’t be avoided: constructing Lisp objects. Almost all macro expansions include constructed lists; that is the whole point of most macros. This is usually safe; there is just one case where you must be careful: when the object you construct is part of a quoted constant in the macro expansion.

If the macro is expanded just once, in compilation, then the object is constructed just once, during compilation. But in interpreted execution, the macro is expanded each time the macro call runs, and this means a new object is constructed each time.

In most clean Lisp code, this difference won’t matter. It can matter only if you perform side-effects on the objects constructed by the macro definition. Thus, to avoid trouble, avoid side effects on objects constructed by macro definitions. Here is an example of how such side effects can get you into trouble:

(defmacro empty-object ()
  (list 'quote (cons nil nil)))

(defun initialize (condition)
  (let ((object (empty-object)))
    (if condition
        (setcar object condition))
    object))

If initialize is interpreted, a new list (nil) is constructed each time initialize is called. Thus, no side effect survives between calls. If initialize is compiled, then the macro empty-object is expanded during compilation, producing a single constant (nil) that is reused and altered each time initialize is called.

One way to avoid pathological cases like this is to think of empty-object as a funny kind of constant, not as a memory allocation construct. You wouldn’t use setcar on a constant such as '(nil), so naturally you won’t use it on (empty-object) either.


14.6 Indenting Macros

Within a macro definition, you can use the declare form (see Defining Macros) to specify how TAB should indent calls to the macro. An indentation specification is written like this:

(declare (indent indent-spec))

This results in the lisp-indent-function property being set on the macro name.

Here are the possibilities for indent-spec:

nil

This is the same as no property—use the standard indentation pattern.

defun

Handle this function like a ‘def’ construct: treat the second line as the start of a body.

an integer, number

The first number arguments of the function are distinguished arguments; the rest are considered the body of the expression. A line in the expression is indented according to whether the first argument on it is distinguished or not. If the argument is part of the body, the line is indented lisp-body-indent more columns than the open-parenthesis starting the containing expression. If the argument is distinguished and is either the first or second argument, it is indented twice that many extra columns. If the argument is distinguished and not the first or second argument, the line uses the standard pattern.

a symbol, symbol

symbol should be a function name; that function is called to calculate the indentation of a line within this expression. The function receives two arguments:

pos

The position at which the line being indented begins.

state

The value returned by parse-partial-sexp (a Lisp primitive for indentation and nesting computation) when it parses up to the beginning of this line.

It should return either a number, which is the number of columns of indentation for that line, or a list whose car is such a number. The difference between returning a number and returning a list is that a number says that all following lines at the same nesting level should be indented just like this one; a list says that following lines might call for different indentations. This makes a difference when the indentation is being computed by C-M-q; if the value is a number, C-M-q need not recalculate indentation for the following lines until the end of the list.


15 Customization Settings

Users of Emacs can customize variables and faces without writing Lisp code, by using the Customize interface. See Easy Customization in The GNU Emacs Manual. This chapter describes how to define customization items that users can interact with through the Customize interface.

Customization items include customizable variables, which are defined with the defcustom macro; customizable faces, which are defined with defface (described separately in Defining Faces); and customization groups, defined with defgroup, which act as containers for groups of related customization items.


15.1 Common Item Keywords

The customization declarations that we will describe in the next few sections—defcustom, defgroup, etc.—all accept keyword arguments (see Variables that Never Change) for specifying various information. This section describes keywords that apply to all types of customization declarations.

All of these keywords, except :tag, can be used more than once in a given item. Each use of the keyword has an independent effect. The keyword :tag is an exception because any given item can only display one name.

:tag label

Use label, a string, instead of the item’s name, to label the item in customization menus and buffers. Don’t use a tag which is substantially different from the item’s real name; that would cause confusion.

:group group

Put this customization item in group group. If this keyword is missing from a customization item, it’ll be placed in the same group that was last defined (in the current file).

When you use :group in a defgroup, it makes the new group a subgroup of group.

If you use this keyword more than once, you can put a single item into more than one group. Displaying any of those groups will show this item. Please don’t overdo this, since the result would be annoying.

Include an external link after the documentation string for this item. This is a sentence containing a button that references some other documentation.

There are several alternatives you can use for link-data:

(custom-manual info-node)

Link to an Info node; info-node is a string which specifies the node name, as in "(emacs)Top". The link appears as ‘[Manual]’ in the customization buffer and enters the built-in Info reader on info-node.

(info-link info-node)

Like custom-manual except that the link appears in the customization buffer with the Info node name.

(url-link url)

Link to a web page; url is a string which specifies the URL. The link appears in the customization buffer as url and invokes the WWW browser specified by browse-url-browser-function.

(emacs-commentary-link library)

Link to the commentary section of a library; library is a string which specifies the library name. See Conventional Headers for Emacs Libraries.

(emacs-library-link library)

Link to an Emacs Lisp library file; library is a string which specifies the library name.

(file-link file)

Link to a file; file is a string which specifies the name of the file to visit with find-file when the user invokes this link.

(function-link function)

Link to the documentation of a function; function is a string which specifies the name of the function to describe with describe-function when the user invokes this link.

(variable-link variable)

Link to the documentation of a variable; variable is a string which specifies the name of the variable to describe with describe-variable when the user invokes this link.

(face-link face)

Link to the documentation of a face; face is a string which specifies the name of the face to describe with describe-face when the user invokes this link.

(custom-group-link group)

Link to another customization group. Invoking it creates a new customization buffer for group.

You can specify the text to use in the customization buffer by adding :tag name after the first element of the link-data; for example, (info-link :tag "foo" "(emacs)Top") makes a link to the Emacs manual which appears in the buffer as ‘foo’.

You can use this keyword more than once, to add multiple links.

:load file

Load file file (a string) before displaying this customization item (see Loading). Loading is done with load, and only if the file is not already loaded.

:require feature

Execute (require 'feature) when your saved customizations set the value of this item. feature should be a symbol.

The most common reason to use :require is when a variable enables a feature such as a minor mode, and just setting the variable won’t have any effect unless the code which implements the mode is loaded.

:version version

This keyword specifies that the item was first introduced in Emacs version version, or that its default value was changed in that version. The value version must be a string.

:package-version '(package . version)

This keyword specifies that the item was first introduced in package version version, or that its meaning or default value was changed in that version. This keyword takes priority over :version.

package should be the official name of the package, as a symbol (e.g., MH-E). version should be a string. If the package package is released as part of Emacs, package and version should appear in the value of customize-package-emacs-version-alist.

Packages distributed as part of Emacs that use the :package-version keyword must also update the customize-package-emacs-version-alist variable.

Variable: customize-package-emacs-version-alist

This alist provides a mapping for the versions of Emacs that are associated with versions of a package listed in the :package-version keyword. Its elements are:

(package (pversion . eversion)…)

For each package, which is a symbol, there are one or more elements that contain a package version pversion with an associated Emacs version eversion. These versions are strings. For example, the MH-E package updates this alist with the following:

(add-to-list 'customize-package-emacs-version-alist
             '(MH-E ("6.0" . "22.1") ("6.1" . "22.1") ("7.0" . "22.1")
                    ("7.1" . "22.1") ("7.2" . "22.1") ("7.3" . "22.1")
                    ("7.4" . "22.1") ("8.0" . "22.1")))

The value of package needs to be unique and it needs to match the package value appearing in the :package-version keyword. Since the user might see the value in an error message, a good choice is the official name of the package, such as MH-E or Gnus.


15.2 Defining Customization Groups

Each Emacs Lisp package should have one main customization group which contains all the options, faces and other groups in the package. If the package has a small number of options and faces, use just one group and put everything in it. When there are more than twenty or so options and faces, then you should structure them into subgroups, and put the subgroups under the package’s main customization group. It is OK to put some of the options and faces in the package’s main group alongside the subgroups.

The package’s main or only group should be a member of one or more of the standard customization groups. (To display the full list of them, use M-x customize.) Choose one or more of them (but not too many), and add your group to each of them using the :group keyword.

The way to declare new customization groups is with defgroup.

Macro: defgroup group members doc [keyword value]…

Declare group as a customization group containing members. Do not quote the symbol group. The argument doc specifies the documentation string for the group.

The argument members is a list specifying an initial set of customization items to be members of the group. However, most often members is nil, and you specify the group’s members by using the :group keyword when defining those members.

If you want to specify group members through members, each element should have the form (name widget). Here name is a symbol, and widget is a widget type for editing that symbol. Useful widgets are custom-variable for a variable, custom-face for a face, and custom-group for a group.

When you introduce a new group into Emacs, use the :version keyword in the defgroup; then you need not use it for the individual members of the group.

In addition to the common keywords (see Common Item Keywords), you can also use this keyword in defgroup:

:prefix prefix

If the name of an item in the group starts with prefix, and the customizable variable custom-unlispify-remove-prefixes is non-nil, the item’s tag will omit prefix. A group can have any number of prefixes.

The variables, faces, and subgroups of a group are stored in the custom-group property of the group’s symbol. See Accessing Symbol Properties. The value of that property is a list of pairs whose car is the symbol of the variable or the face or the subgroup, and the cdr is one of the corresponding symbols custom-variable, custom-face, or custom-group.

User Option: custom-unlispify-remove-prefixes

If this variable is non-nil, the prefixes specified by a group’s :prefix keyword are omitted from tag names, whenever the user customizes the group.

The default value is nil, i.e., the prefix-discarding feature is disabled. This is because discarding prefixes often leads to confusing names for options and faces.


15.3 Defining Customization Variables

Customizable variables, also called user options, are global Lisp variables whose values can be set through the Customize interface. Unlike other global variables, which are defined with defvar (see Defining Global Variables), customizable variables are defined using the defcustom macro. In addition to calling defvar as a subroutine, defcustom states how the variable should be displayed in the Customize interface, the values it is allowed to take, etc.

Macro: defcustom option standard doc [keyword value]…

This macro declares option as a user option (i.e., a customizable variable). You should not quote option.

The argument standard is an expression that specifies the standard value for option. Evaluating the defcustom form evaluates standard, but does not necessarily bind the option to that value. If option already has a default value, it is left unchanged. If the user has already saved a customization for option, the user’s customized value is installed as the default value. Otherwise, the result of evaluating standard is installed as the default value.

Like defvar, this macro marks option as a special variable, meaning that it should always be dynamically bound. If option is already lexically bound, that lexical binding remains in effect until the binding construct exits. See Scoping Rules for Variable Bindings.

The expression standard can be evaluated at various other times, too—whenever the customization facility needs to know option’s standard value. So be sure to use an expression which is harmless to evaluate at any time.

The argument doc specifies the documentation string for the variable.

If a defcustom does not specify any :group, the last group defined with defgroup in the same file will be used. This way, most defcustom do not need an explicit :group.

When you evaluate a defcustom form with C-M-x in Emacs Lisp mode (eval-defun), a special feature of eval-defun arranges to set the variable unconditionally, without testing whether its value is void. (The same feature applies to defvar, see Defining Global Variables.) Using eval-defun on a defcustom that is already defined calls the :set function (see below), if there is one.

If you put a defcustom in a pre-loaded Emacs Lisp file (see Building Emacs), the standard value installed at dump time might be incorrect, e.g., because another variable that it depends on has not been assigned the right value yet. In that case, use custom-reevaluate-setting, described below, to re-evaluate the standard value after Emacs starts up.

In addition to the keywords listed in Common Item Keywords, this macro accepts the following keywords:

:type type

Use type as the data type for this option. It specifies which values are legitimate, and how to display the value (see Customization Types). Every defcustom should specify a value for this keyword.

:options value-list

Specify the list of reasonable values for use in this option. The user is not restricted to using only these values, but they are offered as convenient alternatives.

This is meaningful only for certain types, currently including hook, plist and alist. See the definition of the individual types for a description of how to use :options.

Re-evaluating a defcustom form with a different :options value does not clear the values added by previous evaluations, or added by calls to custom-add-frequent-value (see below).

:set setfunction

Specify setfunction as the way to change the value of this option when using the Customize interface. The function setfunction should take two arguments, a symbol (the option name) and the new value, and should do whatever is necessary to update the value properly for this option (which may not mean simply setting the option as a Lisp variable); preferably, though, it should not modify its value argument destructively. The default for setfunction is set-default-toplevel-value.

If defined, setfunction will also be called when evaluating a defcustom form with C-M-x in Emacs Lisp mode and when the option’s value is changed via the setopt macro (see setopt).

If you specify this keyword, the variable’s documentation string should describe how to do the same job in hand-written Lisp code, either by invoking setfunction directly or by using setopt.

:get getfunction

Specify getfunction as the way to extract the value of this option. The function getfunction should take one argument, a symbol, and should return whatever customize should use as the current value for that symbol (which need not be the symbol’s Lisp value). The default is default-toplevel-value.

You have to really understand the workings of Custom to use :get correctly. It is meant for values that are treated in Custom as variables but are not actually stored in Lisp variables. It is almost surely a mistake to specify getfunction for a value that really is stored in a Lisp variable.

:initialize function

function should be a function used to initialize the variable when the defcustom is evaluated. It should take two arguments, the option name (a symbol) and the value. Here are some predefined functions meant for use in this way:

custom-initialize-set

Use the variable’s :set function to initialize the variable, but do not reinitialize it if it is already non-void.

custom-initialize-default

Like custom-initialize-set, but use the function set-default-toplevel-value to set the variable, instead of the variable’s :set function. This is the usual choice for a variable whose :set function enables or disables a minor mode; with this choice, defining the variable will not call the minor mode function, but customizing the variable will do so.

custom-initialize-reset

Always use the :set function to initialize the variable. If the variable is already non-void, reset it by calling the :set function using the current value (returned by the :get method). This is the default :initialize function.

custom-initialize-changed

Use the :set function to initialize the variable, if it is already set or has been customized; otherwise, just use set-default-toplevel-value.

custom-initialize-delay

This function behaves like custom-initialize-set, but it delays the actual initialization to the next Emacs start. This should be used in files that are preloaded (or for autoloaded variables), so that the initialization is done in the run-time context rather than the build-time context. This also has the side-effect that the (delayed) initialization is performed with the :set function. See Building Emacs.

:local value

If the value is t, mark option as automatically buffer-local; if the value is permanent, also set options permanent-local property to t. See Creating and Deleting Buffer-Local Bindings.

:risky value

Set the variable’s risky-local-variable property to value (see File Local Variables).

:safe function

Set the variable’s safe-local-variable property to function (see File Local Variables).

:set-after variables

When setting variables according to saved customizations, make sure to set the variables variables before this one; i.e., delay setting this variable until after those others have been handled. Use :set-after if setting this variable won’t work properly unless those other variables already have their intended values.

It is useful to specify the :require keyword for an option that turns on a certain feature. This causes Emacs to load the feature, if it is not already loaded, whenever the option is set. See Common Item Keywords. Here is an example:

(defcustom frobnicate-automatically nil
  "Non-nil means automatically frobnicate all buffers."
  :type 'boolean
  :require 'frobnicate-mode
  :group 'frobnicate)

If a customization item has a type such as hook or alist, which supports :options, you can add additional values to the list from outside the defcustom declaration by calling custom-add-frequent-value. For example, if you define a function my-lisp-mode-initialization intended to be called from emacs-lisp-mode-hook, you might want to add that to the list of reasonable values for emacs-lisp-mode-hook, but not by editing its definition. You can do it thus:

(custom-add-frequent-value 'emacs-lisp-mode-hook
   'my-lisp-mode-initialization)
Function: custom-add-frequent-value symbol value

For the customization option symbol, add value to the list of reasonable values.

The precise effect of adding a value depends on the customization type of symbol.

Since evaluating a defcustom form does not clear values added previously, Lisp programs can use this function to add values for user options not yet defined.

Internally, defcustom uses the symbol property standard-value to record the expression for the standard value, saved-value to record the value saved by the user with the customization buffer, and customized-value to record the value set by the user with the customization buffer, but not saved. See Symbol Properties. In addition, there’s themed-value, which is used to record the value set by a theme (see Custom Themes). These properties are lists, the car of which is an expression that evaluates to the value.

Function: custom-reevaluate-setting symbol

This function re-evaluates the standard value of symbol, which should be a user option declared via defcustom. If the variable was customized, this function re-evaluates the saved value instead. Then it sets the user option to that value (using the option’s :set property if that is defined).

This is useful for customizable options that are defined before their value could be computed correctly. For example, during startup Emacs calls this function for some user options that were defined in pre-loaded Emacs Lisp files, but whose initial values depend on information available only at run-time.

Function: custom-variable-p arg

This function returns non-nil if arg is a customizable variable. A customizable variable is either a variable that has a standard-value or custom-autoload property (usually meaning it was declared with defcustom), or an alias for another customizable variable.


15.4 Customization Types

When you define a user option with defcustom, you must specify its customization type. That is a Lisp object which describes (1) which values are legitimate and (2) how to display the value in the customization buffer for editing.

You specify the customization type in defcustom with the :type keyword. The argument of :type is evaluated, but only once when the defcustom is executed, so it isn’t useful for the value to vary. Normally we use a quoted constant. For example:

(defcustom diff-command "diff"
  "The command to use to run diff."
  :type '(string)
  :group 'diff)

In general, a customization type is a list whose first element is a symbol, one of the customization type names defined in the following sections. After this symbol come a number of arguments, depending on the symbol. Between the type symbol and its arguments, you can optionally write keyword-value pairs (see Type Keywords).

Some type symbols do not use any arguments; those are called simple types. For a simple type, if you do not use any keyword-value pairs, you can omit the parentheses around the type symbol. For example just string as a customization type is equivalent to (string).

All customization types are implemented as widgets; see Introduction in The Emacs Widget Library, for details.


15.4.1 Simple Types

This section describes all the simple customization types. For several of these customization types, the customization widget provides inline completion with C-M-i or M-TAB.

sexp

The value may be any Lisp object that can be printed and read back. You can use sexp as a fall-back for any option, if you don’t want to take the time to work out a more specific type to use.

integer

The value must be an integer.

natnum

The value must be a nonnegative integer.

number

The value must be a number (floating point or integer).

float

The value must be floating point.

string

The value must be a string. The customization buffer shows the string without delimiting ‘"’ characters or ‘\’ quotes.

regexp

Like string except that the string must be a valid regular expression.

character

The value must be a character code. A character code is actually an integer, but this type shows the value by inserting the character in the buffer, rather than by showing the number.

file

The value must be a file name. The widget provides completion.

(file :must-match t)

The value must be a file name for an existing file. The widget provides completion.

directory

The value must be a directory. The widget provides completion.

hook

The value must be a list of functions. This customization type is used for hook variables. You can use the :options keyword in a hook variable’s defcustom to specify a list of functions recommended for use in the hook; See Defining Customization Variables.

symbol

The value must be a symbol. It appears in the customization buffer as the symbol name. The widget provides completion.

function

The value must be either a lambda expression or a function name. The widget provides completion for function names.

variable

The value must be a variable name. The widget provides completion.

face

The value must be a symbol which is a face name. The widget provides completion.

boolean

The value is boolean—either nil or t. Note that by using choice and const together (see the next section), you can specify that the value must be nil or t, but also specify the text to describe each value in a way that fits the specific meaning of the alternative.

key

The value is a valid key according to key-valid-p, and suitable for use with, for example keymap-set.

key-sequence

The value is a key sequence. The customization buffer shows the key sequence using the same syntax as the kbd function. See Key Sequences. This is a legacy type; use key instead.

coding-system

The value must be a coding-system name, and you can do completion with M-TAB.

color

The value must be a valid color name. The widget provides completion for color names, as well as a sample and a button for selecting a color name from a list of color names shown in a *Colors* buffer.

fringe-bitmap

The value must be a valid fringe bitmap name. The widget provides completion.


15.4.2 Composite Types

When none of the simple types is appropriate, you can use composite types, which build new types from other types or from specified data. The specified types or data are called the arguments of the composite type. The composite type normally looks like this:

(constructor arguments…)

but you can also add keyword-value pairs before the arguments, like this:

(constructor {keyword value}arguments…)

Here is a table of constructors and how to use them to write composite types:

(cons car-type cdr-type)

The value must be a cons cell, its CAR must fit car-type, and its CDR must fit cdr-type. For example, (cons string symbol) is a customization type which matches values such as ("foo" . foo).

In the customization buffer, the CAR and CDR are displayed and edited separately, each according to their specified type.

(list element-types…)

The value must be a list with exactly as many elements as the element-types given; and each element must fit the corresponding element-type.

For example, (list integer string function) describes a list of three elements; the first element must be an integer, the second a string, and the third a function.

In the customization buffer, each element is displayed and edited separately, according to the type specified for it.

(group element-types…)

This works like list except for the formatting of text in the Custom buffer. list labels each element value with its tag; group does not.

(vector element-types…)

Like list except that the value must be a vector instead of a list. The elements work the same as in list.

(alist :key-type key-type :value-type value-type)

The value must be a list of cons-cells, the CAR of each cell representing a key of customization type key-type, and the CDR of the same cell representing a value of customization type value-type. The user can add and delete key/value pairs, and edit both the key and the value of each pair.

If omitted, key-type and value-type default to sexp.

The user can add any key matching the specified key type, but you can give some keys a preferential treatment by specifying them with the :options (see Defining Customization Variables). The specified keys will always be shown in the customize buffer (together with a suitable value), with a checkbox to include or exclude or disable the key/value pair from the alist. The user will not be able to edit the keys specified by the :options keyword argument.

The argument to the :options keywords should be a list of specifications for reasonable keys in the alist. Ordinarily, they are simply atoms, which stand for themselves. For example:

:options '("foo" "bar" "baz")

specifies that there are three known keys, namely "foo", "bar" and "baz", which will always be shown first.

You may want to restrict the value type for specific keys, for example, the value associated with the "bar" key can only be an integer. You can specify this by using a list instead of an atom in the list. The first element will specify the key, like before, while the second element will specify the value type. For example:

:options '("foo" ("bar" integer) "baz")

Finally, you may want to change how the key is presented. By default, the key is simply shown as a const, since the user cannot change the special keys specified with the :options keyword. However, you may want to use a more specialized type for presenting the key, like function-item if you know it is a symbol with a function binding. This is done by using a customization type specification instead of a symbol for the key.

:options '("foo"
           ((function-item some-function) integer)
           "baz")

Many alists use lists with two elements, instead of cons cells. For example,

(defcustom list-alist
  '(("foo" 1) ("bar" 2) ("baz" 3))
  "Each element is a list of the form (KEY VALUE).")

instead of

(defcustom cons-alist
  '(("foo" . 1) ("bar" . 2) ("baz" . 3))
  "Each element is a cons-cell (KEY . VALUE).")

Because of the way lists are implemented on top of cons cells, you can treat list-alist in the example above as a cons cell alist, where the value type is a list with a single element containing the real value.

(defcustom list-alist '(("foo" 1) ("bar" 2) ("baz" 3))
  "Each element is a list of the form (KEY VALUE)."
  :type '(alist :value-type (group integer)))

The group widget is used here instead of list only because the formatting is better suited for the purpose.

Similarly, you can have alists with more values associated with each key, using variations of this trick:

(defcustom person-data '(("brian"  50 t)
                         ("dorith" 55 nil)
                         ("ken"    52 t))
  "Alist of basic info about people.
Each element has the form (NAME AGE MALE-FLAG)."
  :type '(alist :value-type (group integer boolean)))
(plist :key-type key-type :value-type value-type)

This customization type is similar to alist (see above), except that (i) the information is stored as a property list, (see Property Lists), and (ii) key-type, if omitted, defaults to symbol rather than sexp.

(choice alternative-types…)

The value must fit one of alternative-types. For example, (choice integer string) allows either an integer or a string.

In the customization buffer, the user selects an alternative using a menu, and can then edit the value in the usual way for that alternative.

Normally the strings in this menu are determined automatically from the choices; however, you can specify different strings for the menu by including the :tag keyword in the alternatives. For example, if an integer stands for a number of spaces, while a string is text to use verbatim, you might write the customization type this way,

(choice (integer :tag "Number of spaces")
        (string :tag "Literal text"))

so that the menu offers ‘Number of spaces’ and ‘Literal text’.

In any alternative for which nil is not a valid value, other than a const, you should specify a valid default for that alternative using the :value keyword. See Type Keywords.

If some values are covered by more than one of the alternatives, customize will choose the first alternative that the value fits. This means you should always list the most specific types first, and the most general last. Here’s an example of proper usage:

(choice (const :tag "Off" nil)
        symbol (sexp :tag "Other"))

This way, the special value nil is not treated like other symbols, and symbols are not treated like other Lisp expressions.

(radio element-types…)

This is similar to choice, except that the choices are displayed using radio buttons rather than a menu. This has the advantage of displaying documentation for the choices when applicable and so is often a good choice for a choice between constant functions (function-item customization types).

(const value)

The value must be value—nothing else is allowed.

The main use of const is inside of choice. For example, (choice integer (const nil)) allows either an integer or nil.

:tag is often used with const, inside of choice. For example,

(choice (const :tag "Yes" t)
        (const :tag "No" nil)
        (const :tag "Ask" foo))

describes a variable for which t means yes, nil means no, and foo means “ask”.

(other value)

This alternative can match any Lisp value, but if the user chooses this alternative, that selects the value value.

The main use of other is as the last element of choice. For example,

(choice (const :tag "Yes" t)
        (const :tag "No" nil)
        (other :tag "Ask" foo))

describes a variable for which t means yes, nil means no, and anything else means “ask”. If the user chooses ‘Ask’ from the menu of alternatives, that specifies the value foo; but any other value (not t, nil or foo) displays as ‘Ask’, just like foo.

(function-item function)

Like const, but used for values which are functions. This displays the documentation string as well as the function name. The documentation string is either the one you specify with :doc, or function’s own documentation string.

(variable-item variable)

Like const, but used for values which are variable names. This displays the documentation string as well as the variable name. The documentation string is either the one you specify with :doc, or variable’s own documentation string.

(set types…)

The value must be a list, and each element of the list must match one of the types specified.

This appears in the customization buffer as a checklist, so that each of types may have either one corresponding element or none. It is not possible to specify two different elements that match the same one of types. For example, (set integer symbol) allows one integer and/or one symbol in the list; it does not allow multiple integers or multiple symbols. As a result, it is rare to use nonspecific types such as integer in a set.

Most often, the types in a set are const types, as shown here:

(set (const :bold) (const :italic))

Sometimes they describe possible elements in an alist:

(set (cons :tag "Height" (const height) integer)
     (cons :tag "Width" (const width) integer))

That lets the user specify a height value optionally and a width value optionally.

(repeat element-type)

The value must be a list and each element of the list must fit the type element-type. This appears in the customization buffer as a list of elements, with ‘[INS]’ and ‘[DEL]’ buttons for adding more elements or removing elements.

(restricted-sexp :match-alternatives criteria)

This is the most general composite type construct. The value may be any Lisp object that satisfies one of criteria. criteria should be a list, and each element should be one of these possibilities:

  • A predicate—that is, a function of one argument that returns either nil or non-nil according to the argument. Using a predicate in the list says that objects for which the predicate returns non-nil are acceptable.
  • A quoted constant—that is, 'object. This sort of element in the list says that object itself is an acceptable value.

For example,

(restricted-sexp :match-alternatives
                 (integerp 't 'nil))

allows integers, t and nil as legitimate values.

The customization buffer shows all legitimate values using their read syntax, and the user edits them textually.

Here is a table of the keywords you can use in keyword-value pairs in a composite type:

:tag tag

Use tag as the name of this alternative, for user communication purposes. This is useful for a type that appears inside of a choice.

:match-alternatives criteria

Use criteria to match possible values. This is used only in restricted-sexp.

:args argument-list

Use the elements of argument-list as the arguments of the type construct. For instance, (const :args (foo)) is equivalent to (const foo). You rarely need to write :args explicitly, because normally the arguments are recognized automatically as whatever follows the last keyword-value pair.


15.4.3 Splicing into Lists

The :inline feature lets you splice a variable number of elements into the middle of a list or vector customization type. You use it by adding :inline t to a type specification which is contained in a list or vector specification.

Normally, each entry in a list or vector type specification describes a single element type. But when an entry contains :inline t, the value it matches is merged directly into the containing sequence. For example, if the entry matches a list with three elements, those become three elements of the overall sequence. This is analogous to ‘,@’ in a backquote construct (see Backquote).

For example, to specify a list whose first element must be baz and whose remaining arguments should be zero or more of foo and bar, use this customization type:

(list (const baz) (set :inline t (const foo) (const bar)))

This matches values such as (baz), (baz foo), (baz bar) and (baz foo bar).

When the element-type is a choice, you use :inline not in the choice itself, but in (some of) the alternatives of the choice. For example, to match a list which must start with a file name, followed either by the symbol t or two strings, use this customization type:

(list file
      (choice (const t)
              (list :inline t string string)))

If the user chooses the first alternative in the choice, then the overall list has two elements and the second element is t. If the user chooses the second alternative, then the overall list has three elements and the second and third must be strings.

The widgets can specify predicates to say whether an inline value matches the widget with the :match-inline element.


15.4.4 Type Keywords

You can specify keyword-argument pairs in a customization type after the type name symbol. Here are the keywords you can use, and their meanings:

:value default

Provide a default value.

If nil is not a valid value for the alternative, then it is essential to specify a valid default with :value.

If you use this for a type that appears as an alternative inside of choice; it specifies the default value to use, at first, if and when the user selects this alternative with the menu in the customization buffer.

Of course, if the actual value of the option fits this alternative, it will appear showing the actual value, not default.

:format format-string

This string will be inserted in the buffer to represent the value corresponding to the type. The following ‘%’ escapes are available for use in format-string:

%[button%]

Display the text button marked as a button. The :action attribute specifies what the button will do if the user invokes it; its value is a function which takes two arguments—the widget which the button appears in, and the event.

There is no way to specify two different buttons with different actions.

%{sample%}

Show sample in a special face specified by :sample-face.

%v

Substitute the item’s value. How the value is represented depends on the kind of item, and (for variables) on the customization type.

%d

Substitute the item’s documentation string.

%h

Like ‘%d’, but if the documentation string is more than one line, add a button to control whether to show all of it or just the first line.

%t

Substitute the tag here. You specify the tag with the :tag keyword.

%%

Display a literal ‘%’.

:action action

Perform action if the user clicks on a button.

:button-face face

Use the face face (a face name or a list of face names) for button text displayed with ‘%[…%]’.

:button-prefix prefix
:button-suffix suffix

These specify the text to display before and after a button. Each can be:

nil

No text is inserted.

a string

The string is inserted literally.

a symbol

The symbol’s value is used.

:tag tag

Use tag (a string) as the tag for the value (or part of the value) that corresponds to this type.

:doc doc

Use doc as the documentation string for this value (or part of the value) that corresponds to this type. In order for this to work, you must specify a value for :format, and use ‘%d’ or ‘%h’ in that value.

The usual reason to specify a documentation string for a type is to provide more information about the meanings of alternatives inside a choice type or the parts of some other composite type.

:help-echo motion-doc

When you move to this item with widget-forward or widget-backward, it will display the string motion-doc in the echo area. In addition, motion-doc is used as the mouse help-echo string and may actually be a function or form evaluated to yield a help string. If it is a function, it is called with one argument, the widget.

:match function

Specify how to decide whether a value matches the type. The corresponding value, function, should be a function that accepts two arguments, a widget and a value; it should return non-nil if the value is acceptable.

:match-inline function

Specify how to decide whether an inline value matches the type. The corresponding value, function, should be a function that accepts two arguments, a widget and an inline value; it should return non-nil if the value is acceptable. See Splicing into Lists for more information about inline values.

:validate function

Specify a validation function for input. function takes a widget as an argument, and should return nil if the widget’s current value is valid for the widget. Otherwise, it should return the widget containing the invalid data, and set that widget’s :error property to a string explaining the error.

:type-error string

string should be a string that describes why a value doesn’t match the type, as determined by the :match function. When the :match function returns nil, the widget’s :error property will be set to string.


15.4.5 Defining New Types

In the previous sections we have described how to construct elaborate type specifications for defcustom. In some cases you may want to give such a type specification a name. The obvious case is when you are using the same type for many user options: rather than repeat the specification for each option, you can give the type specification a name, and use that name each defcustom. The other case is when a user option’s value is a recursive data structure. To make it possible for a datatype to refer to itself, it needs to have a name.

Since custom types are implemented as widgets, the way to define a new customize type is to define a new widget. We are not going to describe the widget interface here in details, see Introduction in The Emacs Widget Library, for that. Instead we are going to demonstrate the minimal functionality needed for defining new customize types by a simple example.

(define-widget 'binary-tree-of-string 'lazy
  "A binary tree made of cons-cells and strings."
  :offset 4
  :tag "Node"
  :type '(choice (string :tag "Leaf" :value "")
                 (cons :tag "Interior"
                       :value ("" . "")
                       binary-tree-of-string
                       binary-tree-of-string)))

(defcustom foo-bar ""
  "Sample variable holding a binary tree of strings."
  :type 'binary-tree-of-string)

The function to define a new widget is called define-widget. The first argument is the symbol we want to make a new widget type. The second argument is a symbol representing an existing widget, the new widget is going to be defined in terms of difference from the existing widget. For the purpose of defining new customization types, the lazy widget is perfect, because it accepts a :type keyword argument with the same syntax as the keyword argument to defcustom with the same name. The third argument is a documentation string for the new widget. You will be able to see that string with the M-x widget-browse RET binary-tree-of-string RET command.

After these mandatory arguments follow the keyword arguments. The most important is :type, which describes the data type we want to match with this widget. Here a binary-tree-of-string is described as being either a string, or a cons-cell whose car and cdr are themselves both binary-tree-of-string. Note the reference to the widget type we are currently in the process of defining. The :tag attribute is a string to name the widget in the user interface, and the :offset argument is there to ensure that child nodes are indented four spaces relative to the parent node, making the tree structure apparent in the customization buffer.

The defcustom shows how the new widget can be used as an ordinary customization type.

The reason for the name lazy is that the other composite widgets convert their inferior widgets to internal form when the widget is instantiated in a buffer. This conversion is recursive, so the inferior widgets will convert their inferior widgets. If the data structure is itself recursive, this conversion is an infinite recursion. The lazy widget prevents the recursion: it convert its :type argument only when needed.


15.5 Applying Customizations

The following functions are responsible for installing the user’s customization settings for variables and faces, respectively. When the user invokes ‘Save for future sessions’ in the Customize interface, that takes effect by writing a custom-set-variables and/or a custom-set-faces form into the custom file, to be evaluated the next time Emacs starts.

Function: custom-set-variables &rest args

This function installs the variable customizations specified by args. Each argument in args should have the form

(var expression [now [request [comment]]])

var is a variable name (a symbol), and expression is an expression which evaluates to the desired customized value.

If the defcustom form for var has been evaluated prior to this custom-set-variables call, expression is immediately evaluated, and the variable’s value is set to the result. Otherwise, expression is stored into the variable’s saved-value property, to be evaluated when the relevant defcustom is called (usually when the library defining that variable is loaded into Emacs).

The now, request, and comment entries are for internal use only, and may be omitted. now, if non-nil, means to set the variable’s value now, even if the variable’s defcustom form has not been evaluated. request is a list of features to be loaded immediately (see Features). comment is a string describing the customization.

Function: custom-set-faces &rest args

This function installs the face customizations specified by args. Each argument in args should have the form

(face spec [now [comment]])

face is a face name (a symbol), and spec is the customized face specification for that face (see Defining Faces).

The now and comment entries are for internal use only, and may be omitted. now, if non-nil, means to install the face specification now, even if the defface form has not been evaluated. comment is a string describing the customization.


15.6 Custom Themes

Custom themes are collections of settings that can be enabled or disabled as a unit. See Custom Themes in The GNU Emacs Manual. Each Custom theme is defined by an Emacs Lisp source file, which should follow the conventions described in this section. (Instead of writing a Custom theme by hand, you can also create one using a Customize-like interface; see Creating Custom Themes in The GNU Emacs Manual.)

A Custom theme file should be named foo-theme.el, where foo is the theme name. The first Lisp form in the file should be a call to deftheme, and the last form should be a call to provide-theme.

Macro: deftheme theme &optional doc &rest properties

This macro declares theme (a symbol) as the name of a Custom theme. The optional argument doc should be a string describing the theme; this is the description shown when the user invokes the describe-theme command or types ? in the ‘*Custom Themes*’ buffer. The remaining arguments properties are used pass a property list with theme attributes.

The following attributes are supported:

:family

A symbol designating what “family” a theme belongs to. A family of themes is a set of similar themes that differ by minor aspects, such as face colors that are meant for the light vs dark background of the frame.

:kind

A symbol. If a theme is enabled and this property has the value color-scheme, then the theme-choose-variant command will look for other available themes that belong to the same family in order to switch the themes. Other values are currently unspecified and should not be used.

:background-mode

A symbol, either light or dark. This attribute is currently unused, but should still be specified.

Two special theme names are disallowed (using them causes an error): user is a dummy theme that stores the user’s direct customization settings, and changed is a dummy theme that stores changes made outside of the Customize system.

Macro: provide-theme theme

This macro declares that the theme named theme has been fully specified.

In between deftheme and provide-theme are Lisp forms specifying the theme settings: usually a call to custom-theme-set-variables and/or a call to custom-theme-set-faces.

Function: custom-theme-set-variables theme &rest args

This function specifies the Custom theme theme’s variable settings. theme should be a symbol. Each argument in args should be a list of the form

(var expression [now [request [comment]]])

where the list entries have the same meanings as in custom-set-variables. See Applying Customizations.

Function: custom-theme-set-faces theme &rest args

This function specifies the Custom theme theme’s face settings. theme should be a symbol. Each argument in args should be a list of the form

(face spec [now [comment]])

where the list entries have the same meanings as in custom-set-faces. See Applying Customizations.

In theory, a theme file can also contain other Lisp forms, which would be evaluated when loading the theme, but that is bad form. To protect against loading themes containing malicious code, Emacs displays the source file and asks for confirmation from the user before loading any non-built-in theme for the first time. As such, themes are not ordinarily byte-compiled, and source files usually take precedence when Emacs is looking for a theme to load.

The following functions are useful for programmatically enabling and disabling themes:

Function: custom-theme-p theme

This function return a non-nil value if theme (a symbol) is the name of a Custom theme (i.e., a Custom theme which has been loaded into Emacs, whether or not the theme is enabled). Otherwise, it returns nil.

Variable: custom-known-themes

The value of this variable is a list of themes loaded into Emacs. Each theme is represented by a Lisp symbol (the theme name). The default value of this variable is a list containing two dummy themes: (user changed). The changed theme stores settings made before any Custom themes are applied (e.g., variables set outside of Customize). The user theme stores settings the user has customized and saved. Any additional themes declared with the deftheme macro are added to the front of this list.

Command: load-theme theme &optional no-confirm no-enable

This function loads the Custom theme named theme from its source file, looking for the source file in the directories specified by the variable custom-theme-load-path. See Custom Themes in The GNU Emacs Manual. It also enables the theme (unless the optional argument no-enable is non-nil), causing its variable and face settings to take effect. It prompts the user for confirmation before loading the theme, unless the optional argument no-confirm is non-nil.

Function: require-theme feature &optional noerror

This function searches custom-theme-load-path for a file that provides feature and then loads it. This is like the function require (see Features), except it searches custom-theme-load-path instead of load-path (see Library Search). This can be useful in Custom themes that need to load supporting Lisp files when require is unsuitable for that.

If feature, which should be a symbol, is not already present in the current Emacs session according to featurep, then require-theme searches for a file named feature with an added ‘.elc’ or ‘.el’ suffix, in that order, in the directories specified by custom-theme-load-path.

If a file providing feature is successfully found and loaded, then require-theme returns feature. The optional argument noerror determines what happens if the search or loading fails. If it is nil, the function signals an error; otherwise, it returns nil. If the file loads successfully but does not provide feature, then require-theme signals an error; this cannot be suppressed.

Command: enable-theme theme

This function enables the Custom theme named theme. It signals an error if no such theme has been loaded.

Command: disable-theme theme

This function disables the Custom theme named theme. The theme remains loaded, so that a subsequent call to enable-theme will re-enable it.


16 Loading

Loading a file of Lisp code means bringing its contents into the Lisp environment in the form of Lisp objects. Emacs finds and opens the file, reads the text, evaluates each form, and then closes the file. Such a file is also called a Lisp library.

The load functions evaluate all the expressions in a file just as the eval-buffer function evaluates all the expressions in a buffer. The difference is that the load functions read and evaluate the text in the file as found on disk, not the text in an Emacs buffer.

The loaded file must contain Lisp expressions, either as source code or as byte-compiled code. Each form in the file is called a top-level form. There is no special format for the forms in a loadable file; any form in a file may equally well be typed directly into a buffer and evaluated there. (Indeed, most code is tested this way.) Most often, the forms are function definitions and variable definitions.

Emacs can also load compiled dynamic modules: shared libraries that provide additional functionality for use in Emacs Lisp programs, just like a package written in Emacs Lisp would. When a dynamic module is loaded, Emacs calls a specially-named initialization function which the module needs to implement, and which exposes the additional functions and variables to Emacs Lisp programs.

For on-demand loading of external libraries which are known in advance to be required by certain Emacs primitives, see Dynamically Loaded Libraries.


16.1 How Programs Do Loading

Emacs Lisp has several interfaces for loading. For example, autoload creates a placeholder object for a function defined in a file; trying to call the autoloading function loads the file to get the function’s real definition (see Autoload). require loads a file if it isn’t already loaded (see Features). Ultimately, all these facilities call the load function to do the work.

Function: load filename &optional missing-ok nomessage nosuffix must-suffix

This function finds and opens a file of Lisp code, evaluates all the forms in it, and closes the file.

To find the file, load first looks for a file named filename.elc, that is, for a file whose name is filename with the extension ‘.elc’ appended. If such a file exists, and Emacs was compiled with native-compilation support (see Compilation of Lisp to Native Code), load attempts to find a corresponding ‘.eln’ file, and if found, loads it instead of filename.elc. Otherwise, it loads filename.elc (and starts a background native compilation to produce the missing ‘.eln’ file, followed by loading that file). If there is no filename.elc, then load looks for a file named filename.el. If that file exists, it is loaded. If Emacs was compiled with support for dynamic modules (see Emacs Dynamic Modules), load next looks for a file named filename.ext, where ext is a system-dependent file-name extension of shared libraries (‘.so’ on GNU and Unix systems). Finally, if neither of those names is found, load looks for a file named filename with nothing appended, and loads it if it exists. (The load function is not clever about looking at filename. In the perverse case of a file named foo.el.el, evaluation of (load "foo.el") will indeed find it.)

If Auto Compression mode is enabled, as it is by default, then if load can not find a file, it searches for a compressed version of the file before trying other file names. It decompresses and loads it if it exists. It looks for compressed versions by appending each of the suffixes in jka-compr-load-suffixes to the file name. The value of this variable must be a list of strings. Its standard value is (".gz").

If the optional argument nosuffix is non-nil, then load does not try the suffixes ‘.elc’ and ‘.el’. In this case, you must specify the precise file name you want, except that, if Auto Compression mode is enabled, load will still use jka-compr-load-suffixes to find compressed versions. By specifying the precise file name and using t for nosuffix, you can prevent file names like foo.el.el from being tried.

If the optional argument must-suffix is non-nil, then load insists that the file name used must end in either ‘.el’ or ‘.elc’ (possibly extended with a compression suffix) or the shared-library extension, unless it contains an explicit directory name.

If the option load-prefer-newer is non-nil, then when searching suffixes, load selects whichever version of a file (‘.elc’, ‘.el’, etc.) has been modified most recently. In this case, load doesn’t load the ‘.eln’ natively-compiled file even if it exists.

If filename is a relative file name, such as foo or baz/foo.bar, load searches for the file using the variable load-path. It appends filename to each of the directories listed in load-path, and loads the first file it finds whose name matches. The current default directory is tried only if it is specified in load-path, where nil stands for the default directory. load tries all three possible suffixes in the first directory in load-path, then all three suffixes in the second directory, and so on. See Library Search.

Whatever the name under which the file is eventually found, and the directory where Emacs found it, Emacs sets the value of the variable load-file-name to that file’s name.

If you get a warning that foo.elc is older than foo.el, it means you should consider recompiling foo.el. See Byte Compilation.

When loading a source file (not compiled), load performs character set translation just as Emacs would do when visiting the file. See Coding Systems.

When loading an uncompiled file, Emacs tries to expand any macros that the file contains (see Macros). We refer to this as eager macro expansion. Doing this (rather than deferring the expansion until the relevant code runs) can significantly speed up the execution of uncompiled code. Sometimes, this macro expansion cannot be done, owing to a cyclic dependency. In the simplest example of this, the file you are loading refers to a macro defined in another file, and that file in turn requires the file you are loading. Emacs will issue an error about (‘Eager macro-expansion skipped due to cycle…’) giving details of the problem. You have to restructure your code so that this does not happen. Loading a compiled file does not cause macroexpansion, because this should already have happened during compilation. See Macros and Byte Compilation.

Messages like ‘Loading foo...’ and ‘Loading foo...done’ appear in the echo area during loading unless nomessage is non-nil. If a natively-compiled ‘.eln’ file is loaded, the message says so.

Any unhandled errors while loading a file terminate loading. If the load was done for the sake of autoload, any function definitions made during the loading are undone.

If load can’t find the file to load, then normally it signals a file-error (with ‘Cannot open load file filename’). But if missing-ok is non-nil, then load just returns nil.

You can use the variable load-read-function to specify a function for load to use instead of read for reading expressions. See below.

load returns t if the file loads successfully.

Command: load-file filename

This command loads the file filename. If filename is a relative file name, then the current default directory is assumed. This command does not use load-path, and does not append suffixes. However, it does look for compressed versions (if Auto Compression Mode is enabled). Use this command if you wish to specify precisely the file name to load.

Command: load-library library

This command loads the library named library. It is equivalent to load, except for the way it reads its argument interactively. See Lisp Libraries in The GNU Emacs Manual.

Variable: load-in-progress

This variable is non-nil if Emacs is in the process of loading a file, and it is nil otherwise.

Variable: load-file-name

When Emacs is in the process of loading a file, this variable’s value is the name of that file, as Emacs found it during the search described earlier in this section.

Variable: load-read-function

This variable specifies an alternate expression-reading function for load and eval-region to use instead of read. The function should accept one argument, just as read does.

By default, this variable’s value is read. See Input Functions.

Instead of using this variable, it is cleaner to use another, newer feature: to pass the function as the read-function argument to eval-region. See Eval.

For information about how load is used in building Emacs, see Building Emacs.


16.2 Load Suffixes

We now describe some technical details about the exact suffixes that load tries.

Variable: load-suffixes

This is a list of suffixes indicating (compiled or source) Emacs Lisp files. It should not include the empty string. load uses these suffixes in order when it appends Lisp suffixes to the specified file name. The standard value is (".elc" ".el") which produces the behavior described in the previous section.

Variable: load-file-rep-suffixes

This is a list of suffixes that indicate representations of the same file. This list should normally start with the empty string. When load searches for a file it appends the suffixes in this list, in order, to the file name, before searching for another file.

Enabling Auto Compression mode appends the suffixes in jka-compr-load-suffixes to this list and disabling Auto Compression mode removes them again. The standard value of load-file-rep-suffixes if Auto Compression mode is disabled is (""). Given that the standard value of jka-compr-load-suffixes is (".gz"), the standard value of load-file-rep-suffixes if Auto Compression mode is enabled is ("" ".gz").

Function: get-load-suffixes

This function returns the list of all suffixes that load should try, in order, when its must-suffix argument is non-nil. This takes both load-suffixes and load-file-rep-suffixes into account. If load-suffixes, jka-compr-load-suffixes and load-file-rep-suffixes all have their standard values, this function returns (".elc" ".elc.gz" ".el" ".el.gz") if Auto Compression mode is enabled and (".elc" ".el") if Auto Compression mode is disabled.

To summarize, load normally first tries the suffixes in the value of (get-load-suffixes) and then those in load-file-rep-suffixes. If nosuffix is non-nil, it skips the former group, and if must-suffix is non-nil, it skips the latter group.

User Option: load-prefer-newer

If this option is non-nil, then rather than stopping at the first suffix that exists, load tests them all, and uses whichever file is the newest.


16.4 Loading Non-ASCII Characters

When Emacs Lisp programs contain string constants with non-ASCII characters, these can be represented within Emacs either as unibyte strings or as multibyte strings (see Text Representations). Which representation is used depends on how the file is read into Emacs. If it is read with decoding into multibyte representation, the text of the Lisp program will be multibyte text, and its string constants will be multibyte strings. If a file containing Latin-1 characters (for example) is read without decoding, the text of the program will be unibyte text, and its string constants will be unibyte strings. See Coding Systems.

In most Emacs Lisp programs, the fact that non-ASCII strings are multibyte strings should not be noticeable, since inserting them in unibyte buffers converts them to unibyte automatically. However, if this does make a difference, you can force a particular Lisp file to be interpreted as unibyte by writing ‘coding: raw-text’ in a local variables section. With that designator, the file will unconditionally be interpreted as unibyte. This can matter when making key bindings to non-ASCII characters written as ?vliteral.


16.5 Autoload

The autoload facility lets you register the existence of a function or macro, but put off loading the file that defines it. The first call to the function automatically loads the proper library, in order to install the real definition and other associated code, then runs the real definition as if it had been loaded all along. Autoloading can also be triggered by looking up the documentation of the function or macro (see Documentation Basics), and completion of variable and function names (see Autoload by Prefix below).

There are two ways to set up an autoloaded function: by calling autoload, and by writing a “magic” comment in the source before the real definition. autoload is the low-level primitive for autoloading; any Lisp program can call autoload at any time. Magic comments are the most convenient way to make a function autoload, for packages installed along with Emacs. These comments do nothing on their own, but they serve as a guide for the command loaddefs-generate, which constructs calls to autoload and arranges to execute them when Emacs is built.

Function: autoload function filename &optional docstring interactive type

This function defines the function (or macro) named function so as to load automatically from filename. The string filename specifies the file to load to get the real definition of function.

If filename does not contain either a directory name, or the suffix .el or .elc, this function insists on adding one of these suffixes, and it will not load from a file whose name is just filename with no added suffix. (The variable load-suffixes specifies the exact required suffixes.)

The argument docstring is the documentation string for the function. Specifying the documentation string in the call to autoload makes it possible to look at the documentation without loading the function’s real definition. Normally, this should be identical to the documentation string in the function definition itself. If it isn’t, the function definition’s documentation string takes effect when it is loaded.

If interactive is non-nil, that says function can be called interactively. This lets completion in M-x work without loading function’s real definition. The complete interactive specification is not given here; it’s not needed unless the user actually calls function, and when that happens, it’s time to load the real definition.

If interactive is a list, it is interpreted as a list of modes this command is applicable for.

You can autoload macros and keymaps as well as ordinary functions. Specify type as macro if function is really a macro. Specify type as keymap if function is really a keymap. Various parts of Emacs need to know this information without loading the real definition.

An autoloaded keymap loads automatically during key lookup when a prefix key’s binding is the symbol function. Autoloading does not occur for other kinds of access to the keymap. In particular, it does not happen when a Lisp program gets the keymap from the value of a variable and calls keymap-set; not even if the variable name is the same symbol function.

If function already has a non-void function definition that is not an autoload object, this function does nothing and returns nil. Otherwise, it constructs an autoload object (see Autoload Type), and stores it as the function definition for function. The autoload object has this form:

(autoload filename docstring interactive type)

For example,

(symbol-function 'run-prolog)
     ⇒ (autoload "prolog" 169681 t nil)

In this case, "prolog" is the name of the file to load, 169681 refers to the documentation string in the emacs/etc/DOC file (see Documentation Basics), t means the function is interactive, and nil that it is not a macro or a keymap.

Function: autoloadp object

This function returns non-nil if object is an autoload object. For example, to check if run-prolog is defined as an autoloaded function, evaluate

(autoloadp (symbol-function 'run-prolog))

The autoloaded file usually contains other definitions and may require or provide one or more features. If the file is not completely loaded (due to an error in the evaluation of its contents), any function definitions or provide calls that occurred during the load are undone. This is to ensure that the next attempt to call any function autoloading from this file will try again to load the file. If not for this, then some of the functions in the file might be defined by the aborted load, but fail to work properly for the lack of certain subroutines not loaded successfully because they come later in the file.

If the autoloaded file fails to define the desired Lisp function or macro, then an error is signaled with data "Autoloading failed to define function function-name".

A magic autoload comment (often called an autoload cookie) consists of ‘;;;###autoload’, on a line by itself, just before the real definition of the function in its autoloadable source file. The function loaddefs-generate writes a corresponding autoload call into loaddefs.el. (The string that serves as the autoload cookie and the name of the file generated by loaddefs-generate can be changed from the above defaults, see below.) Building Emacs loads loaddefs.el and thus calls autoload.

The same magic comment can copy any kind of form into loaddefs.el. The form following the magic comment is copied verbatim, except if it is one of the forms which the autoload facility handles specially (e.g., by conversion into an autoload call). The forms which are not copied verbatim are the following:

Definitions for function or function-like objects:

defun and defmacro; also cl-defun and cl-defmacro (see Argument Lists in Common Lisp Extensions), and define-overloadable-function (see the commentary in mode-local.el).

Definitions for major or minor modes:

define-minor-mode, define-globalized-minor-mode, define-generic-mode, define-derived-mode, easy-mmode-define-minor-mode, easy-mmode-define-global-mode, define-compilation-mode, and define-global-minor-mode.

Other definition types:

defcustom, defgroup, deftheme, defclass (see EIEIO in EIEIO), and define-skeleton (see Autotyping in Autotyping).

You can also use a magic comment to execute a form at build time without executing it when the file itself is loaded. To do this, write the form on the same line as the magic comment. Since it is in a comment, it does nothing when you load the source file; but loaddefs-generate copies it to loaddefs.el, where it is executed while building Emacs.

The following example shows how doctor is prepared for autoloading with a magic comment:

;;;###autoload
(defun doctor ()
  "Switch to *doctor* buffer and start giving psychotherapy."
  (interactive)
  (switch-to-buffer "*doctor*")
  (doctor-mode))

Here’s what that produces in loaddefs.el:

(autoload 'doctor "doctor" "\
Switch to *doctor* buffer and start giving psychotherapy.

\(fn)" t nil)

While the loaddefs.el isn’t for editing, we try to keep it somewhat readable for people. For instance, control characters in defvar values are escaped, and we insert a backslash and newline immediately following the double-quote of the doc string to keep the line length down. ‘(fn)’ in the usage part of the documentation string is replaced with the function’s name when the various help functions (see Help Functions) display it.

If you write a function definition with an unusual macro that is not one of the known and recognized function definition methods, use of an ordinary magic autoload comment would copy the whole definition into loaddefs.el. That is not desirable. You can put the desired autoload call into loaddefs.el instead by writing this:

;;;###autoload (autoload 'foo "myfile")
(mydefunmacro foo
  ...)

You can use a non-default string as the autoload cookie and have the corresponding autoload calls written into a file whose name is different from the default loaddefs.el. Emacs provides two variables to control this:

Variable: lisp-mode-autoload-regexp

The value of this constant is a regexp that matches autoload cookies. loaddefs-generate copies the Lisp form that follows the cookie into the autoload file it generates. This will match comments like ‘;;;###autoload’ and ‘;;;###calc-autoload’.

Variable: generated-autoload-file

The value of this variable names an Emacs Lisp file where the autoload calls should go. The default value is loaddefs.el, but you can override that, e.g., in the local variables section of a .el file (see File Local Variables). The autoload file is assumed to contain a trailer starting with a formfeed character.

The following function may be used to explicitly load the library specified by an autoload object:

Function: autoload-do-load autoload &optional name macro-only

This function performs the loading specified by autoload, which should be an autoload object. The optional argument name, if non-nil, should be a symbol whose function value is autoload; in that case, the return value of this function is the symbol’s new function value. If the value of the optional argument macro-only is macro, this function avoids loading a function, only a macro.


16.5.1 Autoload by Prefix

During completion for the commands describe-variable and describe-function, Emacs will try to load files which may contain definitions matching the prefix being completed. The variable definition-prefixes holds a hashtable which maps a prefix to the corresponding list of files to load for it. Entries to this mapping are added by calls to register-definition-prefixes which are generated by loaddefs-generate (see Autoload). Files which don’t contain any definitions worth loading (test files, for example), should set autoload-compute-prefixes to nil as a file-local variable.


16.5.2 When to Use Autoload

Do not add an autoload comment unless it is really necessary. Autoloading code means it is always globally visible. Once an item is autoloaded, there is no compatible way to transition back to it not being autoloaded (after people become accustomed to being able to use it without an explicit load).

  • The most common items to autoload are the interactive entry points to a library. For example, if python.el is a library defining a major-mode for editing Python code, autoload the definition of the python-mode function, so that people can simply use M-x python-mode to load the library.
  • Variables usually don’t need to be autoloaded. An exception is if the variable on its own is generally useful without the whole defining library being loaded. (An example of this might be something like find-exec-terminator.)
  • Don’t autoload a user option just so that a user can set it.
  • Never add an autoload comment to silence a compiler warning in another file. In the file that produces the warning, use (defvar foo) to silence an undefined variable warning, and declare-function (see Telling the Compiler that a Function is Defined) to silence an undefined function warning; or require the relevant library; or use an explicit autoload statement.

16.6 Repeated Loading

You can load a given file more than once in an Emacs session. For example, after you have rewritten and reinstalled a function definition by editing it in a buffer, you may wish to return to the original version; you can do this by reloading the file it came from.

When you load or reload files, bear in mind that the load and load-library functions automatically load a byte-compiled file rather than a non-compiled file of similar name. If you rewrite a file that you intend to save and reinstall, you need to byte-compile the new version; otherwise Emacs will load the older, byte-compiled file instead of your newer, non-compiled file! If that happens, the message displayed when loading the file includes, ‘(compiled; note, source is newer)’, to remind you to recompile it.

When writing the forms in a Lisp library file, keep in mind that the file might be loaded more than once. For example, think about whether each variable should be reinitialized when you reload the library; defvar does not change the value if the variable is already initialized. (See Defining Global Variables.)

The simplest way to add an element to an alist is like this:

(push '(leif-mode " Leif") minor-mode-alist)

But this would add multiple elements if the library is reloaded. To avoid the problem, use add-to-list (see Modifying List Variables):

(add-to-list 'minor-mode-alist '(leif-mode " Leif"))

Occasionally you will want to test explicitly whether a library has already been loaded. If the library uses provide to provide a named feature, you can use featurep earlier in the file to test whether the provide call has been executed before (see Features). Alternatively, you could use something like this:

(defvar foo-was-loaded nil)

(unless foo-was-loaded
  execute-first-time-only
  (setq foo-was-loaded t))

16.7 Features

provide and require are an alternative to autoload for loading files automatically. They work in terms of named features. Autoloading is triggered by calling a specific function, but a feature is loaded the first time another program asks for it by name.

A feature name is a symbol that stands for a collection of functions, variables, etc. The file that defines them should provide the feature. Another program that uses them may ensure they are defined by requiring the feature. This loads the file of definitions if it hasn’t been loaded already.

To require the presence of a feature, call require with the feature name as argument. require looks in the global variable features to see whether the desired feature has been provided already. If not, it loads the feature from the appropriate file. This file should call provide at the top level to add the feature to features; if it fails to do so, require signals an error.

For example, in idlwave.el, the definition for idlwave-complete-filename includes the following code:

(defun idlwave-complete-filename ()
  "Use the comint stuff to complete a file name."
   (require 'comint)
   (let* ((comint-file-name-chars "~/A-Za-z0-9+@:_.$#%={}\\-")
          (comint-completion-addsuffix nil)
          ...)
       (comint-dynamic-complete-filename)))

The expression (require 'comint) loads the file comint.el if it has not yet been loaded, ensuring that comint-dynamic-complete-filename is defined. Features are normally named after the files that provide them, so that require need not be given the file name. (Note that it is important that the require statement be outside the body of the let. Loading a library while its variables are let-bound can have unintended consequences, namely the variables becoming unbound after the let exits.)

The comint.el file contains the following top-level expression:

(provide 'comint)

This adds comint to the global features list, so that (require 'comint) will henceforth know that nothing needs to be done.

When require is used at top level in a file, it takes effect when you byte-compile that file (see Byte Compilation) as well as when you load it. This is in case the required package contains macros that the byte compiler must know about. It also avoids byte compiler warnings for functions and variables defined in the file loaded with require.

Although top-level calls to require are evaluated during byte compilation, provide calls are not. Therefore, you can ensure that a file of definitions is loaded before it is byte-compiled by including a provide followed by a require for the same feature, as in the following example.

(provide 'my-feature)  ; Ignored by byte compiler,
                       ;   evaluated by load.
(require 'my-feature)  ; Evaluated by byte compiler.

The compiler ignores the provide, then processes the require by loading the file in question. Loading the file does execute the provide call, so the subsequent require call does nothing when the file is loaded.

Function: provide feature &optional subfeatures

This function announces that feature is now loaded, or being loaded, into the current Emacs session. This means that the facilities associated with feature are or will be available for other Lisp programs.

The direct effect of calling provide is to add feature to the front of features if it is not already in that list and call any eval-after-load code waiting for it (see Hooks for Loading). The argument feature must be a symbol. provide returns feature.

If provided, subfeatures should be a list of symbols indicating a set of specific subfeatures provided by this version of feature. You can test the presence of a subfeature using featurep. The idea of subfeatures is that you use them when a package (which is one feature) is complex enough to make it useful to give names to various parts or functionalities of the package, which might or might not be loaded, or might or might not be present in a given version. See Testing Availability of Network Features, for an example.

features
     ⇒ (bar bish)

(provide 'foo)
     ⇒ foo
features
     ⇒ (foo bar bish)

When a file is loaded to satisfy an autoload, and it stops due to an error in the evaluation of its contents, any function definitions or provide calls that occurred during the load are undone. See Autoload.

Function: require feature &optional filename noerror

This function checks whether feature is present in the current Emacs session (using (featurep feature); see below). The argument feature must be a symbol.

If the feature is not present, then require loads filename with load. If filename is not supplied, then the name of the symbol feature is used as the base file name to load. However, in this case, require insists on finding feature with an added ‘.el’ or ‘.elc’ suffix (possibly extended with a compression suffix); a file whose name is just feature won’t be used. (The variable load-suffixes specifies the exact required Lisp suffixes.)

If noerror is non-nil, that suppresses errors from actual loading of the file. In that case, require returns nil if loading the file fails. Normally, require returns feature.

If loading the file succeeds but does not provide feature, require signals an error about the missing feature.

Function: featurep feature &optional subfeature

This function returns t if feature has been provided in the current Emacs session (i.e., if feature is a member of features.) If subfeature is non-nil, then the function returns t only if that subfeature is provided as well (i.e., if subfeature is a member of the subfeature property of the feature symbol.)

Variable: features

The value of this variable is a list of symbols that are the features loaded in the current Emacs session. Each symbol was put in this list with a call to provide. The order of the elements in the features list is not significant.

The use-package macro provides a convenient way of loading a feature and configuring it for use. It provides a means to combine requiring a feature, like require does, with code to be run when the feature is actually loaded, similar to load-time hooks (see Hooks for Loading). The declarative syntax of use-package makes it exceptionally easy to use in user init files.

Macro: use-package feature &rest args

This macro specifies how to load the named feature and how to configure and customize it for use. The arguments args are keyword-value pairs. Some of the important keywords and their values are:

:init forms

Specifies forms to execute before feature is loaded.

:config forms

Specifies forms to execute after loading feature.

:defer condition

If condition is non-nil, it specifies to defer loading feature until any of the autoloaded commands or variables of feature are first used. If condition is a number n, it specifies that feature should be loaded after n seconds of idle time.

:commands commands

Specifies commands of feature to be autoloaded.

:bind keybindings

Specifies the keybindings for features commands. Each binding has the form

(key-sequence . command)

or

(:map keymap (key-sequence . command))

where key-sequence is in the form accepted by the kbd macro (see Key Sequences).

For more details about use-package, see use-package User Manual.


16.8 Which File Defined a Certain Symbol

Function: symbol-file symbol &optional type native-p

This function returns the name of the file that defined symbol. If type is nil, then any kind of definition is acceptable. If type is defun, defvar, or defface, that specifies function definition, variable definition, or face definition only.

The value is normally an absolute file name. It can also be nil, if the definition is not associated with any file. If symbol specifies an autoloaded function, the value can be a relative file name without extension.

If the optional third argument native-p is non-nil, and Emacs was built with native compilation support (see Compilation of Lisp to Native Code), this function will try to find the .eln file that defined symbol, instead of the .elc or .el file. If such a .eln file is found and is not outdated, the function will return its absolute file name; otherwise it will report the name of either the source or the byte-compiled file.

The basis for symbol-file is the data in the variable load-history.

Variable: load-history

The value of this variable is an alist that associates the names of loaded library files with the names of the functions and variables they defined, as well as the features they provided or required.

Each element in this alist describes one loaded library (including libraries that are preloaded at startup). It is a list whose CAR is the absolute file name of the library (a string). The rest of the list elements have these forms:

var

The symbol var was defined as a variable.

(defun . fun)

The function fun was defined. (defun . fun), which represents defining fun as a function.

(defface . face)

The face face was defined.

(require . feature)

The feature feature was required.

(provide . feature)

The feature feature was provided.

(cl-defmethod method specializers)

The named method was defined by using cl-defmethod, with specializers as its specializers.

(define-type . type)

The type type was defined.

The value of load-history may have one element whose CAR is nil. This element describes definitions made with eval-buffer on a buffer that is not visiting a file.

The command eval-region updates load-history, but does so by adding the symbols defined to the element for the file being visited, rather than replacing that element. See Eval.

In addition to load-history, every function keeps track of its own history in the symbol property function-history. The reason why functions are treated specially in this respect is that it is common for functions to be defined in two steps in two different files (typically, one of them is an autoload), so in order to be able to properly unload a file, we need to know more precisely what that file did to the function definition.

The symbol property function-history holds a list of the form (file1 def2 file2 def3 ...), where file1 is the last file that changed the definition and def2 was the definition before file1, set by file2, etc. Logically this list should end with the name of the first file that defined this function, but to save space this last element is usually omitted.


16.9 Unloading

You can discard the functions and variables loaded by a library to reclaim memory for other Lisp objects. To do this, use the function unload-feature:

Command: unload-feature feature &optional force

This command unloads the library that provided feature feature. It undefines all functions, macros, and variables defined in that library with defun, defalias, defsubst, defmacro, defconst, defvar, and defcustom. It then restores any autoloads formerly associated with those symbols. (Loading saves these in the function-history property of the symbol.)

Before restoring the previous definitions, unload-feature runs remove-hook to remove functions defined by the library from certain hooks. These hooks include variables whose names end in ‘-hook’ (or the deprecated suffix ‘-hooks’), plus those listed in unload-feature-special-hooks, as well as auto-mode-alist. This is to prevent Emacs from ceasing to function because important hooks refer to functions that are no longer defined.

Standard unloading activities also undo ELP profiling of functions in that library, unprovides any features provided by the library, and cancels timers held in variables defined by the library.

If these measures are not sufficient to prevent malfunction, a library can define an explicit unloader named feature-unload-function. If that symbol is defined as a function, unload-feature calls it with no arguments before doing anything else. It can do whatever is appropriate to unload the library. If it returns nil, unload-feature proceeds to take the normal unload actions. Otherwise it considers the job to be done.

Ordinarily, unload-feature refuses to unload a library on which other loaded libraries depend. (A library a depends on library b if a contains a require for b.) If the optional argument force is non-nil, dependencies are ignored and you can unload any library.

The unload-feature function is written in Lisp; its actions are based on the variable load-history.

Variable: unload-feature-special-hooks

This variable holds a list of hooks to be scanned before unloading a library, to remove functions defined in the library.


16.10 Hooks for Loading

You can ask for code to be executed each time Emacs loads a library, by using the variable after-load-functions:

Variable: after-load-functions

This abnormal hook is run after loading a file. Each function in the hook is called with a single argument, the absolute filename of the file that was just loaded.

If you want code to be executed when a particular library is loaded, use the macro with-eval-after-load:

Macro: with-eval-after-load library body…

This macro arranges to evaluate body at the end of loading the file library, each time library is loaded. If library is already loaded, it evaluates body right away.

You don’t need to give a directory or extension in the file name library. Normally, you just give a bare file name, like this:

(with-eval-after-load "js" (keymap-set js-mode-map "C-c C-c" 'js-eval))

To restrict which files can trigger the evaluation, include a directory or an extension or both in library. Only a file whose absolute true name (i.e., the name with all symbolic links chased out) matches all the given name components will match. In the following example, my_inst.elc or my_inst.elc.gz in some directory ..../foo/bar will trigger the evaluation, but not my_inst.el:

(with-eval-after-load "foo/bar/my_inst.elc" …)

library can also be a feature (i.e., a symbol), in which case body is evaluated at the end of any file where (provide library) is called.

An error in body does not undo the load, but does prevent execution of the rest of body.

Normally, well-designed Lisp programs should not use with-eval-after-load. If you need to examine and set the variables defined in another library (those meant for outside use), you can do it immediately—there is no need to wait until the library is loaded. If you need to call functions defined by that library, you should load the library, preferably with require (see Features).


16.11 Emacs Dynamic Modules

A dynamic Emacs module is a shared library that provides additional functionality for use in Emacs Lisp programs, just like a package written in Emacs Lisp would.

Functions that load Emacs Lisp packages can also load dynamic modules. They recognize dynamic modules by looking at their file-name extension, a.k.a. “suffix”. This suffix is platform-dependent.

Variable: module-file-suffix

This variable holds the system-dependent value of the file-name extension of the module files. Its value is .so on POSIX hosts, .dylib on macOS, and .dll on MS-Windows.

On macOS, dynamic modules can also have the suffix .so in addition to .dylib.

Every dynamic module should export a C-callable function named emacs_module_init, which Emacs will call as part of the call to load or require which loads the module. It should also export a symbol named plugin_is_GPL_compatible to indicate that its code is released under the GPL or compatible license; Emacs will signal an error if your program tries to load modules that don’t export such a symbol.

If a module needs to call Emacs functions, it should do so through the API (Application Programming Interface) defined and documented in the header file emacs-module.h that is part of the Emacs distribution. See Writing Dynamically-Loaded Modules, for details of using that API when writing your own modules.

Modules can create user-ptr Lisp objects that embed pointers to C struct’s defined by the module. This is useful for keeping around complex data structures created by a module, to be passed back to the module’s functions. User-ptr objects can also have associated finalizers – functions to be run when the object is GC’ed; this is useful for freeing any resources allocated for the underlying data structure, such as memory, open file descriptors, etc. See Conversion Between Lisp and Module Values.

Function: user-ptrp object

This function returns t if its argument is a user-ptr object.

Function: module-load file

Emacs calls this low-level primitive to load a module from the specified file and perform the necessary initialization of the module. This is the primitive which makes sure the module exports the plugin_is_GPL_compatible symbol, calls the module’s emacs_module_init function, and signals an error if that function returns an error indication, or if the user typed C-g during the initialization. If the initialization succeeds, module-load returns t. Note that file must already have the proper file-name extension, as this function doesn’t try looking for files with known extensions, unlike load.

Unlike load, module-load doesn’t record the module in load-history, doesn’t print any messages, and doesn’t protect against recursive loads. Most users should therefore use load, load-file, load-library, or require instead of module-load.

Loadable modules in Emacs are enabled by using the --with-modules option at configure time.


17 Byte Compilation

Emacs Lisp has a compiler that translates functions written in Lisp into a special representation called byte-code that can be executed more efficiently. The compiler replaces Lisp function definitions with byte-code. When a byte-code function is called, its definition is evaluated by the byte-code interpreter.

Because the byte-compiled code is evaluated by the byte-code interpreter, instead of being executed directly by the machine’s hardware (as true compiled code is), byte-code is completely transportable from machine to machine without recompilation. It is not, however, as fast as true compiled code.

In general, any version of Emacs can run byte-compiled code produced by recent earlier versions of Emacs, but the reverse is not true.

If you do not want a Lisp file to be compiled, ever, put a file-local variable binding for no-byte-compile into it, like this:

;; -*-no-byte-compile: t; -*-

17.1 Performance of Byte-Compiled Code

A byte-compiled function is not as efficient as a primitive function written in C, but runs much faster than the version written in Lisp. Here is an example:

(defun silly-loop (n)
  "Return the time, in seconds, to run N iterations of a loop."
  (let ((t1 (float-time)))
    (while (> (setq n (1- n)) 0))
    (- (float-time) t1)))
⇒ silly-loop

(silly-loop 50000000)
⇒ 5.200886011123657

(byte-compile 'silly-loop)
⇒ [Compiled code not shown]

(silly-loop 50000000)
⇒ 0.6239290237426758

In this example, the interpreted code required more than 5 seconds to run, whereas the byte-compiled code required less than 1 second. These results are representative, but actual results may vary.


17.2 Byte-Compilation Functions

You can byte-compile an individual function or macro definition with the byte-compile function. You can compile a whole file with byte-compile-file, or several files with byte-recompile-directory or batch-byte-compile.

Sometimes, the byte compiler produces warning and/or error messages (see Compiler Errors, for details). These messages are normally recorded in a buffer called *Compile-Log*, which uses Compilation mode. See Compilation Mode in The GNU Emacs Manual. However, if the variable byte-compile-debug is non-nil, error messages will be signaled as Lisp errors instead (see Errors).

Be careful when writing macro calls in files that you intend to byte-compile. Since macro calls are expanded when they are compiled, the macros need to be loaded into Emacs or the byte compiler will not do the right thing. The usual way to handle this is with require forms which specify the files containing the needed macro definitions (see Features). Normally, the byte compiler does not evaluate the code that it is compiling, but it handles require forms specially, by loading the specified libraries. To avoid loading the macro definition files when someone runs the compiled program, write eval-when-compile around the require calls (see Evaluation During Compilation). For more details, See Macros and Byte Compilation.

Inline (defsubst) functions are less troublesome; if you compile a call to such a function before its definition is known, the call will still work right, it will just run slower.

Function: byte-compile symbol

This function byte-compiles the function definition of symbol, replacing the previous definition with the compiled one. The function definition of symbol must be the actual code for the function; byte-compile does not handle function indirection. The return value is the byte-code function object which is the compiled definition of symbol (see Byte-Code Function Objects).

(defun factorial (integer)
  "Compute factorial of INTEGER."
  (if (= 1 integer) 1
    (* integer (factorial (1- integer)))))
⇒ factorial

(byte-compile 'factorial)
⇒
#[257
  "\211\300U\203^H^@\300\207\211\301^BS!_\207"
  [1 factorial] 4
  "Compute factorial of INTEGER.\n\n(fn INTEGER)"]

If symbol’s definition is a byte-code function object, byte-compile does nothing and returns nil. It does not compile the symbol’s definition again, since the original (non-compiled) code has already been replaced in the symbol’s function cell by the byte-compiled code.

The argument to byte-compile can also be a lambda expression. In that case, the function returns the corresponding compiled code but does not store it anywhere.

Command: compile-defun &optional arg

This command reads the defun containing point, compiles it, and evaluates the result. If you use this on a defun that is actually a function definition, the effect is to install a compiled version of that function.

compile-defun normally displays the result of evaluation in the echo area, but if arg is non-nil, it inserts the result in the current buffer after the form it has compiled.

Command: byte-compile-file filename

This function compiles a file of Lisp code named filename into a file of byte-code. The output file’s name is made by changing the ‘.el’ suffix into ‘.elc’; if filename does not end in ‘.el’, it adds ‘.elc’ to the end of filename.

Compilation works by reading the input file one form at a time. If it is a definition of a function or macro, the compiled function or macro definition is written out. Other forms are batched together, then each batch is compiled, and written so that its compiled code will be executed when the file is read. All comments are discarded when the input file is read.

This command returns t if there were no errors and nil otherwise. When called interactively, it prompts for the file name.

$ ls -l push*
-rw-r--r-- 1 lewis lewis 791 Oct  5 20:31 push.el

(byte-compile-file "~/emacs/push.el")
     ⇒ t

$ ls -l push*
-rw-r--r-- 1 lewis lewis 791 Oct  5 20:31 push.el
-rw-rw-rw- 1 lewis lewis 638 Oct  8 20:25 push.elc
Command: byte-recompile-directory directory &optional flag force follow-symlinks

This command recompiles every ‘.el’ file in directory (or its subdirectories) that needs recompilation. A file needs recompilation if a ‘.elc’ file exists but is older than the ‘.el’ file.

When a ‘.el’ file has no corresponding ‘.elc’ file, flag says what to do. If it is nil, this command ignores these files. If flag is 0, it compiles them. If it is neither nil nor 0, it asks the user whether to compile each such file, and asks about each subdirectory as well.

Interactively, byte-recompile-directory prompts for directory and flag is the prefix argument.

If force is non-nil, this command recompiles every ‘.el’ file that has a ‘.elc’ file.

This command will normally not compile ‘.el’ files that are symlinked. If the optional follow-symlink parameter is non-nil, symlinked ‘.el’ will also be compiled.

The returned value is unpredictable.

Function: batch-byte-compile &optional noforce

This function runs byte-compile-file on files specified on the command line. This function must be used only in a batch execution of Emacs, as it kills Emacs on completion. An error in one file does not prevent processing of subsequent files, but no output file will be generated for it, and the Emacs process will terminate with a nonzero status code.

If noforce is non-nil, this function does not recompile files that have an up-to-date ‘.elc’ file.

$ emacs -batch -f batch-byte-compile *.el

17.3 Documentation Strings and Compilation

When Emacs loads functions and variables from a byte-compiled file, it normally does not load their documentation strings into memory. Each documentation string is dynamically loaded from the byte-compiled file only when needed. This saves memory, and speeds up loading by skipping the processing of the documentation strings.

This feature has a drawback: if you delete, move, or alter the compiled file (such as by compiling a new version), Emacs may no longer be able to access the documentation string of previously-loaded functions or variables. Such a problem normally only occurs if you build Emacs yourself, and happen to edit and/or recompile the Lisp source files. To solve it, just reload each file after recompilation.

Dynamic loading of documentation strings from byte-compiled files is determined, at compile time, for each byte-compiled file. It can be disabled via the option byte-compile-dynamic-docstrings.

User Option: byte-compile-dynamic-docstrings

If this is non-nil, the byte compiler generates compiled files that are set up for dynamic loading of documentation strings.

To disable the dynamic loading feature for a specific file, set this option to nil in its header line (see Local Variables in Files in The GNU Emacs Manual), like this:

-*-byte-compile-dynamic-docstrings: nil;-*-

This is useful mainly if you expect to change the file, and you want Emacs sessions that have already loaded it to keep working when the file changes.

Internally, the dynamic loading of documentation strings is accomplished by writing compiled files with a special Lisp reader construct, ‘#@count’. This construct skips the next count characters. It also uses the ‘#$’ construct, which stands for the name of this file, as a string. Do not use these constructs in Lisp source files; they are not designed to be clear to humans reading the file.


17.4 Dynamic Loading of Individual Functions

When you compile a file, you can optionally enable the dynamic function loading feature (also known as lazy loading). With dynamic function loading, loading the file doesn’t fully read the function definitions in the file. Instead, each function definition contains a place-holder which refers to the file. The first time each function is called, it reads the full definition from the file, to replace the place-holder.

The advantage of dynamic function loading is that loading the file should become faster. This is a good thing for a file which contains many separate user-callable functions, if using one of them does not imply you will probably also use the rest. A specialized mode which provides many keyboard commands often has that usage pattern: a user may invoke the mode, but use only a few of the commands it provides.

The dynamic loading feature has certain disadvantages:

  • If you delete or move the compiled file after loading it, Emacs can no longer load the remaining function definitions not already loaded.
  • If you alter the compiled file (such as by compiling a new version), then trying to load any function not already loaded will usually yield nonsense results.

These problems will never happen in normal circumstances with installed Emacs files. But they are quite likely to happen with Lisp files that you are changing. The easiest way to prevent these problems is to reload the new compiled file immediately after each recompilation.

Experience shows that using dynamic function loading provides benefits that are hardly measurable, so this feature is deprecated since Emacs 27.1.

The byte compiler uses the dynamic function loading feature if the variable byte-compile-dynamic is non-nil at compilation time. Do not set this variable globally, since dynamic loading is desirable only for certain files. Instead, enable the feature for specific source files with file-local variable bindings. For example, you could do it by writing this text in the source file’s first line:

-*-byte-compile-dynamic: t;-*-
Variable: byte-compile-dynamic

If this is non-nil, the byte compiler generates compiled files that are set up for dynamic function loading.

Function: fetch-bytecode function

If function is a byte-code function object, this immediately finishes loading the byte code of function from its byte-compiled file, if it is not fully loaded already. Otherwise, it does nothing. It always returns function.


17.5 Evaluation During Compilation

These features permit you to write code to be evaluated during compilation of a program.

Macro: eval-and-compile body…

This form marks body to be evaluated both when you compile the containing code and when you run it (whether compiled or not).

You can get a similar result by putting body in a separate file and referring to that file with require. That method is preferable when body is large. Effectively require is automatically eval-and-compile, the package is loaded both when compiling and executing.

autoload is also effectively eval-and-compile too. It’s recognized when compiling, so uses of such a function don’t produce “not known to be defined” warnings.

Most uses of eval-and-compile are fairly sophisticated.

If a macro has a helper function to build its result, and that macro is used both locally and outside the package, then eval-and-compile should be used to get the helper both when compiling and then later when running.

If functions are defined programmatically (with fset say), then eval-and-compile can be used to have that done at compile-time as well as run-time, so calls to those functions are checked (and warnings about “not known to be defined” suppressed).

Macro: eval-when-compile body…

This form marks body to be evaluated at compile time but not when the compiled program is loaded. The result of evaluation by the compiler becomes a constant which appears in the compiled program. If you load the source file, rather than compiling it, body is evaluated normally.

If you have a constant that needs some calculation to produce, eval-when-compile can do that at compile-time. For example,

(defvar my-regexp
  (eval-when-compile (regexp-opt '("aaa" "aba" "abb"))))

If you’re using another package, but only need macros from it (the byte compiler will expand those), then eval-when-compile can be used to load it for compiling, but not executing. For example,

(eval-when-compile
  (require 'my-macro-package))

The same sort of thing goes for macros and defsubst functions defined locally and only for use within the file. They are needed for compiling the file, but in most cases they are not needed for execution of the compiled file. For example,

(eval-when-compile
  (unless (fboundp 'some-new-thing)
    (defmacro some-new-thing ()
      (compatibility code))))

This is often good for code that’s only a fallback for compatibility with other versions of Emacs.

Common Lisp Note: At top level, eval-when-compile is analogous to the Common Lisp idiom (eval-when (compile eval) …). Elsewhere, the Common Lisp ‘#.’ reader macro (but not when interpreting) is closer to what eval-when-compile does.


17.6 Compiler Errors

Error and warning messages from byte compilation are printed in a buffer named *Compile-Log*. These messages include file names and line numbers identifying the location of the problem. The usual Emacs commands for operating on compiler output can be used on these messages.

When an error is due to invalid syntax in the program, the byte compiler might get confused about the error’s exact location. One way to investigate is to switch to the buffer  *Compiler Input*. (This buffer name starts with a space, so it does not show up in the Buffer Menu.) This buffer contains the program being compiled, and point shows how far the byte compiler was able to read; the cause of the error might be nearby. See Debugging Invalid Lisp Syntax, for some tips for locating syntax errors.

A common type of warning issued by the byte compiler is for functions and variables that were used but not defined. Such warnings report the line number for the end of the file, not the locations where the missing functions or variables were used; to find these, you must search the file manually.

If you are sure that a warning message about a missing function or variable is unjustified, there are several ways to suppress it:

  • You can suppress the warning for a specific call to a function func by conditionalizing it on an fboundp test, like this:
    (if (fboundp 'func) ...(func ...)...)
    

    The call to func must be in the then-form of the if, and func must appear quoted in the call to fboundp. (This feature operates for cond as well.)

  • Likewise, you can suppress the warning for a specific use of a variable variable by conditionalizing it on a boundp test:
    (if (boundp 'variable) ...variable...)
    

    The reference to variable must be in the then-form of the if, and variable must appear quoted in the call to boundp.

  • You can tell the compiler that a function is defined using declare-function. See Telling the Compiler that a Function is Defined.
  • Likewise, you can tell the compiler that a variable is defined using defvar with no initial value. (Note that this marks the variable as special, i.e. dynamically bound, but only within the current lexical scope, or file if at top-level.) See Defining Global Variables.

You can also suppress compiler warnings within a certain expression using the with-suppressed-warnings macro:

Special Form: with-suppressed-warnings warnings body…

In execution, this is equivalent to (progn body...), but the compiler does not issue warnings for the specified conditions in body. warnings is an association list of warning symbols and function/variable symbols they apply to. For instance, if you wish to call an obsolete function called foo, but want to suppress the compilation warning, say:

(with-suppressed-warnings ((obsolete foo))
  (foo ...))

For more coarse-grained suppression of compiler warnings, you can use the with-no-warnings construct:

Special Form: with-no-warnings body…

In execution, this is equivalent to (progn body...), but the compiler does not issue warnings for anything that occurs inside body.

We recommend that you use with-suppressed-warnings instead, but if you do use this construct, that you use it around the smallest possible piece of code to avoid missing possible warnings other than one you intend to suppress.

Byte compiler warnings can be controlled more precisely by setting the variable byte-compile-warnings. See its documentation string for details.

Sometimes you may wish the byte-compiler warnings to be reported using error. If so, set byte-compile-error-on-warn to a non-nil value.


17.7 Byte-Code Function Objects

Byte-compiled functions have a special data type: they are byte-code function objects. Whenever such an object appears as a function to be called, Emacs uses the byte-code interpreter to execute the byte-code.

Internally, a byte-code function object is much like a vector; its elements can be accessed using aref. Its printed representation is like that for a vector, with an additional ‘#’ before the opening ‘[’. It must have at least four elements; there is no maximum number, but only the first six elements have any normal use. They are:

argdesc

The descriptor of the arguments. This can either be a list of arguments, as described in Features of Argument Lists, or an integer encoding the required number of arguments. In the latter case, the value of the descriptor specifies the minimum number of arguments in the bits zero to 6, and the maximum number of arguments in bits 8 to 14. If the argument list uses &rest, then bit 7 is set; otherwise it’s cleared.

If argdesc is a list, the arguments will be dynamically bound before executing the byte code. If argdesc is an integer, the arguments will be instead pushed onto the stack of the byte-code interpreter, before executing the code.

byte-code

The string containing the byte-code instructions.

constants

The vector of Lisp objects referenced by the byte code. These include symbols used as function names and variable names.

stacksize

The maximum stack size this function needs.

docstring

The documentation string (if any); otherwise, nil. The value may be a number or a list, in case the documentation string is stored in a file. Use the function documentation to get the real documentation string (see Access to Documentation Strings).

interactive

The interactive spec (if any). This can be a string or a Lisp expression. It is nil for a function that isn’t interactive.

Here’s an example of a byte-code function object, in printed representation. It is the definition of the command backward-sexp.

#[256
  "\211\204^G^@\300\262^A\301^A[!\207"
  [1 forward-sexp]
  3
  1793299
  "^p"]

The primitive way to create a byte-code object is with make-byte-code:

Function: make-byte-code &rest elements

This function constructs and returns a byte-code function object with elements as its elements.

You should not try to come up with the elements for a byte-code function yourself, because if they are inconsistent, Emacs may crash when you call the function. Always leave it to the byte compiler to create these objects; it makes the elements consistent (we hope).


17.8 Disassembled Byte-Code

People do not write byte-code; that job is left to the byte compiler. But we provide a disassembler to satisfy a cat-like curiosity. The disassembler converts the byte-compiled code into human-readable form.

The byte-code interpreter is implemented as a simple stack machine. It pushes values onto a stack of its own, then pops them off to use them in calculations whose results are themselves pushed back on the stack. When a byte-code function returns, it pops a value off the stack and returns it as the value of the function.

In addition to the stack, byte-code functions can use, bind, and set ordinary Lisp variables, by transferring values between variables and the stack.

Command: disassemble object &optional buffer-or-name

This command displays the disassembled code for object. In interactive use, or if buffer-or-name is nil or omitted, the output goes in a buffer named *Disassemble*. If buffer-or-name is non-nil, it must be a buffer or the name of an existing buffer. Then the output goes there, at point, and point is left before the output.

The argument object can be a function name, a lambda expression (see Lambda Expressions), or a byte-code object (see Byte-Code Function Objects). If it is a lambda expression, disassemble compiles it and disassembles the resulting compiled code.

Here are two examples of using the disassemble function. We have added explanatory comments to help you relate the byte-code to the Lisp source; these do not appear in the output of disassemble.

(defun factorial (integer)
  "Compute factorial of an integer."
  (if (= 1 integer) 1
    (* integer (factorial (1- integer)))))
     ⇒ factorial

(factorial 4)
     ⇒ 24

(disassemble 'factorial)
     -| byte-code for factorial:
 doc: Compute factorial of an integer.
 args: (arg1)

0   dup                   ; Get the value of integer and
                          ;   push it onto the stack.
1   constant 1            ; Push 1 onto stack.
2   eqlsign               ; Pop top two values off stack, compare
                          ;   them, and push result onto stack.
3   goto-if-nil 1         ; Pop and test top of stack;
                          ;   if nil, go to 1, else continue.
6   constant 1            ; Push 1 onto top of stack.
7   return                ; Return the top element of the stack.
8:1 dup                   ; Push value of integer onto stack.
9   constant factorial    ; Push factorial onto stack.
10  stack-ref 2           ; Push value of integer onto stack.
11  sub1                  ; Pop integer, decrement value,
                          ;   push new value onto stack.
12  call     1            ; Call function factorial using first
                          ;   (i.e., top) stack element as argument;
                          ;   push returned value onto stack.
13  mult                  ; Pop top two values off stack, multiply
                          ;   them, and push result onto stack.
14  return                ; Return the top element of the stack.

The silly-loop function is somewhat more complex:

(defun silly-loop (n)
  "Return time before and after N iterations of a loop."
  (let ((t1 (current-time-string)))
    (while (> (setq n (1- n))
              0))
    (list t1 (current-time-string))))
     ⇒ silly-loop

(disassemble 'silly-loop)
     -| byte-code for silly-loop:
 doc: Return time before and after N iterations of a loop.
 args: (arg1)

0   constant current-time-string  ; Push current-time-string
                                  ;   onto top of stack.
1   call     0            ; Call current-time-string with no
                          ;   argument, push result onto stack as t1.
2:1 stack-ref 1           ; Get value of the argument n
                          ;   and push the value on the stack.
3   sub1                  ; Subtract 1 from top of stack.
4   dup                   ; Duplicate top of stack; i.e., copy the top
                          ;   of the stack and push copy onto stack.
5   stack-set 3           ; Pop the top of the stack,
                          ;   and set n to the value.

;; (In effect, the sequence dup stack-set copies the top of
;; the stack into the value of n without popping it.)

7   constant 0            ; Push 0 onto stack.
8   gtr                   ; Pop top two values off stack,
                          ;   test if n is greater than 0
                          ;   and push result onto stack.
9   goto-if-not-nil 1     ; Goto 1 if n > 0
                          ;   (this continues the while loop)
                          ;   else continue.
12  dup                   ; Push value of t1 onto stack.
13  constant current-time-string  ; Push current-time-string
                                  ;   onto the top of the stack.
14  call     0            ; Call current-time-string again.
15  list2                 ; Pop top two elements off stack, create a
                          ;   list of them, and push it onto stack.
16  return                ; Return value of the top of stack.

18 Compilation of Lisp to Native Code

In addition to the byte-compilation, described in the previous chapter, Emacs can also optionally compile Lisp function definitions into a true compiled code, known as native code. This feature uses the libgccjit library, which is part of the GCC distribution, and requires that Emacs be built with support for using that library. It also requires to have GCC and Binutils (the assembler and linker) available on your system for you to be able to native-compile Lisp code.

To determine whether the current Emacs process can produce and load natively-compiled Lisp code, call native-comp-available-p (see Native-Compilation Functions).

Unlike byte-compiled code, natively-compiled Lisp code is executed directly by the machine’s hardware, and therefore runs at full speed that the host CPU can provide. The resulting speedup generally depends on what the Lisp code does, but is usually 2.5 to 5 times faster than the corresponding byte-compiled code.

Since native code is generally incompatible between different systems, the natively-compiled code is not transportable from one machine to another, it can only be used on the same machine where it was produced or on very similar ones (having the same CPU and run-time libraries). The transportability of natively-compiled code is the same as that of shared libraries (.so or .dll files).

Libraries of natively-compiled code include crucial dependencies on Emacs Lisp primitives (see What Is a Function?) and their calling conventions, and thus Emacs usually won’t load natively-compiled code produced by earlier or later Emacs versions; native compilation of the same Lisp code by a different Emacs version will usually produce a natively-compiled library under a unique file name that only that version of Emacs will be able to load. However, the use of unique file names allows to have in the same directory several versions of the same Lisp library natively-compiled by several different versions of Emacs.

A non-nil file-local variable binding of no-byte-compile (see Byte Compilation) also disables the native compilation of that file. In addition, a similar variable no-native-compile disables just the native compilation of the file. If both no-byte-compile and no-native-compile are specified, the former takes precedence.

Sometimes there could be a need to prevent the native compilation from writing its results, the *.eln files, into a subdirectory of user-emacs-directory (see The Init File). You can do that by either changing the value of native-comp-eln-load-path (see Native-Compilation Variables) or by temporarily pointing the HOME environment variable to a non-existing directory. Note that the latter technique might still produce a small number of *.eln files if Emacs needs to generate trampolines, which are used if Lisp primitives are advised or redefined in your Lisp code that is being natively compiled. See trampolines. Alternatively, you can specify that the *.eln files are written to a non-default directory using the startup-redirect-eln-cache function; see Native-Compilation Functions.


18.1 Native-Compilation Functions

Native-Compilation is implemented as a side effect of byte-compilation (see Byte Compilation). Thus, compiling Lisp code natively always produces its byte code as well, and therefore all the rules and caveats of preparing Lisp code for byte compilation (see Byte-Compilation Functions) are valid for native-compilation as well.

You can natively-compile either a single function or macro definition, or a whole file of Lisp code, with the native-compile function. Natively-compiling a file will produce the .eln file with native code.

Native compilation might produce warning or error messages; these are normally recorded in the buffer called *Native-compile-Log*. In interactive sessions, it uses the special LIMPLE mode (native-comp-limple-mode), which sets up font-lock as appropriate for this log, and is otherwise the same as Fundamental mode. Logging of messages resulting from native-compilation can be controlled by the native-comp-verbose variable (see Native-Compilation Variables).

When Emacs is run non-interactively, messages produced by native-compilation are reported by calling message (see Displaying Messages in the Echo Area), and are usually displayed on the standard error stream of the terminal from which Emacs was invoked.

Function: native-compile function-or-file &optional output

This function compiles function-or-file into native code. The argument function-or-file can be a function symbol, a Lisp form, or a name (a string) of the file which contains the Emacs Lisp source code to compile. If the optional argument output is provided, it must be a string specifying the name of the file to write the compiled code into. Otherwise, if function-or-file is a function or a Lisp form, this function returns the compiled object, and if function-or-file is a file name, the function returns the full absolute name of the file it created for the compiled code. The output file is by default given the .eln extension.

This function runs the final phase of the native compilation, which invokes GCC via libgccjit, in a separate subprocess, which invokes the same Emacs executable as the process that called this function.

Function: batch-native-compile &optional for-tarball

This function runs native-compilation on files specified on the Emacs command line in batch mode. It must be used only in a batch execution of Emacs, as it kills Emacs upon completion of the compilation. If one or more of the files fail to compile, the Emacs process will attempt to compile all the other files, and will terminate with a non-zero status code. The optional argument for-tarball, if non-nil, tells the function to place the resulting .eln files in the last directory mentioned in native-comp-eln-load-path (see Library Search); this is meant to be used as part of building an Emacs source tarball for the first time, when the natively-compiled files, which are absent from the source tarball, should be generated in the build tree instead of the user’s cache directory.

Native compilation can be run entirely asynchronously, in a subprocess of the main Emacs process. This leaves the main Emacs process free to use while the compilation runs in the background. This is the method used by Emacs to natively-compile any Lisp file or byte-compiled Lisp file that is loaded into Emacs, when no natively-compiled file for it is available. Note that because of this use of a subprocess, native compilation may produce warning and errors which byte-compilation does not, and Lisp code may thus need to be modified to work correctly. See native-comp-async-report-warnings-errors in see Native-Compilation Variables for more details.

Function: native-compile-async files &optional recursively load selector

This function compiles the named files asynchronously. The argument files should be a single file name (a string) or a list of one or more file and/or directory names. If directories are present in the list, the optional argument recursively should be non-nil to cause the compilation to recurse into those directories. If load is non-nil, Emacs will load each file that it succeeded to compile. The optional argument selector allows control of which of files will be compiled; it can have one of the following values:

nil or omitted

Select all the files and directories in files.

a regular expression string

Select the files and directories whose names match the regexp.

a function

A predicate function, which will be called with each file and directory in files, and should return non-nil if the file or the directory should be selected for compilation.

On systems with multiple CPU execution units, when files names more than one file, this function will normally start several compilation subprocesses in parallel, under the control of native-comp-async-jobs-number (see Native-Compilation Variables).

Command: emacs-lisp-native-compile

This command compiles the file visited by the current buffer into native code, if the file was changed since the last time it was natively-compiled.

Command: emacs-lisp-native-compile-and-load

This command compiles the file visited by the current buffer into native code, like emacs-lisp-native-compile, but it also loads the native code when the compilation finishes.

The following function allows Lisp programs to test whether native-compilation is available at runtime.

Function: native-comp-available-p

This function returns non-nil if the running Emacs process has the native-compilation support compiled into it. On systems that load libgccjit dynamically, it also makes sure that library is available and can be loaded. Lisp programs that need to know up front whether native-compilation is available should use this predicate.

By default, asynchronous native compilation writes the *.eln files it produces to a subdirectory of the first writable directory specified by the native-comp-eln-load-path variable (see Native-Compilation Variables). You can change this by using the following function in your startup files:

Function: startup-redirect-eln-cache cache-directory

This function arranges for the asynchronous native compilation to write the produced *.eln files to cache-directory, which must be a single directory, a string. It also destructively modifies native-comp-eln-load-path such that its first element is cache-directory. If cache-directory is not an absolute file name, it is interpreted relative to user-emacs-directory (see The Init File).


18.2 Native-Compilation Variables

This section documents the variables that control native-compilation.

User Option: native-comp-speed

This variable specifies the optimization level for native compilation. Its value should be a number between −1 and 3. Values between 0 and 3 specify the optimization levels equivalent to the corresponding compiler -O0, -O1, etc. command-line options of the compiler. The value −1 means disable native-compilation: functions and files will be only byte-compiled; however, the *.eln files will still be produced, they will just contain the compiled code in bytecode form. (This can be achieved at function granularity by using the (declare (speed -1)) form, see The declare Form.) The default value is 2.

User Option: native-comp-debug

This variable specifies the level of debugging information produced by native-compilation. Its value should be a number between zero and 3, with the following meaning:

0

No debugging output. This is the default.

1

Emit debugging symbols with the native code. This allows easier debugging of the native code with debuggers such as gdb.

2

Like 1, and in addition dump pseudo-C code.

3

Like 2, and in addition dump the GCC intermediate passes and libgccjit log file.

User Option: native-comp-verbose

This variable controls the verbosity of native-compilation by suppressing some or all of the log messages emitted by it. If its value is zero, the default, all of the log messages are suppressed. Setting it to a value between 1 and 3 will allow logging of the messages whose level is above the value. The values have the following interpretations:

0

No logging. This is the default.

1

Log the final LIMPLE representation of the code.

2

Log the LAP, the final LIMPLE, and some additional pass info.

3

Maximum verbosity: log everything.

User Option: native-comp-async-jobs-number

This variable determines the maximum number of native-compilation subprocesses that will be started simultaneously. It should be a non-negative number. The default value is zero, which means use half the number of the CPU execution units, or 1 if the CPU has only one execution unit.

User Option: native-comp-async-report-warnings-errors

If this variable’s value is non-nil, warnings and errors from asynchronous native-compilation subprocesses are reported in the main Emacs session in a buffer named *Warnings*. The default value t means display the resulting buffer. To log warnings without popping up the *Warnings* buffer, set this variable to silent.

A common cause for asynchronous native-compilation to produce warnings is compiling a file that is missing some require of a necessary feature. The feature may be loaded into the main emacs, but because native compilation always starts from a subprocess with a pristine environment, that may not be true for the subprocess.

User Option: native-comp-async-query-on-exit

If this variable’s value is non-nil, Emacs will query upon exiting whether to exit and kill any asynchronous native-compilation subprocesses that are still running, thus preventing the corresponding .eln files from being written. If the value is nil, the default, Emacs will kill these subprocesses without querying.

The variable native-comp-eln-load-path holds the list of directories where Emacs looks for the *.eln files (see Library Search); in that role it is the equivalent of load-path used to look for *.el and *.elc files. The directories in this list are also used for writing the *.eln files produced by asynchronous native-compilation; specifically, Emacs will write these files into the first writable directory in the list. Thus, you can control where native-compilation stores the results by changing the value of this variable.

Variable: native-comp-jit-compilation

This variable, if non-nil, enables asynchronous (a.k.a. just-in-time, or JIT) native compilation of the *.elc files loaded by Emacs for which the corresponding *.eln files do not already exist. This JIT compilation uses separate Emacs sub-processes running in batch mode, according to the value of native-comp-async-jobs-number. When the JIT compilation of a Lisp file finishes successfully, the resulting .eln file is loaded and its code replaces the definition of functions provided by the .elc file.

Setting the value of native-comp-jit-compilation to nil disables JIT native compilation. However, even when JIT native compilation is disabled, Emacs might still need to start asynchronous native compilation subprocesses to produce trampolines. To control this, use a separate variable, described below.

Variable: native-comp-enable-subr-trampolines

This variable controls generation of trampolines. A trampoline is a small piece of native code required to allow calling Lisp primitives, which were advised or redefined, from Lisp code that was natively-compiled with native-comp-speed set to 2 or greater. Emacs stores the generated trampolines on separate *.eln files. By default, this variable’s value is t, which enables the generation of trampoline files; setting it to nil disables the generation of trampolines. Note that if a trampoline needed for advising or redefining a primitive is not available and cannot be generated, calls to that primitive from natively-compiled Lisp will ignore redefinitions and advices, and will behave as if the primitive was called directly from C. Therefore, we don’t recommend disabling the trampoline generation, unless you know that all the trampolines needed by your Lisp programs are already compiled and accessible to Emacs.

The value of this variable can also be a string, in which case it specifies the name of a directory in which to store the generated trampoline *.eln files, overriding the directories in native-comp-eln-load-path. This is useful if you want the trampolines to be generated as needed, but don’t want to store them under the user’s HOME directory or in the other public directories where *.eln files are kept. However, unlike with directories in native-comp-eln-load-path, the trampolines will be stored in the directory given by the value of this variable, not in its version-specific subdirectory. If the name of this directory is not absolute, it is interpreted relative to invocation-directory (see Operating System Environment)

If this variable is non-nil, and Emacs needs to produce a trampoline, but it cannot find any writable directory to store the trampoline, it will store it inside temporary-file-directory (see Generating Unique File Names).

Trampolines produced when no writable directory is found to store them, or when this variable is a string, will only be available for the duration of the current Emacs session, because Emacs doesn’t look for trampolines in either of these places.


19 Debugging Lisp Programs

There are several ways to find and investigate problems in an Emacs Lisp program.

Other useful tools for debugging input and output problems are the dribble file (see Terminal Input) and the open-termscript function (see Terminal Output).


19.1 The Lisp Debugger

The ordinary Lisp debugger provides the ability to suspend evaluation of a form. While evaluation is suspended (a state that is commonly known as a break), you may examine the run time stack, examine the values of local or global variables, or change those values. Since a break is a recursive edit, all the usual editing facilities of Emacs are available; you can even run programs that will enter the debugger recursively. See Recursive Editing.


19.1.1 Entering the Debugger on an Error

The most important time to enter the debugger is when a Lisp error happens. This allows you to investigate the immediate causes of the error.

However, entry to the debugger is not a normal consequence of an error. Many commands signal Lisp errors when invoked inappropriately, and during ordinary editing it would be very inconvenient to enter the debugger each time this happens. So if you want errors to enter the debugger, set the variable debug-on-error to non-nil. (The command toggle-debug-on-error provides an easy way to do this.)

Note that, for technical reasons, you cannot use the facilities defined in this subsection to debug errors in Lisp that the redisplay code has invoked. See Debugging Redisplay Errors, for help with these.

User Option: debug-on-error

This variable determines whether the debugger is called when an error is signaled and not handled. If debug-on-error is t, all kinds of errors call the debugger, except those listed in debug-ignored-errors (see below). If it is nil, none call the debugger.

The value can also be a list of error conditions (see How to Signal an Error). Then the debugger is called only for error conditions in this list (except those also listed in debug-ignored-errors). For example, if you set debug-on-error to the list (void-variable), the debugger is only called for errors about a variable that has no value.

Note that eval-expression-debug-on-error overrides this variable in some cases; see below.

When this variable is non-nil, Emacs does not create an error handler around process filter functions and sentinels. Therefore, errors in these functions also invoke the debugger. See Processes.

User Option: debug-ignored-errors

This variable specifies errors which should not enter the debugger, regardless of the value of debug-on-error. Its value is a list of error condition symbols and/or regular expressions. If the error has any of those condition symbols, or if the error message matches any of the regular expressions, then that error does not enter the debugger.

The normal value of this variable includes user-error, as well as several errors that happen often during editing but rarely result from bugs in Lisp programs. However, “rarely” is not “never”; if your program fails with an error that matches this list, you may try changing this list to debug the error. The easiest way is usually to set debug-ignored-errors to nil.

User Option: eval-expression-debug-on-error

If this variable has a non-nil value (the default), running the command eval-expression causes debug-on-error to be temporarily bound to t. See Evaluating Emacs Lisp Expressions in The GNU Emacs Manual.

If eval-expression-debug-on-error is nil, then the value of debug-on-error is not changed during eval-expression.

User Option: debug-on-signal

Normally, errors caught by condition-case never invoke the debugger. The condition-case gets a chance to handle the error before the debugger gets a chance.

If you change debug-on-signal to a non-nil value, the debugger gets the first chance at every error, regardless of the presence of condition-case. (To invoke the debugger, the error must still fulfill the criteria specified by debug-on-error and debug-ignored-errors.)

For example, setting this variable is useful to get a backtrace from code evaluated by emacsclient’s --eval option. If Lisp code evaluated by emacsclient signals an error while this variable is non-nil, the backtrace will popup in the running Emacs.

Warning: Setting this variable to non-nil may have annoying effects. Various parts of Emacs catch errors in the normal course of affairs, and you may not even realize that errors happen there. If you need to debug code wrapped in condition-case, consider using condition-case-unless-debug (see Writing Code to Handle Errors).

User Option: debug-on-event

If you set debug-on-event to a special event (see Special Events), Emacs will try to enter the debugger as soon as it receives this event, bypassing special-event-map. At present, the only supported values correspond to the signals SIGUSR1 and SIGUSR2 (this is the default). This can be helpful when inhibit-quit is set and Emacs is not otherwise responding.

Variable: debug-on-message

If you set debug-on-message to a regular expression, Emacs will enter the debugger if it displays a matching message in the echo area. For example, this can be useful when trying to find the cause of a particular message.

Variable: debug-allow-recursive-debug

You can evaluate forms in the current stack frame in the ‘*Backtrace*’ buffer with the e command, and while edebugging you can use the e and C-x C-e commands to do something similar. By default, the debugger is inhibited by these commands (because (re-)entering the debugger at this point will usually take you out of the debugging context you’re in). Set debug-allow-recursive-debug to a non-nil value to allow these commands to enter the debugger recursively.

To debug an error that happens during loading of the init file, use the option ‘--debug-init’. This binds debug-on-error to t while loading the init file, and bypasses the condition-case which normally catches errors in the init file.


19.1.2 Debugging Redisplay Errors

When an error occurs in Lisp code which redisplay has invoked, Emacs’s usual debugging mechanisms are unusable, for technical reasons. This subsection describes how to get a backtrace from such an error, which should be helpful in debugging it.

These directions apply to Lisp forms used, for example, in :eval mode line constructs (see The Data Structure of the Mode Line), and in all hooks invoked from redisplay, such as:

Note that if you have had an error in a hook function called from redisplay, the error handling might have removed this function from the hook. You will thus need to reinitialize that hook somehow, perhaps with add-hook, to be able to replay the bug.

To generate a backtrace in these circumstances, set the variable backtrace-on-redisplay-error to non-nil. When the error occurs, Emacs will dump the backtrace to the buffer *Redisplay-trace*, but won’t automatically display it in a window. This is to avoid needlessly corrupting the redisplay you are debugging. You will thus need to display the buffer yourself, with a command such as switch-to-buffer-other-frame C-x 5 b.

Variable: backtrace-on-redisplay-error

Set this variable to non-nil to enable the generation of a backtrace when an error occurs in any Lisp called from redisplay.


19.1.3 Debugging Infinite Loops

When a program loops infinitely and fails to return, your first problem is to stop the loop. On most operating systems, you can do this with C-g, which causes a quit. See Quitting.

Ordinary quitting gives no information about why the program was looping. To get more information, you can set the variable debug-on-quit to non-nil. Once you have the debugger running in the middle of the infinite loop, you can proceed from the debugger using the stepping commands. If you step through the entire loop, you may get enough information to solve the problem.

Quitting with C-g is not considered an error, and debug-on-error has no effect on the handling of C-g. Likewise, debug-on-quit has no effect on errors.

User Option: debug-on-quit

This variable determines whether the debugger is called when quit is signaled and not handled. If debug-on-quit is non-nil, then the debugger is called whenever you quit (that is, type C-g). If debug-on-quit is nil (the default), then the debugger is not called when you quit.


19.1.4 Entering the Debugger on a Function Call

To investigate a problem that happens in the middle of a program, one useful technique is to enter the debugger whenever a certain function is called. You can do this to the function in which the problem occurs, and then step through the function, or you can do this to a function called shortly before the problem, step quickly over the call to that function, and then step through its caller.

Command: debug-on-entry function-name

This function requests function-name to invoke the debugger each time it is called.

Any function or macro defined as Lisp code may be set to break on entry, regardless of whether it is interpreted code or compiled code. If the function is a command, it will enter the debugger when called from Lisp and when called interactively (after the reading of the arguments). You can also set debug-on-entry for primitive functions (i.e., those written in C) this way, but it only takes effect when the primitive is called from Lisp code. Debug-on-entry is not allowed for special forms.

When debug-on-entry is called interactively, it prompts for function-name in the minibuffer. If the function is already set up to invoke the debugger on entry, debug-on-entry does nothing. debug-on-entry always returns function-name.

Here’s an example to illustrate use of this function:

(defun fact (n)
  (if (zerop n) 1
      (* n (fact (1- n)))))
     ⇒ fact
(debug-on-entry 'fact)
     ⇒ fact
(fact 3)

------ Buffer: *Backtrace* ------
Debugger entered--entering a function:
* fact(3)
  eval((fact 3))
  eval-last-sexp-1(nil)
  eval-last-sexp(nil)
  call-interactively(eval-last-sexp)
------ Buffer: *Backtrace* ------

Command: cancel-debug-on-entry &optional function-name

This function undoes the effect of debug-on-entry on function-name. When called interactively, it prompts for function-name in the minibuffer. If function-name is omitted or nil, it cancels break-on-entry for all functions. Calling cancel-debug-on-entry does nothing to a function which is not currently set up to break on entry.


19.1.5 Entering the debugger when a variable is modified

Sometimes a problem with a function is due to a wrong setting of a variable. Setting up the debugger to trigger whenever the variable is changed is a quick way to find the origin of the setting.

Command: debug-on-variable-change variable

This function arranges for the debugger to be called whenever variable is modified.

It is implemented using the watchpoint mechanism, so it inherits the same characteristics and limitations: all aliases of variable will be watched together, only dynamic variables can be watched, and changes to the objects referenced by variables are not detected. For details, see Running a function when a variable is changed..

Command: cancel-debug-on-variable-change &optional variable

This function undoes the effect of debug-on-variable-change on variable. When called interactively, it prompts for variable in the minibuffer. If variable is omitted or nil, it cancels break-on-change for all variables. Calling cancel-debug-on-variable-change does nothing to a variable which is not currently set up to break on change.


19.1.6 Explicit Entry to the Debugger

You can cause the debugger to be called at a certain point in your program by writing the expression (debug) at that point. To do this, visit the source file, insert the text ‘(debug)’ at the proper place, and type C-M-x (eval-defun, a Lisp mode key binding). Warning: if you do this for temporary debugging purposes, be sure to undo this insertion before you save the file!

The place where you insert ‘(debug)’ must be a place where an additional form can be evaluated and its value ignored. (If the value of (debug) isn’t ignored, it will alter the execution of the program!) The most common suitable places are inside a progn or an implicit progn (see Sequencing).

If you don’t know exactly where in the source code you want to put the debug statement, but you want to display a backtrace when a certain message is displayed, you can set debug-on-message to a regular expression matching the desired message.


19.1.7 Using the Debugger

When the debugger is entered, it displays the previously selected buffer in one window and a buffer named *Backtrace* in another window. The backtrace buffer contains one line for each level of Lisp function execution currently going on. At the beginning of this buffer is a message describing the reason that the debugger was invoked (such as the error message and associated data, if it was invoked due to an error).

The backtrace buffer is read-only and uses a special major mode, Debugger mode, in which letters are defined as debugger commands. The usual Emacs editing commands are available; thus, you can switch windows to examine the buffer that was being edited at the time of the error, switch buffers, visit files, or do any other sort of editing. However, the debugger is a recursive editing level (see Recursive Editing) and it is wise to go back to the backtrace buffer and exit the debugger (with the q command) when you are finished with it. Exiting the debugger gets out of the recursive edit and buries the backtrace buffer. (You can customize what the q command does with the backtrace buffer by setting the variable debugger-bury-or-kill. For example, set it to kill if you prefer to kill the buffer rather than bury it. Consult the variable’s documentation for more possibilities.)

When the debugger has been entered, the debug-on-error variable is temporarily set according to eval-expression-debug-on-error. If the latter variable is non-nil, debug-on-error will temporarily be set to t. However, further errors that occur while debugging won’t (by default) trigger another debugger, because inhibit-debugger will also be bound to non-nil.

The debugger itself must be run byte-compiled, since it makes assumptions about the state of the Lisp interpreter. These assumptions are false if the debugger is running interpreted.


19.1.8 Backtraces

Debugger mode is derived from Backtrace mode, which is also used to show backtraces by Edebug and ERT. (see Edebug, and the ERT manual in ERT: Emacs Lisp Regression Testing.)

The backtrace buffer shows you the functions that are executing and their argument values. When a backtrace buffer is created, it shows each stack frame on one, possibly very long, line. (A stack frame is the place where the Lisp interpreter records information about a particular invocation of a function.) The most recently called function will be at the top.

In a backtrace you can specify a stack frame by moving point to a line describing that frame. The frame whose line point is on is considered the current frame.

If a function name is underlined, that means Emacs knows where its source code is located. You can click with the mouse on that name, or move to it and type RET, to visit the source code. You can also type RET while point is on any name of a function or variable which is not underlined, to see help information for that symbol in a help buffer, if any exists. The xref-find-definitions command, bound to M-., can also be used on any identifier in a backtrace (see Looking Up Identifiers in The GNU Emacs Manual).

In backtraces, the tails of long lists and the ends of long strings, vectors or structures, as well as objects which are deeply nested, will be printed as underlined “...”. You can click with the mouse on a “...”, or type RET while point is on it, to show the part of the object that was hidden. To control how much abbreviation is done, customize backtrace-line-length.

Here is a list of commands for navigating and viewing backtraces:

v

Toggle the display of local variables of the current stack frame.

p

Move to the beginning of the frame, or to the beginning of the previous frame.

n

Move to the beginning of the next frame.

+

Add line breaks and indentation to the top-level Lisp form at point to make it more readable.

-

Collapse the top-level Lisp form at point back to a single line.

#

Toggle print-circle for the frame at point.

:

Toggle print-gensym for the frame at point.

.

Expand all the forms abbreviated with “...” in the frame at point.


19.1.9 Debugger Commands

The debugger buffer (in Debugger mode) provides special commands in addition to the usual Emacs commands and to the Backtrace mode commands described in the previous section. The most important use of debugger commands is for stepping through code, so that you can see how control flows. The debugger can step through the control structures of an interpreted function, but cannot do so in a byte-compiled function. If you would like to step through a byte-compiled function, replace it with an interpreted definition of the same function. (To do this, visit the source for the function and type C-M-x on its definition.) You cannot use the Lisp debugger to step through a primitive function.

Some of the debugger commands operate on the current frame. If a frame starts with a star, that means that exiting that frame will call the debugger again. This is useful for examining the return value of a function.

Here is a list of Debugger mode commands:

c

Exit the debugger and continue execution. This resumes execution of the program as if the debugger had never been entered (aside from any side-effects that you caused by changing variable values or data structures while inside the debugger).

d

Continue execution, but enter the debugger the next time any Lisp function is called. This allows you to step through the subexpressions of an expression, seeing what values the subexpressions compute, and what else they do.

The stack frame made for the function call which enters the debugger in this way will be flagged automatically so that the debugger will be called again when the frame is exited. You can use the u command to cancel this flag.

b

Flag the current frame so that the debugger will be entered when the frame is exited. Frames flagged in this way are marked with stars in the backtrace buffer.

u

Don’t enter the debugger when the current frame is exited. This cancels a b command on that frame. The visible effect is to remove the star from the line in the backtrace buffer.

j

Flag the current frame like b. Then continue execution like c, but temporarily disable break-on-entry for all functions that are set up to do so by debug-on-entry.

e

Read a Lisp expression in the minibuffer, evaluate it (with the relevant lexical environment, if applicable), and print the value in the echo area. The debugger alters certain important variables, and the current buffer, as part of its operation; e temporarily restores their values from outside the debugger, so you can examine and change them. This makes the debugger more transparent. By contrast, M-: does nothing special in the debugger; it shows you the variable values within the debugger. By default, this command suppresses the debugger during evaluation, so that an error in the evaluated expression won’t add a new error on top of the existing one. Set the debug-allow-recursive-debug user option to a non-nil value to override this.

R

Like e, but also save the result of evaluation in the buffer *Debugger-record*.

q

Terminate the program being debugged; return to top-level Emacs command execution.

If the debugger was entered due to a C-g but you really want to quit, and not debug, use the q command.

r

Return a value from the debugger. The value is computed by reading an expression with the minibuffer and evaluating it.

The r command is useful when the debugger was invoked due to exit from a Lisp call frame (as requested with b or by entering the frame with d); then the value specified in the r command is used as the value of that frame. It is also useful if you call debug and use its return value. Otherwise, r has the same effect as c, and the specified return value does not matter.

You can’t use r when the debugger was entered due to an error.

l

Display a list of functions that will invoke the debugger when called. This is a list of functions that are set to break on entry by means of debug-on-entry.


19.1.10 Invoking the Debugger

Here we describe in full detail the function debug that is used to invoke the debugger.

Command: debug &rest debugger-args

This function enters the debugger. In interactive sessions, it switches to a buffer named *Backtrace* (or *Backtrace*<2> if it is the second recursive entry to the debugger, etc.), and fills it with information about the stack of Lisp function calls. It then enters a recursive edit, showing the backtrace buffer in Debugger mode. In batch mode (more generally, when noninteractive is non-nil, see Batch Mode), this function shows the Lisp backtrace on the standard error stream, and then kills Emacs, causing it to exit with a non-zero exit code (see Killing Emacs). Binding backtrace-on-error-noninteractive to nil suppresses the backtrace in batch mode, see below.

The Debugger mode c, d, j, and r commands exit the recursive edit; then debug switches back to the previous buffer and returns to whatever called debug. This is the only way the function debug can return to its caller.

The use of the debugger-args is that debug displays the rest of its arguments at the top of the *Backtrace* buffer, so that the user can see them. Except as described below, this is the only way these arguments are used.

However, certain values for first argument to debug have a special significance. (Normally, these values are used only by the internals of Emacs, and not by programmers calling debug.) Here is a table of these special values:

lambda

A first argument of lambda means debug was called because of entry to a function when debug-on-next-call was non-nil. The debugger displays ‘Debugger entered--entering a function:’ as a line of text at the top of the buffer.

debug

debug as first argument means debug was called because of entry to a function that was set to debug on entry. The debugger displays the string ‘Debugger entered--entering a function:’, just as in the lambda case. It also marks the stack frame for that function so that it will invoke the debugger when exited.

t

When the first argument is t, this indicates a call to debug due to evaluation of a function call form when debug-on-next-call is non-nil. The debugger displays ‘Debugger entered--beginning evaluation of function call form:’ as the top line in the buffer.

exit

When the first argument is exit, it indicates the exit of a stack frame previously marked to invoke the debugger on exit. The second argument given to debug in this case is the value being returned from the frame. The debugger displays ‘Debugger entered--returning value:’ in the top line of the buffer, followed by the value being returned.

error

When the first argument is error, the debugger indicates that it is being entered because an error or quit was signaled and not handled, by displaying ‘Debugger entered--Lisp error:’ followed by the error signaled and any arguments to signal. For example,

(let ((debug-on-error t))
  (/ 1 0))

------ Buffer: *Backtrace* ------
Debugger entered--Lisp error: (arith-error)
  /(1 0)
...
------ Buffer: *Backtrace* ------

If an error was signaled, presumably the variable debug-on-error is non-nil. If quit was signaled, then presumably the variable debug-on-quit is non-nil.

nil

Use nil as the first of the debugger-args when you want to enter the debugger explicitly. The rest of the debugger-args are printed on the top line of the buffer. You can use this feature to display messages—for example, to remind yourself of the conditions under which debug is called.

Variable: backtrace-on-error-noninteractive

If this variable is non-nil, the default, entering the debugger in batch mode shows the backtrace of Lisp functions calls. Binding the variable to the nil value suppresses the backtrace and shows only the error message.


19.1.11 Internals of the Debugger

This section describes functions and variables used internally by the debugger.

Variable: debugger

The value of this variable is the function to call to invoke the debugger. Its value must be a function of any number of arguments, or, more typically, the name of a function. This function should invoke some kind of debugger. The default value of the variable is debug.

The first argument that Lisp hands to the function indicates why it was called. The convention for arguments is detailed in the description of debug (see Invoking the Debugger).

Function: backtrace

This function prints a trace of Lisp function calls currently active. The trace is identical to the one that debug would show in the *Backtrace* buffer. The return value is always nil.

In the following example, a Lisp expression calls backtrace explicitly. This prints the backtrace to the stream standard-output, which, in this case, is the buffer ‘backtrace-output’.

Each line of the backtrace represents one function call. The line shows the function followed by a list of the values of the function’s arguments if they are all known; if they are still being computed, the line consists of a list containing the function and its unevaluated arguments. Long lists or deeply nested structures may be elided.

(with-output-to-temp-buffer "backtrace-output"
  (let ((var 1))
    (save-excursion
      (setq var (eval '(progn
                         (1+ var)
                         (list 'testing (backtrace))))))))

     ⇒ (testing nil)

----------- Buffer: backtrace-output ------------
  backtrace()
  (list 'testing (backtrace))
  (progn ...)
  eval((progn (1+ var) (list 'testing (backtrace))))
  (setq ...)
  (save-excursion ...)
  (let ...)
  (with-output-to-temp-buffer ...)
  eval((with-output-to-temp-buffer ...))
  eval-last-sexp-1(nil)
  eval-last-sexp(nil)
  call-interactively(eval-last-sexp)
----------- Buffer: backtrace-output ------------
User Option: debugger-stack-frame-as-list

If this variable is non-nil, every stack frame of the backtrace is displayed as a list. This aims at improving the backtrace readability at the cost of special forms no longer being visually different from regular function calls.

With debugger-stack-frame-as-list non-nil, the above example would look as follows:

----------- Buffer: backtrace-output ------------
  (backtrace)
  (list 'testing (backtrace))
  (progn ...)
  (eval (progn (1+ var) (list 'testing (backtrace))))
  (setq ...)
  (save-excursion ...)
  (let ...)
  (with-output-to-temp-buffer ...)
  (eval (with-output-to-temp-buffer ...))
  (eval-last-sexp-1 nil)
  (eval-last-sexp nil)
  (call-interactively eval-last-sexp)
----------- Buffer: backtrace-output ------------
Variable: debug-on-next-call

If this variable is non-nil, it says to call the debugger before the next eval, apply or funcall. Entering the debugger sets debug-on-next-call to nil.

The d command in the debugger works by setting this variable.

Function: backtrace-debug level flag

This function sets the debug-on-exit flag of the stack frame level levels down the stack, giving it the value flag. If flag is non-nil, this will cause the debugger to be entered when that frame later exits. Even a nonlocal exit through that frame will enter the debugger.

This function is used only by the debugger.

Variable: command-debug-status

This variable records the debugging status of the current interactive command. Each time a command is called interactively, this variable is bound to nil. The debugger can set this variable to leave information for future debugger invocations during the same command invocation.

The advantage of using this variable rather than an ordinary global variable is that the data will never carry over to a subsequent command invocation.

This variable is obsolete and will be removed in future versions.

Function: backtrace-frame frame-number &optional base

The function backtrace-frame is intended for use in Lisp debuggers. It returns information about what computation is happening in the stack frame frame-number levels down.

If that frame has not evaluated the arguments yet, or is a special form, the value is (nil function arg-forms…).

If that frame has evaluated its arguments and called its function already, the return value is (t function arg-values…).

In the return value, function is whatever was supplied as the CAR of the evaluated list, or a lambda expression in the case of a macro call. If the function has a &rest argument, that is represented as the tail of the list arg-values.

If base is specified, frame-number counts relative to the topmost frame whose function is base.

If frame-number is out of range, backtrace-frame returns nil.

Function: mapbacktrace function &optional base

The function mapbacktrace calls function once for each frame in the backtrace, starting at the first frame whose function is base (or from the top if base is omitted or nil).

function is called with four arguments: evald, func, args, and flags.

If a frame has not evaluated its arguments yet or is a special form, evald is nil and args is a list of forms.

If a frame has evaluated its arguments and called its function already, evald is t and args is a list of values. flags is a plist of properties of the current frame: currently, the only supported property is :debug-on-exit, which is t if the stack frame’s debug-on-exit flag is set.


19.2 Edebug

Edebug is a source-level debugger for Emacs Lisp programs, with which you can:

  • Step through evaluation, stopping before and after each expression.
  • Set conditional or unconditional breakpoints.
  • Stop when a specified condition is true (the global break event).
  • Trace slow or fast, stopping briefly at each stop point, or at each breakpoint.
  • Display expression results and evaluate expressions as if outside of Edebug.
  • Automatically re-evaluate a list of expressions and display their results each time Edebug updates the display.
  • Output trace information on function calls and returns.
  • Stop when an error occurs.
  • Display a backtrace, omitting Edebug’s own frames.
  • Specify argument evaluation for macros and defining forms.
  • Obtain rudimentary coverage testing and frequency counts.

The first three sections below should tell you enough about Edebug to start using it.


19.2.1 Using Edebug

To debug a Lisp program with Edebug, you must first instrument the Lisp code that you want to debug. A simple way to do this is to first move point into the definition of a function or macro and then do C-u C-M-x (eval-defun with a prefix argument). See Instrumenting for Edebug, for alternative ways to instrument code.

Once a function is instrumented, any call to the function activates Edebug. Depending on which Edebug execution mode you have selected, activating Edebug may stop execution and let you step through the function, or it may update the display and continue execution while checking for debugging commands. The default execution mode is step, which stops execution. See Edebug Execution Modes.

Within Edebug, you normally view an Emacs buffer showing the source of the Lisp code you are debugging. This is referred to as the source code buffer, and it is temporarily read-only.

An arrow in the left fringe indicates the line where the function is executing. Point initially shows where within the line the function is executing, but this ceases to be true if you move point yourself.

If you instrument the definition of fac (shown below) and then execute (fac 3), here is what you would normally see. Point is at the open-parenthesis before if.

(defun fac (n)
=>∗(if (< 0 n)
      (* n (fac (1- n)))
    1))

The places within a function where Edebug can stop execution are called stop points. These occur both before and after each subexpression that is a list, and also after each variable reference. Here we use periods to show the stop points in the function fac:

(defun fac (n)
  .(if .(< 0 n.).
      .(* n. .(fac .(1- n.).).).
    1).)

The special commands of Edebug are available in the source code buffer in addition to the commands of Emacs Lisp mode. For example, you can type the Edebug command SPC to execute until the next stop point. If you type SPC once after entry to fac, here is the display you will see:

(defun fac (n)
=>(if ∗(< 0 n)
      (* n (fac (1- n)))
    1))

When Edebug stops execution after an expression, it displays the expression’s value in the echo area.

Other frequently used commands are b to set a breakpoint at a stop point, g to execute until a breakpoint is reached, and q to exit Edebug and return to the top-level command loop. Type ? to display a list of all Edebug commands.


19.2.2 Instrumenting for Edebug

In order to use Edebug to debug Lisp code, you must first instrument the code. Instrumenting code inserts additional code into it, to invoke Edebug at the proper places.

When you invoke command C-M-x (eval-defun) with a prefix argument on a function definition, it instruments the definition before evaluating it. (This does not modify the source code itself.) If the variable edebug-all-defs is non-nil, that inverts the meaning of the prefix argument: in this case, C-M-x instruments the definition unless it has a prefix argument. The default value of edebug-all-defs is nil. The command M-x edebug-all-defs toggles the value of the variable edebug-all-defs.

If edebug-all-defs is non-nil, then the commands eval-region, and eval-buffer also instrument any definitions they evaluate. Similarly, edebug-all-forms controls whether eval-region should instrument any form, even non-defining forms. This doesn’t apply to loading or evaluations in the minibuffer. The command M-x edebug-all-forms toggles this option.

Another command, M-x edebug-eval-top-level-form, is available to instrument any top-level form regardless of the values of edebug-all-defs and edebug-all-forms. edebug-defun is an alias for edebug-eval-top-level-form.

While Edebug is active, the command I (edebug-instrument-callee) instruments the definition of the function or macro called by the list form after point, if it is not already instrumented. This is possible only if Edebug knows where to find the source for that function; for this reason, after loading Edebug, eval-region records the position of every definition it evaluates, even if not instrumenting it. See also the i command (see Jumping), which steps into the call after instrumenting the function.

Edebug knows how to instrument all the standard special forms, interactive forms with an expression argument, anonymous lambda expressions, and other defining forms. However, Edebug cannot determine on its own what a user-defined macro will do with the arguments of a macro call, so you must provide that information using Edebug specifications; for details, see Edebug and Macros.

When Edebug is about to instrument code for the first time in a session, it runs the hook edebug-setup-hook, then sets it to nil. You can use this to load Edebug specifications associated with a package you are using, but only when you use Edebug.

If Edebug detects a syntax error while instrumenting, it leaves point at the erroneous code and signals an invalid-read-syntax error. Example:

error→ Invalid read syntax: "Expected lambda expression"

One potential reason for such a failure to instrument is that some macro definitions are not yet known to Emacs. To work around this, load the file which defines the function you are about to instrument.

To remove instrumentation from a definition, simply re-evaluate its definition in a way that does not instrument. There are two ways of evaluating forms that never instrument them: from a file with load, and from the minibuffer with eval-expression (M-:).

A different way to remove the instrumentation from a definition is to use the edebug-remove-instrumentation command. It also allows removing the instrumentation from everything that has been instrumented.

See Evaluation, for other evaluation functions available inside of Edebug.


19.2.3 Edebug Execution Modes

Edebug supports several execution modes for running the program you are debugging. We call these alternatives Edebug execution modes; do not confuse them with major or minor modes. The current Edebug execution mode determines how far Edebug continues execution before stopping—whether it stops at each stop point, or continues to the next breakpoint, for example—and how much Edebug displays the progress of the evaluation before it stops.

Normally, you specify the Edebug execution mode by typing a command to continue the program in a certain mode. Here is a table of these commands; all except for S resume execution of the program, at least for a certain distance.

S

Stop: don’t execute any more of the program, but wait for more Edebug commands (edebug-stop).

SPC

Step: stop at the next stop point encountered (edebug-step-mode).

n

Next: stop at the next stop point encountered after an expression (edebug-next-mode). Also see edebug-forward-sexp in Jumping.

t

Trace: pause (normally one second) at each Edebug stop point (edebug-trace-mode).

T

Rapid trace: update the display at each stop point, but don’t actually pause (edebug-Trace-fast-mode).

g

Go: run until the next breakpoint (edebug-go-mode). See Edebug Breakpoints.

c

Continue: pause one second at each breakpoint, and then continue (edebug-continue-mode).

C

Rapid continue: move point to each breakpoint, but don’t pause (edebug-Continue-fast-mode).

G

Go non-stop: ignore breakpoints (edebug-Go-nonstop-mode). You can still stop the program by typing S, or any editing command.

In general, the execution modes earlier in the above list run the program more slowly or stop sooner than the modes later in the list.

When you enter a new Edebug level, Edebug will normally stop at the first instrumented function it encounters. If you prefer to stop only at a break point, or not at all (for example, when gathering coverage data), change the value of edebug-initial-mode from its default step to go, or Go-nonstop, or one of its other values (see Edebug Options). You can do this readily with C-x C-a C-m (edebug-set-initial-mode):

Command: edebug-set-initial-mode

This command, bound to C-x C-a C-m, sets edebug-initial-mode. It prompts you for a key to indicate the mode. You should enter one of the eight keys listed above, which sets the corresponding mode.

Note that you may reenter the same Edebug level several times if, for example, an instrumented function is called several times from one command.

While executing or tracing, you can interrupt the execution by typing any Edebug command. Edebug stops the program at the next stop point and then executes the command you typed. For example, typing t during execution switches to trace mode at the next stop point. You can use S to stop execution without doing anything else.

If your function happens to read input, a character you type intending to interrupt execution may be read by the function instead. You can avoid such unintended results by paying attention to when your program wants input.

Keyboard macros containing the commands in this section do not completely work: exiting from Edebug, to resume the program, loses track of the keyboard macro. This is not easy to fix. Also, defining or executing a keyboard macro outside of Edebug does not affect commands inside Edebug. This is usually an advantage. See also the edebug-continue-kbd-macro option in Edebug Options.

User Option: edebug-sit-for-seconds

This option specifies how many seconds to wait between execution steps in trace mode or continue mode. The default is 1 second.


19.2.4 Jumping

The commands described in this section execute until they reach a specified location. All except i make a temporary breakpoint to establish the place to stop, then switch to go mode. Any other breakpoint reached before the intended stop point will also stop execution. See Edebug Breakpoints, for the details on breakpoints.

These commands may fail to work as expected in case of nonlocal exit, as that can bypass the temporary breakpoint where you expected the program to stop.

h

Proceed to the stop point near where point is (edebug-goto-here).

f

Run the program for one expression (edebug-forward-sexp).

o

Run the program until the end of the containing sexp (edebug-step-out).

i

Step into the function or macro called by the form after point (edebug-step-in).

The h command proceeds to the stop point at or after the current location of point, using a temporary breakpoint.

The f command runs the program forward over one expression. More precisely, it sets a temporary breakpoint at the position that forward-sexp would reach, then executes in go mode so that the program will stop at breakpoints.

With a prefix argument n, the temporary breakpoint is placed n sexps beyond point. If the containing list ends before n more elements, then the place to stop is after the containing expression.

You must check that the position forward-sexp finds is a place that the program will really get to. In cond, for example, this may not be true.

For flexibility, the f command does forward-sexp starting at point, rather than at the stop point. If you want to execute one expression from the current stop point, first type w (edebug-where) to move point there, and then type f.

The o command continues out of an expression. It places a temporary breakpoint at the end of the sexp containing point. If the containing sexp is a function definition itself, o continues until just before the last sexp in the definition. If that is where you are now, it returns from the function and then stops. In other words, this command does not exit the currently executing function unless you are positioned after the last sexp.

Normally, the h, f, and o commands display “Break” and pause for edebug-sit-for-seconds before showing the result of the form just evaluated. You can avoid this pause by setting edebug-sit-on-break to nil. See Edebug Options.

The i command steps into the function or macro called by the list form after point, and stops at its first stop point. Note that the form need not be the one about to be evaluated. But if the form is a function call about to be evaluated, remember to use this command before any of the arguments are evaluated, since otherwise it will be too late.

The i command instruments the function or macro it’s supposed to step into, if it isn’t instrumented already. This is convenient, but keep in mind that the function or macro remains instrumented unless you explicitly arrange to deinstrument it.


19.2.5 Miscellaneous Edebug Commands

Some miscellaneous Edebug commands are described here.

?

Display the help message for Edebug (edebug-help).

a
C-]

Abort one level back to the previous command level (abort-recursive-edit).

q

Return to the top level editor command loop (top-level). This exits all recursive editing levels, including all levels of Edebug activity. However, instrumented code protected with unwind-protect or condition-case forms may resume debugging.

Q

Like q, but don’t stop even for protected code (edebug-top-level-nonstop).

r

Redisplay the most recently known expression result in the echo area (edebug-previous-result).

d

Display a backtrace, excluding Edebug’s own functions for clarity (edebug-pop-to-backtrace).

See Backtraces, for a description of backtraces and the commands which work on them.

If you would like to see Edebug’s functions in the backtrace, use M-x edebug-backtrace-show-instrumentation. To hide them again use M-x edebug-backtrace-hide-instrumentation.

If a backtrace frame starts with ‘>’ that means that Edebug knows where the source code for the frame is located. Use s to jump to the source code for the current frame.

The backtrace buffer is killed automatically when you continue execution.

You can invoke commands from Edebug that activate Edebug again recursively. Whenever Edebug is active, you can quit to the top level with q or abort one recursive edit level with C-]. You can display a backtrace of all the pending evaluations with d.


19.2.6 Breaks

Edebug’s step mode stops execution when the next stop point is reached. There are three other ways to stop Edebug execution once it has started: breakpoints, the global break condition, and source breakpoints.


19.2.6.1 Edebug Breakpoints

While using Edebug, you can specify breakpoints in the program you are testing: these are places where execution should stop. You can set a breakpoint at any stop point, as defined in Using Edebug. For setting and unsetting breakpoints, the stop point that is affected is the first one at or after point in the source code buffer. Here are the Edebug commands for breakpoints:

b

Set a breakpoint at the stop point at or after point (edebug-set-breakpoint). If you use a prefix argument, the breakpoint is temporary—it turns off the first time it stops the program. An overlay with the edebug-enabled-breakpoint or edebug-disabled-breakpoint faces is put at the breakpoint.

u

Unset the breakpoint (if any) at the stop point at or after point (edebug-unset-breakpoint).

U

Unset any breakpoints in the current form (edebug-unset-breakpoints).

D

Toggle whether to disable the breakpoint near point (edebug-toggle-disable-breakpoint). This command is mostly useful if the breakpoint is conditional and it would take some work to recreate the condition.

x condition RET

Set a conditional breakpoint which stops the program only if evaluating condition produces a non-nil value (edebug-set-conditional-breakpoint). With a prefix argument, the breakpoint is temporary.

B

Move point to the next breakpoint in the current definition (edebug-next-breakpoint).

While in Edebug, you can set a breakpoint with b and unset one with u. First move point to the Edebug stop point of your choice, then type b or u to set or unset a breakpoint there. Unsetting a breakpoint where none has been set has no effect.

Re-evaluating or reinstrumenting a definition removes all of its previous breakpoints.

A conditional breakpoint tests a condition each time the program gets there. Any errors that occur as a result of evaluating the condition are ignored, as if the result were nil. To set a conditional breakpoint, use x, and specify the condition expression in the minibuffer. Setting a conditional breakpoint at a stop point that has a previously established conditional breakpoint puts the previous condition expression in the minibuffer so you can edit it.

You can make a conditional or unconditional breakpoint temporary by using a prefix argument with the command to set the breakpoint. When a temporary breakpoint stops the program, it is automatically unset.

Edebug always stops or pauses at a breakpoint, except when the Edebug mode is Go-nonstop. In that mode, it ignores breakpoints entirely.

To find out where your breakpoints are, use the B command, which moves point to the next breakpoint following point, within the same function, or to the first breakpoint if there are no following breakpoints. This command does not continue execution—it just moves point in the buffer.


19.2.6.2 Global Break Condition

A global break condition stops execution when a specified condition is satisfied, no matter where that may occur. Edebug evaluates the global break condition at every stop point; if it evaluates to a non-nil value, then execution stops or pauses depending on the execution mode, as if a breakpoint had been hit. If evaluating the condition gets an error, execution does not stop.

The condition expression is stored in edebug-global-break-condition. You can specify a new expression using the X command from the source code buffer while Edebug is active, or using C-x X X from any buffer at any time, as long as Edebug is loaded (edebug-set-global-break-condition).

The global break condition is the simplest way to find where in your code some event occurs, but it makes code run much more slowly. So you should reset the condition to nil when not using it.


19.2.6.3 Source Breakpoints

All breakpoints in a definition are forgotten each time you reinstrument it. If you wish to make a breakpoint that won’t be forgotten, you can write a source breakpoint, which is simply a call to the function edebug in your source code. You can, of course, make such a call conditional. For example, in the fac function, you can insert the first line as shown below, to stop when the argument reaches zero:

(defun fac (n)
  (if (= n 0) (edebug))
  (if (< 0 n)
      (* n (fac (1- n)))
    1))

When the fac definition is instrumented and the function is called, the call to edebug acts as a breakpoint. Depending on the execution mode, Edebug stops or pauses there.

If no instrumented code is being executed when edebug is called, that function calls debug.


19.2.7 Trapping Errors

Emacs normally displays an error message when an error is signaled and not handled with condition-case. While Edebug is active and executing instrumented code, it normally responds to all unhandled errors. You can customize this with the options edebug-on-error and edebug-on-quit; see Edebug Options.

When Edebug responds to an error, it shows the last stop point encountered before the error. This may be the location of a call to a function which was not instrumented, and within which the error actually occurred. For an unbound variable error, the last known stop point might be quite distant from the offending variable reference. In that case, you might want to display a full backtrace (see Miscellaneous Edebug Commands).

If you change debug-on-error or debug-on-quit while Edebug is active, these changes will be forgotten when Edebug becomes inactive. Furthermore, during Edebug’s recursive edit, these variables are bound to the values they had outside of Edebug.


19.2.8 Edebug Views

These Edebug commands let you view aspects of the buffer and window status as they were before entry to Edebug. The outside window configuration is the collection of windows and contents that were in effect outside of Edebug.

P
v

Switch to viewing the outside window configuration (edebug-view-outside). Type C-x X w to return to Edebug.

p

Temporarily display the outside current buffer with point at its outside position (edebug-bounce-point), pausing for one second before returning to Edebug. With a prefix argument n, pause for n seconds instead.

w

Move point back to the current stop point in the source code buffer (edebug-where).

If you use this command in a different window displaying the same buffer, that window will be used instead to display the current definition in the future.

W

Toggle whether Edebug saves and restores the outside window configuration (edebug-toggle-save-windows).

With a prefix argument, W only toggles saving and restoring of the selected window. To specify a window that is not displaying the source code buffer, you must use C-x X W from the global keymap.

You can view the outside window configuration with v or just bounce to the point in the current buffer with p, even if it is not normally displayed.

After moving point, you may wish to jump back to the stop point. You can do that with w from a source code buffer. You can jump back to the stop point in the source code buffer from any buffer using C-x X w.

Each time you use W to turn saving off, Edebug forgets the saved outside window configuration—so that even if you turn saving back on, the current window configuration remains unchanged when you next exit Edebug (by continuing the program). However, the automatic redisplay of *edebug* and *edebug-trace* may conflict with the buffers you wish to see unless you have enough windows open.


19.2.9 Evaluation

While within Edebug, you can evaluate expressions as if Edebug were not running. Edebug tries to be invisible to the expression’s evaluation and printing. Evaluation of expressions that cause side effects will work as expected, except for changes to data that Edebug explicitly saves and restores. See The Outside Context, for details on this process.

e exp RET

Evaluate expression exp in the context outside of Edebug (edebug-eval-expression). That is, Edebug tries to minimize its interference with the evaluation. The result is shown in the echo area, or, if this command is given a prefix, pop up a new buffer and pretty-print the result there.

By default, this command suppresses the debugger during evaluation, so that an error in the evaluated expression won’t add a new error on top of the existing one. Set the debug-allow-recursive-debug user option to a non-nil value to override this.

M-: exp RET

Evaluate expression exp in the context of Edebug itself (eval-expression).

C-x C-e

Evaluate the expression before point, in the context outside of Edebug (edebug-eval-last-sexp). With the prefix argument of zero (C-u 0 C-x C-e), don’t shorten long items (like strings and lists). Any other prefix will result in the value being pretty-printed in a separate buffer.

Edebug supports evaluation of expressions containing references to lexically bound symbols created by the following constructs in cl.el: lexical-let, macrolet, and symbol-macrolet.


19.2.10 Evaluation List Buffer

You can use the evaluation list buffer, called *edebug*, to evaluate expressions interactively. You can also set up the evaluation list of expressions to be evaluated automatically each time Edebug updates the display.

E

Switch to the evaluation list buffer *edebug* (edebug-visit-eval-list).

In the *edebug* buffer you can use the commands of Lisp Interaction mode (see Lisp Interaction in The GNU Emacs Manual) as well as these special commands:

C-j

Evaluate the expression before point, in the outside context, and insert the value in the buffer (edebug-eval-print-last-sexp). With prefix argument of zero (C-u 0 C-j), don’t shorten long items (like strings and lists).

C-x C-e

Evaluate the expression before point, in the context outside of Edebug (edebug-eval-last-sexp).

C-c C-u

Build a new evaluation list from the contents of the buffer (edebug-update-eval-list).

C-c C-d

Delete the evaluation list group that point is in (edebug-delete-eval-item).

C-c C-w

Switch back to the source code buffer at the current stop point (edebug-where).

You can evaluate expressions in the evaluation list window with C-j or C-x C-e, just as you would in *scratch*; but they are evaluated in the context outside of Edebug.

The expressions you enter interactively (and their results) are lost when you continue execution; but you can set up an evaluation list consisting of expressions to be evaluated each time execution stops.

To do this, write one or more evaluation list groups in the evaluation list buffer. An evaluation list group consists of one or more Lisp expressions. Groups are separated by comment lines.

The command C-c C-u (edebug-update-eval-list) rebuilds the evaluation list, scanning the buffer and using the first expression of each group. (The idea is that the second expression of the group is the value previously computed and displayed.)

Each entry to Edebug redisplays the evaluation list by inserting each expression in the buffer, followed by its current value. It also inserts comment lines so that each expression becomes its own group. Thus, if you type C-c C-u again without changing the buffer text, the evaluation list is effectively unchanged.

If an error occurs during an evaluation from the evaluation list, the error message is displayed in a string as if it were the result. Therefore, expressions using variables that are not currently valid do not interrupt your debugging.

Here is an example of what the evaluation list window looks like after several expressions have been added to it:

(current-buffer)
#<buffer *scratch*>
;---------------------------------------------------------------
(selected-window)
#<window 16 on *scratch*>
;---------------------------------------------------------------
(point)
196
;---------------------------------------------------------------
bad-var
"Symbol's value as variable is void: bad-var"
;---------------------------------------------------------------
(recursion-depth)
0
;---------------------------------------------------------------
this-command
eval-last-sexp
;---------------------------------------------------------------

To delete a group, move point into it and type C-c C-d, or simply delete the text for the group and update the evaluation list with C-c C-u. To add a new expression to the evaluation list, insert the expression at a suitable place, insert a new comment line, then type C-c C-u. You need not insert dashes in the comment line—its contents don’t matter.

After selecting *edebug*, you can return to the source code buffer with C-c C-w. The *edebug* buffer is killed when you continue execution, and recreated next time it is needed.


19.2.11 Printing in Edebug

If an expression in your program produces a value containing circular list structure, you may get an error when Edebug attempts to print it.

One way to cope with circular structure is to set print-length or print-level to truncate the printing. Edebug does this for you; it binds print-length and print-level to the values of the variables edebug-print-length and edebug-print-level (so long as they have non-nil values). See Variables Affecting Output.

User Option: edebug-print-length

If non-nil, Edebug binds print-length to this value while printing results. The default value is 50.

User Option: edebug-print-level

If non-nil, Edebug binds print-level to this value while printing results. The default value is 50.

You can also print circular structures and structures that share elements more informatively by binding print-circle to a non-nil value.

Here is an example of code that creates a circular structure:

(setq a (list 'x 'y))
(setcar a a)

If print-circle is non-nil, printing functions (e.g., prin1) will print a as ‘#1=(#1# y)’. The ‘#1=’ notation labels the structure that follows it with the label ‘1’, and the ‘#1#’ notation references the previously labeled structure. This notation is used for any shared elements of lists or vectors.

User Option: edebug-print-circle

If non-nil, Edebug binds print-circle to this value while printing results. The default value is t.

For further details about how printing can be customized, see see Output Functions.


19.2.12 Trace Buffer

Edebug can record an execution trace, storing it in a buffer named *edebug-trace*. This is a log of function calls and returns, showing the function names and their arguments and values. To enable trace recording, set edebug-trace to a non-nil value.

Making a trace buffer is not the same thing as using trace execution mode (see Edebug Execution Modes).

When trace recording is enabled, each function entry and exit adds lines to the trace buffer. A function entry record consists of ‘::::{’, followed by the function name and argument values. A function exit record consists of ‘::::}’, followed by the function name and result of the function.

The number of ‘:’s in an entry shows its recursion depth. You can use the braces in the trace buffer to find the matching beginning or end of function calls.

You can customize trace recording for function entry and exit by redefining the functions edebug-print-trace-before and edebug-print-trace-after.

Macro: edebug-tracing string body…

This macro requests additional trace information around the execution of the body forms. The argument string specifies text to put in the trace buffer, after the ‘{’ or ‘}’. All the arguments are evaluated, and edebug-tracing returns the value of the last form in body.

Function: edebug-trace format-string &rest format-args

This function inserts text in the trace buffer. It computes the text with (apply 'format format-string format-args). It also appends a newline to separate entries.

edebug-tracing and edebug-trace insert lines in the trace buffer whenever they are called, even if Edebug is not active. Adding text to the trace buffer also scrolls its window to show the last lines inserted.


19.2.13 Coverage Testing

Edebug provides rudimentary coverage testing and display of execution frequency.

Coverage testing works by comparing the result of each expression with the previous result; each form in the program is considered covered if it has returned two different values since you began testing coverage in the current Emacs session. Thus, to do coverage testing on your program, execute it under various conditions and note whether it behaves correctly; Edebug will tell you when you have tried enough different conditions that each form has returned two different values.

Coverage testing makes execution slower, so it is only done if edebug-test-coverage is non-nil. Frequency counting is performed for all executions of an instrumented function, even if the execution mode is Go-nonstop, and regardless of whether coverage testing is enabled.

Use C-x X = (edebug-display-freq-count) to display both the coverage information and the frequency counts for a definition. Just = (edebug-temp-display-freq-count) displays the same information temporarily, only until you type another key.

Command: edebug-display-freq-count

This command displays the frequency count data for each line of the current definition.

It inserts frequency counts as comment lines after each line of code. You can undo all insertions with one undo command. The counts appear under the ‘(’ before an expression or the ‘)’ after an expression, or on the last character of a variable. To simplify the display, a count is not shown if it is equal to the count of an earlier expression on the same line.

The character ‘=’ following the count for an expression says that the expression has returned the same value each time it was evaluated. In other words, it is not yet covered for coverage testing purposes.

To clear the frequency count and coverage data for a definition, simply reinstrument it with eval-defun.

For example, after evaluating (fac 5) with a source breakpoint, and setting edebug-test-coverage to t, when the breakpoint is reached, the frequency data looks like this:

(defun fac (n)
  (if (= n 0) (edebug))
;#6           1      = =5
  (if (< 0 n)
;#5         =
      (* n (fac (1- n)))
;#    5               0
    1))
;#   0

The comment lines show that fac was called 6 times. The first if statement returned 5 times with the same result each time; the same is true of the condition on the second if. The recursive call of fac did not return at all.


19.2.14 The Outside Context

Edebug tries to be transparent to the program you are debugging, but it does not succeed completely. Edebug also tries to be transparent when you evaluate expressions with e or with the evaluation list buffer, by temporarily restoring the outside context. This section explains precisely what context Edebug restores, and how Edebug fails to be completely transparent.


19.2.14.1 Checking Whether to Stop

Whenever Edebug is entered, it needs to save and restore certain data before even deciding whether to make trace information or stop the program.

  • max-lisp-eval-depth (see Eval) is increased to reduce Edebug’s impact on the stack. You could, however, still run out of stack space when using Edebug. You can also enlarge the value of edebug-max-depth if Edebug reaches the limit of recursion depth instrumenting code that contains very large quoted lists.
  • The state of keyboard macro execution is saved and restored. While Edebug is active, executing-kbd-macro is bound to nil unless edebug-continue-kbd-macro is non-nil.

19.2.14.2 Edebug Display Update

When Edebug needs to display something (e.g., in trace mode), it saves the current window configuration from outside Edebug (see Window Configurations). When you exit Edebug, it restores the previous window configuration.

Emacs redisplays only when it pauses. Usually, when you continue execution, the program re-enters Edebug at a breakpoint or after stepping, without pausing or reading input in between. In such cases, Emacs never gets a chance to redisplay the outside configuration. Consequently, what you see is the same window configuration as the last time Edebug was active, with no interruption.

Entry to Edebug for displaying something also saves and restores the following data (though some of them are deliberately not restored if an error or quit signal occurs).

  • Which buffer is current, and the positions of point and the mark in the current buffer, are saved and restored.
  • The outside window configuration is saved and restored if edebug-save-windows is non-nil (see Edebug Options). If the value of edebug-save-windows is a list, only the listed windows are saved and restored.

    The window configuration is not restored on error or quit, but the outside selected window is reselected even on error or quit in case a save-excursion is active.

    The window start and horizontal scrolling of the source code buffer are not restored, however, so that the display remains coherent within Edebug.

    Saving and restoring the outside window configuration can sometimes change the positions of point in the buffers on which the Lisp program you are debugging operates, especially if your program moves point. If this happens and interferes with your debugging, we recommend to set edebug-save-windows to nil (see Edebug Options).

  • The value of point in each displayed buffer is saved and restored if edebug-save-displayed-buffer-points is non-nil.
  • The variables overlay-arrow-position and overlay-arrow-string are saved and restored, so you can safely invoke Edebug from the recursive edit elsewhere in the same buffer.
  • cursor-in-echo-area is locally bound to nil so that the cursor shows up in the window.

19.2.14.3 Edebug Recursive Edit

When Edebug is entered and actually reads commands from the user, it saves (and later restores) these additional data:

  • The current match data. See The Match Data.
  • The variables last-command, this-command, last-command-event, last-input-event, last-event-frame, last-nonmenu-event, and track-mouse. Commands in Edebug do not affect these variables outside of Edebug.

    Executing commands within Edebug can change the key sequence that would be returned by this-command-keys, and there is no way to reset the key sequence from Lisp.

    Edebug cannot save and restore the value of unread-command-events. Entering Edebug while this variable has a nontrivial value can interfere with execution of the program you are debugging.

  • Complex commands executed while in Edebug are added to the variable command-history. In rare cases this can alter execution.
  • Within Edebug, the recursion depth appears one deeper than the recursion depth outside Edebug. This is not true of the automatically updated evaluation list window.
  • standard-output and standard-input are bound to nil by the recursive-edit, but Edebug temporarily restores them during evaluations.
  • The state of keyboard macro definition is saved and restored. While Edebug is active, defining-kbd-macro is bound to edebug-continue-kbd-macro.

19.2.15 Edebug and Macros

To make Edebug properly instrument expressions that call macros, some extra care is needed. This subsection explains the details.


19.2.15.1 Instrumenting Macro Calls

When Edebug instruments an expression that calls a Lisp macro, it needs additional information about the macro to do the job properly. This is because there is no a-priori way to tell which subexpressions of the macro call are forms to be evaluated. (Evaluation may occur explicitly in the macro body, or when the resulting expansion is evaluated, or any time later.)

Therefore, you must define an Edebug specification for each macro that Edebug will encounter, to explain the format of calls to that macro. To do this, add a debug declaration to the macro definition. Here is a simple example that shows the specification for the for example macro (see Evaluating Macro Arguments Repeatedly).

(defmacro for (var from init to final do &rest body)
  "Execute a simple \"for\" loop.
For example, (for i from 1 to 10 do (print i))."
  (declare (debug (symbolp "from" form "to" form "do" &rest form)))
  ...)

The Edebug specification says which parts of a call to the macro are forms to be evaluated. For simple macros, the specification often looks very similar to the formal argument list of the macro definition, but specifications are much more general than macro arguments. See Defining Macros, for more explanation of the declare form.

Take care to ensure that the specifications are known to Edebug when you instrument code. If you are instrumenting a function which uses a macro defined in another file, you may first need to either evaluate the require forms in the file containing your function, or explicitly load the file containing the macro. If the definition of a macro is wrapped by eval-when-compile, you may need to evaluate it.

You can also define an edebug specification for a macro separately from the macro definition with def-edebug-spec. Adding debug declarations is preferred, and more convenient, for macro definitions in Lisp, but def-edebug-spec makes it possible to define Edebug specifications for special forms implemented in C.

Macro: def-edebug-spec macro specification

Specify which expressions of a call to macro macro are forms to be evaluated. specification should be the Edebug specification. Neither argument is evaluated.

The macro argument can actually be any symbol, not just a macro name.

Here is a table of the possibilities for specification and how each directs processing of arguments.

t

All arguments are instrumented for evaluation. This is short for (body).

a symbol

The symbol must have an Edebug specification, which is used instead. This indirection is repeated until another kind of specification is found. This allows you to inherit the specification from another macro.

a list

The elements of the list describe the types of the arguments of a calling form. The possible elements of a specification list are described in the following sections.

If a macro has no Edebug specification, neither through a debug declaration nor through a def-edebug-spec call, the variable edebug-eval-macro-args comes into play.

User Option: edebug-eval-macro-args

This controls the way Edebug treats macro arguments with no explicit Edebug specification. If it is nil (the default), none of the arguments is instrumented for evaluation. Otherwise, all arguments are instrumented.


19.2.15.2 Specification List

A specification list is required for an Edebug specification if some arguments of a macro call are evaluated while others are not. Some elements in a specification list match one or more arguments, but others modify the processing of all following elements. The latter, called specification keywords, are symbols beginning with ‘&’ (such as &optional).

A specification list may contain sublists, which match arguments that are themselves lists, or it may contain vectors used for grouping. Sublists and groups thus subdivide the specification list into a hierarchy of levels. Specification keywords apply only to the remainder of the sublist or group they are contained in.

When a specification list involves alternatives or repetition, matching it against an actual macro call may require backtracking. For more details, see Backtracking in Specifications.

Edebug specifications provide the power of regular expression matching, plus some context-free grammar constructs: the matching of sublists with balanced parentheses, recursive processing of forms, and recursion via indirect specifications.

Here’s a table of the possible elements of a specification list, with their meanings (see Specification Examples, for the referenced examples):

sexp

A single unevaluated Lisp object, which is not instrumented.

form

A single evaluated expression, which is instrumented. If your macro wraps the expression with lambda before it is evaluated, use def-form instead. See def-form below.

place

A generalized variable. See Generalized Variables.

body

Short for &rest form. See &rest below. If your macro wraps its body of code with lambda before it is evaluated, use def-body instead. See def-body below.

lambda-expr

A lambda expression with no quoting.

&optional

All following elements in the specification list are optional; as soon as one does not match, Edebug stops matching at this level.

To make just a few elements optional, followed by non-optional elements, use [&optional specs…]. To specify that several elements must all match or none, use &optional [specs…]. See the defun example.

&rest

All following elements in the specification list are repeated zero or more times. In the last repetition, however, it is not a problem if the expression runs out before matching all of the elements of the specification list.

To repeat only a few elements, use [&rest specs…]. To specify several elements that must all match on every repetition, use &rest [specs…].

&or

Each of the following elements in the specification list is an alternative. One of the alternatives must match, or the &or specification fails.

Each list element following &or is a single alternative. To group two or more list elements as a single alternative, enclose them in […].

&not

Each of the following elements is matched as alternatives as if by using &or, but if any of them match, the specification fails. If none of them match, nothing is matched, but the &not specification succeeds.

&define

Indicates that the specification is for a defining form. Edebug’s definition of a defining form is a form containing one or more code forms which are saved and executed later, after the execution of the defining form.

The defining form itself is not instrumented (that is, Edebug does not stop before and after the defining form), but forms inside it typically will be instrumented. The &define keyword should be the first element in a list specification.

nil

This is successful when there are no more arguments to match at the current argument list level; otherwise it fails. See sublist specifications and the backquote example.

gate

No argument is matched but backtracking through the gate is disabled while matching the remainder of the specifications at this level. This is primarily used to generate more specific syntax error messages. See Backtracking in Specifications, for more details. Also see the let example.

&error

&error should be followed by a string, an error message, in the edebug-spec; it aborts the instrumentation, displaying the message in the minibuffer.

&interpose

Lets a function control the parsing of the remaining code. It takes the form &interpose spec fun args... and means that Edebug will first match spec against the code and then call fun with the code that matched spec, a parsing function pf, and finally args.... The parsing function expects a single argument indicating the specification list to use to parse the remaining code. It should be called exactly once and returns the instrumented code that fun is expected to return. For example (&interpose symbolp pcase--match-pat-args) matches sexps whose first element is a symbol and then lets pcase--match-pat-args lookup the specs associated with that head symbol according to pcase--match-pat-args and pass them to the pf it received as argument.

other-symbol

Any other symbol in a specification list may be a predicate or an indirect specification.

If the symbol has an Edebug specification, this indirect specification should be either a list specification that is used in place of the symbol, or a function that is called to process the arguments. The specification may be defined with def-edebug-elem-spec:

Function: def-edebug-elem-spec element specification

Define the specification to use in place of the symbol element. specification has to be a list.

Otherwise, the symbol should be a predicate. The predicate is called with the argument, and if the predicate returns nil, the specification fails and the argument is not instrumented.

Some suitable predicates include symbolp, integerp, stringp, vectorp, and atom.

[elements…]

A vector of elements groups the elements into a single group specification. Its meaning has nothing to do with vectors.

"string"

The argument should be a symbol named string. This specification is equivalent to the quoted symbol, 'symbol, where the name of symbol is the string, but the string form is preferred.

(vector elements…)

The argument should be a vector whose elements must match the elements in the specification. See the backquote example.

(elements…)

Any other list is a sublist specification and the argument must be a list whose elements match the specification elements.

A sublist specification may be a dotted list and the corresponding list argument may then be a dotted list. Alternatively, the last CDR of a dotted list specification may be another sublist specification (via a grouping or an indirect specification, e.g., (spec . [(more specs…)])) whose elements match the non-dotted list arguments. This is useful in recursive specifications such as in the backquote example. Also see the description of a nil specification above for terminating such recursion.

Note that a sublist specification written as (specs . nil) is equivalent to (specs), and (specs . (sublist-elements…)) is equivalent to (specs sublist-elements…).

Here is a list of additional specifications that may appear only after &define. See the defun example.

&name

Extracts the name of the current defining form from the code. It takes the form &name [prestring] spec [poststring] fun args... and means that Edebug will match spec against the code and then call fun with the concatenation of the current name, args..., prestring, the code that matched spec, and poststring. If fun is absent, it defaults to a function that concatenates the arguments (with an @ between the previous name and the new).

name

The argument, a symbol, is the name of the defining form. Shorthand for [&name symbolp].

A defining form is not required to have a name field; and it may have multiple name fields.

arg

The argument, a symbol, is the name of an argument of the defining form. However, lambda-list keywords (symbols starting with ‘&’) are not allowed.

lambda-list

This matches a lambda list—the argument list of a lambda expression.

def-body

The argument is the body of code in a definition. This is like body, described above, but a definition body must be instrumented with a different Edebug call that looks up information associated with the definition. Use def-body for the highest level list of forms within the definition.

def-form

The argument is a single, highest-level form in a definition. This is like def-body, except it is used to match a single form rather than a list of forms. As a special case, def-form also means that tracing information is not output when the form is executed. See the interactive example.


19.2.15.3 Backtracking in Specifications

If a specification fails to match at some point, this does not necessarily mean a syntax error will be signaled; instead, backtracking will take place until all alternatives have been exhausted. Eventually every element of the argument list must be matched by some element in the specification, and every required element in the specification must match some argument.

When a syntax error is detected, it might not be reported until much later, after higher-level alternatives have been exhausted, and with the point positioned further from the real error. But if backtracking is disabled when an error occurs, it can be reported immediately. Note that backtracking is also reenabled automatically in several situations; when a new alternative is established by &optional, &rest, or &or, or at the start of processing a sublist, group, or indirect specification. The effect of enabling or disabling backtracking is limited to the remainder of the level currently being processed and lower levels.

Backtracking is disabled while matching any of the form specifications (that is, form, body, def-form, and def-body). These specifications will match any form so any error must be in the form itself rather than at a higher level.

Backtracking is also disabled after successfully matching a quoted symbol, string specification, or &define keyword, since this usually indicates a recognized construct. But if you have a set of alternative constructs that all begin with the same symbol, you can usually work around this constraint by factoring the symbol out of the alternatives, e.g., ["foo" &or [first case] [second case] ...].

Most needs are satisfied by these two ways that backtracking is automatically disabled, but occasionally it is useful to explicitly disable backtracking by using the gate specification. This is useful when you know that no higher alternatives could apply. See the example of the let specification.


19.2.15.4 Specification Examples

It may be easier to understand Edebug specifications by studying the examples provided here.

Consider a hypothetical macro my-test-generator that runs tests on supplied lists of data. Although it is Edebug’s default behavior to not instrument arguments as code, as controlled by edebug-eval-macro-args (see Instrumenting Macro Calls), it can be useful to explicitly document that the arguments are data:

(def-edebug-spec my-test-generator (&rest sexp))

A let special form has a sequence of bindings and a body. Each of the bindings is either a symbol or a sublist with a symbol and optional expression. In the specification below, notice the gate inside of the sublist to prevent backtracking once a sublist is found.

(def-edebug-spec let
  ((&rest
    &or symbolp (gate symbolp &optional form))
   body))

Edebug uses the following specifications for defun and the associated argument list and interactive specifications. It is necessary to handle interactive forms specially since an expression argument is actually evaluated outside of the function body. (The specification for defmacro is very similar to that for defun, but allows for the declare statement.)

(def-edebug-spec defun
  (&define name lambda-list
           [&optional stringp]   ; Match the doc string, if present.
           [&optional ("interactive" interactive)]
           def-body))

(def-edebug-elem-spec 'lambda-list
  '(([&rest arg]
     [&optional ["&optional" arg &rest arg]]
     &optional ["&rest" arg]
     )))

(def-edebug-elem-spec 'interactive
  '(&optional &or stringp def-form))    ; Notice: def-form

The specification for backquote below illustrates how to match dotted lists and use nil to terminate recursion. It also illustrates how components of a vector may be matched. (The actual specification defined by Edebug is a little different, and does not support dotted lists because doing so causes very deep recursion that could fail.)

(def-edebug-spec \` (backquote-form))   ; Alias just for clarity.

(def-edebug-elem-spec 'backquote-form
  '(&or ([&or "," ",@"] &or ("quote" backquote-form) form)
        (backquote-form . [&or nil backquote-form])
        (vector &rest backquote-form)
        sexp))

19.2.16 Edebug Options

These options affect the behavior of Edebug:

User Option: edebug-setup-hook

Functions to call before Edebug is used. Each time it is set to a new value, Edebug will call those functions once and then reset edebug-setup-hook to nil. You could use this to load up Edebug specifications associated with a package you are using, but only when you also use Edebug. See Instrumenting for Edebug.

User Option: edebug-all-defs

If this is non-nil, normal evaluation of defining forms such as defun and defmacro instruments them for Edebug. This applies to eval-defun, eval-region and eval-buffer.

Use the command M-x edebug-all-defs to toggle the value of this option. See Instrumenting for Edebug.

User Option: edebug-all-forms

If this is non-nil, the commands eval-defun, eval-region and eval-buffer instrument all forms, even those that don’t define anything. This doesn’t apply to loading or evaluations in the minibuffer.

Use the command M-x edebug-all-forms to toggle the value of this option. See Instrumenting for Edebug.

User Option: edebug-eval-macro-args

When this is non-nil, all macro arguments will be instrumented in the generated code. For any macro, the debug declaration overrides this option. So to specify exceptions for macros that have some arguments evaluated and some not, use the debug declaration specify an Edebug form specification.

User Option: edebug-save-windows

If this is non-nil, Edebug saves and restores the window configuration. That takes some time, so if your program does not care what happens to the window configurations, it is better to set this variable to nil. We also recommend to set this to nil if the default value causes Edebug to overwrite the positions of point in buffers that are involved in the program you are debugging, as result of saving and restoring the window configuration; this could happen if your program moves point in one or more of those buffers. Another option to try to customize in this case is edebug-save-displayed-buffer-points, described below.

If the value of edebug-save-windows is a list, only the listed windows are saved and restored.

You can use the W command in Edebug to change this variable interactively. See Edebug Display Update.

User Option: edebug-save-displayed-buffer-points

If this is non-nil, Edebug saves and restores point in all displayed buffers.

Saving and restoring point in other buffers is necessary if you are debugging code that changes the point of a buffer that is displayed in a non-selected window. If Edebug or the user then selects the window, point in that buffer will move to the window’s value of point.

Saving and restoring point in all buffers is expensive, since it requires selecting each window twice, so enable this only if you need it. See Edebug Display Update.

User Option: edebug-initial-mode

If this variable is non-nil, it specifies the initial execution mode for Edebug when it is first activated. Possible values are step, next, go, Go-nonstop, trace, Trace-fast, continue, and Continue-fast.

The default value is step. This variable can be set interactively with C-x C-a C-m (edebug-set-initial-mode). See Edebug Execution Modes.

User Option: edebug-trace

If this is non-nil, trace each function entry and exit. Tracing output is displayed in a buffer named *edebug-trace*, one function entry or exit per line, indented by the recursion level.

Also see edebug-tracing, in Trace Buffer.

User Option: edebug-test-coverage

If non-nil, Edebug tests coverage of all expressions debugged. See Coverage Testing.

User Option: edebug-continue-kbd-macro

If non-nil, continue defining or executing any keyboard macro that is executing outside of Edebug. Use this with caution since it is not debugged. See Edebug Execution Modes.

User Option: edebug-print-length

If non-nil, the default value of print-length for printing results in Edebug. See Variables Affecting Output.

User Option: edebug-print-level

If non-nil, the default value of print-level for printing results in Edebug. See Variables Affecting Output.

User Option: edebug-print-circle

If non-nil, the default value of print-circle for printing results in Edebug. See Variables Affecting Output.

User Option: edebug-unwrap-results

If non-nil, Edebug tries to remove any of its own instrumentation when showing the results of expressions. This is relevant when debugging macros where the results of expressions are themselves instrumented expressions. As a very artificial example, suppose that the example function fac has been instrumented, and consider a macro of the form:

(defmacro test () "Edebug example."
  (if (symbol-function 'fac)
      …))

If you instrument the test macro and step through it, then by default the result of the symbol-function call has numerous edebug-after and edebug-before forms, which can make it difficult to see the actual result. If edebug-unwrap-results is non-nil, Edebug tries to remove these forms from the result.

User Option: edebug-on-error

Edebug binds debug-on-error to this value, if debug-on-error was previously nil. See Trapping Errors.

User Option: edebug-on-quit

Edebug binds debug-on-quit to this value, if debug-on-quit was previously nil. See Trapping Errors.

If you change the values of edebug-on-error or edebug-on-quit while Edebug is active, their values won’t be used until the next time Edebug is invoked via a new command.

User Option: edebug-global-break-condition

If non-nil, an expression to test for at every stop point. If the result is non-nil, then break. Errors are ignored. See Global Break Condition.

User Option: edebug-sit-for-seconds

Number of seconds to pause when a breakpoint is reached and the execution mode is trace or continue. See Edebug Execution Modes.

User Option: edebug-sit-on-break

Whether or not to pause for edebug-sit-for-seconds on reaching a breakpoint. Set to nil to prevent the pause, non-nil to allow it.

User Option: edebug-behavior-alist

By default, this alist contains one entry with the key edebug and a list of three functions, which are the default implementations of the functions inserted in instrumented code: edebug-enter, edebug-before and edebug-after. To change Edebug’s behavior globally, modify the default entry.

Edebug’s behavior may also be changed on a per-definition basis by adding an entry to this alist, with a key of your choice and three functions. Then set the edebug-behavior symbol property of an instrumented definition to the key of the new entry, and Edebug will call the new functions in place of its own for that definition.

User Option: edebug-new-definition-function

A function run by Edebug after it wraps the body of a definition or closure. After Edebug has initialized its own data, this function is called with one argument, the symbol associated with the definition, which may be the actual symbol defined or one generated by Edebug. This function may be used to set the edebug-behavior symbol property of each definition instrumented by Edebug.

User Option: edebug-after-instrumentation-function

To inspect or modify Edebug’s instrumentation before it is used, set this variable to a function which takes one argument, an instrumented top-level form, and returns either the same or a replacement form, which Edebug will then use as the final result of instrumentation.


19.3 Debugging Invalid Lisp Syntax

The Lisp reader reports invalid syntax, but cannot say where the real problem is. For example, the error ‘End of file during parsing’ in evaluating an expression indicates an excess of open parentheses (or square brackets). The reader detects this imbalance at the end of the file, but it cannot figure out where the close parenthesis should have been. Likewise, ‘Invalid read syntax: ")"’ indicates an excess close parenthesis or missing open parenthesis, but does not say where the missing parenthesis belongs. How, then, to find what to change?

If the problem is not simply an imbalance of parentheses, a useful technique is to try C-M-e (end-of-defun, see Moving by Defuns in The GNU Emacs Manual) at the beginning of each defun, and see if it goes to the place where that defun appears to end. If it does not, there is a problem in that defun.

However, unmatched parentheses are the most common syntax errors in Lisp, and we can give further advice for those cases. (In addition, just moving point through the code with Show Paren mode enabled might find the mismatch.)


19.3.1 Excess Open Parentheses

The first step is to find the defun that is unbalanced. If there is an excess open parenthesis, the way to do this is to go to the end of the file and type C-u C-M-u (backward-up-list, see Moving by Parens in The GNU Emacs Manual). This will move you to the beginning of the first defun that is unbalanced.

The next step is to determine precisely what is wrong. There is no way to be sure of this except by studying the program, but often the existing indentation is a clue to where the parentheses should have been. The easiest way to use this clue is to reindent with C-M-q (indent-pp-sexp, see Multi-line Indent in The GNU Emacs Manual) and see what moves. But don’t do this yet! Keep reading, first.

Before you do this, make sure the defun has enough close parentheses. Otherwise, C-M-q will get an error, or will reindent all the rest of the file until the end. So move to the end of the defun and insert a close parenthesis there. Don’t use C-M-e (end-of-defun) to move there, since that too will fail to work until the defun is balanced.

Now you can go to the beginning of the defun and type C-M-q. Usually all the lines from a certain point to the end of the function will shift to the right. There is probably a missing close parenthesis, or a superfluous open parenthesis, near that point. (However, don’t assume this is true; study the code to make sure.) Once you have found the discrepancy, undo the C-M-q with C-_ (undo), since the old indentation is probably appropriate to the intended parentheses.

After you think you have fixed the problem, use C-M-q again. If the old indentation actually fit the intended nesting of parentheses, and you have put back those parentheses, C-M-q should not change anything.


19.3.2 Excess Close Parentheses

To deal with an excess close parenthesis, first go to the beginning of the file, then type C-u -1 C-M-u (backward-up-list with an argument of −1) to find the end of the first unbalanced defun.

Then find the actual matching close parenthesis by typing C-M-f (forward-sexp, see Expressions in The GNU Emacs Manual) at the beginning of that defun. This will leave you somewhere short of the place where the defun ought to end. It is possible that you will find a spurious close parenthesis in that vicinity.

If you don’t see a problem at that point, the next thing to do is to type C-M-q (indent-pp-sexp) at the beginning of the defun. A range of lines will probably shift left; if so, the missing open parenthesis or spurious close parenthesis is probably near the first of those lines. (However, don’t assume this is true; study the code to make sure.) Once you have found the discrepancy, undo the C-M-q with C-_ (undo), since the old indentation is probably appropriate to the intended parentheses.

After you think you have fixed the problem, use C-M-q again. If the old indentation actually fits the intended nesting of parentheses, and you have put back those parentheses, C-M-q should not change anything.


19.4 Test Coverage

You can do coverage testing for a file of Lisp code by loading the testcover library and using the command M-x testcover-start RET file RET to instrument the code. Then test your code by calling it one or more times. Then use the command M-x testcover-mark-all to display colored highlights on the code to show where coverage is insufficient. The command M-x testcover-next-mark will move point forward to the next highlighted spot.

Normally, a red highlight indicates the form was never completely evaluated; a brown highlight means it always evaluated to the same value (meaning there has been little testing of what is done with the result). However, the red highlight is skipped for forms that can’t possibly complete their evaluation, such as error. The brown highlight is skipped for forms that are expected to always evaluate to the same value, such as (setq x 14).

For difficult cases, you can add do-nothing macros to your code to give advice to the test coverage tool.

Macro: 1value form

Evaluate form and return its value, but inform coverage testing that form’s value should always be the same.

Macro: noreturn form

Evaluate form, informing coverage testing that form should never return. If it ever does return, you get a run-time error.

Edebug also has a coverage testing feature (see Coverage Testing). These features partly duplicate each other, and it would be cleaner to combine them.


19.5 Profiling

If your program is working correctly, but not fast enough, and you want to make it run more quickly or efficiently, the first thing to do is profile your code so that you know where it spends most of the execution time. If you find that one particular function is responsible for a significant portion of the execution time, you can start looking for ways to optimize that piece.

Emacs has built-in support for this. To begin profiling, type M-x profiler-start. You can choose to sample CPU usage periodically (cpu), when memory is allocated (memory), or both. Then run the code you’d like to speed up. After that, type M-x profiler-report to display a summary buffer for CPU usage sampled by each type (cpu and memory) that you chose to profile. The names of the report buffers include the times at which the reports were generated, so you can generate another report later on without erasing previous results. When you have finished profiling, type M-x profiler-stop (there is a small overhead associated with profiling, so we don’t recommend leaving it active except when you are actually running the code you want to examine).

The profiler report buffer shows, on each line, a function that was called, preceded by how much CPU resources it used in absolute and percentage terms since profiling started. If a given line has a ‘+’ symbol to the left of the function name, you can expand that line by typing RET, in order to see the function(s) called by the higher-level function. Use a prefix argument (C-u RET) to see the whole call tree below a function. Pressing RET again will collapse back to the original state.

Press j or mouse-2 to jump to the definition of a function at point. Press d to view a function’s documentation. You can save a profile to a file using C-x C-w. You can compare two profiles using =.

The elp library offers an alternative approach, which is useful when you know in advance which Lisp function(s) you want to profile. Using that library, you begin by setting elp-function-list to the list of function symbols—those are the functions you want to profile. Then type M-x elp-instrument-list RET nil RET to arrange for profiling those functions. After running the code you want to profile, invoke M-x elp-results to display the current results. See the file elp.el for more detailed instructions. This approach is limited to profiling functions written in Lisp, it cannot profile Emacs primitives.

You can measure the time it takes to evaluate individual Emacs Lisp forms using the benchmark library. See the function benchmark-call as well as the macros benchmark-run, benchmark-run-compiled, and benchmark-progn in benchmark.el. You can also use the benchmark command for timing forms interactively.

To profile Emacs at the level of its C code, you can build it using the --enable-profiling option of configure. When Emacs exits, it generates a file gmon.out that you can examine using the gprof utility. This feature is mainly useful for debugging Emacs. It actually stops the Lisp-level M-x profiler-… commands described above from working.


20 Reading and Printing Lisp Objects

Printing and reading are the operations of converting Lisp objects to textual form and vice versa. They use the printed representations and read syntax described in Lisp Data Types.

This chapter describes the Lisp functions for reading and printing. It also describes streams, which specify where to get the text (if reading) or where to put it (if printing).


20.1 Introduction to Reading and Printing

Reading a Lisp object means parsing a Lisp expression in textual form and producing a corresponding Lisp object. This is how Lisp programs get into Lisp from files of Lisp code. We call the text the read syntax of the object. For example, the text ‘(a . 5)’ is the read syntax for a cons cell whose CAR is a and whose CDR is the number 5.

Printing a Lisp object means producing text that represents that object—converting the object to its printed representation (see Printed Representation and Read Syntax). Printing the cons cell described above produces the text ‘(a . 5)’.

Reading and printing are more or less inverse operations: printing the object that results from reading a given piece of text often produces the same text, and reading the text that results from printing an object usually produces a similar-looking object. For example, printing the symbol foo produces the text ‘foo’, and reading that text returns the symbol foo. Printing a list whose elements are a and b produces the text ‘(a b)’, and reading that text produces a list (but not the same list) with elements a and b.

However, these two operations are not precisely inverse to each other. There are three kinds of exceptions:

  • Printing can produce text that cannot be read. For example, buffers, windows, frames, subprocesses and markers print as text that starts with ‘#’; if you try to read this text, you get an error. There is no way to read those data types.
  • One object can have multiple textual representations. For example, ‘1’ and ‘01’ represent the same integer, and ‘(a b)’ and ‘(a . (b))’ represent the same list. Reading will accept any of the alternatives, but printing must choose one of them.
  • Comments can appear at certain points in the middle of an object’s read sequence without affecting the result of reading it.

20.2 Input Streams

Most of the Lisp functions for reading text take an input stream as an argument. The input stream specifies where or how to get the characters of the text to be read. Here are the possible types of input stream:

buffer

The input characters are read from buffer, starting with the character directly after point. Point advances as characters are read.

marker

The input characters are read from the buffer that marker is in, starting with the character directly after the marker. The marker position advances as characters are read. The value of point in the buffer has no effect when the stream is a marker.

string

The input characters are taken from string, starting at the first character in the string and using as many characters as required.

function

The input characters are generated by function, which must support two kinds of calls:

  • When it is called with no arguments, it should return the next character.
  • When it is called with one argument (always a character), function should save the argument and arrange to return it on the next call. This is called unreading the character; it happens when the Lisp reader reads one character too many and wants to put it back where it came from. In this case, it makes no difference what value function returns.
t

t used as a stream means that the input is read from the minibuffer. In fact, the minibuffer is invoked once and the text given by the user is made into a string that is then used as the input stream. If Emacs is running in batch mode (see Batch Mode), standard input is used instead of the minibuffer. For example,

(message "%s" (read t))

will in batch mode read a Lisp expression from standard input and print the result to standard output.

nil

nil supplied as an input stream means to use the value of standard-input instead; that value is the default input stream, and must be a non-nil input stream.

symbol

A symbol as input stream is equivalent to the symbol’s function definition (if any).

Here is an example of reading from a stream that is a buffer, showing where point is located before and after:

---------- Buffer: foo ----------
This∗ is the contents of foo.
---------- Buffer: foo ----------

(read (get-buffer "foo"))
     ⇒ is
(read (get-buffer "foo"))
     ⇒ the

---------- Buffer: foo ----------
This is the∗ contents of foo.
---------- Buffer: foo ----------

Note that the first read skips a space. Reading skips any amount of whitespace preceding the significant text.

Here is an example of reading from a stream that is a marker, initially positioned at the beginning of the buffer shown. The value read is the symbol This.


---------- Buffer: foo ----------
This is the contents of foo.
---------- Buffer: foo ----------

(setq m (set-marker (make-marker) 1 (get-buffer "foo")))
     ⇒ #<marker at 1 in foo>
(read m)
     ⇒ This
m
     ⇒ #<marker at 5 in foo>   ;; Before the first space.

Here we read from the contents of a string:

(read "(When in) the course")
     ⇒ (When in)

The following example reads from the minibuffer. The prompt is: ‘Lisp expression: . (That is always the prompt used when you read from the stream t.) The user’s input is shown following the prompt.

(read t)
     ⇒ 23
---------- Buffer: Minibuffer ----------
Lisp expression: 23 RET
---------- Buffer: Minibuffer ----------

Finally, here is an example of a stream that is a function, named useless-stream. Before we use the stream, we initialize the variable useless-list to a list of characters. Then each call to the function useless-stream obtains the next character in the list or unreads a character by adding it to the front of the list.

(setq useless-list (append "XY()" nil))
     ⇒ (88 89 40 41)

(defun useless-stream (&optional unread)
  (if unread
      (setq useless-list (cons unread useless-list))
    (prog1 (car useless-list)
           (setq useless-list (cdr useless-list)))))
     ⇒ useless-stream

Now we read using the stream thus constructed:

(read 'useless-stream)
     ⇒ XY

useless-list
     ⇒ (40 41)

Note that the open and close parentheses remain in the list. The Lisp reader encountered the open parenthesis, decided that it ended the input, and unread it. Another attempt to read from the stream at this point would read ‘()’ and return nil.


20.3 Input Functions

This section describes the Lisp functions and variables that pertain to reading.

In the functions below, stream stands for an input stream (see the previous section). If stream is nil or omitted, it defaults to the value of standard-input.

An end-of-file error is signaled if reading encounters an unterminated list, vector, or string.

Function: read &optional stream

This function reads one textual Lisp expression from stream, returning it as a Lisp object. This is the basic Lisp input function.

Function: read-from-string string &optional start end

This function reads the first textual Lisp expression from the text in string. It returns a cons cell whose CAR is that expression, and whose CDR is an integer giving the position of the next remaining character in the string (i.e., the first one not read).

If start is supplied, then reading begins at index start in the string (where the first character is at index 0). If you specify end, then reading is forced to stop just before that index, as if the rest of the string were not there.

For example:

(read-from-string "(setq x 55) (setq y 5)")
     ⇒ ((setq x 55) . 11)
(read-from-string "\"A short string\"")
     ⇒ ("A short string" . 16)

;; Read starting at the first character.
(read-from-string "(list 112)" 0)
     ⇒ ((list 112) . 10)
;; Read starting at the second character.
(read-from-string "(list 112)" 1)
     ⇒ (list . 5)
;; Read starting at the seventh character,
;;   and stopping at the ninth.
(read-from-string "(list 112)" 6 8)
     ⇒ (11 . 8)
Function: read-positioning-symbols &optional stream

This function reads one textual expression from stream, like read does, but additionally positions the read symbols to the positions in stream where they occurred. Only the symbol nil is not positioned, this for efficiency reasons. See Symbols with Position. This function is used by the byte compiler.

Variable: standard-input

This variable holds the default input stream—the stream that read uses when the stream argument is nil. The default is t, meaning use the minibuffer.

Variable: read-circle

If non-nil, this variable enables the reading of circular and shared structures. See Read Syntax for Circular Objects. Its default value is t.

When reading or writing from the standard input/output streams of the Emacs process in batch mode, it is sometimes required to make sure any arbitrary binary data will be read/written verbatim, and/or that no translation of newlines to or from CR-LF pairs is performed. This issue does not exist on POSIX hosts, only on MS-Windows and MS-DOS. The following function allows you to control the I/O mode of any standard stream of the Emacs process.

Function: set-binary-mode stream mode

Switch stream into binary or text I/O mode. If mode is non-nil, switch to binary mode, otherwise switch to text mode. The value of stream can be one of stdin, stdout, or stderr. This function flushes any pending output data of stream as a side effect, and returns the previous value of I/O mode for stream. On POSIX hosts, it always returns a non-nil value and does nothing except flushing pending output.

Function: readablep object

This predicate says whether object has readable syntax, i.e., it can be written out and then read back by the Emacs Lisp reader. If it can’t, this function returns nil; if it can, this function returns a printed representation (via prin1, see Output Functions) of object.


20.4 Output Streams

An output stream specifies what to do with the characters produced by printing. Most print functions accept an output stream as an optional argument. Here are the possible types of output stream:

buffer

The output characters are inserted into buffer at point. Point advances as characters are inserted.

marker

The output characters are inserted into the buffer that marker points into, at the marker position. The marker position advances as characters are inserted. The value of point in the buffer has no effect on printing when the stream is a marker, and this kind of printing does not move point (except that if the marker points at or before the position of point, point advances with the surrounding text, as usual).

function

The output characters are passed to function, which is responsible for storing them away. It is called with a single character as argument, as many times as there are characters to be output, and is responsible for storing the characters wherever you want to put them.

t

The output characters are displayed in the echo area. If Emacs is running in batch mode (see Batch Mode), the output is written to the standard output descriptor instead.

nil

nil specified as an output stream means to use the value of the standard-output variable instead; that value is the default output stream, and must not be nil.

symbol

A symbol as output stream is equivalent to the symbol’s function definition (if any).

Many of the valid output streams are also valid as input streams. The difference between input and output streams is therefore more a matter of how you use a Lisp object, than of different types of object.

Here is an example of a buffer used as an output stream. Point is initially located as shown immediately before the ‘h’ in ‘the’. At the end, point is located directly before that same ‘h’.

---------- Buffer: foo ----------
This is t∗he contents of foo.
---------- Buffer: foo ----------

(print "This is the output" (get-buffer "foo"))
     ⇒ "This is the output"

---------- Buffer: foo ----------
This is t
"This is the output"
∗he contents of foo.
---------- Buffer: foo ----------

Now we show a use of a marker as an output stream. Initially, the marker is in buffer foo, between the ‘t’ and the ‘h’ in the word ‘the’. At the end, the marker has advanced over the inserted text so that it remains positioned before the same ‘h’. Note that the location of point, shown in the usual fashion, has no effect.

---------- Buffer: foo ----------
This is the ∗output
---------- Buffer: foo ----------

(setq m (copy-marker 10))
     ⇒ #<marker at 10 in foo>

(print "More output for foo." m)
     ⇒ "More output for foo."

---------- Buffer: foo ----------
This is t
"More output for foo."
he ∗output
---------- Buffer: foo ----------

m
     ⇒ #<marker at 34 in foo>

The following example shows output to the echo area:

(print "Echo Area output" t)
     ⇒ "Echo Area output"
---------- Echo Area ----------
"Echo Area output"
---------- Echo Area ----------

Finally, we show the use of a function as an output stream. The function eat-output takes each character that it is given and conses it onto the front of the list last-output (see Building Cons Cells and Lists). At the end, the list contains all the characters output, but in reverse order.

(setq last-output nil)
     ⇒ nil

(defun eat-output (c)
  (setq last-output (cons c last-output)))
     ⇒ eat-output

(print "This is the output" #'eat-output)
     ⇒ "This is the output"

last-output
     ⇒ (10 34 116 117 112 116 117 111 32 101 104
    116 32 115 105 32 115 105 104 84 34 10)

Now we can put the output in the proper order by reversing the list:

(concat (nreverse last-output))
     ⇒ "
\"This is the output\"
"

Calling concat converts the list to a string so you can see its contents more clearly.

Function: external-debugging-output character

This function can be useful as an output stream when debugging. It writes character to the standard error stream.

For example

(print "This is the output" #'external-debugging-output)
-| This is the output
⇒ "This is the output"

20.5 Output Functions

This section describes the Lisp functions for printing Lisp objects—converting objects into their printed representation.

Some of the Emacs printing functions add quoting characters to the output when necessary so that it can be read properly. The quoting characters used are ‘"’ and ‘\’; they distinguish strings from symbols, and prevent punctuation characters in strings and symbols from being taken as delimiters when reading. See Printed Representation and Read Syntax, for full details. You specify quoting or no quoting by the choice of printing function.

If the text is to be read back into Lisp, then you should print with quoting characters to avoid ambiguity. Likewise, if the purpose is to describe a Lisp object clearly for a Lisp programmer. However, if the purpose of the output is to look nice for humans, then it is usually better to print without quoting.

Lisp objects can refer to themselves. Printing a self-referential object in the normal way would require an infinite amount of text, and the attempt could cause infinite recursion. Emacs detects such recursion and prints ‘#level’ instead of recursively printing an object already being printed. For example, here ‘#0’ indicates a recursive reference to the object at level 0 of the current print operation:

(setq foo (list nil))
     ⇒ (nil)
(setcar foo foo)
     ⇒ (#0)

In the functions below, stream stands for an output stream. (See the previous section for a description of output streams. Also See external-debugging-output, a useful stream value for debugging.) If stream is nil or omitted, it defaults to the value of standard-output.

Function: print object &optional stream

The print function is a convenient way of printing. It outputs the printed representation of object to stream, printing in addition one newline before object and another after it. Quoting characters are used. print returns object. For example:

(progn (print 'The\ cat\ in)
       (print "the hat")
       (print " came back"))
     -|
     -| The\ cat\ in
     -|
     -| "the hat"
     -|
     -| " came back"
     ⇒ " came back"
Function: prin1 object &optional stream overrides

This function outputs the printed representation of object to stream. It does not print newlines to separate output as print does, but it does use quoting characters just like print. It returns object.

(progn (prin1 'The\ cat\ in)
       (prin1 "the hat")
       (prin1 " came back"))
     -| The\ cat\ in"the hat"" came back"
     ⇒ " came back"

If overrides is non-nil, it should either be t (which tells prin1 to use the defaults for all printer related variables), or a list of settings. See Overriding Output Variables, for details.

Function: princ object &optional stream

This function outputs the printed representation of object to stream. It returns object.

This function is intended to produce output that is readable by people, not by read, so it doesn’t insert quoting characters and doesn’t put double-quotes around the contents of strings. It does not add any spacing between calls.

(progn
  (princ 'The\ cat)
  (princ " in the \"hat\""))
     -| The cat in the "hat"
     ⇒ " in the \"hat\""
Function: terpri &optional stream ensure

This function outputs a newline to stream. The name stands for “terminate print”. If ensure is non-nil no newline is printed if stream is already at the beginning of a line. Note in this case stream can not be a function and an error is signaled if it is. This function returns t if a newline is printed.

Function: write-char character &optional stream

This function outputs character to stream. It returns character.

Function: flush-standard-output

If you have Emacs-based batch scripts that send output to the terminal, Emacs will automatically display the output whenever you write a newline characters to standard-output. This function allows you to flush to standard-output without sending a newline character first, which enables you to display incomplete lines.

Function: prin1-to-string object &optional noescape overrides

This function returns a string containing the text that prin1 would have printed for the same argument.

(prin1-to-string 'foo)
     ⇒ "foo"
(prin1-to-string (mark-marker))
     ⇒ "#<marker at 2773 in strings.texi>"

If overrides is non-nil, it should either be t
(which tells prin1 to use the defaults for all printer related
variables), or a list of settings.  See Overriding Output Variables, for details.

If noescape is non-nil, that inhibits use of quoting characters in the output. (This argument is supported in Emacs versions 19 and later.)

(prin1-to-string "foo")
     ⇒ "\"foo\""
(prin1-to-string "foo" t)
     ⇒ "foo"

See format, in Formatting Strings, for other ways to obtain the printed representation of a Lisp object as a string.

Macro: with-output-to-string body…

This macro executes the body forms with standard-output set up to feed output into a string. Then it returns that string.

For example, if the current buffer name is ‘foo’,

(with-output-to-string
  (princ "The buffer is ")
  (princ (buffer-name)))

returns "The buffer is foo".

Function: pp object &optional stream

This function outputs object to stream, just like prin1, but does it in a prettier way. That is, it’ll indent and fill the object to make it more readable for humans.

If you need to use binary I/O in batch mode, e.g., use the functions described in this section to write out arbitrary binary data or avoid conversion of newlines on non-POSIX hosts, see set-binary-mode.


20.6 Variables Affecting Output

Variable: standard-output

The value of this variable is the default output stream—the stream that print functions use when the stream argument is nil. The default is t, meaning display in the echo area.

Variable: print-quoted

If this is non-nil, that means to print quoted forms using abbreviated reader syntax, e.g., (quote foo) prints as 'foo, and (function foo) as #'foo. The default is t.

Variable: print-escape-newlines

If this variable is non-nil, then newline characters in strings are printed as ‘\n’ and formfeeds are printed as ‘\f’. Normally these characters are printed as actual newlines and formfeeds.

This variable affects the print functions prin1 and print that print with quoting. It does not affect princ. Here is an example using prin1:

(prin1 "a\nb")
     -| "a
     -| b"
     ⇒ "a
b"

(let ((print-escape-newlines t))
  (prin1 "a\nb"))
     -| "a\nb"
     ⇒ "a
b"

In the second expression, the local binding of print-escape-newlines is in effect during the call to prin1, but not during the printing of the result.

Variable: print-escape-control-characters

If this variable is non-nil, control characters in strings are printed as backslash sequences by the print functions prin1 and print that print with quoting. If this variable and print-escape-newlines are both non-nil, the latter takes precedences for newlines and formfeeds.

Variable: print-escape-nonascii

If this variable is non-nil, then unibyte non-ASCII characters in strings are unconditionally printed as backslash sequences by the print functions prin1 and print that print with quoting.

Those functions also use backslash sequences for unibyte non-ASCII characters, regardless of the value of this variable, when the output stream is a multibyte buffer or a marker pointing into one.

Variable: print-escape-multibyte

If this variable is non-nil, then multibyte non-ASCII characters in strings are unconditionally printed as backslash sequences by the print functions prin1 and print that print with quoting.

Those functions also use backslash sequences for multibyte non-ASCII characters, regardless of the value of this variable, when the output stream is a unibyte buffer or a marker pointing into one.

Variable: print-charset-text-property

This variable controls printing of ‘charset’ text property on printing a string. The value should be nil, t, or default.

If the value is nil, charset text properties are never printed. If t, they are always printed.

If the value is default, only print charset text properties if there is an “unexpected” charset property. For ascii characters, all charsets are considered “expected”. Otherwise, the expected charset property of a character is given by char-charset.

Variable: print-length

The value of this variable is the maximum number of elements to print in any list, vector or bool-vector. If an object being printed has more than this many elements, it is abbreviated with an ellipsis.

If the value is nil (the default), then there is no limit.

(setq print-length 2)
     ⇒ 2
(print '(1 2 3 4 5))
     -| (1 2 ...)
     ⇒ (1 2 ...)
Variable: print-level

The value of this variable is the maximum depth of nesting of parentheses and brackets when printed. Any list or vector at a depth exceeding this limit is abbreviated with an ellipsis. A value of nil (which is the default) means no limit.

User Option: eval-expression-print-length
User Option: eval-expression-print-level

These are the values for print-length and print-level used by eval-expression, and thus, indirectly, by many interactive evaluation commands (see Evaluating Emacs Lisp Expressions in The GNU Emacs Manual).

These variables are used for detecting and reporting circular and shared structure:

Variable: print-circle

If non-nil, this variable enables detection of circular and shared structure in printing. See Read Syntax for Circular Objects.

Variable: print-unreadable-function

By default, Emacs prints unreadable objects as ‘#<...>"’. For instance:

(prin1-to-string (make-marker))
     ⇒ "#<marker in no buffer>"

If this variable is non-nil, it should be a function that will be called to handle printing of these objects. The function will be called with two arguments: the object and the noescape flag used by the printing functions (see Output Functions).

The function should return either nil (print the object as usual), or a string (which will be printed), or any other object (don’t print the object). For instance:

(let ((print-unreadable-function
       (lambda (object escape) "hello")))
  (prin1-to-string (make-marker)))
     ⇒ "hello"
Variable: print-gensym

If non-nil, this variable enables detection of uninterned symbols (see Creating and Interning Symbols) in printing. When this is enabled, uninterned symbols print with the prefix ‘#:’, which tells the Lisp reader to produce an uninterned symbol.

Variable: print-continuous-numbering

If non-nil, that means number continuously across print calls. This affects the numbers printed for ‘#n=’ labels and ‘#m#’ references. Don’t set this variable with setq; you should only bind it temporarily to t with let. When you do that, you should also bind print-number-table to nil.

Variable: print-number-table

This variable holds a vector used internally by printing to implement the print-circle feature. You should not use it except to bind it to nil when you bind print-continuous-numbering.

Variable: float-output-format

This variable specifies how to print floating-point numbers. The default is nil, meaning use the shortest output that represents the number without losing information.

To control output format more precisely, you can put a string in this variable. The string should hold a ‘%’-specification to be used in the C function sprintf. For further restrictions on what you can use, see the variable’s documentation string.

Variable: print-integers-as-characters

When this variable is non-nil, integers that represent graphic base characters will be printed using Lisp character syntax (see Basic Char Syntax). Other numbers are printed the usual way. For example, the list (4 65 -1 10) would be printed as ‘(4 ?A -1 ?\n)’.

More precisely, values printed in character syntax are those representing characters belonging to the Unicode general categories Letter, Number, Punctuation, Symbol and Private-use (see Character Properties), as well as the control characters having their own escape syntax such as newline.


20.7 Overriding Output Variables

The previous section (see Output Functions) lists the numerous variables that control how the Emacs Lisp printer formats data for outputs. These are generally available for users to change, but sometimes you want to output data in the default format, or override the user settings in some other way. For instance, if you’re storing Emacs Lisp data in a file, you don’t want that data to be shortened by a print-length setting.

The prin1 and prin1-to-string functions therefore have an optional overrides argument. This argument can either be t (which means that all printing variables should be reset to the default values), or a list of settings for some of the variables. Each element in the list can be either t (which means “reset to defaults”, and will usually be the first element of the list), or a pair whose car is a symbol that stands for an output variable and whose cdr is the value for that variable.

For instance, this prints using nothing but defaults:

(prin1 object nil t)

This prints object using the current printing settings, but overrides the value of print-length to be 5:

(prin1 object nil '((length . 5)))

And finally, this prints object using only default settings, but with print-length bound to 5:

(prin1 object nil '(t (length . 5)))

Below is a list of symbols that can be used, and which variables they map to:

length

This overrides print-length.

level

This overrides print-level.

circle

This overrides print-circle.

quoted

This overrides print-quoted.

escape-newlines

This overrides print-escape-newlines.

escape-control-characters

This overrides print-escape-control-characters.

escape-nonascii

This overrides print-escape-nonascii.

escape-multibyte

This overrides print-escape-multibyte.

charset-text-property

This overrides print-charset-text-property.

unreadeable-function

This overrides print-unreadable-function.

gensym

This overrides print-gensym.

continuous-numbering

This overrides print-continuous-numbering.

number-table

This overrides print-number-table.

float-format

This overrides float-output-format.

integers-as-characters

This overrides print-integers-as-characters.

In the future, more overrides may be offered that do not map directly to a variable, but can only be used via this parameter.


21 Minibuffers

A minibuffer is a special buffer that Emacs commands use to read arguments more complicated than the single numeric prefix argument. These arguments include file names, buffer names, and command names (as in M-x). The minibuffer is displayed on the bottom line of the frame, in the same place as the echo area (see The Echo Area), but only while it is in use for reading an argument.


21.1 Introduction to Minibuffers

In most ways, a minibuffer is a normal Emacs buffer. Most operations within a buffer, such as editing commands, work normally in a minibuffer. However, many operations for managing buffers do not apply to minibuffers. The name of a minibuffer always has the form ‘ *Minibuf-number*, and it cannot be changed. Minibuffers are displayed only in special windows used only for minibuffers; these windows always appear at the bottom of a frame. (Sometimes frames have no minibuffer window, and sometimes a special kind of frame contains nothing but a minibuffer window; see Minibuffers and Frames.)

The text in the minibuffer always starts with the prompt string, the text that was specified by the program that is using the minibuffer to tell the user what sort of input to type. This text is marked read-only so you won’t accidentally delete or change it. It is also marked as a field (see Defining and Using Fields), so that certain motion functions, including beginning-of-line, forward-word, forward-sentence, and forward-paragraph, stop at the boundary between the prompt and the actual text.

The minibuffer’s window is normally a single line; it grows automatically if the contents require more space. Whilst the minibuffer is active, you can explicitly resize its window temporarily with the window sizing commands; the window reverts to its normal size when the minibuffer is exited. When the minibuffer is not active, you can resize its window permanently by using the window sizing commands in the frame’s other window, or dragging the mode line with the mouse. (Due to details of the current implementation, for this to work resize-mini-windows must be nil.) If the frame contains just a minibuffer window, you can change its size by changing the frame’s size.

Use of the minibuffer reads input events, and that alters the values of variables such as this-command and last-command (see Information from the Command Loop). Your program should bind them around the code that uses the minibuffer, if you do not want that to change them.

Under some circumstances, a command can use a minibuffer even if there is an active minibuffer; such a minibuffer is called a recursive minibuffer. The first minibuffer is named ‘ *Minibuf-1*. Recursive minibuffers are named by incrementing the number at the end of the name. (The names begin with a space so that they won’t show up in normal buffer lists.) Of several recursive minibuffers, the innermost (or most recently entered) is the active minibuffer–it is the one you can terminate by typing RET (exit-minibuffer) in. We usually call this the minibuffer. You can permit or forbid recursive minibuffers by setting the variable enable-recursive-minibuffers, or by putting properties of that name on command symbols (See Recursive Minibuffers.)

Like other buffers, a minibuffer uses a local keymap (see Keymaps) to specify special key bindings. The function that invokes the minibuffer also sets up its local map according to the job to be done. See Reading Text Strings with the Minibuffer, for the non-completion minibuffer local maps. See Minibuffer Commands that Do Completion, for the minibuffer local maps for completion.

An active minibuffer usually has major mode minibuffer-mode. This is an Emacs internal mode without any special features. To customize the setup of minibuffers, we suggest you use minibuffer-setup-hook (see Minibuffer Miscellany) rather than minibuffer-mode-hook, since the former is run later, after the minibuffer has been fully initialized.

When a minibuffer is inactive, its major mode is minibuffer-inactive-mode, with keymap minibuffer-inactive-mode-map. This is only really useful if the minibuffer is in a separate frame. See Minibuffers and Frames.

When Emacs is running in batch mode, any request to read from the minibuffer actually reads a line from the standard input descriptor that was supplied when Emacs was started. This supports only basic input: none of the special minibuffer features (history, completion, etc.) are available in batch mode.


21.2 Reading Text Strings with the Minibuffer

The most basic primitive for minibuffer input is read-from-minibuffer, which can be used to read either a string or a Lisp object in textual form. The function read-regexp is used for reading regular expressions (see Regular Expressions), which are a special kind of string. There are also specialized functions for reading commands, variables, file names, etc. (see Completion).

In most cases, you should not call minibuffer input functions in the middle of a Lisp function. Instead, do all minibuffer input as part of reading the arguments for a command, in the interactive specification. See Defining Commands.

Function: read-from-minibuffer prompt &optional initial keymap read history default inherit-input-method

This function is the most general way to get input from the minibuffer. By default, it accepts arbitrary text and returns it as a string; however, if read is non-nil, then it uses read to convert the text into a Lisp object (see Input Functions).

The first thing this function does is to activate a minibuffer and display it with prompt (which must be a string) as the prompt. Then the user can edit text in the minibuffer.

When the user types a command to exit the minibuffer, read-from-minibuffer constructs the return value from the text in the minibuffer. Normally it returns a string containing that text. However, if read is non-nil, read-from-minibuffer reads the text and returns the resulting Lisp object, unevaluated. (See Input Functions, for information about reading.)

The argument default specifies default values to make available through the history commands. It should be a string, a list of strings, or nil. The string or strings become the minibuffer’s “future history”, available to the user with M-n. In addition, if the call provides completion (e.g., via the keymap argument), the completion candidates are added to the “future history” when the values in default are exhausted by M-n; see minibuffer-default-add-function.

If read is non-nil, then default is also used as the input to read, if the user enters empty input. If default is a list of strings, the first string is used as the input. If default is nil, empty input results in an end-of-file error. However, in the usual case (where read is nil), read-from-minibuffer ignores default when the user enters empty input and returns an empty string, "". In this respect, it differs from all the other minibuffer input functions in this chapter.

If keymap is non-nil, that keymap is the local keymap to use in the minibuffer. If keymap is omitted or nil, the value of minibuffer-local-map is used as the keymap. Specifying a keymap is the most important way to customize the minibuffer for various applications such as completion.

The argument history specifies a history list variable to use for saving the input and for history commands used in the minibuffer. It defaults to minibuffer-history. If history is the symbol t, history is not recorded. You can optionally specify a starting position in the history list as well. See Minibuffer History.

If the variable minibuffer-allow-text-properties is non-nil, then the string that is returned includes whatever text properties were present in the minibuffer. Otherwise all the text properties are stripped when the value is returned.

The text properties in minibuffer-prompt-properties are applied to the prompt. By default, this property list defines a face to use for the prompt. This face, if present, is applied to the end of the face list and merged before display.

If the user wants to completely control the look of the prompt, the most convenient way to do that is to specify the default face at the end of all face lists. For instance:

(read-from-minibuffer
 (concat
  (propertize "Bold" 'face '(bold default))
  (propertize " and normal: " 'face '(default))))

If the argument inherit-input-method is non-nil, then the minibuffer inherits the current input method (see Input Methods) and the setting of enable-multibyte-characters (see Text Representations) from whichever buffer was current before entering the minibuffer.

Use of initial is mostly deprecated; we recommend using a non-nil value only in conjunction with specifying a cons cell for history. See Initial Input.

Function: read-string prompt &optional initial history default inherit-input-method

This function reads a string from the minibuffer and returns it. The arguments prompt, initial, history and inherit-input-method are used as in read-from-minibuffer. The keymap used is minibuffer-local-map.

The optional argument default is used as in read-from-minibuffer, except that, if non-nil, it also specifies a default value to return if the user enters null input. As in read-from-minibuffer it should be a string, a list of strings, or nil, which is equivalent to an empty string. When default is a string, that string is the default value. When it is a list of strings, the first string is the default value. (All these strings are available to the user in the “future minibuffer history”.)

This function works by calling the read-from-minibuffer function:

(read-string prompt initial history default inherit)
≡
(let ((value
       (read-from-minibuffer prompt initial nil nil
                             history default inherit)))
  (if (and (equal value "") default)
      (if (consp default) (car default) default)
    value))

If you have a long string (for instance, one that is several lines long) that you wish to edit, using read-string may not be ideal. In that case, popping to a new, normal buffer where the user can edit the string may be more convenient, and you can use the read-string-from-buffer function to do that.

Function: read-regexp prompt &optional defaults history

This function reads a regular expression as a string from the minibuffer and returns it. If the minibuffer prompt string prompt does not end in ‘:’ (followed by optional whitespace), the function adds ‘: ’ to the end, preceded by the default return value (see below), if that is non-empty.

The optional argument defaults controls the default value to return if the user enters null input, and should be one of: a string; nil, which is equivalent to an empty string; a list of strings; or a symbol.

If defaults is a symbol, read-regexp consults the value of the variable read-regexp-defaults-function (see below), and if that is non-nil uses it in preference to defaults. The value in this case should be either:

  • regexp-history-last, which means to use the first element of the appropriate minibuffer history list (see below).
  • A function of no arguments, whose return value (which should be nil, a string, or a list of strings) becomes the value of defaults.

read-regexp now ensures that the result of processing defaults is a list (i.e., if the value is nil or a string, it converts it to a list of one element). To this list, read-regexp then appends a few potentially useful candidates for input. These are:

  • The word or symbol at point.
  • The last regexp used in an incremental search.
  • The last string used in an incremental search.
  • The last string or pattern used in query-replace commands.

The function now has a list of regular expressions that it passes to read-from-minibuffer to obtain the user’s input. The first element of the list is the default result in case of empty input. All elements of the list are available to the user as the “future minibuffer history” list (see future list in The GNU Emacs Manual).

The optional argument history, if non-nil, is a symbol specifying a minibuffer history list to use (see Minibuffer History). If it is omitted or nil, the history list defaults to regexp-history.

The user can use the M-s c command to indicate whether case folding should be on or off. If the user has used this command, the returned string will have the text property case-fold set to either fold or inhibit-fold. It is up to the caller of read-regexp to actually use this value, and the convenience function read-regexp-case-fold-search is provided for that. A typical usage pattern here might look like:

(let* ((regexp (read-regexp "Search for: "))
       (case-fold-search (read-regexp-case-fold-search regexp)))
  (re-search-forward regexp))
User Option: read-regexp-defaults-function

The function read-regexp may use the value of this variable to determine its list of default regular expressions. If non-nil, the value of this variable should be either:

  • The symbol regexp-history-last.
  • A function of no arguments that returns either nil, a string, or a list of strings.

See read-regexp above for details of how these values are used.

Variable: minibuffer-allow-text-properties

If this variable is nil, then read-from-minibuffer and read-string strip all text properties from the minibuffer input before returning it. However, read-no-blanks-input (see below), as well as read-minibuffer and related functions (see Reading Lisp Objects With the Minibuffer), and all functions that do minibuffer input with completion, remove the face property unconditionally, regardless of the value of this variable.

If this variable is non-nil, most text properties on strings from the completion table are preserved—but only on the part of the strings that were completed.

(let ((minibuffer-allow-text-properties t))
  (completing-read "String: " (list (propertize "foobar" 'data 'zot))))
=> #("foobar" 3 6 (data zot))

In this example, the user typed ‘foo’ and then hit the TAB key, so the text properties are only preserved on the last three characters.

Variable: minibuffer-local-map

This is the default local keymap for reading from the minibuffer. By default, it makes the following bindings:

C-j

exit-minibuffer

RET

exit-minibuffer

M-<

minibuffer-beginning-of-buffer

C-g

abort-recursive-edit

M-n
DOWN

next-history-element

M-p
UP

previous-history-element

M-s

next-matching-history-element

M-r

previous-matching-history-element

The variable minibuffer-mode-map is an alias for this variable.

Function: read-no-blanks-input prompt &optional initial inherit-input-method

This function reads a string from the minibuffer, but does not allow whitespace characters as part of the input: instead, those characters terminate the input. The arguments prompt, initial, and inherit-input-method are used as in read-from-minibuffer.

This is a simplified interface to the read-from-minibuffer function, and passes the value of the minibuffer-local-ns-map keymap as the keymap argument for that function. Since the keymap minibuffer-local-ns-map does not rebind C-q, it is possible to put a space into the string, by quoting it.

This function discards text properties, regardless of the value of minibuffer-allow-text-properties.

(read-no-blanks-input prompt initial)
≡
(let (minibuffer-allow-text-properties)
  (read-from-minibuffer prompt initial minibuffer-local-ns-map))
Variable: minibuffer-local-ns-map

This built-in variable is the keymap used as the minibuffer local keymap in the function read-no-blanks-input. By default, it makes the following bindings, in addition to those of minibuffer-local-map:

SPC

exit-minibuffer

TAB

exit-minibuffer

?

self-insert-and-exit

Function: format-prompt prompt default &rest format-args

Format prompt with default value default according to the minibuffer-default-prompt-format variable.

minibuffer-default-prompt-format is a format string (defaulting to ‘" (default %s)"’ that says how the “default” bit in prompts like ‘"Local filename (default somefile): "’ are to be formatted.

To allow the users to customize how this is displayed, code that prompts the user for a value (and has a default) should look something along the lines of this code snippet:

(read-file-name
 (format-prompt "Local filename" file)
 nil file)

If format-args is nil, prompt is used as a literal string. If format-args is non-nil, prompt is used as a format control string, and prompt and format-args are passed to format (see Formatting Strings).

minibuffer-default-prompt-format can be ‘""’, in which case no default values are displayed.

If default is nil, there is no default value, and therefore no “default value” string is included in the result value. If default is a non-nil list, the first element of the list is used in the prompt.

Both prompt and minibuffer-default-prompt-format are run through substitute-command-keys (see Substituting Key Bindings in Documentation).

Variable: read-minibuffer-restore-windows

If this option is non-nil (the default), getting input from the minibuffer will restore, on exit, the window configurations of the frame where the minibuffer was entered from and, if it is different, the frame that owns the minibuffer window. This means that if, for example, a user splits a window while getting input from the minibuffer on the same frame, that split will be undone when exiting the minibuffer.

If this option is nil, no such restorations are done. Hence, the window split mentioned above will persist after exiting the minibuffer.


21.3 Reading Lisp Objects with the Minibuffer

This section describes functions for reading Lisp objects with the minibuffer.

Function: read-minibuffer prompt &optional initial

This function reads a Lisp object using the minibuffer, and returns it without evaluating it. The arguments prompt and initial are used as in read-from-minibuffer.

This is a simplified interface to the read-from-minibuffer function:

(read-minibuffer prompt initial)
≡
(let (minibuffer-allow-text-properties)
  (read-from-minibuffer prompt initial nil t))

Here is an example in which we supply the string "(testing)" as initial input:

(read-minibuffer
 "Enter an expression: " (format "%s" '(testing)))

;; Here is how the minibuffer is displayed:

---------- Buffer: Minibuffer ----------
Enter an expression: (testing)∗
---------- Buffer: Minibuffer ----------

The user can type RET immediately to use the initial input as a default, or can edit the input.

Function: eval-minibuffer prompt &optional initial

This function reads a Lisp expression using the minibuffer, evaluates it, then returns the result. The arguments prompt and initial are used as in read-from-minibuffer.

This function simply evaluates the result of a call to read-minibuffer:

(eval-minibuffer prompt initial)
≡
(eval (read-minibuffer prompt initial))
Function: edit-and-eval-command prompt form

This function reads a Lisp expression in the minibuffer, evaluates it, then returns the result. The difference between this command and eval-minibuffer is that here the initial form is not optional and it is treated as a Lisp object to be converted to printed representation rather than as a string of text. It is printed with prin1, so if it is a string, double-quote characters (‘"’) appear in the initial text. See Output Functions.

In the following example, we offer the user an expression with initial text that is already a valid form:

(edit-and-eval-command "Please edit: " '(forward-word 1))

;; After evaluation of the preceding expression,
;;   the following appears in the minibuffer:

---------- Buffer: Minibuffer ----------
Please edit: (forward-word 1)∗
---------- Buffer: Minibuffer ----------

Typing RET right away would exit the minibuffer and evaluate the expression, thus moving point forward one word.


21.4 Minibuffer History

A minibuffer history list records previous minibuffer inputs so the user can reuse them conveniently. It is a variable whose value is a list of strings (previous inputs), most recent first.

There are many separate minibuffer history lists, used for different kinds of inputs. It’s the Lisp programmer’s job to specify the right history list for each use of the minibuffer.

You specify a minibuffer history list with the optional history argument to read-from-minibuffer or completing-read. Here are the possible values for it:

variable

Use variable (a symbol) as the history list.

(variable . startpos)

Use variable (a symbol) as the history list, and assume that the initial history position is startpos (a nonnegative integer).

Specifying 0 for startpos is equivalent to just specifying the symbol variable. previous-history-element will display the most recent element of the history list in the minibuffer. If you specify a positive startpos, the minibuffer history functions behave as if (elt variable (1- startpos)) were the history element currently shown in the minibuffer.

For consistency, you should also specify that element of the history as the initial minibuffer contents, using the initial argument to the minibuffer input function (see Initial Input).

If you don’t specify history, then the default history list minibuffer-history is used. For other standard history lists, see below. You can also create your own history list variable; just initialize it to nil before the first use. If the variable is buffer local, then each buffer will have its own input history list.

Both read-from-minibuffer and completing-read add new elements to the history list automatically, and provide commands to allow the user to reuse items on the list (see Minibuffer Commands). The only thing your program needs to do to use a history list is to initialize it and to pass its name to the input functions when you wish. But it is safe to modify the list by hand when the minibuffer input functions are not using it.

By default, when M-n (next-history-element, see next-history-element) reaches the end of the list of default values provided by the command which initiated reading input from the minibuffer, M-n adds all of the completion candidates, as specified by minibuffer-completion-table (see Minibuffer Commands that Do Completion), to the list of defaults, so that all those candidates are available as “future history”. Your program can control that via the variable minibuffer-default-add-function: if its value is not a function, this automatic addition is disabled, and you can also set this variable to your own function which adds only some candidates, or some other values, to the “future history”.

Emacs functions that add a new element to a history list can also delete old elements if the list gets too long. The variable history-length specifies the maximum length for most history lists. To specify a different maximum length for a particular history list, put the length in the history-length property of the history list symbol. The variable history-delete-duplicates specifies whether to delete duplicates in history.

Function: add-to-history history-var newelt &optional maxelt keep-all

This function adds a new element newelt, if it isn’t the empty string, to the history list stored in the variable history-var, and returns the updated history list. It limits the list length to the value of maxelt (if non-nil) or history-length (described below). The possible values of maxelt have the same meaning as the values of history-length. history-var cannot refer to a lexical variable.

Normally, add-to-history removes duplicate members from the history list if history-delete-duplicates is non-nil. However, if keep-all is non-nil, that says not to remove duplicates, and to add newelt to the list even if it is empty.

Variable: history-add-new-input

If the value of this variable is nil, standard functions that read from the minibuffer don’t add new elements to the history list. This lets Lisp programs explicitly manage input history by using add-to-history. The default value is t.

User Option: history-length

The value of this variable specifies the maximum length for all history lists that don’t specify their own maximum lengths. If the value is t, that means there is no maximum (don’t delete old elements). If a history list variable’s symbol has a non-nil history-length property, it overrides this variable for that particular history list.

User Option: history-delete-duplicates

If the value of this variable is t, that means when adding a new history element, all previous identical elements are deleted.

Here are some of the standard minibuffer history list variables:

Variable: minibuffer-history

The default history list for minibuffer history input.

Variable: query-replace-history

A history list for arguments to query-replace (and similar arguments to other commands).

Variable: file-name-history

A history list for file-name arguments.

Variable: buffer-name-history

A history list for buffer-name arguments.

Variable: regexp-history

A history list for regular expression arguments.

Variable: extended-command-history

A history list for arguments that are names of extended commands.

Variable: shell-command-history

A history list for arguments that are shell commands.

Variable: read-expression-history

A history list for arguments that are Lisp expressions to evaluate.

Variable: face-name-history

A history list for arguments that are faces.

Variable: custom-variable-history

A history list for variable-name arguments read by read-variable.

Variable: read-number-history

A history list for numbers read by read-number.

Variable: goto-line-history

A history list for arguments to goto-line. This variable can be made local in every buffer by customizing the user option goto-line-history-local.


21.5 Initial Input

Several of the functions for minibuffer input have an argument called initial. This is a mostly-deprecated feature for specifying that the minibuffer should start out with certain text, instead of empty as usual.

If initial is a string, the minibuffer starts out containing the text of the string, with point at the end, when the user starts to edit the text. If the user simply types RET to exit the minibuffer, it will use the initial input string to determine the value to return.

We discourage use of a non-nil value for initial, because initial input is an intrusive interface. History lists and default values provide a much more convenient method to offer useful default inputs to the user.

There is just one situation where you should specify a string for an initial argument. This is when you specify a cons cell for the history argument. See Minibuffer History.

initial can also be a cons cell of the form (string . position). This means to insert string in the minibuffer but put point at position within the string’s text.

As a historical accident, position was implemented inconsistently in different functions. In completing-read, position’s value is interpreted as origin-zero; that is, a value of 0 means the beginning of the string, 1 means after the first character, etc. In read-minibuffer, and the other non-completion minibuffer input functions that support this argument, 1 means the beginning of the string, 2 means after the first character, etc.

Use of a cons cell as the value for initial arguments is deprecated.


21.6 Completion

Completion is a feature that fills in the rest of a name starting from an abbreviation for it. Completion works by comparing the user’s input against a list of valid names and determining how much of the name is determined uniquely by what the user has typed. For example, when you type C-x b (switch-to-buffer) and then type the first few letters of the name of the buffer to which you wish to switch, and then type TAB (minibuffer-complete), Emacs extends the name as far as it can.

Standard Emacs commands offer completion for names of symbols, files, buffers, and processes; with the functions in this section, you can implement completion for other kinds of names.

The try-completion function is the basic primitive for completion: it returns the longest determined completion of a given initial string, with a given set of strings to match against.

The function completing-read provides a higher-level interface for completion. A call to completing-read specifies how to determine the list of valid names. The function then activates the minibuffer with a local keymap that binds a few keys to commands useful for completion. Other functions provide convenient simple interfaces for reading certain kinds of names with completion.


21.6.1 Basic Completion Functions

The following completion functions have nothing in themselves to do with minibuffers. We describe them here to keep them near the higher-level completion features that do use the minibuffer.

Function: try-completion string collection &optional predicate

This function returns the longest common substring of all possible completions of string in collection.

collection is called the completion table. Its value must be a list of strings or cons cells, an obarray, a hash table, or a completion function.

try-completion compares string against each of the permissible completions specified by the completion table. If no permissible completions match, it returns nil. If there is just one matching completion, and the match is exact, it returns t. Otherwise, it returns the longest initial sequence common to all possible matching completions.

If collection is a list, the permissible completions are specified by the elements of the list, each of which should be either a string, or a cons cell whose CAR is either a string or a symbol (a symbol is converted to a string using symbol-name). If the list contains elements of any other type, those are ignored.

If collection is an obarray (see Creating and Interning Symbols), the names of all symbols in the obarray form the set of permissible completions.

If collection is a hash table, then the keys that are strings or symbols are the possible completions. Other keys are ignored.

You can also use a function as collection. Then the function is solely responsible for performing completion; try-completion returns whatever this function returns. The function is called with three arguments: string, predicate and nil (the third argument is so that the same function can be used in all-completions and do the appropriate thing in either case). See Programmed Completion.

If the argument predicate is non-nil, then it must be a function of one argument, unless collection is a hash table, in which case it should be a function of two arguments. It is used to test each possible match, and the match is accepted only if predicate returns non-nil. The argument given to predicate is either a string or a cons cell (the CAR of which is a string) from the alist, or a symbol (not a symbol name) from the obarray. If collection is a hash table, predicate is called with two arguments, the string key and the associated value.

In addition, to be acceptable, a completion must also match all the regular expressions in completion-regexp-list. (Unless collection is a function, in which case that function has to handle completion-regexp-list itself.)

In the first of the following examples, the string ‘foo’ is matched by three of the alist CARs. All of the matches begin with the characters ‘fooba’, so that is the result. In the second example, there is only one possible match, and it is exact, so the return value is t.

(try-completion
 "foo"
 '(("foobar1" 1) ("barfoo" 2) ("foobaz" 3) ("foobar2" 4)))
     ⇒ "fooba"

(try-completion "foo" '(("barfoo" 2) ("foo" 3)))
     ⇒ t

In the following example, numerous symbols begin with the characters ‘forw’, and all of them begin with the word ‘forward’. In most of the symbols, this is followed with a ‘-’, but not in all, so no more than ‘forward’ can be completed.

(try-completion "forw" obarray)
     ⇒ "forward"

Finally, in the following example, only two of the three possible matches pass the predicate test (the string ‘foobaz’ is too short). Both of those begin with the string ‘foobar’.

(defun test (s)
  (> (length (car s)) 6))
     ⇒ test
(try-completion
 "foo"
 '(("foobar1" 1) ("barfoo" 2) ("foobaz" 3) ("foobar2" 4))
 'test)
     ⇒ "foobar"
Function: all-completions string collection &optional predicate

This function returns a list of all possible completions of string. The arguments to this function are the same as those of try-completion, and it uses completion-regexp-list in the same way that try-completion does.

If collection is a function, it is called with three arguments: string, predicate and t; then all-completions returns whatever the function returns. See Programmed Completion.

Here is an example, using the function test shown in the example for try-completion:

(defun test (s)
  (> (length (car s)) 6))
     ⇒ test

(all-completions
 "foo"
 '(("foobar1" 1) ("barfoo" 2) ("foobaz" 3) ("foobar2" 4))
 'test)
     ⇒ ("foobar1" "foobar2")
Function: test-completion string collection &optional predicate

This function returns non-nil if string is a valid completion alternative specified by collection and predicate. The arguments are the same as in try-completion. For instance, if collection is a list of strings, this is true if string appears in the list and predicate is satisfied.

This function uses completion-regexp-list in the same way that try-completion does.

If predicate is non-nil and if collection contains several strings that are equal to each other, as determined by compare-strings according to completion-ignore-case, then predicate should accept either all or none of them. Otherwise, the return value of test-completion is essentially unpredictable.

If collection is a function, it is called with three arguments, the values string, predicate and lambda; whatever it returns, test-completion returns in turn.

Function: completion-boundaries string collection predicate suffix

This function returns the boundaries of the field on which collection will operate, assuming that string holds the text before point and suffix holds the text after point.

Normally completion operates on the whole string, so for all normal collections, this will always return (0 . (length suffix)). But more complex completion, such as completion on files, is done one field at a time. For example, completion of "/usr/sh" will include "/usr/share/" but not "/usr/share/doc" even if "/usr/share/doc" exists. Also all-completions on "/usr/sh" will not include "/usr/share/" but only "share/". So if string is "/usr/sh" and suffix is "e/doc", completion-boundaries will return (5 . 1) which tells us that the collection will only return completion information that pertains to the area after "/usr/" and before "/doc". try-completion is not affected by nontrivial boundaries; e.g., try-completion on "/usr/sh" might still return "/usr/share/", not "share/".

If you store a completion alist in a variable, you should mark the variable as risky by giving it a non-nil risky-local-variable property. See File Local Variables.

Variable: completion-ignore-case

If the value of this variable is non-nil, case is not considered significant in completion. Within read-file-name, this variable is overridden by read-file-name-completion-ignore-case (see Reading File Names); within read-buffer, it is overridden by read-buffer-completion-ignore-case (see High-Level Completion Functions).

Variable: completion-regexp-list

This is a list of regular expressions. The completion functions only consider a completion acceptable if it matches all regular expressions in this list, with case-fold-search (see Searching and Case) bound to the value of completion-ignore-case.

Do not set this variable to a non-nil value globally, as that is not safe and will probably cause errors in completion commands. This variable should be only let-bound to non-nil values around calls to basic completion functions: try-completion, test-completion, and all-completions.

Macro: lazy-completion-table var fun

This macro provides a way to initialize the variable var as a collection for completion in a lazy way, not computing its actual contents until they are first needed. You use this macro to produce a value that you store in var. The actual computation of the proper value is done the first time you do completion using var. It is done by calling fun with no arguments. The value fun returns becomes the permanent value of var.

Here is an example:

(defvar foo (lazy-completion-table foo make-my-alist))

There are several functions that take an existing completion table and return a modified version. completion-table-case-fold returns a case-insensitive table. completion-table-in-turn and completion-table-merge combine multiple input tables in different ways. completion-table-subvert alters a table to use a different initial prefix. completion-table-with-quoting returns a table suitable for operating on quoted text. completion-table-with-predicate filters a table with a predicate function. completion-table-with-terminator adds a terminating string.


21.6.2 Completion and the Minibuffer

This section describes the basic interface for reading from the minibuffer with completion.

Function: completing-read prompt collection &optional predicate require-match initial history default inherit-input-method

This function reads a string in the minibuffer, assisting the user by providing completion. It activates the minibuffer with prompt prompt, which must be a string.

The actual completion is done by passing the completion table collection and the completion predicate predicate to the function try-completion (see Basic Completion Functions). This happens in certain commands bound in the local keymaps used for completion. Some of these commands also call test-completion. Thus, if predicate is non-nil, it should be compatible with collection and completion-ignore-case. See Definition of test-completion.

See Programmed Completion, for detailed requirements when collection is a function.

The value of the optional argument require-match determines how the user may exit the minibuffer:

  • If nil, the usual minibuffer exit commands work regardless of the input in the minibuffer.
  • If t, the usual minibuffer exit commands won’t exit unless the input completes to an element of collection.
  • If confirm, the user can exit with any input, but is asked for confirmation if the input is not an element of collection.
  • If confirm-after-completion, the user can exit with any input, but is asked for confirmation if the preceding command was a completion command (i.e., one of the commands in minibuffer-confirm-exit-commands) and the resulting input is not an element of collection. See Minibuffer Commands that Do Completion.
  • If a function, it is called with the input as the only argument. The function should return a non-nil value if the input is acceptable.
  • Any other value of require-match behaves like t, except that the exit commands won’t exit if it performs completion.

However, empty input is always permitted, regardless of the value of require-match; in that case, completing-read returns the first element of default, if it is a list; "", if default is nil; or default. The string or strings in default are also available to the user through the history commands (see Minibuffer Commands). In addition, the completion candidates are added to the “future history” when the values in default are exhausted by M-n; see minibuffer-default-add-function.

The function completing-read uses minibuffer-local-completion-map as the keymap if require-match is nil, and uses minibuffer-local-must-match-map if require-match is non-nil. See Minibuffer Commands that Do Completion.

The argument history specifies which history list variable to use for saving the input and for minibuffer history commands. It defaults to minibuffer-history. If history is the symbol t, history is not recorded. See Minibuffer History.

The argument initial is mostly deprecated; we recommend using a non-nil value only in conjunction with specifying a cons cell for history. See Initial Input. For default input, use default instead.

If the argument inherit-input-method is non-nil, then the minibuffer inherits the current input method (see Input Methods) and the setting of enable-multibyte-characters (see Text Representations) from whichever buffer was current before entering the minibuffer.

If the variable completion-ignore-case is non-nil, completion ignores case when comparing the input against the possible matches. See Basic Completion Functions. In this mode of operation, predicate must also ignore case, or you will get surprising results.

Here’s an example of using completing-read:

(completing-read
 "Complete a foo: "
 '(("foobar1" 1) ("barfoo" 2) ("foobaz" 3) ("foobar2" 4))
 nil t "fo")

;; After evaluation of the preceding expression,
;;   the following appears in the minibuffer:

---------- Buffer: Minibuffer ----------
Complete a foo: fo∗
---------- Buffer: Minibuffer ----------

If the user then types DEL DEL b RET, completing-read returns barfoo.

The completing-read function binds variables to pass information to the commands that actually do completion. They are described in the following section.

Variable: completing-read-function

The value of this variable must be a function, which is called by completing-read to actually do its work. It should accept the same arguments as completing-read. This can be bound to a different function to completely override the normal behavior of completing-read.


21.6.3 Minibuffer Commands that Do Completion

This section describes the keymaps, commands and user options used in the minibuffer to do completion.

Variable: minibuffer-completion-table

The value of this variable is the completion table (see Basic Completion Functions) used for completion in the minibuffer. This is the buffer-local variable that contains what completing-read passes to try-completion. It is used by minibuffer completion commands such as minibuffer-complete.

Variable: minibuffer-completion-predicate

This variable’s value is the predicate that completing-read passes to try-completion. The variable is also used by the other minibuffer completion functions.

Variable: minibuffer-completion-confirm

This variable determines whether Emacs asks for confirmation before exiting the minibuffer; completing-read sets this variable, and the function minibuffer-complete-and-exit checks the value before exiting. If the value is nil, confirmation is not required. If the value is confirm, the user may exit with an input that is not a valid completion alternative, but Emacs asks for confirmation. If the value is confirm-after-completion, the user may exit with an input that is not a valid completion alternative, but Emacs asks for confirmation if the user submitted the input right after any of the completion commands in minibuffer-confirm-exit-commands.

Variable: minibuffer-confirm-exit-commands

This variable holds a list of commands that cause Emacs to ask for confirmation before exiting the minibuffer, if the require-match argument to completing-read is confirm-after-completion. The confirmation is requested if the user attempts to exit the minibuffer immediately after calling any command in this list.

Command: minibuffer-complete-word

This function completes the minibuffer contents by at most a single word. Even if the minibuffer contents have only one completion, minibuffer-complete-word does not add any characters beyond the first character that is not a word constituent. See Syntax Tables.

Command: minibuffer-complete

This function completes the minibuffer contents as far as possible.

Command: minibuffer-complete-and-exit

This function completes the minibuffer contents, and exits if confirmation is not required, i.e., if minibuffer-completion-confirm is nil. If confirmation is required, it is given by repeating this command immediately—the command is programmed to work without confirmation when run twice in succession.

Command: minibuffer-completion-help

This function creates a list of the possible completions of the current minibuffer contents. It works by calling all-completions using the value of the variable minibuffer-completion-table as the collection argument, and the value of minibuffer-completion-predicate as the predicate argument. The list of completions is displayed as text in a buffer named *Completions*.

Function: display-completion-list completions

This function displays completions to the stream in standard-output, usually a buffer. (See Reading and Printing Lisp Objects, for more information about streams.) The argument completions is normally a list of completions just returned by all-completions, but it does not have to be. Each element may be a symbol or a string, either of which is simply printed. It can also be a list of two strings, which is printed as if the strings were concatenated. The first of the two strings is the actual completion, the second string serves as annotation.

This function is called by minibuffer-completion-help. A common way to use it is together with with-output-to-temp-buffer, like this:

(with-output-to-temp-buffer "*Completions*"
  (display-completion-list
    (all-completions (buffer-string) my-alist)))
User Option: completion-auto-help

If this variable is non-nil, the completion commands automatically display a list of possible completions whenever nothing can be completed because the next character is not uniquely determined.

Variable: minibuffer-local-completion-map

completing-read uses this value as the local keymap when an exact match of one of the completions is not required. By default, this keymap makes the following bindings:

?

minibuffer-completion-help

SPC

minibuffer-complete-word

TAB

minibuffer-complete

and uses minibuffer-local-map as its parent keymap (see Definition of minibuffer-local-map).

Variable: minibuffer-local-must-match-map

completing-read uses this value as the local keymap when an exact match of one of the completions is required. Therefore, no keys are bound to exit-minibuffer, the command that exits the minibuffer unconditionally. By default, this keymap makes the following bindings:

C-j

minibuffer-complete-and-exit

RET

minibuffer-complete-and-exit

and uses minibuffer-local-completion-map as its parent keymap.

Variable: minibuffer-local-filename-completion-map

This is a sparse keymap that simply unbinds SPC; because filenames can contain spaces. The function read-file-name combines this keymap with either minibuffer-local-completion-map or minibuffer-local-must-match-map.

Variable: minibuffer-beginning-of-buffer-movement

If non-nil, the M-< command will move to the end of the prompt if point is after the end of the prompt. If point is at or before the end of the prompt, move to the start of the buffer. If this variable is nil, the command behaves like beginning-of-buffer.


21.6.4 High-Level Completion Functions

This section describes the higher-level convenience functions for reading certain sorts of names with completion.

In most cases, you should not call these functions in the middle of a Lisp function. When possible, do all minibuffer input as part of reading the arguments for a command, in the interactive specification. See Defining Commands.

Function: read-buffer prompt &optional default require-match predicate

This function reads the name of a buffer and returns it as a string. It prompts with prompt. The argument default is the default name to use, the value to return if the user exits with an empty minibuffer. If non-nil, it should be a string, a list of strings, or a buffer. If it is a list, the default value is the first element of this list. It is mentioned in the prompt, but is not inserted in the minibuffer as initial input.

The argument prompt should be a string ending with a colon and a space. If default is non-nil, the function inserts it in prompt before the colon to follow the convention for reading from the minibuffer with a default value (see Emacs Programming Tips).

The optional argument require-match has the same meaning as in completing-read. See Completion and the Minibuffer.

The optional argument predicate, if non-nil, specifies a function to filter the buffers that should be considered: the function will be called with every potential candidate as its argument, and should return nil to reject the candidate, non-nil to accept it.

In the following example, the user enters ‘minibuffer.t’, and then types RET. The argument require-match is t, and the only buffer name starting with the given input is ‘minibuffer.texi’, so that name is the value.

(read-buffer "Buffer name: " "foo" t)
;; After evaluation of the preceding expression,
;;   the following prompt appears,
;;   with an empty minibuffer:

---------- Buffer: Minibuffer ----------
Buffer name (default foo): ∗
---------- Buffer: Minibuffer ----------

;; The user types minibuffer.t RET.
     ⇒ "minibuffer.texi"
User Option: read-buffer-function

This variable, if non-nil, specifies a function for reading buffer names. read-buffer calls this function instead of doing its usual work, with the same arguments passed to read-buffer.

User Option: read-buffer-completion-ignore-case

If this variable is non-nil, read-buffer ignores case when performing completion while reading the buffer name.

Function: read-command prompt &optional default

This function reads the name of a command and returns it as a Lisp symbol. The argument prompt is used as in read-from-minibuffer. Recall that a command is anything for which commandp returns t, and a command name is a symbol for which commandp returns t. See Interactive Call.

The argument default specifies what to return if the user enters null input. It can be a symbol, a string or a list of strings. If it is a string, read-command interns it before returning it. If it is a list, read-command interns the first element of this list. If default is nil, that means no default has been specified; then if the user enters null input, the return value is (intern ""), that is, a symbol whose name is an empty string, and whose printed representation is ## (see Symbol Type).

(read-command "Command name? ")

;; After evaluation of the preceding expression,
;;   the following prompt appears with an empty minibuffer:

---------- Buffer: Minibuffer ----------
Command name?
---------- Buffer: Minibuffer ----------

If the user types forward-c RET, then this function returns forward-char.

The read-command function is a simplified interface to completing-read. It uses the variable obarray so as to complete in the set of extant Lisp symbols, and it uses the commandp predicate so as to accept only command names:

(read-command prompt)
≡
(intern (completing-read prompt obarray
                         'commandp t nil))
Function: read-variable prompt &optional default

This function reads the name of a customizable variable and returns it as a symbol. Its arguments have the same form as those of read-command. It behaves just like read-command, except that it uses the predicate custom-variable-p instead of commandp.

Command: read-color &optional prompt convert allow-empty display

This function reads a string that is a color specification, either the color’s name or an RGB hex value such as #RRRGGGBBB. It prompts with prompt (default: "Color (name or #RGB triplet):") and provides completion for color names, but not for hex RGB values. In addition to names of standard colors, completion candidates include the foreground and background colors at point.

Valid RGB values are described in Color Names.

The function’s return value is the string typed by the user in the minibuffer. However, when called interactively or if the optional argument convert is non-nil, it converts any input color name into the corresponding RGB value string and instead returns that. This function requires a valid color specification to be input. Empty color names are allowed when allow-empty is non-nil and the user enters null input.

Interactively, or when display is non-nil, the return value is also displayed in the echo area.

See also the functions read-coding-system and read-non-nil-coding-system, in User-Chosen Coding Systems, and read-input-method-name, in Input Methods.


21.6.5 Reading File Names

The high-level completion functions read-file-name, read-directory-name, and read-shell-command are designed to read file names, directory names, and shell commands, respectively. They provide special features, including automatic insertion of the default directory.

Function: read-file-name prompt &optional directory default require-match initial predicate

This function reads a file name, prompting with prompt and providing completion.

As an exception, this function reads a file name using a graphical file dialog instead of the minibuffer, if all of the following are true:

  1. It is invoked via a mouse command.
  2. The selected frame is on a graphical display supporting such dialogs.
  3. The variable use-dialog-box is non-nil. See Dialog Boxes in The GNU Emacs Manual.
  4. The directory argument, described below, does not specify a remote file. See Remote Files in The GNU Emacs Manual.

The exact behavior when using a graphical file dialog is platform-dependent. Here, we simply document the behavior when using the minibuffer.

read-file-name does not automatically expand the returned file name. You can call expand-file-name yourself if an absolute file name is required.

The optional argument require-match has the same meaning as in completing-read. See Completion and the Minibuffer.

The argument directory specifies the directory to use for completing relative file names. It should be an absolute directory name. If the variable insert-default-directory is non-nil, directory is also inserted in the minibuffer as initial input. It defaults to the current buffer’s value of default-directory.

If you specify initial, that is an initial file name to insert in the buffer (after directory, if that is inserted). In this case, point goes at the beginning of initial. The default for initial is nil—don’t insert any file name. To see what initial does, try the command C-x C-v in a buffer visiting a file. Please note: we recommend using default rather than initial in most cases.

If default is non-nil, then the function returns default if the user exits the minibuffer with the same non-empty contents that read-file-name inserted initially. The initial minibuffer contents are always non-empty if insert-default-directory is non-nil, as it is by default. default is not checked for validity, regardless of the value of require-match. However, if require-match is non-nil, the initial minibuffer contents should be a valid file (or directory) name. Otherwise read-file-name attempts completion if the user exits without any editing, and does not return default. default is also available through the history commands.

If default is nil, read-file-name tries to find a substitute default to use in its place, which it treats in exactly the same way as if it had been specified explicitly. If default is nil, but initial is non-nil, then the default is the absolute file name obtained from directory and initial. If both default and initial are nil and the buffer is visiting a file, read-file-name uses the absolute file name of that file as default. If the buffer is not visiting a file, then there is no default. In that case, if the user types RET without any editing, read-file-name simply returns the pre-inserted contents of the minibuffer.

If the user types RET in an empty minibuffer, this function returns an empty string, regardless of the value of require-match. This is, for instance, how the user can make the current buffer visit no file using M-x set-visited-file-name.

If predicate is non-nil, it specifies a function of one argument that decides which file names are acceptable completion alternatives. A file name is an acceptable value if predicate returns non-nil for it.

Here is an example of using read-file-name:

(read-file-name "The file is ")

;; After evaluation of the preceding expression,
;;   the following appears in the minibuffer:

---------- Buffer: Minibuffer ----------
The file is /gp/gnu/elisp/∗
---------- Buffer: Minibuffer ----------

Typing manual TAB results in the following:

---------- Buffer: Minibuffer ----------
The file is /gp/gnu/elisp/manual.texi∗
---------- Buffer: Minibuffer ----------

If the user types RET, read-file-name returns the file name as the string "/gp/gnu/elisp/manual.texi".

Variable: read-file-name-function

If non-nil, this should be a function that accepts the same arguments as read-file-name. When read-file-name is called, it calls this function with the supplied arguments instead of doing its usual work.

User Option: read-file-name-completion-ignore-case

If this variable is non-nil, read-file-name ignores case when performing completion.

Function: read-directory-name prompt &optional directory default require-match initial

This function is like read-file-name but allows only directory names as completion alternatives.

If default is nil and initial is non-nil, read-directory-name constructs a substitute default by combining directory (or the current buffer’s default directory if directory is nil) and initial. If both default and initial are nil, this function uses directory as substitute default, or the current buffer’s default directory if directory is nil.

User Option: insert-default-directory

This variable is used by read-file-name, and thus, indirectly, by most commands reading file names. (This includes all commands that use the code letters ‘f’ or ‘F’ in their interactive form. See Code Characters for interactive.) Its value controls whether read-file-name starts by placing the name of the default directory in the minibuffer, plus the initial file name, if any. If the value of this variable is nil, then read-file-name does not place any initial input in the minibuffer (unless you specify initial input with the initial argument). In that case, the default directory is still used for completion of relative file names, but is not displayed.

If this variable is nil and the initial minibuffer contents are empty, the user may have to explicitly fetch the next history element to access a default value. If the variable is non-nil, the initial minibuffer contents are always non-empty and the user can always request a default value by immediately typing RET in an unedited minibuffer. (See above.)

For example:

;; Here the minibuffer starts out with the default directory.
(let ((insert-default-directory t))
  (read-file-name "The file is "))

---------- Buffer: Minibuffer ----------
The file is ~lewis/manual/∗
---------- Buffer: Minibuffer ----------

;; Here the minibuffer is empty and only the prompt
;;   appears on its line.
(let ((insert-default-directory nil))
  (read-file-name "The file is "))

---------- Buffer: Minibuffer ----------
The file is ∗
---------- Buffer: Minibuffer ----------
Function: read-shell-command prompt &optional initial history &rest args

This function reads a shell command from the minibuffer, prompting with prompt and providing intelligent completion. It completes the first word of the command using candidates that are appropriate for command names, and the rest of the command words as file names.

This function uses minibuffer-local-shell-command-map as the keymap for minibuffer input. The history argument specifies the history list to use; if is omitted or nil, it defaults to shell-command-history (see shell-command-history). The optional argument initial specifies the initial content of the minibuffer (see Initial Input). The rest of args, if present, are used as the default and inherit-input-method arguments in read-from-minibuffer (see Reading Text Strings with the Minibuffer).

Variable: minibuffer-local-shell-command-map

This keymap is used by read-shell-command for completing command and file names that are part of a shell command. It uses minibuffer-local-map as its parent keymap, and binds TAB to completion-at-point.


21.6.6 Completion Variables

Here are some variables that can be used to alter the default completion behavior.

User Option: completion-styles

The value of this variable is a list of completion style (symbols) to use for performing completion. A completion style is a set of rules for generating completions. Each symbol occurring this list must have a corresponding entry in completion-styles-alist.

Variable: completion-styles-alist

This variable stores a list of available completion styles. Each element in the list has the form

(style try-completion all-completions doc)

Here, style is the name of the completion style (a symbol), which may be used in the completion-styles variable to refer to this style; try-completion is the function that does the completion; all-completions is the function that lists the completions; and doc is a string describing the completion style.

The try-completion and all-completions functions should each accept four arguments: string, collection, predicate, and point. The string, collection, and predicate arguments have the same meanings as in try-completion (see Basic Completion Functions), and the point argument is the position of point within string. Each function should return a non-nil value if it performed its job, and nil if it did not (e.g., if there is no way to complete string according to the completion style).

When the user calls a completion command like minibuffer-complete (see Minibuffer Commands that Do Completion), Emacs looks for the first style listed in completion-styles and calls its try-completion function. If this function returns nil, Emacs moves to the next listed completion style and calls its try-completion function, and so on until one of the try-completion functions successfully performs completion and returns a non-nil value. A similar procedure is used for listing completions, via the all-completions functions.

See Completion Styles in The GNU Emacs Manual, for a description of the available completion styles.

User Option: completion-category-overrides

This variable specifies special completion styles and other completion behaviors to use when completing certain types of text. Its value should be an alist with elements of the form (category . alist). category is a symbol describing what is being completed; currently, the buffer, file, and unicode-name categories are defined, but others can be defined via specialized completion functions (see Programmed Completion). alist is an association list describing how completion should behave for the corresponding category. The following alist keys are supported:

styles

The value should be a list of completion styles (symbols).

cycle

The value should be a value for completion-cycle-threshold (see Completion Options in The GNU Emacs Manual) for this category.

Additional alist entries may be defined in the future.

Variable: completion-extra-properties

This variable is used to specify extra properties of the current completion command. It is intended to be let-bound by specialized completion commands. Its value should be a list of property and value pairs. The following properties are supported:

:annotation-function

The value should be a function to add annotations in the completions buffer. This function must accept one argument, a completion, and should either return nil or a string to be displayed next to the completion. Unless this function puts own face on the annotation suffix string, the completions-annotations face is added by default to that string.

:affixation-function

The value should be a function to add prefixes and suffixes to completions. This function must accept one argument, a list of completions, and should return a list of annotated completions. Each element of the returned list must be a three-element list, the completion, a prefix string, and a suffix string. This function takes priority over :annotation-function.

:exit-function

The value should be a function to run after performing completion. The function should accept two arguments, string and status, where string is the text to which the field was completed, and status indicates what kind of operation happened: finished if text is now complete, sole if the text cannot be further completed but completion is not finished, or exact if the text is a valid completion but may be further completed.


21.6.7 Programmed Completion

Sometimes it is not possible or convenient to create an alist or an obarray containing all the intended possible completions ahead of time. In such a case, you can supply your own function to compute the completion of a given string. This is called programmed completion. Emacs uses programmed completion when completing file names (see File Name Completion), among many other cases.

To use this feature, pass a function as the collection argument to completing-read. The function completing-read arranges to pass your completion function along to try-completion, all-completions, and other basic completion functions, which will then let your function do all the work.

The completion function should accept three arguments:

  • The string to be completed.
  • A predicate function with which to filter possible matches, or nil if none. The function should call the predicate for each possible match, and ignore the match if the predicate returns nil.
  • A flag specifying the type of completion operation to perform; see Basic Completion Functions, for the details of those operations. This flag may be one of the following values.
    nil

    This specifies a try-completion operation. The function should return nil if there are no matches; it should return t if the specified string is a unique and exact match; and it should return the longest common prefix substring of all matches otherwise.

    t

    This specifies an all-completions operation. The function should return a list of all possible completions of the specified string.

    lambda

    This specifies a test-completion operation. The function should return t if the specified string is an exact match for some completion alternative; nil otherwise.

    (boundaries . suffix)

    This specifies a completion-boundaries operation. The function should return (boundaries start . end), where start is the position of the beginning boundary in the specified string, and end is the position of the end boundary in suffix.

    If a Lisp program returns nontrivial boundaries, it should make sure that the all-completions operation is consistent with them. The completions returned by all-completions should only pertain to the piece of the prefix and suffix covered by the completion boundaries. See Basic Completion Functions, for the precise expected semantics of completion boundaries.

    metadata

    This specifies a request for information about the state of the current completion. The return value should have the form (metadata . alist), where alist is an alist whose elements are described below.

    If the flag has any other value, the completion function should return nil.

The following is a list of metadata entries that a completion function may return in response to a metadata flag argument:

category

The value should be a symbol describing what kind of text the completion function is trying to complete. If the symbol matches one of the keys in completion-category-overrides, the usual completion behavior is overridden. See Completion Variables.

annotation-function

The value should be a function for annotating completions. The function should take one argument, string, which is a possible completion. It should return a string, which is displayed after the completion string in the *Completions* buffer. Unless this function puts own face on the annotation suffix string, the completions-annotations face is added by default to that string.

affixation-function

The value should be a function for adding prefixes and suffixes to completions. The function should take one argument, completions, which is a list of possible completions. It should return such a list of completions where each element contains a list of three elements: a completion, a prefix which is displayed before the completion string in the *Completions* buffer, and a suffix displayed after the completion string. This function takes priority over annotation-function.

group-function

The value should be a function for grouping the completion candidates. The function must take two arguments, completion, which is a completion candidate and transform, which is a boolean flag. If transform is nil, the function must return the group title of the group to which the candidate belongs. The returned title can also be nil. Otherwise the function must return the transformed candidate. The transformation can for example remove a redundant prefix, which is displayed in the group title.

display-sort-function

The value should be a function for sorting completions. The function should take one argument, a list of completion strings, and return a sorted list of completion strings. It is allowed to alter the input list destructively.

cycle-sort-function

The value should be a function for sorting completions, when completion-cycle-threshold is non-nil and the user is cycling through completion alternatives. See Completion Options in The GNU Emacs Manual. Its argument list and return value are the same as for display-sort-function.

Function: completion-table-dynamic function &optional switch-buffer

This function is a convenient way to write a function that can act as a programmed completion function. The argument function should be a function that takes one argument, a string, and returns a completion table (see Basic Completion Functions) containing all the possible completions. The table returned by function can also include elements that don’t match the string argument; they are automatically filtered out by completion-table-dynamic. In particular, function can ignore its argument and return a full list of all possible completions. You can think of completion-table-dynamic as a transducer between function and the interface for programmed completion functions.

If the optional argument switch-buffer is non-nil, and completion is performed in the minibuffer, function will be called with current buffer set to the buffer from which the minibuffer was entered.

The return value of completion-table-dynamic is a function that can be used as the 2nd argument to try-completion and all-completions. Note that this function will always return empty metadata and trivial boundaries.

Function: completion-table-with-cache function &optional ignore-case

This is a wrapper for completion-table-dynamic that saves the last argument-result pair. This means that multiple lookups with the same argument only need to call function once. This can be useful when a slow operation is involved, such as calling an external process.


21.6.8 Completion in Ordinary Buffers

Although completion is usually done in the minibuffer, the completion facility can also be used on the text in ordinary Emacs buffers. In many major modes, in-buffer completion is performed by the C-M-i or M-TAB command, bound to completion-at-point. See Symbol Completion in The GNU Emacs Manual. This command uses the abnormal hook variable completion-at-point-functions:

Variable: completion-at-point-functions

The value of this abnormal hook should be a list of functions, which are used to compute a completion table (see Basic Completion Functions) for completing the text at point. It can be used by major modes to provide mode-specific completion tables (see Major Mode Conventions).

When the command completion-at-point runs, it calls the functions in the list one by one, without any argument. Each function should return nil unless it can and wants to take responsibility for the completion data for the text at point. Otherwise it should return a list of the following form:

(start end collection . props)

start and end delimit the text to complete (which should enclose point). collection is a completion table for completing that text, in a form suitable for passing as the second argument to try-completion (see Basic Completion Functions); completion alternatives will be generated from this completion table in the usual way, via the completion styles defined in completion-styles (see Completion Variables). props is a property list for additional information; any of the properties in completion-extra-properties are recognized (see Completion Variables), as well as the following additional ones:

:predicate

The value should be a predicate that completion candidates need to satisfy.

:exclusive

If the value is no, then if the completion table fails to match the text at point, completion-at-point moves on to the next function in completion-at-point-functions instead of reporting a completion failure.

The functions on this hook should generally return quickly, since they may be called very often (e.g., from post-command-hook). Supplying a function for collection is strongly recommended if generating the list of completions is an expensive operation. Emacs may internally call functions in completion-at-point-functions many times, but care about the value of collection for only some of these calls. By supplying a function for collection, Emacs can defer generating completions until necessary. You can use completion-table-dynamic to create a wrapper function:

;; Avoid this pattern.
(let ((beg ...) (end ...) (my-completions (my-make-completions)))
  (list beg end my-completions))

;; Use this instead.
(let ((beg ...) (end ...))
  (list beg
        end
        (completion-table-dynamic
          (lambda (_)
            (my-make-completions)))))

Additionally, the collection should generally not be pre-filtered based on the current text between start and end, because that is the responsibility of the caller of completion-at-point-functions to do that according to the completion styles it decides to use.

A function in completion-at-point-functions may also return a function instead of a list as described above. In that case, that returned function is called, with no argument, and it is entirely responsible for performing the completion. We discourage this usage; it is only intended to help convert old code to using completion-at-point.

The first function in completion-at-point-functions to return a non-nil value is used by completion-at-point. The remaining functions are not called. The exception to this is when there is an :exclusive specification, as described above.

The following function provides a convenient way to perform completion on an arbitrary stretch of text in an Emacs buffer:

Function: completion-in-region start end collection &optional predicate

This function completes the text in the current buffer between the positions start and end, using collection. The argument collection has the same meaning as in try-completion (see Basic Completion Functions).

This function inserts the completion text directly into the current buffer. Unlike completing-read (see Completion and the Minibuffer), it does not activate the minibuffer.

For this function to work, point must be somewhere between start and end.


21.7 Yes-or-No Queries

This section describes functions used to ask the user a yes-or-no question. The function y-or-n-p can be answered with a single character; it is useful for questions where an inadvertent wrong answer will not have serious consequences. yes-or-no-p is suitable for more momentous questions, since it requires three or four characters to answer.

If either of these functions is called in a command that was invoked using the mouse or some other window-system gesture, or in a command invoked via a menu, then they use a dialog box or pop-up menu to ask the question if dialog boxes are supported. Otherwise, they use keyboard input. You can force use either of the mouse or of keyboard input by binding last-nonmenu-event to a suitable value around the call—bind it to t to force keyboard interaction, and to a list to force dialog boxes.

Both yes-or-no-p and y-or-n-p use the minibuffer.

Function: y-or-n-p prompt

This function asks the user a question, expecting input in the minibuffer. It returns t if the user types y, nil if the user types n. This function also accepts SPC to mean yes and DEL to mean no. It accepts C-] and C-g to quit, because the question uses the minibuffer and for that reason the user might try to use C-] to get out. The answer is a single character, with no RET needed to terminate it. Upper and lower case are equivalent.

“Asking the question” means printing prompt in the minibuffer, followed by the string ‘(y or n) . If the input is not one of the expected answers (y, n, SPC, DEL, or something that quits), the function responds ‘Please answer y or n.’, and repeats the request.

This function actually uses the minibuffer, but does not allow editing of the answer. The cursor moves to the minibuffer while the question is being asked.

The answers and their meanings, even ‘y’ and ‘n’, are not hardwired, and are specified by the keymap query-replace-map (see Search and Replace). In particular, if the user enters the special responses recenter, scroll-up, scroll-down, scroll-other-window, or scroll-other-window-down (respectively bound to C-l, C-v, M-v, C-M-v and C-M-S-v in query-replace-map), this function performs the specified window recentering or scrolling operation, and poses the question again.

If you bind help-form (see Help Functions) to a non-nil value while calling y-or-n-p, then pressing help-char causes it to evaluate help-form and display the result. help-char is automatically added to prompt.

Function: y-or-n-p-with-timeout prompt seconds default

Like y-or-n-p, except that if the user fails to answer within seconds seconds, this function stops waiting and returns default. It works by setting up a timer; see Timers for Delayed Execution. The argument seconds should be a number.

Function: yes-or-no-p prompt

This function asks the user a question, expecting input in the minibuffer. It returns t if the user enters ‘yes’, nil if the user types ‘no’. The user must type RET to finalize the response. Upper and lower case are equivalent.

yes-or-no-p starts by displaying prompt in the minibuffer, followed by ‘(yes or no) . The user must type one of the expected responses; otherwise, the function responds ‘Please answer yes or no.’, waits about two seconds and repeats the request.

yes-or-no-p requires more work from the user than y-or-n-p and is appropriate for more crucial decisions.

Here is an example:

(yes-or-no-p "Do you really want to remove everything? ")

;; After evaluation of the preceding expression,
;;   the following prompt appears,
;;   with an empty minibuffer:

---------- Buffer: minibuffer ----------
Do you really want to remove everything? (yes or no)
---------- Buffer: minibuffer ----------

If the user first types y RET, which is invalid because this function demands the entire word ‘yes’, it responds by displaying these prompts, with a brief pause between them:

---------- Buffer: minibuffer ----------
Please answer yes or no.
Do you really want to remove everything? (yes or no)
---------- Buffer: minibuffer ----------

21.8 Asking Multiple-Choice Questions

This section describes facilities for asking the user more complex questions or several similar questions.

When you have a series of similar questions to ask, such as “Do you want to save this buffer?” for each buffer in turn, you should use map-y-or-n-p to ask the collection of questions, rather than asking each question individually. This gives the user certain convenient facilities such as the ability to answer the whole series at once.

Function: map-y-or-n-p prompter actor list &optional help action-alist no-cursor-in-echo-area

This function asks the user a series of questions, reading a single-character answer in the echo area for each one.

The value of list specifies the objects to ask questions about. It should be either a list of objects or a generator function. If it is a function, it will be called with no arguments, and should return either the next object to ask about, or nil, meaning to stop asking questions.

The argument prompter specifies how to ask each question. If prompter is a string, the question text is computed like this:

(format prompter object)

where object is the next object to ask about (as obtained from list). See Formatting Strings, for more information about format.

If prompter is not a string, it should be a function of one argument (the object to ask about) and should return the question text for that object. If the value prompter returns is a string, that is the question to ask the user. The function can also return t, meaning to act on this object without asking the user, or nil, which means to silently ignore this object.

The argument actor says how to act on the objects for which the user answers yes. It should be a function of one argument, and will be called with each object from list for which the user answers yes.

If the argument help is given, it should be a list of this form:

(singular plural action)

where singular is a string containing a singular noun that describes a single object to be acted on, plural is the corresponding plural noun, and action is a transitive verb describing what actor does with the objects.

If you don’t specify help, it defaults to the list ("object" "objects" "act on").

Each time a question is asked, the user can answer as follows:

y, Y, or SPC

act on the object

n, N, or DEL

skip the object

!

act on all the following objects

ESC or q

exit (skip all following objects)

. (period)

act on the object and then exit

C-h

get help

These are the same answers that query-replace accepts. The keymap query-replace-map defines their meaning for map-y-or-n-p as well as for query-replace; see Search and Replace.

You can use action-alist to specify additional possible answers and what they mean. If provided, action-alist should be an alist whose elements are of the form (char function help). Each of the alist elements defines one additional answer. In each element, char is a character (the answer); function is a function of one argument (an object from list); and help is a string. When the user responds with char, map-y-or-n-p calls function. If it returns non-nil, the object is considered to have been acted upon, and map-y-or-n-p advances to the next object in list. If it returns nil, the prompt is repeated for the same object. If the user requests help, the text in help is used to describe these additional answers.

Normally, map-y-or-n-p binds cursor-in-echo-area while prompting. But if no-cursor-in-echo-area is non-nil, it does not do that.

If map-y-or-n-p is called in a command that was invoked using the mouse or some other window-system gesture, or a command invoked via a menu, then it uses a dialog box or pop-up menu to ask the question if dialog boxes are supported. In this case, it does not use keyboard input or the echo area. You can force use either of the mouse or of keyboard input by binding last-nonmenu-event to a suitable value around the call—bind it to t to force keyboard interaction, and to a list to force dialog boxes.

The return value of map-y-or-n-p is the number of objects acted on.

If you need to ask the user a question that might have more than just 2 answers, use read-answer.

Function: read-answer question answers

This function prompts the user with text in question, which should end in the ‘SPC’ character. The function includes in the prompt the possible responses in answers by appending them to the end of question. The possible responses are provided in answers as an alist whose elements are of the following form:

(long-answer short-answer help-message)

where long-answer is the complete text of the user response, a string; short-answer is a short form of the same response, a single character or a function key; and help-message is the text that describes the meaning of the answer. If the variable read-answer-short is non-nil, the prompt will show the short variants of the possible answers and the user is expected to type the single characters/keys shown in the prompt; otherwise the prompt will show the long variants of the answers, and the user is expected to type the full text of one of the answers and end by pressing RET. If use-dialog-box is non-nil, and this function was invoked by mouse events, the question and the answers will be displayed in a GUI dialog box.

The function returns the text of the long-answer selected by the user, regardless of whether long or short answers were shown in the prompt and typed by the user.

Here is an example of using this function:

(let ((read-answer-short t))
  (read-answer "Foo "
     '(("yes"  ?y "perform the action")
       ("no"   ?n "skip to the next")
       ("all"  ?! "perform for the rest without more questions")
       ("help" ?h "show help")
       ("quit" ?q "exit"))))
Function: read-char-from-minibuffer prompt &optional chars history

This function uses the minibuffer to read and return a single character. Optionally, it ignores any input that is not a member of chars, a list of accepted characters. The history argument specifies the history list symbol to use; if it is omitted or nil, this function doesn’t use the history.

If you bind help-form (see Help Functions) to a non-nil value while calling read-char-from-minibuffer, then pressing help-char causes it to evaluate help-form and display the result.


21.9 Reading a Password

To read a password to pass to another program, you can use the function read-passwd.

Function: read-passwd prompt &optional confirm default

This function reads a password, prompting with prompt. It does not echo the password as the user types it; instead, it echoes ‘*’ for each character in the password. If you want to apply another character to hide the password, let-bind the variable read-hide-char with that character.

The optional argument confirm, if non-nil, says to read the password twice and insist it must be the same both times. If it isn’t the same, the user has to type it over and over until the last two times match.

The optional argument default specifies the default password to return if the user enters empty input. If default is nil, then read-passwd returns the null string in that case.


21.10 Minibuffer Commands

This section describes some commands meant for use in the minibuffer.

Command: exit-minibuffer

This command exits the active minibuffer. It is normally bound to keys in minibuffer local keymaps. The command throws an error if the current buffer is a minibuffer, but not the active minibuffer.

Command: self-insert-and-exit

This command exits the active minibuffer after inserting the last character typed on the keyboard (found in last-command-event; see Information from the Command Loop).

Command: previous-history-element n

This command replaces the minibuffer contents with the value of the nth previous (older) history element.

Command: next-history-element n

This command replaces the minibuffer contents with the value of the nth more recent history element. The position in the history can go beyond the current position and invoke “future history” (see Reading Text Strings with the Minibuffer).

Command: previous-matching-history-element pattern n

This command replaces the minibuffer contents with the value of the nth previous (older) history element that matches pattern (a regular expression).

Command: next-matching-history-element pattern n

This command replaces the minibuffer contents with the value of the nth next (newer) history element that matches pattern (a regular expression).

Command: previous-complete-history-element n

This command replaces the minibuffer contents with the value of the nth previous (older) history element that completes the current contents of the minibuffer before the point.

Command: next-complete-history-element n

This command replaces the minibuffer contents with the value of the nth next (newer) history element that completes the current contents of the minibuffer before the point.

Command: goto-history-element nabs

This function puts element of the minibuffer history in the minibuffer. The argument nabs specifies the absolute history position in descending order, where 0 means the current element and a positive number n means the nth previous element. NABS being a negative number -n means the nth entry of “future history”. When this function reaches the end of the default values provided by read-from-minibuffer (see Reading Text Strings with the Minibuffer) and completing-read (see Completion and the Minibuffer), it adds the completion candidates to “future history”, see minibuffer-default-add-function.


21.11 Minibuffer Windows

These functions access and select minibuffer windows, test whether they are active and control how they get resized.

Function: minibuffer-window &optional frame

This function returns the minibuffer window used for frame frame. If frame is nil, that stands for the selected frame.

Note that the minibuffer window used by a frame need not be part of that frame—a frame that has no minibuffer of its own necessarily uses some other frame’s minibuffer window. The minibuffer window of a minibuffer-less frame can be changed by setting that frame’s minibuffer frame parameter (see Buffer Parameters).

Function: set-minibuffer-window window

This function specifies window as the minibuffer window to use. This affects where the minibuffer is displayed if you put text in it without invoking the usual minibuffer commands. It has no effect on the usual minibuffer input functions because they all start by choosing the minibuffer window according to the selected frame.

Function: window-minibuffer-p &optional window

This function returns t if window is a minibuffer window. window defaults to the selected window.

The following function returns the window showing the currently active minibuffer.

Function: active-minibuffer-window

This function returns the window of the currently active minibuffer, or nil if there is no active minibuffer.

It is not sufficient to determine whether a given window shows the currently active minibuffer by comparing it with the result of (minibuffer-window), because there can be more than one minibuffer window if there is more than one frame.

Function: minibuffer-window-active-p window

This function returns non-nil if window shows the currently active minibuffer.

The following two options control whether minibuffer windows are resized automatically and how large they can get in the process.

User Option: resize-mini-windows

This option specifies whether minibuffer windows are resized automatically. The default value is grow-only, which means that a minibuffer window by default expands automatically to accommodate the text it displays and shrinks back to one line as soon as the minibuffer gets empty. If the value is t, Emacs will always try to fit the height of a minibuffer window to the text it displays (with a minimum of one line). If the value is nil, a minibuffer window never changes size automatically. In that case the window resizing commands (see Resizing Windows) can be used to adjust its height.

User Option: max-mini-window-height

This option provides a maximum height for resizing minibuffer windows automatically. A floating-point number specifies the maximum height as a fraction of the frame’s height; an integer specifies the maximum height in units of the frame’s canonical character height (see Frame Font). The default value is 0.25.

Note that the values of the above two variables take effect at display time, so let-binding them around code which produces echo-area messages will not work. If you want to prevent resizing of minibuffer windows when displaying long messages, bind the message-truncate-lines variable instead (see Echo Area Customization).

The option resize-mini-windows does not affect the behavior of minibuffer-only frames (see Frame Layout). The following option allows to automatically resize such frames as well.

User Option: resize-mini-frames

If this is nil, minibuffer-only frames are never resized automatically.

If this is a function, that function is called with the minibuffer-only frame to be resized as sole argument. At the time this function is called, the buffer of the minibuffer window of that frame is the buffer whose contents will be shown the next time that window is redisplayed. The function is expected to fit the frame to the buffer in some appropriate way.

Any other non-nil value means to resize minibuffer-only frames by calling fit-mini-frame-to-buffer, a function that behaves like fit-frame-to-buffer (see Resizing Windows) but does not strip leading or trailing empty lines from the buffer text.


21.12 Minibuffer Contents

These functions access the minibuffer prompt and contents.

Function: minibuffer-prompt

This function returns the prompt string of the currently active minibuffer. If no minibuffer is active, it returns nil.

Function: minibuffer-prompt-end

This function returns the current position of the end of the minibuffer prompt, if a minibuffer is current. Otherwise, it returns the minimum valid buffer position.

Function: minibuffer-prompt-width

This function returns the current display-width of the minibuffer prompt, if a minibuffer is current. Otherwise, it returns zero.

Function: minibuffer-contents

This function returns the editable contents of the minibuffer (that is, everything except the prompt) as a string, if a minibuffer is current. Otherwise, it returns the entire contents of the current buffer.

Function: minibuffer-contents-no-properties

This is like minibuffer-contents, except that it does not copy text properties, just the characters themselves. See Text Properties.

Command: delete-minibuffer-contents

This command erases the editable contents of the minibuffer (that is, everything except the prompt), if a minibuffer is current. Otherwise, it erases the entire current buffer.


21.13 Recursive Minibuffers

These functions and variables deal with recursive minibuffers (see Recursive Editing):

Function: minibuffer-depth

This function returns the current depth of activations of the minibuffer, a nonnegative integer. If no minibuffers are active, it returns zero.

User Option: enable-recursive-minibuffers

If this variable is non-nil, you can invoke commands (such as find-file) that use minibuffers even while the minibuffer is active. Such invocation produces a recursive editing level for a new minibuffer. By default, the outer-level minibuffer is invisible while you are editing the inner one. If you have minibuffer-follows-selected-frame set to nil, you can have minibuffers visible on several frames at the same time. See (emacs)Basic Minibuffer.

If this variable is nil, you cannot invoke minibuffer commands when the minibuffer is active, not even if you switch to another window to do it.

If a command name has a property enable-recursive-minibuffers that is non-nil, then the command can use the minibuffer to read arguments even if it is invoked from the minibuffer. A command can also achieve this by binding enable-recursive-minibuffers to t in the interactive declaration (see Using interactive). The minibuffer command next-matching-history-element (normally M-s in the minibuffer) does the latter.


21.14 Inhibiting Interaction

It’s sometimes useful to be able to run Emacs as a headless server process that responds to commands given over a network connection. However, Emacs is primarily a platform for interactive usage, so many commands prompt the user for feedback in certain anomalous situations. This makes this use case more difficult, since the server process will just hang waiting for user input.

Binding the inhibit-interaction variable to something non-nil makes Emacs signal a inhibited-interaction error instead of prompting, which can then be used by the server process to handle these situations.

Here’s a typical use case:

(let ((inhibit-interaction t))
  (respond-to-client
   (condition-case err
       (my-client-handling-function)
     (inhibited-interaction err))))

If my-client-handling-function ends up calling something that asks the user for something (via y-or-n-p or read-from-minibuffer or the like), an inhibited-interaction error is signaled instead. The server code then catches that error and reports it to the client.


21.15 Minibuffer Miscellany

Function: minibufferp &optional buffer-or-name live

This function returns non-nil if buffer-or-name is a minibuffer. If buffer-or-name is omitted or nil, it tests the current buffer. When live is non-nil, the function returns non-nil only when buffer-or-name is an active minibuffer.

Variable: minibuffer-setup-hook

This is a normal hook that is run whenever a minibuffer is entered. See Hooks.

Macro: minibuffer-with-setup-hook function &rest body

This macro executes body after arranging for the specified function to be called via minibuffer-setup-hook. By default, function is called before the other functions in the minibuffer-setup-hook list, but if function is of the form (:append func), func will be called after the other hook functions.

The body forms should not use the minibuffer more than once. If the minibuffer is re-entered recursively, function will only be called once, for the outermost use of the minibuffer.

Variable: minibuffer-exit-hook

This is a normal hook that is run whenever a minibuffer is exited. See Hooks.

Variable: minibuffer-help-form

The current value of this variable is used to rebind help-form locally inside the minibuffer (see Help Functions).

Variable: minibuffer-scroll-window

If the value of this variable is non-nil, it should be a window object. When the function scroll-other-window is called in the minibuffer, it scrolls this window (see Textual Scrolling).

Function: minibuffer-selected-window

This function returns the window that was selected just before the minibuffer window was selected. If the selected window is not a minibuffer window, it returns nil.

Function: minibuffer-message string &rest args

This function is like message (see Displaying Messages in the Echo Area), but it displays the messages specially when the user types in the minibuffer, typically because Emacs prompted the user for some input. When the minibuffer is the current buffer, this function displays the message specified by string temporarily at the end of the minibuffer text, and thus avoids hiding the minibuffer text by the echo-area display of the message. It leaves the message on display for a few seconds, or until the next input event arrives, whichever comes first. The variable minibuffer-message-timeout specifies the number of seconds to wait in the absence of input. It defaults to 2. If args is non-nil, the actual message is obtained by passing string and args through format-message. See Formatting Strings.

If called when the minibuffer is not the current buffer, this function just calls message, and thus string will be shown in the echo-area.

Command: minibuffer-inactive-mode

This is the major mode used in inactive minibuffers. It uses keymap minibuffer-inactive-mode-map. This can be useful if the minibuffer is in a separate frame. See Minibuffers and Frames.


22 Command Loop

When you run Emacs, it enters the editor command loop almost immediately. This loop reads key sequences, executes their definitions, and displays the results. In this chapter, we describe how these things are done, and the subroutines that allow Lisp programs to do them.


22.1 Command Loop Overview

The first thing the command loop must do is read a key sequence, which is a sequence of input events that translates into a command. It does this by calling the function read-key-sequence. Lisp programs can also call this function (see Key Sequence Input). They can also read input at a lower level with read-key or read-event (see Reading One Event), or discard pending input with discard-input (see Miscellaneous Event Input Features).

The key sequence is translated into a command through the currently active keymaps. See Key Lookup, for information on how this is done. The result should be a keyboard macro or an interactively callable function. If the key is M-x, then it reads the name of another command, which it then calls. This is done by the command execute-extended-command (see Interactive Call).

Prior to executing the command, Emacs runs undo-boundary to create an undo boundary. See Maintaining Undo Lists.

To execute a command, Emacs first reads its arguments by calling command-execute (see Interactive Call). For commands written in Lisp, the interactive specification says how to read the arguments. This may use the prefix argument (see Prefix Command Arguments) or may read with prompting in the minibuffer (see Minibuffers). For example, the command find-file has an interactive specification which says to read a file name using the minibuffer. The function body of find-file does not use the minibuffer, so if you call find-file as a function from Lisp code, you must supply the file name string as an ordinary Lisp function argument.

If the command is a keyboard macro (i.e., a string or vector), Emacs executes it using execute-kbd-macro (see Keyboard Macros).

Variable: pre-command-hook

This normal hook is run by the editor command loop before it executes each command. At that time, this-command contains the command that is about to run, and last-command describes the previous command. See Information from the Command Loop.

Variable: post-command-hook

This normal hook is run by the editor command loop after it executes each command (including commands terminated prematurely by quitting or by errors). At that time, this-command refers to the command that just ran, and last-command refers to the command before that.

This hook is also run when Emacs first enters the command loop (at which point this-command and last-command are both nil).

Quitting is suppressed while running pre-command-hook and post-command-hook. If an error happens while executing one of these hooks, it does not terminate execution of the hook; instead the error is silenced and the function in which the error occurred is removed from the hook.

A request coming into the Emacs server (see Emacs Server in The GNU Emacs Manual) runs these two hooks just as a keyboard command does.

Note that, when the buffer text includes very long lines, these two hooks are called as if they were in a with-restriction form (see Narrowing), with a long-line-optimizations-in-command-hooks label and with the buffer narrowed to a portion around point.


22.2 Defining Commands

The special form interactive turns a Lisp function into a command. The interactive form must be located at top-level in the function body, usually as the first form in the body; this applies to both lambda expressions (see Lambda Expressions) and defun forms (see Defining Functions). This form does nothing during the actual execution of the function; its presence serves as a flag, telling the Emacs command loop that the function can be called interactively. The argument of the interactive form specifies how the arguments for an interactive call should be read.

Alternatively, an interactive form may be specified in a function symbol’s interactive-form property. A non-nil value for this property takes precedence over any interactive form in the function body itself. This feature is seldom used.

Sometimes, a function is only intended to be called interactively, never directly from Lisp. In that case, give the function a non-nil interactive-only property, either directly or via declare (see The declare Form). This causes the byte compiler to warn if the command is called from Lisp. The output of describe-function will include similar information. The value of the property can be: a string, which the byte-compiler will use directly in its warning (it should end with a period, and not start with a capital, e.g., "use (system-name) instead."); t; any other symbol, which should be an alternative function to use in Lisp code.

Generic functions (see Generic Functions) cannot be turned into commands by adding the interactive form to them.


22.2.1 Using interactive

This section describes how to write the interactive form that makes a Lisp function an interactively-callable command, and how to examine a command’s interactive form.

Special Form: interactive &optional arg-descriptor &rest modes

This special form declares that a function is a command, and that it may therefore be called interactively (via M-x or by entering a key sequence bound to it). The argument arg-descriptor declares how to compute the arguments to the command when the command is called interactively.

A command may be called from Lisp programs like any other function, but then the caller supplies the arguments and arg-descriptor has no effect.

The interactive form must be located at top-level in the function body, or in the function symbol’s interactive-form property (see Symbol Properties). It has its effect because the command loop looks for it before calling the function (see Interactive Call). Once the function is called, all its body forms are executed; at this time, if the interactive form occurs within the body, the form simply returns nil without even evaluating its argument.

The modes list allows specifying which modes the command is meant to be used in. See Specifying Modes For Commands for more details about the effect of specifying modes, and when to use it.

By convention, you should put the interactive form in the function body, as the first top-level form. If there is an interactive form in both the interactive-form symbol property and the function body, the former takes precedence. The interactive-form symbol property can be used to add an interactive form to an existing function, or change how its arguments are processed interactively, without redefining the function.

There are three possibilities for the argument arg-descriptor:

  • It may be omitted or nil; then the command is called with no arguments. This leads quickly to an error if the command requires one or more arguments.
  • It may be a string; its contents are a sequence of elements separated by newlines, one for each argument16. Each element consists of a code character (see Code Characters for interactive) optionally followed by a prompt (which some code characters use and some ignore). Here is an example:
    (interactive "P\nbFrobnicate buffer: ")
    

    The code letter ‘P’ sets the command’s first argument to the raw command prefix (see Prefix Command Arguments). ‘bFrobnicate buffer: ’ prompts the user with ‘Frobnicate buffer: ’ to enter the name of an existing buffer, which becomes the second and final argument.

    The prompt string can use ‘%’ to include previous argument values (starting with the first argument) in the prompt. This is done using format-message (see Formatting Strings). For example, here is how you could read the name of an existing buffer followed by a new name to give to that buffer:

    (interactive "bBuffer to rename: \nsRename buffer %s to: ")
    

    If ‘*’ appears at the beginning of the string, then an error is signaled if the buffer is read-only.

    If ‘@’ appears at the beginning of the string, and if the key sequence used to invoke the command includes any mouse events, then the window associated with the first of those events is selected before the command is run.

    If ‘^’ appears at the beginning of the string, and if the command was invoked through shift-translation, set the mark and activate the region temporarily, or extend an already active region, before the command is run. If the command was invoked without shift-translation, and the region is temporarily active, deactivate the region before the command is run. Shift-translation is controlled on the user level by shift-select-mode; see Shift Selection in The GNU Emacs Manual.

    You can use ‘*’, ‘@’, and ^ together; the order does not matter. Actual reading of arguments is controlled by the rest of the prompt string (starting with the first character that is not ‘*’, ‘@’, or ‘^’).

  • It may be a Lisp expression that is not a string; then it should be a form that is evaluated to get a list of arguments to pass to the command. Usually this form will call various functions to read input from the user, most often through the minibuffer (see Minibuffers) or directly from the keyboard (see Reading Input).

    Providing point or the mark as an argument value is also common, but if you do this and read input (whether using the minibuffer or not), be sure to get the integer values of point or the mark after reading. The current buffer may be receiving subprocess output; if subprocess output arrives while the command is waiting for input, it could relocate point and the mark.

    Here’s an example of what not to do:

    (interactive
     (list (region-beginning) (region-end)
           (read-string "Foo: " nil 'my-history)))
    

    Here’s how to avoid the problem, by examining point and the mark after reading the keyboard input:

    (interactive
     (let ((string (read-string "Foo: " nil 'my-history)))
       (list (region-beginning) (region-end) string)))
    

    Warning: the argument values should not include any data types that can’t be printed and then read. Some facilities save command-history in a file to be read in the subsequent sessions; if a command’s arguments contain a data type that prints using ‘#<…>’ syntax, those facilities won’t work.

    There are, however, a few exceptions: it is ok to use a limited set of expressions such as (point), (mark), (region-beginning), and (region-end), because Emacs recognizes them specially and puts the expression (rather than its value) into the command history. To see whether the expression you wrote is one of these exceptions, run the command, then examine (car command-history).

Function: interactive-form function

This function returns the interactive form of function. If function is an interactively callable function (see Interactive Call), the value is the command’s interactive form (interactive spec), which specifies how to compute its arguments. Otherwise, the value is nil. If function is a symbol, its function definition is used. When called on an OClosure, the work is delegated to the generic function oclosure-interactive-form.

Function: oclosure-interactive-form function

Just like interactive-form, this function takes a command and returns its interactive form. The difference is that it is a generic function and it is only called when function is an OClosure (see Open Closures). The purpose is to make it possible for some OClosure types to compute their interactive forms dynamically instead of carrying it in one of their slots.

This is used for example for kmacro functions in order to reduce their memory size, since they all share the same interactive form. It is also used for advice functions, where the interactive form is computed from the interactive forms of its components, so as to make this computation more lazily and to correctly adjust the interactive form when one of its component’s is redefined.


22.2.2 Code Characters for interactive

The code character descriptions below contain a number of key words, defined here as follows:

Completion

Provide completion. TAB, SPC, and RET perform name completion because the argument is read using completing-read (see Completion). ? displays a list of possible completions.

Existing

Require the name of an existing object. An invalid name is not accepted; the commands to exit the minibuffer do not exit if the current input is not valid.

Default

A default value of some sort is used if the user enters no text in the minibuffer. The default depends on the code character.

No I/O

This code letter computes an argument without reading any input. Therefore, it does not use a prompt string, and any prompt string you supply is ignored.

Even though the code letter doesn’t use a prompt string, you must follow it with a newline if it is not the last code character in the string.

Prompt

A prompt immediately follows the code character. The prompt ends either with the end of the string or with a newline.

Special

This code character is meaningful only at the beginning of the interactive string, and it does not look for a prompt or a newline. It is a single, isolated character.

Here are the code character descriptions for use with interactive:

*

Signal an error if the current buffer is read-only. Special.

@

Select the window mentioned in the first mouse event in the key sequence that invoked this command. Special.

^

If the command was invoked through shift-translation, set the mark and activate the region temporarily, or extend an already active region, before the command is run. If the command was invoked without shift-translation, and the region is temporarily active, deactivate the region before the command is run. Special.

a

A function name (i.e., a symbol satisfying fboundp). Existing, Completion, Prompt.

b

The name of an existing buffer. By default, uses the name of the current buffer (see Buffers). Existing, Completion, Default, Prompt.

B

A buffer name. The buffer need not exist. By default, uses the name of a recently used buffer other than the current buffer. Completion, Default, Prompt.

c

A character. The cursor does not move into the echo area. Prompt.

C

A command name (i.e., a symbol satisfying commandp). Existing, Completion, Prompt.

d

The position of point, as an integer (see Point). No I/O.

D

A directory. The default is the current default directory of the current buffer, default-directory (see Functions that Expand Filenames). Existing, Completion, Default, Prompt.

e

The first or next non-keyboard event in the key sequence that invoked the command. More precisely, ‘e’ gets events that are lists, so you can look at the data in the lists. See Input Events. No I/O.

You use ‘e’ for mouse events and for special system events (see Miscellaneous System Events). The event list that the command receives depends on the event. See Input Events, which describes the forms of the list for each event in the corresponding subsections.

You can use ‘e’ more than once in a single command’s interactive specification. If the key sequence that invoked the command has n events that are lists, the nth ‘e’ provides the nth such event. Events that are not lists, such as function keys and ASCII characters, do not count where ‘e’ is concerned.

f

A file name of an existing file (see File Names). See Reading File Names, for details about default values. Existing, Completion, Default, Prompt.

F

A file name. The file need not exist. Completion, Default, Prompt.

G

A file name. The file need not exist. If the user enters just a directory name, then the value is just that directory name, with no file name within the directory added. Completion, Default, Prompt.

i

An irrelevant argument. This code always supplies nil as the argument’s value. No I/O.

k

A key sequence (see Key Sequences). This keeps reading events until a command (or undefined command) is found in the current key maps. The key sequence argument is represented as a string or vector. The cursor does not move into the echo area. Prompt.

If ‘k’ reads a key sequence that ends with a down-event, it also reads and discards the following up-event. You can get access to that up-event with the ‘U’ code character.

This kind of input is used by commands such as describe-key and keymap-global-set.

K

A key sequence on a form that can be used as input to functions like keymap-set. This works like ‘k’, except that it suppresses, for the last input event in the key sequence, the conversions that are normally used (when necessary) to convert an undefined key into a defined one (see Key Sequence Input), so this form is usually used when prompting for a new key sequence that is to be bound to a command.

m

The position of the mark, as an integer. No I/O.

M

Arbitrary text, read in the minibuffer using the current buffer’s input method, and returned as a string (see Input Methods in The GNU Emacs Manual). Prompt.

n

A number, read with the minibuffer. If the input is not a number, the user has to try again. ‘n’ never uses the prefix argument. Prompt.

N

The numeric prefix argument; but if there is no prefix argument, read a number as with n. The value is always a number. See Prefix Command Arguments. Prompt.

p

The numeric prefix argument. (Note that this ‘p’ is lower case.) No I/O.

P

The raw prefix argument. (Note that this ‘P’ is upper case.) No I/O.

r

Point and the mark, as two numeric arguments, smallest first. This is the only code letter that specifies two successive arguments rather than one. This will signal an error if the mark is not set in the buffer which is current when the command is invoked. If Transient Mark mode is turned on (see The Mark) — as it is by default — and user option mark-even-if-inactive is nil, Emacs will signal an error even if the mark is set, but is inactive. No I/O.

s

Arbitrary text, read in the minibuffer and returned as a string (see Reading Text Strings with the Minibuffer). Terminate the input with either C-j or RET. (C-q may be used to include either of these characters in the input.) Prompt.

S

An interned symbol whose name is read in the minibuffer. Terminate the input with either C-j or RET. Other characters that normally terminate a symbol (e.g., whitespace, parentheses and brackets) do not do so here. Prompt.

U

A key sequence or nil. Can be used after a ‘k’ or ‘K’ argument to get the up-event that was discarded (if any) after ‘k’ or ‘K’ read a down-event. If no up-event has been discarded, ‘U’ provides nil as the argument. No I/O.

v

A variable declared to be a user option (i.e., satisfying the predicate custom-variable-p). This reads the variable using read-variable. See Definition of read-variable. Existing, Completion, Prompt.

x

A Lisp object, specified with its read syntax, terminated with a C-j or RET. The object is not evaluated. See Reading Lisp Objects with the Minibuffer. Prompt.

X

A Lisp form’s value. ‘X’ reads as ‘x’ does, then evaluates the form so that its value becomes the argument for the command. Prompt.

z

A coding system name (a symbol). If the user enters null input, the argument value is nil. See Coding Systems. Completion, Existing, Prompt.

Z

A coding system name (a symbol)—but only if this command has a prefix argument. With no prefix argument, ‘Z’ provides nil as the argument value. Completion, Existing, Prompt.


22.2.3 Examples of Using interactive

Here are some examples of interactive:

(defun foo1 ()              ; foo1 takes no arguments,
    (interactive)           ;   just moves forward two words.
    (forward-word 2))
     ⇒ foo1

(defun foo2 (n)             ; foo2 takes one argument,
    (interactive "^p")      ;   which is the numeric prefix.
                            ; under shift-select-mode,
                            ;   will activate or extend region.
    (forward-word (* 2 n)))
     ⇒ foo2

(defun foo3 (n)             ; foo3 takes one argument,
    (interactive "nCount:") ;   which is read with the Minibuffer.
    (forward-word (* 2 n)))
     ⇒ foo3

(defun three-b (b1 b2 b3)
  "Select three existing buffers.
Put them into three windows, selecting the last one."
    (interactive "bBuffer1:\nbBuffer2:\nbBuffer3:")
    (delete-other-windows)
    (split-window (selected-window) 8)
    (switch-to-buffer b1)
    (other-window 1)
    (split-window (selected-window) 8)
    (switch-to-buffer b2)
    (other-window 1)
    (switch-to-buffer b3))
     ⇒ three-b
(three-b "*scratch*" "declarations.texi" "*mail*")
     ⇒ nil

22.2.4 Specifying Modes For Commands

Many commands in Emacs are general, and not tied to any specific mode. For instance, M-x kill-region can be used in pretty much any mode that has editable text, and commands that display information (like M-x list-buffers) can be used in pretty much any context.

Many other commands, however, are specifically tied to a mode, and make no sense outside of that context. For instance, M-x dired-diff will just signal an error if used outside of a Dired buffer.

Emacs therefore has a mechanism for specifying what mode (or modes) a command “belongs” to:

(defun dired-diff (...)
  ...
  (interactive "p" dired-mode)
  ...)

This will mark the command as applicable to dired-mode only (or any modes that are derived from dired-mode). Any number of modes can be added to the interactive form.

Specifying modes affects command completion in M-S-x (execute-extended-command-for-buffer, see Interactive Call). It may also affect completion in M-x, depending on the value of read-extended-command-predicate.

For instance, when using the command-completion-default-include-p predicate as the value of read-extended-command-predicate, M-x won’t list commands that have been marked as being applicable to a specific mode (unless you are in a buffer that uses that mode, of course). This goes for both major and minor modes. (By contrast, M-S-x always omits inapplicable commands from the completion candidates.)

By default, read-extended-command-predicate is nil, and completion in M-x lists all the commands that match what the user has typed, whether those commands are or aren’t marked as applicable to the current buffer’s mode.

Marking commands to be applicable to a mode will also make C-h m list these commands (if they aren’t bound to any keys).

If using this extended interactive form isn’t convenient (because the code is supposed to work in older versions of Emacs that don’t support the extended interactive form), the following equivalent declaration (see The declare Form) can be used instead:

(declare (modes dired-mode))

Which commands to tag with modes is to some degree a matter of taste, but commands that clearly do not work outside of the mode should be tagged. This includes commands that will signal an error if called from somewhere else, but also commands that are destructive when called from an unexpected mode. (This usually includes most of the commands that are written for special (i.e., non-editing) modes.)

Some commands may be harmless, and “work” when called from other modes, but should still be tagged with a mode if they don’t actually make much sense to use elsewhere. For instance, many special modes have commands to exit the buffer bound to q, and may not do anything but issue a message like "Goodbye from this mode" and then call kill-buffer. This command will “work” from any mode, but it is highly unlikely that anybody would actually want to use the command outside the context of this special mode.

Many modes have a set of different commands that start the mode in different ways (e.g., eww-open-in-new-buffer and eww-open-file). Commands like that should never be tagged as mode-specific, as they can be issued by the user from pretty much any context.


22.2.5 Select among Command Alternatives

Sometimes it is useful to define a command that serves as a “generic dispatcher” capable of invoking one of a set of commands according to the user’s needs. For example, imagine that you want to define a command named ‘open’ that can “open” and display several different types of objects. Or you could have a command named ‘mua’ (which stands for Mail User Agent) that can read and send email using one of several email backends, such as Rmail, Gnus, or MH-E. The macro define-alternatives can be used to define such generic commands. A generic command is an interactive function whose implementation can be selected from several alternatives, as a matter of user preference.

Macro: define-alternatives command &rest customizations

This macro defines the new generic command, which can have several alternative implementations. The argument command should be an unquoted symbol.

When invoked, the macro creates an interactive Lisp closure (see Closures). When the user runs M-x command RET for the first time, Emacs asks to select one of the alternative implementations of command, offering completion for the names of these alternatives. These names come from the user option whose name is command-alternatives, which the macro creates (if it didn’t exist before). To be useful, this variable’s value should be an alist whose elements have the form (alt-name . alt-func), where alt-name is the name of the alternative and alt-func is the interactive function to be called if this alternative is selected. When the user selects an alternative, Emacs remembers the selection, and will thereafter automatically call that selected alternative without prompting when the user invokes M-x command again. To choose a different alternative, type C-u M-x command RET–then Emacs will again prompt for one of the alternatives, and the selection will override the previous one.

The variable command-alternatives can be created before calling define-alternatives, with the appropriate values; otherwise the macro creates the variable with a nil value, and it should then be populated with the associations describing the alternatives. Packages that wish to provide their own implementation of an existing generic command can use autoload cookies (see Autoload) to add to the alist, for example:

;;;###autoload (push '("My name" . my-foo-symbol) foo-alternatives

If the optional argument customizations is non-nil, it should consist of alternating defcustom keywords (typically :group and :version) and values to add to the definition of the defcustom command-alternatives.

Here is an example of a simple generic dispatcher command named open with 3 alternative implementations:

(define-alternatives open
  :group 'files
  :version "42.1")
(setq open-alternatives
      '(("file" . find-file)
	("directory" . dired)
	("hexl" . hexl-find-file)))

22.3 Interactive Call

After the command loop has translated a key sequence into a command, it invokes that command using the function command-execute. If the command is a function, command-execute calls call-interactively, which reads the arguments and calls the command. You can also call these functions yourself.

Note that the term “command”, in this context, refers to an interactively callable function (or function-like object), or a keyboard macro. It does not refer to the key sequence used to invoke a command (see Keymaps).

Function: commandp object &optional for-call-interactively

This function returns t if object is a command. Otherwise, it returns nil.

Commands include strings and vectors (which are treated as keyboard macros), lambda expressions that contain a top-level interactive form (see Using interactive), byte-code function objects made from such lambda expressions, autoload objects that are declared as interactive (non-nil fourth argument to autoload), and some primitive functions. Also, a symbol is considered a command if it has a non-nil interactive-form property, or if its function definition satisfies commandp.

If for-call-interactively is non-nil, then commandp returns t only for objects that call-interactively could call—thus, not for keyboard macros.

See documentation in Access to Documentation Strings, for a realistic example of using commandp.

Function: call-interactively command &optional record-flag keys

This function calls the interactively callable function command, providing arguments according to its interactive calling specifications. It returns whatever command returns.

If, for instance, you have a function with the following signature:

(defun foo (begin end)
  (interactive "r")
  ...)

then saying

(call-interactively 'foo)

will call foo with the region (point and mark) as the arguments.

An error is signaled if command is not a function or if it cannot be called interactively (i.e., is not a command). Note that keyboard macros (strings and vectors) are not accepted, even though they are considered commands, because they are not functions. If command is a symbol, then call-interactively uses its function definition.

If record-flag is non-nil, then this command and its arguments are unconditionally added to the list command-history. Otherwise, the command is added only if it uses the minibuffer to read an argument. See Command History.

The argument keys, if given, should be a vector which specifies the sequence of events to supply if the command inquires which events were used to invoke it. If keys is omitted or nil, the default is the return value of this-command-keys-vector. See Definition of this-command-keys-vector.

Function: funcall-interactively function &rest arguments

This function works like funcall (see Calling Functions), but it makes the call look like an interactive invocation: a call to called-interactively-p inside function will return t. If function is not a command, it is called without signaling an error.

Function: command-execute command &optional record-flag keys special

This function executes command. The argument command must satisfy the commandp predicate; i.e., it must be an interactively callable function or a keyboard macro.

A string or vector as command is executed with execute-kbd-macro. A function is passed to call-interactively (see above), along with the record-flag and keys arguments.

If command is a symbol, its function definition is used in its place. A symbol with an autoload definition counts as a command if it was declared to stand for an interactively callable function. Such a definition is handled by loading the specified library and then rechecking the definition of the symbol.

The argument special, if given, means to ignore the prefix argument and not clear it. This is used for executing special events (see Special Events).

Command: execute-extended-command prefix-argument

This function reads a command name from the minibuffer using completing-read (see Completion). Then it uses command-execute to call the specified command. Whatever that command returns becomes the value of execute-extended-command.

If the command asks for a prefix argument, it receives the value prefix-argument. If execute-extended-command is called interactively, the current raw prefix argument is used for prefix-argument, and thus passed on to whatever command is run.

execute-extended-command is the normal definition of M-x, so it uses the string ‘M-x  as a prompt. (It would be better to take the prompt from the events used to invoke execute-extended-command, but that is painful to implement.) A description of the value of the prefix argument, if any, also becomes part of the prompt.

(execute-extended-command 3)
---------- Buffer: Minibuffer ----------
3 M-x forward-word RET
---------- Buffer: Minibuffer ----------
     ⇒ t

This command heeds the read-extended-command-predicate variable, which can filter out commands that are not applicable to the current major mode (or enabled minor modes). By default, the value of this variable is nil, and no commands are filtered out. However, customizing it to invoke the function command-completion-default-include-p will perform mode-dependent filtering. read-extended-command-predicate can be any predicate function; it will be called with two parameters: the command’s symbol and the current buffer. If should return non-nil if the command is to be included when completing in that buffer.

Command: execute-extended-command-for-buffer prefix-argument

This is like execute-extended-command, but limits the commands offered for completion to those commands that are of particular relevance to the current major mode (and enabled minor modes). This includes commands that are tagged with the modes (see Using interactive), and also commands that are bound to locally active keymaps. This command is the normal definition of M-S-x (that’s “meta shift x”).

Both these commands prompt for a command name, but with different completion rules. You can toggle between these two modes by using the M-S-x command while being prompted.


22.4 Distinguish Interactive Calls

Sometimes a command should display additional visual feedback (such as an informative message in the echo area) for interactive calls only. There are three ways to do this. The recommended way to test whether the function was called using call-interactively is to give it an optional argument print-message and use the interactive spec to make it non-nil in interactive calls. Here’s an example:

(defun foo (&optional print-message)
  (interactive "p")
  (when print-message
    (message "foo")))

We use "p" because the numeric prefix argument is never nil. Defined in this way, the function does display the message when called from a keyboard macro.

The above method with the additional argument is usually best, because it allows callers to say “treat this call as interactive”. But you can also do the job by testing called-interactively-p.

Function: called-interactively-p kind

This function returns t when the calling function was called using call-interactively.

The argument kind should be either the symbol interactive or the symbol any. If it is interactive, then called-interactively-p returns t only if the call was made directly by the user—e.g., if the user typed a key sequence bound to the calling function, but not if the user ran a keyboard macro that called the function (see Keyboard Macros). If kind is any, called-interactively-p returns t for any kind of interactive call, including keyboard macros.

If in doubt, use any; the only known proper use of interactive is if you need to decide whether to display a helpful message while a function is running.

A function is never considered to be called interactively if it was called via Lisp evaluation (or with apply or funcall).

Here is an example of using called-interactively-p:

(defun foo ()
  (interactive)
  (when (called-interactively-p 'any)
    (message "Interactive!")
    'foo-called-interactively))

;; Type M-x foo.
     -| Interactive!

(foo)
     ⇒ nil

Here is another example that contrasts direct and indirect calls to called-interactively-p.

(defun bar ()
  (interactive)
  (message "%s" (list (foo) (called-interactively-p 'any))))

;; Type M-x bar.
     -| (nil t)

22.5 Information from the Command Loop

The editor command loop sets several Lisp variables to keep status records for itself and for commands that are run. With the exception of this-command and last-command it’s generally a bad idea to change any of these variables in a Lisp program.

Variable: last-command

This variable records the name of the previous command executed by the command loop (the one before the current command). Normally the value is a symbol with a function definition, but this is not guaranteed.

The value is copied from this-command when a command returns to the command loop, except when the command has specified a prefix argument for the following command.

This variable is always local to the current terminal and cannot be buffer-local. See Multiple Terminals.

Variable: real-last-command

This variable is set up by Emacs just like last-command, but never altered by Lisp programs.

Variable: last-repeatable-command

This variable stores the most recently executed command that was not part of an input event. This is the command repeat will try to repeat, See Repeating in The GNU Emacs Manual.

Variable: this-command

This variable records the name of the command now being executed by the editor command loop. Like last-command, it is normally a symbol with a function definition.

The command loop sets this variable just before running a command, and copies its value into last-command when the command finishes (unless the command specified a prefix argument for the following command).

Some commands set this variable during their execution, as a flag for whatever command runs next. In particular, the functions for killing text set this-command to kill-region so that any kill commands immediately following will know to append the killed text to the previous kill.

If you do not want a particular command to be recognized as the previous command in the case where it got an error, you must code that command to prevent this. One way is to set this-command to t at the beginning of the command, and set this-command back to its proper value at the end, like this:

(defun foo (args…)
  (interactive …)
  (let ((old-this-command this-command))
    (setq this-command t)
    …do the work…
    (setq this-command old-this-command)))

We do not bind this-command with let because that would restore the old value in case of error—a feature of let which in this case does precisely what we want to avoid.

Variable: this-original-command

This has the same value as this-command except when command remapping occurs (see Remapping Commands). In that case, this-command gives the command actually run (the result of remapping), and this-original-command gives the command that was specified to run but remapped into another command.

Variable: current-minibuffer-command

This has the same value as this-command, but is bound recursively when entering a minibuffer. This variable can be used from minibuffer hooks and the like to determine what command opened the current minibuffer session.

Function: this-command-keys

This function returns a string or vector containing the key sequence that invoked the present command. Any events read by the command using read-event without a timeout get tacked on to the end.

However, if the command has called read-key-sequence, it returns the last read key sequence. See Key Sequence Input. The value is a string if all events in the sequence were characters that fit in a string. See Input Events.

(this-command-keys)
;; Now use C-u C-x C-e to evaluate that.
     ⇒ "^X^E"
Function: this-command-keys-vector

Like this-command-keys, except that it always returns the events in a vector, so you don’t need to deal with the complexities of storing input events in a string (see Putting Keyboard Events in Strings).

Function: clear-this-command-keys &optional keep-record

This function empties out the table of events for this-command-keys to return. Unless keep-record is non-nil, it also empties the records that the function recent-keys (see Recording Input) will subsequently return. This is useful after reading a password, to prevent the password from echoing inadvertently as part of the next command in certain cases.

Variable: last-nonmenu-event

This variable holds the last input event read as part of a key sequence, not counting events resulting from mouse menus.

One use of this variable is for telling x-popup-menu where to pop up a menu. It is also used internally by y-or-n-p (see Yes-or-No Queries).

Variable: last-command-event

This variable is set to the last input event that was read by the command loop as part of a command. The principal use of this variable is in self-insert-command, which uses it to decide which character to insert, and in post-self-insert-hook (see User-Level Insertion Commands), which uses it to access the character that was just inserted.

last-command-event
;; Now use C-u C-x C-e to evaluate that.
     ⇒ 5

The value is 5 because that is the ASCII code for C-e.

Variable: last-event-frame

This variable records which frame the last input event was directed to. Usually this is the frame that was selected when the event was generated, but if that frame has redirected input focus to another frame, the value is the frame to which the event was redirected. See Input Focus.

If the last event came from a keyboard macro, the value is macro.

Input events must come from somewhere; sometimes, that is a keyboard macro, a signal, or ‘unread-command-events’, but it is usually a physical input device connected to a computer that is controlled by the user. Those devices are referred to as input devices, and Emacs associates each input event with the input device from which it originated. They are identified by a name that is unique to each input device.

The ability to determine the precise input device used depends on the details of each system. When that information is unavailable, Emacs reports keyboard events as originating from the ‘"Virtual core keyboard"’, and other events as originating from the ‘"Virtual core pointer"’. (These values are used on every platform because the X server reports them when detailed device information is not known.)

Variable: last-event-device

This variable records the name of the input device from which the last input event read was generated. It is nil if no such device exists, i.e., the last input event was read from unread-command-events, or it came from a keyboard macro.

When the X Input Extension is being used on X Windows, the device name is a string that is unique to each physical keyboard, pointing device and touchscreen attached to the X server. Otherwise, it is either the string ‘"Virtual core pointer"’ or ‘"Virtual core keyboard"’, depending on whether the event was generated by a pointing device (such as a mouse) or a keyboard.

Function: device-class frame name

There are various different types of devices, which can be determined from their names. This function can be used to determined the correct type of the device name for an event originating from frame.

The return value is one of the following symbols (“device classes”):

core-keyboard

The core keyboard; this is means the device is a keyboard-like device, but no other characteristics are unknown.

core-pointer

The core pointer; this means the device is a pointing device, but no other characteristics are known.

mouse

A computer mouse.

trackpoint

A trackpoint or joystick (or other similar control.)

eraser

The other end of a stylus on a graphics tablet, or a standalone eraser.

pen

The pointed end of a pen on a graphics tablet, a stylus, or some other similar device.

puck

A device that looks like a computer mouse, but reports absolute coordinates relative to some other surface.

power-button

A power button or volume button (or other similar control.)

keyboard

A computer keyboard.

touchscreen

A computer touchpad.

pad

A collection of sensitive buttons, rings, and strips commonly found around a drawing tablet.

touchpad

An indirect touch device such as a touchpad.

piano

A musical instrument such as an electronic keyboard.

test

A device used by the XTEST extension to report input.


22.6 Adjusting Point After Commands

Emacs cannot display the cursor when point is in the middle of a sequence of text that has the display or composition property, or is invisible. Therefore, after a command finishes and returns to the command loop, if point is within such a sequence, the command loop normally moves point to the edge of the sequence, making this sequence effectively intangible.

A command can inhibit this feature by setting the variable disable-point-adjustment:

Variable: disable-point-adjustment

If this variable is non-nil when a command returns to the command loop, then the command loop does not check for those text properties, and does not move point out of sequences that have them.

The command loop sets this variable to nil before each command, so if a command sets it, the effect applies only to that command.

Variable: global-disable-point-adjustment

If you set this variable to a non-nil value, the feature of moving point out of these sequences is completely turned off.


22.7 Input Events

The Emacs command loop reads a sequence of input events that represent keyboard or mouse activity, or system events sent to Emacs. The events for keyboard activity are characters or symbols; other events are always lists. This section describes the representation and meaning of input events in detail.

Function: eventp object

This function returns non-nil if object is an input event or event type.

Note that any non-nil symbol might be used as an event or an event type; eventp cannot distinguish whether a symbol is intended by Lisp code to be used as an event.


22.7.1 Keyboard Events

There are two kinds of input you can get from the keyboard: ordinary keys, and function keys. Ordinary keys correspond to (possibly modified) characters; the events they generate are represented in Lisp as characters. The event type of a character event is the character itself (an integer), which might have some modifier bits set; see Classifying Events.

An input character event consists of a basic code between 0 and 524287, plus any or all of these modifier bits:

meta

The 2**27 bit in the character code indicates a character typed with the meta key held down.

control

The 2**26 bit in the character code indicates a non-ASCII control character.

ASCII control characters such as C-a have special basic codes of their own, so Emacs needs no special bit to indicate them. Thus, the code for C-a is just 1.

But if you type a control combination not in ASCII, such as % with the control key, the numeric value you get is the code for % plus 2**26 (assuming the terminal supports non-ASCII control characters), i.e. with the 27th bit set.

shift

The 2**25 bit (the 26th bit) in the character event code indicates an ASCII control character typed with the shift key held down.

For letters, the basic code itself indicates upper versus lower case; for digits and punctuation, the shift key selects an entirely different character with a different basic code. In order to keep within the ASCII character set whenever possible, Emacs avoids using the 2**25 bit for those character events.

However, ASCII provides no way to distinguish C-A from C-a, so Emacs uses the 2**25 bit in C-A and not in C-a.

hyper

The 2**24 bit in the character event code indicates a character typed with the hyper key held down.

super

The 2**23 bit in the character event code indicates a character typed with the super key held down.

alt

The 2**22 bit in the character event code indicates a character typed with the alt key held down. (The key labeled Alt on most keyboards is actually treated as the meta key, not this.)

It is best to avoid mentioning specific bit numbers in your program. To test the modifier bits of a character, use the function event-modifiers (see Classifying Events). When making key bindings with keymap-set, you specify these events using strings like ‘C-H-x’ instead (for “control hyper x”) (see Changing Key Bindings).


22.7.2 Function Keys

Most keyboards also have function keys—keys that have names or symbols that are not characters. Function keys are represented in Emacs Lisp as symbols; the symbol’s name is the function key’s label, in lower case. For example, pressing a key labeled F1 generates an input event represented by the symbol f1.

The event type of a function key event is the event symbol itself. See Classifying Events.

Here are a few special cases in the symbol-naming convention for function keys:

backspace, tab, newline, return, delete

These keys correspond to common ASCII control characters that have special keys on most keyboards.

In ASCII, C-i and TAB are the same character. If the terminal can distinguish between them, Emacs conveys the distinction to Lisp programs by representing the former as the integer 9, and the latter as the symbol tab.

Most of the time, it’s not useful to distinguish the two. So normally local-function-key-map (see Keymaps for Translating Sequences of Events) is set up to map tab into 9. Thus, a key binding for character code 9 (the character C-i) also applies to tab. Likewise for the other symbols in this group. The function read-char likewise converts these events into characters.

In ASCII, BS is really C-h. But backspace converts into the character code 127 (DEL), not into code 8 (BS). This is what most users prefer.

left, up, right, down

Cursor arrow keys

kp-add, kp-decimal, kp-divide, …

Keypad keys (to the right of the regular keyboard).

kp-0, kp-1, …

Keypad keys with digits.

kp-f1, kp-f2, kp-f3, kp-f4

Keypad PF keys.

kp-home, kp-left, kp-up, kp-right, kp-down

Keypad arrow keys. Emacs normally translates these into the corresponding non-keypad keys home, left, …

kp-prior, kp-next, kp-end, kp-begin, kp-insert, kp-delete

Additional keypad duplicates of keys ordinarily found elsewhere. Emacs normally translates these into the like-named non-keypad keys.

You can use the modifier keys ALT, CTRL, HYPER, META, SHIFT, and SUPER with function keys. The way to represent them is with prefixes in the symbol name:

A-

The alt modifier.

C-

The control modifier.

H-

The hyper modifier.

M-

The meta modifier.

S-

The shift modifier.

s-

The super modifier.

Thus, the symbol for the key F3 with META held down is M-f3. When you use more than one prefix, we recommend you write them in alphabetical order; but the order does not matter in arguments to the key-binding lookup and modification functions.


22.7.3 Mouse Events

Emacs supports four kinds of mouse events: click events, drag events, button-down events, and motion events. All mouse events are represented as lists. The CAR of the list is the event type; this says which mouse button was involved, and which modifier keys were used with it. The event type can also distinguish double or triple button presses (see Repeat Events). The rest of the list elements give position and time information.

For key lookup, only the event type matters: two events of the same type necessarily run the same command. The command can access the full values of these events using the ‘e’ interactive code. See Code Characters for interactive.

A key sequence that starts with a mouse event is read using the keymaps of the buffer in the window that the mouse was in, not the current buffer. This does not imply that clicking in a window selects that window or its buffer—that is entirely under the control of the command binding of the key sequence.


22.7.4 Click Events

When the user presses a mouse button and releases it at the same location, that generates a click event. Depending on how your window-system reports mouse-wheel events, turning the mouse wheel can generate either a mouse click or a mouse-wheel event. All mouse event share the same format:

(event-type position click-count)
event-type

This is a symbol that indicates which mouse button was used. It is one of the symbols mouse-1, mouse-2, …, where the buttons are numbered left to right. For mouse-wheel event, it can be wheel-up or wheel-down.

You can also use prefixes ‘A-’, ‘C-’, ‘H-’, ‘M-’, ‘S-’ and ‘s-’ for modifiers alt, control, hyper, meta, shift and super, just as you would with function keys.

This symbol also serves as the event type of the event. Key bindings describe events by their types; thus, if there is a key binding for mouse-1, that binding would apply to all events whose event-type is mouse-1.

position

This is a mouse position list specifying where the mouse event occurred; see below for details.

click-count

This is the number of rapid repeated presses so far of the same mouse button or the number of repeated turns of the wheel. See Repeat Events.

To access the contents of a mouse position list in the position slot of a mouse event, you should typically use the functions documented in Accessing Mouse Events.

The explicit format of the list depends on where the event occurred. For clicks in the text area, mode line, header line, tab line, or in the fringe or marginal areas, the mouse position list has the form

(window pos-or-area (x . y) timestamp
 object text-pos (col . row)
 image (dx . dy) (width . height))

The meanings of these list elements are as follows:

window

The window in which the mouse event occurred.

pos-or-area

The buffer position of the character clicked on in the text area; or, if the event was outside the text area, the window area where it occurred. It is one of the symbols mode-line, header-line, tab-line, vertical-line, left-margin, right-margin, left-fringe, or right-fringe.

In one special case, pos-or-area is a list containing a symbol (one of the symbols listed above) instead of just the symbol. This happens after the imaginary prefix keys for the event are registered by Emacs. See Key Sequence Input.

x, y

The relative pixel coordinates of the event. For events in the text area of a window, the coordinate origin (0 . 0) is taken to be the top left corner of the text area. See Window Sizes. For events in a mode line, header line or tab line, the coordinate origin is the top left corner of the window itself. For fringes, margins, and the vertical border, x does not have meaningful data. For fringes and margins, y is relative to the bottom edge of the header line. In all cases, the x and y coordinates increase rightward and downward respectively.

timestamp

The time at which the event occurred, as an integer number of milliseconds since a system-dependent initial time.

object

Either nil, which means the event occurred on buffer text, or a cons cell of the form (string . string-pos) if there is a string from a text property or an overlay at the event position.

string

The string which was clicked on, including any properties.

string-pos

The position in the string where the click occurred.

text-pos

For clicks on a marginal area or on a fringe, this is the buffer position of the first visible character in the corresponding line in the window. For clicks on the mode line, the header line or the tab line, this is nil. For other events, it is the buffer position closest to the click.

col, row

These are the actual column and row coordinate numbers of the glyph under the x, y position. If x lies beyond the last column of actual text on its line, col is reported by adding fictional extra columns that have the default character width. Row 0 is taken to be the header line if the window has one, or Row 1 if the window also has the tab line, or the topmost row of the text area otherwise. Column 0 is taken to be the leftmost column of the text area for clicks on a window text area, or the leftmost mode line or header line column for clicks there. For clicks on fringes or vertical borders, these have no meaningful data. For clicks on margins, col is measured from the left edge of the margin area and row is measured from the top of the margin area.

image

If there is an image at the click location, this is the image object as returned by find-image (see Defining Images); otherwise this is nil.

dx, dy

These are the pixel offsets of the click relative to the top left corner of the object’s glyph that is the nearest one to the click. The relevant objects can be either a buffer, or a string, or an image, see above. If object is nil or a string, the coordinates are relative to the top left corner of the character glyph clicked on. Note that the offsets are always zero on text-mode frames, when object is nil, since each glyph there is considered to have exactly 1x1 pixel dimensions.

width, height

If the click is on a character, either from buffer text or from overlay or display string, these are the pixel width and height of that character’s glyph; otherwise they are dimensions of object clicked on.

For clicks on a scroll bar, position has this form:

(window area (portion . whole) timestamp part)
window

The window whose scroll bar was clicked on.

area

This is the symbol vertical-scroll-bar.

portion

The number of pixels from the top of the scroll bar to the click position. On some toolkits, including GTK+, Emacs cannot extract this data, so the value is always 0.

whole

The total length, in pixels, of the scroll bar. On some toolkits, including GTK+, Emacs cannot extract this data, so the value is always 0.

timestamp

The time at which the event occurred, in milliseconds. On some toolkits, including GTK+, Emacs cannot extract this data, so the value is always 0.

part

The part of the scroll bar on which the click occurred. It is one of the symbols handle (the scroll bar handle), above-handle (the area above the handle), below-handle (the area below the handle), up (the up arrow at one end of the scroll bar), or down (the down arrow at one end of the scroll bar).

For clicks on the frame’s internal border (see Frame Layout), the frame’s tool bar (see Tool bars) or tab bar, position has this form:

 (frame part (X . Y) timestamp)
frame

The frame whose internal border or tool bar or tab bar was clicked on.

part

The part of the frame which was clicked on. This can be one of the following:

tool-bar

The frame has a tool bar, and the event was in the tool-bar area.

tab-bar

The frame has a tab bar, and the event was in the tab-bar area.

left-edge
top-edge
right-edge
bottom-edge

The click was on the corresponding border at an offset of at least one canonical character from the border’s nearest corner.

top-left-corner
top-right-corner
bottom-right-corner
bottom-left-corner

The click was on the corresponding corner of the internal border.

nil

The frame does not have an internal border, and the event was not on the tab bar or the tool bar. This usually happens on text-mode frames. This can also happen on GUI frames with internal border if the frame doesn’t have its drag-internal-border parameter (see Mouse Dragging Parameters) set to a non-nil value.


22.7.5 Drag Events

With Emacs, you can have a drag event without even changing your clothes. A drag event happens every time the user presses a mouse button and then moves the mouse to a different character position before releasing the button. Like all mouse events, drag events are represented in Lisp as lists. The lists record both the starting mouse position and the final position, like this:

(event-type
 (window1 START-POSITION)
 (window2 END-POSITION))

For a drag event, the name of the symbol event-type contains the prefix ‘drag-’. For example, dragging the mouse with button 2 held down generates a drag-mouse-2 event. The second and third elements of the event give the starting and ending position of the drag, as mouse position lists (see Click Events). You can access the second element of any mouse event in the same way. However, the drag event may end outside the boundaries of the frame that was initially selected. In that case, the third element’s position list contains that frame in place of a window.

The ‘drag-’ prefix follows the modifier key prefixes such as ‘C-’ and ‘M-’.

If read-key-sequence receives a drag event that has no key binding, and the corresponding click event does have a binding, it changes the drag event into a click event at the drag’s starting position. This means that you don’t have to distinguish between click and drag events unless you want to.


22.7.6 Button-Down Events

Click and drag events happen when the user releases a mouse button. They cannot happen earlier, because there is no way to distinguish a click from a drag until the button is released.

If you want to take action as soon as a button is pressed, you need to handle button-down events.17 These occur as soon as a button is pressed. They are represented by lists that look exactly like click events (see Click Events), except that the event-type symbol name contains the prefix ‘down-’. The ‘down-’ prefix follows modifier key prefixes such as ‘C-’ and ‘M-’.

The function read-key-sequence ignores any button-down events that don’t have command bindings; therefore, the Emacs command loop ignores them too. This means that you need not worry about defining button-down events unless you want them to do something. The usual reason to define a button-down event is so that you can track mouse motion (by reading motion events) until the button is released. See Motion Events.


22.7.7 Repeat Events

If you press the same mouse button more than once in quick succession without moving the mouse, Emacs generates special repeat mouse events for the second and subsequent presses.

The most common repeat events are double-click events. Emacs generates a double-click event when you click a button twice; the event happens when you release the button (as is normal for all click events).

The event type of a double-click event contains the prefix ‘double-’. Thus, a double click on the second mouse button with meta held down comes to the Lisp program as M-double-mouse-2. If a double-click event has no binding, the binding of the corresponding ordinary click event is used to execute it. Thus, you need not pay attention to the double click feature unless you really want to.

When the user performs a double click, Emacs generates first an ordinary click event, and then a double-click event. Therefore, you must design the command binding of the double click event to assume that the single-click command has already run. It must produce the desired results of a double click, starting from the results of a single click.

This is convenient, if the meaning of a double click somehow builds on the meaning of a single click—which is recommended user interface design practice for double clicks.

If you click a button, then press it down again and start moving the mouse with the button held down, then you get a double-drag event when you ultimately release the button. Its event type contains ‘double-drag’ instead of just ‘drag’. If a double-drag event has no binding, Emacs looks for an alternate binding as if the event were an ordinary drag.

Before the double-click or double-drag event, Emacs generates a double-down event when the user presses the button down for the second time. Its event type contains ‘double-down’ instead of just ‘down’. If a double-down event has no binding, Emacs looks for an alternate binding as if the event were an ordinary button-down event. If it finds no binding that way either, the double-down event is ignored.

To summarize, when you click a button and then press it again right away, Emacs generates a down event and a click event for the first click, a double-down event when you press the button again, and finally either a double-click or a double-drag event.

If you click a button twice and then press it again, all in quick succession, Emacs generates a triple-down event, followed by either a triple-click or a triple-drag. The event types of these events contain ‘triple’ instead of ‘double’. If any triple event has no binding, Emacs uses the binding that it would use for the corresponding double event.

If you click a button three or more times and then press it again, the events for the presses beyond the third are all triple events. Emacs does not have separate event types for quadruple, quintuple, etc. events. However, you can look at the event list to find out precisely how many times the button was pressed.

Function: event-click-count event

This function returns the number of consecutive button presses that led up to event. If event is a double-down, double-click or double-drag event, the value is 2. If event is a triple event, the value is 3 or greater. If event is an ordinary mouse event (not a repeat event), the value is 1.

User Option: double-click-fuzz

To generate repeat events, successive mouse button presses must be at approximately the same screen position. The value of double-click-fuzz specifies the maximum number of pixels the mouse may be moved (horizontally or vertically) between two successive clicks to make a double-click.

This variable is also the threshold for motion of the mouse to count as a drag.

User Option: double-click-time

To generate repeat events, the number of milliseconds between successive button presses must be less than the value of double-click-time. Setting double-click-time to nil disables multi-click detection entirely. Setting it to t removes the time limit; Emacs then detects multi-clicks by position only.


22.7.8 Motion Events

Emacs sometimes generates mouse motion events to describe motion of the mouse without any button activity. Mouse motion events are represented by lists that look like this:

(mouse-movement POSITION)

position is a mouse position list (see Click Events), specifying the current position of the mouse cursor. As with the end-position of a drag event, this position list may represent a location outside the boundaries of the initially selected frame, in which case the list contains that frame in place of a window.

The track-mouse macro enables generation of motion events within its body. Outside of track-mouse body, Emacs does not generate events for mere motion of the mouse, and these events do not appear. See Mouse Tracking.

Variable: mouse-fine-grained-tracking

When non-nil, mouse motion events are generated even for very small movements. Otherwise, motion events are not generated as long as the mouse cursor remains pointing to the same glyph in the text.


22.7.9 Touchscreen Events

Some window systems provide support for input devices that react to the user’s touching the screen and moving fingers while touching the screen. These input devices are known as touchscreens, and Emacs reports the events they generate as touchscreen events.

Most individual events generated by a touchscreen only have meaning as part of a larger sequence of other events: for instance, the simple operation of tapping the touchscreen involves the user placing and raising a finger on the touchscreen, and swiping the display to scroll it involves placing a finger, moving it many times upwards or downwards, and then raising the finger.

While a simplistic model consisting of one finger is adequate for taps and scrolling, more complicated gestures require support for keeping track of multiple fingers, where the position of each finger is represented by a touch point. For example, a “pinch to zoom” gesture might consist of the user placing two fingers and moving them individually in opposite directions, where the distance between the positions of their individual points determine the amount by which to zoom the display, and the center of an imaginary line between those positions determines where to pan the display after zooming.

The low-level touchscreen events described below can be used to implement all the touch sequences described above. In those events, each point is represented by a cons of an arbitrary number identifying the point and a mouse position list (see Click Events) specifying the position of the finger when the event occurred.

(touchscreen-begin point)

This event is sent when point is created by the user pressing a finger against the touchscreen.

(touchscreen-update points)

This event is sent when a point on the touchscreen has changed position. points is a list of touch points containing the up-to-date positions of each touch point currently on the touchscreen.

(touchscreen-end point)

This event is sent when point is no longer present on the display, because another program took the grab, or because the user raised the finger from the touchscreen.


22.7.10 Focus Events

Window systems provide general ways for the user to control which window gets keyboard input. This choice of window is called the focus. When the user does something to switch between Emacs frames, that generates a focus event. The normal definition of a focus event, in the global keymap, is to select a new frame within Emacs, as the user would expect. See Input Focus, which also describes hooks related to focus events.

Focus events are represented in Lisp as lists that look like this:

(switch-frame new-frame)

where new-frame is the frame switched to.

Some X window managers are set up so that just moving the mouse into a window is enough to set the focus there. Usually, there is no need for a Lisp program to know about the focus change until some other kind of input arrives. Emacs generates a focus event only when the user actually types a keyboard key or presses a mouse button in the new frame; just moving the mouse between frames does not generate a focus event.

A focus event in the middle of a key sequence would garble the sequence. So Emacs never generates a focus event in the middle of a key sequence. If the user changes focus in the middle of a key sequence—that is, after a prefix key—then Emacs reorders the events so that the focus event comes either before or after the multi-event key sequence, and not within it.


22.7.11 Xwidget events

Xwidgets (see Embedded Native Widgets) can send events to update Lisp programs on their status. These events are dubbed xwidget-events, and contain various data describing the nature of the change.

(xwidget-event kind xwidget arg)

This event is sent whenever some kind of update occurs in xwidget. There are several types of updates, identified by their kind.

It is a special event (see Special Events), which should be handled by adding a callback to an xwidget that is called whenever an xwidget event for xwidget is received.

You can add a callback by setting the callback of an xwidget’s property list, which should be a function that accepts xwidget and kind as arguments.

load-changed

This xwidget event indicates that the xwidget has reached a particular point of the page-loading process. When these events are sent, arg will contain a string that further describes the status of the widget:

load-started

This means the widget has begun a page-loading operation.

load-finished

This means the xwidget has finished processing whatever page-loading operation that it was previously performing.

load-redirected

This means the xwidget has encountered and followed a redirect during the page-loading operation.

load-committed

This means the xwidget has committed to a given URL during the page-loading operation, i.e. the URL is the final URL that will be rendered by xwidget during the current page-loading operation.

download-callback

This event indicates that a download of some kind has been completed.

In the above events, there can be arguments after arg, which itself indicates the URL from which the download file was retrieved: the first argument after arg indicates the MIME type of the download, as a string, while the second argument contains the full file name of the downloaded file.

download-started

This event indicates that a download has been started. In these events, arg contains the URL of the file that is currently being downloaded.

javascript-callback

This event contains JavaScript callback data. These events are used internally by xwidget-webkit-execute-script.

(xwidget-display-event xwidget source)

This event is sent whenever an xwidget requests that another xwidget be displayed. xwidget is the xwidget that should be displayed, and source is the xwidget that asked to display xwidget.

It is also a special event which should be handled through callbacks. You can add such a callback by setting the display-callback of source’s property list, which should be a function that accepts xwidget and source as arguments.

xwidget’s buffer will be set to a temporary buffer. When displaying the widget, care should be taken to replace the buffer with the buffer in which the xwidget will be displayed, using set-xwidget-buffer (see Embedded Native Widgets).


22.7.12 Miscellaneous System Events

A few other event types represent occurrences within the system.

(delete-frame (frame))

This kind of event indicates that the user gave the window manager a command to delete a particular window, which happens to be an Emacs frame.

The standard definition of the delete-frame event is to delete frame.

(iconify-frame (frame))

This kind of event indicates that the user iconified frame using the window manager. Its standard definition is ignore; since the frame has already been iconified, Emacs has no work to do. The purpose of this event type is so that you can keep track of such events if you want to.

(make-frame-visible (frame))

This kind of event indicates that the user deiconified frame using the window manager. Its standard definition is ignore; since the frame has already been made visible, Emacs has no work to do.

(touch-end (position))

This kind of event indicates that the user’s finger moved off the mouse wheel or the touchpad. The position element is a mouse position list (see Click Events), specifying the position of the mouse cursor when the finger moved off the mouse wheel.

(wheel-up position clicks lines pixel-delta)
(wheel-down position clicks lines pixel-delta)

These events are generated by moving a mouse wheel. The position element is a mouse position list (see Click Events), specifying the position of the mouse cursor when the event occurred.

clicks, if present, is the number of times that the wheel was moved in quick succession. See Repeat Events. lines, if present and not nil, is the positive number of screen lines that should be scrolled (either up, when the event is wheel-up, or down when the event is wheel-down). pixel-delta, if present, is a cons cell of the form (x . y), where x and y are the numbers of pixels by which to scroll in each axis, a.k.a. pixelwise deltas. Usually, only one of the two will be non-zero, the other will be either zero or very close to zero; the larger number indicates the axis to scroll the window. When the variable mwheel-coalesce-scroll-events is nil, the scroll commands ignore the lines element, even if it’s non-nil, and use the pixel-delta data instead; in that case, the direction of scrolling is determined by the sign of the pixelwise deltas, and the direction (up or down) implied by the event kind is ignored.

You can use these x and y pixelwise deltas to determine how much the mouse wheel has actually moved at pixel resolution. For example, the pixelwise deltas could be used to scroll the display at pixel resolution, exactly according to the user’s turning the mouse wheel. This pixelwise scrolling is possible only when mwheel-coalesce-scroll-events is nil, and in general the pixel-delta data is not generated when that variable is non-nil.

The wheel-up and wheel-down events are generated only on some kinds of systems. On other systems, mouse-4 and mouse-5 are used instead. For portable code, use the variables mouse-wheel-up-event, mouse-wheel-up-alternate-event, mouse-wheel-down-event and mouse-wheel-down-alternate-event defined in mwheel.el to determine what event types to expect from the mouse wheel.

Similarly, some mice can generate mouse-wheel-left-event and mouse-wheel-right-event and can be used to scroll if mouse-wheel-tilt-scroll is non-nil. However, some mice also generate other events at the same time as they’re generating these scroll events which may get in the way. The way to fix this is generally to unbind these events (for instance, mouse-6 or mouse-7, but this is very hardware and operating system dependent).

(pinch position dx dy scale angle)

This kind of event is generated by the user performing a “pinch” gesture by placing two fingers on a touchpad and moving them towards or away from each other. position is a mouse position list (see Click Events) that provides the position of the mouse pointer when the event occurred, dx is the change in the horizontal distance between the fingers since the last event in the same sequence, dy is the vertical movement of the fingers since the last event in the same sequence, scale is the ratio of the current distance between the fingers to that distance at the start of the sequence, and angle is the angular difference in degrees between the direction of the line connecting the fingers in this event and the direction of that line in the last event of the same sequence.

As pinch events are only sent at the beginning or during a pinch sequence, they do not report gestures where the user moves two fingers on a touchpad in a rotating fashion without pinching the fingers.

All arguments after position are floating point numbers.

This event is usually sent as part of a sequence, which begins with the user placing two fingers on the touchpad, and ends with the user removing those fingers. dx, dy, and angle will be 0.0 in the first event of a sequence; subsequent events will report non-zero values for these members of the event structure.

dx and dy are reported in imaginary relative units, in which 1.0 is the width and height of the touchpad respectively. They are usually interpreted as being relative to the size of the object beneath the gesture: image, window, etc.

(preedit-text arg)

This event is sent when a system input method tells Emacs to display some text to indicate to the user what will be inserted. The contents of arg are dependent on the window system being used.

On X, arg is a string describing some text to place behind the cursor. It can be nil, which means to remove any text previously displayed.

On PGTK frames (see Frames), arg is a list of strings with information about their color and underline attributes. It has the following form:

   ((string1
     (ul . underline-color)
     (bg . background-color)
     (fg . foreground-color))
    (string2
     (ul . underline-color)
     (bg . background-color)
     (fg . foreground-color))
    …
   )

Color information can be omitted, leaving just the text of the strings. underline-color can be t, meaning underlined text with default underline color, or it can be a string, the name of the color to draw the underline.

This is a special event (see Special Events), which normally should not be bound by the user to any command. Emacs will typically display the text contained in the event in an overlay behind point when it is received.

(drag-n-drop position files)

This kind of event is generated when a group of files is selected in an application outside of Emacs, and then dragged and dropped onto an Emacs frame.

The element position is a list describing the position of the event, in the same format as used in a mouse-click event (see Click Events), and files is the list of file names that were dragged and dropped. The usual way to handle this event is by visiting these files.

This kind of event is generated, at present, only on some kinds of systems.

help-echo

This kind of event is generated when a mouse pointer moves onto a portion of buffer text which has a help-echo text property. The generated event has this form:

(help-echo frame help window object pos)

The precise meaning of the event parameters and the way these parameters are used to display the help-echo text are described in Text help-echo.

sigusr1
sigusr2

These events are generated when the Emacs process receives the signals SIGUSR1 and SIGUSR2. They contain no additional data because signals do not carry additional information. They can be useful for debugging (see Entering the Debugger on an Error).

To catch a user signal, bind the corresponding event to an interactive command in the special-event-map (see Controlling the Active Keymaps). The command is called with no arguments, and the specific signal event is available in last-input-event (see Miscellaneous Event Input Features. For example:

(defun sigusr-handler ()
  (interactive)
  (message "Caught signal %S" last-input-event))

(keymap-set special-event-map "<sigusr1>" 'sigusr-handler)

To test the signal handler, you can make Emacs send a signal to itself:

(signal-process (emacs-pid) 'sigusr1)
language-change

This kind of event is generated on MS-Windows when the input language has changed. This typically means that the keyboard keys will send to Emacs characters from a different language. The generated event has this form:

(language-change frame codepage language-id)

Here frame is the frame which was current when the input language changed; codepage is the new codepage number; and language-id is the numerical ID of the new input language. The coding-system (see Coding Systems) that corresponds to codepage is cpcodepage or windows-codepage. To convert language-id to a string (e.g., to use it for various language-dependent features, such as set-language-environment), use the w32-get-locale-info function, like this:

;; Get the abbreviated language name, such as "ENU" for English
(w32-get-locale-info language-id)
;; Get the full English name of the language,
;; such as "English (United States)"
(w32-get-locale-info language-id 4097)
;; Get the full localized name of the language
(w32-get-locale-info language-id t)
end-session

This event is generated on MS-Windows when the operating system informs Emacs that the user terminated the interactive session, or that the system is shutting down. The standard definition of this event is to invoke the kill-emacs command (see Killing Emacs) so as to shut down Emacs in an orderly fashion; if there are unsaved changes, this will produce auto-save files (see Auto-Saving) that the user can use after restarting the session to restore the unsaved edits.

If one of these events arrives in the middle of a key sequence—that is, after a prefix key—then Emacs reorders the events so that this event comes either before or after the multi-event key sequence, not within it.

Some of these special events, such as delete-frame, invoke Emacs commands by default; others are not bound. If you want to arrange for a special event to invoke a command, you can do that via special-event-map. The command you bind to a function key in that map can then examine the full event which invoked it in last-input-event. See Special Events.


22.7.13 Event Examples

If the user presses and releases the left mouse button over the same location, that generates a sequence of events like this:

(down-mouse-1 (#<window 18 on NEWS> 2613 (0 . 38) -864320))
(mouse-1      (#<window 18 on NEWS> 2613 (0 . 38) -864180))

While holding the control key down, the user might hold down the second mouse button, and drag the mouse from one line to the next. That produces two events, as shown here:

(C-down-mouse-2 (#<window 18 on NEWS> 3440 (0 . 27) -731219))
(C-drag-mouse-2 (#<window 18 on NEWS> 3440 (0 . 27) -731219)
                (#<window 18 on NEWS> 3510 (0 . 28) -729648))

While holding down the meta and shift keys, the user might press the second mouse button on the window’s mode line, and then drag the mouse into another window. That produces a pair of events like these:

(M-S-down-mouse-2 (#<window 18 on NEWS> mode-line (33 . 31) -457844))
(M-S-drag-mouse-2 (#<window 18 on NEWS> mode-line (33 . 31) -457844)
                  (#<window 20 on carlton-sanskrit.tex> 161 (33 . 3)
                   -453816))

The frame with input focus might not take up the entire screen, and the user might move the mouse outside the scope of the frame. Inside the track-mouse macro, that produces an event like this:

(mouse-movement (#<frame *ielm* 0x102849a30> nil (563 . 205) 532301936))

22.7.14 Classifying Events

Every event has an event type, which classifies the event for key binding purposes. For a keyboard event, the event type equals the event value; thus, the event type for a character is the character, and the event type for a function key symbol is the symbol itself. For events that are lists, the event type is the symbol in the CAR of the list. Thus, the event type is always a symbol or a character.

Two events of the same type are equivalent where key bindings are concerned; thus, they always run the same command. That does not necessarily mean they do the same things, however, as some commands look at the whole event to decide what to do. For example, some commands use the location of a mouse event to decide where in the buffer to act.

Sometimes broader classifications of events are useful. For example, you might want to ask whether an event involved the META key, regardless of which other key or mouse button was used.

The functions event-modifiers and event-basic-type are provided to get such information conveniently.

Function: event-modifiers event

This function returns a list of the modifiers that event has. The modifiers are symbols; they include shift, control, meta, alt, hyper and super. In addition, the modifiers list of a mouse event symbol always contains one of click, drag, and down. For double or triple events, it also contains double or triple.

The argument event may be an entire event object, or just an event type. If event is a symbol that has never been used in an event that has been read as input in the current Emacs session, then event-modifiers can return nil, even when event actually has modifiers.

Here are some examples:

(event-modifiers ?a)
     ⇒ nil
(event-modifiers ?A)
     ⇒ (shift)
(event-modifiers ?\C-a)
     ⇒ (control)
(event-modifiers ?\C-%)
     ⇒ (control)
(event-modifiers ?\C-\S-a)
     ⇒ (control shift)
(event-modifiers 'f5)
     ⇒ nil
(event-modifiers 's-f5)
     ⇒ (super)
(event-modifiers 'M-S-f5)
     ⇒ (meta shift)
(event-modifiers 'mouse-1)
     ⇒ (click)
(event-modifiers 'down-mouse-1)
     ⇒ (down)

The modifiers list for a click event explicitly contains click, but the event symbol name itself does not contain ‘click’. Similarly, the modifiers list for an ASCII control character, such as ‘C-a’, contains control, even though reading such an event via read-char will return the value 1 with the control modifier bit removed.

Function: event-basic-type event

This function returns the key or mouse button that event describes, with all modifiers removed. The event argument is as in event-modifiers. For example:

(event-basic-type ?a)
     ⇒ 97
(event-basic-type ?A)
     ⇒ 97
(event-basic-type ?\C-a)
     ⇒ 97
(event-basic-type ?\C-\S-a)
     ⇒ 97
(event-basic-type 'f5)
     ⇒ f5
(event-basic-type 's-f5)
     ⇒ f5
(event-basic-type 'M-S-f5)
     ⇒ f5
(event-basic-type 'down-mouse-1)
     ⇒ mouse-1
Function: mouse-movement-p object

This function returns non-nil if object is a mouse movement event. See Motion Events.


22.7.15 Accessing Mouse Events

This section describes convenient functions for accessing the data in a mouse button or motion event. Keyboard event data can be accessed using the same functions, but data elements that aren’t applicable to keyboard events are zero or nil.

The following two functions return a mouse position list (see Click Events), specifying the position of a mouse event.

Function: event-start event

This returns the starting position of event.

If event is a click or button-down event, this returns the location of the event. If event is a drag event, this returns the drag’s starting position.

Function: event-end event

This returns the ending position of event.

If event is a drag event, this returns the position where the user released the mouse button. If event is a click or button-down event, the value is actually the starting position, which is the only position such events have.

Function: posnp object

This function returns non-nil if object is a mouse position list, in the format documented in Click Events); and nil otherwise.

These functions take a mouse position list as argument, and return various parts of it:

Function: posn-window position

Return the window that position is in. If position represents a location outside the frame where the event was initiated, return that frame instead.

Function: posn-area position

Return the window area recorded in position. It returns nil when the event occurred in the text area of the window; otherwise, it is a symbol identifying the area in which the event occurred.

Function: posn-point position

Return the buffer position in position. When the event occurred in the text area of the window, in a marginal area, or on a fringe, this is an integer specifying a buffer position. Otherwise, the value is undefined.

Function: posn-x-y position

Return the pixel-based x and y coordinates in position, as a cons cell (x . y). These coordinates are relative to the window given by posn-window.

This example shows how to convert the window-relative coordinates in the text area of a window into frame-relative coordinates:

(defun frame-relative-coordinates (position)
  "Return frame-relative coordinates from POSITION.
POSITION is assumed to lie in a window text area."
  (let* ((x-y (posn-x-y position))
         (window (posn-window position))
         (edges (window-inside-pixel-edges window)))
    (cons (+ (car x-y) (car edges))
          (+ (cdr x-y) (cadr edges)))))
Function: posn-col-row position &optional use-window

This function returns a cons cell (col .  row), containing the estimated column and row corresponding to buffer position described by position. The return value is given in units of the frame’s default character width and default line height (including spacing), as computed from the x and y values corresponding to position. (So, if the actual characters have non-default sizes, the actual row and column may differ from these computed values.) If the optional window argument is non-nil, use the default character width in the window indicated by position instead of the frame. (This makes a difference if that window is showing a buffer with a non-default zooming level, for instance.)

Note that row is counted from the top of the text area. If the window given by position possesses a header line (see Window Header Lines) or a tab line, they are not included in the row count.

Function: posn-actual-col-row position

Return the actual row and column in position, as a cons cell (col . row). The values are the actual row and column numbers in the window given by position. See Click Events, for details. The function returns nil if position does not include actual position values; in that case posn-col-row can be used to get approximate values.

Note that this function doesn’t account for the visual width of characters on display, like the number of visual columns taken by a tab character or an image. If you need the coordinates in canonical character units, use posn-col-row instead.

Function: posn-string position

Return the string object described by position, either nil (which means position describes buffer text), or a cons cell (string . string-pos).

Function: posn-image position

Return the image object in position, either nil (if there’s no image at position), or an image spec (image …).

Function: posn-object position

Return the image or string object described by position, either nil (which means position describes buffer text), an image (image …), or a cons cell (string . string-pos).

Function: posn-object-x-y position

Return the pixel-based x and y coordinates relative to the upper left corner of the object described by position, as a cons cell (dx . dy). If the position describes buffer text, return the relative coordinates of the buffer-text character closest to that position.

Function: posn-object-width-height position

Return the pixel width and height of the object described by position, as a cons cell (width . height). If the position describes a buffer position, return the size of the character at that position.

Function: posn-timestamp position

Return the timestamp in position. This is the time at which the event occurred, in milliseconds. Such a timestamp is reported relative to an arbitrary starting time that varies according to the window system in use. On the X Window System, for example, it is the number of milliseconds since the X server was started.

These functions compute a position list given particular buffer position or screen position. You can access the data in this position list with the functions described above.

Function: posn-at-point &optional pos window

This function returns a position list for position pos in window. pos defaults to point in window; window defaults to the selected window.

posn-at-point returns nil if pos is not visible in window.

Function: posn-at-x-y x y &optional frame-or-window whole

This function returns position information corresponding to pixel coordinates x and y in a specified frame or window, frame-or-window, which defaults to the selected window. The coordinates x and y are relative to the text area of the selected window. If whole is non-nil, the x coordinate is relative to the entire window area including scroll bars, margins and fringes.


22.7.16 Accessing Scroll Bar Events

These functions are useful for decoding scroll bar events.

Function: scroll-bar-event-ratio event

This function returns the fractional vertical position of a scroll bar event within the scroll bar. The value is a cons cell (portion . whole) containing two integers whose ratio is the fractional position.

Function: scroll-bar-scale ratio total

This function multiplies (in effect) ratio by total, rounding the result to an integer. The argument ratio is not a number, but rather a pair (num . denom)—typically a value returned by scroll-bar-event-ratio.

This function is handy for scaling a position on a scroll bar into a buffer position. Here’s how to do that:

(+ (point-min)
   (scroll-bar-scale
      (posn-x-y (event-start event))
      (- (point-max) (point-min))))

Recall that scroll bar events have two integers forming a ratio, in place of a pair of x and y coordinates.


22.7.17 Putting Keyboard Events in Strings

In most of the places where strings are used, we conceptualize the string as containing text characters—the same kind of characters found in buffers or files. Occasionally Lisp programs use strings that conceptually contain keyboard characters; for example, they may be key sequences or keyboard macro definitions. However, storing keyboard characters in a string is a complex matter, for reasons of historical compatibility, and it is not always possible.

We recommend that new programs avoid dealing with these complexities by not storing keyboard events in strings containing control characters or the like, but instead store them in the common Emacs format as understood by key-valid-p.

If you read a key sequence with read-key-sequence-vector (or read-key-sequence), or access a key sequence with this-command-keys-vector (or this-command-keys), you can transform this to the recommended format by using key-description.

The complexities stem from the modifier bits that keyboard input characters can include. Aside from the Meta modifier, none of these modifier bits can be included in a string, and the Meta modifier is allowed only in special cases.

The earliest GNU Emacs versions represented meta characters as codes in the range of 128 to 255. At that time, the basic character codes ranged from 0 to 127, so all keyboard character codes did fit in a string. Many Lisp programs used ‘\M-’ in string constants to stand for meta characters, especially in arguments to define-key and similar functions, and key sequences and sequences of events were always represented as strings.

When we added support for larger basic character codes beyond 127, and additional modifier bits, we had to change the representation of meta characters. Now the flag that represents the Meta modifier in a character is 2**27 and such numbers cannot be included in a string.

To support programs with ‘\M-’ in string constants, there are special rules for including certain meta characters in a string. Here are the rules for interpreting a string as a sequence of input characters:

  • If the keyboard character value is in the range of 0 to 127, it can go in the string unchanged.
  • The meta variants of those characters, with codes in the range of 2**27 to 2**27+127, can also go in the string, but you must change their numeric values. You must set the 2**7 bit instead of the 2**27 bit, resulting in a value between 128 and 255. Only a unibyte string can include these codes.
  • Non-ASCII characters above 256 can be included in a multibyte string.
  • Other keyboard character events cannot fit in a string. This includes keyboard events in the range of 128 to 255.

Functions such as read-key-sequence that construct strings of keyboard input characters follow these rules: they construct vectors instead of strings, when the events won’t fit in a string.

When you use the read syntax ‘\M-’ in a string, it produces a code in the range of 128 to 255—the same code that you get if you modify the corresponding keyboard event to put it in the string. Thus, meta events in strings work consistently regardless of how they get into the strings.

However, most programs would do well to avoid these issues by following the recommendations at the beginning of this section.


22.8 Reading Input

The editor command loop reads key sequences using the function read-key-sequence, which uses read-event. These and other functions for event input are also available for use in Lisp programs. See also momentary-string-display in Temporary Displays, and sit-for in Waiting for Elapsed Time or Input. See Terminal Input, for functions and variables for controlling terminal input modes and debugging terminal input.

For higher-level input facilities, see Minibuffers.


22.8.1 Key Sequence Input

The command loop reads input a key sequence at a time, by calling read-key-sequence. Lisp programs can also call this function; for example, describe-key uses it to read the key to describe.

Function: read-key-sequence prompt &optional continue-echo dont-downcase-last switch-frame-ok command-loop

This function reads a key sequence and returns it as a string or vector. It keeps reading events until it has accumulated a complete key sequence; that is, enough to specify a non-prefix command using the currently active keymaps. (Remember that a key sequence that starts with a mouse event is read using the keymaps of the buffer in the window that the mouse was in, not the current buffer.)

If the events are all characters and all can fit in a string, then read-key-sequence returns a string (see Putting Keyboard Events in Strings). Otherwise, it returns a vector, since a vector can hold all kinds of events—characters, symbols, and lists. The elements of the string or vector are the events in the key sequence.

Reading a key sequence includes translating the events in various ways. See Keymaps for Translating Sequences of Events.

The argument prompt is either a string to be displayed in the echo area as a prompt, or nil, meaning not to display a prompt. The argument continue-echo, if non-nil, means to echo this key as a continuation of the previous key.

Normally any upper case event is converted to lower case if the original event is undefined and the lower case equivalent is defined. The argument dont-downcase-last, if non-nil, means do not convert the last event to lower case. This is appropriate for reading a key sequence to be defined.

The argument switch-frame-ok, if non-nil, means that this function should process a switch-frame event if the user switches frames before typing anything. If the user switches frames in the middle of a key sequence, or at the start of the sequence but switch-frame-ok is nil, then the event will be put off until after the current key sequence.

The argument command-loop, if non-nil, means that this key sequence is being read by something that will read commands one after another. It should be nil if the caller will read just one key sequence.

In the following example, Emacs displays the prompt ‘?’ in the echo area, and then the user types C-x C-f.

(read-key-sequence "?")

---------- Echo Area ----------
?C-x C-f
---------- Echo Area ----------

     ⇒ "^X^F"

The function read-key-sequence suppresses quitting: C-g typed while reading with this function works like any other character, and does not set quit-flag. See Quitting.

Function: read-key-sequence-vector prompt &optional continue-echo dont-downcase-last switch-frame-ok command-loop

This is like read-key-sequence except that it always returns the key sequence as a vector, never as a string. See Putting Keyboard Events in Strings.

If an input character is upper-case (or has the shift modifier) and has no key binding, but its lower-case equivalent has one, then read-key-sequence converts the character to lower case. (This behavior can be disabled by setting the translate-upper-case-key-bindings user option to nil.) Note that lookup-key does not perform case conversion in this way.

When reading input results in such a shift-translation, Emacs sets the variable this-command-keys-shift-translated to a non-nil value. Lisp programs can examine this variable if they need to modify their behavior when invoked by shift-translated keys. For example, the function handle-shift-selection examines the value of this variable to determine how to activate or deactivate the region (see handle-shift-selection).

The function read-key-sequence also transforms some mouse events. It converts unbound drag events into click events, and discards unbound button-down events entirely. It also reshuffles focus events and miscellaneous window events so that they never appear in a key sequence with any other events.

When mouse events occur in special parts of a window or frame, such as a mode line or a scroll bar, the event type shows nothing special—it is the same symbol that would normally represent that combination of mouse button and modifier keys. The information about the window part is kept elsewhere in the event—in the coordinates. But read-key-sequence translates this information into imaginary prefix keys, all of which are symbols: tab-line, header-line, horizontal-scroll-bar, menu-bar, tab-bar, mode-line, vertical-line, vertical-scroll-bar, left-margin, right-margin, left-fringe, right-fringe, right-divider, and bottom-divider. You can define meanings for mouse clicks in special window parts by defining key sequences using these imaginary prefix keys.

For example, if you call read-key-sequence and then click the mouse on the window’s mode line, you get two events, like this:

(read-key-sequence "Click on the mode line: ")
     ⇒ [mode-line
         (mouse-1
          (#<window 6 on NEWS> mode-line
           (40 . 63) 5959987))]
Variable: num-input-keys

This variable’s value is the number of key sequences processed so far in this Emacs session. This includes key sequences read from the terminal and key sequences read from keyboard macros being executed.


22.8.2 Reading One Event

The lowest level functions for command input are read-event, read-char, and read-char-exclusive.

If you need a function to read a character using the minibuffer, use read-char-from-minibuffer (see Asking Multiple-Choice Questions).

Function: read-event &optional prompt inherit-input-method seconds

This function reads and returns the next event of command input, waiting if necessary until an event is available.

The returned event may come directly from the user, or from a keyboard macro. It is not decoded by the keyboard’s input coding system (see Terminal I/O Encoding).

If the optional argument prompt is non-nil, it should be a string to display in the echo area as a prompt. If prompt is nil or the string ‘""’, read-event does not display any message to indicate it is waiting for input; instead, it prompts by echoing: it displays descriptions of the events that led to or were read by the current command. See The Echo Area.

If inherit-input-method is non-nil, then the current input method (if any) is employed to make it possible to enter a non-ASCII character. Otherwise, input method handling is disabled for reading this event.

If cursor-in-echo-area is non-nil, then read-event moves the cursor temporarily to the echo area, to the end of any message displayed there. Otherwise read-event does not move the cursor.

If seconds is non-nil, it should be a number specifying the maximum time to wait for input, in seconds. If no input arrives within that time, read-event stops waiting and returns nil. A floating point seconds means to wait for a fractional number of seconds. Some systems support only a whole number of seconds; on these systems, seconds is rounded down. If seconds is nil, read-event waits as long as necessary for input to arrive.

If seconds is nil, Emacs is considered idle while waiting for user input to arrive. Idle timers—those created with run-with-idle-timer (see Idle Timers)—can run during this period. However, if seconds is non-nil, the state of idleness remains unchanged. If Emacs is non-idle when read-event is called, it remains non-idle throughout the operation of read-event; if Emacs is idle (which can happen if the call happens inside an idle timer), it remains idle.

If read-event gets an event that is defined as a help character, then in some cases read-event processes the event directly without returning. See Help Functions. Certain other events, called special events, are also processed directly within read-event (see Special Events).

Here is what happens if you call read-event and then press the right-arrow function key:

(read-event)
     ⇒ right
Function: read-char &optional prompt inherit-input-method seconds

This function reads and returns a character input event. If the user generates an event which is not a character (i.e., a mouse click or function key event), read-char signals an error. The arguments work as in read-event.

If the event has modifiers, Emacs attempts to resolve them and return the code of the corresponding character. For example, if the user types C-a, the function returns 1, which is the ASCII code of the ‘C-a’ character. If some of the modifiers cannot be reflected in the character code, read-char leaves the unresolved modifier bits set in the returned event. For example, if the user types C-M-a, the function returns 134217729, 8000001 in hex, i.e. ‘C-a’ with the Meta modifier bit set. This value is not a valid character code: it fails the characterp test (see Character Codes). Use event-basic-type (see Classifying Events) to recover the character code with the modifier bits removed; use event-modifiers to test for modifiers in the character event returned by read-char.

In the first example below, the user types the character 1 (ASCII code 49). The second example shows a keyboard macro definition that calls read-char from the minibuffer using eval-expression. read-char reads the keyboard macro’s very next character, which is 1. Then eval-expression displays its return value in the echo area.

(read-char)
     ⇒ 49

;; We assume here you use M-: to evaluate this.
(symbol-function 'foo)
     ⇒ "^[:(read-char)^M1"
(execute-kbd-macro 'foo)
     -| 49
     ⇒ nil
Function: read-char-exclusive &optional prompt inherit-input-method seconds

This function reads and returns a character input event. If the user generates an event which is not a character event, read-char-exclusive ignores it and reads another event, until it gets a character. The arguments work as in read-event. The returned value may include modifier bits, as with read-char.

None of the above functions suppress quitting.

Variable: num-nonmacro-input-events

This variable holds the total number of input events received so far from the terminal—not counting those generated by keyboard macros.

We emphasize that, unlike read-key-sequence, the functions read-event, read-char, and read-char-exclusive do not perform the translations described in Keymaps for Translating Sequences of Events. If you wish to read a single key taking these translations into account (for example, to read Function Keys in a terminal or Mouse Events from xterm-mouse-mode), use the function read-key:

Function: read-key &optional prompt disable-fallbacks

This function reads a single key. It is intermediate between read-key-sequence and read-event. Unlike the former, it reads a single key, not a key sequence. Unlike the latter, it does not return a raw event, but decodes and translates the user input according to input-decode-map, local-function-key-map, and key-translation-map (see Keymaps for Translating Sequences of Events).

The argument prompt is either a string to be displayed in the echo area as a prompt, or nil, meaning not to display a prompt.

If argument disable-fallbacks is non-nil then the usual fallback logic for unbound keys in read-key-sequence is not applied. This means that mouse button-down and multi-click events will not be discarded and local-function-key-map and key-translation-map will not get applied. If nil or unspecified, the only fallback disabled is downcasing of the last event.

Function: read-char-choice prompt chars &optional inhibit-quit

This function uses read-from-minibuffer to read and return a single character that is a member of chars, which should be a list of single characters. It discards any input characters that are not members of chars, and shows a message to that effect.

The optional argument inhibit-quit is by default ignored, but if the variable read-char-choice-use-read-key is non-nil, this function uses read-key instead of read-from-minibuffer, and in that case inhibit-quit non-nil means ignore keyboard-quit events while waiting for valid input. In addition, if read-char-choice-use-read-key is non-nil, binding help-form (see Help Functions) to a non-nil value while calling this function causes it to evaluate help-form and display the result when the user presses help-char; it then continues to wait for a valid input character, or for keyboard-quit.

Function: read-multiple-choice prompt choices &optional help-string show-help long-form

Ask user a multiple choice question. prompt should be a string that will be displayed as the prompt.

choices is an alist where the first element in each entry is a character to be entered, the second element is a short name for the entry to be displayed while prompting (if there’s room, it might be shortened), and the third, optional entry is a longer explanation that will be displayed in a help buffer if the user requests more help.

If optional argument help-string is non-nil, it should be a string with a more detailed description of all choices. It will be displayed in a help buffer instead of the default auto-generated description when the user types ?.

If optional argument show-help is non-nil, the help buffer will be displayed immediately, before any user input. If it is a string, use it as the name of the help buffer.

If optional argument long-form is non-nil, the user will have to type in long-form answers (using completing-read) instead of hitting a single key. The answers must be among the second elements of the values in the choices list.

The return value is the matching value from choices.

(read-multiple-choice
 "Continue connecting?"
 '((?a "always" "Accept certificate for this and future sessions.")
   (?s "session only" "Accept certificate this session only.")
   (?n "no" "Refuse to use certificate, close connection.")))

The read-multiple-choice-face face is used to highlight the matching characters in the name string on graphical terminals.


22.8.3 Modifying and Translating Input Events

Emacs modifies every event it reads according to extra-keyboard-modifiers, then translates it through keyboard-translate-table (if applicable), before returning it from read-event.

Variable: extra-keyboard-modifiers

This variable lets Lisp programs “press” the modifier keys on the keyboard. The value is a character. Only the modifiers of the character matter. Each time the user types a keyboard key, it is altered as if those modifier keys were held down. For instance, if you bind extra-keyboard-modifiers to ?\C-\M-a, then all keyboard input characters typed during the scope of the binding will have the control and meta modifiers applied to them. The character ?\C-@, equivalent to the integer 0, does not count as a control character for this purpose, but as a character with no modifiers. Thus, setting extra-keyboard-modifiers to zero cancels any modification.

When using a window system, the program can press any of the modifier keys in this way. Otherwise, only the CTL and META keys can be virtually pressed.

Note that this variable applies only to events that really come from the keyboard, and has no effect on mouse events or any other events.

Variable: keyboard-translate-table

This terminal-local variable is the translate table for keyboard characters. It lets you reshuffle the keys on the keyboard without changing any command bindings. Its value is normally a char-table, or else nil. (It can also be a string or vector, but this is considered obsolete.)

If keyboard-translate-table is a char-table (see Char-Tables), then each character read from the keyboard is looked up in this char-table. If the value found there is non-nil, then it is used instead of the actual input character.

Note that this translation is the first thing that happens to a character after it is read from the terminal. Record-keeping features such as recent-keys and dribble files record the characters after translation.

Note also that this translation is done before the characters are supplied to input methods (see Input Methods). Use translation-table-for-input (see Translation of Characters), if you want to translate characters after input methods operate.

Function: key-translate from to

This function modifies keyboard-translate-table to translate character code from into character code to. It creates the keyboard translate table if necessary.

Here’s an example of using the keyboard-translate-table to make C-x, C-c and C-v perform the cut, copy and paste operations:

(key-translate "C-x" "<control-x>")
(key-translate "C-c" "<control-c>")
(key-translate "C-v" "<control-v>")
(keymap-global-set "<control-x>" 'kill-region)
(keymap-global-set "<control-c>" 'kill-ring-save)
(keymap-global-set "<control-v>" 'yank)

On a graphical terminal that supports extended ASCII input, you can still get the standard Emacs meanings of one of those characters by typing it with the shift key. That makes it a different character as far as keyboard translation is concerned, but it has the same usual meaning.

See Keymaps for Translating Sequences of Events, for mechanisms that translate event sequences at the level of read-key-sequence.


22.8.4 Invoking the Input Method

The event-reading functions invoke the current input method, if any (see Input Methods). If the value of input-method-function is non-nil, it should be a function; when read-event reads a printing character (including SPC) with no modifier bits, it calls that function, passing the character as an argument.

Variable: input-method-function

If this is non-nil, its value specifies the current input method function.

Warning: don’t bind this variable with let. It is often buffer-local, and if you bind it around reading input (which is exactly when you would bind it), switching buffers asynchronously while Emacs is waiting will cause the value to be restored in the wrong buffer.

The input method function should return a list of events which should be used as input. (If the list is nil, that means there is no input, so read-event waits for another event.) These events are processed before the events in unread-command-events (see Miscellaneous Event Input Features). Events returned by the input method function are not passed to the input method function again, even if they are printing characters with no modifier bits.

If the input method function calls read-event or read-key-sequence, it should bind input-method-function to nil first, to prevent recursion.

The input method function is not called when reading the second and subsequent events of a key sequence. Thus, these characters are not subject to input method processing. The input method function should test the values of overriding-local-map and overriding-terminal-local-map; if either of these variables is non-nil, the input method should put its argument into a list and return that list with no further processing.


22.8.5 Quoted Character Input

You can use the function read-quoted-char to ask the user to specify a character, and allow the user to specify a control or meta character conveniently, either literally or as an octal character code. The command quoted-insert uses this function.

Function: read-quoted-char &optional prompt

This function is like read-char, except that if the first character read is an octal digit (0–7), it reads any number of octal digits (but stopping if a non-octal digit is found), and returns the character represented by that numeric character code. If the character that terminates the sequence of octal digits is RET, it is discarded. Any other terminating character is used as input after this function returns.

Quitting is suppressed when the first character is read, so that the user can enter a C-g. See Quitting.

If prompt is supplied, it specifies a string for prompting the user. The prompt string is always displayed in the echo area, followed by a single ‘-’.

In the following example, the user types in the octal number 177 (which is 127 in decimal).

(read-quoted-char "What character")

---------- Echo Area ----------
What character 1 7 7-
---------- Echo Area ----------

     ⇒ 127

22.8.6 Miscellaneous Event Input Features

This section describes how to peek ahead at events without using them up, how to check for pending input, and how to discard pending input. See also the function read-passwd (see Reading a Password).

Variable: unread-command-events

This variable holds a list of events waiting to be read as command input. The events are used in the order they appear in the list, and removed one by one as they are used.

The variable is needed because in some cases a function reads an event and then decides not to use it. Storing the event in this variable causes it to be processed normally, by the command loop or by the functions to read command input.

For example, the function that implements numeric prefix arguments reads any number of digits. When it finds a non-digit event, it must unread the event so that it can be read normally by the command loop. Likewise, incremental search uses this feature to unread events with no special meaning in a search, because these events should exit the search and then execute normally.

The reliable and easy way to extract events from a key sequence so as to put them in unread-command-events is to use listify-key-sequence (see below).

Normally you add events to the front of this list, so that the events most recently unread will be reread first.

Events read from this list are not normally added to the current command’s key sequence (as returned by, e.g., this-command-keys), as the events will already have been added once as they were read for the first time. An element of the form (t . event) forces event to be added to the current command’s key sequence.

Elements read from this list are normally recorded by the record-keeping features (see Recording Input) and while defining a keyboard macro (see Keyboard Macros). However, an element of the form (no-record . event) causes event to be processed normally without recording it.

Function: listify-key-sequence key

This function converts the string or vector key to a list of individual events, which you can put in unread-command-events.

Function: input-pending-p &optional check-timers

This function determines whether any command input is currently available to be read. It returns immediately, with value t if there is available input, nil otherwise. On rare occasions it may return t when no input is available.

If the optional argument check-timers is non-nil, then if no input is available, Emacs runs any timers which are ready. See Timers for Delayed Execution.

Variable: last-input-event

This variable records the last terminal input event read, whether as part of a command or explicitly by a Lisp program.

In the example below, the Lisp program reads the character 1, ASCII code 49. It becomes the value of last-input-event, while C-e (we assume C-x C-e command is used to evaluate this expression) remains the value of last-command-event.

(progn (print (read-char))
       (print last-command-event)
       last-input-event)
     -| 49
     -| 5
     ⇒ 49
Macro: while-no-input body…

This construct runs the body forms and returns the value of the last one—but only if no input arrives. If any input arrives during the execution of the body forms, it aborts them (working much like a quit). The while-no-input form returns nil if aborted by a real quit, and returns t if aborted by arrival of other input.

If a part of body binds inhibit-quit to non-nil, arrival of input during those parts won’t cause an abort until the end of that part.

If you want to be able to distinguish all possible values computed by body from both kinds of abort conditions, write the code like this:

(while-no-input
  (list
    (progn . body)))
Variable: while-no-input-ignore-events

This variable allow setting which special events while-no-input should ignore. It is a list of event symbols (see Event Examples).

Function: discard-input

This function discards the contents of the terminal input buffer and cancels any keyboard macro that might be in the process of definition. It returns nil.

In the following example, the user may type a number of characters right after starting the evaluation of the form. After the sleep-for finishes sleeping, discard-input discards any characters typed during the sleep.

(progn (sleep-for 2)
       (discard-input))
     ⇒ nil

22.9 Special Events

Certain special events are handled at a very low level—as soon as they are read. The read-event function processes these events itself, and never returns them. Instead, it keeps waiting for the first event that is not special and returns that one.

Special events do not echo, they are never grouped into key sequences, and they never appear in the value of last-command-event or (this-command-keys). They do not discard a numeric argument, they cannot be unread with unread-command-events, they may not appear in a keyboard macro, and they are not recorded in a keyboard macro while you are defining one.

Special events do, however, appear in last-input-event immediately after they are read, and this is the way for the event’s definition to find the actual event.

The events types iconify-frame, make-frame-visible, delete-frame, drag-n-drop, language-change, and user signals like sigusr1 are normally handled in this way. The keymap which defines how to handle special events—and which events are special—is in the variable special-event-map (see Controlling the Active Keymaps).


22.10 Waiting for Elapsed Time or Input

The wait functions are designed to wait for a certain amount of time to pass or until there is input. For example, you may wish to pause in the middle of a computation to allow the user time to view the display. sit-for pauses and updates the screen, and returns immediately if input comes in, while sleep-for pauses without updating the screen.

Function: sit-for seconds &optional nodisp

This function performs redisplay (provided there is no pending input from the user), then waits seconds seconds, or until input is available. The usual purpose of sit-for is to give the user time to read text that you display. The value is t if sit-for waited the full time with no input arriving (see Miscellaneous Event Input Features). Otherwise, the value is nil.

The argument seconds need not be an integer. If it is floating point, sit-for waits for a fractional number of seconds. Some systems support only a whole number of seconds; on these systems, seconds is rounded down.

The expression (sit-for 0) is equivalent to (redisplay), i.e., it requests a redisplay, without any delay, if there is no pending input. See Forcing Redisplay.

If nodisp is non-nil, then sit-for does not redisplay, but it still returns as soon as input is available (or when the timeout elapses).

In batch mode (see Batch Mode), sit-for cannot be interrupted, even by input from the standard input descriptor. It is thus equivalent to sleep-for, which is described below.

It is also possible to call sit-for with three arguments, as (sit-for seconds millisec nodisp), but that is considered obsolete.

Function: sleep-for seconds &optional millisec

This function simply pauses for seconds seconds without updating the display. It pays no attention to available input. It returns nil.

The argument seconds need not be an integer. If it is floating point, sleep-for waits for a fractional number of seconds. Some systems support only a whole number of seconds; on these systems, seconds is rounded down.

The optional argument millisec specifies an additional waiting period measured in milliseconds. This adds to the period specified by seconds. If the system doesn’t support waiting fractions of a second, you get an error if you specify nonzero millisec.

Use sleep-for when you wish to guarantee a delay.

See Time of Day, for functions to get the current time.


22.11 Quitting

Typing C-g while a Lisp function is running causes Emacs to quit whatever it is doing. This means that control returns to the innermost active command loop.

Typing C-g while the command loop is waiting for keyboard input does not cause a quit; it acts as an ordinary input character. In the simplest case, you cannot tell the difference, because C-g normally runs the command keyboard-quit, whose effect is to quit. However, when C-g follows a prefix key, they combine to form an undefined key. The effect is to cancel the prefix key as well as any prefix argument.

In the minibuffer, C-g has a different definition: it aborts out of the minibuffer. This means, in effect, that it exits the minibuffer and then quits. (Simply quitting would return to the command loop within the minibuffer.) The reason why C-g does not quit directly when the command reader is reading input is so that its meaning can be redefined in the minibuffer in this way. C-g following a prefix key is not redefined in the minibuffer, and it has its normal effect of canceling the prefix key and prefix argument. This too would not be possible if C-g always quit directly.

When C-g does directly quit, it does so by setting the variable quit-flag to t. Emacs checks this variable at appropriate times and quits if it is not nil. Setting quit-flag non-nil in any way thus causes a quit.

At the level of C code, quitting cannot happen just anywhere; only at the special places that check quit-flag. The reason for this is that quitting at other places might leave an inconsistency in Emacs’s internal state. Because quitting is delayed until a safe place, quitting cannot make Emacs crash.

Certain functions such as read-key-sequence or read-quoted-char prevent quitting entirely even though they wait for input. Instead of quitting, C-g serves as the requested input. In the case of read-key-sequence, this serves to bring about the special behavior of C-g in the command loop. In the case of read-quoted-char, this is so that C-q can be used to quote a C-g.

You can prevent quitting for a portion of a Lisp function by binding the variable inhibit-quit to a non-nil value. Then, although C-g still sets quit-flag to t as usual, the usual result of this—a quit—is prevented. Eventually, inhibit-quit will become nil again, such as when its binding is unwound at the end of a let form. At that time, if quit-flag is still non-nil, the requested quit happens immediately. This behavior is ideal when you wish to make sure that quitting does not happen within a critical section of the program.

In some functions (such as read-quoted-char), C-g is handled in a special way that does not involve quitting. This is done by reading the input with inhibit-quit bound to t, and setting quit-flag to nil before inhibit-quit becomes nil again. This excerpt from the definition of read-quoted-char shows how this is done; it also shows that normal quitting is permitted after the first character of input.

(defun read-quoted-char (&optional prompt)
  "…documentation…"
  (let ((message-log-max nil) done (first t) (code 0) char)
    (while (not done)
      (let ((inhibit-quit first)
            …)
        (and prompt (message "%s-" prompt))
        (setq char (read-event))
        (if inhibit-quit (setq quit-flag nil)))
      …set the variable code)
    code))
Variable: quit-flag

If this variable is non-nil, then Emacs quits immediately, unless inhibit-quit is non-nil. Typing C-g ordinarily sets quit-flag non-nil, regardless of inhibit-quit.

Variable: inhibit-quit

This variable determines whether Emacs should quit when quit-flag is set to a value other than nil. If inhibit-quit is non-nil, then quit-flag has no special effect.

Macro: with-local-quit body…

This macro executes body forms in sequence, but allows quitting, at least locally, within body even if inhibit-quit was non-nil outside this construct. It returns the value of the last form in body, unless exited by quitting, in which case it returns nil.

If inhibit-quit is nil on entry to with-local-quit, it only executes the body, and setting quit-flag causes a normal quit. However, if inhibit-quit is non-nil so that ordinary quitting is delayed, a non-nil quit-flag triggers a special kind of local quit. This ends the execution of body and exits the with-local-quit body with quit-flag still non-nil, so that another (ordinary) quit will happen as soon as that is allowed. If quit-flag is already non-nil at the beginning of body, the local quit happens immediately and the body doesn’t execute at all.

This macro is mainly useful in functions that can be called from timers, process filters, process sentinels, pre-command-hook, post-command-hook, and other places where inhibit-quit is normally bound to t.

Command: keyboard-quit

This function signals the quit condition with (signal 'quit nil). This is the same thing that quitting does. (See signal in Errors.)

To quit without aborting a keyboard macro definition or execution, you can signal the minibuffer-quit condition. This has almost the same effect as the quit condition except that the error handling in the command loop handles it without exiting keyboard macro definition or execution.

You can specify a character other than C-g to use for quitting. See the function set-input-mode in Input Modes.


22.12 Prefix Command Arguments

Most Emacs commands can use a prefix argument, a number specified before the command itself. (Don’t confuse prefix arguments with prefix keys.) The prefix argument is at all times represented by a value, which may be nil, meaning there is currently no prefix argument. Each command may use the prefix argument or ignore it.

There are two representations of the prefix argument: raw and numeric. The editor command loop uses the raw representation internally, and so do the Lisp variables that store the information, but commands can request either representation.

Here are the possible values of a raw prefix argument:

  • nil, meaning there is no prefix argument. Its numeric value is 1, but numerous commands make a distinction between nil and the integer 1.
  • An integer, which stands for itself.
  • A list of one element, which is an integer. This form of prefix argument results from one or a succession of C-us with no digits. The numeric value is the integer in the list, but some commands make a distinction between such a list and an integer alone.
  • The symbol -. This indicates that M-- or C-u - was typed, without following digits. The equivalent numeric value is −1, but some commands make a distinction between the integer −1 and the symbol -.

We illustrate these possibilities by calling the following function with various prefixes:

(defun display-prefix (arg)
  "Display the value of the raw prefix arg."
  (interactive "P")
  (message "%s" arg))

Here are the results of calling display-prefix with various raw prefix arguments:

        M-x display-prefix  -| nil

C-u     M-x display-prefix  -| (4)

C-u C-u M-x display-prefix  -| (16)

C-u 3   M-x display-prefix  -| 3

M-3     M-x display-prefix  -| 3      ; (Same as C-u 3.)

C-u -   M-x display-prefix  -| -

M--     M-x display-prefix  -| -      ; (Same as C-u -.)

C-u - 7 M-x display-prefix  -| -7

M-- 7   M-x display-prefix  -| -7     ; (Same as C-u -7.)

Emacs uses two variables to store the prefix argument: prefix-arg and current-prefix-arg. Commands such as universal-argument that set up prefix arguments for other commands store them in prefix-arg. In contrast, current-prefix-arg conveys the prefix argument to the current command, so setting it has no effect on the prefix arguments for future commands.

Normally, commands specify which representation to use for the prefix argument, either numeric or raw, in the interactive specification. (See Using interactive.) Alternatively, functions may look at the value of the prefix argument directly in the variable current-prefix-arg, but this is less clean.

Function: prefix-numeric-value arg

This function returns the numeric meaning of a valid raw prefix argument value, arg. The argument may be a symbol, a number, or a list. If it is nil, the value 1 is returned; if it is -, the value −1 is returned; if it is a number, that number is returned; if it is a list, the CAR of that list (which should be a number) is returned.

Variable: current-prefix-arg

This variable holds the raw prefix argument for the current command. Commands may examine it directly, but the usual method for accessing it is with (interactive "P").

Variable: prefix-arg

The value of this variable is the raw prefix argument for the next editing command. Commands such as universal-argument that specify prefix arguments for the following command work by setting this variable.

Variable: last-prefix-arg

The raw prefix argument value used by the previous command.

The following commands exist to set up prefix arguments for the following command. Do not call them for any other reason.

Command: universal-argument

This command reads input and specifies a prefix argument for the following command. Don’t call this command yourself unless you know what you are doing.

Command: digit-argument arg

This command adds to the prefix argument for the following command. The argument arg is the raw prefix argument as it was before this command; it is used to compute the updated prefix argument. Don’t call this command yourself unless you know what you are doing.

Command: negative-argument arg

This command adds to the numeric argument for the next command. The argument arg is the raw prefix argument as it was before this command; its value is negated to form the new prefix argument. Don’t call this command yourself unless you know what you are doing.


22.13 Recursive Editing

The Emacs command loop is entered automatically when Emacs starts up. This top-level invocation of the command loop never exits; it keeps running as long as Emacs does. Lisp programs can also invoke the command loop. Since this makes more than one activation of the command loop, we call it recursive editing. A recursive editing level has the effect of suspending whatever command invoked it and permitting the user to do arbitrary editing before resuming that command.

The commands available during recursive editing are the same ones available in the top-level editing loop and defined in the keymaps. Only a few special commands exit the recursive editing level; the others return to the recursive editing level when they finish. (The special commands for exiting are always available, but they do nothing when recursive editing is not in progress.)

All command loops, including recursive ones, set up all-purpose error handlers so that an error in a command run from the command loop will not exit the loop.

Minibuffer input is a special kind of recursive editing. It has a few special wrinkles, such as enabling display of the minibuffer and the minibuffer window, but fewer than you might suppose. Certain keys behave differently in the minibuffer, but that is only because of the minibuffer’s local map; if you switch windows, you get the usual Emacs commands.

To invoke a recursive editing level, call the function recursive-edit. This function contains the command loop; it also contains a call to catch with tag exit, which makes it possible to exit the recursive editing level by throwing to exit (see Explicit Nonlocal Exits: catch and throw). Throwing a t value causes recursive-edit to quit, so that control returns to the command loop one level up. This is called aborting, and is done by C-] (abort-recursive-edit). Similarly, you can throw a string value to make recursive-edit signal an error, printing this string as the message. If you throw a function, recursive-edit will call it without arguments before returning. Throwing any other value, will make recursive-edit return normally to the function that called it. The command C-M-c (exit-recursive-edit) does this.

Most applications should not use recursive editing, except as part of using the minibuffer. Usually it is more convenient for the user if you change the major mode of the current buffer temporarily to a special major mode, which should have a command to go back to the previous mode. (The e command in Rmail uses this technique.) Or, if you wish to give the user different text to edit recursively, create and select a new buffer in a special mode. In this mode, define a command to complete the processing and go back to the previous buffer. (The m command in Rmail does this.)

Recursive edits are useful in debugging. You can insert a call to debug into a function definition as a sort of breakpoint, so that you can look around when the function gets there. debug invokes a recursive edit but also provides the other features of the debugger.

Recursive editing levels are also used when you type C-r in query-replace or use C-x q (kbd-macro-query).

Command: recursive-edit

This function invokes the editor command loop. It is called automatically by the initialization of Emacs, to let the user begin editing. When called from a Lisp program, it enters a recursive editing level.

If the current buffer is not the same as the selected window’s buffer, recursive-edit saves and restores the current buffer. Otherwise, if you switch buffers, the buffer you switched to is current after recursive-edit returns.

In the following example, the function simple-rec first advances point one word, then enters a recursive edit, printing out a message in the echo area. The user can then do any editing desired, and then type C-M-c to exit and continue executing simple-rec.

(defun simple-rec ()
  (forward-word 1)
  (message "Recursive edit in progress")
  (recursive-edit)
  (forward-word 1))
     ⇒ simple-rec
(simple-rec)
     ⇒ nil
Command: exit-recursive-edit

This function exits from the innermost recursive edit (including minibuffer input). Its definition is effectively (throw 'exit nil).

Command: abort-recursive-edit

This function aborts the command that requested the innermost recursive edit (including minibuffer input), by signaling quit after exiting the recursive edit. Its definition is effectively (throw 'exit t). See Quitting.

Command: top-level

This function exits all recursive editing levels; it does not return a value, as it jumps completely out of any computation directly back to the main command loop.

Function: recursion-depth

This function returns the current depth of recursive edits. When no recursive edit is active, it returns 0.


22.14 Disabling Commands

Disabling a command marks the command as requiring user confirmation before it can be executed. Disabling is used for commands which might be confusing to beginning users, to prevent them from using the commands by accident.

The low-level mechanism for disabling a command is to put a non-nil disabled property on the Lisp symbol for the command. These properties are normally set up by the user’s init file (see The Init File) with Lisp expressions such as this:

(put 'upcase-region 'disabled t)

For a few commands, these properties are present by default (you can remove them in your init file if you wish).

If the value of the disabled property is a string, the message saying the command is disabled includes that string. For example:

(put 'delete-region 'disabled
     "Text deleted this way cannot be yanked back!\n")

See Disabling in The GNU Emacs Manual, for the details on what happens when a disabled command is invoked interactively. Disabling a command has no effect on calling it as a function from Lisp programs.

The value of the disabled property can also be a list where the first element is the symbol query. In that case, the user will be queried whether to execute the command. The second element in the list should be nil or non-nil to say whether to use y-or-n-p or yes-or-no-p, respectively, and the third element is the question to use. The command-query convenience function should be used to enable querying for a command.

Command: enable-command command

Allow command (a symbol) to be executed without special confirmation from now on, and alter the user’s init file (see The Init File) so that this will apply to future sessions.

Command: disable-command command

Require special confirmation to execute command from now on, and alter the user’s init file so that this will apply to future sessions.

Variable: disabled-command-function

The value of this variable should be a function. When the user invokes a disabled command interactively, this function is called instead of the disabled command. It can use this-command-keys to determine what the user typed to run the command, and thus find the command itself.

The value may also be nil. Then all commands work normally, even disabled ones.

By default, the value is a function that asks the user whether to proceed.


22.15 Command History

The command loop keeps a history of the complex commands that have been executed, to make it convenient to repeat these commands. A complex command is one for which the interactive argument reading uses the minibuffer. This includes any M-x command, any M-: command, and any command whose interactive specification reads an argument from the minibuffer. Explicit use of the minibuffer during the execution of the command itself does not cause the command to be considered complex.

Variable: command-history

This variable’s value is a list of recent complex commands, each represented as a form to evaluate. It continues to accumulate all complex commands for the duration of the editing session, but when it reaches the maximum size (see Minibuffer History), the oldest elements are deleted as new ones are added.

command-history
⇒ ((switch-to-buffer "chistory.texi")
    (describe-key "^X^[")
    (visit-tags-table "~/emacs/src/")
    (find-tag "repeat-complex-command"))

This history list is actually a special case of minibuffer history (see Minibuffer History), with one special twist: the elements are expressions rather than strings.

There are a number of commands devoted to the editing and recall of previous commands. The commands repeat-complex-command, and list-command-history are described in the user manual (see Repetition in The GNU Emacs Manual). Within the minibuffer, the usual minibuffer history commands are available.


22.16 Keyboard Macros

A keyboard macro is a canned sequence of input events that can be considered a command and made the definition of a key. The Lisp representation of a keyboard macro is a string or vector containing the events. Don’t confuse keyboard macros with Lisp macros (see Macros).

Function: execute-kbd-macro kbdmacro &optional count loopfunc

This function executes kbdmacro as a sequence of events. If kbdmacro is a string or vector, then the events in it are executed exactly as if they had been input by the user. The sequence is not expected to be a single key sequence; normally a keyboard macro definition consists of several key sequences concatenated.

If kbdmacro is a symbol, then its function definition is used in place of kbdmacro. If that is another symbol, this process repeats. Eventually the result should be a string or vector. If the result is not a symbol, string, or vector, an error is signaled.

The argument count is a repeat count; kbdmacro is executed that many times. If count is omitted or nil, kbdmacro is executed once. If it is 0, kbdmacro is executed over and over until it encounters an error or a failing search.

If loopfunc is non-nil, it is a function that is called, without arguments, prior to each iteration of the macro. If loopfunc returns nil, then this stops execution of the macro.

See Reading One Event, for an example of using execute-kbd-macro.

Variable: executing-kbd-macro

This variable contains the string or vector that defines the keyboard macro that is currently executing. It is nil if no macro is currently executing. A command can test this variable so as to behave differently when run from an executing macro. Do not set this variable yourself.

Variable: defining-kbd-macro

This variable is non-nil if and only if a keyboard macro is being defined. A command can test this variable so as to behave differently while a macro is being defined. The value is append while appending to the definition of an existing macro. The commands start-kbd-macro, kmacro-start-macro and end-kbd-macro set this variable—do not set it yourself.

The variable is always local to the current terminal and cannot be buffer-local. See Multiple Terminals.

Variable: last-kbd-macro

This variable is the definition of the most recently defined keyboard macro. Its value is a string or vector, or nil.

The variable is always local to the current terminal and cannot be buffer-local. See Multiple Terminals.

Variable: kbd-macro-termination-hook

This normal hook is run when a keyboard macro terminates, regardless of what caused it to terminate (reaching the macro end or an error which ended the macro prematurely).


23 Keymaps

The command bindings of input events are recorded in data structures called keymaps. Each entry in a keymap associates (or binds) an individual event type, either to another keymap or to a command. When an event type is bound to a keymap, that keymap is used to look up the next input event; this continues until a command is found. The whole process is called key lookup.


23.1 Key Sequences

A key sequence, or key for short, is a sequence of one or more input events that form a unit. Input events include characters, function keys, mouse actions, or system events external to Emacs, such as iconify-frame (see Input Events). The Emacs Lisp representation for a key sequence is a string or vector. Unless otherwise stated, any Emacs Lisp function that accepts a key sequence as an argument can handle both representations.

In the string representation, alphanumeric characters ordinarily stand for themselves; for example, "a" represents a and "2" represents 2. Control character events are prefixed by the substring "\C-", and meta characters by "\M-"; for example, "\C-x" represents the key C-x. In addition, the TAB, RET, ESC, and DEL events are represented by "\t", "\r", "\e", and "\d" respectively. The string representation of a complete key sequence is the concatenation of the string representations of the constituent events; thus, "\C-xl" represents the key sequence C-x l.

Key sequences containing function keys, mouse button events, system events, or non-ASCII characters such as C-= or H-a cannot be represented as strings; they have to be represented as vectors.

In the vector representation, each element of the vector represents an input event, in its Lisp form. See Input Events. For example, the vector [?\C-x ?l] represents the key sequence C-x l.

For examples of key sequences written in string and vector representations, Init Rebinding in The GNU Emacs Manual.

Function: kbd keyseq-text

This function converts the text keyseq-text (a string constant) into a key sequence (a string or vector constant). The contents of keyseq-text should use the same syntax as in the buffer invoked by the C-x C-k RET (kmacro-edit-macro) command; in particular, you must surround function key names with ‘<…>’. See Edit Keyboard Macro in The GNU Emacs Manual.

(kbd "C-x") ⇒ "\C-x"
(kbd "C-x C-f") ⇒ "\C-x\C-f"
(kbd "C-x 4 C-f") ⇒ "\C-x4\C-f"
(kbd "X") ⇒ "X"
(kbd "RET") ⇒ "^M"
(kbd "C-c SPC") ⇒ "\C-c "
(kbd "<f1> SPC") ⇒ [f1 32]
(kbd "C-M-<down>") ⇒ [C-M-down]

The kbd function is very permissive, and will try to return something sensible even if the syntax used isn’t completely conforming. To check whether the syntax is actually valid, use the key-valid-p function.


23.2 Keymap Basics

A keymap is a Lisp data structure that specifies key bindings for various key sequences.

A single keymap directly specifies definitions for individual events. When a key sequence consists of a single event, its binding in a keymap is the keymap’s definition for that event. The binding of a longer key sequence is found by an iterative process: first find the definition of the first event (which must itself be a keymap); then find the second event’s definition in that keymap, and so on until all the events in the key sequence have been processed.

If the binding of a key sequence is a keymap, we call the key sequence a prefix key. Otherwise, we call it a complete key (because no more events can be added to it). If the binding is nil, we call the key undefined. Examples of prefix keys are C-c, C-x, and C-x 4. Examples of defined complete keys are X, RET, and C-x 4 C-f. Examples of undefined complete keys are C-x C-g, and C-c 3. See Prefix Keys, for more details.

The rule for finding the binding of a key sequence assumes that the intermediate bindings (found for the events before the last) are all keymaps; if this is not so, the sequence of events does not form a unit—it is not really one key sequence. In other words, removing one or more events from the end of any valid key sequence must always yield a prefix key. For example, C-f C-n is not a key sequence; C-f is not a prefix key, so a longer sequence starting with C-f cannot be a key sequence.

The set of possible multi-event key sequences depends on the bindings for prefix keys; therefore, it can be different for different keymaps, and can change when bindings are changed. However, a one-event sequence is always a key sequence, because it does not depend on any prefix keys for its well-formedness.

At any time, several primary keymaps are active—that is, in use for finding key bindings. These are the global map, which is shared by all buffers; the local keymap, which is usually associated with a specific major mode; and zero or more minor mode keymaps, which belong to currently enabled minor modes. (Not all minor modes have keymaps.) The local keymap bindings shadow (i.e., take precedence over) the corresponding global bindings. The minor mode keymaps shadow both local and global keymaps. See Active Keymaps, for details.


23.3 Format of Keymaps

Each keymap is a list whose CAR is the symbol keymap. The remaining elements of the list define the key bindings of the keymap. A symbol whose function definition is a keymap is also a keymap. Use the function keymapp (see below) to test whether an object is a keymap.

Several kinds of elements may appear in a keymap, after the symbol keymap that begins it:

(type . binding)

This specifies one binding, for events of type type. Each ordinary binding applies to events of a particular event type, which is always a character or a symbol. See Classifying Events. In this kind of binding, binding is a command.

(type item-name . binding)

This specifies a binding which is also a simple menu item that displays as item-name in the menu. See Simple Menu Items.

(type item-name help-string . binding)

This is a simple menu item with help string help-string.

(type menu-item . details)

This specifies a binding which is also an extended menu item. This allows use of other features. See Extended Menu Items.

(t . binding)

This specifies a default key binding; any event not bound by other elements of the keymap is given binding as its binding. Default bindings allow a keymap to bind all possible event types without having to enumerate all of them. A keymap that has a default binding completely masks any lower-precedence keymap, except for events explicitly bound to nil (see below).

char-table

If an element of a keymap is a char-table, it counts as holding bindings for all character events with no modifier bits (see modifier bits): the element whose index is c is the binding for the character c. This is a compact way to record lots of bindings. A keymap with such a char-table is called a full keymap. Other keymaps are called sparse keymaps.

vector

This kind of element is similar to a char-table: the element whose index is c is the binding for the character c. Since the range of characters that can be bound this way is limited by the vector size, and vector creation allocates space for all character codes from 0 up, this format should not be used except for creating menu keymaps (see Menu Keymaps), where the bindings themselves don’t matter.

string

Aside from elements that specify bindings for keys, a keymap can also have a string as an element. This is called the overall prompt string and makes it possible to use the keymap as a menu. See Defining Menus.

(keymap …)

If an element of a keymap is itself a keymap, it counts as if this inner keymap were inlined in the outer keymap. This is used for multiple-inheritance, such as in make-composed-keymap.

When the binding is nil, it doesn’t constitute a definition but it does take precedence over a default binding or a binding in the parent keymap. On the other hand, a binding of nil does not override lower-precedence keymaps; thus, if the local map gives a binding of nil, Emacs uses the binding from the global map.

Keymaps do not directly record bindings for the meta characters. Instead, meta characters are regarded for purposes of key lookup as sequences of two characters, the first of which is ESC (or whatever is currently the value of meta-prefix-char). Thus, the key M-a is internally represented as ESC a, and its global binding is found at the slot for a in esc-map (see Prefix Keys).

This conversion applies only to characters, not to function keys or other input events; thus, M-end has nothing to do with ESC end.

Here as an example is the local keymap for Lisp mode, a sparse keymap. It defines bindings for DEL, C-c C-z, C-M-q, and C-M-x (the actual value also contains a menu binding, which is omitted here for the sake of brevity).

lisp-mode-map
⇒
(keymap
 (3 keymap
    ;; C-c C-z
    (26 . run-lisp))
 (27 keymap
     ;; C-M-x, treated as ESC C-x
     (24 . lisp-send-defun))
 ;; This part is inherited from lisp-mode-shared-map.
 keymap
 ;; DEL
 (127 . backward-delete-char-untabify)
 (27 keymap
     ;; C-M-q, treated as ESC C-q
     (17 . indent-sexp)))
Function: keymapp object

This function returns t if object is a keymap, nil otherwise. More precisely, this function tests for a list whose CAR is keymap, or for a symbol whose function definition satisfies keymapp.

(keymapp '(keymap))
    ⇒ t
(fset 'foo '(keymap))
(keymapp 'foo)
    ⇒ t
(keymapp (current-global-map))
    ⇒ t

23.4 Creating Keymaps

Here we describe the functions for creating keymaps.

Function: make-sparse-keymap &optional prompt

This function creates and returns a new sparse keymap with no entries. (A sparse keymap is the kind of keymap you usually want.) The new keymap does not contain a char-table, unlike make-keymap, and does not bind any events.

(make-sparse-keymap)
    ⇒ (keymap)

If you specify prompt, that becomes the overall prompt string for the keymap. You should specify this only for menu keymaps (see Defining Menus). A keymap with an overall prompt string will always present a mouse menu or a keyboard menu if it is active for looking up the next input event. Don’t specify an overall prompt string for the main map of a major or minor mode, because that would cause the command loop to present a keyboard menu every time.

Function: make-keymap &optional prompt

This function creates and returns a new full keymap. That keymap contains a char-table (see Char-Tables) with slots for all characters without modifiers. The new keymap initially binds all these characters to nil, and does not bind any other kind of event. The argument prompt specifies a prompt string, as in make-sparse-keymap.

(make-keymap)
    ⇒ (keymap #^[nil nil keymap nil nil nil …])

A full keymap is more efficient than a sparse keymap when it holds lots of bindings; for just a few, the sparse keymap is better.

Function: define-keymap &key options... &rest pairs...

You can create a keymap with the functions described above, and then use keymap-set (see Changing Key Bindings) to specify key bindings in that map. When writing modes, however, you frequently have to bind a large number of keys at once, and using keymap-set on them all can be tedious and error-prone. Instead you can use define-keymap, which creates a keymap and binds a number of keys. Here’s a very basic example:

(define-keymap
  "n" #'forward-line
  "f" #'previous-line
  "C-c C-c" #'quit-window)

This function creates a new sparse keymap, defines the keystrokes in pairs, and returns the new keymap. It signals an error if there are duplicate key bindings in pairs.

pairs is a list of alternating key bindings and key definitions, as accepted by keymap-set. In addition, the key can be the special symbol :menu, in which case the definition should be a menu definition as accepted by easy-menu-define (see Easy Menu). Here’s a brief example of this usage:

(define-keymap :full t
  "g" #'eww-reload
  :menu '("Eww"
          ["Exit" quit-window t]
          ["Reload" eww-reload t]))

A number of keywords can be used before the key/definition pairs to change features of the new keymap. If any of the feature keywords is missing from the define-keymap call, the default value for that feature is nil. Here’s a list of the available feature keywords:

:full

If non-nil, create a char-table keymap (as from make-keymap) instead of a sparse keymap (as from make-sparse-keymap (see Creating Keymaps). A sparse keymap is the default.

:parent

If non-nil, the value should be a keymap to use as the parent (see Inheritance and Keymaps).

:keymap

If non-nil, the value should be a keymap. Instead of creating a new keymap, the specified keymap is modified instead.

:suppress

If non-nil, the keymap will be suppressed with suppress-keymap (see Changing Key Bindings). By default, digits and the minus sign are exempt from suppressing, but if the value is nodigits, this suppresses digits and minus-sign like it does with other characters.

:name

If non-nil, the value should be a string to use as the menu for the keymap if you use it as a menu with x-popup-menu (see Pop-Up Menus).

:prefix

If non-nil, the value should be a symbol to be used as a prefix command (see Prefix Keys). If this is the case, this symbol is returned by define-keymap instead of the map itself.

Macro: defvar-keymap name &key options... &rest pairs...

By far, the most common thing to do with a keymap is to bind it to a variable. This is what virtually all modes do—a mode called foo almost always has a variable called foo-mode-map.

This macro defines name as a variable, passes options and pairs to define-keymap, and uses the result as the default value for the variable. It signals an error if there are duplicate key bindings in pairs.

options is like the keywords in define-keymap, but there’s an additional :doc keyword that provides the doc string for the defined variable.

Here’s an example:

(defvar-keymap eww-textarea-map
  :parent text-mode-map
  :doc "Keymap for the eww text area."
  "RET" #'forward-line
  "TAB" #'shr-next-link)

Each command in the keymap can be marked as ‘repeatable’, i.e. usable in repeat-mode, by putting a repeat-map property on it, e.g.

(put 'undo 'repeat-map 'undo-repeat-map)

where the value of the property is the map to be used by repeat-mode.

To avoid repetitive invocations of put, defvar-keymap also has a :repeat keyword, which can be used to specify which commands in the keymap are usable by repeat-mode. The following values are available:

t

This means all the commands in the keymap are repeatable, and is the most common usage.

(:enter (commands ...) :exit (commands ...))

This specifies that the commands in the :enter list enter repeat-mode, and the commands in the :exit list exit repeat mode.

If the :enter list is empty, then all commands in the map enter repeat-mode. Specifying one or more commands in this list is useful if there is a command which doesn’t exist in the map being defined, but which should have the repeat-map property.

If the :exit list is empty then no commands in the map exit repeat-mode. Specifying one or more commands in this list is useful if the keymap being defined contains a command that should not have the repeat-map property.

In order to make e.g. u repeat the undo command, the following two stanzas are equivalent:

(defvar-keymap undo-repeat-map
  "u" #'undo)
(put 'undo 'repeat-map 'undo-repeat-map)

and

(defvar-keymap undo-repeat-map
  :repeat t
  "u" #'undo)

The latter is preferred when there are many commands in the map, all of which should be repeatable.

Function: copy-keymap keymap

This function returns a copy of keymap. This is almost never needed. If you want a keymap that’s like another yet with a few changes, you should use map inheritance rather than copying. I.e., something like:

(let ((map (make-sparse-keymap)))
  (set-keymap-parent map <theirmap>)
  (keymap-set map ...)
  ...)

When performing copy-keymap, any keymaps that appear directly as bindings in keymap are also copied recursively, and so on to any number of levels. However, recursive copying does not take place when the definition of a character is a symbol whose function definition is a keymap; the same symbol appears in the new copy.

(setq map (copy-keymap (current-local-map)))
⇒ (keymap
     ;; (This implements meta characters.)
     (27 keymap
         (83 . center-paragraph)
         (115 . center-line))
     (9 . tab-to-tab-stop))

(eq map (current-local-map))
    ⇒ nil
(equal map (current-local-map))
    ⇒ t

23.5 Inheritance and Keymaps

A keymap can inherit the bindings of another keymap, which we call the parent keymap. Such a keymap looks like this:

(keymap elements… . parent-keymap)

The effect is that this keymap inherits all the bindings of parent-keymap, whatever they may be at the time a key is looked up, but can add to them or override them with elements.

If you change the bindings in parent-keymap using keymap-set or other key-binding functions, these changed bindings are visible in the inheriting keymap, unless shadowed by the bindings made by elements. The converse is not true: if you use keymap-set to change bindings in the inheriting keymap, these changes are recorded in elements, but have no effect on parent-keymap.

The proper way to construct a keymap with a parent is to use set-keymap-parent; if you have code that directly constructs a keymap with a parent, please convert the program to use set-keymap-parent instead.

Function: keymap-parent keymap

This returns the parent keymap of keymap. If keymap has no parent, keymap-parent returns nil.

Function: set-keymap-parent keymap parent

This sets the parent keymap of keymap to parent, and returns parent. If parent is nil, this function gives keymap no parent at all.

If keymap has submaps (bindings for prefix keys), they too receive new parent keymaps that reflect what parent specifies for those prefix keys.

Here is an example showing how to make a keymap that inherits from text-mode-map:

(let ((map (make-sparse-keymap)))
  (set-keymap-parent map text-mode-map)
  map)

A non-sparse keymap can have a parent too, but this is not very useful. A non-sparse keymap always specifies something as the binding for every numeric character code without modifier bits, even if it is nil, so these character’s bindings are never inherited from the parent keymap.

Sometimes you want to make a keymap that inherits from more than one map. You can use the function make-composed-keymap for this.

Function: make-composed-keymap maps &optional parent

This function returns a new keymap composed of the existing keymap(s) maps, and optionally inheriting from a parent keymap parent. maps can be a single keymap or a list of more than one. When looking up a key in the resulting new map, Emacs searches in each of the maps in turn, and then in parent, stopping at the first match. A nil binding in any one of maps overrides any binding in parent, but it does not override any non-nil binding in any other of the maps.

For example, here is how Emacs sets the parent of help-mode-map, such that it inherits from both button-buffer-map and special-mode-map:

(defvar-keymap help-mode-map
  :parent (make-composed-keymap button-buffer-map
                                special-mode-map)
  ...)

23.6 Prefix Keys

A prefix key is a key sequence whose binding is a keymap. The keymap defines what to do with key sequences that extend the prefix key. For example, C-x is a prefix key, and it uses a keymap that is also stored in the variable ctl-x-map. This keymap defines bindings for key sequences starting with C-x.

Some of the standard Emacs prefix keys use keymaps that are also found in Lisp variables:

  • esc-map is the global keymap for the ESC prefix key. Thus, the global definitions of all meta characters are actually found here. This map is also the function definition of ESC-prefix.
  • help-map is the global keymap for the C-h prefix key.
  • mode-specific-map is the global keymap for the prefix key C-c. This map is actually global, not mode-specific, but its name provides useful information about C-c in the output of C-h b (display-bindings), since the main use of this prefix key is for mode-specific bindings.
  • ctl-x-map is the global keymap used for the C-x prefix key. This map is found via the function cell of the symbol Control-X-prefix.
  • mule-keymap is the global keymap used for the C-x RET prefix key.
  • ctl-x-4-map is the global keymap used for the C-x 4 prefix key.
  • ctl-x-5-map is the global keymap used for the C-x 5 prefix key.
  • 2C-mode-map is the global keymap used for the C-x 6 prefix key.
  • tab-prefix-map is the global keymap used for the C-x t prefix key.
  • vc-prefix-map is the global keymap used for the C-x v prefix key.
  • goto-map is the global keymap used for the M-g prefix key.
  • search-map is the global keymap used for the M-s prefix key.
  • The other Emacs prefix keys are C-x @, C-x a i, C-x ESC and ESC ESC. They use keymaps that have no special names.

The keymap binding of a prefix key is used for looking up the event that follows the prefix key. (It may instead be a symbol whose function definition is a keymap. The effect is the same, but the symbol serves as a name for the prefix key.) Thus, the binding of C-x is the symbol Control-X-prefix, whose function cell holds the keymap for C-x commands. (The same keymap is also the value of ctl-x-map.)

Prefix key definitions can appear in any active keymap. The definitions of C-c, C-x, C-h and ESC as prefix keys appear in the global map, so these prefix keys are always available. Major and minor modes can redefine a key as a prefix by putting a prefix key definition for it in the local map or the minor mode’s map. See Active Keymaps.

If a key is defined as a prefix in more than one active map, then its various definitions are in effect merged: the commands defined in the minor mode keymaps come first, followed by those in the local map’s prefix definition, and then by those from the global map.

In the following example, we make C-p a prefix key in the local keymap, in such a way that C-p is identical to C-x. Then the binding for C-p C-f is the function find-file, just like C-x C-f. By contrast, the key sequence C-p 9 is not found in any active keymap.

(use-local-map (make-sparse-keymap))
    ⇒ nil
(keymap-local-set "C-p" ctl-x-map)
    ⇒ (keymap #^[nil nil keymap …
(keymap-lookup nil "C-p C-f")
    ⇒ find-file
(keymap-lookup nil "C-p 9")
    ⇒ nil
Function: define-prefix-command symbol &optional mapvar prompt

This function prepares symbol for use as a prefix key’s binding: it creates a sparse keymap and stores it as symbol’s function definition. Subsequently binding a key sequence to symbol will make that key sequence into a prefix key. The return value is symbol.

This function also sets symbol as a variable, with the keymap as its value. But if mapvar is non-nil, it sets mapvar as a variable instead.

If prompt is non-nil, that becomes the overall prompt string for the keymap. The prompt string should be given for menu keymaps (see Defining Menus).


23.7 Active Keymaps

Emacs contains many keymaps, but at any time only a few keymaps are active. When Emacs receives user input, it translates the input event (see Keymaps for Translating Sequences of Events), and looks for a key binding in the active keymaps.

Usually, the active keymaps are: (i) the keymap specified by the keymap property, (ii) the keymaps of enabled minor modes, (iii) the current buffer’s local keymap, and (iv) the global keymap, in that order. Emacs searches for each input key sequence in all these keymaps.

Of these usual keymaps, the highest-precedence one is specified by the keymap text or overlay property at point, if any. (For a mouse input event, Emacs uses the event position instead of point; see Searching the Active Keymaps.)

Next in precedence are keymaps specified by enabled minor modes. These keymaps, if any, are specified by the variables emulation-mode-map-alists, minor-mode-overriding-map-alist, and minor-mode-map-alist. See Controlling the Active Keymaps.

Next in precedence is the buffer’s local keymap, containing key bindings specific to the buffer. The minibuffer also has a local keymap (see Introduction to Minibuffers). If there is a local-map text or overlay property at point, that specifies the local keymap to use, in place of the buffer’s default local keymap.

The local keymap is normally set by the buffer’s major mode, and every buffer with the same major mode shares the same local keymap. Hence, if you call keymap-local-set (see Commands for Binding Keys) to change the local keymap in one buffer, that also affects the local keymaps in other buffers with the same major mode.

Finally, the global keymap contains key bindings that are defined regardless of the current buffer, such as C-f. It is always active, and is bound to the variable global-map.

Apart from the above usual keymaps, Emacs provides special ways for programs to make other keymaps active. Firstly, the variable overriding-local-map specifies a keymap that replaces the usual active keymaps, except for the global keymap. Secondly, the terminal-local variable overriding-terminal-local-map specifies a keymap that takes precedence over all other keymaps (including overriding-local-map); this is normally used for modal/transient key bindings (the function set-transient-map provides a convenient interface for this). See Controlling the Active Keymaps, for details.

Making keymaps active is not the only way to use them. Keymaps are also used in other ways, such as for translating events within read-key-sequence. See Keymaps for Translating Sequences of Events.

See Standard Keymaps, for a list of some standard keymaps.

Function: current-active-maps &optional olp position

This returns the list of active keymaps that would be used by the command loop in the current circumstances to look up a key sequence. Normally it ignores overriding-local-map and overriding-terminal-local-map, but if olp is non-nil then it pays attention to them. position can optionally be either an event position as returned by event-start or a buffer position, and may change the keymaps as described for keymap-lookup (see keymap-lookup).


23.8 Searching the Active Keymaps

Here is a pseudo-Lisp summary of how Emacs searches the active keymaps:

(or (if overriding-terminal-local-map
        (find-in overriding-terminal-local-map))
    (if overriding-local-map
        (find-in overriding-local-map)
      (or (find-in (get-char-property (point) 'keymap))
          (find-in-any emulation-mode-map-alists)
          (find-in-any minor-mode-overriding-map-alist)
          (find-in-any minor-mode-map-alist)
          (if (get-char-property (point) 'local-map)
              (find-in (get-char-property (point) 'local-map))
            (find-in (current-local-map)))))
    (find-in (current-global-map)))

Here, find-in and find-in-any are pseudo functions that search in one keymap and in an alist of keymaps, respectively. Note that the set-transient-map function works by setting overriding-terminal-local-map (see Controlling the Active Keymaps).

In the above pseudo-code, if a key sequence starts with a mouse event (see Mouse Events), that event’s position is used instead of point, and the event’s buffer is used instead of the current buffer. In particular, this affects how the keymap and local-map properties are looked up. If a mouse event occurs on a string embedded with a display, before-string, or after-string property (see Properties with Special Meanings), and the string has a non-nil keymap or local-map property, that overrides the corresponding property in the underlying buffer text (i.e., the property specified by the underlying text is ignored).

When a key binding is found in one of the active keymaps, and that binding is a command, the search is over—the command is executed. However, if the binding is a symbol with a value or a string, Emacs replaces the input key sequences with the variable’s value or the string, and restarts the search of the active keymaps. See Key Lookup.

The command which is finally found might also be remapped. See Remapping Commands.


23.9 Controlling the Active Keymaps

Variable: global-map

This variable contains the default global keymap that maps Emacs keyboard input to commands. The global keymap is normally this keymap. The default global keymap is a full keymap that binds self-insert-command to all of the printing characters.

It is normal practice to change the bindings in the global keymap, but you should not assign this variable any value other than the keymap it starts out with.

Function: current-global-map

This function returns the current global keymap. This is the same as the value of global-map unless you change one or the other. The return value is a reference, not a copy; if you use keymap-set or other functions on it you will alter global bindings.

(current-global-map)
⇒ (keymap [set-mark-command beginning-of-line …
            delete-backward-char])
Function: current-local-map

This function returns the current buffer’s local keymap, or nil if it has none. In the following example, the keymap for the *scratch* buffer (using Lisp Interaction mode) is a sparse keymap in which the entry for ESC, ASCII code 27, is another sparse keymap.

(current-local-map)
⇒ (keymap
    (10 . eval-print-last-sexp)
    (9 . lisp-indent-line)
    (127 . backward-delete-char-untabify)
    (27 keymap
        (24 . eval-defun)
        (17 . indent-sexp)))

current-local-map returns a reference to the local keymap, not a copy of it; if you use keymap-set or other functions on it you will alter local bindings.

Function: current-minor-mode-maps

This function returns a list of the keymaps of currently enabled minor modes.

Function: use-global-map keymap

This function makes keymap the new current global keymap. It returns nil.

It is very unusual to change the global keymap.

Function: use-local-map keymap

This function makes keymap the new local keymap of the current buffer. If keymap is nil, then the buffer has no local keymap. use-local-map returns nil. Most major mode commands use this function.

Variable: minor-mode-map-alist

This variable is an alist describing keymaps that may or may not be active according to the values of certain variables. Its elements look like this:

(variable . keymap)

The keymap keymap is active whenever variable has a non-nil value. Typically variable is the variable that enables or disables a minor mode. See Keymaps and Minor Modes.

Note that elements of minor-mode-map-alist do not have the same structure as elements of minor-mode-alist. The map must be the CDR of the element; a list with the map as the second element will not do. The CDR can be either a keymap (a list) or a symbol whose function definition is a keymap.

When more than one minor mode keymap is active, the earlier one in minor-mode-map-alist takes priority. But you should design minor modes so that they don’t interfere with each other. If you do this properly, the order will not matter.

See Keymaps and Minor Modes, for more information about minor modes. See also minor-mode-key-binding (see Functions for Key Lookup).

Variable: minor-mode-overriding-map-alist

This variable allows major modes to override the key bindings for particular minor modes. The elements of this alist look like the elements of minor-mode-map-alist: (variable . keymap).

If a variable appears as an element of minor-mode-overriding-map-alist, the map specified by that element totally replaces any map specified for the same variable in minor-mode-map-alist.

minor-mode-overriding-map-alist is automatically buffer-local in all buffers.

Variable: overriding-local-map

If non-nil, this variable holds a keymap to use instead of the buffer’s local keymap, any text property or overlay keymaps, and any minor mode keymaps. This keymap, if specified, overrides all other maps that would have been active, except for the current global map.

Variable: overriding-terminal-local-map

If non-nil, this variable holds a keymap to use instead of overriding-local-map, the buffer’s local keymap, text property or overlay keymaps, and all the minor mode keymaps.

This variable is always local to the current terminal and cannot be buffer-local. See Multiple Terminals. It is used to implement incremental search mode.

Variable: overriding-local-map-menu-flag

If this variable is non-nil, the value of overriding-local-map or overriding-terminal-local-map can affect the display of the menu bar. The default value is nil, so those map variables have no effect on the menu bar.

Note that these two map variables do affect the execution of key sequences entered using the menu bar, even if they do not affect the menu bar display. So if a menu bar key sequence comes in, you should clear the variables before looking up and executing that key sequence. Modes that use the variables would typically do this anyway; normally they respond to events that they do not handle by “unreading” them and exiting.

Variable: special-event-map

This variable holds a keymap for special events. If an event type has a binding in this keymap, then it is special, and the binding for the event is run directly by read-event. See Special Events.

Variable: emulation-mode-map-alists

This variable holds a list of keymap alists to use for emulation modes. It is intended for modes or packages using multiple minor-mode keymaps. Each element is a keymap alist which has the same format and meaning as minor-mode-map-alist, or a symbol with a variable binding which is such an alist. The active keymaps in each alist are used before minor-mode-map-alist and minor-mode-overriding-map-alist.

Function: set-transient-map keymap &optional keep-pred on-exit

This function adds keymap as a transient keymap, which takes precedence over other keymaps for one (or more) subsequent keys.

Normally, keymap is used just once, to look up the very next key. If the optional argument keep-pred is t, the map stays active as long as the user types keys defined in keymap; when the user types a key that is not in keymap, the transient keymap is deactivated and normal key lookup continues for that key.

The keep-pred argument can also be a function. In that case, the function is called with no arguments, prior to running each command, while keymap is active; it should return non-nil if keymap should stay active.

The optional argument on-exit, if non-nil, specifies a function that is called, with no arguments, after keymap is deactivated.

The optional argument message specifies the message to display after activating the transient map. If message is a string, it is the format string for the message, and any ‘%k’ specifier in that string is replaced with the list of keys from the transient map. Any other non-nil value of message stands for the default message format ‘Repeat with %k’.

If the optional argument timeout is non-nil, it should be a number that specifies how many seconds of idle time to wait before deactivating keymap. The value of the variable set-transient-map-timeout, if non-nil, overrides the value of this argument.

This function works by adding and removing keymap from the variable overriding-terminal-local-map, which takes precedence over all other active keymaps (see Searching the Active Keymaps).


23.10 Key Lookup

Key lookup is the process of finding the binding of a key sequence from a given keymap. The execution or use of the binding is not part of key lookup.

Key lookup uses just the event type of each event in the key sequence; the rest of the event is ignored. In fact, a key sequence used for key lookup may designate a mouse event with just its types (a symbol) instead of the entire event (a list). See Input Events. Such a key sequence is insufficient for command-execute to run, but it is sufficient for looking up or rebinding a key.

When the key sequence consists of multiple events, key lookup processes the events sequentially: the binding of the first event is found, and must be a keymap; then the second event’s binding is found in that keymap, and so on until all the events in the key sequence are used up. (The binding thus found for the last event may or may not be a keymap.) Thus, the process of key lookup is defined in terms of a simpler process for looking up a single event in a keymap. How that is done depends on the type of object associated with the event in that keymap.

Let’s use the term keymap entry to describe the value found by looking up an event type in a keymap. (This doesn’t include the item string and other extra elements in a keymap element for a menu item, because keymap-lookup and other key lookup functions don’t include them in the returned value.) While any Lisp object may be stored in a keymap as a keymap entry, not all make sense for key lookup. Here is a table of the meaningful types of keymap entries:

nil

nil means that the events used so far in the lookup form an undefined key. When a keymap fails to mention an event type at all, and has no default binding, that is equivalent to a binding of nil for that event type.

command

The events used so far in the lookup form a complete key, and command is its binding. See What Is a Function?.

array

The array (either a string or a vector) is a keyboard macro. The events used so far in the lookup form a complete key, and the array is its binding. See Keyboard Macros, for more information.

keymap

The events used so far in the lookup form a prefix key. The next event of the key sequence is looked up in keymap.

list

The meaning of a list depends on what it contains:

  • If the CAR of list is the symbol keymap, then the list is a keymap, and is treated as a keymap (see above).
  • If the CAR of list is lambda, then the list is a lambda expression. This is presumed to be a function, and is treated as such (see above). In order to execute properly as a key binding, this function must be a command—it must have an interactive specification. See Defining Commands.
symbol

The function definition of symbol is used in place of symbol. If that too is a symbol, then this process is repeated, any number of times. Ultimately this should lead to an object that is a keymap, a command, or a keyboard macro.

Note that keymaps and keyboard macros (strings and vectors) are not valid functions, so a symbol with a keymap, string, or vector as its function definition is invalid as a function. It is, however, valid as a key binding. If the definition is a keyboard macro, then the symbol is also valid as an argument to command-execute (see Interactive Call).

The symbol undefined is worth special mention: it means to treat the key as undefined. Strictly speaking, the key is defined, and its binding is the command undefined; but that command does the same thing that is done automatically for an undefined key: it rings the bell (by calling ding) but does not signal an error.

undefined is used in local keymaps to override a global key binding and make the key undefined locally. A local binding of nil would fail to do this because it would not override the global binding.

anything else

If any other type of object is found, the events used so far in the lookup form a complete key, and the object is its binding, but the binding is not executable as a command.

In short, a keymap entry may be a keymap, a command, a keyboard macro, a symbol that leads to one of them, or nil.


23.11 Functions for Key Lookup

Here are the functions and variables pertaining to key lookup.

Function: keymap-lookup keymap key &optional accept-defaults no-remap position

This function returns the definition of key in keymap. All the other functions described in this chapter that look up keys use keymap-lookup. Here are examples:

(keymap-lookup (current-global-map) "C-x C-f")
    ⇒ find-file
(keymap-lookup (current-global-map) "C-x C-f 1 2 3 4 5")
    ⇒ 2

If the string or vector key is not a valid key sequence according to the prefix keys specified in keymap, it must be too long and have extra events at the end that do not fit into a single key sequence. Then the value is a number, the number of events at the front of key that compose a complete key.

If accept-defaults is non-nil, then keymap-lookup considers default bindings as well as bindings for the specific events in key. Otherwise, keymap-lookup reports only bindings for the specific sequence key, ignoring default bindings except when you explicitly ask about them. (To do this, supply t as an element of key; see Format of Keymaps.)

If key contains a meta character (not a function key), that character is implicitly replaced by a two-character sequence: the value of meta-prefix-char, followed by the corresponding non-meta character. Thus, the first example below is handled by conversion into the second example.

(keymap-lookup (current-global-map) "M-f")
    ⇒ forward-word
(keymap-lookup (current-global-map) "ESC f")
    ⇒ forward-word

The keymap argument can be nil, meaning to look up key in the current keymaps (as returned by current-active-maps, see Active Keymaps); or it can be a keymap or a list of keymaps, meaning to look up key only in the specified keymaps.

Unlike read-key-sequence, this function does not modify the specified events in ways that discard information (see Key Sequence Input). In particular, it does not convert letters to lower case and it does not change drag events to clicks.

Like the normal command loop, keymap-lookup will remap the command resulting from looking up key by looking up the command in the current keymaps. However, if the optional third argument no-remap is non-nil, keymap-lookup returns the command without remapping.

If the optional argument position is non-nil, it specifies a mouse position as returned by event-start and event-end, and the lookup occurs in the keymaps associated with that position, instead of in keymap. position can also be a number or a marker, in which case it is interpreted as a buffer position, and the function uses the keymap properties at that position instead of at point.

Command: undefined

Used in keymaps to undefine keys. It calls ding, but does not cause an error.

Function: keymap-local-lookup key &optional accept-defaults

This function returns the binding for key in the current local keymap, or nil if it is undefined there.

The argument accept-defaults controls checking for default bindings, as in keymap-lookup (above).

Function: keymap-global-lookup key &optional accept-defaults

This function returns the binding for command key in the current global keymap, or nil if it is undefined there.

The argument accept-defaults controls checking for default bindings, as in keymap-lookup (above).

Function: minor-mode-key-binding key &optional accept-defaults

This function returns a list of all the active minor mode bindings of key. More precisely, it returns an alist of pairs (modename . binding), where modename is the variable that enables the minor mode, and binding is key’s binding in that mode. If key has no minor-mode bindings, the value is nil.

If the first binding found is not a prefix definition (a keymap or a symbol defined as a keymap), all subsequent bindings from other minor modes are omitted, since they would be completely shadowed. Similarly, the list omits non-prefix bindings that follow prefix bindings.

The argument accept-defaults controls checking for default bindings, as in keymap-lookup (above).

User Option: meta-prefix-char

This variable is the meta-prefix character code. It is used for translating a meta character to a two-character sequence so it can be looked up in a keymap. For useful results, the value should be a prefix event (see Prefix Keys). The default value is 27, which is the ASCII code for ESC.

As long as the value of meta-prefix-char remains 27, key lookup translates M-b into ESC b, which is normally defined as the backward-word command. However, if you were to set meta-prefix-char to 24, the code for C-x, then Emacs will translate M-b into C-x b, whose standard binding is the switch-to-buffer command. (Don’t actually do this!) Here is an illustration of what would happen:

meta-prefix-char                    ; The default value.
     ⇒ 27
(key-binding "\M-b")
     ⇒ backward-word
?\C-x                               ; The print representation
     ⇒ 24                          ;   of a character.
(setq meta-prefix-char 24)
     ⇒ 24
(key-binding "\M-b")
     ⇒ switch-to-buffer            ; Now, typing M-b is
                                    ;   like typing C-x b.

(setq meta-prefix-char 27)          ; Avoid confusion!
     ⇒ 27                          ; Restore the default value!

This translation of one event into two happens only for characters, not for other kinds of input events. Thus, M-F1, a function key, is not converted into ESC F1.


23.12 Changing Key Bindings

The way to rebind a key is to change its entry in a keymap. If you change a binding in the global keymap, the change is effective in all buffers (though it has no direct effect in buffers that shadow the global binding with a local one). If you change the current buffer’s local map, that usually affects all buffers using the same major mode. The keymap-global-set and keymap-local-set functions are convenient interfaces for these operations (see Commands for Binding Keys). You can also use keymap-set, a more general function; then you must explicitly specify the map to change.

When choosing the key sequences for Lisp programs to rebind, please follow the Emacs conventions for use of various keys (see Key Binding Conventions).

The functions below signal an error if keymap is not a keymap, or if key is not a valid key.

key is a string representing a single key or a series of key strokes, and must satisfy key-valid-p. Key strokes are separated by a single space character.

Each key stroke is either a single character, or the name of an event, surrounded by angle brackets. In addition, any key stroke may be preceded by one or more modifier keys. Finally, a limited number of characters have a special shorthand syntax. Here’s some example key sequences:

f

The key f.

S o m

A three key sequence of the keys S, o and m.

C-c o

A two key sequence of the keys c with the control modifier and then the key o

H-<left>

The key named left with the hyper modifier.

M-RET

The return key with a meta modifier.

C-M-<space>

The space key with both the control and meta modifiers.

The only keys that have a special shorthand syntax are NUL, RET, TAB, LFD, ESC, SPC and DEL.

The modifiers have to be specified in alphabetical order: ‘A-C-H-M-S-s’, which is ‘Alt-Control-Hyper-Meta-Shift-super’.

Function: keymap-set keymap key binding

This function sets the binding for key in keymap. (If key is more than one event long, the change is actually made in another keymap reached from keymap.) The argument binding can be any Lisp object, but only certain types are meaningful. (For a list of meaningful types, see Key Lookup.) The value returned by keymap-set is binding.

If key is <t>, this sets the default binding in keymap. When an event has no binding of its own, the Emacs command loop uses the keymap’s default binding, if there is one.

Every prefix of key must be a prefix key (i.e., bound to a keymap) or undefined; otherwise an error is signaled. If some prefix of key is undefined, then keymap-set defines it as a prefix key so that the rest of key can be defined as specified.

If there was previously no binding for key in keymap, the new binding is added at the beginning of keymap. The order of bindings in a keymap makes no difference for keyboard input, but it does matter for menu keymaps (see Menu Keymaps).

Function: keymap-unset keymap key &optional remove

This function is the inverse of keymap-set, it unsets the binding for key in keymap, which is the same as setting the binding to nil. In order to instead remove the binding completely, specify remove as non-nil. This only makes a difference if keymap has a parent keymap: if you just unset a key in a child map, it will still shadow the same key in the parent keymap; using remove instead will allow the key in the parent keymap to be used.

Note: using keymap-unset with remove non-nil is intended for users to put in their init file; Emacs packages should avoid using it if possible, since they have complete control over their own keymaps anyway, and they should not be altering other packages’ keymaps.

This example creates a sparse keymap and makes a number of bindings in it:

(setq map (make-sparse-keymap))
    ⇒ (keymap)
(keymap-set map "C-f" 'forward-char)
    ⇒ forward-char
map
    ⇒ (keymap (6 . forward-char))

;; Build sparse submap for C-x and bind f in that.
(keymap-set map "C-x f" 'forward-word)
    ⇒ forward-word
map
⇒ (keymap
    (24 keymap                ; C-x
        (102 . forward-word)) ;      f
    (6 . forward-char))       ; C-f

;; Bind C-p to the ctl-x-map.
(keymap-set map "C-p" ctl-x-map)
;; ctl-x-map
⇒ [nil … find-file … backward-kill-sentence]

;; Bind C-f to foo in the ctl-x-map.
(keymap-set map "C-p C-f" 'foo)
⇒ 'foo
map
⇒ (keymap     ; Note foo in ctl-x-map.
    (16 keymap [nil … foo … backward-kill-sentence])
    (24 keymap
        (102 . forward-word))
    (6 . forward-char))

Note that storing a new binding for C-p C-f actually works by changing an entry in ctl-x-map, and this has the effect of changing the bindings of both C-p C-f and C-x C-f in the default global map.

keymap-set is the general work horse for defining a key in a keymap. When writing modes, however, you frequently have to bind a large number of keys at once, and using keymap-set on them all can be tedious and error-prone. Instead you can use define-keymap, which creates a keymap and binds a number of keys. See Creating Keymaps, for details.

The function substitute-key-definition scans a keymap for keys that have a certain binding and rebinds them with a different binding. Another feature which is cleaner and can often produce the same results is to remap one command into another (see Remapping Commands).

Function: substitute-key-definition olddef newdef keymap &optional oldmap

This function replaces olddef with newdef for any keys in keymap that were bound to olddef. In other words, olddef is replaced with newdef wherever it appears. The function returns nil.

For example, this redefines C-x C-f, if you do it in an Emacs with standard bindings:

(substitute-key-definition
 'find-file 'find-file-read-only (current-global-map))

If oldmap is non-nil, that changes the behavior of substitute-key-definition: the bindings in oldmap determine which keys to rebind. The rebindings still happen in keymap, not in oldmap. Thus, you can change one map under the control of the bindings in another. For example,

(substitute-key-definition
  'delete-backward-char 'my-funny-delete
  my-map global-map)

puts the special deletion command in my-map for whichever keys are globally bound to the standard deletion command.

Here is an example showing a keymap before and after substitution:

(setq map (list 'keymap
                (cons ?1 olddef-1)
                (cons ?2 olddef-2)
                (cons ?3 olddef-1)))
⇒ (keymap (49 . olddef-1) (50 . olddef-2) (51 . olddef-1))

(substitute-key-definition 'olddef-1 'newdef map)
⇒ nil
map
⇒ (keymap (49 . newdef) (50 . olddef-2) (51 . newdef))
Function: suppress-keymap keymap &optional nodigits

This function changes the contents of the full keymap keymap by remapping self-insert-command to the command undefined (see Remapping Commands). This has the effect of undefining all printing characters, thus making ordinary insertion of text impossible. suppress-keymap returns nil.

If nodigits is nil, then suppress-keymap defines digits to run digit-argument, and - to run negative-argument. Otherwise it makes them undefined like the rest of the printing characters.

The suppress-keymap function does not make it impossible to modify a buffer, as it does not suppress commands such as yank and quoted-insert. To prevent any modification of a buffer, make it read-only (see Read-Only Buffers).

Since this function modifies keymap, you would normally use it on a newly created keymap. Operating on an existing keymap that is used for some other purpose is likely to cause trouble; for example, suppressing global-map would make it impossible to use most of Emacs.

This function can be used to initialize the local keymap of a major mode for which insertion of text is not desirable. But usually such a mode should be derived from special-mode (see Basic Major Modes); then its keymap will automatically inherit from special-mode-map, which is already suppressed. Here is how special-mode-map is defined:

(defvar special-mode-map
  (let ((map (make-sparse-keymap)))
    (suppress-keymap map)
    (keymap-set map "q" 'quit-window)
    …
    map))

23.13 Low-Level Key Binding

Historically, Emacs has supported a number of different syntaxes for defining keys. The documented way to bind a key today is to use the syntax supported by key-valid-p, which is what all the functions like keymap-set and keymap-lookup supports. This section documents the old-style syntax and interface functions; they should not be used in new code.

define-key (and other low-level functions that are used to rebind keys) understand a number of different syntaxes for the keys.

A vector containing lists of keys.

You can use a list containing modifier names plus one base event (a character or function key name). For example, [(control ?a) (meta b)] is equivalent to C-a M-b and [(hyper control left)] is equivalent to C-H-left.

A string of characters with modifiers

Internally, key sequences are often represented as strings using the special escape sequences for shift, control and meta modifiers (see String Type), but this representation can also be used by users when rebinding keys. A string like "\M-x" is read as containing a single M-x, "\C-f" is read as containing a single C-f, and "\M-\C-x" and "\C-\M-x" are both read as containing a single C-M-x.

A vector of characters and key symbols

This is the other internal representation of key sequences. It supports a fuller range of modifiers than the string representation, and also support function keys. An example is ‘[?\C-\H-x home], which represents the C-H-x home key sequence. See Character Type.

Function: define-key keymap key binding &optional remove

This function is like keymap-set (see Changing Key Bindings, but understands only the legacy key syntaxes.

In addition, this function also has a remove argument. If it is non-nil, the definition will be removed. This is almost the same as setting the definition to nil, but makes a difference if the keymap has a parent, and key is shadowing the same binding in the parent. With remove, subsequent lookups will return the binding in the parent, whereas with a nil definition the lookups will return nil.

Here are other legacy key definition functions and commands, with the equivalent modern function to use instead in new code.

Command: global-set-key key binding

This function sets the binding of key in the current global map to binding. Use keymap-global-set instead.

Command: global-unset-key key

This function removes the binding of key from the current global map. Use keymap-global-unset instead.

Command: local-set-key key binding

This function sets the binding of key in the current local keymap to binding. Use keymap-local-set instead.

Command: local-unset-key key

This function removes the binding of key from the current local map. Use keymap-local-unset instead.

Function: substitute-key-definition olddef newdef keymap &optional oldmap

This function replaces olddef with newdef for any keys in keymap that were bound to olddef. In other words, olddef is replaced with newdef wherever it appears. The function returns nil. Use keymap-substitute instead.

Function: define-key-after map key binding &optional after

Define a binding in map for key, with value binding, just like define-key, but position the binding in map after the binding for the event after. The argument key should be of length one—a vector or string with just one element. But after should be a single event type—a symbol or a character, not a sequence. The new binding goes after the binding for after. If after is t or is omitted, then the new binding goes last, at the end of the keymap. However, new bindings are added before any inherited keymap. Use keymap-set-after instead of this function.

Function: keyboard-translate from to

This function modifies keyboard-translate-table to translate character code from into character code to. It creates the keyboard translate table if necessary. Use key-translate instead.

Function: key-binding key &optional accept-defaults no-remap position

This function returns the binding for key according to the current active keymaps. The result is nil if key is undefined in the keymaps. The argument accept-defaults controls checking for default bindings, as in lookup-key (see Functions for Key Lookup). If no-remap is non-nil, key-binding ignores command remappings (see Remapping Commands) and returns the binding directly specified for key. The optional argument position should be either a buffer position or an event position like the value of event-start; it tells the function to consult the maps determined based on that position.

Emacs signals an error if key is not a string or a vector.

Use keymap-lookup instead of this function.

Function: lookup-key keymap key &optional accept-defaults

This function returns the definition of key in keymap. If the string or vector key is not a valid key sequence according to the prefix keys specified in keymap, it must be too long and have extra events at the end that do not fit into a single key sequence. Then the value is a number, the number of events at the front of key that compose a complete key.

If accept-defaults is non-nil, then lookup-key considers default bindings as well as bindings for the specific events in key. Otherwise, lookup-key reports only bindings for the specific sequence key, ignoring default bindings except when you explicitly ask about them.

Use keymap-lookup instead of this function.

Function: local-key-binding key &optional accept-defaults

This function returns the binding for key in the current local keymap, or nil if it is undefined there.

The argument accept-defaults controls checking for default bindings, as in lookup-key (above).

Function: global-key-binding key &optional accept-defaults

This function returns the binding for command key in the current global keymap, or nil if it is undefined there.

The argument accept-defaults controls checking for default bindings, as in lookup-key (above).

Function: event-convert-list list

This function converts a list of modifier names and a basic event type to an event type which specifies all of them. The basic event type must be the last element of the list. For example,

(event-convert-list '(control ?a))
     ⇒ 1
(event-convert-list '(control meta ?a))
     ⇒ -134217727
(event-convert-list '(control super f1))
     ⇒ C-s-f1

23.14 Remapping Commands

A special kind of key binding can be used to remap one command to another, without having to refer to the key sequence(s) bound to the original command. To use this feature, make a key binding for a key sequence that starts with the dummy event remap, followed by the command name you want to remap; for the binding, specify the new definition (usually a command name, but possibly any other valid definition for a key binding).

For example, suppose My mode provides a special command my-kill-line, which should be invoked instead of kill-line. To establish this, its mode keymap should contain the following remapping:

(keymap-set my-mode-map "<remap> <kill-line>" 'my-kill-line)

Then, whenever my-mode-map is active, if the user types C-k (the default global key sequence for kill-line) Emacs will instead run my-kill-line.

Note that remapping only takes place through active keymaps; for example, putting a remapping in a prefix keymap like ctl-x-map typically has no effect, as such keymaps are not themselves active. In addition, remapping only works through a single level; in the following example,

(keymap-set my-mode-map "<remap> <kill-line>" 'my-kill-line)
(keymap-set my-mode-map "<remap> <my-kill-line>" 'my-other-kill-line)

kill-line is not remapped to my-other-kill-line. Instead, if an ordinary key binding specifies kill-line, it is remapped to my-kill-line; if an ordinary binding specifies my-kill-line, it is remapped to my-other-kill-line.

To undo the remapping of a command, remap it to nil; e.g.,

(keymap-set my-mode-map "<remap> <kill-line>" nil)
Function: command-remapping command &optional position keymaps

This function returns the remapping for command (a symbol), given the current active keymaps. If command is not remapped (which is the usual situation), or not a symbol, the function returns nil. position can optionally specify a buffer position or an event position to determine the keymaps to use, as in key-binding.

If the optional argument keymaps is non-nil, it specifies a list of keymaps to search in. This argument is ignored if position is non-nil.


23.15 Keymaps for Translating Sequences of Events

When the read-key-sequence function reads a key sequence (see Key Sequence Input), it uses translation keymaps to translate certain event sequences into others. The translation keymaps are input-decode-map, local-function-key-map, and key-translation-map (in order of priority).

Translation keymaps have the same structure as other keymaps, but are used differently: they specify translations to make while reading key sequences, rather than bindings for complete key sequences. As each key sequence is read, it is checked against each translation keymap. If one of the translation keymaps binds k to a vector v, then whenever k appears as a sub-sequence anywhere in a key sequence, that sub-sequence is replaced with the events in v.

For example, VT100 terminals send ESC O P when the keypad key PF1 is pressed. On such terminals, Emacs must translate that sequence of events into a single event pf1. This is done by binding ESC O P to [pf1] in input-decode-map. Thus, when you type C-c PF1 on the terminal, the terminal emits the character sequence C-c ESC O P, and read-key-sequence translates this back into C-c PF1 and returns it as the vector [?\C-c pf1].

Translation keymaps take effect only after Emacs has decoded the keyboard input (via the input coding system specified by keyboard-coding-system). See Terminal I/O Encoding.

Variable: input-decode-map

This variable holds a keymap that describes the character sequences sent by function keys on an ordinary character terminal.

The value of input-decode-map is usually set up automatically according to the terminal’s Terminfo or Termcap entry, but sometimes those need help from terminal-specific Lisp files. Emacs comes with terminal-specific files for many common terminals; their main purpose is to make entries in input-decode-map beyond those that can be deduced from Termcap and Terminfo. See Terminal-Specific Initialization.

Variable: local-function-key-map

This variable holds a keymap similar to input-decode-map except that it describes key sequences which should be translated to alternative interpretations that are usually preferred. It applies after input-decode-map and before key-translation-map.

Entries in local-function-key-map are ignored if they conflict with bindings made in the minor mode, local, or global keymaps. I.e., the remapping only applies if the original key sequence would otherwise not have any binding.

local-function-key-map inherits from function-key-map. The latter should only be altered if you want the binding to apply in all terminals, so using the former is almost always preferred.

Variable: key-translation-map

This variable is another keymap used just like input-decode-map to translate input events into other events. It differs from input-decode-map in that it goes to work after local-function-key-map is finished rather than before; it receives the results of translation by local-function-key-map.

Just like input-decode-map, but unlike local-function-key-map, this keymap is applied regardless of whether the input key-sequence has a normal binding. Note however that actual key bindings can have an effect on key-translation-map, even though they are overridden by it. Indeed, actual key bindings override local-function-key-map and thus may alter the key sequence that key-translation-map receives. Clearly, it is better to avoid this type of situation.

The intent of key-translation-map is for users to map one character set to another, including ordinary characters normally bound to self-insert-command.

You can use input-decode-map, local-function-key-map, and key-translation-map for more than simple aliases, by using a function, instead of a key sequence, as the translation of a key. Then this function is called to compute the translation of that key.

The key translation function receives one argument, which is the prompt that was specified in read-key-sequence—or nil if the key sequence is being read by the editor command loop. In most cases you can ignore the prompt value.

If the function reads input itself, it can have the effect of altering the event that follows. For example, here’s how to define C-c h to turn the character that follows into a Hyper character:

(defun hyperify (prompt)
  (let ((e (read-event)))
    (vector (if (numberp e)
                (logior (ash 1 24) e)
              (if (memq 'hyper (event-modifiers e))
                  e
                (add-event-modifier "H-" e))))))

(defun add-event-modifier (string e)
  (let ((symbol (if (symbolp e) e (car e))))
    (setq symbol (intern (concat string
                                 (symbol-name symbol))))
    (if (symbolp e)
        symbol
      (cons symbol (cdr e)))))

(keymap-set local-function-key-map "C-c h" 'hyperify)

23.15.1 Interaction with normal keymaps

The end of a key sequence is detected when that key sequence either is bound to a command, or when Emacs determines that no additional event can lead to a sequence that is bound to a command.

This means that, while input-decode-map and key-translation-map apply regardless of whether the original key sequence would have a binding, the presence of such a binding can still prevent translation from taking place. For example, let us return to our VT100 example above and add a binding for C-c ESC to the global map; now when the user hits C-c PF1 Emacs will fail to decode C-c ESC O P into C-c PF1 because it will stop reading keys right after C-c ESC, leaving O P for later. This is in case the user really hit C-c ESC, in which case Emacs should not sit there waiting for the next key to decide whether the user really pressed ESC or PF1.

For that reason, it is better to avoid binding commands to key sequences where the end of the key sequence is a prefix of a key translation. The main such problematic suffixes/prefixes are ESC, M-O (which is really ESC O) and M-[ (which is really ESC [).


23.16 Commands for Binding Keys

This section describes some convenient interactive interfaces for changing key bindings. They work by calling keymap-set (see Changing Key Bindings). In interactive use, these commands prompt for the argument key and expect the user to type a valid key sequence; they also prompt for the binding of the key sequence, and expect the name of a command (i.e., a symbol that satisfies commandp, see Interactive Call). When called from Lisp, these commands expect key to be a string that satisfies key-valid-p (see Key Sequences), and binding to be any Lisp object that is meaningful in a keymap (see Key Lookup).

People often use keymap-global-set in their init files (see The Init File) for simple customization. For example,

(keymap-global-set "C-x C-\\" 'next-line)

redefines C-x C-\ to move down a line.

(keymap-global-set "M-<mouse-1>" 'mouse-set-point)

redefines the first (leftmost) mouse button, entered with the Meta key, to set point where you click.

Be careful when using non-ASCII text characters in Lisp specifications of keys to bind. If these are read as multibyte text, as they usually will be in a Lisp file (see Loading Non-ASCII Characters), you must type the keys as multibyte too. For instance, if you use this:

(keymap-global-set "ö" 'my-function) ; bind o-umlaut

and your language environment is multibyte Latin-1, these commands actually bind the multibyte character with code 246, not the byte code 246 (M-v) sent by a Latin-1 terminal. In order to use this binding, you need to teach Emacs how to decode the keyboard by using an appropriate input method (see Input Methods in The GNU Emacs Manual).

Command: keymap-global-set key binding

This function sets the binding of key in the current global map to binding.

(keymap-global-set key binding)
≡
(keymap-set (current-global-map) key binding)
Command: keymap-global-unset key

This function removes the binding of key from the current global map.

One use of this function is in preparation for defining a longer key that uses key as a prefix—which would not be allowed if key has a non-prefix binding. For example:

(keymap-global-unset "C-l")
    ⇒ nil
(keymap-global-set "C-l C-l" 'redraw-display)
    ⇒ nil
Command: keymap-local-set key binding

This function sets the binding of key in the current local keymap to binding.

(keymap-local-set key binding)
≡
(keymap-set (current-local-map) key binding)
Command: keymap-local-unset key

This function removes the binding of key from the current local map.


23.17 Scanning Keymaps

This section describes functions used to scan all the current keymaps for the sake of printing help information. To display the bindings in a particular keymap, you can use the describe-keymap command (see Other Help Commands in The GNU Emacs Manual)

Function: accessible-keymaps keymap &optional prefix

This function returns a list of all the keymaps that can be reached (via zero or more prefix keys) from keymap. The value is an association list with elements of the form (key . map), where key is a prefix key whose definition in keymap is map.

The elements of the alist are ordered so that the key increases in length. The first element is always ([] . keymap), because the specified keymap is accessible from itself with a prefix of no events.

If prefix is given, it should be a prefix key sequence; then accessible-keymaps includes only the submaps whose prefixes start with prefix. These elements look just as they do in the value of (accessible-keymaps); the only difference is that some elements are omitted.

In the example below, the returned alist indicates that the key ESC, which is displayed as ‘^[’, is a prefix key whose definition is the sparse keymap (keymap (83 . center-paragraph) (115 . foo)).

(accessible-keymaps (current-local-map))
⇒(([] keymap
      (27 keymap   ; Note this keymap for ESC is repeated below.
          (83 . center-paragraph)
          (115 . center-line))
      (9 . tab-to-tab-stop))

   ("^[" keymap
    (83 . center-paragraph)
    (115 . foo)))

In the following example, C-h is a prefix key that uses a sparse keymap starting with (keymap (118 . describe-variable)…). Another prefix, C-x 4, uses a keymap which is also the value of the variable ctl-x-4-map. The event mode-line is one of several dummy events used as prefixes for mouse actions in special parts of a window.

(accessible-keymaps (current-global-map))
⇒ (([] keymap [set-mark-command beginning-of-line …
                   delete-backward-char])
    ("^H" keymap (118 . describe-variable) …
     (8 . help-for-help))
    ("^X" keymap [x-flush-mouse-queue …
     backward-kill-sentence])
    ("^[" keymap [mark-sexp backward-sexp …
     backward-kill-word])
    ("^X4" keymap (15 . display-buffer) …)
    ([mode-line] keymap
     (S-mouse-2 . mouse-split-window-horizontally) …))

These are not all the keymaps you would see in actuality.

Function: map-keymap function keymap

The function map-keymap calls function once for each binding in keymap. It passes two arguments, the event type and the value of the binding. If keymap has a parent, the parent’s bindings are included as well. This works recursively: if the parent has itself a parent, then the grandparent’s bindings are also included and so on.

This function is the cleanest way to examine all the bindings in a keymap.

Function: where-is-internal command &optional keymap firstonly noindirect no-remap

This function is a subroutine used by the where-is command (see Help in The GNU Emacs Manual). It returns a list of all key sequences (of any length) that are bound to command in a set of keymaps.

The argument command can be any object; it is compared with all keymap entries using eq.

If keymap is nil, then the maps used are the current active keymaps, disregarding overriding-local-map (that is, pretending its value is nil). If keymap is a keymap, then the maps searched are keymap and the global keymap. If keymap is a list of keymaps, only those keymaps are searched.

Usually it’s best to use overriding-local-map as the expression for keymap. Then where-is-internal searches precisely the keymaps that are active. To search only the global map, pass the value (keymap) (an empty keymap) as keymap.

If firstonly is non-ascii, then the value is a single vector representing the first key sequence found, rather than a list of all possible key sequences. If firstonly is t, then the value is the first key sequence, except that key sequences consisting entirely of ASCII characters (or meta variants of ASCII characters) are preferred to all other key sequences and that the return value can never be a menu binding.

If noindirect is non-nil, where-is-internal doesn’t look inside menu-items to find their commands. This makes it possible to search for a menu-item itself.

The fifth argument, no-remap, determines how this function treats command remappings (see Remapping Commands). There are two cases of interest:

If a command other-command is remapped to command:

If no-remap is nil, find the bindings for other-command and treat them as though they are also bindings for command. If no-remap is non-nil, include the vector [remap other-command] in the list of possible key sequences, instead of finding those bindings.

If command is remapped to other-command:

If no-remap is nil, return the bindings for other-command rather than command. If no-remap is non-nil, return the bindings for command, ignoring the fact that it is remapped.

If a command maps to a key binding like [some-event], and some-event has a symbol plist containing a non-nil non-key-event property, then that binding is ignored by where-is-internal.

Command: describe-bindings &optional prefix buffer-or-name

This function creates a listing of all current key bindings, and displays it in a buffer named *Help*. The text is grouped by modes—minor modes first, then the major mode, then global bindings.

If prefix is non-nil, it should be a prefix key; then the listing includes only keys that start with prefix.

When several characters with consecutive ASCII codes have the same definition, they are shown together, as ‘firstchar..lastchar’. In this instance, you need to know the ASCII codes to understand which characters this means. For example, in the default global map, the characters ‘SPC .. ~’ are described by a single line. SPC is ASCII 32, ~ is ASCII 126, and the characters between them include all the normal printing characters, (e.g., letters, digits, punctuation, etc.); all these characters are bound to self-insert-command.

If buffer-or-name is non-nil, it should be a buffer or a buffer name. Then describe-bindings lists that buffer’s bindings, instead of the current buffer’s.


24 Major and Minor Modes

A mode is a set of definitions that customize Emacs behavior in useful ways. There are two varieties of modes: minor modes, which provide features that users can turn on and off while editing; and major modes, which are used for editing or interacting with a particular kind of text. Each buffer has exactly one major mode at a time.

This chapter describes how to write both major and minor modes, how to indicate them in the mode line, and how they run hooks supplied by the user. For related topics such as keymaps and syntax tables, see Keymaps, and Syntax Tables.


24.1 Hooks

A hook is a variable where you can store a function or functions (see What Is a Function?) to be called on a particular occasion by an existing program. Emacs provides hooks for the sake of customization. Most often, hooks are set up in the init file (see The Init File), but Lisp programs can set them also. See Standard Hooks, for a list of some standard hook variables.

Most of the hooks in Emacs are normal hooks. These variables contain lists of functions to be called with no arguments. By convention, whenever the hook name ends in ‘-hook’, that tells you it is normal. We try to make all hooks normal, as much as possible, so that you can use them in a uniform way.

Every major mode command is supposed to run a normal hook called the mode hook as one of the last steps of initialization. This makes it easy for a user to customize the behavior of the mode, by overriding the buffer-local variable assignments already made by the mode. Most minor mode functions also run a mode hook at the end. But hooks are used in other contexts too. For example, the hook suspend-hook runs just before Emacs suspends itself (see Suspending Emacs).

If the hook variable’s name does not end with ‘-hook’, that indicates it is probably an abnormal hook. These differ from normal hooks in two ways: they can be called with one or more arguments, and their return values can be used in some way. The hook’s documentation says how the functions are called and how their return values are used. Any functions added to an abnormal hook must follow the hook’s calling convention. By convention, abnormal hook names end in ‘-functions’.

If the name of the variable ends in ‘-predicate’ or ‘-function’ (singular) then its value must be a function, not a list of functions. As with abnormal hooks, the expected arguments and meaning of the return value vary across such single function hooks. The details are explained in each variable’s docstring.

Since hooks (both multi and single function) are variables, their values can be modified with setq or temporarily with let. However, it is often useful to add or remove a particular function from a hook while preserving any other functions it might have. For multi function hooks, the recommended way of doing this is with add-hook and remove-hook (see Setting Hooks). Most normal hook variables are initially void; add-hook knows how to deal with this. You can add hooks either globally or buffer-locally with add-hook. For hooks which hold only a single function, add-hook is not appropriate, but you can use add-function (see Advising Emacs Lisp Functions) to combine new functions with the hook. Note that some single function hooks may be nil which add-function cannot deal with, so you must check for that before calling add-function.


24.1.1 Running Hooks

In this section, we document the run-hooks function, which is used to run a normal hook. We also document the functions for running various kinds of abnormal hooks.

Function: run-hooks &rest hookvars

This function takes one or more normal hook variable names as arguments, and runs each hook in turn. Each argument should be a symbol that is a normal hook variable. These arguments are processed in the order specified.

If a hook variable has a non-nil value, that value should be a list of functions. run-hooks calls all the functions, one by one, with no arguments.

The hook variable’s value can also be a single function—either a lambda expression or a symbol with a function definition—which run-hooks calls. But this usage is obsolete.

If the hook variable is buffer-local, the buffer-local variable will be used instead of the global variable. However, if the buffer-local variable contains the element t, the global hook variable will be run as well.

Function: run-hook-with-args hook &rest args

This function runs an abnormal hook by calling all the hook functions in hook, passing each one the arguments args.

Function: run-hook-with-args-until-failure hook &rest args

This function runs an abnormal hook by calling each hook function in turn, stopping if one of them fails by returning nil. Each hook function is passed the arguments args. If this function stops because one of the hook functions fails, it returns nil; otherwise it returns a non-nil value.

Function: run-hook-with-args-until-success hook &rest args

This function runs an abnormal hook by calling each hook function, stopping if one of them succeeds by returning a non-nil value. Each hook function is passed the arguments args. If this function stops because one of the hook functions returns a non-nil value, it returns that value; otherwise it returns nil.


24.1.2 Setting Hooks

Here’s an example that adds a function to a mode hook to turn on Auto Fill mode when in Lisp Interaction mode:

(add-hook 'lisp-interaction-mode-hook 'auto-fill-mode)

The value of a hook variable should be a list of functions. You can manipulate that list using the normal Lisp facilities, but the modular way is to use the functions add-hook and remove-hook, defined below. They take care to handle some unusual situations and avoid problems.

It works to put a lambda-expression function on a hook, but we recommend avoiding this because it can lead to confusion. If you add the same lambda-expression a second time but write it slightly differently, you will get two equivalent but distinct functions on the hook. If you then remove one of them, the other will still be on it.

Function: add-hook hook function &optional depth local

This function is the handy way to add function function to hook variable hook. You can use it for abnormal hooks as well as for normal hooks. function can be any Lisp function that can accept the proper number of arguments for hook. For example,

(add-hook 'text-mode-hook 'my-text-hook-function)

adds my-text-hook-function to the hook called text-mode-hook.

If function is already present in hook (comparing using equal), then add-hook does not add it a second time.

If function has a non-nil property permanent-local-hook, then kill-all-local-variables (or changing major modes) won’t delete it from the hook variable’s local value.

For a normal hook, hook functions should be designed so that the order in which they are executed does not matter. Any dependence on the order is asking for trouble. However, the order is predictable: normally, function goes at the front of the hook list, so it is executed first (barring another add-hook call).

In some cases, it is important to control the relative ordering of functions on the hook. The optional argument depth lets you indicate where the function should be inserted in the list: it should then be a number between -100 and 100 where the higher the value, the closer to the end of the list the function should go. The depth defaults to 0 and for backward compatibility when depth is a non-nil symbol it is interpreted as a depth of 90. Furthermore, when depth is strictly greater than 0 the function is added after rather than before functions of the same depth. One should never use a depth of 100 (or -100), because one can never be sure that no other function will ever need to come before (or after) us.

add-hook can handle the cases where hook is void or its value is a single function; it sets or changes the value to a list of functions.

If local is non-nil, that says to add function to the buffer-local hook list instead of to the global hook list. This makes the hook buffer-local and adds t to the buffer-local value. The latter acts as a flag to run the hook functions in the default value as well as in the local value.

Function: remove-hook hook function &optional local

This function removes function from the hook variable hook. It compares function with elements of hook using equal, so it works for both symbols and lambda expressions.

If local is non-nil, that says to remove function from the buffer-local hook list instead of from the global hook list.


24.2 Major Modes

Major modes specialize Emacs for editing or interacting with particular kinds of text. Each buffer has exactly one major mode at a time. Every major mode is associated with a major mode command, whose name should end in ‘-mode’. This command takes care of switching to that mode in the current buffer, by setting various buffer-local variables such as a local keymap. See Major Mode Conventions. Note that unlike minor modes there is no way to “turn off” a major mode, instead the buffer must be switched to a different one. However, you can temporarily suspend a major mode and later restore the suspended mode, see below.

The least specialized major mode is called Fundamental mode, which has no mode-specific definitions or variable settings.

Command: fundamental-mode

This is the major mode command for Fundamental mode. Unlike other mode commands, it does not run any mode hooks (see Major Mode Conventions), since you are not supposed to customize this mode.

Function: major-mode-suspend

This function works like fundamental-mode, in that it kills all buffer-local variables, but it also records the major mode in effect, so that it could subsequently be restored. This function and major-mode-restore (described next) are useful when you need to put a buffer under some specialized mode other than the one Emacs chooses for it automatically (see How Emacs Chooses a Major Mode), but would also like to be able to switch back to the original mode later.

Function: major-mode-restore &optional avoided-modes

This function restores the major mode recorded by major-mode-suspend. If no major mode was recorded, this function calls normal-mode (see normal-mode), but tries to force it not to choose any modes in avoided-modes, if that argument is non-nil.

Function: clean-mode

Changing the major mode clears out most local variables, but it doesn’t remove all artifacts in the buffer (like text properties and overlays). It’s rare to change a buffer from one major mode to another (except from fundamental-mode to everything else), so this is usually not a concern. It can sometimes be convenient (mostly when debugging a problem in a buffer) to do a “full reset” of the buffer, and that’s what the clean-mode major mode offers. It will kill all local variables (even the permanently local ones), and also removes all overlays and text properties.

The easiest way to write a major mode is to use the macro define-derived-mode, which sets up the new mode as a variant of an existing major mode. See Defining Derived Modes. We recommend using define-derived-mode even if the new mode is not an obvious derivative of another mode, as it automatically enforces many coding conventions for you. See Basic Major Modes, for common modes to derive from.

The standard GNU Emacs Lisp directory tree contains the code for several major modes, in files such as text-mode.el, texinfo.el, lisp-mode.el, and rmail.el. You can study these libraries to see how modes are written.

User Option: major-mode

The buffer-local value of this variable holds the symbol for the current major mode. Its default value holds the default major mode for new buffers. The standard default value is fundamental-mode.

If the default value is nil, then whenever Emacs creates a new buffer via a command such as C-x b (switch-to-buffer), the new buffer is put in the major mode of the previously current buffer. As an exception, if the major mode of the previous buffer has a mode-class symbol property with value special, the new buffer is put in Fundamental mode (see Major Mode Conventions).


24.2.1 Major Mode Conventions

The code for every major mode should follow various coding conventions, including conventions for local keymap and syntax table initialization, function and variable names, and hooks.

If you use the define-derived-mode macro, it will take care of many of these conventions automatically. See Defining Derived Modes. Note also that Fundamental mode is an exception to many of these conventions, because it represents the default state of Emacs.

The following list of conventions is only partial. Each major mode should aim for consistency in general with other Emacs major modes, as this makes Emacs as a whole more coherent. It is impossible to list here all the possible points where this issue might come up; if the Emacs developers point out an area where your major mode deviates from the usual conventions, please make it compatible.

  • Define a major mode command whose name ends in ‘-mode’. When called with no arguments, this command should switch to the new mode in the current buffer by setting up the keymap, syntax table, and buffer-local variables in an existing buffer. It should not change the buffer’s contents.
  • Write a documentation string for this command that describes the special commands available in this mode. See Getting Help about a Major Mode.

    The documentation string may include the special documentation substrings, ‘\[command]’, ‘\{keymap}’, and ‘\<keymap>’, which allow the help display to adapt automatically to the user’s own key bindings. See Substituting Key Bindings in Documentation.

  • The major mode command should start by calling kill-all-local-variables. This runs the normal hook change-major-mode-hook, then gets rid of the buffer-local variables of the major mode previously in effect. See Creating and Deleting Buffer-Local Bindings.
  • The major mode command should set the variable major-mode to the major mode command symbol. This is how describe-mode discovers which documentation to print.
  • The major mode command should set the variable mode-name to the “pretty” name of the mode, usually a string (but see The Data Structure of the Mode Line, for other possible forms). The name of the mode appears in the mode line.
  • Calling the major mode command twice in direct succession should not fail and should do the same thing as calling the command only once. In other words, the major mode command should be idempotent.
  • Since all global names are in the same name space, all the global variables, constants, and functions that are part of the mode should have names that start with the major mode name (or with an abbreviation of it if the name is long). See Emacs Lisp Coding Conventions.
  • In a major mode for editing some kind of structured text, such as a programming language, indentation of text according to structure is probably useful. So the mode should set indent-line-function to a suitable function, and probably customize other variables for indentation. See Automatic Indentation of code.
  • The major mode should usually have its own keymap, which is used as the local keymap in all buffers in that mode. The major mode command should call use-local-map to install this local map. See Active Keymaps, for more information.

    This keymap should be stored permanently in a global variable named modename-mode-map. Normally the library that defines the mode sets this variable.

    See Tips for Defining Variables Robustly, for advice about how to write the code to set up the mode’s keymap variable.

  • The key sequences bound in a major mode keymap should usually start with C-c, followed by a control character, a digit, or {, }, <, >, : or ;. The other punctuation characters are reserved for minor modes, and ordinary letters are reserved for users.

    A major mode can also rebind the keys M-n, M-p and M-s. The bindings for M-n and M-p should normally be some kind of moving forward and backward, but this does not necessarily mean cursor motion.

    It is legitimate for a major mode to rebind a standard key sequence if it provides a command that does the same job in a way better suited to the text this mode is used for. For example, a major mode for editing a programming language might redefine C-M-a to move to the beginning of a function in a way that works better for that language. The recommended way of tailoring C-M-a to the needs of a major mode is to set beginning-of-defun-function (see Moving over Balanced Expressions) to invoke the function specific to the mode.

    It is also legitimate for a major mode to rebind a standard key sequence whose standard meaning is rarely useful in that mode. For instance, minibuffer modes rebind M-r, whose standard meaning is rarely of any use in the minibuffer. Major modes such as Dired or Rmail that do not allow self-insertion of text can reasonably redefine letters and other printing characters as special commands.

  • Major modes for editing text should not define RET to do anything other than insert a newline. However, it is ok for specialized modes for text that users don’t directly edit, such as Dired and Info modes, to redefine RET to do something entirely different.
  • Major modes should not alter options that are primarily a matter of user preference, such as whether Auto-Fill mode is enabled. Leave this to each user to decide. However, a major mode should customize other variables so that Auto-Fill mode will work usefully if the user decides to use it.
  • The mode may have its own syntax table or may share one with other related modes. If it has its own syntax table, it should store this in a variable named modename-mode-syntax-table. See Syntax Tables.
  • If the mode handles a language that has a syntax for comments, it should set the variables that define the comment syntax. See Options Controlling Comments in The GNU Emacs Manual.
  • The mode may have its own abbrev table or may share one with other related modes. If it has its own abbrev table, it should store this in a variable named modename-mode-abbrev-table. If the major mode command defines any abbrevs itself, it should pass t for the system-flag argument to define-abbrev. See Defining Abbrevs.
  • The mode should specify how to do highlighting for Font Lock mode, by setting up a buffer-local value for the variable font-lock-defaults (see Font Lock Mode).
  • Each face that the mode defines should, if possible, inherit from an existing Emacs face. See Basic Faces, and Faces for Font Lock.
  • Consider adding a mode-specific menu to the menu bar. This should preferably include the most important menu-specific settings and commands that will allow users discovering the main features quickly and efficiently.
  • Consider adding mode-specific context menus for the mode, to be used if and when users activate the context-menu-mode (see Menu Mouse Clicks in The Emacs Manual). To this end, define a mode-specific function which builds one or more menus depending on the location of the mouse-3 click in the buffer, and then add that function to the buffer-local value of context-menu-functions.
  • The mode should specify how Imenu should find the definitions or sections of a buffer, by setting up a buffer-local value for the variable imenu-generic-expression, for the two variables imenu-prev-index-position-function and imenu-extract-index-name-function, or for the variable imenu-create-index-function (see Imenu).
  • The mode can tell ElDoc mode how to retrieve different types of documentation for whatever is at point, by adding one or more buffer-local entries to the special hook eldoc-documentation-functions.
  • The mode can specify how to complete various keywords by adding one or more buffer-local entries to the special hook completion-at-point-functions. See Completion in Ordinary Buffers.
  • To make a buffer-local binding for an Emacs customization variable, use make-local-variable in the major mode command, not make-variable-buffer-local. The latter function would make the variable local to every buffer in which it is subsequently set, which would affect buffers that do not use this mode. It is undesirable for a mode to have such global effects. See Buffer-Local Variables.

    With rare exceptions, the only reasonable way to use make-variable-buffer-local in a Lisp package is for a variable which is used only within that package. Using it on a variable used by other packages would interfere with them.

  • Each major mode should have a normal mode hook named modename-mode-hook. The very last thing the major mode command should do is to call run-mode-hooks. This runs the normal hook change-major-mode-after-body-hook, the mode hook, the function hack-local-variables (when the buffer is visiting a file), and then the normal hook after-change-major-mode-hook. See Mode Hooks.
  • The major mode command may start by calling some other major mode command (called the parent mode) and then alter some of its settings. A mode that does this is called a derived mode. The recommended way to define one is to use the define-derived-mode macro, but this is not required. Such a mode should call the parent mode command inside a delay-mode-hooks form. (Using define-derived-mode does this automatically.) See Defining Derived Modes, and Mode Hooks.
  • If something special should be done if the user switches a buffer from this mode to any other major mode, this mode can set up a buffer-local value for change-major-mode-hook (see Creating and Deleting Buffer-Local Bindings).
  • If this mode is appropriate only for specially-prepared text produced by the mode itself (rather than by the user typing at the keyboard or by an external file), then the major mode command symbol should have a property named mode-class with value special, put on as follows:
    (put 'funny-mode 'mode-class 'special)
    

    This tells Emacs that new buffers created while the current buffer is in Funny mode should not be put in Funny mode, even though the default value of major-mode is nil. By default, the value of nil for major-mode means to use the current buffer’s major mode when creating new buffers (see How Emacs Chooses a Major Mode), but with such special modes, Fundamental mode is used instead. Modes such as Dired, Rmail, and Buffer List use this feature.

    The function view-buffer does not enable View mode in buffers whose mode-class is special, because such modes usually provide their own View-like bindings.

    The define-derived-mode macro automatically marks the derived mode as special if the parent mode is special. Special mode is a convenient parent for such modes to inherit from; See Basic Major Modes.

  • If you want to make the new mode the default for files with certain recognizable names, add an element to auto-mode-alist to select the mode for those file names (see How Emacs Chooses a Major Mode). If you define the mode command to autoload, you should add this element in the same file that calls autoload. If you use an autoload cookie for the mode command, you can also use an autoload cookie for the form that adds the element (see autoload cookie). If you do not autoload the mode command, it is sufficient to add the element in the file that contains the mode definition.
  • The top-level forms in the file defining the mode should be written so that they may be evaluated more than once without adverse consequences. For instance, use defvar or defcustom to set mode-related variables, so that they are not reinitialized if they already have a value (see Defining Global Variables).

24.2.2 How Emacs Chooses a Major Mode

When Emacs visits a file, it automatically selects a major mode for the buffer based on information in the file name or in the file itself. It also processes local variables specified in the file text.

Command: normal-mode &optional find-file

This function establishes the proper major mode and buffer-local variable bindings for the current buffer. It calls set-auto-mode (see below). As of Emacs 26.1, it no longer runs hack-local-variables, this now being done in run-mode-hooks at the initialization of major modes (see Mode Hooks).

If the find-file argument to normal-mode is non-nil, normal-mode assumes that the find-file function is calling it. In this case, it may process local variables in the ‘-*-’ line or at the end of the file. The variable enable-local-variables controls whether to do so. See Local Variables in Files in The GNU Emacs Manual, for the syntax of the local variables section of a file.

If you run normal-mode interactively, the argument find-file is normally nil. In this case, normal-mode unconditionally processes any file local variables.

The function calls set-auto-mode to choose and set a major mode. If this does not specify a mode, the buffer stays in the major mode determined by the default value of major-mode (see below).

normal-mode uses condition-case around the call to the major mode command, so errors are caught and reported as a ‘File mode specification error’, followed by the original error message.

Function: set-auto-mode &optional keep-mode-if-same

This function selects and sets the major mode that is appropriate for the current buffer. It bases its decision (in order of precedence) on the ‘-*- line, on any ‘mode:’ local variable near the end of a file, on the ‘#! line (using interpreter-mode-alist), on the text at the beginning of the buffer (using magic-mode-alist), and finally on the visited file name (using auto-mode-alist). See How Major Modes are Chosen in The GNU Emacs Manual. If enable-local-variables is nil, set-auto-mode does not check the ‘-*- line, or near the end of the file, for any mode tag.

There are some file types where it is not appropriate to scan the file contents for a mode specifier. For example, a tar archive may happen to contain, near the end of the file, a member file that has a local variables section specifying a mode for that particular file. This should not be applied to the containing tar file. Similarly, a tiff image file might just happen to contain a first line that seems to match the ‘-*- pattern. For these reasons, both these file extensions are members of the list inhibit-local-variables-regexps. Add patterns to this list to prevent Emacs searching them for local variables of any kind (not just mode specifiers).

If keep-mode-if-same is non-nil, this function does not call the mode command if the buffer is already in the proper major mode. For instance, set-visited-file-name sets this to t to avoid killing buffer local variables that the user may have set.

Function: set-buffer-major-mode buffer

This function sets the major mode of buffer to the default value of major-mode; if that is nil, it uses the current buffer’s major mode (if that is suitable). As an exception, if buffer’s name is *scratch*, it sets the mode to initial-major-mode.

The low-level primitives for creating buffers do not use this function, but medium-level commands such as switch-to-buffer and find-file-noselect use it whenever they create buffers.

User Option: initial-major-mode

The value of this variable determines the major mode of the initial *scratch* buffer. The value should be a symbol that is a major mode command. The default value is lisp-interaction-mode.

Variable: interpreter-mode-alist

This variable specifies major modes to use for scripts that specify a command interpreter in a ‘#!’ line. Its value is an alist with elements of the form (regexp . mode); this says to use mode mode if the file specifies an interpreter which matches \\`regexp\\'. For example, one of the default elements is ("python[0-9.]*" . python-mode).

Variable: magic-mode-alist

This variable’s value is an alist with elements of the form (regexp . function), where regexp is a regular expression and function is a function or nil. After visiting a file, set-auto-mode calls function if the text at the beginning of the buffer matches regexp and function is non-nil; if function is nil, auto-mode-alist gets to decide the mode.

Variable: magic-fallback-mode-alist

This works like magic-mode-alist, except that it is handled only if auto-mode-alist does not specify a mode for this file.

Variable: auto-mode-alist

This variable contains an association list of file name patterns (regular expressions) and corresponding major mode commands. Usually, the file name patterns test for suffixes, such as ‘.el’ and ‘.c’, but this need not be the case. An ordinary element of the alist looks like (regexp . mode-function).

For example,

(("\\`/tmp/fol/" . text-mode)
 ("\\.texinfo\\'" . texinfo-mode)
 ("\\.texi\\'" . texinfo-mode)
 ("\\.el\\'" . emacs-lisp-mode)
 ("\\.c\\'" . c-mode)
 ("\\.h\\'" . c-mode)
 …)

When you visit a file whose expanded file name (see Functions that Expand Filenames), with version numbers and backup suffixes removed using file-name-sans-versions (see File Name Components), matches a regexp, set-auto-mode calls the corresponding mode-function. This feature enables Emacs to select the proper major mode for most files.

If an element of auto-mode-alist has the form (regexp function t), then after calling function, Emacs searches auto-mode-alist again for a match against the portion of the file name that did not match before. This feature is useful for uncompression packages: an entry of the form ("\\.gz\\'" function t) can uncompress the file and then put the uncompressed file in the proper mode according to the name sans ‘.gz’.

If auto-mode-alist has more than one element whose regexp matches the file name, Emacs will use the first match.

Here is an example of how to prepend several pattern pairs to auto-mode-alist. (You might use this sort of expression in your init file.)

(setq auto-mode-alist
  (append
   ;; File name (within directory) starts with a dot.
   '(("/\\.[^/]*\\'" . fundamental-mode)
     ;; File name has no dot.
     ("/[^\\./]*\\'" . fundamental-mode)
     ;; File name ends in ‘.C’.
     ("\\.C\\'" . c++-mode))
   auto-mode-alist))

24.2.3 Getting Help about a Major Mode

The describe-mode function provides information about major modes. It is normally bound to C-h m. It uses the value of the variable major-mode (see Major Modes), which is why every major mode command needs to set that variable.

Command: describe-mode &optional buffer

This command displays the documentation of the current buffer’s major mode and minor modes. It uses the documentation function to retrieve the documentation strings of the major and minor mode commands (see Access to Documentation Strings).

If called from Lisp with a non-nil buffer argument, this function displays the documentation for that buffer’s major and minor modes, rather than those of the current buffer.


24.2.4 Defining Derived Modes

The recommended way to define a new major mode is to derive it from an existing one using define-derived-mode. If there is no closely related mode, you should inherit from either text-mode, special-mode, or prog-mode. See Basic Major Modes. If none of these are suitable, you can inherit from fundamental-mode (see Major Modes).

Macro: define-derived-mode variant parent name docstring keyword-args… body…

This macro defines variant as a major mode command, using name as the string form of the mode name. variant and parent should be unquoted symbols.

The new command variant is defined to call the function parent, then override certain aspects of that parent mode:

  • The new mode has its own sparse keymap, named variant-map. define-derived-mode makes the parent mode’s keymap the parent of the new map, unless variant-map is already set and already has a parent.
  • The new mode has its own syntax table, kept in the variable variant-syntax-table, unless you override this using the :syntax-table keyword (see below). define-derived-mode makes the parent mode’s syntax-table the parent of variant-syntax-table, unless the latter is already set and already has a parent different from the standard syntax table.
  • The new mode has its own abbrev table, kept in the variable variant-abbrev-table, unless you override this using the :abbrev-table keyword (see below).
  • The new mode has its own mode hook, variant-hook. It runs this hook, after running the hooks of its ancestor modes, with run-mode-hooks, as the last thing it does, apart from running any :after-hook form it may have. See Mode Hooks.

In addition, you can specify how to override other aspects of parent with body. The command variant evaluates the forms in body after setting up all its usual overrides, just before running the mode hooks.

If parent has a non-nil mode-class symbol property, then define-derived-mode sets the mode-class property of variant to the same value. This ensures, for example, that if parent is a special mode, then variant is also a special mode (see Major Mode Conventions).

You can also specify nil for parent. This gives the new mode no parent. Then define-derived-mode behaves as described above, but, of course, omits all actions connected with parent.

The argument docstring specifies the documentation string for the new mode. define-derived-mode adds some general information about the mode’s hook, followed by the mode’s keymap, at the end of this documentation string. If you omit docstring, define-derived-mode generates a documentation string.

The keyword-args are pairs of keywords and values. The values, except for :after-hook’s, are evaluated. The following keywords are currently supported:

:syntax-table

You can use this to explicitly specify a syntax table for the new mode. If you specify a nil value, the new mode uses the same syntax table as parent, or the standard syntax table if parent is nil. (Note that this does not follow the convention used for non-keyword arguments that a nil value is equivalent with not specifying the argument.)

:abbrev-table

You can use this to explicitly specify an abbrev table for the new mode. If you specify a nil value, the new mode uses the same abbrev table as parent, or fundamental-mode-abbrev-table if parent is nil. (Again, a nil value is not equivalent to not specifying this keyword.)

:interactive

Modes are interactive commands by default. If you specify a nil value, the mode defined here won’t be interactive. This is useful for modes that are never meant to be activated by users manually, but are only supposed to be used in some specially-formatted buffer.

:group

If this is specified, the value should be the customization group for this mode. (Not all major modes have one.) The command customize-mode uses this. define-derived-mode does not automatically define the specified customization group.

:after-hook

This optional keyword specifies a single Lisp form to evaluate as the final act of the mode function, after the mode hooks have been run. It should not be quoted. Since the form might be evaluated after the mode function has terminated, it should not access any element of the mode function’s local state. An :after-hook form is useful for setting up aspects of the mode which depend on the user’s settings, which in turn may have been changed in a mode hook.

Here is a hypothetical example:

(defvar-keymap hypertext-mode-map
  "<down-mouse-3>" #'do-hyper-link)

(define-derived-mode hypertext-mode
  text-mode "Hypertext"
  "Major mode for hypertext."
  (setq-local case-fold-search nil))

Do not write an interactive spec in the definition; define-derived-mode does that automatically.

Function: derived-mode-p &rest modes

This function returns non-nil if the current major mode is derived from any of the major modes given by the symbols modes.


24.2.5 Basic Major Modes

Apart from Fundamental mode, there are three major modes that other major modes commonly derive from: Text mode, Prog mode, and Special mode. While Text mode is useful in its own right (e.g., for editing files ending in .txt), Prog mode and Special mode exist mainly to let other modes derive from them.

As far as possible, new major modes should be derived, either directly or indirectly, from one of these three modes. One reason is that this allows users to customize a single mode hook (e.g., prog-mode-hook) for an entire family of relevant modes (e.g., all programming language modes).

Command: text-mode

Text mode is a major mode for editing human languages. It defines the ‘"’ and ‘\’ characters as having punctuation syntax (see Table of Syntax Classes), and binds M-TAB to ispell-complete-word (see Spelling in The GNU Emacs Manual).

An example of a major mode derived from Text mode is HTML mode. See SGML and HTML Modes in The GNU Emacs Manual.

Command: prog-mode

Prog mode is a basic major mode for buffers containing programming language source code. Most of the programming language major modes built into Emacs are derived from it.

Prog mode binds parse-sexp-ignore-comments to t (see Motion Commands Based on Parsing) and bidi-paragraph-direction to left-to-right (see Bidirectional Display).

Command: special-mode

Special mode is a basic major mode for buffers containing text that is produced specially by Emacs, rather than directly from a file. Major modes derived from Special mode are given a mode-class property of special (see Major Mode Conventions).

Special mode sets the buffer to read-only. Its keymap defines several common bindings, including q for quit-window and g for revert-buffer (see Reverting).

An example of a major mode derived from Special mode is Buffer Menu mode, which is used by the *Buffer List* buffer. See Listing Existing Buffers in The GNU Emacs Manual.

In addition, modes for buffers of tabulated data can inherit from Tabulated List mode, which is in turn derived from Special mode. See Tabulated List mode.


24.2.6 Mode Hooks

Every major mode command should finish by running the mode-independent normal hook change-major-mode-after-body-hook, its mode hook, and the normal hook after-change-major-mode-hook. It does this by calling run-mode-hooks. If the major mode is a derived mode, that is if it calls another major mode (the parent mode) in its body, it should do this inside delay-mode-hooks so that the parent won’t run these hooks itself. Instead, the derived mode’s call to run-mode-hooks runs the parent’s mode hook too. See Major Mode Conventions.

Emacs versions before Emacs 22 did not have delay-mode-hooks. Versions before 24 did not have change-major-mode-after-body-hook. When user-implemented major modes do not use run-mode-hooks and have not been updated to use these newer features, they won’t entirely follow these conventions: they may run the parent’s mode hook too early, or fail to run after-change-major-mode-hook. If you encounter such a major mode, please correct it to follow these conventions.

When you define a major mode using define-derived-mode, it automatically makes sure these conventions are followed. If you define a major mode “by hand”, not using define-derived-mode, use the following functions to handle these conventions automatically.

Function: run-mode-hooks &rest hookvars

Major modes should run their mode hook using this function. It is similar to run-hooks (see Hooks), but it also runs change-major-mode-after-body-hook, hack-local-variables (when the buffer is visiting a file) (see File Local Variables), and after-change-major-mode-hook. The last thing it does is to evaluate any :after-hook forms declared by parent modes (see Defining Derived Modes).

When this function is called during the execution of a delay-mode-hooks form, it does not run the hooks or hack-local-variables or evaluate the forms immediately. Instead, it arranges for the next call to run-mode-hooks to run them.

Macro: delay-mode-hooks body…

When one major mode command calls another, it should do so inside of delay-mode-hooks.

This macro executes body, but tells all run-mode-hooks calls during the execution of body to delay running their hooks. The hooks will actually run during the next call to run-mode-hooks after the end of the delay-mode-hooks construct.

Variable: change-major-mode-after-body-hook

This is a normal hook run by run-mode-hooks. It is run before the mode hooks.

Variable: after-change-major-mode-hook

This is a normal hook run by run-mode-hooks. It is run at the very end of every properly-written major mode command.


24.2.7 Tabulated List mode

Tabulated List mode is a major mode for displaying tabulated data, i.e., data consisting of entries, each entry occupying one row of text with its contents divided into columns. Tabulated List mode provides facilities for pretty-printing rows and columns, and sorting the rows according to the values in each column. It is derived from Special mode (see Basic Major Modes).

Tabulated List mode is geared towards displaying text using monospaced fonts, using a single font and text size. If you want to display a table using variable pitch fonts or images, make-vtable can be used instead. vtable also support having more than a single table in a buffer, or having a buffer that contains both a table and additional text in it. See (vtable)Introduction, for more information.

Tabulated List mode is intended to be used as a parent mode by a more specialized major mode. Examples include Process Menu mode (see Process Information) and Package Menu mode (see Package Menu in The GNU Emacs Manual).

Such a derived mode should use define-derived-mode in the usual way, specifying tabulated-list-mode as the second argument (see Defining Derived Modes). The body of the define-derived-mode form should specify the format of the tabulated data, by assigning values to the variables documented below; optionally, it can then call the function tabulated-list-init-header, which will populate a header with the names of the columns.

The derived mode should also define a listing command. This, not the mode command, is what the user calls (e.g., M-x list-processes). The listing command should create or switch to a buffer, turn on the derived mode, specify the tabulated data, and finally call tabulated-list-print to populate the buffer.

User Option: tabulated-list-gui-sort-indicator-asc

This variable specifies the character to be used on GUI frames as an indication that the column is sorted in the ascending order.

Whenever you change the sort direction in Tabulated List buffers, this indicator toggles between ascending (“asc”) and descending (“desc”).

User Option: tabulated-list-gui-sort-indicator-desc

Like tabulated-list-gui-sort-indicator-asc, but used when the column is sorted in the descending order.

User Option: tabulated-list-tty-sort-indicator-asc

Like tabulated-list-gui-sort-indicator-asc, but used for text-mode frames.

User Option: tabulated-list-tty-sort-indicator-desc

Like tabulated-list-tty-sort-indicator-asc, but used when the column is sorted in the descending order.

Variable: tabulated-list-format

This buffer-local variable specifies the format of the Tabulated List data. Its value should be a vector. Each element of the vector represents a data column, and should be a list (name width sort), where

  • name is the column’s name (a string).
  • width is the width to reserve for the column (an integer). This is meaningless for the last column, which runs to the end of each line.
  • sort specifies how to sort entries by the column. If nil, the column cannot be used for sorting. If t, the column is sorted by comparing string values. Otherwise, this should be a predicate function for sort (see Functions that Rearrange Lists), which accepts two arguments with the same form as the elements of tabulated-list-entries (see below).
Variable: tabulated-list-entries

This buffer-local variable specifies the entries displayed in the Tabulated List buffer. Its value should be either a list, or a function.

If the value is a list, each list element corresponds to one entry, and should have the form (id contents), where

  • id is either nil, or a Lisp object that identifies the entry. If the latter, the cursor stays on the same entry when re-sorting entries. Comparison is done with equal.
  • contents is a vector with the same number of elements as tabulated-list-format. Each vector element is either a string, which is inserted into the buffer as-is; an image descriptor, which is used to insert an image (see Image Descriptors); or a list (label . properties), which means to insert a text button by calling insert-text-button with label and properties as arguments (see Making Buttons).

    There should be no newlines in any of these strings.

Otherwise, the value should be a function which returns a list of the above form when called with no arguments.

Variable: tabulated-list-revert-hook

This normal hook is run prior to reverting a Tabulated List buffer. A derived mode can add a function to this hook to recompute tabulated-list-entries.

Variable: tabulated-list-printer

The value of this variable is the function called to insert an entry at point, including its terminating newline. The function should accept two arguments, id and contents, having the same meanings as in tabulated-list-entries. The default value is a function which inserts an entry in a straightforward way; a mode which uses Tabulated List mode in a more complex way can specify another function.

Variable: tabulated-list-sort-key

The value of this variable specifies the current sort key for the Tabulated List buffer. If it is nil, no sorting is done. Otherwise, it should have the form (name . flip), where name is a string matching one of the column names in tabulated-list-format, and flip, if non-nil, means to invert the sort order.

Function: tabulated-list-init-header

This function computes and sets header-line-format for the Tabulated List buffer (see Window Header Lines), and assigns a keymap to the header line to allow sorting entries by clicking on column headers.

Modes derived from Tabulated List mode should call this after setting the above variables (in particular, only after setting tabulated-list-format).

Function: tabulated-list-print &optional remember-pos update

This function populates the current buffer with entries. It should be called by the listing command. It erases the buffer, sorts the entries specified by tabulated-list-entries according to tabulated-list-sort-key, then calls the function specified by tabulated-list-printer to insert each entry.

If the optional argument remember-pos is non-nil, this function looks for the id element on the current line, if any, and tries to move to that entry after all the entries are (re)inserted.

If the optional argument update is non-nil, this function will only erase or add entries that have changed since the last print. This is several times faster if most entries haven’t changed since the last time this function was called. The only difference in outcome is that tags placed via tabulated-list-put-tag will not be removed from entries that haven’t changed (normally all tags are removed).

Function: tabulated-list-delete-entry

This function deletes the entry at point.

It returns a list (id cols), where id is the ID of the deleted entry and cols is a vector of its column descriptors. It moves point to the beginning of the current line. It returns nil if there is no entry at point.

Note that this function only changes the buffer contents; it does not alter tabulated-list-entries.

Function: tabulated-list-get-id &optional pos

This defsubst returns the ID object from tabulated-list-entries (if that is a list) or from the list returned by tabulated-list-entries (if it is a function). If omitted or nil, pos defaults to point.

Function: tabulated-list-get-entry &optional pos

This defsubst returns the entry object from tabulated-list-entries (if that is a list) or from the list returned by tabulated-list-entries (if it is a function). This will be a vector for the ID at pos. If there is no entry at pos, then the function returns nil.

Function: tabulated-list-header-overlay-p &optional POS

This defsubst returns non-nil if there is a fake header at pos. A fake header is used if tabulated-list-use-header-line is nil to put the column names at the beginning of the buffer. If omitted or nil, pos defaults to point-min.

Function: tabulated-list-put-tag tag &optional advance

This function puts tag in the padding area of the current line. The padding area can be empty space at the beginning of the line, the width of which is governed by tabulated-list-padding. tag should be a string, with a length less than or equal to tabulated-list-padding. If advance is non-nil, this function advances point by one line.

Function: tabulated-list-clear-all-tags

This function clears all tags from the padding area in the current buffer.

Function: tabulated-list-set-col col desc &optional change-entry-data

This function changes the tabulated list entry at point, setting col to desc. col is the column number to change, or the name of the column to change. desc is the new column descriptor, which is inserted via tabulated-list-print-col.

If change-entry-data is non-nil, this function modifies the underlying data (usually the column descriptor in the list tabulated-list-entries) by setting the column descriptor of the vector to desc.


24.2.8 Generic Modes

Generic modes are simple major modes with basic support for comment syntax and Font Lock mode. To define a generic mode, use the macro define-generic-mode. See the file generic-x.el for some examples of the use of define-generic-mode.

Macro: define-generic-mode mode comment-list keyword-list font-lock-list auto-mode-list function-list &optional docstring

This macro defines a generic mode command named mode (a symbol, not quoted). The optional argument docstring is the documentation for the mode command. If you do not supply it, define-generic-mode generates one by default.

The argument comment-list is a list in which each element is either a character, a string of one or two characters, or a cons cell. A character or a string is set up in the mode’s syntax table as a comment starter. If the entry is a cons cell, the CAR is set up as a comment starter and the CDR as a comment ender. (Use nil for the latter if you want comments to end at the end of the line.) Note that the syntax table mechanism has limitations about what comment starters and enders are actually possible. See Syntax Tables.

The argument keyword-list is a list of keywords to highlight with font-lock-keyword-face. Each keyword should be a string. Meanwhile, font-lock-list is a list of additional expressions to highlight. Each element of this list should have the same form as an element of font-lock-keywords. See Search-based Fontification.

The argument auto-mode-list is a list of regular expressions to add to the variable auto-mode-alist. They are added by the execution of the define-generic-mode form, not by expanding the macro call.

Finally, function-list is a list of functions for the mode command to call for additional setup. It calls these functions just before it runs the mode hook variable mode-hook.


24.2.9 Major Mode Examples

Text mode is perhaps the simplest mode besides Fundamental mode. Here are excerpts from text-mode.el that illustrate many of the conventions listed above:

;; Create the syntax table for this mode.
(defvar text-mode-syntax-table
  (let ((st (make-syntax-table)))
    (modify-syntax-entry ?\" ".   " st)
    (modify-syntax-entry ?\\ ".   " st)
    ;; Add 'p' so M-c on 'hello' leads to 'Hello', not 'hello'.
    (modify-syntax-entry ?' "w p" st)
    …
    st)
  "Syntax table used while in `text-mode'.")

;; Create the keymap for this mode.
(defvar-keymap text-mode-map
  :doc "Keymap for `text-mode'.
Many other modes, such as `mail-mode' and `outline-mode', inherit all
the commands defined in this map."
  "C-M-i" #'ispell-complete-word)

Here is how the actual mode command is defined now:

(define-derived-mode text-mode nil "Text"
  "Major mode for editing text written for humans to read.
In this mode, paragraphs are delimited only by blank or white lines.
You can thus get the full benefit of adaptive filling
 (see the variable `adaptive-fill-mode').
\\{text-mode-map}
Turning on Text mode runs the normal hook `text-mode-hook'."
  (setq-local text-mode-variant t)
  (setq-local require-final-newline mode-require-final-newline))

The three Lisp modes (Lisp mode, Emacs Lisp mode, and Lisp Interaction mode) have more features than Text mode and the code is correspondingly more complicated. Here are excerpts from lisp-mode.el that illustrate how these modes are written.

Here is how the Lisp mode syntax and abbrev tables are defined:

;; Create mode-specific table variables.
(define-abbrev-table 'lisp-mode-abbrev-table ()
  "Abbrev table for Lisp mode.")

(defvar lisp-mode-syntax-table
  (let ((table (make-syntax-table lisp--mode-syntax-table)))
    (modify-syntax-entry ?\[ "_   " table)
    (modify-syntax-entry ?\] "_   " table)
    (modify-syntax-entry ?# "' 14" table)
    (modify-syntax-entry ?| "\" 23bn" table)
    table)
  "Syntax table used in `lisp-mode'.")

The three modes for Lisp share much of their code. For instance, Lisp mode and Emacs Lisp mode inherit from Lisp Data mode and Lisp Interaction Mode inherits from Emacs Lisp mode.

Amongst other things, Lisp Data mode sets up the comment-start variable to handle Lisp comments:

  (setq-local comment-start ";")
  …

Each of the different Lisp modes has a slightly different keymap. For example, Lisp mode binds C-c C-z to run-lisp, but the other Lisp modes do not. However, all Lisp modes have some commands in common. The following code sets up the common commands:

(defvar-keymap lisp-mode-shared-map
  :parent prog-mode-map
  :doc "Keymap for commands shared by all sorts of Lisp modes."
  "C-M-q" #'indent-sexp
  "DEL" #'backward-delete-char-untabify)

And here is the code to set up the keymap for Lisp mode:

(defvar-keymap lisp-mode-map
  :doc "Keymap for ordinary Lisp mode.
All commands in `lisp-mode-shared-map' are inherited by this map."
  :parent lisp-mode-shared-map
  "C-M-x" #'lisp-eval-defun
  "C-c C-z" #'run-lisp)

Finally, here is the major mode command for Lisp mode:

(define-derived-mode lisp-mode lisp-data-mode "Lisp"
  "Major mode for editing Lisp code for Lisps other than GNU Emacs Lisp.
Commands:
Delete converts tabs to spaces as it moves back.
Blank lines separate paragraphs.  Semicolons start comments.

\\{lisp-mode-map}
Note that `run-lisp' may be used either to start an inferior Lisp job
or to switch back to an existing one."
  (setq-local find-tag-default-function 'lisp-find-tag-default)
  (setq-local comment-start-skip
              "\\(\\(^\\|[^\\\n]\\)\\(\\\\\\\\\\)*\\)\\(;+\\|#|\\) *")
  (setq imenu-case-fold-search t))

24.3 Minor Modes

A minor mode provides optional features that users may enable or disable independently of the choice of major mode. Minor modes can be enabled individually or in combination.

Most minor modes implement features that are independent of the major mode, and can thus be used with most major modes. For example, Auto Fill mode works with any major mode that permits text insertion. A few minor modes, however, are specific to a particular major mode. For example, Diff Auto Refine mode is a minor mode that is intended to be used only with Diff mode.

Ideally, a minor mode should have its desired effect regardless of the other minor modes in effect. It should be possible to activate and deactivate minor modes in any order.

Variable: local-minor-modes

This buffer-local variable lists the currently enabled minor modes in the current buffer, and is a list of symbols.

Variable: global-minor-modes

This variable lists the currently enabled global minor modes, and is a list of symbols.

Variable: minor-mode-list

The value of this variable is a list of all minor mode commands.


24.3.1 Conventions for Writing Minor Modes

There are conventions for writing minor modes just as there are for major modes (see Major Modes). These conventions are described below. The easiest way to follow them is to use the macro define-minor-mode. See Defining Minor Modes.

  • Define a variable whose name ends in ‘-mode’. We call this the mode variable. The minor mode command should set this variable. The value will be nil if the mode is disabled, and non-nil if the mode is enabled. The variable should be buffer-local if the minor mode is buffer-local.

    This variable is used in conjunction with the minor-mode-alist to display the minor mode name in the mode line. It also determines whether the minor mode keymap is active, via minor-mode-map-alist (see Controlling the Active Keymaps). Individual commands or hooks can also check its value.

  • Define a command, called the mode command, whose name is the same as the mode variable. Its job is to set the value of the mode variable, plus anything else that needs to be done to actually enable or disable the mode’s features.

    The mode command should accept one optional argument. If called interactively with no prefix argument, it should toggle the mode (i.e., enable if it is disabled, and disable if it is enabled). If called interactively with a prefix argument, it should enable the mode if the argument is positive and disable it otherwise.

    If the mode command is called from Lisp (i.e., non-interactively), it should enable the mode if the argument is omitted or nil; it should toggle the mode if the argument is the symbol toggle; otherwise it should treat the argument in the same way as for an interactive call with a numeric prefix argument, as described above.

    The following example shows how to implement this behavior (it is similar to the code generated by the define-minor-mode macro):

    (interactive (list (or current-prefix-arg 'toggle)))
    (let ((enable
           (if (eq arg 'toggle)
               (not foo-mode) ; this is the mode’s mode variable
             (> (prefix-numeric-value arg) 0))))
      (if enable
          do-enable
        do-disable))
    

    The reason for this somewhat complex behavior is that it lets users easily toggle the minor mode interactively, and also lets the minor mode be easily enabled in a mode hook, like this:

    (add-hook 'text-mode-hook 'foo-mode)
    

    This behaves correctly whether or not foo-mode was already enabled, since the foo-mode mode command unconditionally enables the minor mode when it is called from Lisp with no argument. Disabling a minor mode in a mode hook is a little uglier:

    (add-hook 'text-mode-hook (lambda () (foo-mode -1)))
    

    However, this is not very commonly done.

    Enabling or disabling a minor mode twice in direct succession should not fail and should do the same thing as enabling or disabling it only once. In other words, the minor mode command should be idempotent.

  • Add an element to minor-mode-alist for each minor mode (see Definition of minor-mode-alist), if you want to indicate the minor mode in the mode line. This element should be a list of the following form:
    (mode-variable string)
    

    Here mode-variable is the variable that controls enabling of the minor mode, and string is a short string, starting with a space, to represent the mode in the mode line. These strings must be short so that there is room for several of them at once.

    When you add an element to minor-mode-alist, use assq to check for an existing element, to avoid duplication. For example:

    (unless (assq 'leif-mode minor-mode-alist)
      (push '(leif-mode " Leif") minor-mode-alist))
    

    or like this, using add-to-list (see Modifying List Variables):

    (add-to-list 'minor-mode-alist '(leif-mode " Leif"))
    

In addition, several major mode conventions (see Major Mode Conventions) apply to minor modes as well: those regarding the names of global symbols, the use of a hook at the end of the initialization function, and the use of keymaps and other tables.

The minor mode should, if possible, support enabling and disabling via Custom (see Customization Settings). To do this, the mode variable should be defined with defcustom, usually with :type 'boolean. If just setting the variable is not sufficient to enable the mode, you should also specify a :set method which enables the mode by invoking the mode command. Note in the variable’s documentation string that setting the variable other than via Custom may not take effect. Also, mark the definition with an autoload cookie (see autoload cookie), and specify a :require so that customizing the variable will load the library that defines the mode. For example:

;;;###autoload
(defcustom msb-mode nil
  "Toggle msb-mode.
Setting this variable directly does not take effect;
use either \\[customize] or the function `msb-mode'."
  :set 'custom-set-minor-mode
  :initialize 'custom-initialize-default
  :version "20.4"
  :type    'boolean
  :group   'msb
  :require 'msb)

24.3.2 Keymaps and Minor Modes

Each minor mode can have its own keymap, which is active when the mode is enabled. To set up a keymap for a minor mode, add an element to the alist minor-mode-map-alist. See Definition of minor-mode-map-alist.

One use of minor mode keymaps is to modify the behavior of certain self-inserting characters so that they do something else as well as self-insert. (Another way to customize self-insert-command is through post-self-insert-hook, see User-Level Insertion Commands. Apart from this, the facilities for customizing self-insert-command are limited to special cases, designed for abbrevs and Auto Fill mode. Do not try substituting your own definition of self-insert-command for the standard one. The editor command loop handles this function specially.)

Minor modes may bind commands to key sequences consisting of C-c followed by a punctuation character. However, sequences consisting of C-c followed by one of {}<>:;, or a control character or digit, are reserved for major modes. Also, C-c letter is reserved for users. See Key Binding Conventions.


24.3.3 Defining Minor Modes

The macro define-minor-mode offers a convenient way of implementing a mode in one self-contained definition.

Macro: define-minor-mode mode doc keyword-args… body…

This macro defines a new minor mode whose name is mode (a symbol). It defines a command named mode to toggle the minor mode, with doc as its documentation string.

The toggle command takes one optional (prefix) argument. If called interactively with no argument it toggles the mode on or off. A positive prefix argument enables the mode, any other prefix argument disables it. From Lisp, an argument of toggle toggles the mode, whereas an omitted or nil argument enables the mode. This makes it easy to enable the minor mode in a major mode hook, for example. If doc is nil, the macro supplies a default documentation string explaining the above.

By default, it also defines a variable named mode, which is set to t or nil by enabling or disabling the mode.

The keyword-args consist of keywords followed by corresponding values. A few keywords have special meanings:

:global global

If non-nil, this specifies that the minor mode should be global rather than buffer-local. It defaults to nil.

One of the effects of making a minor mode global is that the mode variable becomes a customization variable. Toggling it through the Customize interface turns the mode on and off, and its value can be saved for future Emacs sessions (see Saving Customizations in The GNU Emacs Manual. For the saved variable to work, you should ensure that the minor mode function is available each time Emacs starts; usually this is done by marking the define-minor-mode form as autoloaded.

:init-value init-value

This is the value to which the mode variable is initialized. Except in unusual circumstances (see below), this value must be nil.

:lighter lighter

The string lighter says what to display in the mode line when the mode is enabled; if it is nil, the mode is not displayed in the mode line.

:keymap keymap

The optional argument keymap specifies the keymap for the minor mode. If non-nil, it should be a variable name (whose value is a keymap), a keymap, or an alist of the form

(key-sequence . definition)

where each key-sequence and definition are arguments suitable for passing to define-key (see Changing Key Bindings). If keymap is a keymap or an alist, this also defines the variable mode-map.

:variable place

This replaces the default variable mode, used to store the state of the mode. If you specify this, the mode variable is not defined, and any init-value argument is unused. place can be a different named variable (which you must define yourself), or anything that can be used with the setf function (see Generalized Variables). place can also be a cons (get . set), where get is an expression that returns the current state, and set is a function of one argument (a state) which should be assigned to place.

:after-hook after-hook

This defines a single Lisp form which is evaluated after the mode hooks have run. It should not be quoted.

:interactive value

Minor modes are interactive commands by default. If value is nil, this is inhibited. If value is a list of symbols, it’s used to say which major modes this minor mode is useful in.

Any other keyword arguments are passed directly to the defcustom generated for the variable mode. See Defining Customization Variables, for the description of those keywords and their values.

The command named mode first performs the standard actions such as setting the variable named mode and then executes the body forms, if any. It then runs the mode hook variable mode-hook and finishes by evaluating any form in :after-hook. (Note that all of this, including running the hook, is done both when the mode is enabled and disabled.)

The initial value must be nil except in cases where (1) the mode is preloaded in Emacs, or (2) it is painless for loading to enable the mode even though the user did not request it. For instance, if the mode has no effect unless something else is enabled, and will always be loaded by that time, enabling it by default is harmless. But these are unusual circumstances. Normally, the initial value must be nil.

The name easy-mmode-define-minor-mode is an alias for this macro.

Here is an example of using define-minor-mode:

(define-minor-mode hungry-mode
  "Toggle Hungry mode.
Interactively with no argument, this command toggles the mode.
A positive prefix argument enables the mode, any other prefix
argument disables it.  From Lisp, argument omitted or nil enables
the mode, `toggle' toggles the state.

When Hungry mode is enabled, the control delete key
gobbles all preceding whitespace except the last.
See the command \\[hungry-electric-delete]."
 ;; The initial value.
 nil
 ;; The indicator for the mode line.
 " Hungry"
 ;; The minor mode bindings.
 '(([C-backspace] . hungry-electric-delete)))

This defines a minor mode named “Hungry mode”, a command named hungry-mode to toggle it, a variable named hungry-mode which indicates whether the mode is enabled, and a variable named hungry-mode-map which holds the keymap that is active when the mode is enabled. It initializes the keymap with a key binding for C-DEL. There are no body forms—many minor modes don’t need any.

Here’s an equivalent way to write it:

(define-minor-mode hungry-mode
  "Toggle Hungry mode.
...rest of documentation as before..."
 ;; The initial value.
 :init-value nil
 ;; The indicator for the mode line.
 :lighter " Hungry"
 ;; The minor mode bindings.
 :keymap
 '(([C-backspace] . hungry-electric-delete)
   ([C-M-backspace]
    . (lambda ()
        (interactive)
        (hungry-electric-delete t)))))
Macro: define-globalized-minor-mode global-mode mode turn-on keyword-args… body…

This defines a global toggle named global-mode whose meaning is to enable or disable the buffer-local minor mode mode in all (or some; see below) buffers. It also executes the body forms. To turn on the minor mode in a buffer, it uses the function turn-on; to turn off the minor mode, it calls mode with −1 as argument. (The function turn-on is a separate function so it could determine whether to enable the minor mode or not when it is not a priori clear that it should always be enabled.)

Globally enabling the mode also affects buffers subsequently created by visiting files, and buffers that use a major mode other than Fundamental mode; but it does not detect the creation of a new buffer in Fundamental mode.

This macro defines the customization option global-mode (see Customization Settings), which can be toggled via the Customize interface to turn the minor mode on and off. As with define-minor-mode, you should ensure that the define-globalized-minor-mode form is evaluated each time Emacs starts, for example by providing a :require keyword.

Use :group group in keyword-args to specify the custom group for the mode variable of the global minor mode.

By default, the buffer-local minor mode variable that says whether the mode is switched on or off is the same as the name of the mode itself. Use :variable variable if that’s not the case–some minor modes use a different variable to store this state information.

Generally speaking, when you define a globalized minor mode, you should also define a non-globalized version, so that people could use it (or disable it) in individual buffers. This also allows them to disable a globally enabled minor mode in a specific major mode, by using that mode’s hook.

If the macro is given a :predicate keyword, it will create a user option called the same as the global mode variable, but with -modes instead of -mode at the end, i.e. global-modes. This variable will be used in a predicate function that determines whether the minor mode should be activated in a particular major mode, and users can customize the value of the variable to control the modes in which the minor mode will be switched on. Valid values of :predicate (and thus valid values of the user option it creates) include t (use in all major modes), nil (don’t use in any major modes), or a list of mode names, optionally preceded with not (as in (not mode-name …)). These elements can be mixed, as shown in the following examples.

(c-mode (not mail-mode message-mode) text-mode)

This means “use in modes derived from c-mode, and not in modes derived from message-mode or mail-mode, but do use in modes derived from text-mode, and otherwise no other modes”.

((not c-mode) t)

This means “don’t use in modes derived from c-mode, but do use everywhere else”.

(text-mode)

This means “use in modes derived from text-mode, but nowhere else”. (There’s an implicit nil element at the end.)

Macro: buffer-local-set-state variable value...

Minor modes often set buffer-local variables that affect some features in Emacs. When a minor mode is switched off, the mode is expected to restore the previous state of these variables. This convenience macro helps with doing that: It works much like setq-local, but returns an object that can be used to restore these values back to their previous values/states (using the companion function buffer-local-restore-state).


24.4 Mode Line Format

Each Emacs window (aside from minibuffer windows) typically has a mode line at the bottom, which displays status information about the buffer displayed in the window. The mode line contains information about the buffer, such as its name, associated file, depth of recursive editing, and major and minor modes. A window can also have a header line, which is much like the mode line but appears at the top of the window.

This section describes how to control the contents of the mode line and header line. We include it in this chapter because much of the information displayed in the mode line relates to the enabled major and minor modes.


24.4.1 Mode Line Basics

The contents of each mode line are specified by the buffer-local variable mode-line-format (see The Top Level of Mode Line Control). This variable holds a mode line construct: a template that controls what is displayed on the buffer’s mode line. The value of header-line-format specifies the buffer’s header line in the same way. All windows for the same buffer use the same mode-line-format and header-line-format unless a mode-line-format or header-line-format parameter has been specified for that window (see Window Parameters).

For efficiency, Emacs does not continuously recompute each window’s mode line and header line. It does so when circumstances appear to call for it—for instance, if you change the window configuration, switch buffers, narrow or widen the buffer, scroll, or modify the buffer. If you alter any of the variables referenced by mode-line-format or header-line-format (see Variables Used in the Mode Line), or any other data structures that affect how text is displayed (see Emacs Display), you should use the function force-mode-line-update to update the display.

Function: force-mode-line-update &optional all

This function forces Emacs to update the current buffer’s mode line and header line, based on the latest values of all relevant variables, during its next redisplay cycle. If the optional argument all is non-nil, it forces an update for all mode lines and header lines.

This function also forces an update of the menu bar and frame title.

The selected window’s mode line is usually displayed in a different color using the face mode-line-active. Other windows’ mode lines appear in the face mode-line-inactive instead. See Faces.

Function: mode-line-window-selected-p

If you want to have more extensive differences between the mode lines in selected and non-selected windows, you can use this predicate in an :eval construct. For instance, if you want to display the buffer name in bold in selected windows, but in italics in the other windows, you can say something like:

(setq-default
 mode-line-buffer-identification
 '(:eval (propertize "%12b"
		     'face (if (mode-line-window-selected-p)
			       'bold
			     'italic))))

Some modes put a lot of data in the mode line, pushing elements at the end of the mode line off to the right. Emacs can “compress” the mode line if the mode-line-compact variable is non-nil by turning stretches of spaces into a single space. If this variable is long, this is only done when the mode line is wider than the currently selected window. (This computation is approximate, based on the number of characters, and not their displayed width.) This variable can be buffer-local to only compress mode-lines in certain buffers.


24.4.2 The Data Structure of the Mode Line

The mode line contents are controlled by a data structure called a mode line construct, made up of lists, strings, symbols, and numbers kept in buffer-local variables. Each data type has a specific meaning for the mode line appearance, as described below. The same data structure is used for constructing frame titles (see Frame Titles) and header lines (see Window Header Lines).

A mode line construct may be as simple as a fixed string of text, but it usually specifies how to combine fixed strings with variables’ values to construct the text. Many of these variables are themselves defined to have mode line constructs as their values.

Here are the meanings of various data types as mode line constructs:

string

A string as a mode line construct appears verbatim except for %-constructs in it. These stand for substitution of other data; see %-Constructs in the Mode Line.

If parts of the string have face properties, they control display of the text just as they would text in the buffer. Any characters which have no face properties are displayed, by default, in the face mode-line or mode-line-inactive (see Standard Faces in The GNU Emacs Manual). The help-echo and keymap properties in string have special meanings. See Properties in the Mode Line.

symbol

A symbol as a mode line construct stands for its value. The value of symbol is used as a mode line construct, in place of symbol. However, the symbols t and nil are ignored, as is any symbol whose value is void.

There is one exception: if the value of symbol is a string, it is displayed verbatim: the %-constructs are not recognized.

Unless symbol is marked as risky (i.e., it has a non-nil risky-local-variable property), all text properties specified in symbol’s value are ignored. This includes the text properties of strings in symbol’s value, as well as all :eval and :propertize forms in it. (The reason for this is security: non-risky variables could be set automatically from file variables without prompting the user.)

(string rest…)
(list rest…)

A list whose first element is a string or list means to process all the elements recursively and concatenate the results. This is the most common form of mode line construct. (Note that text properties are handled specially (for reasons of efficiency) when displaying strings in the mode line: Only the text property on the first character of the string are considered, and they are then used over the entire string. If you need a string with different text properties, you have to use the special :propertize mode line construct.)

(:eval form)

A list whose first element is the symbol :eval says to evaluate form, and use the result as a string to display. Make sure this evaluation cannot load any files, as doing so could cause infinite recursion.

(:propertize elt props…)

A list whose first element is the symbol :propertize says to process the mode line construct elt recursively, then add the text properties specified by props to the result. The argument props should consist of zero or more pairs text-property value. If elt is or produces a string with text properties, all the characters of that string should have the same properties, or else some of them might be removed by :propertize.

(symbol then else)

A list whose first element is a symbol that is not a keyword specifies a conditional. Its meaning depends on the value of symbol. If symbol has a non-nil value, the second element, then, is processed recursively as a mode line construct. Otherwise, the third element, else, is processed recursively. You may omit else; then the mode line construct displays nothing if the value of symbol is nil or void.

(width rest…)

A list whose first element is an integer specifies truncation or padding of the results of rest. The remaining elements rest are processed recursively as mode line constructs and concatenated together. When width is positive, the result is space filled on the right if its width is less than width. When width is negative, the result is truncated on the right to −width columns if its width exceeds −width.

For example, the usual way to show what percentage of a buffer is above the top of the window is to use a list like this: (-3 "%p").


24.4.3 The Top Level of Mode Line Control

The variable in overall control of the mode line is mode-line-format.

User Option: mode-line-format

The value of this variable is a mode line construct that controls the contents of the mode-line. It is always buffer-local in all buffers.

If you set this variable to nil in a buffer, that buffer does not have a mode line. (A window that is just one line tall also does not display a mode line.)

The default value of mode-line-format is designed to use the values of other variables such as mode-line-position and mode-line-modes (which in turn incorporates the values of the variables mode-name and minor-mode-alist). Very few modes need to alter mode-line-format itself. For most purposes, it is sufficient to alter some of the variables that mode-line-format either directly or indirectly refers to.

If you do alter mode-line-format itself, the new value should use the same variables that appear in the default value (see Variables Used in the Mode Line), rather than duplicating their contents or displaying the information in another fashion. This way, customizations made by the user or by Lisp programs (such as display-time and major modes) via changes to those variables remain effective.

Here is a hypothetical example of a mode-line-format that might be useful for Shell mode (in reality, Shell mode does not set mode-line-format):

(setq mode-line-format
  (list "-"
   'mode-line-mule-info
   'mode-line-modified
   'mode-line-frame-identification
   "%b--"
   ;; Note that this is evaluated while making the list.
   ;; It makes a mode line construct which is just a string.
   (getenv "HOST")
   ":"
   'default-directory
   "   "
   'global-mode-string
   "   %[("
   '(:eval (format-time-string "%F"))
   'mode-line-process
   'minor-mode-alist
   "%n"
   ")%]--"
   '(which-function-mode ("" which-func-format "--"))
   '(line-number-mode "L%l--")
   '(column-number-mode "C%c--")
   '(-3 "%p")))

(The variables line-number-mode, column-number-mode and which-function-mode enable particular minor modes; as usual, these variable names are also the minor mode command names.)


24.4.4 Variables Used in the Mode Line

This section describes variables incorporated by the standard value of mode-line-format into the text of the mode line. There is nothing inherently special about these variables; any other variables could have the same effects on the mode line if the value of mode-line-format is changed to use them. However, various parts of Emacs set these variables on the understanding that they will control parts of the mode line; therefore, practically speaking, it is essential for the mode line to use them. Also see Optional Mode Line in The GNU Emacs Manual.

Variable: mode-line-mule-info

This variable holds the value of the mode line construct that displays information about the language environment, buffer coding system, and current input method. See Non-ASCII Characters.

Variable: mode-line-modified

This variable holds the value of the mode line construct that displays whether the current buffer is modified. Its default value displays ‘**’ if the buffer is modified, ‘--’ if the buffer is not modified, ‘%%’ if the buffer is read only, and ‘%*’ if the buffer is read only and modified.

Changing this variable does not force an update of the mode line.

Variable: mode-line-frame-identification

This variable identifies the current frame. Its default value displays " " if you are using a window system which can show multiple frames, or "-%F " on an ordinary terminal which shows only one frame at a time.

Variable: mode-line-buffer-identification

This variable identifies the buffer being displayed in the window. Its default value displays the buffer name, padded with spaces to at least 12 columns.

Variable: mode-line-position

This variable indicates the position in the buffer. Its default value displays the buffer percentage and, optionally, the buffer size, the line number and the column number.

User Option: mode-line-percent-position

This option is used in mode-line-position. Its value specifies both the buffer percentage to display (one of nil, "%o", "%p", "%P" or "%q", see %-Constructs in the Mode Line) and a width to space-fill or truncate to. You are recommended to set this option with the customize-variable facility.

Variable: vc-mode

The variable vc-mode, buffer-local in each buffer, records whether the buffer’s visited file is maintained with version control, and, if so, which kind. Its value is a string that appears in the mode line, or nil for no version control.

Variable: mode-line-modes

This variable displays the buffer’s major and minor modes. Its default value also displays the recursive editing level, information on the process status, and whether narrowing is in effect.

Variable: mode-line-remote

This variable is used to show whether default-directory for the current buffer is remote.

Variable: mode-line-client

This variable is used to identify emacsclient frames.

The following three variables are used in mode-line-modes:

Variable: mode-name

This buffer-local variable holds the “pretty” name of the current buffer’s major mode. Each major mode should set this variable so that the mode name will appear in the mode line. The value does not have to be a string, but can use any of the data types valid in a mode-line construct (see The Data Structure of the Mode Line). To compute the string that will identify the mode name in the mode line, use format-mode-line (see Emulating Mode Line Formatting).

Variable: mode-line-process

This buffer-local variable contains the mode line information on process status in modes used for communicating with subprocesses. It is displayed immediately following the major mode name, with no intervening space. For example, its value in the *shell* buffer is (":%s"), which allows the shell to display its status along with the major mode as: ‘(Shell:run)’. Normally this variable is nil.

Variable: mode-line-front-space

This variable is displayed at the front of the mode line. By default, this construct is displayed right at the beginning of the mode line, except that if there is a memory-full message, it is displayed first.

Variable: mode-line-end-spaces

This variable is displayed at the end of the mode line.

Variable: mode-line-misc-info

Mode line construct for miscellaneous information. By default, this shows the information specified by global-mode-string.

Variable: mode-line-position-line-format

The format used to display line numbers when line-number-mode (see Optional Mode Line in The GNU Emacs Manual) is switched on. ‘%l’ in the format will be replaced with the line number.

Variable: mode-line-position-column-format

The format used to display column numbers when column-number-mode (see Optional Mode Line in The GNU Emacs Manual) is switched on. ‘%c’ in the format will be replaced with a zero-based column number, and ‘%C’ will be replaced with a one-based column number.

Variable: mode-line-position-column-line-format

The format used to display column numbers when both line-number-mode and column-number-mode are switched on. See the previous two variables for the meaning of the ‘%l’, ‘%c’ and ‘%C’ format specs.

Variable: minor-mode-alist

This variable holds an association list whose elements specify how the mode line should indicate that a minor mode is active. Each element of the minor-mode-alist should be a two-element list:

(minor-mode-variable mode-line-string)

More generally, mode-line-string can be any mode line construct. It appears in the mode line when the value of minor-mode-variable is non-nil, and not otherwise. These strings should begin with spaces so that they don’t run together. Conventionally, the minor-mode-variable for a specific mode is set to a non-nil value when that minor mode is activated.

minor-mode-alist itself is not buffer-local. Each variable mentioned in the alist should be buffer-local if its minor mode can be enabled separately in each buffer.

Variable: global-mode-string

This variable holds a mode line construct that, by default, appears in the mode line as part of mode-line-misc-info, just after the which-function-mode information if that minor mode is enabled, else after mode-line-modes. Elements that are added to this construct should normally end in a space (to ensure that consecutive global-mode-string elements display properly). For instance, the command display-time sets global-mode-string to refer to the variable display-time-string, which holds a string containing the time and load information.

The ‘%M’ construct substitutes the value of global-mode-string. This construct is not used by the default mode line, as the variable itself is used in mode-line-misc-info.

Here is a simplified version of the default value of mode-line-format. The real default value also specifies addition of text properties.

("-"
 mode-line-mule-info
 mode-line-modified
 mode-line-frame-identification
 mode-line-buffer-identification
 "   "
 mode-line-position
 (vc-mode vc-mode)
 "   "
 mode-line-modes
 (which-function-mode ("" which-func-format "--"))
 (global-mode-string ("--" global-mode-string))
 "-%-")

24.4.5 %-Constructs in the Mode Line

Strings used as mode line constructs can use certain %-constructs to substitute various kinds of data. The following is a list of the defined %-constructs, and what they mean.

In any construct except ‘%%’, you can add a decimal integer after the ‘%’ to specify a minimum field width. If the width is less, the field is padded to that width. Purely numeric constructs (‘c’, ‘i’, ‘I’, and ‘l’) are padded by inserting spaces to the left, and others are padded by inserting spaces to the right.

%b

The current buffer name, obtained with the buffer-name function. See Buffer Names.

%c

The current column number of point, counting from zero starting at the left margin of the window.

%C

The current column number of point, counting from one starting at the left margin of the window.

%e

When Emacs is nearly out of memory for Lisp objects, a brief message saying so. Otherwise, this is empty.

%f

The visited file name, obtained with the buffer-file-name function. See Buffer File Name.

%F

The title (only on a window system) or the name of the selected frame. See Basic Parameters.

%i

The size of the accessible part of the current buffer; basically (- (point-max) (point-min)).

%I

Like ‘%i’, but the size is printed in a more readable way by using ‘k’ for 10^3, ‘M’ for 10^6, ‘G’ for 10^9, etc., to abbreviate.

%l

The current line number of point, counting within the accessible portion of the buffer.

%M

The value of global-mode-string (which is part of mode-line-misc-info by default).

%n

Narrow’ when narrowing is in effect; nothing otherwise (see narrow-to-region in Narrowing).

%o

The degree of travel of the window through (the visible portion of) the buffer, i.e. the size of the text above the top of the window expressed as a percentage of all the text outside the window, or ‘Top’, ‘Bottom’ or ‘All’.

%p

The percentage of the buffer text above the top of window, or ‘Top’, ‘Bottom’ or ‘All’. Note that the default mode line construct truncates this to three characters.

%P

The percentage of the buffer text that is above the bottom of the window (which includes the text visible in the window, as well as the text above the top), plus ‘Top’ if the top of the buffer is visible on screen; or ‘Bottom’ or ‘All’.

%q

The percentages of text above both the top and the bottom of the window, separated by ‘-’, or ‘All’.

%s

The status of the subprocess belonging to the current buffer, obtained with process-status. See Process Information.

%z

The mnemonics of keyboard, terminal, and buffer coding systems.

%Z

Like ‘%z’, but including the end-of-line format.

%*

%’ if the buffer is read only (see buffer-read-only);
*’ if the buffer is modified (see buffer-modified-p);
-’ otherwise. See Buffer Modification.

%+

*’ if the buffer is modified (see buffer-modified-p);
%’ if the buffer is read only (see buffer-read-only);
-’ otherwise. This differs from ‘%*’ only for a modified read-only buffer. See Buffer Modification.

%&

*’ if the buffer is modified, and ‘-’ otherwise.

%@

@’ if the buffer’s default-directory (see Functions that Expand Filenames) is on a remote machine, and ‘-’ otherwise.

%[

An indication of the depth of recursive editing levels (not counting minibuffer levels): one ‘[’ for each editing level. See Recursive Editing.

%]

One ‘]’ for each recursive editing level (not counting minibuffer levels).

%-

Dashes sufficient to fill the remainder of the mode line.

%%

The character ‘%’—this is how to include a literal ‘%’ in a string in which %-constructs are allowed.

Obsolete %-Constructs

The following constructs should no longer be used.

%m

Obsolete; use the mode-name variable instead. The %m construct is inadequate, as it produces an empty string if the value of mode-name is a non-string mode-line construct (as in emacs-lisp-mode, for example).


24.4.6 Properties in the Mode Line

Certain text properties are meaningful in the mode line. The face property affects the appearance of text; the help-echo property associates help strings with the text, and keymap can make the text mouse-sensitive.

There are four ways to specify text properties for text in the mode line:

  1. Put a string with a text property directly into the mode line data structure, but see The Data Structure of the Mode Line for caveats for that.
  2. Put a text property on a mode line %-construct such as ‘%12b’; then the expansion of the %-construct will have that same text property.
  3. Use a (:propertize elt props…) construct to give elt a text property specified by props.
  4. Use a list containing :eval form in the mode line data structure, and make form evaluate to a string that has a text property.

You can use the keymap property to specify a keymap. This keymap only takes real effect for mouse clicks; binding character keys and function keys to it has no effect, since it is impossible to move point into the mode line.

When the mode line refers to a variable which does not have a non-nil risky-local-variable property, any text properties given or specified within that variable’s values are ignored. This is because such properties could otherwise specify functions to be called, and those functions could come from file local variables.


24.4.7 Window Header Lines

A window can have a header line at the top, just as it can have a mode line at the bottom. The header line feature works just like the mode line feature, except that it’s controlled by header-line-format:

Variable: header-line-format

This variable, local in every buffer, specifies how to display the header line, for windows displaying the buffer. The format of the value is the same as for mode-line-format (see The Data Structure of the Mode Line). It is normally nil, so that ordinary buffers have no header line.

If display-line-numbers-mode is turned on in a buffer (see display-line-numbers-mode in The GNU Emacs Manual), the buffer text is indented on display by the amount of screen space needed to show the line numbers. By contrast, text of the header line is not automatically indented, because a header line never displays a line number, and because the text of the header line is not necessarily directly related to buffer text below it. If a Lisp program needs the header-line text to be aligned with buffer text (for example, if the buffer displays columnar data, like tabulated-list-mode does, see Tabulated List mode), it should turn on the minor mode header-line-indent-mode.

Command: header-line-indent-mode

This buffer-local minor mode tracks the changes of the width of the line-number display on screen (which may vary depending on the range of line numbers shown in the window), and allows Lisp programs to arrange that header-line text is always aligned with buffer text when the line-number width changes. Such Lisp programs should turn on this mode in the buffer, and use the variables header-line-indent and header-line-indent-width in the header-line-format to ensure it is adjusted to the text indentation at all times.

Variable: header-line-indent

This variable’s value is a whitespace string whose width is kept equal to the current width of line-numbers on display, provided that header-line-indent-mode is turned on in the buffer shown in the window. The number of spaces is calculated under the assumption that the face of the header-line text uses the same font, including size, as the frame’s default font; if that assumption is false, use header-line-indent-width, described below, instead. This variable is intended to be used in simple situations where the header-line text needs to be indented as a whole to be realigned with buffer text, by prepending this variable’s value to the actual header-line text. For example, the following definition of header-line-format:

(setq header-line-format
      `("" header-line-indent ,my-header-line))

where my-header-line is the format string that produces the actual text of the header line, will make sure the header-line text is always indented like the buffer text below it.

Variable: header-line-indent-width

This variable’s value is kept updated to provide the current width, in units of the frame’s canonical character width, used for displaying the line numbers, provided that header-line-indent-mode is turned on in the buffer shown in the window. It can be used for aligning the header-line text with the buffer text when header-line-indent is not flexible enough. For example, if the header line uses a font whose metrics is different from the default face’s font, your Lisp program can calculate the width of line-number display in pixels, by multiplying the value of this variable by the value returned by frame-char-width (see Frame Font), and then use the result to align header-line text using the :align-to display property spec (see Specified Spaces) in pixels on the relevant parts of header-line-format.

Function: window-header-line-height &optional window

This function returns the height in pixels of window’s header line. window must be a live window, and defaults to the selected window.

A window that is just one line tall never displays a header line. A window that is two lines tall cannot display both a mode line and a header line at once; if it has a mode line, then it does not display a header line.


24.4.8 Emulating Mode Line Formatting

You can use the function format-mode-line to compute the text that would appear in a mode line or header line based on a certain mode line construct.

Function: format-mode-line format &optional face window buffer

This function formats a line of text according to format as if it were generating the mode line for window, but it also returns the text as a string. The argument window defaults to the selected window. If buffer is non-nil, all the information used is taken from buffer; by default, it comes from window’s buffer.

The value string normally has text properties that correspond to the faces, keymaps, etc., that the mode line would have. Any character for which no face property is specified by format gets a default value determined by face. If face is t, that stands for either mode-line if window is selected, otherwise mode-line-inactive. If face is nil or omitted, that stands for the default face. If face is an integer, the value returned by this function will have no text properties.

You can also specify other valid faces as the value of face. If specified, that face provides the face property for characters whose face is not specified by format.

Note that using mode-line, mode-line-inactive, or header-line as face will actually redisplay the mode line or the header line, respectively, using the current definitions of the corresponding face, in addition to returning the formatted string. (Other faces do not cause redisplay.)

For example, (format-mode-line header-line-format) returns the text that would appear in the selected window’s header line ("" if it has no header line). (format-mode-line header-line-format 'header-line) returns the same text, with each character carrying the face that it will have in the header line itself, and also redraws the header line.


24.5 Imenu

Imenu is a feature that lets users select a definition or section in the buffer, from a menu which lists all of them, to go directly to that location in the buffer. Imenu works by constructing a buffer index which lists the names and buffer positions of the definitions, or other named portions of the buffer; then the user can choose one of them and move point to it. Major modes can add a menu bar item to use Imenu using imenu-add-to-menubar.

Command: imenu-add-to-menubar name

This function defines a local menu bar item named name to run Imenu.

The user-level commands for using Imenu are described in the Emacs Manual (see Imenu in the Emacs Manual). This section explains how to customize Imenu’s method of finding definitions or buffer portions for a particular major mode.

The usual and simplest way is to set the variable imenu-generic-expression:

Variable: imenu-generic-expression

This variable, if non-nil, is a list that specifies regular expressions for finding definitions for Imenu. Simple elements of imenu-generic-expression look like this:

(menu-title regexp index)

Here, if menu-title is non-nil, it says that the matches for this element should go in a submenu of the buffer index; menu-title itself specifies the name for the submenu. If menu-title is nil, the matches for this element go directly in the top level of the buffer index.

The second item in the list, regexp, is a regular expression (see Regular Expressions); anything in the buffer that it matches is considered a definition, something to mention in the buffer index. The third item, index, is a non-negative integer that indicates which subexpression in regexp matches the definition’s name.

An element can also look like this:

(menu-title regexp index function arguments…)

Each match for this element creates an index item, and when the index item is selected by the user, it calls function with arguments consisting of the item name, the buffer position, and arguments.

For Emacs Lisp mode, imenu-generic-expression could look like this:

((nil "^\\s-*(def\\(un\\|subst\\|macro\\|advice\\)\
\\s-+\\([-A-Za-z0-9+]+\\)" 2)
 ("*Vars*" "^\\s-*(def\\(var\\|const\\)\
\\s-+\\([-A-Za-z0-9+]+\\)" 2)
 ("*Types*"
  "^\\s-*\
(def\\(type\\|struct\\|class\\|ine-condition\\)\
\\s-+\\([-A-Za-z0-9+]+\\)" 2))

Setting this variable makes it buffer-local in the current buffer.

Variable: imenu-case-fold-search

This variable controls whether matching against the regular expressions in the value of imenu-generic-expression is case-sensitive: t, the default, means matching should ignore case.

Setting this variable makes it buffer-local in the current buffer.

Variable: imenu-syntax-alist

This variable is an alist of syntax table modifiers to use while processing imenu-generic-expression, to override the syntax table of the current buffer. Each element should have this form:

(characters . syntax-description)

The CAR, characters, can be either a character or a string. The element says to give that character or characters the syntax specified by syntax-description, which is passed to modify-syntax-entry (see Syntax Table Functions).

This feature is typically used to give word syntax to characters which normally have symbol syntax, and thus to simplify imenu-generic-expression and speed up matching. For example, Fortran mode uses it this way:

(setq imenu-syntax-alist '(("_$" . "w")))

The imenu-generic-expression regular expressions can then use ‘\\sw+’ instead of ‘\\(\\sw\\|\\s_\\)+’. Note that this technique may be inconvenient when the mode needs to limit the initial character of a name to a smaller set of characters than are allowed in the rest of a name.

Setting this variable makes it buffer-local in the current buffer.

Another way to customize Imenu for a major mode is to set the variables imenu-prev-index-position-function and imenu-extract-index-name-function:

Variable: imenu-prev-index-position-function

If this variable is non-nil, its value should be a function that finds the next definition to put in the buffer index, scanning backward in the buffer from point. It should return nil if it doesn’t find another definition before point. Otherwise it should leave point at the place it finds a definition and return any non-nil value.

Setting this variable makes it buffer-local in the current buffer.

Variable: imenu-extract-index-name-function

If this variable is non-nil, its value should be a function to return the name for a definition, assuming point is in that definition as the imenu-prev-index-position-function function would leave it.

Setting this variable makes it buffer-local in the current buffer.

The last way to customize Imenu for a major mode is to set the variable imenu-create-index-function:

Variable: imenu-create-index-function

This variable specifies the function to use for creating a buffer index. The function should take no arguments, and return an index alist for the current buffer. It is called within save-excursion, so where it leaves point makes no difference.

The index alist can have three types of elements. Simple elements look like this:

(index-name . index-position)

Selecting a simple element has the effect of moving to position index-position in the buffer. Special elements look like this:

(index-name index-position function arguments…)

Selecting a special element performs:

(funcall function
         index-name index-position arguments…)

A nested sub-alist element looks like this:

(menu-title . sub-alist)

It creates the submenu menu-title specified by sub-alist.

The default value of imenu-create-index-function is imenu-default-create-index-function. This function calls the value of imenu-prev-index-position-function and the value of imenu-extract-index-name-function to produce the index alist. However, if either of these two variables is nil, the default function uses imenu-generic-expression instead.

Setting this variable makes it buffer-local in the current buffer.

If built with tree-sitter, Emacs can automatically generate an Imenu index if the major mode sets relevant variables.

Variable: treesit-simple-imenu-settings

This variable instructs Emacs how to generate Imenu indexes. It should be a list of (category regexp pred name-fn).

category should be the name of a category, like "Function", "Class", etc. regexp should be a regexp matching the type of nodes that belong to category. pred should be either nil or a function that takes a node as the argument. It should return non-nil if the node is a valid node for category, or nil if not.

category could also be nil, in which case the entries matched by regexp and pred are not grouped under category.

name-fn should be either nil or a function that takes a defun node and returns the name of that defun, e.g., the function name for a function definition. If name-fn is nil, treesit-defun-name (see Developing major modes with tree-sitter) is used instead.

treesit-major-mode-setup (see Developing major modes with tree-sitter) automatically sets up Imenu if this variable is non-nil.


24.6 Font Lock Mode

Font Lock mode is a buffer-local minor mode that automatically attaches face properties to certain parts of the buffer based on their syntactic role. How it parses the buffer depends on the major mode; most major modes define syntactic criteria for which faces to use in which contexts. This section explains how to customize Font Lock for a particular major mode.

Font Lock mode finds text to highlight in three ways: through parsing based on a full-blown parser (usually, via an external library or program), through syntactic parsing based on the Emacs’s built-in syntax table, or through searching (usually for regular expressions). If enabled, parser-based fontification happens first (see Parser-based Font Lock). Syntactic fontification happens next; it finds comments and string constants and highlights them. Search-based fontification happens last.


24.6.1 Font Lock Basics

The Font Lock functionality is based on several basic functions. Each of these calls the function specified by the corresponding variable. This indirection allows major and minor modes to modify the way fontification works in the buffers of that mode, and even use the Font Lock mechanisms for features that have nothing to do with fontification. (This is why the description below says “should” when it describes what the functions do: the mode can customize the values of the corresponding variables to do something entirely different.) The variables mentioned below are described in Other Font Lock Variables.

font-lock-fontify-buffer

This function should fontify the current buffer’s accessible portion, by calling the function specified by font-lock-fontify-buffer-function.

font-lock-unfontify-buffer

Used when turning Font Lock off to remove the fontification. Calls the function specified by font-lock-unfontify-buffer-function.

font-lock-fontify-region beg end &optional loudly

Should fontify the region between beg and end. If loudly is non-nil, should display status messages while fontifying. Calls the function specified by font-lock-fontify-region-function.

font-lock-unfontify-region beg end

Should remove fontification from the region between beg and end. Calls the function specified by font-lock-unfontify-region-function.

font-lock-flush &optional beg end

This function should mark the fontification of the region between beg and end as outdated. If not specified or nil, beg and end default to the beginning and end of the buffer’s accessible portion. Calls the function specified by font-lock-flush-function.

font-lock-ensure &optional beg end

This function should make sure the region between beg and end has been fontified. The optional arguments beg and end default to the beginning and the end of the buffer’s accessible portion. Calls the function specified by font-lock-ensure-function.

font-lock-debug-fontify

This is a convenience command meant to be used when developing font locking for a mode, and should not be called from Lisp code. It recomputes all the relevant variables and then calls font-lock-fontify-region on the entire buffer.

There are several variables that control how Font Lock mode highlights text. But major modes should not set any of these variables directly. Instead, they should set font-lock-defaults as a buffer-local variable. The value assigned to this variable is used, if and when Font Lock mode is enabled, to set all the other variables.

Variable: font-lock-defaults

This variable is set by modes to specify how to fontify text in that mode. It automatically becomes buffer-local when set. If its value is nil, Font Lock mode does no highlighting, and you can use the ‘Faces’ menu (under ‘Edit’ and then ‘Text Properties’ in the menu bar) to assign faces explicitly to text in the buffer.

If non-nil, the value should look like this:

(keywords [keywords-only [case-fold
 [syntax-alist other-vars…]]])

The first element, keywords, indirectly specifies the value of font-lock-keywords which directs search-based fontification. It can be a symbol, a variable or a function whose value is the list to use for font-lock-keywords. It can also be a list of several such symbols, one for each possible level of fontification. The first symbol specifies the ‘mode default’ level of fontification, the next symbol level 1 fontification, the next level 2, and so on. The ‘mode default’ level is normally the same as level 1. It is used when font-lock-maximum-decoration has a nil value. See Levels of Font Lock.

The second element, keywords-only, specifies the value of the variable font-lock-keywords-only. If this is omitted or nil, syntactic fontification (of strings and comments) is also performed. If this is non-nil, syntactic fontification is not performed. See Syntactic Font Lock.

The third element, case-fold, specifies the value of font-lock-keywords-case-fold-search. If it is non-nil, Font Lock mode ignores case during search-based fontification.

If the fourth element, syntax-alist, is non-nil, it should be a list of cons cells of the form (char-or-string . string). These are used to set up a syntax table for syntactic fontification; the resulting syntax table is stored in font-lock-syntax-table. If syntax-alist is omitted or nil, syntactic fontification uses the syntax table returned by the syntax-table function. See Syntax Table Functions.

All the remaining elements (if any) are collectively called other-vars. Each of these elements should have the form (variable . value)—which means, make variable buffer-local and then set it to value. You can use these other-vars to set other variables that affect fontification, aside from those you can control with the first five elements. See Other Font Lock Variables.

If your mode fontifies text explicitly by adding font-lock-face properties, it can specify (nil t) for font-lock-defaults to turn off all automatic fontification. However, this is not required; it is possible to fontify some things using font-lock-face properties and set up automatic fontification for other parts of the text.


24.6.2 Search-based Fontification

The variable which directly controls search-based fontification is font-lock-keywords, which is typically specified via the keywords element in font-lock-defaults.

Variable: font-lock-keywords

The value of this variable is a list of the keywords to highlight. Lisp programs should not set this variable directly. Normally, the value is automatically set by Font Lock mode, using the keywords element in font-lock-defaults. The value can also be altered using the functions font-lock-add-keywords and font-lock-remove-keywords (see Customizing Search-Based Fontification).

Each element of font-lock-keywords specifies how to find certain cases of text, and how to highlight those cases. Font Lock mode processes the elements of font-lock-keywords one by one, and for each element, it finds and handles all matches. Ordinarily, once part of the text has been fontified already, this cannot be overridden by a subsequent match in the same text; but you can specify different behavior using the override element of a subexp-highlighter.

Each element of font-lock-keywords should have one of these forms:

regexp

Highlight all matches for regexp using font-lock-keyword-face. For example,

;; Highlight occurrences of the word ‘foo
;; using font-lock-keyword-face.
"\\<foo\\>"

Be careful when composing these regular expressions; a poorly written pattern can dramatically slow things down! The function regexp-opt (see Regular Expression Functions) is useful for calculating optimal regular expressions to match several keywords.

function

Find text by calling function, and highlight the matches it finds using font-lock-keyword-face.

When function is called, it receives one argument, the limit of the search; it should begin searching at point, and not search beyond the limit. It should return non-nil if it succeeds, and set the match data to describe the match that was found. Returning nil indicates failure of the search.

Fontification will call function repeatedly with the same limit, and with point where the previous invocation left it, until function fails. On failure, function need not reset point in any particular way.

(matcher . subexp)

In this kind of element, matcher is either a regular expression or a function, as described above. The CDR, subexp, specifies which subexpression of matcher should be highlighted (instead of the entire text that matcher matched).

;; Highlight the ‘bar’ in each occurrence of ‘fubar’,
;; using font-lock-keyword-face.
("fu\\(bar\\)" . 1)
(matcher . facespec)

In this kind of element, facespec is an expression whose value specifies the face to use for highlighting. In the simplest case, facespec is a Lisp variable (a symbol) whose value is a face name.

;; Highlight occurrences of ‘fubar’,
;; using the face which is the value of fubar-face.
("fubar" . fubar-face)

However, facespec can also evaluate to a list of this form:

(subexp
(face face prop1 val1 prop2 val2…))

to specify the face face and various additional text properties to put on the text that matches. If you do this, be sure to add the other text property names that you set in this way to the value of font-lock-extra-managed-props so that the properties will also be cleared out when they are no longer appropriate. Alternatively, you can set the variable font-lock-unfontify-region-function to a function that clears these properties. See Other Font Lock Variables.

(matcher . subexp-highlighter)

In this kind of element, subexp-highlighter is a list which specifies how to highlight matches found by matcher. It has the form:

(subexp facespec [override [laxmatch]])

The CAR, subexp, is an integer specifying which subexpression of the match to fontify (0 means the entire matching text). The second subelement, facespec, is an expression whose value specifies the face, as described above.

The last two values in subexp-highlighter, override and laxmatch, are optional flags. If override is t, this element can override existing fontification made by previous elements of font-lock-keywords. If it is keep, then each character is fontified if it has not been fontified already by some other element. If it is prepend, the face specified by facespec is added to the beginning of the font-lock-face property. If it is append, the face is added to the end of the font-lock-face property.

If laxmatch is non-nil, it means there should be no error if there is no subexpression numbered subexp in matcher. Obviously, fontification of the subexpression numbered subexp will not occur. However, fontification of other subexpressions (and other regexps) will continue. If laxmatch is nil, and the specified subexpression is missing, then an error is signaled which terminates search-based fontification.

Here are some examples of elements of this kind, and what they do:

;; Highlight occurrences of either ‘foo’ or ‘bar’, using
;; foo-bar-face, even if they have already been highlighted.
;; foo-bar-face should be a variable whose value is a face.
("foo\\|bar" 0 foo-bar-face t)

;; Highlight the first subexpression within each occurrence
;; that the function fubar-match finds,
;; using the face which is the value of fubar-face.
(fubar-match 1 fubar-face)
(matcher . anchored-highlighter)

In this kind of element, anchored-highlighter specifies how to highlight text that follows a match found by matcher. So a match found by matcher acts as the anchor for further searches specified by anchored-highlighter. anchored-highlighter is a list of the following form:

(anchored-matcher pre-form post-form subexp-highlighters…)

Here, anchored-matcher, like matcher, is either a regular expression or a function. After a match of matcher is found, point is at the end of the match. Now, Font Lock evaluates the form pre-form. Then it searches for matches of anchored-matcher and uses subexp-highlighters to highlight these. A subexp-highlighter is as described above. Finally, Font Lock evaluates post-form.

The forms pre-form and post-form can be used to initialize before, and cleanup after, anchored-matcher is used. Typically, pre-form is used to move point to some position relative to the match of matcher, before starting with anchored-matcher. post-form might be used to move back, before resuming with matcher.

After Font Lock evaluates pre-form, it does not search for anchored-matcher beyond the end of the line. However, if pre-form returns a buffer position that is greater than the position of point after pre-form is evaluated, then the position returned by pre-form is used as the limit of the search instead. It is generally a bad idea to return a position greater than the end of the line; in other words, the anchored-matcher search should not span lines.

For example,

;; Highlight occurrences of the word ‘item’ following
;; an occurrence of the word ‘anchor’ (on the same line)
;; in the value of item-face.
("\\<anchor\\>" "\\<item\\>" nil nil (0 item-face))

Here, pre-form and post-form are nil. Therefore searching for ‘item’ starts at the end of the match of ‘anchor’, and searching for subsequent instances of ‘anchor’ resumes from where searching for ‘item’ concluded.

(matcher highlighters…)

This sort of element specifies several highlighter lists for a single matcher. A highlighter list can be of the type subexp-highlighter or anchored-highlighter as described above.

For example,

;; Highlight occurrences of the word ‘anchor’ in the value
;; of anchor-face, and subsequent occurrences of the word
;; item’ (on the same line) in the value of item-face.
("\\<anchor\\>" (0 anchor-face)
                ("\\<item\\>" nil nil (0 item-face)))
(eval . form)

Here form is an expression to be evaluated the first time this value of font-lock-keywords is used in a buffer. Its value should have one of the forms described in this table.

Warning: Do not design an element of font-lock-keywords to match text which spans lines; this does not work reliably. For details, see Multiline Font Lock Constructs.

You can use case-fold in font-lock-defaults to specify the value of font-lock-keywords-case-fold-search which says whether search-based fontification should be case-insensitive.

Variable: font-lock-keywords-case-fold-search

Non-nil means that regular expression matching for the sake of font-lock-keywords should be case-insensitive.


24.6.3 Customizing Search-Based Fontification

You can use font-lock-add-keywords to add additional search-based fontification rules to a major mode, and font-lock-remove-keywords to remove rules. You can also customize the font-lock-ignore option to selectively disable fontification rules for keywords that match certain criteria.

Function: font-lock-add-keywords mode keywords &optional how

This function adds highlighting keywords, for the current buffer or for major mode mode. The argument keywords should be a list with the same format as the variable font-lock-keywords.

If mode is a symbol which is a major mode command name, such as c-mode, the effect is that enabling Font Lock mode in mode will add keywords to font-lock-keywords. Calling with a non-nil value of mode is correct only in your ~/.emacs file.

If mode is nil, this function adds keywords to font-lock-keywords in the current buffer. This way of calling font-lock-add-keywords is usually used in mode hook functions.

By default, keywords are added at the beginning of font-lock-keywords. If the optional argument how is set, they are used to replace the value of font-lock-keywords. If how is any other non-nil value, they are added at the end of font-lock-keywords.

Some modes provide specialized support you can use in additional highlighting patterns. See the variables c-font-lock-extra-types, c++-font-lock-extra-types, and java-font-lock-extra-types, for example.

Warning: Major mode commands must not call font-lock-add-keywords under any circumstances, either directly or indirectly, except through their mode hooks. (Doing so would lead to incorrect behavior for some minor modes.) They should set up their rules for search-based fontification by setting font-lock-keywords.

Function: font-lock-remove-keywords mode keywords

This function removes keywords from font-lock-keywords for the current buffer or for major mode mode. As in font-lock-add-keywords, mode should be a major mode command name or nil. All the caveats and requirements for font-lock-add-keywords apply here too. The argument keywords must exactly match the one used by the corresponding font-lock-add-keywords.

For example, the following code adds two fontification patterns for C mode: one to fontify the word ‘FIXME’, even in comments, and another to fontify the words ‘and’, ‘or’ and ‘not’ as keywords.

(font-lock-add-keywords 'c-mode
 '(("\\<\\(FIXME\\):" 1 font-lock-warning-face prepend)
   ("\\<\\(and\\|or\\|not\\)\\>" . font-lock-keyword-face)))

This example affects only C mode proper. To add the same patterns to C mode and all modes derived from it, do this instead:

(add-hook 'c-mode-hook
 (lambda ()
  (font-lock-add-keywords nil
   '(("\\<\\(FIXME\\):" 1 font-lock-warning-face prepend)
     ("\\<\\(and\\|or\\|not\\)\\>" .
      font-lock-keyword-face)))))
User Option: font-lock-ignore

This option defines conditions for selectively disabling fontifications due to certain Font Lock keywords. If non-nil, its value is a list of elements of the following form:

(symbol condition …)

Here, symbol is a symbol, usually a major or minor mode. The subsequent conditions of a symbol’s list element will be in effect if symbol is bound and its value is non-nil. For a mode’s symbol, it means that the current major mode is derived from that mode, or that minor mode is enabled in the buffer. When a condition is in effect, any fontifications caused by font-lock-keywords elements that match the condition will be disabled.

Each condition can be one of the following:

a symbol

This condition matches any element of Font Lock keywords that references the symbol. This is usually a face, but can be any symbol referenced by an element of the font-lock-keywords list. The symbol can contain wildcards: * matches any string in the symbol’ss name, ? matches a single character, and [char-set], where char-set is a string of one or more characters, matches a single character from the set.

a string

This condition matches any element of Font Lock keywords whose matcher is a regexp which matches the string. In other words, this condition matches a Font Lock rule which highlights the string. Thus, the string could be a specific program keyword whose highlighting you want to disable.

(pred function)

This condition matches any element of Font Lock keywords for which function, when called with the element as the argument, returns non-nil.

(not condition)

This matches if condition doesn’t.

(and condition …)

This matches if each of the conditions matches.

(or condition …)

This matches if at least one of the conditions matches.

(except condition)

This condition can only be used at top level or inside an or clause. It undoes the effect of a previously matching condition on the same level.

As an example, consider the following setting:

(setq font-lock-ignore
      '((prog-mode font-lock-*-face
                   (except help-echo))
        (emacs-lisp-mode (except ";;;###autoload)")
        (whitespace-mode whitespace-empty-at-bob-regexp)
        (makefile-mode (except *))))

Line by line, this does the following:

  1. In all programming modes, disable fontifications due to all font-lock keywords that apply one of the standard font-lock faces (excluding strings and comments, which are covered by syntactic Font Lock).
  2. However, keep any keywords that add a help-echo text property.
  3. In Emacs Lisp mode, also keep the highlighting of autoload cookies, which would have been excluded by the first condition.
  4. When whitespace-mode (a minor mode) is enabled, also don’t highlight an empty line at beginning of buffer.
  5. Finally, in Makefile mode, don’t apply any conditions.

24.6.4 Other Font Lock Variables

This section describes additional variables that a major mode can set by means of other-vars in font-lock-defaults (see Font Lock Basics).

Variable: font-lock-mark-block-function

If this variable is non-nil, it should be a function that is called with no arguments, to choose an enclosing range of text for refontification for the command M-x font-lock-fontify-block.

The function should report its choice by placing the region around it. A good choice is a range of text large enough to give proper results, but not too large so that refontification becomes slow. Typical values are mark-defun for programming modes or mark-paragraph for textual modes.

Variable: font-lock-extra-managed-props

This variable specifies additional properties (other than font-lock-face) that are being managed by Font Lock mode. It is used by font-lock-default-unfontify-region, which normally only manages the font-lock-face property. If you want Font Lock to manage other properties as well, you must specify them in a facespec in font-lock-keywords as well as add them to this list. See Search-based Fontification.

Variable: font-lock-fontify-buffer-function

Function to use for fontifying the buffer. The default value is font-lock-default-fontify-buffer.

Variable: font-lock-unfontify-buffer-function

Function to use for unfontifying the buffer. This is used when turning off Font Lock mode. The default value is font-lock-default-unfontify-buffer.

Variable: font-lock-fontify-region-function

Function to use for fontifying a region. It should take two arguments, the beginning and end of the region, and an optional third argument verbose. If verbose is non-nil, the function should print status messages. The default value is font-lock-default-fontify-region.

Variable: font-lock-unfontify-region-function

Function to use for unfontifying a region. It should take two arguments, the beginning and end of the region. The default value is font-lock-default-unfontify-region.

Variable: font-lock-flush-function

Function to use for declaring that a region’s fontification is out of date. It takes two arguments, the beginning and end of the region. The default value of this variable is font-lock-after-change-function.

Variable: font-lock-ensure-function

Function to use for making sure a region of the current buffer has been fontified. It is called with two arguments, the beginning and end of the region. The default value of this variable is a function that calls font-lock-default-fontify-buffer if the buffer is not fontified; the effect is to make sure the entire accessible portion of the buffer is fontified.

Function: jit-lock-register function &optional contextual

This function tells Font Lock mode to run the Lisp function function any time it has to fontify or refontify part of the current buffer. It calls function before calling the default fontification functions, and gives it two arguments, start and end, which specify the region to be fontified or refontified. If function performs fontifications, it can return a list of the form (jit-lock-bounds beg . end), to indicate the bounds of the region it actually fontified; Just-In-Time (a.k.a. “JIT”) font-lock will use this information to optimize subsequent redisplay cycles and regions of buffer text it will pass to future calls to function.

The optional argument contextual, if non-nil, forces Font Lock mode to always refontify a syntactically relevant part of the buffer, and not just the modified lines. This argument can usually be omitted.

When Font Lock is activated in a buffer, it calls this function with a non-nil value of contextual if the value of font-lock-keywords-only (see Syntactic Font Lock) is nil.

Function: jit-lock-unregister function

If function was previously registered as a fontification function using jit-lock-register, this function unregisters it.

Command: jit-lock-debug-mode &optional arg

This is a minor mode whose purpose is to help in debugging code that is run by JIT font-lock. When this mode is enabled, most of the code that JIT font-lock normally runs during redisplay cycles, where Lisp errors are suppressed, is instead run by a timer. Thus, this mode allows using debugging aids such as debug-on-error (see Entering the Debugger on an Error) and Edebug (see Edebug) for finding and fixing problems in font-lock code and any other code run by JIT font-lock. Another command that could be useful when developing and debugging font-lock is font-lock-debug-fontify, see Font Lock Basics.


24.6.5 Levels of Font Lock

Some major modes offer three different levels of fontification. You can define multiple levels by using a list of symbols for keywords in font-lock-defaults. Each symbol specifies one level of fontification; it is up to the user to choose one of these levels, normally by setting font-lock-maximum-decoration (see Font Lock in the GNU Emacs Manual). The chosen level’s symbol value is used to initialize font-lock-keywords.

Here are the conventions for how to define the levels of fontification:

  • Level 1: highlight function declarations, file directives (such as include or import directives), strings and comments. The idea is speed, so only the most important and top-level components are fontified.
  • Level 2: in addition to level 1, highlight all language keywords, including type names that act like keywords, as well as named constant values. The idea is that all keywords (either syntactic or semantic) should be fontified appropriately.
  • Level 3: in addition to level 2, highlight the symbols being defined in function and variable declarations, and all builtin function names, wherever they appear.

24.6.6 Precalculated Fontification

Some major modes such as list-buffers and occur construct the buffer text programmatically. The easiest way for them to support Font Lock mode is to specify the faces of text when they insert the text in the buffer.

The way to do this is to specify the faces in the text with the special text property font-lock-face (see Properties with Special Meanings). When Font Lock mode is enabled, this property controls the display, just like the face property. When Font Lock mode is disabled, font-lock-face has no effect on the display.

It is ok for a mode to use font-lock-face for some text and also use the normal Font Lock machinery. But if the mode does not use the normal Font Lock machinery, it should not set the variable font-lock-defaults. In this case the face property will not be overridden, so using the face property could work too. However, using font-lock-face is generally preferable as it allows the user to control the fontification by toggling font-lock-mode, and lets the code work regardless of whether the mode uses Font Lock machinery or not.


24.6.7 Faces for Font Lock

Font Lock mode can highlight using any face, but Emacs defines several faces specifically for Font Lock to use to highlight text. These Font Lock faces are listed below. They can also be used by major modes for syntactic highlighting outside of Font Lock mode (see Major Mode Conventions).

Each of these symbols is both a face name, and a variable whose default value is the symbol itself. Thus, the default value of font-lock-comment-face is font-lock-comment-face.

The faces are listed with descriptions of their typical usage, and in order of greater to lesser prominence. If a mode’s syntactic categories do not fit well with the usage descriptions, the faces can be assigned using the ordering as a guide.

font-lock-warning-face

for a construct that is peculiar (e.g., an unescaped confusable quote in an Emacs Lisp symbol like ‘‘foo’), or that greatly changes the meaning of other text, like ‘;;;###autoload’ in Emacs Lisp and ‘#error’ in C.

font-lock-function-name-face

for the name of a function being defined or declared.

font-lock-function-call-face

for the name of a function being called. This face inherits, by default, from font-lock-function-name-face.

font-lock-variable-name-face

for the name of a variable being defined or declared.

font-lock-variable-use-face

for the name of a variable being referenced. This face inherits, by default, from font-lock-variable-name-face.

font-lock-keyword-face

for a keyword with special syntactic significance, like ‘for’ and ‘if’ in C.

font-lock-comment-face

for comments.

font-lock-comment-delimiter-face

for comments delimiters, like ‘/*’ and ‘*/’ in C. On most terminals, this inherits from font-lock-comment-face.

font-lock-type-face

for the names of user-defined data types.

font-lock-constant-face

for the names of constants, like ‘NULL’ in C.

font-lock-builtin-face

for the names of built-in functions.

font-lock-preprocessor-face

for preprocessor commands. This inherits, by default, from font-lock-builtin-face.

font-lock-string-face

for string constants.

font-lock-doc-face

for documentation embedded in program code inside specially-formed comments or strings. This face inherits, by default, from font-lock-string-face.

font-lock-doc-markup-face

for mark-up elements in text using font-lock-doc-face. It is typically used for the mark-up constructs in documentation embedded in program code, following conventions such as Haddock, Javadoc or Doxygen. This face inherits, by default, from font-lock-constant-face.

font-lock-negation-char-face

for easily-overlooked negation characters.

font-lock-escape-face

for escape sequences in strings. This face inherits, by default, from font-lock-regexp-grouping-backslash.

Here is an example in Python, where the escape sequence \n is used:

print('Hello world!\n')
font-lock-number-face

for numbers.

font-lock-operator-face

for operators.

font-lock-property-name-face

for properties of an object, such as the declaration of fields in a struct. This face inherits, by default, from font-lock-variable-name-face.

font-lock-property-use-face

for properties of an object, such as use of fields in a struct. This face inherits, by default, from font-lock-property-name-face.

For example,

typedef struct
{
  int prop;
//    ^ property
} obj;

int main()
{
  obj o;
  o.prop = 3;
//  ^ property
}
font-lock-punctuation-face

for punctuation such as brackets and delimiters.

font-lock-bracket-face

for brackets (e.g., (), [], {}). This face inherits, by default, from font-lock-punctuation-face.

font-lock-delimiter-face

for delimiters (e.g., ;, :, ,). This face inherits, by default, from font-lock-punctuation-face.

font-lock-misc-punctuation-face

for punctuation that is not a bracket or delimiter. This face inherits, by default, from font-lock-punctuation-face.


24.6.8 Syntactic Font Lock

Syntactic fontification uses a syntax table (see Syntax Tables) to find and highlight syntactically relevant text. If enabled, it runs prior to search-based fontification. The variable font-lock-syntactic-face-function, documented below, determines which syntactic constructs to highlight. There are several variables that affect syntactic fontification; you should set them by means of font-lock-defaults (see Font Lock Basics).

Whenever Font Lock mode performs syntactic fontification on a stretch of text, it first calls the function specified by syntax-propertize-function. Major modes can use this to apply syntax-table text properties to override the buffer’s syntax table in special cases. See Syntax Properties.

Variable: font-lock-keywords-only

If the value of this variable is non-nil, Font Lock does not do syntactic fontification, only search-based fontification based on font-lock-keywords. It is normally set by Font Lock mode based on the keywords-only element in font-lock-defaults. If the value is nil, Font Lock will call jit-lock-register (see Other Font Lock Variables) to set up for automatic refontification of buffer text following a modified line to reflect the new syntactic context due to the change.

To use only syntactic fontification, this variable should be non-nil, while font-lock-keywords should be set to nil (see Font Lock Basics).

Variable: font-lock-syntax-table

This variable holds the syntax table to use for fontification of comments and strings. It is normally set by Font Lock mode based on the syntax-alist element in font-lock-defaults. If this value is nil, syntactic fontification uses the buffer’s syntax table (the value returned by the function syntax-table; see Syntax Table Functions).

Variable: font-lock-syntactic-face-function

If this variable is non-nil, it should be a function to determine which face to use for a given syntactic element (a string or a comment).

The function is called with one argument, the parse state at point returned by parse-partial-sexp, and should return a face. The default value returns font-lock-comment-face for comments and font-lock-string-face for strings (see Faces for Font Lock).

This variable is normally set through the “other” elements in font-lock-defaults:

(setq-local font-lock-defaults
            `(,python-font-lock-keywords
              nil nil nil nil
              (font-lock-syntactic-face-function
               . python-font-lock-syntactic-face-function)))

24.6.9 Multiline Font Lock Constructs

Normally, elements of font-lock-keywords should not match across multiple lines; that doesn’t work reliably, because Font Lock usually scans just part of the buffer, and it can miss a multi-line construct that crosses the line boundary where the scan starts. (The scan normally starts at the beginning of a line.)

Making elements that match multiline constructs work properly has two aspects: correct identification and correct rehighlighting. The first means that Font Lock finds all multiline constructs. The second means that Font Lock will correctly rehighlight all the relevant text when a multiline construct is changed—for example, if some of the text that was previously part of a multiline construct ceases to be part of it. The two aspects are closely related, and often getting one of them to work will appear to make the other also work. However, for reliable results you must attend explicitly to both aspects.

There are three ways to ensure correct identification of multiline constructs:

  • Add a function to font-lock-extend-region-functions that does the identification and extends the scan so that the scanned text never starts or ends in the middle of a multiline construct.
  • Use the font-lock-fontify-region-function hook similarly to extend the scan so that the scanned text never starts or ends in the middle of a multiline construct.
  • Somehow identify the multiline construct right when it gets inserted into the buffer (or at any point after that but before font-lock tries to highlight it), and mark it with a font-lock-multiline which will instruct font-lock not to start or end the scan in the middle of the construct.

There are several ways to do rehighlighting of multiline constructs:

  • Place a font-lock-multiline property on the construct. This will rehighlight the whole construct if any part of it is changed. In some cases you can do this automatically by setting the font-lock-multiline variable, which see.
  • Make sure jit-lock-contextually is set and rely on it doing its job. This will only rehighlight the part of the construct that follows the actual change, and will do it after a short delay. This only works if the highlighting of the various parts of your multiline construct never depends on text in subsequent lines. Since jit-lock-contextually is activated by default, this can be an attractive solution.
  • Place a jit-lock-defer-multiline property on the construct. This works only if jit-lock-contextually is used, and with the same delay before rehighlighting, but like font-lock-multiline, it also handles the case where highlighting depends on subsequent lines.
  • If parsing the syntax of a construct depends on it being parsed in one single chunk, you can add the syntax-multiline text property over the construct in question. The most common use for this is when the syntax property to apply to ‘FOO’ depend on some later text ‘BAR’: By placing this text property over the whole of ‘FOO...BAR’, you make sure that any change of ‘BAR’ will also cause the syntax property of ‘FOO’ to be recomputed. Note: For this to work, the mode needs to add syntax-propertize-multiline to syntax-propertize-extend-region-functions.

24.6.9.1 Font Lock Multiline

One way to ensure reliable rehighlighting of multiline Font Lock constructs is to put on them the text property font-lock-multiline. It should be present and non-nil for text that is part of a multiline construct.

When Font Lock is about to highlight a range of text, it first extends the boundaries of the range as necessary so that they do not fall within text marked with the font-lock-multiline property. Then it removes any font-lock-multiline properties from the range, and highlights it. The highlighting specification (mostly font-lock-keywords) must reinstall this property each time, whenever it is appropriate.

Warning: don’t use the font-lock-multiline property on large ranges of text, because that will make rehighlighting slow.

Variable: font-lock-multiline

If the font-lock-multiline variable is set to t, Font Lock will try to add the font-lock-multiline property automatically on multiline constructs. This is not a universal solution, however, since it slows down Font Lock somewhat. It can miss some multiline constructs, or make the property larger or smaller than necessary.

For elements whose matcher is a function, the function should ensure that submatch 0 covers the whole relevant multiline construct, even if only a small subpart will be highlighted. It is often just as easy to add the font-lock-multiline property by hand.

The font-lock-multiline property is meant to ensure proper refontification; it does not automatically identify new multiline constructs. Identifying them requires that Font Lock mode operate on large enough chunks at a time. This will happen by accident on many cases, which may give the impression that multiline constructs magically work. If you set the font-lock-multiline variable non-nil, this impression will be even stronger, since the highlighting of those constructs which are found will be properly updated from then on. But that does not work reliably.

To find multiline constructs reliably, you must either manually place the font-lock-multiline property on the text before Font Lock mode looks at it, or use font-lock-fontify-region-function.


24.6.9.2 Region to Fontify after a Buffer Change

When a buffer is changed, the region that Font Lock refontifies is by default the smallest sequence of whole lines that spans the change. While this works well most of the time, sometimes it doesn’t—for example, when a change alters the syntactic meaning of text on an earlier line.

You can enlarge (or even reduce) the region to refontify by setting the following variable:

Variable: font-lock-extend-after-change-region-function

This buffer-local variable is either nil or a function for Font Lock mode to call to determine the region to scan and fontify.

The function is given three parameters, the standard beg, end, and old-len from after-change-functions (see Change Hooks). It should return either a cons of the beginning and end buffer positions (in that order) of the region to fontify, or nil (which means choose the region in the standard way). This function needs to preserve point, the match-data, and the current restriction. The region it returns may start or end in the middle of a line.

Since this function is called after every buffer change, it should be reasonably fast.


24.6.10 Parser-based Font Lock

Besides simple syntactic font lock and regexp-based font lock, Emacs also provides complete syntactic font lock with the help of a parser. Currently, Emacs uses the tree-sitter library (see Parsing Program Source) for this purpose.

Parser-based font lock and other font lock mechanisms are not mutually exclusive. By default, if enabled, parser-based font lock runs first, replacing syntactic font lock, followed by regexp-based font lock.

Although parser-based font lock doesn’t share the same customization variables with regexp-based font lock, it uses similar customization schemes. The tree-sitter counterpart of font-lock-keywords is treesit-font-lock-settings.

In general, tree-sitter fontification works as follows:

  • A Lisp program (usually, part of a major mode) provides a query consisting of patterns, each pattern associated with a capture name.
  • The tree-sitter library finds the nodes in the parse tree that match these patterns, tags the nodes with the corresponding capture names, and returns them to the Lisp program.
  • The Lisp program uses the returned nodes to highlight the portions of buffer text corresponding to each node as appropriate, using the tagged capture names of the nodes to determine the correct fontification. For example, a node tagged font-lock-keyword would be highlighted in font-lock-keyword face.

For more information about queries, patterns, and capture names, see Pattern Matching Tree-sitter Nodes.

To set up tree-sitter fontification, a major mode should first set treesit-font-lock-settings with the output of treesit-font-lock-rules, then call treesit-major-mode-setup.

Function: treesit-font-lock-rules &rest query-specs

This function is used to set treesit-font-lock-settings. It takes care of compiling queries and other post-processing, and outputs a value that treesit-font-lock-settings accepts. Here’s an example:

(treesit-font-lock-rules
 :language 'javascript
 :feature 'constant
 :override t
 '((true) @font-lock-constant-face
   (false) @font-lock-constant-face)
 :language 'html
 :feature 'script
 "(script_element) @font-lock-builtin-face")

This function takes a series of query-specs, where each query-spec is a query preceded by one or more keyword/value pairs. Each query is a tree-sitter query in either the string, s-expression, or compiled form.

For each query, the keyword/value pairs that precede it add meta information to it. The :language keyword declares query’s language. The :feature keyword sets the feature name of query. Users can control which features are enabled with treesit-font-lock-level and treesit-font-lock-feature-list (described below). These two keywords are mandatory.

Other keywords are optional:

KeywordValueDescription
:overridenilIf the region already has a face, discard the new face
tAlways apply the new face
appendAppend the new face to existing ones
prependPrepend the new face to existing ones
keepFill-in regions without an existing face

Lisp programs mark patterns in query with capture names (names that start with @), and tree-sitter will return matched nodes tagged with those same capture names. For the purpose of fontification, capture names in query should be face names like font-lock-keyword-face. The captured node will be fontified with that face.

A capture name can also be a function name, in which case the function is called with 4 arguments: node and override, start and end, where node is the node itself, override is the :override property of the rule which captured this node, and start and end limit the region which this function should fontify. (If this function wants to respect the override argument, it can use treesit-fontify-with-override.)

Beyond the 4 arguments presented, this function should accept more arguments as optional arguments for future extensibility.

If a capture name is both a face and a function, the face takes priority. If a capture name is neither a face nor a function, it is ignored.

Variable: treesit-font-lock-feature-list

This is a list of lists of feature symbols. Each element of the list is a list that represents a decoration level. treesit-font-lock-level controls which levels are activated.

Each element of the list is a list of the form (feature …), where each feature corresponds to the :feature value of a query defined in treesit-font-lock-rules. Removing a feature symbol from this list disables the corresponding query during font-lock.

Common feature names, for many programming languages, include definition, type, assignment, builtin, constant, keyword, string-interpolation, comment, doc, string, operator, preprocessor, escape-sequence, and key. Major modes are free to subdivide or extend these common features.

Some of these features warrant some explanation: definition highlights whatever is being defined, e.g., the function name in a function definition, the struct name in a struct definition, the variable name in a variable definition; assignment highlights whatever is being assigned to, e.g., the variable or field in an assignment statement; key highlights keys in key-value pairs, e.g., keys in a JSON object or Python dictionary; doc highlights docstrings or doc-comments.

For example, the value of this variable could be:

((comment string doc) ; level 1
 (function-name keyword type builtin constant) ; level 2
 (variable-name string-interpolation key)) ; level 3

Major modes should set this variable before calling treesit-major-mode-setup.

For this variable to take effect, a Lisp program should call treesit-font-lock-recompute-features (which resets treesit-font-lock-settings accordingly), or treesit-major-mode-setup (which calls treesit-font-lock-recompute-features).

Variable: treesit-font-lock-settings

A list of settings for tree-sitter based font lock. The exact format of each setting is considered internal. One should always use treesit-font-lock-rules to set this variable.

Multi-language major modes should provide range functions in treesit-range-functions, and Emacs will set the ranges accordingly before fontifing a region (see Parsing Text in Multiple Languages).


24.7 Automatic Indentation of code

For programming languages, an important feature of a major mode is to provide automatic indentation. There are two parts: one is to decide what is the right indentation of a line, and the other is to decide when to reindent a line. By default, Emacs reindents a line whenever you type a character in electric-indent-chars, which by default only includes Newline. Major modes can add chars to electric-indent-chars according to the syntax of the language.

Deciding what is the right indentation is controlled in Emacs by indent-line-function (see Indentation Controlled by Major Mode). For some modes, the right indentation cannot be known reliably, typically because indentation is significant so several indentations are valid but with different meanings. In that case, the mode should set electric-indent-inhibit to make sure the line is not constantly re-indented against the user’s wishes.

Writing a good indentation function can be difficult and to a large extent it is still a black art. Many major mode authors will start by writing a simple indentation function that works for simple cases, for example by comparing with the indentation of the previous text line. For most programming languages that are not really line-based, this tends to scale very poorly: improving such a function to let it handle more diverse situations tends to become more and more difficult, resulting in the end with a large, complex, unmaintainable indentation function which nobody dares to touch.

A good indentation function will usually need to actually parse the text, according to the syntax of the language. Luckily, it is not necessary to parse the text in as much detail as would be needed for a compiler, but on the other hand, the parser embedded in the indentation code will want to be somewhat friendly to syntactically incorrect code.

Good maintainable indentation functions usually fall into two categories: either parsing forward from some safe starting point until the position of interest, or parsing backward from the position of interest. Neither of the two is a clearly better choice than the other: parsing backward is often more difficult than parsing forward because programming languages are designed to be parsed forward, but for the purpose of indentation it has the advantage of not needing to guess a safe starting point, and it generally enjoys the property that only a minimum of text will be analyzed to decide the indentation of a line, so indentation will tend to be less affected by syntax errors in some earlier unrelated piece of code. Parsing forward on the other hand is usually easier and has the advantage of making it possible to reindent efficiently a whole region at a time, with a single parse.

Rather than write your own indentation function from scratch, it is often preferable to try and reuse some existing ones or to rely on a generic indentation engine. There are sadly few such engines. The CC-mode indentation code (used with C, C++, Java, Awk and a few other such modes) has been made more generic over the years, so if your language seems somewhat similar to one of those languages, you might try to use that engine. Another one is SMIE which takes an approach in the spirit of Lisp sexps and adapts it to non-Lisp languages. Yet another one is to rely on a full-blown parser, for example, the tree-sitter library.


24.7.1 Simple Minded Indentation Engine

SMIE is a package that provides a generic navigation and indentation engine. Based on a very simple parser using an operator precedence grammar, it lets major modes extend the sexp-based navigation of Lisp to non-Lisp languages as well as provide a simple to use but reliable auto-indentation.

Operator precedence grammar is a very primitive technology for parsing compared to some of the more common techniques used in compilers. It has the following characteristics: its parsing power is very limited, and it is largely unable to detect syntax errors, but it has the advantage of being algorithmically efficient and able to parse forward just as well as backward. In practice that means that SMIE can use it for indentation based on backward parsing, that it can provide both forward-sexp and backward-sexp functionality, and that it will naturally work on syntactically incorrect code without any extra effort. The downside is that it also means that most programming languages cannot be parsed correctly using SMIE, at least not without resorting to some special tricks (see Living With a Weak Parser).


24.7.1.1 SMIE Setup and Features

SMIE is meant to be a one-stop shop for structural navigation and various other features which rely on the syntactic structure of code, in particular automatic indentation. The main entry point is smie-setup which is a function typically called while setting up a major mode.

Function: smie-setup grammar rules-function &rest keywords

Setup SMIE navigation and indentation. grammar is a grammar table generated by smie-prec2->grammar. rules-function is a set of indentation rules for use on smie-rules-function. keywords are additional arguments, which can include the following keywords:

  • :forward-token fun: Specify the forward lexer to use.
  • :backward-token fun: Specify the backward lexer to use.

Calling this function is sufficient to make commands such as forward-sexp, backward-sexp, and transpose-sexps be able to properly handle structural elements other than just the paired parentheses already handled by syntax tables. For example, if the provided grammar is precise enough, transpose-sexps can correctly transpose the two arguments of a + operator, taking into account the precedence rules of the language.

Calling smie-setup is also sufficient to make TAB indentation work in the expected way, extends blink-matching-paren to apply to elements like begin...end, and provides some commands that you can bind in the major mode keymap.

Command: smie-close-block

This command closes the most recently opened (and not yet closed) block.

Command: smie-down-list &optional arg

This command is like down-list but it also pays attention to nesting of tokens other than parentheses, such as begin...end.


24.7.1.2 Operator Precedence Grammars

SMIE’s precedence grammars simply give to each token a pair of precedences: the left-precedence and the right-precedence. We say T1 < T2 if the right-precedence of token T1 is less than the left-precedence of token T2. A good way to read this < is as a kind of parenthesis: if we find ... T1 something T2 ... then that should be parsed as ... T1 (something T2 ... rather than as ... T1 something) T2 .... The latter interpretation would be the case if we had T1 > T2. If we have T1 = T2, it means that token T2 follows token T1 in the same syntactic construction, so typically we have "begin" = "end". Such pairs of precedences are sufficient to express left-associativity or right-associativity of infix operators, nesting of tokens like parentheses and many other cases.

Function: smie-prec2->grammar table

This function takes a prec2 grammar table and returns an alist suitable for use in smie-setup. The prec2 table is itself meant to be built by one of the functions below.

Function: smie-merge-prec2s &rest tables

This function takes several prec2 tables and merges them into a new prec2 table.

Function: smie-precs->prec2 precs

This function builds a prec2 table from a table of precedences precs. precs should be a list, sorted by precedence (for example "+" will come before "*"), of elements of the form (assoc op ...), where each op is a token that acts as an operator; assoc is their associativity, which can be either left, right, assoc, or nonassoc. All operators in a given element share the same precedence level and associativity.

Function: smie-bnf->prec2 bnf &rest resolvers

This function lets you specify the grammar using a BNF notation. It accepts a bnf description of the grammar along with a set of conflict resolution rules resolvers, and returns a prec2 table.

bnf is a list of nonterminal definitions of the form (nonterm rhs1 rhs2 ...) where each rhs is a (non-empty) list of terminals (aka tokens) or non-terminals.

Not all grammars are accepted:

  • An rhs cannot be an empty list (an empty list is never needed, since SMIE allows all non-terminals to match the empty string anyway).
  • An rhs cannot have 2 consecutive non-terminals: each pair of non-terminals needs to be separated by a terminal (aka token). This is a fundamental limitation of operator precedence grammars.

Additionally, conflicts can occur:

  • The returned prec2 table holds constraints between pairs of tokens, and for any given pair only one constraint can be present: T1 < T2, T1 = T2, or T1 > T2.
  • A token can be an opener (something similar to an open-paren), a closer (like a close-paren), or neither of the two (e.g., an infix operator, or an inner token like "else").

Precedence conflicts can be resolved via resolvers, which is a list of precs tables (see smie-precs->prec2): for each precedence conflict, if those precs tables specify a particular constraint, then the conflict is resolved by using this constraint instead, else a conflict is reported and one of the conflicting constraints is picked arbitrarily and the others are simply ignored.


24.7.1.3 Defining the Grammar of a Language

The usual way to define the SMIE grammar of a language is by defining a new global variable that holds the precedence table by giving a set of BNF rules. For example, the grammar definition for a small Pascal-like language could look like:

(require 'smie)
(defvar sample-smie-grammar
  (smie-prec2->grammar
   (smie-bnf->prec2
    '((id)
      (inst ("begin" insts "end")
            ("if" exp "then" inst "else" inst)
            (id ":=" exp)
            (exp))
      (insts (insts ";" insts) (inst))
      (exp (exp "+" exp)
           (exp "*" exp)
           ("(" exps ")"))
      (exps (exps "," exps) (exp)))
    '((assoc ";"))
    '((assoc ","))
    '((assoc "+") (assoc "*")))))

A few things to note:

  • The above grammar does not explicitly mention the syntax of function calls: SMIE will automatically allow any sequence of sexps, such as identifiers, balanced parentheses, or begin ... end blocks to appear anywhere anyway.
  • The grammar category id has no right hand side: this does not mean that it can match only the empty string, since as mentioned any sequence of sexps can appear anywhere anyway.
  • Because non terminals cannot appear consecutively in the BNF grammar, it is difficult to correctly handle tokens that act as terminators, so the above grammar treats ";" as a statement separator instead, which SMIE can handle very well.
  • Separators used in sequences (such as "," and ";" above) are best defined with BNF rules such as (foo (foo "separator" foo) ...) which generate precedence conflicts which are then resolved by giving them an explicit (assoc "separator").
  • The ("(" exps ")") rule was not needed to pair up parens, since SMIE will pair up any characters that are marked as having paren syntax in the syntax table. What this rule does instead (together with the definition of exps) is to make it clear that "," should not appear outside of parentheses.
  • Rather than have a single precs table to resolve conflicts, it is preferable to have several tables, so as to let the BNF part of the grammar specify relative precedences where possible.
  • Unless there is a very good reason to prefer left or right, it is usually preferable to mark operators as associative, using assoc. For that reason "+" and "*" are defined above as assoc, although the language defines them formally as left associative.

24.7.1.4 Defining Tokens

SMIE comes with a predefined lexical analyzer which uses syntax tables in the following way: any sequence of characters that have word or symbol syntax is considered a token, and so is any sequence of characters that have punctuation syntax. This default lexer is often a good starting point but is rarely actually correct for any given language. For example, it will consider "2,+3" to be composed of 3 tokens: "2", ",+", and "3".

To describe the lexing rules of your language to SMIE, you need 2 functions, one to fetch the next token, and another to fetch the previous token. Those functions will usually first skip whitespace and comments and then look at the next chunk of text to see if it is a special token. If so it should skip the token and return a description of this token. Usually this is simply the string extracted from the buffer, but it can be anything you want. For example:

(defvar sample-keywords-regexp
  (regexp-opt '("+" "*" "," ";" ">" ">=" "<" "<=" ":=" "=")))
(defun sample-smie-forward-token ()
  (forward-comment (point-max))
  (cond
   ((looking-at sample-keywords-regexp)
    (goto-char (match-end 0))
    (match-string-no-properties 0))
   (t (buffer-substring-no-properties
       (point)
       (progn (skip-syntax-forward "w_")
              (point))))))
(defun sample-smie-backward-token ()
  (forward-comment (- (point)))
  (cond
   ((looking-back sample-keywords-regexp (- (point) 2) t)
    (goto-char (match-beginning 0))
    (match-string-no-properties 0))
   (t (buffer-substring-no-properties
       (point)
       (progn (skip-syntax-backward "w_")
              (point))))))

Notice how those lexers return the empty string when in front of parentheses. This is because SMIE automatically takes care of the parentheses defined in the syntax table. More specifically if the lexer returns nil or an empty string, SMIE tries to handle the corresponding text as a sexp according to syntax tables.


24.7.1.5 Living With a Weak Parser

The parsing technique used by SMIE does not allow tokens to behave differently in different contexts. For most programming languages, this manifests itself by precedence conflicts when converting the BNF grammar.

Sometimes, those conflicts can be worked around by expressing the grammar slightly differently. For example, for Modula-2 it might seem natural to have a BNF grammar that looks like this:

  ...
  (inst ("IF" exp "THEN" insts "ELSE" insts "END")
        ("CASE" exp "OF" cases "END")
        ...)
  (cases (cases "|" cases)
         (caselabel ":" insts)
         ("ELSE" insts))
  ...

But this will create conflicts for "ELSE": on the one hand, the IF rule implies (among many other things) that "ELSE" = "END"; but on the other hand, since "ELSE" appears within cases, which appears left of "END", we also have "ELSE" > "END". We can solve the conflict either by using:

  ...
  (inst ("IF" exp "THEN" insts "ELSE" insts "END")
        ("CASE" exp "OF" cases "END")
        ("CASE" exp "OF" cases "ELSE" insts "END")
        ...)
  (cases (cases "|" cases) (caselabel ":" insts))
  ...

or

  ...
  (inst ("IF" exp "THEN" else "END")
        ("CASE" exp "OF" cases "END")
        ...)
  (else (insts "ELSE" insts))
  (cases (cases "|" cases) (caselabel ":" insts) (else))
  ...

Reworking the grammar to try and solve conflicts has its downsides, tho, because SMIE assumes that the grammar reflects the logical structure of the code, so it is preferable to keep the BNF closer to the intended abstract syntax tree.

Other times, after careful consideration you may conclude that those conflicts are not serious and simply resolve them via the resolvers argument of smie-bnf->prec2. Usually this is because the grammar is simply ambiguous: the conflict does not affect the set of programs described by the grammar, but only the way those programs are parsed. This is typically the case for separators and associative infix operators, where you want to add a resolver like '((assoc "|")). Another case where this can happen is for the classic dangling else problem, where you will use '((assoc "else" "then")). It can also happen for cases where the conflict is real and cannot really be resolved, but it is unlikely to pose a problem in practice.

Finally, in many cases some conflicts will remain despite all efforts to restructure the grammar. Do not despair: while the parser cannot be made more clever, you can make the lexer as smart as you want. So, the solution is then to look at the tokens involved in the conflict and to split one of those tokens into 2 (or more) different tokens. E.g., if the grammar needs to distinguish between two incompatible uses of the token "begin", make the lexer return different tokens (say "begin-fun" and "begin-plain") depending on which kind of "begin" it finds. This pushes the work of distinguishing the different cases to the lexer, which will thus have to look at the surrounding text to find ad-hoc clues.


24.7.1.6 Specifying Indentation Rules

Based on the provided grammar, SMIE will be able to provide automatic indentation without any extra effort. But in practice, this default indentation style will probably not be good enough. You will want to tweak it in many different cases.

SMIE indentation is based on the idea that indentation rules should be as local as possible. To this end, it relies on the idea of virtual indentation, which is the indentation that a particular program point would have if it were at the beginning of a line. Of course, if that program point is indeed at the beginning of a line, its virtual indentation is its current indentation. But if not, then SMIE uses the indentation algorithm to compute the virtual indentation of that point. Now in practice, the virtual indentation of a program point does not have to be identical to the indentation it would have if we inserted a newline before it. To see how this works, the SMIE rule for indentation after a { in C does not care whether the { is standing on a line of its own or is at the end of the preceding line. Instead, these different cases are handled in the indentation rule that decides how to indent before a {.

Another important concept is the notion of parent: The parent of a token, is the head token of the nearest enclosing syntactic construct. For example, the parent of an else is the if to which it belongs, and the parent of an if, in turn, is the lead token of the surrounding construct. The command backward-sexp jumps from a token to its parent, but there are some caveats: for openers (tokens which start a construct, like if), you need to start with point before the token, while for others you need to start with point after the token. backward-sexp stops with point before the parent token if that is the opener of the token of interest, and otherwise it stops with point after the parent token.

SMIE indentation rules are specified using a function that takes two arguments method and arg where the meaning of arg and the expected return value depend on method.

method can be:

  • :after, in which case arg is a token and the function should return the offset to use for indentation after arg.
  • :before, in which case arg is a token and the function should return the offset to use to indent arg itself.
  • :elem, in which case the function should return either the offset to use to indent function arguments (if arg is the symbol arg) or the basic indentation step (if arg is the symbol basic).
  • :list-intro, in which case arg is a token and the function should return non-nil if the token is followed by a list of expressions (not separated by any token) rather than an expression.

When arg is a token, the function is called with point just before that token. A return value of nil always means to fallback on the default behavior, so the function should return nil for arguments it does not expect.

offset can be:

  • nil: use the default indentation rule.
  • (column . column): indent to column column.
  • number: offset by number, relative to a base token which is the current token for :after and its parent for :before.

24.7.1.7 Helper Functions for Indentation Rules

SMIE provides various functions designed specifically for use in the indentation rules function (several of those functions break if used in another context). These functions all start with the prefix smie-rule-.

Function: smie-rule-bolp

Return non-nil if the current token is the first on the line.

Function: smie-rule-hanging-p

Return non-nil if the current token is hanging. A token is hanging if it is the last token on the line and if it is preceded by other tokens: a lone token on a line is not hanging.

Function: smie-rule-next-p &rest tokens

Return non-nil if the next token is among tokens.

Function: smie-rule-prev-p &rest tokens

Return non-nil if the previous token is among tokens.

Function: smie-rule-parent-p &rest parents

Return non-nil if the current token’s parent is among parents.

Function: smie-rule-sibling-p

Return non-nil if the current token’s parent is actually a sibling. This is the case for example when the parent of a "," is just the previous ",".

Function: smie-rule-parent &optional offset

Return the proper offset to align the current token with the parent. If non-nil, offset should be an integer giving an additional offset to apply.

Function: smie-rule-separator method

Indent current token as a separator.

By separator, we mean here a token whose sole purpose is to separate various elements within some enclosing syntactic construct, and which does not have any semantic significance in itself (i.e., it would typically not exist as a node in an abstract syntax tree).

Such a token is expected to have an associative syntax and be closely tied to its syntactic parent. Typical examples are "," in lists of arguments (enclosed inside parentheses), or ";" in sequences of instructions (enclosed in a {...} or begin...end block).

method should be the method name that was passed to smie-rules-function.


24.7.1.8 Sample Indentation Rules

Here is an example of an indentation function:

(defun sample-smie-rules (kind token)
  (pcase (cons kind token)
    (`(:elem . basic) sample-indent-basic)
    (`(,_ . ",") (smie-rule-separator kind))
    (`(:after . ":=") sample-indent-basic)
    (`(:before . ,(or `"begin" `"(" `"{"))
     (if (smie-rule-hanging-p) (smie-rule-parent)))
    (`(:before . "if")
     (and (not (smie-rule-bolp)) (smie-rule-prev-p "else")
          (smie-rule-parent)))))

A few things to note:

  • The first case indicates the basic indentation increment to use. If sample-indent-basic is nil, then SMIE uses the global setting smie-indent-basic. The major mode could have set smie-indent-basic buffer-locally instead, but that is discouraged.
  • The rule for the token "," make SMIE try to be more clever when the comma separator is placed at the beginning of lines. It tries to outdent the separator so as to align the code after the comma; for example:
    x = longfunctionname (
            arg1
          , arg2
        );
    
  • The rule for indentation after ":=" exists because otherwise SMIE would treat ":=" as an infix operator and would align the right argument with the left one.
  • The rule for indentation before "begin" is an example of the use of virtual indentation: This rule is used only when "begin" is hanging, which can happen only when "begin" is not at the beginning of a line. So this is not used when indenting "begin" itself but only when indenting something relative to this "begin". Concretely, this rule changes the indentation from:
        if x > 0 then begin
                dosomething(x);
            end
    

    to

        if x > 0 then begin
            dosomething(x);
        end
    
  • The rule for indentation before "if" is similar to the one for "begin", but where the purpose is to treat "else if" as a single unit, so as to align a sequence of tests rather than indent each test further to the right. This function does this only in the case where the "if" is not placed on a separate line, hence the smie-rule-bolp test.

    If we know that the "else" is always aligned with its "if" and is always at the beginning of a line, we can use a more efficient rule:

    ((equal token "if")
     (and (not (smie-rule-bolp))
          (smie-rule-prev-p "else")
          (save-excursion
            (sample-smie-backward-token)
            (cons 'column (current-column)))))
    

    The advantage of this formulation is that it reuses the indentation of the previous "else", rather than going all the way back to the first "if" of the sequence.


24.7.1.9 Customizing Indentation

If you are using a mode whose indentation is provided by SMIE, you can customize the indentation to suit your preferences. You can do this on a per-mode basis (using the option smie-config), or a per-file basis (using the function smie-config-local in a file-local variable specification).

User Option: smie-config

This option lets you customize indentation on a per-mode basis. It is an alist with elements of the form (mode . rules). For the precise form of rules, see the variable’s documentation; but you may find it easier to use the command smie-config-guess.

Command: smie-config-guess

This command tries to work out appropriate settings to produce your preferred style of indentation. Simply call the command while visiting a file that is indented with your style.

Command: smie-config-save

Call this command after using smie-config-guess, to save your settings for future sessions.

Command: smie-config-show-indent &optional move

This command displays the rules that are used to indent the current line.

Command: smie-config-set-indent

This command adds a local rule to adjust the indentation of the current line.

Function: smie-config-local rules

This function adds rules as indentation rules for the current buffer. These add to any mode-specific rules defined by the smie-config option. To specify custom indentation rules for a specific file, add an entry to the file’s local variables of the form: eval: (smie-config-local '(rules)).


24.7.2 Parser-based Indentation

When built with the tree-sitter library (see Parsing Program Source), Emacs is capable of parsing the program source and producing a syntax tree. This syntax tree can be used for guiding the program source indentation commands. For maximum flexibility, it is possible to write a custom indentation function that queries the syntax tree and indents accordingly for each language, but that is a lot of work. It is more convenient to use the simple indentation engine described below: then the major mode needs only write some indentation rules, and the engine takes care of the rest.

To enable the parser-based indentation engine, either set treesit-simple-indent-rules and call treesit-major-mode-setup, or equivalently, set the value of indent-line-function to treesit-indent.

Variable: treesit-indent-function

This variable stores the actual function called by treesit-indent. By default, its value is treesit-simple-indent. In the future we might add other, more complex indentation engines.

Writing indentation rules

Variable: treesit-simple-indent-rules

This local variable stores indentation rules for every language. It is an alist with elements of the form (language . rules), where language is a language symbol, and rules is a list with elements of the form (matcher anchor offset).

First, Emacs passes the smallest tree-sitter node at the beginning of the current line to matcher; if it returns non-nil, this rule is applicable. Then Emacs passes the node to anchor, which returns a buffer position. Emacs takes the column number of that position, adds offset to it, and the result is the indentation column for the current line.

The matcher and anchor are functions, and Emacs provides convenient defaults for them.

Each matcher or anchor is a function that takes three arguments: node, parent, and bol. The argument bol is the buffer position whose indentation is required: the position of the first non-whitespace character after the beginning of the line. The argument node is the largest node that starts at that position (and is not a root node); and parent is the parent of node. However, when that position is in a whitespace or inside a multi-line string, no node can start at that position, so node is nil. In that case, parent would be the smallest node that spans that position.

matcher should return non-nil if the rule is applicable, and anchor should return a buffer position.

offset can be an integer, a variable whose value is an integer, or a function that returns an integer. If it is a function, it is passed node, parent, and bol, like matchers and anchors.

Variable: treesit-simple-indent-presets

This is a list of defaults for matchers and anchors in treesit-simple-indent-rules. Each of them represents a function that takes 3 arguments: node, parent, and bol. The available default functions are:

no-node

This matcher is a function that is called with 3 arguments: node, parent, and bol. It returns non-nil, indicating a match, if node is nil, i.e., there is no node that starts at bol. This is the case when bol is on an empty line or inside a multi-line string, etc.

parent-is

This matcher is a function of one argument, type; it returns a function that is called with 3 arguments: node, parent, and bol, and returns non-nil (i.e., a match) if parent’s type matches regexp type.

node-is

This matcher is a function of one argument, type; it returns a function that is called with 3 arguments: node, parent, and bol, and returns non-nil if node’s type matches regexp type.

field-is

This matcher is a function of one argument, name; it returns a function that is called with 3 arguments: node, parent, and bol, and returns non-nil if node’s field name in parent matches regexp name.

query

This matcher is a function of one argument, query; it returns a function that is called with 3 arguments: node, parent, and bol, and returns non-nil if querying parent with query captures node (see Pattern Matching Tree-sitter Nodes).

match

This matcher is a function of 5 arguments: node-type, parent-type, node-field, node-index-min, and node-index-max). It returns a function that is called with 3 arguments: node, parent, and bol, and returns non-nil if node’s type matches regexp node-type, parent’s type matches regexp parent-type, node’s field name in parent matches regexp node-field, and node’s index among its siblings is between node-index-min and node-index-max. If the value of an argument is nil, this matcher doesn’t check that argument. For example, to match the first child where parent is argument_list, use

(match nil "argument_list" nil 0 0)

In addition, node-type can be a special value null, which matches when the value of node is nil.

n-p-gp

Short for “node-parent-grandparent”, this matcher is a function of 3 arguments: node-type, parent-type, and grandparent-type. It returns a function that is called with 3 arguments: node, parent, and bol, and returns non-nil if: (1) node-type matches node’s type, and (2) parent-type matches parent’s type, and (3) grandparent-type matches parent’s parent’s type. If any of node-type, parent-type, and grandparent-type is nil, this function doesn’t check for it.

comment-end

This matcher is a function that is called with 3 arguments: node, parent, and bol, and returns non-nil if point is before a comment-ending token. Comment-ending tokens are defined by regexp comment-end-skip.

catch-all

This matcher is a function that is called with 3 arguments: node, parent, and bol. It always returns non-nil, indicating a match.

first-sibling

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the start of the first child of parent.

nth-sibling

This anchor is a function of two arguments: n, and an optional argument named. It returns a function that is called with 3 arguments: node, parent, and bol, and returns the start of the nth child of parent. If named is non-nil, only named children are counted (see named node).

parent

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the start of parent.

grand-parent

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the start of parent’s parent.

great-grand-parent

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the start of parent’s parent’s parent.

parent-bol

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the first non-space character on the line which parent’s start is on.

standalone-parent

This anchor is a function that is called with 3 arguments: node, parent, and bol. It finds the first ancestor node (parent, grandparent, etc.) of node that starts on its own line, and return the start of that node. “Starting on its own line” means there is only whitespace character before the node on the line which the node’s start is on.

prev-sibling

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the start of the previous sibling of node.

no-indent

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the start of node.

prev-line

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the first non-whitespace character on the previous line.

column-0

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the beginning of the current line, which is at column 0.

comment-start

This anchor is a function that is called with 3 arguments: node, parent, and bol, and returns the position after the comment-start token. Comment-start tokens are defined by regular expression comment-start-skip. This function assumes parent is the comment node.

prev-adaptive-prefix

This anchor is a function that is called with 3 arguments: node, parent, and bol. It tries to match adaptive-fill-regexp to the text at the beginning of the previous non-empty line. If there is a match, this function returns the end of the match, otherwise it returns nil. However, if the current line begins with a prefix (e.g., ‘-’), return the beginning of the prefix of the previous line instead, so that the two prefixes align. This anchor is useful for an indent-relative-like indent behavior for block comments.

Indentation utilities

Here are some utility functions that can help writing parser-based indentation rules.

Command: treesit-check-indent mode

This command checks the current buffer’s indentation against major mode mode. It indents the current buffer according to mode and compares the results with the current indentation. Then it pops up a buffer showing the differences. Correct indentation (target) is shown in green color, current indentation is shown in red color.

It is also helpful to use treesit-inspect-mode (see Tree-sitter Language Grammar) when writing indentation rules.


24.8 Desktop Save Mode

Desktop Save Mode is a feature to save the state of Emacs from one session to another. The user-level commands for using Desktop Save Mode are described in the GNU Emacs Manual (see Saving Emacs Sessions in the GNU Emacs Manual). Modes whose buffers visit a file, don’t have to do anything to use this feature.

For buffers not visiting a file to have their state saved, the major mode must bind the buffer local variable desktop-save-buffer to a non-nil value.

Variable: desktop-save-buffer

If this buffer-local variable is non-nil, the buffer will have its state saved in the desktop file at desktop save. If the value is a function, it is called at desktop save with argument desktop-dirname, and its value is saved in the desktop file along with the state of the buffer for which it was called. When file names are returned as part of the auxiliary information, they should be formatted using the call

(desktop-file-name file-name desktop-dirname)

For buffers not visiting a file to be restored, the major mode must define a function to do the job, and that function must be listed in the alist desktop-buffer-mode-handlers.

Variable: desktop-buffer-mode-handlers

Alist with elements

(major-mode . restore-buffer-function)

The function restore-buffer-function will be called with argument list

(buffer-file-name buffer-name desktop-buffer-misc)

and it should return the restored buffer. Here desktop-buffer-misc is the value returned by the function optionally bound to desktop-save-buffer.


25 Documentation

GNU Emacs has convenient built-in help facilities, most of which derive their information from documentation strings associated with functions and variables. This chapter describes how to access documentation strings in Lisp programs.

The contents of a documentation string should follow certain conventions. In particular, its first line should be a complete sentence (or two complete sentences) that briefly describes what the function or variable does. See Tips for Documentation Strings, for how to write good documentation strings.

Note that the documentation strings for Emacs are not the same thing as the Emacs manual. Manuals have their own source files, written in the Texinfo language; documentation strings are specified in the definitions of the functions and variables they apply to. A collection of documentation strings is not sufficient as a manual because a good manual is not organized in that fashion; it is organized in terms of topics of discussion.

For commands to display documentation strings, see Help in The GNU Emacs Manual.


25.1 Documentation Basics

A documentation string is written using the Lisp syntax for strings, with double-quote characters surrounding the text. It is, in fact, an actual Lisp string. When the string appears in the proper place in a function or variable definition, it serves as the function’s or variable’s documentation.

In a function definition (a lambda or defun form), the documentation string is specified after the argument list, and is normally stored directly in the function object. See Documentation Strings of Functions. You can also put function documentation in the function-documentation property of a function name (see Access to Documentation Strings).

In a variable definition (a defvar form), the documentation string is specified after the initial value. See Defining Global Variables. The string is stored in the variable’s variable-documentation property.

Sometimes, Emacs does not keep documentation strings in memory. There are two such circumstances. Firstly, to save memory, the documentation for primitive functions (see What Is a Function?) and built-in variables is kept in a file named DOC, in the directory specified by doc-directory (see Access to Documentation Strings). Secondly, when a function or variable is loaded from a byte-compiled file, Emacs avoids loading its documentation string (see Documentation Strings and Compilation). In both cases, Emacs looks up the documentation string from the file only when needed, such as when the user calls C-h f (describe-function) for a function.

Documentation strings can contain special key substitution sequences, referring to key bindings which are looked up only when the user views the documentation. This allows the help commands to display the correct keys even if a user rearranges the default key bindings. See Substituting Key Bindings in Documentation.

In the documentation string of an autoloaded command (see Autoload), these key-substitution sequences have an additional special effect: they cause C-h f on the command to trigger autoloading. (This is needed for correctly setting up the hyperlinks in the *Help* buffer.)


25.2 Access to Documentation Strings

Function: documentation-property symbol property &optional verbatim

This function returns the documentation string recorded in symbol’s property list under property property. It is most often used to look up the documentation strings of variables, for which property is variable-documentation. However, it can also be used to look up other kinds of documentation, such as for customization groups (but for function documentation, use the documentation function, below).

If the property value refers to a documentation string stored in the DOC file or a byte-compiled file, this function looks up that string and returns it.

If the property value isn’t nil, isn’t a string, and doesn’t refer to text in a file, then it is evaluated as a Lisp expression to obtain a string.

Finally, this function passes the string through substitute-command-keys to substitute key bindings (see Substituting Key Bindings in Documentation). It skips this step if verbatim is non-nil.

(documentation-property 'command-line-processed
   'variable-documentation)
     ⇒ "Non-nil once command line has been processed"
(symbol-plist 'command-line-processed)
     ⇒ (variable-documentation 188902)
(documentation-property 'emacs 'group-documentation)
     ⇒ "Customization of the One True Editor."
Function: documentation function &optional verbatim

This function returns the documentation string of function. It handles macros, named keyboard macros, and special forms, as well as ordinary functions.

If function is a symbol, this function first looks for the function-documentation property of that symbol; if that has a non-nil value, the documentation comes from that value (if the value is not a string, it is evaluated).

If function is not a symbol, or if it has no function-documentation property, then documentation extracts the documentation string from the actual function definition, reading it from a file if called for.

Finally, unless verbatim is non-nil, this function calls substitute-command-keys. The result is the documentation string to return.

The documentation function signals a void-function error if function has no function definition. However, it is OK if the function definition has no documentation string. In that case, documentation returns nil.

Function: function-documentation function

Generic function used by documentation to extract the raw docstring from a function object. You can specify how to get the docstring of a specific function type by adding a corresponding method to it.

Function: face-documentation face

This function returns the documentation string of face as a face.

Here is an example of using the two functions, documentation and documentation-property, to display the documentation strings for several symbols in a *Help* buffer.

(defun describe-symbols (pattern)
  "Describe the Emacs Lisp symbols matching PATTERN.
All symbols that have PATTERN in their name are described
in the *Help* buffer."
  (interactive "sDescribe symbols matching: ")
  (let ((describe-func
         (lambda (s)
           ;; Print description of symbol.
           (if (fboundp s)             ; It is a function.
               (princ
                (format "%s\t%s\n%s\n\n" s
                  (if (commandp s)
                      (let ((keys (where-is-internal s)))
                        (if keys
                            (concat
                             "Keys: "
                             (mapconcat 'key-description
                                        keys " "))
                          "Keys: none"))
                    "Function")
                  (or (documentation s)
                      "not documented"))))

           (if (boundp s)              ; It is a variable.
               (princ
                (format "%s\t%s\n%s\n\n" s
                  (if (custom-variable-p s)
                      "Option " "Variable")
                  (or (documentation-property
                        s 'variable-documentation)
                      "not documented"))))))
        sym-list)

    ;; Build a list of symbols that match pattern.
    (mapatoms (lambda (sym)
                (if (string-match pattern (symbol-name sym))
                    (setq sym-list (cons sym sym-list)))))

    ;; Display the data.
    (help-setup-xref (list 'describe-symbols pattern)
                 (called-interactively-p 'interactive))
    (with-help-window (help-buffer)
      (mapcar describe-func (sort sym-list 'string<)))))

The describe-symbols function works like apropos, but provides more information.

(describe-symbols "goal")

---------- Buffer: *Help* ----------
goal-column     Option
Semipermanent goal column for vertical motion, as set by …

minibuffer-temporary-goal-position      Variable
not documented

set-goal-column Keys: C-x C-n
Set the current horizontal position as a goal for C-n and C-p.
Those commands will move to this position in the line moved to
rather than trying to keep the same horizontal position.
With a non-nil argument ARG, clears out the goal column
so that C-n and C-p resume vertical motion.
The goal column is stored in the variable ‘goal-column’.

(fn ARG)

temporary-goal-column   Variable
Current goal column for vertical motion.
It is the column where point was at the start of the current run
of vertical motion commands.

When moving by visual lines via the function ‘line-move-visual’, it is a cons
cell (COL . HSCROLL), where COL is the x-position, in pixels,
divided by the default column width, and HSCROLL is the number of
columns by which window is scrolled from left margin.

When the ‘track-eol’ feature is doing its job, the value is
‘most-positive-fixnum’.
---------- Buffer: *Help* ----------
Function: Snarf-documentation filename

This function is used when building Emacs, just before the runnable Emacs is dumped. It finds the positions of the documentation strings stored in the file filename, and records those positions into memory in the function definitions and variable property lists. See Building Emacs.

Emacs reads the file filename from the emacs/etc directory. When the dumped Emacs is later executed, the same file will be looked for in the directory doc-directory. Usually filename is "DOC".

Variable: doc-directory

This variable holds the name of the directory which should contain the file "DOC" that contains documentation strings for built-in functions and variables.

In most cases, this is the same as data-directory. They may be different when you run Emacs from the directory where you built it, without actually installing it. See Definition of data-directory.


25.3 Substituting Key Bindings in Documentation

When documentation strings refer to key sequences, they should use the current, actual key bindings. They can do so using certain special text sequences described below. Accessing documentation strings in the usual way substitutes current key binding information for these special sequences. This works by calling substitute-command-keys. You can also call that function yourself.

Here is a list of the special sequences and what they mean:

\[command]

stands for a key sequence that will invoke command, or ‘M-x command’ if command has no key bindings.

\{mapvar}

stands for a summary of the keymap which is the value of the variable mapvar. The summary is made using describe-bindings.

\<mapvar>

stands for no text itself. It is used only for a side effect: it specifies mapvar’s value as the keymap for any following ‘\[command]’ sequences in this documentation string.

\`KEYSEQ'

stands for a key sequence KEYSEQ, which will use the same face as a command substitution. This should be used only when a key sequence has no corresponding command, for example when it is read directly with read-key-sequence. It must be a valid key sequence according to key-valid-p. It can also be used with command names, like ‘\`M-x foo'’, where you want this to be fontified like a keyboard sequence, but you want to inhibit translating it into a key sequence like ‘\[foo]’ does.

`

(grave accent) stands for a left quote. This generates a left single quotation mark, an apostrophe, or a grave accent depending on the value of text-quoting-style. See Text Quoting Style.

'

(apostrophe) stands for a right quote. This generates a right single quotation mark or an apostrophe depending on the value of text-quoting-style.

\=

quotes the following character and is discarded; thus, ‘\=`’ puts ‘`’ into the output, ‘\=\[’ puts ‘\[’ into the output, and ‘\=\=’ puts ‘\=’ into the output.

\+

This indicates that the symbol directly following should not be marked as link in the *Help* buffer.

Please note: Each ‘\’ must be doubled when written in a string in Emacs Lisp.

Function: substitute-command-keys string &optional no-face include-menus

This function scans string for the above special sequences and replaces them by what they stand for, returning the result as a string. This permits display of documentation that refers accurately to the user’s own customized key bindings. By default, the key bindings are given a special face help-key-binding, but if the optional argument no-face is non-nil, the function doesn’t add this face to the produced string.

Function: substitute-quotes string

This function works like substitute-command-keys, but only replaces quote characters.

If a command has multiple bindings, this function normally uses the first one it finds. You can specify one particular key binding by assigning an :advertised-binding symbol property to the command, like this:

(put 'undo :advertised-binding [?\C-/])

The :advertised-binding property also affects the binding shown in menu items (see The Menu Bar). The property is ignored if it specifies a key binding that the command does not actually have.

Here are examples of the special sequences:

(substitute-command-keys
   "To abort recursive edit, type `\\[abort-recursive-edit]'.")
⇒ "To abort recursive edit, type ‘C-]’."

(substitute-command-keys
   "The keys that are defined for the minibuffer here are:
  \\{minibuffer-local-must-match-map}")
⇒ "The keys that are defined for the minibuffer here are:

?               minibuffer-completion-help
SPC             minibuffer-complete-word
TAB             minibuffer-complete
C-j             minibuffer-complete-and-exit
RET             minibuffer-complete-and-exit
C-g             abort-recursive-edit
"

The keymap description will normally exclude menu items, but if
include-menus is non-nil, include them.

(substitute-command-keys
   "To abort a recursive edit from the minibuffer, type \
`\\<minibuffer-local-must-match-map>\\[abort-recursive-edit]'.")
⇒ "To abort a recursive edit from the minibuffer, type ‘C-g’."

There are other special conventions for the text in documentation strings—for instance, you can refer to functions, variables, and sections of this manual. See Tips for Documentation Strings, for details.


25.4 Text Quoting Style

Typically, grave accents and apostrophes are treated specially in documentation strings and diagnostic messages, and translate to matching single quotation marks (also called “curved quotes”). For example, the documentation string "Alias for `foo'." and the function call (message "Alias for `foo'.") both translate to "Alias for ‘foo’.". Less commonly, Emacs displays grave accents and apostrophes as themselves, or as apostrophes only (e.g., "Alias for 'foo'."). Documentation strings and message formats should be written so that they display well with any of these styles. For example, the documentation string "Alias for 'foo'." is probably not what you want, as it can display as "Alias for ’foo’.", an unusual style in English.

Sometimes you may need to display a grave accent or apostrophe without translation, regardless of text quoting style. In a documentation string, you can do this with escapes. For example, in the documentation string "\\=`(a ,(sin 0)) ==> (a 0.0)" the grave accent is intended to denote Lisp code, so it is escaped and displays as itself regardless of quoting style. In a call to message or error, you can avoid translation by using a format "%s" with an argument that is a call to format. For example, (message "%s" (format "`(a ,(sin %S)) ==> (a %S)" x (sin x))) displays a message that starts with grave accent regardless of text quoting style.

User Option: text-quoting-style

The value of this user option is a symbol that specifies the style Emacs should use for single quotes in the wording of help and messages. If the option’s value is curve, the style is ‘like this’ with curved single quotes. If the value is straight, the style is 'like this' with straight apostrophes. If the value is grave, quotes are not translated and the style is `like this' with grave accent and apostrophe, the standard style before Emacs version 25. The default value nil acts like curve if curved single quotes seem to be displayable, and like grave otherwise.

This option is useful on platforms that have problems with curved quotes. You can customize it freely according to your personal preference.

Function: text-quoting-style

You should not read the value of the variable text-quoting-style directly. Instead, use this function with the same name to dynamically compute the correct quoting style on the current terminal in the nil case described above.


25.5 Describing Characters for Help Messages

These functions convert events, key sequences, or characters to textual descriptions. These descriptions are useful for including arbitrary text characters or key sequences in messages, because they convert non-printing and whitespace characters to sequences of printing characters. The description of a non-whitespace printing character is the character itself.

Function: key-description sequence &optional prefix

This function returns a string containing the Emacs standard notation for the input events in sequence. If prefix is non-nil, it is a sequence of input events leading up to sequence and is included in the return value. Both arguments may be strings, vectors or lists. See Input Events, for more information about valid events.

(key-description [?\M-3 delete])
     ⇒ "M-3 <delete>"
(key-description [delete] "\M-3")
     ⇒ "M-3 <delete>"

See also the examples for single-key-description, below.

Function: single-key-description event &optional no-angles

This function returns a string describing event in the standard Emacs notation for keyboard input. A normal printing character appears as itself, but a control character turns into a string starting with ‘C-’, a meta character turns into a string starting with ‘M-’, and space, tab, etc., appear as ‘SPC’, ‘TAB’, etc. A function key symbol appears inside angle brackets ‘<…>’. An event that is a list appears as the name of the symbol in the CAR of the list, inside angle brackets.

If the optional argument no-angles is non-nil, the angle brackets around function keys and event symbols are omitted; this is for compatibility with old versions of Emacs which didn’t use the brackets.

(single-key-description ?\C-x)
     ⇒ "C-x"
(key-description "\C-x \M-y \n \t \r \f123")
     ⇒ "C-x SPC M-y SPC C-j SPC TAB SPC RET SPC C-l 1 2 3"
(single-key-description 'delete)
     ⇒ "<delete>"
(single-key-description 'C-mouse-1)
     ⇒ "C-<mouse-1>"
(single-key-description 'C-mouse-1 t)
     ⇒ "C-mouse-1"
Function: text-char-description character

This function returns a string describing character in the standard Emacs notation for characters that can appear in text—similar to single-key-description, except that the argument must be a valid character code that passes a characterp test (see Character Codes). The function produces descriptions of control characters with a leading caret (which is how Emacs usually displays control characters in buffers). Characters with modifier bits will cause this function to signal an error (ASCII characters with the Control modifier are an exception, they are represented as control characters).

(text-char-description ?\C-c)
     ⇒ "^C"
(text-char-description ?\M-m)
     error→ Wrong type argument: characterp, 134217837
Command: read-kbd-macro string &optional need-vector

This function is used mainly for operating on keyboard macros, but it can also be used as a rough inverse for key-description. You call it with a string containing key descriptions, separated by spaces; it returns a string or vector containing the corresponding events. (This may or may not be a single valid key sequence, depending on what events you use; see Key Sequences.) If need-vector is non-nil, the return value is always a vector.


25.6 Help Functions

Emacs provides a variety of built-in help functions, all accessible to the user as subcommands of the prefix C-h. For more information about them, see Help in The GNU Emacs Manual. Here we describe some program-level interfaces to the same information.

Command: apropos pattern &optional do-all

This function finds all meaningful symbols whose names contain a match for the apropos pattern pattern. An apropos pattern is either a word to match, a space-separated list of words of which at least two must match, or a regular expression (if any special regular expression characters occur). A symbol is meaningful if it has a definition as a function, variable, or face, or has properties.

The function returns a list of elements that look like this:

(symbol score function-doc variable-doc
 plist-doc widget-doc face-doc group-doc)

Here, score is an integer measure of how important the symbol seems to be as a match. Each of the remaining elements is a documentation string, or nil, for symbol as a function, variable, etc.

It also displays the symbols in a buffer named *Apropos*, each with a one-line description taken from the beginning of its documentation string.

If do-all is non-nil, or if the user option apropos-do-all is non-nil, then apropos also shows key bindings for the functions that are found; it also shows all interned symbols, not just meaningful ones (and it lists them in the return value as well).

Variable: help-map

The value of this variable is a local keymap for characters following the Help key, C-h.

Prefix Command: help-command

This symbol is not a function; its function definition cell holds the keymap known as help-map. It is defined in help.el as follows:

(keymap-set global-map (key-description (string help-char)) 'help-command)
(fset 'help-command help-map)
User Option: help-char

The value of this variable is the help character—the character that Emacs recognizes as meaning Help. By default, its value is 8, which stands for C-h. When Emacs reads this character, if help-form is a non-nil Lisp expression, it evaluates that expression, and displays the result in a window if it is a string.

Usually the value of help-form is nil. Then the help character has no special meaning at the level of command input, and it becomes part of a key sequence in the normal way. The standard key binding of C-h is a prefix key for several general-purpose help features.

The help character is special after prefix keys, too. If it has no binding as a subcommand of the prefix key, it runs describe-prefix-bindings, which displays a list of all the subcommands of the prefix key.

User Option: help-event-list

The value of this variable is a list of event types that serve as alternative help characters. These events are handled just like the event specified by help-char.

Variable: help-form

If this variable is non-nil, its value is a form to evaluate whenever the character help-char is read. If evaluating the form produces a string, that string is displayed.

A command that calls read-event, read-char-choice, read-char, read-char-from-minibuffer, or y-or-n-p probably should bind help-form to a non-nil expression while it does input. (The time when you should not do this is when C-h has some other meaning.) Evaluating this expression should result in a string that explains what the input is for and how to enter it properly.

Entry to the minibuffer binds this variable to the value of minibuffer-help-form (see Definition of minibuffer-help-form).

Variable: prefix-help-command

This variable holds a function to print help for a prefix key. The function is called when the user types a prefix key followed by the help character, and the help character has no binding after that prefix. The variable’s default value is describe-prefix-bindings.

Command: describe-prefix-bindings

This function calls describe-bindings to display a list of all the subcommands of the prefix key of the most recent key sequence. The prefix described consists of all but the last event of that key sequence. (The last event is, presumably, the help character.)

The following two functions are meant for modes that want to provide help without relinquishing control, such as the electric modes. Their names begin with ‘Helper’ to distinguish them from the ordinary help functions.

Command: Helper-describe-bindings

This command pops up a window displaying a help buffer containing a listing of all of the key bindings from both the local and global keymaps. It works by calling describe-bindings.

Command: Helper-help

This command provides help for the current mode. It prompts the user in the minibuffer with the message ‘Help (Type ? for further options)’, and then provides assistance in finding out what the key bindings are, and what the mode is intended for. It returns nil.

This can be customized by changing the map Helper-help-map.

Variable: data-directory

This variable holds the name of the directory in which Emacs finds certain documentation and text files that come with Emacs.

Function: help-buffer

This function returns the name of the help buffer, which is normally *Help*; if such a buffer does not exist, it is first created.

Macro: with-help-window buffer-or-name body…

This macro evaluates body like with-output-to-temp-buffer (see Temporary Displays), inserting any output produced by its forms into a buffer specified by buffer-or-name, which can be a buffer or the name of a buffer. (Frequently, buffer-or-name is the value returned by the function help-buffer.) This macro puts the specified buffer into Help mode and displays a message telling the user how to quit and scroll the help window. It selects the help window if the current value of the user option help-window-select has been set accordingly. It returns the last value in body.

Function: help-setup-xref item interactive-p

This function updates the cross reference data in the *Help* buffer, which is used to regenerate the help information when the user clicks on the ‘Back’ or ‘Forward’ buttons. Most commands that use the *Help* buffer should invoke this function before clearing the buffer. The item argument should have the form (function . args), where function is a function to call, with argument list args, to regenerate the help buffer. The interactive-p argument is non-nil if the calling command was invoked interactively; in that case, the stack of items for the *Help* buffer’s ‘Back’ buttons is cleared.

See describe-symbols example, for an example of using help-buffer, with-help-window, and help-setup-xref.

Macro: make-help-screen fname help-line help-text help-map

This macro defines a help command named fname that acts like a prefix key that shows a list of the subcommands it offers.

When invoked, fname displays help-text in a window, then reads and executes a key sequence according to help-map. The string help-text should describe the bindings available in help-map.

The command fname is defined to handle a few events itself, by scrolling the display of help-text. When fname reads one of those special events, it does the scrolling and then reads another event. When it reads an event that is not one of those few, and which has a binding in help-map, it executes that key’s binding and then returns.

The argument help-line should be a single-line summary of the alternatives in help-map. In the current version of Emacs, this argument is used only if you set the option three-step-help to t.

This macro is used in the command help-for-help which is the binding of C-h C-h.

User Option: three-step-help

If this variable is non-nil, commands defined with make-help-screen display their help-line strings in the echo area at first, and display the longer help-text strings only if the user types the help character again.


25.7 Documentation Groups

Emacs can list functions based on various groupings. For instance, string-trim and mapconcat are “string” functions, so M-x shortdoc RET string RET will give an overview of functions that operate on strings.

The documentation groups are created with the define-short-documentation-group macro.

Macro: define-short-documentation-group group &rest functions

Define group as a group of functions, and provide short summaries of using those functions. The optional argument functions is a list whose elements are of the form:

(func [keyword val]…)

The following keywords are recognized:

:eval

The value should be a form that has no side effect when evaluated. The form will be used in the documentation by printing it with prin1 (see Output Functions). However, if the form is a string, it will be inserted as-is, and the string will then be read to yield the form. In any case, the form will then be evaluated, and the result used. For instance:

:eval (concat "foo" "bar" "zot")
:eval "(make-string 5 ?x)"

will result in:

(concat "foo" "bar" "zot")
⇒ "foobarzot"
(make-string 5 ?x)
⇒ "xxxxx"

(The reason for allowing both Lisp forms and strings here is so that printing could be controlled in the few cases where a certain presentation of the form is wished for. In the example, ‘?x’ would otherwise have been printed as ‘120’ if it hadn’t been included in a string.)

:no-eval

This is like :eval, except that the form will not be evaluated. In these cases, a :result element of some kind (see below) should be included.

:no-eval (file-symlink-p "/tmp/foo")
:eg-result t
:no-eval*

Like :no-eval, but always inserts ‘[it depends]’ as the result. For instance:

:no-eval* (buffer-string)

will result in:

(buffer-string)
→ [it depends]
:no-value

Like :no-eval, but is used when the function in question has no well-defined return value, and is used for side effect only.

:result

Used to output the result from non-evaluating example forms.

:no-eval (setcar list 'c)
:result c
:eg-result

Used to output an example result from non-evaluating example forms. For instance:

:no-eval (looking-at "f[0-9]")
:eg-result t

will result in:

(looking-at "f[0-9]")
eg. → t
:result-string
:eg-result-string

These two are the same as :result and :eg-result, respectively, but are inserted as is. This is useful when the result is unreadable or should be of a particular form:

:no-eval (find-file "/tmp/foo")
:eg-result-string "#<buffer foo>"
:no-eval (default-file-modes)
:eg-result-string "#o755"
:no-manual

Indicates that this function is not documented in the manual.

:args

By default, the function’s actual argument list is shown. If :args is present, they are used instead.

:args (regexp string)

Here’s a very short example:

(define-short-documentation-group string
  "Creating Strings"
  (substring
   :eval (substring "foobar" 0 3)
   :eval (substring "foobar" 3))
  (concat
   :eval (concat "foo" "bar" "zot")))

The first argument is the name of the group to be defined, and then follows any number of function descriptions.

A function can belong to any number of documentation groups.

In addition to function descriptions, the list can also have string elements, which are used to divide a documentation group into sections.

Function: shortdoc-add-function group section elem

Lisp packages can add functions to groups with this function. Each elem should be a function description, as described above. group is the function group, and section is what section in the function group to insert the function into.

If group doesn’t exist, it will be created. If section doesn’t exist, it will be added to the end of the function group.


26 Files

This chapter describes the Emacs Lisp functions and variables to find, create, view, save, and otherwise work with files and directories. A few other file-related functions are described in Buffers, and those related to backups and auto-saving are described in Backups and Auto-Saving.

Many of the file functions take one or more arguments that are file names. A file name is a string. Most of these functions expand file name arguments using the function expand-file-name, so that ~ is handled correctly, as are relative file names (including ../ and the empty string). See Functions that Expand Filenames.

In addition, certain magic file names are handled specially. For example, when a remote file name is specified, Emacs accesses the file over the network via an appropriate protocol. See Remote Files in The GNU Emacs Manual. This handling is done at a very low level, so you may assume that all the functions described in this chapter accept magic file names as file name arguments, except where noted. See Making Certain File Names “Magic”, for details.

When file I/O functions signal Lisp errors, they usually use the condition file-error (see Writing Code to Handle Errors). The error message is in most cases obtained from the operating system, according to locale system-messages-locale, and decoded using coding system locale-coding-system (see Locales).


26.1 Visiting Files

Visiting a file means reading a file into a buffer. Once this is done, we say that the buffer is visiting that file, and call the file the visited file of the buffer.

A file and a buffer are two different things. A file is information recorded permanently in the computer (unless you delete it). A buffer, on the other hand, is information inside of Emacs that will vanish at the end of the editing session (or when you kill the buffer). When a buffer is visiting a file, it contains information copied from the file. The copy in the buffer is what you modify with editing commands. Changes to the buffer do not change the file; to make the changes permanent, you must save the buffer, which means copying the altered buffer contents back into the file.

Despite the distinction between files and buffers, people often refer to a file when they mean a buffer and vice-versa. Indeed, we say, “I am editing a file”, rather than, “I am editing a buffer that I will soon save as a file of the same name”. Humans do not usually need to make the distinction explicit. When dealing with a computer program, however, it is good to keep the distinction in mind.


26.1.1 Functions for Visiting Files

This section describes the functions normally used to visit files. For historical reasons, these functions have names starting with ‘find-’ rather than ‘visit-’. See Buffer File Name, for functions and variables that access the visited file name of a buffer or that find an existing buffer by its visited file name.

In a Lisp program, if you want to look at the contents of a file but not alter it, the fastest way is to use insert-file-contents in a temporary buffer. Visiting the file is not necessary and takes longer. See Reading from Files.

Command: find-file filename &optional wildcards

This command selects a buffer visiting the file filename, using an existing buffer if there is one, and otherwise creating a new buffer and reading the file into it. It also returns that buffer.

Aside from some technical details, the body of the find-file function is basically equivalent to:

(switch-to-buffer (find-file-noselect filename nil nil wildcards))

(See switch-to-buffer in Switching to a Buffer in a Window.)

If wildcards is non-nil, which is always true in an interactive call, then find-file expands wildcard characters in filename and visits all the matching files.

When find-file is called interactively, it prompts for filename in the minibuffer.

Command: find-file-literally filename

This command visits filename, like find-file does, but it does not perform any format conversions (see File Format Conversion), character code conversions (see Coding Systems), or end-of-line conversions (see End of line conversion). The buffer visiting the file is made unibyte, and its major mode is Fundamental mode, regardless of the file name. File local variable specifications in the file (see File Local Variables) are ignored, and automatic decompression and adding a newline at the end of the file due to require-final-newline (see require-final-newline) are also disabled.

Note that if Emacs already has a buffer visiting the same file non-literally, it will not visit the same file literally, but instead just switch to the existing buffer. If you want to be sure of accessing a file’s contents literally, you should create a temporary buffer and then read the file contents into it using insert-file-contents-literally (see Reading from Files).

Function: find-file-noselect filename &optional nowarn rawfile wildcards

This function is the guts of all the file-visiting functions. It returns a buffer visiting the file filename. You may make the buffer current or display it in a window if you wish, but this function does not do so.

The function returns an existing buffer if there is one; otherwise it creates a new buffer and reads the file into it. When find-file-noselect uses an existing buffer, it first verifies that the file has not changed since it was last visited or saved in that buffer. If the file has changed, this function asks the user whether to reread the changed file. If the user says ‘yes’, any edits previously made in the buffer are lost.

Reading the file involves decoding the file’s contents (see Coding Systems), including end-of-line conversion, and format conversion (see File Format Conversion). If wildcards is non-nil, then find-file-noselect expands wildcard characters in filename and visits all the matching files.

This function displays warning or advisory messages in various peculiar cases, unless the optional argument nowarn is non-nil. For example, if it needs to create a buffer, and there is no file named filename, it displays the message ‘(New file)’ in the echo area, and leaves the buffer empty.

The find-file-noselect function normally calls after-find-file after reading the file (see Subroutines of Visiting). That function sets the buffer major mode, parses local variables, warns the user if there exists an auto-save file more recent than the file just visited, and finishes by running the functions in find-file-hook.

If the optional argument rawfile is non-nil, then after-find-file is not called, and the find-file-not-found-functions are not run in case of failure. What’s more, a non-nil rawfile value suppresses coding system conversion and format conversion.

The find-file-noselect function usually returns the buffer that is visiting the file filename. But, if wildcards are actually used and expanded, it returns a list of buffers that are visiting the various files.

(find-file-noselect "/etc/fstab")
     ⇒ #<buffer fstab>
Command: find-file-other-window filename &optional wildcards

This command selects a buffer visiting the file filename, but does so in a window other than the selected window. It may use another existing window or split a window; see Switching to a Buffer in a Window.

When this command is called interactively, it prompts for filename.

Command: find-file-read-only filename &optional wildcards

This command selects a buffer visiting the file filename, like find-file, but it marks the buffer as read-only. See Read-Only Buffers, for related functions and variables.

When this command is called interactively, it prompts for filename.

User Option: find-file-wildcards

If this variable is non-nil, then the various find-file commands check for wildcard characters and visit all the files that match them (when invoked interactively or when their wildcards argument is non-nil). If this option is nil, then the find-file commands ignore their wildcards argument and never treat wildcard characters specially.

User Option: find-file-hook

The value of this variable is a list of functions to be called after a file is visited. The file’s local-variables specification (if any) will have been processed before the hooks are run. The buffer visiting the file is current when the hook functions are run.

This variable is a normal hook. See Hooks.

Variable: find-file-not-found-functions

The value of this variable is a list of functions to be called when find-file or find-file-noselect is passed a nonexistent file name. find-file-noselect calls these functions as soon as it detects a nonexistent file. It calls them in the order of the list, until one of them returns non-nil. buffer-file-name is already set up.

This is not a normal hook because the values of the functions are used, and in many cases only some of the functions are called.

Variable: find-file-literally

This buffer-local variable, if set to a non-nil value, makes save-buffer behave as if the buffer were visiting its file literally, i.e., without conversions of any kind. The command find-file-literally sets this variable’s local value, but other equivalent functions and commands can do that as well, e.g., to avoid automatic addition of a newline at the end of the file. This variable is permanent local, so it is unaffected by changes of major modes.


26.1.2 Subroutines of Visiting

The find-file-noselect function uses two important subroutines which are sometimes useful in user Lisp code: create-file-buffer and after-find-file. This section explains how to use them.

Function: create-file-buffer filename

This function creates a suitably named buffer for visiting filename, and returns it. It uses filename (sans directory) as the name if that name is free; otherwise, it appends a string such as ‘<2>’ to get an unused name. See also Creating Buffers. Note that the uniquify library affects the result of this function. See Uniquify in The GNU Emacs Manual.

Please note: create-file-buffer does not associate the new buffer with a file and does not select the buffer. It also does not use the default major mode.

(create-file-buffer "foo")
     ⇒ #<buffer foo>
(create-file-buffer "foo")
     ⇒ #<buffer foo<2>>
(create-file-buffer "foo")
     ⇒ #<buffer foo<3>>

This function is used by find-file-noselect. It uses generate-new-buffer (see Creating Buffers).

Function: after-find-file &optional error warn noauto after-find-file-from-revert-buffer nomodes

This function sets the buffer major mode, and parses local variables (see How Emacs Chooses a Major Mode). It is called by find-file-noselect and by the default revert function (see Reverting).

If reading the file got an error because the file does not exist, but its directory does exist, the caller should pass a non-nil value for error. In that case, after-find-file issues a warning: ‘(New file)’. For more serious errors, the caller should usually not call after-find-file.

If warn is non-nil, then this function issues a warning if an auto-save file exists and is more recent than the visited file.

If noauto is non-nil, that says not to enable or disable Auto-Save mode. The mode remains enabled if it was enabled before.

If after-find-file-from-revert-buffer is non-nil, that means this call was from revert-buffer. This has no direct effect, but some mode functions and hook functions check the value of this variable.

If nomodes is non-nil, that means don’t alter the buffer’s major mode, don’t process local variables specifications in the file, and don’t run find-file-hook. This feature is used by revert-buffer in some cases.

The last thing after-find-file does is call all the functions in the list find-file-hook.


26.2 Saving Buffers

When you edit a file in Emacs, you are actually working on a buffer that is visiting that file—that is, the contents of the file are copied into the buffer and the copy is what you edit. Changes to the buffer do not change the file until you save the buffer, which means copying the contents of the buffer into the file. Buffers which are not visiting a file can still be “saved”, in a sense, using functions in the buffer-local write-contents-functions hook.

Command: save-buffer &optional backup-option

This function saves the contents of the current buffer in its visited file if the buffer has been modified since it was last visited or saved. Otherwise it does nothing.

save-buffer is responsible for making backup files. Normally, backup-option is nil, and save-buffer makes a backup file only if this is the first save since visiting the file. Other values for backup-option request the making of backup files in other circumstances:

  • With an argument of 4 or 64, reflecting 1 or 3 C-u’s, the save-buffer function marks this version of the file to be backed up when the buffer is next saved.
  • With an argument of 16 or 64, reflecting 2 or 3 C-u’s, the save-buffer function unconditionally backs up the previous version of the file before saving it.
  • With an argument of 0, unconditionally do not make any backup file.
Command: save-some-buffers &optional save-silently-p pred

This command saves some modified file-visiting buffers. Normally it asks the user about each buffer. But if save-silently-p is non-nil, it saves all the file-visiting buffers without querying the user.

The optional pred argument provides a predicate that controls which buffers to ask about (or to save silently if save-silently-p is non-nil). If pred is nil, that means to use the value of save-some-buffers-default-predicate instead of pred. If the result is nil, it means ask only about file-visiting buffers. If it is t, that means also offer to save certain other non-file buffers—those that have a non-nil buffer-local value of buffer-offer-save (see Killing Buffers). A user who says ‘yes’ to saving a non-file buffer is asked to specify the file name to use. The save-buffers-kill-emacs function passes the value t for pred.

If the predicate is neither t nor nil, then it should be a function of no arguments. It will be called in each buffer to decide whether to offer to save that buffer. If it returns a non-nil value in a certain buffer, that means do offer to save that buffer.

Command: write-file filename &optional confirm

This function writes the current buffer into file filename, makes the buffer visit that file, and marks it not modified. Then it renames the buffer based on filename, appending a string like ‘<2>’ if necessary to make a unique buffer name. It does most of this work by calling set-visited-file-name (see Buffer File Name) and save-buffer.

If confirm is non-nil, that means to ask for confirmation before overwriting an existing file. Interactively, confirmation is required, unless the user supplies a prefix argument.

If filename is a directory name (see Directory Names), write-file uses the name of the visited file, in directory filename. If the buffer is not visiting a file, it uses the buffer name instead.

Saving a buffer runs several hooks. It also performs format conversion (see File Format Conversion). Note that these hooks, described below, are only run by save-buffer, they are not run by other primitives and functions that write buffer text to files, and in particular auto-saving (see Auto-Saving) doesn’t run these hooks.

Variable: write-file-functions

The value of this variable is a list of functions to be called before writing out a buffer to its visited file. If one of them returns non-nil, the file is considered already written and the rest of the functions are not called, nor is the usual code for writing the file executed.

If a function in write-file-functions returns non-nil, it is responsible for making a backup file (if that is appropriate). To do so, execute the following code:

(or buffer-backed-up (backup-buffer))

You might wish to save the file modes value returned by backup-buffer and use that (if non-nil) to set the mode bits of the file that you write. This is what save-buffer normally does. See Making Backup Files.

The hook functions in write-file-functions are also responsible for encoding the data (if desired): they must choose a suitable coding system and end-of-line conversion (see Coding Systems in Lisp), perform the encoding (see Explicit Encoding and Decoding), and set last-coding-system-used to the coding system that was used (see Encoding and I/O).

If you set this hook locally in a buffer, it is assumed to be associated with the file or the way the contents of the buffer were obtained. Thus the variable is marked as a permanent local, so that changing the major mode does not alter a buffer-local value. On the other hand, calling set-visited-file-name will reset it. If this is not what you want, you might like to use write-contents-functions instead.

Even though this is not a normal hook, you can use add-hook and remove-hook to manipulate the list. See Hooks.

Variable: write-contents-functions

This works just like write-file-functions, but it is intended for hooks that pertain to the buffer’s contents, not to the particular visited file or its location, and can be used to create arbitrary save processes for buffers that aren’t visiting files at all. Such hooks are usually set up by major modes, as buffer-local bindings for this variable. This variable automatically becomes buffer-local whenever it is set; switching to a new major mode always resets this variable, but calling set-visited-file-name does not.

If any of the functions in this hook returns non-nil, the file is considered already written and the rest are not called and neither are the functions in write-file-functions.

When using this hook to save buffers that are not visiting files (for instance, special-mode buffers), keep in mind that, if the function fails to save correctly and returns a nil value, save-buffer will go on to prompt the user for a file to save the buffer in. If this is undesirable, consider having the function fail by raising an error.

User Option: before-save-hook

This normal hook runs before a buffer is saved in its visited file, regardless of whether that is done normally or by one of the hooks described above. For instance, the copyright.el program uses this hook to make sure the file you are saving has the current year in its copyright notice.

User Option: after-save-hook

This normal hook runs after a buffer has been saved in its visited file.

User Option: file-precious-flag

If this variable is non-nil, then save-buffer protects against I/O errors while saving by writing the new file to a temporary name instead of the name it is supposed to have, and then renaming it to the intended name after it is clear there are no errors. This procedure prevents problems such as a lack of disk space from resulting in an invalid file.

As a side effect, backups are necessarily made by copying. See Backup by Renaming or by Copying?. Yet, at the same time, saving a precious file always breaks all hard links between the file you save and other file names.

Some modes give this variable a non-nil buffer-local value in particular buffers.

User Option: require-final-newline

This variable determines whether files may be written out that do not end with a newline. If the value of the variable is t, then save-buffer silently adds a newline at the end of the buffer whenever it does not already end in one. If the value is visit, Emacs adds a missing newline just after it visits the file. If the value is visit-save, Emacs adds a missing newline both on visiting and on saving. For any other non-nil value, save-buffer asks the user whether to add a newline each time the case arises.

If the value of the variable is nil, then save-buffer doesn’t add newlines at all. nil is the default value, but a few major modes set it to t in particular buffers.

See also the function set-visited-file-name (see Buffer File Name).


26.3 Reading from Files

To copy the contents of a file into a buffer, use the function insert-file-contents. (Don’t use the command insert-file in a Lisp program, as that sets the mark.)

Function: insert-file-contents filename &optional visit beg end replace

This function inserts the contents of file filename into the current buffer after point. It returns a list of the absolute file name and the length of the data inserted. An error is signaled if filename is not the name of a file that can be read.

This function checks the file contents against the defined file formats, and converts the file contents if appropriate and also calls the functions in the list after-insert-file-functions. See File Format Conversion. Normally, one of the functions in the after-insert-file-functions list determines the coding system (see Coding Systems) used for decoding the file’s contents, including end-of-line conversion. However, if the file contains null bytes, it is by default visited without any code conversions. See inhibit-null-byte-detection.

If visit is non-nil, this function additionally marks the buffer as unmodified and sets up various fields in the buffer so that it is visiting the file filename: these include the buffer’s visited file name and its last save file modtime. This feature is used by find-file-noselect and you probably should not use it yourself.

If beg and end are non-nil, they should be numbers that are byte offsets specifying the portion of the file to insert. In this case, visit must be nil. For example,

(insert-file-contents filename nil 0 500)

inserts the characters coded by the first 500 bytes of a file.

If beg or end happens to be in the middle of a character’s multibyte sequence, Emacs’s character code conversion will insert one or more eight-bit characters (a.k.a. “raw bytes”) (see Character Sets) into the buffer. If you want to read part of a file this way, we recommend to bind coding-system-for-read to a suitable value around the call to this function (see Specifying a Coding System for One Operation), and to write Lisp code which will check for raw bytes at the boundaries, read the entire sequence of these bytes, and convert them back to valid characters.

If the argument replace is non-nil, it means to replace the contents of the buffer (actually, just the accessible portion) with the contents of the file. This is better than simply deleting the buffer contents and inserting the whole file, because (1) it preserves some marker positions and (2) it puts less data in the undo list.

It is possible to read a special file (such as a FIFO or an I/O device) with insert-file-contents, as long as replace, and visit and beg are nil. However, you should normally use an end argument for these files to avoid inserting (potentially) unlimited data into the buffer (for instance, when inserting data from /dev/urandom).

Function: insert-file-contents-literally filename &optional visit beg end replace

This function works like insert-file-contents except that each byte in the file is handled separately, being converted into an eight-bit character if needed. It does not run after-insert-file-functions, and does not do format decoding, character code conversion, automatic uncompression, and so on.

If you want to pass a file name to another process so that another program can read the file, use the function file-local-copy; see Making Certain File Names “Magic”.


26.4 Writing to Files

You can write the contents of a buffer, or part of a buffer, directly to a file on disk using the append-to-file and write-region functions. Don’t use these functions to write to files that are being visited; that could cause confusion in the mechanisms for visiting.

Command: append-to-file start end filename

This function appends the contents of the region delimited by start and end in the current buffer to the end of file filename. If that file does not exist, it is created. This function returns nil.

An error is signaled if you cannot write or create filename.

When called from Lisp, this function is completely equivalent to:

(write-region start end filename t)
Command: write-region start end filename &optional append visit lockname mustbenew

This function writes the region delimited by start and end in the current buffer into the file specified by filename.

If start is nil, then the command writes the entire buffer contents (not just the accessible portion) to the file and ignores end.

If start is a string, then write-region writes or appends that string, rather than text from the buffer. end is ignored in this case.

If append is non-nil, then the specified text is appended to the existing file contents (if any). If append is a number, write-region seeks to that byte offset from the start of the file and writes the data from there.

If mustbenew is non-nil, then write-region asks for confirmation if filename names an existing file. If mustbenew is the symbol excl, then write-region does not ask for confirmation, but instead it signals an error file-already-exists if the file already exists. Although write-region normally follows a symbolic link and creates the pointed-to file if the symbolic link is dangling, it does not follow symbolic links if mustbenew is excl.

The test for an existing file, when mustbenew is excl, uses a special system feature. At least for files on a local disk, there is no chance that some other program could create a file of the same name before Emacs does, without Emacs’s noticing.

If visit is t, then Emacs establishes an association between the buffer and the file: the buffer is then visiting that file. It also sets the last file modification time for the current buffer to filename’s modtime, and marks the buffer as not modified. This feature is used by save-buffer, but you probably should not use it yourself.

If visit is a string, it specifies the file name to visit. This way, you can write the data to one file (filename) while recording the buffer as visiting another file (visit). The argument visit is used in the echo area message and also for file locking; visit is stored in buffer-file-name. This feature is used to implement file-precious-flag; don’t use it yourself unless you really know what you’re doing.

The optional argument lockname, if non-nil, specifies the file name to use for purposes of locking and unlocking, overriding filename and visit for that purpose.

The function write-region converts the data which it writes to the appropriate file formats specified by buffer-file-format and also calls the functions in the list write-region-annotate-functions. See File Format Conversion.

Normally, write-region displays the message ‘Wrote filename’ in the echo area. This message is inhibited if visit is neither t nor nil nor a string, or if Emacs is operating in batch mode (see Batch Mode). This feature is useful for programs that use files for internal purposes, files that the user does not need to know about.

Variable: write-region-inhibit-fsync

If this variable’s value is nil, write-region uses the fsync system call after writing a file. Although this slows Emacs down, it lessens the risk of data loss after power failure. If the value is t, Emacs does not use fsync. The default value is nil when Emacs is interactive, and t when Emacs runs in batch mode. See Files and Secondary Storage.

Macro: with-temp-file file body…

The with-temp-file macro evaluates the body forms with a temporary buffer as the current buffer; then, at the end, it writes the buffer contents into file file. It kills the temporary buffer when finished, restoring the buffer that was current before the with-temp-file form. Then it returns the value of the last form in body.

The current buffer is restored even in case of an abnormal exit via throw or error (see Nonlocal Exits).

Like with-temp-buffer (see Current Buffer), the temporary buffer used by this macro does not run the hooks kill-buffer-hook, kill-buffer-query-functions (see Killing Buffers), and buffer-list-update-hook (see The Buffer List).


26.5 File Locks

When two users edit the same file at the same time, they are likely to interfere with each other. Emacs tries to prevent this situation from arising by recording a file lock when a file is being modified. Emacs can then detect the first attempt to modify a buffer visiting a file that is locked by another Emacs job, and ask the user what to do. The file lock is really a file, a symbolic link with a special name, stored in the same directory as the file you are editing. The name is constructed by prepending .# to the file name of the buffer. The target of the symbolic link will be of the form user@host.pid:boot, where user is replaced with the current username (from user-login-name), host with the name of the host where Emacs is running (from system-name), pid with Emacs’s process id, and boot with the time since the last reboot. :boot is omitted if the boot time is unavailable. (On file systems that do not support symbolic links, a regular file is used instead, with contents of the form user@host.pid:boot.)

When you access files using NFS, there may be a small probability that you and another user will both lock the same file simultaneously. If this happens, it is possible for the two users to make changes simultaneously, but Emacs will still warn the user who saves second. Also, the detection of modification of a buffer visiting a file changed on disk catches some cases of simultaneous editing; see Buffer Modification Time.

Function: file-locked-p filename

This function returns nil if the file filename is not locked. It returns t if it is locked by this Emacs process, and it returns the name of the user who has locked it if it is locked by some other job.

(file-locked-p "foo")
     ⇒ nil
Function: lock-buffer &optional filename

This function locks the file filename, if the current buffer is modified. The argument filename defaults to the current buffer’s visited file. Nothing is done if the current buffer is not visiting a file, or is not modified, or if the option create-lockfiles is nil.

Function: unlock-buffer

This function unlocks the file being visited in the current buffer, if the buffer is modified. If the buffer is not modified, then the file should not be locked, so this function does nothing. It also does nothing if the current buffer is not visiting a file, or is not locked. This function handles file system errors by calling display-warning and otherwise ignores the error.

User Option: create-lockfiles

If this variable is nil, Emacs does not lock files.

User Option: lock-file-name-transforms

By default, Emacs creates the lock files in the same directory as the files that are being locked. This can be changed by customizing this variable. Is has the same syntax as auto-save-file-name-transforms (see Auto-Saving). For instance, to make Emacs write all the lock files to /var/tmp/, you could say something like:

(setq lock-file-name-transforms
      '(("\\`/.*/\\([^/]+\\)\\'" "/var/tmp/\\1" t)))
Function: ask-user-about-lock file other-user

This function is called when the user tries to modify file, but it is locked by another user named other-user. The default definition of this function asks the user to say what to do. The value this function returns determines what Emacs does next:

  • A value of t says to grab the lock on the file. Then this user may edit the file and other-user loses the lock.
  • A value of nil says to ignore the lock and let this user edit the file anyway.
  • This function may instead signal a file-locked error, in which case the change that the user was about to make does not take place.

    The error message for this error looks like this:

    error→ File is locked: file other-user
    

    where file is the name of the file and other-user is the name of the user who has locked the file.

If you wish, you can replace the ask-user-about-lock function with your own version that makes the decision in another way.

User Option: remote-file-name-inhibit-locks

You can prevent the creation of remote lock files by setting the variable remote-file-name-inhibit-locks to t.

Command: lock-file-mode

This command, called interactively, toggles the local value of create-lockfiles in the current buffer.


26.6 Information about Files

This section describes the functions for retrieving various types of information about files (or directories or symbolic links), such as whether a file is readable or writable, and its size. These functions all take arguments which are file names. Except where noted, these arguments need to specify existing files, or an error is signaled.

Be careful with file names that end in spaces. On some filesystems (notably, MS-Windows), trailing whitespace characters in file names are silently and automatically ignored.


26.6.1 Testing Accessibility

These functions test for permission to access a file for reading, writing, or execution. Unless explicitly stated otherwise, they follow symbolic links. See Distinguishing Kinds of Files.

On some operating systems, more complex sets of access permissions can be specified, via mechanisms such as Access Control Lists (ACLs). See Extended File Attributes, for how to query and set those permissions.

Function: file-exists-p filename

This function returns t if a file named filename appears to exist. This does not mean you can necessarily read the file, only that you can probably find out its attributes. (On GNU and other POSIX-like systems, this is true if the file exists and you have execute permission on the containing directories, regardless of the permissions of the file itself.)

If the file does not exist, or if there was trouble determining whether the file exists, this function returns nil.

Since a file name that is an empty string is interpreted relative to the current buffer’s default directory (see Absolute and Relative File Names), calling file-exists-p with an argument that is an empty string will report about the buffer’s default directory.

Directories are files, so file-exists-p can return t when given a directory. However, because file-exists-p follows symbolic links, it returns t for a symbolic link name only if the target of the link exists; if your Lisp program needs to consider dangling symlinks whose target doesn’t exist as existing files, use file-attributes (see File Attributes) instead of file-exists-p.

Function: file-readable-p filename

This function returns t if a file named filename exists and you can read it. It returns nil otherwise.

Function: file-executable-p filename

This function returns t if a file named filename exists and you can execute it. It returns nil otherwise. On GNU and other POSIX-like systems, if the file is a directory, execute permission means you can check the existence and attributes of files inside the directory, and open those files if their modes permit.

Function: file-writable-p filename

This function returns t if the file filename can be written or created by you, and nil otherwise. A file is writable if the file exists and you can write it. It is creatable if it does not exist, but its parent directory does exist and you can write in that directory.

In the example below, foo is not writable because the parent directory does not exist, even though the user could create such a directory.

(file-writable-p "~/no-such-dir/foo")
     ⇒ nil
Function: file-accessible-directory-p dirname

This function returns t if you have permission to open existing files in the directory whose name as a file is dirname; otherwise (e.g., if there is no such directory), it returns nil. The value of dirname may be either a directory name (such as /foo/) or the file name of a file which is a directory (such as /foo, without the final slash).

For example, from the following we deduce that any attempt to read a file in /foo/ will give an error:

(file-accessible-directory-p "/foo")
     ⇒ nil
Macro: with-existing-directory body…

This macro ensures that default-directory is bound to an existing directory before executing body. If default-directory already exists, that’s preferred, and otherwise some other directory is used. This macro can be useful, for instance, when calling an external command that requires that it’s running in a directory that exists. The chosen directory is not guaranteed to be writable.

Function: access-file filename string

If you can read filename this function returns nil; otherwise it signals an error using string as the error message text.

Function: file-ownership-preserved-p filename &optional group

This function returns t if deleting the file filename and then creating it anew would keep the file’s owner unchanged. It also returns t for nonexistent files.

If the optional argument group is non-nil, this function also checks that the file’s group would be unchanged.

This function does not follow symbolic links.

Function: file-modes filename &optional flag

This function returns the mode bits of filename—an integer summarizing its read, write, and execution permissions. This function follows symbolic links. If the file does not exist, the return value is nil.

See File permissions in The GNU Coreutils Manual, for a description of mode bits. For example, if the low-order bit is 1, the file is executable by all users; if the second-lowest-order bit is 1, the file is writable by all users; etc. The highest possible value is 4095 (7777 octal), meaning that everyone has read, write, and execute permission, the SUID bit is set for both others and group, and the sticky bit is set.

By default this function follows symbolic links. However, if the optional argument flag is the symbol nofollow, this function does not follow filename if it is a symbolic link; this can help prevent inadvertently obtaining the mode bits of a file somewhere else, and is more consistent with file-attributes (see File Attributes).

See Changing File Names and Attributes, for the set-file-modes function, which can be used to set these permissions.

(file-modes "~/junk/diffs" 'nofollow)
     ⇒ 492               ; Decimal integer.
(format "%o" 492)
     ⇒ "754"             ; Convert to octal.

(set-file-modes "~/junk/diffs" #o666 'nofollow)
     ⇒ nil

$ ls -l diffs
-rw-rw-rw- 1 lewis lewis 3063 Oct 30 16:00 diffs

MS-DOS note: On MS-DOS, there is no such thing as an executable file mode bit. So file-modes considers a file executable if its name ends in one of the standard executable extensions, such as .com, .bat, .exe, and some others. Files that begin with the POSIX-standard ‘#!’ signature, such as shell and Perl scripts, are also considered executable. Directories are also reported as executable, for compatibility with POSIX. These conventions are also followed by file-attributes (see File Attributes).


26.6.2 Distinguishing Kinds of Files

This section describes how to distinguish various kinds of files, such as directories, symbolic links, and ordinary files.

Symbolic links are ordinarily followed wherever they appear. For example, to interpret the file name a/b/c, any of a, a/b, and a/b/c can be symbolic links that are followed, possibly recursively if the link targets are themselves symbolic links. However, a few functions do not follow symbolic links at the end of a file name (a/b/c in this example). Such a function is said to not follow symbolic links.

If the file filename is a symbolic link, this function does not follow it and instead returns its link target as a string. (The link target string is not necessarily the full absolute file name of the target; determining the full file name that the link points to is nontrivial, see below.)

If the file filename is not a symbolic link, or does not exist, or if there is trouble determining whether it is a symbolic link, file-symlink-p returns nil.

Here are a few examples of using this function:

(file-symlink-p "not-a-symlink")
     ⇒ nil
(file-symlink-p "sym-link")
     ⇒ "not-a-symlink"
(file-symlink-p "sym-link2")
     ⇒ "sym-link"
(file-symlink-p "/bin")
     ⇒ "/pub/bin"

Note that in the third example, the function returned sym-link, but did not proceed to resolve it, although that file is itself a symbolic link. That is because this function does not follow symbolic links—the process of following the symbolic links does not apply to the last component of the file name.

The string that this function returns is what is recorded in the symbolic link; it may or may not include any leading directories. This function does not expand the link target to produce a fully-qualified file name, and in particular does not use the leading directories, if any, of the filename argument if the link target is not an absolute file name. Here’s an example:

(file-symlink-p "/foo/bar/baz")
     ⇒ "some-file"

Here, although /foo/bar/baz was given as a fully-qualified file name, the result is not, and in fact does not have any leading directories at all. And since some-file might itself be a symbolic link, you cannot simply prepend leading directories to it, nor even naively use expand-file-name (see Functions that Expand Filenames) to produce its absolute file name.

For this reason, this function is seldom useful if you need to determine more than just the fact that a file is or isn’t a symbolic link. If you actually need the file name of the link target, use file-chase-links or file-truename, described in Truenames.

Function: file-directory-p filename

This function returns t if filename is the name of an existing directory. It returns nil if filename does not name a directory, or if there is trouble determining whether it is a directory. This function follows symbolic links.

(file-directory-p "~rms")
     ⇒ t
(file-directory-p "~rms/lewis/files.texi")
     ⇒ nil
(file-directory-p "~rms/lewis/no-such-file")
     ⇒ nil
(file-directory-p "$HOME")
     ⇒ nil
(file-directory-p
 (substitute-in-file-name "$HOME"))
     ⇒ t
Function: file-regular-p filename

This function returns t if the file filename exists and is a regular file (not a directory, named pipe, terminal, or other I/O device). It returns nil if filename does not exist or is not a regular file, or if there is trouble determining whether it is a regular file. This function follows symbolic links.


26.6.3 Truenames

The truename of a file is the name that you get by following symbolic links at all levels until none remain, then simplifying away ‘.’ and ‘..’ appearing as name components. This results in a sort of canonical name for the file. A file does not always have a unique truename; the number of distinct truenames a file has is equal to the number of hard links to the file. However, truenames are useful because they eliminate symbolic links as a cause of name variation.

Function: file-truename filename

This function returns the truename of the file filename. If the argument is not an absolute file name, this function first expands it against default-directory.

This function does not expand environment variables. Only substitute-in-file-name does that. See Definition of substitute-in-file-name.

If you may need to follow symbolic links preceding ‘..’ appearing as a name component, call file-truename without prior direct or indirect calls to expand-file-name. Otherwise, the file name component immediately preceding ‘..’ will be simplified away before file-truename is called. To eliminate the need for a call to expand-file-name, file-truename handles ‘~’ in the same way that expand-file-name does.

If the target of a symbolic links has remote file name syntax, file-truename returns it quoted. See Functions that Expand Filenames.

This function follows symbolic links, starting with filename, until it finds a file name which is not the name of a symbolic link. Then it returns that file name. This function does not follow symbolic links at the level of parent directories.

If you specify a number for limit, then after chasing through that many links, the function just returns what it has even if that is still a symbolic link.

To illustrate the difference between file-chase-links and file-truename, suppose that /usr/foo is a symbolic link to the directory /home/foo, and /home/foo/hello is an ordinary file (or at least, not a symbolic link) or nonexistent. Then we would have:

(file-chase-links "/usr/foo/hello")
     ;; This does not follow the links in the parent directories.
     ⇒ "/usr/foo/hello"
(file-truename "/usr/foo/hello")
     ;; Assuming that /home is not a symbolic link.
     ⇒ "/home/foo/hello"
Function: file-equal-p file1 file2

This function returns t if the files file1 and file2 name the same file. This is similar to comparing their truenames, except that remote file names are also handled in an appropriate manner. If file1 or file2 does not exist, the return value is unspecified.

Function: file-name-case-insensitive-p filename

Sometimes file names or their parts need to be compared as strings, in which case it’s important to know whether the underlying filesystem is case-insensitive. This function returns t if file filename is on a case-insensitive filesystem. It always returns t on MS-DOS and MS-Windows. On Cygwin and macOS, filesystems may or may not be case-insensitive, and the function tries to determine case-sensitivity by a runtime test. If the test is inconclusive, the function returns t on Cygwin and nil on macOS.

Currently this function always returns nil on platforms other than MS-DOS, MS-Windows, Cygwin, and macOS. It does not detect case-insensitivity of mounted filesystems, such as Samba shares or NFS-mounted Windows volumes. On remote hosts, it assumes t for the ‘smb’ method. For all other connection methods, runtime tests are performed.

Function: vc-responsible-backend file

This function determines the responsible VC backend of the given file. For example, if emacs.c is a file tracked by Git, (vc-responsible-backend "emacs.c") returns ‘Git’. Note that if file is a symbolic link, vc-responsible-backend will not resolve it—the backend of the symbolic link file itself is reported. To get the backend VC of the file to which file refers, wrap file with a symbolic link resolving function such as file-chase-links:

(vc-responsible-backend (file-chase-links "emacs.c"))

26.6.4 File Attributes

This section describes the functions for getting detailed information about a file, including the owner and group numbers, the number of names, the inode number, the size, and the times of access and modification.

Function: file-newer-than-file-p filename1 filename2

This function returns t if the file filename1 is newer than file filename2. If filename1 does not exist, it returns nil. If filename1 does exist, but filename2 does not, it returns t.

In the following example, assume that the file aug-19 was written on the 19th, aug-20 was written on the 20th, and the file no-file doesn’t exist at all.

(file-newer-than-file-p "aug-19" "aug-20")
     ⇒ nil
(file-newer-than-file-p "aug-20" "aug-19")
     ⇒ t
(file-newer-than-file-p "aug-19" "no-file")
     ⇒ t
(file-newer-than-file-p "no-file" "aug-19")
     ⇒ nil
Function: file-has-changed-p filename tag

This function returns non-nil if the time stamp of filename has changed since the last call. When called for the first time for some filename, it records the last modification time and size of the file, and returns non-nil when filename exists. Thereafter, when called for the same filename, it compares the current time stamp and size with the recorded ones, and returns non-nil only if either the time stamp or the size (or both) are different. This is useful when a Lisp program wants to re-read a file whenever it changes. With an optional argument tag, which must be a symbol, the size and modification time comparisons are limited to calls with the same tag.

Function: file-attributes filename &optional id-format

This function returns a list of attributes of file filename. If the specified file does not exist, it returns nil. This function does not follow symbolic links. The optional parameter id-format specifies the preferred format of attributes UID and GID (see below)—the valid values are 'string and 'integer. The latter is the default, but we plan to change that, so you should specify a non-nil value for id-format if you use the returned UID or GID.

On GNU platforms when operating on a local file, this function is atomic: if the filesystem is simultaneously being changed by some other process, this function returns the file’s attributes either before or after the change. Otherwise this function is not atomic, and might return nil if it detects the race condition, or might return a hodgepodge of the previous and current file attributes.

Accessor functions are provided to access the elements in this list. The accessors are mentioned along with the descriptions of the elements below.

The elements of the list, in order, are:

  1. t for a directory, a string for a symbolic link (the name linked to), or nil for a text file (file-attribute-type).
  2. The number of names the file has (file-attribute-link-number). Alternate names, also known as hard links, can be created by using the add-name-to-file function (see Changing File Names and Attributes).
  3. The file’s UID, normally as a string (file-attribute-user-id). However, if it does not correspond to a named user, the value is an integer.
  4. The file’s GID, likewise (file-attribute-group-id).
  5. The time of last access as a Lisp timestamp (file-attribute-access-time). The timestamp is in the style of current-time (see Time of Day) and is truncated to that of the filesystem’s timestamp resolution; for example, on some FAT-based filesystems, only the date of last access is recorded, so this time will always hold the midnight of the day of the last access.
  6. The time of last modification as a Lisp timestamp (file-attribute-modification-time). This is the last time when the file’s contents were modified.
  7. The time of last status change as a Lisp timestamp (file-attribute-status-change-time). This is the time of the last change to the file’s access mode bits, its owner and group, and other information recorded in the filesystem for the file, beyond the file’s contents.
  8. The size of the file in bytes (file-attribute-size).
  9. The file’s modes, as a string of ten letters or dashes, as in ‘ls -l’ (file-attribute-modes).
  10. An unspecified value, present for backward compatibility.
  11. The file’s inode number (file-attribute-inode-number), a nonnegative integer.
  12. The filesystem’s identifier of the device that the file is on (file-attribute-device-number), an integer or a cons cell of two integers. The latter is sometimes used by remote files, in order to distinguish remote filesystems from local ones.

The file’s inode and device together give enough information to distinguish any two files on the system—no two files can have the same values for both of these attributes. This tuple that uniquely identifies the file is returned by file-attribute-file-identifier.

For example, here are the file attributes for files.texi:

(file-attributes "files.texi" 'string)
     ⇒  (nil 1 "lh" "users"
          (20614 64019 50040 152000)
          (20000 23 0 0)
          (20614 64555 902289 872000)
          122295 "-rw-rw-rw-"
          t 6473924464520138
          1014478468)

and here is how the result is interpreted:

nil

is neither a directory nor a symbolic link.

1

has only one name (the name files.texi in the current default directory).

"lh"

is owned by the user with name ‘lh’.

"users"

is in the group with name ‘users’.

(20614 64019 50040 152000)

was last accessed on October 23, 2012, at 20:12:03.050040152 UTC. (This timestamp is (1351023123050040152 . 1000000000) if current-time-list is nil.)

(20000 23 0 0)

was last modified on July 15, 2001, at 08:53:43.000000000 UTC. (This timestamp is (1310720023000000000 . 1000000000) if current-time-list is nil.)

(20614 64555 902289 872000)

last had its status changed on October 23, 2012, at 20:20:59.902289872 UTC. (This timestamp is (1351023659902289872 . 1000000000) if current-time-list is nil.)

122295

is 122295 bytes long. (It may not contain 122295 characters, though, if some of the bytes belong to multibyte sequences, and also if the end-of-line format is CR-LF.)

"-rw-rw-rw-"

has a mode of read and write access for the owner, group, and world.

t

is merely a placeholder; it carries no information.

6473924464520138

has an inode number of 6473924464520138.

1014478468

is on the file-system device whose number is 1014478468.

This function returns the number of names (i.e., hard links) that file filename has. If the file does not exist, this function returns nil. Note that symbolic links have no effect on this function, because they are not considered to be names of the files they link to. This function does not follow symbolic links.

$ ls -l foo*
-rw-rw-rw- 2 rms rms 4 Aug 19 01:27 foo
-rw-rw-rw- 2 rms rms 4 Aug 19 01:27 foo1

(file-nlinks "foo")
     ⇒ 2
(file-nlinks "doesnt-exist")
     ⇒ nil

26.6.5 Extended File Attributes

On some operating systems, each file can be associated with arbitrary extended file attributes. At present, Emacs supports querying and setting two specific sets of extended file attributes: Access Control Lists (ACLs) and SELinux contexts. These extended file attributes are used, on some systems, to impose more sophisticated file access controls than the basic Unix-style permissions discussed in the previous sections.

A detailed explanation of ACLs and SELinux is beyond the scope of this manual. For our purposes, each file can be associated with an ACL, which specifies its properties under an ACL-based file control system, and/or an SELinux context, which specifies its properties under the SELinux system.

Function: file-acl filename

This function returns the ACL for the file filename. The exact Lisp representation of the ACL is unspecified (and may change in future Emacs versions), but it is the same as what set-file-acl takes for its acl argument (see Changing File Names and Attributes).

The underlying ACL implementation is platform-specific; on GNU/Linux and BSD, Emacs uses the POSIX ACL interface, while on MS-Windows Emacs emulates the POSIX ACL interface with native file security APIs.

If ACLs are not supported or the file does not exist, then the return value is nil.

Function: file-selinux-context filename

This function returns the SELinux context of the file filename, as a list of the form (user role type range). The list elements are the context’s user, role, type, and range respectively, as Lisp strings; see the SELinux documentation for details about what these actually mean. The return value has the same form as what set-file-selinux-context takes for its context argument (see Changing File Names and Attributes).

If SELinux is not supported or the file does not exist, then the return value is (nil nil nil nil).

Function: file-extended-attributes filename

This function returns an alist of the Emacs-recognized extended attributes of file filename. Currently, it serves as a convenient way to retrieve both the ACL and SELinux context; you can then call the function set-file-extended-attributes, with the returned alist as its second argument, to apply the same file access attributes to another file (see Changing File Names and Attributes).

One of the elements is (acl . acl), where acl has the same form returned by file-acl.

Another element is (selinux-context . context), where context is the SELinux context, in the same form returned by file-selinux-context.


26.6.6 Locating Files in Standard Places

This section explains how to search for a file in a list of directories (a path), or for an executable file in the standard list of executable file directories.

To search for a user-specific configuration file, See Standard File Names, for the locate-user-emacs-file function.

Function: locate-file filename path &optional suffixes predicate

This function searches for a file whose name is filename in a list of directories given by path, trying the suffixes in suffixes. If it finds such a file, it returns the file’s absolute file name (see Absolute and Relative File Names); otherwise it returns nil.

The optional argument suffixes gives the list of file-name suffixes to append to filename when searching. locate-file tries each possible directory with each of these suffixes. If suffixes is nil, or (""), then there are no suffixes, and filename is used only as-is. Typical values of suffixes are exec-suffixes (see Functions that Create Subprocesses), load-suffixes, load-file-rep-suffixes and the return value of the function get-load-suffixes (see Load Suffixes).

Typical values for path are exec-path (see Functions that Create Subprocesses) when looking for executable programs, or load-path (see Library Search) when looking for Lisp files. If filename is absolute, path has no effect, but the suffixes in suffixes are still tried.

The optional argument predicate, if non-nil, specifies a predicate function for testing whether a candidate file is suitable. The predicate is passed the candidate file name as its single argument. If predicate is nil or omitted, locate-file uses file-readable-p as the predicate. See Distinguishing Kinds of Files, for other useful predicates, e.g., file-executable-p and file-directory-p.

This function will normally skip directories, so if you want it to find directories, make sure the predicate function returns dir-ok for them. For example:

(locate-file "html" '("/var/www" "/srv") nil
             (lambda (f) (if (file-directory-p f) 'dir-ok)))

For compatibility, predicate can also be one of the symbols executable, readable, writable, exists, or a list of one or more of these symbols.

Function: executable-find program &optional remote

This function searches for the executable file of the named program and returns the absolute file name of the executable, including its file-name extensions, if any. It returns nil if the file is not found. The function searches in all the directories in exec-path, and tries all the file-name extensions in exec-suffixes (see Functions that Create Subprocesses).

If remote is non-nil, and default-directory is a remote directory, program is searched on the respective remote host.


26.7 Changing File Names and Attributes

The functions in this section rename, copy, delete, link, and set the modes (permissions) of files. Typically, they signal a file-error error if they fail to perform their function, reporting the system-dependent error message that describes the reason for the failure. If they fail because a file is missing, they signal a file-missing error instead.

For performance, the operating system may cache or alias changes made by these functions instead of writing them immediately to secondary storage. See Files and Secondary Storage.

In the functions that have an argument newname, if this argument is a directory name it is treated as if the nondirectory part of the source name were appended. Typically, a directory name is one that ends in ‘/’ (see Directory Names). For example, if the old name is a/b/c, the newname d/e/f/ is treated as if it were d/e/f/c. This special treatment does not apply if newname is not a directory name but names a file that is a directory; for example, the newname d/e/f is left as-is even if d/e/f happens to be a directory.

In the functions that have an argument newname, if a file by the name of newname already exists, the actions taken depend on the value of the argument ok-if-already-exists:

  • Signal a file-already-exists error if ok-if-already-exists is nil.
  • Request confirmation if ok-if-already-exists is a number.
  • Replace the old file without confirmation if ok-if-already-exists is any other value.
Command: add-name-to-file oldname newname &optional ok-if-already-exists

This function gives the file named oldname the additional name newname. This means that newname becomes a new hard link to oldname.

If newname is a symbolic link, its directory entry is replaced, not the directory entry it points to. If oldname is a symbolic link, this function might or might not follow the link; it does not follow the link on GNU platforms. If oldname is a directory, this function typically fails, although for the superuser on a few old-fashioned non-GNU platforms it can succeed and create a filesystem that is not tree-structured.

In the first part of the following example, we list two files, foo and foo3.

$ ls -li fo*
81908 -rw-rw-rw- 1 rms rms 29 Aug 18 20:32 foo
84302 -rw-rw-rw- 1 rms rms 24 Aug 18 20:31 foo3

Now we create a hard link, by calling add-name-to-file, then list the files again. This shows two names for one file, foo and foo2.

(add-name-to-file "foo" "foo2")
     ⇒ nil

$ ls -li fo*
81908 -rw-rw-rw- 2 rms rms 29 Aug 18 20:32 foo
81908 -rw-rw-rw- 2 rms rms 29 Aug 18 20:32 foo2
84302 -rw-rw-rw- 1 rms rms 24 Aug 18 20:31 foo3

Finally, we evaluate the following:

(add-name-to-file "foo" "foo3" t)

and list the files again. Now there are three names for one file: foo, foo2, and foo3. The old contents of foo3 are lost.

(add-name-to-file "foo1" "foo3")
     ⇒ nil

$ ls -li fo*
81908 -rw-rw-rw- 3 rms rms 29 Aug 18 20:32 foo
81908 -rw-rw-rw- 3 rms rms 29 Aug 18 20:32 foo2
81908 -rw-rw-rw- 3 rms rms 29 Aug 18 20:32 foo3

This function is meaningless on operating systems where multiple names for one file are not allowed. Some systems implement multiple names by copying the file instead.

See also file-nlinks in File Attributes.

Command: rename-file filename newname &optional ok-if-already-exists

This command renames the file filename as newname.

If filename has additional names aside from filename, it continues to have those names. In fact, adding the name newname with add-name-to-file and then deleting filename has the same effect as renaming, aside from momentary intermediate states and treatment of errors, directories and symbolic links.

This command does not follow symbolic links. If filename is a symbolic link, this command renames the symbolic link, not the file it points to. If newname is a symbolic link, its directory entry is replaced, not the directory entry it points to.

This command does nothing if filename and newname are the same directory entry, i.e., if they refer to the same parent directory and give the same name within that directory. Otherwise, if filename and newname name the same file, this command does nothing on POSIX-conforming systems, and removes filename on some non-POSIX systems.

If newname exists, then it must be an empty directory if oldname is a directory and a non-directory otherwise.

Command: copy-file oldname newname &optional ok-if-already-exists time preserve-uid-gid preserve-permissions

This command copies the file oldname to newname. An error is signaled if oldname is not a regular file. If newname names a directory, it copies oldname into that directory, preserving its final name component.

This function follows symbolic links, except that it does not follow a dangling symbolic link to create newname.

If time is non-nil, then this function gives the new file the same last-modified time that the old one has. (This works on only some operating systems.) If setting the time gets an error, copy-file signals a file-date-error error. In an interactive call, a prefix argument specifies a non-nil value for time.

If argument preserve-uid-gid is nil, we let the operating system decide the user and group ownership of the new file (this is usually set to the user running Emacs). If preserve-uid-gid is non-nil, we attempt to copy the user and group ownership of the file. This works only on some operating systems, and only if you have the correct permissions to do so.

If the optional argument preserve-permissions is non-nil, this function copies the file modes (or “permissions”) of oldname to newname, as well as the Access Control List and SELinux context (if any). See Information about Files.

Otherwise, the file modes of newname are left unchanged if it is an existing file, and set to those of oldname, masked by the default file permissions (see set-default-file-modes below), if newname is to be newly created. The Access Control List or SELinux context are not copied over in either case.

This command makes a symbolic link to target, named linkname. This is like the shell command ‘ln -s target linkname’. The target argument is treated only as a string; it need not name an existing file. If ok-if-already-exists is an integer, indicating interactive use, then leading ‘~’ is expanded and leading ‘/:’ is stripped in the target string.

If target is a relative file name, the resulting symbolic link is interpreted relative to the directory containing the symbolic link. See Absolute and Relative File Names.

If both target and linkname have remote file name syntax, and if both remote identifications are equal, the symbolic link points to the local file name part of target.

This function is not available on systems that don’t support symbolic links.

Command: delete-file filename &optional trash

This command deletes the file filename. If the file has multiple names, it continues to exist under the other names. If filename is a symbolic link, delete-file deletes only the symbolic link and not its target.

The command signals a suitable kind of file-error error if filename cannot be deleted. (On GNU and other POSIX-like systems, a file can be deleted if its directory is writable.) If the file does not exist, this command will not signal any error.

If the optional argument trash is non-nil and the variable delete-by-moving-to-trash is non-nil, this command moves the file into the system Trash instead of deleting it. See Miscellaneous File Operations in The GNU Emacs Manual. When called interactively, trash is t if no prefix argument is given, and nil otherwise.

See also delete-directory in Creating, Copying and Deleting Directories.

Command: set-file-modes filename mode &optional flag

This function sets the file mode (or permissions) of filename to mode.

By default this function follows symbolic links. However, if the optional argument flag is the symbol nofollow, this function does not follow filename if it is a symbolic link; this can help prevent inadvertently changing the mode bits of a file somewhere else. On platforms that do not support changing mode bits on a symbolic link, this function signals an error when filename is a symbolic link and flag is nofollow.

If called non-interactively, mode must be an integer. Only the lowest 12 bits of the integer are used; on most systems, only the lowest 9 bits are meaningful. You can use the Lisp construct for octal numbers to enter mode. For example,

(set-file-modes "myfile" #o644 'nofollow)

specifies that the file should be readable and writable for its owner, readable for group members, and readable for all other users. See File permissions in The GNU Coreutils Manual, for a description of mode bit specifications.

Interactively, mode is read from the minibuffer using read-file-modes (see below), which lets the user type in either an integer or a string representing the permissions symbolically.

See Testing Accessibility, for the function file-modes, which returns the permissions of a file.

Function: set-default-file-modes mode

This function sets the default permissions for new files created by Emacs and its subprocesses. Every file created with Emacs initially has these permissions, or a subset of them (write-region will not grant execute permissions even if the default file permissions allow execution). On GNU and other POSIX-like systems, the default permissions are given by the bitwise complement of the ‘umask’ value, i.e. each bit that is set in the argument mode will be reset in the default permissions with which Emacs creates files.

The argument mode should be an integer which specifies the permissions, similar to set-file-modes above. Only the lowest 9 bits are meaningful.

The default file permissions have no effect when you save a modified version of an existing file; saving a file preserves its existing permissions.

Macro: with-file-modes mode body…

This macro evaluates the body forms with the default permissions for new files temporarily set to modes (whose value is as for set-file-modes above). When finished, it restores the original default file permissions, and returns the value of the last form in body.

This is useful for creating private files, for example.

Function: default-file-modes

This function returns the default file permissions, as an integer.

Function: read-file-modes &optional prompt base-file

This function reads a set of file mode bits from the minibuffer. The first optional argument prompt specifies a non-default prompt. Second second optional argument base-file is the name of a file on whose permissions to base the mode bits that this function returns, if what the user types specifies mode bits relative to permissions of an existing file.

If user input represents an octal number, this function returns that number. If it is a complete symbolic specification of mode bits, as in "u=rwx", the function converts it to the equivalent numeric value using file-modes-symbolic-to-number and returns the result. If the specification is relative, as in "o+g", then the permissions on which the specification is based are taken from the mode bits of base-file. If base-file is omitted or nil, the function uses 0 as the base mode bits. The complete and relative specifications can be combined, as in "u+r,g+rx,o+r,g-w". See File permissions in The GNU Coreutils Manual, for a description of file mode specifications.

Function: file-modes-symbolic-to-number modes &optional base-modes

This function converts a symbolic file mode specification in modes into the equivalent integer. If the symbolic specification is based on an existing file, that file’s mode bits are taken from the optional argument base-modes; if that argument is omitted or nil, it defaults to 0, i.e., no access rights at all.

Function: file-modes-number-to-symbolic modes

This function converts a numeric file mode specification in modes into the equivalent string form. The string which this function returns is in the same format produced by the shell command ls -l and by file-attributes, not the symbolic form accepted by file-modes-symbolic-to-number and the chmod shell command.

Function: set-file-times filename &optional time flag

This function sets the access and modification times of filename to time. The return value is t if the times are successfully set, otherwise it is nil. time defaults to the current time and must be a time value (see Time of Day).

By default this function follows symbolic links. However, if the optional argument flag is the symbol nofollow, this function does not follow filename if it is a symbolic link; this can help prevent inadvertently changing the times of a file somewhere else. On platforms that do not support changing times on a symbolic link, this function signals an error when filename is a symbolic link and flag is nofollow.

Function: set-file-extended-attributes filename attribute-alist

This function sets the Emacs-recognized extended file attributes for filename. The second argument attribute-alist should be an alist of the same form returned by file-extended-attributes. The return value is t if the attributes are successfully set, otherwise it is nil. See Extended File Attributes.

Function: set-file-selinux-context filename context

This function sets the SELinux security context for filename to context. The context argument should be a list (user role type range), where each element is a string. See Extended File Attributes.

The function returns t if it succeeds in setting the SELinux context of filename. It returns nil if the context was not set (e.g., if SELinux is disabled, or if Emacs was compiled without SELinux support).

Function: set-file-acl filename acl

This function sets the Access Control List for filename to acl. The acl argument should have the same form returned by the function file-acl. See Extended File Attributes.

The function returns t if it successfully sets the ACL of filename, nil otherwise.


26.8 Files and Secondary Storage

After Emacs changes a file, there are two reasons the changes might not survive later failures of power or media, both having to do with efficiency. First, the operating system might alias written data with data already stored elsewhere on secondary storage until one file or the other is later modified; this will lose both files if the only copy on secondary storage is lost due to media failure. Second, the operating system might not write data to secondary storage immediately, which will lose the data if power is lost.

Although both sorts of failures can largely be avoided by a suitably configured file system, such systems are typically more expensive or less efficient. In more-typical systems, to survive media failure you can copy the file to a different device, and to survive a power failure you can use the write-region function with the write-region-inhibit-fsync variable set to nil. See Writing to Files.


26.9 File Names

Files are generally referred to by their names, in Emacs as elsewhere. File names in Emacs are represented as strings. The functions that operate on a file all expect a file name argument.

In addition to operating on files themselves, Emacs Lisp programs often need to operate on file names; i.e., to take them apart and to use part of a name to construct related file names. This section describes how to manipulate file names.

The functions in this section do not actually access files, so they can operate on file names that do not refer to an existing file or directory.

On MS-DOS and MS-Windows, these functions (like the function that actually operate on files) accept MS-DOS or MS-Windows file-name syntax, where backslashes separate the components, as well as POSIX syntax; but they always return POSIX syntax. This enables Lisp programs to specify file names in POSIX syntax and work properly on all systems without change.19


26.9.1 File Name Components

The operating system groups files into directories. To specify a file, you must specify the directory and the file’s name within that directory. Therefore, Emacs considers a file name as having two main parts: the directory name part, and the nondirectory part (or file name within the directory). Either part may be empty. Concatenating these two parts reproduces the original file name. 20

On most systems, the directory part is everything up to and including the last slash (backslash is also allowed in input on MS-DOS or MS-Windows); the nondirectory part is the rest.

For some purposes, the nondirectory part is further subdivided into the name proper and the version number. On most systems, only backup files have version numbers in their names.

Function: file-name-directory filename

This function returns the directory part of filename, as a directory name (see Directory Names), or nil if filename does not include a directory part.

On GNU and other POSIX-like systems, a string returned by this function always ends in a slash. On MS-DOS it can also end in a colon.

(file-name-directory "lewis/foo")  ; GNU example
     ⇒ "lewis/"
(file-name-directory "foo")        ; GNU example
     ⇒ nil
Function: file-name-nondirectory filename

This function returns the nondirectory part of filename.

(file-name-nondirectory "lewis/foo")
     ⇒ "foo"
(file-name-nondirectory "foo")
     ⇒ "foo"
(file-name-nondirectory "lewis/")
     ⇒ ""
Function: file-name-sans-versions filename &optional keep-backup-version

This function returns filename with any file version numbers, backup version numbers, or trailing tildes discarded.

If keep-backup-version is non-nil, then true file version numbers understood as such by the file system are discarded from the return value, but backup version numbers are kept.

(file-name-sans-versions "~rms/foo.~1~")
     ⇒ "~rms/foo"
(file-name-sans-versions "~rms/foo~")
     ⇒ "~rms/foo"
(file-name-sans-versions "~rms/foo")
     ⇒ "~rms/foo"
Function: file-name-extension filename &optional period

This function returns filename’s final extension, if any, after applying file-name-sans-versions to remove any version/backup part. The extension, in a file name, is the part that follows the last ‘.’ in the last name component (minus any version/backup part).

This function returns nil for extensionless file names such as foo. It returns "" for null extensions, as in foo.. If the last component of a file name begins with a ‘.’, that ‘.’ doesn’t count as the beginning of an extension. Thus, .emacs’s extension is nil, not ‘.emacs’.

If period is non-nil, then the returned value includes the period that delimits the extension, and if filename has no extension, the value is "".

Function: file-name-with-extension filename extension

This function returns filename with its extension set to extension. A single leading dot in the extension will be stripped if there is one. For example:

(file-name-with-extension "file" "el")
     ⇒ "file.el"
(file-name-with-extension "file" ".el")
     ⇒ "file.el"
(file-name-with-extension "file.c" "el")
     ⇒ "file.el"

Note that this function will error if filename or extension are empty, or if the filename is shaped like a directory (i.e., if directory-name-p returns non-nil).

Function: file-name-sans-extension filename

This function returns filename minus its extension, if any. The version/backup part, if present, is only removed if the file has an extension. For example,

(file-name-sans-extension "foo.lose.c")
     ⇒ "foo.lose"
(file-name-sans-extension "big.hack/foo")
     ⇒ "big.hack/foo"
(file-name-sans-extension "/my/home/.emacs")
     ⇒ "/my/home/.emacs"
(file-name-sans-extension "/my/home/.emacs.el")
     ⇒ "/my/home/.emacs"
(file-name-sans-extension "~/foo.el.~3~")
     ⇒ "~/foo"
(file-name-sans-extension "~/foo.~3~")
     ⇒ "~/foo.~3~"

Note that the ‘.~3~’ in the two last examples is the backup part, not an extension.

Function: file-name-base filename

This function is the composition of file-name-sans-extension and file-name-nondirectory. For example,

(file-name-base "/my/home/foo.c")
    ⇒ "foo"
Function: file-name-split filename

This function splits a file name into its components, and can be thought of as the inverse of string-join with the appropriate directory separator. For example,

(file-name-split "/tmp/foo.txt")
    ⇒ ("" "tmp" "foo.txt")
(string-join (file-name-split "/tmp/foo.txt") "/")
    ⇒ "/tmp/foo.txt"

26.9.2 Absolute and Relative File Names

All the directories in the file system form a tree starting at the root directory. A file name can specify all the directory names starting from the root of the tree; then it is called an absolute file name. Or it can specify the position of the file in the tree relative to a default directory; then it is called a relative file name. On GNU and other POSIX-like systems, after any leading ‘~’ has been expanded, an absolute file name starts with a ‘/’ (see abbreviate-file-name), and a relative one does not. On MS-DOS and MS-Windows, an absolute file name starts with a slash or a backslash, or with a drive specification ‘x:/’, where x is the drive letter.

Function: file-name-absolute-p filename

This function returns t if file filename is an absolute file name, nil otherwise. A file name is considered to be absolute if its first component is ‘~’, or is ‘~user’ where user is a valid login name. In the following examples, assume that there is a user named ‘rms’ but no user named ‘nosuchuser’.

(file-name-absolute-p "~rms/foo")
     ⇒ t
(file-name-absolute-p "~nosuchuser/foo")
     ⇒ nil
(file-name-absolute-p "rms/foo")
     ⇒ nil
(file-name-absolute-p "/user/rms/foo")
     ⇒ t

Given a possibly relative file name, you can expand any leading ‘~’ and convert the result to an absolute name using expand-file-name (see Functions that Expand Filenames). This function converts absolute file names to relative names:

Function: file-relative-name filename &optional directory

This function tries to return a relative name that is equivalent to filename, assuming the result will be interpreted relative to directory (an absolute directory name or directory file name). If directory is omitted or nil, it defaults to the current buffer’s default directory.

On some operating systems, an absolute file name begins with a device name. On such systems, filename has no relative equivalent based on directory if they start with two different device names. In this case, file-relative-name returns filename in absolute form.

(file-relative-name "/foo/bar" "/foo/")
     ⇒ "bar"
(file-relative-name "/foo/bar" "/hack/")
     ⇒ "../foo/bar"

A file name that is an empty string stands for the current buffer’s default directory.


26.9.3 Directory Names

A directory name is a string that must name a directory if it names any file at all. A directory is actually a kind of file, and it has a file name (called the directory file name), which is related to the directory name but is typically not identical. (This is not quite the same as the usual POSIX terminology.) These two names for the same entity are related by a syntactic transformation. On GNU and other POSIX-like systems, this is simple: to obtain a directory name, append a ‘/’ to a directory file name that does not already end in ‘/’. On MS-DOS the relationship is more complicated.

The difference between a directory name and a directory file name is subtle but crucial. When an Emacs variable or function argument is described as being a directory name, a directory file name is not acceptable. When file-name-directory returns a string, that is always a directory name.

The following two functions convert between directory names and directory file names. They do nothing special with environment variable substitutions such as ‘$HOME’, and the constructs ‘~’, ‘.’ and ‘..’.

Function: file-name-as-directory filename

This function returns a string representing filename in a form that the operating system will interpret as the name of a directory (a directory name). On most systems, this means appending a slash to the string (if it does not already end in one).

(file-name-as-directory "~rms/lewis")
     ⇒ "~rms/lewis/"
Function: directory-name-p filename

This function returns non-nil if filename ends with a directory separator character. This is the forward slash ‘/’ on GNU and other POSIX-like systems; MS-Windows and MS-DOS recognize both the forward slash and the backslash ‘\’ as directory separators.

Function: directory-file-name dirname

This function returns a string representing dirname in a form that the operating system will interpret as the name of a file (a directory file name). On most systems, this means removing the final directory separators from the string, unless the string consists entirely of directory separators.

(directory-file-name "~lewis/")
     ⇒ "~lewis"
Function: file-name-concat directory &rest components

Concatenate components to directory, inserting a slash before the components if directory or the preceding component didn’t end with a slash.

(file-name-concat "/tmp" "foo")
     ⇒ "/tmp/foo"

A directory or components that are nil or the empty string are ignored—they are filtered out first and do not affect the results in any way.

This is almost the same as using concat, but dirname (and the non-final components) may or may not end with slash characters, and this function will not double those characters.

To convert a directory name to its abbreviation, use this function:

Function: abbreviate-file-name filename

This function returns an abbreviated form of filename. It applies the abbreviations specified in directory-abbrev-alist (see File Aliases in The GNU Emacs Manual), then substitutes ‘~’ for the user’s home directory if the argument names a file in the home directory or one of its subdirectories. If the home directory is a root directory, it is not replaced with ‘~’, because this does not make the result shorter on many systems.

You can use this function for directory names and for file names, because it recognizes abbreviations even as part of the name.

Function: file-name-parent-directory filename

This function returns the directory name of the parent directory of filename. If filename is at the root directory of the filesystem, it returns nil. A relative filename is assumed to be relative to default-directory, and the return value will also be relative in that case. If the return value is non-nil, it ends in a slash.


26.9.4 Functions that Expand Filenames

Expanding a file name means converting a relative file name to an absolute one. Since this is done relative to a default directory, you must specify the default directory as well as the file name to be expanded. It also involves expanding abbreviations like ~/ (see abbreviate-file-name), and eliminating redundancies like ./ and name/../.

Function: expand-file-name filename &optional directory

This function converts filename to an absolute file name. If directory is supplied, it is the default directory to start with if filename is relative and does not start with ‘~’. (The value of directory should itself be an absolute directory name or directory file name; it may start with ‘~’.) Otherwise, the current buffer’s value of default-directory is used. For example:

(expand-file-name "foo")
     ⇒ "/xcssun/users/rms/lewis/foo"
(expand-file-name "../foo")
     ⇒ "/xcssun/users/rms/foo"
(expand-file-name "foo" "/usr/spool/")
     ⇒ "/usr/spool/foo"

If the part of filename before the first slash is ‘~’, it expands to your home directory, which is typically specified by the value of the HOME environment variable (see General Variables in The GNU Emacs Manual). If the part before the first slash is ‘~user’ and if user is a valid login name, it expands to user’s home directory. If you do not want this expansion for a relative filename that might begin with a literal ‘~’, you can use (concat (file-name-as-directory directory) filename) instead of (expand-file-name filename directory).

File names containing ‘.’ or ‘..’ are simplified to their canonical form:

(expand-file-name "bar/../foo")
     ⇒ "/xcssun/users/rms/lewis/foo"

In some cases, a leading ‘..’ component can remain in the output:

(expand-file-name "../home" "/")
     ⇒ "/../home"

This is for the sake of filesystems that have the concept of a superroot above the root directory /. On other filesystems, /../ is interpreted exactly the same as /.

Expanding . or the empty string returns the default directory:

(expand-file-name "." "/usr/spool/")
     ⇒ "/usr/spool"
(expand-file-name "" "/usr/spool/")
     ⇒ "/usr/spool"

Note that expand-file-name does not expand environment variables; only substitute-in-file-name does that:

(expand-file-name "$HOME/foo")
     ⇒ "/xcssun/users/rms/lewis/$HOME/foo"

Note also that expand-file-name does not follow symbolic links at any level. This results in a difference between the way file-truename and expand-file-name treat ‘..’. Assuming that ‘/tmp/bar’ is a symbolic link to the directory ‘/tmp/foo/bar’ we get:

(file-truename "/tmp/bar/../myfile")
     ⇒ "/tmp/foo/myfile"
(expand-file-name "/tmp/bar/../myfile")
     ⇒ "/tmp/myfile"

If you may need to follow symbolic links preceding ‘..’, you should make sure to call file-truename without prior direct or indirect calls to expand-file-name. See Truenames.

Variable: default-directory

The value of this buffer-local variable is the default directory for the current buffer. It should be an absolute directory name; it may start with ‘~’. This variable is buffer-local in every buffer.

expand-file-name uses the default directory when its second argument is nil.

The value is always a string ending with a slash.

default-directory
     ⇒ "/user/lewis/manual/"
Function: substitute-in-file-name filename

This function replaces environment variable references in filename with the environment variable values. Following standard Unix shell syntax, ‘$’ is the prefix to substitute an environment variable value. If the input contains ‘$$’, that is converted to ‘$’; this gives the user a way to quote a ‘$’.

The environment variable name is the series of alphanumeric characters (including underscores) that follow the ‘$’. If the character following the ‘$’ is a ‘{’, then the variable name is everything up to the matching ‘}’.

Calling substitute-in-file-name on output produced by substitute-in-file-name tends to give incorrect results. For instance, use of ‘$$’ to quote a single ‘$’ won’t work properly, and ‘$’ in an environment variable’s value could lead to repeated substitution. Therefore, programs that call this function and put the output where it will be passed to this function need to double all ‘$’ characters to prevent subsequent incorrect results.

Here we assume that the environment variable HOME, which holds the user’s home directory, has value ‘/xcssun/users/rms’.

(substitute-in-file-name "$HOME/foo")
     ⇒ "/xcssun/users/rms/foo"

After substitution, if a ‘~’ or a ‘/’ appears immediately after another ‘/’, the function discards everything before it (up through the immediately preceding ‘/’).

(substitute-in-file-name "bar/~/foo")
     ⇒ "~/foo"
(substitute-in-file-name "/usr/local/$HOME/foo")
     ⇒ "/xcssun/users/rms/foo"
     ;; /usr/local/ has been discarded.

Sometimes, it is not desired to expand file names. In such cases, the file name can be quoted to suppress the expansion, and to handle the file name literally. Quoting happens by prefixing the file name with ‘/:’.

Macro: file-name-quote name

This macro adds the quotation prefix ‘/:’ to the file name. For a local file name, it prefixes name with ‘/:’. If name is a remote file name, the local part of name (see Making Certain File Names “Magic”) is quoted. If name is already a quoted file name, name is returned unchanged.

(substitute-in-file-name (file-name-quote "bar/~/foo"))
     ⇒ "/:bar/~/foo"

(substitute-in-file-name (file-name-quote "/ssh:host:bar/~/foo"))
     ⇒ "/ssh:host:/:bar/~/foo"

The macro cannot be used to suppress file name handlers from magic file names (see Making Certain File Names “Magic”).

Macro: file-name-unquote name

This macro removes the quotation prefix ‘/:’ from the file name, if any. If name is a remote file name, the local part of name is unquoted.

Macro: file-name-quoted-p name

This macro returns non-nil, when name is quoted with the prefix ‘/:’. If name is a remote file name, the local part of name is checked.


26.9.5 Generating Unique File Names

Some programs need to write temporary files. Here is the usual way to construct a name for such a file:

(make-temp-file name-of-application)

The job of make-temp-file is to prevent two different users or two different jobs from trying to use the exact same file name.

Function: make-temp-file prefix &optional dir-flag suffix text

This function creates a temporary file and returns its name. Emacs creates the temporary file’s name by adding to prefix some random characters that are different in each Emacs job. The result is guaranteed to be a newly created file, containing text if that’s given as a string and empty otherwise. On MS-DOS, this function can truncate prefix to fit into the 8+3 file-name limits. If prefix is a relative file name, it is expanded against temporary-file-directory.

(make-temp-file "foo")
     ⇒ "/tmp/foo232J6v"

When make-temp-file returns, the file has been created and is empty. At that point, you should write the intended contents into the file.

If dir-flag is non-nil, make-temp-file creates an empty directory instead of an empty file. It returns the file name, not the directory name, of that directory. See Directory Names.

If suffix is non-nil, make-temp-file adds it at the end of the file name.

If text is a string, make-temp-file inserts it in the file.

To prevent conflicts among different libraries running in the same Emacs, each Lisp program that uses make-temp-file should have its own prefix. The number added to the end of prefix distinguishes between the same application running in different Emacs jobs. Additional added characters permit a large number of distinct names even in one Emacs job.

The default directory for temporary files is controlled by the variable temporary-file-directory. This variable gives the user a uniform way to specify the directory for all temporary files. Some programs use small-temporary-file-directory instead, if that is non-nil. To use it, you should expand the prefix against the proper directory before calling make-temp-file.

User Option: temporary-file-directory

This variable specifies the directory name for creating temporary files. Its value should be a directory name (see Directory Names), but it is good for Lisp programs to cope if the value is a directory’s file name instead. Using the value as the second argument to expand-file-name is a good way to achieve that.

The default value is determined in a reasonable way for your operating system; it is based on the TMPDIR, TMP and TEMP environment variables, with a fall-back to a system-dependent name if none of these variables is defined.

Even if you do not use make-temp-file to create the temporary file, you should still use this variable to decide which directory to put the file in. However, if you expect the file to be small, you should use small-temporary-file-directory first if that is non-nil.

User Option: small-temporary-file-directory

This variable specifies the directory name for creating certain temporary files, which are likely to be small.

If you want to write a temporary file which is likely to be small, you should compute the directory like this:

(make-temp-file
  (expand-file-name prefix
                    (or small-temporary-file-directory
                        temporary-file-directory)))
Function: make-temp-name base-name

This function generates a string that might be a unique file name. The name starts with base-name, and has several random characters appended to it, which are different in each Emacs job. It is like make-temp-file except that (i) it just constructs a name and does not create a file, (ii) base-name should be an absolute file name that is not magic, and (iii) if the returned file name is magic, it might name an existing file. See Making Certain File Names “Magic”.

Warning: In most cases, you should not use this function; use make-temp-file instead! This function is susceptible to a race condition, between the make-temp-name call and the creation of the file, which in some cases may cause a security hole.

Sometimes, it is necessary to create a temporary file on a remote host or a mounted directory. The following two functions support this.

Function: make-nearby-temp-file prefix &optional dir-flag suffix

This function is similar to make-temp-file, but it creates a temporary file as close as possible to default-directory. If prefix is a relative file name, and default-directory is a remote file name or located on a mounted file systems, the temporary file is created in the directory returned by the function temporary-file-directory. Otherwise, the function make-temp-file is used. prefix, dir-flag and suffix have the same meaning as in make-temp-file.

(let ((default-directory "/ssh:remotehost:"))
  (make-nearby-temp-file "foo"))
     ⇒ "/ssh:remotehost:/tmp/foo232J6v"
Function: temporary-file-directory

The directory for writing temporary files via make-nearby-temp-file. In case of a remote default-directory, this is a directory for temporary files on that remote host. If such a directory does not exist, or default-directory ought to be located on a mounted file system (see mounted-file-systems), the function returns default-directory. For a non-remote and non-mounted default-directory, the value of the variable temporary-file-directory is returned.

In order to extract the local part of the file’s name of a temporary file, use file-local-name (see Making Certain File Names “Magic”).


26.9.6 File Name Completion

This section describes low-level subroutines for completing a file name. For higher level functions, see Reading File Names.

Function: file-name-all-completions partial-filename directory

This function returns a list of all possible completions for a file whose name starts with partial-filename in directory directory. The order of the completions is the order of the files in the directory, which is unpredictable and conveys no useful information.

The argument partial-filename must be a file name containing no directory part and no slash (or backslash on some systems). The current buffer’s default directory is prepended to directory, if directory is not absolute.

In the following example, suppose that ~rms/lewis is the current default directory, and has five files whose names begin with ‘f’: foo, file~, file.c, file.c.~1~, and file.c.~2~.

(file-name-all-completions "f" "")
     ⇒ ("foo" "file~" "file.c.~2~"
                "file.c.~1~" "file.c")

(file-name-all-completions "fo" "")
     ⇒ ("foo")
Function: file-name-completion filename directory &optional predicate

This function completes the file name filename in directory directory. It returns the longest prefix common to all file names in directory directory that start with filename. If predicate is non-nil then it ignores possible completions that don’t satisfy predicate, after calling that function with one argument, the expanded absolute file name.

If only one match exists and filename matches it exactly, the function returns t. The function returns nil if directory directory contains no name starting with filename.

In the following example, suppose that the current default directory has five files whose names begin with ‘f’: foo, file~, file.c, file.c.~1~, and file.c.~2~.

(file-name-completion "fi" "")
     ⇒ "file"

(file-name-completion "file.c.~1" "")
     ⇒ "file.c.~1~"

(file-name-completion "file.c.~1~" "")
     ⇒ t

(file-name-completion "file.c.~3" "")
     ⇒ nil
User Option: completion-ignored-extensions

file-name-completion usually ignores file names that end in any string in this list. It does not ignore them when all the possible completions end in one of these suffixes. This variable has no effect on file-name-all-completions.

A typical value might look like this:

completion-ignored-extensions
     ⇒ (".o" ".elc" "~" ".dvi")

If an element of completion-ignored-extensions ends in a slash ‘/’, it signals a directory. The elements which do not end in a slash will never match a directory; thus, the above value will not filter out a directory named foo.elc.


26.9.7 Standard File Names

Sometimes, an Emacs Lisp program needs to specify a standard file name for a particular use—typically, to hold configuration data specified by the current user. Usually, such files should be located in the directory specified by user-emacs-directory, which is typically ~/.config/emacs/ or ~/.emacs.d/ by default (see How Emacs Finds Your Init File in The GNU Emacs Manual). For example, abbrev definitions are stored by default in ~/.config/emacs/abbrev_defs or ~/.emacs.d/abbrev_defs. The easiest way to specify such a file name is to use the function locate-user-emacs-file.

Function: locate-user-emacs-file base-name &optional old-name

This function returns an absolute file name for an Emacs-specific configuration or data file. The argument base-name should be a relative file name. The return value is the absolute name of a file in the directory specified by user-emacs-directory; if that directory does not exist, this function creates it.

If the optional argument old-name is non-nil, it specifies a file in the user’s home directory, ~/old-name. If such a file exists, the return value is the absolute name of that file, instead of the file specified by base-name. This argument is intended to be used by Emacs packages to provide backward compatibility. For instance, prior to the introduction of user-emacs-directory, the abbrev file was located in ~/.abbrev_defs. Here is the definition of abbrev-file-name:

(defcustom abbrev-file-name
  (locate-user-emacs-file "abbrev_defs" ".abbrev_defs")
  "Default name of file from which to read abbrevs."
  …
  :type 'file)

A lower-level function for standardizing file names, which locate-user-emacs-file uses as a subroutine, is convert-standard-filename.

Function: convert-standard-filename filename

This function returns a file name based on filename, which fits the conventions of the current operating system.

On GNU and other POSIX-like systems, this simply returns filename. On other operating systems, it may enforce system-specific file name conventions; for example, on MS-DOS this function performs a variety of changes to enforce MS-DOS file name limitations, including converting any leading ‘.’ to ‘_’ and truncating to three characters after the ‘.’.

The recommended way to use this function is to specify a name which fits the conventions of GNU and Unix systems, and pass it to convert-standard-filename.


26.10 Contents of Directories

A directory is a kind of file that contains other files entered under various names. Directories are a feature of the file system.

Emacs can list the names of the files in a directory as a Lisp list, or display the names in a buffer using the ls shell command. In the latter case, it can optionally display information about each file, depending on the options passed to the ls command.

Function: directory-files directory &optional full-name match-regexp nosort count

This function returns a list of the names of the files in the directory directory. By default, the list is in alphabetical order.

If full-name is non-nil, the function returns the files’ absolute file names. Otherwise, it returns the names relative to the specified directory.

If match-regexp is non-nil, this function returns only those file names whose non-directory part contain a match for that regular expression—the other file names are excluded from the list. On case-insensitive filesystems, the regular expression matching is case-insensitive.

If nosort is non-nil, directory-files does not sort the list, so you get the file names in no particular order. Use this if you want the utmost possible speed and don’t care what order the files are processed in. If the order of processing is visible to the user, then the user will probably be happier if you do sort the names.

If count is non-nil, the function will return names of first count number of files, or names of all files, whichever occurs first. count has to be an integer greater than zero.

(directory-files "~lewis")
     ⇒ ("#foo#" "#foo.el#" "." ".."
         "dired-mods.el" "files.texi"
         "files.texi.~1~")

An error is signaled if directory is not the name of a directory that can be read.

Function: directory-empty-p directory

This utility function returns t if given directory is an accessible directory and it does not contain any files, i.e., is an empty directory. It will ignore ‘.’ and ‘..’ on systems that return them as files in a directory.

Symbolic links to directories count as directories. See file-symlink-p to distinguish symlinks.

Function: directory-files-recursively directory regexp &optional include-directories predicate follow-symlinks

Return all files under directory whose names match regexp. This function searches the specified directory and its sub-directories, recursively, for files whose basenames (i.e., without the leading directories) match the specified regexp, and returns a list of the absolute file names of the matching files (see absolute file names). The file names are returned in depth-first order, meaning that files in some sub-directory are returned before the files in its parent directory. In addition, matching files found in each subdirectory are sorted alphabetically by their basenames. By default, directories whose names match regexp are omitted from the list, but if the optional argument include-directories is non-nil, they are included.

By default, all subdirectories are descended into. If predicate is t, errors when trying to descend into a subdirectory (for instance, if it’s not readable by this user) are ignored. If it’s neither nil nor t, it should be a function that takes one parameter (the subdirectory name) and should return non-nil if the directory is to be descended into.

Symbolic links to subdirectories are not followed by default, but if follow-symlinks is non-nil, they are followed.

Function: locate-dominating-file file name

Starting at file, go up the directory tree hierarchy looking for the first directory where name, a string, exists, and return that directory. If file is a file, its directory will serve as the starting point for the search; otherwise file should be a directory from which to start. The function looks in the starting directory, then in its parent, then in its parent’s parent, etc., until it either finds a directory with name or reaches the root directory of the filesystem without finding name – in the latter case the function returns nil.

The argument name can also be a predicate function. The predicate is called for every directory examined by the function, starting from file (even if file is not a directory). It is called with one argument (the file or directory) and should return non-nil if that directory is the one it is looking for.

Function: file-in-directory-p file dir

This function returns t if file is a file in directory dir, or in a subdirectory of dir. It also returns t if file and dir are the same directory. It compares the truenames of the two directories. If dir does not name an existing directory, the return value is nil.

Function: directory-files-and-attributes directory &optional full-name match-regexp nosort id-format count

This is similar to directory-files in deciding which files to report on and how to report their names. However, instead of returning a list of file names, it returns for each file a list (filename . attributes), where attributes is what file-attributes returns for that file. The optional argument id-format has the same meaning as the corresponding argument to file-attributes (see Definition of file-attributes).

Constant: directory-files-no-dot-files-regexp

This regular expression matches any file name except ‘.’ and ‘..’. More precisely, it matches parts of any nonempty string except those two. It is useful as the match-regexp argument to directory-files and directory-files-and-attributes:

(directory-files "/foo" nil directory-files-no-dot-files-regexp)

returns nil, if directory ‘/foo’ is empty.

Function: file-expand-wildcards pattern &optional full regexp

This function expands the wildcard pattern pattern, returning a list of file names that match it.

pattern is, by default, a “glob”/wildcard string, e.g., ‘"/tmp/*.png"’ or ‘"/*/*/foo.png"’, but can also be a regular expression if the optional regexp parameter is non-nil. In any case, the matches are applied per sub-directory, so a match can’t span a parent/sub directory.

If pattern is written as an absolute file name, the values are absolute also.

If pattern is written as a relative file name, it is interpreted relative to the current default directory. The file names returned are normally also relative to the current default directory. However, if full is non-nil, they are absolute.

Function: insert-directory file switches &optional wildcard full-directory-p

This function inserts (in the current buffer) a directory listing for directory file, formatted with ls according to switches. It leaves point after the inserted text. switches may be a string of options, or a list of strings representing individual options.

The argument file may be either a directory or a file specification including wildcard characters. If wildcard is non-nil, that means treat file as a file specification with wildcards.

If full-directory-p is non-nil, that means the directory listing is expected to show the full contents of a directory. You should specify t when file is a directory and switches do not contain ‘-d’. (The ‘-d’ option to ls says to describe a directory itself as a file, rather than showing its contents.)

On most systems, this function works by running a directory listing program whose name is in the variable insert-directory-program. If wildcard is non-nil, it also runs the shell specified by shell-file-name, to expand the wildcards.

MS-DOS and MS-Windows systems usually lack the standard Unix program ls, so this function emulates the standard Unix program ls with Lisp code.

As a technical detail, when switches contains the long ‘--dired’ option, insert-directory treats it specially, for the sake of dired. However, the normally equivalent short ‘-D’ option is just passed on to insert-directory-program, as any other option.

Variable: insert-directory-program

This variable’s value is the program to run to generate a directory listing for the function insert-directory. It is ignored on systems which generate the listing with Lisp code.


26.11 Creating, Copying and Deleting Directories

Most Emacs Lisp file-manipulation functions get errors when used on files that are directories. For example, you cannot delete a directory with delete-file. These special functions exist to create and delete directories.

Command: make-directory dirname &optional parents

This command creates a directory named dirname. If parents is non-nil, as is always the case in an interactive call, that means to create the parent directories first, if they don’t already exist. As a function, make-directory returns non-nil if dirname already exists as a directory and parents is non-nil, and returns nil if it successfully created dirname. mkdir is an alias for this.

Command: make-empty-file filename &optional parents

This command creates an empty file named filename. As make-directory, this command creates parent directories if parents is non-nil. If filename already exists, this command signals an error.

Command: copy-directory dirname newname &optional keep-time parents copy-contents

This command copies the directory named dirname to newname. If newname is a directory name, dirname will be copied to a subdirectory there. See Directory Names.

It always sets the file modes of the copied files to match the corresponding original file.

The third argument keep-time non-nil means to preserve the modification time of the copied files. A prefix arg makes keep-time non-nil.

The fourth argument parents says whether to create parent directories if they don’t exist. Interactively, this happens by default.

The fifth argument copy-contents, if non-nil, means to copy the contents of dirname directly into newname if the latter is a directory name, instead of copying dirname into it as a subdirectory.

Command: delete-directory dirname &optional recursive trash

This command deletes the directory named dirname. The function delete-file does not work for files that are directories; you must use delete-directory for them. If recursive is nil, and the directory contains any files, delete-directory signals an error. If recursive is non-nil, there is no error merely because the directory or its files are deleted by some other process before delete-directory gets to them.

delete-directory only follows symbolic links at the level of parent directories.

If the optional argument trash is non-nil and the variable delete-by-moving-to-trash is non-nil, this command moves the file into the system Trash instead of deleting it. See Miscellaneous File Operations in The GNU Emacs Manual. When called interactively, trash is t if no prefix argument is given, and nil otherwise.


26.12 Making Certain File Names “Magic”

You can implement special handling for certain file names. This is called making those names magic. The principal use for this feature is in implementing access to remote files (see Remote Files in The GNU Emacs Manual).

To define a kind of magic file name, you must supply a regular expression to define the class of names (all those that match the regular expression), plus a handler that implements all the primitive Emacs file operations for file names that match.

The variable file-name-handler-alist holds a list of handlers, together with regular expressions that determine when to apply each handler. Each element has this form:

(regexp . handler)

All the Emacs primitives for file access and file name transformation check the given file name against file-name-handler-alist. If the file name matches regexp, the primitives handle that file by calling handler.

The first argument given to handler is the name of the primitive, as a symbol; the remaining arguments are the arguments that were passed to that primitive. (The first of these arguments is most often the file name itself.) For example, if you do this:

(file-exists-p filename)

and filename has handler handler, then handler is called like this:

(funcall handler 'file-exists-p filename)

When a function takes two or more arguments that must be file names, it checks each of those names for a handler. For example, if you do this:

(expand-file-name filename dirname)

then it checks for a handler for filename and then for a handler for dirname. In either case, the handler is called like this:

(funcall handler 'expand-file-name filename dirname)

The handler then needs to figure out whether to handle filename or dirname.

If the specified file name matches more than one handler, the one whose match starts last in the file name gets precedence. This rule is chosen so that handlers for jobs such as uncompression are handled first, before handlers for jobs such as remote file access.

Here are the operations that a magic file name handler gets to handle:

abbreviate-file-name, access-file, add-name-to-file, byte-compiler-base-file-name,
copy-directory, copy-file, delete-directory, delete-file, diff-latest-backup-file, directory-file-name, directory-files, directory-files-and-attributes, dired-compress-file, dired-uncache, exec-path, expand-file-name,
file-accessible-directory-p, file-acl, file-attributes, file-directory-p, file-equal-p, file-executable-p, file-exists-p, file-in-directory-p, file-local-copy, file-locked-p, file-modes, file-name-all-completions, file-name-as-directory, file-name-case-insensitive-p, file-name-completion, file-name-directory, file-name-nondirectory, file-name-sans-versions, file-newer-than-file-p, file-notify-add-watch, file-notify-rm-watch, file-notify-valid-p, file-ownership-preserved-p, file-readable-p, file-regular-p, file-remote-p, file-selinux-context, file-symlink-p, file-system-info, file-truename, file-writable-p, find-backup-file-name,
get-file-buffer, insert-directory, insert-file-contents,
list-system-processes, load, lock-file, make-auto-save-file-name, make-directory, make-lock-file-name, make-nearby-temp-file, make-process, make-symbolic-link,
memory-info, process-attributes, process-file, rename-file, set-file-acl, set-file-modes, set-file-selinux-context, set-file-times, set-visited-file-modtime, shell-command, start-file-process, substitute-in-file-name,
temporary-file-directory, unhandled-file-name-directory, unlock-file, vc-registered, verify-visited-file-modtime,
write-region.

Handlers for insert-file-contents typically need to clear the buffer’s modified flag, with (set-buffer-modified-p nil), if the visit argument is non-nil. This also has the effect of unlocking the buffer if it is locked.

The handler function must handle all of the above operations, and possibly others to be added in the future. It need not implement all these operations itself—when it has nothing special to do for a certain operation, it can reinvoke the primitive, to handle the operation in the usual way. It should always reinvoke the primitive for an operation it does not recognize. Here’s one way to do this:

(defun my-file-handler (operation &rest args)
  ;; First check for the specific operations
  ;; that we have special handling for.
  (cond ((eq operation 'insert-file-contents) …)
        ((eq operation 'write-region) …)
        …
        ;; Handle any operation we don’t know about.
        (t (let ((inhibit-file-name-handlers
                  (cons 'my-file-handler
                        (and (eq inhibit-file-name-operation operation)
                             inhibit-file-name-handlers)))
                 (inhibit-file-name-operation operation))
             (apply operation args)))))

When a handler function decides to call the ordinary Emacs primitive for the operation at hand, it needs to prevent the primitive from calling the same handler once again, thus leading to an infinite recursion. The example above shows how to do this, with the variables inhibit-file-name-handlers and inhibit-file-name-operation. Be careful to use them exactly as shown above; the details are crucial for proper behavior in the case of multiple handlers, and for operations that have two file names that may each have handlers.

Handlers that don’t really do anything special for actual access to the file—such as the ones that implement completion of host names for remote file names—should have a non-nil safe-magic property. For instance, Emacs normally protects directory names it finds in PATH from becoming magic, if they look like magic file names, by prefixing them with ‘/:’. But if the handler that would be used for them has a non-nil safe-magic property, the ‘/:’ is not added.

A file name handler can have an operations property to declare which operations it handles in a nontrivial way. If this property has a non-nil value, it should be a list of operations; then only those operations will call the handler. This avoids inefficiency, but its main purpose is for autoloaded handler functions, so that they won’t be loaded except when they have real work to do.

Simply deferring all operations to the usual primitives does not work. For instance, if the file name handler applies to file-exists-p, then it must handle load itself, because the usual load code won’t work properly in that case. However, if the handler uses the operations property to say it doesn’t handle file-exists-p, then it need not handle load nontrivially.

Variable: inhibit-file-name-handlers

This variable holds a list of handlers whose use is presently inhibited for a certain operation.

Variable: inhibit-file-name-operation

The operation for which certain handlers are presently inhibited.

Function: find-file-name-handler file operation

This function returns the handler function for file name file, or nil if there is none. The argument operation should be the operation to be performed on the file—the value you will pass to the handler as its first argument when you call it. If operation equals inhibit-file-name-operation, or if it is not found in the operations property of the handler, this function returns nil.

Function: file-local-copy filename

This function copies file filename to an ordinary non-magic file on the local machine, if it isn’t on the local machine already. Magic file names should handle the file-local-copy operation if they refer to files on other machines. A magic file name that is used for other purposes than remote file access should not handle file-local-copy; then this function will treat the file as local.

If filename is local, whether magic or not, this function does nothing and returns nil. Otherwise it returns the file name of the local copy file.

Function: file-remote-p filename &optional identification connected

This function tests whether filename is a remote file. If filename is local (not remote), the return value is nil. If filename is indeed remote, the return value is a string that identifies the remote system.

This identifier string can include a host name and a user name, as well as characters designating the method used to access the remote system. For example, the remote identifier string for the file name /sudo::/some/file is /sudo:root@localhost:.

If file-remote-p returns the same identifier for two different file names, that means they are stored on the same file system and can be accessed locally with respect to each other. This means, for example, that it is possible to start a remote process accessing both files at the same time. Implementers of file name handlers need to ensure this principle is valid.

identification specifies which part of the identifier shall be returned as string. identification can be the symbol method, user or host; any other value is handled like nil and means to return the complete identifier string. In the example above, the remote user identifier string would be root.

If connected is non-nil, this function returns nil even if filename is remote, if Emacs has no network connection to its host. This is useful when you want to avoid the delay of making connections when they don’t exist.

Function: unhandled-file-name-directory filename

This function returns the name of a directory that is not magic. For a non-magic filename it returns the corresponding directory name (see Directory Names). For a magic filename, it invokes the file name handler, which therefore decides what value to return. If filename is not accessible from a local process, then the file name handler should indicate that by returning nil.

This is useful for running a subprocess; every subprocess must have a non-magic directory to serve as its current directory, and this function is a good way to come up with one.

Function: file-local-name filename

This function returns the local part of filename. This is the part of the file’s name that identifies it on the remote host, and is typically obtained by removing from the remote file name the parts that specify the remote host and the method of accessing it. For example:

(file-local-name "/ssh:user@host:/foo/bar")
     ⇒ "/foo/bar"

For a remote filename, this function returns a file name which could be used directly as an argument of a remote process (see Creating an Asynchronous Process, and see Creating a Synchronous Process), and as the program to run on the remote host. If filename is local, this function returns it unchanged.

User Option: remote-file-name-inhibit-cache

The attributes of remote files can be cached for better performance. If they are changed outside of Emacs’s control, the cached values become invalid, and must be reread.

When this variable is set to nil, cached values are never expired. Use this setting with caution, only if you are sure nothing other than Emacs ever changes the remote files. If it is set to t, cached values are never used. This is the safest value, but could result in performance degradation.

A compromise is to set it to a positive number. This means that cached values are used for that amount of seconds since they were cached. If a remote file is checked regularly, it might be a good idea to let-bind this variable to a value less than the time period between consecutive checks. For example:

(defun display-time-file-nonempty-p (file)
  (let ((remote-file-name-inhibit-cache
         (- display-time-interval 5)))
    (and (file-exists-p file)
         (< 0 (file-attribute-size
               (file-attributes
                (file-chase-links file)))))))

26.13 File Format Conversion

Emacs performs several steps to convert the data in a buffer (text, text properties, and possibly other information) to and from a representation suitable for storing into a file. This section describes the fundamental functions that perform this format conversion, namely insert-file-contents for reading a file into a buffer, and write-region for writing a buffer into a file.


26.13.1 Overview

The function insert-file-contents:

  • initially, inserts bytes from the file into the buffer;
  • decodes bytes to characters as appropriate;
  • processes formats as defined by entries in format-alist; and
  • calls functions in after-insert-file-functions.

The function write-region:

  • initially, calls functions in write-region-annotate-functions;
  • processes formats as defined by entries in format-alist;
  • encodes characters to bytes as appropriate; and
  • modifies the file with the bytes.

This shows the symmetry of the lowest-level operations; reading and writing handle things in opposite order. The rest of this section describes the two facilities surrounding the three variables named above, as well as some related functions. Coding Systems, for details on character encoding and decoding.


26.13.2 Round-Trip Specification

The most general of the two facilities is controlled by the variable format-alist, a list of file format specifications, which describe textual representations used in files for the data in an Emacs buffer. The descriptions for reading and writing are paired, which is why we call this “round-trip” specification (see Piecemeal Specification, for non-paired specification).

Variable: format-alist

This list contains one format definition for each defined file format. Each format definition is a list of this form:

(name doc-string regexp from-fn to-fn modify mode-fn preserve)

Here is what the elements in a format definition mean:

name

The name of this format.

doc-string

A documentation string for the format.

regexp

A regular expression which is used to recognize files represented in this format. If nil, the format is never applied automatically.

from-fn

A shell command or function to decode data in this format (to convert file data into the usual Emacs data representation).

A shell command is represented as a string; Emacs runs the command as a filter to perform the conversion.

If from-fn is a function, it is called with two arguments, begin and end, which specify the part of the buffer it should convert. It should convert the text by editing it in place. Since this can change the length of the text, from-fn should return the modified end position.

One responsibility of from-fn is to make sure that the beginning of the file no longer matches regexp. Otherwise it is likely to get called again. Also, from-fn must not involve buffers or files other than the one being decoded, otherwise the internal buffer used for formatting might be overwritten.

to-fn

A shell command or function to encode data in this format—that is, to convert the usual Emacs data representation into this format.

If to-fn is a string, it is a shell command; Emacs runs the command as a filter to perform the conversion.

If to-fn is a function, it is called with three arguments: begin and end, which specify the part of the buffer it should convert, and buffer, which specifies which buffer. There are two ways it can do the conversion:

  • By editing the buffer in place. In this case, to-fn should return the end-position of the range of text, as modified.
  • By returning a list of annotations. This is a list of elements of the form (position . string), where position is an integer specifying the relative position in the text to be written, and string is the annotation to add there. The list must be sorted in order of position when to-fn returns it.

    When write-region actually writes the text from the buffer to the file, it intermixes the specified annotations at the corresponding positions. All this takes place without modifying the buffer.

to-fn must not involve buffers or files other than the one being encoded, otherwise the internal buffer used for formatting might be overwritten.

modify

A flag, t if the encoding function modifies the buffer, and nil if it works by returning a list of annotations.

mode-fn

A minor-mode function to call after visiting a file converted from this format. The function is called with one argument, the integer 1; that tells a minor-mode function to enable the mode.

preserve

A flag, t if format-write-file should not remove this format from buffer-file-format.

The function insert-file-contents automatically recognizes file formats when it reads the specified file. It checks the text of the beginning of the file against the regular expressions of the format definitions, and if it finds a match, it calls the decoding function for that format. Then it checks all the known formats over again. It keeps checking them until none of them is applicable.

Visiting a file, with find-file-noselect or the commands that use it, performs conversion likewise (because it calls insert-file-contents); it also calls the mode function for each format that it decodes. It stores a list of the format names in the buffer-local variable buffer-file-format.

Variable: buffer-file-format

This variable states the format of the visited file. More precisely, this is a list of the file format names that were decoded in the course of visiting the current buffer’s file. It is always buffer-local in all buffers.

When write-region writes data into a file, it first calls the encoding functions for the formats listed in buffer-file-format, in the order of appearance in the list.

Command: format-write-file file format &optional confirm

This command writes the current buffer contents into the file file in a format based on format, which is a list of format names. It constructs the actual format starting from format, then appending any elements from the value of buffer-file-format with a non-nil preserve flag (see above), if they are not already present in format. It then updates buffer-file-format with this format, making it the default for future saves. Except for the format argument, this command is similar to write-file. In particular, confirm has the same meaning and interactive treatment as the corresponding argument to write-file. See Definition of write-file.

Command: format-find-file file format

This command finds the file file, converting it according to format format. It also makes format the default if the buffer is saved later.

The argument format is a list of format names. If format is nil, no conversion takes place. Interactively, typing just RET for format specifies nil.

Command: format-insert-file file format &optional beg end

This command inserts the contents of file file, converting it according to format format. If beg and end are non-nil, they specify which part of the file to read, as in insert-file-contents (see Reading from Files).

The return value is like what insert-file-contents returns: a list of the absolute file name and the length of the data inserted (after conversion).

The argument format is a list of format names. If format is nil, no conversion takes place. Interactively, typing just RET for format specifies nil.

Variable: buffer-auto-save-file-format

This variable specifies the format to use for auto-saving. Its value is a list of format names, just like the value of buffer-file-format; however, it is used instead of buffer-file-format for writing auto-save files. If the value is t, the default, auto-saving uses the same format as a regular save in the same buffer. This variable is always buffer-local in all buffers.


26.13.3 Piecemeal Specification

In contrast to the round-trip specification described in the previous subsection (see Round-Trip Specification), you can use the variables after-insert-file-functions and write-region-annotate-functions to separately control the respective reading and writing conversions.

Conversion starts with one representation and produces another representation. When there is only one conversion to do, there is no conflict about what to start with. However, when there are multiple conversions involved, conflict may arise when two conversions need to start with the same data.

This situation is best understood in the context of converting text properties during write-region. For example, the character at position 42 in a buffer is ‘X’ with a text property foo. If the conversion for foo is done by inserting into the buffer, say, ‘FOO:’, then that changes the character at position 42 from ‘X’ to ‘F’. The next conversion will start with the wrong data straight away.

To avoid conflict, cooperative conversions do not modify the buffer, but instead specify annotations, a list of elements of the form (position . string), sorted in order of increasing position.

If there is more than one conversion, write-region merges their annotations destructively into one sorted list. Later, when the text from the buffer is actually written to the file, it intermixes the specified annotations at the corresponding positions. All this takes place without modifying the buffer.

In contrast, when reading, the annotations intermixed with the text are handled immediately. insert-file-contents sets point to the beginning of some text to be converted, then calls the conversion functions with the length of that text. These functions should always return with point at the beginning of the inserted text. This approach makes sense for reading because annotations removed by the first converter can’t be mistakenly processed by a later converter. Each conversion function should scan for the annotations it recognizes, remove the annotation, modify the buffer text (to set a text property, for example), and return the updated length of the text, as it stands after those changes. The value returned by one function becomes the argument to the next function.

Variable: write-region-annotate-functions

A list of functions for write-region to call. Each function in the list is called with two arguments: the start and end of the region to be written. These functions should not alter the contents of the buffer. Instead, they should return annotations.

As a special case, a function may return with a different buffer current. Emacs takes this to mean that the current buffer contains altered text to be output. It therefore changes the start and end arguments of the write-region call, giving them the values of point-min and point-max in the new buffer, respectively. It also discards all previous annotations, because they should have been dealt with by this function.

Variable: write-region-post-annotation-function

The value of this variable, if non-nil, should be a function. This function is called, with no arguments, after write-region has completed.

If any function in write-region-annotate-functions returns with a different buffer current, Emacs calls write-region-post-annotation-function more than once. Emacs calls it with the last buffer that was current, and again with the buffer before that, and so on back to the original buffer.

Thus, a function in write-region-annotate-functions can create a buffer, give this variable the local value of kill-buffer in that buffer, set up the buffer with altered text, and make the buffer current. The buffer will be killed after write-region is done.

Variable: after-insert-file-functions

Each function in this list is called by insert-file-contents with one argument, the number of characters inserted, and with point at the beginning of the inserted text. Each function should leave point unchanged, and return the new character count describing the inserted text as modified by the function.

We invite users to write Lisp programs to store and retrieve text properties in files, using these hooks, and thus to experiment with various data formats and find good ones. Eventually we hope users will produce good, general extensions we can install in Emacs.

We suggest not trying to handle arbitrary Lisp objects as text property names or values—because a program that general is probably difficult to write, and slow. Instead, choose a set of possible data types that are reasonably flexible, and not too hard to encode.


27 Backups and Auto-Saving

Backup files and auto-save files are two methods by which Emacs tries to protect the user from the consequences of crashes or of the user’s own errors. Auto-saving preserves the text from earlier in the current editing session; backup files preserve file contents prior to the current session.


27.1 Backup Files

A backup file is a copy of the old contents of a file you are editing. Emacs makes a backup file the first time you save a buffer into its visited file. Thus, normally, the backup file contains the contents of the file as it was before the current editing session. The contents of the backup file normally remain unchanged once it exists.

Backups are usually made by renaming the visited file to a new name. Optionally, you can specify that backup files should be made by copying the visited file. This choice makes a difference for files with multiple names; it also can affect whether the edited file remains owned by the original owner or becomes owned by the user editing it.

By default, Emacs makes a single backup file for each file edited. You can alternatively request numbered backups; then each new backup file gets a new name. You can delete old numbered backups when you don’t want them any more, or Emacs can delete them automatically.

For performance, the operating system may not write the backup file’s contents to secondary storage immediately, or may alias the backup data with the original until one or the other is later modified. See Files and Secondary Storage.


27.1.1 Making Backup Files

Function: backup-buffer

This function makes a backup of the file visited by the current buffer, if appropriate. It is called by save-buffer before saving the buffer the first time.

If a backup was made by renaming, the return value is a cons cell of the form (modes extra-alist backupname), where modes are the mode bits of the original file, as returned by file-modes (see Testing Accessibility), extra-alist is an alist describing the original file’s extended attributes, as returned by file-extended-attributes (see Extended File Attributes), and backupname is the name of the backup.

In all other cases (i.e., if a backup was made by copying or if no backup was made), this function returns nil.

Variable: buffer-backed-up

This buffer-local variable says whether this buffer’s file has been backed up on account of this buffer. If it is non-nil, the backup file has been written. Otherwise, the file should be backed up when it is next saved (if backups are enabled). This is a permanent local; kill-all-local-variables does not alter it.

User Option: make-backup-files

This variable determines whether or not to make backup files. If it is non-nil, then Emacs creates a backup of each file when it is saved for the first time—provided that backup-inhibited is nil (see below).

The following example shows how to change the make-backup-files variable only in the Rmail buffers and not elsewhere. Setting it nil stops Emacs from making backups of these files, which may save disk space. (You would put this code in your init file.)

(add-hook 'rmail-mode-hook
          (lambda () (setq-local make-backup-files nil)))
Variable: backup-enable-predicate

This variable’s value is a function to be called on certain occasions to decide whether a file should have backup files. The function receives one argument, an absolute file name to consider. If the function returns nil, backups are disabled for that file. Otherwise, the other variables in this section say whether and how to make backups.

The default value is normal-backup-enable-predicate, which checks for files in temporary-file-directory and small-temporary-file-directory.

Variable: backup-inhibited

If this variable is non-nil, backups are inhibited. It records the result of testing backup-enable-predicate on the visited file name. It can also coherently be used by other mechanisms that inhibit backups based on which file is visited. For example, VC sets this variable non-nil to prevent making backups for files managed with a version control system.

This is a permanent local, so that changing the major mode does not lose its value. Major modes should not set this variable—they should set make-backup-files instead.

User Option: backup-directory-alist

This variable’s value is an alist of filename patterns and backup directories. Each element looks like

(regexp . directory)

Backups of files with names matching regexp will be made in directory. directory may be relative or absolute. If it is absolute, so that all matching files are backed up into the same directory, the file names in this directory will be the full name of the file backed up with all directory separators changed to ‘!’ to prevent clashes. This will not work correctly if your filesystem truncates the resulting name.

For the common case of all backups going into one directory, the alist should contain a single element pairing ‘"."’ with the appropriate directory.

If this variable is nil (the default), or it fails to match a filename, the backup is made in the original file’s directory.

On MS-DOS filesystems without long names this variable is always ignored.

User Option: make-backup-file-name-function

This variable’s value is a function to use for making backup file names. The function make-backup-file-name calls it. See Naming Backup Files.

This could be buffer-local to do something special for specific files. If you change it, you may need to change backup-file-name-p and file-name-sans-versions too.


27.1.2 Backup by Renaming or by Copying?

There are two ways that Emacs can make a backup file:

  • Emacs can rename the original file so that it becomes a backup file, and then write the buffer being saved into a new file. After this procedure, any other names (i.e., hard links) of the original file now refer to the backup file. The new file is owned by the user doing the editing, and its group is the default for new files written by the user in that directory.
  • Emacs can copy the original file into a backup file, and then overwrite the original file with new contents. After this procedure, any other names (i.e., hard links) of the original file continue to refer to the current (updated) version of the file. The file’s owner and group will be unchanged.

The first method, renaming, is the default.

The variable backup-by-copying, if non-nil, says to use the second method, which is to copy the original file and overwrite it with the new buffer contents. The variable file-precious-flag, if non-nil, also has this effect (as a sideline of its main significance). See Saving Buffers.

User Option: backup-by-copying

If this variable is non-nil, Emacs always makes backup files by copying. The default is nil.

The following three variables, when non-nil, cause the second method to be used in certain special cases. They have no effect on the treatment of files that don’t fall into the special cases.

User Option: backup-by-copying-when-linked

If this variable is non-nil, Emacs makes backups by copying for files with multiple names (hard links). The default is nil.

This variable is significant only if backup-by-copying is nil, since copying is always used when that variable is non-nil.

User Option: backup-by-copying-when-mismatch

If this variable is non-nil (the default), Emacs makes backups by copying in cases where renaming would change either the owner or the group of the file.

The value has no effect when renaming would not alter the owner or group of the file; that is, for files which are owned by the user and whose group matches the default for a new file created there by the user.

This variable is significant only if backup-by-copying is nil, since copying is always used when that variable is non-nil.

User Option: backup-by-copying-when-privileged-mismatch

This variable, if non-nil, specifies the same behavior as backup-by-copying-when-mismatch, but only for certain user-id and group-id values: namely, those less than or equal to a certain number. You set this variable to that number.

Thus, if you set backup-by-copying-when-privileged-mismatch to 0, backup by copying is done for the superuser and group 0 only, when necessary to prevent a change in the owner of the file.

The default is 200.


27.1.3 Making and Deleting Numbered Backup Files

If a file’s name is foo, the names of its numbered backup versions are foo.~v~, for various integers v, like this: foo.~1~, foo.~2~, foo.~3~, …, foo.~259~, and so on.

User Option: version-control

This variable controls whether to make a single non-numbered backup file or multiple numbered backups.

nil

Make numbered backups if the visited file already has numbered backups; otherwise, do not. This is the default.

never

Do not make numbered backups.

anything else

Make numbered backups.

The use of numbered backups ultimately leads to a large number of backup versions, which must then be deleted. Emacs can do this automatically or it can ask the user whether to delete them.

User Option: kept-new-versions

The value of this variable is the number of newest versions to keep when a new numbered backup is made. The newly made backup is included in the count. The default value is 2.

User Option: kept-old-versions

The value of this variable is the number of oldest versions to keep when a new numbered backup is made. The default value is 2.

If there are backups numbered 1, 2, 3, 5, and 7, and both of these variables have the value 2, then the backups numbered 1 and 2 are kept as old versions and those numbered 5 and 7 are kept as new versions; backup version 3 is excess. The function find-backup-file-name (see Naming Backup Files) is responsible for determining which backup versions to delete, but does not delete them itself.

User Option: delete-old-versions

If this variable is t, then saving a file deletes excess backup versions silently. If it is nil, that means to ask for confirmation before deleting excess backups. Otherwise, they are not deleted at all.

User Option: dired-kept-versions

This variable specifies how many of the newest backup versions to keep in the Dired command . (dired-clean-directory). That’s the same thing kept-new-versions specifies when you make a new backup file. The default is 2.


27.1.4 Naming Backup Files

The functions in this section are documented mainly because you can customize the naming conventions for backup files by redefining them. If you change one, you probably need to change the rest.

Function: backup-file-name-p filename

This function returns a non-nil value if filename is a possible name for a backup file. It just checks the name, not whether a file with the name filename exists.

(backup-file-name-p "foo")
     ⇒ nil
(backup-file-name-p "foo~")
     ⇒ 3

The standard definition of this function is as follows:

(defun backup-file-name-p (file)
  "Return non-nil if FILE is a backup file \
name (numeric or not)..."
  (string-match "~\\'" file))

Thus, the function returns a non-nil value if the file name ends with a ‘~’. (We use a backslash to split the documentation string’s first line into two lines in the text, but produce just one line in the string itself.)

This simple expression is placed in a separate function to make it easy to redefine for customization.

Function: make-backup-file-name filename

This function returns a string that is the name to use for a non-numbered backup file for file filename. On Unix, this is just filename with a tilde appended.

The standard definition of this function, on most operating systems, is as follows:

(defun make-backup-file-name (file)
  "Create the non-numeric backup file name for FILE..."
  (concat file "~"))

You can change the backup-file naming convention by redefining this function. The following example redefines make-backup-file-name to prepend a ‘.’ in addition to appending a tilde:

(defun make-backup-file-name (filename)
  (expand-file-name
    (concat "." (file-name-nondirectory filename) "~")
    (file-name-directory filename)))

(make-backup-file-name "backups.texi")
     ⇒ ".backups.texi~"

Some parts of Emacs, including some Dired commands, assume that backup file names end with ‘~’. If you do not follow that convention, it will not cause serious problems, but these commands may give less-than-desirable results.

Function: find-backup-file-name filename

This function computes the file name for a new backup file for filename. It may also propose certain existing backup files for deletion. find-backup-file-name returns a list whose CAR is the name for the new backup file and whose CDR is a list of backup files whose deletion is proposed. The value can also be nil, which means not to make a backup.

Two variables, kept-old-versions and kept-new-versions, determine which backup versions should be kept. This function keeps those versions by excluding them from the CDR of the value. See Making and Deleting Numbered Backup Files.

In this example, the value says that ~rms/foo.~5~ is the name to use for the new backup file, and ~rms/foo.~3~ is an excess version that the caller should consider deleting now.

(find-backup-file-name "~rms/foo")
     ⇒ ("~rms/foo.~5~" "~rms/foo.~3~")
Function: file-backup-file-names filename

This function returns a list of all the backup file names for filename, or nil if there are none. The files are sorted by modification time, descending, so that the most recent files are first.

Function: file-newest-backup filename

This function returns the first element of the list returned by file-backup-file-names.

Some file comparison commands use this function so that they can automatically compare a file with its most recent backup.


27.2 Auto-Saving

Emacs periodically saves all files that you are visiting; this is called auto-saving. Auto-saving prevents you from losing more than a limited amount of work if the system crashes. By default, auto-saves happen every 300 keystrokes, or after around 30 seconds of idle time. See Auto-Saving: Protection Against Disasters in The GNU Emacs Manual, for information on auto-save for users. Here we describe the functions used to implement auto-saving and the variables that control them.

Variable: buffer-auto-save-file-name

This buffer-local variable is the name of the file used for auto-saving the current buffer. It is nil if the buffer should not be auto-saved.

buffer-auto-save-file-name
     ⇒ "/xcssun/users/rms/lewis/#backups.texi#"
Command: auto-save-mode arg

This is the mode command for Auto Save mode, a buffer-local minor mode. When Auto Save mode is enabled, auto-saving is enabled in the buffer. The calling convention is the same as for other minor mode commands (see Conventions for Writing Minor Modes).

Unlike most minor modes, there is no auto-save-mode variable. Auto Save mode is enabled if buffer-auto-save-file-name is non-nil and buffer-saved-size (see below) is non-zero.

Variable: auto-save-file-name-transforms

This variable lists transforms to apply to buffer’s file name before making the auto-save file name.

Each transform is a list of the form (regexp replacement [uniquify]). regexp is a regular expression to match against the file name; if it matches, replace-match is used to replace the matching part with replacement. If the optional element uniquify is non-nil, the auto-save file name is constructed by concatenating the directory part of the transformed file name with the buffer’s file name in which all directory separators were changed to ‘!’ to prevent clashes. (This will not work correctly if your filesystem truncates the resulting name.)

If uniquify is one of the members of secure-hash-algorithms, Emacs constructs the nondirectory part of the auto-save file name by applying that secure-hash to the buffer file name. This avoids any risk of excessively long file names.

All the transforms in the list are tried, in the order they are listed. When one transform applies, its result is final; no further transforms are tried.

The default value is set up to put the auto-save files of remote files into the temporary directory (see Generating Unique File Names).

On MS-DOS filesystems without long names this variable is always ignored.

Function: auto-save-file-name-p filename

This function returns a non-nil value if filename is a string that could be the name of an auto-save file. It assumes the usual naming convention for auto-save files: a name that begins and ends with hash marks (‘#’) is a possible auto-save file name. The argument filename should not contain a directory part.

(make-auto-save-file-name)
     ⇒ "/xcssun/users/rms/lewis/#backups.texi#"
(auto-save-file-name-p "#backups.texi#")
     ⇒ 0
(auto-save-file-name-p "backups.texi")
     ⇒ nil
Function: make-auto-save-file-name

This function returns the file name to use for auto-saving the current buffer. This is just the file name with hash marks (‘#’) prepended and appended to it. This function does not look at the variable auto-save-visited-file-name (described below); callers of this function should check that variable first.

(make-auto-save-file-name)
     ⇒ "/xcssun/users/rms/lewis/#backups.texi#"
User Option: auto-save-visited-file-name

If this variable is non-nil, Emacs auto-saves buffers in the files they are visiting. That is, the auto-save is done in the same file that you are editing. Normally, this variable is nil, so auto-save files have distinct names that are created by make-auto-save-file-name.

When you change the value of this variable, the new value does not take effect in an existing buffer until the next time auto-save mode is reenabled in it. If auto-save mode is already enabled, auto-saves continue to go in the same file name until auto-save-mode is called again.

Note that setting this variable to a non-nil value does not change the fact that auto-saving is different from saving the buffer; e.g., the hooks described in Saving Buffers are not run when a buffer is auto-saved.

Function: recent-auto-save-p

This function returns t if the current buffer has been auto-saved since the last time it was read in or saved.

Function: set-buffer-auto-saved

This function marks the current buffer as auto-saved. The buffer will not be auto-saved again until the buffer text is changed again. The function returns nil.

User Option: auto-save-interval

The value of this variable specifies how often to do auto-saving, in terms of number of input events. Each time this many additional input events are read, Emacs does auto-saving for all buffers in which that is enabled. Setting this to zero disables autosaving based on the number of characters typed.

User Option: auto-save-timeout

The value of this variable is the number of seconds of idle time that should cause auto-saving. Each time the user pauses for this long, Emacs does auto-saving for all buffers in which that is enabled. (If the current buffer is large, the specified timeout is multiplied by a factor that increases as the size increases; for a million-byte buffer, the factor is almost 4.)

If the value is zero or nil, then auto-saving is not done as a result of idleness, only after a certain number of input events as specified by auto-save-interval.

Variable: auto-save-hook

This normal hook is run whenever an auto-save is about to happen.

User Option: auto-save-default

If this variable is non-nil, buffers that are visiting files have auto-saving enabled by default. Otherwise, they do not.

Command: do-auto-save &optional no-message current-only

This function auto-saves all buffers that need to be auto-saved. It saves all buffers for which auto-saving is enabled and that have been changed since the previous auto-save.

If any buffers are auto-saved, do-auto-save normally displays a message saying ‘Auto-saving...’ in the echo area while auto-saving is going on. However, if no-message is non-nil, the message is inhibited.

If current-only is non-nil, only the current buffer is auto-saved.

Function: delete-auto-save-file-if-necessary &optional force

This function deletes the current buffer’s auto-save file if delete-auto-save-files is non-nil. It is called every time a buffer is saved.

Unless force is non-nil, this function only deletes the file if it was written by the current Emacs session since the last true save.

User Option: delete-auto-save-files

This variable is used by the function delete-auto-save-file-if-necessary. If it is non-nil, Emacs deletes auto-save files when a true save is done (in the visited file). This saves disk space and unclutters your directory.

Function: rename-auto-save-file

This function adjusts the current buffer’s auto-save file name if the visited file name has changed. It also renames an existing auto-save file, if it was made in the current Emacs session. If the visited file name has not changed, this function does nothing.

Variable: buffer-saved-size

The value of this buffer-local variable is the length of the current buffer, when it was last read in, saved, or auto-saved. This is used to detect a substantial decrease in size, and turn off auto-saving in response.

If it is −1, that means auto-saving is temporarily shut off in this buffer due to a substantial decrease in size. Explicitly saving the buffer stores a positive value in this variable, thus reenabling auto-saving. Turning auto-save mode off or on also updates this variable, so that the substantial decrease in size is forgotten.

If it is −2, that means this buffer should disregard changes in buffer size; in particular, it should not shut off auto-saving temporarily due to changes in buffer size.

Variable: auto-save-list-file-name

This variable (if non-nil) specifies a file for recording the names of all the auto-save files. Each time Emacs does auto-saving, it writes two lines into this file for each buffer that has auto-saving enabled. The first line gives the name of the visited file (it’s empty if the buffer has none), and the second gives the name of the auto-save file.

When Emacs exits normally, it deletes this file; if Emacs crashes, you can look in the file to find all the auto-save files that might contain work that was otherwise lost. The recover-session command uses this file to find them.

The default name for this file specifies your home directory and starts with ‘.saves-’. It also contains the Emacs process ID and the host name.

User Option: auto-save-list-file-prefix

After Emacs reads your init file, it initializes auto-save-list-file-name (if you have not already set it non-nil) based on this prefix, adding the host name and process ID. If you set this to nil in your init file, then Emacs does not initialize auto-save-list-file-name.


27.3 Reverting

If you have made extensive changes to a file and then change your mind about them, you can get rid of them by reading in the previous version of the file with the revert-buffer command. See Reverting a Buffer in The GNU Emacs Manual.

Command: revert-buffer &optional ignore-auto noconfirm preserve-modes

This command replaces the buffer text with the text of the visited file on disk. This action undoes all changes since the file was visited or saved.

By default, if the latest auto-save file is more recent than the visited file, and the argument ignore-auto is nil, revert-buffer asks the user whether to use that auto-save instead. When you invoke this command interactively, ignore-auto is t if there is no numeric prefix argument; thus, the interactive default is not to check the auto-save file.

Normally, revert-buffer asks for confirmation before it changes the buffer; but if the argument noconfirm is non-nil, revert-buffer does not ask for confirmation.

Normally, this command reinitializes the buffer’s major and minor modes using normal-mode. But if preserve-modes is non-nil, the modes remain unchanged.

Reverting tries to preserve marker positions in the buffer by using the replacement feature of insert-file-contents. If the buffer contents and the file contents are identical before the revert operation, reverting preserves all the markers. If they are not identical, reverting does change the buffer; in that case, it preserves the markers in the unchanged text (if any) at the beginning and end of the buffer. Preserving any additional markers would be problematic.

When reverting from non-file sources, markers are usually not preserved, but this is up to the specific revert-buffer-function implementation.

Variable: revert-buffer-in-progress-p

revert-buffer binds this variable to a non-nil value while it is working.

You can customize how revert-buffer does its work by setting the variables described in the rest of this section.

User Option: revert-without-query

This variable holds a list of files that should be reverted without query. The value is a list of regular expressions. If the visited file name matches one of these regular expressions, and the file has changed on disk but the buffer is not modified, then revert-buffer reverts the file without asking the user for confirmation.

Some major modes customize revert-buffer by making buffer-local bindings for these variables:

Variable: revert-buffer-function

The value of this variable is the function to use to revert this buffer. It should be a function with two optional arguments to do the work of reverting. The two optional arguments, ignore-auto and noconfirm, are the arguments that revert-buffer received.

Modes such as Dired mode, in which the text being edited does not consist of a file’s contents but can be regenerated in some other fashion, can give this variable a buffer-local value that is a special function to regenerate the contents.

Variable: revert-buffer-insert-file-contents-function

The value of this variable specifies the function to use to insert the updated contents when reverting this buffer. The function receives two arguments: first the file name to use; second, t if the user has asked to read the auto-save file.

The reason for a mode to change this variable instead of revert-buffer-function is to avoid duplicating or replacing the rest of what revert-buffer does: asking for confirmation, clearing the undo list, deciding the proper major mode, and running the hooks listed below.

Variable: before-revert-hook

This normal hook is run by the default revert-buffer-function before inserting the modified contents. A custom revert-buffer-function may or may not run this hook.

Variable: after-revert-hook

This normal hook is run by the default revert-buffer-function after inserting the modified contents. A custom revert-buffer-function may or may not run this hook.

Emacs can revert buffers automatically. It does that by default for buffers visiting files. The following describes how to add support for auto-reverting new types of buffers.

First, such buffers must have a suitable revert-buffer-function and buffer-stale-function defined.

Variable: buffer-stale-function

The value of this variable specifies a function to call to check whether a buffer needs reverting. The default value only handles buffers that are visiting files, by checking their modification time. Buffers that are not visiting files require a custom function of one optional argument noconfirm. The function should return non-nil if the buffer should be reverted. The buffer is current when this function is called.

While this function is mainly intended for use in auto-reverting, it could be used for other purposes as well. For instance, if auto-reverting is not enabled, it could be used to warn the user that the buffer needs reverting. The idea behind the noconfirm argument is that it should be t if the buffer is going to be reverted without asking the user and nil if the function is just going to be used to warn the user that the buffer is out of date. In particular, for use in auto-reverting, noconfirm is t. If the function is only going to be used for auto-reverting, you can ignore the noconfirm argument.

If you just want to automatically auto-revert every auto-revert-interval seconds (like the Buffer Menu), use:

(setq-local buffer-stale-function
     (lambda (&optional noconfirm) 'fast))

in the buffer’s mode function.

The special return value ‘fast’ tells the caller that the need for reverting was not checked, but that reverting the buffer is fast. It also tells Auto Revert not to print any revert messages, even if auto-revert-verbose is non-nil. This is important, as getting revert messages every auto-revert-interval seconds can be very annoying. The information provided by this return value could also be useful if the function is consulted for purposes other than auto-reverting.

Once the buffer has a suitable revert-buffer-function and buffer-stale-function, several problems usually remain.

The buffer will only auto-revert if it is marked unmodified. Hence, you will have to make sure that various functions mark the buffer modified if and only if either the buffer contains information that might be lost by reverting, or there is reason to believe that the user might be inconvenienced by auto-reverting, because he is actively working on the buffer. The user can always override this by manually adjusting the modified status of the buffer. To support this, calling the revert-buffer-function on a buffer that is marked unmodified should always keep the buffer marked unmodified.

It is important to assure that point does not continuously jump around as a consequence of auto-reverting. Of course, moving point might be inevitable if the buffer radically changes.

You should make sure that the revert-buffer-function does not print messages that unnecessarily duplicate Auto Revert’s own messages, displayed if auto-revert-verbose is t, and effectively override a nil value for auto-revert-verbose. Hence, adapting a mode for auto-reverting often involves getting rid of such messages. This is especially important for buffers that automatically revert every auto-revert-interval seconds.

If the new auto-reverting is part of Emacs, you should mention it in the documentation string of global-auto-revert-non-file-buffers.

Similarly, you should document the additions in the Emacs manual.


28 Buffers

A buffer is a Lisp object containing text to be edited. Buffers are used to hold the contents of files that are being visited; there may also be buffers that are not visiting files. While several buffers may exist at one time, only one buffer is designated the current buffer at any time. Most editing commands act on the contents of the current buffer. Each buffer, including the current buffer, may or may not be displayed in any windows.


28.1 Buffer Basics

A buffer is a Lisp object containing text to be edited. Buffers are used to hold the contents of files that are being visited; there may also be buffers that are not visiting files. Although several buffers normally exist, only one buffer is designated the current buffer at any time. Most editing commands act on the contents of the current buffer. Each buffer, including the current buffer, may or may not be displayed in any windows.

Buffers in Emacs editing are objects that have distinct names and hold text that can be edited. Buffers appear to Lisp programs as a special data type. You can think of the contents of a buffer as a string that you can extend; insertions and deletions may occur in any part of the buffer. See Text.

A Lisp buffer object contains numerous pieces of information. Some of this information is directly accessible to the programmer through variables, while other information is accessible only through special-purpose functions. For example, the visited file name is directly accessible through a variable, while the value of point is accessible only through a primitive function.

Buffer-specific information that is directly accessible is stored in buffer-local variable bindings, which are variable values that are effective only in a particular buffer. This feature allows each buffer to override the values of certain variables. Most major modes override variables such as fill-column or comment-column in this way. For more information about buffer-local variables and functions related to them, see Buffer-Local Variables.

For functions and variables related to visiting files in buffers, see Visiting Files and Saving Buffers. For functions and variables related to the display of buffers in windows, see Buffers and Windows.

Function: bufferp object

This function returns t if object is a buffer, nil otherwise.


28.2 The Current Buffer

There are, in general, many buffers in an Emacs session. At any time, one of them is designated the current buffer—the buffer in which most editing takes place. Most of the primitives for examining or changing text operate implicitly on the current buffer (see Text).

Normally, the buffer displayed in the selected window (see Selecting Windows) is the current buffer, but this is not always so: a Lisp program can temporarily designate any buffer as current in order to operate on its contents, without changing what is displayed on the screen. The most basic function for designating a current buffer is set-buffer.

Function: current-buffer

This function returns the current buffer.

(current-buffer)
     ⇒ #<buffer buffers.texi>
Function: set-buffer buffer-or-name

This function makes buffer-or-name the current buffer. buffer-or-name must be an existing buffer or the name of an existing buffer. The return value is the buffer made current.

This function does not display the buffer in any window, so the user cannot necessarily see the buffer. But Lisp programs will now operate on it.

When an editing command returns to the editor command loop, Emacs automatically calls set-buffer on the buffer shown in the selected window (see Selecting Windows). This is to prevent confusion: it ensures that the buffer that the cursor is in, when Emacs reads a command, is the buffer to which that command applies (see Command Loop). Thus, you should not use set-buffer to switch visibly to a different buffer; for that, use the functions described in Switching to a Buffer in a Window.

When writing a Lisp function, do not rely on this behavior of the command loop to restore the current buffer after an operation. Editing commands can also be called as Lisp functions by other programs, not just from the command loop; it is convenient for the caller if the subroutine does not change which buffer is current (unless, of course, that is the subroutine’s purpose).

To operate temporarily on another buffer, put the set-buffer within a save-current-buffer form. Here, as an example, is a simplified version of the command append-to-buffer:

(defun append-to-buffer (buffer start end)
  "Append the text of the region to BUFFER."
  (interactive "BAppend to buffer: \nr")
  (let ((oldbuf (current-buffer)))
    (save-current-buffer
      (set-buffer (get-buffer-create buffer))
      (insert-buffer-substring oldbuf start end))))

Here, we bind a local variable to record the current buffer, and then save-current-buffer arranges to make it current again later. Next, set-buffer makes the specified buffer current, and insert-buffer-substring copies the string from the original buffer to the specified (and now current) buffer.

Alternatively, we can use the with-current-buffer macro:

(defun append-to-buffer (buffer start end)
  "Append the text of the region to BUFFER."
  (interactive "BAppend to buffer: \nr")
  (let ((oldbuf (current-buffer)))
    (with-current-buffer (get-buffer-create buffer)
      (insert-buffer-substring oldbuf start end))))

In either case, if the buffer appended to happens to be displayed in some window, the next redisplay will show how its text has changed. If it is not displayed in any window, you will not see the change immediately on the screen. The command causes the buffer to become current temporarily, but does not cause it to be displayed.

If you make local bindings (with let or function arguments) for a variable that may also have buffer-local bindings, make sure that the same buffer is current at the beginning and at the end of the local binding’s scope. Otherwise you might bind it in one buffer and unbind it in another!

Do not rely on using set-buffer to change the current buffer back, because that won’t do the job if a quit happens while the wrong buffer is current. For instance, in the previous example, it would have been wrong to do this:

  (let ((oldbuf (current-buffer)))
    (set-buffer (get-buffer-create buffer))
    (insert-buffer-substring oldbuf start end)
    (set-buffer oldbuf))

Using save-current-buffer or with-current-buffer, as we did, correctly handles quitting, errors, and throw, as well as ordinary evaluation.

Special Form: save-current-buffer body…

The save-current-buffer special form saves the identity of the current buffer, evaluates the body forms, and finally restores that buffer as current. The return value is the value of the last form in body. The current buffer is restored even in case of an abnormal exit via throw or error (see Nonlocal Exits).

If the buffer that used to be current has been killed by the time of exit from save-current-buffer, then it is not made current again, of course. Instead, whichever buffer was current just before exit remains current.

Macro: with-current-buffer buffer-or-name body…

The with-current-buffer macro saves the identity of the current buffer, makes buffer-or-name current, evaluates the body forms, and finally restores the current buffer. buffer-or-name must specify an existing buffer or the name of an existing buffer.

The return value is the value of the last form in body. The current buffer is restored even in case of an abnormal exit via throw or error (see Nonlocal Exits).

Macro: with-temp-buffer body…

The with-temp-buffer macro evaluates the body forms with a temporary buffer as the current buffer. It saves the identity of the current buffer, creates a temporary buffer and makes it current, evaluates the body forms, and finally restores the previous current buffer while killing the temporary buffer.

By default, undo information (see Undo) is not recorded in the buffer created by this macro (but body can enable that, if needed). The temporary buffer also does not run the hooks kill-buffer-hook, kill-buffer-query-functions (see Killing Buffers), and buffer-list-update-hook (see The Buffer List).

The return value is the value of the last form in body. You can return the contents of the temporary buffer by using (buffer-string) as the last form.

The current buffer is restored even in case of an abnormal exit via throw or error (see Nonlocal Exits).

See also with-temp-file in Writing to Files.


28.3 Buffer Names

Each buffer has a unique name, which is a string. Many of the functions that work on buffers accept either a buffer or a buffer name as an argument. Any argument called buffer-or-name is of this sort, and an error is signaled if it is neither a string nor a buffer. Any argument called buffer must be an actual buffer object, not a name.

Buffers that are ephemeral and generally uninteresting to the user have names starting with a space, so that the list-buffers and buffer-menu commands don’t mention them (but if such a buffer visits a file, it is mentioned). A name starting with space also initially disables recording undo information; see Undo.

Function: buffer-name &optional buffer

This function returns the name of buffer as a string. buffer defaults to the current buffer.

If buffer-name returns nil, it means that buffer has been killed. See Killing Buffers.

(buffer-name)
     ⇒ "buffers.texi"

(setq foo (get-buffer "temp"))
     ⇒ #<buffer temp>
(kill-buffer foo)
     ⇒ nil
(buffer-name foo)
     ⇒ nil
foo
     ⇒ #<killed buffer>
Command: rename-buffer newname &optional unique

This function renames the current buffer to newname. An error is signaled if newname is not a string.

Ordinarily, rename-buffer signals an error if newname is already in use. However, if unique is non-nil, it modifies newname to make a name that is not in use. Interactively, you can make unique non-nil with a numeric prefix argument. (This is how the command rename-uniquely is implemented.)

This function returns the name actually given to the buffer.

Function: get-buffer buffer-or-name

This function returns the buffer specified by buffer-or-name. If buffer-or-name is a string and there is no buffer with that name, the value is nil. If buffer-or-name is a buffer, it is returned as given; that is not very useful, so the argument is usually a name. For example:

(setq b (get-buffer "lewis"))
     ⇒ #<buffer lewis>
(get-buffer b)
     ⇒ #<buffer lewis>
(get-buffer "Frazzle-nots")
     ⇒ nil

See also the function get-buffer-create in Creating Buffers.

Function: generate-new-buffer-name starting-name &optional ignore

This function returns a name that would be unique for a new buffer—but does not create the buffer. It starts with starting-name, and produces a name not currently in use for any buffer by appending a number inside of ‘<…>’. It starts at 2 and keeps incrementing the number until it is not the name of an existing buffer.

If the optional second argument ignore is non-nil, it should be a string, a potential buffer name. It means to consider that potential buffer acceptable, if it is tried, even if it is the name of an existing buffer (which would normally be rejected). Thus, if buffers named ‘foo’, ‘foo<2>’, ‘foo<3>’ and ‘foo<4>’ exist,

(generate-new-buffer-name "foo")
     ⇒ "foo<5>"
(generate-new-buffer-name "foo" "foo<3>")
     ⇒ "foo<3>"
(generate-new-buffer-name "foo" "foo<6>")
     ⇒ "foo<5>"

See the related function generate-new-buffer in Creating Buffers.


28.4 Buffer File Name

The buffer file name is the name of the file that is visited in that buffer. When a buffer is not visiting a file, its buffer file name is nil. Most of the time, the buffer name is the same as the nondirectory part of the buffer file name, but the buffer file name and the buffer name are distinct and can be set independently. See Visiting Files.

Function: buffer-file-name &optional buffer

This function returns the absolute file name of the file that buffer is visiting. If buffer is not visiting any file, buffer-file-name returns nil. If buffer is not supplied, it defaults to the current buffer.

(buffer-file-name (other-buffer))
     ⇒ "/usr/user/lewis/manual/files.texi"
Variable: buffer-file-name

This buffer-local variable contains the name of the file being visited in the current buffer, or nil if it is not visiting a file. It is a permanent local variable, unaffected by kill-all-local-variables.

buffer-file-name
     ⇒ "/usr/user/lewis/manual/buffers.texi"

It is risky to change this variable’s value without doing various other things. Normally it is better to use set-visited-file-name (see below); some of the things done there, such as changing the buffer name, are not strictly necessary, but others are essential to avoid confusing Emacs.

Variable: buffer-file-truename

This buffer-local variable holds the abbreviated truename of the file visited in the current buffer, or nil if no file is visited. It is a permanent local, unaffected by kill-all-local-variables. See Truenames, and abbreviate-file-name.

Variable: buffer-file-number

This buffer-local variable holds the inode number and device identifier of the file visited in the current buffer, or nil if no file or a nonexistent file is visited. It is a permanent local, unaffected by kill-all-local-variables.

The value is normally a list of the form (inodenum device). This tuple uniquely identifies the file among all files accessible on the system. See the function file-attributes, in File Attributes, for more information about them.

If buffer-file-name is the name of a symbolic link, then both inodenum and device refer to the recursive target of the link.

Function: get-file-buffer filename

This function returns the buffer visiting file filename. If there is no such buffer, it returns nil. The argument filename, which must be a string, is expanded (see Functions that Expand Filenames), then compared against the visited file names of all live buffers. Note that the buffer’s buffer-file-name must match the expansion of filename exactly. This function will not recognize other names for the same file.

(get-file-buffer "buffers.texi")
    ⇒ #<buffer buffers.texi>

In unusual circumstances, there can be more than one buffer visiting the same file name. In such cases, this function returns the first such buffer in the buffer list.

Function: find-buffer-visiting filename &optional predicate

This is like get-file-buffer, except that it can return any buffer visiting the file possibly under a different name. That is, the buffer’s buffer-file-name does not need to match the expansion of filename exactly, it only needs to refer to the same file. If predicate is non-nil, it should be a function of one argument, a buffer visiting filename. The buffer is only considered a suitable return value if predicate returns non-nil. If it can not find a suitable buffer to return, find-buffer-visiting returns nil.

Command: set-visited-file-name filename &optional no-query along-with-file

If filename is a non-empty string, this function changes the name of the file visited in the current buffer to filename. (If the buffer had no visited file, this gives it one.) The next time the buffer is saved it will go in the newly-specified file.

This command marks the buffer as modified, since it does not (as far as Emacs knows) match the contents of filename, even if it matched the former visited file. It also renames the buffer to correspond to the new file name, unless the new name is already in use.

If filename is nil or the empty string, that stands for “no visited file”. In this case, set-visited-file-name marks the buffer as having no visited file, without changing the buffer’s modified flag.

Normally, this function asks the user for confirmation if there already is a buffer visiting filename. If no-query is non-nil, that prevents asking this question. If there already is a buffer visiting filename, and the user confirms or no-query is non-nil, this function makes the new buffer name unique by appending a number inside of ‘<…>’ to filename.

If along-with-file is non-nil, that means to assume that the former visited file has been renamed to filename. In this case, the command does not change the buffer’s modified flag, nor the buffer’s recorded last file modification time as reported by visited-file-modtime (see Buffer Modification Time). If along-with-file is nil, this function clears the recorded last file modification time, after which visited-file-modtime returns zero.

When the function set-visited-file-name is called interactively, it prompts for filename in the minibuffer.

Variable: list-buffers-directory

This buffer-local variable specifies a string to display in a buffer listing where the visited file name would go, for buffers that don’t have a visited file name. Dired buffers use this variable.


28.5 Buffer Modification

Emacs keeps a flag called the modified flag for each buffer, to record whether you have changed the text of the buffer. This flag is set to t whenever you alter the contents of the buffer, and cleared to nil when you save it. Thus, the flag shows whether there are unsaved changes. The flag value is normally shown in the mode line (see Variables Used in the Mode Line), and controls saving (see Saving Buffers) and auto-saving (see Auto-Saving).

Some Lisp programs set the flag explicitly. For example, the function set-visited-file-name sets the flag to t, because the text does not match the newly-visited file, even if it is unchanged from the file formerly visited.

The functions that modify the contents of buffers are described in Text.

Function: buffer-modified-p &optional buffer

This function returns non-nil if buffer has been modified since it was last read in from a file or saved, or nil otherwise. If buffer has been auto-saved since the time it was last modified, this function returns the symbol autosaved. If buffer is nil or omitted, it defaults to the current buffer.

Function: set-buffer-modified-p flag

This function marks the current buffer as modified if flag is non-nil, or as unmodified if the flag is nil.

Another effect of calling this function is to cause unconditional redisplay of the mode line for the current buffer. In fact, the function force-mode-line-update works by doing this:

(set-buffer-modified-p (buffer-modified-p))
Function: restore-buffer-modified-p flag

Like set-buffer-modified-p, but does not force redisplay of mode lines. This function also allows flag’s value to be the symbol autosaved, which marks the buffer as modified and auto-saved after the last modification.

Command: not-modified &optional arg

This command marks the current buffer as unmodified, and not needing to be saved. If arg is non-nil, it marks the buffer as modified, so that it will be saved at the next suitable occasion. Interactively, arg is the prefix argument.

Don’t use this function in programs, since it prints a message in the echo area; use set-buffer-modified-p (above) instead.

Function: buffer-modified-tick &optional buffer

This function returns buffer’s modification-count. This is a counter that increments every time the buffer is modified. If buffer is nil (or omitted), the current buffer is used.

Function: buffer-chars-modified-tick &optional buffer

This function returns buffer’s character-change modification-count. Changes to text properties leave this counter unchanged; however, each time text is inserted or removed from the buffer, the counter is reset to the value that would be returned by buffer-modified-tick. By comparing the values returned by two buffer-chars-modified-tick calls, you can tell whether a character change occurred in that buffer in between the calls. If buffer is nil (or omitted), the current buffer is used.

Sometimes there’s a need for modifying buffer in a way that doesn’t really change its text, like if only its text properties are changed. If your program needs to modify a buffer without triggering any hooks and features that react to buffer modifications, use the with-silent-modifications macro.

Macro: with-silent-modifications body…

Execute body pretending it does not modify the buffer. This includes checking whether the buffer’s file is locked (see File Locks), running buffer modification hooks (see Change Hooks), etc. Note that if body actually modifies the buffer text (as opposed to its text properties), its undo data may become corrupted.


28.6 Buffer Modification Time

Suppose that you visit a file and make changes in its buffer, and meanwhile the file itself is changed on disk. At this point, saving the buffer would overwrite the changes in the file. Occasionally this may be what you want, but usually it would lose valuable information. Emacs therefore checks the file’s modification time using the functions described below before saving the file. (See File Attributes, for how to examine a file’s modification time.)

Function: verify-visited-file-modtime &optional buffer

This function compares what buffer (by default, the current-buffer) has recorded for the modification time of its visited file against the actual modification time of the file as recorded by the operating system. The two should be the same unless some other process has written the file since Emacs visited or saved it.

The function returns t if the last actual modification time and Emacs’s recorded modification time are the same, nil otherwise. It also returns t if the buffer has no recorded last modification time, that is if visited-file-modtime would return zero.

It always returns t for buffers that are not visiting a file, even if visited-file-modtime returns a non-zero value. For instance, it always returns t for dired buffers. It returns t for buffers that are visiting a file that does not exist and never existed, but nil for file-visiting buffers whose file has been deleted.

Function: clear-visited-file-modtime

This function clears out the record of the last modification time of the file being visited by the current buffer. As a result, the next attempt to save this buffer will not complain of a discrepancy in file modification times.

This function is called in set-visited-file-name and other exceptional places where the usual test to avoid overwriting a changed file should not be done.

Function: visited-file-modtime

This function returns the current buffer’s recorded last file modification time, as a Lisp timestamp (see Time of Day).

If the buffer has no recorded last modification time, this function returns zero. This case occurs, for instance, if the buffer is not visiting a file or if the time has been explicitly cleared by clear-visited-file-modtime. Note, however, that visited-file-modtime returns a timestamp for some non-file buffers too. For instance, in a Dired buffer listing a directory, it returns the last modification time of that directory, as recorded by Dired.

If the buffer is visiting a file that doesn’t exist, this function returns −1.

Function: set-visited-file-modtime &optional time

This function updates the buffer’s record of the last modification time of the visited file, to the value specified by time if time is not nil, and otherwise to the last modification time of the visited file.

If time is neither nil nor an integer flag returned by visited-file-modtime, it should be a Lisp time value (see Time of Day).

This function is useful if the buffer was not read from the file normally, or if the file itself has been changed for some known benign reason.

Function: ask-user-about-supersession-threat filename

This function is used to ask a user how to proceed after an attempt to modify a buffer visiting file filename when the file is newer than the buffer text. Emacs detects this because the modification time of the file on disk is newer than the last save-time and its contents have changed. This means some other program has probably altered the file.

Depending on the user’s answer, the function may return normally, in which case the modification of the buffer proceeds, or it may signal a file-supersession error with data (filename), in which case the proposed buffer modification is not allowed.

This function is called automatically by Emacs on the proper occasions. It exists so you can customize Emacs by redefining it. See the file userlock.el for the standard definition.

See also the file locking mechanism in File Locks.


28.7 Read-Only Buffers

If a buffer is read-only, then you cannot change its contents, although you may change your view of the contents by scrolling and narrowing.

Read-only buffers are used in two kinds of situations:

  • A buffer visiting a write-protected file is normally read-only.

    Here, the purpose is to inform the user that editing the buffer with the aim of saving it in the file may be futile or undesirable. The user who wants to change the buffer text despite this can do so after clearing the read-only flag with C-x C-q.

  • Modes such as Dired and Rmail make buffers read-only when altering the contents with the usual editing commands would probably be a mistake.

    The special commands of these modes bind buffer-read-only to nil (with let) or bind inhibit-read-only to t around the places where they themselves change the text.

Variable: buffer-read-only

This buffer-local variable specifies whether the buffer is read-only. The buffer is read-only if this variable is non-nil. However, characters that have the inhibit-read-only text property can still be modified. See inhibit-read-only.

Variable: inhibit-read-only

If this variable is non-nil, then read-only buffers and, depending on the actual value, some or all read-only characters may be modified. Read-only characters in a buffer are those that have a non-nil read-only text property. See Properties with Special Meanings, for more information about text properties.

If inhibit-read-only is t, all read-only character properties have no effect. If inhibit-read-only is a list, then read-only character properties have no effect if they are members of the list (comparison is done with eq).

Command: read-only-mode &optional arg

This is the mode command for Read Only minor mode, a buffer-local minor mode. When the mode is enabled, buffer-read-only is non-nil in the buffer; when disabled, buffer-read-only is nil in the buffer. The calling convention is the same as for other minor mode commands (see Conventions for Writing Minor Modes).

This minor mode mainly serves as a wrapper for buffer-read-only; unlike most minor modes, there is no separate read-only-mode variable. Even when Read Only mode is disabled, characters with non-nil read-only text properties remain read-only. To temporarily ignore all read-only states, bind inhibit-read-only, as described above.

When enabling Read Only mode, this mode command also enables View mode if the option view-read-only is non-nil. See Miscellaneous Buffer Operations in The GNU Emacs Manual. When disabling Read Only mode, it disables View mode if View mode was enabled.

Function: barf-if-buffer-read-only &optional position

This function signals a buffer-read-only error if the current buffer is read-only. If the text at position (which defaults to point) has the inhibit-read-only text property set, the error will not be raised.

See Using interactive, for another way to signal an error if the current buffer is read-only.


28.8 The Buffer List

The buffer list is a list of all live buffers. The order of the buffers in this list is based primarily on how recently each buffer has been displayed in a window. Several functions, notably other-buffer, use this ordering. A buffer list displayed for the user also follows this order.

Creating a buffer adds it to the end of the buffer list, and killing a buffer removes it from that list. A buffer moves to the front of this list whenever it is chosen for display in a window (see Switching to a Buffer in a Window) or a window displaying it is selected (see Selecting Windows). A buffer moves to the end of the list when it is buried (see bury-buffer, below). There are no functions available to the Lisp programmer which directly manipulate the buffer list.

In addition to the fundamental buffer list just described, Emacs maintains a local buffer list for each frame, in which the buffers that have been displayed (or had their windows selected) in that frame come first. (This order is recorded in the frame’s buffer-list frame parameter; see Buffer Parameters.) Buffers never displayed in that frame come afterward, ordered according to the fundamental buffer list.

Function: buffer-list &optional frame

This function returns the buffer list, including all buffers, even those whose names begin with a space. The elements are actual buffers, not their names.

If frame is a frame, this returns frame’s local buffer list. If frame is nil or omitted, the fundamental buffer list is used: the buffers appear in order of most recent display or selection, regardless of which frames they were displayed on.

(buffer-list)
     ⇒ (#<buffer buffers.texi>
         #<buffer  *Minibuf-1*> #<buffer buffer.c>
         #<buffer *Help*> #<buffer TAGS>)

;; Note that the name of the minibuffer
;;   begins with a space!
(mapcar #'buffer-name (buffer-list))
    ⇒ ("buffers.texi" " *Minibuf-1*"
        "buffer.c" "*Help*" "TAGS")

The list returned by buffer-list is constructed specifically; it is not an internal Emacs data structure, and modifying it has no effect on the order of buffers. If you want to change the order of buffers in the fundamental buffer list, here is an easy way:

(defun reorder-buffer-list (new-list)
  (while new-list
    (bury-buffer (car new-list))
    (setq new-list (cdr new-list))))

With this method, you can specify any order for the list, but there is no danger of losing a buffer or adding something that is not a valid live buffer.

To change the order or value of a specific frame’s buffer list, set that frame’s buffer-list parameter with modify-frame-parameters (see Access to Frame Parameters).

Function: other-buffer &optional buffer visible-ok frame

This function returns the first buffer in the buffer list other than buffer. Usually, this is the buffer appearing in the most recently selected window (in frame frame or else the selected frame, see Input Focus), aside from buffer. Buffers whose names start with a space are not considered at all.

If buffer is not supplied (or if it is not a live buffer), then other-buffer returns the first buffer in the selected frame’s local buffer list. (If frame is non-nil, it returns the first buffer in frame’s local buffer list instead.)

If frame has a non-nil buffer-predicate parameter, then other-buffer uses that predicate to decide which buffers to consider. It calls the predicate once for each buffer, and if the value is nil, that buffer is ignored. See Buffer Parameters.

If visible-ok is nil, other-buffer avoids returning a buffer visible in any window on any visible frame, except as a last resort. If visible-ok is non-nil, then it does not matter whether a buffer is displayed somewhere or not.

If no suitable buffer exists, the buffer *scratch* is returned (and created, if necessary).

Function: last-buffer &optional buffer visible-ok frame

This function returns the last buffer in frame’s buffer list other than buffer. If frame is omitted or nil, it uses the selected frame’s buffer list.

The argument visible-ok is handled as with other-buffer, see above. If no suitable buffer can be found, the buffer *scratch* is returned.

Command: bury-buffer &optional buffer-or-name

This command puts buffer-or-name at the end of the buffer list, without changing the order of any of the other buffers on the list. This buffer therefore becomes the least desirable candidate for other-buffer to return. The argument can be either a buffer itself or the name of one.

This function operates on each frame’s buffer-list parameter as well as the fundamental buffer list; therefore, the buffer that you bury will come last in the value of (buffer-list frame) and in the value of (buffer-list). In addition, it also puts the buffer at the end of the list of buffers of the selected window (see Window History) provided it is shown in that window.

If buffer-or-name is nil or omitted, this means to bury the current buffer. In addition, if the current buffer is displayed in the selected window (see Selecting Windows), this makes sure that the window is either deleted or another buffer is shown in it. More precisely, if the selected window is dedicated (see Dedicated Windows) and there are other windows on its frame, the window is deleted. If it is the only window on its frame and that frame is not the only frame on its terminal, the frame is dismissed by calling the function specified by frame-auto-hide-function (see Quitting Windows). Otherwise, it calls switch-to-prev-buffer (see Window History) to show another buffer in that window. If buffer-or-name is displayed in some other window, it remains displayed there.

To replace a buffer in all the windows that display it, use replace-buffer-in-windows, See Buffers and Windows.

Command: unbury-buffer

This command switches to the last buffer in the local buffer list of the selected frame. More precisely, it calls the function switch-to-buffer (see Switching to a Buffer in a Window), to display the buffer returned by last-buffer (see above), in the selected window.

Variable: buffer-list-update-hook

This is a normal hook run whenever the buffer list changes. Functions (implicitly) running this hook are get-buffer-create (see Creating Buffers), rename-buffer (see Buffer Names), kill-buffer (see Killing Buffers), bury-buffer (see above), and select-window (see Selecting Windows). This hook is not run for internal or temporary buffers created by get-buffer-create or generate-new-buffer with a non-nil argument inhibit-buffer-hooks.

Functions run by this hook should avoid calling select-window with a nil norecord argument since this may lead to infinite recursion.

Function: buffer-match-p condition buffer-or-name &optional arg

This function checks if a buffer designated by buffer-or-name satisfies the specified condition. Optional third argument arg is passed to the predicate function in condition. A valid condition can be one of the following:

  • A string, interpreted as a regular expression. The buffer satisfies the condition if the regular expression matches the buffer name.
  • A predicate function, which should return non-nil if the buffer matches. If the function expects one argument, it is called with buffer-or-name as the argument; if it expects 2 arguments, the first argument is buffer-or-name and the second is arg (or nil if arg is omitted).
  • A cons-cell (oper . expr) where oper is one of
    (not cond)

    Satisfied if cond doesn’t satisfy buffer-match-p with the same buffer and arg.

    (or conds…)

    Satisfied if any condition in conds satisfies buffer-match-p, with the same buffer and arg.

    (and conds…)

    Satisfied if all the conditions in conds satisfy buffer-match-p, with the same buffer and arg.

    derived-mode

    Satisfied if the buffer’s major mode derives from expr. Note that this condition might fail to report a match if buffer-match-p is invoked before the major mode of the buffer has been established.

    major-mode

    Satisfied if the buffer’s major mode is equal to expr. Prefer using derived-mode instead, when both can work. Note that this condition might fail to report a match if buffer-match-p is invoked before the major mode of the buffer has been established.

  • t Satisfied by any buffer. A convenient alternative to "" (empty string) or (and) (empty conjunction).
Function: match-buffers condition &optional buffer-list arg

This function returns a list of all buffers that satisfy the condition. If no buffers match, the function returns nil. The argument condition is as defined in buffer-match-p above. By default, all the buffers are considered, but this can be restricted via the optional argument buffer-list, which should be a list of buffers to consider. Optional third argument arg will be passed to condition in the same way as buffer-match-p does.


28.9 Creating Buffers

This section describes the two primitives for creating buffers. get-buffer-create creates a buffer if it finds no existing buffer with the specified name; generate-new-buffer always creates a new buffer and gives it a unique name.

Both functions accept an optional argument inhibit-buffer-hooks. If it is non-nil, the buffer they create does not run the hooks kill-buffer-hook, kill-buffer-query-functions (see Killing Buffers), and buffer-list-update-hook (see The Buffer List). This avoids slowing down internal or temporary buffers that are never presented to users or passed on to other applications.

Other functions you can use to create buffers include with-output-to-temp-buffer (see Temporary Displays) and create-file-buffer (see Visiting Files). Starting a subprocess can also create a buffer (see Processes).

Function: get-buffer-create buffer-or-name &optional inhibit-buffer-hooks

This function returns a buffer named buffer-or-name. The buffer returned does not become the current buffer—this function does not change which buffer is current.

buffer-or-name must be either a string or an existing buffer. If it is a string and a live buffer with that name already exists, get-buffer-create returns that buffer. If no such buffer exists, it creates a new buffer. If buffer-or-name is a buffer instead of a string, it is returned as given, even if it is dead.

(get-buffer-create "foo")
     ⇒ #<buffer foo>

The major mode for a newly created buffer is set to Fundamental mode. (The default value of the variable major-mode is handled at a higher level; see How Emacs Chooses a Major Mode.) If the name begins with a space, the buffer initially disables undo information recording (see Undo).

Function: generate-new-buffer name &optional inhibit-buffer-hooks

This function returns a newly created, empty buffer, but does not make it current. The name of the buffer is generated by passing name to the function generate-new-buffer-name (see Buffer Names). Thus, if there is no buffer named name, then that is the name of the new buffer; if that name is in use, a suffix of the form ‘<n>’, where n is an integer, is appended to name.

An error is signaled if name is not a string.

(generate-new-buffer "bar")
     ⇒ #<buffer bar>
(generate-new-buffer "bar")
     ⇒ #<buffer bar<2>>
(generate-new-buffer "bar")
     ⇒ #<buffer bar<3>>

The major mode for the new buffer is set to Fundamental mode. The default value of the variable major-mode is handled at a higher level. See How Emacs Chooses a Major Mode.


28.10 Killing Buffers

Killing a buffer makes its name unknown to Emacs and makes the memory space it occupied available for other use.

The buffer object for the buffer that has been killed remains in existence as long as anything refers to it, but it is specially marked so that you cannot make it current or display it. Killed buffers retain their identity, however; if you kill two distinct buffers, they remain distinct according to eq although both are dead.

If you kill a buffer that is current or displayed in a window, Emacs automatically selects or displays some other buffer instead. This means that killing a buffer can change the current buffer. Therefore, when you kill a buffer, you should also take the precautions associated with changing the current buffer (unless you happen to know that the buffer being killed isn’t current). See The Current Buffer.

If you kill a buffer that is the base buffer of one or more indirect buffers (see Indirect Buffers), the indirect buffers are automatically killed as well.

The buffer-name of a buffer is nil if, and only if, the buffer is killed. A buffer that has not been killed is called a live buffer. To test whether a buffer is live or killed, use the function buffer-live-p (see below).

Command: kill-buffer &optional buffer-or-name

This function kills the buffer buffer-or-name, freeing all its memory for other uses or to be returned to the operating system. If buffer-or-name is nil or omitted, it kills the current buffer.

Any processes that have this buffer as the process-buffer are sent the SIGHUP (hangup) signal, which normally causes them to terminate. See Sending Signals to Processes.

If the buffer is visiting a file and contains unsaved changes, kill-buffer asks the user to confirm before the buffer is killed. It does this even if not called interactively. To prevent the request for confirmation, clear the modified flag before calling kill-buffer. See Buffer Modification.

This function calls replace-buffer-in-windows for cleaning up all windows currently displaying the buffer to be killed.

Killing a buffer that is already dead has no effect.

This function returns t if it actually killed the buffer. It returns nil if the user refuses to confirm or if buffer-or-name was already dead.

(kill-buffer "foo.unchanged")
     ⇒ t
(kill-buffer "foo.changed")

---------- Buffer: Minibuffer ----------
Buffer foo.changed modified; kill anyway? (yes or no) yes
---------- Buffer: Minibuffer ----------

     ⇒ t
Variable: kill-buffer-query-functions

Before confirming unsaved changes, kill-buffer calls the functions in the list kill-buffer-query-functions, in order of appearance, with no arguments. The buffer being killed is the current buffer when they are called. The idea of this feature is that these functions will ask for confirmation from the user. If any of them returns nil, kill-buffer spares the buffer’s life.

This hook is not run for internal or temporary buffers created by get-buffer-create or generate-new-buffer with a non-nil argument inhibit-buffer-hooks.

Variable: kill-buffer-hook

This is a normal hook run by kill-buffer after asking all the questions it is going to ask, just before actually killing the buffer. The buffer to be killed is current when the hook functions run. See Hooks. This variable is a permanent local, so its local binding is not cleared by changing major modes.

This hook is not run for internal or temporary buffers created by get-buffer-create or generate-new-buffer with a non-nil argument inhibit-buffer-hooks.

User Option: buffer-offer-save

This variable, if non-nil in a particular buffer, tells save-buffers-kill-emacs to offer to save that buffer, just as it offers to save file-visiting buffers. If save-some-buffers is called with the second optional argument set to t, it will also offer to save the buffer. Lastly, if this variable is set to the symbol always, both save-buffers-kill-emacs and save-some-buffers will always offer to save. See Definition of save-some-buffers. The variable buffer-offer-save automatically becomes buffer-local when set for any reason. See Buffer-Local Variables.

Variable: buffer-save-without-query

This variable, if non-nil in a particular buffer, tells save-buffers-kill-emacs and save-some-buffers to save this buffer (if it’s modified) without asking the user. The variable automatically becomes buffer-local when set for any reason.

Function: buffer-live-p object

This function returns t if object is a live buffer (a buffer which has not been killed), nil otherwise.


28.11 Indirect Buffers

An indirect buffer shares the text of some other buffer, which is called the base buffer of the indirect buffer. In some ways it is the analogue, for buffers, of a symbolic link among files. The base buffer may not itself be an indirect buffer.

The text of the indirect buffer is always identical to the text of its base buffer; changes made by editing either one are visible immediately in the other. This includes the text properties as well as the characters themselves.

In all other respects, the indirect buffer and its base buffer are completely separate. They have different names, independent values of point, independent narrowing, independent markers and overlays (though inserting or deleting text in either buffer relocates the markers and overlays for both), independent major modes, and independent buffer-local variable bindings.

An indirect buffer cannot visit a file, but its base buffer can. If you try to save the indirect buffer, that actually saves the base buffer.

Killing an indirect buffer has no effect on its base buffer. Killing the base buffer effectively kills the indirect buffer in that it cannot ever again be the current buffer.

Command: make-indirect-buffer base-buffer name &optional clone inhibit-buffer-hooks

This creates and returns an indirect buffer named name whose base buffer is base-buffer. The argument base-buffer may be a live buffer or the name (a string) of an existing buffer. If name is the name of an existing buffer, an error is signaled.

If clone is non-nil, then the indirect buffer originally shares the state of base-buffer such as major mode, minor modes, buffer local variables and so on. If clone is omitted or nil the indirect buffer’s state is set to the default state for new buffers.

If base-buffer is an indirect buffer, its base buffer is used as the base for the new buffer. If, in addition, clone is non-nil, the initial state is copied from the actual base buffer, not from base-buffer.

See Creating Buffers, for the meaning of inhibit-buffer-hooks.

Command: clone-indirect-buffer newname display-flag &optional norecord

This function creates and returns a new indirect buffer that shares the current buffer’s base buffer and copies the rest of the current buffer’s attributes. (If the current buffer is not indirect, it is used as the base buffer.)

If display-flag is non-nil, as it always is in interactive calls, that means to display the new buffer by calling pop-to-buffer. If norecord is non-nil, that means not to put the new buffer to the front of the buffer list.

Function: buffer-base-buffer &optional buffer

This function returns the base buffer of buffer, which defaults to the current buffer. If buffer is not indirect, the value is nil. Otherwise, the value is another buffer, which is never an indirect buffer.


28.12 Swapping Text Between Two Buffers

Specialized modes sometimes need to let the user access from the same buffer several vastly different types of text. For example, you may need to display a summary of the buffer text, in addition to letting the user access the text itself.

This could be implemented with multiple buffers (kept in sync when the user edits the text), or with narrowing (see Narrowing). But these alternatives might sometimes become tedious or prohibitively expensive, especially if each type of text requires expensive buffer-global operations in order to provide correct display and editing commands.

Emacs provides another facility for such modes: you can quickly swap buffer text between two buffers with buffer-swap-text. This function is very fast because it doesn’t move any text, it only changes the internal data structures of the buffer object to point to a different chunk of text. Using it, you can pretend that a group of two or more buffers are actually a single virtual buffer that holds the contents of all the individual buffers together.

Function: buffer-swap-text buffer

This function swaps the text of the current buffer and that of its argument buffer. It signals an error if one of the two buffers is an indirect buffer (see Indirect Buffers) or is a base buffer of an indirect buffer.

All the buffer properties that are related to the buffer text are swapped as well: the positions of point and mark, all the markers, the overlays, the text properties, the undo list, the value of the enable-multibyte-characters flag (see enable-multibyte-characters), etc.

Warning: If this function is called from within a save-excursion form, the current buffer will be set to buffer upon leaving the form, since the marker used by save-excursion to save the position and buffer will be swapped as well.

If you use buffer-swap-text on a file-visiting buffer, you should set up a hook to save the buffer’s original text rather than what it was swapped with. write-region-annotate-functions works for this purpose. You should probably set buffer-saved-size to −2 in the buffer, so that changes in the text it is swapped with will not interfere with auto-saving.


28.13 The Buffer Gap

Emacs buffers are implemented using an invisible gap to make insertion and deletion faster. Insertion works by filling in part of the gap, and deletion adds to the gap. Of course, this means that the gap must first be moved to the locus of the insertion or deletion. Emacs moves the gap only when you try to insert or delete. This is why your first editing command in one part of a large buffer, after previously editing in another far-away part, sometimes involves a noticeable delay.

This mechanism works invisibly, and Lisp code should never be affected by the gap’s current location, but these functions are available for getting information about the gap status.

Function: gap-position

This function returns the current gap position in the current buffer.

Function: gap-size

This function returns the current gap size of the current buffer.


29 Windows

This chapter describes the functions and variables related to Emacs windows. See Frames, for how windows are assigned an area of screen available for Emacs to use. See Emacs Display, for information on how text is displayed in windows.


29.1 Basic Concepts of Emacs Windows

A window is an area of the screen that can be used to display a buffer (see Buffers). Windows are grouped into frames (see Frames). Each frame contains at least one window; the user can subdivide a frame into multiple, non-overlapping windows to view several buffers at once. Lisp programs can use multiple windows for a variety of purposes. In Rmail, for example, you can view a summary of message titles in one window, and the contents of the selected message in another window.

Emacs uses the term “window” with a different meaning than in graphical desktop environments and window systems, such as the X Window System. When Emacs is run on X, each graphical X window owned by the Emacs process corresponds to one Emacs frame. When Emacs is run on a text terminal, each Emacs frame fills the entire terminal screen. In either case, the frame may contain one or more Emacs windows. For disambiguation, we use the term window-system window when we mean the window-system window corresponding to an Emacs frame.

Unlike X windows, Emacs windows are tiled; they never overlap within the area of their frame. When a window is created, resized, or deleted, the change in window space is taken from or given to other windows on the same frame, so that the total area of the frame is unchanged.

In Emacs Lisp, windows are represented by a special Lisp object type (see Window Type).

Function: windowp object

This function returns t if object is a window (whether or not it displays a buffer). Otherwise, it returns nil.

A live window is one that is actually displaying a buffer in a frame.

Function: window-live-p object

This function returns t if object is a live window and nil otherwise. A live window is one that displays a buffer.

The windows in each frame are organized into a window tree. See Windows and Frames. The leaf nodes of each window tree are live windows—the ones actually displaying buffers. The internal nodes of the window tree are internal windows, which are not live.

A valid window is one that is either live or internal. A valid window can be deleted, i.e., removed from its frame (see Deleting Windows); then it is no longer valid, but the Lisp object representing it might be still referenced from other Lisp objects. A deleted window may be made valid again by restoring a saved window configuration (see Window Configurations).

You can distinguish valid windows from deleted windows with window-valid-p.

Function: window-valid-p object

This function returns t if object is a live window, or an internal window in a window tree. Otherwise, it returns nil, including for the case where object is a deleted window.

The following schematic shows the structure of a live window:

        ____________________________________________
       |________________ Tab Line _______________|RD| ^
       |______________ Header Line ______________|  | |
     ^ |LS|LM|LF|                       |RF|RM|RS|  | |
     | |  |  |  |                       |  |  |  |  | |
Window |  |  |  |                       |  |  |  |  | Window
Body | |  |  |  |      Window Body      |  |  |  |  | Total
Height |  |  |  |                       |  |  |  |  | Height
     | |  |  |  |<- Window Body Width ->|  |  |  |  | |
     v |__|__|__|_______________________|__|__|__|  | |
       |_________ Horizontal Scroll Bar _________|  | |
       |_______________ Mode Line _______________|__| |
       |_____________ Bottom Divider _______________| v
        <---------- Window Total Width ------------>

At the center of that window is the body, where the buffer text is displayed. The body can be surrounded by a series of optional areas which we will call window decorations. On the left and right, from innermost to outermost, these are the left and right fringes, denoted by LF and RF (see Fringes); the left and right margins, denoted by LM and RM in the schematic (see Displaying in the Margins); the left or right vertical scroll bar, only one of which is present at any time, denoted by LS and RS (see Scroll Bars); and the right divider, denoted by RD (see Window Dividers). Together these are the window’s left and right decorations.

At the top of the window are the tab line and the header line (see Window Header Lines). The text area of the window includes the header line and the tab line, if they are present in the window. At the bottom of the window are the horizontal scroll bar (see Scroll Bars); the mode line (see Mode Line Format); and the bottom divider (see Window Dividers). Together these form the window’s top and bottom decorations.

There are two special areas omitted in the schematic:

  • When any of the fringes is missing, the display engine may use one character cell in its place for showing a continuation or truncation glyph provided a text line doesn’t fit in a window.
  • When both, the vertical scroll bar and the right divider are missing, the display engine usurps one pixel for drawing a vertical divider line between this window and the window on its right, provided such a window exists. On a text terminal, this divider always occupies an entire character cell.

In either case, the resulting artifact is considered part of the window’s body although its screen space cannot be used for displaying buffer text.

Note also, that line numbers (and their surrounding whitespace) as displayed by display-line-numbers-mode (see Display Custom in The GNU Emacs Manual) do not count as decorations either; they are part of the window’s body too.

Internal windows neither show any text nor do they have decorations. Hence, the concept of “body” does not make sense for them. In fact, most functions operating on the body of a window will yield an error when applied to an internal window.

By default, an Emacs frame exhibits one special live window that is used for displaying messages and accepting user input—the minibuffer window (see Minibuffer Windows). Since the minibuffer window is used for displaying text, it has a body but it does not have a tab or header line or any margins. Finally, a tooltip window which is used for displaying a tooltip in a tooltip frame (see Tooltips) has a body too but no decorations at all.


29.2 Windows and Frames

Each window belongs to exactly one frame (see Frames). For all windows belonging to a specific frame, we sometimes also say that these windows are owned by that frame or simply that they are on that frame.

Function: window-frame &optional window

This function returns the specified window’s frame—the frame that window belongs to. If window is omitted or nil, it defaults to the selected window (see Selecting Windows).

Function: window-list &optional frame minibuffer window

This function returns a list of all live windows owned by the specified frame. If frame is omitted or nil, it defaults to the selected frame (see Input Focus).

The optional argument minibuffer specifies whether to include the minibuffer window (see Minibuffer Windows) in that list. If minibuffer is t, the minibuffer window is included. If nil or omitted, the minibuffer window is included only if it is active. If minibuffer is neither nil nor t, the minibuffer window is never included.

The optional argument window, if non-nil, must be a live window on the specified frame; then window will be the first element in the returned list. If window is omitted or nil, the window selected within frame (see Selecting Windows) is the first element.

Windows on the same frame are organized into a window tree, whose leaf nodes are the live windows. The internal nodes of a window tree are not live; they exist for the purpose of organizing the relationships between live windows. The root node of a window tree is called the root window. It is either a live window or an internal window. If it is a live window, then the frame has just one window besides the minibuffer window, or the frame is a minibuffer-only frame, see Frame Layout.

A minibuffer window (see Minibuffer Windows) that is not alone on its frame does not have a parent window, so it strictly speaking is not part of its frame’s window tree. Nonetheless, it is a sibling window of the frame’s root window, and thus can be reached from the root window via window-next-sibling, see below. Also, the function window-tree described at the end of this section lists the minibuffer window alongside the actual window tree.

Function: frame-root-window &optional frame-or-window

This function returns the root window for frame-or-window. The argument frame-or-window should be either a window or a frame; if omitted or nil, it defaults to the selected frame. If frame-or-window is a window, the return value is the root window of that window’s frame.

When a live window is split (see Splitting Windows), there are two live windows where previously there was one. One of these is represented by the same Lisp window object as the original window, and the other is represented by a newly-created Lisp window object. Both of these live windows become leaf nodes of the window tree, as child windows of a single internal window. If necessary, Emacs automatically creates this internal window, which is also called the parent window, and assigns it to the appropriate position in the window tree. The set of windows that share the same parent are called siblings.

Function: window-parent &optional window

This function returns the parent window of window. If window is omitted or nil, it defaults to the selected window. The return value is nil if window has no parent (i.e., it is a minibuffer window or the root window of its frame).

A parent window always has at least two child windows. If this number were to fall to one as a result of window deletion (see Deleting Windows), Emacs automatically deletes the parent window too, and its sole remaining child window takes its place in the window tree.

A child window can be either a live window, or an internal window (which in turn would have its own child windows). Therefore, each internal window can be thought of as occupying a certain rectangular screen area—the union of the areas occupied by the live windows that are ultimately descended from it.

For each internal window, the screen areas of the immediate children are arranged either vertically or horizontally (never both). If the child windows are arranged one above the other, they are said to form a vertical combination; if they are arranged side by side, they are said to form a horizontal combination. Consider the following example:

     ______________________________________
    | ______  ____________________________ |
    ||      || __________________________ ||
    ||      |||                          |||
    ||      |||                          |||
    ||      |||                          |||
    ||      |||____________W4____________|||
    ||      || __________________________ ||
    ||      |||                          |||
    ||      |||                          |||
    ||      |||____________W5____________|||
    ||__W2__||_____________W3_____________ |
    |__________________W1__________________|

The root window of this frame is an internal window, W1. Its child windows form a horizontal combination, consisting of the live window W2 and the internal window W3. The child windows of W3 form a vertical combination, consisting of the live windows W4 and W5. Hence, the live windows in this window tree are W2, W4, and W5.

The following functions can be used to retrieve a child window of an internal window, and the siblings of a child window. Their window argument always defaults to the selected window (see Selecting Windows).

Function: window-top-child &optional window

This function returns the topmost child window of window, if window is an internal window whose children form a vertical combination. For any other type of window, the return value is nil.

Function: window-left-child &optional window

This function returns the leftmost child window of window, if window is an internal window whose children form a horizontal combination. For any other type of window, the return value is nil.

Function: window-child window

This function returns the first child window of the internal window window—the topmost child window for a vertical combination, or the leftmost child window for a horizontal combination. If window is a live window, the return value is nil.

Function: window-combined-p &optional window horizontal

This function returns a non-nil value if and only if window is part of a vertical combination.

If the optional argument horizontal is non-nil, this means to return non-nil if and only if window is part of a horizontal combination.

Function: window-next-sibling &optional window

This function returns the next sibling of the specified window. The return value is nil if window is the last child of its parent.

Function: window-prev-sibling &optional window

This function returns the previous sibling of the specified window. The return value is nil if window is the first child of its parent.

The functions window-next-sibling and window-prev-sibling should not be confused with the functions next-window and previous-window, which return the next and previous window in the cyclic ordering of windows (see Cyclic Ordering of Windows).

The following functions can be useful to locate a window within its frame.

Function: frame-first-window &optional frame-or-window

This function returns the live window at the upper left corner of the frame specified by frame-or-window. The argument frame-or-window must denote a window or a live frame and defaults to the selected frame. If frame-or-window specifies a window, this function returns the first window on that window’s frame. Under the assumption that the frame from our canonical example is selected (frame-first-window) returns W2.

Function: window-at-side-p &optional window side

This function returns t if window is located at side of its containing frame. The argument window must be a valid window and defaults to the selected one. The argument side can be any of the symbols left, top, right or bottom. The default value nil is handled like bottom.

Note that this function disregards the minibuffer window (see Minibuffer Windows). Hence, with side equal to bottom it may return t also when the minibuffer window appears right below window.

Function: window-in-direction direction &optional window ignore sign wrap minibuf

This function returns the nearest live window in direction direction as seen from the position of window-point in window window. The argument direction must be one of above, below, left or right. The optional argument window must denote a live window and defaults to the selected one.

This function does not return a window whose no-other-window parameter is non-nil (see Window Parameters). If the nearest window’s no-other-window parameter is non-nil, this function tries to find another window in the indicated direction whose no-other-window parameter is nil. If the optional argument ignore is non-nil, a window may be returned even if its no-other-window parameter is non-nil.

If the optional argument sign is a negative number, it means to use the right or bottom edge of window as reference position instead of window-point. If sign is a positive number, it means to use the left or top edge of window as reference position.

If the optional argument wrap is non-nil, this means to wrap direction around frame borders. For example, if window is at the top of the frame and direction is above, then this function usually returns the frame’s minibuffer window if it’s active and a window at the bottom of the frame otherwise.

If the optional argument minibuf is t, this function may return the minibuffer window even when it’s not active. If the optional argument minibuf is nil, this means to return the minibuffer window if and only if it is currently active. If minibuf is neither nil nor t, this function never returns the minibuffer window. However, if wrap is non-nil, it always acts as if minibuf were nil.

If it doesn’t find a suitable window, this function returns nil.

Don’t use this function to check whether there is no window in direction. Calling window-at-side-p described above is a much more efficient way to do that.

The following function retrieves the entire window tree of a frame:

Function: window-tree &optional frame

This function returns a list representing the window tree for frame frame. If frame is omitted or nil, it defaults to the selected frame.

The return value is a list of the form (root mini), where root represents the window tree of the frame’s root window, and mini is the frame’s minibuffer window.

If the root window is live, root is that window itself. Otherwise, root is a list (dir edges w1 w2 ...) where dir is nil for a horizontal combination and t for a vertical combination, edges gives the size and position of the combination, and the remaining elements are the child windows. Each child window may again be a window object (for a live window) or a list with the same format as above (for an internal window). The edges element is a list (left top right bottom), similar to the value returned by window-edges (see Coordinates and Windows).


29.3 Selecting Windows

In each frame, at any time, exactly one Emacs window is designated as selected within the frame. For the selected frame, that window is called the selected window—the one in which most editing takes place, and in which the cursor for selected windows appears (see Cursor Parameters). Keyboard input that inserts or deletes text is also normally directed to this window. The selected window’s buffer is usually also the current buffer, except when set-buffer has been used (see The Current Buffer). As for non-selected frames, the window selected within the frame becomes the selected window if the frame is ever selected.

Function: selected-window

This function returns the selected window (which is always a live window).

The following function explicitly selects a window and its frame.

Function: select-window window &optional norecord

This function makes window the selected window and the window selected within its frame, and selects that frame. It also makes window’s buffer (see Buffers and Windows) current and sets that buffer’s value of point to the value of window-point (see Windows and Point) in window. window must be a live window. The return value is window.

By default, this function also moves window’s buffer to the front of the buffer list (see The Buffer List) and makes window the most recently selected window. If the optional argument norecord is non-nil, these additional actions are omitted.

In addition, this function by default also tells the display engine to update the display of window when its frame gets redisplayed the next time. If norecord is non-nil, such updates are usually not performed. If, however, norecord equals the special symbol mark-for-redisplay, the additional actions mentioned above are omitted but window’s display will be nevertheless updated.

Note that sometimes selecting a window is not enough to show it, or make its frame the top-most frame on display: you may also need to raise the frame or make sure input focus is directed to that frame. See Input Focus.

For historical reasons, Emacs does not run a separate hook whenever a window gets selected. Applications and internal routines often temporarily select a window to perform a few actions on it. They do that either to simplify coding—because many functions by default operate on the selected window when no window argument is specified—or because some functions did not (and still do not) take a window as argument and always operate(d) on the selected window instead. Running a hook every time a window gets selected for a short time and once more when the previously selected window gets restored is not useful.

However, when its norecord argument is nil, select-window updates the buffer list and thus indirectly runs the normal hook buffer-list-update-hook (see The Buffer List). Consequently, that hook provides one way to run a function whenever a window gets selected more “permanently”.

Since buffer-list-update-hook is also run by functions that are not related to window management, it will usually make sense to save the value of the selected window somewhere and compare it with the value of selected-window while running that hook. Also, to avoid false positives when using buffer-list-update-hook, it is good practice that every select-window call supposed to select a window only temporarily passes a non-nil norecord argument. If possible, the macro with-selected-window (see below) should be used in such cases.

Emacs also runs the hook window-selection-change-functions whenever the redisplay routine detects that another window has been selected since last redisplay. See Hooks for Window Scrolling and Changes, for a detailed explanation. window-state-change-functions (described in the same section) is another abnormal hook run after a different window has been selected but is triggered by other window changes as well.

The sequence of calls to select-window with a non-nil norecord argument determines an ordering of windows by their selection or use time, see below. The function get-lru-window, for example, can then be used to retrieve the least recently selected window (see Cyclic Ordering of Windows).

Function: frame-selected-window &optional frame

This function returns the window on frame that is selected within that frame. frame should be a live frame; if omitted or nil, it defaults to the selected frame.

Function: set-frame-selected-window frame window &optional norecord

This function makes window the window selected within the frame frame. frame should be a live frame; if nil, it defaults to the selected frame. window must denote a live window.

If frame is the selected frame, this makes window the selected window.

If the optional argument norecord is non-nil, this function does not alter the ordering of the most recently selected windows, nor the buffer list.

The following macros are useful to temporarily select a window without affecting the ordering of recently selected windows or the buffer list.

Macro: save-selected-window forms…

This macro records the selected frame, as well as the selected window of each frame, executes forms in sequence, then restores the earlier selected frame and windows. It also saves and restores the current buffer. It returns the value of the last form in forms.

This macro does not save or restore anything about the sizes, arrangement or contents of windows; therefore, if forms change them, the change persists. If the previously selected window of some frame is no longer live at the time of exit from forms, that frame’s selected window is left alone. If the previously selected window is no longer live, then whatever window is selected at the end of forms remains selected. The current buffer is restored if and only if it is still live when exiting forms.

This macro changes neither the ordering of recently selected windows nor the buffer list.

Macro: with-selected-window window forms…

This macro selects window, executes forms in sequence, then restores the previously selected window and current buffer. The ordering of recently selected windows and the buffer list remain unchanged unless you deliberately change them within forms; for example, by calling select-window with argument norecord nil. Hence, this macro is the preferred way to temporarily work with window as the selected window without needlessly running buffer-list-update-hook.

Note that this macro temporarily puts the window management code in an unstable state. In particular, the most recently used window (see below) will not necessarily match the selected one. Hence, functions like get-lru-window and get-mru-window may return unexpected results when called from the body of this macro.

Macro: with-selected-frame frame forms…

This macro executes forms with frame as the selected frame. The value returned is the value of the last form in forms. This macro saves and restores the selected frame, and changes the order of neither the recently selected windows nor the buffers in the buffer list.

Function: window-use-time &optional window

This function returns the use time of window window. window must be a live window and defaults to the selected one.

The use time of a window is not really a time value, but an integer that does increase monotonically with each call of select-window with a nil norecord argument. The window with the lowest use time is usually called the least recently used window. The window with the highest use time is called the most recently used one (see Cyclic Ordering of Windows) and is usually the selected window unless with-selected-window has been used.

Function: window-bump-use-time &optional window

This function marks window as being the second most recently used one (after the selected window). It does nothing if window is the selected window or the selected window does not have the highest use time among all windows which may happen within the scope of with-selected-window.

Sometimes several windows collectively and cooperatively display a buffer, for example, under the management of Follow Mode (see (emacs)Follow Mode), where the windows together display a bigger portion of the buffer than one window could alone. It is often useful to consider such a window group as a single entity. Several functions such as window-group-start (see The Window Start and End Positions) allow you to do this by supplying, as an argument, one of the windows as a stand-in for the whole group.

Function: selected-window-group

When the selected window is a member of a group of windows, this function returns a list of the windows in the group, ordered such that the first window in the list is displaying the earliest part of the buffer, and so on. Otherwise the function returns a list containing just the selected window.

The selected window is considered part of a group when the buffer local variable selected-window-group-function is set to a function. In this case, selected-window-group calls it with no arguments and returns its result (which should be the list of windows in the group).


29.4 Window Sizes

Emacs provides miscellaneous functions for finding the height and width of a window. The return value of many of these functions can be specified either in units of pixels or in units of lines and columns. On a graphical display, the latter actually correspond to the height and width of a default character specified by the frame’s default font as returned by frame-char-height and frame-char-width (see Frame Font). Thus, if a window is displaying text with a different font or size, the reported line height and column width for that window may differ from the actual number of text lines or columns displayed within it.

The total height of a window is the number of lines comprising its body and its top and bottom decorations (see Basic Concepts of Emacs Windows).

Function: window-total-height &optional window round

This function returns the total height, in lines, of the window window. If window is omitted or nil, it defaults to the selected window. If window is an internal window, the return value is the total height occupied by its descendant windows.

If a window’s pixel height is not an integral multiple of its frame’s default character height, the number of lines occupied by the window is rounded internally. This is done in a way such that, if the window is a parent window, the sum of the total heights of all its child windows internally equals the total height of their parent. This means that although two windows have the same pixel height, their internal total heights may differ by one line. This means also, that if window is vertically combined and has a next sibling, the topmost row of that sibling can be calculated as the sum of this window’s topmost row and total height (see Coordinates and Windows)

If the optional argument round is ceiling, this function returns the smallest integer larger than window’s pixel height divided by the character height of its frame; if it is floor, it returns the largest integer smaller than said value; with any other round it returns the internal value of windows’s total height.

The total width of a window is the number of columns comprising its body and its left and right decorations (see Basic Concepts of Emacs Windows).

Function: window-total-width &optional window round

This function returns the total width, in columns, of the window window. If window is omitted or nil, it defaults to the selected window. If window is internal, the return value is the total width occupied by its descendant windows.

If a window’s pixel width is not an integral multiple of its frame’s character width, the number of columns occupied by the window is rounded internally. This is done in a way such that, if the window is a parent window, the sum of the total widths of all its children internally equals the total width of their parent. This means that although two windows have the same pixel width, their internal total widths may differ by one column. This means also, that if this window is horizontally combined and has a next sibling, the leftmost column of that sibling can be calculated as the sum of this window’s leftmost column and total width (see Coordinates and Windows). The optional argument round behaves as it does for window-total-height.

Function: window-total-size &optional window horizontal round

This function returns either the total height in lines or the total width in columns of the window window. If horizontal is omitted or nil, this is equivalent to calling window-total-height for window; otherwise it is equivalent to calling window-total-width for window. The optional argument round behaves as it does for window-total-height.

The following two functions can be used to return the total size of a window in units of pixels.

Function: window-pixel-height &optional window

This function returns the total height of window window in pixels. window must be a valid window and defaults to the selected one.

The return value includes the heights of window’s top and bottom decorations. If window is an internal window, its pixel height is the pixel height of the screen areas spanned by its children.

Function: window-pixel-width &optional window

This function returns the width of window window in pixels. window must be a valid window and defaults to the selected one.

The return value includes the widths of window’s left and right decorations. If window is an internal window, its pixel width is the width of the screen areas spanned by its children.

The following functions can be used to determine whether a given window has any adjacent windows.

Function: window-full-height-p &optional window

This function returns non-nil if window has no other window above or below it in its frame. More precisely, this means that the total height of window equals the total height of the root window on that frame. The minibuffer window does not count in this regard. If window is omitted or nil, it defaults to the selected window.

Function: window-full-width-p &optional window

This function returns non-nil if window has no other window to the left or right in its frame, i.e., its total width equals that of the root window on that frame. If window is omitted or nil, it defaults to the selected window.

The body height of a window is the height of its body, which does not include any of its top or bottom decorations (see Basic Concepts of Emacs Windows).

Function: window-body-height &optional window pixelwise

This function returns the height, in lines, of the body of window window. If window is omitted or nil, it defaults to the selected window; otherwise it must be a live window.

The optional argument pixelwise defines the units to use for the height. If nil, return the body height of window in characters, rounded down to the nearest integer, if necessary. This means that if a line at the bottom of the text area is only partially visible, that line is not counted. It also means that the height of a window’s body can never exceed its total height as returned by window-total-height.

If pixelwise is remap and the default face is remapped (see Face Remapping), use the remapped face to determine the character height. For any other non-nil value, return the height in pixels.

The body width of a window is the width of its body and of the text area, which does not include any of its left or right decorations (see Basic Concepts of Emacs Windows).

Note that when one or both fringes are removed (by setting their width to zero), the display engine reserves two character cells, one on each side of the window, for displaying the continuation and truncation glyphs, which leaves 2 columns less for text display. (The function window-max-chars-per-line, described below, takes this peculiarity into account.)

Function: window-body-width &optional window pixelwise

This function returns the width, in columns, of the body of window window. If window is omitted or nil, it defaults to the selected window; otherwise it must be a live window.

The optional argument pixelwise defines the units to use for the width. If nil, return the body width of window in characters, rounded down to the nearest integer, if necessary. This means that if a column on the right of the text area is only partially visible, that column is not counted. It also means that the width of a window’s body can never exceed its total width as returned by window-total-width.

If pixelwise is remap and the default face is remapped (see Face Remapping), use the remapped face to determine the character width. For any other non-nil value, return the width in pixels.

Function: window-body-size &optional window horizontal pixelwise

This function returns the body height or body width of window. If horizontal is omitted or nil, it is equivalent to calling window-body-height for window; otherwise it is equivalent to calling window-body-width. In either case, the optional argument pixelwise is passed to the function called.

The pixel heights of a window’s mode, tab and header line can be retrieved with the functions given below. Their return value is usually accurate unless the window has not been displayed before: In that case, the return value is based on an estimate of the font used for the window’s frame.

Function: window-mode-line-height &optional window

This function returns the height in pixels of window’s mode line. window must be a live window and defaults to the selected one. If window has no mode line, the return value is zero.

Function: window-tab-line-height &optional window

This function returns the height in pixels of window’s tab line. window must be a live window and defaults to the selected one. If window has no tab line, the return value is zero.

Function: window-header-line-height &optional window

This function returns the height in pixels of window’s header line. window must be a live window and defaults to the selected one. If window has no header line, the return value is zero.

Functions for retrieving the height and/or width of window dividers (see Window Dividers), fringes (see Fringes), scroll bars (see Scroll Bars), and display margins (see Displaying in the Margins) are described in the corresponding sections.

If your Lisp program needs to make layout decisions, you will find the following function useful:

Function: window-max-chars-per-line &optional window face

This function returns the number of characters displayed in the specified face face in the specified window window (which must be a live window). If face was remapped (see Face Remapping), the information is returned for the remapped face. If omitted or nil, face defaults to the default face, and window defaults to the selected window.

Unlike window-body-width, this function accounts for the actual size of face’s font, instead of working in units of the canonical character width of window’s frame (see Frame Font). It also accounts for space used by the continuation glyph, if window lacks one or both of its fringes.

Commands that change the size of windows (see Resizing Windows), or split them (see Splitting Windows), obey the variables window-min-height and window-min-width, which specify the smallest allowable window height and width. They also obey the variable window-size-fixed, with which a window can be fixed in size (see Preserving Window Sizes).

User Option: window-min-height

This option specifies the minimum total height, in lines, of any window. Its value has to accommodate at least one text line and any top or bottom decorations.

User Option: window-min-width

This option specifies the minimum total width, in columns, of any window. Its value has to accommodate at least two text columns and any left or right decorations.

The following function tells how small a specific window can get taking into account the sizes of its areas and the values of window-min-height, window-min-width and window-size-fixed (see Preserving Window Sizes).

Function: window-min-size &optional window horizontal ignore pixelwise

This function returns the minimum size of window. window must be a valid window and defaults to the selected one. The optional argument horizontal non-nil means to return the minimum number of columns of window; otherwise return the minimum number of window’s lines.

The return value makes sure that all components of window remain fully visible if window’s size were actually set to it. With horizontal nil it includes any top or bottom decorations. With horizontal non-nil it includes any left or right decorations of window.

The optional argument ignore, if non-nil, means ignore restrictions imposed by fixed size windows, window-min-height or window-min-width settings. If ignore equals safe, live windows may get as small as window-safe-min-height lines and window-safe-min-width columns. If ignore is a window, ignore restrictions for that window only. Any other non-nil value means ignore all of the above restrictions for all windows.

The optional argument pixelwise non-nil means to return the minimum size of window counted in pixels.


29.5 Resizing Windows

This section describes functions for resizing a window without changing the size of its frame. Because live windows do not overlap, these functions are meaningful only on frames that contain two or more windows: resizing a window also changes the size of at least one other window. If there is just one window on a frame, its size cannot be changed except by resizing the frame (see Frame Size).

Except where noted, these functions also accept internal windows as arguments. Resizing an internal window causes its child windows to be resized to fit the same space.

Function: window-resizable window delta &optional horizontal ignore pixelwise

This function returns delta if the size of window can be changed vertically by delta lines. If the optional argument horizontal is non-nil, it instead returns delta if window can be resized horizontally by delta columns. It does not actually change the window size.

If window is nil, it defaults to the selected window.

A positive value of delta means to check whether the window can be enlarged by that number of lines or columns; a negative value of delta means to check whether the window can be shrunk by that many lines or columns. If delta is non-zero, a return value of 0 means that the window cannot be resized.

Normally, the variables window-min-height and window-min-width specify the smallest allowable window size (see Window Sizes). However, if the optional argument ignore is non-nil, this function ignores window-min-height and window-min-width, as well as window-size-fixed. Instead, it considers the minimum height of a window as the sum of its top and bottom decorations plus the text of one line; and its minimum width as the sum of its left and right decorations plus text that takes two columns.

If the optional argument pixelwise is non-nil, delta is interpreted as pixels.

Function: window-resize window delta &optional horizontal ignore pixelwise

This function resizes window by delta increments. If horizontal is nil, it changes the height by delta lines; otherwise, it changes the width by delta columns. A positive delta means to enlarge the window, and a negative delta means to shrink it.

If window is nil, it defaults to the selected window. If the window cannot be resized as demanded, an error is signaled.

The optional argument ignore has the same meaning as for the function window-resizable above.

If the optional argument pixelwise is non-nil, delta will be interpreted as pixels.

The choice of which window edges this function alters depends on the values of the option window-combination-resize and the combination limits of the involved windows; in some cases, it may alter both edges. See Recombining Windows. To resize by moving only the bottom or right edge of a window, use the function adjust-window-trailing-edge.

Function: adjust-window-trailing-edge window delta &optional horizontal pixelwise

This function moves window’s bottom edge by delta lines. If optional argument horizontal is non-nil, it instead moves the right edge by delta columns. If window is nil, it defaults to the selected window.

If the optional argument pixelwise is non-nil, delta is interpreted as pixels.

A positive delta moves the edge downwards or to the right; a negative delta moves it upwards or to the left. If the edge cannot be moved as far as specified by delta, this function moves it as far as possible but does not signal an error.

This function tries to resize windows adjacent to the edge that is moved. If this is not possible for some reason (e.g., if that adjacent window is fixed-size), it may resize other windows.

User Option: window-resize-pixelwise

If the value of this option is non-nil, Emacs resizes windows in units of pixels. This currently affects functions like split-window (see Splitting Windows), maximize-window, minimize-window, fit-window-to-buffer, fit-frame-to-buffer and shrink-window-if-larger-than-buffer (all listed below).

Note that when a frame’s pixel size is not a multiple of its character size, at least one window may get resized pixelwise even if this option is nil. The default value is nil.

The following commands resize windows in more specific ways. When called interactively, they act on the selected window.

Command: fit-window-to-buffer &optional window max-height min-height max-width min-width preserve-size

This command adjusts the height or width of window to fit the text in it. It returns non-nil if it was able to resize window, and nil otherwise. If window is omitted or nil, it defaults to the selected window. Otherwise, it should be a live window.

If window is part of a vertical combination, this function adjusts window’s height. The new height is calculated from the actual height of the accessible portion of its buffer. The optional argument max-height, if non-nil, specifies the maximum total height that this function can give window. The optional argument min-height, if non-nil, specifies the minimum total height that it can give, which overrides the variable window-min-height. Both max-height and min-height are specified in lines and include any top or bottom decorations of window.

If window is part of a horizontal combination and the value of the option fit-window-to-buffer-horizontally (see below) is non-nil, this function adjusts window’s width. The new width of window is calculated from the maximum length of its buffer’s lines that follow the current start position of window. The optional argument max-width specifies a maximum width and defaults to the width of window’s frame. The optional argument min-width specifies a minimum width and defaults to window-min-width. Both max-width and min-width are specified in columns and include any left or right decorations of window.

The optional argument preserve-size, if non-nil, will install a parameter to preserve the size of window during future resize operations (see Preserving Window Sizes).

If the option fit-frame-to-buffer (see below) is non-nil, this function will try to resize the frame of window to fit its contents by calling fit-frame-to-buffer (see below).

User Option: fit-window-to-buffer-horizontally

If this is non-nil, fit-window-to-buffer can resize windows horizontally. If this is nil (the default) fit-window-to-buffer never resizes windows horizontally. If this is only, it can resize windows horizontally only. Any other value means fit-window-to-buffer can resize windows in both dimensions.

User Option: fit-frame-to-buffer

If this option is non-nil, fit-window-to-buffer can fit a frame to its buffer. A frame is fit if and only if its root window is a live window and this option is non-nil. If this is horizontally, frames are fit horizontally only. If this is vertically, frames are fit vertically only. Any other non-nil value means frames can be resized in both dimensions.

If you have a frame that displays only one window, you can fit that frame to its buffer using the command fit-frame-to-buffer.

Command: fit-frame-to-buffer &optional frame max-height min-height max-width min-width only

This command adjusts the size of frame to display the contents of its buffer exactly. frame can be any live frame and defaults to the selected one. Fitting is done only if frame’s root window is live.

The arguments max-height, min-height, max-width and min-width, if non-nil, specify bounds on the new body size of frame’s root window. A non-nil value specified by any of these arguments overrides the corresponding value specified by the option fit-frame-to-buffer-sizes described below.

If the optional argument only is vertically, this function may resize the frame vertically only. If only is horizontally, it may resize the frame horizontally only.

The behavior of fit-frame-to-buffer can be controlled with the help of the two options listed next.

User Option: fit-frame-to-buffer-margins

This option can be used to specify margins around frames to be fit by fit-frame-to-buffer. Such margins can be useful to avoid, for example, that the resized frame overlaps the taskbar or parts of its parent frame.

It specifies the numbers of pixels to be left free on the left, above, the right, and below a frame that shall be fit. The default specifies nil for each which means to use no margins. The value specified here can be overridden for a specific frame by that frame’s fit-frame-to-buffer-margins parameter, if present.

User Option: fit-frame-to-buffer-sizes

This option specifies size boundaries for fit-frame-to-buffer. It specifies the maximum and minimum lines and maximum and minimum columns of the root window’s body of any frame that shall be fit to its buffer. Any value this option specifies will be overridden by the corresponding argument of fit-frame-to-buffer, if non-nil.

Command: shrink-window-if-larger-than-buffer &optional window

This command attempts to reduce window’s height as much as possible while still showing its full buffer, but no less than window-min-height lines. The return value is non-nil if the window was resized, and nil otherwise. If window is omitted or nil, it defaults to the selected window. Otherwise, it should be a live window.

This command does nothing if the window is already too short to display all of its buffer, or if any of the buffer is scrolled off-screen, or if the window is the only live window in its frame.

This command calls fit-window-to-buffer (see above) to do its work.

Command: balance-windows &optional window-or-frame

This function balances windows in a way that gives more space to full-width and/or full-height windows. If window-or-frame specifies a frame, it balances all windows on that frame. If window-or-frame specifies a window, it balances only that window and its siblings (see Windows and Frames).

Command: balance-windows-area

This function attempts to give all windows on the selected frame approximately the same share of the screen area. Full-width or full-height windows are not given more space than other windows.

Command: maximize-window &optional window

This function attempts to make window as large as possible, in both dimensions, without resizing its frame or deleting other windows. If window is omitted or nil, it defaults to the selected window.

Command: minimize-window &optional window

This function attempts to make window as small as possible, in both dimensions, without deleting it or resizing its frame. If window is omitted or nil, it defaults to the selected window.


29.6 Preserving Window Sizes

A window can get resized explicitly by using one of the functions from the preceding section or implicitly, for example, when resizing an adjacent window, when splitting or deleting a window (see Splitting Windows, see Deleting Windows) or when resizing the window’s frame (see Frame Size).

It is possible to avoid implicit resizing of a specific window when there are one or more other resizable windows on the same frame. For this purpose, Emacs must be advised to preserve the size of that window. There are two basic ways to do that.

Variable: window-size-fixed

If this buffer-local variable is non-nil, the size of any window displaying the buffer cannot normally be changed. Deleting a window or changing the frame’s size may still change the window’s size, if there is no choice.

If the value is height, then only the window’s height is fixed; if the value is width, then only the window’s width is fixed. Any other non-nil value fixes both the width and the height.

If this variable is nil, this does not necessarily mean that any window showing the buffer can be resized in the desired direction. To determine that, use the function window-resizable. See Resizing Windows.

Often window-size-fixed is overly aggressive because it inhibits any attempt to explicitly resize or split an affected window as well. This may even happen after the window has been resized implicitly, for example, when deleting an adjacent window or resizing the window’s frame. The following function tries hard to never disallow resizing such a window explicitly:

Function: window-preserve-size &optional window horizontal preserve

This function (un-)marks the height of window window as preserved for future resize operations. window must be a live window and defaults to the selected one. If the optional argument horizontal is non-nil, it (un-)marks the width of window as preserved.

If the optional argument preserve is t, this means to preserve the current height/width of window’s body. The height/width of window will change only if Emacs has no better choice. Resizing a window whose height/width is preserved by this function never throws an error.

If preserve is nil, this means to stop preserving the height/width of window, lifting any respective restraint induced by a previous call of this function for window. Calling enlarge-window, shrink-window or fit-window-to-buffer with window as argument may also remove the respective restraint.

window-preserve-size is currently invoked by the following functions:

fit-window-to-buffer

If the optional argument preserve-size of that function (see Resizing Windows) is non-nil, the size established by that function is preserved.

display-buffer

If the alist argument of that function (see Choosing a Window for Displaying a Buffer) contains a preserve-size entry, the size of the window produced by that function is preserved.

window-preserve-size installs a window parameter (see Window Parameters) called window-preserved-size which is consulted by the window resizing functions. This parameter will not prevent resizing the window when the window shows another buffer than the one when window-preserve-size was invoked or if its size has changed since then.

The following function can be used to check whether the height of a particular window is preserved:

Function: window-preserved-size &optional window horizontal

This function returns the preserved height of window window in pixels. window must be a live window and defaults to the selected one. If the optional argument horizontal is non-nil, it returns the preserved width of window. It returns nil if the size of window is not preserved.


29.7 Splitting Windows

This section describes functions for creating a new window by splitting an existing one. Note that some windows are special in the sense that these functions may fail to split them as described here. Examples of such windows are side windows (see Side Windows) and atomic windows (see Atomic Windows).

Function: split-window &optional window size side pixelwise

This function creates a new live window next to the window window. If window is omitted or nil, it defaults to the selected window. That window is split, and reduced in size. The space is taken up by the new window, which is returned.

The optional second argument size determines the sizes of window and/or the new window. If it is omitted or nil, both windows are given equal sizes; if there is an odd line, it is allocated to the new window. If size is a positive number, window is given size lines (or columns, depending on the value of side). If size is a negative number, the new window is given −size lines (or columns).

If size is nil, this function obeys the variables window-min-height and window-min-width (see Window Sizes). Thus, it signals an error if splitting would result in making a window smaller than those variables specify. However, a non-nil value for size causes those variables to be ignored; in that case, the smallest allowable window is considered to be one that has space for a text that is one line tall and/or two columns wide.

Hence, if size is specified, it’s the caller’s responsibility to check whether the emanating windows are large enough to encompass all of their decorations like a mode line or a scroll bar. The function window-min-size (see Window Sizes) can be used to determine the minimum requirements of window in this regard. Since the new window usually inherits areas like the mode line or the scroll bar from window, that function is also a good guess for the minimum size of the new window. The caller should specify a smaller size only if it correspondingly removes an inherited area before the next redisplay.

The optional third argument side determines the position of the new window relative to window. If it is nil or below, the new window is placed below window. If it is above, the new window is placed above window. In both these cases, size specifies a total window height, in lines.

If side is t or right, the new window is placed on the right of window. If side is left, the new window is placed on the left of window. In both these cases, size specifies a total window width, in columns.

The optional fourth argument pixelwise, if non-nil, means to interpret size in units of pixels, instead of lines and columns.

If window is a live window, the new window inherits various properties from it, including margins and scroll bars. If window is an internal window, the new window inherits the properties of the window selected within window’s frame.

The behavior of this function may be altered by the window parameters of window, so long as the variable ignore-window-parameters is nil. If the value of the split-window window parameter is t, this function ignores all other window parameters. Otherwise, if the value of the split-window window parameter is a function, that function is called with the arguments window, size, and side, in lieu of the usual action of split-window. Otherwise, this function obeys the window-atom or window-side window parameter, if any. See Window Parameters.

As an example, here is a sequence of split-window calls that yields the window configuration discussed in Windows and Frames. This example demonstrates splitting a live window as well as splitting an internal window. We begin with a frame containing a single window (a live root window), which we denote by W4. Calling (split-window W4) yields this window configuration:

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W4_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W5_________________||
    |__________________W3__________________|

The split-window call has created a new live window, denoted by W5. It has also created a new internal window, denoted by W3, which becomes the root window and the parent of both W4 and W5.

Next, we call (split-window W3 nil 'left), passing the internal window W3 as the argument. The result:

     ______________________________________
    | ______  ____________________________ |
    ||      || __________________________ ||
    ||      |||                          |||
    ||      |||                          |||
    ||      |||                          |||
    ||      |||____________W4____________|||
    ||      || __________________________ ||
    ||      |||                          |||
    ||      |||                          |||
    ||      |||____________W5____________|||
    ||__W2__||_____________W3_____________ |
    |__________________W1__________________|

A new live window W2 is created, to the left of the internal window W3. A new internal window W1 is created, becoming the new root window.

For interactive use, Emacs provides two commands which always split the selected window. These call split-window internally.

Command: split-window-right &optional size window-to-split

This function splits the window window-to-split into two side-by-side windows, putting window-to-split on the left. window-to-split defaults to the selected window. If size is positive, the left window gets size columns; if size is negative, the right window gets −size columns.

Command: split-window-below &optional size window-to-split

This function splits the window window-to-split into two windows, one above the other, leaving the upper window selected. window-to-split defaults to the selected window. If size is positive, the upper window gets size lines; if size is negative, the lower window gets −size lines.

Command: split-root-window-below &optional size

This function splits the whole frame in two. The current window configuration is retained on the top, and a new window is created below, taking up the whole width of the frame. size is treated as by split-window-below.

Command: split-root-window-right &optional size

This function splits the whole frame in two. The current window configuration is retained on the left, and a new window is created on the right, taking up the whole height of the frame. size is treated as by split-window-right.

User Option: split-window-keep-point

If the value of this variable is non-nil (the default), split-window-below behaves as described above.

If it is nil, split-window-below adjusts point in each of the two windows to minimize redisplay. (This is useful on slow terminals.) It selects whichever window contains the screen line that point was previously on. Note that this only affects split-window-below, not the lower-level split-window function.


29.8 Deleting Windows

Deleting a window removes it from the frame’s window tree. If the window is a live window, it disappears from the screen. If the window is an internal window, its child windows are deleted too.

Even after a window is deleted, it continues to exist as a Lisp object, until there are no more references to it. Window deletion can be reversed, by restoring a saved window configuration (see Window Configurations).

Command: delete-window &optional window

This function removes window from display and returns nil. If window is omitted or nil, it defaults to the selected window.

If deleting the window would leave no more windows in the window tree (e.g., if it is the only live window in the frame) or all remaining windows on window’s frame are side windows (see Side Windows), an error is signaled. If window is part of an atomic window (see Atomic Windows), this function tries to delete the root of that atomic window instead.

By default, the space taken up by window is given to one of its adjacent sibling windows, if any. However, if the variable window-combination-resize is non-nil, the space is proportionally distributed among any remaining windows in the same window combination. See Recombining Windows.

The behavior of this function may be altered by the window parameters of window, so long as the variable ignore-window-parameters is nil. If the value of the delete-window window parameter is t, this function ignores all other window parameters. Otherwise, if the value of the delete-window window parameter is a function, that function is called with the argument window, in lieu of the usual action of delete-window. See Window Parameters.

When delete-window deletes the selected window of its frame, it has to make another window the new selected window of that frame. The following option allows configuring which window is chosen.

User Option: delete-window-choose-selected

This option allows specifying which window should become a frame’s selected window after delete-window has deleted the previously selected one. Possible choices are

  • mru (the default) choose the most recently used window on that frame.
  • pos choose the window comprising the frame coordinates of point of the previously selected window on that frame.
  • nil choose the first window (the window returned by frame-first-window) on that frame.

A window with a non-nil no-other-window parameter is chosen only if all other windows on that frame have that parameter set to a non-nil value too.

Command: delete-other-windows &optional window

This function makes window fill its frame, deleting other windows as necessary. If window is omitted or nil, it defaults to the selected window. An error is signaled if window is a side window (see Side Windows). If window is part of an atomic window (see Atomic Windows), this function tries to make the root of that atomic window fill its frame. The return value is nil.

The behavior of this function may be altered by the window parameters of window, so long as the variable ignore-window-parameters is nil. If the value of the delete-other-windows window parameter is t, this function ignores all other window parameters. Otherwise, if the value of the delete-other-windows window parameter is a function, that function is called with the argument window, in lieu of the usual action of delete-other-windows. See Window Parameters.

Also, if ignore-window-parameters is nil, this function does not delete any window whose no-delete-other-windows parameter is non-nil.

Command: delete-windows-on &optional buffer-or-name frame

This function deletes all windows showing buffer-or-name, by calling delete-window on those windows. buffer-or-name should be a buffer, or the name of a buffer; if omitted or nil, it defaults to the current buffer. If there are no windows showing the specified buffer, this function does nothing. If the specified buffer is a minibuffer, an error is signaled.

If there is a dedicated window showing the buffer, and that window is the only one on its frame, this function also deletes that frame if it is not the only frame on the terminal.

The optional argument frame specifies which frames to operate on:

  • nil means operate on all frames.
  • t means operate on the selected frame.
  • visible means operate on all visible frames.
  • 0 means operate on all visible or iconified frames.
  • A frame means operate on that frame.

Note that this argument does not have the same meaning as in other functions which scan all live windows (see Cyclic Ordering of Windows). Specifically, the meanings of t and nil here are the opposite of what they are in those other functions.


29.9 Recombining Windows

When deleting the last sibling of a window W, its parent window is deleted too, with W replacing it in the window tree. This means that W must be recombined with its parent’s siblings to form a new window combination (see Windows and Frames). In some occasions, deleting a live window may even entail the deletion of two internal windows.

     ______________________________________
    | ______  ____________________________ |
    ||      || __________________________ ||
    ||      ||| ___________  ___________ |||
    ||      ||||           ||           ||||
    ||      ||||____W6_____||_____W7____||||
    ||      |||____________W4____________|||
    ||      || __________________________ ||
    ||      |||                          |||
    ||      |||                          |||
    ||      |||____________W5____________|||
    ||__W2__||_____________W3_____________ |
    |__________________W1__________________|

Deleting W5 in this configuration normally causes the deletion of W3 and W4. The remaining live windows W2, W6 and W7 are recombined to form a new horizontal combination with parent W1.

Sometimes, however, it makes sense to not delete a parent window like W4. In particular, a parent window should not be removed when it was used to preserve a combination embedded in a combination of the same type. Such embeddings make sense to assure that when you split a window and subsequently delete the new window, Emacs reestablishes the layout of the associated frame as it existed before the splitting.

Consider a scenario starting with two live windows W2 and W3 and their parent W1.

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W2_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W3_________________||
    |__________________W1__________________|

Split W2 to make a new window W4 as follows.

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W2_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W4_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W3_________________||
    |__________________W1__________________|

Now, when enlarging a window vertically, Emacs tries to obtain the corresponding space from its lower sibling, provided such a window exists. In our scenario, enlarging W4 will steal space from W3.

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W2_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W4_________________||
    | ____________________________________ |
    ||_________________W3_________________||
    |__________________W1__________________|

Deleting W4 will now give its entire space to W2, including the space earlier stolen from W3.

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W2_________________||
    | ____________________________________ |
    ||_________________W3_________________||
    |__________________W1__________________|

This can be counterintuitive, in particular if W4 were used for displaying a buffer only temporarily (see Temporary Displays), and you want to continue working with the initial layout.

The behavior can be fixed by making a new parent window when splitting W2. The variable described next allows that to be done.

User Option: window-combination-limit

This variable controls whether splitting a window shall make a new parent window. The following values are recognized:

nil

This means that the new live window is allowed to share the existing parent window, if one exists, provided the split occurs in the same direction as the existing window combination (otherwise, a new internal window is created anyway).

window-size

This means that display-buffer makes a new parent window when it splits a window and is passed a window-height or window-width entry in the alist argument (see Action Functions for Buffer Display). Otherwise, window splitting behaves as for a value of nil.

temp-buffer-resize

In this case with-temp-buffer-window makes a new parent window when it splits a window and temp-buffer-resize-mode is enabled (see Temporary Displays). Otherwise, window splitting behaves as for nil.

temp-buffer

In this case with-temp-buffer-window always makes a new parent window when it splits an existing window (see Temporary Displays). Otherwise, window splitting behaves as for nil.

display-buffer

This means that when display-buffer (see Choosing a Window for Displaying a Buffer) splits a window it always makes a new parent window. Otherwise, window splitting behaves as for nil.

t

This means that splitting a window always creates a new parent window. Thus, if the value of this variable is at all times t, then at all times every window tree is a binary tree (a tree where each window except the root window has exactly one sibling).

The default is window-size. Other values are reserved for future use.

If, as a consequence of this variable’s setting, split-window makes a new parent window, it also calls set-window-combination-limit (see below) on the newly-created internal window. This affects how the window tree is rearranged when the child windows are deleted (see below).

If window-combination-limit is t, splitting W2 in the initial configuration of our scenario would have produced this:

     ______________________________________
    | ____________________________________ |
    || __________________________________ ||
    |||                                  |||
    |||________________W2________________|||
    || __________________________________ ||
    |||                                  |||
    |||________________W4________________|||
    ||_________________W5_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W3_________________||
    |__________________W1__________________|

A new internal window W5 has been created; its children are W2 and the new live window W4. Now, W2 is the only sibling of W4, so enlarging W4 will try to shrink W2, leaving W3 unaffected. Observe that W5 represents a vertical combination of two windows embedded in the vertical combination W1.

Function: set-window-combination-limit window limit

This function sets the combination limit of the window window to limit. This value can be retrieved via the function window-combination-limit. See below for its effects; note that it is only meaningful for internal windows. The split-window function automatically calls this function, passing it t as limit, provided the value of the variable window-combination-limit is t when it is called.

Function: window-combination-limit window

This function returns the combination limit for window.

The combination limit is meaningful only for an internal window. If it is nil, then Emacs is allowed to automatically delete window, in response to a window deletion, in order to group the child windows of window with its sibling windows to form a new window combination. If the combination limit is t, the child windows of window are never automatically recombined with its siblings.

If, in the configuration shown at the beginning of this section, the combination limit of W4 (the parent window of W6 and W7) is t, deleting W5 will not implicitly delete W4 too.

Alternatively, the problems sketched above can be avoided by always resizing all windows in the same combination whenever one of its windows is split or deleted. This also permits splitting windows that would be otherwise too small for such an operation.

User Option: window-combination-resize

If this variable is nil, split-window can only split a window (denoted by window) if window’s screen area is large enough to accommodate both itself and the new window.

If this variable is t, split-window tries to resize all windows that are part of the same combination as window, in order to accommodate the new window. In particular, this may allow split-window to succeed even if window is a fixed-size window or too small to ordinarily split. Furthermore, subsequently resizing or deleting window may resize all other windows in its combination.

The default is nil. Other values are reserved for future use. A specific split operation may ignore the value of this variable if it is affected by a non-nil value of window-combination-limit.

To illustrate the effect of window-combination-resize, consider the following frame layout.

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W2_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W3_________________||
    |__________________W1__________________|

If window-combination-resize is nil, splitting window W3 leaves the size of W2 unchanged:

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||                                    ||
    ||_________________W2_________________||
    | ____________________________________ |
    ||                                    ||
    ||_________________W3_________________||
    | ____________________________________ |
    ||                                    ||
    ||_________________W4_________________||
    |__________________W1__________________|

If window-combination-resize is t, splitting W3 instead leaves all three live windows with approximately the same height:

     ______________________________________
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W2_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W3_________________||
    | ____________________________________ |
    ||                                    ||
    ||                                    ||
    ||_________________W4_________________||
    |__________________W1__________________|

Deleting any of the live windows W2, W3 or W4 will distribute its space proportionally among the two remaining live windows.


29.10 Cyclic Ordering of Windows

When you use the command C-x o (other-window) to select some other window, it moves through live windows in a specific order. For any given configuration of windows, this order never varies. It is called the cyclic ordering of windows.

The ordering is determined by a depth-first traversal of each frame’s window tree, retrieving the live windows which are the leaf nodes of the tree (see Windows and Frames). If the minibuffer is active, the minibuffer window is included too. The ordering is cyclic, so the last window in the sequence is followed by the first one.

Function: next-window &optional window minibuf all-frames

This function returns a live window, the one following window in the cyclic ordering of windows. window should be a live window; if omitted or nil, it defaults to the selected window.

The optional argument minibuf specifies whether minibuffer windows should be included in the cyclic ordering. Normally, when minibuf is nil, a minibuffer window is included only if it is currently active; this matches the behavior of C-x o. (Note that a minibuffer window is active as long as its minibuffer is in use; see Minibuffers).

If minibuf is t, the cyclic ordering includes all minibuffer windows. If minibuf is neither t nor nil, minibuffer windows are not included even if they are active.

The optional argument all-frames specifies which frames to consider:

  • nil means to consider windows on window’s frame. If the minibuffer window is considered (as specified by the minibuf argument), then frames that share the minibuffer window are considered too.
  • t means to consider windows on all existing frames.
  • visible means to consider windows on all visible frames.
  • 0 means to consider windows on all visible or iconified frames.
  • A frame means to consider windows on that specific frame.
  • Anything else means to consider windows on window’s frame, and no others.

If more than one frame is considered, the cyclic ordering is obtained by appending the orderings for those frames, in the same order as the list of all live frames (see Finding All Frames).

Function: previous-window &optional window minibuf all-frames

This function returns a live window, the one preceding window in the cyclic ordering of windows. The other arguments are handled like in next-window.

Command: other-window count &optional all-frames

This function selects a live window, one count places from the selected window in the cyclic ordering of windows. If count is a positive number, it skips count windows forwards; if count is negative, it skips −count windows backwards; if count is zero, that simply re-selects the selected window. When called interactively, count is the numeric prefix argument.

The optional argument all-frames has the same meaning as in next-window, like a nil minibuf argument to next-window.

This function does not select a window that has a non-nil no-other-window window parameter (see Window Parameters), provided that ignore-window-parameters is nil.

If the other-window parameter of the selected window is a function, and ignore-window-parameters is nil, that function will be called with the arguments count and all-frames instead of the normal operation of this function.

Function: walk-windows fun &optional minibuf all-frames

This function calls the function fun once for each live window, with the window as the argument.

It follows the cyclic ordering of windows. The optional arguments minibuf and all-frames specify the set of windows included; these have the same arguments as in next-window. If all-frames specifies a frame, the first window walked is the first window on that frame (the one returned by frame-first-window), not necessarily the selected window.

If fun changes the window configuration by splitting or deleting windows, that does not alter the set of windows walked, which is determined prior to calling fun for the first time.

Function: one-window-p &optional no-mini all-frames

This function returns t if the selected window is the only live window, and nil otherwise.

If the minibuffer window is active, it is normally considered (so that this function returns nil). However, if the optional argument no-mini is non-nil, the minibuffer window is ignored even if active. The optional argument all-frames has the same meaning as for next-window.

The following functions return a window which satisfies some criterion, without selecting it:

Function: get-lru-window &optional all-frames dedicated not-selected no-other

This function returns a live window which is heuristically the least recently used one. The least recently used window is the least recently selected one—the window whose use time is less than the use time of all other live windows (see Selecting Windows). The optional argument all-frames has the same meaning as in next-window.

If any full-width windows are present, only those windows are considered. A minibuffer window is never a candidate. A dedicated window (see Dedicated Windows) is never a candidate unless the optional argument dedicated is non-nil. The selected window is never returned, unless it is the only candidate. However, if the optional argument not-selected is non-nil, this function returns nil in that case. The optional argument no-other, if non-nil, means to never return a window whose no-other-window parameter is non-nil.

Function: get-mru-window &optional all-frames dedicated not-selected no-other

This function is like get-lru-window, but it returns the most recently used window instead. The most recently used window is the most recently selected one—the window whose use time exceeds the use time of all other live windows (see Selecting Windows). The meaning of the arguments is the same as for get-lru-window.

Since in practice the most recently used window is always the selected one, it usually makes sense to call this function with a non-nil not-selected argument only.

Function: get-largest-window &optional all-frames dedicated not-selected no-other

This function returns the window with the largest area (height times width). If there are two candidate windows of the same size, it prefers the one that comes first in the cyclic ordering of windows, starting from the selected window. The meaning of the arguments is the same as for get-lru-window.

Function: get-window-with-predicate predicate &optional minibuf all-frames default

This function calls the function predicate for each of the windows in the cyclic order of windows in turn, passing it the window as an argument. If the predicate returns non-nil for any window, this function stops and returns that window. If no such window is found, the return value is default (which defaults to nil).

The optional arguments minibuf and all-frames specify the windows to search, and have the same meanings as in next-window.


29.11 Buffers and Windows

This section describes low-level functions for examining and setting the contents of windows. See Switching to a Buffer in a Window, for higher-level functions for displaying a specific buffer in a window.

Function: window-buffer &optional window

This function returns the buffer that window is displaying. If window is omitted or nil it defaults to the selected window. If window is an internal window, this function returns nil.

Function: set-window-buffer window buffer-or-name &optional keep-margins

This function makes window display buffer-or-name. window should be a live window; if nil, it defaults to the selected window. buffer-or-name should be a buffer, or the name of an existing buffer. This function does not change which window is selected, nor does it directly change which buffer is current (see The Current Buffer). Its return value is nil.

If window is strongly dedicated to a buffer and buffer-or-name does not specify that buffer, this function signals an error. See Dedicated Windows.

By default, this function resets window’s position, display margins, fringe widths, and scroll bar settings, based on the local variables in the specified buffer. However, if the optional argument keep-margins is non-nil, it leaves window’s display margins, fringes and scroll bar settings alone.

When writing an application, you should normally use display-buffer (see Choosing a Window for Displaying a Buffer) or the higher-level functions described in Switching to a Buffer in a Window, instead of calling set-window-buffer directly.

This runs window-scroll-functions, followed by window-configuration-change-hook. See Hooks for Window Scrolling and Changes.

Variable: buffer-display-count

This buffer-local variable records the number of times a buffer has been displayed in a window. It is incremented each time set-window-buffer is called for the buffer.

Variable: buffer-display-time

This buffer-local variable records the time at which a buffer was last displayed in a window. The value is nil if the buffer has never been displayed. It is updated each time set-window-buffer is called for the buffer, with the value returned by current-time (see Time of Day).

Function: get-buffer-window &optional buffer-or-name all-frames

This function returns the first window displaying buffer-or-name in the cyclic ordering of windows, starting from the selected window (see Cyclic Ordering of Windows). If no such window exists, the return value is nil.

buffer-or-name should be a buffer or the name of a buffer; if omitted or nil, it defaults to the current buffer. The optional argument all-frames specifies which windows to consider:

  • t means consider windows on all existing frames.
  • visible means consider windows on all visible frames.
  • 0 means consider windows on all visible or iconified frames.
  • A frame means consider windows on that frame only.
  • Any other value means consider windows on the selected frame.

Note that these meanings differ slightly from those of the all-frames argument to next-window (see Cyclic Ordering of Windows). This function may be changed in a future version of Emacs to eliminate this discrepancy.

Function: get-buffer-window-list &optional buffer-or-name minibuf all-frames

This function returns a list of all windows currently displaying buffer-or-name. buffer-or-name should be a buffer or the name of an existing buffer. If omitted or nil, it defaults to the current buffer. If the currently selected window displays buffer-or-name, it will be the first in the list returned by this function.

The arguments minibuf and all-frames have the same meanings as in the function next-window (see Cyclic Ordering of Windows). Note that the all-frames argument does not behave exactly like in get-buffer-window.

Command: replace-buffer-in-windows &optional buffer-or-name

This command replaces buffer-or-name with some other buffer, in all windows displaying it. buffer-or-name should be a buffer, or the name of an existing buffer; if omitted or nil, it defaults to the current buffer.

The replacement buffer in each window is chosen via switch-to-prev-buffer (see Window History). With the exception of side windows (see Side Windows), any dedicated window displaying buffer-or-name is deleted if possible (see Dedicated Windows). If such a window is the only window on its frame and there are other frames on the same terminal, the frame is deleted as well. If the dedicated window is the only window on the only frame on its terminal, the buffer is replaced anyway.


29.12 Switching to a Buffer in a Window

This section describes high-level functions for switching to a specified buffer in some window. In general, “switching to a buffer” means to (1) show the buffer in some window, (2) make that window the selected window (and its frame the selected frame), and (3) make the buffer the current buffer.

Do not use these functions to make a buffer temporarily current just so a Lisp program can access or modify it. They have side-effects, such as changing window histories (see Window History), which will surprise the user if used that way. If you want to make a buffer current to modify it in Lisp, use with-current-buffer, save-current-buffer, or set-buffer. See The Current Buffer.

Command: switch-to-buffer buffer-or-name &optional norecord force-same-window

This command attempts to display buffer-or-name in the selected window and make it the current buffer. It is often used interactively (as the binding of C-x b), as well as in Lisp programs. The return value is the buffer switched to.

If buffer-or-name is nil, it defaults to the buffer returned by other-buffer (see The Buffer List). If buffer-or-name is a string that is not the name of any existing buffer, this function creates a new buffer with that name; the new buffer’s major mode is determined by the variable major-mode (see Major Modes).

Normally, the specified buffer is put at the front of the buffer list—both the global buffer list and the selected frame’s buffer list (see The Buffer List). However, this is not done if the optional argument norecord is non-nil.

Sometimes, the selected window may not be suitable for displaying the buffer. This happens if the selected window is a minibuffer window, or if the selected window is strongly dedicated to its buffer (see Dedicated Windows). In such cases, the command normally tries to display the buffer in some other window, by invoking pop-to-buffer (see below).

If the optional argument force-same-window is non-nil and the selected window is not suitable for displaying the buffer, this function always signals an error when called non-interactively. In interactive use, if the selected window is a minibuffer window, this function will try to use some other window instead. If the selected window is strongly dedicated to its buffer, the option switch-to-buffer-in-dedicated-window described next can be used to proceed.

User Option: switch-to-buffer-in-dedicated-window

This option, if non-nil, allows switch-to-buffer to proceed when called interactively and the selected window is strongly dedicated to its buffer.

The following values are respected:

nil

Disallows switching and signals an error as in non-interactive use.

prompt

Prompts the user whether to allow switching.

pop

Invokes pop-to-buffer to proceed.

t

Marks the selected window as non-dedicated and proceeds.

This option does not affect non-interactive calls of switch-to-buffer.

By default, switch-to-buffer tries to preserve window-point. This behavior can be tuned using the following option.

User Option: switch-to-buffer-preserve-window-point

If this variable is nil, switch-to-buffer displays the buffer specified by buffer-or-name at the position of that buffer’s point. If this variable is already-displayed, it tries to display the buffer at its previous position in the selected window, provided the buffer is currently displayed in some other window on any visible or iconified frame. If this variable is t, switch-to-buffer unconditionally tries to display the buffer at its previous position in the selected window.

This variable is ignored if the buffer is already displayed in the selected window or never appeared in it before, or if switch-to-buffer calls pop-to-buffer to display the buffer.

User Option: switch-to-buffer-obey-display-actions

If this variable is non-nil, switch-to-buffer respects display actions specified by display-buffer-overriding-action, display-buffer-alist and other display related variables.

The next two commands are similar to switch-to-buffer, except for the described features.

Command: switch-to-buffer-other-window buffer-or-name &optional norecord

This function displays the buffer specified by buffer-or-name in some window other than the selected window. It uses the function pop-to-buffer internally (see below).

If the selected window already displays the specified buffer, it continues to do so, but another window is nonetheless found to display it as well.

The buffer-or-name and norecord arguments have the same meanings as in switch-to-buffer.

Command: switch-to-buffer-other-frame buffer-or-name &optional norecord

This function displays the buffer specified by buffer-or-name in a new frame. It uses the function pop-to-buffer internally (see below).

If the specified buffer is already displayed in another window, in any frame on the current terminal, this switches to that window instead of creating a new frame. However, the selected window is never used for this.

The buffer-or-name and norecord arguments have the same meanings as in switch-to-buffer.

The above commands use the function pop-to-buffer, which flexibly displays a buffer in some window and selects that window for editing. In turn, pop-to-buffer uses display-buffer for displaying the buffer. Hence, all the variables affecting display-buffer will affect it as well. See Choosing a Window for Displaying a Buffer, for the documentation of display-buffer.

Command: pop-to-buffer buffer-or-name &optional action norecord

This function makes buffer-or-name the current buffer and displays it in some window, preferably not the window currently selected. It then selects the displaying window. If that window is on a different graphical frame, that frame is given input focus if possible (see Input Focus).

If buffer-or-name is nil, it defaults to the buffer returned by other-buffer (see The Buffer List). If buffer-or-name is a string that is not the name of any existing buffer, this function creates a new buffer with that name; the new buffer’s major mode is determined by the variable major-mode (see Major Modes). In any case, that buffer is made current and returned, even when no suitable window was found to display it.

If action is non-nil, it should be a display action to pass to display-buffer (see Choosing a Window for Displaying a Buffer). Alternatively, a non-nil, non-list value means to pop to a window other than the selected one—even if the buffer is already displayed in the selected window.

Like switch-to-buffer, this function updates the buffer list unless norecord is non-nil.


29.13 Displaying a Buffer in a Suitable Window

This section describes lower-level functions Emacs uses to find or create a window for displaying a specified buffer. The common workhorse of these functions is display-buffer which eventually handles all incoming requests for buffer display (see Choosing a Window for Displaying a Buffer).

display-buffer delegates the task of finding a suitable window to so-called action functions (see Action Functions for Buffer Display). First, display-buffer compiles a so-called action alist—a special association list that action functions can use to fine-tune their behavior. Then it passes that alist on to each action function it calls (see Action Alists for Buffer Display).

The behavior of display-buffer is highly customizable. To understand how customizations are used in practice, you may wish to study examples illustrating the order of precedence which display-buffer uses to call action functions (see Precedence of Action Functions). To avoid conflicts between Lisp programs calling display-buffer and user customizations of its behavior, it may make sense to follow a number of guidelines which are sketched in the final part of this section (see The Zen of Buffer Display).


29.13.1 Choosing a Window for Displaying a Buffer

The command display-buffer flexibly chooses a window for display, and displays a specified buffer in that window. It can be called interactively, via the key binding C-x 4 C-o. It is also used as a subroutine by many functions and commands, including switch-to-buffer and pop-to-buffer (see Switching to a Buffer in a Window).

This command performs several complex steps to find a window to display in. These steps are described by means of display actions, which have the form (functions . alist). Here, functions is either a single function or a list of functions, referred to as “action functions” (see Action Functions for Buffer Display); and alist is an association list, referred to as “action alist” (see Action Alists for Buffer Display). See The Zen of Buffer Display, for samples of display actions.

An action function accepts two arguments: the buffer to display and an action alist. It attempts to display the buffer in some window, picking or creating a window according to its own criteria. If successful, it returns the window; otherwise, it returns nil.

display-buffer works by combining display actions from several sources, and calling the action functions in turn, until one of them manages to display the buffer and returns a non-nil value.

Command: display-buffer buffer-or-name &optional action frame

This command makes buffer-or-name appear in some window, without selecting the window or making the buffer current. The argument buffer-or-name must be a buffer or the name of an existing buffer. The return value is the window chosen to display the buffer, or nil if no suitable window was found.

The optional argument action, if non-nil, should normally be a display action (described above). display-buffer builds a list of action functions and an action alist, by consolidating display actions from the following sources (in order of their precedence, from highest to lowest):

  • The variable display-buffer-overriding-action.
  • The user option display-buffer-alist.
  • The action argument.
  • The user option display-buffer-base-action.
  • The constant display-buffer-fallback-action.

In practice this means that display-buffer builds a list of all action functions specified by these display actions. The first element of this list is the first action function specified by display-buffer-overriding-action, if any. Its last element is display-buffer-pop-up-frame—the last action function specified by display-buffer-fallback-action. Duplicates are not removed from this list—hence one and the same action function may be called multiple times during one call of display-buffer.

display-buffer calls the action functions specified by this list in turn, passing the buffer as the first argument and the combined action alist as the second argument, until one of the functions returns non-nil. See Precedence of Action Functions, for examples how display actions specified by different sources are processed by display-buffer.

Note that the second argument is always the list of all action alist entries specified by the sources named above. Hence, the first element of that list is the first action alist entry specified by display-buffer-overriding-action, if any. Its last element is the last alist entry of display-buffer-base-action, if any (the action alist of display-buffer-fallback-action is empty).

Note also, that the combined action alist may contain duplicate entries and entries for the same key with different values. As a rule, action functions always use the first association of a key they find. Hence, the association an action function uses is not necessarily the association provided by the display action that specified that action function,

The argument action can also have a non-nil, non-list value. This has the special meaning that the buffer should be displayed in a window other than the selected one, even if the selected window is already displaying it. If called interactively with a prefix argument, action is t. Lisp programs should always supply a list value.

The optional argument frame, if non-nil, specifies which frames to check when deciding whether the buffer is already displayed. It is equivalent to adding an element (reusable-frames . frame) to the action alist of action (see Action Alists for Buffer Display). The frame argument is provided for compatibility reasons, Lisp programs should not use it.

Variable: display-buffer-overriding-action

The value of this variable should be a display action, which is treated with the highest priority by display-buffer. The default value is an empty display action, i.e., (nil . nil).

User Option: display-buffer-alist

The value of this option is an alist mapping conditions to display actions. Each condition is passed to buffer-match-p (see The Buffer List), along with the buffer name and the action argument passed to display-buffer. If it returns a non-nil value, then display-buffer uses the corresponding display action to display the buffer. Caveat: if you use derived-mode or major-mode as condition, buffer-match-p could fail to report a match if display-buffer is called before the major mode of the buffer is set.

User Option: display-buffer-base-action

The value of this option should be a display action. This option can be used to define a standard display action for calls to display-buffer.

Constant: display-buffer-fallback-action

This display action specifies the fallback behavior for display-buffer if no other display actions are given.


29.13.2 Action Functions for Buffer Display

An action function is a function display-buffer calls for choosing a window to display a buffer. Action functions take two arguments: buffer, the buffer to display, and alist, an action alist (see Action Alists for Buffer Display). They are supposed to return a window displaying buffer if they succeed and nil if they fail.

The following basic action functions are defined in Emacs.

Function: display-buffer-same-window buffer alist

This function tries to display buffer in the selected window. It fails if the selected window is a minibuffer window or is dedicated to another buffer (see Dedicated Windows). It also fails if alist has a non-nil inhibit-same-window entry.

Function: display-buffer-reuse-window buffer alist

This function tries to display buffer by finding a window that is already displaying it. Windows on the selected frame are preferred to windows on other frames.

If alist has a non-nil inhibit-same-window entry, the selected window is not eligible for reuse. The set of frames to search for a window already displaying buffer can be specified with the help of the reusable-frames action alist entry. If alist contains no reusable-frames entry, this function searches just the selected frame.

If this function chooses a window on another frame, it makes that frame visible and, unless alist contains an inhibit-switch-frame entry, raises that frame if necessary.

Function: display-buffer-reuse-mode-window buffer alist

This function tries to display buffer by finding a window that is displaying a buffer in a given mode.

If alist contains a mode entry, its value specifies a major mode (a symbol) or a list of major modes. If alist contains no mode entry, the current major mode of buffer is used instead. A window is a candidate if it displays a buffer whose mode derives from one of the modes specified thusly.

The behavior is also controlled by alist entries for inhibit-same-window, reusable-frames and inhibit-switch-frame, like display-buffer-reuse-window does.

Function: display-buffer-pop-up-window buffer alist

This function tries to display buffer by splitting the largest or least recently-used window (usually located on the selected frame). It actually performs the split by calling the function specified by split-window-preferred-function (see Additional Options for Displaying Buffers).

The size of the new window can be adjusted by supplying window-height and window-width entries in alist. If alist contains a preserve-size entry, Emacs will also try to preserve the size of the new window during future resize operations (see Preserving Window Sizes).

This function fails if no window can be split. More often than not, this happens because no window is large enough to allow splitting. Setting split-height-threshold or split-width-threshold to lower values may help in this regard. Splitting also fails when the selected frame has an unsplittable frame parameter; see Buffer Parameters.

Function: display-buffer-in-previous-window buffer alist

This function tries to display buffer in a window where it was displayed previously.

If alist contains a non-nil inhibit-same-window entry, the selected window is not eligible for use. A dedicated window is usable only if it already shows buffer. If alist contains a previous-window entry, the window specified by that entry is usable even if it never showed buffer before.

If alist contains a reusable-frames entry (see Action Alists for Buffer Display), its value determines which frames to search for a suitable window. If alist contains no reusable-frames entry, this function searches just the selected frame if display-buffer-reuse-frames and pop-up-frames are both nil; it searches all frames on the current terminal if either of those variables is non-nil.

If more than one window qualifies as usable according to these rules, this function makes a choice in the following order of preference:

  • The window specified by any previous-window alist entry, provided it is not the selected window.
  • A window that showed buffer before, provided it is not the selected window.
  • The selected window if it is either specified by a previous-window alist entry or showed buffer before.
Function: display-buffer-use-some-window buffer alist

This function tries to display buffer by choosing an existing window and displaying the buffer in that window. It first tries to find a window that has not been used recently (see Cyclic Ordering of Windows) on any frame specified by a lru-frames alist entry, falling back to the selected frame if no such entry exists. It also prefers windows that satisfy the constraints specified by window-min-width and window-min-height alist entries; preferring full-width windows if no window-min-width entry is found. Finally, it will not return a window whose use time is higher than that specified by any lru-time entry provided by alist.

If no less recently used window is found, this function will try to use some other window, preferably a large window on some visible frame. It can fail if all windows are dedicated to other buffers (see Dedicated Windows).

Function: display-buffer-use-least-recent-window buffer alist

This function is similar to display-buffer-use-some-window, but will try harder to not use the a recently used window. In particular, it does not use the selected window. In addition, it will first try to reuse a window that shows buffer already, base the decision whether it should use a window showing another buffer on that window’s use time alone and pop up a new window if no usable window is found.

Finally, this function will bump the use time (see Selecting Windows) of any window it returns in order to avoid that further invocations will use that window for showing another buffer. An application that wants to display several buffers in a row can help this function by providing a lru-time alist entry it has initially set to the value of the selected window’s use time. Each invocation of this function will then bump the use time of the window returned to a value higher than that and a subsequent invocation will inhibit this function to use a window it returned earlier.

Function: display-buffer-in-direction buffer alist

This function tries to display buffer at a location specified by alist. For this purpose, alist should contain a direction entry whose value is one of left, above (or up), right and below (or down). Other values are usually interpreted as below.

If alist also contains a window entry, its value specifies a reference window. That value can be a special symbol like main which stands for the selected frame’s main window (see Side Window Options and Functions) or root standing for the selected frame’s root window (see Windows and Frames). It can also specify an arbitrary valid window. Any other value (or omitting the window entry entirely) means to use the selected window as reference window.

This function first tries to reuse a window in the specified direction that already shows buffer. If no such window exists, it tries to split the reference window in order to produce a new window in the specified direction. If this fails as well, it will try to display buffer in an existing window in the specified direction. In either case, the window chosen will appear on the side of the reference window specified by the direction entry, sharing at least one edge with the reference window.

If the reference window is live, the edge the chosen window will share with it is always the opposite of the one specified by the direction entry. For example, if the value of the direction entry is left, the chosen window’s right edge coordinate (see Coordinates and Windows) will equal the reference window’s left edge coordinate.

If the reference window is internal, a reused window must share with it the edge specified by the direction entry. Hence if, for example, the reference window is the frame’s root window and the value of the direction entry is left, a reused window must be on the left of the frame. This means that the left edge coordinate of the chosen window and that of the reference window are the same.

A new window, however, will be created by splitting the reference window such that the chosen window will share the opposite edge with the reference window. In our example, a new root window would be created with a new live window and the reference window as its children. The chosen window’s right edge coordinate would then equal the left edge coordinate of the reference window. Its left edge coordinate would equal the left edge coordinate of the frame’s new root window.

Four special values for direction entries allow to implicitly specify the selected frame’s main window as the reference window: leftmost, top, rightmost and bottom. This means that instead of, for example, (direction . left) (window . main) one can just specify (direction . leftmost). An existing window alist entry is ignored in such cases.

Function: display-buffer-below-selected buffer alist

This function tries to display buffer in a window below the selected window. If there is a window below the selected one and that window already displays buffer, it reuses that window.

If there is no such window, this function tries to create a new window by splitting the selected one, and displays buffer there. It will also try to adjust that window’s size provided alist contains a suitable window-height or window-width entry, see above.

If splitting the selected window fails and there is a non-dedicated window below the selected one showing some other buffer, this function tries to use that window for showing buffer.

If alist contains a window-min-height entry, this function ensures that the window used is or can become at least as high as specified by that entry’s value. Note that this is only a guarantee. In order to actually resize the window used, alist must also provide an appropriate window-height entry.

Function: display-buffer-at-bottom buffer alist

This function tries to display buffer in a window at the bottom of the selected frame.

This either tries to split the window at the bottom of the frame or the frame’s root window, or to reuse an existing window at the bottom of the selected frame.

Function: display-buffer-pop-up-frame buffer alist

This function creates a new frame, and displays the buffer in that frame’s window. It actually performs the frame creation by calling the function specified in pop-up-frame-function (see Additional Options for Displaying Buffers). If alist contains a pop-up-frame-parameters entry, the associated value is added to the newly created frame’s parameters.

Function: display-buffer-full-frame buffer alist

This function displays the buffer on the current frame, deleting all other windows so that it takes up the full frame.

Function: display-buffer-in-child-frame buffer alist

This function tries to display buffer in a child frame (see Child Frames) of the selected frame, either reusing an existing child frame or by making a new one. If alist has a non-nil child-frame-parameters entry, the corresponding value is an alist of frame parameters to give the new frame. A parent-frame parameter specifying the selected frame is provided by default. If the child frame should become the child of another frame, a corresponding entry must be added to alist.

The appearance of child frames is largely dependent on the parameters provided via alist. It is advisable to use at least ratios to specify the size (see Size Parameters) and the position (see Position Parameters) of the child frame, and to add a keep-ratio parameter (see Frame Interaction Parameters), in order to make sure that the child frame remains visible. For other parameters that should be considered see Child Frames.

Function: display-buffer-use-some-frame buffer alist

This function tries to display buffer by finding a frame that meets a predicate (by default any frame other than the selected frame).

If this function chooses a window on another frame, it makes that frame visible and, unless alist contains an inhibit-switch-frame entry, raises that frame if necessary.

If alist has a non-nil frame-predicate entry, its value is a function taking one argument (a frame), returning non-nil if the frame is a candidate; this function replaces the default predicate.

If alist has a non-nil inhibit-same-window entry, the selected window is not used; thus if the selected frame has a single window, it is not used.

Function: display-buffer-no-window buffer alist

If alist has a non-nil allow-no-window entry, then this function does not display buffer and returns the symbol fail. This constitutes the only exception to the convention that an action function returns either nil or a window showing buffer. If alist has no such allow-no-window entry, this function returns nil.

If this function returns fail, display-buffer will skip the execution of any further display actions and return nil immediately. If this function returns nil, display-buffer will continue with the next display action, if any.

It is assumed that when a caller of display-buffer specifies a non-nil allow-no-window entry, it is also able to handle a nil return value.

Two other action functions are described in their proper sections—display-buffer-in-side-window (see Displaying Buffers in Side Windows) and display-buffer-in-atom-window (see Atomic Windows).


29.13.3 Action Alists for Buffer Display

An action alist is an association list mapping predefined symbols recognized by action functions to values these functions are supposed to interpret accordingly. In each call, display-buffer constructs a new, possibly empty action alist and passes that entire list on to any action function it calls.

By design, action functions are free in their interpretation of action alist entries. In fact, some entries like allow-no-window or previous-window have a meaning only for one or a few action functions, and are ignored by the rest. Other entries, like inhibit-same-window or window-parameters, are supposed to be respected by most action functions, including those provided by application programs and external packages.

In the previous subsection we have described in detail how individual action functions interpret the action alist entries they care about. Here we give a reference list of all known action alist entries according to their symbols, together with their values and action functions (see Action Functions for Buffer Display) that recognize them. Throughout this list, the terms “buffer” will refer to the buffer display-buffer is supposed to display, and “value” refers to the entry’s value.

inhibit-same-window

If the value is non-nil, this signals that the selected window must not be used for displaying the buffer. All action functions that (re-)use an existing window should respect this entry.

previous-window

The value must specify a window that may have displayed the buffer previously. display-buffer-in-previous-window will give preference to such a window provided it is still live and not dedicated to another buffer.

mode

The value is either a major mode or a list of major modes. display-buffer-reuse-mode-window may reuse a window whenever the value specified by this entry matches the major mode of that window’s buffer. Other action functions ignore such entries.

frame-predicate

The value must be a function taking one argument (a frame), supposed to return non-nil if that frame is a candidate for displaying the buffer. This entry is used by display-buffer-use-some-frame.

reusable-frames

The value specifies the set of frames to search for a window that can be reused because it already displays the buffer. It can be set as follows:

  • nil means consider only windows on the selected frame. (Actually, the last frame used that is not a minibuffer-only frame.)
  • visible means consider windows on all visible frames.
  • 0 means consider windows on all visible or iconified frames.
  • A frame means consider windows on that frame only.
  • t means consider windows on all frames. (Note that this value is rarely the right thing to use—it might also return a tooltip frame.)

Note that the meaning of nil differs slightly from that of the all-frames argument to next-window (see Cyclic Ordering of Windows).

A major client of this is display-buffer-reuse-window, but all other action functions that try to reuse a window are affected as well. display-buffer-in-previous-window consults it when searching for a window that previously displayed the buffer on another frame.

inhibit-switch-frame

A non-nil value prevents another frame from being raised or selected, if the window chosen by display-buffer is displayed there. Primarily affected by this are display-buffer-use-some-frame and display-buffer-reuse-window. Ideally, display-buffer-pop-up-frame should be affected as well, but there is no guarantee that the window manager will comply.

window-parameters

The value specifies an alist of window parameters to give the chosen window. All action functions that choose a window should process this entry.

window-min-width

The value specifies a minimum width of the window used, in canonical frame columns. The special value full-width means the chosen window should be one that has no other windows on the left or right of it in its frame.

This entry is currently honored by display-buffer-use-some-window and display-buffer-use-least-recent-window, which try hard to avoid returning a less recently used window that does not satisfy the entry.

Note that providing such an entry alone does not necessarily make the window as wide as specified by its value. To actually resize an existing window or make a new window as wide as specified by this entry’s value, a window-width entry specifying that value should be provided as well. Such a window-width entry can, however, specify a completely different value, or ask the window width to fit that of its buffer, in which case the window-min-width entry provides the guaranteed minimum width of the window.

window-min-height

The value specifies a minimum height of the window used, in canonical frame lines. The special value full-height means the chosen window should be a full-height window, one that has no other windows above or below it in its frame.

This entry is currently honored by display-buffer-below-selected which does not use a window that is not as high as specified by this entry. It’s also honored by display-buffer-use-some-window and display-buffer-use-least-recent-window which try hard to avoid returning a less recently used window if it does not satisfy this constraint.

Note that providing such an entry alone does not necessarily make the window as tall as specified by its value. To actually resize an existing window or make a new window as tall as specified by that value, a window-height entry specifying that value should be provided as well. Such a window-height entry can, however, specify a completely different value or ask the window height to be fit to that of its buffer in which case the window-min-height entry provides the guaranteed minimum height of the window used.

window-height

The value specifies whether and how to adjust the height of the chosen window and can be one of the following:

  • nil means to leave the height of the chosen window alone.
  • An integer number specifies the desired total height of the chosen window in lines.
  • A floating-point number specifies the fraction of the chosen window’s desired total height with respect to the total height of its frame’s root window.
  • A cons cell whose CAR is body-lines and whose CDR is an integer that specifies the height of the chosen window’s body in frame lines.
  • If the value specifies a function, that function is called with one argument—the chosen window. The function is supposed to adjust the height of the window; its return value is ignored. Suitable functions are fit-window-to-buffer and shrink-window-if-larger-than-buffer, see Resizing Windows.

By convention, the height of the chosen window is adjusted only if the window is part of a vertical combination (see Windows and Frames) to avoid changing the height of other, unrelated windows. Also, this entry should be processed only under certain conditions which are specified right below this list.

window-width

This entry is similar to the window-height entry described before, but used to adjust the chosen window’s width instead. The value can be one of the following:

  • nil means to leave the width of the chosen window alone.
  • An integer specifies the desired total width of the chosen window in columns.
  • A floating-point number specifies the fraction of the chosen window’s desired total width with respect to the total width of the frame’s root window.
  • A cons cell whose CAR is body-columns and whose CDR is an integer that specifies the width of the chosen window’s body in frame columns.
  • If the value specifies a function, that function is called with one argument—the chosen window. The function is supposed to adjust the width of the window; its return value is ignored.
window-size

This entry is a combination of the two preceding ones and can be used to adjust the chosen window’s height and width. Since windows can be resized in one direction only without affecting other windows, window-size is effective only to set up the size of a window appearing alone on a frame. The value can be one of the following:

  • nil means to leave the size of the chosen window alone.
  • A cons cell of two integers specifies the desired total width and height of the chosen window in lines and columns. It’s effect is to adjust the size of the frame accordingly.
  • A cons cell whose CAR equals body-chars and whose CDR is a cons cell of two integers—the desired body width and height of the chosen window in frame columns and lines. It’s effect is to adjust the size of the frame accordingly.
  • If the value specifies a function, that function is called with one argument—the chosen window. The function is supposed to adjust the size of the window’s frame; its return value is ignored.

This entry should be processed under only certain conditions which are specified right below this list.

dedicated

If non-nil, such an entry tells display-buffer to mark any window it creates as dedicated to its buffer (see Dedicated Windows). It does that by calling set-window-dedicated-p with the chosen window as first argument and the entry’s value as second. Side windows are by default dedicated with the value side (see Side Window Options and Functions).

preserve-size

If non-nil such an entry tells Emacs to preserve the size of the window chosen (see Preserving Window Sizes). The value should be either (t . nil) to preserve the width of the window, (nil . t) to preserve its height or (t . t) to preserve both, its width and its height. This entry should be processed only under certain conditions which are specified right after this list.

lru-frames

The value specifies the set of frames to search for a window that can be used to display the buffer. It is honored by display-buffer-use-some-window and display-buffer-use-least-recent-window when trying to find a less recently used window showing some other buffer. Its values are the same as for the reusable-frames entry described above.

lru-time

The value is supposed to specify a use time (see Selecting Windows). This entry is honored by display-buffer-use-some-window and display-buffer-use-least-recent-window when trying to find a less recently used window showing some other buffer. If a window’s use time is higher than the value specified by this option, these action functions will not consider such a window for displaying the buffer.

bump-use-time

If non-nil, such an entry will cause display-buffer to bump the use time (see Selecting Windows) of the window it uses. This should avoid later use of this window by action functions like display-buffer-use-some-window and display-buffer-use-least-recent-window for showing another buffer.

There is a fine difference between using this entry and using the action function display-buffer-use-least-recent-window. Calling the latter means to only bump the use times of windows that function uses for displaying the buffer. The entry described here will cause display-buffer to bump the use time of any window used for displaying a buffer.

pop-up-frame-parameters

The value specifies an alist of frame parameters to give a new frame, if one is created. display-buffer-pop-up-frame is its one and only addressee.

parent-frame

The value specifies the parent frame to be used when the buffer is displayed on a child frame. This entry is used only by display-buffer-in-child-frame.

child-frame-parameters

The value specifies an alist of frame parameters to use when the buffer is displayed on a child frame. This entry is used only by display-buffer-in-child-frame.

side

The value denotes the side of the frame or window where a new window displaying the buffer shall be created. This entry is used by display-buffer-in-side-window to indicate the side of the frame where a new side window shall be placed (see Displaying Buffers in Side Windows). It is also used by display-buffer-in-atom-window to indicate the side of an existing window where the new window shall be located (see Atomic Windows).

slot

If non-nil, the value specifies the slot of the side window supposed to display the buffer. This entry is used only by display-buffer-in-side-window.

direction

The value specifies a direction which, together with a window entry, allows display-buffer-in-direction to determine the location of the window to display the buffer.

window

The value specifies a window that is in some way related to the window chosen by display-buffer. This entry is currently used by display-buffer-in-atom-window to indicate the window on whose side the new window shall be created. It is also used by display-buffer-in-direction to specify the reference window on whose side the resulting window shall appear.

allow-no-window

If the value is non-nil, display-buffer does not necessarily have to display the buffer and the caller is prepared to accept that. This entry is not intended for user customizations, since there is no guarantee that an arbitrary caller of display-buffer will be able to handle the case that no window will display the buffer. display-buffer-no-window is the only action function that cares about this entry.

body-function

The value must be a function taking one argument (a displayed window). This function can be used to fill the displayed window’s body with some contents that might depend on dimensions of the displayed window. It is called after the buffer is displayed, and before the entries window-height, window-width and preserve-size are applied that could resize the window to fit it to the inserted contents.

By convention, the entries window-height, window-width and preserve-size are applied after the chosen window’s buffer has been set up and if and only if that window never showed another buffer before. More precisely, the latter means that the window must have been either created by the current display-buffer call or the window was created earlier by display-buffer to show the buffer and never was used to show another buffer until it was reused by the current invocation of display-buffer.

If no window-height, window-width or window-size entry was specified, the window may still be resized automatically when the buffer is temporary and temp-buffer-resize-mode has been enabled, Temporary Displays. In that case, the CDR of a window-height, window-width or window-size entry can be used to inhibit or override the default behavior of temp-buffer-resize-mode for specific buffers or invocations of display-buffer.


29.13.4 Additional Options for Displaying Buffers

The behavior of buffer display actions (see Choosing a Window for Displaying a Buffer) can be further modified by the following user options.

User Option: pop-up-windows

If the value of this variable is non-nil, display-buffer is allowed to split an existing window to make a new window for displaying in. This is the default.

This variable is provided for backward compatibility only. It is obeyed by display-buffer via a special mechanism in display-buffer-fallback-action, which calls the action function display-buffer-pop-up-window (see Action Functions for Buffer Display) when the value of this option is non-nil. It is not consulted by display-buffer-pop-up-window itself, which the user may specify directly in display-buffer-alist etc.

User Option: split-window-preferred-function

This variable specifies a function for splitting a window, in order to make a new window for displaying a buffer. It is used by the display-buffer-pop-up-window action function to actually split the window.

The value must be a function that takes one argument, a window, and returns either a new window (which will be used to display the desired buffer) or nil (which means the splitting failed). The default value is split-window-sensibly, which is documented next.

Function: split-window-sensibly &optional window

This function tries to split window and return the newly created window. If window cannot be split, it returns nil. If window is omitted or nil, it defaults to the selected window.

This function obeys the usual rules that determine when a window may be split (see Splitting Windows). It first tries to split by placing the new window below, subject to the restriction imposed by split-height-threshold (see below), in addition to any other restrictions. If that fails, it tries to split by placing the new window to the right, subject to split-width-threshold (see below). If that also fails, and the window is the only window on its frame, this function again tries to split and place the new window below, disregarding split-height-threshold. If this fails as well, this function gives up and returns nil.

User Option: split-height-threshold

This variable specifies whether split-window-sensibly is allowed to split the window placing the new window below. If it is an integer, that means to split only if the original window has at least that many lines. If it is nil, that means not to split this way.

User Option: split-width-threshold

This variable specifies whether split-window-sensibly is allowed to split the window placing the new window to the right. If the value is an integer, that means to split only if the original window has at least that many columns. If the value is nil, that means not to split this way.

User Option: even-window-sizes

This variable, if non-nil, causes display-buffer to even window sizes whenever it reuses an existing window, and that window is adjacent to the selected one.

If its value is width-only, sizes are evened only if the reused window is on the left or right of the selected one and the selected window is wider than the reused one. If its value is height-only sizes are evened only if the reused window is above or beneath the selected window and the selected window is higher than the reused one. Any other non-nil value means to even sizes in any of these cases provided the selected window is larger than the reused one in the sense of their combination.

User Option: pop-up-frames

If the value of this variable is non-nil, that means display-buffer may display buffers by making new frames. The default is nil.

A non-nil value also means that when display-buffer is looking for a window already displaying buffer-or-name, it can search any visible or iconified frame, not just the selected frame.

This variable is provided mainly for backward compatibility. It is obeyed by display-buffer via a special mechanism in display-buffer-fallback-action, which calls the action function display-buffer-pop-up-frame (see Action Functions for Buffer Display) if the value is non-nil. (This is done before attempting to split a window.) This variable is not consulted by display-buffer-pop-up-frame itself, which the user may specify directly in display-buffer-alist etc.

User Option: pop-up-frame-function

This variable specifies a function for creating a new frame, in order to make a new window for displaying a buffer. It is used by the display-buffer-pop-up-frame action function.

The value should be a function that takes no arguments and returns a frame, or nil if no frame could be created. The default value is a function that creates a frame using the parameters specified by pop-up-frame-alist (see below).

User Option: pop-up-frame-alist

This variable holds an alist of frame parameters (see Frame Parameters), which is used by the function specified by pop-up-frame-function to make a new frame. The default is nil.

This option is provided for backward compatibility only. Note, that when display-buffer-pop-up-frame calls the function specified by pop-up-frame-function, it prepends the value of all pop-up-frame-parameters action alist entries to pop-up-frame-alist so that the values specified by the action alist entry effectively override any corresponding values of pop-up-frame-alist.

Hence, users should set up a pop-up-frame-parameters action alist entry in display-buffer-alist instead of customizing pop-up-frame-alist. Only this will guarantee that the value of a parameter specified by the user overrides the value of that parameter specified by the caller of display-buffer.

Many efforts in the design of display-buffer have been given to maintain compatibility with code that uses older options like pop-up-windows, pop-up-frames, pop-up-frame-alist, same-window-buffer-names and same-window-regexps. Lisp Programs and users should refrain from using these options. Above we already warned against customizing pop-up-frame-alist. Here we describe how to convert the remaining options to use display actions instead.

pop-up-windows

This variable is t by default. Instead of customizing it to nil and thus telling display-buffer what not to do, it’s much better to list in display-buffer-base-action the action functions it should try instead as, for example:

(setopt
 display-buffer-base-action
 '((display-buffer-reuse-window display-buffer-same-window
    display-buffer-in-previous-window
    display-buffer-use-some-window)))
pop-up-frames

Instead of customizing this variable to t, customize display-buffer-base-action, for example, as follows:

(setopt
 display-buffer-base-action
 '((display-buffer-reuse-window display-buffer-pop-up-frame)
   (reusable-frames . 0)))
same-window-buffer-names
same-window-regexps

Instead of adding a buffer name or a regular expression to one of these options use a display-buffer-alist entry for that buffer specifying the action function display-buffer-same-window.

(setopt
 display-buffer-alist
 (cons '("\\*foo\\*" (display-buffer-same-window))
        display-buffer-alist))

29.13.5 Precedence of Action Functions

From the past subsections we already know that display-buffer must be supplied with a number of display actions (see Choosing a Window for Displaying a Buffer) in order to display a buffer. In a completely uncustomized Emacs, these actions are specified by display-buffer-fallback-action in the following order of precedence: Reuse a window, pop up a new window on the same frame, use a window previously showing the buffer, use some window and pop up a new frame. (Note that the remaining actions named by display-buffer-fallback-action are void in an uncustomized Emacs).

Consider the following form:

(display-buffer (get-buffer-create "*foo*"))

Evaluating this form in the buffer *scratch* of an uncustomized Emacs session will usually fail to reuse a window that shows *foo* already, but succeed in popping up a new window. Evaluating the same form again will now not cause any visible changes—display-buffer reused the window already showing *foo* because that action was applicable and had the highest precedence among all applicable actions.

Popping up a new window will fail if there is not enough space on the selected frame. In an uncustomized Emacs it typically fails when there are already two windows on a frame. For example, if you now type C-x 1 followed by C-x 2 and evaluate the form once more, *foo* should show up in the lower window—display-buffer just used “some” window. If, before typing C-x 2 you had typed C-x o, *foo* would have been shown in the upper window because “some” window stands for the “least recently used” window and the selected window has been least recently used if and only if it is alone on its frame.

Let’s assume you did not type C-x o and *foo* is shown in the lower window. Type C-x o to get there followed by C-x left and evaluate the form again. This should display *foo* in the same, lower window because that window had already shown *foo* previously and was therefore chosen instead of some other window.

So far we have only observed the default behavior in an uncustomized Emacs session. To see how this behavior can be customized, let’s consider the option display-buffer-base-action. It provides a very coarse customization which conceptually affects the display of any buffer. It can be used to supplement the actions supplied by display-buffer-fallback-action by reordering them or by adding actions that are not present there but fit more closely the user’s editing practice. However, it can also be used to change the default behavior in a more profound way.

Let’s consider a user who, as a rule, prefers to display buffers on another frame. Such a user might provide the following customization:

(setopt
 display-buffer-base-action
 '((display-buffer-reuse-window display-buffer-pop-up-frame)
   (reusable-frames . 0)))

This setting will cause display-buffer to first try to find a window showing the buffer on a visible or iconified frame and, if no such frame exists, pop up a new frame. You can observe this behavior on a graphical system by typing C-x 1 in the window showing *scratch* and evaluating our canonical display-buffer form. This will usually create (and give focus to) a new frame whose root window shows *foo*. Iconify that frame and evaluate the canonical form again: display-buffer will reuse the window on the new frame (usually raising the frame and giving it focus too).

Only if creating a new frame fails, display-buffer will apply the actions supplied by display-buffer-fallback-action which means to again try reusing a window, popping up a new window and so on. A trivial way to make frame creation fail is supplied by the following form:

(let ((pop-up-frame-function 'ignore))
  (display-buffer (get-buffer-create "*foo*")))

We will forget about that form immediately after observing that it fails to create a new frame and uses a fallback action instead.

Note that display-buffer-reuse-window appears redundant in the customization of display-buffer-base-action because it is already part of display-buffer-fallback-action and should be tried there anyway. However, that would fail because due to the precedence of display-buffer-base-action over display-buffer-fallback-action, at that time display-buffer-pop-up-frame would have already won the race. In fact, this:

(setopt
 display-buffer-base-action
 '(display-buffer-pop-up-frame (reusable-frames . 0)))

would cause display-buffer to always pop up a new frame which is probably not what our user wants.

So far, we have only shown how users can customize the default behavior of display-buffer. Let us now see how applications can change the course of display-buffer. The canonical way to do that is to use the action argument of display-buffer or a function that calls it, like, for example, pop-to-buffer (see Switching to a Buffer in a Window).

Suppose an application wants to display *foo* preferably below the selected window (to immediately attract the attention of the user to the new window) or, if that fails, in a window at the bottom of the frame. It could do that with a call like this:

(display-buffer
 (get-buffer-create "*foo*")
 '((display-buffer-below-selected display-buffer-at-bottom)))

In order to see how this new, modified form works, delete any frame showing *foo*, type C-x 1 followed by C-x 2 in the window showing *scratch*, and subsequently evaluate that form. display-buffer should split the upper window, and show *foo* in the new window. Alternatively, if after C-x 2 you had typed C-x o, display-buffer would have split the window at the bottom instead.

Suppose now that, before evaluating the new form, you have made the selected window as small as possible, for example, by evaluating the form (fit-window-to-buffer) in that window. In that case, display-buffer would have failed to split the selected window and would have split the frame’s root window instead, effectively displaying *foo* at the bottom of the frame.

In either case, evaluating the new form a second time should reuse the window already showing *foo* since both functions supplied by the action argument try to reuse such a window first.

By setting the action argument, an application effectively overrules any customization of display-buffer-base-action. Our user can now either accept the choice of the application, or redouble by customizing the option display-buffer-alist as follows:

(setopt
 display-buffer-alist
 '(("\\*foo\\*"
    (display-buffer-reuse-window display-buffer-pop-up-frame))))

Trying this with the new, modified form above in a configuration that does not show *foo* anywhere, will display *foo* on a separate frame, completely ignoring the action argument of display-buffer.

Note that we didn’t care to specify a reusable-frames action alist entry in our specification of display-buffer-alist. display-buffer always takes the first one it finds—in our case the one specified by display-buffer-base-action. If we wanted to use a different specification, for example, to exclude iconified frames showing *foo* from the list of reusable ones, we would have to specify that separately, however:

(setopt
 display-buffer-alist
 '(("\\*foo\\*"
    (display-buffer-reuse-window display-buffer-pop-up-frame)
    (reusable-frames . visible))))

If you try this, you will notice that repeated attempts to display *foo* will succeed to reuse a frame only if that frame is visible.

The above example would allow the conclusion that users customize display-buffer-alist for the sole purpose to overrule the action argument chosen by applications. Such a conclusion would be incorrect. display-buffer-alist is the standard option for users to direct the course of display of specific buffers in a preferred way regardless of whether the display is also guided by an action argument.

We can, however, reasonably conclude that customizing display-buffer-alist differs from customizing display-buffer-base-action in two major aspects: it is stronger because it overrides the action argument of display-buffer, and it allows to explicitly specify the affected buffers. In fact, displaying other buffers is not affected in any way by a customization for *foo*. For example,

(display-buffer (get-buffer-create "*bar*"))

continues being governed by the settings of display-buffer-base-action and display-buffer-fallback-action only.

We could stop with our examples here but Lisp programs still have an ace up their sleeves which they can use to overrule any customization of display-buffer-alist. It’s the variable display-buffer-overriding-action which they can bind around display-buffer calls as follows:

(let ((display-buffer-overriding-action
       '((display-buffer-same-window))))
  (display-buffer
   (get-buffer-create "*foo*")
   '((display-buffer-below-selected display-buffer-at-bottom))))

Evaluating this form will usually display *foo* in the selected window regardless of the action argument and any user customizations. (Usually, an application will not bother to also provide an action argument. Here it just serves to illustrate the fact that it gets overridden.)

It might be illustrative to look at the list of action functions display-buffer would have tried to display *foo* with the customizations we provided here. The list (including comments explaining who added this and the subsequent elements) is:

(display-buffer-same-window  ;; `display-buffer-overriding-action'
 display-buffer-reuse-window ;; `display-buffer-alist'
 display-buffer-pop-up-frame
 display-buffer-below-selected ;; ACTION argument
 display-buffer-at-bottom
 display-buffer-reuse-window ;; `display-buffer-base-action'
 display-buffer-pop-up-frame
 display-buffer--maybe-same-window ;; `display-buffer-fallback-action'
 display-buffer-reuse-window
 display-buffer--maybe-pop-up-frame-or-window
 display-buffer-in-previous-window
 display-buffer-use-some-window
 display-buffer-pop-up-frame)

Note that among the internal functions listed here, display-buffer--maybe-same-window is effectively ignored while display-buffer--maybe-pop-up-frame-or-window actually runs display-buffer-pop-up-window.

The action alist passed in each function call is:

((reusable-frames . visible)
 (reusable-frames . 0))

which shows that we have used the second specification of display-buffer-alist above, overriding the specification supplied by display-buffer-base-action. Suppose our user had written that as

(setopt
 display-buffer-alist
 '(("\\*foo\\*"
    (display-buffer-reuse-window display-buffer-pop-up-frame)
    (inhibit-same-window . t)
    (reusable-frames . visible))))

In this case the inhibit-same-window alist entry will successfully invalidate the display-buffer-same-window specification from display-buffer-overriding-action and display-buffer will show *foo* on another frame. To make display-buffer-overriding-action more robust in this regard, the application would have to specify an appropriate inhibit-same-window entry too, for example, as follows:

(let ((display-buffer-overriding-action
       '(display-buffer-same-window (inhibit-same-window . nil))))
  (display-buffer (get-buffer-create "*foo*")))

This last example shows that while the precedence order of action functions is fixed, as described in Choosing a Window for Displaying a Buffer, an action alist entry specified by a display action ranked lower in that order can affect the execution of a higher ranked display action.


29.13.6 The Zen of Buffer Display

In its most simplistic form, a frame accommodates always one single window that can be used for displaying a buffer. As a consequence, it is always the latest call of display-buffer that will have succeeded in placing its buffer there.

Since working with such a frame is not very practical, Emacs by default allows for more complex frame layouts controlled by the default values of the frame size and the split-height-threshold and split-width-threshold options. Displaying a buffer not yet shown on a frame then either splits the single window on that frame or (re-)uses one of its two windows.

The default behavior is abandoned as soon as the user customizes one of these thresholds or manually changes the frame’s layout. The default behavior is also abandoned when calling display-buffer with a non-nil action argument or the user customizes one of the options mentioned in the previous subsections. Mastering display-buffer soon may become a frustrating experience due to the plethora of applicable display actions and the resulting frame layouts.

However, refraining from using buffer display functions and falling back on a split & delete windows metaphor is not a good idea either. Buffer display functions give Lisp programs and users a framework to reconcile their different needs; no comparable framework exists for splitting and deleting windows. Buffer display functions also allow to at least partially restore the layout of a frame when removing a buffer from it later (see Quitting Windows).

Below we will give a number of guidelines to redeem the frustration mentioned above and thus to avoid literally losing buffers in-between the windows of a frame.

Write display actions without stress

Writing display actions can be a pain because one has to lump together action functions and action alists in one huge list. (Historical reasons prevented us from having display-buffer support separate arguments for these.) It might help to memorize some basic forms like the ones listed below:

'(nil (inhibit-same-window . t))

specifies an action alist entry only and no action function. Its sole purpose is to inhibit a display-buffer-same-window function specified elsewhere from showing the buffer in the same window, see also the last example of the preceding subsection.

'(display-buffer-below-selected)

on the other hand, specifies one action function and an empty action alist. To combine the effects of the above two specifications one would write the form

'(display-buffer-below-selected (inhibit-same-window . t))

to add another action function one would write

'((display-buffer-below-selected display-buffer-at-bottom)
  (inhibit-same-window . t))

and to add another alist entry one would write

'((display-buffer-below-selected display-buffer-at-bottom)
  (inhibit-same-window . t)
  (window-height . fit-window-to-buffer))

That last form can be used as action argument of display-buffer in the following way:

(display-buffer
 (get-buffer-create "*foo*")
 '((display-buffer-below-selected display-buffer-at-bottom)
   (inhibit-same-window . t)
   (window-height . fit-window-to-buffer)))

In a customization of display-buffer-alist it would be used as follows:

(setopt
 display-buffer-alist
 '(("\\*foo\\*"
    (display-buffer-below-selected display-buffer-at-bottom)
    (inhibit-same-window . t)
    (window-height . fit-window-to-buffer))))

To add a customization for a second buffer one would then write:

(setopt
 display-buffer-alist
 '(("\\*foo\\*"
    (display-buffer-below-selected display-buffer-at-bottom)
    (inhibit-same-window . t)
    (window-height . fit-window-to-buffer))
   ("\\*bar\\*"
    (display-buffer-reuse-window display-buffer-pop-up-frame)
    (reusable-frames . visible))))
Treat each other with respect

display-buffer-alist and display-buffer-base-action are user options—Lisp programs must never set or rebind them. display-buffer-overriding-action, on the other hand, is reserved for applications—who seldom use that option and if they use it, then with utmost care.

Older implementations of display-buffer frequently caused users and applications to fight over the settings of user options like pop-up-frames and pop-up-windows (see Additional Options for Displaying Buffers). This was one major reason for redesigning display-buffer—to provide a clear framework specifying what users and applications should be allowed to do.

Lisp programs must be prepared that user customizations may cause buffers to get displayed in an unexpected way. They should never assume in their subsequent behavior, that the buffer has been shown precisely the way they asked for in the action argument of display-buffer.

Users should not pose too many and too severe restrictions on how arbitrary buffers get displayed. Otherwise, they will risk to lose the characteristics of showing a buffer for a certain purpose. Suppose a Lisp program has been written to compare different versions of a buffer in two windows side-by-side. If the customization of display-buffer-alist prescribes that any such buffer should be always shown in or below the selected window, the program will have a hard time to set up the desired window configuration via display-buffer.

To specify a preference for showing an arbitrary buffer, users should customize display-buffer-base-action. An example of how users who prefer working with multiple frames would do that was given in the previous subsection. display-buffer-alist should be reserved for displaying specific buffers in a specific way.

Consider reusing a window that already shows the buffer

Generally, it’s always a good idea for users and Lisp programmers to be prepared for the case that a window already shows the buffer in question and to reuse that window. In the preceding subsection we have shown that failing to do so properly may cause display-buffer to continuously pop up a new frame although a frame showing that buffer existed already. In a few cases only, it might be undesirable to reuse a window, for example, when a different portion of the buffer should be shown in that window.

Hence, display-buffer-reuse-window is one action function that should be used as often as possible, both in action arguments and customizations. An inhibit-same-window entry in the action argument usually takes care of the most common case where reusing a window showing the buffer should be avoided—that where the window in question is the selected one.

Attract focus to the window chosen

This is a no-brainer for people working with multiple frames—the frame showing the buffer will automatically raise and get focus unless an inhibit-switch-frame entry forbids it. For single frame users this task can be considerably more difficult. In particular, display-buffer-pop-up-window and display-buffer-use-some-window can become obtrusive in this regard. They split or use a seemingly arbitrary (often the largest or least recently used) window, distracting the user’s attention.

Some Lisp programs therefore try to choose a window at the bottom of the frame, for example, in order to display the buffer in vicinity of the minibuffer window where the user is expected to answer a question related to the new window. For non-input related actions display-buffer-below-selected might be preferable because the selected window usually already has the user’s attention.

Take care which window is selected

Many applications call display-buffer from within window excursions produced by with-selected-window or select-window calls with a non-nil norecord argument. This is almost always a bad idea because the window selected within such an excursion is usually not the window selected in the configuration presented to the user.

If, for example, a user had added an inhibit-same-window alist entry, that entry would have avoided the window selected within the scope of the excursion and not the window selected in the resulting configuration. Even if no such entry has been added, the resulting behavior might be strange. While in a frame containing one live window, evaluating the following form

(progn
  (split-window)
  (display-buffer "*Messages*"))

will display a window showing the *Messages* buffer at the bottom and leave the other window selected. Evaluating the next form

(with-selected-window (split-window)
  (display-buffer "*Messages*"))

will display *Messages* in a window on the top and select it which is usually not what display-buffer is supposed to do.

On the other hand, while evaluating the following form

(progn
  (split-window)
  (pop-to-buffer "*Messages*"))

will correctly select the *Messages* buffer, the next form

(progn
  (split-window)
  (with-selected-window (selected-window)
    (pop-to-buffer "*Messages*")))

will not.

Also, invocations of action functions like display-buffer-use-some-window and display-buffer-use-least-recent-window that expect the selected window to have the highest use time among all windows, may fail to produce a window according to their specifications.

Hence, an application that relies on using a window excursion should try to postpone the display-buffer call until after the excursion has terminated.


29.14 Window History

Each window remembers in a list the buffers it has previously displayed, and the order in which these buffers were removed from it. This history is used, for example, by replace-buffer-in-windows (see Buffers and Windows), and when quitting windows (see Quitting Windows). The list is automatically maintained by Emacs, but you can use the following functions to explicitly inspect or alter it:

Function: window-prev-buffers &optional window

This function returns a list specifying the previous contents of window. The optional argument window should be a live window and defaults to the selected one.

Each list element has the form (buffer window-start window-pos), where buffer is a buffer previously shown in the window, window-start is the window start position (see The Window Start and End Positions) when that buffer was last shown, and window-pos is the point position (see Windows and Point) when that buffer was last shown in window.

The list is ordered so that earlier elements correspond to more recently-shown buffers, and the first element usually corresponds to the buffer most recently removed from the window.

Function: set-window-prev-buffers window prev-buffers

This function sets window’s previous buffers to the value of prev-buffers. The argument window must be a live window and defaults to the selected one. The argument prev-buffers should be a list of the same form as that returned by window-prev-buffers.

In addition, each window maintains a list of next buffers, which is a list of buffers re-shown by switch-to-prev-buffer (see below). This list is mainly used by switch-to-prev-buffer and switch-to-next-buffer for choosing buffers to switch to.

Function: window-next-buffers &optional window

This function returns the list of buffers recently re-shown in window via switch-to-prev-buffer. The window argument must denote a live window or nil (meaning the selected window).

Function: set-window-next-buffers window next-buffers

This function sets the next buffer list of window to next-buffers. The window argument should be a live window or nil (meaning the selected window). The argument next-buffers should be a list of buffers.

The following commands can be used to cycle through the global buffer list, much like bury-buffer and unbury-buffer. However, they cycle according to the specified window’s history list, rather than the global buffer list. In addition, they restore window-specific window start and point positions, and may show a buffer even if it is already shown in another window. The switch-to-prev-buffer command, in particular, is used by replace-buffer-in-windows, bury-buffer and quit-window to find a replacement buffer for a window.

Command: switch-to-prev-buffer &optional window bury-or-kill

This command displays the previous buffer in window. The argument window should be a live window or nil (meaning the selected window). If the optional argument bury-or-kill is non-nil, this means that the buffer currently shown in window is about to be buried or killed and consequently should not be switched to in future invocations of this command.

The previous buffer is usually the buffer shown before the buffer currently shown in window. However, a buffer that has been buried or killed, or has been already shown by a recent invocation of switch-to-prev-buffer, does not qualify as previous buffer.

If repeated invocations of this command have already shown all buffers previously shown in window, further invocations will show buffers from the buffer list of the frame window appears on (see The Buffer List).

The option switch-to-prev-buffer-skip described below can be used to inhibit switching to certain buffers, for example, to those already shown in another window. Also, if window’s frame has a buffer-predicate parameter (see Buffer Parameters), that predicate may inhibit switching to certain buffers.

Command: switch-to-next-buffer &optional window

This command switches to the next buffer in window, thus undoing the effect of the last switch-to-prev-buffer command in window. The argument window must be a live window and defaults to the selected one.

If there is no recent invocation of switch-to-prev-buffer that can be undone, this function tries to show a buffer from the buffer list of the frame window appears on (see The Buffer List).

The option switch-to-prev-buffer-skip and the buffer-predicate (see Buffer Parameters) of window’s frame affect this command as they do for switch-to-prev-buffer.

By default switch-to-prev-buffer and switch-to-next-buffer can switch to a buffer that is already shown in another window. The following option can be used to override this behavior.

User Option: switch-to-prev-buffer-skip

If this variable is nil, switch-to-prev-buffer may switch to any buffer, including those already shown in other windows.

If this variable is non-nil, switch-to-prev-buffer will refrain from switching to certain buffers. The following values can be used:

  • this means do not switch to a buffer shown on the frame that hosts the window switch-to-prev-buffer is acting upon.
  • visible means do not switch to a buffer shown on any visible frame.
  • 0 (the number zero) means do not switch to a buffer shown on any visible or iconified frame.
  • t means do not switch to a buffer shown on any live frame.
  • A function that takes three arguments—the window argument of switch-to-prev-buffer, a buffer switch-to-prev-buffer intends to switch to and the bury-or-kill argument of switch-to-prev-buffer. If that function returns non-nil, switch-to-prev-buffer will refrain from switching to the buffer specified by the second argument.

The command switch-to-next-buffer obeys this option in a similar way. If this option specifies a function, switch-to-next-buffer will call that function with the third argument always nil.

Note that since switch-to-prev-buffer is called by bury-buffer, replace-buffer-in-windows and quit-restore-window as well, customizing this option may also affect the behavior of Emacs when a window is quit or a buffer gets buried or killed.

Note also that under certain circumstances switch-to-prev-buffer and switch-to-next-buffer may ignore this option, for example, when there is only one buffer left these functions can switch to.

User Option: switch-to-prev-buffer-skip-regexp

This user option should be either a regular expression or a list of regular expressions. Buffers whose names match one of those regular expressions will be ignored by switch-to-prev-buffer and switch-to-next-buffer (except when there’s no other buffer to switch to).


29.15 Dedicated Windows

Functions for displaying a buffer can be told to not use specific windows by marking these windows as dedicated to their buffers. display-buffer (see Choosing a Window for Displaying a Buffer) never uses a dedicated window for displaying another buffer in it. get-lru-window and get-largest-window (see Cyclic Ordering of Windows) do not consider dedicated windows as candidates when their dedicated argument is non-nil. The behavior of set-window-buffer (see Buffers and Windows) with respect to dedicated windows is slightly different, see below.

Functions supposed to remove a buffer from a window or a window from a frame can behave specially when a window they operate on is dedicated. We will distinguish four basic cases, namely where (1) the window is not the only window on its frame, (2) the window is the only window on its frame but there are other frames on the same terminal left, (3) the window is the only window on the only frame on the same terminal, and (4) the dedication’s value is side (see Displaying Buffers in Side Windows).

In particular, delete-windows-on (see Deleting Windows) handles case (2) by deleting the associated frame and cases (3) and (4) by showing another buffer in that frame’s only window. The function replace-buffer-in-windows (see Buffers and Windows) which is called when a buffer gets killed, deletes the window in case (1) and behaves like delete-windows-on otherwise.

When bury-buffer (see The Buffer List) operates on the selected window (which shows the buffer that shall be buried), it handles case (2) by calling frame-auto-hide-function (see Quitting Windows) to deal with the selected frame. The other two cases are handled as with replace-buffer-in-windows.

Function: window-dedicated-p &optional window

This function returns non-nil if window is dedicated to its buffer and nil otherwise. More precisely, the return value is the value assigned by the last call of set-window-dedicated-p for window, or nil if that function was never called with window as its argument. The default for window is the selected window.

Function: set-window-dedicated-p window flag

This function marks window as dedicated to its buffer if flag is non-nil, and non-dedicated otherwise.

As a special case, if flag is t, window becomes strongly dedicated to its buffer. set-window-buffer signals an error when the window it acts upon is strongly dedicated to its buffer and does not already display the buffer it is asked to display. Other functions do not treat t differently from any non-nil value.

You can also tell display-buffer to mark a window it creates as dedicated to its buffer by providing a suitable dedicated action alist entry (see Action Alists for Buffer Display).


29.16 Quitting Windows

After a command uses display-buffer to put a buffer on the screen, the user may decide to hide it and return to the previous configuration of the Emacs display. We call that quitting the window. The way to do this is to call quit-window while the window used by display-buffer is the selected window.

The right way to restore the previous configuration of the display depends on what was done to the window where the buffer now appears. It might be right to delete that window, or delete its frame, or just display another buffer in that window. One complication is that the user may have changed the window configuration since the act of displaying that buffer, and it would be undesirable to undo the user’s explicitly requested changes.

To enable quit-window to do the right thing, display-buffer saves information about what it did in the window’s quit-restore parameter (see Window Parameters).

Command: quit-window &optional kill window

This command quits window and buries its buffer. The argument window must be a live window and defaults to the selected one. With prefix argument kill non-nil, it kills the buffer instead of burying it.

The function quit-window first runs quit-window-hook. Then it calls the function quit-restore-window, described next, which does the hard work.

You can get more control by calling quit-restore-window instead.

Function: quit-restore-window &optional window bury-or-kill

This function handles window and its buffer after quitting. The optional argument window must be a live window and defaults to the selected one. The function takes account of the window’s quit-restore parameter.

The optional argument bury-or-kill specifies how to deal with window’s buffer. The following values are meaningful:

nil

This means to not deal with the buffer in any particular way. As a consequence, if window is not deleted, invoking switch-to-prev-buffer will usually show the buffer again.

append

This means that if window is not deleted, its buffer is moved to the end of window’s list of previous buffers (see Window History), so it’s less likely that future invocations of switch-to-prev-buffer will switch to it. Also, it moves the buffer to the end of the frame’s buffer list (see The Buffer List).

bury

This means that if window is not deleted, its buffer is removed from window’s list of previous buffers. Also, it moves the buffer to the end of the frame’s buffer list. This is the most reliable way to prevent switch-to-prev-buffer from switching to this buffer again, short of killing the buffer.

kill

This means to kill window’s buffer.

The argument bury-or-kill also specifies what to do with window’s frame when window should be deleted, if it is the only window on its frame, and there are other frames on that frame’s terminal. If bury-or-kill equals kill, it means to delete the frame. Otherwise, the fate of the frame is determined by calling frame-auto-hide-function (see below) with that frame as sole argument.

This function always sets window’s quit-restore parameter to nil unless it deletes the window.

The window window’s quit-restore parameter (see Window Parameters) should be nil or a list of four elements:

(method obuffer owindow this-buffer)

The first element, method, is one of the four symbols window, frame, same and other. frame and window control how to delete window, while same and other control displaying some other buffer in it.

Specifically, window means that the window has been specially created by display-buffer; frame means that a separate frame has been created; same, that the window has only ever displayed this buffer; other, that the window showed another buffer before.

The second element, obuffer, is either one of the symbols window or frame, or a list of the form

(prev-buffer prev-window-start prev-window-point height)

which says which buffer was shown in window before, that buffer’s window start (see The Window Start and End Positions) and window point (see Windows and Point) positions at that time, and window’s height at that time. If prev-buffer is still live when quitting window, quitting the window may reuse window to display prev-buffer.

The third element, owindow, is the window that was selected just before the displaying was done. If quitting deletes window, it tries to select owindow.

The fourth element, this-buffer, is the buffer whose displaying set the quit-restore parameter. Quitting window may delete that window only if it still shows that buffer.

Quitting window tries to delete it if and only if (1) method is either window or frame, (2) the window has no history of previously-displayed buffers and (3) this-buffer equals the buffer currently displayed in window. If window is part of an atomic window (see Atomic Windows), quitting will try to delete the root of that atomic window instead. In either case, it tries to avoid signaling an error when window cannot be deleted.

If obuffer is a list, and prev-buffer is still live, quitting displays prev-buffer in window according to the rest of the elements of obuffer. This includes resizing the window to height if it was temporarily resized to display this-buffer.

Otherwise, if window was previously used for displaying other buffers (see Window History), the most recent buffer in that history will be displayed.

The following option specifies a function to do the right thing with a frame containing one window when quitting that window.

User Option: frame-auto-hide-function

The function specified by this option is called to automatically hide frames. This function is called with one argument—a frame.

The function specified here is called by bury-buffer (see The Buffer List) when the selected window is dedicated and shows the buffer to bury. It is also called by quit-restore-window (see above) when the frame of the window to quit has been specially created for displaying that window’s buffer and the buffer is not killed.

The default is to call iconify-frame (see Visibility of Frames). Alternatively, you may specify either delete-frame (see Deleting Frames) to remove the frame from its display, make-frame-invisible to make the frame invisible, ignore to leave the frame unchanged, or any other function that can take a frame as its sole argument.

Note that the function specified by this option is called only if the specified frame contains just one live window and there is at least one other frame on the same terminal.

For a particular frame, the value specified here may be overridden by that frame’s auto-hide-function frame parameter (see Frame Interaction Parameters).


29.17 Side Windows

Side windows are special windows positioned at any of the four sides of a frame’s root window (see Windows and Frames). In practice, this means that the area of the frame’s root window is subdivided into a main window and a number of side windows surrounding that main window. The main window is either a “normal” live window or specifies the area containing all the normal windows.

In their most simple form of use, side windows allow to display specific buffers always in the same area of a frame. Hence they can be regarded as a generalization of the concept provided by display-buffer-at-bottom (see Action Functions for Buffer Display) to the remaining sides of a frame. With suitable customizations, however, side windows can be also used to provide frame layouts similar to those found in so-called integrated development environments (IDEs).


29.17.1 Displaying Buffers in Side Windows

The following action function for display-buffer (see Action Functions for Buffer Display) creates or reuses a side window for displaying the specified buffer.

Function: display-buffer-in-side-window buffer alist

This function displays buffer in a side window of the selected frame. It returns the window used for displaying buffer, nil if no such window can be found or created.

alist is an association list of symbols and values as for display-buffer. The following symbols in alist are special for this function:

side

Denotes the side of the frame where the window shall be located. Valid values are left, top, right and bottom. If unspecified, the window is located at the bottom of the frame.

slot

Denotes a slot at the specified side where to locate the window. A value of zero means to preferably position the window in the middle of the specified side. A negative value means to use a slot preceding (that is, above or on the left of) the middle slot. A positive value means to use a slot following (that is, below or on the right of) the middle slot. Hence, all windows on a specific side are ordered by their slot value. If unspecified, the window is located in the middle of the specified side.

dedicated

The dedicated flag (see Dedicated Windows) has a slightly different meaning for side windows. When a side window is created, that flag is set to the value side to prevent display-buffer to use the window in other action functions. Its value persists across invocations of quit-window, kill-buffer, previous-buffer and next-buffer.

In particular, these commands will refrain from showing, in a side window, buffers that have not been displayed in that window before. They will also refrain from having a normal, non-side window show a buffer that has been already displayed in a side window. A notable exception to the latter rule occurs when an application, after displaying a buffer, resets that buffer’s local variables. To override these rules and always delete a side window with quit-window or kill-buffer, and eventually prevent the use of previous-buffer and next-buffer, set this value to t or specify a value via display-buffer-mark-dedicated.

If you specify the same slot on the same side for two or more different buffers, the buffer displayed last is shown in the corresponding window. Hence, slots can be used for sharing the same side window between buffers.

This function installs the window-side and window-slot parameters (see Window Parameters) and makes them persistent. It does not install any other window parameters unless they have been explicitly provided via a window-parameters entry in alist.

By default, side windows cannot be split via split-window (see Splitting Windows). Also, a side window is not reused or split by any buffer display action (see Action Functions for Buffer Display) unless it is explicitly specified as target of that action. Note also that delete-other-windows cannot make a side window the only window on its frame (see Deleting Windows).


29.17.2 Side Window Options and Functions

The following options provide additional control over the placement of side windows.

User Option: window-sides-vertical

If non-nil, the side windows on the left and right of a frame occupy the frame’s full height. Otherwise, the side windows on the top and bottom of the frame occupy the frame’s full width.

User Option: window-sides-slots

This option specifies the maximum number of side windows on each side of a frame. The value is a list of four elements specifying the number of side window slots on (in this order) the left, top, right and bottom of each frame. If an element is a number, it means to display at most that many windows on the corresponding side. If an element is nil, it means there’s no bound on the number of slots on that side.

If any of the specified values is zero, no window can be created on the corresponding side. display-buffer-in-side-window will not signal an error in that case, but will return nil. If a specified value just forbids the creation of an additional side window, the most suitable window on that side is reused and may have its window-slot parameter changed accordingly.

User Option: window-sides-reversed

This option specifies whether top/bottom side windows should appear in reverse order. When this is nil, side windows on the top and bottom of a frame are always drawn from left to right with increasing slot values. When this is t, the drawing order is reversed and side windows on the top and bottom of a frame are drawn from right to left with increasing slot values.

When this is bidi, the drawing order is reversed if and only if the value of bidi-paragraph-direction (see Bidirectional Display) is right-to-left in the buffer displayed in the window most recently selected within the main window area of this frame. Sometimes that window may be hard to find, so heuristics are used to avoid that the drawing order changes inadvertently when another window gets selected.

The layout of side windows on the left or right of a frame is not affected by the value of this variable.

When a frame has side windows, the following function returns the main window of that frame.

Function: window-main-window &optional frame

This function returns the main window of the specified frame. The optional argument frame must be a live frame and defaults to the selected one.

If frame has no side windows, it returns frame’s root window. Otherwise, it returns either an internal non-side window such that all other non-side windows on frame descend from it, or the single live non-side window of frame. Note that the main window of a frame cannot be deleted via delete-window.

The following command is handy to toggle the appearance of all side windows on a specified frame.

Command: window-toggle-side-windows &optional frame

This command toggles side windows on the specified frame. The optional argument frame must be a live frame and defaults to the selected one.

If frame has at least one side window, this command saves the state of frame’s root window in the frame’s window-state frame parameter and deletes all side windows on frame afterwards.

If frame has no side windows, but does have a window-state parameter, this command uses that parameter’s value to restore the side windows on frame leaving frame’s main window alone.

An error is signaled if frame has no side windows and no saved state is found for it.


29.17.3 Frame Layouts with Side Windows

Side windows can be used to create more complex frame layouts like those provided by integrated development environments (IDEs). In such layouts, the area of the main window is where the normal editing activities take place. Side windows are not conceived for editing in the usual sense. Rather, they are supposed to display information complementary to the current editing activity, like lists of files, tags or buffers, help information, search or grep results or shell output.

The layout of such a frame might appear as follows:

     ___________________________________
    |          *Buffer List*            |
    |___________________________________|
    |     |                       |     |
    |  *  |                       |  *  |
    |  d  |                       |  T  |
    |  i  |                       |  a  |
    |  r  |   Main Window Area    |  g  |
    |  e  |                       |  s  |
    |  d  |                       |  *  |
    |  *  |                       |     |
    |_____|_______________________|_____|
    | *help*/*grep*/  |  *shell*/       |
    | *Completions*   |  *compilation*  |
    |_________________|_________________|
    |             Echo Area             |
    |___________________________________|


The following example illustrates how window parameters (see Window Parameters) can be used with display-buffer-in-side-window (see Displaying Buffers in Side Windows) to set up code for producing the frame layout sketched above.

(defvar parameters
  '(window-parameters . ((no-other-window . t)
                         (no-delete-other-windows . t))))

(setq fit-window-to-buffer-horizontally t)
(setq window-resize-pixelwise t)

(setq
 display-buffer-alist
 `(("\\*Buffer List\\*" display-buffer-in-side-window
    (side . top) (slot . 0) (window-height . fit-window-to-buffer)
    (preserve-size . (nil . t)) ,parameters)
   ("\\*Tags List\\*" display-buffer-in-side-window
    (side . right) (slot . 0) (window-width . fit-window-to-buffer)
    (preserve-size . (t . nil)) ,parameters)
   ("\\*\\(?:help\\|grep\\|Completions\\)\\*"
    display-buffer-in-side-window
    (side . bottom) (slot . -1) (preserve-size . (nil . t))
    ,parameters)
   ("\\*\\(?:shell\\|compilation\\)\\*" display-buffer-in-side-window
    (side . bottom) (slot . 1) (preserve-size . (nil . t))
    ,parameters)))

This specifies display-buffer-alist entries (see Choosing a Window for Displaying a Buffer) for buffers with fixed names. In particular, it asks for showing *Buffer List* with adjustable height at the top of the frame and *Tags List* with adjustable width on the frame’s right. It also asks for having the *help*, *grep* and *Completions* buffers share a window on the bottom left side of the frame and the *shell* and *compilation* buffers appear in a window on the bottom right side of the frame.

Note that the option fit-window-to-buffer-horizontally must have a non-nil value in order to allow horizontal adjustment of windows. Entries are also added that ask for preserving the height of side windows at the top and bottom of the frame and the width of side windows at the left or right of the frame. To assure that side windows retain their respective sizes when maximizing the frame, the variable window-resize-pixelwise is set to a non-nil value. See Resizing Windows.

The last form also makes sure that none of the created side windows are accessible via C-x o by installing the no-other-window parameter for each of these windows. In addition, it makes sure that side windows are not deleted via C-x 1 by installing the no-delete-other-windows parameter for each of these windows.

Since dired buffers have no fixed names, we use a special function dired-default-directory-on-left in order to display a lean directory buffer on the left side of the frame.

(defun dired-default-directory-on-left ()
  "Display `default-directory' in side window on left, hiding details."
  (interactive)
  (let ((buffer (dired-noselect default-directory)))
    (with-current-buffer buffer (dired-hide-details-mode t))
    (display-buffer-in-side-window
     buffer `((side . left) (slot . 0)
              (window-width . fit-window-to-buffer)
              (preserve-size . (t . nil)) ,parameters))))

Evaluating the preceding forms and typing, in any order, M-x list-buffers, C-h f, M-x shell, M-x list-tags, and M-x dired-default-directory-on-left should now reproduce the frame layout sketched above.


29.18 Atomic Windows

Atomic windows are rectangular compositions of at least two live windows. They have the following distinctive characteristics:

  • The function split-window (see Splitting Windows), when applied to a constituent of an atomic window, will try to create the new window outside of the atomic window.
  • The function delete-window (see Deleting Windows), when applied to a constituent of an atomic window, will try to delete the entire atomic window instead.
  • The function delete-other-windows (see Deleting Windows), when applied to a constituent of an atomic window, will try to make the atomic window fill its frame or main window (see Side Windows).

This means that the basic groups of functions that alter the window structure treat an atomic window like a live one, thus preserving the internal structure of the atomic window.

Atomic windows are useful to construct and preserve window layouts that are meaningful only when all involved buffers are shown simultaneously in a specific manner, such as when showing differences between file revisions, or the same text in different languages or markups. They can also be used to permanently display information pertinent to a specific window in bars on that window’s sides.

Atomic windows are implemented with the help of the reserved window-atom window parameter (see Window Parameters) and an internal window (see Basic Concepts of Emacs Windows) called the root window of the atomic window. All windows that are part of the same atomic window have this root window as their common ancestor and are assigned a non-nil window-atom parameter.

The following function returns the root of the atomic window a specified window is part of:

Function: window-atom-root &optional window

This functions returns the root of the atomic window window is a part of. The specified window must be a valid window and defaults to the selected one. It returns nil if window is not part of an atomic window.

The most simple approach to make a new atomic window is to take an existing internal window and apply the following function:

Function: window-make-atom window

This function converts window into an atomic window. The specified window must be an internal window. All this function does is to set the window-atom parameter of each descendant of window to t.

To create a new atomic window from an existing live window or to add a new window to an existing atomic window, the following buffer display action function (see Action Functions for Buffer Display) can be used:

Function: display-buffer-in-atom-window buffer alist

This function tries to display buffer in a new window that will be combined with an existing window to form an atomic window. If the existing window is already part of an atomic window, it adds the new window to that atomic window.

The specified alist is an association list of symbols and values. The following symbols have a special meaning:

window

The value of such an element specifies an existing window the new window shall be combined with. If it specifies an internal window, all children of that window become part of the atomic window too. If no window is specified, the new window becomes a sibling of the selected window. The window-atom parameter of the existing window is set to main provided that window is live and its window-atom parameter was not already set.

side

The value of such an element denotes the side of the existing window where the new window shall be located. Valid values are below, right, above and left. The default is below. The window-atom parameter of the new window is set to this value.

The return value is the new window, nil when creating that window failed.

Note that the value of the window-atom parameter does not really matter as long as it is non-nil. The values assigned by display-buffer-in-atom-window just allow for easy retrieval of the original and the new window after that function has been applied. Note also that the window-atom parameter is the only window parameter assigned by display-buffer-in-atom-window. Further parameters have to be set by the application explicitly via a window-parameters entry in alist.

Atomic windows automatically cease to exist when one of their constituents gets deleted. To dissolve an atomic window manually, reset the window-atom parameter of its constituents—the root of the atomic window and all its descendants.

The following code snippet, when applied to a single-window frame, first splits the selected window and makes the selected and the new window constituents of an atomic window with their parent as root. It then displays the buffer *Messages* in a new window at the frame’s bottom and makes that new window part of the atomic window just created.

(let ((window (split-window-right)))
  (window-make-atom (window-parent window))
  (display-buffer-in-atom-window
   (get-buffer-create "*Messages*")
   `((window . ,(window-parent window)) (window-height . 5))))

At this moment typing C-x 2 in any window of that frame produces a new window at the bottom of the frame. Typing C-x 3 instead will put the new window at the frame’s right. In either case, typing now C-x 1 in any window of the atomic window will remove the new window only. Typing C-x 0 in any window of the atomic window will make that new window fill the frame.


29.19 Windows and Point

Each window has its own value of point (see Point), independent of the value of point in other windows displaying the same buffer. This makes it useful to have multiple windows showing one buffer.

  • The window point is established when a window is first created; it is initialized from the buffer’s point, or from the window point of another window opened on the buffer if such a window exists.
  • Selecting a window sets the value of point in its buffer from the window’s value of point. Conversely, deselecting a window sets the window’s value of point from that of the buffer. Thus, when you switch between windows that display a given buffer, the point value for the selected window is in effect in the buffer, while the point values for the other windows are stored in those windows.
  • As long as the selected window displays the current buffer, the window’s point and the buffer’s point always move together; they remain equal.

Emacs displays the cursor, by default as a rectangular block, in each window at the position of that window’s point. When the user switches to another buffer in a window, Emacs moves that window’s cursor to where point is in that buffer. If the exact position of point is hidden behind some display element, such as a display string or an image, Emacs displays the cursor immediately before or after that display element.

Function: window-point &optional window

This function returns the current position of point in window. For a nonselected window, this is the value point would have (in that window’s buffer) if that window were selected. The default for window is the selected window.

When window is the selected window, the value returned is the value of point in that window’s buffer. Strictly speaking, it would be more correct to return the top-level value of point, outside of any save-excursion forms. But that value is hard to find.

Function: set-window-point window position

This function positions point in window at position position in window’s buffer. It returns position.

If window is selected, this simply does goto-char in window’s buffer.

Variable: window-point-insertion-type

This variable specifies the marker insertion type (see Marker Insertion Types) of window-point. The default is nil, so window-point will stay behind text inserted there.


29.20 The Window Start and End Positions

Each window maintains a marker used to keep track of a buffer position that specifies where in the buffer display should start. This position is called the display-start position of the window (or just the start). The character after this position is the one that appears at the upper left corner of the window. It is usually, but not inevitably, at the beginning of a text line.

After switching windows or buffers, and in some other cases, if the window start is in the middle of a line, Emacs adjusts the window start to the start of a line. This prevents certain operations from leaving the window start at a meaningless point within a line. This feature may interfere with testing some Lisp code by executing it using the commands of Lisp mode, because they trigger this readjustment. To test such code, put it into a command and bind the command to a key.

Function: window-start &optional window

This function returns the display-start position of window window. If window is nil, the selected window is used.

When you create a window, or display a different buffer in it, the display-start position is set to a display-start position recently used for the same buffer, or to point-min if the buffer doesn’t have any.

Redisplay updates the window-start position (if you have not specified it explicitly since the previous redisplay)—to make sure point appears on the screen. Nothing except redisplay automatically changes the window-start position; if you move point, do not expect the window-start position to change in response until after the next redisplay.

Function: window-group-start &optional window

This function is like window-start, except that when window is a part of a group of windows (see Window Group), window-group-start returns the start position of the entire group. This condition holds when the buffer local variable window-group-start-function is set to a function. In this case, window-group-start calls the function with the single argument window, then returns its result.

Function: window-end &optional window update

This function returns the position where display of its buffer ends in window. The default for window is the selected window.

Simply changing the buffer text or moving point does not update the value that window-end returns. The value is updated only when Emacs redisplays and redisplay completes without being preempted.

If the last redisplay of window was preempted, and did not finish, Emacs does not know the position of the end of display in that window. In that case, this function returns nil.

If update is non-nil, window-end always returns an up-to-date value for where display ends, based on the current window-start value. If a previously saved value of that position is still valid, window-end returns that value; otherwise it computes the correct value by scanning the buffer text.

Even if update is non-nil, window-end does not attempt to scroll the display if point has moved off the screen, the way real redisplay would do. It does not alter the window-start value. In effect, it reports where the displayed text will end if scrolling is not required. Note that the position it returns might be only partially visible.

Function: window-group-end &optional window update

This function is like window-end, except that when window is a part of a group of windows (see Window Group), window-group-end returns the end position of the entire group. This condition holds when the buffer local variable window-group-end-function is set to a function. In this case, window-group-end calls the function with the two arguments window and update, then returns its result. The argument update has the same meaning as in window-end.

Function: set-window-start window position &optional noforce

This function sets the display-start position of window to position in window’s buffer. It returns position.

The display routines insist that the position of point be visible when a buffer is displayed. Normally, they select the display-start position according to their internal logic (and scroll the window if necessary) to make point visible. However, if you specify the start position with this function using nil for noforce, it means you want display to start at position even if that would put the location of point off the screen. If this does place point off screen, the display routines attempt to move point to the left margin on the middle line in the window.

For example, if point is 1 and you set the start of the window to 37, the start of the next line, point will be above the top of the window. The display routines will automatically move point if it is still 1 when redisplay occurs. Here is an example:

;; Here is what ‘foo’ looks like before executing
;;   the set-window-start expression.

---------- Buffer: foo ----------
∗This is the contents of buffer foo.
2
3
4
5
6
---------- Buffer: foo ----------

(set-window-start
 (selected-window)
 (save-excursion
   (goto-char 1)
   (forward-line 1)
   (point)))
⇒ 37

;; Here is what ‘foo’ looks like after executing
;;   the set-window-start expression.
---------- Buffer: foo ----------
2
3
∗4
5
6
---------- Buffer: foo ----------

If the attempt to make point visible (i.e., in a fully-visible screen line) fails, the display routines will disregard the requested window-start position and compute a new one anyway. Thus, for reliable results Lisp programs that call this function should always move point to be inside the window whose display starts at position.

If noforce is non-nil, and position would place point off screen at the next redisplay, then redisplay computes a new window-start position that works well with point, and thus position is not used.

Function: set-window-group-start window position &optional noforce

This function is like set-window-start, except that when window is a part of a group of windows (see Window Group), set-window-group-start sets the start position of the entire group. This condition holds when the buffer local variable set-window-group-start-function is set to a function. In this case, set-window-group-start calls the function with the three arguments window, position, and noforce, then returns its result. The arguments position and noforce in this function have the same meaning as in set-window-start.

Function: pos-visible-in-window-p &optional position window partially

This function returns non-nil if position is within the range of text currently visible on the screen in window. It returns nil if position is scrolled vertically out of view. Locations that are partially obscured are not considered visible unless partially is non-nil. The argument position defaults to the current position of point in window; window defaults to the selected window. If position is t, that means to check either the first visible position of the last screen line in window, or the end-of-buffer position, whichever comes first.

This function considers only vertical scrolling. If position is out of view only because window has been scrolled horizontally, pos-visible-in-window-p returns non-nil anyway. See Horizontal Scrolling.

If position is visible, pos-visible-in-window-p returns t if partially is nil; if partially is non-nil, and the character following position is fully visible, it returns a list of the form (x y), where x and y are the pixel coordinates relative to the top left corner of the window; otherwise it returns an extended list of the form (x y rtop rbot rowh vpos), where rtop and rbot specify the number of off-window pixels at the top and bottom of the row at position, rowh specifies the visible height of that row, and vpos specifies the vertical position (zero-based row number) of that row.

Here is an example:

;; If point is off the screen now, recenter it now.
(or (pos-visible-in-window-p
     (point) (selected-window))
    (recenter 0))
Function: pos-visible-in-window-group-p &optional position window partially

This function is like pos-visible-in-window-p, except that when window is a part of a group of windows (see Window Group), pos-visible-in-window-group-p tests the visibility of pos in the entire group, not just in the single window. This condition holds when the buffer local variable pos-visible-in-window-group-p-function is set to a function. In this case pos-visible-in-window-group-p calls the function with the three arguments position, window, and partially, then returns its result. The arguments position and partially have the same meaning as in pos-visible-in-window-p.

Function: window-line-height &optional line window

This function returns the height of text line line in window. If line is one of header-line or mode-line, window-line-height returns information about the corresponding line of the window. Otherwise, line is a text line number starting from 0. A negative number counts from the end of the window. The default for line is the current line in window; the default for window is the selected window.

If the display is not up to date, window-line-height returns nil. In that case, pos-visible-in-window-p may be used to obtain related information.

If there is no line corresponding to the specified line, window-line-height returns nil. Otherwise, it returns a list (height vpos ypos offbot), where height is the height in pixels of the visible part of the line, vpos and ypos are the vertical position in lines and pixels of the line relative to the top of the first text line, and offbot is the number of off-window pixels at the bottom of the text line. If there are off-window pixels at the top of the (first) text line, ypos is negative.


29.21 Textual Scrolling

Textual scrolling means moving the text up or down through a window. It works by changing the window’s display-start location. It may also change the value of window-point to keep point on the screen (see Windows and Point).

The basic textual scrolling functions are scroll-up (which scrolls forward) and scroll-down (which scrolls backward). In these function names, “up” and “down” refer to the direction of motion of the buffer text relative to the window. Imagine that the text is written on a long roll of paper and that the scrolling commands move the paper up and down. Thus, if you are looking at the middle of a buffer and repeatedly call scroll-down, you will eventually see the beginning of the buffer.

Unfortunately, this sometimes causes confusion, because some people tend to think in terms of the opposite convention: they imagine the window moving over text that remains in place, so that “down” commands take you to the end of the buffer. This convention is consistent with fact that such a command is bound to a key named PageDown on modern keyboards.

Textual scrolling functions (aside from scroll-other-window) have unpredictable results if the current buffer is not the one displayed in the selected window. See The Current Buffer.

If the window contains a row taller than the height of the window (for example in the presence of a large image), the scroll functions will adjust the window’s vertical scroll position to scroll the partially visible row. Lisp callers can disable this feature by binding the variable auto-window-vscroll to nil (see Vertical Fractional Scrolling).

Command: scroll-up &optional count

This function scrolls forward by count lines in the selected window.

If count is negative, it scrolls backward instead. If count is nil (or omitted), the distance scrolled is next-screen-context-lines lines less than the height of the window’s body.

If the selected window cannot be scrolled any further, this function signals an error. Otherwise, it returns nil.

Command: scroll-down &optional count

This function scrolls backward by count lines in the selected window.

If count is negative, it scrolls forward instead. In other respects, it behaves the same way as scroll-up does.

Command: scroll-up-command &optional count

This behaves like scroll-up, except that if the selected window cannot be scrolled any further and the value of the variable scroll-error-top-bottom is t, it tries to move to the end of the buffer instead. If point is already there, it signals an error.

Command: scroll-down-command &optional count

This behaves like scroll-down, except that if the selected window cannot be scrolled any further and the value of the variable scroll-error-top-bottom is t, it tries to move to the beginning of the buffer instead. If point is already there, it signals an error.

Command: scroll-other-window &optional count

This function scrolls the text in another window upward count lines. Negative values of count, or nil, are handled as in scroll-up.

You can specify which buffer to scroll by setting the variable other-window-scroll-buffer to a buffer. If that buffer isn’t already displayed, scroll-other-window displays it in some window.

When the selected window is the minibuffer, the next window is normally the leftmost one immediately above it. You can specify a different window to scroll, when the minibuffer is selected, by setting the variable minibuffer-scroll-window. This variable has no effect when any other window is selected. When it is non-nil and the minibuffer is selected, it takes precedence over other-window-scroll-buffer. See Definition of minibuffer-scroll-window.

Command: scroll-other-window-down &optional count

This function scrolls the text in another window downward count lines. Negative values of count, or nil, are handled as in scroll-down. In other respects, it behaves the same way as scroll-other-window does.

Variable: other-window-scroll-buffer

If this variable is non-nil, it tells scroll-other-window which buffer’s window to scroll.

User Option: scroll-margin

This option specifies the size of the scroll margin—a minimum number of lines between point and the top or bottom of a window. Whenever point gets within this many lines of the top or bottom of the window, redisplay scrolls the text automatically (if possible) to move point out of the margin, closer to the center of the window.

User Option: maximum-scroll-margin

This variable limits the effective value of scroll-margin to a fraction of the current window line height. For example, if the current window has 20 lines and maximum-scroll-margin is 0.1, then the scroll margins will never be larger than 2 lines, no matter how big scroll-margin is.

maximum-scroll-margin itself has a maximum value of 0.5, which allows setting margins large to keep the cursor at the middle line of the window (or two middle lines if the window has an even number of lines). If it’s set to a larger value (or any value other than a float between 0.0 and 0.5) then the default value of 0.25 will be used instead.

User Option: scroll-conservatively

This variable controls how scrolling is done automatically when point moves off the screen (or into the scroll margin). If the value is a positive integer n, then redisplay scrolls the text up to n lines in either direction, if that will bring point back into proper view. This behavior is called conservative scrolling. Otherwise, scrolling happens in the usual way, under the control of other variables such as scroll-up-aggressively and scroll-down-aggressively.

The default value is zero, which means that conservative scrolling never happens.

User Option: scroll-down-aggressively

The value of this variable should be either nil or a fraction f between 0 and 1. If it is a fraction, that specifies where on the screen to put point when scrolling down. More precisely, when a window scrolls down because point is above the window start, the new start position is chosen to put point f part of the window height from the top. The larger f, the more aggressive the scrolling.

A value of nil is equivalent to .5, since its effect is to center point. This variable automatically becomes buffer-local when set in any fashion.

User Option: scroll-up-aggressively

Likewise, for scrolling up. The value, f, specifies how far point should be placed from the bottom of the window; thus, as with scroll-down-aggressively, a larger value scrolls more aggressively.

User Option: scroll-step

This variable is an older variant of scroll-conservatively. The difference is that if its value is n, that permits scrolling only by precisely n lines, not a smaller number. This feature does not work with scroll-margin. The default value is zero.

User Option: scroll-preserve-screen-position

If this option is t, whenever a scrolling command moves point off-window, Emacs tries to adjust point to keep the cursor at its old vertical position in the window, rather than the window edge.

If the value is non-nil and not t, Emacs adjusts point to keep the cursor at the same vertical position, even if the scrolling command didn’t move point off-window.

This option affects all scroll commands that have a non-nil scroll-command symbol property.

User Option: next-screen-context-lines

The value of this variable is the number of lines of continuity to retain when scrolling by full screens. For example, scroll-up with an argument of nil scrolls so that this many lines at the bottom of the window appear instead at the top. The default value is 2.

User Option: scroll-error-top-bottom

If this option is nil (the default), scroll-up-command and scroll-down-command simply signal an error when no more scrolling is possible.

If the value is t, these commands instead move point to the beginning or end of the buffer (depending on scrolling direction); only if point is already on that position do they signal an error.

Command: recenter &optional count redisplay

This function scrolls the text in the selected window so that point is displayed at a specified vertical position within the window. It does not move point with respect to the text.

If count is a non-negative number, that puts the line containing point count lines down from the top of the window. If count is a negative number, then it counts upward from the bottom of the window, so that −1 stands for the last usable line in the window.

If count is nil (or a non-nil list), recenter puts the line containing point in the middle of the window. If count is nil and redisplay is non-nil, this function may redraw the frame, according to the value of recenter-redisplay. Thus, omitting the second argument can be used to countermand the effect of recenter-redisplay being non-nil. Interactive calls pass non-nil for redisplay.

When recenter is called interactively, count is the raw prefix argument. Thus, typing C-u as the prefix sets the count to a non-nil list, while typing C-u 4 sets count to 4, which positions the current line four lines from the top.

With an argument of zero, recenter positions the current line at the top of the window. The command recenter-top-bottom offers a more convenient way to achieve this.

Function: recenter-window-group &optional count

This function is like recenter, except that when the selected window is part of a group of windows (see Window Group), recenter-window-group scrolls the entire group. This condition holds when the buffer local variable recenter-window-group-function is set to a function. In this case, recenter-window-group calls the function with the argument count, then returns its result. The argument count has the same meaning as in recenter, but with respect to the entire window group.

User Option: recenter-redisplay

If this variable is non-nil, calling recenter with a nil count argument and non-nil redisplay argument redraws the frame. The default value is tty, which means only redraw the frame if it is a tty frame.

Command: recenter-top-bottom &optional count

This command, which is the default binding for C-l, acts like recenter, except if called with no argument. In that case, successive calls place point according to the cycling order defined by the variable recenter-positions.

User Option: recenter-positions

This variable controls how recenter-top-bottom behaves when called with no argument. The default value is (middle top bottom), which means that successive calls of recenter-top-bottom with no argument cycle between placing point at the middle, top, and bottom of the window.


29.22 Vertical Fractional Scrolling

Vertical fractional scrolling means shifting text in a window up or down by a specified multiple or fraction of a line. Emacs uses it, for example, on images and screen lines which are taller than the window. Each window has a vertical scroll position, which is a number, never less than zero. It specifies how far to raise the contents of the window when displaying them. Raising the window contents generally makes all or part of some lines disappear off the top, and all or part of some other lines appear at the bottom. The usual value is zero.

The vertical scroll position is measured in units of the normal line height, which is the height of the default font. Thus, if the value is .5, that means the window contents will be scrolled up half the normal line height. If it is 3.3, that means the window contents are scrolled up somewhat over three times the normal line height.

What fraction of a line the vertical scrolling covers, or how many lines, depends on what the lines contain. A value of .5 could scroll a line whose height is very short off the screen, while a value of 3.3 could scroll just part of the way through a tall line or an image.

Function: window-vscroll &optional window pixels-p

This function returns the current vertical scroll position of window. The default for window is the selected window. If pixels-p is non-nil, the return value is measured in pixels, rather than in units of the normal line height.

(window-vscroll)
     ⇒ 0
Function: set-window-vscroll window lines &optional pixels-p preserve-vscroll-p

This function sets window’s vertical scroll position to lines. If window is nil, the selected window is used. The argument lines should be zero or positive; if not, it is taken as zero.

The actual vertical scroll position must always correspond to an integral number of pixels, so the value you specify is rounded accordingly.

The return value is the result of this rounding.

(set-window-vscroll (selected-window) 1.2)
     ⇒ 1.13

If pixels-p is non-nil, lines specifies a number of pixels. In this case, the return value is lines.

Normally, the vscroll does not take effect on windows that aren’t the minibuffer-scroll-window or the selected window when the mini-window is resized (see Minibuffer Windows). This “frozen” behavior is disabled when the preserve-vscroll-p parameter is non-nil, which means to set the vscroll as usual.

Variable: auto-window-vscroll

If this variable is non-nil, the line-move, scroll-up, and scroll-down functions will automatically modify the vertical scroll position to scroll through display rows that are taller than the height of the window, for example in the presence of large images.


29.23 Horizontal Scrolling

Horizontal scrolling means shifting the image in the window left or right by a specified multiple of the normal character width. Each window has a horizontal scroll position, which is a number, never less than zero. It specifies how far to shift the contents left. Shifting the window contents left generally makes all or part of some characters disappear off the left, and all or part of some other characters appear at the right. The usual value is zero.

The horizontal scroll position is measured in units of the normal character width, which is the width of space in the default font. Thus, if the value is 5, that means the window contents are scrolled left by 5 times the normal character width. How many characters actually disappear off to the left depends on their width, and could vary from line to line.

Because we read from side to side in the inner loop, and from top to bottom in the outer loop, the effect of horizontal scrolling is not like that of textual or vertical scrolling. Textual scrolling involves selection of a portion of text to display, and vertical scrolling moves the window contents contiguously; but horizontal scrolling causes part of each line to go off screen.

Usually, no horizontal scrolling is in effect; then the leftmost column is at the left edge of the window. In this state, scrolling to the right is meaningless, since there is no data to the left of the edge to be revealed by it; so this is not allowed. Scrolling to the left is allowed; it scrolls the first columns of text off the edge of the window and can reveal additional columns on the right that were truncated before. Once a window has a nonzero amount of leftward horizontal scrolling, you can scroll it back to the right, but only so far as to reduce the net horizontal scroll to zero. There is no limit to how far left you can scroll, but eventually all the text will disappear off the left edge.

If auto-hscroll-mode is set, redisplay automatically alters the horizontal scrolling of a window as necessary to ensure that point is always visible. However, you can still set the horizontal scrolling value explicitly. The value you specify serves as a lower bound for automatic scrolling, i.e., automatic scrolling will not scroll a window to a column less than the specified one.

The default value of auto-hscroll-mode is t; setting it to current-line activates a variant of automatic horizontal scrolling whereby only the line showing the cursor is horizontally scrolled to make point visible, the rest of the window is left either unscrolled, or at the minimum scroll amount set by scroll-left and scroll-right, see below.

Command: scroll-left &optional count set-minimum

This function scrolls the selected window count columns to the left (or to the right if count is negative). The default for count is the window width, minus 2.

The return value is the total amount of leftward horizontal scrolling in effect after the change—just like the value returned by window-hscroll (below).

Note that text in paragraphs whose base direction is right-to-left (see Bidirectional Display) moves in the opposite direction: e.g., it moves to the right when scroll-left is invoked with a positive value of count.

Once you scroll a window as far right as it can go, back to its normal position where the total leftward scrolling is zero, attempts to scroll any farther right have no effect.

If set-minimum is non-nil, the new scroll amount becomes the lower bound for automatic scrolling; that is, automatic scrolling will not scroll a window to a column less than the value returned by this function. Interactive calls pass non-nil for set-minimum.

Command: scroll-right &optional count set-minimum

This function scrolls the selected window count columns to the right (or to the left if count is negative). The default for count is the window width, minus 2. Aside from the direction of scrolling, this works just like scroll-left.

Function: window-hscroll &optional window

This function returns the total leftward horizontal scrolling of window—the number of columns by which the text in window is scrolled left past the left margin. (In right-to-left paragraphs, the value is the total amount of the rightward scrolling instead.) The default for window is the selected window.

The return value is never negative. It is zero when no horizontal scrolling has been done in window (which is usually the case).

(window-hscroll)
     ⇒ 0
(scroll-left 5)
     ⇒ 5
(window-hscroll)
     ⇒ 5
Function: set-window-hscroll window columns

This function sets horizontal scrolling of window. The value of columns specifies the amount of scrolling, in terms of columns from the left margin (right margin in right-to-left paragraphs). The argument columns should be zero or positive; if not, it is taken as zero. Fractional values of columns are not supported at present.

Note that set-window-hscroll may appear not to work if you test it by evaluating a call with M-: in a simple way. What happens is that the function sets the horizontal scroll value and returns, but then redisplay adjusts the horizontal scrolling to make point visible, and this overrides what the function did. You can observe the function’s effect if you call it while point is sufficiently far from the left margin that it will remain visible.

The value returned is columns.

(set-window-hscroll (selected-window) 10)
     ⇒ 10

Here is how you can determine whether a given position position is off the screen due to horizontal scrolling:

(defun hscroll-on-screen (window position)
  (save-excursion
    (goto-char position)
    (and
     (>= (- (current-column) (window-hscroll window)) 0)
     (< (- (current-column) (window-hscroll window))
        (window-width window)))))

29.24 Coordinates and Windows

This section describes functions that report positions of and within a window. Most of these functions report positions relative to an origin at the native position of the window’s frame (see Frame Geometry). Some functions report positions relative to the origin of the display of the window’s frame. In any case, the origin has the coordinates (0, 0) and X and Y coordinates increase rightward and downward respectively.

For the following functions, X and Y coordinates are reported in integer character units, i.e., numbers of lines and columns respectively. On a graphical display, each “line” and “column” corresponds to the height and width of the default character specified by the frame’s default font (see Frame Font).

Function: window-edges &optional window body absolute pixelwise

This function returns a list of the edge coordinates of window. If window is omitted or nil, it defaults to the selected window.

The return value has the form (left top right bottom). These list elements are, respectively, the X coordinate of the leftmost column occupied by the window, the Y coordinate of the topmost row, the X coordinate one column to the right of the rightmost column, and the Y coordinate one row down from the bottommost row.

Note that these are the actual outer edges of the window, including any of its decorations. On a text terminal, if the window has a neighbor on its right, its right edge includes the separator line between the window and its neighbor.

If the optional argument body is nil, this means to return the edges corresponding to the total size of window. body non-nil means to return the edges of window’s body. If body is non-nil, window must specify a live window.

If the optional argument absolute is nil, this means to return edges relative to the native position of window’s frame. absolute non-nil means to return coordinates relative to the origin (0, 0) of window’s display. On non-graphical systems this argument has no effect.

If the optional argument pixelwise is nil, this means to return the coordinates in terms of the default character width and height of window’s frame (see Frame Font), rounded if necessary. pixelwise non-nil means to return the coordinates in pixels. Note that the pixel specified by right and bottom is immediately outside of these edges. If absolute is non-nil, pixelwise is implicitly non-nil too.

Function: window-body-edges &optional window

This function returns the edges of window’s body (see Window Sizes). Calling (window-body-edges window) is equivalent to calling (window-edges window t), see above.

The following functions can be used to relate a set of frame-relative coordinates to a window:

Function: window-at x y &optional frame

This function returns the live window at the coordinates x and y given in default character sizes (see Frame Font) relative to the native position of frame (see Frame Geometry).

If there is no window at that position, the return value is nil. If frame is omitted or nil, it defaults to the selected frame.

Function: coordinates-in-window-p coordinates window

This function checks whether a window window occupies the frame relative coordinates coordinates, and if so, which part of the window that is. window should be a live window.

coordinates should be a cons cell of the form (x . y), where x and y are given in default character sizes (see Frame Font) relative to the native position of window’s frame (see Frame Geometry).

If there is no window at the specified position, the return value is nil . Otherwise, the return value is one of the following:

(relx . rely)

The coordinates are inside window. The numbers relx and rely are the equivalent window-relative coordinates for the specified position, counting from 0 at the top left corner of the window.

mode-line

The coordinates are in the mode line of window.

header-line

The coordinates are in the header line of window.

tab-line

The coordinates are in the tab line of window.

right-divider

The coordinates are in the divider separating window from a window on the right.

bottom-divider

The coordinates are in the divider separating window from a window beneath.

vertical-line

The coordinates are in the vertical line between window and its neighbor to the right. This value occurs only if the window doesn’t have a scroll bar; positions in a scroll bar are considered outside the window for these purposes.

left-fringe
right-fringe

The coordinates are in the left or right fringe of the window.

left-margin
right-margin

The coordinates are in the left or right margin of the window.

nil

The coordinates are not in any part of window.

The function coordinates-in-window-p does not require a frame as argument because it always uses the frame that window is on.

The following functions return window positions in pixels, rather than character units. Though mostly useful on graphical displays, they can also be called on text terminals, where the screen area of each text character is taken to be one pixel.

Function: window-pixel-edges &optional window

This function returns a list of pixel coordinates for the edges of window. Calling (window-pixel-edges window) is equivalent to calling (window-edges window nil nil t), see above.

Function: window-body-pixel-edges &optional window

This function returns the pixel edges of window’s body. Calling (window-body-pixel-edges window) is equivalent to calling (window-edges window t nil t), see above.

The following functions return window positions in pixels, relative to the origin of the display screen rather than that of the frame:

Function: window-absolute-pixel-edges &optional window

This function returns the pixel coordinates of window relative to an origin at (0, 0) of the display of window’s frame. Calling (window-absolute-pixel-edges) is equivalent to calling (window-edges window nil t t), see above.

Function: window-absolute-body-pixel-edges &optional window

This function returns the pixel coordinates of window’s body relative to an origin at (0, 0) of the display of window’s frame. Calling (window-absolute-body-pixel-edges window) is equivalent to calling (window-edges window t t t), see above.

Combined with set-mouse-absolute-pixel-position, this function can be used to move the mouse pointer to an arbitrary buffer position visible in some window:

(let ((edges (window-absolute-body-pixel-edges))
      (position (pos-visible-in-window-p nil nil t)))
  (set-mouse-absolute-pixel-position
   (+ (nth 0 edges) (nth 0 position))
   (+ (nth 1 edges) (nth 1 position))))

On a graphical terminal this form “warps” the mouse cursor to the upper left corner of the glyph at the selected window’s point. A position calculated this way can be also used to show a tooltip window there.

The following function returns the screen coordinates of a buffer position visible in a window:

Function: window-absolute-pixel-position &optional position window

If the buffer position position is visible in window window, this function returns the display coordinates of the upper/left corner of the glyph at position. The return value is a cons of the X- and Y-coordinates of that corner, relative to an origin at (0, 0) of window’s display. It returns nil if position is not visible in window.

window must be a live window and defaults to the selected window. position defaults to the value of window-point of window.

This means that in order to move the mouse pointer to the position of point in the selected window, it’s sufficient to write:

(let ((position (window-absolute-pixel-position)))
  (set-mouse-absolute-pixel-position
   (car position) (cdr position)))

The following function returns the largest rectangle that can be inscribed in a window without covering text displayed in that window.

Function: window-largest-empty-rectangle &optional window count min-width min-height positions left

This function calculates the dimensions of the largest empty rectangle that can be inscribed in the specified window’s text area. window must be a live window and defaults to the selected one.

The return value is a triple of the width and the start and end y-coordinates of the largest rectangle that can be inscribed into the empty space (space not displaying any text) of the text area of window. No x-coordinates are returned by this function—any such rectangle is assumed to end at the right edge of window’s text area. If no empty space can be found, the return value is nil.

The optional argument count, if non-nil, specifies a maximum number of rectangles to return. This means that the return value is a list of triples specifying rectangles with the largest rectangle first. count can be also a cons cell whose car specifies the number of rectangles to return and whose CDR, if non-nil, states that all rectangles returned must be disjoint.

The optional arguments min-width and min-height, if non-nil, specify the minimum width and height of any rectangle returned.

The optional argument positions, if non-nil, is a cons cell whose CAR specifies the uppermost and whose CDR specifies the lowermost pixel position that must be covered by any rectangle returned. These positions measure from the start of the text area of window.

The optional argument left, if non-nil, means to return values suitable for buffers displaying right to left text. In that case, any rectangle returned is assumed to start at the left edge of window’s text area.

Note that this function has to retrieve the dimensions of each line of window’s glyph matrix via window-lines-pixel-dimensions (see Size of Displayed Text). Hence, this function may also return nil when the current glyph matrix of window is not up-to-date.


29.25 Mouse Window Auto-selection

The following option allows to automatically select the window under the mouse pointer. This accomplishes a policy similar to that of window managers that give focus to a frame (and thus trigger its subsequent selection) whenever the mouse pointer enters its window-system window (see Input Focus).

User Option: mouse-autoselect-window

If this variable is non-nil, Emacs will try to automatically select the window under the mouse pointer. The following values are meaningful:

A positive number

This specifies a delay in seconds after which auto-selection triggers. The window under the mouse pointer is selected after the mouse has remained in it for the entire duration of the delay.

A negative number

A negative number has a similar effect as a positive number, but selects the window under the mouse pointer only after the mouse pointer has remained in it for the entire duration of the absolute value of that number and in addition has stopped moving.

Other value

Any other non-nil value means to select a window instantaneously as soon as the mouse pointer enters it.

In either case, the mouse pointer must enter the text area of a window in order to trigger its selection. Dragging the scroll bar slider or the mode line of a window conceptually should not cause its auto-selection.

Mouse auto-selection selects the minibuffer window only if it is active, and never deselects the active minibuffer window.

Mouse auto-selection can be used to emulate a focus follows mouse policy for child frames (see Child Frames) which usually are not tracked by the window manager. This requires to set the value of focus-follows-mouse (see Input Focus) to a non-nil value. If the value of focus-follows-mouse is auto-raise, entering a child frame with the mouse will raise it automatically above all other child frames of that frame’s parent frame.


29.26 Window Configurations

A window configuration records the entire layout of one frame—all windows, their sizes, their decorations, which buffers they contain, how those buffers are scrolled, and their value of point, It also includes the value of minibuffer-scroll-window. As a special exception, the window configuration does not record the value of point in the selected window for the current buffer.

You can bring back an entire frame layout by restoring a previously saved window configuration. If you want to record the layout of all frames instead of just one, use a frame configuration instead of a window configuration. See Frame Configurations.

Function: current-window-configuration &optional frame

This function returns a new object representing frame’s current window configuration. The default for frame is the selected frame. The variable window-persistent-parameters specifies which window parameters (if any) are saved by this function. See Window Parameters.

Function: set-window-configuration configuration &optional dont-set-frame dont-set-miniwindow

This function restores the configuration of windows and buffers as specified by configuration, for the frame that configuration was created for, regardless of whether that frame is selected or not. The argument configuration must be a value that was previously returned by current-window-configuration for that frame. Normally the function also selects the frame which is recorded in the configuration, but if dont-set-frame is non-nil, it leaves selected the frame which was already selected at the start of the function.

Normally the function restores the saved minibuffer (if any), but if dont-set-miniwindow is non-nil, the minibuffer current at the start of the function (if any) remains in the mini-window.

If the frame from which configuration was saved is dead, all this function does is to restore the value of the variable minibuffer-scroll-window and to adjust the value returned by minibuffer-selected-window. In this case, the function returns nil. Otherwise, it returns t.

If the buffer of a window of configuration has been killed since configuration was made, that window is, as a rule, removed from the restored configuration. However, if that window is the last window remaining in the restored configuration, another live buffer is shown in it.

Here is a way of using this function to get the same effect as save-window-excursion:

(let ((config (current-window-configuration)))
  (unwind-protect
      (progn (split-window-below nil)
             …)
    (set-window-configuration config)))
Macro: save-window-excursion forms…

This macro records the window configuration of the selected frame, executes forms in sequence, then restores the earlier window configuration. The return value is the value of the final form in forms.

Most Lisp code should not use this macro; save-selected-window is typically sufficient. In particular, this macro cannot reliably prevent the code in forms from opening new windows, because new windows might be opened in other frames (see Choosing a Window for Displaying a Buffer), and save-window-excursion only saves and restores the window configuration on the current frame.

Function: window-configuration-p object

This function returns t if object is a window configuration.

Function: window-configuration-equal-p config1 config2

This function says whether two window configurations have the same window layout, but ignores the values of point and the saved scrolling positions—it can return t even if those aspects differ.

Function: window-configuration-frame config

This function returns the frame for which the window configuration config was made.

Other primitives to look inside of window configurations would make sense, but are not implemented because we did not need them. See the file winner.el for some more operations on windows configurations.

The objects returned by current-window-configuration die together with the Emacs process. In order to store a window configuration on disk and read it back in another Emacs session, you can use the functions described next. These functions are also useful to clone the state of a frame into an arbitrary live window (set-window-configuration effectively clones the windows of a frame into the root window of that very frame only).

Function: window-state-get &optional window writable

This function returns the state of window as a Lisp object. The argument window must be a valid window and defaults to the root window of the selected frame.

If the optional argument writable is non-nil, this means to not use markers for sampling positions like window-point or window-start. This argument should be non-nil when the state will be written to disk and read back in another session.

Together, the argument writable and the variable window-persistent-parameters specify which window parameters are saved by this function. See Window Parameters.

The value returned by window-state-get can be used in the same session to make a clone of a window in another window. It can be also written to disk and read back in another session. In either case, use the following function to restore the state of the window.

Function: window-state-put state &optional window ignore

This function puts the window state state into window. The argument state should be the state of a window returned by an earlier invocation of window-state-get, see above. The optional argument window can be either a live window or an internal window (see Windows and Frames). If window is not a live window, it is replaced by a new live window created on the same frame before putting state into it. If window is nil, it puts the window state into a new window.

If the optional argument ignore is non-nil, it means to ignore minimum window sizes and fixed-size restrictions. If ignore is safe, this means windows can get as small as one line and/or two columns.

The functions window-state-get and window-state-put also allow to exchange the contents of two live windows. The following function does precisely that:

Command: window-swap-states &optional window-1 window-2 size

This command swaps the states of the two live windows window-1 and window-2. window-1 must specify a live window and defaults to the selected one. window-2 must specify a live window and defaults to the window following window-1 in the cyclic ordering of windows, excluding minibuffer windows and including live windows on all visible frames.

Optional argument size non-nil means to try swapping the sizes of window-1 and window-2 as well. A value of height means to swap heights only, a value of width means to swap widths only, while t means to swap both widths and heights, if possible. Frames are not resized by this function.


29.27 Window Parameters

This section describes the window parameters that can be used to associate additional information with windows.

Function: window-parameter window parameter

This function returns window’s value for parameter. The default for window is the selected window. If window has no setting for parameter, this function returns nil.

Function: window-parameters &optional window

This function returns all parameters of window and their values. The default for window is the selected window. The return value is either nil, or an association list whose elements have the form (parameter . value).

Function: set-window-parameter window parameter value

This function sets window’s value of parameter to value and returns value. The default for window is the selected window.

By default, the functions that save and restore window configurations or the states of windows (see Window Configurations) do not care about window parameters. This means that when you change the value of a parameter within the body of a save-window-excursion, the previous value is not restored when that macro exits. It also means that when you restore via window-state-put a window state saved earlier by window-state-get, all cloned windows have their parameters reset to nil. The following variable allows you to override the standard behavior:

Variable: window-persistent-parameters

This variable is an alist specifying which parameters get saved by current-window-configuration and window-state-get, and subsequently restored by set-window-configuration and window-state-put. See Window Configurations.

The CAR of each entry of this alist is a symbol specifying the parameter. The CDR should be one of the following:

nil

This value means the parameter is saved neither by window-state-get nor by current-window-configuration.

t

This value specifies that the parameter is saved by current-window-configuration and (provided its writable argument is nil) by window-state-get.

writable

This means that the parameter is saved unconditionally by both current-window-configuration and window-state-get. This value should not be used for parameters whose values do not have a read syntax. Otherwise, invoking window-state-put in another session may fail with an invalid-read-syntax error.

Some functions (notably delete-window, delete-other-windows and split-window), may behave specially when the window specified by their window argument has a parameter whose name is equal to the function’s name. You can override such special behavior by binding the following variable to a non-nil value:

Variable: ignore-window-parameters

If this variable is non-nil, some standard functions do not process window parameters. The functions currently affected by this are split-window, delete-window, delete-other-windows, and other-window.

An application can bind this variable to a non-nil value around calls to these functions. If it does so, the application is fully responsible for correctly assigning the parameters of all involved windows when exiting that function.

The following parameters are currently used by the window management code:

delete-window

This parameter affects the execution of delete-window (see Deleting Windows).

delete-other-windows

This parameter affects the execution of delete-other-windows (see Deleting Windows).

no-delete-other-windows

This parameter marks the window as not deletable by delete-other-windows (see Deleting Windows).

split-window

This parameter affects the execution of split-window (see Splitting Windows).

other-window

This parameter affects the execution of other-window (see Cyclic Ordering of Windows).

no-other-window

This parameter marks the window as not selectable by other-window (see Cyclic Ordering of Windows).

clone-of

This parameter specifies the window that this one has been cloned from. It is installed by window-state-get (see Window Configurations).

window-preserved-size

This parameter specifies a buffer, a direction where nil means vertical and t horizontal, and a size in pixels. If this window displays the specified buffer and its size in the indicated direction equals the size specified by this parameter, then Emacs will try to preserve the size of this window in the indicated direction. This parameter is installed and updated by the function window-preserve-size (see Preserving Window Sizes).

quit-restore

This parameter is installed by the buffer display functions (see Choosing a Window for Displaying a Buffer) and consulted by quit-restore-window (see Quitting Windows). It is a list of four elements, see the description of quit-restore-window in Quitting Windows for details.

window-side
window-slot

These parameters are used internally for implementing side windows (see Side Windows).

window-atom

This parameter is used internally for implementing atomic windows, see Atomic Windows.

mode-line-format

This parameter replaces the value of the buffer-local variable mode-line-format (see Mode Line Basics) of this window’s buffer whenever this window is displayed. The symbol none means to suppress display of a mode line for this window. Display and contents of the mode line on other windows showing this buffer are not affected.

header-line-format

This parameter replaces the value of the buffer-local variable header-line-format (see Mode Line Basics) of this window’s buffer whenever this window is displayed. The symbol none means to suppress display of a header line for this window. Display and contents of the header line on other windows showing this buffer are not affected.

tab-line-format

This parameter replaces the value of the buffer-local variable tab-line-format (see Mode Line Basics) of this window’s buffer whenever this window is displayed. The symbol none means to suppress display of a tab line for this window. Display and contents of the tab line on other windows showing this buffer are not affected.

min-margins

The value of this parameter is a cons cell whose CAR and CDR, if non-nil, specify the minimum values (in columns) for the left and right margin of this window (see Displaying in the Margins. When present, Emacs will use these values instead of the actual margin widths for determining whether a window can be split or shrunk horizontally.

Emacs never auto-adjusts the margins of any window after splitting or resizing it. It is the sole responsibility of any application setting this parameter to adjust the margins of this window as well as those of any new window that inherits this window’s margins due to a split. Both window-configuration-change-hook and window-size-change-functions (see Hooks for Window Scrolling and Changes) should be employed for this purpose.

This parameter was introduced in Emacs version 25.1 to support applications that use large margins to center buffer text within a window and should be used, with due care, exclusively by those applications. It might be replaced by an improved solution in future versions of Emacs.


29.28 Hooks for Window Scrolling and Changes

This section describes how Lisp programs can take action after a window has been scrolled or other window modifications occurred. We first consider the case where a window shows a different part of its buffer.

Variable: window-scroll-functions

This variable holds a list of functions that Emacs should call before redisplaying a window with scrolling. Displaying a different buffer in a window and making a new window also call these functions.

This variable is not a normal hook, because each function is called with two arguments: the window, and its new display-start position. At the time of the call, the display-start position of the argument window is already set to its new value, and the buffer to be displayed in the window is set as the current buffer.

These functions must take care when using window-end (see The Window Start and End Positions); if you need an up-to-date value, you must use the update argument to ensure you get it.

Warning: don’t use this feature to alter the way the window is scrolled. It’s not designed for that, and such use probably won’t work.

In addition, you can use jit-lock-register to register a Font Lock fontification function, which will be called whenever parts of a buffer are (re)fontified because a window was scrolled or its size changed. See Other Font Lock Variables.

The remainder of this section covers six hooks that are called during redisplay provided a significant, non-scrolling change of a window has been detected. For simplicity, these hooks and the functions they call will be collectively referred to as window change functions. As any hook, these hooks can be set either globally or buffer-locally via the local argument of add-hook (see Setting Hooks) when the hook is installed.

The first of these hooks is run after a window buffer change is detected, which means that a window was created, deleted or assigned another buffer.

Variable: window-buffer-change-functions

This variable specifies functions called during redisplay when window buffers have changed. The value should be a list of functions that take one argument.

Functions specified buffer-locally are called for any window showing the corresponding buffer if that window has been created or assigned that buffer since the last time window change functions were run. In this case the window is passed as argument.

Functions specified by the default value are called for a frame if at least one window on that frame has been added, deleted or assigned another buffer since the last time window change functions were run. In this case the frame is passed as argument.

The second of these hooks is run when a window size change has been detected which means that a window was created, assigned another buffer, or changed its total size or that of its text area.

Variable: window-size-change-functions

This variable specifies functions called during redisplay when a window size change occurred. The value should be a list of functions that take one argument.

Functions specified buffer-locally are called for any window showing the corresponding buffer if that window has been added or assigned another buffer or changed its total or body size since the last time window change functions were run. In this case the window is passed as argument.

Functions specified by the default value are called for a frame if at least one window on that frame has been added or assigned another buffer or changed its total or body size since the last time window change functions were run. In this case the frame is passed as argument.

The third of these hooks is run when a window selection change has selected another window since the last redisplay.

Variable: window-selection-change-functions

This variable specifies functions called during redisplay when the selected window or a frame’s selected window has changed. The value should be a list of functions that take one argument.

Functions specified buffer-locally are called for any window showing the corresponding buffer if that window has been selected or deselected (among all windows or among all windows on its frame) since the last time window change functions were run. In this case the window is passed as argument.

Functions specified by the default value are called for a frame if that frame has been selected or deselected or the frame’s selected window has changed since the last time window change functions were run. In this case the frame is passed as argument.

The fourth of these hooks is run when a window state change has been detected, which means that at least one of the three preceding window changes has occurred.

Variable: window-state-change-functions

This variable specifies functions called during redisplay when a window buffer or size change occurred or the selected window or a frame’s selected window has changed. The value should be a list of functions that take one argument.

Functions specified buffer-locally are called for any window showing the corresponding buffer if that window has been added or assigned another buffer, changed its total or body size or has been selected or deselected (among all windows or among all windows on its frame) since the last time window change functions were run. In this case the window is passed as argument.

Functions specified by the default value are called for a frame if at least one window on that frame has been added, deleted or assigned another buffer, changed its total or body size or that frame has been selected or deselected or the frame’s selected window has changed since the last time window change functions were run. In this case the frame is passed as argument.

Functions specified by the default value are also run for a frame when that frame’s window state change flag (see below) has been set since last redisplay.

The fifth of these hooks is run when a window configuration change has been detected which means that either the buffer or the size of a window changed. It differs from the four preceding hooks in the way it is run.

Variable: window-configuration-change-hook

This variable specifies functions called during redisplay when either the buffer or the size of a window has changed. The value should be a list of functions that take no argument.

Functions specified buffer-locally are called for any window showing the corresponding buffer if at least one window on that frame has been added, deleted or assigned another buffer or changed its total or body size since the last time window change functions were run. Each call is performed with the window showing the buffer temporarily selected and its buffer current.

Functions specified by the default value are called for each frame if at least one window on that frame has been added, deleted or assigned another buffer or changed its total or body size since the last time window change functions were run. Each call is performed with the frame temporarily selected and the selected window’s buffer current.

Finally, Emacs runs a normal hook that generalizes the behavior of window-state-change-functions.

Variable: window-state-change-hook

The default value of this variable specifies functions called during redisplay when a window state change has been detected or the window state change flag has been set on at least one frame. The value should be a list of functions that take no argument.

Applications should put a function on this hook only if they want to react to changes that happened on (or have been signaled for) two or more frames since last redisplay. In every other case, putting the function on window-state-change-functions should be preferred.

Window change functions are called during redisplay for each frame as follows: First, any buffer-local window buffer change function, window size change function, selected window change and window state change functions are called in this order. Next, the default values for these functions are called in the same order. Then any buffer-local window configuration change functions are called followed by functions specified by the default value of those functions. Finally, functions on window-state-change-hook are run.

Window change functions are run for a specific frame only if a corresponding change was registered for that frame earlier. Such changes include the creation or deletion of a window or the assignment of another buffer or size to a window. Note that even when such a change has been registered, this does not mean that any of the hooks described above is run. If, for example, a change was registered within the scope of a window excursion (see Window Configurations), this will trigger a call of window change functions only if that excursion still persists at the time change functions are run. If it is exited earlier, hooks will be run only if registered by a change outside the scope of that excursion.

The window state change flag of a frame, if set, will cause the default values of window-state-change-functions (for that frame) and window-state-change-hook to be run during next redisplay regardless of whether a window state change actually occurred for that frame or not. After running any functions on these hooks, the flag is reset for each frame. Applications can set that flag and inspect its value using the following functions.

Function: set-frame-window-state-change &optional frame arg

This function sets frame’s window state change flag if arg is non-nil and resets it otherwise. frame must be a live frame and defaults to the selected one.

Function: frame-window-state-change &optional frame

This functions returns t if frame’s window state change flag is set and nil otherwise. frame must be a live frame and defaults to the selected one.

While window change functions are run, the functions described next can be called to get more insight into what has changed for a specific window or frame since the last redisplay. All these functions take a live window as single, optional argument, defaulting to the selected window.

Function: window-old-buffer &optional window

This function returns the buffer shown in window at the last time window change functions were run for window’s frame. If it returns nil, window has been created after that. If it returns t, window was not shown at that time but has been restored from a previously saved window configuration afterwards. Otherwise, the return value is the buffer shown by window at that time.

Function: window-old-pixel-width &optional window

This function returns the total pixel width of window the last time window change functions found window live on its frame. It is zero if window was created after that.

Function: window-old-pixel-height &optional window

This function returns the total pixel height of window the last time window change functions found window live on its frame. It is zero if window was created after that.

Function: window-old-body-pixel-width &optional window

This function returns the pixel width of window’s text area the last time window change functions found window live on its frame. It is zero if window was created after that.

Function: window-old-body-pixel-height &optional window

This function returns the pixel height of window’s text area the last time window change functions found window live on its frame. It is zero if window was created after that.

In order to find out which window or frame was selected the last time window change functions were run, the following functions can be used:

Function: frame-old-selected-window &optional frame

This function returns the selected window of frame at the last time window change functions were run. If omitted or nil frame defaults to the selected frame.

Function: old-selected-window

This function returns the selected window at the last time window change functions were run.

Function: old-selected-frame

This function returns the selected frame at the last time window change functions were run.

Note that window change functions provide no information about which windows have been deleted since the last time they were run. If necessary, applications should remember any window showing a specific buffer in a local variable of that buffer and update it in a function run by the default values of any of the hooks that are run when a window buffer change was detected.

The following caveats should be considered when adding a function to window change functions:

  • Some operations will not trigger a call of window change functions. These include showing another buffer in a minibuffer window or any change of a tooltip window.
  • Window change functions should not create or delete windows or change the buffer, size or selection status of any window because there is no guarantee that the information about such a change will be propagated to other window change functions. If at all, any such change should be executed only by the last function listed by the default value of window-state-change-hook.
  • Macros like save-window-excursion, with-selected-window or with-current-buffer can be used when running window change functions.
  • Running window change functions does not save and restore match data. Unless running window-configuration-change-hook it does not save or restore the selected window or frame or the current buffer either.
  • Any redisplay triggering the run of window change functions may be aborted. If the abort occurs before window change functions have run to their completion, they will be run again with the previous values, that is, as if redisplay had not been performed. If aborted later, they will be run with the new values, that is, as if redisplay had been actually performed.

30 Frames

A frame is a screen object that contains one or more Emacs windows (see Windows). It is the kind of object called a “window” in the terminology of graphical environments; but we can’t call it a “window” here, because Emacs uses that word in a different way. In Emacs Lisp, a frame object is a Lisp object that represents a frame on the screen. See Frame Type.

A frame initially contains a single main window and/or a minibuffer window; you can subdivide the main window vertically or horizontally into smaller windows. See Splitting Windows.

A terminal is a display device capable of displaying one or more Emacs frames. In Emacs Lisp, a terminal object is a Lisp object that represents a terminal. See Terminal Type.

There are two classes of terminals: text terminals and graphical terminals. Text terminals are non-graphics-capable displays, including xterm and other terminal emulators. On a text terminal, each Emacs frame occupies the terminal’s entire screen; although you can create additional frames and switch between them, the terminal only shows one frame at a time. Graphical terminals, on the other hand, are managed by graphical display systems such as the X Window System, which allow Emacs to show multiple frames simultaneously on the same display.

On GNU and Unix systems, you can create additional frames on any available terminal, within a single Emacs session, regardless of whether Emacs was started on a text or graphical terminal. Emacs can display on both graphical and text terminals simultaneously. This comes in handy, for instance, when you connect to the same session from several remote locations. See Multiple Terminals.

Function: framep object

This predicate returns a non-nil value if object is a frame, and nil otherwise. For a frame, the value indicates which kind of display the frame uses:

t

The frame is displayed on a text terminal.

x

The frame is displayed on an X graphical terminal.

w32

The frame is displayed on a MS-Windows graphical terminal.

ns

The frame is displayed on a GNUstep or Macintosh Cocoa graphical terminal.

pc

The frame is displayed on an MS-DOS terminal.

pgtk

The frame is displayed using pure GTK facilities.

Function: frame-terminal &optional frame

This function returns the terminal object that displays frame. If frame is nil or unspecified, it defaults to the selected frame.

Function: terminal-live-p object

This predicate returns a non-nil value if object is a terminal that is live (i.e., not deleted), and nil otherwise. For live terminals, the return value indicates what kind of frames are displayed on that terminal; the list of possible values is the same as for framep above.

On a graphical terminal we distinguish two types of frames: A normal top-level frame is a frame whose window-system window is a child of the window-system’s root window for that terminal. A child frame is a frame whose window-system window is the child of the window-system window of another Emacs frame. See Child Frames.


30.1 Creating Frames

To create a new frame, call the function make-frame.

Command: make-frame &optional parameters

This function creates and returns a new frame, displaying the current buffer.

The parameters argument is an alist that specifies frame parameters for the new frame. See Frame Parameters. If you specify the terminal parameter in parameters, the new frame is created on that terminal. Otherwise, if you specify the window-system frame parameter in parameters, that determines whether the frame should be displayed on a text terminal or a graphical terminal. See Window Systems. If neither is specified, the new frame is created in the same terminal as the selected frame.

Any parameters not mentioned in parameters default to the values in the alist default-frame-alist (see Initial Frame Parameters); parameters not specified there default from the X resources or its equivalent on your operating system (see X Resources in The GNU Emacs Manual). After the frame is created, this function applies any parameters specified in frame-inherited-parameters (see below) it has no assigned yet, taking the values from the frame that was selected when make-frame was called.

Note that on multi-monitor displays (see Multiple Terminals), the window manager might position the frame differently than specified by the positional parameters in parameters (see Position Parameters). For example, some window managers have a policy of displaying the frame on the monitor that contains the largest part of the window (a.k.a. the dominating monitor).

This function itself does not make the new frame the selected frame. See Input Focus. The previously selected frame remains selected. On graphical terminals, however, the window system may select the new frame for its own reasons.

Variable: before-make-frame-hook

A normal hook run by make-frame before it creates the frame.

Variable: after-make-frame-functions

An abnormal hook run by make-frame after it created the frame. Each function in after-make-frame-functions receives one argument, the frame just created.

Note that any functions added to these hooks by your initial file are usually not run for the initial frame, since Emacs reads the initial file only after creating that frame. However, if the initial frame is specified to use a separate minibuffer frame (see Minibuffers and Frames), the functions will be run for both, the minibuffer-less and the minibuffer frame. Alternatively, you can add functions to these hooks in your “early init file” (see The Init File), in which case they will be in effect for the initial frame as well.

Variable: frame-inherited-parameters

This variable specifies the list of frame parameters that a newly created frame inherits from the currently selected frame. For each parameter (a symbol) that is an element in this list and has not been assigned earlier when processing make-frame, the function sets the value of that parameter in the created frame to its value in the selected frame.

User Option: server-after-make-frame-hook

A normal hook run when the Emacs server starts using a client frame. When this hook is called, the client frame is the selected one. Note that, depending on how emacsclient was invoked (see Invoking emacsclient in The GNU Emacs Manual), this client frame could be a new frame created for the client, or it could be an existing frame that the server reused for handling the client commands. See Emacs Server in The GNU Emacs Manual.


30.2 Multiple Terminals

Emacs represents each terminal as a terminal object data type (see Terminal Type). On GNU and Unix systems, Emacs can use multiple terminals simultaneously in each session. On other systems, it can only use a single terminal. Each terminal object has the following attributes:

  • The name of the device used by the terminal (e.g., ‘:0.0’ or /dev/tty).
  • The terminal and keyboard coding systems used on the terminal. See Terminal I/O Encoding.
  • The kind of display associated with the terminal. This is the symbol returned by the function terminal-live-p (i.e., x, t, w32, ns, pc, haiku, or pgtk). See Frames.
  • A list of terminal parameters. See Terminal Parameters.

There is no primitive for creating terminal objects. Emacs creates them as needed, such as when you call make-frame-on-display (described below).

Function: terminal-name &optional terminal

This function returns the file name of the device used by terminal. If terminal is omitted or nil, it defaults to the selected frame’s terminal. terminal can also be a frame, meaning that frame’s terminal.

Function: terminal-list

This function returns a list of all live terminal objects.

Function: get-device-terminal device

This function returns a terminal whose device name is given by device. If device is a string, it can be either the file name of a terminal device, or the name of an X display of the form ‘host:server.screen’. If device is a frame, this function returns that frame’s terminal; nil means the selected frame. Finally, if device is a terminal object that represents a live terminal, that terminal is returned. The function signals an error if its argument is none of the above.

Function: delete-terminal &optional terminal force

This function deletes all frames on terminal and frees the resources used by it. It runs the abnormal hook delete-terminal-functions, passing terminal as the argument to each function.

If terminal is omitted or nil, it defaults to the selected frame’s terminal. terminal can also be a frame, meaning that frame’s terminal.

Normally, this function signals an error if you attempt to delete the sole active terminal, but if force is non-nil, you are allowed to do so. Emacs automatically calls this function when the last frame on a terminal is deleted (see Deleting Frames).

Variable: delete-terminal-functions

An abnormal hook run by delete-terminal. Each function receives one argument, the terminal argument passed to delete-terminal. Due to technical details, the functions may be called either just before the terminal is deleted, or just afterwards.

A few Lisp variables are terminal-local; that is, they have a separate binding for each terminal. The binding in effect at any time is the one for the terminal that the currently selected frame belongs to. These variables include default-minibuffer-frame, defining-kbd-macro, last-kbd-macro, and system-key-alist. They are always terminal-local, and can never be buffer-local (see Buffer-Local Variables).

On GNU and Unix systems, each X display is a separate graphical terminal. When Emacs is started from within the X window system, it uses the X display specified by the DISPLAY environment variable, or by the ‘--display’ option (see Initial Options in The GNU Emacs Manual). Emacs can connect to other X displays via the command make-frame-on-display. Each X display has its own selected frame and its own minibuffer windows; however, only one of those frames is the selected frame at any given moment (see Input Focus). Emacs can even connect to other text terminals, by interacting with the emacsclient program. See Emacs Server in The GNU Emacs Manual.

A single X server can handle more than one display. Each X display has a three-part name, ‘hostname:displaynumber.screennumber’. The first part, hostname, specifies the name of the machine to which the display is physically connected. The second part, displaynumber, is a zero-based number that identifies one or more monitors connected to that machine that share a common keyboard and pointing device (mouse, tablet, etc.). The third part, screennumber, identifies a zero-based screen number (a separate monitor) that is part of a single monitor collection on that X server. When you use two or more screens belonging to one server, Emacs knows by the similarity in their names that they share a single keyboard.

Systems that don’t use the X window system, such as MS-Windows, don’t support the notion of X displays, and have only one display on each host. The display name on these systems doesn’t follow the above 3-part format; for example, the display name on MS-Windows systems is a constant string ‘w32’, and exists for compatibility, so that you could pass it to functions that expect a display name.

Command: make-frame-on-display display &optional parameters

This function creates and returns a new frame on display, taking the other frame parameters from the alist parameters. display should be the name of an X display (a string).

Before creating the frame, this function ensures that Emacs is set up to display graphics. For instance, if Emacs has not processed X resources (e.g., if it was started on a text terminal), it does so at this time. In all other respects, this function behaves like make-frame (see Creating Frames).

Function: x-display-list

This function returns a list that indicates which X displays Emacs has a connection to. The elements of the list are strings, and each one is a display name.

Function: x-open-connection display &optional xrm-string must-succeed

This function opens a connection to the X display display, without creating a frame on that display. Normally, Emacs Lisp programs need not call this function, as make-frame-on-display calls it automatically. The only reason for calling it is to check whether communication can be established with a given X display.

The optional argument xrm-string, if not nil, is a string of resource names and values, in the same format used in the .Xresources file. See X Resources in The GNU Emacs Manual. These values apply to all Emacs frames created on this display, overriding the resource values recorded in the X server. Here’s an example of what this string might look like:

"*BorderWidth: 3\n*InternalBorder: 2\n"

If must-succeed is non-nil, failure to open the connection terminates Emacs. Otherwise, it is an ordinary Lisp error.

Function: x-close-connection display

This function closes the connection to display display. Before you can do this, you must first delete all the frames that were open on that display (see Deleting Frames).

On some multi-monitor setups, a single X display outputs to more than one physical monitor. You can use the functions display-monitor-attributes-list and frame-monitor-attributes to obtain information about such setups.

Function: display-monitor-attributes-list &optional display

This function returns a list of physical monitor attributes on display, which can be a display name (a string), a terminal, or a frame; if omitted or nil, it defaults to the selected frame’s display. Each element of the list is an association list, representing the attributes of a physical monitor. The first element corresponds to the primary monitor. The attribute keys and values are:

geometry

Position of the top-left corner of the monitor’s screen and its size, in pixels, as ‘(x y width height)’. Note that, if the monitor is not the primary monitor, some of the coordinates might be negative.

workarea

Position of the top-left corner and size of the work area (usable space) in pixels as ‘(x y width height)’. This may be different from ‘geometry’ in that space occupied by various window manager features (docks, taskbars, etc.) may be excluded from the work area. Whether or not such features actually subtract from the work area depends on the platform and environment. Again, if the monitor is not the primary monitor, some of the coordinates might be negative.

mm-size

Width and height in millimeters as ‘(width height)

frames

List of frames that this physical monitor dominates (see below).

name

Name of the physical monitor as string.

source

Source of the multi-monitor information as string; e.g., ‘XRandR 1.5’, ‘XRandr’ or ‘Xinerama’.

x, y, width, and height are integers. ‘name’ and ‘source’ may be absent.

A frame is dominated by a physical monitor when either the largest area of the frame resides in that monitor, or (if the frame does not intersect any physical monitors) that monitor is the closest to the frame. Every (non-tooltip) frame (whether visible or not) in a graphical display is dominated by exactly one physical monitor at a time, though the frame can span multiple (or no) physical monitors.

Here’s an example of the data produced by this function on a 2-monitor display:

  (display-monitor-attributes-list)
  ⇒
  (((geometry 0 0 1920 1080) ;; Left-hand, primary monitor
    (workarea 0 0 1920 1050) ;; A taskbar occupies some of the height
    (mm-size 677 381)
    (name . "DISPLAY1")
    (frames #<frame emacs@host *Messages* 0x11578c0>
            #<frame emacs@host *scratch* 0x114b838>))
   ((geometry 1920 0 1680 1050) ;; Right-hand monitor
    (workarea 1920 0 1680 1050) ;; Whole screen can be used
    (mm-size 593 370)
    (name . "DISPLAY2")
    (frames)))
Function: frame-monitor-attributes &optional frame

This function returns the attributes of the physical monitor dominating (see above) frame, which defaults to the selected frame.

On multi-monitor displays it is possible to use the command make-frame-on-monitor to make frames on the specified monitor.

Command: make-frame-on-monitor monitor &optional display parameters

This function creates and returns a new frame on monitor located on display, taking the other frame parameters from the alist parameters. monitor should be the name of the physical monitor, the same string as returned by the function display-monitor-attributes-list in the attribute name. display should be the name of an X display (a string).

Variable: display-monitors-changed-functions

This variable is an abnormal hook run when the monitor configuration changes, which can happen if a monitor is rotated, moved, added or removed from a multiple-monitor setup, if the primary monitor changes, or if the resolution of a monitor changes. It is called with a single argument consisting of the terminal on which the monitor configuration changed. Programs should call display-monitor-attributes-list with the terminal as the argument to retrieve the new monitor configuration on that terminal.


30.3 Frame Geometry

The geometry of a frame depends on the toolkit that was used to build this instance of Emacs and the terminal that displays the frame. This chapter describes these dependencies and some of the functions to deal with them. Note that the frame argument of all of these functions has to specify a live frame (see Deleting Frames). If omitted or nil, it specifies the selected frame (see Input Focus).


30.3.1 Frame Layout

A visible frame occupies a rectangular area on its terminal’s display. This area may contain a number of nested rectangles, each serving a different purpose. The drawing below sketches the layout of a frame on a graphical terminal:


        <------------ Outer Frame Width ----------->
        ____________________________________________
     ^(0)  ________ External/Outer Border _______   |
     | |  |_____________ Title Bar ______________|  |
     | | (1)_____________ Menu Bar ______________|  | ^
     | | (2)_____________ Tool Bar ______________|  | ^
     | | (3)_____________ Tab Bar _______________|  | ^
     | |  |  _________ Internal Border ________  |  | ^
     | |  | |   ^                              | |  | |
     | |  | |   |                              | |  | |
Outer  |  | | Inner                            | |  | Native
Frame  |  | | Frame                            | |  | Frame
Height |  | | Height                           | |  | Height
     | |  | |   |                              | |  | |
     | |  | |<--+--- Inner Frame Width ------->| |  | |
     | |  | |   |                              | |  | |
     | |  | |___v______________________________| |  | |
     | |  |___________ Internal Border __________|  | v
     v |___________ External/Outer Border __________|
           <-------- Native Frame Width -------->

In practice not all of the areas shown in the drawing will or may be present. The meaning of these areas is described below.

Outer Frame

The outer frame is a rectangle comprising all areas shown in the drawing. The edges of that rectangle are called the outer edges of the frame. Together, the outer width and outer height of the frame specify the outer size of that rectangle.

Knowing the outer size of a frame is useful for fitting a frame into the working area of its display (see Multiple Terminals) or for placing two frames adjacent to each other on the screen. Usually, the outer size of a frame is available only after the frame has been mapped (made visible, see Visibility of Frames) at least once. For the initial frame or a frame that has not been created yet, the outer size can be only estimated or must be calculated from the window-system’s or window manager’s defaults. One workaround is to obtain the differences of the outer and native (see below) sizes of a mapped frame and use them for calculating the outer size of the new frame.

The position of the upper left corner of the outer frame (indicated by ‘(0)’ in the drawing above) is the outer position of the frame. The outer position of a graphical frame is also referred to as “the position” of the frame because it usually remains unchanged on its display whenever the frame is resized or its layout is changed.

The outer position is specified by and can be set via the left and top frame parameters (see Position Parameters). For a normal, top-level frame these parameters usually represent its absolute position (see below) with respect to its display’s origin. For a child frame (see Child Frames) these parameters represent its position relative to the native position (see below) of its parent frame. For frames on text terminals the values of these parameters are meaningless and always zero.

External Border

The external border is part of the decorations supplied by the window manager. It is typically used for resizing the frame with the mouse and is therefore not shown on “fullboth” and maximized frames (see Size Parameters). Its width is determined by the window manager and cannot be changed by Emacs’ functions.

External borders don’t exist on text terminal frames. For graphical frames, their display can be suppressed by setting the override-redirect or undecorated frame parameter (see Window Management Parameters).

Outer Border

The outer border is a separate border whose width can be specified with the border-width frame parameter (see Layout Parameters). In practice, either the external or the outer border of a frame are displayed but never both at the same time. Usually, the outer border is shown only for special frames that are not (fully) controlled by the window manager like tooltip frames (see Tooltips), child frames (see Child Frames) and undecorated or override-redirect frames (see Window Management Parameters).

Outer borders are never shown on text terminal frames and on frames generated by GTK+ routines. On MS-Windows, the outer border is emulated with the help of a one pixel wide external border. Non-toolkit builds on X allow to change the color of the outer border by setting the border-color frame parameter (see Layout Parameters).

Title Bar

The title bar, a.k.a. caption bar, is also part of the window manager’s decorations and typically displays the title of the frame (see Frame Titles) as well as buttons for minimizing, maximizing and deleting the frame. It can be also used for dragging the frame with the mouse. The title bar is usually not displayed for fullboth (see Size Parameters), tooltip (see Tooltips) and child frames (see Child Frames) and doesn’t exist for terminal frames. Display of the title bar can be suppressed by setting the override-redirect or the undecorated frame parameters (see Window Management Parameters).

Menu Bar

The menu bar (see The Menu Bar) can be either internal (drawn by Emacs itself) or external (drawn by the toolkit). Most builds (GTK+, Lucid, Motif and MS-Windows) rely on an external menu bar. NS also uses an external menu bar which, however, is not part of the outer frame. Non-toolkit builds can provide an internal menu bar. On text terminal frames, the menu bar is part of the frame’s root window (see Windows and Frames). As a rule, menu bars are never shown on child frames (see Child Frames). Display of the menu bar can be suppressed by setting the menu-bar-lines parameter (see Layout Parameters) to zero.

Whether the menu bar is wrapped or truncated whenever its width becomes too large to fit on its frame depends on the toolkit . Usually, only Motif and MS-Windows builds can wrap the menu bar. When they (un-)wrap the menu bar, they try to keep the outer height of the frame unchanged, so the native height of the frame (see below) will change instead.

Tool Bar

Like the menu bar, the tool bar (see Tool bars) can be either internal (drawn by Emacs itself) or external (drawn by a toolkit). The GTK+ and NS builds have the tool bar drawn by the toolkit. The remaining builds use internal tool bars. With GTK+ the tool bar can be located on either side of the frame, immediately outside the internal border, see below. Tool bars are usually not shown for child frames (see Child Frames). Display of the tool bar can be suppressed by setting the tool-bar-lines parameter (see Layout Parameters) to zero.

If the variable auto-resize-tool-bars is non-nil, Emacs wraps the internal tool bar when its width becomes too large for its frame. If and when Emacs (un-)wraps the internal tool bar, it by default keeps the outer height of the frame unchanged, so the native height of the frame (see below) will change instead. Emacs built with GTK+, on the other hand, never wraps the tool bar but may automatically increase the outer width of a frame in order to accommodate an overlong tool bar.

Tab Bar

The tab bar (see Tab Bars in The GNU Emacs Manual) is always drawn by Emacs itself. The tab bar appears above the tool bar in Emacs built with an internal tool bar, and below the tool bar in builds with an external tool bar. Display of the tab bar can be suppressed by setting the tab-bar-lines parameter (see Layout Parameters) to zero.

Native Frame

The native frame is a rectangle located entirely within the outer frame. It excludes the areas occupied by an external or outer border, the title bar and any external menu or tool bar. The edges of the native frame are called the native edges of the frame. Together, the native width and native height of a frame specify the native size of the frame.

The native size of a frame is the size Emacs passes to the window-system or window manager when creating or resizing the frame from within Emacs. It is also the size Emacs receives from the window-system or window manager whenever these resize the frame’s window-system window, for example, after maximizing the frame by clicking on the corresponding button in the title bar or when dragging its external border with the mouse.

The position of the top left corner of the native frame specifies the native position of the frame. (1)–(3) in the drawing above indicate that position for the various builds:

  • (1) non-toolkit, Haiku, and terminal frames
  • (2) Lucid, Motif, and MS-Windows frames
  • (3) GTK+ and NS frames

Accordingly, the native height of a frame may include the height of the tool bar but not that of the menu bar (Lucid, Motif, MS-Windows) or those of the menu bar and the tool bar (non-toolkit and text terminal frames).

The native position of a frame is the reference position for functions that set or return the current position of the mouse (see Mouse Position) and for functions dealing with the position of windows like window-edges, window-at or coordinates-in-window-p (see Coordinates and Windows). It also specifies the (0, 0) origin for locating and positioning child frames within this frame (see Child Frames).

Note also that the native position of a frame usually remains unaltered on its display when removing or adding the window manager decorations by changing the frame’s override-redirect or undecorated parameter (see Window Management Parameters).

Internal Border

The internal border is a border drawn by Emacs around the inner frame (see below). The specification of its appearance depends on whether or not the given frame is a child frame (see Child Frames).

For normal frames its width is specified by the internal-border-width frame parameter (see Layout Parameters), and its color is specified by the background of the internal-border face.

For child frames its width is specified by the child-frame-border-width frame parameter (but will use the internal-border-width parameter as fallback), and its color is specified by the background of the child-frame-border face.

Inner Frame

The inner frame is the rectangle reserved for the frame’s windows. It’s enclosed by the internal border which, however, is not part of the inner frame. Its edges are called the inner edges of the frame. The inner width and inner height specify the inner size of the rectangle. The inner frame is sometimes also referred to as the display area of the frame.

As a rule, the inner frame is subdivided into the frame’s root window (see Windows and Frames) and the frame’s minibuffer window (see Minibuffer Windows). There are two notable exceptions to this rule: A minibuffer-less frame contains a root window only and does not contain a minibuffer window. A minibuffer-only frame contains only a minibuffer window which also serves as that frame’s root window. See Initial Frame Parameters for how to create such frame configurations.

Text Area

The text area of a frame is a somewhat fictitious area that can be embedded in the native frame. Its position is unspecified. Its width can be obtained by removing from that of the native width the widths of the internal border, one vertical scroll bar, and one left and one right fringe if they are specified for this frame, see Layout Parameters. Its height can be obtained by removing from that of the native height the widths of the internal border and the heights of the frame’s internal menu and tool bars, the tab bar and one horizontal scroll bar if specified for this frame.

The absolute position of a frame is given as a pair (X, Y) of horizontal and vertical pixel offsets relative to an origin (0, 0) of the frame’s display. Correspondingly, the absolute edges of a frame are given as pixel offsets from that origin.

Note that with multiple monitors, the origin of the display does not necessarily coincide with the top-left corner of the entire usable display area of the terminal. Hence the absolute position of a frame can be negative in such an environment even when that frame is completely visible.

By convention, vertical offsets increase “downwards”. This means that the height of a frame is obtained by subtracting the offset of its top edge from that of its bottom edge. Horizontal offsets increase “rightwards”, as expected, so a frame’s width is calculated by subtracting the offset of its left edge from that of its right edge.

For a frame on a graphical terminal the following function returns the sizes of the areas described above:

Function: frame-geometry &optional frame

This function returns geometric attributes of frame. The return value is an association list of the attributes listed below. All coordinate, height and width values are integers counting pixels. Note that if frame has not been mapped yet, (see Visibility of Frames) some of the return values may only represent approximations of the actual values—those that can be seen after the frame has been mapped.

outer-position

A cons representing the absolute position of the outer frame, relative to the origin at position (0, 0) of frame’s display.

outer-size

A cons of the outer width and height of frame.

external-border-size

A cons of the horizontal and vertical width of frame’s external borders as supplied by the window manager. If the window manager doesn’t supply these values, Emacs will try to guess them from the coordinates of the outer and inner frame.

outer-border-width

The width of the outer border of frame. The value is meaningful for non-GTK+ X builds only.

title-bar-size

A cons of the width and height of the title bar of frame as supplied by the window manager or operating system. If both of them are zero, the frame has no title bar. If only the width is zero, Emacs was not able to retrieve the width information.

menu-bar-external

If non-nil, this means the menu bar is external (not part of the native frame of frame).

menu-bar-size

A cons of the width and height of the menu bar of frame.

tool-bar-external

If non-nil, this means the tool bar is external (not part of the native frame of frame).

tool-bar-position

This tells on which side the tool bar on frame is and can be one of left, top, right or bottom. The only toolkit that currently supports a value other than top is GTK+.

tool-bar-size

A cons of the width and height of the tool bar of frame.

internal-border-width

The width of the internal border of frame.

The following function can be used to retrieve the edges of the outer, native and inner frame.

Function: frame-edges &optional frame type

This function returns the absolute edges of the outer, native or inner frame of frame. frame must be a live frame and defaults to the selected one. The returned list has the form (left top right bottom) where all values are in pixels relative to the origin of frame’s display. For terminal frames the values returned for left and top are always zero.

Optional argument type specifies the type of the edges to return: outer-edges means to return the outer edges of frame, native-edges (or nil) means to return its native edges and inner-edges means to return its inner edges.

By convention, the pixels of the display at the values returned for left and top are considered to be inside (part of) frame. Hence, if left and top are both zero, the pixel at the display’s origin is part of frame. The pixels at bottom and right, on the other hand, are considered to lie immediately outside frame. This means that if you have, for example, two side-by-side frames positioned such that the right outer edge of the frame on the left equals the left outer edge of the frame on the right, the pixels at that edge show a part of the frame on the right.


30.3.2 Frame Font

Each frame has a default font which specifies the default character size for that frame. This size is meant when retrieving or changing the size of a frame in terms of columns or lines (see Size Parameters). It is also used when resizing (see Window Sizes) or splitting (see Splitting Windows) windows.

The terms line height and canonical character height are sometimes used instead of “default character height”. Similarly, the terms column width and canonical character width are used instead of “default character width”.

Function: frame-char-height &optional frame
Function: frame-char-width &optional frame

These functions return the default height and width of a character in frame, measured in pixels. Together, these values establish the size of the default font on frame. The values depend on the choice of font for frame, see Font and Color Parameters.

The default font can be also set directly with the following function:

Command: set-frame-font font &optional keep-size frames

This sets the default font to font. When called interactively, it prompts for the name of a font, and uses that font on the selected frame. When called from Lisp, font should be a font name (a string), a font object, font entity, or a font spec.

If the optional argument keep-size is nil, this keeps the number of frame lines and columns fixed. (If non-nil, the option frame-inhibit-implied-resize described in the next section will override this.) If keep-size is non-nil (or with a prefix argument), it tries to keep the size of the display area of the current frame fixed by adjusting the number of lines and columns.

If the optional argument frames is nil, this applies the font to the selected frame only. If frames is non-nil, it should be a list of frames to act upon, or t meaning all existing and all future graphical frames.


30.3.3 Frame Position

On graphical systems, the position of a normal top-level frame is specified as the absolute position of its outer frame (see Frame Geometry). The position of a child frame (see Child Frames) is specified via pixel offsets of its outer edges relative to the native position of its parent frame.

You can access or change the position of a frame using the frame parameters left and top (see Position Parameters). Here are two additional functions for working with the positions of an existing, visible frame. For both functions, the argument frame must denote a live frame and defaults to the selected frame.

Function: frame-position &optional frame

For a normal, non-child frame this function returns a cons of the pixel coordinates of its outer position (see Frame Layout) with respect to the origin (0, 0) of its display. For a child frame (see Child Frames) this function returns the pixel coordinates of its outer position with respect to an origin (0, 0) at the native position of frame’s parent.

Negative values never indicate an offset from the right or bottom edge of frame’s display or parent frame. Rather, they mean that frame’s outer position is on the left and/or above the origin of its display or the native position of its parent frame. This usually means that frame is only partially visible (or completely invisible). However, on systems where the display’s origin does not coincide with its top-left corner, the frame may be visible on a secondary monitor.

On a text terminal frame both values are zero.

Function: set-frame-position frame x y

This function sets the outer frame position of frame to (x, y). The latter arguments specify pixels and normally count from the origin at the position (0, 0) of frame’s display. For child frames, they count from the native position of frame’s parent frame.

Negative parameter values position the right edge of the outer frame by -x pixels left from the right edge of the screen (or the parent frame’s native rectangle) and the bottom edge by -y pixels up from the bottom edge of the screen (or the parent frame’s native rectangle).

Note that negative values do not permit to align the right or bottom edge of frame exactly at the right or bottom edge of its display or parent frame. Neither do they allow to specify a position that does not lie within the edges of the display or parent frame. The frame parameters left and top (see Position Parameters) allow to do that, but may still fail to provide good results for the initial or a new frame.

This function has no effect on text terminal frames.

Variable: move-frame-functions

This hook specifies the functions that are run when an Emacs frame is moved (assigned a new position) by the window-system or window manager. The functions are run with one argument, the frame that moved. For a child frame (see Child Frames), the functions are run only when the position of the frame changes in relation to that of its parent frame.


30.3.4 Frame Size

The canonical way to specify the size of a frame from within Emacs is by specifying its text size—a tuple of the width and height of the frame’s text area (see Frame Layout). It can be measured either in pixels or in terms of the frame’s canonical character size (see Frame Font).

For frames with an internal menu or tool bar, the frame’s native height cannot be told exactly before the frame has been actually drawn. This means that in general you cannot use the native size to specify the initial size of a frame. As soon as you know the native size of a visible frame, you can calculate its outer size (see Frame Layout) by adding in the remaining components from the return value of frame-geometry. For invisible frames or for frames that have yet to be created, however, the outer size can only be estimated. This also means that calculating an exact initial position of a frame specified via offsets from the right or bottom edge of the screen (see Frame Position) is impossible.

The text size of any frame can be set and retrieved with the help of the height and width frame parameters (see Size Parameters). The text size of the initial frame can be also set with the help of an X-style geometry specification. See Command Line Arguments for Emacs Invocation in The GNU Emacs Manual. Below we list some functions to access and set the size of an existing, visible frame, by default the selected one.

Function: frame-height &optional frame
Function: frame-width &optional frame

These functions return the height and width of the text area of frame, measured in units of the default font height and width of frame (see Frame Font). These functions are plain shorthands for writing (frame-parameter frame 'height) and (frame-parameter frame 'width).

If the text area of frame measured in pixels is not a multiple of its default font size, the values returned by these functions are rounded down to the number of characters of the default font that fully fit into the text area.

The functions following next return the pixel widths and heights of the native, outer and inner frame and the text area (see Frame Layout) of a given frame. For a text terminal, the results are in characters rather than pixels.

Function: frame-outer-width &optional frame
Function: frame-outer-height &optional frame

These functions return the outer width and height of frame in pixels.

Function: frame-native-height &optional frame
Function: frame-native-width &optional frame

These functions return the native width and height of frame in pixels.

Function: frame-inner-width &optional frame
Function: frame-inner-height &optional frame

These functions return the inner width and height of frame in pixels.

Function: frame-text-width &optional frame
Function: frame-text-height &optional frame

These functions return the width and height of the text area of frame in pixels.

On window systems that support it, Emacs tries by default to make the text size of a frame measured in pixels a multiple of the frame’s character size. This, however, usually means that a frame can be resized only in character size increments when dragging its external borders. It also may break attempts to truly maximize the frame or making it “fullheight” or “fullwidth” (see Size Parameters) leaving some empty space below and/or on the right of the frame. The following option may help in that case.

User Option: frame-resize-pixelwise

If this option is nil (the default), a frame’s text pixel size is usually rounded to a multiple of the current values of that frame’s frame-char-height and frame-char-width whenever the frame is resized. If this is non-nil, no rounding occurs, hence frame sizes can increase/decrease by one pixel.

Setting this variable usually causes the next resize operation to pass the corresponding size hints to the window manager. This means that this variable should be set only in a user’s initial file; applications should never bind it temporarily.

The precise meaning of a value of nil for this option depends on the toolkit used. Dragging the external border with the mouse is done character-wise provided the window manager is willing to process the corresponding size hints. Calling set-frame-size (see below) with arguments that do not specify the frame size as an integer multiple of its character size, however, may: be ignored, cause a rounding (GTK+), or be accepted (Lucid, Motif, MS-Windows).

With some window managers you may have to set this to non-nil in order to make a frame appear truly maximized or full-screen.

Function: set-frame-size frame width height &optional pixelwise

This function sets the size of the text area of frame, measured in terms of the canonical height and width of a character on frame (see Frame Font).

The optional argument pixelwise non-nil means to measure the new width and height in units of pixels instead. Note that if frame-resize-pixelwise is nil, some toolkits may refuse to truly honor the request if it does not increase/decrease the frame size to a multiple of its character size.

Function: set-frame-height frame height &optional pretend pixelwise

This function resizes the text area of frame to a height of height lines. The sizes of existing windows in frame are altered proportionally to fit.

If pretend is non-nil, then Emacs displays height lines of output in frame, but does not change its value for the actual height of the frame. This is only useful on text terminals. Using a smaller height than the terminal actually implements may be useful to reproduce behavior observed on a smaller screen, or if the terminal malfunctions when using its whole screen. Setting the frame height directly does not always work, because knowing the correct actual size may be necessary for correct cursor positioning on text terminals.

The optional fourth argument pixelwise non-nil means that frame should be height pixels high. Note that if frame-resize-pixelwise is nil, some window managers may refuse to truly honor the request if it does not increase/decrease the frame height to a multiple of its character height.

When used interactively, this command will ask the user for the number of lines to set the height of the currently selected frame. You can also provide this value with a numeric prefix.

Function: set-frame-width frame width &optional pretend pixelwise

This function sets the width of the text area of frame, measured in characters. The argument pretend has the same meaning as in set-frame-height.

The optional fourth argument pixelwise non-nil means that frame should be width pixels wide. Note that if frame-resize-pixelwise is nil, some window managers may refuse to fully honor the request if it does not increase/decrease the frame width to a multiple of its character width.

When used interactively, this command will ask the user for the number of columns to set the width of the currently selected frame. You can also provide this value with a numeric prefix.

None of these three functions will make a frame smaller than needed to display all of its windows together with their scroll bars, fringes, margins, dividers, mode and header lines. This contrasts with requests by the window manager triggered, for example, by dragging the external border of a frame with the mouse. Such requests are always honored by clipping, if necessary, portions that cannot be displayed at the right, bottom corner of the frame. The parameters min-width and min-height (see Size Parameters) can be used to obtain a similar behavior when changing the frame size from within Emacs.

The abnormal hook window-size-change-functions (see Hooks for Window Scrolling and Changes) tracks all changes of the inner size of a frame including those induced by request of the window-system or window manager. To rule out false positives that might occur when changing only the sizes of a frame’s windows without actually changing the size of the inner frame, use the following function.

Function: frame-size-changed-p &optional frame

This function returns non-nil when the inner width or height of frame has changed since window-size-change-functions was run the last time for frame. It always returns nil immediately after running window-size-change-functions for frame.


30.3.5 Implied Frame Resizing

By default, Emacs tries to keep the number of lines and columns of a frame’s text area unaltered when, for example, toggling its menu or tool bar, changing its default font or setting the width of any of its scroll bars. This means that in such case Emacs must ask the window manager to resize the frame’s window in order to accommodate the size change.

Occasionally, such implied frame resizing may be unwanted, for example, when a frame has been maximized or made full-screen (where it’s turned off by default). In general, users can disable implied resizing with the following option:

User Option: frame-inhibit-implied-resize

If this option is nil, changing a frame’s font, menu bar, tool bar, internal borders, fringes or scroll bars may resize its outer frame in order to keep the number of columns or lines of its text area unaltered. If this option is t, no such resizing is done.

The value of this option can be also a list of frame parameters. In that case, implied resizing is inhibited for the change of a parameter that appears in this list. Parameters currently handled by this option are font, font-backend, internal-border-width, menu-bar-lines and tool-bar-lines.

Changing any of the scroll-bar-width, scroll-bar-height, vertical-scroll-bars, horizontal-scroll-bars, left-fringe and right-fringe frame parameters is handled as if the frame contained just one live window. This means, for example, that removing vertical scroll bars on a frame containing several side by side windows will shrink the outer frame width by the width of one scroll bar provided this option is nil and keep it unchanged if this option is t or a list containing vertical-scroll-bars.

The default value is (tab-bar-lines tool-bar-lines) for Lucid, Motif and MS-Windows (which means that adding/removing a tool or tab bar there does not change the outer frame height), (tab-bar-lines) on all other window systems including GTK+ (which means that changing any of the parameters listed above with the exception of tab-bar-lines may change the size of the outer frame), and t otherwise (which means the outer frame size never changes implicitly when there’s no window system support).

Note that when a frame is not large enough to accommodate a change of any of the parameters listed above, Emacs may try to enlarge the frame even if this option is non-nil.

Note also that window managers usually do not ask for resizing a frame when they change the number of lines occupied by an external menu or tool bar. Typically, such “wrappings” occur when a user shrinks a frame horizontally, making it impossible to display all elements of its menu or tool bar. They may also result from a change of the major mode altering the number of items of a menu or tool bar. Any such wrappings may implicitly alter the number of lines of a frame’s text area and are unaffected by the setting of this option.


30.4 Frame Parameters

A frame has many parameters that control its appearance and behavior. Just what parameters a frame has depends on what display mechanism it uses.

Frame parameters exist mostly for the sake of graphical displays. Most frame parameters have no effect when applied to a frame on a text terminal; only the height, width, name, title, menu-bar-lines, buffer-list and buffer-predicate parameters do something special. If the terminal supports colors, the parameters foreground-color, background-color, background-mode and display-type are also meaningful. If the terminal supports frame transparency, the parameter alpha is also meaningful.

By default, frame parameters are saved and restored by the desktop library functions (see Desktop Save Mode) when the variable desktop-restore-frames is non-nil. It’s the responsibility of applications that their parameters are included in frameset-persistent-filter-alist to avoid that they get meaningless or even harmful values in restored sessions.


30.4.1 Access to Frame Parameters

These functions let you read and change the parameter values of a frame.

Function: frame-parameter frame parameter

This function returns the value of the parameter parameter (a symbol) of frame. If frame is nil, it returns the selected frame’s parameter. If frame has no setting for parameter, this function returns nil.

Function: frame-parameters &optional frame

The function frame-parameters returns an alist listing all the parameters of frame and their values. If frame is nil or omitted, this returns the selected frame’s parameters

Function: modify-frame-parameters frame alist

This function alters the frame frame based on the elements of alist. Each element of alist has the form (parm . value), where parm is a symbol naming a parameter. If you don’t mention a parameter in alist, its value doesn’t change. If frame is nil, it defaults to the selected frame.

Some parameters are only meaningful for frames on certain kinds of display (see Frames). If alist includes parameters that are not meaningful for the frame’s display, this function will change its value in the frame’s parameter list, but will otherwise ignore it.

When alist specifies more than one parameter whose value can affect the new size of frame, the final size of the frame may differ according to the toolkit used. For example, specifying that a frame should from now on have a menu and/or tool bar instead of none and simultaneously specifying the new height of the frame will inevitably lead to a recalculation of the frame’s height. Conceptually, in such case, this function will try to have the explicit height specification prevail. It cannot be excluded, however, that the addition (or removal) of the menu or tool bar, when eventually performed by the toolkit, will defeat this intention.

Sometimes, binding frame-inhibit-implied-resize (see Implied Frame Resizing) to a non-nil value around calls to this function may fix the problem sketched here. Sometimes, however, exactly such binding may be hit by the problem.

Function: set-frame-parameter frame parm value

This function sets the frame parameter parm to the specified value. If frame is nil, it defaults to the selected frame.

Function: modify-all-frames-parameters alist

This function alters the frame parameters of all existing frames according to alist, then modifies default-frame-alist (and, if necessary, initial-frame-alist) to apply the same parameter values to frames that will be created henceforth.


30.4.2 Initial Frame Parameters

You can specify the parameters for the initial startup frame by setting initial-frame-alist in your init file (see The Init File).

User Option: initial-frame-alist

This variable’s value is an alist of parameter values used when creating the initial frame. You can set this variable to specify the appearance of the initial frame without altering subsequent frames. Each element has the form:

(parameter . value)

Emacs creates the initial frame before it reads your init file. After reading that file, Emacs checks initial-frame-alist, and applies the parameter settings in the altered value to the already created initial frame.

If these settings affect the frame geometry and appearance, you’ll see the frame appear with the wrong ones and then change to the specified ones. If that bothers you, you can specify the same geometry and appearance with X resources; those do take effect before the frame is created. See X Resources in The GNU Emacs Manual.

X resource settings typically apply to all frames. If you want to specify some X resources solely for the sake of the initial frame, and you don’t want them to apply to subsequent frames, here’s how to achieve this. Specify parameters in default-frame-alist to override the X resources for subsequent frames; then, to prevent these from affecting the initial frame, specify the same parameters in initial-frame-alist with values that match the X resources.

If these parameters include (minibuffer . nil), that indicates that the initial frame should have no minibuffer. In this case, Emacs creates a separate minibuffer-only frame as well.

User Option: minibuffer-frame-alist

This variable’s value is an alist of parameter values used when creating an initial minibuffer-only frame (i.e., the minibuffer-only frame that Emacs creates if initial-frame-alist specifies a frame with no minibuffer).

User Option: default-frame-alist

This is an alist specifying default values of frame parameters for all Emacs frames—the first frame, and subsequent frames. When using the X Window System, you can get the same results by means of X resources in many cases.

Setting this variable does not affect existing frames. Furthermore, functions that display a buffer in a separate frame may override the default parameters by supplying their own parameters.

If you invoke Emacs with command-line options that specify frame appearance, those options take effect by adding elements to either initial-frame-alist or default-frame-alist. Options which affect just the initial frame, such as ‘--geometry’ and ‘--maximized’, add to initial-frame-alist; the others add to default-frame-alist. see Command Line Arguments for Emacs Invocation in The GNU Emacs Manual.


30.4.3 Window Frame Parameters

Just what parameters a frame has depends on what display mechanism it uses. This section describes the parameters that have special meanings on some or all kinds of terminals. Of these, name, title, height, width, buffer-list and buffer-predicate provide meaningful information in terminal frames, and tty-color-mode is meaningful only for frames on text terminals.


30.4.3.1 Basic Parameters

These frame parameters give the most basic information about the frame. title and name are meaningful on all terminals.

display

The display on which to open this frame. It should be a string of the form ‘host:dpy.screen’, just like the DISPLAY environment variable. See Multiple Terminals, for more details about display names.

display-type

This parameter describes the range of possible colors that can be used in this frame. Its value is color, grayscale or mono.

title

If a frame has a non-nil title, that title appears in the window system’s title bar at the top of the frame, and also in the mode line of windows in that frame if mode-line-frame-identification uses ‘%F’ (see %-Constructs in the Mode Line). This is normally the case when Emacs is not using a window system, and can only display one frame at a time. When Emacs is using a window system, this parameter, if non-nil, overrides the title determined by the name parameter and the implicit title calculated according to frame-title-format. It also overrides the title determined by icon-title-format for iconified frames. See Frame Titles.

name

The name of the frame. If you don’t specify a name via this parameter, Emacs sets the frame name automatically, as specified by frame-title-format and icon-title-format, and that is the frame’s title that will appear on display when Emacs uses a window system (unless the title parameter overrides it).

If you specify the frame name explicitly when you create the frame, the name is also used (instead of the name of the Emacs executable) when looking up X resources for the frame.

explicit-name

If the frame name was specified explicitly when the frame was created, this parameter will be that name. If the frame wasn’t explicitly named, this parameter will be nil.


30.4.3.2 Position Parameters

Parameters describing the X- and Y-offsets of a frame are always measured in pixels. For a normal, non-child frame they specify the frame’s outer position (see Frame Geometry) relative to its display’s origin. For a child frame (see Child Frames) they specify the frame’s outer position relative to the native position of the frame’s parent frame. (Note that none of these parameters is meaningful on TTY frames.)

left

The position, in pixels, of the left outer edge of the frame with respect to the left edge of the frame’s display or parent frame. It can be specified in one of the following ways.

an integer

A positive integer always relates the left edge of the frame to the left edge of its display or parent frame. A negative integer relates the right frame edge to the right edge of the display or parent frame.

(+ pos)

This specifies the position of the left frame edge relative to the left edge of its display or parent frame. The integer pos may be positive or negative; a negative value specifies a position outside the screen or parent frame or on a monitor other than the primary one (for multi-monitor displays).

(- pos)

This specifies the position of the right frame edge relative to the right edge of the display or parent frame. The integer pos may be positive or negative; a negative value specifies a position outside the screen or parent frame or on a monitor other than the primary one (for multi-monitor displays).

a floating-point value

A floating-point value in the range 0.0 to 1.0 specifies the left edge’s offset via the left position ratio of the frame—the ratio of the left edge of its outer frame to the width of the frame’s workarea (see Multiple Terminals) or its parent’s native frame (see Child Frames) minus the width of the outer frame. Thus, a left position ratio of 0.0 flushes a frame to the left, a ratio of 0.5 centers it and a ratio of 1.0 flushes it to the right of its display or parent frame. Similarly, the top position ratio of a frame is the ratio of the frame’s top position to the height of its workarea or parent frame minus the height of the frame.

Emacs will try to keep the position ratios of a child frame unaltered if that frame has a non-nil keep-ratio parameter (see Frame Interaction Parameters) and its parent frame is resized.

Since the outer size of a frame (see Frame Geometry) is usually unavailable before a frame has been made visible, it is generally not advisable to use floating-point values when creating decorated frames. Floating-point values are more suited for ensuring that an (undecorated) child frame is positioned nicely within the area of its parent frame.

Some window managers ignore program-specified positions. If you want to be sure the position you specify is not ignored, specify a non-nil value for the user-position parameter as in the following example:

(modify-frame-parameters
  nil '((user-position . t) (left . (+ -4))))

In general, it is not a good idea to position a frame relative to the right or bottom edge of its display. Positioning the initial or a new frame is either not accurate (because the size of the outer frame is not yet fully known before the frame has been made visible) or will cause additional flicker (if the frame has to be repositioned after becoming visible).

Note also, that positions specified relative to the right/bottom edge of a display, workarea or parent frame as well as floating-point offsets are stored internally as integer offsets relative to the left/top edge of the display, workarea or parent frame edge. They are also returned as such by functions like frame-parameters and restored as such by the desktop saving routines.

top

The screen position of the top (or bottom) edge, in pixels, with respect to the top (or bottom) edge of the display or parent frame. It works just like left, except vertically instead of horizontally.

icon-left

The screen position of the left edge of the frame’s icon, in pixels, counting from the left edge of the screen. This takes effect when the frame is iconified, if the window manager supports this feature. If you specify a value for this parameter, then you must also specify a value for icon-top and vice versa.

icon-top

The screen position of the top edge of the frame’s icon, in pixels, counting from the top edge of the screen. This takes effect when the frame is iconified, if the window manager supports this feature.

user-position

When you create a frame and specify its screen position with the left and top parameters, use this parameter to say whether the specified position was user-specified (explicitly requested in some way by a human user) or merely program-specified (chosen by a program). A non-nil value says the position was user-specified.

Window managers generally heed user-specified positions, and some heed program-specified positions too. But many ignore program-specified positions, placing the window in a default fashion or letting the user place it with the mouse. Some window managers, including twm, let the user specify whether to obey program-specified positions or ignore them.

When you call make-frame, you should specify a non-nil value for this parameter if the values of the left and top parameters represent the user’s stated preference; otherwise, use nil.

z-group

This parameter specifies a relative position of the frame’s window-system window in the stacking (Z-) order of the frame’s display.

If this is above, the window-system will display the window that corresponds to the frame above all other window-system windows that do not have the above property set. If this is nil, the frame’s window is displayed below all windows that have the above property set and above all windows that have the below property set. If this is below, the frame’s window is displayed below all windows that do not have the below property set.

To position the frame above or below a specific other frame use the function frame-restack (see Raising, Lowering and Restacking Frames).


30.4.3.3 Size Parameters

Frame parameters usually specify frame sizes in character units. On graphical displays, the default face determines the actual pixel sizes of these character units (see Face Attributes).

width

This parameter specifies the width of the frame. It can be specified as in the following ways:

an integer

A positive integer specifies the width of the frame’s text area (see Frame Geometry) in characters.

a cons cell

If this is a cons cell with the symbol text-pixels in its CAR, the CDR of that cell specifies the width of the frame’s text area in pixels.

a floating-point value

A floating-point number between 0.0 and 1.0 can be used to specify the width of a frame via its width ratio—the ratio of its outer width (see Frame Geometry) to the width of the frame’s workarea (see Multiple Terminals) or its parent frame’s (see Child Frames) native frame. Thus, a value of 0.5 makes the frame occupy half of the width of its workarea or parent frame, a value of 1.0 the full width. Similarly, the height ratio of a frame is the ratio of its outer height to the height of its workarea or its parent’s native frame.

Emacs will try to keep the width and height ratio of a child frame unaltered if that frame has a non-nil keep-ratio parameter (see Frame Interaction Parameters) and its parent frame is resized.

Since the outer size of a frame is usually unavailable before a frame has been made visible, it is generally not advisable to use floating-point values when creating decorated frames. Floating-point values are more suited to ensure that a child frame always fits within the area of its parent frame as, for example, when customizing display-buffer-alist (see Choosing a Window for Displaying a Buffer) via display-buffer-in-child-frame.

Regardless of how this parameter was specified, functions reporting the value of this parameter like frame-parameters always report the width of the frame’s text area in characters as an integer rounded, if necessary, to a multiple of the frame’s default character width. That value is also used by the desktop saving routines.

height

This parameter specifies the height of the frame. It works just like width, except vertically instead of horizontally.

user-size

This does for the size parameters height and width what the user-position parameter (see user-position) does for the position parameters top and left.

min-width

This parameter specifies the minimum native width (see Frame Geometry) of the frame, in characters. Normally, the functions that establish a frame’s initial width or resize a frame horizontally make sure that all the frame’s windows, vertical scroll bars, fringes, margins and vertical dividers can be displayed. This parameter, if non-nil allows to make a frame narrower than that with the consequence that any components that do not fit will be clipped by the window manager.

min-height

This parameter specifies the minimum native height (see Frame Geometry) of the frame, in characters. Normally, the functions that establish a frame’s initial size or resize a frame make sure that all the frame’s windows, horizontal scroll bars and dividers, mode and header lines, the echo area and the internal menu and tool bar can be displayed. This parameter, if non-nil allows to make a frame smaller than that with the consequence that any components that do not fit will be clipped by the window manager.

fullscreen

This parameter specifies whether to maximize the frame’s width, height or both. Its value can be fullwidth, fullheight, fullboth, or maximized.21 A fullwidth frame is as wide as possible, a fullheight frame is as tall as possible, and a fullboth frame is both as wide and as tall as possible. A maximized frame is like a “fullboth” frame, except that it usually keeps its title bar and the buttons for resizing and closing the frame. Also, maximized frames typically avoid hiding any task bar or panels displayed on the desktop. A “fullboth” frame, on the other hand, usually omits the title bar and occupies the entire available screen space.

Full-height and full-width frames are more similar to maximized frames in this regard. However, these typically display an external border which might be absent with maximized frames. Hence the heights of maximized and full-height frames and the widths of maximized and full-width frames often differ by a few pixels.

With some window managers you may have to customize the variable frame-resize-pixelwise (see Frame Size) in order to make a frame truly appear maximized or full-screen. Moreover, some window managers might not support smooth transition between the various full-screen or maximization states. Customizing the variable x-frame-normalize-before-maximize can help to overcome that.

Full-screen on macOS hides both the tool-bar and the menu-bar, however both will be displayed if the mouse pointer is moved to the top of the screen.

fullscreen-restore

This parameter specifies the desired fullscreen state of the frame after invoking the toggle-frame-fullscreen command (see Frame Commands in The GNU Emacs Manual) in the “fullboth” state. Normally this parameter is installed automatically by that command when toggling the state to fullboth. If, however, you start Emacs in the “fullboth” state, you have to specify the desired behavior in your initial file as, for example

(setq default-frame-alist
    '((fullscreen . fullboth)
      (fullscreen-restore . fullheight)))

This will give a new frame full height after typing in it F11 for the first time.

fit-frame-to-buffer-margins

This parameter allows to override the value of the option fit-frame-to-buffer-margins when fitting this frame to the buffer of its root window with fit-frame-to-buffer (see Resizing Windows).

fit-frame-to-buffer-sizes

This parameter allows to override the value of the option fit-frame-to-buffer-sizes when fitting this frame to the buffer of its root window with fit-frame-to-buffer (see Resizing Windows).


30.4.3.4 Layout Parameters

These frame parameters enable or disable various parts of the frame, or control their sizes.

border-width

The width in pixels of the frame’s outer border (see Frame Geometry).

internal-border-width

The width in pixels of the frame’s internal border (see Frame Geometry).

child-frame-border-width

The width in pixels of the frame’s internal border (see Frame Geometry) if the given frame is a child frame (see Child Frames). If this is nil, the value specified by the internal-border-width parameter is used instead.

vertical-scroll-bars

Whether the frame has scroll bars (see Scroll Bars) for vertical scrolling, and which side of the frame they should be on. The possible values are left, right, and nil for no scroll bars.

horizontal-scroll-bars

Whether the frame has scroll bars for horizontal scrolling (t and bottom mean yes, nil means no).

scroll-bar-width

The width of vertical scroll bars, in pixels, or nil meaning to use the default width.

scroll-bar-height

The height of horizontal scroll bars, in pixels, or nil meaning to use the default height.

left-fringe
right-fringe

The default width of the left and right fringes of windows in this frame (see Fringes). If either of these is zero, that effectively removes the corresponding fringe.

When you use frame-parameter to query the value of either of these two frame parameters, the return value is always an integer. When using set-frame-parameter, passing a nil value imposes an actual default value of 8 pixels.

right-divider-width

The width (thickness) reserved for the right divider (see Window Dividers) of any window on the frame, in pixels. A value of zero means to not draw right dividers.

bottom-divider-width

The width (thickness) reserved for the bottom divider (see Window Dividers) of any window on the frame, in pixels. A value of zero means to not draw bottom dividers.

menu-bar-lines

The number of lines to allocate at the top of the frame for a menu bar (see The Menu Bar). The default is one if Menu Bar mode is enabled and zero otherwise. See Menu Bars in The GNU Emacs Manual. For an external menu bar (see Frame Layout), this value remains unchanged even when the menu bar wraps to two or more lines. In that case, the menu-bar-size value returned by frame-geometry (see Frame Geometry) allows to derive whether the menu bar actually occupies one or more lines.

tool-bar-lines

The number of lines to use for the tool bar (see Tool bars). The default is one if Tool Bar mode is enabled and zero otherwise. See Tool Bars in The GNU Emacs Manual. This value may change whenever the tool bar wraps (see Frame Layout).

tool-bar-position

The position of the tool bar when Emacs was built with GTK+. Its value can be one of top, bottom left, right. The default is top.

tab-bar-lines

The number of lines to use for the tab bar (see Tab Bars in The GNU Emacs Manual). The default is one if Tab Bar mode is enabled and zero otherwise. This value may change whenever the tab bar wraps (see Frame Layout).

line-spacing

Additional space to leave below each text line, in pixels (a positive integer). See Line Height, for more information.

no-special-glyphs

If this is non-nil, it suppresses the display of any truncation and continuation glyphs (see Truncation) for all buffers displayed by this frame. This is useful to eliminate such glyphs when fitting a frame to its buffer via fit-frame-to-buffer (see Resizing Windows).


30.4.3.5 Buffer Parameters

These frame parameters, meaningful on all kinds of terminals, deal with which buffers have been, or should, be displayed in the frame.

minibuffer

Whether this frame has its own minibuffer. The value t means yes, nil means no, only means this frame is just a minibuffer. If the value is a minibuffer window (in some other frame), the frame uses that minibuffer.

This parameter takes effect when the frame is created. If specified as nil, Emacs will try to set it to the minibuffer window of default-minibuffer-frame (see Minibuffers and Frames). For an existing frame, this parameter can be used exclusively to specify another minibuffer window. It is not allowed to change it from a minibuffer window to t and vice-versa, or from t to nil. If the parameter specifies a minibuffer window already, setting it to nil has no effect.

The special value child-frame means to make a minibuffer-only child frame (see Child Frames) whose parent becomes the frame created. As if specified as nil, Emacs will set this parameter to the minibuffer window of the child frame but will not select the child frame after its creation.

buffer-predicate

The buffer-predicate function for this frame. The function other-buffer uses this predicate (from the selected frame) to decide which buffers it should consider, if the predicate is not nil. It calls the predicate with one argument, a buffer, once for each buffer; if the predicate returns a non-nil value, it considers that buffer.

buffer-list

A list of buffers that have been selected in this frame, ordered most-recently-selected first.

unsplittable

If non-nil, this frame’s window is never split automatically.


30.4.3.6 Frame Interaction Parameters

These parameters supply forms of interactions between different frames.

parent-frame

If non-nil, this means that this frame is a child frame (see Child Frames), and this parameter specifies its parent frame. If nil, this means that this frame is a normal, top-level frame.

delete-before

If non-nil, this parameter specifies another frame whose deletion will automatically trigger the deletion of this frame. See Deleting Frames.

mouse-wheel-frame

If non-nil, this parameter specifies the frame whose windows will be scrolled whenever the mouse wheel is scrolled with the mouse pointer hovering over this frame, see Mouse Commands in The GNU Emacs Manual.

no-other-frame

If this is non-nil, then this frame is not eligible as candidate for the functions next-frame, previous-frame (see Finding All Frames) and other-frame, see Frame Commands in The GNU Emacs Manual.

auto-hide-function

When this parameter specifies a function, that function will be called instead of the function specified by the variable frame-auto-hide-function when quitting the frame’s only window (see Quitting Windows) and there are other frames left.

minibuffer-exit

When this parameter is non-nil, Emacs will by default make this frame invisible whenever the minibuffer (see Minibuffers) is exited. Alternatively, it can specify the functions iconify-frame and delete-frame. This parameter is useful to make a child frame disappear automatically (similar to how Emacs deals with a window) when exiting the minibuffer.

keep-ratio

This parameter is currently meaningful for child frames (see Child Frames) only. If it is non-nil, then Emacs will try to keep the frame’s size (width and height) ratios (see Size Parameters) as well as its left and right position ratios (see Position Parameters) unaltered whenever its parent frame is resized.

If the value of this parameter is nil, the frame’s position and size remain unaltered when the parent frame is resized, so the position and size ratios may change. If the value of this parameter is t, Emacs will try to preserve the frame’s size and position ratios, hence the frame’s size and position relative to its parent frame may change.

More individual control is possible by using a cons cell: In that case the frame’s width ratio is preserved if the CAR of the cell is either t or width-only. The height ratio is preserved if the CAR of the cell is either t or height-only. The left position ratio is preserved if the CDR of the cell is either t or left-only. The top position ratio is preserved if the CDR of the cell is either t or top-only.


30.4.3.7 Mouse Dragging Parameters

The parameters described below provide support for resizing a frame by dragging its internal borders with the mouse. They also allow moving a frame with the mouse by dragging the header or tab line of its topmost or the mode line of its bottommost window.

These parameters are mostly useful for child frames (see Child Frames) that come without window manager decorations. If necessary, they can be used for undecorated top-level frames as well.

drag-internal-border

If non-nil, the frame can be resized by dragging its internal borders, if present, with the mouse.

drag-with-header-line

If non-nil, the frame can be moved with the mouse by dragging the header line of its topmost window.

drag-with-tab-line

If non-nil, the frame can be moved with the mouse by dragging the tab line of its topmost window.

drag-with-mode-line

If non-nil, the frame can be moved with the mouse by dragging the mode line of its bottommost window. Note that such a frame is not allowed to have its own minibuffer window.

snap-width

A frame that is moved with the mouse will “snap” at the border(s) of the display or its parent frame whenever it is dragged as near to such an edge as the number of pixels specified by this parameter.

top-visible

If this parameter is a number, the top edge of the frame never appears above the top edge of its display or parent frame. Moreover, as many pixels of the frame as specified by that number will remain visible when the frame is moved against any of the remaining edges of its display or parent frame. Setting this parameter is useful to guard against dragging a child frame with a non-nil drag-with-header-line parameter completely out of the area of its parent frame.

bottom-visible

If this parameter is a number, the bottom edge of the frame never appears below the bottom edge of its display or parent frame. Moreover, as many pixels of the frame as specified by that number will remain visible when the frame is moved against any of the remaining edges of its display or parent frame. Setting this parameter is useful to guard against dragging a child frame with a non-nil drag-with-mode-line parameter completely out of the area of its parent frame.


30.4.3.8 Window Management Parameters

The following frame parameters control various aspects of the frame’s interaction with the window manager or window system. They have no effect on text terminals.

visibility

The state of visibility of the frame. There are three possibilities: nil for invisible, t for visible, and icon for iconified. See Visibility of Frames.

auto-raise

If non-nil, Emacs automatically raises the frame when it is selected. Some window managers do not allow this.

auto-lower

If non-nil, Emacs automatically lowers the frame when it is deselected. Some window managers do not allow this.

icon-type

The type of icon to use for this frame. If the value is a string, that specifies a file containing a bitmap to use; nil specifies no icon (in which case the window manager decides what to show); any other non-nil value specifies the default Emacs icon.

icon-name

The name to use in the icon for this frame, when and if the icon appears. If this is nil, the frame’s title is used.

window-id

The ID number which the graphical display uses for this frame. Emacs assigns this parameter when the frame is created; changing the parameter has no effect on the actual ID number.

outer-window-id

The ID number of the outermost window-system window in which the frame exists. As with window-id, changing this parameter has no actual effect.

wait-for-wm

If non-nil, tell Xt to wait for the window manager to confirm geometry changes. Some window managers, including versions of Fvwm2 and KDE, fail to confirm, so Xt hangs. Set this to nil to prevent hanging with those window managers.

sticky

If non-nil, the frame is visible on all virtual desktops on systems with virtual desktops.

shaded

If non-nil, tell the window manager to display the frame in a way that its contents are hidden, leaving only the title bar.

use-frame-synchronization

If non-nil, synchronize the frame redisplay with the refresh rate of the monitor to avoid graphics tearing. At present, this is only implemented on Haiku and the X window system inside no-toolkit and X toolkit builds, does not work correctly with toolkit scroll bars, and requires a compositing manager supporting the relevant display synchronization protocols. The synchronizeResize X resource must also be set to the string "extended".

inhibit-double-buffering

If non-nil, the frame is drawn to the screen without double buffering. Emacs normally attempts to use double buffering, where available, to reduce flicker. Set this property if you experience display bugs or pine for that retro, flicker-y feeling.

skip-taskbar

If non-nil, this tells the window manager to remove the frame’s icon from the taskbar associated with the frame’s display and inhibit switching to the frame’s window via the combination Alt-TAB. On MS-Windows, iconifying such a frame will "roll in" its window-system window at the bottom of the desktop. Some window managers may not honor this parameter.

no-focus-on-map

If non-nil, this means that the frame does not want to receive input focus when it is mapped (see Visibility of Frames). Some window managers may not honor this parameter.

no-accept-focus

If non-nil, this means that the frame does not want to receive input focus via explicit mouse clicks or when moving the mouse into it either via focus-follows-mouse (see Input Focus) or mouse-autoselect-window (see Mouse Window Auto-selection). This may have the unwanted side-effect that a user cannot scroll a non-selected frame with the mouse. Some window managers may not honor this parameter. On Haiku, it also has the side-effect that the window will not be able to receive any keyboard input from the user, not even if the user switches to the frame using the key combination Alt-TAB.

undecorated

If non-nil, this frame’s window-system window is drawn without decorations, like the title, minimize/maximize boxes and external borders. This usually means that the window cannot be dragged, resized, iconified, maximized or deleted with the mouse. If nil, the frame’s window is usually drawn with all the elements listed above unless their display has been suspended via window manager settings.

Under X, Emacs uses the Motif window manager hints to turn off decorations. Some window managers may not honor these hints.

NS builds consider the tool bar to be a decoration, and therefore hide it on an undecorated frame.

override-redirect

If non-nil, this means that this is an override redirect frame—a frame not handled by window managers under X. Override redirect frames have no window manager decorations, can be positioned and resized only via Emacs’ positioning and resizing functions and are usually drawn on top of all other frames. Setting this parameter has no effect on MS-Windows.

ns-appearance

Only available on macOS, if set to dark draw this frame’s window-system window using the “vibrant dark” theme, and if set to light use the “aqua” theme, otherwise use the system default. The “vibrant dark” theme can be used to set the toolbar and scrollbars to a dark appearance when using an Emacs theme with a dark background.

ns-transparent-titlebar

Only available on macOS, if non-nil, set the titlebar and toolbar to be transparent. This effectively sets the background color of both to match the Emacs background color.


30.4.3.9 Cursor Parameters

This frame parameter controls the way the cursor looks.

cursor-type

How to display the cursor. Legitimate values are:

box

Display a filled box. (This is the default.)

(box . size)

Display a filled box. However, display it as a hollow box if point is under masked image larger than size pixels in either dimension.

hollow

Display a hollow box.

nil

Don’t display a cursor.

bar

Display a vertical bar between characters.

(bar . width)

Display a vertical bar width pixels wide between characters.

hbar

Display a horizontal bar.

(hbar . height)

Display a horizontal bar height pixels high.

The cursor-type frame parameter may be overridden by the variables cursor-type and cursor-in-non-selected-windows:

User Option: cursor-type

This buffer-local variable controls how the cursor looks in a selected window showing the buffer. If its value is t, that means to use the cursor specified by the cursor-type frame parameter. Otherwise, the value should be one of the cursor types listed above, and it overrides the cursor-type frame parameter.

User Option: cursor-in-non-selected-windows

This buffer-local variable controls how the cursor looks in a window that is not selected. It supports the same values as the cursor-type frame parameter; also, nil means don’t display a cursor in nonselected windows, and t (the default) means use a standard modification of the usual cursor type (solid box becomes hollow box, and bar becomes a narrower bar).

User Option: x-stretch-cursor

This variable controls the width of the block cursor displayed on extra-wide glyphs such as a tab or a stretch of white space. By default, the block cursor is only as wide as the font’s default character, and will not cover all of the width of the glyph under it if that glyph is extra-wide. A non-nil value of this variable means draw the block cursor as wide as the glyph under it. The default value is nil.

This variable has no effect on text-mode frames, since the text-mode cursor is drawn by the terminal out of Emacs’s control.

This variable specifies how to blink the cursor. Each element has the form (on-state . off-state). Whenever the cursor type equals on-state (comparing using equal), the corresponding off-state specifies what the cursor looks like when it blinks off. Both on-state and off-state should be suitable values for the cursor-type frame parameter.

There are various defaults for how to blink each type of cursor, if the type is not mentioned as an on-state here. Changes in this variable do not take effect immediately, only when you specify the cursor-type frame parameter.


30.4.3.10 Font and Color Parameters

These frame parameters control the use of fonts and colors.

font-backend

A list of symbols, specifying the font backends to use for drawing characters on the frame, in order of priority. In Emacs built without Cairo drawing on X, there are currently three potentially available font backends: x (the X core font driver), xft (the Xft font driver), and xfthb (the Xft font driver with HarfBuzz text shaping). If built with Cairo drawing, there are also three potentially available font backends on X: x, ftcr (the FreeType font driver on Cairo), and ftcrhb (the FreeType font driver on Cairo with HarfBuzz text shaping). When Emacs is built with HarfBuzz, the default font driver is ftcrhb, although use of the ftcr driver is still possible, but not recommended. On MS-Windows, there are currently three available font backends: gdi (the core MS-Windows font driver), uniscribe (font driver for OTF and TTF fonts with text shaping by the Uniscribe engine), and harfbuzz (font driver for OTF and TTF fonts with HarfBuzz text shaping) (see Windows Fonts in The GNU Emacs Manual). The harfbuzz driver is similarly recommended. On Haiku, there can be several font drivers (see Haiku Fonts in The GNU Emacs Manual).

On other systems, there is only one available font backend, so it does not make sense to modify this frame parameter.

background-mode

This parameter is either dark or light, according to whether the background color is a light one or a dark one.

tty-color-mode

This parameter overrides the terminal’s color support as given by the system’s terminal capabilities database in that this parameter’s value specifies the color mode to use on a text terminal. The value can be either a symbol or a number. A number specifies the number of colors to use (and, indirectly, what commands to issue to produce each color). For example, (tty-color-mode . 8) specifies use of the ANSI escape sequences for 8 standard text colors. A value of −1 turns off color support.

If the parameter’s value is a symbol, it specifies a number through the value of tty-color-mode-alist, and the associated number is used instead.

screen-gamma

If this is a number, Emacs performs gamma correction which adjusts the brightness of all colors. The value should be the screen gamma of your display.

Usual PC monitors have a screen gamma of 2.2, so color values in Emacs, and in X windows generally, are calibrated to display properly on a monitor with that gamma value. If you specify 2.2 for screen-gamma, that means no correction is needed. Other values request correction, designed to make the corrected colors appear on your screen the way they would have appeared without correction on an ordinary monitor with a gamma value of 2.2.

If your monitor displays colors too light, you should specify a screen-gamma value smaller than 2.2. This requests correction that makes colors darker. A screen gamma value of 1.5 may give good results for LCD color displays.

alpha

This parameter specifies the opacity of the frame, on graphical displays that support variable opacity. It should be an integer between 0 and 100, where 0 means completely transparent and 100 means completely opaque. It can also have a nil value, which tells Emacs not to set the frame opacity (leaving it to the window manager).

To prevent the frame from disappearing completely from view, the variable frame-alpha-lower-limit defines a lower opacity limit. If the value of the frame parameter is less than the value of this variable, Emacs uses the latter. By default, frame-alpha-lower-limit is 20.

The alpha frame parameter can also be a cons cell (active . inactive), where active is the opacity of the frame when it is selected, and inactive is the opacity when it is not selected.

Some window systems do not support the alpha parameter for child frames (see Child Frames).

alpha-background

Sets the background transparency of the frame. Unlike the alpha frame parameter, this only controls the transparency of the background while keeping foreground elements such as text fully opaque. It should be an integer between 0 and 100, where 0 means completely transparent and 100 means completely opaque (default).

The following frame parameters are semi-obsolete in that they are automatically equivalent to particular face attributes of particular faces (see Standard Faces in The Emacs Manual):

font

The name of the font for displaying text in the frame. This is a string, either a valid font name for your system or the name of an Emacs fontset (see Fontsets). It is equivalent to the font attribute of the default face.

foreground-color

The color to use for characters. It is equivalent to the :foreground attribute of the default face.

background-color

The color to use for the background of characters. It is equivalent to the :background attribute of the default face.

mouse-color

The color for the mouse pointer. It is equivalent to the :background attribute of the mouse face.

cursor-color

The color for the cursor that shows point. It is equivalent to the :background attribute of the cursor face.

border-color

The color for the border of the frame. It is equivalent to the :background attribute of the border face.

scroll-bar-foreground

If non-nil, the color for the foreground of scroll bars. It is equivalent to the :foreground attribute of the scroll-bar face.

scroll-bar-background

If non-nil, the color for the background of scroll bars. It is equivalent to the :background attribute of the scroll-bar face.


30.4.4 Geometry

Here’s how to examine the data in an X-style window geometry specification:

Function: x-parse-geometry geom

The function x-parse-geometry converts a standard X window geometry string to an alist that you can use as part of the argument to make-frame.

The alist describes which parameters were specified in geom, and gives the values specified for them. Each element looks like (parameter . value). The possible parameter values are left, top, width, and height.

For the size parameters, the value must be an integer. The position parameter names left and top are not totally accurate, because some values indicate the position of the right or bottom edges instead. The value possibilities for the position parameters are: an integer, a list (+ pos), or a list (- pos); as previously described (see Position Parameters).

Here is an example:

(x-parse-geometry "35x70+0-0")
     ⇒ ((height . 70) (width . 35)
         (top - 0) (left . 0))

30.5 Terminal Parameters

Each terminal has a list of associated parameters. These terminal parameters are mostly a convenient way of storage for terminal-local variables, but some terminal parameters have a special meaning.

This section describes functions to read and change the parameter values of a terminal. They all accept as their argument either a terminal or a frame; the latter means use that frame’s terminal. An argument of nil means the selected frame’s terminal.

Function: terminal-parameters &optional terminal

This function returns an alist listing all the parameters of terminal and their values.

Function: terminal-parameter terminal parameter

This function returns the value of the parameter parameter (a symbol) of terminal. If terminal has no setting for parameter, this function returns nil.

Function: set-terminal-parameter terminal parameter value

This function sets the parameter parameter of terminal to the specified value, and returns the previous value of that parameter.

Here’s a list of a few terminal parameters that have a special meaning:

background-mode

The classification of the terminal’s background color, either light or dark.

normal-erase-is-backspace

Value is either 1 or 0, depending on whether normal-erase-is-backspace-mode is turned on or off on this terminal. See DEL Does Not Delete in The Emacs Manual.

terminal-initted

After the terminal is initialized, this is set to the terminal-specific initialization function.

tty-mode-set-strings

When present, a list of strings containing escape sequences that Emacs will output while configuring a tty for rendering. Emacs emits these strings only when configuring a terminal: if you want to enable a mode on a terminal that is already active (for example, while in tty-setup-hook), explicitly output the necessary escape sequence using send-string-to-terminal in addition to adding the sequence to tty-mode-set-strings.

tty-mode-reset-strings

When present, a list of strings that undo the effects of the strings in tty-mode-set-strings. Emacs emits these strings when exiting, deleting a terminal, or suspending itself.


30.6 Frame Titles

Every frame has a name parameter; this serves as the default for the frame title which window systems typically display at the top of the frame. You can specify a name explicitly by setting the name frame property.

Normally you don’t specify the name explicitly, and Emacs computes the frame name automatically based on a template stored in the variable frame-title-format. Emacs recomputes the name each time the frame is redisplayed.

Variable: frame-title-format

This variable specifies how to compute a name for a frame when you have not explicitly specified one (via the frame’s parameters; see Basic Parameters). The variable’s value is actually a mode line construct, just like mode-line-format, except that the ‘%c’, ‘%C’, and ‘%l’ constructs are ignored. See The Data Structure of the Mode Line.

Variable: icon-title-format

This variable specifies how to compute the name for an iconified frame when you have not explicitly specified the frame’s name via the frame’s parameters. The resulting title appears in the frame’s icon itself. If the value is a string, is should be a mode line construct like that of frame-title-format. The value can also be t, which means to use frame-title-format instead; this avoids problems with some window managers and desktop environments, where a change in a frame’s title (when a frame is iconified) is interpreted as a request to raise the frame and/or give it input focus. It is also useful if you want the frame’s title to be the same no matter if the frame is iconified or not. The default value is a string identical to the default value of frame-title-format.

Variable: multiple-frames

This variable is set automatically by Emacs. Its value is t when there are two or more frames (not counting minibuffer-only frames or invisible frames). The default value of frame-title-format uses multiple-frames so as to put the buffer name in the frame title only when there is more than one frame.

The value of this variable is not guaranteed to be accurate except while processing frame-title-format or icon-title-format.


30.7 Deleting Frames

A live frame is one that has not been deleted. When a frame is deleted, it is removed from its terminal display, although it may continue to exist as a Lisp object until there are no more references to it.

Command: delete-frame &optional frame force

This function deletes the frame frame. The argument frame must specify a live frame (see below) and defaults to the selected frame.

It first deletes any child frame of frame (see Child Frames) and any frame whose delete-before frame parameter (see Frame Interaction Parameters) specifies frame. All such deletions are performed recursively; so this step makes sure that no other frames with frame as their ancestor will exist. Then, unless frame specifies a tooltip, this function runs the hook delete-frame-functions (each function getting one argument, frame) before actually killing the frame. After actually killing the frame and removing the frame from the frame list, delete-frame runs after-delete-frame-functions.

Note that a frame cannot be deleted as long as its minibuffer serves as surrogate minibuffer for another frame (see Minibuffers and Frames). Normally, you cannot delete a frame if all other frames are invisible, but if force is non-nil, then you are allowed to do so.

Function: frame-live-p frame

This function returns non-nil if the frame frame has not been deleted. The possible non-nil return values are like those of framep. See Frames.

Some window managers provide a command to delete a window. These work by sending a special message to the program that operates the window. When Emacs gets one of these commands, it generates a delete-frame event, whose normal definition is a command that calls the function delete-frame. See Miscellaneous System Events.

Command: delete-other-frames &optional frame iconify

This command deletes all frames on frame’s terminal, except frame. If frame uses another frame’s minibuffer, that minibuffer frame is left untouched. The argument frame must specify a live frame and defaults to the selected frame. Internally, this command works by calling delete-frame with force nil for all frames that shall be deleted.

This function does not delete any of frame’s child frames (see Child Frames). If frame is a child frame, it deletes frame’s siblings only.

With the prefix argument iconify, the frames are iconified rather than deleted.


30.8 Finding All Frames

Function: frame-list

This function returns a list of all the live frames, i.e., those that have not been deleted. It is analogous to buffer-list for buffers, and includes frames on all terminals. The list that you get is newly created, so modifying the list doesn’t have any effect on the internals of Emacs.

Function: visible-frame-list

This function returns a list of just the currently visible frames. See Visibility of Frames. Frames on text terminals always count as visible, even though only the selected one is actually displayed.

Function: frame-list-z-order &optional display

This function returns a list of Emacs’ frames, in Z (stacking) order (see Raising, Lowering and Restacking Frames). The optional argument display specifies which display to poll. display should be either a frame or a display name (a string). If omitted or nil, that stands for the selected frame’s display. It returns nil if display contains no Emacs frame.

Frames are listed from topmost (first) to bottommost (last). As a special case, if display is non-nil and specifies a live frame, it returns the child frames of that frame in Z (stacking) order.

This function is not meaningful on text terminals.

Function: next-frame &optional frame minibuf

This function lets you cycle conveniently through all the frames on a specific terminal from an arbitrary starting point. It returns the frame following frame, in the list of all live frames, on frame’s terminal. The argument frame must specify a live frame and defaults to the selected frame. It never returns a frame whose no-other-frame parameter (see Frame Interaction Parameters) is non-nil.

The second argument, minibuf, says which frames to consider when deciding what the next frame should be:

nil

Consider all frames except minibuffer-only frames.

visible

Consider only visible frames.

0

Consider only visible or iconified frames.

a window

Consider only the frames using that particular window as their minibuffer window.

anything else

Consider all frames.

Function: previous-frame &optional frame minibuf

Like next-frame, but cycles through all frames in the opposite direction.

See also next-window and previous-window, in Cyclic Ordering of Windows.


30.9 Minibuffers and Frames

Normally, each frame has its own minibuffer window at the bottom, which is used whenever that frame is selected. You can get that window with the function minibuffer-window (see Minibuffer Windows).

However, you can also create a frame without a minibuffer. Such a frame must use the minibuffer window of some other frame. That other frame will serve as surrogate minibuffer frame for this frame and cannot be deleted via delete-frame (see Deleting Frames) as long as this frame is live.

When you create the frame, you can explicitly specify its minibuffer window (in some other frame) with the minibuffer frame parameter (see Buffer Parameters). If you don’t, then the minibuffer is found in the frame which is the value of the variable default-minibuffer-frame. Its value should be a frame that does have a minibuffer.

If you use a minibuffer-only frame, you might want that frame to raise when you enter the minibuffer. If so, set the variable minibuffer-auto-raise to t. See Raising, Lowering and Restacking Frames.

Variable: default-minibuffer-frame

This variable specifies the frame to use for the minibuffer window, by default. It does not affect existing frames. It is always local to the current terminal and cannot be buffer-local. See Multiple Terminals.


30.10 Input Focus

At any time, one frame in Emacs is the selected frame. The selected window (see Selecting Windows) always resides on the selected frame.

When Emacs displays its frames on several terminals (see Multiple Terminals), each terminal has its own selected frame. But only one of these is the selected frame: it’s the frame that belongs to the terminal from which the most recent input came. That is, when Emacs runs a command that came from a certain terminal, the selected frame is the one of that terminal. Since Emacs runs only a single command at any given time, it needs to consider only one selected frame at a time; this frame is what we call the selected frame in this manual. The display on which the selected frame is shown is the selected frame’s display.

Function: selected-frame

This function returns the selected frame.

Some window systems and window managers direct keyboard input to the window object that the mouse is in; others require explicit clicks or commands to shift the focus to various window objects. Either way, Emacs automatically keeps track of which frames have focus. To explicitly switch to a different frame from a Lisp function, call select-frame-set-input-focus.

The plural “frames” in the previous paragraph is deliberate: while Emacs itself has only one selected frame, Emacs can have frames on many different terminals (recall that a connection to a window system counts as a terminal), and each terminal has its own idea of which frame has input focus. Under the X Window System, where user input is organized into individual “seats” of input, each seat in turn can have its own specific input focus. When you set the input focus to a frame, you set the focus for that frame’s terminal on the last seat which interacted with Emacs, but frames on other terminals and seats may still remain focused.

If the input focus is set before any user interaction has occurred on the specified terminal, then the X server picks a random seat (normally the one with the lowest number) and sets the input focus there.

Lisp programs can switch frames temporarily by calling the function select-frame. This does not alter the window system’s concept of focus; rather, it escapes from the window manager’s control until that control is somehow reasserted.

When using a text terminal, only one frame can be displayed at a time on the terminal, so after a call to select-frame, the next redisplay actually displays the newly selected frame. This frame remains selected until a subsequent call to select-frame. Each frame on a text terminal has a number which appears in the mode line before the buffer name (see Variables Used in the Mode Line).

Function: select-frame-set-input-focus frame &optional norecord

This function selects frame, raises it (should it happen to be obscured by other frames) and tries to give it the window system’s focus. On a text terminal, the next redisplay displays the new frame on the entire terminal screen. The optional argument norecord has the same meaning as for select-frame (see below). The return value of this function is not significant.

Ideally, the function described next should focus a frame without also raising it above other frames. Unfortunately, many window-systems or window managers may refuse to comply.

Function: x-focus-frame frame &optional noactivate

This function gives frame the focus of the X server without necessarily raising it. frame nil means use the selected frame. Under X, the optional argument noactivate, if non-nil, means to avoid making frame’s window-system window the “active” window which should insist a bit more on avoiding to raise frame above other frames.

On MS-Windows the noactivate argument has no effect. However, if frame is a child frame (see Child Frames), this function usually focuses frame without raising it above other child frames.

If there is no window system support, this function does nothing.

Command: select-frame frame &optional norecord

This function selects frame frame, temporarily disregarding the focus of the X server if any. The selection of frame lasts until the next time the user does something to select a different frame, or until the next time this function is called. (If you are using a window system, the previously selected frame may be restored as the selected frame after return to the command loop, because it still may have the window system’s input focus.)

The specified frame becomes the selected frame, and its terminal becomes the selected terminal. This function then calls select-window as a subroutine, passing the window selected within frame as its first argument and norecord as its second argument (hence, if norecord is non-nil, this avoids changing the order of recently selected windows and the buffer list). See Selecting Windows.

This function returns frame, or nil if frame has been deleted.

In general, you should never use select-frame in a way that could switch to a different terminal without switching back when you’re done.

Emacs cooperates with the window system by arranging to select frames as the server and window manager request. When a window system informs Emacs that one of its frames has been selected, Emacs internally generates a focus-in event. When an Emacs frame is displayed on a text-terminal emulator, such as xterm, which supports reporting of focus-change notification, the focus-in and focus-out events are available even for text-mode frames. Focus events are normally handled by handle-focus-in.

Command: handle-focus-in event

This function handles focus-in events from window systems and terminals that support explicit focus notifications. It updates the per-frame focus flags that frame-focus-state queries and calls after-focus-change-function. In addition, it generates a switch-frame event in order to switch the Emacs notion of the selected frame to the frame most recently focused in some terminal. It’s important to note that this switching of the Emacs selected frame to the most recently focused frame does not mean that other frames do not continue to have the focus in their respective terminals. Do not invoke this function yourself: instead, attach logic to after-focus-change-function.

Command: handle-switch-frame frame

This function handles a switch-frame event, which Emacs generates for itself upon focus notification or under various other circumstances involving an input event arriving at a different frame from the last event. Do not invoke this function yourself.

Function: redirect-frame-focus frame &optional focus-frame

This function redirects focus from frame to focus-frame. This means that focus-frame will receive subsequent keystrokes and events intended for frame. After such an event, the value of last-event-frame will be focus-frame. Also, switch-frame events specifying frame will instead select focus-frame.

If focus-frame is omitted or nil, that cancels any existing redirection for frame, which therefore once again receives its own events.

One use of focus redirection is for frames that don’t have minibuffers. These frames use minibuffers on other frames. Activating a minibuffer on another frame redirects focus to that frame. This puts the focus on the minibuffer’s frame, where it belongs, even though the mouse remains in the frame that activated the minibuffer.

Selecting a frame can also change focus redirections. Selecting frame bar, when foo had been selected, changes any redirections pointing to foo so that they point to bar instead. This allows focus redirection to work properly when the user switches from one frame to another using select-window.

This means that a frame whose focus is redirected to itself is treated differently from a frame whose focus is not redirected. select-frame affects the former but not the latter.

The redirection lasts until redirect-frame-focus is called to change it.

Function: frame-focus-state frame

This function retrieves the last known focus state of frame.

It returns nil if the frame is known not to be focused, t if the frame is known to be focused, or unknown if Emacs does not know the focus state of the frame. (You may see this last state in TTY frames running on terminals that do not support explicit focus notifications.)

Variable: after-focus-change-function

This function is called with no arguments when Emacs notices that a frame may have gotten or lost focus. Focus events are delivered asynchronously, and may not be delivered in the expected order, so code that wants to do something depending on the state of focused frames have go through all the frames and check.

For instance, here’s a simple example function that sets the background color based on whether the frame has focus or not:

(add-function :after after-focus-change-function
              #'my-change-background)
(defun my-change-background ()
  (dolist (frame (frame-list))
    (pcase (frame-focus-state frame)
      (`t (set-face-background 'default "black" frame))
      (`nil (set-face-background 'default "#404040" frame)))))

Multiple frames may appear to have input focus simultaneously due to focus event delivery differences, the presence of multiple Emacs terminals, and other factors, and code should be robust in the face of this situation.

Depending on window system, focus events may also be delivered repeatedly and with different focus states before settling to the expected values. Code relying on focus notifications should “debounce” any user-visible updates arising from focus changes, perhaps by deferring work until redisplay.

This function may be called in arbitrary contexts, including from inside read-event, so take the same care as you might when writing a process filter.

User Option: focus-follows-mouse

This option informs Emacs whether and how the window manager transfers focus when you move the mouse pointer into a frame. It can have three meaningful values:

nil

The default value nil should be used when your window manager follows a “click-to-focus” policy where you have to click the mouse inside of a frame in order for that frame to gain focus.

t

The value t should be used when your window manager has the focus automatically follow the position of the mouse pointer but a frame that gains focus is not raised automatically and may even remain occluded by other window-system windows.

auto-raise

The value auto-raise should be used when your window manager has the focus automatically follow the position of the mouse pointer and a frame that gains focus is raised automatically.

If this option is non-nil, Emacs moves the mouse pointer to the frame selected by select-frame-set-input-focus. That function is used by a number of commands like, for example, other-frame and pop-to-buffer.

The distinction between the values t and auto-raise is not needed for “normal” frames because the window manager usually takes care of raising them. It is useful to automatically raise child frames via mouse-autoselect-window (see Mouse Window Auto-selection).

Note that this option does not distinguish “sloppy” focus (where the frame that previously had focus retains focus as long as the mouse pointer does not move into another window-system window) from “strict” focus (where a frame immediately loses focus when it’s left by the mouse pointer). Neither does it recognize whether your window manager supports delayed focusing or auto-raising where you can explicitly specify the time until a new frame gets focus or is auto-raised.

You can supply a “focus follows mouse” policy for individual Emacs windows by customizing the variable mouse-autoselect-window (see Mouse Window Auto-selection).


30.11 Visibility of Frames

A frame on a graphical display may be visible, invisible, or iconified. If it is visible, its contents are displayed in the usual manner. If it is iconified, its contents are not displayed, but there is a little icon somewhere to bring the frame back into view (some window managers refer to this state as minimized rather than iconified, but from Emacs’ point of view they are the same thing). If a frame is invisible, it is not displayed at all.

The concept of visibility is strongly related to that of (un-)mapped frames. A frame (or, more precisely, its window-system window) is and becomes mapped when it is displayed for the first time and whenever it changes its state of visibility from iconified or invisible to visible. Conversely, a frame is and becomes unmapped whenever it changes its status from visible to iconified or invisible.

Visibility is meaningless on text terminals, since only the selected frame is actually displayed in any case.

Function: frame-visible-p frame

This function returns the visibility status of frame frame. The value is t if frame is visible, nil if it is invisible, and icon if it is iconified.

On a text terminal, all frames are considered visible for the purposes of this function, even though only one frame is displayed. See Raising, Lowering and Restacking Frames.

Command: iconify-frame &optional frame

This function iconifies frame frame. If you omit frame, it iconifies the selected frame. This usually makes all child frames of frame (and their descendants) invisible (see Child Frames).

Command: make-frame-visible &optional frame

This function makes frame frame visible. If you omit frame, it makes the selected frame visible. This does not raise the frame, but you can do that with raise-frame if you wish (see Raising, Lowering and Restacking Frames).

Making a frame visible usually makes all its child frames (and their descendants) visible as well (see Child Frames).

Command: make-frame-invisible &optional frame force

This function makes frame frame invisible. If you omit frame, it makes the selected frame invisible. Usually, this makes all child frames of frame (and their descendants) invisible too (see Child Frames).

Unless force is non-nil, this function refuses to make frame invisible if all other frames are invisible.

The visibility status of a frame is also available as a frame parameter. You can read or change it as such. See Window Management Parameters. The user can also iconify and deiconify frames with the window manager. This happens below the level at which Emacs can exert any control, but Emacs does provide events that you can use to keep track of such changes. See Miscellaneous System Events.

Function: x-double-buffered-p &optional frame

This function returns non-nil if frame is currently being rendered with double buffering. frame defaults to the selected frame.


30.12 Raising, Lowering and Restacking Frames

Most window systems use a desktop metaphor. Part of this metaphor is the idea that system-level windows (representing, e.g., Emacs frames) are stacked in a notional third dimension perpendicular to the screen surface. The order induced by stacking is total and usually referred to as stacking (or Z-) order. Where the areas of two windows overlap, the one higher up in that order will (partially) cover the one underneath.

You can raise a frame to the top of that order or lower a frame to its bottom by using the functions raise-frame and lower-frame. You can restack a frame directly above or below another frame using the function frame-restack.

Note that all functions described below will respect the adherence of frames (and all other window-system windows) to their respective z-group (see Position Parameters). For example, you usually cannot lower a frame below that of the desktop window and you cannot raise a frame whose z-group parameter is nil above the window-system’s taskbar or tooltip window.

Command: raise-frame &optional frame

This function raises frame frame (default, the selected frame) above all other frames belonging to the same or a lower z-group as frame. If frame is invisible or iconified, this makes it visible. If frame is a child frame (see Child Frames), this raises frame above all other child frames of its parent.

Command: lower-frame &optional frame

This function lowers frame frame (default, the selected frame) below all other frames belonging to the same or a higher z-group as frame. If frame is a child frame (see Child Frames), this lowers frame below all other child frames of its parent.

Function: frame-restack frame1 frame2 &optional above

This function restacks frame1 below frame2. This implies that if both frames are visible and their display areas overlap, frame2 will (partially) obscure frame1. If the optional third argument above is non-nil, this function restacks frame1 above frame2. This means that if both frames are visible and their display areas overlap, frame1 will (partially) obscure frame2.

Technically, this function may be thought of as an atomic action performed in two steps: The first step removes frame1’s window-system window from the display. The second step reinserts frame1’s window into the display below (above if above is true) that of frame2. Hence the position of frame2 in its display’s Z (stacking) order relative to all other frames excluding frame1 remains unaltered.

Some window managers may refuse to restack windows.

Note that the effect of restacking will only hold as long as neither of the involved frames is iconified or made invisible. You can use the z-group (see Position Parameters) frame parameter to add a frame to a group of frames permanently shown above or below other frames. As long as a frame belongs to one of these groups, restacking it will only affect its relative stacking position within that group. The effect of restacking frames belonging to different z-groups is undefined. You can list frames in their current stacking order with the function frame-list-z-order (see Finding All Frames).

User Option: minibuffer-auto-raise

If this is non-nil, activation of the minibuffer raises the frame that the minibuffer window is in.

On window systems, you can also enable auto-raising (on frame selection) or auto-lowering (on frame deselection) using frame parameters. See Window Management Parameters.

The concept of raising and lowering frames also applies to text terminal frames. On each text terminal, only the top frame is displayed at any one time.

Function: tty-top-frame &optional terminal

This function returns the top frame on terminal. terminal should be a terminal object, a frame (meaning that frame’s terminal), or nil (meaning the selected frame’s terminal). If it does not refer to a text terminal, the return value is nil.


30.13 Frame Configurations

A frame configuration records the current arrangement of frames, all their properties, and the window configuration of each one. (See Window Configurations.)

Function: current-frame-configuration

This function returns a frame configuration list that describes the current arrangement of frames and their contents.

Function: set-frame-configuration configuration &optional nodelete

This function restores the state of frames described in configuration. However, this function does not restore deleted frames.

Ordinarily, this function deletes all existing frames not listed in configuration. But if nodelete is non-nil, the unwanted frames are iconified instead.


30.14 Child Frames

Child frames are objects halfway between windows (see Windows) and “normal” frames. Like windows, they are attached to an owning frame. Unlike windows, they may overlap each other—changing the size or position of one child frame does not change the size or position of any of its sibling child frames.

By design, operations to make or modify child frames are implemented with the help of frame parameters (see Frame Parameters) without any specialized functions or customizable variables. Note that child frames are meaningful on graphical terminals only.

To create a new child frame or to convert a normal frame into a child frame, set that frame’s parent-frame parameter (see Frame Interaction Parameters) to that of an already existing frame. The frame specified by that parameter will then be the frame’s parent frame as long as the parameter is not changed or reset. Technically, this makes the child frame’s window-system window a child window of the parent frame’s window-system window.

The parent-frame parameter can be changed at any time. Setting it to another frame reparents the child frame. Setting it to another child frame makes the frame a nested child frame. Setting it to nil restores the frame’s status as a top-level frame—a frame whose window-system window is a child of its display’s root window.22

Since child frames can be arbitrarily nested, a frame can be both a child and a parent frame. Also, the relative roles of child and parent frame may be reversed at any time (though it’s usually a good idea to keep the size of a child frame sufficiently smaller than that of its parent). An error will be signaled for the attempt to make a frame an ancestor of itself.

Most window-systems clip a child frame at the native edges (see Frame Geometry) of its parent frame—everything outside these edges is usually invisible. A child frame’s left and top parameters specify a position relative to the top-left corner of its parent’s native frame. When the parent frame is resized, this position remains conceptually unaltered.

NS builds do not clip child frames at the parent frame’s edges, allowing them to be positioned so they do not obscure the parent frame while still being visible themselves.

Usually, moving a parent frame moves along all its child frames and their descendants as well, keeping their relative positions unaltered. Note that the hook move-frame-functions (see Frame Position) is run for a child frame only when the position of the child frame relative to its parent frame changes.

When a parent frame is resized, its child frames conceptually retain their previous sizes and their positions relative to the left upper corner of the parent. This means that a child frame may become (partially) invisible when its parent frame shrinks. The parameter keep-ratio (see Frame Interaction Parameters) can be used to resize and reposition a child frame proportionally whenever its parent frame is resized. This may avoid obscuring parts of a frame when its parent frame is shrunk.

A visible child frame always appears on top of its parent frame thus obscuring parts of it, except on NS builds where it may be positioned beneath the parent. This is comparable to the window-system window of a top-level frame which also always appears on top of its parent window—the desktop’s root window. When a parent frame is iconified or made invisible (see Visibility of Frames), its child frames are made invisible. When a parent frame is deiconified or made visible, its child frames are made visible.

When a parent frame is about to be deleted (see Deleting Frames), its child frames are recursively deleted before it. There is one exception to this rule: When the child frame serves as a surrogate minibuffer frame (see Minibuffers and Frames) for another frame, it is retained until the parent frame has been deleted. If, at this time, no remaining frame uses the child frame as its minibuffer frame, Emacs will try to delete the child frame too. If that deletion fails for whatever reason, the child frame is made a top-level frame.

Whether a child frame can have a menu or tool bar is window-system or window manager dependent. Most window-systems explicitly disallow menu bars for child frames. It seems advisable to disable both, menu and tool bars, via the frame’s initial parameters settings.

Usually, child frames do not exhibit window manager decorations like a title bar or external borders (see Frame Geometry). When the child frame does not show a menu or tool bar, any other of the frame’s borders (see Layout Parameters) can be used instead of the external borders.

In particular, under X (but not when building with GTK+), the frame’s outer border can be used. On MS-Windows, specifying a non-zero outer border width will show a one-pixel wide external border. Under all window-systems, the internal border can be used. In either case, it’s advisable to disable a child frame’s window manager decorations with the undecorated frame parameter (see Window Management Parameters).

To resize or move an undecorated child frame with the mouse, special frame parameters (see Mouse Dragging Parameters) have to be used. The internal border of a child frame, if present, can be used to resize the frame with the mouse, provided that frame has a non-nil drag-internal-border parameter. If set, the snap-width parameter indicates the number of pixels where the frame snaps at the respective edge or corner of its parent frame.

There are two ways to drag an entire child frame with the mouse: The drag-with-mode-line parameter, if non-nil, allows to drag a frame without minibuffer window (see Minibuffer Windows) via the mode line area of its bottommost window. The drag-with-header-line parameter, if non-nil, allows to drag the frame via the header line area of its topmost window.

In order to give a child frame a draggable header or mode line, the window parameters mode-line-format and header-line-format are handy (see Window Parameters). These allow to remove an unwanted mode line (when drag-with-header-line is chosen) and to remove mouse-sensitive areas which might interfere with frame dragging.

When the user drags a frame with a mouse and overshoots, it’s easy to drag a frame out of the screen area of its parent. Retrieving such a frame can be hairy once the mouse button has been released. To prevent such a situation, it is advisable to set the frame’s top-visible or bottom-visible parameter (see Mouse Dragging Parameters).

Set the top-visible parameter of a child frame to a number when you intend to allow the user dragging that frame by its header line. Setting top-visible to a number inhibits dragging the top edge of the child frame above the top edge of its parent. Set the bottom-visible parameter to a number when you intend to drag that frame via its mode line; this inhibits dragging the bottom edge of the child frame beneath the bottom edge of its parent. In either case, that number also specifies width and height (in pixels) of the area of the child frame that remains visible during dragging.

When a child frame is used for displaying a buffer via display-buffer-in-child-frame (see Action Functions for Buffer Display), the frame’s auto-hide-function parameter (see Frame Interaction Parameters) can be set to a function, in order to appropriately deal with the frame when the window displaying the buffer shall be quit.

When a child frame is used during minibuffer interaction, for example, to display completions in a separate window, the minibuffer-exit parameter (see Frame Interaction Parameters) is useful in order to deal with the frame when the minibuffer is exited.

The behavior of child frames deviates from that of top-level frames in a number of other ways as well. Here we sketch a few of them:

  • The semantics of maximizing and iconifying child frames is highly window-system dependent. As a rule, applications should never invoke these operations on child frames. By default, invoking iconify-frame on a child frame will try to iconify the top-level frame corresponding to that child frame instead. To obtain a different behavior, users may customize the option iconify-child-frame described below.
  • Raising, lowering and restacking child frames (see Raising, Lowering and Restacking Frames) or changing the z-group (see Position Parameters) of a child frame changes only the stacking order of child frames with the same parent.
  • Many window-systems are not able to change the opacity (see Font and Color Parameters) of child frames.
  • Transferring focus from a child frame to an ancestor that is not its parent by clicking with the mouse in a visible part of that ancestor’s window may fail with some window-systems. You may have to click into the direct parent’s window-system window first.
  • Window managers might not bother to extend their focus follows mouse policy to child frames. Customizing mouse-autoselect-window can help in this regard (see Mouse Window Auto-selection).
  • Dropping (see Drag and Drop) on child frames is not guaranteed to work on all window-systems. Some will drop the object on the parent frame or on some ancestor instead.

The following two functions can be useful when working with child and parent frames:

Function: frame-parent &optional frame

This function returns the parent frame of frame. The parent frame of frame is the Emacs frame whose window-system window is the parent window of frame’s window-system window. If such a frame exists, frame is considered a child frame of that frame.

This function returns nil if frame has no parent frame.

Function: frame-ancestor-p ancestor descendant

This functions returns non-nil if ancestor is an ancestor of descendant. ancestor is an ancestor of descendant when it is either descendant’s parent frame or it is an ancestor of descendant’s parent frame. Both, ancestor and descendant must specify live frames.

Note also the function window-largest-empty-rectangle (see Coordinates and Windows) which can be used to inscribe a child frame in the largest empty area of an existing window. This can be useful to avoid that a child frame obscures any text shown in that window.

Customizing the following option can be useful to tweak the behavior of iconify-frame for child frames.

User Option: iconify-child-frame

This option tells Emacs how to proceed when it is asked to iconify a child frame. If it is nil, iconify-frame will do nothing when invoked on a child frame. If it is iconify-top-level, Emacs will try to iconify the top-level frame that is the ancestor of this child frame instead. If it is make-invisible, Emacs will try to make this child frame invisible instead of iconifying it.

Any other value means to try iconifying the child frame. Since such an attempt may not be honored by all window managers and can even lead to making the child frame unresponsive to user actions, the default is to iconify the top level frame instead.


30.15 Mouse Tracking

Sometimes it is useful to track the mouse, which means to display something to indicate where the mouse is and move the indicator as the mouse moves. For efficient mouse tracking, you need a way to wait until the mouse actually moves.

The convenient way to track the mouse is to ask for events to represent mouse motion. Then you can wait for motion by waiting for an event. In addition, you can easily handle any other sorts of events that may occur. That is useful, because normally you don’t want to track the mouse forever—only until some other event, such as the release of a button.

Macro: track-mouse body…

This macro executes body, with generation of mouse motion events enabled. Typically, body would use read-event to read the motion events and modify the display accordingly. See Motion Events, for the format of mouse motion events.

The value of track-mouse is that of the last form in body. You should design body to return when it sees the up-event that indicates the release of the button, or whatever kind of event means it is time to stop tracking. Its value also controls how mouse events are reported while a mouse button is held down: if it is dropping or drag-source, the motion events are reported relative to the frame underneath the pointer. If there is no such frame, the events will be reported relative to the frame the mouse buttons were first pressed on. In addition, the posn-window of the mouse position list will be nil if the value is drag-source. This is useful to determine if a frame is not directly visible underneath the mouse pointer.

The track-mouse macro causes Emacs to generate mouse motion events by binding the variable track-mouse to a non-nil value. If that variable has the special value dragging, it additionally instructs the display engine to refrain from changing the shape of the mouse pointer. This is desirable in Lisp programs that require mouse dragging across large portions of Emacs display, which might otherwise cause the mouse pointer to change its shape according to the display portion it hovers on (see Pointer Shape). Therefore, Lisp programs that need the mouse pointer to retain its original shape during dragging should bind track-mouse to the value dragging at the beginning of their body.

The usual purpose of tracking mouse motion is to indicate on the screen the consequences of pushing or releasing a button at the current position.

In many cases, you can avoid the need to track the mouse by using the mouse-face text property (see Properties with Special Meanings). That works at a much lower level and runs more smoothly than Lisp-level mouse tracking.


30.16 Mouse Position

The functions mouse-position and set-mouse-position give access to the current position of the mouse.

Function: mouse-position

This function returns a description of the position of the mouse. The value looks like (frame x . y), where x and y are integers giving the (possibly rounded) position in multiples of the default character size of frame (see Frame Font) relative to the native position of frame (see Frame Geometry).

Variable: mouse-position-function

If non-nil, the value of this variable is a function for mouse-position to call. mouse-position calls this function just before returning, with its normal return value as the sole argument, and it returns whatever this function returns to it.

This abnormal hook exists for the benefit of packages like xt-mouse.el that need to do mouse handling at the Lisp level.

Variable: tty-menu-calls-mouse-position-function

If non-nil, TTY menus will call mouse-position-function as described above. This exists for cases where mouse-position-function is not safe to be called by the TTY menus, such as if it could trigger redisplay.

Function: set-mouse-position frame x y

This function warps the mouse to position x, y in frame frame. The arguments x and y are integers, giving the position in multiples of the default character size of frame (see Frame Font) relative to the native position of frame (see Frame Geometry).

The resulting mouse position is constrained to the native frame of frame. If frame is not visible, this function does nothing. The return value is not significant.

Function: mouse-pixel-position

This function is like mouse-position except that it returns coordinates in units of pixels rather than units of characters.

Function: set-mouse-pixel-position frame x y

This function warps the mouse like set-mouse-position except that x and y are in units of pixels rather than units of characters.

The resulting mouse position is not constrained to the native frame of frame. If frame is not visible, this function does nothing. The return value is not significant.

On a graphical terminal the following two functions allow the absolute position of the mouse cursor to be retrieved and set.

Function: mouse-absolute-pixel-position

This function returns a cons cell (x . y) of the coordinates of the mouse cursor position in pixels, relative to a position (0, 0) of the selected frame’s display.

Function: set-mouse-absolute-pixel-position x y

This function moves the mouse cursor to the position (x, y). The coordinates x and y are interpreted in pixels relative to a position (0, 0) of the selected frame’s display.

The following function can tell whether the mouse cursor is currently visible on a frame:

Function: frame-pointer-visible-p &optional frame

This predicate function returns non-nil if the mouse pointer displayed on frame is visible; otherwise it returns nil. frame omitted or nil means the selected frame. This is useful when make-pointer-invisible is set to t: it allows you to know if the pointer has been hidden. See Mouse Avoidance in The Emacs Manual.


30.17 Pop-Up Menus

A Lisp program can pop up a menu so that the user can choose an alternative with the mouse. On a text terminal, if the mouse is not available, the user can choose an alternative using the keyboard motion keys—C-n, C-p, or up- and down-arrow keys.

Function: x-popup-menu position menu

This function displays a pop-up menu and returns an indication of what selection the user makes.

The argument position specifies where on the screen to put the top left corner of the menu. It can be either a mouse button event (which says to put the menu where the user actuated the button) or a list of this form:

((xoffset yoffset) window)

where xoffset and yoffset are coordinates, measured in pixels, counting from the top left corner of window. window may be a window or a frame.

If position is t, it means to use the current mouse position (or the top-left corner of the frame if the mouse is not available on a text terminal). If position is nil, it means to precompute the key binding equivalents for the keymaps specified in menu, without actually displaying or popping up the menu.

The argument menu says what to display in the menu. It can be a keymap or a list of keymaps (see Menu Keymaps). In this case, the return value is the list of events corresponding to the user’s choice. This list has more than one element if the choice occurred in a submenu. (Note that x-popup-menu does not actually execute the command bound to that sequence of events.) On text terminals and toolkits that support menu titles, the title is taken from the prompt string of menu if menu is a keymap, or from the prompt string of the first keymap in menu if it is a list of keymaps (see Defining Menus).

Alternatively, menu can have the following form:

(title pane1 pane2...)

where each pane is a list of form

(title item1 item2...)

Each item should be a cons cell, (line . value), where line is a string and value is the value to return if that line is chosen. Unlike in a menu keymap, a nil value does not make the menu item non-selectable. Alternatively, each item can be a string rather than a cons cell; this makes a non-selectable menu item.

If the user gets rid of the menu without making a valid choice, for instance by clicking the mouse away from a valid choice or by typing C-g, then this normally results in a quit and x-popup-menu does not return. But if position is a mouse button event (indicating that the user invoked the menu with the mouse) then no quit occurs and x-popup-menu returns nil.

Usage note: Don’t use x-popup-menu to display a menu if you could do the job with a prefix key defined with a menu keymap. If you use a menu keymap to implement a menu, C-h c and C-h a can see the individual items in that menu and provide help for them. If instead you implement the menu by defining a command that calls x-popup-menu, the help facilities cannot know what happens inside that command, so they cannot give any help for the menu’s items.

The menu bar mechanism, which lets you switch between submenus by moving the mouse, cannot look within the definition of a command to see that it calls x-popup-menu. Therefore, if you try to implement a submenu using x-popup-menu, it cannot work with the menu bar in an integrated fashion. This is why all menu bar submenus are implemented with menu keymaps within the parent menu, and never with x-popup-menu. See The Menu Bar.

If you want a menu bar submenu to have contents that vary, you should still use a menu keymap to implement it. To make the contents vary, add a hook function to menu-bar-update-hook to update the contents of the menu keymap as necessary.

Variable: x-pre-popup-menu-hook

A normal hook run immediately before a pop-up menu is displayed, either directly by calling x-popup-menu, or through a menu keymap. It won’t be called if x-popup-menu returns for some other reason without displaying a pop-up menu.


30.18 Dialog Boxes

A dialog box is a variant of a pop-up menu—it looks a little different, it always appears in the center of a frame, and it has just one level and one or more buttons. The main use of dialog boxes is for asking questions that the user can answer with “yes”, “no”, and a few other alternatives. With a single button, they can also force the user to acknowledge important information. The functions y-or-n-p and yes-or-no-p use dialog boxes instead of the keyboard, when called from commands invoked by mouse clicks.

Function: x-popup-dialog position contents &optional header

This function displays a pop-up dialog box and returns an indication of what selection the user makes. The argument contents specifies the alternatives to offer; it has this format:

(title (string . value)…)

which looks like the list that specifies a single pane for x-popup-menu.

The return value is value from the chosen alternative.

As for x-popup-menu, an element of the list may be just a string instead of a cons cell (string . value). That makes a box that cannot be selected.

If nil appears in the list, it separates the left-hand items from the right-hand items; items that precede the nil appear on the left, and items that follow the nil appear on the right. If you don’t include a nil in the list, then approximately half the items appear on each side.

Dialog boxes always appear in the center of a frame; the argument position specifies which frame. The possible values are as in x-popup-menu, but the precise coordinates or the individual window don’t matter; only the frame matters.

If header is non-nil, the frame title for the box is ‘Information’, otherwise it is ‘Question’. The former is used for message-box (see message-box). (On text terminals, the box title is not displayed.)

In some configurations, Emacs cannot display a real dialog box; so instead it displays the same items in a pop-up menu in the center of the frame.

If the user gets rid of the dialog box without making a valid choice, for instance using the window manager, then this produces a quit and x-popup-dialog does not return.


30.19 Pointer Shape

You can specify the mouse pointer style for particular text or images using the pointer text property, and for images with the :pointer and :map image properties. The values you can use in these properties are in the table below. The actual shapes may vary between systems; the descriptions are examples.

text
nil

The usual mouse pointer style used over text (an “I”-like shape).

arrow
vdrag
modeline

An arrow that points north-west.

hand

A hand that points upwards.

hdrag

A right-left arrow.

nhdrag

An up-down arrow.

hourglass

A rotating ring.

Over void parts of the window (parts that do not correspond to any of the buffer contents), the mouse pointer usually uses the arrow style, but you can specify a different style (one of those above) by setting void-text-area-pointer.

User Option: void-text-area-pointer

This variable specifies the mouse pointer style for void text areas. These include the areas after the end of a line or below the last line in the buffer. The default is to use the arrow (non-text) pointer style.

When using some window systems, you can specify what the text pointer style really looks like by setting the variable x-pointer-shape.

Variable: x-pointer-shape

This variable specifies the pointer shape to use ordinarily in the Emacs frame, for the text pointer style.

Variable: x-sensitive-text-pointer-shape

This variable specifies the pointer shape to use when the mouse is over mouse-sensitive text.

These variables affect newly created frames. They do not normally affect existing frames; however, if you set the mouse color of a frame, that also installs the current value of those two variables. See Font and Color Parameters.

The values you can use, to specify either of these pointer shapes, are defined in the file lisp/term/x-win.el. Use M-x apropos RET x-pointer RET to see a list of them.


30.20 Window System Selections

In window systems, such as X, data can be transferred between different applications by means of selections. X defines an arbitrary number of selection types, each of which can store its own data; however, only three are commonly used: the clipboard, primary selection, and secondary selection. Other window systems support only the clipboard. See Cut and Paste in The GNU Emacs Manual, for Emacs commands that make use of these selections. This section documents the low-level functions for reading and setting window-system selections.

Command: gui-set-selection type data

This function sets a window-system selection. It takes two arguments: a selection type type, and the value to assign to it, data.

type should be a symbol; it is usually one of PRIMARY, SECONDARY or CLIPBOARD. These are symbols with upper-case names, in accord with X Window System conventions. If type is nil, that stands for PRIMARY.

If data is nil, it means to clear out the selection. Otherwise, data may be a string, a symbol, an integer, an overlay, or a cons of two markers pointing to the same buffer. An overlay or a pair of markers stands for text in the overlay or between the markers. The argument data may also be a vector of valid non-vector selection values.

If data is a string, then its text properties can specify values used for individual data types. For example, if data has a property named text/uri-list, then a call to gui-get-selection with the data type text/uri-list will result in the value of that property being used instead of data itself.

This function returns data.

Function: gui-get-selection &optional type data-type

This function accesses selections set up by Emacs or by other programs. It takes two optional arguments, type and data-type. The default for type, the selection type, is PRIMARY.

The data-type argument specifies the form of data conversion to use, to convert the raw data obtained from another program into Lisp data. Meaningful values include TEXT, STRING, UTF8_STRING, TARGETS, LENGTH, DELETE, FILE_NAME, CHARACTER_POSITION, NAME, LINE_NUMBER, COLUMN_NUMBER, OWNER_OS, HOST_NAME, USER, CLASS, ATOM, and INTEGER. (These are symbols with upper-case names in accord with X conventions.) The default for data-type is STRING. Window systems other than X usually support only a small subset of these types, in addition to STRING.

User Option: selection-coding-system

This variable specifies the coding system to use when reading and writing selections or the clipboard. See Coding Systems. The default is compound-text-with-extensions, which converts to the text representation that X11 normally uses.

When Emacs runs on MS-Windows, it does not implement X selections in general, but it does support the clipboard. gui-get-selection and gui-set-selection on MS-Windows support the text data type only; if the clipboard holds other types of data, Emacs treats the clipboard as empty. The supported data type is STRING.

For backward compatibility, there are obsolete aliases x-get-selection and x-set-selection, which were the names of gui-get-selection and gui-set-selection before Emacs 25.1.


30.21 Yanking Media

If you choose, for instance, “Copy Image” in a web browser, that image is put onto the clipboard, and Emacs can access it via gui-get-selection. But in general, inserting image data into an arbitrary buffer isn’t very useful—you can’t really do much with it by default.

So Emacs has a system to let modes register handlers for these “complicated” selections.

Function: yank-media-handler types handler

types can be a MIME media type symbol, a regexp to match these, or a list of these symbols and regexps. For instance:

(yank-media-handler 'text/html #'my-html-handler)
(yank-media-handler "image/.*" #'my-image-handler)

A mode can register as many handlers as required.

The handler function is called with two parameters: The MIME media type symbol and the data (as a string). The handler should then insert the object into the buffer, or save it, or do whatever is appropriate for the mode.

The yank-media command will consult the registered handlers in the current buffer, compare that with the available media types on the clipboard, and then pass on the matching selection to the handler (if any). If there’s more than one matching selection, the user is queried first.

The yank-media-types command can be used to explore the clipboard/primary selection. It lists all the media types that are currently available, and can be handy when creating handlers—to see what data is actually available. Some applications put a surprising amount of different data types on the clipboard.


30.22 Drag and Drop

When the user drops something from another application over Emacs, Emacs will try to insert any text and open any URL that was dropped. If text was dropped, then it will always be inserted at the location of the mouse pointer where the drop happened, or saved in the kill ring if insertion failed, which could happen if the buffer was read-only. If a URL was dropped instead, then Emacs will first try to call an appropriate handler function by matching the URL against regexps defined in the variable dnd-protocol-alist, and then against those defined in the variables browse-url-handlers and browse-url-default-handlers. Should no suitable handler be located, Emacs will fall back to inserting the URL as plain text.

Variable: dnd-protocol-alist

This variable is a list of cons cells of the form (pattern . action). pattern is a regexp that URLs are matched against after being dropped. action is a function that is called with two arguments, should a URL being dropped match pattern: the URL being dropped, and the action being performed for the drop, which is one of the symbols copy, move, link, private or ask.

If action is private, then it means the program that initiated the drop wants Emacs to perform an unspecified action with the URL; a reasonable action to perform in that case is to open the URL or copy its contents into the current buffer. Otherwise, action has the same meaning as the action argument to dnd-begin-file-drag.

Emacs implements receiving text and URLs individually for each window system, and does not by default support receiving other kinds of data as drops. To support receiving other kinds of data, use the X-specific interface described below.

When a user drags something from another application over Emacs under the X Window System, that other application expects Emacs to tell it if Emacs understands the data being dragged. The function in the variable x-dnd-test-function is called by Emacs to determine what to reply to any such inquiry. The default value is x-dnd-default-test-function, which accepts drops if the type of the data to be dropped is present in x-dnd-known-types. Changing the variables x-dnd-test-function and x-dnd-known-types can make Emacs accept or reject drops based on some other criteria.

If you want to change the way Emacs receives drops of different data types, or you want to enable it to understand a new type, change the variable x-dnd-types-alist. Doing so correctly requires detailed knowledge of what data types other applications use for drag and drop.

These data types are typically implemented as special data types that can be obtained from an X selection provided by the other application. In most cases, they are either the same data types that are typically accepted by gui-set-selection, or MIME types, depending on the specific drag-and-drop protocol being used. For example, the data type used for plain text may be either "STRING" or "text/plain".

When Emacs runs on X window system, it supports the X Direct Save (XDS) protocol, which allows users to save a file by dragging and dropping it onto an Emacs window, such as a Dired window. To comply with the unique requirements of XDS, these drag-and-drop requests are processed specially: instead of being handled according to x-dnd-types-alist, they are handled by the direct-save function that is the value of the variable x-dnd-direct-save-function. The value should be a function of two arguments, need-name and filename. The XDS protocol uses a two-step procedure for dragging files:

  1. The application from which the file is dragged asks Emacs to provide the full file name under which to save the file. For this purpose, the direct-save function is called with its first argument need-name non-nil, and the second argument filename set to the basename of the file to be saved. It should return the fully-expanded absolute file name under which to save the file. For example, if a file is dragged to a Dired window, the natural directory for the file is the directory of the file shown at location of the drop. If saving the file is not possible for some reason, the function should return nil, which will cancel the drag-and-drop operation.
  2. The application from which the file is dragged saves the file under the name returned by the first call to the direct-save function. If it succeeds in saving the file, the direct-save function is called again, this time with the first argument need-name set to nil and the second argument filename set to the full absolute name of the saved file. The function is then expected to do whatever is needed given the fact that file was saved. For example, Dired should update the directory on display by showing the new file there.

The default value of x-dnd-direct-save-function is x-dnd-save-direct.

Function: x-dnd-save-direct need-name filename

When called with the need-name argument non-nil, this function prompts the user for the absolute file name under which it should be saved. If the specified file already exists, it additionally asks the user whether to overwrite it, and returns the absolute file name only if the user confirms the overwriting.

When called with the need-name argument nil, it reverts the Dired listing if the current buffer is in Dired mode or one of its descendants, and otherwise visits the file by calling find-file (see Functions for Visiting Files).

Function: x-dnd-save-direct-immediately need-name filename

This function works like x-dnd-save-direct, but when called with its need-name argument non-nil, it doesn’t prompt the user for the full name of the file to be saved; instead, it returns its argument filename expanded against the current buffer’s default directory (see Functions that Expand Filenames). (It still asks for confirmation if a file by that name already exists in the default directory.)

On capable window systems, Emacs also supports dragging contents from its frames to windows of other applications.

Function: dnd-begin-text-drag text &optional frame action allow-same-frame

This function begins dragging text from frame to another program (known as the drop target), and returns the result of drag-and-drop operation when the text is dropped or the drag-and-drop operation is canceled. text is the text that will be inserted by the drop target.

action must be one of the symbols copy or move, where copy means that text should be inserted by the drop target, and move means the same as copy, but in addition the caller may have to delete text from its source as explained below.

frame is the frame where the mouse is currently held down, or nil, which means to use the selected frame. This function may return immediately if no mouse buttons are held down, so it should be only called immediately after a down-mouse-1 or similar event (see Mouse Events), with frame set to the frame where that event was generated (see Click Events).

allow-same-frame specifies whether or not drops on top of frame itself are to be ignored.

The return value specifies the action that the drop target actually performed, and optionally what the caller should do. It can be one of the following symbols:

copy

The drop target inserted the dropped text.

move

The drop target inserted the dropped text, but in addition the caller should delete text from wherever it originated, such as its buffer.

private

The drop target performed some other unspecified action.

nil

The drag-and-drop operation was canceled.

Function: dnd-begin-file-drag file &optional frame action allow-same-frame

This function begins dragging file from frame to another program, and returns the result of the drag-and-drop operation when the file is dropped or the drag-and-drop operation is canceled.

If file is a remote file, then a temporary copy will be made.

action must be one of the symbols copy, move or link, where copy means that file should be opened or copied by the drop target, move means the drop target should move the file to another location, and link means the drop target should create a symbolic link to file. It is an error to specify link as the action if file is a remote file.

frame and allow-same-frame have the same meaning as in dnd-begin-text-drag.

The return value is the action that the drop target actually performed, which can be one of the following symbols:

copy

The drop target opened or copied file to a different location.

move

The drop target moved file to a different location.

link

The drop target (usually a file manager) created a symbolic link to file.

private

The drop target performed some other unspecified action.

nil

The drag-and-drop operation was canceled.

Function: dnd-begin-drag-files files &optional frame action allow-same-frame

This function is like dnd-begin-file-drag, except that files is a list of files. If the drop target doesn’t support dropping multiple files, then the first file will be used instead.

Function: dnd-direct-save file name &optional frame allow-same-frame

This function is similar to dnd-begin-file-drag (with the default action of copy), but instead of specifying the action you specify the name of the copy created by the target program in name.

The high-level interfaces described above are implemented on top of a lower-level primitive. If you need to drag content other than files or text, use the low-level interface x-begin-drag instead. However, using it will require detailed knowledge of the data types and actions used by the programs to transfer content via drag-and-drop on each platform you want to support.

Function: x-begin-drag targets &optional action frame return-frame allow-current-frame follow-tooltip

This function begins a drag from frame, and returns when the drag-and-drop operation ends, either because the drop was successful, or because the drop was rejected. The drop occurs when all mouse buttons are released on top of an X window other than frame (the drop target), or any X window if allow-current-frame is non-nil. If no mouse buttons are held down when the drag-and-drop operation begins, this function may immediately return nil.

targets is a list of strings describing selection targets, much like the data-type argument to gui-get-selection, that the drop target can request from Emacs (see Window System Selections).

action is a symbol describing the action recommended to the target. It can either be XdndActionCopy, which means to copy the contents of the selection XdndSelection to the drop target; or XdndActionMove, which means copy as with XdndActionCopy, and in addition the caller should delete whatever was stored in that selection after copying it.

action may also be an alist which associates between symbols describing the available actions, and strings that the drop target is expected to present to the user to choose between the available actions.

If return-frame is non-nil and the mouse moves over an Emacs frame after first moving out of frame, then the frame to which the mouse moves will be returned immediately. If return-frame is the symbol now, then any frame underneath the mouse pointer will be returned without waiting for the mouse to first move out of frame. return-frame is useful when you want to treat dragging content from one frame to another specially, while also being able to drag content to other programs, but it is not guaranteed to work on all systems and with all window managers.

If follow-tooltip is non-nil, the position of any tooltip (such as one shown by tooltip-show) will follow the location of the mouse pointer whenever it moves during the drag-and-drop operation. The tooltip will be hidden once all mouse buttons are released.

If the drop was rejected or no drop target was found, this function returns nil. Otherwise, it returns a symbol describing the action the target chose to perform, which can differ from action if that isn’t supported by the drop target. XdndActionPrivate is also a valid return value in addition to XdndActionCopy and XdndActionMove; it means that the drop target chose to perform an unspecified action, and no further processing is required by the caller.

The caller must cooperate with the target to fully perform the action chosen by the target. For example, callers should delete the buffer text that was dragged if this function returns XdndActionMove.

On X Windows, several different drag-and-drop protocols are supported by x-begin-drag. When dragging content that is known to not be supported by a specific drag-and-drop protocol, it might be desirable to turn that protocol off, by changing the values of the following variables:

Variable: x-dnd-disable-motif-protocol

When this is non-nil, the Motif drag and drop protocols are disabled, and dropping onto programs that only understand them will not work.

Variable: x-dnd-use-offix-drop

When this is nil, the OffiX (old KDE) drag and drop protocol is disabled. When this is the symbol files, the OffiX protocol will only be used if "FILE_NAME" is one of the targets given to x-begin-drag. Any other value means to use the OffiX protocol to drop all supported content.

Variable: x-dnd-use-unsupported-drop

When one of the "STRING", "UTF8_STRING", "COMPOUND_TEXT" or "TEXT" targets is present in the list given to x-begin-drag, Emacs will try to use synthesized mouse events and the primary selection to insert the text if the drop target doesn’t support any drag-and-drop protocol at all.

A side effect is that Emacs will become the owner of the primary selection upon such a drop. If that is not desired, then the drop emulation can be disabled by setting this variable to nil.


30.23 Color Names

A color name is text (usually in a string) that specifies a color. Symbolic names such as ‘black’, ‘white’, ‘red’, etc., are allowed; use M-x list-colors-display to see a list of defined names. You can also specify colors numerically in forms such as ‘#rgb’ and ‘RGB:r/g/b’, where r specifies the red level, g specifies the green level, and b specifies the blue level. You can use either one, two, three, or four hex digits for r; then you must use the same number of hex digits for all g and b as well, making either 3, 6, 9 or 12 hex digits in all. (See the documentation of the X Window System for more details about numerical RGB specification of colors.)

These functions provide a way to determine which color names are valid, and what they look like. In some cases, the value depends on the selected frame, as described below; see Input Focus, for the meaning of the term “selected frame”.

To read user input of color names with completion, use read-color (see read-color).

Function: color-defined-p color &optional frame

This function reports whether a color name is meaningful. It returns t if so; otherwise, nil. The argument frame says which frame’s display to ask about; if frame is omitted or nil, the selected frame is used.

Note that this does not tell you whether the display you are using really supports that color. When using X, you can ask for any defined color on any kind of display, and you will get some result—typically, the closest it can do. To determine whether a frame can really display a certain color, use color-supported-p (see below).

This function used to be called x-color-defined-p, and that name is still supported as an alias.

Function: defined-colors &optional frame

This function returns a list of the color names that are defined and supported on frame frame (default, the selected frame). If frame does not support colors, the value is nil.

This function used to be called x-defined-colors, and that name is still supported as an alias.

Function: color-supported-p color &optional frame background-p

This returns t if frame can really display the color color (or at least something close to it). If frame is omitted or nil, the question applies to the selected frame.

Some terminals support a different set of colors for foreground and background. If background-p is non-nil, that means you are asking whether color can be used as a background; otherwise you are asking whether it can be used as a foreground.

The argument color must be a valid color name.

Function: color-gray-p color &optional frame

This returns t if color is a shade of gray, as defined on frame’s display. If frame is omitted or nil, the question applies to the selected frame. If color is not a valid color name, this function returns nil.

Function: color-values color &optional frame

This function returns a value that describes what color should ideally look like on frame. If color is defined, the value is a list of three integers, which give the amount of red, the amount of green, and the amount of blue. Each integer ranges in principle from 0 to 65535, but some displays may not use the full range. This three-element list is called the rgb values of the color.

If color is not defined, the value is nil.

(color-values "black")
     ⇒ (0 0 0)
(color-values "white")
     ⇒ (65535 65535 65535)
(color-values "red")
     ⇒ (65535 0 0)
(color-values "pink")
     ⇒ (65535 49344 52171)
(color-values "hungry")
     ⇒ nil

The color values are returned for frame’s display. If frame is omitted or nil, the information is returned for the selected frame’s display. If the frame cannot display colors, the value is nil.

This function used to be called x-color-values, and that name is still supported as an alias.

Function: color-name-to-rgb color &optional frame

This function does the same as color-values, but it returns color values as floating-point numbers between 0.0 and 1.0 inclusive.

Function: color-dark-p rgb

This function returns non-nil if the color described by its RGB triplet rgb is more readable against white background than against dark background. The argument rgb should be a list of the form (r g b), with each component a floating-point number in the range 0.0 to 1.0 inclusive. You can use color-name-to-rgb to convert a color’s name to such a list.


30.24 Text Terminal Colors

Text terminals usually support only a small number of colors, and the computer uses small integers to select colors on the terminal. This means that the computer cannot reliably tell what the selected color looks like; instead, you have to inform your application which small integers correspond to which colors. However, Emacs does know the standard set of colors and will try to use them automatically.

The functions described in this section control how terminal colors are used by Emacs.

Several of these functions use or return rgb values, described in Color Names.

These functions accept a display (either a frame or the name of a terminal) as an optional argument. We hope in the future to make Emacs support different colors on different text terminals; then this argument will specify which terminal to operate on (the default being the selected frame’s terminal; see Input Focus). At present, though, the frame argument has no effect.

Function: tty-color-define name number &optional rgb frame

This function associates the color name name with color number number on the terminal.

The optional argument rgb, if specified, is an rgb value, a list of three numbers that specify what the color actually looks like. If you do not specify rgb, then this color cannot be used by tty-color-approximate to approximate other colors, because Emacs will not know what it looks like.

Function: tty-color-clear &optional frame

This function clears the table of defined colors for a text terminal.

Function: tty-color-alist &optional frame

This function returns an alist recording the known colors supported by a text terminal.

Each element has the form (name number . rgb) or (name number). Here, name is the color name, number is the number used to specify it to the terminal. If present, rgb is a list of three color values (for red, green, and blue) that says what the color actually looks like.

Function: tty-color-approximate rgb &optional frame

This function finds the closest color, among the known colors supported for display, to that described by the rgb value rgb (a list of color values). The return value is an element of tty-color-alist.

Function: tty-color-translate color &optional frame

This function finds the closest color to color among the known colors supported for display and returns its index (an integer). If the name color is not defined, the value is nil.


30.25 X Resources

This section describes some of the functions and variables for querying and using X resources, or their equivalent on your operating system. See X Resources in The GNU Emacs Manual, for more information about X resources.

Function: x-get-resource attribute class &optional component subclass

The function x-get-resource retrieves a resource value from the X Window defaults database.

Resources are indexed by a combination of a key and a class. This function searches using a key of the form ‘instance.attribute’ (where instance is the name under which Emacs was invoked), and using ‘Emacs.class’ as the class.

The optional arguments component and subclass add to the key and the class, respectively. You must specify both of them or neither. If you specify them, the key is ‘instance.component.attribute’, and the class is ‘Emacs.class.subclass’.

Variable: x-resource-class

This variable specifies the application name that x-get-resource should look up. The default value is "Emacs". You can examine X resources for other application names by binding this variable to some other string, around a call to x-get-resource.

Variable: x-resource-name

This variable specifies the instance name that x-get-resource should look up. The default value is the name Emacs was invoked with, or the value specified with the ‘-name’ or ‘-rn’ switches.

To illustrate some of the above, suppose that you have the line:

xterm.vt100.background: yellow

in your X resources file (whose name is usually ~/.Xdefaults or ~/.Xresources). Then:

(let ((x-resource-class "XTerm") (x-resource-name "xterm"))
  (x-get-resource "vt100.background" "VT100.Background"))
     ⇒ "yellow"
(let ((x-resource-class "XTerm") (x-resource-name "xterm"))
  (x-get-resource "background" "VT100" "vt100" "Background"))
     ⇒ "yellow"
Variable: inhibit-x-resources

If this variable is non-nil, Emacs does not look up X resources, and X resources do not have any effect when creating new frames.


30.26 Display Feature Testing

The functions in this section describe the basic capabilities of a particular display. Lisp programs can use them to adapt their behavior to what the display can do. For example, a program that ordinarily uses a popup menu could use the minibuffer if popup menus are not supported.

The optional argument display in these functions specifies which display to ask the question about. It can be a display name, a frame (which designates the display that frame is on), or nil (which refers to the selected frame’s display, see Input Focus).

See Color Names, Text Terminal Colors, for other functions to obtain information about displays.

Function: display-popup-menus-p &optional display

This function returns t if popup menus are supported on display, nil if not. Support for popup menus requires that the mouse be available, since the menu is popped up by clicking the mouse on some portion of the Emacs display.

Function: display-graphic-p &optional display

This function returns t if display is a graphic display capable of displaying several frames and several different fonts at once. This is true for displays that use a window system such as X, and false for text terminals.

Function: display-mouse-p &optional display

This function returns t if display has a mouse available, nil if not.

Function: display-color-p &optional display

This function returns t if the screen is a color screen. It used to be called x-display-color-p, and that name is still supported as an alias.

Function: display-grayscale-p &optional display

This function returns t if the screen can display shades of gray. (All color displays can do this.)

Function: display-supports-face-attributes-p attributes &optional display

This function returns non-nil if all the face attributes in attributes are supported (see Face Attributes).

The definition of “supported” is somewhat heuristic, but basically means that a face containing all the attributes in attributes, when merged with the default face for display, can be represented in a way that’s

  1. different in appearance than the default face, and
  2. close in spirit to what the attributes specify, if not exact.

Point (2) implies that a :weight black attribute will be satisfied by any display that can display bold, as will :foreground "yellow" as long as some yellowish color can be displayed, but :slant italic will not be satisfied by the tty display code’s automatic substitution of a dim face for italic.

Function: display-selections-p &optional display

This function returns t if display supports selections. Windowed displays normally support selections, but they may also be supported in some other cases.

Function: display-images-p &optional display

This function returns t if display can display images. Windowed displays ought in principle to handle images, but some systems lack the support for that. On a display that does not support images, Emacs cannot display a tool bar.

Function: display-screens &optional display

This function returns the number of screens associated with the display.

Function: display-pixel-height &optional display

This function returns the height of the screen in pixels. On a character terminal, it gives the height in characters.

For graphical terminals, note that on multi-monitor setups this refers to the pixel height for all physical monitors associated with display. See Multiple Terminals.

Function: display-pixel-width &optional display

This function returns the width of the screen in pixels. On a character terminal, it gives the width in characters.

For graphical terminals, note that on multi-monitor setups this refers to the pixel width for all physical monitors associated with display. See Multiple Terminals.

Function: display-mm-height &optional display

This function returns the height of the screen in millimeters, or nil if Emacs cannot get that information.

For graphical terminals, note that on multi-monitor setups this refers to the height for all physical monitors associated with display. See Multiple Terminals.

Function: display-mm-width &optional display

This function returns the width of the screen in millimeters, or nil if Emacs cannot get that information.

For graphical terminals, note that on multi-monitor setups this refers to the width for all physical monitors associated with display. See Multiple Terminals.

User Option: display-mm-dimensions-alist

This variable allows the user to specify the dimensions of graphical displays returned by display-mm-height and display-mm-width in case the system provides incorrect values.

Function: display-backing-store &optional display

This function returns the backing store capability of the display. Backing store means recording the pixels of windows (and parts of windows) that are not exposed, so that when exposed they can be displayed very quickly.

Values can be the symbols always, when-mapped, or not-useful. The function can also return nil when the question is inapplicable to a certain kind of display.

Function: display-save-under &optional display

This function returns non-nil if the display supports the SaveUnder feature. That feature is used by pop-up windows to save the pixels they obscure, so that they can pop down quickly.

Function: display-planes &optional display

This function returns the number of planes the display supports. This is typically the number of bits per pixel. For a tty display, it is log to base two of the number of colors supported.

Function: display-visual-class &optional display

This function returns the visual class for the screen. The value is one of the symbols static-gray (a limited, unchangeable number of grays), gray-scale (a full range of grays), static-color (a limited, unchangeable number of colors), pseudo-color (a limited number of colors), true-color (a full range of colors), and direct-color (a full range of colors).

Function: display-color-cells &optional display

This function returns the number of color cells the screen supports.

These functions obtain additional information about the window system in use where Emacs shows the specified display. (Their names begin with x- for historical reasons.)

Function: x-server-version &optional display

This function returns the list of version numbers of the GUI window system running on display, such as the X server on GNU and Unix systems. The value is a list of three integers: the major and minor version numbers of the protocol, and the distributor-specific release number of the window system software itself. On GNU and Unix systems, these are normally the version of the X protocol and the distributor-specific release number of the X server software. On MS-Windows, this is the version of the Windows OS.

Function: x-server-vendor &optional display

This function returns the vendor that provided the window system software (as a string). On GNU and Unix systems this really means whoever distributes the X server. On MS-Windows this is the vendor ID string of the Windows OS (Microsoft).

When the developers of X labeled software distributors as “vendors”, they showed their false assumption that no system could ever be developed and distributed noncommercially.


31 Positions

A position is the index of a character in the text of a buffer. More precisely, a position identifies the place between two characters (or before the first character, or after the last character), so we can speak of the character before or after a given position. However, we often speak of the character “at” a position, meaning the character after that position.

Positions are usually represented as integers starting from 1, but can also be represented as markers—special objects that relocate automatically when text is inserted or deleted so they stay with the surrounding characters. Functions that expect an argument to be a position (an integer), but accept a marker as a substitute, normally ignore which buffer the marker points into; they convert the marker to an integer, and use that integer, exactly as if you had passed the integer as the argument, even if the marker points to the wrong buffer. A marker that points nowhere cannot convert to an integer; using it instead of an integer causes an error. See Markers.

See also the field feature (see Defining and Using Fields), which provides functions that are used by many cursor-motion commands.


31.1 Point

Point is a special buffer position used by many editing commands, including the self-inserting typed characters and text insertion functions. Other commands move point through the text to allow editing and insertion at different places.

Like other positions, point designates a place between two characters (or before the first character, or after the last character), rather than a particular character. Usually terminals display the cursor over the character that immediately follows point; point is actually before the character on which the cursor sits.

The value of point is a number no less than 1, and no greater than the buffer size plus 1. If narrowing is in effect (see Narrowing), then point is constrained to fall within the accessible portion of the buffer (possibly at one end of it).

Each buffer has its own value of point, which is independent of the value of point in other buffers. Each window also has a value of point, which is independent of the value of point in other windows on the same buffer. This is why point can have different values in various windows that display the same buffer. When a buffer appears in only one window, the buffer’s point and the window’s point normally have the same value, so the distinction is rarely important. See Windows and Point, for more details.

Function: point

This function returns the value of point in the current buffer, as an integer.

(point)
     ⇒ 175
Function: point-min

This function returns the minimum accessible value of point in the current buffer. This is normally 1, but if narrowing is in effect, it is the position of the start of the region that you narrowed to. (See Narrowing.)

Function: point-max

This function returns the maximum accessible value of point in the current buffer. This is (1+ (buffer-size)), unless narrowing is in effect, in which case it is the position of the end of the region that you narrowed to. (See Narrowing.)

Function: buffer-end flag

This function returns (point-max) if flag is greater than 0, (point-min) otherwise. The argument flag must be a number.

Function: buffer-size &optional buffer

This function returns the total number of characters in the current buffer. In the absence of any narrowing (see Narrowing), point-max returns a value one larger than this.

If you specify a buffer, buffer, then the value is the size of buffer.

(buffer-size)
     ⇒ 35
(point-max)
     ⇒ 36

31.2 Motion

Motion functions change the value of point, either relative to the current value of point, relative to the beginning or end of the buffer, or relative to the edges of the selected window. See Point.


31.2.1 Motion by Characters

These functions move point based on a count of characters. goto-char is the fundamental primitive; the other functions use that.

Command: goto-char position

This function sets point in the current buffer to the value position.

If narrowing is in effect, position still counts from the beginning of the buffer, but point cannot go outside the accessible portion. If position is out of range, goto-char moves point to the beginning or the end of the accessible portion.

When this function is called interactively, position is the numeric prefix argument, if provided; otherwise it is read from the minibuffer.

goto-char returns position.

Command: forward-char &optional count

This function moves point count characters forward, towards the end of the buffer (or backward, towards the beginning of the buffer, if count is negative). If count is nil, the default is 1.

If this attempts to move past the beginning or end of the buffer (or the limits of the accessible portion, when narrowing is in effect), it signals an error with error symbol beginning-of-buffer or end-of-buffer.

In an interactive call, count is the numeric prefix argument.

Command: backward-char &optional count

This is just like forward-char except that it moves in the opposite direction.


31.2.2 Motion by Words

The functions for parsing words described below use the syntax table and char-script-table to decide whether a given character is part of a word. See Syntax Tables, and see Character Properties.

Command: forward-word &optional count

This function moves point forward count words (or backward if count is negative). If count is omitted or nil, it defaults to 1. In an interactive call, count is specified by the numeric prefix argument.

“Moving one word” means moving until point crosses a word-constituent character, which indicates the beginning of a word, and then continue moving until the word ends. By default, characters that begin and end words, known as word boundaries, are defined by the current buffer’s syntax table (see Table of Syntax Classes), but modes can override that by setting up a suitable find-word-boundary-function-table, described below. Characters that belong to different scripts (as defined by char-script-table), also define a word boundary (see Character Properties). In any case, this function cannot move point past the boundary of the accessible portion of the buffer, or across a field boundary (see Defining and Using Fields). The most common case of a field boundary is the end of the prompt in the minibuffer.

If it is possible to move count words, without being stopped prematurely by the buffer boundary or a field boundary, the value is t. Otherwise, the return value is nil and point stops at the buffer boundary or field boundary.

If inhibit-field-text-motion is non-nil, this function ignores field boundaries.

Command: backward-word &optional count

This function is just like forward-word, except that it moves backward until encountering the front of a word, rather than forward.

User Option: words-include-escapes

This variable affects the behavior of forward-word and backward-word, and everything that uses them. If it is non-nil, then characters in the escape and character-quote syntax classes count as part of words. Otherwise, they do not.

Variable: inhibit-field-text-motion

If this variable is non-nil, certain motion functions including forward-word, forward-sentence, and forward-paragraph ignore field boundaries.

Variable: find-word-boundary-function-table

This variable affects the behavior of forward-word and backward-word, and everything that uses them. Its value is a char-table (see Char-Tables) of functions to search for word boundaries. If a character has a non-nil entry in this table, then when a word starts or ends with that character, the corresponding function will be called with 2 arguments: pos and limit. The function should return the position of the other word boundary. Specifically, if pos is smaller than limit, then pos is at the beginning of a word, and the function should return the position after the last character of the word; otherwise, pos is at the last character of a word, and the function should return the position of that word’s first character.

Function: forward-word-strictly &optional count

This function is like forward-word, but it is not affected by find-word-boundary-function-table. Lisp programs that should not change behavior when word movement is modified by modes which set that table, such as subword-mode, should use this function instead of forward-word.

Function: backward-word-strictly &optional count

This function is like backward-word, but it is not affected by find-word-boundary-function-table. Like with forward-word-strictly, use this function instead of backward-word when movement by words should only consider syntax tables.


31.2.3 Motion to an End of the Buffer

To move point to the beginning of the buffer, write:

(goto-char (point-min))

Likewise, to move to the end of the buffer, use:

(goto-char (point-max))

Here are two commands that users use to do these things. They are documented here to warn you not to use them in Lisp programs, because they set the mark and display messages in the echo area.

Command: beginning-of-buffer &optional n

This function moves point to the beginning of the buffer (or the limits of the accessible portion, when narrowing is in effect), setting the mark at the previous position (except in Transient Mark mode, if the mark is already active, it does not set the mark.)

If n is non-nil, then it puts point n tenths of the way from the beginning of the accessible portion of the buffer. In an interactive call, n is the numeric prefix argument, if provided; otherwise n defaults to nil.

Warning: Don’t use this function in Lisp programs!

Command: end-of-buffer &optional n

This function moves point to the end of the buffer (or the limits of the accessible portion, when narrowing is in effect), setting the mark at the previous position (except in Transient Mark mode when the mark is already active). If n is non-nil, then it puts point n tenths of the way from the end of the accessible portion of the buffer.

In an interactive call, n is the numeric prefix argument, if provided; otherwise n defaults to nil.

Warning: Don’t use this function in Lisp programs!


31.2.4 Motion by Text Lines

Text lines are portions of the buffer delimited by newline characters, which are regarded as part of the previous line. The first text line begins at the beginning of the buffer, and the last text line ends at the end of the buffer whether or not the last character is a newline. The division of the buffer into text lines is not affected by the width of the window, by line continuation in display, or by how tabs and control characters are displayed.

Command: beginning-of-line &optional count

This function moves point to the beginning of the current line. With an argument count not nil or 1, it moves forward count−1 lines and then to the beginning of the line.

This function does not move point across a field boundary (see Defining and Using Fields) unless doing so would move beyond there to a different line; therefore, if count is nil or 1, and point starts at a field boundary, point does not move. To ignore field boundaries, either bind inhibit-field-text-motion to t, or use the forward-line function instead. For instance, (forward-line 0) does the same thing as (beginning-of-line), except that it ignores field boundaries.

If this function reaches the end of the buffer (or of the accessible portion, if narrowing is in effect), it positions point there. No error is signaled.

Function: line-beginning-position &optional count

Return the position that (beginning-of-line count) would move to.

Command: end-of-line &optional count

This function moves point to the end of the current line. With an argument count not nil or 1, it moves forward count−1 lines and then to the end of the line.

This function does not move point across a field boundary (see Defining and Using Fields) unless doing so would move beyond there to a different line; therefore, if count is nil or 1, and point starts at a field boundary, point does not move. To ignore field boundaries, bind inhibit-field-text-motion to t.

If this function reaches the end of the buffer (or of the accessible portion, if narrowing is in effect), it positions point there. No error is signaled.

Function: line-end-position &optional count

Return the position that (end-of-line count) would move to.

Function: pos-bol &optional count

Like line-beginning-position, but ignores fields (and is more efficient).

Function: pos-eol &optional count

Like line-end-position, but ignores fields (and is more efficient).

Command: forward-line &optional count

This function moves point forward count lines, to the beginning of the line following that. If count is negative, it moves point −count lines backward, to the beginning of a line preceding that. If count is zero, it moves point to the beginning of the current line. If count is nil, that means 1.

If forward-line encounters the beginning or end of the buffer (or of the accessible portion) before finding that many lines, it sets point there. No error is signaled.

forward-line returns the difference between count and the number of lines actually moved. If you attempt to move down five lines from the beginning of a buffer that has only three lines, point stops at the end of the last line, and the value will be 2. As an explicit exception, if the last accessible line is non-empty, but has no newline (e.g., if the buffer ends without a newline), the function sets point to the end of that line, and the value returned by the function counts that line as one line successfully moved.

In an interactive call, count is the numeric prefix argument.

Function: count-lines start end &optional ignore-invisible-lines

This function returns the number of lines between the positions start and end in the current buffer. If start and end are equal, then it returns 0. Otherwise it returns at least 1, even if start and end are on the same line. This is because the text between them, considered in isolation, must contain at least one line unless it is empty.

If the optional ignore-invisible-lines is non-nil, invisible lines will not be included in the count.

Command: count-words start end

This function returns the number of words between the positions start and end in the current buffer.

This function can also be called interactively. In that case, it prints a message reporting the number of lines, words, and characters in the buffer, or in the region if the region is active.

Function: line-number-at-pos &optional pos absolute

This function returns the line number in the current buffer corresponding to the buffer position pos. If pos is nil or omitted, the current buffer position is used. If absolute is nil, the default, counting starts at (point-min), so the value refers to the contents of the accessible portion of the (potentially narrowed) buffer. If absolute is non-nil, ignore any narrowing and return the absolute line number.

Also see the functions bolp and eolp in Examining Text Near Point. These functions do not move point, but test whether it is already at the beginning or end of a line.


31.2.5 Motion by Screen Lines

The line functions in the previous section count text lines, delimited only by newline characters. By contrast, these functions count screen lines, which are defined by the way the text appears on the screen. A text line is a single screen line if it is short enough to fit the width of the selected window, but otherwise it may occupy several screen lines.

In some cases, text lines are truncated on the screen rather than continued onto additional screen lines. In these cases, vertical-motion moves point much like forward-line. See Truncation.

Because the width of a given string depends on the flags that control the appearance of certain characters, vertical-motion behaves differently, for a given piece of text, depending on the buffer it is in, and even on the selected window (because the width, the truncation flag, and display table may vary between windows). See Usual Display Conventions.

These functions scan text to determine where screen lines break, and thus take time proportional to the distance scanned.

Function: vertical-motion count &optional window cur-col

This function moves point to the start of the screen line count screen lines down from the screen line containing point. If count is negative, it moves up instead. If count is zero, point moves to the visual start of the current screen line.

The count argument can be a cons cell, (cols . lines), instead of an integer. Then the function moves by lines screen lines, as described for count above, and puts point cols columns from the visual start of that screen line. The value of cols can be a float, and is interpreted in units of the frame’s canonical character width (see Frame Font); this allows specifying accurate horizontal position of point when the target screen line uses variable fonts. Note that cols are counted from the visual start of the line; if the window is scrolled horizontally (see Horizontal Scrolling), the column where point will end is in addition to the number of columns by which the text is scrolled, and if the target line is a continuation line, its leftmost column is considered column zero (unlike column-oriented functions, see Counting Columns).

The return value is the number of screen lines over which point was moved. The value may be less in absolute value than count if the beginning or end of the buffer was reached.

The window window is used for obtaining parameters such as the width, the horizontal scrolling, and the display table. But vertical-motion always operates on the current buffer, even if window currently displays some other buffer.

The optional argument cur-col specifies the current column when the function is called. This is the window-relative horizontal coordinate of point, measured in units of font width of the frame’s default face. Providing it speeds up the function, especially in very long lines, because the function doesn’t have to go back in the buffer in order to determine the current column. Note that cur-col is also counted from the visual start of the line.

Function: count-screen-lines &optional beg end count-final-newline window

This function returns the number of screen lines in the text from beg to end. The number of screen lines may be different from the number of actual lines, due to line continuation, the display table, etc. If beg and end are nil or omitted, they default to the beginning and end of the accessible portion of the buffer.

If the region ends with a newline, that is ignored unless the optional third argument count-final-newline is non-nil.

The optional fourth argument window specifies the window for obtaining parameters such as width, horizontal scrolling, and so on. The default is to use the selected window’s parameters.

Like vertical-motion, count-screen-lines always uses the current buffer, regardless of which buffer is displayed in window. This makes possible to use count-screen-lines in any buffer, whether or not it is currently displayed in some window.

Command: move-to-window-line count

This function moves point with respect to the text currently displayed in the selected window. It moves point to the beginning of the screen line count screen lines from the top of the window; zero means the topmost line. If count is negative, that specifies a position −count lines from the bottom (or the last line of the buffer, if the buffer ends above the specified screen position); thus, count of −1 specifies the last fully visible screen line of the window.

If count is nil, then point moves to the beginning of the line in the middle of the window. If the absolute value of count is greater than the size of the window, then point moves to the place that would appear on that screen line if the window were tall enough. This will probably cause the next redisplay to scroll to bring that location onto the screen.

In an interactive call, count is the numeric prefix argument.

The value returned is the screen line number point has moved to, relative to the top line of the window.

Function: move-to-window-group-line count

This function is like move-to-window-line, except that when the selected window is a part of a group of windows (see Window Group), move-to-window-group-line will move to a position with respect to the entire group, not just the single window. This condition holds when the buffer local variable move-to-window-group-line-function is set to a function. In this case, move-to-window-group-line calls the function with the argument count, then returns its result.

Function: compute-motion from frompos to topos width offsets window

This function scans the current buffer, calculating screen positions. It scans the buffer forward from position from, assuming that is at screen coordinates frompos, to position to or coordinates topos, whichever comes first. It returns the ending buffer position and screen coordinates.

The coordinate arguments frompos and topos are cons cells of the form (hpos . vpos).

The argument width is the number of columns available to display text; this affects handling of continuation lines. nil means the actual number of usable text columns in the window, which is equivalent to the value returned by (window-width window).

The argument offsets is either nil or a cons cell of the form (hscroll . tab-offset). Here hscroll is the number of columns not being displayed at the left margin; most callers get this by calling window-hscroll. Meanwhile, tab-offset is the offset between column numbers on the screen and column numbers in the buffer. This can be nonzero in a continuation line, when the previous screen lines’ widths do not add up to a multiple of tab-width. It is always zero in a non-continuation line.

The window window serves only to specify which display table to use. compute-motion always operates on the current buffer, regardless of what buffer is displayed in window.

The return value is a list of five elements:

(pos hpos vpos prevhpos contin)

Here pos is the buffer position where the scan stopped, vpos is the vertical screen position, and hpos is the horizontal screen position.

The result prevhpos is the horizontal position one character back from pos. The result contin is t if the last line was continued after (or within) the previous character.

For example, to find the buffer position of column col of screen line line of a certain window, pass the window’s display start location as from and the window’s upper-left coordinates as frompos. Pass the buffer’s (point-max) as to, to limit the scan to the end of the accessible portion of the buffer, and pass line and col as topos. Here’s a function that does this:

(defun coordinates-of-position (col line)
  (car (compute-motion (window-start)
                       '(0 . 0)
                       (point-max)
                       (cons col line)
                       (window-width)
                       (cons (window-hscroll) 0)
                       (selected-window))))

When you use compute-motion for the minibuffer, you need to use minibuffer-prompt-width to get the horizontal position of the beginning of the first screen line. See Minibuffer Contents.


31.2.6 Moving over Balanced Expressions

Here are several functions concerned with balanced-parenthesis expressions (also called sexps in connection with moving across them in Emacs). The syntax table controls how these functions interpret various characters; see Syntax Tables. See Parsing Expressions, for lower-level primitives for scanning sexps or parts of sexps. For user-level commands, see Commands for Editing with Parentheses in The GNU Emacs Manual.

Command: forward-list &optional arg

This function moves forward across arg (default 1) balanced groups of parentheses. (Other syntactic entities such as words or paired string quotes are ignored.)

Command: backward-list &optional arg

This function moves backward across arg (default 1) balanced groups of parentheses. (Other syntactic entities such as words or paired string quotes are ignored.)

Command: up-list &optional arg escape-strings no-syntax-crossing

This function moves forward out of arg (default 1) levels of parentheses. A negative argument means move backward but still to a less deep spot. If escape-strings is non-nil (as it is interactively), move out of enclosing strings as well. If no-syntax-crossing is non-nil (as it is interactively), prefer to break out of any enclosing string instead of moving to the start of a list broken across multiple strings. On error, location of point is unspecified.

Command: backward-up-list &optional arg escape-strings no-syntax-crossing

This function is just like up-list, but with a negated argument.

Command: down-list &optional arg

This function moves forward into arg (default 1) levels of parentheses. A negative argument means move backward but still go deeper in parentheses (−arg levels).

Command: forward-sexp &optional arg

This function moves forward across arg (default 1) balanced expressions. Balanced expressions include both those delimited by parentheses and other kinds, such as words and string constants. See Parsing Expressions. For example,

---------- Buffer: foo ----------
(concat∗ "foo " (car x) y z)
---------- Buffer: foo ----------

(forward-sexp 3)
     ⇒ nil

---------- Buffer: foo ----------
(concat "foo " (car x) y∗ z)
---------- Buffer: foo ----------
Command: backward-sexp &optional arg

This function moves backward across arg (default 1) balanced expressions.

Command: beginning-of-defun &optional arg

This function moves back to the argth beginning of a defun. If arg is negative, this actually moves forward, but it still moves to the beginning of a defun, not to the end of one. arg defaults to 1.

Command: end-of-defun &optional arg

This function moves forward to the argth end of a defun. If arg is negative, this actually moves backward, but it still moves to the end of a defun, not to the beginning of one. arg defaults to 1.

User Option: defun-prompt-regexp

If non-nil, this buffer-local variable holds a regular expression that specifies what text can appear before the open-parenthesis that starts a defun. That is to say, a defun begins on a line that starts with a match for this regular expression, followed by a character with open-parenthesis syntax.

User Option: open-paren-in-column-0-is-defun-start

If this variable’s value is non-nil, an open parenthesis in column 0 is considered to be the start of a defun. If it is nil, an open parenthesis in column 0 has no special meaning. The default is t. If a string literal happens to have a parenthesis in column 0, escape it with a backslash to avoid a false positive.

Variable: beginning-of-defun-function

If non-nil, this variable holds a function for finding the beginning of a defun. The function beginning-of-defun calls this function instead of using its normal method, passing it its optional argument. If the argument is non-nil, the function should move back by that many functions, like beginning-of-defun does.

Variable: end-of-defun-function

If non-nil, this variable holds a function for finding the end of a defun. The function end-of-defun calls this function instead of using its normal method.

If Emacs is compiled with tree-sitter, it can use the tree-sitter parser information to move across syntax constructs. Since what exactly is considered a defun varies between languages, a major mode should set treesit-defun-type-regexp to determine that. Then the mode can get navigation-by-defun functionality for free, by using treesit-beginning-of-defun and treesit-end-of-defun.

Variable: treesit-defun-type-regexp

This variable determines which nodes are considered defuns by Emacs. It can be a regexp that matches the type of defun nodes. (For “node” and “node type”, see Parsing Program Source.)

For example, python-mode sets this variable to a regexp that matches either ‘function_definition’ or ‘class_definition’.

Sometimes not all nodes matched by the regexp are valid defuns. Therefore, this variable can also be a cons cell of the form (regexp . pred), where pred should be a function that takes a node as its argument, and returns non-nil if the node is a valid defun, or nil if it is not valid.

Variable: treesit-defun-tactic

This variable determines how Emacs treats nested defuns. If the value is top-level, navigation functions only move across top-level defuns. If the value is nested, navigation functions recognize nested defuns.


31.2.7 Skipping Characters

The following two functions move point over a specified set of characters. For example, they are often used to skip whitespace. For related functions, see Motion and Syntax.

These functions convert the set string to multibyte if the buffer is multibyte, and they convert it to unibyte if the buffer is unibyte, as the search functions do (see Searching and Matching).

Function: skip-chars-forward character-set &optional limit

This function moves point in the current buffer forward, skipping over a given set of characters. It examines the character following point, then advances point if the character matches character-set. This continues until it reaches a character that does not match. The function returns the number of characters moved over.

The argument character-set is a string, like the inside of a ‘[…]’ in a regular expression except that ‘]’ does not terminate it, and ‘\’ quotes ‘^’, ‘-’ or ‘\’. Thus, "a-zA-Z" skips over all letters, stopping before the first nonletter, and "^a-zA-Z" skips nonletters stopping before the first letter (see Regular Expressions). Character classes can also be used, e.g., "[:alnum:]" (see Character Classes).

If limit is supplied (it must be a number or a marker), it specifies the maximum position in the buffer that point can be skipped to. Point will stop at or before limit.

In the following example, point is initially located directly before the ‘T’. After the form is evaluated, point is located at the end of that line (between the ‘t’ of ‘hat’ and the newline). The function skips all letters and spaces, but not newlines.

---------- Buffer: foo ----------
I read "∗The cat in the hat
comes back" twice.
---------- Buffer: foo ----------

(skip-chars-forward "a-zA-Z ")
     ⇒ 18

---------- Buffer: foo ----------
I read "The cat in the hat∗
comes back" twice.
---------- Buffer: foo ----------
Function: skip-chars-backward character-set &optional limit

This function moves point backward, skipping characters that match character-set, until limit. It is just like skip-chars-forward except for the direction of motion.

The return value indicates the distance traveled. It is an integer that is zero or less.


31.3 Excursions

It is often useful to move point temporarily within a localized portion of the program. This is called an excursion, and it is done with the save-excursion special form. This construct remembers the initial identity of the current buffer, and its value of point, and restores them after the excursion completes. It is the standard way to move point within one part of a program and avoid affecting the rest of the program, and is used thousands of times in the Lisp sources of Emacs.

If you only need to save and restore the identity of the current buffer, use save-current-buffer or with-current-buffer instead (see The Current Buffer). If you need to save or restore window configurations, see the forms described in Window Configurations and in Frame Configurations.

Special Form: save-excursion body…

This special form saves the identity of the current buffer and the value of point in it, evaluates body, and finally restores the buffer and its saved value of point. Both saved values are restored even in case of an abnormal exit via throw or error (see Nonlocal Exits).

The value returned by save-excursion is the result of the last form in body, or nil if no body forms were given.

Because save-excursion only saves point for the buffer that was current at the start of the excursion, any changes made to point in other buffers, during the excursion, will remain in effect afterward. This frequently leads to unintended consequences, so the byte compiler warns if you call set-buffer during an excursion:

Warning: Use ‘with-current-buffer’ rather than
         save-excursion+set-buffer

To avoid such problems, you should call save-excursion only after setting the desired current buffer, as in the following example:

(defun append-string-to-buffer (string buffer)
  "Append STRING to the end of BUFFER."
  (with-current-buffer buffer
    (save-excursion
      (goto-char (point-max))
      (insert string))))

Likewise, save-excursion does not restore window-buffer correspondences altered by functions such as switch-to-buffer.

Warning: Ordinary insertion of text adjacent to the saved point value relocates the saved value, just as it relocates all markers. More precisely, the saved value is a marker with insertion type nil. See Marker Insertion Types. Therefore, when the saved point value is restored, it normally comes before the inserted text.

Macro: save-mark-and-excursion body…

This macro is like save-excursion, but also saves and restores the mark location and mark-active. This macro does what save-excursion did before Emacs 25.1.


31.4 Narrowing

Narrowing means limiting the text addressable by Emacs editing commands to a limited range of characters in a buffer. The text that remains addressable is called the accessible portion of the buffer.

Narrowing is specified with two buffer positions, which become the beginning and end of the accessible portion. For most editing commands and primitives, these positions replace the values of the beginning and end of the buffer. While narrowing is in effect, no text outside the accessible portion is displayed, and point cannot move outside the accessible portion. Note that narrowing does not alter actual buffer positions (see Point); it only determines which positions are considered the accessible portion of the buffer. Most functions refuse to operate on text that is outside the accessible portion.

Commands for saving buffers are unaffected by narrowing; they save the entire buffer regardless of any narrowing.

If you need to display in a single buffer several very different types of text, consider using an alternative facility described in Swapping Text Between Two Buffers.

Command: narrow-to-region start end

This function sets the accessible portion of the current buffer to start at start and end at end. Both arguments should be character positions.

In an interactive call, start and end are set to the bounds of the current region (point and the mark, with the smallest first).

However, when the narrowing has been set by with-restriction with a label argument (see below), narrow-to-region can be used only within the limits of that narrowing. If start or end are outside these limits, the corresponding limit set by with-restriction is used instead. To gain access to other portions of the buffer, use without-restriction with the same label.

Command: narrow-to-page &optional move-count

This function sets the accessible portion of the current buffer to include just the current page. An optional first argument move-count non-nil means to move forward or backward by move-count pages and then narrow to one page. The variable page-delimiter specifies where pages start and end (see Standard Regular Expressions Used in Editing).

In an interactive call, move-count is set to the numeric prefix argument.

Command: widen

This function cancels any narrowing in the current buffer, so that the entire contents are accessible. This is called widening. It is equivalent to the following expression:

(narrow-to-region 1 (1+ (buffer-size)))

However, when a narrowing has been set by with-restriction with a label argument (see below), the limits set by with-restriction are restored, instead of canceling the narrowing. To gain access to other portions of the buffer, use without-restriction with the same label.

Function: buffer-narrowed-p

This function returns non-nil if the buffer is narrowed, and nil otherwise.

Special Form: save-restriction body…

This special form saves the current bounds of the accessible portion, evaluates the body forms, and finally restores the saved bounds, thus restoring the same state of narrowing (or absence thereof) formerly in effect. The state of narrowing is restored even in the event of an abnormal exit via throw or error (see Nonlocal Exits). Therefore, this construct is a clean way to narrow a buffer temporarily.

This construct also saves and restores the narrowings that were set by with-restriction with a label argument (see below).

The value returned by save-restriction is that returned by the last form in body, or nil if no body forms were given.

Caution: it is easy to make a mistake when using the save-restriction construct. Read the entire description here before you try it.

If body changes the current buffer, save-restriction still restores the restrictions on the original buffer (the buffer whose restrictions it saved from), but it does not restore the identity of the current buffer.

save-restriction does not restore point; use save-excursion for that. If you use both save-restriction and save-excursion together, save-excursion should come first (on the outside). Otherwise, the old point value would be restored with temporary narrowing still in effect. If the old point value were outside the limits of the temporary narrowing, this would fail to restore it accurately.

Here is a simple example of correct use of save-restriction:

---------- Buffer: foo ----------
This is the contents of foo
This is the contents of foo
This is the contents of foo∗
---------- Buffer: foo ----------

(save-excursion
  (save-restriction
    (goto-char 1)
    (forward-line 2)
    (narrow-to-region 1 (point))
    (goto-char (point-min))
    (replace-string "foo" "bar")))

---------- Buffer: foo ----------
This is the contents of bar
This is the contents of bar
This is the contents of foo∗
---------- Buffer: foo ----------
Special Form: with-restriction start end [:label label] body

This special form saves the current bounds of the accessible portion of the buffer, sets the accessible portion to start at start and end at end, evaluates the body forms, and restores the saved bounds. In that case it is equivalent to

(save-restriction
  (narrow-to-region start end)
  body)

When the optional argument label, which is evaluated to get the label to use and must yield a non-nil value, is present, the narrowing is labeled. A labeled narrowing differs from a non-labeled one in several ways:

  • During the evaluation of the body form, narrow-to-region and widen can be used only within the start and end limits.
  • To lift the restriction introduced by with-restriction and gain access to other portions of the buffer, use without-restriction with the same label argument. (Another way to gain access to other portions of the buffer is to use an indirect buffer (see Indirect Buffers).)
  • Labeled narrowings can be nested.
  • Labeled narrowings can only be used in Lisp programs: they are never visible on display, and never interfere with narrowings set by the user.

If you use with-restriction with the optional label argument, we recommend documenting the label in the doc strings of the functions which use it, so that other Lisp programs your code calls could lift the labeled narrowing if and when it needs.

Special Form: without-restriction [:label label] body

This special form saves the current bounds of the accessible portion of the buffer, widens the buffer, evaluates the body forms, and restores the saved bounds. In that case it is equivalent to

(save-restriction
  (widen)
  body)

When the optional argument label is present, the narrowing set by with-restriction with the same label argument is lifted.


32 Markers

A marker is a Lisp object used to specify a position in a buffer relative to the surrounding text. A marker changes its offset from the beginning of the buffer automatically whenever text is inserted or deleted, so that it stays with the two characters on either side of it.


32.1 Overview of Markers

A marker specifies a buffer and a position in that buffer. A marker can be used to represent a position in functions that require one, just as an integer could be used. In that case, the marker’s buffer is normally ignored. Of course, a marker used in this way usually points to a position in the buffer that the function operates on, but that is entirely the programmer’s responsibility. See Positions, for a complete description of positions.

A marker has three attributes: the marker position, the marker buffer, and the insertion type. The marker position is an integer that is equivalent (at a given time) to the marker as a position in that buffer. But the marker’s position value can change during the life of the marker, and often does. Insertion and deletion of text in the buffer relocate the marker. The idea is that a marker positioned between two characters remains between those two characters despite insertion and deletion elsewhere in the buffer. Relocation changes the integer equivalent of the marker.

Deleting text around a marker’s position leaves the marker between the characters immediately before and after the deleted text. Inserting text at the position of a marker normally leaves the marker either in front of or after the new text, depending on the marker’s insertion type (see Marker Insertion Types)—unless the insertion is done with insert-before-markers (see Inserting Text).

Insertion and deletion in a buffer must check all the markers and relocate them if necessary. This slows processing in a buffer with a large number of markers. For this reason, it is a good idea to make a marker point nowhere if you are sure you don’t need it any more. Markers that can no longer be accessed are eventually removed (see Garbage Collection).

Because it is common to perform arithmetic operations on a marker position, most of these operations (including + and -) accept markers as arguments. In such cases, the marker stands for its current position.

Here are examples of creating markers, setting markers, and moving point to markers:

;; Make a new marker that initially does not point anywhere:
(setq m1 (make-marker))
     ⇒ #<marker in no buffer>

;; Set m1 to point between the 99th and 100th characters
;;   in the current buffer:
(set-marker m1 100)
     ⇒ #<marker at 100 in markers.texi>

;; Now insert one character at the beginning of the buffer:
(goto-char (point-min))
     ⇒ 1
(insert "Q")
     ⇒ nil

;; m1 is updated appropriately.
m1
     ⇒ #<marker at 101 in markers.texi>

;; Two markers that point to the same position
;;   are not eq, but they are equal.
(setq m2 (copy-marker m1))
     ⇒ #<marker at 101 in markers.texi>
(eq m1 m2)
     ⇒ nil
(equal m1 m2)
     ⇒ t

;; When you are finished using a marker, make it point nowhere.
(set-marker m1 nil)
     ⇒ #<marker in no buffer>

32.2 Predicates on Markers

You can test an object to see whether it is a marker, or whether it is either an integer or a marker. The latter test is useful in connection with the arithmetic functions that work with both markers and integers.

Function: markerp object

This function returns t if object is a marker, nil otherwise. Note that integers are not markers, even though many functions will accept either a marker or an integer.

Function: integer-or-marker-p object

This function returns t if object is an integer or a marker, nil otherwise.

Function: number-or-marker-p object

This function returns t if object is a number (either integer or floating point) or a marker, nil otherwise.


32.3 Functions that Create Markers

When you create a new marker, you can make it point nowhere, or point to the present position of point, or to the beginning or end of the accessible portion of the buffer, or to the same place as another given marker.

The next four functions all return markers with insertion type nil. See Marker Insertion Types.

Function: make-marker

This function returns a newly created marker that does not point anywhere.

(make-marker)
     ⇒ #<marker in no buffer>
Function: point-marker

This function returns a new marker that points to the present position of point in the current buffer. See Point. For an example, see copy-marker, below.

Function: point-min-marker

This function returns a new marker that points to the beginning of the accessible portion of the buffer. This will be the beginning of the buffer unless narrowing is in effect. See Narrowing.

Function: point-max-marker

This function returns a new marker that points to the end of the accessible portion of the buffer. This will be the end of the buffer unless narrowing is in effect. See Narrowing.

Here are examples of this function and point-min-marker, shown in a buffer containing a version of the source file for the text of this chapter.

(point-min-marker)
     ⇒ #<marker at 1 in markers.texi>
(point-max-marker)
     ⇒ #<marker at 24080 in markers.texi>

(narrow-to-region 100 200)
     ⇒ nil
(point-min-marker)
     ⇒ #<marker at 100 in markers.texi>
(point-max-marker)
     ⇒ #<marker at 200 in markers.texi>
Function: copy-marker &optional marker-or-integer insertion-type

If passed a marker as its argument, copy-marker returns a new marker that points to the same place and the same buffer as does marker-or-integer. If passed an integer as its argument, copy-marker returns a new marker that points to position marker-or-integer in the current buffer.

The new marker’s insertion type is specified by the argument insertion-type. See Marker Insertion Types.

(copy-marker 0)
     ⇒ #<marker at 1 in markers.texi>

(copy-marker 90000)
     ⇒ #<marker at 24080 in markers.texi>

An error is signaled if marker is neither a marker nor an integer.

Two distinct markers are considered equal (even though not eq) to each other if they have the same position and buffer, or if they both point nowhere.

(setq p (point-marker))
     ⇒ #<marker at 2139 in markers.texi>

(setq q (copy-marker p))
     ⇒ #<marker at 2139 in markers.texi>

(eq p q)
     ⇒ nil

(equal p q)
     ⇒ t

32.4 Information from Markers

This section describes the functions for accessing the components of a marker object.

Function: marker-position marker

This function returns the position that marker points to, or nil if it points nowhere.

Function: marker-buffer marker

This function returns the buffer that marker points into, or nil if it points nowhere.

(setq m (make-marker))
     ⇒ #<marker in no buffer>
(marker-position m)
     ⇒ nil
(marker-buffer m)
     ⇒ nil

(set-marker m 3770 (current-buffer))
     ⇒ #<marker at 3770 in markers.texi>
(marker-buffer m)
     ⇒ #<buffer markers.texi>
(marker-position m)
     ⇒ 3770

32.5 Marker Insertion Types

When you insert text directly at the place where a marker points, there are two possible ways to relocate that marker: it can point before the inserted text, or point after it. You can specify which one a given marker should do by setting its insertion type. Note that use of insert-before-markers ignores markers’ insertion types, always relocating a marker to point after the inserted text.

Function: set-marker-insertion-type marker type

This function sets the insertion type of marker marker to type. If type is t, marker will advance when text is inserted at its position. If type is nil, marker does not advance when text is inserted there.

Function: marker-insertion-type marker

This function reports the current insertion type of marker.

All functions that create markers without accepting an argument that specifies the insertion type, create them with insertion type nil (see Functions that Create Markers). Also, the mark has, by default, insertion type nil.


32.6 Moving Marker Positions

This section describes how to change the position of an existing marker. When you do this, be sure you know whether the marker is used outside of your program, and, if so, what effects will result from moving it—otherwise, confusing things may happen in other parts of Emacs.

Function: set-marker marker position &optional buffer

This function moves marker to position in buffer. If buffer is not provided, it defaults to the current buffer.

If position is nil or a marker that points nowhere, then marker is set to point nowhere.

The value returned is marker.

(setq m (point-marker))
     ⇒ #<marker at 4714 in markers.texi>
(set-marker m 55)
     ⇒ #<marker at 55 in markers.texi>
(setq b (get-buffer "foo"))
     ⇒ #<buffer foo>
(set-marker m 0 b)
     ⇒ #<marker at 1 in foo>
Function: move-marker marker position &optional buffer

This is another name for set-marker.


32.7 The Mark

Each buffer has a special marker, which is designated the mark. When a buffer is newly created, this marker exists but does not point anywhere; this means that the mark doesn’t exist in that buffer yet. Subsequent commands can set the mark.

The mark specifies a position to bound a range of text for many commands, such as kill-region and indent-rigidly. These commands typically act on the text between point and the mark, which is called the region. If you are writing a command that operates on the region, don’t examine the mark directly; instead, use interactive with the ‘r’ specification. This provides the values of point and the mark as arguments to the command in an interactive call, but permits other Lisp programs to specify arguments explicitly. See Code Characters for interactive.

Some commands set the mark as a side-effect. Commands should do this only if it has a potential use to the user, and never for their own internal purposes. For example, the replace-regexp command sets the mark to the value of point before doing any replacements, because this enables the user to move back there conveniently after the replace is finished.

Once the mark exists in a buffer, it normally never ceases to exist. However, it may become inactive, if Transient Mark mode is enabled. The buffer-local variable mark-active, if non-nil, means that the mark is active. A command can call the function deactivate-mark to deactivate the mark directly, or it can request deactivation of the mark upon return to the editor command loop by setting the variable deactivate-mark to a non-nil value.

If Transient Mark mode is enabled, certain editing commands that normally apply to text near point, apply instead to the region when the mark is active. This is the main motivation for using Transient Mark mode. (Another is that this enables highlighting of the region when the mark is active. See Emacs Display.)

In addition to the mark, each buffer has a mark ring which is a list of markers containing previous values of the mark. When editing commands change the mark, they should normally save the old value of the mark on the mark ring. The variable mark-ring-max specifies the maximum number of entries in the mark ring; once the list becomes this long, adding a new element deletes the last element.

There is also a separate global mark ring, but that is used only in a few particular user-level commands, and is not relevant to Lisp programming. So we do not describe it here.

Function: mark &optional force

This function returns the current buffer’s mark position as an integer, or nil if no mark has ever been set in this buffer.

If Transient Mark mode is enabled, and mark-even-if-inactive is nil, mark signals an error if the mark is inactive. However, if force is non-nil, then mark disregards inactivity of the mark, and returns the mark position (or nil) anyway.

Function: mark-marker

This function returns the marker that represents the current buffer’s mark. It is not a copy, it is the marker used internally. Therefore, changing this marker’s position will directly affect the buffer’s mark. Don’t do that unless that is the effect you want.

(setq m (mark-marker))
     ⇒ #<marker at 3420 in markers.texi>
(set-marker m 100)
     ⇒ #<marker at 100 in markers.texi>
(mark-marker)
     ⇒ #<marker at 100 in markers.texi>

Like any marker, this marker can be set to point at any buffer you like. If you make it point at any buffer other than the one of which it is the mark, it will yield perfectly consistent, but rather odd, results. We recommend that you not do it!

Function: set-mark position

This function sets the mark to position, and activates the mark. The old value of the mark is not pushed onto the mark ring.

Please note: Use this function only if you want the user to see that the mark has moved, and you want the previous mark position to be lost. Normally, when a new mark is set, the old one should go on the mark-ring. For this reason, most applications should use push-mark and pop-mark, not set-mark.

Novice Emacs Lisp programmers often try to use the mark for the wrong purposes. The mark saves a location for the user’s convenience. An editing command should not alter the mark unless altering the mark is part of the user-level functionality of the command. (And, in that case, this effect should be documented.) To remember a location for internal use in the Lisp program, store it in a Lisp variable. For example:

(let ((beg (point)))
  (forward-line 1)
  (delete-region beg (point))).
Function: push-mark &optional position nomsg activate

This function sets the current buffer’s mark to position, and pushes a copy of the previous mark onto mark-ring. If position is nil, then the value of point is used.

The function push-mark normally does not activate the mark. To do that, specify t for the argument activate.

A ‘Mark set’ message is displayed unless nomsg is non-nil.

Function: pop-mark

This function pops off the top element of mark-ring and makes that mark become the buffer’s actual mark. This does not move point in the buffer, and it does nothing if mark-ring is empty. It deactivates the mark.

User Option: transient-mark-mode

This variable, if non-nil, enables Transient Mark mode. In Transient Mark mode, every buffer-modifying primitive sets deactivate-mark. As a consequence, most commands that modify the buffer also deactivate the mark.

When Transient Mark mode is enabled and the mark is active, many commands that normally apply to the text near point instead apply to the region. Such commands should use the function use-region-p to test whether they should operate on the region. See The Region.

Lisp programs can set transient-mark-mode to non-nil, non-t values to enable Transient Mark mode temporarily. If the value is lambda, Transient Mark mode is automatically turned off after any action, such as buffer modification, that would normally deactivate the mark. If the value is (only . oldval), then transient-mark-mode is set to the value oldval after any subsequent command that moves point and is not shift-translated (see shift-translation), or after any other action that would normally deactivate the mark. (Marking a region with the mouse will temporarily enable transient-mark-mode in this way.)

User Option: mark-even-if-inactive

If this is non-nil, Lisp programs and the Emacs user can use the mark even when it is inactive. This option affects the behavior of Transient Mark mode. When the option is non-nil, deactivation of the mark turns off region highlighting, but commands that use the mark behave as if the mark were still active.

Variable: deactivate-mark

If an editor command sets this variable non-nil, then the editor command loop deactivates the mark after the command returns (if Transient Mark mode is enabled). All the primitives that change the buffer set deactivate-mark, to deactivate the mark when the command is finished. Setting this variable makes it buffer-local.

To write Lisp code that modifies the buffer without causing deactivation of the mark at the end of the command, bind deactivate-mark to nil around the code that does the modification. For example:

(let (deactivate-mark)
  (insert " "))
Function: deactivate-mark &optional force

If Transient Mark mode is enabled or force is non-nil, this function deactivates the mark and runs the normal hook deactivate-mark-hook. Otherwise, it does nothing.

Variable: mark-active

The mark is active when this variable is non-nil. This variable is always buffer-local in each buffer. Do not use the value of this variable to decide whether a command that normally operates on text near point should operate on the region instead. Use the function use-region-p for that (see The Region).

Variable: activate-mark-hook
Variable: deactivate-mark-hook

These normal hooks are run, respectively, when the mark becomes active and when it becomes inactive. The hook activate-mark-hook is also run when the region is reactivated, for instance after using a command that switches back to a buffer that has an active mark.

Function: handle-shift-selection

This function implements the shift-selection behavior of point-motion commands. See Shift Selection in The GNU Emacs Manual. It is called automatically by the Emacs command loop whenever a command with a ‘^’ character in its interactive spec is invoked, before the command itself is executed (see ^).

If shift-select-mode is non-nil and the current command was invoked via shift translation (see shift-translation), this function sets the mark and temporarily activates the region, unless the region was already temporarily activated in this way. Otherwise, if the region has been activated temporarily, it deactivates the mark and restores the variable transient-mark-mode to its earlier value.

Variable: mark-ring

The value of this buffer-local variable is the list of saved former marks of the current buffer, most recent first.

mark-ring
⇒ (#<marker at 11050 in markers.texi>
    #<marker at 10832 in markers.texi>
    …)
User Option: mark-ring-max

The value of this variable is the maximum size of mark-ring. If more marks than this are pushed onto the mark-ring, push-mark discards an old mark when it adds a new one.

When Delete Selection mode (see Delete Selection in The GNU Emacs Manual) is enabled, commands that operate on the active region (a.k.a. “selection”) behave slightly differently. This works by adding the function delete-selection-pre-hook to the pre-command-hook (see Command Loop Overview). That function calls delete-selection-helper to delete the selection as appropriate for the command. If you want to adapt a command to Delete Selection mode, put the delete-selection property on the function’s symbol (see Accessing Symbol Properties); commands that don’t have this property on their symbol won’t delete the selection. This property can have one of several values to tailor the behavior to what the command is supposed to do; see the doc strings of delete-selection-pre-hook and delete-selection-helper for the details.


32.8 The Region

The text between point and the mark is known as the region. Various functions operate on text delimited by point and the mark, but only those functions specifically related to the region itself are described here.

The next two functions signal an error if the mark does not point anywhere. If Transient Mark mode is enabled and mark-even-if-inactive is nil, they also signal an error if the mark is inactive.

Function: region-beginning

This function returns the position of the beginning of the region (as an integer). This is the position of either point or the mark, whichever is smaller.

Function: region-end

This function returns the position of the end of the region (as an integer). This is the position of either point or the mark, whichever is larger.

Instead of using region-beginning and region-end, a command designed to operate on a region should normally use interactive with the ‘r’ specification to find the beginning and end of the region. This lets other Lisp programs specify the bounds explicitly as arguments. See Code Characters for interactive.

Function: use-region-p

This function returns t if Transient Mark mode is enabled, the mark is active, and there is a valid region in the buffer. This function is intended to be used by commands that operate on the region, instead of on text near point, when the mark is active.

A region is valid if it has a non-zero size, or if the user option use-empty-active-region is non-nil (by default, it is nil). The function region-active-p is similar to use-region-p, but considers all regions as valid. In most cases, you should not use region-active-p, since if the region is empty it is often more appropriate to operate on point.


33 Text

This chapter describes the functions that deal with the text in a buffer. Most examine, insert, or delete text in the current buffer, often operating at point or on text adjacent to point. Many are interactive. All the functions that change the text provide for undoing the changes (see Undo).

Many text-related functions operate on a region of text defined by two buffer positions passed in arguments named start and end. These arguments should be either markers (see Markers) or numeric character positions (see Positions). The order of these arguments does not matter; it is all right for start to be the end of the region and end the beginning. For example, (delete-region 1 10) and (delete-region 10 1) are equivalent. An args-out-of-range error is signaled if either start or end is outside the accessible portion of the buffer. In an interactive call, point and the mark are used for these arguments.

Throughout this chapter, “text” refers to the characters in the buffer, together with their properties (when relevant). Keep in mind that point is always between two characters, and the cursor appears on the character after point.


33.1 Examining Text Near Point

Many functions are provided to look at the characters around point. Several simple functions are described here. See also looking-at in Regular Expression Searching.

In the following four functions, “beginning” or “end” of buffer refers to the beginning or end of the accessible portion.

Function: char-after &optional position

This function returns the character in the current buffer at (i.e., immediately after) position position. If position is out of range for this purpose, either before the beginning of the buffer, or at or beyond the end, then the value is nil. The default for position is point.

In the following example, assume that the first character in the buffer is ‘@’:

(string (char-after 1))
     ⇒ "@"
Function: char-before &optional position

This function returns the character in the current buffer immediately before position position. If position is out of range for this purpose, either at or before the beginning of the buffer, or beyond the end, then the value is nil. The default for position is point.

Function: following-char

This function returns the character following point in the current buffer. This is similar to (char-after (point)). However, if point is at the end of the buffer, then following-char returns 0.

Remember that point is always between characters, and the cursor normally appears over the character following point. Therefore, the character returned by following-char is the character the cursor is over.

In this example, point is between the ‘a’ and the ‘c’.

---------- Buffer: foo ----------
Gentlemen may cry ``Pea∗ce! Peace!,''
but there is no peace.
---------- Buffer: foo ----------

(string (preceding-char))
     ⇒ "a"
(string (following-char))
     ⇒ "c"
Function: preceding-char

This function returns the character preceding point in the current buffer. See above, under following-char, for an example. If point is at the beginning of the buffer, preceding-char returns 0.

Function: bobp

This function returns t if point is at the beginning of the buffer. If narrowing is in effect, this means the beginning of the accessible portion of the text. See also point-min in Point.

Function: eobp

This function returns t if point is at the end of the buffer. If narrowing is in effect, this means the end of accessible portion of the text. See also point-max in See Point.

Function: bolp

This function returns t if point is at the beginning of a line. See Motion by Text Lines. The beginning of the buffer (or of its accessible portion) always counts as the beginning of a line.

Function: eolp

This function returns t if point is at the end of a line. The end of the buffer (or of its accessible portion) is always considered the end of a line.


33.2 Examining Buffer Contents

This section describes functions that allow a Lisp program to convert any portion of the text in the buffer into a string.

Function: buffer-substring start end

This function returns a string containing a copy of the text of the region defined by positions start and end in the current buffer. If the arguments are not positions in the accessible portion of the buffer, buffer-substring signals an args-out-of-range error.

Here’s an example which assumes Font-Lock mode is not enabled:

---------- Buffer: foo ----------
This is the contents of buffer foo

---------- Buffer: foo ----------

(buffer-substring 1 10)
     ⇒ "This is t"
(buffer-substring (point-max) 10)
     ⇒ "he contents of buffer foo\n"

If the text being copied has any text properties, these are copied into the string along with the characters they belong to. See Text Properties. However, overlays (see Overlays) in the buffer and their properties are ignored, not copied.

For example, if Font-Lock mode is enabled, you might get results like these:

(buffer-substring 1 10)
     ⇒ #("This is t" 0 1 (fontified t) 1 9 (fontified t))
Function: buffer-substring-no-properties start end

This is like buffer-substring, except that it does not copy text properties, just the characters themselves. See Text Properties.

Function: buffer-string

This function returns the contents of the entire accessible portion of the current buffer, as a string. If the text being copied has any text properties, these are copied into the string along with the characters they belong to.

If you need to make sure the resulting string, when copied to a different location, will not change its visual appearance due to reordering of bidirectional text, use the buffer-substring-with-bidi-context function (see buffer-substring-with-bidi-context).

Function: filter-buffer-substring start end &optional delete

This function filters the buffer text between start and end using a function specified by the variable filter-buffer-substring-function, and returns the result.

The default filter function consults the obsolete wrapper hook filter-buffer-substring-functions (see the documentation string of the macro with-wrapper-hook for the details about this obsolete facility). If it is nil, it returns the unaltered text from the buffer, i.e., what buffer-substring would return.

If delete is non-nil, the function deletes the text between start and end after copying it, like delete-and-extract-region.

Lisp code should use this function instead of buffer-substring, buffer-substring-no-properties, or delete-and-extract-region when copying into user-accessible data structures such as the kill-ring, X clipboard, and registers. Major and minor modes can modify filter-buffer-substring-function to alter such text as it is copied out of the buffer.

Variable: filter-buffer-substring-function

The value of this variable is a function that filter-buffer-substring will call to do the actual work. The function receives three arguments, the same as those of filter-buffer-substring, which it should treat as per the documentation of that function. It should return the filtered text (and optionally delete the source text).

The following two variables are obsoleted by filter-buffer-substring-function, but are still supported for backward compatibility.

Variable: filter-buffer-substring-functions

This obsolete variable is a wrapper hook, whose members should be functions that accept four arguments: fun, start, end, and delete. fun is a function that takes three arguments (start, end, and delete), and returns a string. In both cases, the start, end, and delete arguments are the same as those of filter-buffer-substring.

The first hook function is passed a fun that is equivalent to the default operation of filter-buffer-substring, i.e., it returns the buffer-substring between start and end and optionally deletes the original text from the buffer. In most cases, the hook function will call fun once, and then do its own processing of the result. The next hook function receives a fun equivalent to this, and so on. The actual return value is the result of all the hook functions acting in sequence.

Function: current-word &optional strict really-word

This function returns the symbol (or word) at or near point, as a string. The return value includes no text properties.

If the optional argument really-word is non-nil, it finds a word; otherwise, it finds a symbol (which includes both word characters and symbol constituent characters).

If the optional argument strict is non-nil, then point must be in or next to the symbol or word—if no symbol or word is there, the function returns nil. Otherwise, a nearby symbol or word on the same line is acceptable.

Function: thing-at-point thing &optional no-properties

Return the thing around or next to point, as a string.

The argument thing is a symbol which specifies a kind of syntactic entity. Possibilities include symbol, list, sexp, defun, filename, existing-filename, url, word, sentence, whitespace, line, page, string, and others.

When the optional argument no-properties is non-nil, this function strips text properties from the return value.

---------- Buffer: foo ----------
Gentlemen may cry ``Pea∗ce! Peace!,''
but there is no peace.
---------- Buffer: foo ----------

(thing-at-point 'word)
     ⇒ "Peace"
(thing-at-point 'line)
     ⇒ "Gentlemen may cry ``Peace! Peace!,''\n"
(thing-at-point 'whitespace)
     ⇒ nil
Variable: thing-at-point-provider-alist

This variable allows users and modes to tweak how thing-at-point works. It’s an association list of things and functions (called with zero parameters) to return that thing. Entries for thing will be evaluated in turn until a non-nil result is returned.

For instance, a major mode could say:

(setq-local thing-at-point-provider-alist
            (append thing-at-point-provider-alist
                    '((url . my-mode--url-at-point))))

If no providers have a non-nil return, the thing will be computed the standard way.


33.3 Comparing Text

This function lets you compare portions of the text in a buffer, without copying them into strings first.

Function: compare-buffer-substrings buffer1 start1 end1 buffer2 start2 end2

This function lets you compare two substrings of the same buffer or two different buffers. The first three arguments specify one substring, giving a buffer (or a buffer name) and two positions within the buffer. The last three arguments specify the other substring in the same way. You can use nil for buffer1, buffer2, or both to stand for the current buffer.

The value is negative if the first substring is less, positive if the first is greater, and zero if they are equal. The absolute value of the result is one plus the index of the first differing characters within the substrings.

This function ignores case when comparing characters if case-fold-search is non-nil. It always ignores text properties.

Suppose you have the text ‘foobarbar haha!rara! in the current buffer; then in this example the two substrings are ‘rbar ’ and ‘rara!’. The value is 2 because the first substring is greater at the second character.

(compare-buffer-substrings nil 6 11 nil 16 21)
     ⇒ 2

33.4 Inserting Text

Insertion means adding new text to a buffer. The inserted text goes at point—between the character before point and the character after point. Some insertion functions leave point before the inserted text, while other functions leave it after. We call the former insertion after point and the latter insertion before point.

Insertion moves markers located at positions after the insertion point, so that they stay with the surrounding text (see Markers). When a marker points at the place of insertion, insertion may or may not relocate the marker, depending on the marker’s insertion type (see Marker Insertion Types). Certain special functions such as insert-before-markers relocate all such markers to point after the inserted text, regardless of the markers’ insertion type.

Insertion functions signal an error if the current buffer is read-only (see Read-Only Buffers) or if they insert within read-only text (see Properties with Special Meanings).

These functions copy text characters from strings and buffers along with their properties. The inserted characters have exactly the same properties as the characters they were copied from. By contrast, characters specified as separate arguments, not part of a string or buffer, inherit their text properties from the neighboring text.

The insertion functions convert text from unibyte to multibyte in order to insert in a multibyte buffer, and vice versa—if the text comes from a string or from a buffer. However, they do not convert unibyte character codes 128 through 255 to multibyte characters, not even if the current buffer is a multibyte buffer. See Converting Text Representations.

Function: insert &rest args

This function inserts the strings and/or characters args into the current buffer, at point, moving point forward. In other words, it inserts the text before point. An error is signaled unless all args are either strings or characters. The value is nil.

Function: insert-before-markers &rest args

This function inserts the strings and/or characters args into the current buffer, at point, moving point forward. An error is signaled unless all args are either strings or characters. The value is nil.

This function is unlike the other insertion functions in that it relocates markers initially pointing at the insertion point, to point after the inserted text. If an overlay begins at the insertion point, the inserted text falls outside the overlay; if a nonempty overlay ends at the insertion point, the inserted text falls inside that overlay.

Command: insert-char character &optional count inherit

This command inserts count instances of character into the current buffer before point. The argument count must be an integer, and character must be a character.

If called interactively, this command prompts for character using its Unicode name or its code point. See Inserting Text in The GNU Emacs Manual.

This function does not convert unibyte character codes 128 through 255 to multibyte characters, not even if the current buffer is a multibyte buffer. See Converting Text Representations.

If inherit is non-nil, the inserted characters inherit sticky text properties from the two characters before and after the insertion point. See Stickiness of Text Properties.

Function: insert-buffer-substring from-buffer-or-name &optional start end

This function inserts a portion of buffer from-buffer-or-name into the current buffer before point. The text inserted is the region between start (inclusive) and end (exclusive). (These arguments default to the beginning and end of the accessible portion of that buffer.) This function returns nil.

In this example, the form is executed with buffer ‘bar’ as the current buffer. We assume that buffer ‘bar’ is initially empty.

---------- Buffer: foo ----------
We hold these truths to be self-evident, that all
---------- Buffer: foo ----------

(insert-buffer-substring "foo" 1 20)
     ⇒ nil

---------- Buffer: bar ----------
We hold these truth∗
---------- Buffer: bar ----------
Function: insert-buffer-substring-no-properties from-buffer-or-name &optional start end

This is like insert-buffer-substring except that it does not copy any text properties.

Function: insert-into-buffer to-buffer &optional start end

This is like insert-buffer-substring, but works in the opposite direction: The text is copied from the current buffer into to-buffer. The block of text is copied to the current point in to-buffer, and point (in that buffer) is advanced to after the end of the copied text. Is start/end is nil, the entire text in the current buffer is copied over.

See Stickiness of Text Properties, for other insertion functions that inherit text properties from the nearby text in addition to inserting it. Whitespace inserted by indentation functions also inherits text properties.


33.5 User-Level Insertion Commands

This section describes higher-level commands for inserting text, commands intended primarily for the user but useful also in Lisp programs.

Command: insert-buffer from-buffer-or-name

This command inserts the entire accessible contents of from-buffer-or-name (which must exist) into the current buffer after point. It leaves the mark after the inserted text. The value is nil.

Command: self-insert-command count &optional char

This command inserts the character char (the last character typed); it does so count times, before point, and returns nil. Most printing characters are bound to this command. In routine use, self-insert-command is the most frequently called function in Emacs, but programs rarely use it except to install it on a keymap.

In an interactive call, count is the numeric prefix argument.

Self-insertion translates the input character through translation-table-for-input. See Translation of Characters.

This command calls auto-fill-function whenever that is non-nil and the character inserted is in the table auto-fill-chars (see Auto Filling).

This command performs abbrev expansion if Abbrev mode is enabled and the inserted character does not have word-constituent syntax. (See Abbrevs and Abbrev Expansion, and Table of Syntax Classes.) It is also responsible for calling blink-paren-function when the inserted character has close parenthesis syntax (see Blinking Parentheses).

The final thing this command does is to run the hook post-self-insert-hook. You could use this to automatically reindent text as it is typed, for example. The functions on this hook can use last-command-event (see Information from the Command Loop) to access the character just inserted.

If any function on this hook needs to act on the region (see The Region), it should make sure Delete Selection mode (see Delete Selection in The GNU Emacs Manual) doesn’t delete the region before post-self-insert-hook functions are invoked. The way to do so is to add a function that returns nil to self-insert-uses-region-functions, a special hook that tells Delete Selection mode it should not delete the region.

Do not try substituting your own definition of self-insert-command for the standard one. The editor command loop handles this function specially.

Command: newline &optional number-of-newlines interactive

This command inserts newlines into the current buffer before point. If number-of-newlines is supplied, that many newline characters are inserted. In an interactive call, number-of-newlines is the numeric prefix argument.

This command calls self-insert-command to insert newlines, which may subsequently break the preceding line by calling auto-fill-function (see Auto Filling). Typically what auto-fill-function does is insert a newline; thus, the overall result in this case is to insert two newlines at different places: one at point, and another earlier in the line. newline does not auto-fill if number-of-newlines is non-nil.

This command does not run the hook post-self-insert-hook unless called interactively or interactive is non-nil.

This command indents to the left margin if that is not zero. See Margins for Filling.

The value returned is nil.

Command: ensure-empty-lines &optional number-of-empty-lines

This command can be used to ensure that you have a specific number of empty lines before point. (An “empty line” is here defined as a line with no characters on it—a line with space characters isn’t an empty line.) It defaults to ensuring that there’s a single empty line before point.

If point isn’t at the beginning of a line, a newline character is inserted first. If there’s more empty lines before point than specified, the number of empty lines is reduced. Otherwise it’s increased to the specified number.

Variable: overwrite-mode

This variable controls whether overwrite mode is in effect. The value should be overwrite-mode-textual, overwrite-mode-binary, or nil. overwrite-mode-textual specifies textual overwrite mode (treats newlines and tabs specially), and overwrite-mode-binary specifies binary overwrite mode (treats newlines and tabs like any other characters).


33.6 Deleting Text

Deletion means removing part of the text in a buffer, without saving it in the kill ring (see The Kill Ring). Deleted text can’t be yanked, but can be reinserted using the undo mechanism (see Undo). Some deletion functions do save text in the kill ring in some special cases.

All of the deletion functions operate on the current buffer.

Command: erase-buffer

This function deletes the entire text of the current buffer (not just the accessible portion), leaving it empty. If the buffer is read-only, it signals a buffer-read-only error; if some of the text in it is read-only, it signals a text-read-only error. Otherwise, it deletes the text without asking for any confirmation. It returns nil.

Normally, deleting a large amount of text from a buffer inhibits further auto-saving of that buffer because it has shrunk. However, erase-buffer does not do this, the idea being that the future text is not really related to the former text, and its size should not be compared with that of the former text.

Command: delete-region start end

This command deletes the text between positions start and end in the current buffer, and returns nil. If point was inside the deleted region, its value afterward is start. Otherwise, point relocates with the surrounding text, as markers do.

Function: delete-and-extract-region start end

This function deletes the text between positions start and end in the current buffer, and returns a string containing the text just deleted.

If point was inside the deleted region, its value afterward is start. Otherwise, point relocates with the surrounding text, as markers do.

Command: delete-char count &optional killp

This command deletes count characters directly after point, or before point if count is negative. If killp is non-nil, then it saves the deleted characters in the kill ring.

In an interactive call, count is the numeric prefix argument, and killp is the unprocessed prefix argument. Therefore, if a prefix argument is supplied, the text is saved in the kill ring. If no prefix argument is supplied, then one character is deleted, but not saved in the kill ring.

The value returned is always nil.

Command: delete-backward-char count &optional killp

This command deletes count characters directly before point, or after point if count is negative. If killp is non-nil, then it saves the deleted characters in the kill ring.

In an interactive call, count is the numeric prefix argument, and killp is the unprocessed prefix argument. Therefore, if a prefix argument is supplied, the text is saved in the kill ring. If no prefix argument is supplied, then one character is deleted, but not saved in the kill ring.

The value returned is always nil.

Command: backward-delete-char-untabify count &optional killp

This command deletes count characters backward, changing tabs into spaces. When the next character to be deleted is a tab, it is first replaced with the proper number of spaces to preserve alignment and then one of those spaces is deleted instead of the tab. If killp is non-nil, then the command saves the deleted characters in the kill ring.

Conversion of tabs to spaces happens only if count is positive. If it is negative, exactly −count characters after point are deleted.

In an interactive call, count is the numeric prefix argument, and killp is the unprocessed prefix argument. Therefore, if a prefix argument is supplied, the text is saved in the kill ring. If no prefix argument is supplied, then one character is deleted, but not saved in the kill ring.

The value returned is always nil.

User Option: backward-delete-char-untabify-method

This option specifies how backward-delete-char-untabify should deal with whitespace. Possible values include untabify, the default, meaning convert a tab to many spaces and delete one; hungry, meaning delete all tabs and spaces before point with one command; all meaning delete all tabs, spaces and newlines before point, and nil, meaning do nothing special for whitespace characters.


33.7 User-Level Deletion Commands

This section describes higher-level commands for deleting text, commands intended primarily for the user but useful also in Lisp programs.

Command: delete-horizontal-space &optional backward-only

This function deletes all spaces and tabs around point. It returns nil.

If backward-only is non-nil, the function deletes spaces and tabs before point, but not after point.

In the following examples, we call delete-horizontal-space four times, once on each line, with point between the second and third characters on the line each time.

---------- Buffer: foo ----------
I ∗thought
I ∗     thought
We∗ thought
Yo∗u thought
---------- Buffer: foo ----------

(delete-horizontal-space)   ; Four times.
     ⇒ nil

---------- Buffer: foo ----------
Ithought
Ithought
Wethought
You thought
---------- Buffer: foo ----------
Command: delete-indentation &optional join-following-p beg end

This function joins the line point is on to the previous line, deleting any whitespace at the join and in some cases replacing it with one space. If join-following-p is non-nil, delete-indentation joins this line to the following line instead. Otherwise, if beg and end are non-nil, this function joins all lines in the region they define.

In an interactive call, join-following-p is the prefix argument, and beg and end are, respectively, the start and end of the region if it is active, else nil. The function returns nil.

If there is a fill prefix, and the second of the lines being joined starts with the prefix, then delete-indentation deletes the fill prefix before joining the lines. See Margins for Filling.

In the example below, point is located on the line starting ‘events’, and it makes no difference if there are trailing spaces in the preceding line.

---------- Buffer: foo ----------
When in the course of human
∗    events, it becomes necessary
---------- Buffer: foo ----------

(delete-indentation)
     ⇒ nil

---------- Buffer: foo ----------
When in the course of human∗ events, it becomes necessary
---------- Buffer: foo ----------

After the lines are joined, the function fixup-whitespace is responsible for deciding whether to leave a space at the junction.

Command: fixup-whitespace

This function replaces all the horizontal whitespace surrounding point with either one space or no space, according to the context. It returns nil.

At the beginning or end of a line, the appropriate amount of space is none. Before a character with close parenthesis syntax, or after a character with open parenthesis or expression-prefix syntax, no space is also appropriate. Otherwise, one space is appropriate. See Table of Syntax Classes.

In the example below, fixup-whitespace is called the first time with point before the word ‘spaces’ in the first line. For the second invocation, point is directly after the ‘(’.

---------- Buffer: foo ----------
This has too many     ∗spaces
This has too many spaces at the start of (∗   this list)
---------- Buffer: foo ----------

(fixup-whitespace)
     ⇒ nil
(fixup-whitespace)
     ⇒ nil

---------- Buffer: foo ----------
This has too many spaces
This has too many spaces at the start of (this list)
---------- Buffer: foo ----------
Command: just-one-space &optional n

This command replaces any spaces and tabs around point with a single space, or n spaces if n is specified. It returns nil.

Command: delete-blank-lines

This function deletes blank lines surrounding point. If point is on a blank line with one or more blank lines before or after it, then all but one of them are deleted. If point is on an isolated blank line, then it is deleted. If point is on a nonblank line, the command deletes all blank lines immediately following it.

A blank line is defined as a line containing only tabs and spaces.

delete-blank-lines returns nil.

Command: delete-trailing-whitespace &optional start end

Delete trailing whitespace in the region defined by start and end.

This command deletes whitespace characters after the last non-whitespace character in each line in the region.

If this command acts on the entire buffer (i.e., if called interactively with the mark inactive, or called from Lisp with end nil), it also deletes all trailing lines at the end of the buffer if the variable delete-trailing-lines is non-nil.


33.8 The Kill Ring

Kill functions delete text like the deletion functions, but save it so that the user can reinsert it by yanking. Most of these functions have ‘kill-’ in their name. By contrast, the functions whose names start with ‘delete-’ normally do not save text for yanking (though they can still be undone); these are deletion functions.

Most of the kill commands are primarily for interactive use, and are not described here. What we do describe are the functions provided for use in writing such commands. You can use these functions to write commands for killing text. When you need to delete text for internal purposes within a Lisp function, you should normally use deletion functions, so as not to disturb the kill ring contents. See Deleting Text.

Killed text is saved for later yanking in the kill ring. This is a list that holds a number of recent kills, not just the last text kill. We call this a “ring” because yanking treats it as having elements in a cyclic order. The list is kept in the variable kill-ring, and can be operated on with the usual functions for lists; there are also specialized functions, described in this section, that treat it as a ring.

Some people think this use of the word “kill” is unfortunate, since it refers to operations that specifically do not destroy the entities killed. This is in sharp contrast to ordinary life, in which death is permanent and killed entities do not come back to life. Therefore, other metaphors have been proposed. For example, the term “cut ring” makes sense to people who, in pre-computer days, used scissors and paste to cut up and rearrange manuscripts. However, it would be difficult to change the terminology now.


33.8.1 Kill Ring Concepts

The kill ring records killed text as strings in a list, most recent first. A short kill ring, for example, might look like this:

("some text" "a different piece of text" "even older text")

When the list reaches kill-ring-max entries in length, adding a new entry automatically deletes the last entry.

When kill commands are interwoven with other commands, each kill command makes a new entry in the kill ring. Multiple kill commands in succession build up a single kill ring entry, which would be yanked as a unit; the second and subsequent consecutive kill commands add text to the entry made by the first one.

For yanking, one entry in the kill ring is designated the front of the ring. Some yank commands rotate the ring by designating a different element as the front. But this virtual rotation doesn’t change the list itself—the most recent entry always comes first in the list.


33.8.2 Functions for Killing

kill-region is the usual subroutine for killing text. Any command that calls this function is a kill command (and should probably have ‘kill’ in its name). kill-region puts the newly killed text in a new element at the beginning of the kill ring or adds it to the most recent element. It determines automatically (using last-command) whether the previous command was a kill command, and if so appends the killed text to the most recent entry.

The commands described below can filter the killed text before they save it in the kill ring. They call filter-buffer-substring (see Examining Buffer Contents) to perform the filtering. By default, there’s no filtering, but major and minor modes and hook functions can set up filtering, so that text saved in the kill ring is different from what was in the buffer.

Command: kill-region start end &optional region

This function kills the stretch of text between start and end; but if the optional argument region is non-nil, it ignores start and end, and kills the text in the current region instead. The text is deleted but saved in the kill ring, along with its text properties. The value is always nil.

In an interactive call, start and end are point and the mark, and region is always non-nil, so the command always kills the text in the current region.

If the buffer or text is read-only, kill-region modifies the kill ring just the same, then signals an error without modifying the buffer. This is convenient because it lets the user use a series of kill commands to copy text from a read-only buffer into the kill ring.

User Option: kill-read-only-ok

If this option is non-nil, kill-region does not signal an error if the buffer or text is read-only. Instead, it simply returns, updating the kill ring but not changing the buffer.

Command: copy-region-as-kill start end &optional region

This function saves the stretch of text between start and end on the kill ring (including text properties), but does not delete the text from the buffer. However, if the optional argument region is non-nil, the function ignores start and end, and saves the current region instead. It always returns nil.

In an interactive call, start and end are point and the mark, and region is always non-nil, so the command always saves the text in the current region.

The command does not set this-command to kill-region, so a subsequent kill command does not append to the same kill ring entry.


33.8.3 Yanking

Yanking means inserting text from the kill ring, but it does not insert the text blindly. The yank command, and related commands, use insert-for-yank to perform special processing on the text before it is inserted.

Function: insert-for-yank string

This function works like insert, except that it processes the text in string according to the yank-handler text property, as well as the variables yank-handled-properties and yank-excluded-properties (see below), before inserting the result into the current buffer.

string will be run through yank-transform-functions (see below) before inserting.

Function: insert-buffer-substring-as-yank buf &optional start end

This function resembles insert-buffer-substring, except that it processes the text according to yank-handled-properties and yank-excluded-properties. (It does not handle the yank-handler property, which does not normally occur in buffer text anyway.)

If you put a yank-handler text property on all or part of a string, that alters how insert-for-yank inserts the string. If different parts of the string have different yank-handler values (comparison being done with eq), each substring is handled separately. The property value must be a list of one to four elements, with the following format (where elements after the first may be omitted):

(function param noexclude undo)

Here is what the elements do:

function

When function is non-nil, it is called instead of insert to insert the string, with one argument—the string to insert.

param

If param is present and non-nil, it replaces string (or the substring of string being processed) as the object passed to function (or insert). For example, if function is yank-rectangle, param should be a list of strings to insert as a rectangle.

noexclude

If noexclude is present and non-nil, that disables the normal action of yank-handled-properties and yank-excluded-properties on the inserted string.

undo

If undo is present and non-nil, it is a function that will be called by yank-pop to undo the insertion of the current object. It is called with two arguments, the start and end of the current region. function can set yank-undo-function to override the undo value.

User Option: yank-handled-properties

This variable specifies special text property handling conditions for yanked text. It takes effect after the text has been inserted (either normally, or via the yank-handler property), and prior to yank-excluded-properties taking effect.

The value should be an alist of elements (prop . fun). Each alist element is handled in order. The inserted text is scanned for stretches of text having text properties eq to prop; for each such stretch, fun is called with three arguments: the value of the property, and the start and end positions of the text.

User Option: yank-excluded-properties

The value of this variable is the list of properties to remove from inserted text. Its default value contains properties that might lead to annoying results, such as causing the text to respond to the mouse or specifying key bindings. It takes effect after yank-handled-properties.

Variable: yank-transform-functions

This variable is a list of functions. Each function is called (in order) with the string to be yanked as the argument, and should return a (possibly transformed) string. This variable can be set globally, but can also be used to create new commands that are variations on yank. For instance, to create a command that works like yank, but cleans up whitespace before inserting, you could say something like:

(defun yank-with-clean-whitespace ()
  (interactive)
  (let ((yank-transform-functions
	 '(string-clean-whitespace)))
    (call-interactively #'yank)))

33.8.4 Functions for Yanking

This section describes higher-level commands for yanking, which are intended primarily for the user but useful also in Lisp programs. Both yank and yank-pop honor the yank-excluded-properties variable and yank-handler text property (see Yanking).

Command: yank &optional arg

This command inserts before point the text at the front of the kill ring. It sets the mark at the beginning of that text, using push-mark (see The Mark), and puts point at the end.

If arg is a non-nil list (which occurs interactively when the user types C-u with no digits), then yank inserts the text as described above, but puts point before the yanked text and sets the mark after it.

If arg is a number, then yank inserts the argth most recently killed text—the argth element of the kill ring list, counted cyclically from the front, which is considered the first element for this purpose.

yank does not alter the contents of the kill ring, unless it used text provided by another program, in which case it pushes that text onto the kill ring. However if arg is an integer different from one, it rotates the kill ring to place the yanked string at the front.

yank returns nil.

Command: yank-pop &optional arg

When invoked immediately after a yank or another yank-pop, this command replaces the just-yanked entry from the kill ring with a different entry from the kill ring. When this command is invoked like that, the region contains text that was just inserted by another yank command. yank-pop deletes that text and inserts in its place a different piece of killed text. It does not add the deleted text to the kill ring, since it is already in the kill ring somewhere. It does however rotate the kill ring to place the newly yanked string at the front.

If arg is nil, then the replacement text is the previous element of the kill ring. If arg is numeric, the replacement is the argth previous kill. If arg is negative, a more recent kill is the replacement.

The sequence of kills in the kill ring wraps around, so if yank-pop is invoked repeatedly and reaches the oldest kill, the one that comes after it is the newest one, and the one before the newest one is the oldest one.

This command can also be invoked after a command that is not a yank command. In that case, it prompts in the minibuffer for a kill-ring entry, with completion, and uses the kill ring elements as the minibuffer history (see Minibuffer History). This allows the user to interactively select one of the previous kills recorded in the kill ring.

The return value is always nil.

Variable: yank-undo-function

If this variable is non-nil, the function yank-pop uses its value instead of delete-region to delete the text inserted by the previous yank or yank-pop command. The value must be a function of two arguments, the start and end of the current region.

The function insert-for-yank automatically sets this variable according to the undo element of the yank-handler text property, if there is one.


33.8.5 Low-Level Kill Ring

These functions and variables provide access to the kill ring at a lower level, but are still convenient for use in Lisp programs, because they take care of interaction with window system selections (see Window System Selections).

Function: current-kill n &optional do-not-move

The function current-kill rotates the yanking pointer, which designates the front of the kill ring, by n places (from newer kills to older ones), and returns the text at that place in the ring.

If the optional second argument do-not-move is non-nil, then current-kill doesn’t alter the yanking pointer; it just returns the nth kill, counting from the current yanking pointer.

If n is zero, indicating a request for the latest kill, current-kill calls the value of interprogram-paste-function (documented below) before consulting the kill ring. If that value is a function and calling it returns a string or a list of several strings, current-kill pushes the strings onto the kill ring and returns the first string. It also sets the yanking pointer to point to the kill-ring entry of the first string returned by interprogram-paste-function, regardless of the value of do-not-move. Otherwise, current-kill does not treat a zero value for n specially: it returns the entry pointed at by the yanking pointer and does not move the yanking pointer.

Function: kill-new string &optional replace

This function pushes the text string onto the kill ring and makes the yanking pointer point to it. It discards the oldest entry if appropriate. It also invokes the values of interprogram-paste-function (subject to the user option save-interprogram-paste-before-kill) and interprogram-cut-function (see below).

If replace is non-nil, then kill-new replaces the first element of the kill ring with string, rather than pushing string onto the kill ring.

Function: kill-append string before-p

This function appends the text string to the first entry in the kill ring and makes the yanking pointer point to the combined entry. Normally string goes at the end of the entry, but if before-p is non-nil, it goes at the beginning. This function calls kill-new as a subroutine, thus causing the values of interprogram-cut-function and possibly interprogram-paste-function (see below) to be invoked by extension.

Variable: interprogram-paste-function

This variable provides a way of transferring killed text from other programs, when you are using a window system. Its value should be nil or a function of no arguments.

If the value is a function, current-kill calls it to get the most recent kill. If the function returns a non-nil value, then that value is used as the most recent kill. If it returns nil, then the front of the kill ring is used.

To facilitate support for window systems that support multiple selections, this function may also return a list of strings. In that case, the first string is used as the most recent kill, and all the other strings are pushed onto the kill ring, for easy access by yank-pop.

The normal use of this function is to get the window system’s clipboard as the most recent kill, even if the selection belongs to another application. See Window System Selections. However, if the clipboard contents come from the current Emacs session, this function should return nil.

Variable: interprogram-cut-function

This variable provides a way of communicating killed text to other programs, when you are using a window system. Its value should be nil or a function of one required argument.

If the value is a function, kill-new and kill-append call it with the new first element of the kill ring as the argument.

The normal use of this function is to put newly killed text in the window system’s clipboard. See Window System Selections.


33.8.6 Internals of the Kill Ring

The variable kill-ring holds the kill ring contents, in the form of a list of strings. The most recent kill is always at the front of the list.

The kill-ring-yank-pointer variable points to a link in the kill ring list, whose CAR is the text to yank next. We say it identifies the front of the ring. Moving kill-ring-yank-pointer to a different link is called rotating the kill ring. We call the kill ring a “ring” because the functions that move the yank pointer wrap around from the end of the list to the beginning, or vice-versa. Rotation of the kill ring is virtual; it does not change the value of kill-ring.

Both kill-ring and kill-ring-yank-pointer are Lisp variables whose values are normally lists. The word “pointer” in the name of the kill-ring-yank-pointer indicates that the variable’s purpose is to identify one element of the list for use by the next yank command.

The value of kill-ring-yank-pointer is always eq to one of the links in the kill ring list. The element it identifies is the CAR of that link. Kill commands, which change the kill ring, also set this variable to the value of kill-ring. The effect is to rotate the ring so that the newly killed text is at the front.

Here is a diagram that shows the variable kill-ring-yank-pointer pointing to the second entry in the kill ring ("some text" "a different piece of text" "yet older text").

kill-ring                  ---- kill-ring-yank-pointer
  |                       |
  |                       v
  |     --- ---          --- ---      --- ---
   --> |   |   |------> |   |   |--> |   |   |--> nil
        --- ---          --- ---      --- ---
         |                |            |
         |                |            |
         |                |             -->"yet older text"
         |                |
         |                 --> "a different piece of text"
         |
          --> "some text"

This state of affairs might occur after C-y (yank) immediately followed by M-y (yank-pop).

Variable: kill-ring

This variable holds the list of killed text sequences, most recently killed first.

Variable: kill-ring-yank-pointer

This variable’s value indicates which element of the kill ring is at the front of the ring for yanking. More precisely, the value is a tail of the value of kill-ring, and its CAR is the kill string that C-y should yank.

User Option: kill-ring-max

The value of this variable is the maximum length to which the kill ring can grow, before elements are thrown away at the end. The default value for kill-ring-max is 120.


33.9 Undo

Most buffers have an undo list, which records all changes made to the buffer’s text so that they can be undone. (The buffers that don’t have one are usually special-purpose buffers for which Emacs assumes that undoing is not useful. In particular, any buffer whose name begins with a space has its undo recording off by default; see Buffer Names.) All the primitives that modify the text in the buffer automatically add elements to the front of the undo list, which is in the variable buffer-undo-list.

Variable: buffer-undo-list

This buffer-local variable’s value is the undo list of the current buffer. A value of t disables the recording of undo information.

Here are the kinds of elements an undo list can have:

position

This kind of element records a previous value of point; undoing this element moves point to position. Ordinary cursor motion does not make any sort of undo record, but deletion operations use these entries to record where point was before the command.

(beg . end)

This kind of element indicates how to delete text that was inserted. Upon insertion, the text occupied the range begend in the buffer.

(text . position)

This kind of element indicates how to reinsert text that was deleted. The deleted text itself is the string text. The place to reinsert it is (abs position). If position is positive, point was at the beginning of the deleted text, otherwise it was at the end. Zero or more (marker . adjustment) elements follow immediately after this element.

(t . time-flag)

This kind of element indicates that an unmodified buffer became modified. A time-flag that is a non-integer Lisp timestamp represents the visited file’s modification time as of when it was previously visited or saved, using the same format as current-time; see Time of Day. A time-flag of 0 means the buffer does not correspond to any file; −1 means the visited file previously did not exist. primitive-undo uses these values to determine whether to mark the buffer as unmodified once again; it does so only if the file’s status matches that of time-flag.

(nil property value beg . end)

This kind of element records a change in a text property. Here’s how you might undo the change:

(put-text-property beg end property value)
(marker . adjustment)

This kind of element records the fact that the marker marker was relocated due to deletion of surrounding text, and that it moved adjustment character positions. If the marker’s location is consistent with the (text . position) element preceding it in the undo list, then undoing this element moves markeradjustment characters.

(apply funname . args)

This is an extensible undo item, which is undone by calling funname with arguments args.

(apply delta beg end funname . args)

This is an extensible undo item, which records a change limited to the range beg to end, which increased the size of the buffer by delta characters. It is undone by calling funname with arguments args.

This kind of element enables undo limited to a region to determine whether the element pertains to that region.

nil

This element is a boundary. The elements between two boundaries are called a change group; normally, each change group corresponds to one keyboard command, and undo commands normally undo an entire group as a unit.

Function: undo-boundary

This function places a boundary element in the undo list. The undo command stops at such a boundary, and successive undo commands undo to earlier and earlier boundaries. This function returns nil.

Calling this function explicitly is useful for splitting the effects of a command into more than one unit. For example, query-replace calls undo-boundary after each replacement, so that the user can undo individual replacements one by one.

Mostly, however, this function is called automatically at an appropriate time.

Function: undo-auto-amalgamate

The editor command loop automatically calls undo-boundary just before executing each key sequence, so that each undo normally undoes the effects of one command. A few exceptional commands are amalgamating: these commands generally cause small changes to buffers, so with these a boundary is inserted only every 20th command, allowing the changes to be undone as a group. By default, the commands self-insert-command, which produces self-inserting input characters (see User-Level Insertion Commands), and delete-char, which deletes characters (see Deleting Text), are amalgamating. Where a command affects the contents of several buffers, as may happen, for example, when a function on the post-command-hook affects a buffer other than the current-buffer, then undo-boundary will be called in each of the affected buffers.

This function can be called before an amalgamating command. It removes the previous undo-boundary if a series of such calls have been made.

The maximum number of changes that can be amalgamated is controlled by the amalgamating-undo-limit variable. If this variable is 1, no changes are amalgamated.

A Lisp program can amalgamate a series of changes into a single change group by calling undo-amalgamate-change-group (see Atomic Change Groups). Note that amalgamating-undo-limit has no effect on the groups produced by that function.

Variable: undo-auto-current-boundary-timer

Some buffers, such as process buffers, can change even when no commands are executing. In these cases, undo-boundary is normally called periodically by the timer in this variable. Setting this variable to non-nil prevents this behavior.

Variable: undo-in-progress

This variable is normally nil, but the undo commands bind it to t. This is so that various kinds of change hooks can tell when they’re being called for the sake of undoing.

Function: primitive-undo count list

This is the basic function for undoing elements of an undo list. It undoes the first count elements of list, returning the rest of list.

primitive-undo adds elements to the buffer’s undo list when it changes the buffer. Undo commands avoid confusion by saving the undo list value at the beginning of a sequence of undo operations. Then the undo operations use and update the saved value. The new elements added by undoing are not part of this saved value, so they don’t interfere with continuing to undo.

This function does not bind undo-in-progress.

Macro: with-undo-amalgamate body…

This macro removes all the undo boundaries inserted during the execution of body so that it can be undone as a single step.

Some commands leave the region active after execution in such a way that it interferes with selective undo of that command. To make undo ignore the active region when invoked immediately after such a command, set the property undo-inhibit-region of the command’s function symbol to a non-nil value. See Standard Symbol Properties.


33.10 Maintaining Undo Lists

This section describes how to enable and disable undo information for a given buffer. It also explains how the undo list is truncated automatically so it doesn’t get too big.

Recording of undo information in a newly created buffer is normally enabled to start with; but if the buffer name starts with a space, the undo recording is initially disabled. You can explicitly enable or disable undo recording with the following two functions, or by setting buffer-undo-list yourself.

Command: buffer-enable-undo &optional buffer-or-name

This command enables recording undo information for buffer buffer-or-name, so that subsequent changes can be undone. If no argument is supplied, then the current buffer is used. This function does nothing if undo recording is already enabled in the buffer. It returns nil.

In an interactive call, buffer-or-name is the current buffer. You cannot specify any other buffer.

Command: buffer-disable-undo &optional buffer-or-name

This function discards the undo list of buffer-or-name, and disables further recording of undo information. As a result, it is no longer possible to undo either previous changes or any subsequent changes. If the undo list of buffer-or-name is already disabled, this function has no effect.

In an interactive call, BUFFER-OR-NAME is the current buffer. You cannot specify any other buffer. This function returns nil.

As editing continues, undo lists get longer and longer. To prevent them from using up all available memory space, garbage collection trims them back to size limits you can set. (For this purpose, the size of an undo list measures the cons cells that make up the list, plus the strings of deleted text.) Three variables control the range of acceptable sizes: undo-limit, undo-strong-limit and undo-outer-limit. In these variables, size is counted as the number of bytes occupied, which includes both saved text and other data.

User Option: undo-limit

This is the soft limit for the acceptable size of an undo list. The change group at which this size is exceeded is the last one kept.

User Option: undo-strong-limit

This is the upper limit for the acceptable size of an undo list. The change group at which this size is exceeded is discarded itself (along with all older change groups). There is one exception: the very latest change group is only discarded if it exceeds undo-outer-limit.

User Option: undo-outer-limit

If at garbage collection time the undo info for the current command exceeds this limit, Emacs discards the info and displays a warning. This is a last ditch limit to prevent memory overflow.

User Option: undo-ask-before-discard

If this variable is non-nil, when the undo info exceeds undo-outer-limit, Emacs asks in the echo area whether to discard the info. The default value is nil, which means to discard it automatically.

This option is mainly intended for debugging. Garbage collection is inhibited while the question is asked, which means that Emacs might leak memory if the user waits too long before answering the question.


33.11 Filling

Filling means adjusting the lengths of lines (by moving the line breaks) so that they are nearly (but no greater than) a specified maximum width. Additionally, lines can be justified, which means inserting spaces to make the left and/or right margins line up precisely. The width is controlled by the variable fill-column. For ease of reading, lines should be no longer than 70 or so columns.

You can use Auto Fill mode (see Auto Filling) to fill text automatically as you insert it, but changes to existing text may leave it improperly filled. Then you must fill the text explicitly.

Most of the commands in this section return values that are not meaningful. All the functions that do filling take note of the current left margin, current right margin, and current justification style (see Margins for Filling). If the current justification style is none, the filling functions don’t actually do anything.

Several of the filling functions have an argument justify. If it is non-nil, that requests some kind of justification. It can be left, right, full, or center, to request a specific style of justification. If it is t, that means to use the current justification style for this part of the text (see current-justification, below). Any other value is treated as full.

When you call the filling functions interactively, using a prefix argument implies the value full for justify.

Command: fill-paragraph &optional justify region

This command fills the paragraph at or after point. If justify is non-nil, each line is justified as well. It uses the ordinary paragraph motion commands to find paragraph boundaries. See Paragraphs in The GNU Emacs Manual.

When region is non-nil, then if Transient Mark mode is enabled and the mark is active, this command calls fill-region to fill all the paragraphs in the region, instead of filling only the current paragraph. When this command is called interactively, region is t.

Command: fill-region start end &optional justify nosqueeze to-eop

This command fills each of the paragraphs in the region from start to end. It justifies as well if justify is non-nil.

If nosqueeze is non-nil, that means to leave whitespace other than line breaks untouched. If to-eop is non-nil, that means to keep filling to the end of the paragraph—or the next hard newline, if use-hard-newlines is enabled (see below).

The variable paragraph-separate controls how to distinguish paragraphs. See Standard Regular Expressions Used in Editing.

Function: pixel-fill-region start end pixel-width

Most Emacs buffers use monospaced text, so all the filling functions (like fill-region) work based on the number of characters and char-width. However, Emacs can render other types of things, like text that contains images and using proportional fonts, and the pixel-fill-region exists to handle that. It fills the region of text between start and end at pixel granularity, so text using variable-pitch fonts or several different fonts looks filled regardless of different character sizes. The argument pixel-width specifies the maximum pixel width a line is allowed to have after filling; it is the pixel-resolution equivalent of the fill-column in fill-region. For instance, this Lisp snippet will insert text using a proportional font, and then fill this to be no wider than 300 pixels:

(insert (propertize
	 "This is a sentence that's ends here."
	 'face 'variable-pitch))
(pixel-fill-region (point) (point-max) 300)

If start isn’t at the start of a line, the horizontal position of start, converted to pixel units, will be used as the indentation prefix on subsequent lines.

The pixel-fill-width helper function can be used to compute the pixel width to use. If given no arguments, it’ll return a value slightly less than the width of the current window. The first optional value, columns, specifies the number of columns using the standard, monospaced fonts, e.g. fill-column. The second optional value is the window to use. You’d typically use it like this:

(pixel-fill-region
  start end (pixel-fill-width fill-column))
Command: fill-individual-paragraphs start end &optional justify citation-regexp

This command fills each paragraph in the region according to its individual fill prefix. Thus, if the lines of a paragraph were indented with spaces, the filled paragraph will remain indented in the same fashion.

The first two arguments, start and end, are the beginning and end of the region to be filled. The third and fourth arguments, justify and citation-regexp, are optional. If justify is non-nil, the paragraphs are justified as well as filled. If citation-regexp is non-nil, it means the function is operating on a mail message and therefore should not fill the header lines. If citation-regexp is a string, it is used as a regular expression; if it matches the beginning of a line, that line is treated as a citation marker.

Ordinarily, fill-individual-paragraphs regards each change in indentation as starting a new paragraph. If fill-individual-varying-indent is non-nil, then only separator lines separate paragraphs. That mode can handle indented paragraphs with additional indentation on the first line.

User Option: fill-individual-varying-indent

This variable alters the action of fill-individual-paragraphs as described above.

Command: fill-region-as-paragraph start end &optional justify nosqueeze squeeze-after

This command considers a region of text as a single paragraph and fills it. If the region was made up of many paragraphs, the blank lines between paragraphs are removed. This function justifies as well as filling when justify is non-nil.

If nosqueeze is non-nil, that means to leave whitespace other than line breaks untouched. If squeeze-after is non-nil, it specifies a position in the region, and means that whitespace other than line breaks should be left untouched before that position.

In Adaptive Fill mode, this command calls fill-context-prefix to choose a fill prefix by default. See Adaptive Fill Mode.

Command: justify-current-line &optional how eop nosqueeze

This command inserts spaces between the words of the current line so that the line ends exactly at fill-column. It returns nil.

The argument how, if non-nil specifies explicitly the style of justification. It can be left, right, full, center, or none. If it is t, that means to follow specified justification style (see current-justification, below). nil means to do full justification.

If eop is non-nil, that means do only left-justification if current-justification specifies full justification. This is used for the last line of a paragraph; even if the paragraph as a whole is fully justified, the last line should not be.

If nosqueeze is non-nil, that means do not change interior whitespace.

User Option: default-justification

This variable’s value specifies the style of justification to use for text that doesn’t specify a style with a text property. The possible values are left, right, full, center, or none. The default value is left.

Function: current-justification

This function returns the proper justification style to use for filling the text around point.

This returns the value of the justification text property at point, or the variable default-justification if there is no such text property. However, it returns nil rather than none to mean “don’t justify”.

User Option: sentence-end-double-space

If this variable is non-nil, a period followed by just one space does not count as the end of a sentence, and the filling functions avoid breaking the line at such a place.

User Option: sentence-end-without-period

If this variable is non-nil, a sentence can end without a period. This is used for languages like Thai, where sentences end with a double space but without a period.

User Option: sentence-end-without-space

If this variable is non-nil, it should be a string of characters that can end a sentence without following spaces.

User Option: fill-separate-heterogeneous-words-with-space

If this variable is non-nil, two words of different kind (e.g., English and CJK) will be separated with a space when concatenating one that is in the end of a line and the other that is in the beginning of the next line for filling.

Variable: fill-paragraph-function

This variable provides a way to override the filling of paragraphs. If its value is non-nil, fill-paragraph calls this function to do the work. If the function returns a non-nil value, fill-paragraph assumes the job is done, and immediately returns that value.

The usual use of this feature is to fill comments in programming language modes. If the function needs to fill a paragraph in the usual way, it can do so as follows:

(let ((fill-paragraph-function nil))
  (fill-paragraph arg))
Variable: fill-forward-paragraph-function

This variable provides a way to override how the filling functions, such as fill-region and fill-paragraph, move forward to the next paragraph. Its value should be a function, which is called with a single argument n, the number of paragraphs to move, and should return the difference between n and the number of paragraphs actually moved. The default value of this variable is forward-paragraph. See Paragraphs in The GNU Emacs Manual.

Variable: use-hard-newlines

If this variable is non-nil, the filling functions do not delete newlines that have the hard text property. These hard newlines act as paragraph separators. See Hard and Soft Newlines in The GNU Emacs Manual.


33.12 Margins for Filling

User Option: fill-prefix

This buffer-local variable, if non-nil, specifies a string of text that appears at the beginning of normal text lines and should be disregarded when filling them. Any line that fails to start with the fill prefix is considered the start of a paragraph; so is any line that starts with the fill prefix followed by additional whitespace. Lines that start with the fill prefix but no additional whitespace are ordinary text lines that can be filled together. The resulting filled lines also start with the fill prefix.

The fill prefix follows the left margin whitespace, if any.

User Option: fill-column

This buffer-local variable specifies the maximum width of filled lines. Its value should be an integer, which is a number of columns. All the filling, justification, and centering commands are affected by this variable, including Auto Fill mode (see Auto Filling).

As a practical matter, if you are writing text for other people to read, you should set fill-column to no more than 70. Otherwise the line will be too long for people to read comfortably, and this can make the text seem clumsy.

The default value for fill-column is 70. To disable Auto Fill mode in a specific mode, you could say something like:

(add-hook 'foo-mode-hook (lambda () (auto-fill-mode -1))
Command: set-left-margin from to margin

This sets the left-margin property on the text from from to to to the value margin. If Auto Fill mode is enabled, this command also refills the region to fit the new margin.

Command: set-right-margin from to margin

This sets the right-margin property on the text from from to to to the value margin. If Auto Fill mode is enabled, this command also refills the region to fit the new margin.

Function: current-left-margin

This function returns the proper left margin value to use for filling the text around point. The value is the sum of the left-margin property of the character at the start of the current line (or zero if none), and the value of the variable left-margin.

Function: current-fill-column

This function returns the proper fill column value to use for filling the text around point. The value is the value of the fill-column variable, minus the value of the right-margin property of the character after point.

Command: move-to-left-margin &optional n force

This function moves point to the left margin of the current line. The column moved to is determined by calling the function current-left-margin. If the argument n is non-nil, move-to-left-margin moves forward n−1 lines first.

If force is non-nil, that says to fix the line’s indentation if that doesn’t match the left margin value.

Function: delete-to-left-margin &optional from to

This function removes left margin indentation from the text between from and to. The amount of indentation to delete is determined by calling current-left-margin. In no case does this function delete non-whitespace. If from and to are omitted, they default to the whole buffer.

Function: indent-to-left-margin

This function adjusts the indentation at the beginning of the current line to the value specified by the variable left-margin. (That may involve either inserting or deleting whitespace.) This function is value of indent-line-function in Paragraph-Indent Text mode.

User Option: left-margin

This variable specifies the base left margin column. In Fundamental mode, RET indents to this column. This variable automatically becomes buffer-local when set in any fashion.

User Option: fill-nobreak-predicate

This variable gives major modes a way to specify not to break a line at certain places. Its value should be a list of functions. Whenever filling considers breaking the line at a certain place in the buffer, it calls each of these functions with no arguments and with point located at that place. If any of the functions returns non-nil, then the line won’t be broken there.


33.13 Adaptive Fill Mode

When Adaptive Fill Mode is enabled, Emacs determines the fill prefix automatically from the text in each paragraph being filled rather than using a predetermined value. During filling, this fill prefix gets inserted at the start of the second and subsequent lines of the paragraph as described in Filling, and in Auto Filling.

User Option: adaptive-fill-mode

Adaptive Fill mode is enabled when this variable is non-nil. It is t by default.

Function: fill-context-prefix from to

This function implements the heart of Adaptive Fill mode; it chooses a fill prefix based on the text between from and to, typically the start and end of a paragraph. It does this by looking at the first two lines of the paragraph, based on the variables described below.

Usually, this function returns the fill prefix, a string. However, before doing this, the function makes a final check (not specially mentioned in the following) that a line starting with this prefix wouldn’t look like the start of a paragraph. Should this happen, the function signals the anomaly by returning nil instead.

In detail, fill-context-prefix does this:

  1. It takes a candidate for the fill prefix from the first line—it tries first the function in adaptive-fill-function (if any), then the regular expression adaptive-fill-regexp (see below). The first non-nil result of these, or the empty string if they’re both nil, becomes the first line’s candidate.
  2. If the paragraph has as yet only one line, the function tests the validity of the prefix candidate just found. The function then returns the candidate if it’s valid, or a string of spaces otherwise. (see the description of adaptive-fill-first-line-regexp below).
  3. When the paragraph already has two lines, the function next looks for a prefix candidate on the second line, in just the same way it did for the first line. If it doesn’t find one, it returns nil.
  4. The function now compares the two candidate prefixes heuristically: if the non-whitespace characters in the line 2 candidate occur in the same order in the line 1 candidate, the function returns the line 2 candidate. Otherwise, it returns the largest initial substring which is common to both candidates (which might be the empty string).
User Option: adaptive-fill-regexp

Adaptive Fill mode matches this regular expression against the text starting after the left margin whitespace (if any) on a line; the characters it matches are that line’s candidate for the fill prefix.

The default value matches whitespace with certain punctuation characters intermingled.

User Option: adaptive-fill-first-line-regexp

Used only in one-line paragraphs, this regular expression acts as an additional check of the validity of the one available candidate fill prefix: the candidate must match this regular expression, or match comment-start-skip. If it doesn’t, fill-context-prefix replaces the candidate with a string of spaces of the same width as it.

The default value of this variable is "\\`[ \t]*\\'", which matches only a string of whitespace. The effect of this default is to force the fill prefixes found in one-line paragraphs always to be pure whitespace.

User Option: adaptive-fill-function

You can specify more complex ways of choosing a fill prefix automatically by setting this variable to a function. The function is called with point after the left margin (if any) of a line, and it must preserve point. It should return either that line’s fill prefix or nil, meaning it has failed to determine a prefix.


33.14 Auto Filling

Auto Fill mode is a minor mode that fills lines automatically as text is inserted. See Auto Fill in The GNU Emacs Manual. This section describes some variables used by Auto Fill mode. For a description of functions that you can call explicitly to fill and justify existing text, see Filling.

Auto Fill mode also enables the functions that change the margins and justification style to refill portions of the text. See Margins for Filling.

Variable: auto-fill-function

The value of this buffer-local variable should be a function (of no arguments) to be called after self-inserting a character from the table auto-fill-chars, see below. It may be nil, in which case nothing special is done in that case.

The value of auto-fill-function is do-auto-fill when Auto Fill mode is enabled. That is a function whose sole purpose is to implement the usual strategy for breaking a line.

Variable: normal-auto-fill-function

This variable specifies the function to use for auto-fill-function, if and when Auto Fill is turned on. Major modes can set buffer-local values for this variable to alter how Auto Fill works.

Variable: auto-fill-chars

A char table of characters which invoke auto-fill-function when self-inserted—space and newline in most language environments. They have an entry t in the table.

User Option: comment-auto-fill-only-comments

This variable, if non-nil, means to fill lines automatically within comments only. More precisely, this means that if a comment syntax was defined for the current buffer, then self-inserting a character outside of a comment will not call auto-fill-function.


33.15 Sorting Text

The sorting functions described in this section all rearrange text in a buffer. This is in contrast to the function sort, which rearranges the order of the elements of a list (see Functions that Rearrange Lists). The values returned by these functions are not meaningful.

Function: sort-subr reverse nextrecfun endrecfun &optional startkeyfun endkeyfun predicate

This function is the general text-sorting routine that subdivides a buffer into records and then sorts them. Most of the commands in this section use this function.

To understand how sort-subr works, consider the whole accessible portion of the buffer as being divided into disjoint pieces called sort records. The records may or may not be contiguous, but they must not overlap. A portion of each sort record (perhaps all of it) is designated as the sort key. Sorting rearranges the records in order by their sort keys.

Usually, the records are rearranged in order of ascending sort key. If the first argument to the sort-subr function, reverse, is non-nil, the sort records are rearranged in order of descending sort key.

The next four arguments to sort-subr are functions that are called to move point across a sort record. They are called many times from within sort-subr.

  1. nextrecfun is called with point at the end of a record. This function moves point to the start of the next record. The first record is assumed to start at the position of point when sort-subr is called. Therefore, you should usually move point to the beginning of the buffer before calling sort-subr.

    This function can indicate there are no more sort records by leaving point at the end of the buffer.

  2. endrecfun is called with point within a record. It moves point to the end of the record.
  3. startkeyfun is called to move point from the start of a record to the start of the sort key. This argument is optional; if it is omitted, the whole record is the sort key. If supplied, the function should either return a non-nil value to be used as the sort key, or return nil to indicate that the sort key is in the buffer starting at point. In the latter case, endkeyfun is called to find the end of the sort key.
  4. endkeyfun is called to move point from the start of the sort key to the end of the sort key. This argument is optional. If startkeyfun returns nil and this argument is omitted (or nil), then the sort key extends to the end of the record. There is no need for endkeyfun if startkeyfun returns a non-nil value.

The argument predicate is the function to use to compare keys. It is called with two arguments, the keys to compare, and should return non-nil if the first key should come before the second in the sorting order. What exactly are the key arguments depends on what startkeyfun and endkeyfun return. If predicate is omitted or nil, it defaults to < if the keys are numbers, to compare-buffer-substrings if the keys are cons cells (whose car and cdr are start and end buffer positions of the key), and to string< otherwise (with keys assumed to be strings).

As an example of sort-subr, here is the complete function definition for sort-lines:

;; Note that the first two lines of doc string
;; are effectively one line when viewed by a user.
(defun sort-lines (reverse beg end)
  "Sort lines in region alphabetically;\
 argument means descending order.
Called from a program, there are three arguments:
REVERSE (non-nil means reverse order),\
 BEG and END (region to sort).
The variable `sort-fold-case' determines\
 whether alphabetic case affects
the sort order."
  (interactive "P\nr")
  (save-excursion
    (save-restriction
      (narrow-to-region beg end)
      (goto-char (point-min))
      (let ((inhibit-field-text-motion t))
        (sort-subr reverse 'forward-line 'end-of-line)))))

Here forward-line moves point to the start of the next record, and end-of-line moves point to the end of record. We do not pass the arguments startkeyfun and endkeyfun, because the entire record is used as the sort key.

The sort-paragraphs function is very much the same, except that its sort-subr call looks like this:

(sort-subr reverse
           (lambda ()
             (while (and (not (eobp))
                         (looking-at paragraph-separate))
               (forward-line 1)))
           'forward-paragraph)

Markers pointing into any sort records are left with no useful position after sort-subr returns.

User Option: sort-fold-case

If this variable is non-nil, sort-subr and the other buffer sorting functions ignore case when comparing strings.

Command: sort-regexp-fields reverse record-regexp key-regexp start end

This command sorts the region between start and end alphabetically as specified by record-regexp and key-regexp. If reverse is a negative integer, then sorting is in reverse order.

Alphabetical sorting means that two sort keys are compared by comparing the first characters of each, the second characters of each, and so on. If a mismatch is found, it means that the sort keys are unequal; the sort key whose character is less at the point of first mismatch is the lesser sort key. The individual characters are compared according to their numerical character codes in the Emacs character set.

The value of the record-regexp argument specifies how to divide the buffer into sort records. At the end of each record, a search is done for this regular expression, and the text that matches it is taken as the next record. For example, the regular expression ‘^.+$’, which matches lines with at least one character besides a newline, would make each such line into a sort record. See Regular Expressions, for a description of the syntax and meaning of regular expressions.

The value of the key-regexp argument specifies what part of each record is the sort key. The key-regexp could match the whole record, or only a part. In the latter case, the rest of the record has no effect on the sorted order of records, but it is carried along when the record moves to its new position.

The key-regexp argument can refer to the text matched by a subexpression of record-regexp, or it can be a regular expression on its own.

If key-regexp is:

\digit

then the text matched by the digitth ‘\(...\)’ parenthesis grouping in record-regexp is the sort key.

\&

then the whole record is the sort key.

a regular expression

then sort-regexp-fields searches for a match for the regular expression within the record. If such a match is found, it is the sort key. If there is no match for key-regexp within a record then that record is ignored, which means its position in the buffer is not changed. (The other records may move around it.)

For example, if you plan to sort all the lines in the region by the first word on each line starting with the letter ‘f’, you should set record-regexp to ‘^.*$’ and set key-regexp to ‘\<f\w*\>’. The resulting expression looks like this:

(sort-regexp-fields nil "^.*$" "\\<f\\w*\\>"
                    (region-beginning)
                    (region-end))

If you call sort-regexp-fields interactively, it prompts for record-regexp and key-regexp in the minibuffer.

Command: sort-lines reverse start end

This command alphabetically sorts lines in the region between start and end. If reverse is non-nil, the sort is in reverse order.

Command: sort-paragraphs reverse start end

This command alphabetically sorts paragraphs in the region between start and end. If reverse is non-nil, the sort is in reverse order.

Command: sort-pages reverse start end

This command alphabetically sorts pages in the region between start and end. If reverse is non-nil, the sort is in reverse order.

Command: sort-fields field start end

This command sorts lines in the region between start and end, comparing them alphabetically by the fieldth field of each line. Fields are separated by whitespace and numbered starting from 1. If field is negative, sorting is by the −fieldth field from the end of the line. This command is useful for sorting tables.

Command: sort-numeric-fields field start end

This command sorts lines in the region between start and end, comparing them numerically by the fieldth field of each line. Fields are separated by whitespace and numbered starting from 1. The specified field must contain a number in each line of the region. Numbers starting with 0 are treated as octal, and numbers starting with ‘0x’ are treated as hexadecimal.

If field is negative, sorting is by the −fieldth field from the end of the line. This command is useful for sorting tables.

User Option: sort-numeric-base

This variable specifies the default radix for sort-numeric-fields to parse numbers.

Command: sort-columns reverse &optional beg end

This command sorts the lines in the region between beg and end, comparing them alphabetically by a certain range of columns. The column positions of beg and end bound the range of columns to sort on.

If reverse is non-nil, the sort is in reverse order.

One unusual thing about this command is that the entire line containing position beg, and the entire line containing position end, are included in the region sorted.

Note that sort-columns rejects text that contains tabs, because tabs could be split across the specified columns. Use M-x untabify to convert tabs to spaces before sorting.

When possible, this command actually works by calling the sort utility program.


33.16 Counting Columns

The column functions convert between a character position (counting characters from the beginning of the buffer) and a column position (counting screen characters from the beginning of a line).

These functions count each character according to the number of columns it occupies on the screen. This means control characters count as occupying 2 or 4 columns, depending upon the value of ctl-arrow, and tabs count as occupying a number of columns that depends on the value of tab-width and on the column where the tab begins. See Usual Display Conventions.

Column number computations ignore the width of the window and the amount of horizontal scrolling. Consequently, a column value can be arbitrarily high. The first (or leftmost) column is numbered 0. They also ignore overlays and text properties, aside from invisibility. Invisible text is considered as having zero width, unless buffer-invisibility-spec specifies that invisible text should be displayed as ellipsis (see Invisible Text).

Function: current-column

This function returns the horizontal position of point, measured in columns, counting from 0 at the left margin. The column position is the sum of the widths of all the displayed representations of the characters between the start of the current line and point.

Command: move-to-column column &optional force

This function moves point to column in the current line. The calculation of column takes into account the widths of the displayed representations of the characters between the start of the line and point.

When called interactively, column is the value of prefix numeric argument. If column is not an integer, an error is signaled.

If it is impossible to move to column column because that is in the middle of a multicolumn character such as a tab, point moves to the end of that character. However, if force is non-nil, and column is in the middle of a tab, then move-to-column either converts the tab into spaces (when indent-tabs-mode is nil), or inserts enough spaces before it (otherwise), so that point can move precisely to column column. Other multicolumn characters can cause anomalies despite force, since there is no way to split them.

The argument force also has an effect if the line isn’t long enough to reach column column; if it is t, that means to add whitespace at the end of the line to reach that column.

The return value is the column number actually moved to.


33.17 Indentation

The indentation functions are used to examine, move to, and change whitespace that is at the beginning of a line. Some of the functions can also change whitespace elsewhere on a line. Columns and indentation count from zero at the left margin.


33.17.1 Indentation Primitives

This section describes the primitive functions used to count and insert indentation. The functions in the following sections use these primitives. See Size of Displayed Text, for related functions.

Function: current-indentation

This function returns the indentation of the current line, which is the horizontal position of the first nonblank character. If the contents are entirely blank, then this is the horizontal position of the end of the line.

This function considers invisible text as having zero width, unless buffer-invisibility-spec specifies that invisible text should be displayed as ellipsis. See Invisible Text.

Command: indent-to column &optional minimum

This function indents from point with tabs and spaces until column is reached. If minimum is specified and non-nil, then at least that many spaces are inserted even if this requires going beyond column. Otherwise the function does nothing if point is already beyond column. The value is the column at which the inserted indentation ends.

The inserted whitespace characters inherit text properties from the surrounding text (usually, from the preceding text only). See Stickiness of Text Properties.

User Option: indent-tabs-mode

If this variable is non-nil, indentation functions can insert tabs as well as spaces. Otherwise, they insert only spaces. Setting this variable automatically makes it buffer-local in the current buffer.


33.17.2 Indentation Controlled by Major Mode

An important function of each major mode is to customize the TAB key to indent properly for the language being edited. This section describes the mechanism of the TAB key and how to control it. The functions in this section return unpredictable values.

Command: indent-for-tab-command &optional rigid

This is the command bound to TAB in most editing modes. Its usual action is to indent the current line, but it can alternatively insert a tab character or indent a region.

Here is what it does:

  • First, it checks whether Transient Mark mode is enabled and the region is active. If so, it calls indent-region to indent all the text in the region (see Indenting an Entire Region).
  • Otherwise, if the indentation function in indent-line-function is indent-to-left-margin (a trivial command that inserts a tab character), or if the variable tab-always-indent specifies that a tab character ought to be inserted (see below), then it inserts a tab character.
  • Otherwise, it indents the current line; this is done by calling the function in indent-line-function. If the line is already indented, and the value of tab-always-indent is complete (see below), it tries completing the text at point.

If rigid is non-nil (interactively, with a prefix argument), then after this command indents a line or inserts a tab, it also rigidly indents the entire balanced expression which starts at the beginning of the current line, in order to reflect the new indentation. This argument is ignored if the command indents the region.

Variable: indent-line-function

This variable’s value is the function to be used by indent-for-tab-command, and various other indentation commands, to indent the current line. It is usually assigned by the major mode; for instance, Lisp mode sets it to lisp-indent-line, C mode sets it to c-indent-line, and so on. The default value is indent-relative. See Automatic Indentation of code.

Command: indent-according-to-mode

This command calls the function in indent-line-function to indent the current line in a way appropriate for the current major mode.

Command: newline-and-indent

This function inserts a newline, then indents the new line (the one following the newline just inserted) according to the major mode. It does indentation by calling indent-according-to-mode.

Command: reindent-then-newline-and-indent

This command reindents the current line, inserts a newline at point, and then indents the new line (the one following the newline just inserted). It does indentation on both lines by calling indent-according-to-mode.

User Option: tab-always-indent

This variable can be used to customize the behavior of the TAB (indent-for-tab-command) command. If the value is t (the default), the command normally just indents the current line. If the value is nil, the command indents the current line only if point is at the left margin or in the line’s indentation; otherwise, it inserts a tab character. If the value is complete, the command first tries to indent the current line, and if the line was already indented, it calls completion-at-point to complete the text at point (see Completion in Ordinary Buffers).

User Option: tab-first-completion

If tab-always-indent is complete, whether to expand or indent can be further customized via the tab-first-completion variable. The following values can be used:

eol

Only complete if point is at the end of a line.

word

Complete unless the next character has word syntax.

word-or-paren

Complete unless the next character has word syntax or is a parenthesis.

word-or-paren-or-punct

Complete unless the next character has word syntax, or is a parenthesis, or is punctuation.

In any case, typing TAB a second time always results in completion.

Some major modes need to support embedded regions of text whose syntax belongs to a different major mode. Examples include literate programming source files that combine documentation and snippets of source code, Yacc/Bison programs that include snippets of Python or JS code, etc. To correctly indent the embedded chunks, the primary mode needs to delegate the indentation to another mode’s indentation engine (e.g., call js-indent-line for JS code or python-indent-line for Python), while providing it with some context to guide the indentation. Major modes, for their part, should avoid calling widen in their indentation code and obey prog-first-column.

Variable: prog-indentation-context

This variable, when non-nil, holds the indentation context for the sub-mode’s indentation engine provided by the superior major mode. The value should be a list of the form (first-column . rest. The members of the list have the following meaning:

first-column

The column to be used for top-level constructs. This replaces the default value of the top-level column used by the sub-mode, usually zero.

rest

This value is currently unused.

The following convenience function should be used by major mode’s indentation engine in support of invocations as sub-modes of another major mode.

Function: prog-first-column

Call this function instead of using a literal value (usually, zero) of the column number for indenting top-level program constructs. The function’s value is the column number to use for top-level constructs. When no superior mode is in effect, this function returns zero.


33.17.3 Indenting an Entire Region

This section describes commands that indent all the lines in the region. They return unpredictable values.

Command: indent-region start end &optional to-column

This command indents each nonblank line starting between start (inclusive) and end (exclusive). If to-column is nil, indent-region indents each nonblank line by calling the current mode’s indentation function, the value of indent-line-function.

If to-column is non-nil, it should be an integer specifying the number of columns of indentation; then this function gives each line exactly that much indentation, by either adding or deleting whitespace.

If there is a fill prefix, indent-region indents each line by making it start with the fill prefix.

Variable: indent-region-function

The value of this variable is a function that can be used by indent-region as a short cut. It should take two arguments, the start and end of the region. You should design the function so that it will produce the same results as indenting the lines of the region one by one, but presumably faster.

If the value is nil, there is no short cut, and indent-region actually works line by line.

A short-cut function is useful in modes such as C mode and Lisp mode, where the indent-line-function must scan from the beginning of the function definition: applying it to each line would be quadratic in time. The short cut can update the scan information as it moves through the lines indenting them; this takes linear time. In a mode where indenting a line individually is fast, there is no need for a short cut.

indent-region with a non-nil argument to-column has a different meaning and does not use this variable.

Command: indent-rigidly start end count

This function indents all lines starting between start (inclusive) and end (exclusive) sideways by count columns. This preserves the shape of the affected region, moving it as a rigid unit.

This is useful not only for indenting regions of unindented text, but also for indenting regions of formatted code. For example, if count is 3, this command adds 3 columns of indentation to every line that begins in the specified region.

If called interactively with no prefix argument, this command invokes a transient mode for adjusting indentation rigidly. See Indentation Commands in The GNU Emacs Manual.

Command: indent-code-rigidly start end columns &optional nochange-regexp

This is like indent-rigidly, except that it doesn’t alter lines that start within strings or comments.

In addition, it doesn’t alter a line if nochange-regexp matches at the beginning of the line (if nochange-regexp is non-nil).


33.17.4 Indentation Relative to Previous Lines

This section describes two commands that indent the current line based on the contents of previous lines.

Command: indent-relative &optional first-only unindented-ok

This command inserts whitespace at point, extending to the same column as the next indent point of the previous nonblank line. An indent point is a non-whitespace character following whitespace. The next indent point is the first one at a column greater than the current column of point. For example, if point is underneath and to the left of the first non-blank character of a line of text, it moves to that column by inserting whitespace.

If the previous nonblank line has no next indent point (i.e., none at a great enough column position), indent-relative either does nothing (if unindented-ok is non-nil) or calls tab-to-tab-stop. Thus, if point is underneath and to the right of the last column of a short line of text, this command ordinarily moves point to the next tab stop by inserting whitespace.

If first-only is non-nil, only the first indent point is considered.

The return value of indent-relative is unpredictable.

In the following example, point is at the beginning of the second line:

            This line is indented twelve spaces.
∗The quick brown fox jumped.

Evaluation of the expression (indent-relative nil) produces the following:

            This line is indented twelve spaces.
            ∗The quick brown fox jumped.

In this next example, point is between the ‘m’ and ‘p’ of ‘jumped’:

            This line is indented twelve spaces.
The quick brown fox jum∗ped.

Evaluation of the expression (indent-relative nil) produces the following:

            This line is indented twelve spaces.
The quick brown fox jum  ∗ped.
Command: indent-relative-first-indent-point

This command indents the current line like the previous nonblank line, by calling indent-relative with t as the first-only argument. The return value is unpredictable.

If the previous nonblank line has no indent points beyond the current column, this command does nothing.


33.17.5 Adjustable Tab Stops

This section explains the mechanism for user-specified tab stops and the mechanisms that use and set them. The name “tab stops” is used because the feature is similar to that of the tab stops on a typewriter. The feature works by inserting an appropriate number of spaces and tab characters to reach the next tab stop column; it does not affect the display of tab characters in the buffer (see Usual Display Conventions). Note that the TAB character as input uses this tab stop feature only in a few major modes, such as Text mode. See Tab Stops in The GNU Emacs Manual.

Command: tab-to-tab-stop

This command inserts spaces or tabs before point, up to the next tab stop column defined by tab-stop-list.

User Option: tab-stop-list

This variable defines the tab stop columns used by tab-to-tab-stop. It should be either nil, or a list of increasing integers, which need not be evenly spaced. The list is implicitly extended to infinity through repetition of the interval between the last and penultimate elements (or tab-width if the list has fewer than two elements). A value of nil means a tab stop every tab-width columns.

Use M-x edit-tab-stops to edit the location of tab stops interactively.


33.17.6 Indentation-Based Motion Commands

These commands, primarily for interactive use, act based on the indentation in the text.

Command: back-to-indentation

This command moves point to the first non-whitespace character in the current line (which is the line in which point is located).

Command: backward-to-indentation &optional arg

This command moves point backward arg lines and then to the first nonblank character on that line. If arg is omitted or nil, it defaults to 1.

Command: forward-to-indentation &optional arg

This command moves point forward arg lines and then to the first nonblank character on that line. If arg is omitted or nil, it defaults to 1.


33.18 Case Changes

The case change commands described here work on text in the current buffer. See Case Conversion in Lisp, for case conversion functions that work on strings and characters. See The Case Table, for how to customize which characters are upper or lower case and how to convert them.

Command: capitalize-region start end

This function capitalizes all words in the region defined by start and end. To capitalize means to convert each word’s first character to upper case and convert the rest of each word to lower case. The function returns nil.

If one end of the region is in the middle of a word, the part of the word within the region is treated as an entire word.

When capitalize-region is called interactively, start and end are point and the mark, with the smallest first.

---------- Buffer: foo ----------
This is the contents of the 5th foo.
---------- Buffer: foo ----------

(capitalize-region 1 37)
⇒ nil

---------- Buffer: foo ----------
This Is The Contents Of The 5th Foo.
---------- Buffer: foo ----------
Command: downcase-region start end

This function converts all of the letters in the region defined by start and end to lower case. The function returns nil.

When downcase-region is called interactively, start and end are point and the mark, with the smallest first.

Command: upcase-region start end

This function converts all of the letters in the region defined by start and end to upper case. The function returns nil.

When upcase-region is called interactively, start and end are point and the mark, with the smallest first.

Command: capitalize-word count

This function capitalizes count words after point, moving point over as it does. To capitalize means to convert each word’s first character to upper case and convert the rest of each word to lower case. If count is negative, the function capitalizes the −count previous words but does not move point. The value is nil.

If point is in the middle of a word, the part of the word before point is ignored when moving forward. The rest is treated as an entire word.

When capitalize-word is called interactively, count is set to the numeric prefix argument.

Command: downcase-word count

This function converts the count words after point to all lower case, moving point over as it does. If count is negative, it converts the −count previous words but does not move point. The value is nil.

When downcase-word is called interactively, count is set to the numeric prefix argument.

Command: upcase-word count

This function converts the count words after point to all upper case, moving point over as it does. If count is negative, it converts the −count previous words but does not move point. The value is nil.

When upcase-word is called interactively, count is set to the numeric prefix argument.


33.19 Text Properties

Each character position in a buffer or a string can have a text property list, much like the property list of a symbol (see Property Lists). The properties belong to a particular character at a particular place, such as, the letter ‘T’ at the beginning of this sentence or the first ‘o’ in ‘foo’—if the same character occurs in two different places, the two occurrences in general have different properties.

Each property has a name and a value. Both of these can be any Lisp object, but the name is normally a symbol. Typically each property name symbol is used for a particular purpose; for instance, the text property face specifies the faces for displaying the character (see Properties with Special Meanings). The usual way to access the property list is to specify a name and ask what value corresponds to it.

If a character has a category property, we call it the property category of the character. It should be a symbol. The properties of the symbol serve as defaults for the properties of the character.

Copying text between strings and buffers preserves the properties along with the characters; this includes such diverse functions as substring, insert, and buffer-substring. Killing and then yanking text (see The Kill Ring) also preserves the properties, except that some properties are handled specially and might be removed when text is yanked; see Yanking.


33.19.1 Examining Text Properties

The simplest way to examine text properties is to ask for the value of a particular property of a particular character. For that, use get-text-property. Use text-properties-at to get the entire property list of a character. See Text Property Search Functions, for functions to examine the properties of a number of characters at once.

These functions handle both strings and buffers. Keep in mind that positions in a string start from 0, whereas positions in a buffer start from 1. Passing a buffer other than the current buffer may be slow.

Function: get-text-property pos prop &optional object

This function returns the value of the prop property of the character after position pos in object (a buffer or string). The argument object is optional and defaults to the current buffer.

If position is at the end of object, the value is nil, but note that buffer narrowing does not affect the value. That is, if object is a buffer or nil, and the buffer is narrowed and position is at the end of the narrowed buffer, the result may be non-nil.

If there is no prop property strictly speaking, but the character has a property category that is a symbol, then get-text-property returns the prop property of that symbol.

Function: get-char-property position prop &optional object

This function is like get-text-property, except that it checks overlays first and then text properties. See Overlays.

The argument object may be a string, a buffer, or a window. If it is a window, then the buffer displayed in that window is used for text properties and overlays, but only the overlays active for that window are considered. If object is a buffer, then overlays in that buffer are considered first, in order of decreasing priority, followed by the text properties. If object is a string, only text properties are considered, since strings never have overlays.

Function: get-pos-property position prop &optional object

This function is like get-char-property, except that it pays attention to properties’ stickiness and overlays’ advancement settings instead of the property of the character at (i.e., right after) position.

Function: get-char-property-and-overlay position prop &optional object

This is like get-char-property, but gives extra information about the overlay that the property value comes from.

Its value is a cons cell whose CAR is the property value, the same value get-char-property would return with the same arguments. Its CDR is the overlay in which the property was found, or nil, if it was found as a text property or not found at all.

If position is at the end of object, both the CAR and the CDR of the value are nil.

Variable: char-property-alias-alist

This variable holds an alist which maps property names to a list of alternative property names. If a character does not specify a direct value for a property, the alternative property names are consulted in order; the first non-nil value is used. This variable takes precedence over default-text-properties, and category properties take precedence over this variable.

Function: text-properties-at position &optional object

This function returns the entire property list of the character at position in the string or buffer object. If object is nil, it defaults to the current buffer.

If position is at the end of object, the value is nil, but note that buffer narrowing does not affect the value. That is, if object is a buffer or nil, and the buffer is narrowed and position is at the end of the narrowed buffer, the result may be non-nil.

Variable: default-text-properties

This variable holds a property list giving default values for text properties. Whenever a character does not specify a value for a property, neither directly, through a category symbol, or through char-property-alias-alist, the value stored in this list is used instead. Here is an example:

(setq default-text-properties '(foo 69)
      char-property-alias-alist nil)
;; Make sure character 1 has no properties of its own.
(set-text-properties 1 2 nil)
;; What we get, when we ask, is the default value.
(get-text-property 1 'foo)
     ⇒ 69
Function: object-intervals OBJECT

This function returns a copy of the intervals (i.e., text properties) in object as a list of intervals. object must be a string or a buffer. Altering the structure of this list does not change the intervals in the object.

(object-intervals (propertize "foo" 'face 'bold))
     ⇒ ((0 3 (face bold)))

Each element in the returned list represents one interval. Each interval has three parts: The first is the start, the second is the end, and the third part is the text property itself.


33.19.2 Changing Text Properties

The primitives for changing properties apply to a specified range of text in a buffer or string. The function set-text-properties (see end of section) sets the entire property list of the text in that range; more often, it is useful to add, change, or delete just certain properties specified by name.

Since text properties are considered part of the contents of the buffer (or string), and can affect how a buffer looks on the screen, any change in buffer text properties marks the buffer as modified. Buffer text property changes are undoable also (see Undo). Positions in a string start from 0, whereas positions in a buffer start from 1.

Function: put-text-property start end prop value &optional object

This function sets the prop property to value for the text between start and end in the string or buffer object. If object is nil, it defaults to the current buffer.

Function: add-text-properties start end props &optional object

This function adds or overrides text properties for the text between start and end in the string or buffer object. If object is nil, it defaults to the current buffer.

The argument props specifies which properties to add. It should have the form of a property list (see Property Lists): a list whose elements include the property names followed alternately by the corresponding values.

The return value is t if the function actually changed some property’s value; nil otherwise (if props is nil or its values agree with those in the text).

For example, here is how to set the comment and face properties of a range of text:

(add-text-properties start end
                     '(comment t face highlight))
Function: remove-text-properties start end props &optional object

This function deletes specified text properties from the text between start and end in the string or buffer object. If object is nil, it defaults to the current buffer.

The argument props specifies which properties to delete. It should have the form of a property list (see Property Lists): a list whose elements are property names alternating with corresponding values. But only the names matter—the values that accompany them are ignored. For example, here’s how to remove the face property.

(remove-text-properties start end '(face nil))

The return value is t if the function actually changed some property’s value; nil otherwise (if props is nil or if no character in the specified text had any of those properties).

To remove all text properties from certain text, use set-text-properties and specify nil for the new property list.

Function: remove-list-of-text-properties start end list-of-properties &optional object

Like remove-text-properties except that list-of-properties is a list of property names only, not an alternating list of property names and values.

Function: set-text-properties start end props &optional object

This function completely replaces the text property list for the text between start and end in the string or buffer object. If object is nil, it defaults to the current buffer.

The argument props is the new property list. It should be a list whose elements are property names alternating with corresponding values.

After set-text-properties returns, all the characters in the specified range have identical properties.

If props is nil, the effect is to get rid of all properties from the specified range of text. Here’s an example:

(set-text-properties start end nil)

Do not rely on the return value of this function.

Function: add-face-text-property start end face &optional appendp object

This function acts on the text between start and end, adding the face face to the face text property. face should be a valid value for the face property (see Properties with Special Meanings), such as a face name or an anonymous face (see Faces).

If any text in the region already has a non-nil face property, those face(s) are retained. This function sets the face property to a list of faces, with face as the first element (by default) and the pre-existing faces as the remaining elements. If the optional argument appendp is non-nil, face is appended to the end of the list instead. Note that in a face list, the first occurring value for each attribute takes precedence.

For example, the following code would assign an italicized green face to the text between start and end:

(add-face-text-property start end 'italic)
(add-face-text-property start end '(:foreground "red"))
(add-face-text-property start end '(:foreground "green"))

The optional argument object, if non-nil, specifies a buffer or string to act on, rather than the current buffer. If object is a string, then start and end are zero-based indices into the string.

The easiest way to make a string with text properties is with propertize:

Function: propertize string &rest properties

This function returns a copy of string with the text properties properties added. These properties apply to all the characters in the string that is returned. Here is an example that constructs a string with a face property and a mouse-face property:

(propertize "foo" 'face 'italic
            'mouse-face 'bold-italic)
     ⇒ #("foo" 0 3 (mouse-face bold-italic face italic))

To put different properties on various parts of a string, you can construct each part with propertize and then combine them with concat:

(concat
 (propertize "foo" 'face 'italic
             'mouse-face 'bold-italic)
 " and "
 (propertize "bar" 'face 'italic
             'mouse-face 'bold-italic))
     ⇒ #("foo and bar"
                 0 3 (face italic mouse-face bold-italic)
                 3 8 nil
                 8 11 (face italic mouse-face bold-italic))

See Examining Buffer Contents, for the function buffer-substring-no-properties, which copies text from the buffer but does not copy its properties.

If you wish to add text properties to a buffer or remove them without marking the buffer as modified, you can wrap the calls above in the with-silent-modifications macro. See Buffer Modification.


33.19.4 Properties with Special Meanings

Here is a table of text property names that have special built-in meanings. The following sections list a few additional special property names that control filling and property inheritance. All other names have no standard meaning, and you can use them as you like.

Note: the properties composition, display, invisible and intangible can also cause point to move to an acceptable place, after each Emacs command. See Adjusting Point After Commands.

category

If a character has a category property, we call it the property category of the character. It should be a symbol. The properties of this symbol serve as defaults for the properties of the character.

face

The face property controls the appearance of the character (see Faces). The value of the property can be the following:

  • A face name (a symbol or string).
  • An anonymous face: a property list of the form (keyword value …), where each keyword is a face attribute name and value is a value for that attribute.
  • A list of faces. Each list element should be either a face name or an anonymous face. This specifies a face which is an aggregate of the attributes of each of the listed faces. Faces occurring earlier in the list have higher priority.
  • A cons cell of the form (foreground-color . color-name) or (background-color . color-name). This specifies the foreground or background color, similar to (:foreground color-name) or (:background color-name). This form is supported for backward compatibility only, and should be avoided.
  • A cons cell of the form (:filtered filter face-spec), that specifies the face given by face-spec, but only if filter matches when the face is used for display. The face-spec can use any of the forms mentioned above. The filter should be of the form (:window param value), which matches for windows whose parameter param is eq to value. If the variable face-filters-always-match is non-nil, all face filters are deemed to have matched.

Font Lock mode (see Font Lock Mode) works in most buffers by dynamically updating the face property of characters based on the context.

The add-face-text-property function provides a convenient way to set this text property. See Changing Text Properties.

font-lock-face

This property specifies a value for the face property that Font Lock mode should apply to the underlying text. It is one of the fontification methods used by Font Lock mode, and is useful for special modes that implement their own highlighting. See Precalculated Fontification. When Font Lock mode is disabled, font-lock-face has no effect.

mouse-face

This property is used instead of face when the mouse pointer hovers over the text which has this property. When this happens, the entire stretch of text that has the same mouse-face property value, not just the character under the mouse, is highlighted.

Emacs ignores all face attributes from the mouse-face property that alter the text size (e.g., :height, :weight, and :slant). Those attributes are always the same as for the unhighlighted text.

cursor-face

This property is similar to mouse-face, but it is used when point (not the mouse) is inside text that has this property. The highlighting happens only if the mode cursor-face-highlight-mode is enabled. When the variable cursor-face-highlight-nonselected-window is non-nil, the text with this face is highlighted even if the window is not selected, similarly to what highlight-nonselected-windows does for the region (see The Mark and the Region in The GNU Emacs Manual).

fontified

This property says whether the text is ready for display. If nil, Emacs’s redisplay routine calls the functions in fontification-functions (see Automatic Face Assignment) to prepare this part of the buffer before it is displayed. It is used internally by the just-in-time font locking code.

display

This property activates various features that change the way text is displayed. For example, it can make text appear taller or shorter, higher or lower, wider or narrow, or replaced with an image. See The display Property.

help-echo

If text has a string as its help-echo property, then when you move the mouse onto that text, Emacs displays that string in the echo area, or in the tooltip window (see Tooltips), after passing it through substitute-command-keys.

If the value of the help-echo property is a function, that function is called with three arguments, window, object and pos and should return a help string or nil for none. The first argument, window is the window in which the help was found. The second, object, is the buffer, overlay or string which had the help-echo property. The pos argument is as follows:

  • If object is a buffer, pos is the position in the buffer.
  • If object is an overlay, that overlay has a help-echo property, and pos is the position in the overlay’s buffer.
  • If object is a string (an overlay string or a string displayed with the display property), pos is the position in that string.

If the value of the help-echo property is neither a function nor a string, it is evaluated to obtain a help string.

You can alter the way help text is displayed by setting the variable show-help-function (see Help display).

This feature is used in the mode line and for other active text.

help-echo-inhibit-substitution

If the first character of a help-echo string has a non-nil help-echo-inhibit-substitution property, then it is displayed as-is by show-help-function, without being passed through substitute-command-keys.

keymap

The keymap property specifies an additional keymap for commands. When this keymap applies, it is used for key lookup before the minor mode keymaps and before the buffer’s local map. See Active Keymaps. If the property value is a symbol, the symbol’s function definition is used as the keymap.

The property’s value for the character before point applies if it is non-nil and rear-sticky, and the property’s value for the character after point applies if it is non-nil and front-sticky. (For mouse clicks, the position of the click is used instead of the position of point.)

local-map

This property works like keymap except that it specifies a keymap to use instead of the buffer’s local map. For most purposes (perhaps all purposes), it is better to use the keymap property.

syntax-table

The syntax-table property overrides what the syntax table says about this particular character. See Syntax Properties.

read-only

If a character has the property read-only, then modifying that character is not allowed. Any command that would do so gets an error, text-read-only. If the property value is a string, that string is used as the error message.

Insertion next to a read-only character is an error if inserting ordinary text there would inherit the read-only property due to stickiness. Thus, you can control permission to insert next to read-only text by controlling the stickiness. See Stickiness of Text Properties.

Since changing properties counts as modifying the buffer, it is not possible to remove a read-only property unless you know the special trick: bind inhibit-read-only to a non-nil value and then remove the property. See Read-Only Buffers.

inhibit-read-only

Characters that have the property inhibit-read-only can be edited even in read-only buffers. See Read-Only Buffers.

invisible

A non-nil invisible property can make a character invisible on the screen. See Invisible Text, for details.

inhibit-isearch

A non-nil inhibit-isearch property will make isearch skip the text.

intangible

If a group of consecutive characters have equal and non-nil intangible properties, then you cannot place point between them. If you try to move point forward into the group, point actually moves to the end of the group. If you try to move point backward into the group, point actually moves to the start of the group.

If consecutive characters have unequal non-nil intangible properties, they belong to separate groups; each group is separately treated as described above.

When the variable inhibit-point-motion-hooks is non-nil (as it is by default), the intangible property is ignored.

Beware: this property operates at a very low level, and affects a lot of code in unexpected ways. So use it with extreme caution. A common misuse is to put an intangible property on invisible text, which is actually unnecessary since the command loop will move point outside of the invisible text at the end of each command anyway. See Adjusting Point After Commands. For these reasons, this property is obsolete; use the cursor-intangible property instead.

cursor-intangible

When the minor mode cursor-intangible-mode is turned on, point is moved away from any position that has a non-nil cursor-intangible property, just before redisplay happens. Note that “stickiness” of the property (see Stickiness of Text Properties) is taken into account when computing allowed cursor positions, so (for instance) to insert a stretch of five ‘x’ characters into which the cursor can’t enter, you should do something like:

(insert
 (propertize "xxxx" 'cursor-intangible t)
 (propertize "x" 'cursor-intangible t 'rear-nonsticky t))

When the variable cursor-sensor-inhibit is non-nil, the cursor-intangible property and the cursor-sensor-functions property (described below) are ignored.

field

Consecutive characters with the same field property constitute a field. Some motion functions including forward-word and beginning-of-line stop moving at a field boundary. See Defining and Using Fields.

cursor

Normally, the cursor is displayed at the beginning or the end of any overlay and text property strings that “hide” (i.e., are displayed instead of) the current buffer position. You can instead tell Emacs to place the cursor on any desired character of these strings by giving that character a non-nil cursor text property. In addition, if the value of the cursor property is an integer, it specifies the number of buffer’s character positions, starting with the position where the overlay or the display property begins, for which the cursor should be displayed on that character. Specifically, if the value of the cursor property of a character is the number n, the cursor will be displayed on this character for any buffer position in the range [ovpos..ovpos+n), where ovpos is the overlay’s starting position given by overlay-start (see Managing Overlays), or the position where the display text property begins in the buffer.

In other words, the string character with the cursor property of any non-nil value is the character where to display the cursor when the overlay or display string make point not visible on display. The value of the property says for which buffer positions to display the cursor there. If the value is an integer n, the cursor is displayed there when point is anywhere between the beginning of the overlay or display property and n positions after that. If the value is anything else and non-nil, the cursor is displayed there only when point is at the buffer position that is the beginning of the display property, or at overlay-start if that position is not visible on display. Note that an integer value of the cursor property could mean that the cursor is displayed on that character even when point is visible on display.

One subtlety of this property is that it doesn’t work to put this property on a newline character that is part of a display or overlay string. That’s because the newline doesn’t have a graphic representation on the screen for Emacs to find when it looks for a character on display with that cursor property.

When the buffer has many overlay strings (e.g., see before-string) that conceal some of the buffer text or display properties that are strings, it is a good idea to use the cursor property on these strings to cue the Emacs display about the places where to put the cursor while traversing these strings. This directly communicates to the display engine where the Lisp program wants to put the cursor, or where the user would expect the cursor, when point is located on some buffer position that is “covered” by the display or overlay string.

pointer

This specifies a specific pointer shape when the mouse pointer is over this text or image. See Pointer Shape, for possible pointer shapes.

line-spacing

A newline can have a line-spacing text or overlay property that controls the height of the display line ending with that newline. The property value overrides the default frame line spacing and the buffer local line-spacing variable. See Line Height.

line-height

A newline can have a line-height text or overlay property that controls the total height of the display line ending in that newline. See Line Height.

wrap-prefix

If text has a wrap-prefix property, the prefix it defines will be added at display time to the beginning of every continuation line due to text wrapping (so if lines are truncated, the wrap-prefix is never used). It may be a string or an image (see Other Display Specifications), or a stretch of whitespace such as specified by the :width or :align-to display properties (see Specified Spaces).

A wrap-prefix may also be specified for an entire buffer using the wrap-prefix buffer-local variable (however, a wrap-prefix text-property takes precedence over the value of the wrap-prefix variable). See Truncation.

line-prefix

If text has a line-prefix property, the prefix it defines will be added at display time to the beginning of every non-continuation line. It may be a string or an image (see Other Display Specifications), or a stretch of whitespace such as specified by the :width or :align-to display properties (see Specified Spaces).

A line-prefix may also be specified for an entire buffer using the line-prefix buffer-local variable (however, a line-prefix text-property takes precedence over the value of the line-prefix variable). See Truncation.

modification-hooks

If a character has the property modification-hooks, then its value should be a list of functions; modifying that character calls all of those functions before the actual modification. Each function receives two arguments: the beginning and end of the part of the buffer being modified. Note that if a particular modification hook function appears on several characters being modified by a single primitive, you can’t predict how many times the function will be called. Furthermore, insertion will not modify any existing character, so this hook will only be run when removing some characters, replacing them with others, or changing their text-properties.

Unlike with other similar hooks, when Emacs calls these functions, inhibit-modification-hooks does not get bound to non-nil. If the functions modify the buffer, you should consider binding this variable to non-nil to prevent any buffer changes running the change hooks. Otherwise, you must be prepared for recursive calls. See Change Hooks.

Overlays also support the modification-hooks property, but the details are somewhat different (see Overlay Properties).

insert-in-front-hooks
insert-behind-hooks

The operation of inserting text in a buffer also calls the functions listed in the insert-in-front-hooks property of the following character and in the insert-behind-hooks property of the preceding character. These functions receive two arguments, the beginning and end of the inserted text. The functions are called after the actual insertion takes place.

When these functions are called, inhibit-modification-hooks is bound to non-nil. If the functions modify the buffer, you might want to bind inhibit-modification-hooks to nil, so as to cause the change hooks to run for these modifications. However, doing this may call your own change hook recursively, so be sure to prepare for that.

See also Change Hooks, for other hooks that are called when you change text in a buffer.

point-entered
point-left

The special properties point-entered and point-left record hook functions that report motion of point. Each time point moves, Emacs compares these two property values:

  • the point-left property of the character after the old location, and
  • the point-entered property of the character after the new location.

If these two values differ, each of them is called (if not nil) with two arguments: the old value of point, and the new one.

The same comparison is made for the characters before the old and new locations. The result may be to execute two point-left functions (which may be the same function) and/or two point-entered functions (which may be the same function). In any case, all the point-left functions are called first, followed by all the point-entered functions.

It is possible to use char-after to examine characters at various buffer positions without moving point to those positions. Only an actual change in the value of point runs these hook functions.

The variable inhibit-point-motion-hooks by default inhibits running the point-left and point-entered hooks, see Inhibit point motion hooks.

These properties are obsolete; please use cursor-sensor-functions instead.

cursor-sensor-functions

This special property records a list of functions that react to cursor motion. Each function in the list is called, just before redisplay, with 3 arguments: the affected window, the previous known position of the cursor, and one of the symbols entered or left, depending on whether the cursor is entering the text that has this property or leaving it. The functions are called only when the minor mode cursor-sensor-mode is turned on.

When the variable cursor-sensor-inhibit is non-nil, the cursor-sensor-functions property is ignored.

composition

This text property is used to display a sequence of characters as a single glyph composed from components. But the value of the property itself is completely internal to Emacs and should not be manipulated directly by, for instance, put-text-property.

minibuffer-message

This text property tells where to display temporary messages in an active minibuffer. Specifically, the first character of the minibuffer text which has this property will have the temporary message displayed before it. The default is to display temporary messages at the end of the minibuffer text. This text property is used by the function that is the default value of set-message-function (see Displaying Messages in the Echo Area).

Variable: inhibit-point-motion-hooks

When this obsolete variable is non-nil, point-left and point-entered hooks are not run, and the intangible property has no effect. Do not set this variable globally; bind it with let. Since the affected properties are obsolete, this variable’s default value is t, to effectively disable them.

Variable: show-help-function

If this variable is non-nil, it specifies a function called to display help strings. These may be help-echo properties, menu help strings (see Simple Menu Items, see Extended Menu Items), or tool bar help strings (see Tool bars). The specified function is called with one argument, the help string to display, which is passed through substitute-command-keys before being given to the function, unless the help string has a non-nil help-echo-inhibit-substitution property on its first character; see Substituting Key Bindings in Documentation. See the code of Tooltip mode (see Tooltips in The GNU Emacs Manual) for an example of a mode that uses show-help-function.

Variable: face-filters-always-match

If this variable is non-nil, face filters that specify attributes applied only when certain conditions are met will be deemed to match always.


33.19.5 Formatted Text Properties

These text properties affect the behavior of the fill commands. They are used for representing formatted text. See Filling, and Margins for Filling.

hard

If a newline character has this property, it is a “hard” newline. The fill commands do not alter hard newlines and do not move words across them. However, this property takes effect only if the use-hard-newlines minor mode is enabled. See Hard and Soft Newlines in The GNU Emacs Manual.

right-margin

This property specifies an extra right margin for filling this part of the text.

left-margin

This property specifies an extra left margin for filling this part of the text.

justification

This property specifies the style of justification for filling this part of the text.


33.19.6 Stickiness of Text Properties

Self-inserting characters, the ones that get inserted into a buffer when the user types them (see User-Level Insertion Commands), normally take on the same properties as the preceding character. This is called inheritance of properties.

By contrast, a Lisp program can do insertion with inheritance or without, depending on the choice of insertion primitive. The ordinary text insertion functions, such as insert, do not inherit any properties. They insert text with precisely the properties of the string being inserted, and no others. This is correct for programs that copy text from one context to another—for example, into or out of the kill ring. To insert with inheritance, use the special primitives described in this section. Self-inserting characters inherit properties because they work using these primitives.

When you do insertion with inheritance, which properties are inherited, and from where, depends on which properties are sticky. Insertion after a character inherits those of its properties that are rear-sticky. Insertion before a character inherits those of its properties that are front-sticky. When both sides offer different sticky values for the same property, the previous character’s value takes precedence.

By default, a text property is rear-sticky but not front-sticky; thus, the default is to inherit all the properties of the preceding character, and nothing from the following character.

You can control the stickiness of various text properties with two specific text properties, front-sticky and rear-nonsticky, and with the variable text-property-default-nonsticky. You can use the variable to specify a different default for a given property. You can use those two text properties to make any specific properties sticky or nonsticky in any particular part of the text.

If a character’s front-sticky property is t, then all its properties are front-sticky. If the front-sticky property is a list, then the sticky properties of the character are those whose names are in the list. For example, if a character has a front-sticky property whose value is (face read-only), then insertion before the character can inherit its face property and its read-only property, but no others.

The rear-nonsticky property works the opposite way. Most properties are rear-sticky by default, so the rear-nonsticky property says which properties are not rear-sticky. If a character’s rear-nonsticky property is t, then none of its properties are rear-sticky. If the rear-nonsticky property is a list, properties are rear-sticky unless their names are in the list.

Variable: text-property-default-nonsticky

This variable holds an alist which defines the default rear-stickiness of various text properties. Each element has the form (property . nonstickiness), and it defines the stickiness of a particular text property, property.

If nonstickiness is non-nil, this means that the property property is rear-nonsticky by default. Since all properties are front-nonsticky by default, this makes property nonsticky in both directions by default.

The text properties front-sticky and rear-nonsticky, when used, take precedence over the default nonstickiness specified in text-property-default-nonsticky.

Here are the functions that insert text with inheritance of properties:

Function: insert-and-inherit &rest strings

Insert the strings strings, just like the function insert, but inherit any sticky properties from the adjoining text.

Function: insert-before-markers-and-inherit &rest strings

Insert the strings strings, just like the function insert-before-markers, but inherit any sticky properties from the adjoining text.

See Inserting Text, for the ordinary insertion functions which do not inherit.


33.19.7 Lazy Computation of Text Properties

Instead of computing text properties for all the text in the buffer, you can arrange to compute the text properties for parts of the text when and if something depends on them.

The primitive that extracts text from the buffer along with its properties is buffer-substring. Before examining the properties, this function runs the abnormal hook buffer-access-fontify-functions.

Variable: buffer-access-fontify-functions

This variable holds a list of functions for computing text properties. Before buffer-substring copies the text and text properties for a portion of the buffer, it calls all the functions in this list. Each of the functions receives two arguments that specify the range of the buffer being accessed. (The buffer itself is always the current buffer.)

The function buffer-substring-no-properties does not call these functions, since it ignores text properties anyway.

In order to prevent the hook functions from being called more than once for the same part of the buffer, you can use the variable buffer-access-fontified-property.

Variable: buffer-access-fontified-property

If this variable’s value is non-nil, it is a symbol which is used as a text property name. A non-nil value for that text property means the other text properties for this character have already been computed.

If all the characters in the range specified for buffer-substring have a non-nil value for this property, buffer-substring does not call the buffer-access-fontify-functions functions. It assumes these characters already have the right text properties, and just copies the properties they already have.

The normal way to use this feature is that the buffer-access-fontify-functions functions add this property, as well as others, to the characters they operate on. That way, they avoid being called over and over for the same text.


33.19.8 Defining Clickable Text

Clickable text is text that can be clicked, with either the mouse or via a keyboard command, to produce some result. Many major modes use clickable text to implement textual hyper-links, or links for short.

The easiest way to insert and manipulate links is to use the button package. See Buttons. In this section, we will explain how to manually set up clickable text in a buffer, using text properties. For simplicity, we will refer to the clickable text as a link.

Implementing a link involves three separate steps: (1) indicating clickability when the mouse moves over the link; (2) making RET or mouse-2 on that link do something; and (3) setting up a follow-link condition so that the link obeys mouse-1-click-follows-link.

To indicate clickability, add the mouse-face text property to the text of the link; then Emacs will highlight the link when the mouse moves over it. In addition, you should define a tooltip or echo area message, using the help-echo text property. See Properties with Special Meanings. For instance, here is how Dired indicates that file names are clickable:

 (if (dired-move-to-filename)
     (add-text-properties
       (point)
       (save-excursion
         (dired-move-to-end-of-filename)
         (point))
       '(mouse-face highlight
         help-echo "mouse-2: visit this file in other window")))

To make the link clickable, bind RET and mouse-2 to commands that perform the desired action. Each command should check to see whether it was called on a link, and act accordingly. For instance, Dired’s major mode keymap binds mouse-2 to the following command:

(defun dired-mouse-find-file-other-window (event)
  "In Dired, visit the file or directory name you click on."
  (interactive "e")
  (let ((window (posn-window (event-end event)))
        (pos (posn-point (event-end event)))
        file)
    (if (not (windowp window))
        (error "No file chosen"))
    (with-current-buffer (window-buffer window)
      (goto-char pos)
      (setq file (dired-get-file-for-visit)))
    (if (file-directory-p file)
        (or (and (cdr dired-subdir-alist)
                 (dired-goto-subdir file))
            (progn
              (select-window window)
              (dired-other-window file)))
      (select-window window)
      (find-file-other-window (file-name-sans-versions file t)))))

This command uses the functions posn-window and posn-point to determine where the click occurred, and dired-get-file-for-visit to determine which file to visit.

Instead of binding the mouse command in a major mode keymap, you can bind it within the link text, using the keymap text property (see Properties with Special Meanings). For instance:

(let ((map (make-sparse-keymap)))
  (define-key map [mouse-2] 'operate-this-button)
  (put-text-property link-start link-end 'keymap map))

With this method, you can easily define different commands for different links. Furthermore, the global definition of RET and mouse-2 remain available for the rest of the text in the buffer.

The basic Emacs command for clicking on links is mouse-2. However, for compatibility with other graphical applications, Emacs also recognizes mouse-1 clicks on links, provided the user clicks on the link quickly without moving the mouse. This behavior is controlled by the user option mouse-1-click-follows-link. See Mouse References in The GNU Emacs Manual.

To set up the link so that it obeys mouse-1-click-follows-link, you must either (1) apply a follow-link text or overlay property to the link text, or (2) bind the follow-link event to a keymap (which can be a major mode keymap or a local keymap specified via the keymap text property). The value of the follow-link property, or the binding for the follow-link event, acts as a condition for the link action. This condition tells Emacs two things: the circumstances under which a mouse-1 click should be regarded as occurring inside the link, and how to compute an action code that says what to translate the mouse-1 click into. The link action condition can be one of the following:

mouse-face

If the condition is the symbol mouse-face, a position is inside a link if there is a non-nil mouse-face property at that position. The action code is always t.

For example, here is how Info mode handles mouse-1:

(keymap-set Info-mode-map "<follow-link>" 'mouse-face)
a function

If the condition is a function, func, then a position pos is inside a link if (func pos) evaluates to non-nil. The value returned by func serves as the action code.

For example, here is how pcvs enables mouse-1 to follow links on file names only:

(keymap-set map "<follow-link>"
            (lambda (pos)
              (eq (get-char-property pos 'face) 'cvs-filename-face)))
anything else

If the condition value is anything else, then the position is inside a link and the condition itself is the action code. Clearly, you should specify this kind of condition only when applying the condition via a text or overlay property on the link text (so that it does not apply to the entire buffer).

The action code tells mouse-1 how to follow the link:

a string or vector

If the action code is a string or vector, the mouse-1 event is translated into the first element of the string or vector; i.e., the action of the mouse-1 click is the local or global binding of that character or symbol. Thus, if the action code is "foo", mouse-1 translates into f. If it is [foo], mouse-1 translates into foo.

anything else

For any other non-nil action code, the mouse-1 event is translated into a mouse-2 event at the same position.

To define mouse-1 to activate a button defined with define-button-type, give the button a follow-link property. The property value should be a link action condition, as described above. See Buttons. For example, here is how Help mode handles mouse-1:

(define-button-type 'help-xref
  'follow-link t
  'action #'help-button-action)

To define mouse-1 on a widget defined with define-widget, give the widget a :follow-link property. The property value should be a link action condition, as described above. For example, here is how the link widget specifies that a mouse-1 click shall be translated to RET:

(define-widget 'link 'item
  "An embedded link."
  :button-prefix 'widget-link-prefix
  :button-suffix 'widget-link-suffix
  :follow-link "\C-m"
  :help-echo "Follow the link."
  :format "%[%t%]")

This function returns non-nil if position pos in the current buffer is on a link. pos can also be a mouse event location, as returned by event-start (see Accessing Mouse Events).


33.19.9 Defining and Using Fields

A field is a range of consecutive characters in the buffer that are identified by having the same value (comparing with eq) of the field property (either a text-property or an overlay property). This section describes special functions that are available for operating on fields.

You specify a field with a buffer position, pos. We think of each field as containing a range of buffer positions, so the position you specify stands for the field containing that position.

When the characters before and after pos are part of the same field, there is no doubt which field contains pos: the one those characters both belong to. When pos is at a boundary between fields, which field it belongs to depends on the stickiness of the field properties of the two surrounding characters (see Stickiness of Text Properties). The field whose property would be inherited by text inserted at pos is the field that contains pos.

There is an anomalous case where newly inserted text at pos would not inherit the field property from either side. This happens if the previous character’s field property is not rear-sticky, and the following character’s field property is not front-sticky. In this case, pos belongs to neither the preceding field nor the following field; the field functions treat it as belonging to an empty field whose beginning and end are both at pos.

In all of these functions, if pos is omitted or nil, the value of point is used by default. If narrowing is in effect, then pos should fall within the accessible portion. See Narrowing.

Function: field-beginning &optional pos escape-from-edge limit

This function returns the beginning of the field specified by pos.

If pos is at the beginning of its field, and escape-from-edge is non-nil, then the return value is always the beginning of the preceding field that ends at pos, regardless of the stickiness of the field properties around pos.

If limit is non-nil, it is a buffer position; if the beginning of the field is before limit, then limit will be returned instead.

Function: field-end &optional pos escape-from-edge limit

This function returns the end of the field specified by pos.

If pos is at the end of its field, and escape-from-edge is non-nil, then the return value is always the end of the following field that begins at pos, regardless of the stickiness of the field properties around pos.

If limit is non-nil, it is a buffer position; if the end of the field is after limit, then limit will be returned instead.

Function: field-string &optional pos

This function returns the contents of the field specified by pos, as a string.

Function: field-string-no-properties &optional pos

This function returns the contents of the field specified by pos, as a string, discarding text properties.

Function: delete-field &optional pos

This function deletes the text of the field specified by pos.

Function: constrain-to-field new-pos old-pos &optional escape-from-edge only-in-line inhibit-capture-property

This function constrains new-pos to the field that old-pos belongs to—in other words, it returns the position closest to new-pos that is in the same field as old-pos.

If new-pos is nil, then constrain-to-field uses the value of point instead, and moves point to the resulting position in addition to returning that position.

If old-pos is at the boundary of two fields, then the acceptable final positions depend on the argument escape-from-edge. If escape-from-edge is nil, then new-pos must be in the field whose field property equals what new characters inserted at old-pos would inherit. (This depends on the stickiness of the field property for the characters before and after old-pos.) If escape-from-edge is non-nil, new-pos can be anywhere in the two adjacent fields. Additionally, if two fields are separated by another field with the special value boundary, then any point within this special field is also considered to be on the boundary.

Commands like C-a with no argument, that normally move backward to a specific kind of location and stay there once there, probably should specify nil for escape-from-edge. Other motion commands that check fields should probably pass t.

If the optional argument only-in-line is non-nil, and constraining new-pos in the usual way would move it to a different line, new-pos is returned unconstrained. This used in commands that move by line, such as next-line and beginning-of-line, so that they respect field boundaries only in the case where they can still move to the right line.

If the optional argument inhibit-capture-property is non-nil, and old-pos has a non-nil property of that name, then any field boundaries are ignored.

You can cause constrain-to-field to ignore all field boundaries (and so never constrain anything) by binding the variable inhibit-field-text-motion to a non-nil value.


33.19.10 Why Text Properties are not Intervals

Some editors that support adding attributes to text in the buffer do so by letting the user specify intervals within the text, and adding the properties to the intervals. Those editors permit the user or the programmer to determine where individual intervals start and end. We deliberately provided a different sort of interface in Emacs Lisp to avoid certain paradoxical behavior associated with text modification.

If the actual subdivision into intervals is meaningful, that means you can distinguish between a buffer that is just one interval with a certain property, and a buffer containing the same text subdivided into two intervals, both of which have that property.

Suppose you take the buffer with just one interval and kill part of the text. The text remaining in the buffer is one interval, and the copy in the kill ring (and the undo list) becomes a separate interval. Then if you yank back the killed text, you get two intervals with the same properties. Thus, editing does not preserve the distinction between one interval and two.

Suppose we attempt to fix this problem by coalescing the two intervals when the text is inserted. That works fine if the buffer originally was a single interval. But suppose instead that we have two adjacent intervals with the same properties, and we kill the text of one interval and yank it back. The same interval-coalescence feature that rescues the other case causes trouble in this one: after yanking, we have just one interval. Once again, editing does not preserve the distinction between one interval and two.

Insertion of text at the border between intervals also raises questions that have no satisfactory answer.

However, it is easy to arrange for editing to behave consistently for questions of the form, “What are the properties of text at this buffer or string position?” So we have decided these are the only questions that make sense; we have not implemented asking questions about where intervals start or end.

In practice, you can usually use the text property search functions in place of explicit interval boundaries. You can think of them as finding the boundaries of intervals, assuming that intervals are always coalesced whenever possible. See Text Property Search Functions.

Emacs also provides explicit intervals as a presentation feature; see Overlays.


33.20 Substituting for a Character Code

The following functions replace characters within a specified region based on their character codes.

Function: subst-char-in-region start end old-char new-char &optional noundo

This function replaces all occurrences of the character old-char with the character new-char in the region of the current buffer defined by start and end. Both characters must have the same length of their multibyte form.

If noundo is non-nil, then subst-char-in-region does not record the change for undo and does not mark the buffer as modified. This was useful for controlling the old selective display feature (see Selective Display).

subst-char-in-region does not move point and returns nil.

---------- Buffer: foo ----------
This is the contents of the buffer before.
---------- Buffer: foo ----------

(subst-char-in-region 1 20 ?i ?X)
     ⇒ nil

---------- Buffer: foo ----------
ThXs Xs the contents of the buffer before.
---------- Buffer: foo ----------
Function: subst-char-in-string fromchar tochar string &optional inplace

This function replaces all occurrences of the character fromchar with tochar in string. By default, substitution occurs in a copy of string, but if the optional argument inplace is non-nil, the function modifies the string itself. In any case, the function returns the resulting string.

Command: translate-region start end table

This function applies a translation table to the characters in the buffer between positions start and end.

The translation table table is a string or a char-table; (aref table ochar) gives the translated character corresponding to ochar. If table is a string, any characters with codes larger than the length of table are not altered by the translation.

The return value of translate-region is the number of characters that were actually changed by the translation. This does not count characters that were mapped into themselves in the translation table.


33.21 Registers

A register is a sort of variable used in Emacs editing that can hold a variety of different kinds of values. Each register is named by a single character. All ASCII characters and their meta variants (but with the exception of C-g) can be used to name registers. Thus, there are 255 possible registers. A register is designated in Emacs Lisp by the character that is its name.

Variable: register-alist

This variable is an alist of elements of the form (name . contents). Normally, there is one element for each Emacs register that has been used.

The object name is a character (an integer) identifying the register.

The contents of a register can have several possible types:

a number

A number stands for itself. If insert-register finds a number in the register, it converts the number to decimal.

a marker

A marker represents a buffer position to jump to.

a string

A string is text saved in the register.

a rectangle

A rectangle is represented by a list of strings.

(window-configuration position)

This represents a window configuration to restore in one frame, and a position to jump to in the current buffer.

(frame-configuration position)

This represents a frame configuration to restore, and a position to jump to in the current buffer. Frame configurations are also known as framesets.

(file filename)

This represents a file to visit; jumping to this value visits file filename.

(file-query filename position)

This represents a file to visit and a position in it; jumping to this value visits file filename and goes to buffer position position. Restoring this type of position asks the user for confirmation first.

(buffer buffer-name)

This represents a buffer; jumping to this value switches to buffer buffer-name.

The functions in this section return unpredictable values unless otherwise stated.

Function: get-register reg

This function returns the contents of the register reg, or nil if it has no contents.

Function: set-register reg value

This function sets the contents of register reg to value. A register can be set to any value, but the other register functions expect only certain data types. The return value is value.

Command: view-register reg

This command displays what is contained in register reg.

Command: insert-register reg &optional beforep

This command inserts contents of register reg into the current buffer.

Normally, this command puts point before the inserted text, and the mark after it. However, if the optional second argument beforep is non-nil, it puts the mark before and point after.

When called interactively, the command defaults to putting point after text, and a prefix argument inverts this behavior.

If the register contains a rectangle, then the rectangle is inserted with its upper left corner at point. This means that text is inserted in the current line and underneath it on successive lines.

If the register contains something other than saved text (a string) or a rectangle (a list), currently useless things happen. This may be changed in the future.

Function: register-read-with-preview prompt

This function reads and returns a register name, prompting with prompt and possibly showing a preview of the existing registers and their contents. The preview is shown in a temporary window, after the delay specified by the user option register-preview-delay, if its value and register-alist are both non-nil. The preview is also shown if the user requests help (e.g., by typing the help character). We recommend that all interactive commands which read register names use this function.


33.22 Transposition of Text

This function can be used to transpose stretches of text:

Function: transpose-regions start1 end1 start2 end2 &optional leave-markers

This function exchanges two nonoverlapping portions of the buffer (if they overlap, the function signals an error). Arguments start1 and end1 specify the bounds of one portion and arguments start2 and end2 specify the bounds of the other portion.

Normally, transpose-regions relocates markers with the transposed text; a marker previously positioned within one of the two transposed portions moves along with that portion, thus remaining between the same two characters in their new position. However, if leave-markers is non-nil, transpose-regions does not do this—it leaves all markers unrelocated.


33.23 Replacing Buffer Text

You can use the following function to replace the text of one buffer with the text of another buffer:

Command: replace-buffer-contents source &optional max-secs max-costs

This function replaces the accessible portion of the current buffer with the accessible portion of the buffer source. source may either be a buffer object or the name of a buffer. When replace-buffer-contents succeeds, the text of the accessible portion of the current buffer will be equal to the text of the accessible portion of the source buffer.

This function attempts to keep point, markers, text properties, and overlays in the current buffer intact. One potential case where this behavior is useful is external code formatting programs: they typically write the reformatted text into a temporary buffer or file, and using delete-region and insert-buffer-substring would destroy these properties. However, the latter combination is typically faster (See Deleting Text, and Inserting Text).

For its working, replace-buffer-contents needs to compare the contents of the original buffer with that of source which is a costly operation if the buffers are huge and there is a high number of differences between them. In order to keep replace-buffer-contents’s runtime in bounds, it has two optional arguments.

max-secs defines a hard boundary in terms of seconds. If given and exceeded, it will fall back to delete-region and insert-buffer-substring.

max-costs defines the quality of the difference computation. If the actual costs exceed this limit, heuristics are used to provide a faster but suboptimal solution. The default value is 1000000.

replace-buffer-contents returns t if a non-destructive replacement could be performed. Otherwise, i.e., if max-secs was exceeded, it returns nil.

Function: replace-region-contents beg end replace-fn &optional max-secs max-costs

This function replaces the region between beg and end using the given replace-fn. The function replace-fn is run in the current buffer narrowed to the specified region and it should return either a string or a buffer replacing the region.

The replacement is performed using replace-buffer-contents (see above) which also describes the max-secs and max-costs arguments and the return value.

Note: If the replacement is a string, it will be placed in a temporary buffer so that replace-buffer-contents can operate on it. Therefore, if you already have the replacement in a buffer, it makes no sense to convert it to a string using buffer-substring or similar.


33.24 Dealing With Compressed Data

When auto-compression-mode is enabled, Emacs automatically uncompresses compressed files when you visit them, and automatically recompresses them if you alter and save them. See Compressed Files in The GNU Emacs Manual.

The above feature works by calling an external executable (e.g., gzip). Emacs can also be compiled with support for built-in decompression using the zlib library, which is faster than calling an external program.

Function: zlib-available-p

This function returns non-nil if built-in zlib decompression is available.

Function: zlib-decompress-region start end &optional allow-partial

This function decompresses the region between start and end, using built-in zlib decompression. The region should contain data that were compressed with gzip or zlib. On success, the function replaces the contents of the region with the decompressed data. If allow-partial is nil or omitted, then on failure, the function leaves the region unchanged and returns nil. Otherwise, it returns the number of bytes that were not decompressed and replaces the region text by whatever data was successfully decompressed. This function can be called only in unibyte buffers.


33.25 Base 64 Encoding

Base 64 code is used in email to encode a sequence of 8-bit bytes as a longer sequence of ASCII graphic characters. It is defined in Internet RFC232045 and also in RFC 4648. This section describes the functions for converting to and from this code.

Command: base64-encode-region beg end &optional no-line-break

This function converts the region from beg to end into base 64 code. It returns the length of the encoded text. An error is signaled if a character in the region is multibyte, i.e., in a multibyte buffer the region must contain only ASCII characters or raw bytes.

Normally, this function inserts newline characters into the encoded text, to avoid overlong lines. However, if the optional argument no-line-break is non-nil, these newlines are not added, so the output is just one long line.

Command: base64url-encode-region beg end &optional no-pad

This function is like base64-encode-region, but it implements the URL variant of base 64 encoding, per RFC 4648, and it doesn’t insert newline characters into the encoded text, so the output is just one long line.

If the optional argument no-pad is non-nil then this function doesn’t generate the padding (=).

Function: base64-encode-string string &optional no-line-break

This function converts the string string into base 64 code. It returns a string containing the encoded text. As for base64-encode-region, an error is signaled if a character in the string is multibyte.

Normally, this function inserts newline characters into the encoded text, to avoid overlong lines. However, if the optional argument no-line-break is non-nil, these newlines are not added, so the result string is just one long line.

Function: base64url-encode-string string &optional no-pad

Like base64-encode-string, but generates the URL variant of base 64, and doesn’t insert newline characters into the encoded text, so the result is just one long line.

If the optional argument no-pad is non-nil then this function doesn’t generate the padding.

Command: base64-decode-region beg end &optional base64url ignore-invalid

This function converts the region from beg to end from base 64 code into the corresponding decoded text. It returns the length of the decoded text.

The decoding functions ignore newline characters in the encoded text.

If optional argument base64url is non-nil, then padding is optional, and the URL variant of base 64 encoding is used. If optional argument ignore-invalid is non-nil, then any unrecognized characters are ignored.

Function: base64-decode-string string &optional base64url ignore-invalid

This function converts the string string from base 64 code into the corresponding decoded text. It returns a unibyte string containing the decoded text.

The decoding functions ignore newline characters in the encoded text.

If optional argument base64url is non-nil, then padding is optional, and the URL variant of base 64 encoding is used. If optional argument ignore-invalid is non-nil, then any unrecognized characters are ignored.


33.26 Checksum/Hash

Emacs has built-in support for computing cryptographic hashes. A cryptographic hash, or checksum, is a digital fingerprint of a piece of data (e.g., a block of text) which can be used to check that you have an unaltered copy of that data.

Emacs supports several common cryptographic hash algorithms: MD5, SHA-1, SHA-2, SHA-224, SHA-256, SHA-384 and SHA-512. MD5 is the oldest of these algorithms, and is commonly used in message digests to check the integrity of messages transmitted over a network. MD5 and SHA-1 are not collision resistant (i.e., it is possible to deliberately design different pieces of data which have the same MD5 or SHA-1 hash), so you should not use them for anything security-related. For security-related applications you should use the other hash types, such as SHA-2 (e.g. sha256 or sha512).

Function: secure-hash-algorithms

This function returns a list of symbols representing algorithms that secure-hash can use.

Function: secure-hash algorithm object &optional start end binary

This function returns a hash for object. The argument algorithm is a symbol stating which hash to compute: one of md5, sha1, sha224, sha256, sha384 or sha512. The argument object should be a buffer or a string.

The optional arguments start and end are character positions specifying the portion of object to compute the message digest for. If they are nil or omitted, the hash is computed for the whole of object.

If the argument binary is omitted or nil, the function returns the text form of the hash, as an ordinary Lisp string. If binary is non-nil, it returns the hash in binary form, as a sequence of bytes stored in a unibyte string. The length of the returned string depends on algorithm:

  • For md5: 32 characters (32 bytes if binary is non-nil).
  • For sha1: 40 characters (40 bytes if binary is non-nil).
  • For sha224: 56 characters (56 bytes if binary is non-nil).
  • For sha256: 64 characters (64 bytes if binary is non-nil).
  • For sha384: 96 characters (96 bytes if binary is non-nil).
  • For sha512: 128 characters (128 bytes if binary is non-nil).

This function does not compute the hash directly from the internal representation of object’s text (see Text Representations). Instead, it encodes the text using a coding system (see Coding Systems), and computes the hash from that encoded text. If object is a buffer, the coding system used is the one which would be chosen by default for writing the text of that buffer into a file. If object is a string, the user’s preferred coding system is used (see Recognize Coding in GNU Emacs Manual).

Function: md5 object &optional start end coding-system noerror

This function returns an MD5 hash. It is semi-obsolete, since for most purposes it is equivalent to calling secure-hash with md5 as the algorithm argument. The object, start and end arguments have the same meanings as in secure-hash. The function returns a 32-character string.

If coding-system is non-nil, it specifies a coding system to use to encode the text; if omitted or nil, the default coding system is used, like in secure-hash.

Normally, md5 signals an error if the text can’t be encoded using the specified or chosen coding system. However, if noerror is non-nil, it silently uses raw-text coding instead.

Function: buffer-hash &optional buffer-or-name

Return a hash of buffer-or-name. If nil, this defaults to the current buffer. As opposed to secure-hash, this function computes the hash based on the internal representation of the buffer, disregarding any coding systems. It’s therefore only useful when comparing two buffers running in the same Emacs, and is not guaranteed to return the same hash between different Emacs versions. It should be somewhat more efficient on larger buffers than secure-hash is, and should not allocate more memory.

Function: sha1 object &optional start end binary

This function is equivalent to calling secure-hash like this:

(secure-hash 'sha1 object start end binary)

It returns a 40-character string if binary is nil, or a 40-byte unibyte string otherwise.


33.27 Suspicious Text

Emacs can display text from many external sources, like email and Web sites. Attackers may attempt to confuse the user reading this text by using obfuscated URLs or email addresses, and tricking the user into visiting a web page they didn’t intend to visit, or sending an email to the wrong address.

This usually involves using characters from scripts that visually look like ASCII characters (i.e., are homoglyphs), but there are also other techniques used, like using bidirectional overrides, or having an HTML link text that says one thing, while the underlying URL points somewhere else.

To help identify these suspicious text strings, Emacs provides a library to do a number of checks on text. (See UTS #39: Unicode Security Mechanisms for the rationale behind the checks that are available and more details about them.) Packages that present data that might be suspicious should use this library to flag suspicious text on display.

Function: textsec-suspicious-p object type

This function is the high-level interface function that packages should use. It respects the textsec-check user option, which allows the user to disable the checks.

This function checks object (whose data type depends on type) to see if it looks suspicious when interpreted as a thing of type. The available types and the corresponding object data types are:

domain

Check whether a domain (e.g., ‘www.gnu.org’ looks suspicious. object should be a string, the domain name.

url

Check whether an URL (e.g., ‘http://gnu.org/foo/bar’) looks suspicious. object should be a string, the URL to check.

link

Check whether an HTML link (e.g., ‘<a href='http://gnu.org'>fsf.org</a>’ looks suspicious. In this case, object should be a cons cell where the car is the URL string, and the cdr is the link text. The link is deemed suspicious if the link text contains a domain name, and that domain name points to something other than the URL.

email-address

Check whether an email address (e.g., ‘foo@example.org’) looks suspicious. object should be a string.

local-address

Check whether the local part of an email address (the bit before the ‘@’ sign) looks suspicious. object should be a string.

name

Check whether a name (used in an email address header) looks suspicious. object should be a string.

email-address-header

Check whether a full RFC2822 email address header (e.g., ‘=?utf-8?Q?=C3=81?= <foo@example.com>’) looks suspicious. object should be a string.

If object is suspicious, this function returns a string that explains why it is suspicious. If object is not suspicious, the function returns nil.

If the text is suspicious, the application should mark the suspicious text with the textsec-suspicious face, and make the explanation returned by textsec-suspicious-p available to the user in some way (for example, in a tooltip). The application might also prompt the user for confirmation before taking any action on a suspicious string (like sending an email to a suspicious email address).


33.28 GnuTLS Cryptography

If compiled with GnuTLS, Emacs offers built-in cryptographic support. Following the GnuTLS API terminology, the available tools are digests, MACs, symmetric ciphers, and AEAD ciphers.

The terms used herein, such as IV (Initialization Vector), require some familiarity with cryptography and will not be defined in detail. Please consult https://www.gnutls.org/ for specific documentation which may help you understand the terminology and structure of the GnuTLS library.


33.28.1 Format of GnuTLS Cryptography Inputs

The inputs to GnuTLS cryptographic functions can be specified in several ways, both as primitive Emacs Lisp types or as lists.

The list form is currently similar to how md5 and secure-hash operate.

buffer

Simply passing a buffer as input means the whole buffer should be used.

string

A string as input will be used directly. It may be modified by the function (unlike most other Emacs Lisp functions) to reduce the chance of exposing sensitive data after the function does its work.

(buffer-or-string start end coding-system noerror)

This specifies a buffer or a string as described above, but an optional range can be specified with start and end.

In addition an optional coding-system can be specified if needed.

The last optional item, noerror, overrides the normal error when the text can’t be encoded using the specified or chosen coding system. When noerror is non-nil, this function silently uses raw-text coding instead.

(iv-auto length)

This generates a random IV (Initialization Vector) of the specified length and passes it to the function. This ensures that the IV is unpredictable and unlikely to be reused in the same session.


33.28.2 GnuTLS Cryptographic Functions

Function: gnutls-digests

This function returns the alist of the GnuTLS digest algorithms.

Each entry has a key which represents the algorithm, followed by a plist with internal details about the algorithm. The plist will have :type gnutls-digest-algorithm and also will have the key :digest-algorithm-length 64 to indicate the size, in bytes, of the resulting digest.

There is a name parallel between GnuTLS MAC and digest algorithms but they are separate things internally and should not be mixed.

Function: gnutls-hash-digest digest-method input

The digest-method can be the whole plist from gnutls-digests, or just the symbol key, or a string with the name of that symbol.

The input can be specified as a buffer or string or in other ways (see Format of GnuTLS Cryptography Inputs).

This function returns nil on error, and signals a Lisp error if the digest-method or input are invalid. On success, it returns a list of a binary string (the output) and the IV used.

Function: gnutls-macs

This function returns the alist of the GnuTLS MAC algorithms.

Each entry has a key which represents the algorithm, followed by a plist with internal details about the algorithm. The plist will have :type gnutls-mac-algorithm and also will have the keys :mac-algorithm-length :mac-algorithm-keysize :mac-algorithm-noncesize to indicate the size, in bytes, of the resulting hash, the key, and the nonce respectively.

The nonce is currently unused and only some MACs support it.

There is a name parallel between GnuTLS MAC and digest algorithms but they are separate things internally and should not be mixed.

Function: gnutls-hash-mac hash-method key input

The hash-method can be the whole plist from gnutls-macs, or just the symbol key, or a string with the name of that symbol.

The key can be specified as a buffer or string or in other ways (see Format of GnuTLS Cryptography Inputs). The key will be wiped after use if it’s a string.

The input can be specified as a buffer or string or in other ways (see Format of GnuTLS Cryptography Inputs).

This function returns nil on error, and signals a Lisp error if the hash-method or key or input are invalid.

On success, it returns a list of a binary string (the output) and the IV used.

Function: gnutls-ciphers

This function returns the alist of the GnuTLS ciphers.

Each entry has a key which represents the cipher, followed by a plist with internal details about the algorithm. The plist will have :type gnutls-symmetric-cipher and also will have the keys :cipher-aead-capable set to nil or t to indicate AEAD capability; and :cipher-tagsize :cipher-blocksize :cipher-keysize :cipher-ivsize to indicate the size, in bytes, of the tag, block size of the resulting data, the key, and the IV respectively.

Function: gnutls-symmetric-encrypt cipher key iv input &optional aead_auth

The cipher can be the whole plist from gnutls-ciphers, or just the symbol key, or a string with the name of that symbol.

The key can be specified as a buffer or string or in other ways (see Format of GnuTLS Cryptography Inputs). The key will be wiped after use if it’s a string.

The iv and input and the optional aead_auth can be specified as a buffer or string or in other ways (see Format of GnuTLS Cryptography Inputs).

aead_auth is only checked with AEAD ciphers, that is, ciphers whose plist has :cipher-aead-capable t. Otherwise it’s ignored.

This function returns nil on error, and signals a Lisp error if the cipher or key, iv, or input are invalid, or if aead_auth was specified with an AEAD cipher and was invalid.

On success, it returns a list of a binary string (the output) and the IV used.

Function: gnutls-symmetric-decrypt cipher key iv input &optional aead_auth

The cipher can be the whole plist from gnutls-ciphers, or just the symbol key, or a string with the name of that symbol.

The key can be specified as a buffer or string or in other ways (see Format of GnuTLS Cryptography Inputs). The key will be wiped after use if it’s a string.

The iv and input and the optional aead_auth can be specified as a buffer or string or in other ways (see Format of GnuTLS Cryptography Inputs).

aead_auth is only checked with AEAD ciphers, that is, ciphers whose plist has :cipher-aead-capable t. Otherwise it’s ignored.

This function returns nil on decryption error, and signals a Lisp error if the cipher or key, iv, or input are invalid, or if aead_auth was specified with an AEAD cipher and was invalid.

On success, it returns a list of a binary string (the output) and the IV used.


33.29 Database

Emacs can be compiled with built-in support for accessing SQLite databases. This section describes the facilities available for accessing SQLite databases from Lisp programs.

Function: sqlite-available-p

The function returns non-nil if built-in SQLite support is available in this Emacs session.

When SQLite support is available, the following functions can be used.

Function: sqlite-open &optional file

This function opens file as an SQLite database file. If file doesn’t exist, a new database will be created and stored in that file. If file is omitted or nil, a new in-memory database is created instead.

The return value is a database object that can be used as the argument to most of the subsequent functions described below.

Function: sqlitep object

This predicate returns non-nil if object is an SQLite database object. The database object returned by the sqlite-open function satisfies this predicate.

Function: sqlite-close db

Close the database db. It’s usually not necessary to call this function explicitly—the database will automatically be closed if Emacs shuts down or the database object is garbage collected.

Function: sqlite-execute db statement &optional values

Execute the SQL statement. For instance:

(sqlite-execute db "insert into foo values ('bar', 2)")

If the optional values parameter is present, it should be either a list or a vector of values to bind while executing the statement. For instance:

(sqlite-execute db "insert into foo values (?, ?)" '("bar" 2))

This has exactly the same effect as the previous example, but is more efficient and safer (because it doesn’t involve any string parsing or interpolation).

sqlite-execute usually returns the number of affected rows. For instance, an ‘insert’ statement will typically return ‘1’, whereas an ‘update’ statement may return zero or a higher number. However, when using SQL statements like ‘insert into … returning … and the like, the values specified by ‘returning … will be returned instead.

Strings in SQLite are, by default, stored as utf-8, and selecting a text column will decode the string using that charset. Selecting a blob column will return the raw data without any decoding (i.e., it will return a unibyte string containing the bytes as stored in the database). Inserting binary data into blob columns, however, requires some care, as sqlite-execute will, by default, interpret all strings as utf-8.

So if you have, for instance, GIF data in a unibyte string called gif, you have to mark it specially to let sqlite-execute know this:

(put-text-property 0 1 'coding-system 'binary gif)
(sqlite-execute db "insert into foo values (?, ?)" (list gif 2))
Function: sqlite-select db query &optional values return-type

Select some data from db and return them. For instance:

(sqlite-select db "select * from foo where key = 2")
  ⇒ (("bar" 2))

As with the sqlite-execute, you can optionally pass in a list or a vector of values that will be bound before executing the select:

(sqlite-select db "select * from foo where key = ?" [2])
  ⇒ (("bar" 2))

This is usually more efficient and safer than the method used by the previous example.

By default, this function returns a list of matching rows, where each row is a list of column values. If return-type is full, the names of the columns (as a list of strings) will be returned as the first element in the return value.

If return-type is set, this function will return a statement object instead. This object can be examined by using the sqlite-next, sqlite-columns and sqlite-more-p functions. If the result set is small, it’s often more convenient to just return the data directly, but if the result set is large (or if you won’t be using all the data from the set), using the set method will allocate a lot less memory, and is therefore more memory-efficient.

Function: sqlite-next statement

This function returns the next row in the result set statement, typically an object returned by sqlite-select.

(sqlite-next stmt)
    ⇒ ("bar" 2)
Function: sqlite-columns statement

This function returns the column names of the result set statement, typically an object returned by sqlite-select.

(sqlite-columns stmt)
    ⇒ ("name" "issue")
Function: sqlite-more-p statement

This predicate says whether there is more data to be fetched from the result set statement, typically an object returned by sqlite-select.

Function: sqlite-finalize statement

If statement is not going to be used any more, calling this function will free the resources used by statement. This is usually not necessary—when the statement object is garbage-collected, Emacs will automatically free its resources.

Function: sqlite-transaction db

Start a transaction in db. When in a transaction, other readers of the database won’t access the results until the transaction has been committed by sqlite-commit.

Function: sqlite-commit db

End a transaction in db and write the data out to its file.

Function: sqlite-rollback db

End a transaction in db and discard any changes that have been made by the transaction.

Macro: with-sqlite-transaction db body…

Like progn (see Sequencing), but executes body with a transaction held, and commits the transaction at the end if body completes normally. If body signals an error, or committing the transaction fails, the changes in db performed by body are rolled back. The macro returns the value of body if it completes normally and commit succeeds.

Function: sqlite-pragma db pragma

Execute pragma in db. A pragma is usually a command that affects the database overall, instead of any particular table. For instance, to make SQLite automatically garbage collect data that’s no longer needed, you can say:

(sqlite-pragma db "auto_vacuum = FULL")

This function returns non-nil on success and nil if the pragma failed. Many pragmas can only be issued when the database is brand new and empty.

Function: sqlite-load-extension db module

Load the named extension module into the database db. Extensions are usually shared-library files; on GNU and Unix systems, they have the .so file-name extension.

Function: sqlite-version

Return a string denoting the version of the SQLite library in use.

If you wish to list the contents of an SQLite file, you can use the sqlite-mode-open-file command. This will pop to a buffer using sqlite-mode, which allows you to examine (and alter) the contents of an SQLite database.


33.30 Parsing HTML and XML

Emacs can be compiled with built-in libxml2 support.

Function: libxml-available-p

This function returns non-nil if built-in libxml2 support is available in this Emacs session.

When libxml2 support is available, the following functions can be used to parse HTML or XML text into Lisp object trees.

Function: libxml-parse-html-region &optional start end base-url discard-comments

This function parses the text between start and end as HTML, and returns a list representing the HTML parse tree. It attempts to handle real-world HTML by robustly coping with syntax mistakes.

If start or end are nil, they default to the values from point-min and point-max, respectively.

The optional argument base-url, if non-nil, should be used for warnings and errors reported by the libxml2 library, but Emacs currently calls the library with errors and warnings disabled, so this argument is not used.

If the optional argument discard-comments is non-nil, any top-level comment is discarded. (This argument is obsolete and will be removed in future Emacs versions. To remove comments, use the xml-remove-comments utility function on the data before you call the parsing function.)

In the parse tree, each HTML node is represented by a list in which the first element is a symbol representing the node name, the second element is an alist of node attributes, and the remaining elements are the subnodes.

The following example demonstrates this. Given this (malformed) HTML document:

<html><head></head><body width=101><div class=thing>Foo<div>Yes

A call to libxml-parse-html-region returns this DOM (document object model):

(html nil
 (head nil)
 (body ((width . "101"))
  (div ((class . "thing"))
   "Foo"
   (div nil
    "Yes"))))
Function: shr-insert-document dom

This function renders the parsed HTML in dom into the current buffer. The argument dom should be a list as generated by libxml-parse-html-region. This function is, e.g., used by EWW in The Emacs Web Wowser Manual.

Function: libxml-parse-xml-region &optional start end base-url discard-comments

This function is the same as libxml-parse-html-region, except that it parses the text as XML rather than HTML (so it is stricter about syntax).


33.30.1 Document Object Model

The DOM returned by libxml-parse-html-region (and the other XML parsing functions) is a tree structure where each node has a node name (called a tag), and optional key/value attribute list, and then a list of child nodes. The child nodes are either strings or DOM objects.

(body ((width . "101"))
 (div ((class . "thing"))
  "Foo"
  (div nil
   "Yes")))
Function: dom-node tag &optional attributes &rest children

This function creates a DOM node of type tag. If given, attributes should be a key/value pair list. If given, children should be DOM nodes.

The following functions can be used to work with this structure. Each function takes a DOM node, or a list of nodes. In the latter case, only the first node in the list is used.

Simple accessors:

dom-tag node

Return the tag (also called “node name”) of the node.

dom-attr node attribute

Return the value of attribute in the node. A common usage would be:

(dom-attr img 'href)
=> "https://fsf.org/logo.png"
dom-children node

Return all the children of the node.

dom-non-text-children node

Return all the non-string children of the node.

dom-attributes node

Return the key/value pair list of attributes of the node.

dom-text node

Return all the textual elements of the node as a concatenated string.

dom-texts node

Return all the textual elements of the node, as well as the textual elements of all the children of the node, recursively, as a concatenated string. This function also takes an optional separator to be inserted between the textual elements.

dom-parent dom node

Return the parent of node in dom.

dom-remove dom node

Remove node from dom.

The following are functions for altering the DOM.

dom-set-attribute node attribute value

Set the attribute of the node to value.

dom-remove-attribute node attribute

Remove attribute from node.

dom-append-child node child

Append child as the last child of node.

dom-add-child-before node child before

Add child to node’s child list before the before node. If before is nil, make child the first child.

dom-set-attributes node attributes

Replace all the attributes of the node with a new key/value list.

The following are functions for searching for elements in the DOM. They all return lists of matching nodes.

dom-by-tag dom tag

Return all nodes in dom that are of type tag. A typical use would be:

(dom-by-tag dom 'td)
=> '((td ...) (td ...) (td ...))
dom-by-class dom match

Return all nodes in dom that have class names that match match, which is a regular expression.

dom-by-style dom style

Return all nodes in dom that have styles that match match, which is a regular expression.

dom-by-id dom style

Return all nodes in dom that have IDs that match match, which is a regular expression.

dom-search dom predicate

Return all nodes in dom where predicate returns a non-nil value. predicate is called with the node to be tested as its parameter.

dom-strings dom

Return all strings in dom.

Utility functions:

dom-pp dom &optional remove-empty

Pretty-print dom at point. If remove-empty, don’t print textual nodes that just contain white-space.

dom-print dom &optional pretty xml

Print dom at point. If xml is non-nil, print as XML; otherwise, print as HTML. If pretty is non-nil, indent the HTML/XML logically.


33.31 Parsing and generating JSON values

When Emacs is compiled with JSON (JavaScript Object Notation) support, it provides several functions to convert between Lisp objects and JSON values. Any JSON value can be converted to a Lisp object, but not vice versa. Specifically:

  • JSON uses three keywords: true, null, false. true is represented by the symbol t. By default, the remaining two are represented, respectively, by the symbols :null and :false.
  • JSON only has floating-point numbers. They can represent both Lisp integers and Lisp floating-point numbers.
  • JSON strings are always Unicode strings encoded in UTF-8. Lisp strings can contain non-Unicode characters.
  • JSON has only one sequence type, the array. JSON arrays are represented using Lisp vectors.
  • JSON has only one map type, the object. JSON objects are represented using Lisp hashtables, alists or plists. When an alist or plist contains several elements with the same key, Emacs uses only the first element for serialization, in accordance with the behavior of assq.

Note that nil, being both a valid alist and a valid plist, represents {}, the empty JSON object; not null, false, or an empty array, all of which are different JSON values.

Function: json-available-p

This predicate returns non-nil if Emacs has been built with JSON support, and the library is available on the current system.

If some Lisp object can’t be represented in JSON, the serialization functions will signal an error of type wrong-type-argument. The parsing functions can also signal the following errors:

json-unavailable

Signaled when the parsing library isn’t available.

json-end-of-file

Signaled when encountering a premature end of the input text.

json-trailing-content

Signaled when encountering unexpected input after the first JSON object parsed.

json-parse-error

Signaled when encountering invalid JSON syntax.

Top-level values and the subobjects within these top-level values can be serialized to JSON. Likewise, the parsing functions will return any of the possible types described above.

Function: json-serialize object &rest args

This function returns a new Lisp string which contains the JSON representation of object. The argument args is a list of keyword/argument pairs. The following keywords are accepted:

:null-object

The value decides which Lisp object to use to represent the JSON keyword null. It defaults to the symbol :null.

:false-object

The value decides which Lisp object to use to represent the JSON keyword false. It defaults to the symbol :false.

Function: json-insert object &rest args

This function inserts the JSON representation of object into the current buffer before point. The argument args are interpreted as in json-parse-string.

Function: json-parse-string string &rest args

This function parses the JSON value in string, which must be a Lisp string. If string doesn’t contain a valid JSON object, this function signals the json-parse-error error.

The argument args is a list of keyword/argument pairs. The following keywords are accepted:

:object-type

The value decides which Lisp object to use for representing the key-value mappings of a JSON object. It can be either hash-table, the default, to make hashtables with strings as keys; alist to use alists with symbols as keys; or plist to use plists with keyword symbols as keys.

:array-type

The value decides which Lisp object to use for representing a JSON array. It can be either array, the default, to use Lisp arrays; or list to use lists.

:null-object

The value decides which Lisp object to use to represent the JSON keyword null. It defaults to the symbol :null.

:false-object

The value decides which Lisp object to use to represent the JSON keyword false. It defaults to the symbol :false.

Function: json-parse-buffer &rest args

This function reads the next JSON value from the current buffer, starting at point. It moves point to the position immediately after the value if contains a valid JSON object; otherwise it signals the json-parse-error error and doesn’t move point. The arguments args are interpreted as in json-parse-string.


33.32 JSONRPC communication

The jsonrpc library implements the JSONRPC specification, version 2.0, as it is described in https://www.jsonrpc.org/. As the name suggests, JSONRPC is a generic Remote Procedure Call protocol designed around JSON objects, which you can convert to and from Lisp objects (see Parsing and generating JSON values).


33.32.1 Overview

Quoting from the spec, JSONRPC "is transport agnostic in that the concepts can be used within the same process, over sockets, over http, or in many various message passing environments."

To model this agnosticism, the jsonrpc library uses objects of a jsonrpc-connection class, which represent a connection to a remote JSON endpoint (for details on Emacs’s object system, see EIEIO in EIEIO). In modern object-oriented parlance, this class is “abstract”, i.e. the actual class of a useful connection object is always a subclass of jsonrpc-connection. Nevertheless, we can define two distinct APIs around the jsonrpc-connection class:

  1. A user interface for building JSONRPC applications

    In this scenario, the JSONRPC application selects a concrete subclass of jsonrpc-connection, and proceeds to create objects of that subclass using make-instance. To initiate a contact to the remote endpoint, the JSONRPC application passes this object to the functions jsonrpc-notify, jsonrpc-request, and/or jsonrpc-async-request. For handling remotely initiated contacts, which generally come in asynchronously, the instantiation should include :request-dispatcher and :notification-dispatcher initargs, which are both functions of 3 arguments: the connection object; a symbol naming the JSONRPC method invoked remotely; and a JSONRPC params object.

    The function passed as :request-dispatcher is responsible for handling the remote endpoint’s requests, which expect a reply from the local endpoint (in this case, the program you’re building). Inside that function, you may either return locally (a normal return) or non-locally (an error return). A local return value must be a Lisp object that can be serialized as JSON (see Parsing and generating JSON values). This determines a success response, and the object is forwarded to the server as the JSONRPC result object. A non-local return, achieved by calling the function jsonrpc-error, causes an error response to be sent to the server. The details of the accompanying JSONRPC error are filled out with whatever was passed to jsonrpc-error. A non-local return triggered by an unexpected error of any other type also causes an error response to be sent (unless you have set debug-on-error, in which case this calls the Lisp debugger, see Entering the Debugger on an Error).

  2. A inheritance interface for building JSONRPC transport implementations

    In this scenario, jsonrpc-connection is subclassed to implement a different underlying transport strategy (for details on how to subclass, see (eieio)Inheritance.). Users of the application-building interface can then instantiate objects of this concrete class (using the make-instance function) and connect to JSONRPC endpoints using that strategy.

    This API has mandatory and optional parts.

    To allow its users to initiate JSONRPC contacts (notifications or requests) or reply to endpoint requests, the subclass must have an implementation of the jsonrpc-connection-send method.

    Likewise, for handling the three types of remote contacts (requests, notifications, and responses to local requests), the transport implementation must arrange for the function jsonrpc-connection-receive to be called after noticing a new JSONRPC message on the wire (whatever that "wire" may be).

    Finally, and optionally, the jsonrpc-connection subclass should implement the jsonrpc-shutdown and jsonrpc-running-p methods if these concepts apply to the transport. If they do, then any system resources (e.g. processes, timers, etc.) used to listen for messages on the wire should be released in jsonrpc-shutdown, i.e. they should only be needed while jsonrpc-running-p is non-nil.


33.32.2 Process-based JSONRPC connections

For convenience, the jsonrpc library comes with a built-in jsonrpc-process-connection transport implementation that can talk to local subprocesses (using the standard input and standard output); or TCP hosts (using sockets); or any other remote endpoint that Emacs’s process object can represent (see Processes).

Using this transport, the JSONRPC messages are encoded on the wire as plain text and prefaced by some basic HTTP-style enveloping headers, such as “Content-Length”.

For an example of an application using this transport scheme on top of JSONRPC, see the Language Server Protocol.

Along with the mandatory :request-dispatcher and :notification-dispatcher initargs, users of the jsonrpc-process-connection class should pass the following initargs as keyword-value pairs to make-instance:

:process

Value must be a live process object or a function of no arguments producing one such object. If passed a process object, the object is expected to contain a pre-established connection; otherwise, the function is called immediately after the object is made.

:on-shutdown

Value must be a function of a single argument, the jsonrpc-process-connection object. The function is called after the underlying process object has been deleted (either deliberately by jsonrpc-shutdown, or unexpectedly, because of some external cause).


33.32.3 JSONRPC JSON object format

JSONRPC JSON objects are exchanged as Lisp plists (see Property Lists): JSON-compatible plists are handed to the dispatcher functions and, likewise, JSON-compatible plists should be given to jsonrpc-notify, jsonrpc-request, and jsonrpc-async-request.

To facilitate handling plists, this library makes liberal use of cl-lib library (see cl-lib in Common Lisp Extensions for GNU Emacs Lisp) and suggests (but doesn’t force) its clients to do the same. A macro jsonrpc-lambda can be used to create a lambda for destructuring a JSON-object like in this example:

(jsonrpc-async-request
 myproc :frobnicate `(:foo "trix")
 :success-fn (jsonrpc-lambda (&key bar baz &allow-other-keys)
               (message "Server replied back with %s and %s!"
                        bar baz))
 :error-fn (jsonrpc-lambda (&key code message _data)
             (message "Sadly, server reports %s: %s"
                      code message)))

33.32.4 Deferred JSONRPC requests

In many RPC situations, synchronization between the two communicating endpoints is a matter of correctly designing the RPC application: when synchronization is needed, requests (which are blocking) should be used; when it isn’t, notifications should suffice. However, when Emacs acts as one of these endpoints, asynchronous events (e.g. timer- or process-related) may be triggered while there is still uncertainty about the state of the remote endpoint. Furthermore, acting on these events may only sometimes demand synchronization, depending on the event’s specific nature.

The :deferred keyword argument to jsonrpc-request and jsonrpc-async-request is designed to let the caller indicate that the specific request needs synchronization and its actual issuance may be delayed to the future, until some condition is satisfied. Specifying :deferred for a request doesn’t mean it will be delayed, only that it can be. If the request isn’t sent immediately, jsonrpc will make renewed efforts to send it at certain key times during communication, such as when receiving or sending other messages to the endpoint.

Before any attempt to send the request, the application-specific conditions are checked. Since the jsonrpc library can’t know what these conditions are, the program can use the jsonrpc-connection-ready-p generic function (see Generic Functions) to specify them. The default method for this function returns t, but you can add overriding methods that return nil in some situations, based on the arguments passed to it, which are the jsonrpc-connection object (see Overview) and whichever value you passed as the :deferred keyword argument.


33.33 Atomic Change Groups

In database terminology, an atomic change is an indivisible change—it can succeed entirely or it can fail entirely, but it cannot partly succeed. A Lisp program can make a series of changes to one or several buffers as an atomic change group, meaning that either the entire series of changes will be installed in their buffers or, in case of an error, none of them will be.

To do this for one buffer, the one already current, simply write a call to atomic-change-group around the code that makes the changes, like this:

(atomic-change-group
  (insert foo)
  (delete-region x y))

If an error (or other nonlocal exit) occurs inside the body of atomic-change-group, it unmakes all the changes in that buffer that were during the execution of the body. This kind of change group has no effect on any other buffers—any such changes remain.

If you need something more sophisticated, such as to make changes in various buffers constitute one atomic group, you must directly call lower-level functions that atomic-change-group uses.

Function: prepare-change-group &optional buffer

This function sets up a change group for buffer buffer, which defaults to the current buffer. It returns a handle that represents the change group. You must use this handle to activate the change group and subsequently to finish it.

To use the change group, you must activate it. You must do this before making any changes in the text of buffer.

Function: activate-change-group handle

This function activates the change group that handle designates.

After you activate the change group, any changes you make in that buffer become part of it. Once you have made all the desired changes in the buffer, you must finish the change group. There are two ways to do this: you can either accept (and finalize) all the changes, or cancel them all.

Function: accept-change-group handle

This function accepts all the changes in the change group specified by handle, making them final.

Function: cancel-change-group handle

This function cancels and undoes all the changes in the change group specified by handle.

You can cause some or all of the changes in a change group to be considered as a single unit for the purposes of the undo commands (see Undo) by using undo-amalgamate-change-group.

Function: undo-amalgamate-change-group

Amalgamate all the changes made in the change-group since the state identified by handle. This function removes all undo boundaries between undo records of changes since the state described by handle. Usually, handle is the handle returned by prepare-change-group, in which case all the changes since the beginning of the change-group are amalgamated into a single undo unit.

Your code should use unwind-protect to make sure the group is always finished. The call to activate-change-group should be inside the unwind-protect, in case the user types C-g just after it runs. (This is one reason why prepare-change-group and activate-change-group are separate functions, because normally you would call prepare-change-group before the start of that unwind-protect.) Once you finish the group, don’t use the handle again—in particular, don’t try to finish the same group twice.

To make a multibuffer change group, call prepare-change-group once for each buffer you want to cover, then use nconc to combine the returned values, like this:

(nconc (prepare-change-group buffer-1)
       (prepare-change-group buffer-2))

You can then activate the multibuffer change group with a single call to activate-change-group, and finish it with a single call to accept-change-group or cancel-change-group.

Nested use of several change groups for the same buffer works as you would expect. Non-nested use of change groups for the same buffer will get Emacs confused, so don’t let it happen; the first change group you start for any given buffer should be the last one finished.


33.34 Change Hooks

These hook variables let you arrange to take notice of changes in buffers (or in a particular buffer, if you make them buffer-local). See also Properties with Special Meanings, for how to detect changes to specific parts of the text.

The functions you use in these hooks should save and restore the match data if they do anything that uses regular expressions; otherwise, they will interfere in bizarre ways with the editing operations that call them.

Variable: before-change-functions

This variable holds a list of functions to call when Emacs is about to modify a buffer. Each function gets two arguments, the beginning and end of the region that is about to change, represented as integers. The buffer that is about to change is always the current buffer when the function is called.

Variable: after-change-functions

This variable holds a list of functions to call after Emacs modifies a buffer. Each function receives three arguments: the beginning and end of the region just changed, and the length of the text that existed before the change. All three arguments are integers. The buffer that has been changed is always the current buffer when the function is called.

The length of the old text is the difference between the buffer positions before and after that text as it was before the change. As for the changed text, its length is simply the difference between the first two arguments.

Output of messages into the *Messages* buffer does not call these functions, and neither do certain internal buffer changes, such as changes in buffers created by Emacs internally for certain jobs, that should not be visible to Lisp programs.

The vast majority of buffer changing primitives will call before-change-functions and after-change-functions in balanced pairs, once for each change, where the arguments to these hooks exactly delimit the change being made. Yet, hook functions should not rely on this always being the case, because some complex primitives call before-change-functions once before making changes, and then call after-change-functions zero or more times, depending on how many individual changes the primitive is making. When that happens, the arguments to before-change-functions will enclose a region in which the individual changes are made, but won’t necessarily be the minimal such region, and the arguments to each successive call of after-change-functions will then delimit the part of text being changed exactly. In general, we advise using either the before- or the after-change hook, but not both.

Macro: combine-after-change-calls body…

The macro executes body normally, but arranges to call the after-change functions just once for a series of several changes—if that seems safe.

If a program makes several text changes in the same area of the buffer, using the macro combine-after-change-calls around that part of the program can make it run considerably faster when after-change hooks are in use. When the after-change hooks are ultimately called, the arguments specify a portion of the buffer including all of the changes made within the combine-after-change-calls body.

Warning: You must not alter the values of after-change-functions within the body of a combine-after-change-calls form.

Warning: If the changes you combine occur in widely scattered parts of the buffer, this will still work, but it is not advisable, because it may lead to inefficient behavior for some change hook functions.

Macro: combine-change-calls beg end body…

This executes body normally, except any buffer changes it makes do not trigger the calls to before-change-functions and after-change-functions. Instead there is a single call of each of these hooks for the region enclosed by beg and end, the parameters supplied to after-change-functions reflecting the changes made to the size of the region by body.

The result of this macro is the result returned by body.

This macro is useful when a function makes a possibly large number of repetitive changes to the buffer, and the change hooks would otherwise take a long time to run, were they to be run for each individual buffer modification. Emacs itself uses this macro, for example, in the commands comment-region and uncomment-region.

Warning: You must not alter the values of before-change-functions or after-change-function within body.

Warning: You must not make any buffer changes outside of the region specified by beg and end.

Variable: first-change-hook

This variable is a normal hook that is run whenever a buffer is changed that was previously in the unmodified state.

Variable: inhibit-modification-hooks

If this variable is non-nil, all of the change hooks are disabled; none of them run. This affects all the hook variables described above in this section, as well as the hooks attached to certain special text properties (see Properties with Special Meanings) and overlay properties (see Overlay Properties).

Also, this variable is bound to non-nil while running those same hook variables, so that by default modifying the buffer from a modification hook does not cause other modification hooks to be run. If you do want modification hooks to be run in a particular piece of code that is itself run from a modification hook, then rebind locally inhibit-modification-hooks to nil. However, doing this may cause recursive calls to the modification hooks, so be sure to prepare for that (for example, by binding some variable which tells your hook to do nothing).

We recommend that you only bind this variable for modifications that do not result in lasting changes to buffer text contents (for example face changes or temporary modifications). If you need to delay change hooks during a series of changes (typically for performance reasons), use combine-change-calls or combine-after-change-calls instead.


34 Non-ASCII Characters

This chapter covers the special issues relating to characters and how they are stored in strings and buffers.


34.1 Text Representations

Emacs buffers and strings support a large repertoire of characters from many different scripts, allowing users to type and display text in almost any known written language.

To support this multitude of characters and scripts, Emacs closely follows the Unicode Standard. The Unicode Standard assigns a unique number, called a codepoint, to each and every character. The range of codepoints defined by Unicode, or the Unicode codespace, is 0..#x10FFFF (in hexadecimal notation), inclusive. Emacs extends this range with codepoints in the range #x110000..#x3FFFFF, which it uses for representing characters that are not unified with Unicode and raw 8-bit bytes that cannot be interpreted as characters. Thus, a character codepoint in Emacs is a 22-bit integer.

To conserve memory, Emacs does not hold fixed-length 22-bit numbers that are codepoints of text characters within buffers and strings. Rather, Emacs uses a variable-length internal representation of characters, that stores each character as a sequence of 1 to 5 8-bit bytes, depending on the magnitude of its codepoint24. For example, any ASCII character takes up only 1 byte, a Latin-1 character takes up 2 bytes, etc. We call this representation of text multibyte.

Outside Emacs, characters can be represented in many different encodings, such as ISO-8859-1, GB-2312, Big-5, etc. Emacs converts between these external encodings and its internal representation, as appropriate, when it reads text into a buffer or a string, or when it writes text to a disk file or passes it to some other process.

Occasionally, Emacs needs to hold and manipulate encoded text or binary non-text data in its buffers or strings. For example, when Emacs visits a file, it first reads the file’s text verbatim into a buffer, and only then converts it to the internal representation. Before the conversion, the buffer holds encoded text.

Encoded text is not really text, as far as Emacs is concerned, but rather a sequence of raw 8-bit bytes. We call buffers and strings that hold encoded text unibyte buffers and strings, because Emacs treats them as a sequence of individual bytes. Usually, Emacs displays unibyte buffers and strings as octal codes such as \237. We recommend that you never use unibyte buffers and strings except for manipulating encoded text or binary non-text data.

In a buffer, the buffer-local value of the variable enable-multibyte-characters specifies the representation used. The representation for a string is determined and recorded in the string when the string is constructed.

Variable: enable-multibyte-characters

This variable specifies the current buffer’s text representation. If it is non-nil, the buffer contains multibyte text; otherwise, it contains unibyte encoded text or binary non-text data.

You cannot set this variable directly; instead, use the function set-buffer-multibyte to change a buffer’s representation.

Function: position-bytes position

Buffer positions are measured in character units. This function returns the byte-position corresponding to buffer position position in the current buffer. This is 1 at the start of the buffer, and counts upward in bytes. If position is out of range, the value is nil.

Function: byte-to-position byte-position

Return the buffer position, in character units, corresponding to given byte-position in the current buffer. If byte-position is out of range, the value is nil. In a multibyte buffer, an arbitrary value of byte-position can be not at character boundary, but inside a multibyte sequence representing a single character; in this case, this function returns the buffer position of the character whose multibyte sequence includes byte-position. In other words, the value does not change for all byte positions that belong to the same character.

The following two functions are useful when a Lisp program needs to map buffer positions to byte offsets in a file visited by the buffer.

Function: bufferpos-to-filepos position &optional quality coding-system

This function is similar to position-bytes, but instead of byte position in the current buffer it returns the offset from the beginning of the current buffer’s file of the byte that corresponds to the given character position in the buffer. The conversion requires to know how the text is encoded in the buffer’s file; this is what the coding-system argument is for, defaulting to the value of buffer-file-coding-system. The optional argument quality specifies how accurate the result should be; it should be one of the following:

exact

The result must be accurate. The function may need to encode and decode a large part of the buffer, which is expensive and can be slow.

approximate

The value can be an approximation. The function may avoid expensive processing and return an inexact result.

nil

If the exact result needs expensive processing, the function will return nil rather than an approximation. This is the default if the argument is omitted.

Function: filepos-to-bufferpos byte &optional quality coding-system

This function returns the buffer position corresponding to a file position specified by byte, a zero-base byte offset from the file’s beginning. The function performs the conversion opposite to what bufferpos-to-filepos does. Optional arguments quality and coding-system have the same meaning and values as for bufferpos-to-filepos.

Function: multibyte-string-p string

Return t if string is a multibyte string, nil otherwise. This function also returns nil if string is some object other than a string.

Function: string-bytes string

This function returns the number of bytes in string. If string is a multibyte string, this can be greater than (length string).

Function: unibyte-string &rest bytes

This function concatenates all its argument bytes and makes the result a unibyte string.


34.2 Disabling Multibyte Characters

By default, Emacs starts in multibyte mode: it stores the contents of buffers and strings using an internal encoding that represents non-ASCII characters using multi-byte sequences. Multibyte mode allows you to use all the supported languages and scripts without limitations.

Under very special circumstances, you may want to disable multibyte character support, for a specific buffer. When multibyte characters are disabled in a buffer, we call that unibyte mode. In unibyte mode, each character in the buffer has a character code ranging from 0 through 255 (0377 octal); 0 through 127 (0177 octal) represent ASCII characters, and 128 (0200 octal) through 255 (0377 octal) represent non-ASCII characters.

To edit a particular file in unibyte representation, visit it using find-file-literally. See Functions for Visiting Files. You can convert a multibyte buffer to unibyte by saving it to a file, killing the buffer, and visiting the file again with find-file-literally. Alternatively, you can use C-x RET c (universal-coding-system-argument) and specify ‘raw-text’ as the coding system with which to visit or save a file. See Specifying a Coding System for File Text in GNU Emacs Manual. Unlike find-file-literally, finding a file as ‘raw-text’ doesn’t disable format conversion, uncompression, or auto mode selection.

The buffer-local variable enable-multibyte-characters is non-nil in multibyte buffers, and nil in unibyte ones. The mode line also indicates whether a buffer is multibyte or not. With a graphical display, in a multibyte buffer, the portion of the mode line that indicates the character set has a tooltip that (amongst other things) says that the buffer is multibyte. In a unibyte buffer, the character set indicator is absent. Thus, in a unibyte buffer (when using a graphical display) there is normally nothing before the indication of the visited file’s end-of-line convention (colon, backslash, etc.), unless you are using an input method.

You can turn off multibyte support in a specific buffer by invoking the command toggle-enable-multibyte-characters in that buffer.


34.3 Converting Text Representations

Emacs can convert unibyte text to multibyte; it can also convert multibyte text to unibyte, provided that the multibyte text contains only ASCII and 8-bit raw bytes. In general, these conversions happen when inserting text into a buffer, or when putting text from several strings together in one string. You can also explicitly convert a string’s contents to either representation.

Emacs chooses the representation for a string based on the text from which it is constructed. The general rule is to convert unibyte text to multibyte text when combining it with other multibyte text, because the multibyte representation is more general and can hold whatever characters the unibyte text has.

When inserting text into a buffer, Emacs converts the text to the buffer’s representation, as specified by enable-multibyte-characters in that buffer. In particular, when you insert multibyte text into a unibyte buffer, Emacs converts the text to unibyte, even though this conversion cannot in general preserve all the characters that might be in the multibyte text. The other natural alternative, to convert the buffer contents to multibyte, is not acceptable because the buffer’s representation is a choice made by the user that cannot be overridden automatically.

Converting unibyte text to multibyte text leaves ASCII characters unchanged, and converts bytes with codes 128 through 255 to the multibyte representation of raw eight-bit bytes.

Converting multibyte text to unibyte converts all ASCII and eight-bit characters to their single-byte form, but loses information for non-ASCII characters by discarding all but the low 8 bits of each character’s codepoint. Converting unibyte text to multibyte and back to unibyte reproduces the original unibyte text.

The next two functions either return the argument string, or a newly created string with no text properties.

Function: string-to-multibyte string

This function returns a multibyte string containing the same sequence of characters as string. If string is a multibyte string, it is returned unchanged. The function assumes that string includes only ASCII characters and raw 8-bit bytes; the latter are converted to their multibyte representation corresponding to the codepoints #x3FFF80 through #x3FFFFF, inclusive (see codepoints).

Function: string-to-unibyte string

This function returns a unibyte string containing the same sequence of characters as string. If string is a unibyte string, it is returned unchanged. Otherwise, ASCII characters and characters in the eight-bit charset are converted to their corresponding byte values. Use this function for string arguments that contain only ASCII and eight-bit characters; the function signals an error if any other characters are encountered.

Function: byte-to-string byte

This function returns a unibyte string containing a single byte of character data, byte. It signals an error if byte is not an integer between 0 and 255.

Function: multibyte-char-to-unibyte char

This converts the multibyte character char to a unibyte character, and returns that character. If char is neither ASCII nor eight-bit, the function returns −1.

Function: unibyte-char-to-multibyte char

This converts the unibyte character char to a multibyte character, assuming char is either ASCII or raw 8-bit byte.


34.4 Selecting a Representation

Sometimes it is useful to examine an existing buffer or string as multibyte when it was unibyte, or vice versa.

Function: set-buffer-multibyte multibyte

Set the representation type of the current buffer. If multibyte is non-nil, the buffer becomes multibyte. If multibyte is nil, the buffer becomes unibyte.

This function leaves the buffer contents unchanged when viewed as a sequence of bytes. As a consequence, it can change the contents viewed as characters; for instance, a sequence of three bytes which is treated as one character in multibyte representation will count as three characters in unibyte representation. Eight-bit characters representing raw bytes are an exception. They are represented by one byte in a unibyte buffer, but when the buffer is set to multibyte, they are converted to two-byte sequences, and vice versa.

This function sets enable-multibyte-characters to record which representation is in use. It also adjusts various data in the buffer (including overlays, text properties and markers) so that they cover the same text as they did before.

This function signals an error if the buffer is narrowed, since the narrowing might have occurred in the middle of multibyte character sequences.

This function also signals an error if the buffer is an indirect buffer. An indirect buffer always inherits the representation of its base buffer.

Function: string-as-unibyte string

If string is already a unibyte string, this function returns string itself. Otherwise, it returns a new string with the same bytes as string, but treating each byte as a separate character (so that the value may have more characters than string); as an exception, each eight-bit character representing a raw byte is converted into a single byte. The newly-created string contains no text properties.

Function: string-as-multibyte string

If string is a multibyte string, this function returns string itself. Otherwise, it returns a new string with the same bytes as string, but treating each multibyte sequence as one character. This means that the value may have fewer characters than string has. If a byte sequence in string is invalid as a multibyte representation of a single character, each byte in the sequence is treated as a raw 8-bit byte. The newly-created string contains no text properties.


34.5 Character Codes

The unibyte and multibyte text representations use different character codes. The valid character codes for unibyte representation range from 0 to #xFF (255)—the values that can fit in one byte. The valid character codes for multibyte representation range from 0 to #x3FFFFF. In this code space, values 0 through #x7F (127) are for ASCII characters, and values #x80 (128) through #x3FFF7F (4194175) are for non-ASCII characters.

Emacs character codes are a superset of the Unicode standard. Values 0 through #x10FFFF (1114111) correspond to Unicode characters of the same codepoint; values #x110000 (1114112) through #x3FFF7F (4194175) represent characters that are not unified with Unicode; and values #x3FFF80 (4194176) through #x3FFFFF (4194303) represent eight-bit raw bytes.

Function: characterp charcode

This returns t if charcode is a valid character, and nil otherwise.

(characterp 65)
     ⇒ t
(characterp 4194303)
     ⇒ t
(characterp 4194304)
     ⇒ nil
Function: max-char &optional unicode

This function returns the largest value that a valid character codepoint can have in Emacs. If the optional argument unicode is non-nil, it returns the largest character codepoint defined by the Unicode Standard (which is smaller than the maximum codepoint supported by Emacs).

(characterp (max-char))
     ⇒ t
(characterp (1+ (max-char)))
     ⇒ nil
Function: char-from-name string &optional ignore-case

This function returns the character whose Unicode name is string. If ignore-case is non-nil, case is ignored in string. This function returns nil if string does not name a character.

;; U+03A3
(= (char-from-name "GREEK CAPITAL LETTER SIGMA") #x03A3)
     ⇒ t
Function: get-byte &optional pos string

This function returns the byte at character position pos in the current buffer. If the current buffer is unibyte, this is literally the byte at that position. If the buffer is multibyte, byte values of ASCII characters are the same as character codepoints, whereas eight-bit raw bytes are converted to their 8-bit codes. The function signals an error if the character at pos is non-ASCII.

The optional argument string means to get a byte value from that string instead of the current buffer.


34.6 Character Properties

A character property is a named attribute of a character that specifies how the character behaves and how it should be handled during text processing and display. Thus, character properties are an important part of specifying the character’s semantics.

On the whole, Emacs follows the Unicode Standard in its implementation of character properties. In particular, Emacs supports the Unicode Character Property Model, and the Emacs character property database is derived from the Unicode Character Database (UCD). See the Character Properties chapter of the Unicode Standard, for a detailed description of Unicode character properties and their meaning. This section assumes you are already familiar with that chapter of the Unicode Standard, and want to apply that knowledge to Emacs Lisp programs.

In Emacs, each property has a name, which is a symbol, and a set of possible values, whose types depend on the property; if a character does not have a certain property, the value is nil. As a general rule, the names of character properties in Emacs are produced from the corresponding Unicode properties by downcasing them and replacing each ‘_’ character with a dash ‘-’. For example, Canonical_Combining_Class becomes canonical-combining-class. However, sometimes we shorten the names to make their use easier.

Some codepoints are left unassigned by the UCD—they don’t correspond to any character. The Unicode Standard defines default values of properties for such codepoints; they are mentioned below for each property.

Here is the full list of value types for all the character properties that Emacs knows about:

name

Corresponds to the Name Unicode property. The value is a string consisting of upper-case Latin letters A to Z, digits, spaces, and hyphen ‘-’ characters. For unassigned codepoints, the value is nil.

general-category

Corresponds to the General_Category Unicode property. The value is a symbol whose name is a 2-letter abbreviation of the character’s classification. For unassigned codepoints, the value is Cn.

canonical-combining-class

Corresponds to the Canonical_Combining_Class Unicode property. The value is an integer. For unassigned codepoints, the value is zero.

bidi-class

Corresponds to the Unicode Bidi_Class property. The value is a symbol whose name is the Unicode directional type of the character. Emacs uses this property when it reorders bidirectional text for display (see Bidirectional Display). For unassigned codepoints, the value depends on the code blocks to which the codepoint belongs: most unassigned codepoints get the value of L (strong L), but some get values of AL (Arabic letter) or R (strong R).

decomposition

Corresponds to the Unicode properties Decomposition_Type and Decomposition_Value. The value is a list, whose first element may be a symbol representing a compatibility formatting tag, such as small25; the other elements are characters that give the compatibility decomposition sequence of this character. For characters that don’t have decomposition sequences, and for unassigned codepoints, the value is a list with a single member, the character itself.

decimal-digit-value

Corresponds to the Unicode Numeric_Value property for characters whose Numeric_Type is ‘Decimal’. The value is an integer, or nil if the character has no decimal digit value. For unassigned codepoints, the value is nil, which means NaN, or “not a number”.

digit-value

Corresponds to the Unicode Numeric_Value property for characters whose Numeric_Type is ‘Digit’. The value is an integer. Examples of such characters include compatibility subscript and superscript digits, for which the value is the corresponding number. For characters that don’t have any numeric value, and for unassigned codepoints, the value is nil, which means NaN.

numeric-value

Corresponds to the Unicode Numeric_Value property for characters whose Numeric_Type is ‘Numeric’. The value of this property is a number. Examples of characters that have this property include fractions, subscripts, superscripts, Roman numerals, currency numerators, and encircled numbers. For example, the value of this property for the character U+2155 VULGAR FRACTION ONE FIFTH is 0.2. For characters that don’t have any numeric value, and for unassigned codepoints, the value is nil, which means NaN.

mirrored

Corresponds to the Unicode Bidi_Mirrored property. The value of this property is a symbol, either Y or N. For unassigned codepoints, the value is N.

mirroring

Corresponds to the Unicode Bidi_Mirroring_Glyph property. The value of this property is a character whose glyph represents the mirror image of the character’s glyph, or nil if there’s no defined mirroring glyph. All the characters whose mirrored property is N have nil as their mirroring property; however, some characters whose mirrored property is Y also have nil for mirroring, because no appropriate characters exist with mirrored glyphs. Emacs uses this property to display mirror images of characters when appropriate (see Bidirectional Display). For unassigned codepoints, the value is nil.

paired-bracket

Corresponds to the Unicode Bidi_Paired_Bracket property. The value of this property is the codepoint of a character’s paired bracket, or nil if the character is not a bracket character. This establishes a mapping between characters that are treated as bracket pairs by the Unicode Bidirectional Algorithm; Emacs uses this property when it decides how to reorder for display parentheses, braces, and other similar characters (see Bidirectional Display).

bracket-type

Corresponds to the Unicode Bidi_Paired_Bracket_Type property. For characters whose paired-bracket property is non-nil, the value of this property is a symbol, either o (for opening bracket characters) or c (for closing bracket characters). For characters whose paired-bracket property is nil, the value is the symbol n (None). Like paired-bracket, this property is used for bidirectional display.

old-name

Corresponds to the Unicode Unicode_1_Name property. The value is a string. For unassigned codepoints, and characters that have no value for this property, the value is nil.

iso-10646-comment

Corresponds to the Unicode ISO_Comment property. The value is either a string or nil. For unassigned codepoints, the value is nil.

uppercase

Corresponds to the Unicode Simple_Uppercase_Mapping property. The value of this property is a single character. For unassigned codepoints, the value is nil, which means the character itself.

lowercase

Corresponds to the Unicode Simple_Lowercase_Mapping property. The value of this property is a single character. For unassigned codepoints, the value is nil, which means the character itself.

titlecase

Corresponds to the Unicode Simple_Titlecase_Mapping property. Title case is a special form of a character used when the first character of a word needs to be capitalized. The value of this property is a single character. For unassigned codepoints, the value is nil, which means the character itself.

special-uppercase

Corresponds to Unicode language- and context-independent special upper-casing rules. The value of this property is a string (which may be empty). For example mapping for U+00DF LATIN SMALL LETTER SHARP S is "SS". For characters with no special mapping, the value is nil which means uppercase property needs to be consulted instead.

special-lowercase

Corresponds to Unicode language- and context-independent special lower-casing rules. The value of this property is a string (which may be empty). For example mapping for U+0130 LATIN CAPITAL LETTER I WITH DOT ABOVE the value is "i\u0307" (i.e. 2-character string consisting of LATIN SMALL LETTER I followed by U+0307 COMBINING DOT ABOVE). For characters with no special mapping, the value is nil which means lowercase property needs to be consulted instead.

special-titlecase

Corresponds to Unicode unconditional special title-casing rules. The value of this property is a string (which may be empty). For example mapping for U+FB01 LATIN SMALL LIGATURE FI the value is "Fi". For characters with no special mapping, the value is nil which means titlecase property needs to be consulted instead.

Function: get-char-code-property char propname

This function returns the value of char’s propname property.

(get-char-code-property ?\s 'general-category)
     ⇒ Zs
(get-char-code-property ?1 'general-category)
     ⇒ Nd
;; U+2084
(get-char-code-property ?\N{SUBSCRIPT FOUR}
                        'digit-value)
     ⇒ 4
;; U+2155
(get-char-code-property ?\N{VULGAR FRACTION ONE FIFTH}
                        'numeric-value)
     ⇒ 0.2
;; U+2163
(get-char-code-property ?\N{ROMAN NUMERAL FOUR}
                        'numeric-value)
     ⇒ 4
(get-char-code-property ?\( 'paired-bracket)
     ⇒ 41  ; closing parenthesis
(get-char-code-property ?\) 'bracket-type)
     ⇒ c
Function: char-code-property-description prop value

This function returns the description string of property prop’s value, or nil if value has no description.

(char-code-property-description 'general-category 'Zs)
     ⇒ "Separator, Space"
(char-code-property-description 'general-category 'Nd)
     ⇒ "Number, Decimal Digit"
(char-code-property-description 'numeric-value '1/5)
     ⇒ nil
Function: put-char-code-property char propname value

This function stores value as the value of the property propname for the character char.

Variable: unicode-category-table

The value of this variable is a char-table (see Char-Tables) that specifies, for each character, its Unicode General_Category property as a symbol.

Variable: char-script-table

The value of this variable is a char-table that specifies, for each character, a symbol whose name is the script to which the character belongs, according to the Unicode Standard classification of the Unicode code space into script-specific blocks. This char-table has a single extra slot whose value is the list of all script symbols. Note that Emacs’ classification of characters into scripts is not a 1-for-1 reflection of the Unicode standard, e.g. there is no ‘symbol’ script in Unicode.

Variable: char-width-table

The value of this variable is a char-table that specifies the width of each character in columns that it will occupy on the screen.

Variable: printable-chars

The value of this variable is a char-table that specifies, for each character, whether it is printable or not. That is, if evaluating (aref printable-chars char) results in t, the character is printable, and if it results in nil, it is not.


34.7 Character Sets

An Emacs character set, or charset, is a set of characters in which each character is assigned a numeric code point. (The Unicode Standard calls this a coded character set.) Each Emacs charset has a name which is a symbol. A single character can belong to any number of different character sets, but it will generally have a different code point in each charset. Examples of character sets include ascii, iso-8859-1, greek-iso8859-7, and windows-1255. The code point assigned to a character in a charset is usually different from its code point used in Emacs buffers and strings.

Emacs defines several special character sets. The character set unicode includes all the characters whose Emacs code points are in the range 0..#x10FFFF. The character set emacs includes all ASCII and non-ASCII characters. Finally, the eight-bit charset includes the 8-bit raw bytes; Emacs uses it to represent raw bytes encountered in text.

Function: charsetp object

Returns t if object is a symbol that names a character set, nil otherwise.

Variable: charset-list

The value is a list of all defined character set names.

Function: charset-priority-list &optional highestp

This function returns a list of all defined character sets ordered by their priority. If highestp is non-nil, the function returns a single character set of the highest priority.

Function: set-charset-priority &rest charsets

This function makes charsets the highest priority character sets.

Function: char-charset character &optional restriction

This function returns the name of the character set of highest priority that character belongs to. ASCII characters are an exception: for them, this function always returns ascii.

If restriction is non-nil, it should be a list of charsets to search. Alternatively, it can be a coding system, in which case the returned charset must be supported by that coding system (see Coding Systems).

Function: charset-plist charset

This function returns the property list of the character set charset. Although charset is a symbol, this is not the same as the property list of that symbol. Charset properties include important information about the charset, such as its documentation string, short name, etc.

Function: put-charset-property charset propname value

This function sets the propname property of charset to the given value.

Function: get-charset-property charset propname

This function returns the value of charsets property propname.

Command: list-charset-chars charset

This command displays a list of characters in the character set charset.

Emacs can convert between its internal representation of a character and the character’s codepoint in a specific charset. The following two functions support these conversions.

Function: decode-char charset code-point

This function decodes a character that is assigned a code-point in charset, to the corresponding Emacs character, and returns it. If charset doesn’t contain a character of that code point, the value is nil.

For backward compatibility, if code-point doesn’t fit in a Lisp fixnum (see most-positive-fixnum), it can be specified as a cons cell (high . low), where low are the lower 16 bits of the value and high are the high 16 bits. This usage is obsolescent.

Function: encode-char char charset

This function returns the code point assigned to the character char in charset. If charset doesn’t have a codepoint for char, the value is nil.

The following function comes in handy for applying a certain function to all or part of the characters in a charset:

Function: map-charset-chars function charset &optional arg from-code to-code

Call function for characters in charset. function is called with two arguments. The first one is a cons cell (from . to), where from and to indicate a range of characters contained in charset. The second argument passed to function is arg, or nil if arg is omitted.

By default, the range of codepoints passed to function includes all the characters in charset, but optional arguments from-code and to-code limit that to the range of characters between these two codepoints of charset. If either of them is nil, it defaults to the first or last codepoint of charset, respectively. Note that from-code and to-code are charset’s codepoints, not the Emacs codes of characters; by contrast, the values from and to in the cons cell passed to function are Emacs character codes. Those Emacs character codes are either Unicode code points, or Emacs internal code points that extend Unicode and are beyond the Unicode range of characters 0..#x10FFFF (see Text Representations). The latter happens rarely, with legacy CJK charsets for codepoints of charset which specify characters not yet unified with Unicode.


34.8 Scanning for Character Sets

Sometimes it is useful to find out which character set a particular character belongs to. One use for this is in determining which coding systems (see Coding Systems) are capable of representing all of the text in question; another is to determine the font(s) for displaying that text.

Function: charset-after &optional pos

This function returns the charset of highest priority containing the character at position pos in the current buffer. If pos is omitted or nil, it defaults to the current value of point. If pos is out of range, the value is nil.

Function: find-charset-region beg end &optional translation

This function returns a list of the character sets of highest priority that contain characters in the current buffer between positions beg and end.

The optional argument translation specifies a translation table to use for scanning the text (see Translation of Characters). If it is non-nil, then each character in the region is translated through this table, and the value returned describes the translated characters instead of the characters actually in the buffer.

Function: find-charset-string string &optional translation

This function returns a list of character sets of highest priority that contain characters in string. It is just like find-charset-region, except that it applies to the contents of string instead of part of the current buffer.


34.9 Translation of Characters

A translation table is a char-table (see Char-Tables) that specifies a mapping of characters into characters. These tables are used in encoding and decoding, and for other purposes. Some coding systems specify their own particular translation tables; there are also default translation tables which apply to all other coding systems.

A translation table has two extra slots. The first is either nil or a translation table that performs the reverse translation; the second is the maximum number of characters to look up for translating sequences of characters (see the description of make-translation-table-from-alist below).

Function: make-translation-table &rest translations

This function returns a translation table based on the argument translations. Each element of translations should be a list of elements of the form (from . to); this says to translate the character from into to.

The arguments and the forms in each argument are processed in order, and if a previous form already translates to to some other character, say to-alt, from is also translated to to-alt.

During decoding, the translation table’s translations are applied to the characters that result from ordinary decoding. If a coding system has the property :decode-translation-table, that specifies the translation table to use, or a list of translation tables to apply in sequence. (This is a property of the coding system, as returned by coding-system-get, not a property of the symbol that is the coding system’s name. See Basic Concepts of Coding Systems.) Finally, if standard-translation-table-for-decode is non-nil, the resulting characters are translated by that table.

During encoding, the translation table’s translations are applied to the characters in the buffer, and the result of translation is actually encoded. If a coding system has property :encode-translation-table, that specifies the translation table to use, or a list of translation tables to apply in sequence. In addition, if the variable standard-translation-table-for-encode is non-nil, it specifies the translation table to use for translating the result.

Variable: standard-translation-table-for-decode

This is the default translation table for decoding. If a coding system specifies its own translation tables, the table that is the value of this variable, if non-nil, is applied after them.

Variable: standard-translation-table-for-encode

This is the default translation table for encoding. If a coding system specifies its own translation tables, the table that is the value of this variable, if non-nil, is applied after them.

Variable: translation-table-for-input

Self-inserting characters are translated through this translation table before they are inserted. Search commands also translate their input through this table, so they can compare more reliably with what’s in the buffer.

This variable automatically becomes buffer-local when set.

Function: make-translation-table-from-vector vec

This function returns a translation table made from vec that is an array of 256 elements to map bytes (values 0 through #xFF) to characters. Elements may be nil for untranslated bytes. The returned table has a translation table for reverse mapping in the first extra slot, and the value 1 in the second extra slot.

This function provides an easy way to make a private coding system that maps each byte to a specific character. You can specify the returned table and the reverse translation table using the properties :decode-translation-table and :encode-translation-table respectively in the props argument to define-coding-system.

Function: make-translation-table-from-alist alist

This function is similar to make-translation-table but returns a complex translation table rather than a simple one-to-one mapping. Each element of alist is of the form (from . to), where from and to are either characters or vectors specifying a sequence of characters. If from is a character, that character is translated to to (i.e., to a character or a character sequence). If from is a vector of characters, that sequence is translated to to. The returned table has a translation table for reverse mapping in the first extra slot, and the maximum length of all the from character sequences in the second extra slot.


34.10 Coding Systems

When Emacs reads or writes a file, and when Emacs sends text to a subprocess or receives text from a subprocess, it normally performs character code conversion and end-of-line conversion as specified by a particular coding system.

How to define a coding system is an arcane matter, and is not documented here.


34.10.1 Basic Concepts of Coding Systems

Character code conversion involves conversion between the internal representation of characters used inside Emacs and some other encoding. Emacs supports many different encodings, in that it can convert to and from them. For example, it can convert text to or from encodings such as Latin 1, Latin 2, Latin 3, Latin 4, Latin 5, and several variants of ISO 2022. In some cases, Emacs supports several alternative encodings for the same characters; for example, there are three coding systems for the Cyrillic (Russian) alphabet: ISO, Alternativnyj, and KOI8.

Every coding system specifies a particular set of character code conversions, but the coding system undecided is special: it leaves the choice unspecified, to be chosen heuristically for each file or string, based on the file’s or string’s data, when they are decoded or encoded. The coding system prefer-utf-8 is like undecided, but it prefers to choose utf-8 when possible.

In general, a coding system doesn’t guarantee roundtrip identity: decoding a byte sequence using a coding system, then encoding the resulting text in the same coding system, can produce a different byte sequence. But some coding systems do guarantee that the byte sequence will be the same as what you originally decoded. Here are a few examples:

iso-8859-1, utf-8, big5, shift_jis, euc-jp

Encoding buffer text and then decoding the result can also fail to reproduce the original text. For instance, if you encode a character with a coding system which does not support that character, the result is unpredictable, and thus decoding it using the same coding system may produce a different text. Currently, Emacs can’t report errors that result from encoding unsupported characters.

End of line conversion handles three different conventions used on various systems for representing end of line in files. The Unix convention, used on GNU and Unix systems, is to use the linefeed character (also called newline). The DOS convention, used on MS-Windows and MS-DOS systems, is to use a carriage return and a linefeed at the end of a line. The Mac convention is to use just carriage return. (This was the convention used in Classic Mac OS.)

Base coding systems such as latin-1 leave the end-of-line conversion unspecified, to be chosen based on the data. Variant coding systems such as latin-1-unix, latin-1-dos and latin-1-mac specify the end-of-line conversion explicitly as well. Most base coding systems have three corresponding variants whose names are formed by adding ‘-unix’, ‘-dos’ and ‘-mac’.

The coding system raw-text is special in that it prevents character code conversion, and causes the buffer visited with this coding system to be a unibyte buffer. For historical reasons, you can save both unibyte and multibyte text with this coding system. When you use raw-text to encode multibyte text, it does perform one character code conversion: it converts eight-bit characters to their single-byte external representation. raw-text does not specify the end-of-line conversion, allowing that to be determined as usual by the data, and has the usual three variants which specify the end-of-line conversion.

no-conversion (and its alias binary) is equivalent to raw-text-unix: it specifies no conversion of either character codes or end-of-line.

The coding system utf-8-emacs specifies that the data is represented in the internal Emacs encoding (see Text Representations). This is like raw-text in that no code conversion happens, but different in that the result is multibyte data. The name emacs-internal is an alias for utf-8-emacs-unix (so it forces no conversion of end-of-line, unlike utf-8-emacs, which can decode all 3 kinds of end-of-line conventions).

Function: coding-system-get coding-system property

This function returns the specified property of the coding system coding-system. Most coding system properties exist for internal purposes, but one that you might find useful is :mime-charset. That property’s value is the name used in MIME for the character coding which this coding system can read and write. Examples:

(coding-system-get 'iso-latin-1 :mime-charset)
     ⇒ iso-8859-1
(coding-system-get 'iso-2022-cn :mime-charset)
     ⇒ iso-2022-cn
(coding-system-get 'cyrillic-koi8 :mime-charset)
     ⇒ koi8-r

The value of the :mime-charset property is also defined as an alias for the coding system.

Function: coding-system-aliases coding-system

This function returns the list of aliases of coding-system.


34.10.2 Encoding and I/O

The principal purpose of coding systems is for use in reading and writing files. The function insert-file-contents uses a coding system to decode the file data, and write-region uses one to encode the buffer contents.

You can specify the coding system to use either explicitly (see Specifying a Coding System for One Operation), or implicitly using a default mechanism (see Default Coding Systems). But these methods may not completely specify what to do. For example, they may choose a coding system such as undecided which leaves the character code conversion to be determined from the data. In these cases, the I/O operation finishes the job of choosing a coding system. Very often you will want to find out afterwards which coding system was chosen.

Variable: buffer-file-coding-system

This buffer-local variable records the coding system used for saving the buffer and for writing part of the buffer with write-region. If the text to be written cannot be safely encoded using the coding system specified by this variable, these operations select an alternative encoding by calling the function select-safe-coding-system (see User-Chosen Coding Systems). If selecting a different encoding requires to ask the user to specify a coding system, buffer-file-coding-system is updated to the newly selected coding system.

buffer-file-coding-system does not affect sending text to a subprocess.

Variable: save-buffer-coding-system

This variable specifies the coding system for saving the buffer (by overriding buffer-file-coding-system). Note that it is not used for write-region.

When a command to save the buffer starts out to use buffer-file-coding-system (or save-buffer-coding-system), and that coding system cannot handle the actual text in the buffer, the command asks the user to choose another coding system (by calling select-safe-coding-system). After that happens, the command also updates buffer-file-coding-system to represent the coding system that the user specified.

Variable: last-coding-system-used

I/O operations for files and subprocesses set this variable to the coding system name that was used. The explicit encoding and decoding functions (see Explicit Encoding and Decoding) set it too.

Warning: Since receiving subprocess output sets this variable, it can change whenever Emacs waits; therefore, you should copy the value shortly after the function call that stores the value you are interested in.

The variable selection-coding-system specifies how to encode selections for the window system. See Window System Selections.

Variable: file-name-coding-system

The variable file-name-coding-system specifies the coding system to use for encoding file names. Emacs encodes file names using that coding system for all file operations. If file-name-coding-system is nil, Emacs uses a default coding system determined by the selected language environment. In the default language environment, any non-ASCII characters in file names are not encoded specially; they appear in the file system using the internal Emacs representation.

Warning: if you change file-name-coding-system (or the language environment) in the middle of an Emacs session, problems can result if you have already visited files whose names were encoded using the earlier coding system and are handled differently under the new coding system. If you try to save one of these buffers under the visited file name, saving may use the wrong file name, or it may get an error. If such a problem happens, use C-x C-w to specify a new file name for that buffer.

On Windows 2000 and later, Emacs by default uses Unicode APIs to pass file names to the OS, so the value of file-name-coding-system is largely ignored. Lisp applications that need to encode or decode file names on the Lisp level should use utf-8 coding-system when system-type is windows-nt; the conversion of UTF-8 encoded file names to the encoding appropriate for communicating with the OS is performed internally by Emacs.


34.10.3 Coding Systems in Lisp

Here are the Lisp facilities for working with coding systems:

Function: coding-system-list &optional base-only

This function returns a list of all coding system names (symbols). If base-only is non-nil, the value includes only the base coding systems. Otherwise, it includes alias and variant coding systems as well.

Function: coding-system-p object

This function returns t if object is a coding system name or nil.

Function: check-coding-system coding-system

This function checks the validity of coding-system. If that is valid, it returns coding-system. If coding-system is nil, the function returns nil. For any other values, it signals an error whose error-symbol is coding-system-error (see signal).

Function: coding-system-eol-type coding-system

This function returns the type of end-of-line (a.k.a. eol) conversion used by coding-system. If coding-system specifies a certain eol conversion, the return value is an integer 0, 1, or 2, standing for unix, dos, and mac, respectively. If coding-system doesn’t specify eol conversion explicitly, the return value is a vector of coding systems, each one with one of the possible eol conversion types, like this:

(coding-system-eol-type 'latin-1)
     ⇒ [latin-1-unix latin-1-dos latin-1-mac]

If this function returns a vector, Emacs will decide, as part of the text encoding or decoding process, what eol conversion to use. For decoding, the end-of-line format of the text is auto-detected, and the eol conversion is set to match it (e.g., DOS-style CRLF format will imply dos eol conversion). For encoding, the eol conversion is taken from the appropriate default coding system (e.g., default value of buffer-file-coding-system for buffer-file-coding-system), or from the default eol conversion appropriate for the underlying platform.

Function: coding-system-change-eol-conversion coding-system eol-type

This function returns a coding system which is like coding-system except for its eol conversion, which is specified by eol-type. eol-type should be unix, dos, mac, or nil. If it is nil, the returned coding system determines the end-of-line conversion from the data.

eol-type may also be 0, 1 or 2, standing for unix, dos and mac, respectively.

Function: coding-system-change-text-conversion eol-coding text-coding

This function returns a coding system which uses the end-of-line conversion of eol-coding, and the text conversion of text-coding. If text-coding is nil, it returns undecided, or one of its variants according to eol-coding.

Function: find-coding-systems-region from to

This function returns a list of coding systems that could be used to encode a text between from and to. All coding systems in the list can safely encode any multibyte characters in that portion of the text.

If the text contains no multibyte characters, the function returns the list (undecided).

Function: find-coding-systems-string string

This function returns a list of coding systems that could be used to encode the text of string. All coding systems in the list can safely encode any multibyte characters in string. If the text contains no multibyte characters, this returns the list (undecided).

Function: find-coding-systems-for-charsets charsets

This function returns a list of coding systems that could be used to encode all the character sets in the list charsets.

Function: check-coding-systems-region start end coding-system-list

This function checks whether coding systems in the list coding-system-list can encode all the characters in the region between start and end. If all of the coding systems in the list can encode the specified text, the function returns nil. If some coding systems cannot encode some of the characters, the value is an alist, each element of which has the form (coding-system1 pos1 pos2 …), meaning that coding-system1 cannot encode characters at buffer positions pos1, pos2, ....

start may be a string, in which case end is ignored and the returned value references string indices instead of buffer positions.

Function: detect-coding-region start end &optional highest

This function chooses a plausible coding system for decoding the text from start to end. This text should be a byte sequence, i.e., unibyte text or multibyte text with only ASCII and eight-bit characters (see Explicit Encoding and Decoding).

Normally this function returns a list of coding systems that could handle decoding the text that was scanned. They are listed in order of decreasing priority. But if highest is non-nil, then the return value is just one coding system, the one that is highest in priority.

If the region contains only ASCII characters except for such ISO-2022 control characters ISO-2022 as ESC, the value is undecided or (undecided), or a variant specifying end-of-line conversion, if that can be deduced from the text.

If the region contains null bytes, the value is no-conversion, even if the region contains text encoded in some coding system.

Function: detect-coding-string string &optional highest

This function is like detect-coding-region except that it operates on the contents of string instead of bytes in the buffer.

Variable: inhibit-null-byte-detection

If this variable has a non-nil value, null bytes are ignored when detecting the encoding of a region or a string. This allows the encoding of text that contains null bytes to be correctly detected, such as Info files with Index nodes.

Variable: inhibit-iso-escape-detection

If this variable has a non-nil value, ISO-2022 escape sequences are ignored when detecting the encoding of a region or a string. The result is that no text is ever detected as encoded in some ISO-2022 encoding, and all escape sequences become visible in a buffer. Warning: Use this variable with extreme caution, because many files in the Emacs distribution use ISO-2022 encoding.

Function: coding-system-charset-list coding-system

This function returns the list of character sets (see Character Sets) supported by coding-system. Some coding systems that support too many character sets to list them all yield special values:

  • If coding-system supports all Emacs characters, the value is (emacs).
  • If coding-system supports all Unicode characters, the value is (unicode).
  • If coding-system supports all ISO-2022 charsets, the value is iso-2022.
  • If coding-system supports all the characters in the internal coding system used by Emacs version 21 (prior to the implementation of internal Unicode support), the value is emacs-mule.

See Process Information, in particular the description of the functions process-coding-system and set-process-coding-system, for how to examine or set the coding systems used for I/O to a subprocess.


34.10.4 User-Chosen Coding Systems

Function: select-safe-coding-system from to &optional default-coding-system accept-default-p file

This function selects a coding system for encoding specified text, asking the user to choose if necessary. Normally the specified text is the text in the current buffer between from and to. If from is a string, the string specifies the text to encode, and to is ignored.

If the specified text includes raw bytes (see Text Representations), select-safe-coding-system suggests raw-text for its encoding.

If default-coding-system is non-nil, that is the first coding system to try; if that can handle the text, select-safe-coding-system returns that coding system. It can also be a list of coding systems; then the function tries each of them one by one. After trying all of them, it next tries the current buffer’s value of buffer-file-coding-system (if it is not undecided), then the default value of buffer-file-coding-system and finally the user’s most preferred coding system, which the user can set using the command prefer-coding-system (see Recognizing Coding Systems in The GNU Emacs Manual).

If one of those coding systems can safely encode all the specified text, select-safe-coding-system chooses it and returns it. Otherwise, it asks the user to choose from a list of coding systems which can encode all the text, and returns the user’s choice.

default-coding-system can also be a list whose first element is t and whose other elements are coding systems. Then, if no coding system in the list can handle the text, select-safe-coding-system queries the user immediately, without trying any of the three alternatives described above. This is handy for checking only the coding systems in the list.

The optional argument accept-default-p determines whether a coding system selected without user interaction is acceptable. If it’s omitted or nil, such a silent selection is always acceptable. If it is non-nil, it should be a function; select-safe-coding-system calls this function with one argument, the base coding system of the selected coding system. If the function returns nil, select-safe-coding-system rejects the silently selected coding system, and asks the user to select a coding system from a list of possible candidates.

If the variable select-safe-coding-system-accept-default-p is non-nil, it should be a function taking a single argument. It is used in place of accept-default-p, overriding any value supplied for this argument.

As a final step, before returning the chosen coding system, select-safe-coding-system checks whether that coding system is consistent with what would be selected if the contents of the region were read from a file. (If not, this could lead to data corruption in a file subsequently re-visited and edited.) Normally, select-safe-coding-system uses buffer-file-name as the file for this purpose, but if file is non-nil, it uses that file instead (this can be relevant for write-region and similar functions). If it detects an apparent inconsistency, select-safe-coding-system queries the user before selecting the coding system.

Variable: select-safe-coding-system-function

This variable names the function to be called to request the user to select a proper coding system for encoding text when the default coding system for an output operation cannot safely encode that text. The default value of this variable is select-safe-coding-system. Emacs primitives that write text to files, such as write-region, or send text to other processes, such as process-send-region, normally call the value of this variable, unless coding-system-for-write is bound to a non-nil value (see Specifying a Coding System for One Operation).

Here are two functions you can use to let the user specify a coding system, with completion. See Completion.

Function: read-coding-system prompt &optional default

This function reads a coding system using the minibuffer, prompting with string prompt, and returns the coding system name as a symbol. If the user enters null input, default specifies which coding system to return. It should be a symbol or a string.

Function: read-non-nil-coding-system prompt

This function reads a coding system using the minibuffer, prompting with string prompt, and returns the coding system name as a symbol. If the user tries to enter null input, it asks the user to try again. See Coding Systems.


34.10.5 Default Coding Systems

This section describes variables that specify the default coding system for certain files or when running certain subprograms, and the function that I/O operations use to access them.

The idea of these variables is that you set them once and for all to the defaults you want, and then do not change them again. To specify a particular coding system for a particular operation in a Lisp program, don’t change these variables; instead, override them using coding-system-for-read and coding-system-for-write (see Specifying a Coding System for One Operation).

User Option: auto-coding-regexp-alist

This variable is an alist of text patterns and corresponding coding systems. Each element has the form (regexp . coding-system); a file whose first few kilobytes match regexp is decoded with coding-system when its contents are read into a buffer. The settings in this alist take priority over coding: tags in the files and the contents of file-coding-system-alist (see below). The default value is set so that Emacs automatically recognizes mail files in Babyl format and reads them with no code conversions.

User Option: file-coding-system-alist

This variable is an alist that specifies the coding systems to use for reading and writing particular files. Each element has the form (pattern . coding), where pattern is a regular expression that matches certain file names. The element applies to file names that match pattern.

The CDR of the element, coding, should be either a coding system, a cons cell containing two coding systems, or a function name (a symbol with a function definition). If coding is a coding system, that coding system is used for both reading the file and writing it. If coding is a cons cell containing two coding systems, its CAR specifies the coding system for decoding, and its CDR specifies the coding system for encoding.

If coding is a function name, the function should take one argument, a list of all arguments passed to find-operation-coding-system. It must return a coding system or a cons cell containing two coding systems. This value has the same meaning as described above.

If coding (or what returned by the above function) is undecided, the normal code-detection is performed.

User Option: auto-coding-alist

This variable is an alist that specifies the coding systems to use for reading and writing particular files. Its form is like that of file-coding-system-alist, but, unlike the latter, this variable takes priority over any coding: tags in the file.

Variable: process-coding-system-alist

This variable is an alist specifying which coding systems to use for a subprocess, depending on which program is running in the subprocess. It works like file-coding-system-alist, except that pattern is matched against the program name used to start the subprocess. The coding system or systems specified in this alist are used to initialize the coding systems used for I/O to the subprocess, but you can specify other coding systems later using set-process-coding-system.

Warning: Coding systems such as undecided, which determine the coding system from the data, do not work entirely reliably with asynchronous subprocess output. This is because Emacs handles asynchronous subprocess output in batches, as it arrives. If the coding system leaves the character code conversion unspecified, or leaves the end-of-line conversion unspecified, Emacs must try to detect the proper conversion from one batch at a time, and this does not always work.

Therefore, with an asynchronous subprocess, if at all possible, use a coding system which determines both the character code conversion and the end of line conversion—that is, one like latin-1-unix, rather than undecided or latin-1.

Variable: network-coding-system-alist

This variable is an alist that specifies the coding system to use for network streams. It works much like file-coding-system-alist, with the difference that the pattern in an element may be either a port number or a regular expression. If it is a regular expression, it is matched against the network service name used to open the network stream.

Variable: default-process-coding-system

This variable specifies the coding systems to use for subprocess (and network stream) input and output, when nothing else specifies what to do.

The value should be a cons cell of the form (input-coding . output-coding). Here input-coding applies to input from the subprocess, and output-coding applies to output to it.

User Option: auto-coding-functions

This variable holds a list of functions that try to determine a coding system for a file based on its undecoded contents.

Each function in this list should be written to look at text in the current buffer, but should not modify it in any way. The buffer will contain the text of parts of the file. Each function should take one argument, size, which tells it how many characters to look at, starting from point. If the function succeeds in determining a coding system for the file, it should return that coding system. Otherwise, it should return nil.

The functions in this list could be called either when the file is visited and Emacs wants to decode its contents, and/or when the file’s buffer is about to be saved and Emacs wants to determine how to encode its contents.

If a file has a ‘coding:’ tag, that takes precedence, so these functions won’t be called.

Function: find-auto-coding filename size

This function tries to determine a suitable coding system for filename. It examines the buffer visiting the named file, using the variables documented above in sequence, until it finds a match for one of the rules specified by these variables. It then returns a cons cell of the form (coding . source), where coding is the coding system to use and source is a symbol, one of auto-coding-alist, auto-coding-regexp-alist, :coding, or auto-coding-functions, indicating which one supplied the matching rule. The value :coding means the coding system was specified by the coding: tag in the file (see coding tag in The GNU Emacs Manual). The order of looking for a matching rule is auto-coding-alist first, then auto-coding-regexp-alist, then the coding: tag, and lastly auto-coding-functions. If no matching rule was found, the function returns nil.

The second argument size is the size of text, in characters, following point. The function examines text only within size characters after point. Normally, the buffer should be positioned at the beginning when this function is called, because one of the places for the coding: tag is the first one or two lines of the file; in that case, size should be the size of the buffer.

Function: set-auto-coding filename size

This function returns a suitable coding system for file filename. It uses find-auto-coding to find the coding system. If no coding system could be determined, the function returns nil. The meaning of the argument size is like in find-auto-coding.

Function: find-operation-coding-system operation &rest arguments

This function returns the coding system to use (by default) for performing operation with arguments. The value has this form:

(decoding-system . encoding-system)

The first element, decoding-system, is the coding system to use for decoding (in case operation does decoding), and encoding-system is the coding system for encoding (in case operation does encoding).

The argument operation is a symbol; it should be one of write-region, start-process, call-process, call-process-region, insert-file-contents, or open-network-stream. These are the names of the Emacs I/O primitives that can do character code and eol conversion.

The remaining arguments should be the same arguments that might be given to the corresponding I/O primitive. Depending on the primitive, one of those arguments is selected as the target. For example, if operation does file I/O, whichever argument specifies the file name is the target. For subprocess primitives, the process name is the target. For open-network-stream, the target is the service name or port number.

Depending on operation, this function looks up the target in file-coding-system-alist, process-coding-system-alist, or network-coding-system-alist. If the target is found in the alist, find-operation-coding-system returns its association in the alist; otherwise it returns nil.

If operation is insert-file-contents, the argument corresponding to the target may be a cons cell of the form (filename . buffer). In that case, filename is a file name to look up in file-coding-system-alist, and buffer is a buffer that contains the file’s contents (not yet decoded). If file-coding-system-alist specifies a function to call for this file, and that function needs to examine the file’s contents (as it usually does), it should examine the contents of buffer instead of reading the file.


34.10.6 Specifying a Coding System for One Operation

You can specify the coding system for a specific operation by binding the variables coding-system-for-read and/or coding-system-for-write.

Variable: coding-system-for-read

If this variable is non-nil, it specifies the coding system to use for reading a file, or for input from a synchronous subprocess.

It also applies to any asynchronous subprocess or network stream, but in a different way: the value of coding-system-for-read when you start the subprocess or open the network stream specifies the input decoding method for that subprocess or network stream. It remains in use for that subprocess or network stream unless and until overridden.

The right way to use this variable is to bind it with let for a specific I/O operation. Its global value is normally nil, and you should not globally set it to any other value. Here is an example of the right way to use the variable:

;; Read the file with no character code conversion.
(let ((coding-system-for-read 'no-conversion))
  (insert-file-contents filename))

When its value is non-nil, this variable takes precedence over all other methods of specifying a coding system to use for input, including file-coding-system-alist, process-coding-system-alist and network-coding-system-alist.

Variable: coding-system-for-write

This works much like coding-system-for-read, except that it applies to output rather than input. It affects writing to files, as well as sending output to subprocesses and net connections. It also applies to encoding command-line arguments with which Emacs invokes subprocesses.

When a single operation does both input and output, as do call-process-region and start-process, both coding-system-for-read and coding-system-for-write affect it.

Variable: coding-system-require-warning

Binding coding-system-for-write to a non-nil value prevents output primitives from calling the function specified by select-safe-coding-system-function (see User-Chosen Coding Systems). This is because C-x RET c (universal-coding-system-argument) works by binding coding-system-for-write, and Emacs should obey user selection. If a Lisp program binds coding-system-for-write to a value that might not be safe for encoding the text to be written, it can also bind coding-system-require-warning to a non-nil value, which will force the output primitives to check the encoding by calling the value of select-safe-coding-system-function even though coding-system-for-write is non-nil. Alternatively, call select-safe-coding-system explicitly before using the specified encoding.

User Option: inhibit-eol-conversion

When this variable is non-nil, no end-of-line conversion is done, no matter which coding system is specified. This applies to all the Emacs I/O and subprocess primitives, and to the explicit encoding and decoding functions (see Explicit Encoding and Decoding).

Sometimes, you need to prefer several coding systems for some operation, rather than fix a single one. Emacs lets you specify a priority order for using coding systems. This ordering affects the sorting of lists of coding systems returned by functions such as find-coding-systems-region (see Coding Systems in Lisp).

Function: coding-system-priority-list &optional highestp

This function returns the list of coding systems in the order of their current priorities. Optional argument highestp, if non-nil, means return only the highest priority coding system.

Function: set-coding-system-priority &rest coding-systems

This function puts coding-systems at the beginning of the priority list for coding systems, thus making their priority higher than all the rest.

Macro: with-coding-priority coding-systems &rest body

This macro executes body, like progn does (see progn), with coding-systems at the front of the priority list for coding systems. coding-systems should be a list of coding systems to prefer during execution of body.


34.10.7 Explicit Encoding and Decoding

All the operations that transfer text in and out of Emacs have the ability to use a coding system to encode or decode the text. You can also explicitly encode and decode text using the functions in this section.

The result of encoding, and the input to decoding, are not ordinary text. They logically consist of a series of byte values; that is, a series of ASCII and eight-bit characters. In unibyte buffers and strings, these characters have codes in the range 0 through #xFF (255). In a multibyte buffer or string, eight-bit characters have character codes higher than #xFF (see Text Representations), but Emacs transparently converts them to their single-byte values when you encode or decode such text.

The usual way to read a file into a buffer as a sequence of bytes, so you can decode the contents explicitly, is with insert-file-contents-literally (see Reading from Files); alternatively, specify a non-nil rawfile argument when visiting a file with find-file-noselect. These methods result in a unibyte buffer.

The usual way to use the byte sequence that results from explicitly encoding text is to copy it to a file or process—for example, to write it with write-region (see Writing to Files), and suppress encoding by binding coding-system-for-write to no-conversion.

Here are the functions to perform explicit encoding or decoding. The encoding functions produce sequences of bytes; the decoding functions are meant to operate on sequences of bytes. All of these functions discard text properties. They also set last-coding-system-used to the precise coding system they used.

Command: encode-coding-region start end coding-system &optional destination

This command encodes the text from start to end according to coding system coding-system. Normally, the encoded text replaces the original text in the buffer, but the optional argument destination can change that. If destination is a buffer, the encoded text is inserted in that buffer after point (point does not move); if it is t, the command returns the encoded text as a unibyte string without inserting it.

If encoded text is inserted in some buffer, this command returns the length of the encoded text.

The result of encoding is logically a sequence of bytes, but the buffer remains multibyte if it was multibyte before, and any 8-bit bytes are converted to their multibyte representation (see Text Representations).

Do not use undecided for coding-system when encoding text, since that may lead to unexpected results. Instead, use select-safe-coding-system (see select-safe-coding-system) to suggest a suitable encoding, if there’s no obvious pertinent value for coding-system.

Function: encode-coding-string string coding-system &optional nocopy buffer

This function encodes the text in string according to coding system coding-system. It returns a new string containing the encoded text, except when nocopy is non-nil, in which case the function may return string itself if the encoding operation is trivial. The result of encoding is a unibyte string.

Command: decode-coding-region start end coding-system &optional destination

This command decodes the text from start to end according to coding system coding-system. To make explicit decoding useful, the text before decoding ought to be a sequence of byte values, but both multibyte and unibyte buffers are acceptable (in the multibyte case, the raw byte values should be represented as eight-bit characters). Normally, the decoded text replaces the original text in the buffer, but the optional argument destination can change that. If destination is a buffer, the decoded text is inserted in that buffer after point (point does not move); if it is t, the command returns the decoded text as a multibyte string without inserting it.

If decoded text is inserted in some buffer, this command returns the length of the decoded text. If that buffer is a unibyte buffer (see Selecting a Representation), the internal representation of the decoded text (see Text Representations) is inserted into the buffer as individual bytes.

This command puts a charset text property on the decoded text. The value of the property states the character set used to decode the original text.

This command detects the encoding of the text if necessary. If coding-system is undecided, the command detects the encoding of the text based on the byte sequences it finds in the text, and also detects the type of end-of-line convention used by the text (see eol type). If coding-system is undecided-eol-type, where eol-type is unix, dos, or mac, then the command detects only the encoding of the text. Any coding-system that doesn’t specify eol-type, as in utf-8, causes the command to detect the end-of-line convention; specify the encoding completely, as in utf-8-unix, if the EOL convention used by the text is known in advance, to prevent any automatic detection.

Function: decode-coding-string string coding-system &optional nocopy buffer

This function decodes the text in string according to coding-system. It returns a new string containing the decoded text, except when nocopy is non-nil, in which case the function may return string itself if the decoding operation is trivial. To make explicit decoding useful, the contents of string ought to be a unibyte string with a sequence of byte values, but a multibyte string is also acceptable (assuming it contains 8-bit bytes in their multibyte form).

This function detects the encoding of the string if needed, like decode-coding-region does.

If optional argument buffer specifies a buffer, the decoded text is inserted in that buffer after point (point does not move). In this case, the return value is the length of the decoded text. If that buffer is a unibyte buffer, the internal representation of the decoded text is inserted into it as individual bytes.

This function puts a charset text property on the decoded text. The value of the property states the character set used to decode the original text:

(decode-coding-string "Gr\374ss Gott" 'latin-1)
     ⇒ #("Grüss Gott" 0 9 (charset iso-8859-1))
Function: decode-coding-inserted-region from to filename &optional visit beg end replace

This function decodes the text from from to to as if it were being read from file filename using insert-file-contents using the rest of the arguments provided.

The normal way to use this function is after reading text from a file without decoding, if you decide you would rather have decoded it. Instead of deleting the text and reading it again, this time with decoding, you can call this function.


34.10.8 Terminal I/O Encoding

Emacs can use coding systems to decode keyboard input and encode terminal output. This is useful for terminals that transmit or display text using a particular encoding, such as Latin-1. Emacs does not set last-coding-system-used when encoding or decoding terminal I/O.

Function: keyboard-coding-system &optional terminal

This function returns the coding system used for decoding keyboard input from terminal. A value of no-conversion means no decoding is done. If terminal is omitted or nil, it means the selected frame’s terminal. See Multiple Terminals.

Command: set-keyboard-coding-system coding-system &optional terminal

This command specifies coding-system as the coding system to use for decoding keyboard input from terminal. If coding-system is nil, that means not to decode keyboard input. If terminal is a frame, it means that frame’s terminal; if it is nil, that means the currently selected frame’s terminal. See Multiple Terminals. Note that on modern MS-Windows systems Emacs always uses Unicode input when decoding keyboard input, so the encoding set by this command has no effect on Windows.

Function: terminal-coding-system &optional terminal

This function returns the coding system that is in use for encoding terminal output from terminal. A value of no-conversion means no encoding is done. If terminal is a frame, it means that frame’s terminal; if it is nil, that means the currently selected frame’s terminal.

Command: set-terminal-coding-system coding-system &optional terminal

This command specifies coding-system as the coding system to use for encoding terminal output from terminal. If coding-system is nil, that means not to encode terminal output. If terminal is a frame, it means that frame’s terminal; if it is nil, that means the currently selected frame’s terminal.


34.11 Input Methods

Input methods provide convenient ways of entering non-ASCII characters from the keyboard. Unlike coding systems, which translate non-ASCII characters to and from encodings meant to be read by programs, input methods provide human-friendly commands. (See Input Methods in The GNU Emacs Manual, for information on how users use input methods to enter text.) How to define input methods is not yet documented in this manual, but here we describe how to use them.

Each input method has a name, which is currently a string; in the future, symbols may also be usable as input method names.

Variable: current-input-method

This variable holds the name of the input method now active in the current buffer. (It automatically becomes local in each buffer when set in any fashion.) It is nil if no input method is active in the buffer now.

User Option: default-input-method

This variable holds the default input method for commands that choose an input method. Unlike current-input-method, this variable is normally global.

Command: set-input-method input-method

This command activates input method input-method for the current buffer. It also sets default-input-method to input-method. If input-method is nil, this command deactivates any input method for the current buffer.

Function: read-input-method-name prompt &optional default inhibit-null

This function reads an input method name with the minibuffer, prompting with prompt. If default is non-nil, that is returned by default, if the user enters empty input. However, if inhibit-null is non-nil, empty input signals an error.

The returned value is a string.

Variable: input-method-alist

This variable defines all the supported input methods. Each element defines one input method, and should have the form:

(input-method language-env activate-func
 title description args...)

Here input-method is the input method name, a string; language-env is another string, the name of the language environment this input method is recommended for. (That serves only for documentation purposes.)

activate-func is a function to call to activate this method. The args, if any, are passed as arguments to activate-func. All told, the arguments to activate-func are input-method and the args.

title is a string to display in the mode line while this method is active. description is a string describing this method and what it is good for.

The fundamental interface to input methods is through the variable input-method-function. See Reading One Event, and Invoking the Input Method.


34.12 Locales

In POSIX, locales control which language to use in language-related features. These Emacs variables control how Emacs interacts with these features.

Variable: locale-coding-system

This variable specifies the coding system to use for decoding system error messages and—on X Window system only—keyboard input, for sending batch output to the standard output and error streams, for encoding the format argument to format-time-string, and for decoding the return value of format-time-string.

Variable: system-messages-locale

This variable specifies the locale to use for generating system error messages. Changing the locale can cause messages to come out in a different language or in a different orthography. If the variable is nil, the locale is specified by environment variables in the usual POSIX fashion.

Variable: system-time-locale

This variable specifies the locale to use for formatting time values. Changing the locale can cause messages to appear according to the conventions of a different language. If the variable is nil, the locale is specified by environment variables in the usual POSIX fashion.

Function: locale-info item

This function returns locale data item for the current POSIX locale, if available. item should be one of these symbols:

codeset

Return the character set as a string (locale item CODESET).

days

Return a 7-element vector of day names (locale items DAY_1 through DAY_7);

months

Return a 12-element vector of month names (locale items MON_1 through MON_12).

paper

Return a list (width height) of 2 integers, for the default paper size measured in millimeters (locale items _NL_PAPER_WIDTH and _NL_PAPER_HEIGHT).

If the system can’t provide the requested information, or if item is not one of those symbols, the value is nil. All strings in the return value are decoded using locale-coding-system. See Locales in The GNU Libc Manual, for more information about locales and locale items.


35 Searching and Matching

GNU Emacs provides two ways to search through a buffer for specified text: exact string searches and regular expression searches. After a regular expression search, you can examine the match data to determine which text matched the whole regular expression or various portions of it.

The ‘skip-chars…’ functions also perform a kind of searching. See Skipping Characters. To search for changes in character properties, see Text Property Search Functions.


35.2 Searching and Case

By default, searches in Emacs ignore the case of the text they are searching through; if you specify searching for ‘FOO’, then ‘Foo’ or ‘foo’ is also considered a match. This applies to regular expressions, too; thus, ‘[aB]’ would match ‘a’ or ‘A’ or ‘b’ or ‘B’.

If you do not want this feature, set the variable case-fold-search to nil. Then all letters must match exactly, including case. This is a buffer-local variable; altering the variable affects only the current buffer. (See Introduction to Buffer-Local Variables.) Alternatively, you may change the default value. In Lisp code, you will more typically use let to bind case-fold-search to the desired value.

Note that the user-level incremental search feature handles case distinctions differently. When the search string contains only lower case letters, the search ignores case, but when the search string contains one or more upper case letters, the search becomes case-sensitive. But this has nothing to do with the searching functions used in Lisp code. See Incremental Search in The GNU Emacs Manual.

User Option: case-fold-search

This buffer-local variable determines whether searches should ignore case. If the variable is nil they do not ignore case; otherwise (and by default) they do ignore case.

User Option: case-replace

This variable determines whether the higher-level replacement functions should preserve case. If the variable is nil, that means to use the replacement text verbatim. A non-nil value means to convert the case of the replacement text according to the text being replaced.

This variable is used by passing it as an argument to the function replace-match. See Replacing the Text that Matched.


35.3 Regular Expressions

A regular expression, or regexp for short, is a pattern that denotes a (possibly infinite) set of strings. Searching for matches for a regexp is a very powerful operation. This section explains how to write regexps; the following section says how to search for them.

For interactive development of regular expressions, you can use the M-x re-builder command. It provides a convenient interface for creating regular expressions, by giving immediate visual feedback in a separate buffer. As you edit the regexp, all its matches in the target buffer are highlighted. Each parenthesized sub-expression of the regexp is shown in a distinct face, which makes it easier to verify even very complex regexps.

Note that by default Emacs search ignores case (see Searching and Case). To enable case-sensitive regexp search and match, bind case-fold-search to nil around the code you want to be case-sensitive.


35.3.1 Syntax of Regular Expressions

Regular expressions have a syntax in which a few characters are special constructs and the rest are ordinary. An ordinary character is a simple regular expression that matches that character and nothing else. The special characters are ‘.’, ‘*’, ‘+’, ‘?’, ‘[’, ‘^’, ‘$’, and ‘\’; no new special characters will be defined in the future. The character ‘]’ is special if it ends a character alternative (see later). The character ‘-’ is special inside a character alternative. A ‘[:’ and balancing ‘:]’ enclose a character class inside a character alternative. Any other character appearing in a regular expression is ordinary, unless a ‘\’ precedes it.

For example, ‘f’ is not a special character, so it is ordinary, and therefore ‘f’ is a regular expression that matches the string ‘f’ and no other string. (It does not match the string ‘fg’, but it does match a part of that string.) Likewise, ‘o’ is a regular expression that matches only ‘o’.

Any two regular expressions a and b can be concatenated. The result is a regular expression that matches a string if a matches some amount of the beginning of that string and b matches the rest of the string.

As a simple example, we can concatenate the regular expressions ‘f’ and ‘o’ to get the regular expression ‘fo’, which matches only the string ‘fo’. Still trivial. To do something more powerful, you need to use one of the special regular expression constructs.


35.3.1.1 Special Characters in Regular Expressions

Here is a list of the characters that are special in a regular expression.

.(Period)

is a special character that matches any single character except a newline. Using concatenation, we can make regular expressions like ‘a.b’, which matches any three-character string that begins with ‘a’ and ends with ‘b’.

*

is not a construct by itself; it is a postfix operator that means to match the preceding regular expression repetitively as many times as possible. Thus, ‘o*’ matches any number of ‘o’s (including no ‘o’s).

*’ always applies to the smallest possible preceding expression. Thus, ‘fo*’ has a repeating ‘o’, not a repeating ‘fo’. It matches ‘f’, ‘fo’, ‘foo’, and so on.

The matcher processes a ‘*’ construct by matching, immediately, as many repetitions as can be found. Then it continues with the rest of the pattern. If that fails, backtracking occurs, discarding some of the matches of the ‘*’-modified construct in the hope that this will make it possible to match the rest of the pattern. For example, in matching ‘ca*ar’ against the string ‘caaar’, the ‘a*’ first tries to match all three ‘a’s; but the rest of the pattern is ‘ar’ and there is only ‘r’ left to match, so this try fails. The next alternative is for ‘a*’ to match only two ‘a’s. With this choice, the rest of the regexp matches successfully.

+

is a postfix operator, similar to ‘*’ except that it must match the preceding expression at least once. So, for example, ‘ca+r’ matches the strings ‘car’ and ‘caaaar’ but not the string ‘cr’, whereas ‘ca*r’ matches all three strings.

?

is a postfix operator, similar to ‘*’ except that it must match the preceding expression either once or not at all. For example, ‘ca?r’ matches ‘car’ or ‘cr’; nothing else.

*?’, ‘+?’, ‘??

are non-greedy variants of the operators ‘*’, ‘+’ and ‘?’. Where those operators match the largest possible substring (consistent with matching the entire containing expression), the non-greedy variants match the smallest possible substring (consistent with matching the entire containing expression).

For example, the regular expression ‘c[ad]*a’ when applied to the string ‘cdaaada’ matches the whole string; but the regular expression ‘c[ad]*?a’, applied to that same string, matches just ‘cda’. (The smallest possible match here for ‘[ad]*?’ that permits the whole expression to match is ‘d’.)

[ … ]

is a character alternative, which begins with ‘[’ and is terminated by ‘]’. In the simplest case, the characters between the two brackets are what this character alternative can match.

Thus, ‘[ad]’ matches either one ‘a’ or one ‘d’, and ‘[ad]*’ matches any string composed of just ‘a’s and ‘d’s (including the empty string). It follows that ‘c[ad]*r’ matches ‘cr’, ‘car’, ‘cdr’, ‘caddaar’, etc.

You can also include character ranges in a character alternative, by writing the starting and ending characters with a ‘-’ between them. Thus, ‘[a-z]’ matches any lower-case ASCII letter. Ranges may be intermixed freely with individual characters, as in ‘[a-z$%.]’, which matches any lower case ASCII letter or ‘$’, ‘%’ or period. However, the ending character of one range should not be the starting point of another one; for example, ‘[a-m-z]’ should be avoided.

A character alternative can also specify named character classes (see Character Classes). For example, ‘[[:ascii:]]’ matches any ASCII character. Using a character class is equivalent to mentioning each of the characters in that class; but the latter is not feasible in practice, since some classes include thousands of different characters. A character class should not appear as the lower or upper bound of a range.

The usual regexp special characters are not special inside a character alternative. A completely different set of characters is special: ‘]’, ‘-’ and ‘^’. To include ‘]’ in a character alternative, put it at the beginning. To include ‘^’, put it anywhere but at the beginning. To include ‘-’, put it at the end. Thus, ‘[]^-]’ matches all three of these special characters. You cannot use ‘\’ to escape these three characters, since ‘\’ is not special here.

The following aspects of ranges are specific to Emacs, in that POSIX allows but does not require this behavior and programs other than Emacs may behave differently:

  1. If case-fold-search is non-nil, ‘[a-z]’ also matches upper-case letters.
  2. A range is not affected by the locale’s collation sequence: it always represents the set of characters with codepoints ranging between those of its bounds, so that ‘[a-z]’ matches only ASCII letters, even outside the C or POSIX locale.
  3. If the lower bound of a range is greater than its upper bound, the range is empty and represents no characters. Thus, ‘[z-a]’ always fails to match, and ‘[^z-a]’ matches any character, including newline. However, a reversed range should always be from the letter ‘z’ to the letter ‘a’ to make it clear that it is not a typo; for example, ‘[+-*/]’ should be avoided, because it matches only ‘/’ rather than the likely-intended four characters.
  4. If the end points of a range are raw 8-bit bytes (see Text Representations), or if the range start is ASCII and the end is a raw byte (as in ‘[a-\377]’), the range will match only ASCII characters and raw 8-bit bytes, but not non-ASCII characters. This feature is intended for searching text in unibyte buffers and strings.

Some kinds of character alternatives are not the best style even though they have a well-defined meaning in Emacs. They include:

  1. Although a range’s bound can be almost any character, it is better style to stay within natural sequences of ASCII letters and digits because most people have not memorized character code tables. For example, ‘[.-9]’ is less clear than ‘[./0-9]’, and ‘[`-~]’ is less clear than ‘[`a-z{|}~]’. Unicode character escapes can help here; for example, for most programmers ‘[ก-ฺ฿-๛]’ is less clear than ‘[\u0E01-\u0E3A\u0E3F-\u0E5B]’.
  2. Although a character alternative can include duplicates, it is better style to avoid them. For example, ‘[XYa-yYb-zX]’ is less clear than ‘[XYa-z]’.
  3. Although a range can denote just one, two, or three characters, it is simpler to list the characters. For example, ‘[a-a0]’ is less clear than ‘[a0]’, ‘[i-j]’ is less clear than ‘[ij]’, and ‘[i-k]’ is less clear than ‘[ijk]’.
  4. Although a ‘-’ can appear at the beginning of a character alternative or as the upper bound of a range, it is better style to put ‘-’ by itself at the end of a character alternative. For example, although ‘[-a-z]’ is valid, ‘[a-z-]’ is better style; and although ‘[*--]’ is valid, ‘[*+,-]’ is clearer.
[^ … ]

[^’ begins a complemented character alternative. This matches any character except the ones specified. Thus, ‘[^a-z0-9A-Z]’ matches all characters except ASCII letters and digits.

^’ is not special in a character alternative unless it is the first character. The character following the ‘^’ is treated as if it were first (in other words, ‘-’ and ‘]’ are not special there).

A complemented character alternative can match a newline, unless newline is mentioned as one of the characters not to match. This is in contrast to the handling of regexps in programs such as grep.

You can specify named character classes, just like in character alternatives. For instance, ‘[^[:ascii:]]’ matches any non-ASCII character. See Character Classes.

^

When matching a buffer, ‘^’ matches the empty string, but only at the beginning of a line in the text being matched (or the beginning of the accessible portion of the buffer). Otherwise it fails to match anything. Thus, ‘^foo’ matches a ‘foo’ that occurs at the beginning of a line.

When matching a string instead of a buffer, ‘^’ matches at the beginning of the string or after a newline character.

For historical compatibility reasons, ‘^’ can be used only at the beginning of the regular expression, or after ‘\(’, ‘\(?:’ or ‘\|’.

$

is similar to ‘^’ but matches only at the end of a line (or the end of the accessible portion of the buffer). Thus, ‘x+$’ matches a string of one ‘x’ or more at the end of a line.

When matching a string instead of a buffer, ‘$’ matches at the end of the string or before a newline character.

For historical compatibility reasons, ‘$’ can be used only at the end of the regular expression, or before ‘\)’ or ‘\|’.

\

has two functions: it quotes the special characters (including ‘\’), and it introduces additional special constructs.

Because ‘\’ quotes special characters, ‘\$’ is a regular expression that matches only ‘$’, and ‘\[’ is a regular expression that matches only ‘[’, and so on.

Note that ‘\’ also has special meaning in the read syntax of Lisp strings (see String Type), and must be quoted with ‘\’. For example, the regular expression that matches the ‘\’ character is ‘\\’. To write a Lisp string that contains the characters ‘\\’, Lisp syntax requires you to quote each ‘\’ with another ‘\’. Therefore, the read syntax for a regular expression matching ‘\’ is "\\\\".

Please note: For historical compatibility, special characters are treated as ordinary ones if they are in contexts where their special meanings make no sense. For example, ‘*foo’ treats ‘*’ as ordinary since there is no preceding expression on which the ‘*’ can act. It is poor practice to depend on this behavior; quote the special character anyway, regardless of where it appears.

As a ‘\’ is not special inside a character alternative, it can never remove the special meaning of ‘-’, ‘^’ or ‘]’. You should not quote these characters when they have no special meaning. This would not clarify anything, since backslashes can legitimately precede these characters where they have special meaning, as in ‘[^\]’ ("[^\\]" for Lisp string syntax), which matches any single character except a backslash.

In practice, most ‘]’ that occur in regular expressions close a character alternative and hence are special. However, occasionally a regular expression may try to match a complex pattern of literal ‘[’ and ‘]’. In such situations, it sometimes may be necessary to carefully parse the regexp from the start to determine which square brackets enclose a character alternative. For example, ‘[^][]]’ consists of the complemented character alternative ‘[^][]’ (which matches any single character that is not a square bracket), followed by a literal ‘]’.

The exact rules are that at the beginning of a regexp, ‘[’ is special and ‘]’ not. This lasts until the first unquoted ‘[’, after which we are in a character alternative; ‘[’ is no longer special (except when it starts a character class) but ‘]’ is special, unless it immediately follows the special ‘[’ or that ‘[’ followed by a ‘^’. This lasts until the next special ‘]’ that does not end a character class. This ends the character alternative and restores the ordinary syntax of regular expressions; an unquoted ‘[’ is special again and a ‘]’ not.


35.3.1.2 Character Classes

Below is a table of the classes you can use in a character alternative, and what they mean. Note that the ‘[’ and ‘]’ characters that enclose the class name are part of the name, so a regular expression using these classes needs one more pair of brackets. For example, a regular expression matching a sequence of one or more letters and digits would be ‘[[:alnum:]]+’, not ‘[:alnum:]+’.

[:ascii:]

This matches any ASCII character (codes 0–127).

[:alnum:]

This matches any letter or digit. For multibyte characters, it matches characters whose Unicode ‘general-category’ property (see Character Properties) indicates they are alphabetic or decimal number characters.

[:alpha:]

This matches any letter. For multibyte characters, it matches characters whose Unicode ‘general-category’ property (see Character Properties) indicates they are alphabetic characters.

[:blank:]

This matches horizontal whitespace, as defined by Annex C of the Unicode Technical Standard #18. In particular, it matches spaces, tabs, and other characters whose Unicode ‘general-category’ property (see Character Properties) indicates they are spacing separators.

[:cntrl:]

This matches any character whose code is in the range 0–31.

[:digit:]

This matches ‘0’ through ‘9’. Thus, ‘[-+[:digit:]]’ matches any digit, as well as ‘+’ and ‘-’.

[:graph:]

This matches graphic characters—everything except spaces, ASCII and non-ASCII control characters, surrogates, and codepoints unassigned by Unicode, as indicated by the Unicode ‘general-category’ property (see Character Properties).

[:lower:]

This matches any lower-case letter, as determined by the current case table (see The Case Table). If case-fold-search is non-nil, this also matches any upper-case letter. Note that a buffer can have its own local case table different from the default one.

[:multibyte:]

This matches any multibyte character (see Text Representations).

[:nonascii:]

This matches any non-ASCII character.

[:print:]

This matches any printing character—either spaces or graphic characters matched by ‘[:graph:]’.

[:punct:]

This matches any punctuation character. (At present, for multibyte characters, it matches anything that has non-word syntax, and thus its exact definition can vary from one major mode to another, since the syntax of a character depends on the major mode.)

[:space:]

This matches any character that has whitespace syntax (see Table of Syntax Classes). Note that the syntax of a character, and thus which characters are considered “whitespace”, depends on the major mode.

[:unibyte:]

This matches any unibyte character (see Text Representations).

[:upper:]

This matches any upper-case letter, as determined by the current case table (see The Case Table). If case-fold-search is non-nil, this also matches any lower-case letter. Note that a buffer can have its own local case table different from the default one.

[:word:]

This matches any character that has word syntax (see Table of Syntax Classes). Note that the syntax of a character, and thus which characters are considered “word-constituent”, depends on the major mode.

[:xdigit:]

This matches the hexadecimal digits: ‘0’ through ‘9’, ‘a’ through ‘f’ and ‘A’ through ‘F’.


35.3.1.3 Backslash Constructs in Regular Expressions

For the most part, ‘\’ followed by any character matches only that character. However, there are several exceptions: certain sequences starting with ‘\’ that have special meanings. Here is a table of the special ‘\’ constructs.

\|

specifies an alternative. Two regular expressions a and b with ‘\|’ in between form an expression that matches anything that either a or b matches.

Thus, ‘foo\|bar’ matches either ‘foo’ or ‘bar’ but no other string.

\|’ applies to the largest possible surrounding expressions. Only a surrounding ‘\( … \)’ grouping can limit the grouping power of ‘\|’.

If you need full backtracking capability to handle multiple uses of ‘\|’, use the POSIX regular expression functions (see POSIX Regular Expression Searching).

\{m\}

is a postfix operator that repeats the previous pattern exactly m times. Thus, ‘x\{5\}’ matches the string ‘xxxxx’ and nothing else. ‘c[ad]\{3\}r’ matches string such as ‘caaar’, ‘cdddr’, ‘cadar’, and so on.

\{m,n\}

is a more general postfix operator that specifies repetition with a minimum of m repeats and a maximum of n repeats. If m is omitted, the minimum is 0; if n is omitted, there is no maximum. For both forms, m and n, if specified, may be no larger than 2**16 − 1 .

For example, ‘c[ad]\{1,2\}r’ matches the strings ‘car’, ‘cdr’, ‘caar’, ‘cadr’, ‘cdar’, and ‘cddr’, and nothing else.
\{0,1\}’ or ‘\{,1\}’ is equivalent to ‘?’.
\{0,\}’ or ‘\{,\}’ is equivalent to ‘*’.
\{1,\}’ is equivalent to ‘+’.

\( … \)

is a grouping construct that serves three purposes:

  1. To enclose a set of ‘\|’ alternatives for other operations. Thus, the regular expression ‘\(foo\|bar\)x’ matches either ‘foox’ or ‘barx’.
  2. To enclose a complicated expression for the postfix operators ‘*’, ‘+’ and ‘?’ to operate on. Thus, ‘ba\(na\)*’ matches ‘ba’, ‘bana’, ‘banana’, ‘bananana’, etc., with any number (zero or more) of ‘na’ strings.
  3. To record a matched substring for future reference with ‘\digit’ (see below).

This last application is not a consequence of the idea of a parenthetical grouping; it is a separate feature that was assigned as a second meaning to the same ‘\( … \)’ construct because, in practice, there was usually no conflict between the two meanings. But occasionally there is a conflict, and that led to the introduction of shy groups.

\(?: … \)

is the shy group construct. A shy group serves the first two purposes of an ordinary group (controlling the nesting of other operators), but it does not get a number, so you cannot refer back to its value with ‘\digit’. Shy groups are particularly useful for mechanically-constructed regular expressions, because they can be added automatically without altering the numbering of ordinary, non-shy groups.

Shy groups are also called non-capturing or unnumbered groups.

\(?num: … \)

is the explicitly numbered group construct. Normal groups get their number implicitly, based on their position, which can be inconvenient. This construct allows you to force a particular group number. There is no particular restriction on the numbering, e.g., you can have several groups with the same number in which case the last one to match (i.e., the rightmost match) will win. Implicitly numbered groups always get the smallest integer larger than the one of any previous group.

\digit

matches the same text that matched the digitth occurrence of a grouping (‘\( … \)’) construct.

In other words, after the end of a group, the matcher remembers the beginning and end of the text matched by that group. Later on in the regular expression you can use ‘\’ followed by digit to match that same text, whatever it may have been.

The strings matching the first nine grouping constructs appearing in the entire regular expression passed to a search or matching function are assigned numbers 1 through 9 in the order that the open parentheses appear in the regular expression. So you can use ‘\1’ through ‘\9’ to refer to the text matched by the corresponding grouping constructs.

For example, ‘\(.*\)\1’ matches any newline-free string that is composed of two identical halves. The ‘\(.*\)’ matches the first half, which may be anything, but the ‘\1’ that follows must match the same exact text.

If a ‘\( … \)’ construct matches more than once (which can happen, for instance, if it is followed by ‘*’), only the last match is recorded.

If a particular grouping construct in the regular expression was never matched—for instance, if it appears inside of an alternative that wasn’t used, or inside of a repetition that repeated zero times—then the corresponding ‘\digit’ construct never matches anything. To use an artificial example, ‘\(foo\(b*\)\|lose\)\2’ cannot match ‘lose’: the second alternative inside the larger group matches it, but then ‘\2’ is undefined and can’t match anything. But it can match ‘foobb’, because the first alternative matches ‘foob’ and ‘\2’ matches ‘b’.

\w

matches any word-constituent character. The editor syntax table determines which characters these are. See Syntax Tables.

\W

matches any character that is not a word constituent.

\scode

matches any character whose syntax is code. Here code is a character that represents a syntax code: thus, ‘w’ for word constituent, ‘-’ for whitespace, ‘(’ for open parenthesis, etc. To represent whitespace syntax, use either ‘-’ or a space character. See Table of Syntax Classes, for a list of syntax codes and the characters that stand for them.

\Scode

matches any character whose syntax is not code.

\ccode

matches any character whose category is code. Here code is a character that represents a category: for example, in the standard category table, ‘c’ stands for Chinese characters and ‘g’ stands for Greek characters. You can see the list of all the currently defined categories with M-x describe-categories RET. You can also define your own categories in addition to the standard ones using the define-category function (see Categories).

\Ccode

matches any character whose category is not code.

The following regular expression constructs match the empty string—that is, they don’t consume any characters—but whether they match depends on the context. For all, the beginning and end of the accessible portion of the buffer are treated as if they were the actual beginning and end of the buffer.

\`

matches the empty string, but only at the beginning of the buffer or string being matched against.

\'

matches the empty string, but only at the end of the buffer or string being matched against.

\=

matches the empty string, but only at point. (This construct is not defined when matching against a string.)

\b

matches the empty string, but only at the beginning or end of a word. Thus, ‘\bfoo\b’ matches any occurrence of ‘foo’ as a separate word. ‘\bballs?\b’ matches ‘ball’ or ‘balls’ as a separate word.

\b’ matches at the beginning or end of the buffer (or string) regardless of what text appears next to it.

\B

matches the empty string, but not at the beginning or end of a word, nor at the beginning or end of the buffer (or string).

\<

matches the empty string, but only at the beginning of a word. ‘\<’ matches at the beginning of the buffer (or string) only if a word-constituent character follows.

\>

matches the empty string, but only at the end of a word. ‘\>’ matches at the end of the buffer (or string) only if the contents end with a word-constituent character.

\_<

matches the empty string, but only at the beginning of a symbol. A symbol is a sequence of one or more word or symbol constituent characters. ‘\_<’ matches at the beginning of the buffer (or string) only if a symbol-constituent character follows.

\_>

matches the empty string, but only at the end of a symbol. ‘\_>’ matches at the end of the buffer (or string) only if the contents end with a symbol-constituent character.

Not every string is a valid regular expression. For example, a string that ends inside a character alternative without a terminating ‘]’ is invalid, and so is a string that ends with a single ‘\’. If an invalid regular expression is passed to any of the search functions, an invalid-regexp error is signaled.


35.3.2 Complex Regexp Example

Here is a complicated regexp which was formerly used by Emacs to recognize the end of a sentence together with any whitespace that follows. (Nowadays Emacs uses a similar but more complex default regexp constructed by the function sentence-end. See Standard Regular Expressions Used in Editing.)

Below, we show first the regexp as a string in Lisp syntax (to distinguish spaces from tab characters), and then the result of evaluating it. The string constant begins and ends with a double-quote. ‘\"’ stands for a double-quote as part of the string, ‘\\’ for a backslash as part of the string, ‘\t’ for a tab and ‘\n’ for a newline.

"[.?!][]\"')}]*\\($\\| $\\|\t\\|  \\)[ \t\n]*"
     ⇒ "[.?!][]\"')}]*\\($\\| $\\|  \\|  \\)[
]*"

In the output, tab and newline appear as themselves.

This regular expression contains four parts in succession and can be deciphered as follows:

[.?!]

The first part of the pattern is a character alternative that matches any one of three characters: period, question mark, and exclamation mark. The match must begin with one of these three characters. (This is one point where the new default regexp used by Emacs differs from the old. The new value also allows some non-ASCII characters that end a sentence without any following whitespace.)

[]\"')}]*

The second part of the pattern matches any closing braces and quotation marks, zero or more of them, that may follow the period, question mark or exclamation mark. The \" is Lisp syntax for a double-quote in a string. The ‘*’ at the end indicates that the immediately preceding regular expression (a character alternative, in this case) may be repeated zero or more times.

\\($\\| $\\|\t\\|  \\)

The third part of the pattern matches the whitespace that follows the end of a sentence: the end of a line (optionally with a space), or a tab, or two spaces. The double backslashes mark the parentheses and vertical bars as regular expression syntax; the parentheses delimit a group and the vertical bars separate alternatives. The dollar sign is used to match the end of a line.

[ \t\n]*

Finally, the last part of the pattern matches any additional whitespace beyond the minimum needed to end a sentence.

In the rx notation (see The rx Structured Regexp Notation), the regexp could be written

(rx (any ".?!")                    ; Punctuation ending sentence.
    (zero-or-more (any "\"')]}"))  ; Closing quotes or brackets.
    (or line-end
        (seq " " line-end)
        "\t"
        "  ")                      ; Two spaces.
    (zero-or-more (any "\t\n ")))  ; Optional extra whitespace.

Since rx regexps are just S-expressions, they can be formatted and commented as such.


35.3.3 The rx Structured Regexp Notation

As an alternative to the string-based syntax, Emacs provides the structured rx notation based on Lisp S-expressions. This notation is usually easier to read, write and maintain than regexp strings, and can be indented and commented freely. It requires a conversion into string form since that is what regexp functions expect, but that conversion typically takes place during byte-compilation rather than when the Lisp code using the regexp is run.

Here is an rx regexp26 that matches a block comment in the C programming language:

(rx "/*"                          ; Initial /*
    (zero-or-more
     (or (not (any "*"))          ;  Either non-*,
         (seq "*"                 ;  or * followed by
              (not (any "/")))))  ;  non-/
    (one-or-more "*")             ; At least one star,
    "/")                          ; and the final /

or, using shorter synonyms and written more compactly,

(rx "/*"
    (* (| (not "*")
          (: "*" (not "/"))))
    (+ "*") "/")

In conventional string syntax, it would be written

"/\\*\\(?:[^*]\\|\\*[^/]\\)*\\*+/"

The rx notation is mainly useful in Lisp code; it cannot be used in most interactive situations where a regexp is requested, such as when running query-replace-regexp or in variable customization.


35.3.3.1 Constructs in rx regexps

The various forms in rx regexps are described below. The shorthand rx represents any rx form. rx… means zero or more rx forms and, unless stated otherwise, matches these forms in sequence as if wrapped in a (seq …) subform.

These are all valid arguments to the rx macro. All forms are defined by their described semantics; the corresponding string regexps are provided for ease of understanding only. A, B, … denote (suitably bracketed) string regexp subexpressions therein.

Literals

"some-string"

Match the string ‘some-string’ literally. There are no characters with special meaning, unlike in string regexps.

?C

Match the character ‘C’ literally.

Sequence and alternative

(seq rx…)
(sequence rx…)
(: rx…)
(and rx…)

Match the rxs in sequence. Without arguments, the expression matches the empty string.
Corresponding string regexp: ‘AB’ (subexpressions in sequence).

(or rx…)
(| rx…)

Match exactly one of the rxs. If all arguments are strings, characters, or or forms so constrained, the longest possible match will always be used. Otherwise, either the longest match or the first (in left-to-right order) will be used. Without arguments, the expression will not match anything at all.
Corresponding string regexp: ‘A\|B\|…’.

unmatchable

Refuse any match. Equivalent to (or). See regexp-unmatchable.

Repetition

Normally, repetition forms are greedy, in that they attempt to match as many times as possible. Some forms are non-greedy; they try to match as few times as possible (see Non-greedy repetition).

(zero-or-more rx…)
(0+ rx…)

Match the rxs zero or more times. Greedy by default.
Corresponding string regexp: ‘A*’ (greedy), ‘A*?’ (non-greedy)

(one-or-more rx…)
(1+ rx…)

Match the rxs one or more times. Greedy by default.
Corresponding string regexp: ‘A+’ (greedy), ‘A+?’ (non-greedy)

(zero-or-one rx…)
(optional rx…)
(opt rx…)

Match the rxs once or an empty string. Greedy by default.
Corresponding string regexp: ‘A?’ (greedy), ‘A??’ (non-greedy).

(* rx…)

Match the rxs zero or more times. Greedy.
Corresponding string regexp: ‘A*

(+ rx…)

Match the rxs one or more times. Greedy.
Corresponding string regexp: ‘A+

(? rx…)

Match the rxs once or an empty string. Greedy.
Corresponding string regexp: ‘A?

(*? rx…)

Match the rxs zero or more times. Non-greedy.
Corresponding string regexp: ‘A*?

(+? rx…)

Match the rxs one or more times. Non-greedy.
Corresponding string regexp: ‘A+?

(?? rx…)

Match the rxs or an empty string. Non-greedy.
Corresponding string regexp: ‘A??

(= n rx…)
(repeat n rx)

Match the rxs exactly n times.
Corresponding string regexp: ‘A\{n\}

(>= n rx…)

Match the rxs n or more times. Greedy.
Corresponding string regexp: ‘A\{n,\}

(** n m rx…)
(repeat n m rx…)

Match the rxs at least n but no more than m times. Greedy.
Corresponding string regexp: ‘A\{n,m\}

The greediness of some repetition forms can be controlled using the following constructs. However, it is usually better to use the explicit non-greedy forms above when such matching is required.

(minimal-match rx)

Match rx, with zero-or-more, 0+, one-or-more, 1+, zero-or-one, opt and optional using non-greedy matching.

(maximal-match rx)

Match rx, with zero-or-more, 0+, one-or-more, 1+, zero-or-one, opt and optional using greedy matching. This is the default.

Matching single characters

(any set…)
(char set…)
(in set…)

Match a single character from one of the sets. Each set is a character, a string representing the set of its characters, a range or a character class (see below). A range is either a hyphen-separated string like "A-Z", or a cons of characters like (?A . ?Z).

Note that hyphen (-) is special in strings in this construct, since it acts as a range separator. To include a hyphen, add it as a separate character or single-character string.
Corresponding string regexp: ‘[…]

(not charspec)

Match a character not included in charspec. charspec can be a character, a single-character string, an any, not, or, intersection, syntax or category form, or a character class. If charspec is an or form, its arguments have the same restrictions as those of intersection; see below.
Corresponding string regexp: ‘[^…]’, ‘\Scode’, ‘\Ccode

(intersection charset…)

Match a character included in all of the charsets. Each charset can be a character, a single-character string, an any form without character classes, or an intersection, or or not form whose arguments are also charsets.

not-newline, nonl

Match any character except a newline.
Corresponding string regexp: ‘.’ (dot)

anychar, anything

Match any character.
Corresponding string regexp: ‘.\|\n’ (for example)

character class

Match a character from a named character class:

alpha, alphabetic, letter

Match alphabetic characters. More precisely, match characters whose Unicode ‘general-category’ property indicates that they are alphabetic.

alnum, alphanumeric

Match alphabetic characters and digits. More precisely, match characters whose Unicode ‘general-category’ property indicates that they are alphabetic or decimal digits.

digit, numeric, num

Match the digits ‘0’–‘9’.

xdigit, hex-digit, hex

Match the hexadecimal digits ‘0’–‘9’, ‘A’–‘F’ and ‘a’–‘f’.

cntrl, control

Match any character whose code is in the range 0–31.

blank

Match horizontal whitespace. More precisely, match characters whose Unicode ‘general-category’ property indicates that they are spacing separators.

space, whitespace, white

Match any character that has whitespace syntax (see Table of Syntax Classes).

lower, lower-case

Match anything lower-case, as determined by the current case table. If case-fold-search is non-nil, this also matches any upper-case letter.

upper, upper-case

Match anything upper-case, as determined by the current case table. If case-fold-search is non-nil, this also matches any lower-case letter.

graph, graphic

Match any character except whitespace, ASCII and non-ASCII control characters, surrogates, and codepoints unassigned by Unicode, as indicated by the Unicode ‘general-category’ property.

print, printing

Match whitespace or a character matched by graph.

punct, punctuation

Match any punctuation character. (At present, for multibyte characters, anything that has non-word syntax.)

word, wordchar

Match any character that has word syntax (see Table of Syntax Classes).

ascii

Match any ASCII character (codes 0–127).

nonascii

Match any non-ASCII character (but not raw bytes).

Corresponding string regexp: ‘[[:class:]]

(syntax syntax)

Match a character with syntax syntax, being one of the following names:

Syntax nameSyntax character
whitespace-
punctuation.
wordw
symbol_
open-parenthesis(
close-parenthesis)
expression-prefix'
string-quote"
paired-delimiter$
escape\
character-quote/
comment-start<
comment-end>
string-delimiter|
comment-delimiter!

For details, see Table of Syntax Classes. Please note that (syntax punctuation) is not equivalent to the character class punctuation.
Corresponding string regexp: ‘\schar’ where char is the syntax character.

(category category)

Match a character in category category, which is either one of the names below or its category character.

Category nameCategory character
space-for-indentspace
base.
consonant0
base-vowel1
upper-diacritical-mark2
lower-diacritical-mark3
tone-mark4
symbol5
digit6
vowel-modifying-diacritical-mark7
vowel-sign8
semivowel-lower9
not-at-end-of-line<
not-at-beginning-of-line>
alpha-numeric-two-byteA
chinese-two-byteC
greek-two-byteG
japanese-hiragana-two-byteH
indian-two-byteI
japanese-katakana-two-byteK
strong-left-to-rightL
korean-hangul-two-byteN
strong-right-to-leftR
cyrillic-two-byteY
combining-diacritic^
asciia
arabicb
chinesec
ethiopice
greekg
koreanh
indiani
japanesej
japanese-katakanak
latinl
laoo
tibetanq
japanese-romanr
thait
vietnamesev
hebreww
cyrillicy
can-break|

For more information about currently defined categories, run the command M-x describe-categories RET. For how to define new categories, see Categories.
Corresponding string regexp: ‘\cchar’ where char is the category character.

Zero-width assertions

These all match the empty string, but only in specific places.

line-start, bol

Match at the beginning of a line.
Corresponding string regexp: ‘^

line-end, eol

Match at the end of a line.
Corresponding string regexp: ‘$

string-start, bos, buffer-start, bot

Match at the start of the string or buffer being matched against.
Corresponding string regexp: ‘\`

string-end, eos, buffer-end, eot

Match at the end of the string or buffer being matched against.
Corresponding string regexp: ‘\'

point

Match at point.
Corresponding string regexp: ‘\=

word-start, bow

Match at the beginning of a word.
Corresponding string regexp: ‘\<

word-end, eow

Match at the end of a word.
Corresponding string regexp: ‘\>

word-boundary

Match at the beginning or end of a word.
Corresponding string regexp: ‘\b

not-word-boundary

Match anywhere but at the beginning or end of a word.
Corresponding string regexp: ‘\B

symbol-start

Match at the beginning of a symbol.
Corresponding string regexp: ‘\_<

symbol-end

Match at the end of a symbol.
Corresponding string regexp: ‘\_>

Capture groups

(group rx…)
(submatch rx…)

Match the rxs, making the matched text and position accessible in the match data. The first group in a regexp is numbered 1; subsequent groups will be numbered one above the previously highest-numbered group in the pattern so far.
Corresponding string regexp: ‘\(…\)

(group-n n rx…)
(submatch-n n rx…)

Like group, but explicitly assign the group number n. n must be positive.
Corresponding string regexp: ‘\(?n:…\)

(backref n)

Match the text previously matched by group number n. n must be in the range 1–9.
Corresponding string regexp: ‘\n

Dynamic inclusion

(literal expr)

Match the literal string that is the result from evaluating the Lisp expression expr. The evaluation takes place at call time, in the current lexical environment.

(regexp expr)
(regex expr)

Match the string regexp that is the result from evaluating the Lisp expression expr. The evaluation takes place at call time, in the current lexical environment.

(eval expr)

Match the rx form that is the result from evaluating the Lisp expression expr. The evaluation takes place at macro-expansion time for rx, at call time for rx-to-string, in the current global environment.


35.3.3.2 Functions and macros using rx regexps

Macro: rx rx-form…

Translate the rx-forms to a string regexp, as if they were the body of a (seq …) form. The rx macro expands to a string constant, or, if literal or regexp forms are used, a Lisp expression that evaluates to a string. Example:

(rx (+ alpha) "=" (+ digit))
  ⇒ "[[:alpha:]]+=[[:digit:]]+"
Function: rx-to-string rx-expr &optional no-group

Translate rx-expr to a string regexp which is returned. If no-group is absent or nil, bracket the result in a non-capturing group, ‘\(?:…\)’, if necessary to ensure that a postfix operator appended to it will apply to the whole expression. Example:

(rx-to-string '(seq (+ alpha) "=" (+ digit)) t)
  ⇒ "[[:alpha:]]+=[[:digit:]]+"

Arguments to literal and regexp forms in rx-expr must be string literals.

The pcase macro can use rx expressions as patterns directly; see rx in pcase.

For mechanisms to add user-defined extensions to the rx notation, see Defining new rx forms.


35.3.3.3 Defining new rx forms

The rx notation can be extended by defining new symbols and parameterized forms in terms of other rx expressions. This is handy for sharing parts between several regexps, and for making complex ones easier to build and understand by putting them together from smaller pieces.

For example, you could define name to mean (one-or-more letter), and (quoted x) to mean (seq ?' x ?') for any x. These forms could then be used in rx expressions like any other: (rx (quoted name)) would match a nonempty sequence of letters inside single quotes.

The Lisp macros below provide different ways of binding names to definitions. Common to all of them are the following rules:

  • Built-in rx forms, like digit and group, cannot be redefined.
  • The definitions live in a name space of their own, separate from that of Lisp variables. There is thus no need to attach a suffix like -regexp to names; they cannot collide with anything else.
  • Definitions cannot refer to themselves recursively, directly or indirectly. If you find yourself needing this, you want a parser, not a regular expression.
  • Definitions are only ever expanded in calls to rx or rx-to-string, not merely by their presence in definition macros. This means that the order of definitions doesn’t matter, even when they refer to each other, and that syntax errors only show up when they are used, not when they are defined.
  • User-defined forms are allowed wherever arbitrary rx expressions are expected; for example, in the body of a zero-or-one form, but not inside any or category forms. They are also allowed inside not and intersection forms.
Macro: rx-define name [arglist] rx-form

Define name globally in all subsequent calls to rx and rx-to-string. If arglist is absent, then name is defined as a plain symbol to be replaced with rx-form. Example:

(rx-define haskell-comment (seq "--" (zero-or-more nonl)))
(rx haskell-comment)
     ⇒ "--.*"

If arglist is present, it must be a list of zero or more argument names, and name is then defined as a parameterized form. When used in an rx expression as (name arg…), each arg will replace the corresponding argument name inside rx-form.

arglist may end in &rest and one final argument name, denoting a rest parameter. The rest parameter will expand to all extra actual argument values not matched by any other parameter in arglist, spliced into rx-form where it occurs. Example:

(rx-define moan (x y &rest r) (seq x (one-or-more y) r "!"))
(rx (moan "MOO" "A" "MEE" "OW"))
     ⇒ "MOOA+MEEOW!"

Since the definition is global, it is recommended to give name a package prefix to avoid name clashes with definitions elsewhere, as is usual when naming non-local variables and functions.

Forms defined this way only perform simple template substitution. For arbitrary computations, use them together with the rx forms eval, regexp or literal. Example:

(defun n-tuple-rx (n element)
  `(seq "<"
        (group-n 1 ,element)
        ,@(mapcar (lambda (i) `(seq ?, (group-n ,i ,element)))
                  (number-sequence 2 n))
        ">"))
(rx-define n-tuple (n element) (eval (n-tuple-rx n 'element)))
(rx (n-tuple 3 (+ (in "0-9"))))
  ⇒ "<\\(?1:[0-9]+\\),\\(?2:[0-9]+\\),\\(?3:[0-9]+\\)>"
Macro: rx-let (bindings…) body…

Make the rx definitions in bindings available locally for rx macro invocations in body, which is then evaluated.

Each element of bindings is on the form (name [arglistrx-form), where the parts have the same meaning as in rx-define above. Example:

(rx-let ((comma-separated (item) (seq item (0+ "," item)))
         (number (1+ digit))
         (numbers (comma-separated number)))
  (re-search-forward (rx "(" numbers ")")))

The definitions are only available during the macro-expansion of body, and are thus not present during execution of compiled code.

rx-let can be used not only inside a function, but also at top level to include global variable and function definitions that need to share a common set of rx forms. Since the names are local inside body, there is no need for any package prefixes. Example:

(rx-let ((phone-number (seq (opt ?+) (1+ (any digit ?-)))))
  (defun find-next-phone-number ()
    (re-search-forward (rx phone-number)))
  (defun phone-number-p (string)
    (string-match-p (rx bos phone-number eos) string)))

The scope of the rx-let bindings is lexical, which means that they are not visible outside body itself, even in functions called from body.

Macro: rx-let-eval bindings body…

Evaluate bindings to a list of bindings as in rx-let, and evaluate body with those bindings in effect for calls to rx-to-string.

This macro is similar to rx-let, except that the bindings argument is evaluated (and thus needs to be quoted if it is a list literal), and the definitions are substituted at run time, which is required for rx-to-string to work. Example:

(rx-let-eval
    '((ponder (x) (seq "Where have all the " x " gone?")))
  (looking-at (rx-to-string
               '(ponder (or "flowers" "young girls"
                            "left socks")))))

Another difference from rx-let is that the bindings are dynamically scoped, and thus also available in functions called from body. However, they are not visible inside functions defined in body.


35.3.4 Regular Expression Functions

These functions operate on regular expressions.

Function: regexp-quote string

This function returns a regular expression whose only exact match is string. Using this regular expression in looking-at will succeed only if the next characters in the buffer are string; using it in a search function will succeed if the text being searched contains string. See Regular Expression Searching.

This allows you to request an exact string match or search when calling a function that wants a regular expression.

(regexp-quote "^The cat$")
     ⇒ "\\^The cat\\$"

One use of regexp-quote is to combine an exact string match with context described as a regular expression. For example, this searches for the string that is the value of string, surrounded by whitespace:

(re-search-forward
 (concat "\\s-" (regexp-quote string) "\\s-"))

The returned string may be string itself if it does not contain any special characters.

Function: regexp-opt strings &optional paren

This function returns an efficient regular expression that will match any of the strings in the list strings. This is useful when you need to make matching or searching as fast as possible—for example, for Font Lock mode27.

If strings is the empty list, the return value is a regexp that never matches anything.

The optional argument paren can be any of the following:

a string

The resulting regexp is preceded by paren and followed by ‘\)’, e.g. use ‘"\\(?1:"’ to produce an explicitly numbered group.

words

The resulting regexp is surrounded by ‘\<\(’ and ‘\)\>’.

symbols

The resulting regexp is surrounded by ‘\_<\(’ and ‘\)\_>’ (this is often appropriate when matching programming-language keywords and the like).

non-nil

The resulting regexp is surrounded by ‘\(’ and ‘\)’.

nil

The resulting regexp is surrounded by ‘\(?:’ and ‘\)’, if it is necessary to ensure that a postfix operator appended to it will apply to the whole expression.

The returned regexp is ordered in such a way that it will always match the longest string possible.

Up to reordering, the resulting regexp of regexp-opt is equivalent to but usually more efficient than that of a simplified version:

(defun simplified-regexp-opt (strings &optional paren)
 (let ((parens
        (cond
         ((stringp paren)       (cons paren "\\)"))
         ((eq paren 'words)    '("\\<\\(" . "\\)\\>"))
         ((eq paren 'symbols) '("\\_<\\(" . "\\)\\_>"))
         ((null paren)          '("\\(?:" . "\\)"))
         (t                       '("\\(" . "\\)")))))
   (concat (car parens)
           (mapconcat 'regexp-quote strings "\\|")
           (cdr parens))))
Function: regexp-opt-depth regexp

This function returns the total number of grouping constructs (parenthesized expressions) in regexp. This does not include shy groups (see Backslash Constructs in Regular Expressions).

Function: regexp-opt-charset chars

This function returns a regular expression matching a character in the list of characters chars.

(regexp-opt-charset '(?a ?b ?c ?d ?e))
     ⇒ "[a-e]"
Variable: regexp-unmatchable

This variable contains a regexp that is guaranteed not to match any string at all. It is particularly useful as default value for variables that may be set to a pattern that actually matches something.


35.3.5 Problems with Regular Expressions

The Emacs regexp implementation, like many of its kind, is generally robust but occasionally causes trouble in either of two ways: matching may run out of internal stack space and signal an error, and it can take a long time to complete. The advice below will make these symptoms less likely and help alleviate problems that do arise.

  • Anchor regexps at the beginning of a line, string or buffer using zero-width assertions (‘^’ and \`). This takes advantage of fast paths in the implementation and can avoid futile matching attempts. Other zero-width assertions may also bring benefits by causing a match to fail early.
  • Avoid or-patterns in favor of character alternatives: write ‘[ab]’ instead of ‘a\|b’. Recall that ‘\s-’ and ‘\sw’ are equivalent to ‘[[:space:]]’ and ‘[[:word:]]’, respectively.
  • Since the last branch of an or-pattern does not add a backtrack point on the stack, consider putting the most likely matched pattern last. For example, ‘^\(?:a\|.b\)*c’ will run out of stack if trying to match a very long string of ‘a’s, but the equivalent ‘^\(?:.b\|a\)*c’ will not.

    (It is a trade-off: successfully matched or-patterns run faster with the most frequently matched pattern first.)

  • Try to ensure that any part of the text can only match in a single way. For example, ‘a*a*’ will match the same set of strings as ‘a*’, but the former can do so in many ways and will therefore cause slow backtracking if the match fails later on. Make or-pattern branches mutually exclusive if possible, so that matching will not go far into more than one branch before failing.

    Be especially careful with nested repetitions: they can easily result in very slow matching in the presence of ambiguities. For example, ‘\(?:a*b*\)+c’ will take a long time attempting to match even a moderately long string of ‘a’s before failing. The equivalent ‘\(?:a\|b\)*c’ is much faster, and ‘[ab]*c’ better still.

  • Don’t use capturing groups unless they are really needed; that is, use ‘\(?:…\)’ instead of ‘\(…\)’ for bracketing purposes.
  • Consider using rx (see The rx Structured Regexp Notation); it can optimize some or-patterns automatically and will never introduce capturing groups unless explicitly requested.

If you run into regexp stack overflow despite following the above advice, don’t be afraid of performing the matching in multiple function calls, each using a simpler regexp where backtracking can more easily be contained.


35.5 POSIX Regular Expression Searching

The usual regular expression functions do backtracking when necessary to handle the ‘\|’ and repetition constructs, but they continue this only until they find some match. Then they succeed and report the first match found.

This section describes alternative search functions which perform the full backtracking specified by the POSIX standard for regular expression matching. They continue backtracking until they have tried all possibilities and found all matches, so they can report the longest match, as required by POSIX. This is much slower, so use these functions only when you really need the longest match.

The POSIX search and match functions do not properly support the non-greedy repetition operators (see non-greedy). This is because POSIX backtracking conflicts with the semantics of non-greedy repetition.

Command: posix-search-forward regexp &optional limit noerror count

This is like re-search-forward except that it performs the full backtracking specified by the POSIX standard for regular expression matching.

Command: posix-search-backward regexp &optional limit noerror count

This is like re-search-backward except that it performs the full backtracking specified by the POSIX standard for regular expression matching.

Function: posix-looking-at regexp &optional inhibit-modify

This is like looking-at except that it performs the full backtracking specified by the POSIX standard for regular expression matching.

Function: posix-string-match regexp string &optional start inhibit-modify

This is like string-match except that it performs the full backtracking specified by the POSIX standard for regular expression matching.


35.6 The Match Data

Emacs keeps track of the start and end positions of the segments of text found during a search; this is called the match data. Thanks to the match data, you can search for a complex pattern, such as a date in a mail message, and then extract parts of the match under control of the pattern.

Because the match data normally describe the most recent search only, you must be careful not to do another search inadvertently between the search you wish to refer back to and the use of the match data. If you can’t avoid another intervening search, you must save and restore the match data around it, to prevent it from being overwritten.

Notice that all functions are allowed to overwrite the match data unless they’re explicitly documented not to do so. A consequence is that functions that are run implicitly in the background (see Timers for Delayed Execution, and Idle Timers) should likely save and restore the match data explicitly.


35.6.1 Replacing the Text that Matched

This function replaces all or part of the text matched by the last search. It works by means of the match data.

Function: replace-match replacement &optional fixedcase literal string subexp

This function performs a replacement operation on a buffer or string.

If you did the last search in a buffer, you should omit the string argument or specify nil for it, and make sure that the current buffer is the one in which you performed the last search. Then this function edits the buffer, replacing the matched text with replacement. It leaves point at the end of the replacement text.

If you performed the last search on a string, pass the same string as string. Then this function returns a new string, in which the matched text is replaced by replacement.

If fixedcase is non-nil, then replace-match uses the replacement text without case conversion; otherwise, it converts the replacement text depending upon the capitalization of the text to be replaced. If the original text is all upper case, this converts the replacement text to upper case. If all words of the original text are capitalized, this capitalizes all the words of the replacement text. If all the words are one-letter and they are all upper case, they are treated as capitalized words rather than all-upper-case words.

If literal is non-nil, then replacement is inserted exactly as it is, the only alterations being case changes as needed. If it is nil (the default), then the character ‘\’ is treated specially. If a ‘\’ appears in replacement, then it must be part of one of the following sequences:

\&

This stands for the entire text being replaced.

\n’, where n is a digit

This stands for the text that matched the nth subexpression in the original regexp. Subexpressions are those expressions grouped inside ‘\(…\)’. If the nth subexpression never matched, an empty string is substituted.

\\

This stands for a single ‘\’ in the replacement text.

\?

This stands for itself (for compatibility with replace-regexp and related commands; see Regexp Replace in The GNU Emacs Manual).

Any other character following ‘\’ signals an error.

The substitutions performed by ‘\&’ and ‘\n’ occur after case conversion, if any. Therefore, the strings they substitute are never case-converted.

If subexp is non-nil, that says to replace just subexpression number subexp of the regexp that was matched, not the entire match. For example, after matching ‘foo \(ba*r\)’, calling replace-match with 1 as subexp means to replace just the text that matched ‘\(ba*r\)’.

Function: match-substitute-replacement replacement &optional fixedcase literal string subexp

This function returns the text that would be inserted into the buffer by replace-match, but without modifying the buffer. It is useful if you want to present the user with actual replacement result, with constructs like ‘\n’ or ‘\&’ substituted with matched groups. Arguments replacement and optional fixedcase, literal, string and subexp have the same meaning as for replace-match.


35.6.2 Simple Match Data Access

This section explains how to use the match data to find out what was matched by the last search or match operation, if it succeeded.

You can ask about the entire matching text, or about a particular parenthetical subexpression of a regular expression. The count argument in the functions below specifies which. If count is zero, you are asking about the entire match. If count is positive, it specifies which subexpression you want.

Recall that the subexpressions of a regular expression are those expressions grouped with escaped parentheses, ‘\(…\)’. The countth subexpression is found by counting occurrences of ‘\(’ from the beginning of the whole regular expression. The first subexpression is numbered 1, the second 2, and so on. Only regular expressions can have subexpressions—after a simple string search, the only information available is about the entire match.

Every successful search sets the match data. Therefore, you should query the match data immediately after searching, before calling any other function that might perform another search. Alternatively, you may save and restore the match data (see Saving and Restoring the Match Data) around the call to functions that could perform another search. Or use the functions that explicitly do not modify the match data; e.g., string-match-p.

A search which fails may or may not alter the match data. In the current implementation, it does not, but we may change it in the future. Don’t try to rely on the value of the match data after a failing search.

Function: match-string count &optional in-string

This function returns, as a string, the text matched in the last search or match operation. It returns the entire text if count is zero, or just the portion corresponding to the countth parenthetical subexpression, if count is positive.

If the last such operation was done against a string with string-match, then you should pass the same string as the argument in-string. After a buffer search or match, you should omit in-string or pass nil for it; but you should make sure that the current buffer when you call match-string is the one in which you did the searching or matching. Failure to follow this advice will lead to incorrect results.

The value is nil if count is out of range, or for a subexpression inside a ‘\|’ alternative that wasn’t used or a repetition that repeated zero times.

Function: match-string-no-properties count &optional in-string

This function is like match-string except that the result has no text properties.

Function: match-beginning count

If the last regular expression search found a match, this function returns the position of the start of the matching text or of a subexpression of it.

If count is zero, then the value is the position of the start of the entire match. Otherwise, count specifies a subexpression in the regular expression, and the value of the function is the starting position of the match for that subexpression.

The value is nil for a subexpression inside a ‘\|’ alternative that wasn’t used or a repetition that repeated zero times.

Function: match-end count

This function is like match-beginning except that it returns the position of the end of the match, rather than the position of the beginning.

Here is an example of using the match data, with a comment showing the positions within the text:

(string-match "\\(qu\\)\\(ick\\)"
              "The quick fox jumped quickly.")
              ;0123456789
     ⇒ 4

(match-string 0 "The quick fox jumped quickly.")
     ⇒ "quick"
(match-string 1 "The quick fox jumped quickly.")
     ⇒ "qu"
(match-string 2 "The quick fox jumped quickly.")
     ⇒ "ick"

(match-beginning 1)       ; The beginning of the match
     ⇒ 4                 ;   with ‘qu’ is at index 4.

(match-beginning 2)       ; The beginning of the match
     ⇒ 6                 ;   with ‘ick’ is at index 6.

(match-end 1)             ; The end of the match
     ⇒ 6                 ;   with ‘qu’ is at index 6.

(match-end 2)             ; The end of the match
     ⇒ 9                 ;   with ‘ick’ is at index 9.

Here is another example. Point is initially located at the beginning of the line. Searching moves point to between the space and the word ‘in’. The beginning of the entire match is at the 9th character of the buffer (‘T’), and the beginning of the match for the first subexpression is at the 13th character (‘c’).

(list
  (re-search-forward "The \\(cat \\)")
  (match-beginning 0)
  (match-beginning 1))
    ⇒ (17 9 13)

---------- Buffer: foo ----------
I read "The cat ∗in the hat comes back" twice.
        ^   ^
        9  13
---------- Buffer: foo ----------

(In this case, the index returned is a buffer position; the first character of the buffer counts as 1.)


35.6.3 Accessing the Entire Match Data

The functions match-data and set-match-data read or write the entire match data, all at once.

Function: match-data &optional integers reuse reseat

This function returns a list of positions (markers or integers) that record all the information on the text that the last search matched. Element zero is the position of the beginning of the match for the whole expression; element one is the position of the end of the match for the expression. The next two elements are the positions of the beginning and end of the match for the first subexpression, and so on. In general, element number 2n corresponds to (match-beginning n); and element number 2n + 1 corresponds to (match-end n).

Normally all the elements are markers or nil, but if integers is non-nil, that means to use integers instead of markers. (In that case, the buffer itself is appended as an additional element at the end of the list, to facilitate complete restoration of the match data.) If the last match was done on a string with string-match, then integers are always used, since markers can’t point into a string.

If reuse is non-nil, it should be a list. In that case, match-data stores the match data in reuse. That is, reuse is destructively modified. reuse does not need to have the right length. If it is not long enough to contain the match data, it is extended. If it is too long, the length of reuse stays the same, but the elements that were not used are set to nil. The purpose of this feature is to reduce the need for garbage collection.

If reseat is non-nil, all markers on the reuse list are reseated to point to nowhere.

As always, there must be no possibility of intervening searches between the call to a search function and the call to match-data that is intended to access the match data for that search.

(match-data)
     ⇒  (#<marker at 9 in foo>
          #<marker at 17 in foo>
          #<marker at 13 in foo>
          #<marker at 17 in foo>)
Function: set-match-data match-list &optional reseat

This function sets the match data from the elements of match-list, which should be a list that was the value of a previous call to match-data. (More precisely, anything that has the same format will work.)

If match-list refers to a buffer that doesn’t exist, you don’t get an error; that sets the match data in a meaningless but harmless way.

If reseat is non-nil, all markers on the match-list list are reseated to point to nowhere.

store-match-data is a semi-obsolete alias for set-match-data.


35.6.4 Saving and Restoring the Match Data

When you call a function that may search, you may need to save and restore the match data around that call, if you want to preserve the match data from an earlier search for later use. Here is an example that shows the problem that arises if you fail to save the match data:

(re-search-forward "The \\(cat \\)")
     ⇒ 48
(foo)                   ; foo does more searching.
(match-end 0)
     ⇒ 61              ; Unexpected result—not 48!

You can save and restore the match data with save-match-data:

Macro: save-match-data body…

This macro executes body, saving and restoring the match data around it. The return value is the value of the last form in body.

You could use set-match-data together with match-data to imitate the effect of the special form save-match-data. Here is how:

(let ((data (match-data)))
  (unwind-protect
      …   ; Ok to change the original match data.
    (set-match-data data)))

Emacs automatically saves and restores the match data when it runs process filter functions (see Process Filter Functions) and process sentinels (see Sentinels: Detecting Process Status Changes).


35.7 Search and Replace

If you want to find all matches for a regexp in part of the buffer and replace them, the most flexible way is to write an explicit loop using re-search-forward and replace-match, like this:

(while (re-search-forward "foo[ \t]+bar" nil t)
  (replace-match "foobar"))

See Replacing the Text that Matched, for a description of replace-match.

It may be more convenient to limit the replacements to a specific region. The function replace-regexp-in-region does that.

Function: replace-regexp-in-region regexp replacement &optional start end

This function replaces all the occurrences of regexp with replacement in the region of buffer text between start and end; start defaults to position of point, and end defaults to the last accessible position of the buffer. The search for regexp is case-sensitive, and replacement is inserted without changing its letter-case. The replacement string can use the same special elements starting with ‘\’ as replace-match does. The function returns the number of replaced occurrences, or nil if regexp is not found. The function preserves the position of point.

(replace-regexp-in-region "foo[ \t]+bar" "foobar")
Function: replace-string-in-region string replacement &optional start end

This function works similarly to replace-regexp-in-region, but searches for, and replaces, literal strings instead of regular expressions.

Emacs also has special functions for replacing matches in a string.

Function: replace-regexp-in-string regexp rep string &optional fixedcase literal subexp start

This function copies string and searches it for matches for regexp, and replaces them with rep. It returns the modified copy. If start is non-nil, the search for matches starts at that index in string, and the returned value does not include the first start characters of string. To get the whole transformed string, concatenate the first start characters of string with the return value.

This function uses replace-match to do the replacement, and it passes the optional arguments fixedcase, literal and subexp along to replace-match.

Instead of a string, rep can be a function. In that case, replace-regexp-in-string calls rep for each match, passing the text of the match as its sole argument. It collects the value rep returns and passes that to replace-match as the replacement string. The match data at this point are the result of matching regexp against a substring of string.

Function: string-replace from-string to-string in-string

This function replaces all occurrences of from-string with to-string in in-string and returns the result. It may return one of its arguments unchanged, a constant string or a new string. Case is significant, and text properties are ignored.

If you want to write a command along the lines of query-replace, you can use perform-replace to do the work.

Function: perform-replace from-string replacements query-flag regexp-flag delimited-flag &optional repeat-count map start end backward region-noncontiguous-p

This function is the guts of query-replace and related commands. It searches for occurrences of from-string in the text between positions start and end and replaces some or all of them. If start is nil (or omitted), point is used instead, and the end of the buffer’s accessible portion is used for end. (If the optional argument backward is non-nil, the search starts at end and goes backward.)

If query-flag is nil, it replaces all occurrences; otherwise, it asks the user what to do about each one.

If regexp-flag is non-nil, then from-string is considered a regular expression; otherwise, it must match literally. If delimited-flag is non-nil, then only replacements surrounded by word boundaries are considered.

The argument replacements specifies what to replace occurrences with. If it is a string, that string is used. It can also be a list of strings, to be used in cyclic order.

If replacements is a cons cell, (function . data), this means to call function after each match to get the replacement text. This function is called with two arguments: data, and the number of replacements already made.

If repeat-count is non-nil, it should be an integer. Then it specifies how many times to use each of the strings in the replacements list before advancing cyclically to the next one.

If from-string contains upper-case letters, then perform-replace binds case-fold-search to nil, and it uses the replacements without altering their case.

Normally, the keymap query-replace-map defines the possible user responses for queries. The argument map, if non-nil, specifies a keymap to use instead of query-replace-map.

Non-nil region-noncontiguous-p means that the region between start and end is composed of noncontiguous pieces. The most common example of this is a rectangular region, where the pieces are separated by newline characters.

This function uses one of two functions to search for the next occurrence of from-string. These functions are specified by the values of two variables: replace-re-search-function and replace-search-function. The former is called when the argument regexp-flag is non-nil, the latter when it is nil.

Variable: query-replace-map

This variable holds a special keymap that defines the valid user responses for perform-replace and the commands that use it, as well as y-or-n-p and map-y-or-n-p. This map is unusual in two ways:

  • The key bindings are not commands, just symbols that are meaningful to the functions that use this map.
  • Prefix keys are not supported; each key binding must be for a single-event key sequence. This is because the functions don’t use read-key-sequence to get the input; instead, they read a single event and look it up “by hand”.

Here are the meaningful bindings for query-replace-map. Several of them are meaningful only for query-replace and friends.

act

Do take the action being considered—in other words, “yes”.

skip

Do not take action for this question—in other words, “no”.

exit

Answer this question “no”, and give up on the entire series of questions, assuming that the answers will be “no”.

exit-prefix

Like exit, but add the key that was pressed to unread-command-events (see Miscellaneous Event Input Features).

act-and-exit

Answer this question “yes”, and give up on the entire series of questions, assuming that subsequent answers will be “no”.

act-and-show

Answer this question “yes”, but show the results—don’t advance yet to the next question.

automatic

Answer this question and all subsequent questions in the series with “yes”, without further user interaction.

backup

Move back to the previous place that a question was asked about.

undo

Undo last replacement and move back to the place where that replacement was performed.

undo-all

Undo all replacements and move back to the place where the first replacement was performed.

edit

Enter a recursive edit to deal with this question—instead of any other action that would normally be taken.

edit-replacement

Edit the replacement for this question in the minibuffer.

delete-and-edit

Delete the text being considered, then enter a recursive edit to replace it.

recenter
scroll-up
scroll-down
scroll-other-window
scroll-other-window-down

Perform the specified window scroll operation, then ask the same question again. Only y-or-n-p and related functions use this answer.

quit

Perform a quit right away. Only y-or-n-p and related functions use this answer.

help

Display some help, then ask again.

Variable: multi-query-replace-map

This variable holds a keymap that extends query-replace-map by providing additional key bindings that are useful in multi-buffer replacements. The additional bindings are:

automatic-all

Answer this question and all subsequent questions in the series with “yes”, without further user interaction, for all remaining buffers.

exit-current

Answer this question “no”, and give up on the entire series of questions for the current buffer. Continue to the next buffer in the sequence.

Variable: replace-search-function

This variable specifies a function that perform-replace calls to search for the next string to replace. Its default value is search-forward. Any other value should name a function of 3 arguments: the first 3 arguments of search-forward (see Searching for Strings).

Variable: replace-re-search-function

This variable specifies a function that perform-replace calls to search for the next regexp to replace. Its default value is re-search-forward. Any other value should name a function of 3 arguments: the first 3 arguments of re-search-forward (see Regular Expression Searching).


35.8 Standard Regular Expressions Used in Editing

This section describes some variables that hold regular expressions used for certain purposes in editing:

User Option: page-delimiter

This is the regular expression describing line-beginnings that separate pages. The default value is "^\014" (i.e., "^^L" or "^\C-l"); this matches a line that starts with a formfeed character.

The following two regular expressions should not assume the match always starts at the beginning of a line; they should not use ‘^’ to anchor the match. Most often, the paragraph commands do check for a match only at the beginning of a line, which means that ‘^’ would be superfluous. When there is a nonzero left margin, they accept matches that start after the left margin. In that case, a ‘^’ would be incorrect. However, a ‘^’ is harmless in modes where a left margin is never used.

User Option: paragraph-separate

This is the regular expression for recognizing the beginning of a line that separates paragraphs. (If you change this, you may have to change paragraph-start also.) The default value is "[ \t\f]*$", which matches a line that consists entirely of spaces, tabs, and form feeds (after its left margin).

User Option: paragraph-start

This is the regular expression for recognizing the beginning of a line that starts or separates paragraphs. The default value is "\f\\|[ \t]*$", which matches a line containing only whitespace or starting with a form feed (after its left margin).

User Option: sentence-end

If non-nil, the value should be a regular expression describing the end of a sentence, including the whitespace following the sentence. (All paragraph boundaries also end sentences, regardless.)

If the value is nil, as it is by default, then the function sentence-end constructs the regexp. That is why you should always call the function sentence-end to obtain the regexp to be used to recognize the end of a sentence.

Function: sentence-end

This function returns the value of the variable sentence-end, if non-nil. Otherwise it returns a default value based on the values of the variables sentence-end-double-space (see Definition of sentence-end-double-space), sentence-end-without-period, and sentence-end-without-space.


36 Syntax Tables

A syntax table specifies the syntactic role of each character in a buffer. It can be used to determine where words, symbols, and other syntactic constructs begin and end. This information is used by many Emacs facilities, including Font Lock mode (see Font Lock Mode) and the various complex movement commands (see Motion).


36.1 Syntax Table Concepts

A syntax table is a data structure which can be used to look up the syntax class and other syntactic properties of each character. Syntax tables are used by Lisp programs for scanning and moving across text.

Internally, a syntax table is a char-table (see Char-Tables). The element at index c describes the character with code c; its value is a cons cell which specifies the syntax of the character in question. See Syntax Table Internals, for details. However, instead of using aset and aref to modify and inspect syntax table contents, you should usually use the higher-level functions char-syntax and modify-syntax-entry, which are described in Syntax Table Functions.

Function: syntax-table-p object

This function returns t if object is a syntax table.

Each buffer has its own major mode, and each major mode has its own idea of the syntax class of various characters. For example, in Lisp mode, the character ‘;’ begins a comment, but in C mode, it terminates a statement. To support these variations, the syntax table is local to each buffer. Typically, each major mode has its own syntax table, which it installs in all buffers that use that mode. For example, the variable emacs-lisp-mode-syntax-table holds the syntax table used by Emacs Lisp mode, and c-mode-syntax-table holds the syntax table used by C mode. Changing a major mode’s syntax table alters the syntax in all of that mode’s buffers, as well as in any buffers subsequently put in that mode. Occasionally, several similar modes share one syntax table. See Major Mode Examples, for an example of how to set up a syntax table.

A syntax table can inherit from another syntax table, which is called its parent syntax table. A syntax table can leave the syntax class of some characters unspecified, by giving them the “inherit” syntax class; such a character then acquires the syntax class specified by the parent syntax table (see Table of Syntax Classes). Emacs defines a standard syntax table, which is the default parent syntax table, and is also the syntax table used by Fundamental mode.

Function: standard-syntax-table

This function returns the standard syntax table, which is the syntax table used in Fundamental mode.

Syntax tables are not used by the Emacs Lisp reader, which has its own built-in syntactic rules which cannot be changed. (Some Lisp systems provide ways to redefine the read syntax, but we decided to leave this feature out of Emacs Lisp for simplicity.)


36.2 Syntax Descriptors

The syntax class of a character describes its syntactic role. Each syntax table specifies the syntax class of each character. There is no necessary relationship between the class of a character in one syntax table and its class in any other table.

Each syntax class is designated by a mnemonic character, which serves as the name of the class when you need to specify a class. Usually, this designator character is one that is often assigned that class; however, its meaning as a designator is unvarying and independent of what syntax that character currently has. Thus, ‘\’ as a designator character always stands for escape character syntax, regardless of whether the ‘\’ character actually has that syntax in the current syntax table. See Table of Syntax Classes, for a list of syntax classes and their designator characters.

A syntax descriptor is a Lisp string that describes the syntax class and other syntactic properties of a character. When you want to modify the syntax of a character, that is done by calling the function modify-syntax-entry and passing a syntax descriptor as one of its arguments (see Syntax Table Functions).

The first character in a syntax descriptor must be a syntax class designator character. The second character, if present, specifies a matching character (e.g., in Lisp, the matching character for ‘(’ is ‘)’); a space specifies that there is no matching character. Then come characters specifying additional syntax properties (see Syntax Flags).

If no matching character or flags are needed, only one character (specifying the syntax class) is sufficient.

For example, the syntax descriptor for the character ‘*’ in C mode is ". 23" (i.e., punctuation, matching character slot unused, second character of a comment-starter, first character of a comment-ender), and the entry for ‘/’ is ‘. 14’ (i.e., punctuation, matching character slot unused, first character of a comment-starter, second character of a comment-ender).

Emacs also defines raw syntax descriptors, which are used to describe syntax classes at a lower level. See Syntax Table Internals.


36.2.1 Table of Syntax Classes

Here is a table of syntax classes, the characters that designate them, their meanings, and examples of their use.

Whitespace characters: ‘ ’ or ‘-

Characters that separate symbols and words from each other. Typically, whitespace characters have no other syntactic significance, and multiple whitespace characters are syntactically equivalent to a single one. Space, tab, and formfeed are classified as whitespace in almost all major modes.

This syntax class can be designated by either ‘  or ‘-’. Both designators are equivalent.

Word constituents: ‘w

Parts of words in human languages. These are typically used in variable and command names in programs. All upper- and lower-case letters, and the digits, are typically word constituents.

Symbol constituents: ‘_

Extra characters used in variable and command names along with word constituents. Examples include the characters ‘$&*+-_<>’ in Lisp mode, which may be part of a symbol name even though they are not part of English words. In standard C, the only non-word-constituent character that is valid in symbols is underscore (‘_’).

Punctuation characters: ‘.

Characters used as punctuation in a human language, or used in a programming language to separate symbols from one another. Some programming language modes, such as Emacs Lisp mode, have no characters in this class since the few characters that are not symbol or word constituents all have other uses. Other programming language modes, such as C mode, use punctuation syntax for operators.

Open parenthesis characters: ‘(
Close parenthesis characters: ‘)

Characters used in dissimilar pairs to surround sentences or expressions. Such a grouping is begun with an open parenthesis character and terminated with a close. Each open parenthesis character matches a particular close parenthesis character, and vice versa. Normally, Emacs indicates momentarily the matching open parenthesis when you insert a close parenthesis. See Blinking Parentheses.

In human languages, and in C code, the parenthesis pairs are ‘()’, ‘[]’, and ‘{}’. In Emacs Lisp, the delimiters for lists and vectors (‘()’ and ‘[]’) are classified as parenthesis characters.

String quotes: ‘"

Characters used to delimit string constants. The same string quote character appears at the beginning and the end of a string. Such quoted strings do not nest.

The parsing facilities of Emacs consider a string as a single token. The usual syntactic meanings of the characters in the string are suppressed.

The Lisp modes have two string quote characters: double-quote (‘"’) and vertical bar (‘|’). ‘|’ is not used in Emacs Lisp, but it is used in Common Lisp. C also has two string quote characters: double-quote for strings, and apostrophe (‘'’) for character constants.

Human text has no string quote characters. We do not want quotation marks to turn off the usual syntactic properties of other characters in the quotation.

Escape-syntax characters: ‘\

Characters that start an escape sequence, such as is used in string and character constants. The character ‘\’ belongs to this class in both C and Lisp. (In C, it is used thus only inside strings, but it turns out to cause no trouble to treat it this way throughout C code.)

Characters in this class count as part of words if words-include-escapes is non-nil. See Motion by Words.

Character quotes: ‘/

Characters used to quote the following character so that it loses its normal syntactic meaning. This differs from an escape character in that only the character immediately following is ever affected.

Characters in this class count as part of words if words-include-escapes is non-nil. See Motion by Words.

This class is used for backslash in TeX mode.

Paired delimiters: ‘$

Similar to string quote characters, except that the syntactic properties of the characters between the delimiters are not suppressed. Only TeX mode uses a paired delimiter presently—the ‘$’ that both enters and leaves math mode.

Expression prefixes: ‘'

Characters used for syntactic operators that are considered as part of an expression if they appear next to one. In Lisp modes, these characters include the apostrophe, ‘'’ (used for quoting), the comma, ‘,’ (used in macros), and ‘#’ (used in the read syntax for certain data types).

Comment starters: ‘<
Comment enders: ‘>

Characters used in various languages to delimit comments. Human text has no comment characters. In Lisp, the semicolon (‘;’) starts a comment and a newline or formfeed ends one.

Inherit standard syntax: ‘@

This syntax class does not specify a particular syntax. It says to look in the parent syntax table to find the syntax of this character.

Generic comment delimiters: ‘!

(This syntax class is also known as “comment-fence”.) Characters that start or end a special kind of comment. Any generic comment delimiter matches any generic comment delimiter, but they cannot match a comment starter or comment ender; generic comment delimiters can only match each other.

This syntax class is primarily meant for use with the syntax-table text property (see Syntax Properties). You can mark any range of characters as forming a comment, by giving the first and last characters of the range syntax-table properties identifying them as generic comment delimiters.

Generic string delimiters: ‘|

(This syntax class is also known as “string-fence”.) Characters that start or end a string. This class differs from the string quote class in that any generic string delimiter can match any other generic string delimiter; but they do not match ordinary string quote characters.

This syntax class is primarily meant for use with the syntax-table text property (see Syntax Properties). You can mark any range of characters as forming a string constant, by giving the first and last characters of the range syntax-table properties identifying them as generic string delimiters.


36.2.2 Syntax Flags

In addition to the classes, entries for characters in a syntax table can specify flags. There are eight possible flags, represented by the characters ‘1’, ‘2’, ‘3’, ‘4’, ‘b’, ‘c’, ‘n’, and ‘p’.

All the flags except ‘p’ are used to describe comment delimiters. The digit flags are used for comment delimiters made up of 2 characters. They indicate that a character can also be part of a comment sequence, in addition to the syntactic properties associated with its character class. The flags are independent of the class and each other for the sake of characters such as ‘*’ in C mode, which is a punctuation character, and the second character of a start-of-comment sequence (‘/*’), and the first character of an end-of-comment sequence (‘*/’). The flags ‘b’, ‘c’, and ‘n’ are used to qualify the corresponding comment delimiter.

Here is a table of the possible flags for a character c, and what they mean:

  • 1’ means c is the start of a two-character comment-start sequence.
  • 2’ means c is the second character of such a sequence.
  • 3’ means c is the start of a two-character comment-end sequence.
  • 4’ means c is the second character of such a sequence.
  • b’ means that c as a comment delimiter belongs to the alternative “b” comment style. For a two-character comment starter, this flag is only significant on the second char, and for a 2-character comment ender it is only significant on the first char.
  • c’ means that c as a comment delimiter belongs to the alternative “c” comment style. For a two-character comment delimiter, ‘c’ on either character makes it of style “c”.
  • n’ on a comment delimiter character specifies that this kind of comment can be nested. Inside such a comment, only comments of the same style will be recognized. For a two-character comment delimiter, ‘n’ on either character makes it nestable.

    Emacs supports several comment styles simultaneously in any one syntax table. A comment style is a set of flags ‘b’, ‘c’, and ‘n’, so there can be up to 8 different comment styles, each one named by the set of its flags. Each comment delimiter has a style and only matches comment delimiters of the same style. Thus if a comment starts with the comment-start sequence of style “bn”, it will extend until the next matching comment-end sequence of style “bn”. When the set of flags has neither flag ‘b’ nor flag ‘c’ set, the resulting style is called the “a” style.

    The appropriate comment syntax settings for C++ can be as follows:

    /

    124

    *

    23b

    newline

    >

    This defines four comment-delimiting sequences:

    /*

    This is a comment-start sequence for “b” style because the second character, ‘*’, has the ‘b’ flag.

    //

    This is a comment-start sequence for “a” style because the second character, ‘/’, does not have the ‘b’ flag.

    */

    This is a comment-end sequence for “b” style because the first character, ‘*’, has the ‘b’ flag.

    newline

    This is a comment-end sequence for “a” style, because the newline character does not have the ‘b’ flag.

  • p’ identifies an additional prefix character for Lisp syntax. These characters are treated as whitespace when they appear between expressions. When they appear within an expression, they are handled according to their usual syntax classes.

    The function backward-prefix-chars moves back over these characters, as well as over characters whose primary syntax class is prefix (‘'’). See Motion and Syntax.


36.3 Syntax Table Functions

In this section we describe functions for creating, accessing and altering syntax tables.

Function: make-syntax-table &optional table

This function creates a new syntax table. If table is non-nil, the parent of the new syntax table is table; otherwise, the parent is the standard syntax table.

In the new syntax table, all characters are initially given the “inherit” (‘@’) syntax class, i.e., their syntax is inherited from the parent table (see Table of Syntax Classes).

Function: copy-syntax-table &optional table

This function constructs a copy of table and returns it. If table is omitted or nil, it returns a copy of the standard syntax table. Otherwise, an error is signaled if table is not a syntax table.

Command: modify-syntax-entry char syntax-descriptor &optional table

This function sets the syntax entry for char according to syntax-descriptor. char must be a character, or a cons cell of the form (min . max); in the latter case, the function sets the syntax entries for all characters in the range between min and max, inclusive.

The syntax is changed only for table, which defaults to the current buffer’s syntax table, and not in any other syntax table.

The argument syntax-descriptor is a syntax descriptor, i.e., a string whose first character is a syntax class designator and whose second and subsequent characters optionally specify a matching character and syntax flags. See Syntax Descriptors. An error is signaled if syntax-descriptor is not a valid syntax descriptor.

This function always returns nil. The old syntax information in the table for this character is discarded.

Examples:

;; Put the space character in class whitespace.
(modify-syntax-entry ?\s " ")
     ⇒ nil

;; Make ‘$’ an open parenthesis character,
;;   with ‘^’ as its matching close.
(modify-syntax-entry ?$ "(^")
     ⇒ nil

;; Make ‘^’ a close parenthesis character,
;;   with ‘$’ as its matching open.
(modify-syntax-entry ?^ ")$")
     ⇒ nil

;; Make ‘/’ a punctuation character,
;;   the first character of a start-comment sequence,
;;   and the second character of an end-comment sequence.
;;   This is used in C mode.
(modify-syntax-entry ?/ ". 14")
     ⇒ nil
Function: char-syntax character

This function returns the syntax class of character, represented by its designator character (see Table of Syntax Classes). This returns only the class, not its matching character or syntax flags.

The following examples apply to C mode. (We use string to make it easier to see the character returned by char-syntax.)

;; Space characters have whitespace syntax class.
(string (char-syntax ?\s))
     ⇒ " "

;; Forward slash characters have punctuation syntax.
;; Note that this char-syntax call does not reveal
;; that it is also part of comment-start and -end sequences.
(string (char-syntax ?/))
     ⇒ "."

;; Open parenthesis characters have open parenthesis syntax.
;; Note that this char-syntax call does not reveal that
;; it has a matching character, ‘)’.
(string (char-syntax ?\())
     ⇒ "("
Function: set-syntax-table table

This function makes table the syntax table for the current buffer. It returns table.

Function: syntax-table

This function returns the current syntax table, which is the table for the current buffer.

Command: describe-syntax &optional buffer

This command displays the contents of the syntax table of buffer (by default, the current buffer) in a help buffer.

Macro: with-syntax-table table body…

This macro executes body using table as the current syntax table. It returns the value of the last form in body, after restoring the old current syntax table.

Since each buffer has its own current syntax table, we should make that more precise: with-syntax-table temporarily alters the current syntax table of whichever buffer is current at the time the macro execution starts. Other buffers are not affected.


36.4 Syntax Properties

When the syntax table is not flexible enough to specify the syntax of a language, you can override the syntax table for specific character occurrences in the buffer, by applying a syntax-table text property. See Text Properties, for how to apply text properties.

The valid values of syntax-table text property are:

syntax-table

If the property value is a syntax table, that table is used instead of the current buffer’s syntax table to determine the syntax for the underlying text character.

(syntax-code . matching-char)

A cons cell of this format is a raw syntax descriptor (see Syntax Table Internals), which directly specifies a syntax class for the underlying text character.

nil

If the property is nil, the character’s syntax is determined from the current syntax table in the usual way.

Variable: parse-sexp-lookup-properties

If this is non-nil, the syntax scanning functions, like forward-sexp, pay attention to syntax-table text properties. Otherwise they use only the current syntax table.

Variable: syntax-propertize-function

This variable, if non-nil, should store a function for applying syntax-table properties to a specified stretch of text. It is intended to be used by major modes to install a function which applies syntax-table properties in some mode-appropriate way.

The function is called by syntax-ppss (see Finding the Parse State for a Position), and by Font Lock mode during syntactic fontification (see Syntactic Font Lock). It is called with two arguments, start and end, which are the starting and ending positions of the text on which it should act. It is allowed to arbitrarily move point within the region delimited by start and end; such motions don’t need to use save-excursion (see Excursions). It is also allowed to call syntax-ppss on any position before end, but if a Lisp program calls syntax-ppss on some position and later modifies the buffer at some earlier position, then it is that program’s responsibility to call syntax-ppss-flush-cache to flush the now obsolete info from the cache.

Caution: When this variable is non-nil, Emacs removes syntax-table text properties arbitrarily and relies on syntax-propertize-function to reapply them. Thus if this facility is used at all, the function must apply all syntax-table text properties used by the major mode. In particular, Modes derived from a CC Mode mode must not use this variable, since CC Mode uses other means to apply and remove these text properties.

Variable: syntax-propertize-extend-region-functions

This abnormal hook is run by the syntax parsing code prior to calling syntax-propertize-function. Its role is to help locate safe starting and ending buffer positions for passing to syntax-propertize-function. For example, a major mode can add a function to this hook to identify multi-line syntactic constructs, and ensure that the boundaries do not fall in the middle of one.

Each function in this hook should accept two arguments, start and end. It should return either a cons cell of two adjusted buffer positions, (new-start . new-end), or nil if no adjustment is necessary. The hook functions are run in turn, repeatedly, until they all return nil.


36.5 Motion and Syntax

This section describes functions for moving across characters that have certain syntax classes.

Function: skip-syntax-forward syntaxes &optional limit

This function moves point forward across characters having syntax classes mentioned in syntaxes (a string of syntax class characters). It stops when it encounters the end of the buffer, or position limit (if specified), or a character it is not supposed to skip.

If syntaxes starts with ‘^’, then the function skips characters whose syntax is not in syntaxes.

The return value is the distance traveled, which is a nonnegative integer.

Function: skip-syntax-backward syntaxes &optional limit

This function moves point backward across characters whose syntax classes are mentioned in syntaxes. It stops when it encounters the beginning of the buffer, or position limit (if specified), or a character it is not supposed to skip.

If syntaxes starts with ‘^’, then the function skips characters whose syntax is not in syntaxes.

The return value indicates the distance traveled. It is an integer that is zero or less.

Function: backward-prefix-chars

This function moves point backward over any number of characters with expression prefix syntax. This includes both characters in the expression prefix syntax class, and characters with the ‘p’ flag.


36.6 Parsing Expressions

This section describes functions for parsing and scanning balanced expressions. We will refer to such expressions as sexps, following the terminology of Lisp, even though these functions can act on languages other than Lisp. Basically, a sexp is either a balanced parenthetical grouping, a string, or a symbol (i.e., a sequence of characters whose syntax is either word constituent or symbol constituent). However, characters in the expression prefix syntax class (see Table of Syntax Classes) are treated as part of the sexp if they appear next to it.

The syntax table controls the interpretation of characters, so these functions can be used for Lisp expressions when in Lisp mode and for C expressions when in C mode. See Moving over Balanced Expressions, for convenient higher-level functions for moving over balanced expressions.

A character’s syntax controls how it changes the state of the parser, rather than describing the state itself. For example, a string delimiter character toggles the parser state between in-string and in-code, but the syntax of characters does not directly say whether they are inside a string. For example (note that 15 is the syntax code for generic string delimiters),

(put-text-property 1 9 'syntax-table '(15 . nil))

does not tell Emacs that the first eight chars of the current buffer are a string, but rather that they are all string delimiters. As a result, Emacs treats them as four consecutive empty string constants.


36.6.1 Motion Commands Based on Parsing

This section describes simple point-motion functions that operate based on parsing expressions.

Function: scan-lists from count depth

This function scans forward count balanced parenthetical groupings from position from. It returns the position where the scan stops. If count is negative, the scan moves backwards.

If depth is nonzero, treat the starting position as being depth parentheses deep. The scanner moves forward or backward through the buffer until the depth changes to zero count times. Hence, a positive value for depth has the effect of moving out depth levels of parenthesis from the starting position, while a negative depth has the effect of moving deeper by -depth levels of parenthesis.

Scanning ignores comments if parse-sexp-ignore-comments is non-nil.

If the scan reaches the beginning or end of the accessible part of the buffer before it has scanned over count parenthetical groupings, the return value is nil if the depth at that point is zero; if the depth is non-zero, a scan-error error is signaled.

Function: scan-sexps from count

This function scans forward count sexps from position from. It returns the position where the scan stops. If count is negative, the scan moves backwards.

Scanning ignores comments if parse-sexp-ignore-comments is non-nil.

If the scan reaches the beginning or end of (the accessible part of) the buffer while in the middle of a parenthetical grouping, an error is signaled. If it reaches the beginning or end between groupings but before count is used up, nil is returned.

Function: forward-comment count

This function moves point forward across count complete comments (that is, including the starting delimiter and the terminating delimiter if any), plus any whitespace encountered on the way. It moves backward if count is negative. If it encounters anything other than a comment or whitespace, it stops, leaving point at the place where it stopped. This includes (for instance) finding the end of a comment when moving forward and expecting the beginning of one. The function also stops immediately after moving over the specified number of complete comments. If count comments are found as expected, with nothing except whitespace between them, it returns t; otherwise it returns nil.

This function cannot tell whether the comments it traverses are embedded within a string. If they look like comments, it treats them as comments.

To move forward over all comments and whitespace following point, use (forward-comment (buffer-size)). (buffer-size) is a good argument to use, because the number of comments in the buffer cannot exceed that many.


36.6.2 Finding the Parse State for a Position

For syntactic analysis, such as in indentation, often the useful thing is to compute the syntactic state corresponding to a given buffer position. This function does that conveniently.

Function: syntax-ppss &optional pos

This function returns the parser state that the parser would reach at position pos starting from the beginning of the visible portion of the buffer. See Parser State, for a description of the parser state.

The return value is the same as if you call the low-level parsing function parse-partial-sexp to parse from the beginning of the visible portion of the buffer to pos (see Low-Level Parsing). However, syntax-ppss uses caches to speed up the computation. Due to this optimization, the second value (previous complete subexpression) and sixth value (minimum parenthesis depth) in the returned parser state are not meaningful.

This function has a side effect: it adds a buffer-local entry to before-change-functions (see Change Hooks) for syntax-ppss-flush-cache (see below). This entry keeps the cache consistent as the buffer is modified. However, the cache might not be updated if syntax-ppss is called while before-change-functions is temporarily let-bound, or if the buffer is modified without running the hook, such as when using inhibit-modification-hooks. In those cases, it is necessary to call syntax-ppss-flush-cache explicitly.

Function: syntax-ppss-flush-cache beg &rest ignored-args

This function flushes the cache used by syntax-ppss, starting at position beg. The remaining arguments, ignored-args, are ignored; this function accepts them so that it can be directly used on hooks such as before-change-functions (see Change Hooks).


36.6.3 Parser State

A parser state is a list of (currently) eleven elements describing the state of the syntactic parser, after it parses the text between a specified starting point and a specified end point in the buffer using parse-partial-sexp (see Low-Level Parsing). Parsing functions such as syntax-ppss (see Finding the Parse State for a Position) also return a parser state as the value. parse-partial-sexp can accept a parser state as an argument, for resuming parsing.

Here are the meanings of the elements of the parser state:

  1. The depth in parentheses, counting from 0. Warning: this can be negative if there are more close parens than open parens between the parser’s starting point and end point.
  2. The character position of the start of the innermost parenthetical grouping containing the stopping point; nil if none.
  3. The character position of the start of the last complete subexpression terminated; nil if none.
  4. Non-nil if inside a string. More precisely, this is the character that will terminate the string, or t if a generic string delimiter character should terminate it.
  5. t if inside a non-nestable comment (of any comment style; see Syntax Flags); or the comment nesting level if inside a comment that can be nested.
  6. t if the end point is just after a quote character.
  7. The minimum parenthesis depth encountered during this scan.
  8. What kind of comment is active: nil if not in a comment or in a comment of style ‘a’; 1 for a comment of style ‘b’; 2 for a comment of style ‘c’; and syntax-table for a comment that should be ended by a generic comment delimiter character.
  9. The string or comment start position. While inside a comment, this is the position where the comment began; while inside a string, this is the position where the string began. When outside of strings and comments, this element is nil.
  10. The list of the positions of the currently open parentheses, starting with the outermost.
  11. When the last buffer position scanned was the (potential) first character of a two character construct (comment delimiter or escaped/char-quoted character pair), the syntax-code (see Syntax Table Internals) of that position. Otherwise nil.

Elements 1, 2, and 6 are ignored in a state which you pass as an argument to parse-partial-sexp to continue parsing. Elements 9 and 10 are mainly used internally by the parser code.

Some additional useful information is available from a parser state using these functions:

Function: syntax-ppss-toplevel-pos state

This function extracts, from parser state state, the last position scanned in the parse which was at top level in grammatical structure. “At top level” means outside of any parentheses, comments, or strings.

The value is nil if state represents a parse which has arrived at a top level position.

Function: syntax-ppss-context state

Return string if the end position of the scan returning state is in a string, and comment if it’s in a comment. Otherwise return nil.


36.6.4 Low-Level Parsing

The most basic way to use the expression parser is to tell it to start at a given position with a certain state, and parse up to a specified end position.

Function: parse-partial-sexp start limit &optional target-depth stop-before state stop-comment

This function parses a sexp in the current buffer starting at start, not scanning past limit. It stops at position limit or when certain criteria described below are met, and sets point to the location where parsing stops. It returns a parser state describing the status of the parse at the point where it stops.

If the third argument target-depth is non-nil, parsing stops if the depth in parentheses becomes equal to target-depth. The depth starts at 0, or at whatever is given in state.

If the fourth argument stop-before is non-nil, parsing stops when it comes to any character that starts a sexp. If stop-comment is non-nil, parsing stops after the start of an unnested comment. If stop-comment is the symbol syntax-table, parsing stops after the start of an unnested comment or a string, or after the end of an unnested comment or a string, whichever comes first.

If state is nil, start is assumed to be at the top level of parenthesis structure, such as the beginning of a function definition. Alternatively, you might wish to resume parsing in the middle of the structure. To do this, you must provide a state argument that describes the initial status of parsing. The value returned by a previous call to parse-partial-sexp will do nicely.


36.6.5 Parameters to Control Parsing

Variable: multibyte-syntax-as-symbol

If this variable is non-nil, scan-sexps treats all non-ASCII characters as symbol constituents regardless of what the syntax table says about them. (However, syntax-table text properties can still override the syntax.)

User Option: parse-sexp-ignore-comments

If the value is non-nil, then comments are treated as whitespace by the functions in this section and by forward-sexp, scan-lists and scan-sexps.

The behavior of parse-partial-sexp is also affected by parse-sexp-lookup-properties (see Syntax Properties).

Variable: comment-end-can-be-escaped

If this buffer local variable is non-nil, a single character which usually terminates a comment doesn’t do so when that character is escaped. This is used in C and C++ Modes, where line comments starting with ‘//’ can be continued onto the next line by escaping the newline with ‘\’.

You can use forward-comment to move forward or backward over one comment or several comments.


36.7 Syntax Table Internals

Syntax tables are implemented as char-tables (see Char-Tables), but most Lisp programs don’t work directly with their elements. Syntax tables do not store syntax data as syntax descriptors (see Syntax Descriptors); they use an internal format, which is documented in this section. This internal format can also be assigned as syntax properties (see Syntax Properties).

Each entry in a syntax table is a raw syntax descriptor: a cons cell of the form (syntax-code . matching-char). syntax-code is an integer which encodes the syntax class and syntax flags, according to the table below. matching-char, if non-nil, specifies a matching character (similar to the second character in a syntax descriptor).

Use aref (see Functions that Operate on Arrays) to get the raw syntax descriptor of a character, e.g. (aref (syntax-table) ch).

Here are the syntax codes corresponding to the various syntax classes:

CodeClassCodeClass
0whitespace8paired delimiter
1punctuation9escape
2word10character quote
3symbol11comment-start
4open parenthesis12comment-end
5close parenthesis13inherit
6expression prefix14generic comment
7string quote15generic string

For example, in the standard syntax table, the entry for ‘(’ is (4 . 41). 41 is the character code for ‘)’.

Syntax flags are encoded in higher order bits, starting 16 bits from the least significant bit. This table gives the power of two which corresponds to each syntax flag.

PrefixFlagPrefixFlag
1(ash 1 16)p(ash 1 20)
2(ash 1 17)b(ash 1 21)
3(ash 1 18)n(ash 1 22)
4(ash 1 19)c(ash 1 23)
Function: string-to-syntax desc

Given a syntax descriptor desc (a string), this function returns the corresponding raw syntax descriptor.

Function: syntax-class-to-char syntax

Given a raw syntax descriptor syntax (an integer), this function returns the corresponding syntax descriptor (a character).

Function: syntax-after pos

This function returns the raw syntax descriptor for the character in the buffer after position pos, taking account of syntax properties as well as the syntax table. If pos is outside the buffer’s accessible portion (see accessible portion), the return value is nil.

Function: syntax-class syntax

This function returns the syntax code for the raw syntax descriptor syntax. More precisely, it takes the raw syntax descriptor’s syntax-code component, masks off the high 16 bits which record the syntax flags, and returns the resulting integer.

If syntax is nil, the return value is nil. This is so that the expression

(syntax-class (syntax-after pos))

evaluates to nil if pos is outside the buffer’s accessible portion, without throwing errors or returning an incorrect code.


36.8 Categories

Categories provide an alternate way of classifying characters syntactically. You can define several categories as needed, then independently assign each character to one or more categories. Unlike syntax classes, categories are not mutually exclusive; it is normal for one character to belong to several categories.

Each buffer has a category table which records which categories are defined and also which characters belong to each category. Each category table defines its own categories, but normally these are initialized by copying from the standard categories table, so that the standard categories are available in all modes.

Each category has a name, which is an ASCII printing character in the range ‘  to ‘~’. You specify the name of a category when you define it with define-category.

The category table is actually a char-table (see Char-Tables). The element of the category table at index c is a category set—a bool-vector—that indicates which categories character c belongs to. In this category set, if the element at index cat is t, that means category cat is a member of the set, and that character c belongs to category cat.

For the next three functions, the optional argument table defaults to the current buffer’s category table.

Function: define-category char docstring &optional table

This function defines a new category, with name char and documentation docstring, for the category table table.

Here’s an example of defining a new category for characters that have strong right-to-left directionality (see Bidirectional Display) and using it in a special category table. To obtain the information about the directionality of characters, the example code uses the ‘bidi-class’ Unicode property (see bidi-class).

(defvar special-category-table-for-bidi
  ;;     Make an empty category-table.
  (let ((category-table (make-category-table))
        ;; Create a char-table which gives the 'bidi-class' Unicode
        ;; property for each character.
        (uniprop-table
         (unicode-property-table-internal 'bidi-class)))
    (define-category ?R "Characters of bidi-class R, AL, or RLO"
                     category-table)
    ;; Modify the category entry of each character whose
    ;; 'bidi-class' Unicode property is R, AL, or RLO --
    ;; these have a right-to-left directionality.
    (map-char-table
     (lambda (key val)
       (if (memq val '(R AL RLO))
           (modify-category-entry key ?R category-table)))
     uniprop-table)
    category-table))
Function: category-docstring category &optional table

This function returns the documentation string of category category in category table table.

(category-docstring ?a)
     ⇒ "ASCII"
(category-docstring ?l)
     ⇒ "Latin"
Function: get-unused-category &optional table

This function returns a category name (a character) which is not currently defined in table. If all possible categories are in use in table, it returns nil.

Function: category-table

This function returns the current buffer’s category table.

Function: category-table-p object

This function returns t if object is a category table, otherwise nil.

Function: standard-category-table

This function returns the standard category table.

Function: copy-category-table &optional table

This function constructs a copy of table and returns it. If table is not supplied (or is nil), it returns a copy of the standard category table. Otherwise, an error is signaled if table is not a category table.

Function: set-category-table table

This function makes table the category table for the current buffer. It returns table.

Function: make-category-table

This creates and returns an empty category table. In an empty category table, no categories have been allocated, and no characters belong to any categories.

Function: make-category-set categories

This function returns a new category set—a bool-vector—whose initial contents are the categories listed in the string categories. The elements of categories should be category names; the new category set has t for each of those categories, and nil for all other categories.

(make-category-set "al")
     ⇒ #&128"\0\0\0\0\0\0\0\0\0\0\0\0\2\20\0\0"
Function: char-category-set char

This function returns the category set for character char in the current buffer’s category table. This is the bool-vector which records which categories the character char belongs to. The function char-category-set does not allocate storage, because it returns the same bool-vector that exists in the category table.

(char-category-set ?a)
     ⇒ #&128"\0\0\0\0\0\0\0\0\0\0\0\0\2\20\0\0"
Function: category-set-mnemonics category-set

This function converts the category set category-set into a string containing the characters that designate the categories that are members of the set.

(category-set-mnemonics (char-category-set ?a))
     ⇒ "al"
Function: modify-category-entry char category &optional table reset

This function modifies the category set of char in category table table (which defaults to the current buffer’s category table). char can be a character, or a cons cell of the form (min . max); in the latter case, the function modifies the category sets of all characters in the range between min and max, inclusive.

Normally, it modifies a category set by adding category to it. But if reset is non-nil, then it deletes category instead.

Command: describe-categories &optional buffer-or-name

This function describes the category specifications in the current category table. It inserts the descriptions in a buffer, and then displays that buffer. If buffer-or-name is non-nil, it describes the category table of that buffer instead.


37 Parsing Program Source

Emacs provides various ways to parse program source text and produce a syntax tree. In a syntax tree, text is no longer considered a one-dimensional stream of characters, but a structured tree of nodes, where each node represents a piece of text. Thus, a syntax tree can enable interesting features like precise fontification, indentation, navigation, structured editing, etc.

Emacs has a simple facility for parsing balanced expressions (see Parsing Expressions). There is also the SMIE library for generic navigation and indentation (see Simple Minded Indentation Engine).

In addition to those, Emacs also provides integration with the tree-sitter library if support for it was compiled in. The tree-sitter library implements an incremental parser and has support for a wide range of programming languages.

Function: treesit-available-p

This function returns non-nil if tree-sitter features are available for the current Emacs session.

To be able to parse the program source using the tree-sitter library and access the syntax tree of the program, a Lisp program needs to load a language grammar library, and create a parser for that language and the current buffer. After that, the Lisp program can query the parser about specific nodes of the syntax tree. Then, it can access various kinds of information about each node, and search for nodes using a powerful pattern-matching syntax. This chapter explains how to do all this, and also how a Lisp program can work with source files that mix multiple programming languages.


37.1 Tree-sitter Language Grammar

Loading a language grammar

Tree-sitter relies on language grammar to parse text in that language. In Emacs, a language grammar is represented by a symbol. For example, the C language grammar is represented as the symbol c, and c can be passed to tree-sitter functions as the language argument.

Tree-sitter language grammars are distributed as dynamic libraries. In order to use a language grammar in Emacs, you need to make sure that the dynamic library is installed on the system. Emacs looks for language grammars in several places, in the following order:

  • first, in the list of directories specified by the variable treesit-extra-load-path;
  • then, in the tree-sitter subdirectory of the directory specified by user-emacs-directory (see The Init File);
  • and finally, in the system’s default locations for dynamic libraries.

In each of these directories, Emacs looks for a file with file-name extensions specified by the variable dynamic-library-suffixes.

If Emacs cannot find the library or has problems loading it, Emacs signals the treesit-load-language-error error. The data of that signal could be one of the following:

(not-found error-msg …)

This means that Emacs could not find the language grammar library.

(symbol-error error-msg)

This means that Emacs could not find in the library the expected function that every language grammar library should export.

(version-mismatch error-msg)

This means that the version of the language grammar library is incompatible with that of the tree-sitter library.

In all of these cases, error-msg might provide additional details about the failure.

Function: treesit-language-available-p language &optional detail

This function returns non-nil if the language grammar for language exists and can be loaded.

If detail is non-nil, return (t . nil) when language is available, and (nil . data) when it’s unavailable. data is the signal data of treesit-load-language-error.

By convention, the file name of the dynamic library for language is libtree-sitter-language.ext, where ext is the system-specific extension for dynamic libraries. Also by convention, the function provided by that library is named tree_sitter_language. If a language grammar library doesn’t follow this convention, you should add an entry

(language library-base-name function-name)

to the list in the variable treesit-load-name-override-list, where library-base-name is the basename of the dynamic library’s file name (usually, libtree-sitter-language), and function-name is the function provided by the library (usually, tree_sitter_language). For example,

(cool-lang "libtree-sitter-coool" "tree_sitter_cooool")

for a language that considers itself too “cool” to abide by conventions.

Function: treesit-library-abi-version &optional min-compatible

This function returns the version of the language grammar Application Binary Interface (ABI) supported by the tree-sitter library. By default, it returns the latest ABI version supported by the library, but if min-compatible is non-nil, it returns the oldest ABI version which the library still can support. Language grammar libraries must be built for ABI versions between the oldest and the latest versions supported by the tree-sitter library, otherwise the library will be unable to load them.

Function: treesit-language-abi-version language

This function returns the ABI version of the language grammar library loaded by Emacs for language. If language is unavailable, this function returns nil.

Concrete syntax tree

A syntax tree is what a parser generates. In a syntax tree, each node represents a piece of text, and is connected to each other by a parent-child relationship. For example, if the source text is

1 + 2

its syntax tree could be

                  +--------------+
                  | root "1 + 2" |
                  +--------------+
                         |
        +--------------------------------+
        |       expression "1 + 2"       |
        +--------------------------------+
           |             |            |
+------------+   +--------------+   +------------+
| number "1" |   | operator "+" |   | number "2" |
+------------+   +--------------+   +------------+

We can also represent it as an s-expression:

(root (expression (number) (operator) (number)))

Node types

Names like root, expression, number, and operator specify the type of the nodes. However, not all nodes in a syntax tree have a type. Nodes that don’t have a type are known as anonymous nodes, and nodes with a type are named nodes. Anonymous nodes are tokens with fixed spellings, including punctuation characters like bracket ‘]’, and keywords like return.

Field names

To make the syntax tree easier to analyze, many language grammar assign field names to child nodes. For example, a function_definition node could have a declarator and a body:

(function_definition
 declarator: (declaration)
 body: (compound_statement))

Exploring the syntax tree

To aid in understanding the syntax of a language and in debugging Lisp programs that use the syntax tree, Emacs provides an “explore” mode, which displays the syntax tree of the source in the current buffer in real time. Emacs also comes with an “inspect mode”, which displays information of the nodes at point in the mode-line.

Command: treesit-explore-mode

This mode pops up a window displaying the syntax tree of the source in the current buffer. Selecting text in the source buffer highlights the corresponding nodes in the syntax tree display. Clicking on nodes in the syntax tree highlights the corresponding text in the source buffer.

Command: treesit-inspect-mode

This minor mode displays on the mode-line the node that starts at point. For example, the mode-line can display

parent field: (node (child (…)))

where node, child, etc., are nodes which begin at point. parent is the parent of node. node is displayed in a bold typeface. field-names are field names of node and of child, etc.

If no node starts at point, i.e., point is in the middle of a node, then the mode line displays the earliest node that spans point, and its immediate parent.

This minor mode doesn’t create parsers on its own. It uses the first parser in (treesit-parser-list) (see Using Tree-sitter Parser).

Reading the grammar definition

Authors of language grammars define the grammar of a programming language, which determines how a parser constructs a concrete syntax tree out of the program text. In order to use the syntax tree effectively, you need to consult the grammar file.

The grammar file is usually grammar.js in a language grammar’s project repository. The link to a language grammar’s home page can be found on tree-sitter’s homepage.

The grammar definition is written in JavaScript. For example, the rule matching a function_definition node may look like

function_definition: $ => seq(
  $.declaration_specifiers,
  field('declarator', $.declaration),
  field('body', $.compound_statement)
)

The rules are represented by functions that take a single argument $, representing the whole grammar. The function itself is constructed by other functions: the seq function puts together a sequence of children; the field function annotates a child with a field name. If we write the above definition in the so-called Backus-Naur Form (BNF) syntax, it would look like

function_definition :=
  <declaration_specifiers> <declaration> <compound_statement>

and the node returned by the parser would look like

(function_definition
  (declaration_specifier)
  declarator: (declaration)
  body: (compound_statement))

Below is a list of functions that one can see in a grammar definition. Each function takes other rules as arguments and returns a new rule.

seq(rule1, rule2, …)

matches each rule one after another.

choice(rule1, rule2, …)

matches one of the rules in its arguments.

repeat(rule)

matches rule zero or more times. This is like the ‘*’ operator in regular expressions.

repeat1(rule)

matches rule one or more times. This is like the ‘+’ operator in regular expressions.

optional(rule)

matches rule zero or one times. This is like the ‘?’ operator in regular expressions.

field(name, rule)

assigns field name name to the child node matched by rule.

alias(rule, alias)

makes nodes matched by rule appear as alias in the syntax tree generated by the parser. For example,

alias(preprocessor_call_exp, call_expression)

makes any node matched by preprocessor_call_exp appear as call_expression.

Below are grammar functions of lesser importance for reading a language grammar.

token(rule)

marks rule to produce a single leaf node. That is, instead of generating a parent node with individual child nodes under it, everything is combined into a single leaf node. See Retrieving Nodes.

token.immediate(rule)

Normally, grammar rules ignore preceding whitespace; this changes rule to match only when there is no preceding whitespace.

prec(n, rule)

gives rule the level-n precedence.

prec.left([n,] rule)

marks rule as left-associative, optionally with level n.

prec.right([n,] rule)

marks rule as right-associative, optionally with level n.

prec.dynamic(n, rule)

this is like prec, but the precedence is applied at runtime instead.

The documentation of the tree-sitter project has more about writing a grammar. Read especially “The Grammar DSL” section.


37.2 Using Tree-sitter Parser

This section describes how to create and configure a tree-sitter parser. In Emacs, each tree-sitter parser is associated with a buffer. As the user edits the buffer, the associated parser and syntax tree are automatically kept up-to-date.

Variable: treesit-max-buffer-size

This variable contains the maximum size of buffers in which tree-sitter can be activated. Major modes should check this value when deciding whether to enable tree-sitter features.

Function: treesit-parser-create language &optional buffer no-reuse

Create a parser for the specified buffer and language (see Tree-sitter Language Grammar). If buffer is omitted or nil, it stands for the current buffer.

By default, this function reuses a parser if one already exists for language in buffer, but if no-reuse is non-nil, this function always creates a new parser.

If that buffer is an indirect buffer, its base buffer is used instead. That is, indirect buffers use their base buffer’s parsers. If the base buffer is narrowed, an indirect buffer might not be able to retrieve information of the portion of the buffer text that is invisible in the base buffer. Lisp programs should widen as necessary should they want to use a parser in an indirect buffer.

Given a parser, we can query information about it.

Function: treesit-parser-buffer parser

This function returns the buffer associated with parser.

Function: treesit-parser-language parser

This function returns the language used by parser.

Function: treesit-parser-p object

This function checks if object is a tree-sitter parser, and returns non-nil if it is, and nil otherwise.

There is no need to explicitly parse a buffer, because parsing is done automatically and lazily. A parser only parses when a Lisp program queries for a node in its syntax tree. Therefore, when a parser is first created, it doesn’t parse the buffer; it waits until the Lisp program queries for a node for the first time. Similarly, when some change is made in the buffer, a parser doesn’t re-parse immediately.

When a parser does parse, it checks for the size of the buffer. Tree-sitter can only handle buffers no larger than about 4GB. If the size exceeds that, Emacs signals the treesit-buffer-too-large error with signal data being the buffer size.

Once a parser is created, Emacs automatically adds it to the internal parser list. Every time a change is made to the buffer, Emacs updates parsers in this list so they can update their syntax tree incrementally.

Function: treesit-parser-list &optional buffer

This function returns the parser list of buffer. If buffer is nil or omitted, it defaults to the current buffer. If that buffer is an indirect buffer, its base buffer is used instead. That is, indirect buffers use their base buffer’s parsers.

Function: treesit-parser-delete parser

This function deletes parser.

Normally, a parser “sees” the whole buffer, but when the buffer is narrowed (see Narrowing), the parser will only see the accessible portion of the buffer. As far as the parser can tell, the hidden region was deleted. When the buffer is later widened, the parser thinks text is inserted at the beginning and at the end. Although parsers respect narrowing, modes should not use narrowing as a means to handle a multi-language buffer; instead, set the ranges in which the parser should operate. See Parsing Text in Multiple Languages.

Because a parser parses lazily, when the user or a Lisp program narrows the buffer, the parser is not affected immediately; as long as the mode doesn’t query for a node while the buffer is narrowed, the parser is oblivious of the narrowing.

Besides creating a parser for a buffer, a Lisp program can also parse a string. Unlike a buffer, parsing a string is a one-off operation, and there is no way to update the result.

Function: treesit-parse-string string language

This function parses string using language, and returns the root node of the generated syntax tree.

Be notified by changes to the parse tree

A Lisp program might want to be notified of text affected by incremental parsing. For example, inserting a comment-closing token converts text before that token into a comment. Even though the text is not directly edited, it is deemed to be “changed” nevertheless.

Emacs lets a Lisp program register callback functions (a.k.a. notifiers) for these kinds of changes. A notifier function takes two arguments: ranges and parser. ranges is a list of cons cells of the form (start . end), where start and end mark the start and the end positions of a range. parser is the parser issuing the notification.

Every time a parser reparses a buffer, it compares the old and new parse-tree, computes the ranges in which nodes have changed, and passes the ranges to notifier functions. Note that the initial parse is also considered a “change”, so notifier functions are called on the initial parse, with range being the whole buffer.

Function: treesit-parser-add-notifier parser function

This function adds function to parser’s list of after-change notifier functions. function must be a function symbol, not a lambda function (see Anonymous Functions).

Function: treesit-parser-remove-notifier parser function

This function removes function from the list of parser’s after-change notifier functions. function must be a function symbol, rather than a lambda function.

Function: treesit-parser-notifiers parser

This function returns the list of parser’s notifier functions.


37.3 Retrieving Nodes

Here are some terms and conventions we use when documenting tree-sitter functions.

A node in a syntax tree spans some portion of the program text in the buffer. We say that a node is “smaller” or “larger” than another if it spans, respectively, a smaller or larger portion of buffer text than the other node. Since nodes that are deeper (“lower”) in the tree are children of the nodes that are “higher” in the tree, it follows that a lower node will always be smaller than a node that is higher in the node hierarchy. A node that is higher up in the syntax tree contains one or more smaller nodes as its children, and therefore spans a larger portion of buffer text.

When a function cannot find a node, it returns nil. For convenience, all functions that take a node as argument and return a node, also accept the node argument of nil and in that case just return nil.

Nodes are not automatically updated when the associated buffer is modified, and there is no way to update a node once it is retrieved. Using an outdated node signals the treesit-node-outdated error.

Retrieving nodes from syntax tree

Function: treesit-node-at pos &optional parser-or-lang named

This function returns a leaf node at buffer position pos. A leaf node is a node that doesn’t have any child nodes.

This function tries to return a node whose span covers pos: the node’s beginning position is less than or equal to pos, and the node’s end position is greater than or equal to pos.

If no leaf node’s span covers pos (e.g., pos is in the whitespace between two leaf nodes), this function returns the first leaf node after pos.

Finally, if there is no leaf node after pos, return the first leaf node before pos.

If parser-or-lang is a parser object, this function uses that parser; if parser-or-lang is a language, this function uses the first parser for that language in the current buffer, or creates one if none exists; if parser-or-lang is nil, this function tries to guess the language at pos by calling treesit-language-at (see Parsing Text in Multiple Languages).

If this function cannot find a suitable node to return, it returns nil.

If named is non-nil, this function looks only for named nodes (see named node).

Example:

;; Find the node at point in a C parser's syntax tree.
(treesit-node-at (point) 'c)
  ⇒ #<treesit-node (primitive_type) in 23-27>
Function: treesit-node-on beg end &optional parser-or-lang named

This function returns the smallest node that covers the region of buffer text between beg and end. In other words, the start of the node is before or at beg, and the end of the node is at or after end.

Beware: calling this function on an empty line that is not inside any top-level construct (function definition, etc.) most probably will give you the root node, because the root node is the smallest node that covers that empty line. Most of the time, you want to use treesit-node-at instead.

If parser-or-lang is a parser object, this function uses that parser; if parser-or-lang is a language, this function uses the first parser for that language in the current buffer, or creates one if none exists; if parser-or-lang is nil, this function tries to guess the language at beg by calling treesit-language-at.

If named is non-nil, this function looks for a named node only (see named node).

Function: treesit-parser-root-node parser

This function returns the root node of the syntax tree generated by parser.

Function: treesit-buffer-root-node &optional language

This function finds the first parser for language in the current buffer, or creates one if none exists, and returns the root node generated by that parser. If language is omitted, it uses the first parser in the parser list. If it cannot find an appropriate parser, it returns nil.

Given a node, a Lisp program can retrieve other nodes starting from it, or query for information about this node.

Retrieving nodes from other nodes

By kinship

Function: treesit-node-parent node

This function returns the immediate parent of node.

If node is more than 1000 levels deep in a parse tree, the return value is undefined. Currently it returns nil, but that could change in the future.

Function: treesit-node-child node n &optional named

This function returns the n’th child of node. If named is non-nil, it counts only named nodes (see named node).

For example, in a node that represents a string "text", there are three children nodes: the opening quote ", the string text text, and the closing quote ". Among these nodes, the first child is the opening quote ", and the first named child is the string text.

This function returns nil if there is no n’th child. n could be negative, e.g., −1 represents the last child.

Function: treesit-node-children node &optional named

This function returns all of node’s children as a list. If named is non-nil, it retrieves only named nodes.

Function: treesit-node-next-sibling node &optional named

This function finds the next sibling of node. If named is non-nil, it finds the next named sibling.

Function: treesit-node-prev-sibling node &optional named

This function finds the previous sibling of node. If named is non-nil, it finds the previous named sibling.

By field name

To make the syntax tree easier to analyze, many language grammars assign field names to child nodes (see field name). For example, a function_definition node could have a declarator node and a body node.

Function: treesit-node-child-by-field-name node field-name

This function finds the child of node whose field name is field-name, a string.

;; Get the child that has "body" as its field name.
(treesit-node-child-by-field-name node "body")
  ⇒ #<treesit-node (compound_statement) in 45-89>

By position

Function: treesit-node-first-child-for-pos node pos &optional named

This function finds the first child of node that extends beyond buffer position pos. “Extends beyond” means the end of the child node is greater or equal to pos. This function only looks for immediate children of node, and doesn’t look in its grandchildren. If named is non-nil, it looks for the first named child (see named node).

Function: treesit-node-descendant-for-range node beg end &optional named

This function finds the smallest descendant node of node that spans the region of text between positions beg and end. It is similar to treesit-node-at. If named is non-nil, it looks for the smallest named child.

Searching for node

Function: treesit-search-subtree node predicate &optional backward all depth

This function traverses the subtree of node (including node itself), looking for a node for which predicate returns non-nil. predicate is a regexp that is matched against each node’s type, or a predicate function that takes a node and returns non-nil if the node matches. The function returns the first node that matches, or nil if none does.

By default, this function only traverses named nodes, but if all is non-nil, it traverses all the nodes. If backward is non-nil, it traverses backwards (i.e., it visits the last child first when traversing down the tree). If depth is non-nil, it must be a number that limits the tree traversal to that many levels down the tree. If depth is nil, it defaults to 1000.

Function: treesit-search-forward start predicate &optional backward all

Like treesit-search-subtree, this function also traverses the parse tree and matches each node with predicate (except for start), where predicate can be a regexp or a function. For a tree like the one below where start is marked ‘S’, this function traverses as numbered from 1 to 12:

              12
              |
     S--------3----------11
     |        |          |
o--o-+--o  1--+--2    6--+-----10
|  |                  |        |
o  o                +-+-+   +--+--+
                    |   |   |  |  |
                    4   5   7  8  9

Note that this function doesn’t traverse the subtree of start, and it always traverses leaf nodes first, before moving upwards.

Like treesit-search-subtree, this function only searches for named nodes by default, but if all is non-nil, it searches for all nodes. If backward is non-nil, it searches backwards.

While treesit-search-subtree traverses the subtree of a node, this function starts with node start and traverses every node that comes after it in the buffer position order, i.e., nodes with start positions greater than the end position of start.

In the tree shown above, treesit-search-subtree traverses node ‘S’ (start) and nodes marked with o, where this function traverses the nodes marked with numbers. This function is useful for answering questions like “what is the first node after start in the buffer that satisfies some condition?”

Function: treesit-search-forward-goto node predicate &optional start backward all

This function moves point to the start or end of the next node after node in the buffer that matches predicate. If start is non-nil, stop at the beginning rather than the end of a node.

This function guarantees that the matched node it returns makes progress in terms of buffer position: the start/end position of the returned node is always greater than that of node.

Arguments predicate, backward, and all are the same as in treesit-search-forward.

Function: treesit-induce-sparse-tree root predicate &optional process-fn depth

This function creates a sparse tree from root’s subtree.

It takes the subtree under root, and combs it so only the nodes that match predicate are left. Like previous functions, the predicate can be a regexp string that matches against each node’s type, or a function that takes a node and returns non-nil if it matches.

For example, given the subtree on the left that consists of both numbers and letters, if predicate is “letter only”, the returned tree is the one on the right.

    a                 a              a
    |                 |              |
+---+---+         +---+---+      +---+---+
|   |   |         |   |   |      |   |   |
b   1   2         b   |   |      b   c   d
    |   |     =>      |   |  =>      |
    c   +--+          c   +          e
    |   |  |          |   |
 +--+   d  4       +--+   d
 |  |              |
 e  5              e

If process-fn is non-nil, instead of returning the matched nodes, this function passes each node to process-fn and uses the returned value instead. If non-nil, depth limits the number of levels to go down from root. If depth is nil, it defaults to 1000.

Each node in the returned tree looks like (tree-sitter-node . (child …)). The tree-sitter-node of the root of this tree will be nil if root doesn’t match predicate. If no node matches predicate, the function returns nil.

More convenience functions

Function: treesit-filter-child node predicate &optional named

This function finds immediate children of node that satisfy predicate.

The predicate function takes a node as argument and should return non-nil to indicate that the node should be kept. If named is non-nil, this function only examines named nodes.

Function: treesit-parent-until node predicate &optional include-node

This function repeatedly finds the parents of node, and returns the parent that satisfies pred, a function that takes a node as argument and returns a boolean that indicates a match. If no parent satisfies pred, this function returns nil.

Normally this function only looks at the parents of node but not node itself. But if include-node is non-nil, this function returns node if node satisfies pred.

Function: treesit-parent-while node pred

This function goes up the tree starting from node, and keeps doing so as long as the nodes satisfy pred, a function that takes a node as argument. That is, this function returns the highest parent of node that still satisfies pred. Note that if node satisfies pred but its immediate parent doesn’t, node itself is returned.

Function: treesit-node-top-level node &optional type

This function returns the highest parent of node that has the same type as node. If no such parent exists, it returns nil. Therefore this function is also useful for testing whether node is top-level.

If type is non-nil, this function matches each parent’s type with type as a regexp, rather than using node’s type.


37.4 Accessing Node Information

Basic information of Node

Every node is associated with a parser, and that parser is associated with a buffer. The following functions retrieve them.

Function: treesit-node-parser node

This function returns node’s associated parser.

Function: treesit-node-buffer node

This function returns node’s parser’s associated buffer.

Function: treesit-node-language node

This function returns node’s parser’s associated language.

Each node represents a portion of text in the buffer. Functions below find relevant information about that text.

Function: treesit-node-start node

Return the start position of node.

Function: treesit-node-end node

Return the end position of node.

Function: treesit-node-text node &optional object

Return the buffer text that node represents, as a string. (If node is retrieved from parsing a string, it will be the text from that string.)

Here are some predicates on tree-sitter nodes:

Function: treesit-node-p object

Checks if object is a tree-sitter syntax node.

Function: treesit-node-eq node1 node2

Checks if node1 and node2 refer to the same node in a tree-sitter syntax tree. This function uses the same equivalence metric as equal. You can also compare nodes using equal (see Equality Predicates).

Property information

In general, nodes in a concrete syntax tree fall into two categories: named nodes and anonymous nodes. Whether a node is named or anonymous is determined by the language grammar (see named node).

Apart from being named or anonymous, a node can have other properties. A node can be “missing”: such nodes are inserted by the parser in order to recover from certain kinds of syntax errors, i.e., something should probably be there according to the grammar, but is not there. This can happen during editing of the program source, when the source is not yet in its final form.

A node can be “extra”: such nodes represent things like comments, which can appear anywhere in the text.

A node can be “outdated”, if its parser has reparsed at least once after the node was created.

A node “has error” if the text it spans contains a syntax error. It can be that the node itself has an error, or one of its descendants has an error.

A node is considered live if its parser is not deleted, and the buffer to which it belongs is a live buffer (see Killing Buffers).

Function: treesit-node-check node property

This function returns non-nil if node has the specified property. property can be named, missing, extra, outdated, has-error, or live.

Function: treesit-node-type node

Named nodes have “types” (see node type). For example, a named node can be a string_literal node, where string_literal is its type. The type of an anonymous node is just the text that the node represents; e.g., the type of a ‘,’ node is just ‘,’.

This function returns node’s type as a string.

Information as a child or parent

Function: treesit-node-index node &optional named

This function returns the index of node as a child node of its parent. If named is non-nil, it only counts named nodes (see named node).

Function: treesit-node-field-name node

A child of a parent node could have a field name (see field name). This function returns the field name of node as a child of its parent.

Function: treesit-node-field-name-for-child node n

This function returns the field name of the n’th child of node. It returns nil if there is no n’th child, or the n’th child doesn’t have a field name.

Note that n counts both named and anonymous children, and n can be negative, e.g., −1 represents the last child.

Function: treesit-node-child-count node &optional named

This function returns the number of children of node. If named is non-nil, it only counts named children (see named node).


37.5 Pattern Matching Tree-sitter Nodes

Tree-sitter lets Lisp programs match patterns using a small declarative language. This pattern matching consists of two steps: first tree-sitter matches a pattern against nodes in the syntax tree, then it captures specific nodes that matched the pattern and returns the captured nodes.

We describe first how to write the most basic query pattern and how to capture nodes in a pattern, then the pattern-matching function, and finally the more advanced pattern syntax.

Basic query syntax

A query consists of multiple patterns. Each pattern is an s-expression that matches a certain node in the syntax node. A pattern has the form (type (child…)).

For example, a pattern that matches a binary_expression node that contains number_literal child nodes would look like

(binary_expression (number_literal))

To capture a node using the query pattern above, append @capture-name after the node pattern you want to capture. For example,

(binary_expression (number_literal) @number-in-exp)

captures number_literal nodes that are inside a binary_expression node with the capture name number-in-exp.

We can capture the binary_expression node as well, with, for example, the capture name biexp:

(binary_expression
 (number_literal) @number-in-exp) @biexp

Query function

Now we can introduce the query functions.

Function: treesit-query-capture node query &optional beg end node-only

This function matches patterns in query within node. The argument query can be either an s-expression, a string, or a compiled query object. For now, we focus on the s-expression syntax; string syntax and compiled queries are described at the end of the section.

The argument node can also be a parser or a language symbol. A parser means use its root node, a language symbol means find or create a parser for that language in the current buffer, and use the root node.

The function returns all the captured nodes in an alist with elements of the form (capture_name . node). If node-only is non-nil, it returns the list of nodes instead. By default the entire text of node is searched, but if beg and end are both non-nil, they specify the region of buffer text where this function should match nodes. Any matching node whose span overlaps with the region between beg and end is captured; it doesn’t have to be completely contained in the region.

This function raises the treesit-query-error error if query is malformed. The signal data contains a description of the specific error. You can use treesit-query-validate to validate and debug the query.

For example, suppose node’s text is 1 + 2, and query is

(setq query
      '((binary_expression
         (number_literal) @number-in-exp) @biexp)

Matching that query would return

(treesit-query-capture node query)
    ⇒ ((biexp . <node for "1 + 2">)
       (number-in-exp . <node for "1">)
       (number-in-exp . <node for "2">))

As mentioned earlier, query could contain multiple patterns. For example, it could have two top-level patterns:

(setq query
      '((binary_expression) @biexp
        (number_literal) @number @biexp)
Function: treesit-query-string string query language

This function parses string as language, matches its root node with query, and returns the result.

More query syntax

Besides node type and capture name, tree-sitter’s pattern syntax can express anonymous node, field name, wildcard, quantification, grouping, alternation, anchor, and predicate.

Anonymous node

An anonymous node is written verbatim, surrounded by quotes. A pattern matching (and capturing) keyword return would be

"return" @keyword

Wild card

In a pattern, ‘(_)’ matches any named node, and ‘_’ matches any named or anonymous node. For example, to capture any named child of a binary_expression node, the pattern would be

(binary_expression (_) @in-biexp)

Field name

It is possible to capture child nodes that have specific field names. In the pattern below, declarator and body are field names, indicated by the colon following them.

(function_definition
  declarator: (_) @func-declarator
  body: (_) @func-body)

It is also possible to capture a node that doesn’t have a certain field, say, a function_definition without a body field:

(function_definition !body) @func-no-body

Quantify node

Tree-sitter recognizes quantification operators ‘:*’, ‘:+’, and ‘:?’. Their meanings are the same as in regular expressions: ‘:*’ matches the preceding pattern zero or more times, ‘:+’ matches one or more times, and ‘:?’ matches zero or one times.

For example, the following pattern matches type_declaration nodes that have zero or more long keywords.

(type_declaration "long" :*) @long-type

The following pattern matches a type declaration that may or may not have a long keyword:

(type_declaration "long" :?) @long-type

Grouping

Similar to groups in regular expressions, we can bundle patterns into groups and apply quantification operators to them. For example, to express a comma-separated list of identifiers, one could write

(identifier) ("," (identifier)) :*

Alternation

Again, similar to regular expressions, we can express “match any one of these patterns” in a pattern. The syntax is a vector of patterns. For example, to capture some keywords in C, the pattern would be

[
  "return"
  "break"
  "if"
  "else"
] @keyword

Anchor

The anchor operator :anchor can be used to enforce juxtaposition, i.e., to enforce two things to be directly next to each other. The two “things” can be two nodes, or a child and the end of its parent. For example, to capture the first child, the last child, or two adjacent children:

;; Anchor the child with the end of its parent.
(compound_expression (_) @last-child :anchor)

;; Anchor the child with the beginning of its parent.
(compound_expression :anchor (_) @first-child)

;; Anchor two adjacent children.
(compound_expression
 (_) @prev-child
 :anchor
 (_) @next-child)

Note that the enforcement of juxtaposition ignores any anonymous nodes.

Predicate

It is possible to add predicate constraints to a pattern. For example, with the following pattern:

(
 (array :anchor (_) @first (_) @last :anchor)
 (:equal @first @last)
)

tree-sitter only matches arrays where the first element is equal to the last element. To attach a predicate to a pattern, we need to group them together. Currently there are three predicates: :equal, :match, and :pred.

Predicate: :equal arg1 arg2

Matches if arg1 is equal to arg2. Arguments can be either strings or capture names. Capture names represent the text that the captured node spans in the buffer.

Predicate: :match regexp capture-name

Matches if the text that capture-name’s node spans in the buffer matches regular expression regexp, given as a string literal. Matching is case-sensitive.

Predicate: :pred fn &rest nodes

Matches if function fn returns non-nil when passed each node in nodes as arguments.

Note that a predicate can only refer to capture names that appear in the same pattern. Indeed, it makes little sense to refer to capture names in other patterns.

String patterns

Besides s-expressions, Emacs allows the tree-sitter’s native query syntax to be used by writing them as strings. It largely resembles the s-expression syntax. For example, the following query

(treesit-query-capture
 node '((addition_expression
         left: (_) @left
         "+" @plus-sign
         right: (_) @right) @addition

         ["return" "break"] @keyword))

is equivalent to

(treesit-query-capture
 node "(addition_expression
        left: (_) @left
        \"+\" @plus-sign
        right: (_) @right) @addition

        [\"return\" \"break\"] @keyword")

Most patterns can be written directly as s-expressions inside a string. Only a few of them need modification:

  • Anchor :anchor is written as ‘.’.
  • :?’ is written as ‘?’.
  • :*’ is written as ‘*’.
  • :+’ is written as ‘+’.
  • :equal, :match and :pred are written as #equal, #match and #pred, respectively. In general, predicates change their ‘:’ to ‘#’.

For example,

'((
   (compound_expression :anchor (_) @first (_) :* @rest)
   (:match "love" @first)
   ))

is written in string form as

"(
  (compound_expression . (_) @first (_)* @rest)
  (#match \"love\" @first)
  )"

Compiling queries

If a query is intended to be used repeatedly, especially in tight loops, it is important to compile that query, because a compiled query is much faster than an uncompiled one. A compiled query can be used anywhere a query is accepted.

Function: treesit-query-compile language query

This function compiles query for language into a compiled query object and returns it.

This function raises the treesit-query-error error if query is malformed. The signal data contains a description of the specific error. You can use treesit-query-validate to validate and debug the query.

Function: treesit-query-language query

This function returns the language of query.

Function: treesit-query-expand query

This function converts the s-expression query into the string format.

Function: treesit-pattern-expand pattern

This function converts the s-expression pattern into the string format.

For more details, read the tree-sitter project’s documentation about pattern-matching, which can be found at https://tree-sitter.github.io/tree-sitter/using-parsers#pattern-matching-with-queries.


37.6 Parsing Text in Multiple Languages

Sometimes, the source of a programming language could contain snippets of other languages; HTML + CSS + JavaScript is one example. In that case, text segments written in different languages need to be assigned different parsers. Traditionally, this is achieved by using narrowing. While tree-sitter works with narrowing (see narrowing), the recommended way is instead to specify regions of buffer text (i.e., ranges) in which a parser will operate. This section describes functions for setting and getting ranges for a parser.

Lisp programs should call treesit-update-ranges to make sure the ranges for each parser are correct before using parsers in a buffer, and call treesit-language-at to figure out the language responsible for the text at some position. These two functions don’t work by themselves, they need major modes to set treesit-range-settings and treesit-language-at-point-function, which do the actual work. These functions and variables are explained in more detail towards the end of the section.

Getting and setting ranges

Function: treesit-parser-set-included-ranges parser ranges

This function sets up parser to operate on ranges. The parser will only read the text of the specified ranges. Each range in ranges is a pair of the form (beg . end).

The ranges in ranges must come in order and must not overlap. That is, in pseudo code:

(cl-loop for idx from 1 to (1- (length ranges))
         for prev = (nth (1- idx) ranges)
         for next = (nth idx ranges)
         should (<= (car prev) (cdr prev)
                    (car next) (cdr next)))

If ranges violates this constraint, or something else went wrong, this function signals the treesit-range-invalid error. The signal data contains a specific error message and the ranges we are trying to set.

This function can also be used for disabling ranges. If ranges is nil, the parser is set to parse the whole buffer.

Example:

(treesit-parser-set-included-ranges
 parser '((1 . 9) (16 . 24) (24 . 25)))
Function: treesit-parser-included-ranges parser

This function returns the ranges set for parser. The return value is the same as the ranges argument of treesit-parser-included-ranges: a list of cons cells of the form (beg . end). If parser doesn’t have any ranges, the return value is nil.

(treesit-parser-included-ranges parser)
    ⇒ ((1 . 9) (16 . 24) (24 . 25))
Function: treesit-query-range source query &optional beg end

This function matches source with query and returns the ranges of captured nodes. The return value is a list of cons cells of the form (beg . end), where beg and end specify the beginning and the end of a region of text.

For convenience, source can be a language symbol, a parser, or a node. If it’s a language symbol, this function matches in the root node of the first parser using that language; if a parser, this function matches in the root node of that parser; if a node, this function matches in that node.

The argument query is the query used to capture nodes (see Pattern Matching Tree-sitter Nodes). The capture names don’t matter. The arguments beg and end, if both non-nil, limit the range in which this function queries.

Like other query functions, this function raises the treesit-query-error error if query is malformed.

Supporting multiple languages in Lisp programs

It should suffice for general Lisp programs to call the following two functions in order to support program sources that mix multiple languages.

Function: treesit-update-ranges &optional beg end

This function updates ranges for parsers in the buffer. It makes sure the parsers’ ranges are set correctly between beg and end, according to treesit-range-settings. If omitted, beg defaults to the beginning of the buffer, and end defaults to the end of the buffer.

For example, fontification functions use this function before querying for nodes in a region.

Function: treesit-language-at pos

This function returns the language of the text at buffer position pos. Under the hood it calls treesit-language-at-point-function and returns its return value. If treesit-language-at-point-function is nil, this function returns the language of the first parser in the returned value of treesit-parser-list. If there is no parser in the buffer, it returns nil.

Supporting multiple languages in major modes

Normally, in a set of languages that can be mixed together, there is a host language and one or more embedded languages. A Lisp program usually first parses the whole document with the host language’s parser, retrieves some information, sets ranges for the embedded languages with that information, and then parses the embedded languages.

Take a buffer containing HTML, CSS, and JavaScript as an example. A Lisp program will first parse the whole buffer with an HTML parser, then query the parser for style_element and script_element nodes, which correspond to CSS and JavaScript text, respectively. Then it sets the range of the CSS and JavaScript parsers to the range which their corresponding nodes span.

Given a simple HTML document:

<html>
  <script>1 + 2</script>
  <style>body { color: "blue"; }</style>
</html>

a Lisp program will first parse with a HTML parser, then set ranges for CSS and JavaScript parsers:

;; Create parsers.
(setq html (treesit-parser-create 'html))
(setq css (treesit-parser-create 'css))
(setq js (treesit-parser-create 'javascript))

;; Set CSS ranges.
(setq css-range
      (treesit-query-range
       'html
       '((style_element (raw_text) @capture))))
(treesit-parser-set-included-ranges css css-range)

;; Set JavaScript ranges.
(setq js-range
      (treesit-query-range
       'html
       '((script_element (raw_text) @capture))))
(treesit-parser-set-included-ranges js js-range)

Emacs automates this process in treesit-update-ranges. A multi-language major mode should set treesit-range-settings so that treesit-update-ranges knows how to perform this process automatically. Major modes should use the helper function treesit-range-rules to generate a value that can be assigned to treesit-range-settings. The settings in the following example directly translate into operations shown above.

(setq treesit-range-settings
      (treesit-range-rules
       :embed 'javascript
       :host 'html
       '((script_element (raw_text) @capture))

       :embed 'css
       :host 'html
       '((style_element (raw_text) @capture))))
Function: treesit-range-rules &rest query-specs

This function is used to set treesit-range-settings. It takes care of compiling queries and other post-processing, and outputs a value that treesit-range-settings can have.

It takes a series of query-specs, where each query-spec is a query preceded by zero or more keyword/value pairs. Each query is a tree-sitter query in either the string, s-expression, or compiled form, or a function.

If query is a tree-sitter query, it should be preceded by two keyword/value pairs, where the :embed keyword specifies the embedded language, and the :host keyword specifies the host language.

treesit-update-ranges uses query to figure out how to set the ranges for parsers for the embedded language. It queries query in a host language parser, computes the ranges which the captured nodes span, and applies these ranges to embedded language parsers.

If query is a function, it doesn’t need any keyword and value pair. It should be a function that takes 2 arguments, start and end, and sets the ranges for parsers in the current buffer in the region between start and end. It is fine for this function to set ranges in a larger region that encompasses the region between start and end.

Variable: treesit-range-settings

This variable helps treesit-update-ranges in updating the ranges for parsers in the buffer. It is a list of settings where the exact format of a setting is considered internal. You should use treesit-range-rules to generate a value that this variable can have.

Variable: treesit-language-at-point-function

This variable’s value should be a function that takes a single argument, pos, which is a buffer position, and returns the language of the buffer text at pos. This variable is used by treesit-language-at.


37.7 Developing major modes with tree-sitter

This section covers some general guidelines on developing tree-sitter integration for a major mode.

A major mode supporting tree-sitter features should roughly follow this pattern:

(define-derived-mode woomy-mode prog-mode "Woomy"
  "A mode for Woomy programming language."
  (when (treesit-ready-p 'woomy)
    (setq-local treesit-variables ...)
    ...
    (treesit-major-mode-setup)))

treesit-ready-p automatically emits a warning if conditions for enabling tree-sitter aren’t met.

If a tree-sitter major mode shares setup with its “native” counterpart, one can create a “base mode” that contains the common setup, like this:

(define-derived-mode woomy--base-mode prog-mode "Woomy"
  "An internal mode for Woomy programming language."
  (common-setup)
  ...)

(define-derived-mode woomy-mode woomy--base-mode "Woomy"
  "A mode for Woomy programming language."
  (native-setup)
  ...)

(define-derived-mode woomy-ts-mode woomy--base-mode "Woomy"
  "A mode for Woomy programming language."
  (when (treesit-ready-p 'woomy)
    (setq-local treesit-variables ...)
    ...
    (treesit-major-mode-setup)))
Function: treesit-ready-p language &optional quiet

This function checks for conditions for activating tree-sitter. It checks whether Emacs was built with tree-sitter, whether the buffer’s size is not too large for tree-sitter to handle, and whether the grammar for language is available on the system (see Tree-sitter Language Grammar).

This function emits a warning if tree-sitter cannot be activated. If quiet is message, the warning is turned into a message; if quiet is t, no warning or message is displayed.

If all the necessary conditions are met, this function returns non-nil; otherwise it returns nil.

Function: treesit-major-mode-setup

This function activates some tree-sitter features for a major mode.

Currently, it sets up the following features:

  • If treesit-font-lock-settings (see Parser-based Font Lock) is non-nil, it sets up fontification.
  • If treesit-simple-indent-rules (see Parser-based Indentation) is non-nil, it sets up indentation.
  • If treesit-defun-type-regexp is non-nil, it sets up navigation functions for beginning-of-defun and end-of-defun.
  • If treesit-defun-name-function is non-nil, it sets up add-log functions used by add-log-current-defun.
  • If treesit-simple-imenu-settings (see Imenu) is non-nil, it sets up Imenu.

For more information on these built-in tree-sitter features, see Parser-based Font Lock, see Parser-based Indentation, and see Moving over Balanced Expressions.

For supporting mixing of multiple languages in a major mode, see Parsing Text in Multiple Languages.

Besides beginning-of-defun and end-of-defun, Emacs provides some additional functions for working with defuns: treesit-defun-at-point returns the defun node at point, and treesit-defun-name returns the name of a defun node.

Function: treesit-defun-at-point

This function returns the defun node at point, or nil if none is found. It respects treesit-defun-tactic: if its value is top-level, this function returns the top-level defun, and if its value is nested, it returns the immediate enclosing defun.

This function requires treesit-defun-type-regexp to work. If it is nil, this function simply returns nil.

Function: treesit-defun-name node

This function returns the defun name of node. It returns nil if there is no defun name for node, or if node is not a defun node, or if node is nil.

Depending on the language and major mode, the defun names are names like function name, class name, struct name, etc.

If treesit-defun-name-function is nil, this function always returns nil.

Variable: treesit-defun-name-function

If non-nil, this variable’s value should be a function that is called with a node as its argument, and returns the defun name of the node. The function should have the same semantics as treesit-defun-name: if the node is not a defun node, or the node is a defun node but doesn’t have a name, or the node is nil, it should return nil.


37.8 Tree-sitter C API Correspondence

Emacs’ tree-sitter integration doesn’t expose every feature provided by tree-sitter’s C API. Missing features include:

  • Creating a tree cursor and navigating the syntax tree with it.
  • Setting timeout and cancellation flag for a parser.
  • Setting the logger for a parser.
  • Printing a DOT graph of the syntax tree to a file.
  • Copying and modifying a syntax tree. (Emacs doesn’t expose a tree object.)
  • Using (row, column) coordinates as position.
  • Updating a node with changes. (In Emacs, retrieve a new node instead of updating the existing one.)
  • Querying statics of a language grammar.

In addition, Emacs makes some changes to the C API to make the API more convenient and idiomatic:

  • Instead of using byte positions, the Emacs Lisp API uses character positions.
  • Null nodes are converted to nil.

Below is the correspondence between all C API functions and their ELisp counterparts. Sometimes one ELisp function corresponds to multiple C functions, and many C functions don’t have an ELisp counterpart.

ts_parser_new                           treesit-parser-create
ts_parser_delete
ts_parser_set_language
ts_parser_language                      treesit-parser-language
ts_parser_set_included_ranges           treesit-parser-set-included-ranges
ts_parser_included_ranges               treesit-parser-included-ranges
ts_parser_parse
ts_parser_parse_string                  treesit-parse-string
ts_parser_parse_string_encoding
ts_parser_reset
ts_parser_set_timeout_micros
ts_parser_timeout_micros
ts_parser_set_cancellation_flag
ts_parser_cancellation_flag
ts_parser_set_logger
ts_parser_logger
ts_parser_print_dot_graphs
ts_tree_copy
ts_tree_delete
ts_tree_root_node
ts_tree_language
ts_tree_edit
ts_tree_get_changed_ranges
ts_tree_print_dot_graph
ts_node_type                            treesit-node-type
ts_node_symbol
ts_node_start_byte                      treesit-node-start
ts_node_start_point
ts_node_end_byte                        treesit-node-end
ts_node_end_point
ts_node_string                          treesit-node-string
ts_node_is_null
ts_node_is_named                        treesit-node-check
ts_node_is_missing                      treesit-node-check
ts_node_is_extra                        treesit-node-check
ts_node_has_changes
ts_node_has_error                       treesit-node-check
ts_node_parent                          treesit-node-parent
ts_node_child                           treesit-node-child
ts_node_field_name_for_child            treesit-node-field-name-for-child
ts_node_child_count                     treesit-node-child-count
ts_node_named_child                     treesit-node-child
ts_node_named_child_count               treesit-node-child-count
ts_node_child_by_field_name             treesit-node-child-by-field-name
ts_node_child_by_field_id
ts_node_next_sibling                    treesit-node-next-sibling
ts_node_prev_sibling                    treesit-node-prev-sibling
ts_node_next_named_sibling              treesit-node-next-sibling
ts_node_prev_named_sibling              treesit-node-prev-sibling
ts_node_first_child_for_byte            treesit-node-first-child-for-pos
ts_node_first_named_child_for_byte      treesit-node-first-child-for-pos
ts_node_descendant_for_byte_range       treesit-node-descendant-for-range
ts_node_descendant_for_point_range
ts_node_named_descendant_for_byte_range treesit-node-descendant-for-range
ts_node_named_descendant_for_point_range
ts_node_edit
ts_node_eq                              treesit-node-eq
ts_tree_cursor_new
ts_tree_cursor_delete
ts_tree_cursor_reset
ts_tree_cursor_current_node
ts_tree_cursor_current_field_name
ts_tree_cursor_current_field_id
ts_tree_cursor_goto_parent
ts_tree_cursor_goto_next_sibling
ts_tree_cursor_goto_first_child
ts_tree_cursor_goto_first_child_for_byte
ts_tree_cursor_goto_first_child_for_point
ts_tree_cursor_copy
ts_query_new
ts_query_delete
ts_query_pattern_count
ts_query_capture_count
ts_query_string_count
ts_query_start_byte_for_pattern
ts_query_predicates_for_pattern
ts_query_step_is_definite
ts_query_capture_name_for_id
ts_query_string_value_for_id
ts_query_disable_capture
ts_query_disable_pattern
ts_query_cursor_new
ts_query_cursor_delete
ts_query_cursor_exec                    treesit-query-capture
ts_query_cursor_did_exceed_match_limit
ts_query_cursor_match_limit
ts_query_cursor_set_match_limit
ts_query_cursor_set_byte_range
ts_query_cursor_set_point_range
ts_query_cursor_next_match
ts_query_cursor_remove_match
ts_query_cursor_next_capture
ts_language_symbol_count
ts_language_symbol_name
ts_language_symbol_for_name
ts_language_field_count
ts_language_field_name_for_id
ts_language_field_id_for_name
ts_language_symbol_type
ts_language_version

38 Abbrevs and Abbrev Expansion

An abbreviation or abbrev is a string of characters that may be expanded to a longer string. The user can insert the abbrev string and find it replaced automatically with the expansion of the abbrev. This saves typing.

The set of abbrevs currently in effect is recorded in an abbrev table. Each buffer has a local abbrev table, but normally all buffers in the same major mode share one abbrev table. There is also a global abbrev table. Normally both are used.

An abbrev table is represented as an obarray. See Creating and Interning Symbols, for information about obarrays. Each abbreviation is represented by a symbol in the obarray. The symbol’s name is the abbreviation; its value is the expansion; its function definition is the hook function for performing the expansion (see Defining Abbrevs); and its property list cell contains various additional properties, including the use count and the number of times the abbreviation has been expanded (see Abbrev Properties).

Certain abbrevs, called system abbrevs, are defined by a major mode instead of the user. A system abbrev is identified by its non-nil :system property (see Abbrev Properties). When abbrevs are saved to an abbrev file, system abbrevs are omitted. See Saving Abbrevs in Files.

Because the symbols used for abbrevs are not interned in the usual obarray, they will never appear as the result of reading a Lisp expression; in fact, normally they are never used except by the code that handles abbrevs. Therefore, it is safe to use them in a nonstandard way.

If the minor mode Abbrev mode is enabled, the buffer-local variable abbrev-mode is non-nil, and abbrevs are automatically expanded in the buffer. For the user-level commands for abbrevs, see Abbrev Mode in The GNU Emacs Manual.


38.1 Abbrev Tables

This section describes how to create and manipulate abbrev tables.

Function: make-abbrev-table &optional props

This function creates and returns a new, empty abbrev table—an obarray containing no symbols. It is a vector filled with zeros. props is a property list that is applied to the new table (see Abbrev Table Properties).

Function: abbrev-table-p object

This function returns a non-nil value if object is an abbrev table.

Function: clear-abbrev-table abbrev-table

This function undefines all the abbrevs in abbrev-table, leaving it empty.

Function: copy-abbrev-table abbrev-table

This function returns a copy of abbrev-table—a new abbrev table containing the same abbrev definitions. It does not copy any property lists; only the names, values, and functions.

Function: define-abbrev-table tabname definitions &optional docstring &rest props

This function defines tabname (a symbol) as an abbrev table name, i.e., as a variable whose value is an abbrev table. It defines abbrevs in the table according to definitions, a list of elements of the form (abbrevname expansion [hook] [props...]). These elements are passed as arguments to define-abbrev.

The optional string docstring is the documentation string of the variable tabname. The property list props is applied to the abbrev table (see Abbrev Table Properties).

If this function is called more than once for the same tabname, subsequent calls add the definitions in definitions to tabname, rather than overwriting the entire original contents. (A subsequent call only overrides abbrevs explicitly redefined or undefined in definitions.)

Variable: abbrev-table-name-list

This is a list of symbols whose values are abbrev tables. define-abbrev-table adds the new abbrev table name to this list.

Function: insert-abbrev-table-description name &optional human

This function inserts before point a description of the abbrev table named name. The argument name is a symbol whose value is an abbrev table.

If human is non-nil, the description is human-oriented. System abbrevs are listed and identified as such. Otherwise the description is a Lisp expression—a call to define-abbrev-table that would define name as it is currently defined, but without the system abbrevs. (The mode or package using name is supposed to add these to name separately.)


38.2 Defining Abbrevs

define-abbrev is the low-level basic function for defining an abbrev in an abbrev table.

When a major mode defines a system abbrev, it should call define-abbrev and specify t for the :system property. Be aware that any saved non-system abbrevs are restored at startup, i.e., before some major modes are loaded. Therefore, major modes should not assume that their abbrev tables are empty when they are first loaded.

Function: define-abbrev abbrev-table name expansion &optional hook &rest props

This function defines an abbrev named name, in abbrev-table, to expand to expansion and call hook, with properties props (see Abbrev Properties). The return value is name. The :system property in props is treated specially here: if it has the value force, then it will overwrite an existing definition even for a non-system abbrev of the same name.

name should be a string. The argument expansion is normally the desired expansion (a string), or nil to undefine the abbrev. If it is anything but a string or nil, then the abbreviation expands solely by running hook.

The argument hook is a function or nil. If hook is non-nil, then it is called with no arguments after the abbrev is replaced with expansion; point is located at the end of expansion when hook is called.

If hook is a non-nil symbol whose no-self-insert property is non-nil, hook can explicitly control whether to insert the self-inserting input character that triggered the expansion. If hook returns non-nil in this case, that inhibits insertion of the character. By contrast, if hook returns nil, expand-abbrev (or abbrev-insert) also returns nil, as if expansion had not really occurred.

Normally, define-abbrev sets the variable abbrevs-changed to t, if it actually changes the abbrev. This is so that some commands will offer to save the abbrevs. It does not do this for a system abbrev, since those aren’t saved anyway.

User Option: only-global-abbrevs

If this variable is non-nil, it means that the user plans to use global abbrevs only. This tells the commands that define mode-specific abbrevs to define global ones instead. This variable does not alter the behavior of the functions in this section; it is examined by their callers.


38.3 Saving Abbrevs in Files

A file of saved abbrev definitions is actually a file of Lisp code. The abbrevs are saved in the form of a Lisp program to define the same abbrev tables with the same contents. Therefore, you can load the file with load (see How Programs Do Loading). However, the function quietly-read-abbrev-file is provided as a more convenient interface. Emacs automatically calls this function at startup.

User-level facilities such as save-some-buffers can save abbrevs in a file automatically, under the control of variables described here.

User Option: abbrev-file-name

This is the default file name for reading and saving abbrevs. By default, Emacs will look for ~/.emacs.d/abbrev_defs, and, if not found, for ~/.abbrev_defs; if neither file exists, Emacs will create ~/.emacs.d/abbrev_defs.

Function: quietly-read-abbrev-file &optional filename

This function reads abbrev definitions from a file named filename, previously written with write-abbrev-file. If filename is omitted or nil, the file specified in abbrev-file-name is used.

As the name implies, this function does not display any messages.

User Option: save-abbrevs

A non-nil value for save-abbrevs means that Emacs should offer to save abbrevs (if any have changed) when files are saved. If the value is silently, Emacs saves the abbrevs without asking the user. abbrev-file-name specifies the file to save the abbrevs in. The default value is t.

Variable: abbrevs-changed

This variable is set non-nil by defining or altering any abbrevs (except system abbrevs). This serves as a flag for various Emacs commands to offer to save your abbrevs.

Command: write-abbrev-file &optional filename

Save all abbrev definitions (except system abbrevs), for all abbrev tables listed in abbrev-table-name-list, in the file filename, in the form of a Lisp program that when loaded will define the same abbrevs. Tables that do not have any abbrevs to save are omitted. If filename is nil or omitted, abbrev-file-name is used. This function returns nil.


38.4 Looking Up and Expanding Abbreviations

Abbrevs are usually expanded by certain interactive commands, including self-insert-command. This section describes the subroutines used in writing such commands, as well as the variables they use for communication.

Function: abbrev-symbol abbrev &optional table

This function returns the symbol representing the abbrev named abbrev. It returns nil if that abbrev is not defined. The optional second argument table is the abbrev table in which to look it up. If table is nil, this function tries first the current buffer’s local abbrev table, and second the global abbrev table.

Function: abbrev-expansion abbrev &optional table

This function returns the string that abbrev would expand into (as defined by the abbrev tables used for the current buffer). It returns nil if abbrev is not a valid abbrev. The optional argument table specifies the abbrev table to use, as in abbrev-symbol.

Command: expand-abbrev

This command expands the abbrev before point, if any. If point does not follow an abbrev, this command does nothing. To do the expansion, it calls the function that is the value of the abbrev-expand-function variable, with no arguments, and returns whatever that function does.

The default expansion function returns the abbrev symbol if it did expansion, and nil otherwise. If the abbrev symbol has a hook function that is a symbol whose no-self-insert property is non-nil, and if the hook function returns nil as its value, then the default expansion function returns nil, even though expansion did occur.

Function: abbrev-insert abbrev &optional name start end

This function inserts the abbrev expansion of abbrev, replacing the text between start and end. If start is omitted, it defaults to point. name, if non-nil, should be the name by which this abbrev was found (a string); it is used to figure out whether to adjust the capitalization of the expansion. The function returns abbrev if the abbrev was successfully inserted, otherwise it returns nil.

Command: abbrev-prefix-mark &optional arg

This command marks the current location of point as the beginning of an abbrev. The next call to expand-abbrev will use the text from here to point (where it is then) as the abbrev to expand, rather than using the previous word as usual.

First, this command expands any abbrev before point, unless arg is non-nil. (Interactively, arg is the prefix argument.) Then it inserts a hyphen before point, to indicate the start of the next abbrev to be expanded. The actual expansion removes the hyphen.

User Option: abbrev-all-caps

When this is set non-nil, an abbrev entered entirely in upper case is expanded using all upper case. Otherwise, an abbrev entered entirely in upper case is expanded by capitalizing each word of the expansion.

Variable: abbrev-start-location

The value of this variable is a buffer position (an integer or a marker) for expand-abbrev to use as the start of the next abbrev to be expanded. The value can also be nil, which means to use the word before point instead. abbrev-start-location is set to nil each time expand-abbrev is called. This variable is also set by abbrev-prefix-mark.

Variable: abbrev-start-location-buffer

The value of this variable is the buffer for which abbrev-start-location has been set. Trying to expand an abbrev in any other buffer clears abbrev-start-location. This variable is set by abbrev-prefix-mark.

Variable: last-abbrev

This is the abbrev-symbol of the most recent abbrev expanded. This information is left by expand-abbrev for the sake of the unexpand-abbrev command (see Expanding Abbrevs in The GNU Emacs Manual).

Variable: last-abbrev-location

This is the location of the most recent abbrev expanded. This contains information left by expand-abbrev for the sake of the unexpand-abbrev command.

Variable: last-abbrev-text

This is the exact expansion text of the most recent abbrev expanded, after case conversion (if any). Its value is nil if the abbrev has already been unexpanded. This contains information left by expand-abbrev for the sake of the unexpand-abbrev command.

Variable: abbrev-expand-function

The value of this variable is a function that expand-abbrev will call with no arguments to do the expansion. The function can do anything it wants before and after performing the expansion. It should return the abbrev symbol if expansion took place.

The following sample code shows a simple use of abbrev-expand-function. It assumes that foo-mode is a mode for editing certain files in which lines that start with ‘#’ are comments. You want to use Text mode abbrevs for those lines. The regular local abbrev table, foo-mode-abbrev-table is appropriate for all other lines. See Standard Abbrev Tables, for the definitions of local-abbrev-table and text-mode-abbrev-table. See Advising Emacs Lisp Functions, for details of add-function.

(defun foo-mode-abbrev-expand-function (expand)
  (if (not (save-excursion (forward-line 0) (eq (char-after) ?#)))
      ;; Performs normal expansion.
      (funcall expand)
    ;; We're inside a comment: use the text-mode abbrevs.
    (let ((local-abbrev-table text-mode-abbrev-table))
      (funcall expand))))

(add-hook 'foo-mode-hook
          (lambda ()
            (add-function :around (local 'abbrev-expand-function)
                          #'foo-mode-abbrev-expand-function)))

38.5 Standard Abbrev Tables

Here we list the variables that hold the abbrev tables for the preloaded major modes of Emacs.

Variable: global-abbrev-table

This is the abbrev table for mode-independent abbrevs. The abbrevs defined in it apply to all buffers. Each buffer may also have a local abbrev table, whose abbrev definitions take precedence over those in the global table.

Variable: local-abbrev-table

The value of this buffer-local variable is the (mode-specific) abbreviation table of the current buffer. It can also be a list of such tables.

Variable: abbrev-minor-mode-table-alist

The value of this variable is a list of elements of the form (mode . abbrev-table) where mode is the name of a variable: if the variable is bound to a non-nil value, then the abbrev-table is active, otherwise it is ignored. abbrev-table can also be a list of abbrev tables.

Variable: fundamental-mode-abbrev-table

This is the local abbrev table used in Fundamental mode; in other words, it is the local abbrev table in all buffers in Fundamental mode.

Variable: text-mode-abbrev-table

This is the local abbrev table used in Text mode.

Variable: lisp-mode-abbrev-table

This is the local abbrev table used in Lisp mode. It is the parent of the local abbrev table used in Emacs Lisp mode. See Abbrev Table Properties.


38.6 Abbrev Properties

Abbrevs have properties, some of which influence the way they work. You can provide them as arguments to define-abbrev, and manipulate them with the following functions:

Function: abbrev-put abbrev prop val

Set the property prop of abbrev to value val.

Function: abbrev-get abbrev prop

Return the property prop of abbrev, or nil if the abbrev has no such property.

The following properties have special meanings:

:count

This property counts the number of times the abbrev has been expanded. If not explicitly set, it is initialized to 0 by define-abbrev.

:system

If non-nil, this property marks the abbrev as a system abbrev. Such abbrevs are not saved (see Saving Abbrevs in Files).

:enable-function

If non-nil, this property should be a function of no arguments which returns nil if the abbrev should not be used and t otherwise.

:case-fixed

If non-nil, this property indicates that the case of the abbrev’s name is significant and should only match a text with the same pattern of capitalization. It also disables the code that modifies the capitalization of the expansion.


38.7 Abbrev Table Properties

Like abbrevs, abbrev tables have properties, some of which influence the way they work. You can provide them as arguments to define-abbrev-table, and manipulate them with the functions:

Function: abbrev-table-put table prop val

Set the property prop of abbrev table table to value val.

Function: abbrev-table-get table prop

Return the property prop of abbrev table table, or nil if table has no such property.

The following properties have special meaning:

:enable-function

This is like the :enable-function abbrev property except that it applies to all abbrevs in the table. It is used before even trying to find the abbrev before point, so it can dynamically modify the abbrev table.

:case-fixed

This is like the :case-fixed abbrev property except that it applies to all abbrevs in the table.

:regexp

If non-nil, this property is a regular expression that indicates how to extract the name of the abbrev before point, before looking it up in the table. When the regular expression matches before point, the abbrev name is expected to be in submatch 1. If this property is nil, the default is to use backward-word and forward-word to find the name. This property allows the use of abbrevs whose name contains characters of non-word syntax.

:parents

This property holds a list of tables from which to inherit other abbrevs.

:abbrev-table-modiff

This property holds a counter incremented each time a new abbrev is added to the table.


39 Threads

Emacs Lisp provides a limited form of concurrency, called threads. All the threads in a given instance of Emacs share the same memory. Concurrency in Emacs Lisp is “mostly cooperative”, meaning that Emacs will only switch execution between threads at well-defined times. However, the Emacs thread support has been designed in a way to later allow more fine-grained concurrency, and correct programs should not rely on cooperative threading.

Currently, thread switching will occur upon explicit request via thread-yield, when waiting for keyboard input or for process output from asynchronous processes (e.g., during accept-process-output), or during blocking operations relating to threads, such as mutex locking or thread-join.

Emacs Lisp provides primitives to create and control threads, and also to create and control mutexes and condition variables, useful for thread synchronization.

While global variables are shared among all Emacs Lisp threads, local variables are not—a dynamic let binding is local. Each thread also has its own current buffer (see The Current Buffer) and its own match data (see The Match Data).

Note that let bindings are treated specially by the Emacs Lisp implementation. There is no way to duplicate this unwinding and rewinding behavior other than by using let. For example, a manual implementation of let written using unwind-protect cannot arrange for variable values to be thread-specific.

In the case of lexical bindings (see Scoping Rules for Variable Bindings), a closure is an object like any other in Emacs Lisp, and bindings in a closure are shared by any threads invoking the closure.


39.1 Basic Thread Functions

Threads can be created and waited for. A thread cannot be exited directly, but the current thread can be exited implicitly, and other threads can be signaled.

Function: make-thread function &optional name

Create a new thread of execution which invokes function. When function returns, the thread exits.

The new thread is created with no local variable bindings in effect. The new thread’s current buffer is inherited from the current thread.

name can be supplied to give a name to the thread. The name is used for debugging and informational purposes only; it has no meaning to Emacs. If name is provided, it must be a string.

This function returns the new thread.

Function: threadp object

This function returns t if object represents an Emacs thread, nil otherwise.

Function: thread-join thread

Block until thread exits, or until the current thread is signaled. It returns the result of the thread function. If thread has already exited, this returns immediately.

Function: thread-signal thread error-symbol data

Like signal (see How to Signal an Error), but the signal is delivered in the thread thread. If thread is the current thread, then this just calls signal immediately. Otherwise, thread will receive the signal as soon as it becomes current. If thread was blocked by a call to mutex-lock, condition-wait, or thread-join; thread-signal will unblock it.

If thread is the main thread, the signal is not propagated there. Instead, it is shown as message in the main thread.

Function: thread-yield

Yield execution to the next runnable thread.

Function: thread-name thread

Return the name of thread, as specified to make-thread.

Function: thread-live-p thread

Return t if thread is alive, or nil if it is not. A thread is alive as long as its function is still executing.

Function: thread--blocker thread

Return the object that thread is waiting on. This function is primarily intended for debugging, and is given a “double hyphen” name to indicate that.

If thread is blocked in thread-join, this returns the thread for which it is waiting.

If thread is blocked in mutex-lock, this returns the mutex.

If thread is blocked in condition-wait, this returns the condition variable.

Otherwise, this returns nil.

Function: current-thread

Return the current thread.

Function: all-threads

Return a list of all the live thread objects. A new list is returned by each invocation.

Variable: main-thread

This variable keeps the main thread Emacs is running, or nil if Emacs is compiled without thread support.

When code run by a thread signals an error that is unhandled, the thread exits. Other threads can access the error form which caused the thread to exit using the following function.

Function: thread-last-error &optional cleanup

This function returns the last error form recorded when a thread exited due to an error. Each thread that exits abnormally overwrites the form stored by the previous thread’s error with a new value, so only the last one can be accessed. If cleanup is non-nil, the stored form is reset to nil.


39.2 Mutexes

A mutex is an exclusive lock. At any moment, zero or one threads may own a mutex. If a thread attempts to acquire a mutex, and the mutex is already owned by some other thread, then the acquiring thread will block until the mutex becomes available.

Emacs Lisp mutexes are of a type called recursive, which means that a thread can re-acquire a mutex it owns any number of times. A mutex keeps a count of how many times it has been acquired, and each acquisition of a mutex must be paired with a release. The last release by a thread of a mutex reverts it to the unowned state, potentially allowing another thread to acquire the mutex.

Function: mutexp object

This function returns t if object represents an Emacs mutex, nil otherwise.

Function: make-mutex &optional name

Create a new mutex and return it. If name is specified, it is a name given to the mutex. It must be a string. The name is for debugging purposes only; it has no meaning to Emacs.

Function: mutex-name mutex

Return the name of mutex, as specified to make-mutex.

Function: mutex-lock mutex

This will block until this thread acquires mutex, or until this thread is signaled using thread-signal. If mutex is already owned by this thread, this simply returns.

Function: mutex-unlock mutex

Release mutex. If mutex is not owned by this thread, this will signal an error.

Macro: with-mutex mutex body…

This macro is the simplest and safest way to evaluate forms while holding a mutex. It acquires mutex, invokes body, and then releases mutex. It returns the result of body.


39.3 Condition Variables

A condition variable is a way for a thread to block until some event occurs. A thread can wait on a condition variable, to be woken up when some other thread notifies the condition.

A condition variable is associated with a mutex and, conceptually, with some condition. For proper operation, the mutex must be acquired, and then a waiting thread must loop, testing the condition and waiting on the condition variable. For example:

(with-mutex mutex
  (while (not global-variable)
    (condition-wait cond-var)))

The mutex ensures atomicity, and the loop is for robustness—there may be spurious notifications.

Similarly, the mutex must be held before notifying the condition. The typical, and best, approach is to acquire the mutex, make the changes associated with this condition, and then notify it:

(with-mutex mutex
  (setq global-variable (some-computation))
  (condition-notify cond-var))
Function: make-condition-variable mutex &optional name

Make a new condition variable associated with mutex. If name is specified, it is a name given to the condition variable. It must be a string. The name is for debugging purposes only; it has no meaning to Emacs.

Function: condition-variable-p object

This function returns t if object represents a condition variable, nil otherwise.

Function: condition-wait cond

Wait for another thread to notify cond, a condition variable. This function will block until the condition is notified, or until a signal is delivered to this thread using thread-signal.

It is an error to call condition-wait without holding the condition’s associated mutex.

condition-wait releases the associated mutex while waiting. This allows other threads to acquire the mutex in order to notify the condition.

Function: condition-notify cond &optional all

Notify cond. The mutex with cond must be held before calling this. Ordinarily a single waiting thread is woken by condition-notify; but if all is not nil, then all threads waiting on cond are notified.

condition-notify releases the associated mutex while waiting. This allows other threads to acquire the mutex in order to wait on the condition.

Function: condition-name cond

Return the name of cond, as passed to make-condition-variable.

Function: condition-mutex cond

Return the mutex associated with cond. Note that the associated mutex cannot be changed.


39.4 The Thread List

The list-threads command lists all the currently alive threads. In the resulting buffer, each thread is identified either by the name passed to make-thread (see Basic Thread Functions), or by its unique internal identifier if it was not created with a name. The status of each thread at the time of the creation or last update of the buffer is shown, in addition to the object the thread was blocked on at the time, if it was blocked.

Variable: thread-list-refresh-seconds

The *Threads* buffer will automatically update twice per second. You can make the refresh rate faster or slower by customizing this variable.

Here are the commands available in the thread list buffer:

b

Show a backtrace of the thread at point. This will show where in its code the thread had yielded or was blocked at the moment you pressed b. Be aware that the backtrace is a snapshot; the thread could have meanwhile resumed execution, and be in a different state, or could have exited.

You may use g in the thread’s backtrace buffer to get an updated backtrace, as backtrace buffers do not automatically update. See Backtraces, for a description of backtraces and the other commands which work on them.

s

Signal the thread at point. After s, type q to send a quit signal or e to send an error signal. Threads may implement handling of signals, but the default behavior is to exit on any signal. Therefore you should only use this command if you understand how to restart the target thread, because your Emacs session may behave incorrectly if necessary threads are killed.

g

Update the list of threads and their statuses.


40 Processes

In the terminology of operating systems, a process is a space in which a program can execute. Emacs runs in a process. Emacs Lisp programs can invoke other programs in processes of their own. These are called subprocesses or child processes of the Emacs process, which is their parent process.

A subprocess of Emacs may be synchronous or asynchronous, depending on how it is created. When you create a synchronous subprocess, the Lisp program waits for the subprocess to terminate before continuing execution. When you create an asynchronous subprocess, it can run in parallel with the Lisp program. This kind of subprocess is represented within Emacs by a Lisp object which is also called a “process”. Lisp programs can use this object to communicate with the subprocess or to control it. For example, you can send signals, obtain status information, receive output from the process, or send input to it.

In addition to processes that run programs, Lisp programs can open connections of several types to devices or processes running on the same machine or on other machines. The supported connection types are: TCP and UDP network connections, serial port connections, and pipe connections. Each such connection is also represented by a process object.

Function: processp object

This function returns t if object represents an Emacs process object, nil otherwise. The process object can represent a subprocess running a program or a connection of any supported type.

In addition to subprocesses of the current Emacs session, you can also access other processes running on your machine. See Accessing Other Processes.


40.1 Functions that Create Subprocesses

There are three primitives that create a new subprocess in which to run a program. One of them, make-process, creates an asynchronous process and returns a process object (see Creating an Asynchronous Process). The other two, call-process and call-process-region, create a synchronous process and do not return a process object (see Creating a Synchronous Process). There are various higher-level functions that make use of these primitives to run particular types of process.

Synchronous and asynchronous processes are explained in the following sections. Since the three functions are all called in a similar fashion, their common arguments are described here.

In all cases, the functions specify the program to be run. An error is signaled if the file is not found or cannot be executed. If the file name is relative, the variable exec-path contains a list of directories to search. Emacs initializes exec-path when it starts up, based on the value of the environment variable PATH. The standard file name constructs, ‘~’, ‘.’, and ‘..’, are interpreted as usual in exec-path, but environment variable substitutions (‘$HOME’, etc.) are not recognized; use substitute-in-file-name to perform them (see Functions that Expand Filenames). nil in this list refers to default-directory.

Executing a program can also try adding suffixes to the specified name:

User Option: exec-suffixes

This variable is a list of suffixes (strings) to try adding to the specified program file name. The list should include "" if you want the name to be tried exactly as specified. The default value is system-dependent.

Please note: The argument program contains only the name of the program file; it may not contain any command-line arguments. You must use a separate argument, args, to provide those, as described below.

Each of the subprocess-creating functions has a buffer-or-name argument that specifies where the output from the program will go. It should be a buffer or a buffer name; if it is a buffer name, that will create the buffer if it does not already exist. It can also be nil, which says to discard the output, unless a custom filter function handles it. (See Process Filter Functions, and Reading and Printing Lisp Objects.) Normally, you should avoid having multiple processes send output to the same buffer because their output would be intermixed randomly. For synchronous processes, you can send the output to a file instead of a buffer (and the corresponding argument is therefore more appropriately called destination). By default, both standard output and standard error streams go to the same destination, but all the 3 primitives allow optionally to direct the standard error stream to a different destination.

All three of the subprocess-creating functions allow to specify command-line arguments for the process to run. For call-process and call-process-region, these come in the form of a &rest argument, args. For make-process, both the program to run and its command-line arguments are specified as a list of strings. The command-line arguments must all be strings, and they are supplied to the program as separate argument strings. Wildcard characters and other shell constructs have no special meanings in these strings, since the strings are passed directly to the specified program.

The subprocess inherits its environment from Emacs, but you can specify overrides for it with process-environment. See Operating System Environment. The subprocess gets its current directory from the value of default-directory.

Variable: exec-directory

The value of this variable is a string, the name of a directory that contains programs that come with GNU Emacs and are intended for Emacs to invoke. The program movemail is an example of such a program; Rmail uses it to fetch new mail from an inbox.

User Option: exec-path

The value of this variable is a list of directories to search for programs to run in subprocesses. Each element is either the name of a directory (i.e., a string), or nil, which stands for the default directory (which is the value of default-directory). See executable-find, for the details of this search.

The value of exec-path is used by call-process and start-process when the program argument is not an absolute file name.

Generally, you should not modify exec-path directly. Instead, ensure that your PATH environment variable is set appropriately before starting Emacs. Trying to modify exec-path independently of PATH can lead to confusing results.

Function: exec-path

This function is an extension of the variable exec-path. If default-directory indicates a remote directory, this function returns a list of directories used for searching programs on the respective remote host. In case of a local default-directory, the function returns just the value of the variable exec-path.


40.2 Shell Arguments

Lisp programs sometimes need to run a shell and give it a command that contains file names that were specified by the user. These programs ought to be able to support any valid file name. But the shell gives special treatment to certain characters, and if these characters occur in the file name, they will confuse the shell. To handle these characters, use the function shell-quote-argument:

Function: shell-quote-argument argument &optional posix

This function returns a string that represents, in shell syntax, an argument whose actual contents are argument. It should work reliably to concatenate the return value into a shell command and then pass it to a shell for execution.

Precisely what this function does depends on your operating system. The function is designed to work with the syntax of your system’s standard shell; if you use an unusual shell, you will need to redefine this function. See Security Considerations.

;; This example shows the behavior on GNU and Unix systems.
(shell-quote-argument "foo > bar")
     ⇒ "foo\\ \\>\\ bar"

;; This example shows the behavior on MS-DOS and MS-Windows.
(shell-quote-argument "foo > bar")
     ⇒ "\"foo > bar\""

Here’s an example of using shell-quote-argument to construct a shell command:

(concat "diff -u "
        (shell-quote-argument oldfile)
        " "
        (shell-quote-argument newfile))

If the optional posix argument is non-nil, argument is quoted according to POSIX shell quoting rules, regardless of the system’s shell. This is useful when your shell could run on a remote host, which requires a POSIX shell in general.

(shell-quote-argument "foo > bar" (file-remote-p default-directory))

The following two functions are useful for combining a list of individual command-line argument strings into a single string, and taking a string apart into a list of individual command-line arguments. These functions are mainly intended for converting user input in the minibuffer, a Lisp string, into a list of string arguments to be passed to make-process, call-process or start-process, or for converting such lists of arguments into a single Lisp string to be presented in the minibuffer or echo area. Note that if a shell is involved (e.g., if using call-process-shell-command), arguments should still be protected by shell-quote-argument; combine-and-quote-strings is not intended to protect special characters from shell evaluation.

Function: split-string-shell-command string

This function splits string into substrings, respecting double and single quotes, as well as backslash quoting.

(split-string-shell-command "ls /tmp/'foo bar'")
     ⇒ ("ls" "/tmp/foo bar")
Function: split-string-and-unquote string &optional separators

This function splits string into substrings at matches for the regular expression separators, like split-string does (see Creating Strings); in addition, it removes quoting from the substrings. It then makes a list of the substrings and returns it.

If separators is omitted or nil, it defaults to "\\s-+", which is a regular expression that matches one or more characters with whitespace syntax (see Table of Syntax Classes).

This function supports two types of quoting: enclosing a whole string in double quotes "…", and quoting individual characters with a backslash escape ‘\’. The latter is also used in Lisp strings, so this function can handle those as well.

Function: combine-and-quote-strings list-of-strings &optional separator

This function concatenates list-of-strings into a single string, quoting each string as necessary. It also sticks the separator string between each pair of strings; if separator is omitted or nil, it defaults to " ". The return value is the resulting string.

The strings in list-of-strings that need quoting are those that include separator as their substring. Quoting a string encloses it in double quotes "…". In the simplest case, if you are consing a command from the individual command-line arguments, every argument that includes embedded blanks will be quoted.


40.3 Creating a Synchronous Process

After a synchronous process is created, Emacs waits for the process to terminate before continuing. Starting Dired on GNU or Unix28 is an example of this: it runs ls in a synchronous process, then modifies the output slightly. Because the process is synchronous, the entire directory listing arrives in the buffer before Emacs tries to do anything with it.

While Emacs waits for the synchronous subprocess to terminate, the user can quit by typing C-g. The first C-g tries to kill the subprocess with a SIGINT signal; but it waits until the subprocess actually terminates before quitting. If during that time the user types another C-g, that kills the subprocess instantly with SIGKILL and quits immediately (except on MS-DOS, where killing other processes doesn’t work). See Quitting.

The synchronous subprocess functions return an indication of how the process terminated.

The output from a synchronous subprocess is generally decoded using a coding system, much like text read from a file. The input sent to a subprocess by call-process-region is encoded using a coding system, much like text written into a file. See Coding Systems.

Function: call-process program &optional infile destination display &rest args

This function calls program and waits for it to finish.

The current working directory of the subprocess is set to the current buffer’s value of default-directory if that is local (as determined by unhandled-file-name-directory), or "~" otherwise. If you want to run a process in a remote directory use process-file.

The standard input for the new process comes from file infile if infile is not nil, and from the null device otherwise. The argument destination says where to put the process output. Here are the possibilities:

a buffer

Insert the output in that buffer, before point. This includes both the standard output stream and the standard error stream of the process.

a buffer name (a string)

Insert the output in a buffer with that name, before point.

t

Insert the output in the current buffer, before point.

nil

Discard the output.

0

Discard the output, and return nil immediately without waiting for the subprocess to finish.

In this case, the process is not truly synchronous, since it can run in parallel with Emacs; but you can think of it as synchronous in that Emacs is essentially finished with the subprocess as soon as this function returns.

MS-DOS doesn’t support asynchronous subprocesses, so this option doesn’t work there.

(:file file-name)

Send the output to the file name specified, overwriting it if it already exists.

(real-destination error-destination)

Keep the standard output stream separate from the standard error stream; deal with the ordinary output as specified by real-destination, and dispose of the error output according to error-destination. If error-destination is nil, that means to discard the error output, t means mix it with the ordinary output, and a string specifies a file name to redirect error output into.

You can’t directly specify a buffer to put the error output in; that is too difficult to implement. But you can achieve this result by sending the error output to a temporary file and then inserting the file into a buffer when the subprocess finishes.

If display is non-nil, then call-process redisplays the buffer as output is inserted. (However, if the coding system chosen for decoding output is undecided, meaning deduce the encoding from the actual data, then redisplay sometimes cannot continue once non-ASCII characters are encountered. There are fundamental reasons why it is hard to fix this; see Receiving Output from Processes.)

Otherwise the function call-process does no redisplay, and the results become visible on the screen only when Emacs redisplays that buffer in the normal course of events.

The remaining arguments, args, are strings that specify command line arguments for the program. Each string is passed to program as a separate argument.

The value returned by call-process (unless you told it not to wait) indicates the reason for process termination. A number gives the exit status of the subprocess; 0 means success, and any other value means failure. If the process terminated with a signal, call-process returns a string describing the signal. If you told call-process not to wait, it returns nil.

In the examples below, the buffer ‘foo’ is current.

(call-process "pwd" nil t)
     ⇒ 0

---------- Buffer: foo ----------
/home/lewis/manual
---------- Buffer: foo ----------

(call-process "grep" nil "bar" nil "lewis" "/etc/passwd")
     ⇒ 0

---------- Buffer: bar ----------
lewis:x:1001:1001:Bil Lewis,,,,:/home/lewis:/bin/bash

---------- Buffer: bar ----------

Here is an example of the use of call-process, as used to be found in the definition of the insert-directory function:

(call-process insert-directory-program nil t nil switches
              (if full-directory-p
                  (concat (file-name-as-directory file) ".")
                file))
Function: process-file program &optional infile buffer display &rest args

This function processes files synchronously in a separate process. It is similar to call-process, but may invoke a file name handler based on the value of the variable default-directory, which specifies the current working directory of the subprocess.

The arguments are handled in almost the same way as for call-process, with the following differences:

Some file name handlers may not support all combinations and forms of the arguments infile, buffer, and display. For example, some file name handlers might behave as if display were nil, regardless of the value actually passed. As another example, some file name handlers might not support separating standard output and error output by way of the buffer argument.

If a file name handler is invoked, it determines the program to run based on the first argument program. For instance, suppose that a handler for remote files is invoked. Then the path that is used for searching for the program might be different from exec-path.

The second argument infile may invoke a file name handler. The file name handler could be different from the handler chosen for the process-file function itself. (For example, default-directory could be on one remote host, and infile on a different remote host. Or default-directory could be non-special, whereas infile is on a remote host.)

If buffer is a list of the form (real-destination error-destination), and error-destination names a file, then the same remarks as for infile apply.

The remaining arguments (args) will be passed to the process verbatim. Emacs is not involved in processing file names that are present in args. To avoid confusion, it may be best to avoid absolute file names in args, but rather to specify all file names as relative to default-directory. The function file-relative-name is useful for constructing such relative file names. Alternatively, you can use file-local-name (see Making Certain File Names “Magic”) to obtain an absolute file name as seen from the remote host’s perspective.

Variable: process-file-side-effects

This variable indicates whether a call of process-file changes remote files.

By default, this variable is always set to t, meaning that a call of process-file could potentially change any file on a remote host. When set to nil, a file name handler could optimize its behavior with respect to remote file attribute caching.

You should only ever change this variable with a let-binding; never with setq.

User Option: process-file-return-signal-string

This user option indicates whether a call of process-file returns a string describing the signal interrupting a remote process.

When a process returns an exit code greater than 128, it is interpreted as a signal. process-file requires to return a string describing this signal.

Since there are processes violating this rule, returning exit codes greater than 128 which are not bound to a signal, process-file returns always the exit code as natural number for remote processes. Setting this user option to non-nil forces process-file to interpret such exit codes as signals, and to return a corresponding string.

Function: call-process-region start end program &optional delete destination display &rest args

This function sends the text from start to end as standard input to a process running program. It deletes the text sent if delete is non-nil; this is useful when destination is t, to insert the output in the current buffer in place of the input.

The arguments destination and display control what to do with the output from the subprocess, and whether to update the display as it comes in. For details, see the description of call-process, above. If destination is the integer 0, call-process-region discards the output and returns nil immediately, without waiting for the subprocess to finish (this only works if asynchronous subprocesses are supported; i.e., not on MS-DOS).

The remaining arguments, args, are strings that specify command line arguments for the program.

The return value of call-process-region is just like that of call-process: nil if you told it to return without waiting; otherwise, a number or string which indicates how the subprocess terminated.

In the following example, we use call-process-region to run the cat utility, with standard input being the first five characters in buffer ‘foo’ (the word ‘input’). cat copies its standard input into its standard output. Since the argument destination is t, this output is inserted in the current buffer.

---------- Buffer: foo ----------
input∗
---------- Buffer: foo ----------

(call-process-region 1 6 "cat" nil t)
     ⇒ 0

---------- Buffer: foo ----------
inputinput∗
---------- Buffer: foo ----------

For example, the shell-command-on-region command uses call-shell-region in a manner similar to this:

(call-shell-region
 start end
 command              ; shell command
 nil                  ; do not delete region
 buffer)              ; send output to buffer
Function: call-process-shell-command command &optional infile destination display

This function executes the shell command command synchronously. The other arguments are handled as in call-process. An old calling convention allowed passing any number of additional arguments after display, which were concatenated to command; this is still supported, but strongly discouraged.

Function: process-file-shell-command command &optional infile destination display

This function is like call-process-shell-command, but uses process-file internally. Depending on default-directory, command can be executed also on remote hosts. An old calling convention allowed passing any number of additional arguments after display, which were concatenated to command; this is still supported, but strongly discouraged.

Function: call-shell-region start end command &optional delete destination

This function sends the text from start to end as standard input to an inferior shell running command. This function is similar than call-process-region, with process being a shell. The arguments delete, destination and the return value are like in call-process-region. Note that this function doesn’t accept additional arguments.

Function: shell-command-to-string command

This function executes command (a string) as a shell command, then returns the command’s output as a string.

Function: process-lines program &rest args

This function runs program, waits for it to finish, and returns its output as a list of strings. Each string in the list holds a single line of text output by the program; the end-of-line characters are stripped from each line. The arguments beyond program, args, are strings that specify command-line arguments with which to run the program.

If program exits with a non-zero exit status, this function signals an error.

This function works by calling call-process, so program output is decoded in the same way as for call-process.

Function: process-lines-ignore-status program &rest args

This function is just like process-lines, but does not signal an error if program exits with a non-zero exit status.


40.4 Creating an Asynchronous Process

In this section, we describe how to create an asynchronous process. After an asynchronous process is created, it runs in parallel with Emacs, and Emacs can communicate with it using the functions described in the following sections (see Sending Input to Processes, and see Receiving Output from Processes). Note that process communication is only partially asynchronous: Emacs sends and receives data to and from a process only when those functions are called.

An asynchronous process is controlled either via a pty (pseudo-terminal) or a pipe. The choice of pty or pipe is made when creating the process, by default based on the value of the variable process-connection-type (see below). If available, ptys are usually preferable for processes visible to the user, as in Shell mode, because they allow for job control (C-c, C-z, etc.) between the process and its children, and because interactive programs treat ptys as terminal devices, whereas pipes don’t support these features. However, for subprocesses used by Lisp programs for internal purposes (i.e., no user interaction with the subprocess is required), where significant amounts of data need to be exchanged between the subprocess and the Lisp program, it is often better to use a pipe, because pipes are more efficient. Also, the total number of ptys is limited on many systems, and it is good not to waste them unnecessarily.

Function: make-process &rest args

This function is the basic low-level primitive for starting asynchronous subprocesses. It returns a process object representing the subprocess. Compared to the more high-level start-process, described below, it takes keyword arguments, is more flexible, and allows to specify process filters and sentinels in a single call.

The arguments args are a list of keyword/argument pairs. Omitting a keyword is always equivalent to specifying it with value nil. Here are the meaningful keywords:

:name name

Use the string name as the process name; if a process with this name already exists, then name is modified (by appending ‘<1>’, etc.) to be unique.

:buffer buffer

Use buffer as the process buffer. If the value is nil, the subprocess is not associated with any buffer.

:command command

Use command as the command line of the process. The value should be a list starting with the program’s executable file name, followed by strings to give to the program as its arguments. If the first element of the list is nil, Emacs opens a new pseudoterminal (pty) and associates its input and output with buffer, without actually running any program; the rest of the list elements are ignored in that case.

:coding coding

If coding is a symbol, it specifies the coding system to be used for both reading and writing of data from and to the connection. If coding is a cons cell (decoding . encoding), then decoding will be used for reading and encoding for writing. The coding system used for encoding the data written to the program is also used for encoding the command-line arguments (but not the program itself, whose file name is encoded as any other file name; see file-name-coding-system).

If coding is nil, the default rules for finding the coding system will apply. See Default Coding Systems.

:connection-type type

Initialize the type of device used to communicate with the subprocess. Possible values are pty to use a pty, pipe to use a pipe, or nil to use the default derived from the value of the process-connection-type variable. If type is a cons cell (input . output), then input will be used for standard input and output for standard output (and standard error if :stderr is nil).

On systems where ptys are not available (MS-Windows), this parameter is ignored, and pipes are used unconditionally.

:noquery query-flag

Initialize the process query flag to query-flag. See Querying Before Exit.

:stop stopped

If provided, stopped must be nil; it is an error to use any non-nil value. The :stop key is ignored otherwise and is retained for compatibility with other process types such as pipe processes. Asynchronous subprocesses never start in the stopped state.

:filter filter

Initialize the process filter to filter. If not specified, a default filter will be provided, which can be overridden later. See Process Filter Functions.

:sentinel sentinel

Initialize the process sentinel to sentinel. If not specified, a default sentinel will be used, which can be overridden later. See Sentinels: Detecting Process Status Changes.

:stderr stderr

Associate stderr with the standard error of the process. A non-nil value should be either a buffer or a pipe process created with make-pipe-process, described below. If stderr is nil, standard error is mixed with standard output, and both are sent to buffer or filter.

If stderr is a buffer, Emacs will create a pipe process, the standard error process. This process will have the default filter (see Process Filter Functions), sentinel (see Sentinels: Detecting Process Status Changes), and coding systems (see Default Coding Systems). On the other hand, it will use query-flag as its query-on-exit flag (see Querying Before Exit). It will be associated with the stderr buffer (see Process Buffers) and send its output (which is the standard error of the main process) there. To get the process object for the standard error process, pass the stderr buffer to get-buffer-process.

If stderr is a pipe process, Emacs will use it as standard error process for the new process.

:file-handler file-handler

If file-handler is non-nil, then look for a file name handler for the current buffer’s default-directory, and invoke that file name handler to make the process. If there is no such handler, proceed as if file-handler were nil.

The original argument list, modified with the actual connection information, is available via the process-contact function.

The current working directory of the subprocess is set to the current buffer’s value of default-directory if that is local (as determined by unhandled-file-name-directory), or ~ otherwise. If you want to run a process in a remote directory, pass :file-handler t to make-process. In that case, the current working directory is the local name component of default-directory (as determined by file-local-name).

Depending on the implementation of the file name handler, it might not be possible to apply filter or sentinel to the resulting process object. The :stderr argument cannot be a pipe process, file name handlers do not support pipe processes for this. A buffer as :stderr argument is accepted, its contents is shown without the use of pipe processes. See Process Filter Functions, Sentinels: Detecting Process Status Changes, and Accepting Output from Processes.

Some file name handlers may not support make-process. In such cases, this function does nothing and returns nil.

Function: make-pipe-process &rest args

This function creates a bidirectional pipe which can be attached to a child process. This is useful with the :stderr keyword of make-process. The function returns a process object.

The arguments args are a list of keyword/argument pairs. Omitting a keyword is always equivalent to specifying it with value nil.

Here are the meaningful keywords:

:name name

Use the string name as the process name. As with make-process, it is modified if necessary to make it unique.

:buffer buffer

Use buffer as the process buffer.

:coding coding

If coding is a symbol, it specifies the coding system to be used for both reading and writing of data from and to the connection. If coding is a cons cell (decoding . encoding), then decoding will be used for reading and encoding for writing.

If coding is nil, the default rules for finding the coding system will apply. See Default Coding Systems.

:noquery query-flag

Initialize the process query flag to query-flag. See Querying Before Exit.

:stop stopped

If stopped is non-nil, start the process in the stopped state. In the stopped state, a pipe process does not accept incoming data, but you can send outgoing data. The stopped state is set by stop-process and cleared by continue-process (see Sending Signals to Processes).

:filter filter

Initialize the process filter to filter. If not specified, a default filter will be provided, which can be changed later. See Process Filter Functions.

:sentinel sentinel

Initialize the process sentinel to sentinel. If not specified, a default sentinel will be used, which can be changed later. See Sentinels: Detecting Process Status Changes.

The original argument list, modified with the actual connection information, is available via the process-contact function.

Function: start-process name buffer-or-name program &rest args

This function is a higher-level wrapper around make-process, exposing an interface that is similar to call-process. It creates a new asynchronous subprocess and starts the specified program running in it. It returns a process object that stands for the new subprocess in Lisp. The argument name specifies the name for the process object; as with make-process, it is modified if necessary to make it unique. The buffer buffer-or-name is the buffer to associate with the process.

If program is nil, Emacs opens a new pseudoterminal (pty) and associates its input and output with buffer-or-name, without creating a subprocess. In that case, the remaining arguments args are ignored.

The rest of args are strings that specify command line arguments for the subprocess.

In the example below, the first process is started and runs (rather, sleeps) for 100 seconds (the output buffer ‘foo’ is created immediately). Meanwhile, the second process is started, and given the name ‘my-process<1>’ for the sake of uniqueness. It inserts the directory listing at the end of the buffer ‘foo’, before the first process finishes. Then it finishes, and a message to that effect is inserted in the buffer. Much later, the first process finishes, and another message is inserted in the buffer for it.

(start-process "my-process" "foo" "sleep" "100")
     ⇒ #<process my-process>

(start-process "my-process" "foo" "ls" "-l" "/bin")
     ⇒ #<process my-process<1>>

---------- Buffer: foo ----------
total 8336
-rwxr-xr-x 1 root root 971384 Mar 30 10:14 bash
-rwxr-xr-x 1 root root 146920 Jul  5  2011 bsd-csh
…
-rwxr-xr-x 1 root root 696880 Feb 28 15:55 zsh4

Process my-process<1> finished

Process my-process finished
---------- Buffer: foo ----------
Function: start-file-process name buffer-or-name program &rest args

Like start-process, this function starts a new asynchronous subprocess running program in it, and returns its process object.

The difference from start-process is that this function may invoke a file name handler based on the value of default-directory. This handler ought to run program, perhaps on the local host, perhaps on a remote host that corresponds to default-directory. In the latter case, the local part of default-directory becomes the working directory of the process.

This function does not try to invoke file name handlers for program or for the rest of args. For that reason, if program or any of args use the remote-file syntax (see Making Certain File Names “Magic”), they must be converted either to file names relative to default-directory, or to names that identify the files locally on the remote host, by running them through file-local-name.

Depending on the implementation of the file name handler, it might not be possible to apply process-filter or process-sentinel to the resulting process object. See Process Filter Functions, and Sentinels: Detecting Process Status Changes.

Some file name handlers may not support start-file-process (for example the function ange-ftp-hook-function). In such cases, this function does nothing and returns nil.

Function: start-process-shell-command name buffer-or-name command

This function is like start-process, except that it uses a shell to execute the specified command. The argument command is a shell command string. The variable shell-file-name specifies which shell to use.

The point of running a program through the shell, rather than directly with make-process or start-process, is so that you can employ shell features such as wildcards in the arguments. It follows that if you include any arbitrary user-specified arguments in the command, you should quote them with shell-quote-argument first, so that any special shell characters do not have their special shell meanings. See Shell Arguments. Of course, when executing commands based on user input you should also consider the security implications.

Function: start-file-process-shell-command name buffer-or-name command

This function is like start-process-shell-command, but uses start-file-process internally. Because of this, command can also be executed on remote hosts, depending on default-directory.

Variable: process-connection-type

This variable controls the type of device used to communicate with asynchronous subprocesses. If it is non-nil, then ptys are used, when available. Otherwise, pipes are used.

The value of process-connection-type takes effect when make-process or start-process is called. So you can specify how to communicate with one subprocess by binding the variable around the call to these functions.

Note that the value of this variable is ignored when make-process is called with a non-nil value of the :stderr parameter; in that case, Emacs will communicate with the process using pipes. It is also ignored if ptys are unavailable (MS-Windows).

(let ((process-connection-type nil))  ; use a pipe
  (start-process …))

To determine whether a given subprocess actually got a pipe or a pty, use the function process-tty-name (see Process Information).

Variable: process-error-pause-time

If a process sentinel/filter function has an error, Emacs will (by default) pause Emacs for process-error-pause-time seconds after displaying this error, so that users will see the error in question. However, this can lead to situations where Emacs becomes unresponsive (if there’s a lot of these errors happening), so this can be disabled by setting process-error-pause-time to 0.


40.5 Deleting Processes

Deleting a process disconnects Emacs immediately from the subprocess. Processes are deleted automatically after they terminate, but not necessarily right away. You can delete a process explicitly at any time. If you explicitly delete a terminated process before it is deleted automatically, no harm results. Deleting a running process sends a signal to terminate it (and its child processes, if any), and calls the process sentinel. See Sentinels: Detecting Process Status Changes.

When a process is deleted, the process object itself continues to exist as long as other Lisp objects point to it. All the Lisp primitives that work on process objects accept deleted processes, but those that do I/O or send signals will report an error. The process mark continues to point to the same place as before, usually into a buffer where output from the process was being inserted.

User Option: delete-exited-processes

This variable controls automatic deletion of processes that have terminated (due to calling exit or to a signal). If it is nil, then they continue to exist until the user runs list-processes. Otherwise, they are deleted immediately after they exit.

Function: delete-process &optional process

This function deletes a process, killing it with a SIGKILL signal if the process was running a program. The argument may be a process, the name of a process, a buffer, or the name of a buffer. (A buffer or buffer-name stands for the process that get-buffer-process returns, and a missing or nil process means that the current buffer’s process should be killed.) Calling delete-process on a running process terminates it, updates the process status, and runs the sentinel immediately. If the process has already terminated, calling delete-process has no effect on its status, or on the running of its sentinel (which will happen sooner or later).

If the process object represents a network, serial, or pipe connection, its status changes to closed; otherwise, it changes to signal, unless the process already exited. See process-status.

(delete-process "*shell*")
     ⇒ nil

40.6 Process Information

Several functions return information about processes.

Command: list-processes &optional query-only buffer

This command displays a listing of all living processes. In addition, it finally deletes any process whose status was ‘Exited’ or ‘Signaled’. It returns nil.

The processes are shown in a buffer named *Process List* (unless you specify otherwise using the optional argument buffer), whose major mode is Process Menu mode.

If query-only is non-nil, it only lists processes whose query flag is non-nil. See Querying Before Exit.

Function: process-list

This function returns a list of all processes that have not been deleted.

(process-list)
     ⇒ (#<process display-time> #<process shell>)
Function: num-processors &optional query

This function returns the number of processors, a positive integer. Each usable thread execution unit counts as a processor. By default, the count includes the number of available processors, which you can override by setting the OMP_NUM_THREADS environment variable of OpenMP. If the optional argument query is current, this function ignores OMP_NUM_THREADS; if query is all, this function also counts processors that are on the system but are not available to the current process.

Function: get-process name

This function returns the process named name (a string), or nil if there is none. The argument name can also be a process object, in which case it is returned.

(get-process "shell")
     ⇒ #<process shell>
Function: process-command process

This function returns the command that was executed to start process. This is a list of strings, the first string being the program executed and the rest of the strings being the arguments that were given to the program. For a network, serial, or pipe connection, this is either nil, which means the process is running or t (process is stopped).

(process-command (get-process "shell"))
     ⇒ ("bash" "-i")
Function: process-contact process &optional key no-block

This function returns information about how a network, a serial, or a pipe connection was set up. When key is nil, it returns (hostname service) for a network connection, (port speed) for a serial connection, and t for a pipe connection. For an ordinary child process, this function always returns t when called with a nil key.

If key is t, the value is the complete status information for the connection, server, serial port, or pipe; that is, the list of keywords and values specified in make-network-process, make-serial-process, or make-pipe-process, except that some of the values represent the current status instead of what you specified.

For a network process, the values include (see make-network-process for a complete list):

:buffer

The associated value is the process buffer.

:filter

The associated value is the process filter function. See Process Filter Functions.

:sentinel

The associated value is the process sentinel function. See Sentinels: Detecting Process Status Changes.

:remote

In a connection, the address in internal format of the remote peer.

:local

The local address, in internal format.

:service

In a server, if you specified t for service, this value is the actual port number.

:local and :remote are included even if they were not specified explicitly in make-network-process.

For a serial connection, see make-serial-process and serial-process-configure for the list of keys. For a pipe connection, see make-pipe-process for the list of keys.

If key is a keyword, the function returns the value corresponding to that keyword.

If process is a non-blocking network stream that hasn’t been fully set up yet, then this function will block until that has happened. If given the optional no-block parameter, this function will return nil instead of blocking.

Function: process-id process

This function returns the PID of process. This is an integral number that distinguishes the process process from all other processes running on the same computer at the current time. The PID of a process is chosen by the operating system kernel when the process is started and remains constant as long as the process exists. For network, serial, and pipe connections, this function returns nil.

Function: process-name process

This function returns the name of process, as a string.

Function: process-status process-name

This function returns the status of process-name as a symbol. The argument process-name must be a process, a buffer, or a process name (a string).

The possible values for an actual subprocess are:

run

for a process that is running.

stop

for a process that is stopped but continuable.

exit

for a process that has exited.

signal

for a process that has received a fatal signal.

open

for a network, serial, or pipe connection that is open.

closed

for a network, serial, or pipe connection that is closed. Once a connection is closed, you cannot reopen it, though you might be able to open a new connection to the same place.

connect

for a non-blocking connection that is waiting to complete.

failed

for a non-blocking connection that has failed to complete.

listen

for a network server that is listening.

nil

if process-name is not the name of an existing process.

(process-status (get-buffer "*shell*"))
     ⇒ run

For a network, serial, or pipe connection, process-status returns one of the symbols open, stop, or closed. The latter means that the other side closed the connection, or Emacs did delete-process. The value stop means that stop-process was called on the connection.

Function: process-live-p process

This function returns non-nil if process is alive. A process is considered alive if its status is run, open, listen, connect or stop.

Function: process-type process

This function returns the symbol network for a network connection or server, serial for a serial port connection, pipe for a pipe connection, or real for a subprocess created for running a program.

Function: process-exit-status process

This function returns the exit status of process or the signal number that killed it. (Use the result of process-status to determine which of those it is.) If process has not yet terminated, the value is 0. For network, serial, and pipe connections that are already closed, the value is either 0 or 256, depending on whether the connection was closed normally or abnormally.

Function: process-tty-name process &optional stream

This function returns the terminal name that process is using for its communication with Emacs—or nil if it is using pipes instead of a pty (see process-connection-type in Creating an Asynchronous Process). By default, this function returns the terminal name if any of process’s standard streams use a terminal. If stream is one of stdin, stdout, or stderr, this function returns the terminal name (or nil, as above) that process uses for that stream specifically. You can use this to determine whether a particular stream uses a pipe or a pty.

If process represents a program running on a remote host, this function returns the local terminal name that communicates with process; you can get the terminal name used by that program on the remote host with the process property remote-tty. If process represents a network, serial, or pipe connection, this function always returns nil.

Function: process-coding-system process

This function returns a cons cell (decode . encode), describing the coding systems in use for decoding output from, and encoding input to, process (see Coding Systems).

Function: set-process-coding-system process &optional decoding-system encoding-system

This function specifies the coding systems to use for subsequent output from and input to process. It will use decoding-system to decode subprocess output, and encoding-system to encode subprocess input.

Every process also has a property list that you can use to store miscellaneous values associated with the process.

Function: process-get process propname

This function returns the value of the propname property of process.

Function: process-put process propname value

This function sets the value of the propname property of process to value.

Function: process-plist process

This function returns the process plist of process.

Function: set-process-plist process plist

This function sets the process plist of process to plist.


40.7 Sending Input to Processes

Asynchronous subprocesses receive input when it is sent to them by Emacs, which is done with the functions in this section. You must specify the process to send input to, and the input data to send. If the subprocess runs a program, the data appears on the standard input of that program; for connections, the data is sent to the connected device or program.

Some operating systems have limited space for buffered input in a pty. On these systems, Emacs sends an EOF periodically amidst the other characters, to force them through. For most programs, these EOFs do no harm.

Subprocess input is normally encoded using a coding system before the subprocess receives it, much like text written into a file. You can use set-process-coding-system to specify which coding system to use (see Process Information). Otherwise, the coding system comes from coding-system-for-write, if that is non-nil; or else from the defaulting mechanism (see Default Coding Systems).

Sometimes the system is unable to accept input for that process, because the input buffer is full. When this happens, the send functions wait a short while, accepting output from subprocesses, and then try again. This gives the subprocess a chance to read more of its pending input and make space in the buffer. It also allows filters (including the one currently running), sentinels and timers to run—so take account of that in writing your code.

In these functions, the process argument can be a process or the name of a process, or a buffer or buffer name (which stands for a process via get-buffer-process). nil means the current buffer’s process.

Function: process-send-string process string

This function sends process the contents of string as standard input. It returns nil. For example, to make a Shell buffer list files:

(process-send-string "shell<1>" "ls\n")
     ⇒ nil
Function: process-send-region process start end

This function sends the text in the region defined by start and end as standard input to process.

An error is signaled unless both start and end are integers or markers that indicate positions in the current buffer. (It is unimportant which number is larger.)

Function: process-send-eof &optional process

This function makes process see an end-of-file in its input. The EOF comes after any text already sent to it. The function returns process.

(process-send-eof "shell")
     ⇒ "shell"
Function: process-running-child-p &optional process

This function will tell you whether a process, which must not be a connection but a real subprocess, has given control of its terminal to a child process of its own. If this is true, the function returns the numeric ID of the foreground process group of process; it returns nil if Emacs can be certain that this is not so. The value is t if Emacs cannot tell whether this is true. This function signals an error if process is a network, serial, or pipe connection, or if the subprocess is not active.


40.8 Sending Signals to Processes

Sending a signal to a subprocess is a way of interrupting its activities. There are several different signals, each with its own meaning. The set of signals and their names is defined by the operating system. For example, the signal SIGINT means that the user has typed C-c, or that some analogous thing has happened.

Each signal has a standard effect on the subprocess. Most signals kill the subprocess, but some stop (or resume) execution instead. Most signals can optionally be handled by programs; if the program handles the signal, then we can say nothing in general about its effects.

You can send signals explicitly by calling the functions in this section. Emacs also sends signals automatically at certain times: killing a buffer sends a SIGHUP signal to all its associated processes; killing Emacs sends a SIGHUP signal to all remaining processes. (SIGHUP is a signal that usually indicates that the user “hung up the phone”, i.e., disconnected.)

Each of the signal-sending functions takes two optional arguments: process and current-group.

The argument process must be either a process, a process name, a buffer, a buffer name, or nil. A buffer or buffer name stands for a process through get-buffer-process. nil stands for the process associated with the current buffer. Except with stop-process and continue-process, an error is signaled if process does not identify an active process, or if it represents a network, serial, or pipe connection.

The argument current-group is a flag that makes a difference when you are running a job-control shell as an Emacs subprocess. If it is non-nil, then the signal is sent to the current process-group of the terminal that Emacs uses to communicate with the subprocess. If the process is a job-control shell, this means the shell’s current subjob. If current-group is nil, the signal is sent to the process group of the immediate subprocess of Emacs. If the subprocess is a job-control shell, this is the shell itself. If current-group is lambda, the signal is sent to the process-group that owns the terminal, but only if it is not the shell itself.

The flag current-group has no effect when a pipe is used to communicate with the subprocess, because the operating system does not support the distinction in the case of pipes. For the same reason, job-control shells won’t work when a pipe is used. See process-connection-type in Creating an Asynchronous Process.

Function: interrupt-process &optional process current-group

This function interrupts the process process by sending the signal SIGINT. Outside of Emacs, typing the interrupt character (normally C-c on some systems, and DEL on others) sends this signal. When the argument current-group is non-nil, you can think of this function as typing C-c on the terminal by which Emacs talks to the subprocess.

Command: kill-process &optional process current-group

This command kills the process process by sending the signal SIGKILL. This signal kills the subprocess immediately, and cannot be handled by the subprocess. Interactively, it’ll prompt the user for a process name, defaulting to the process (if any) in the current buffer.

Function: quit-process &optional process current-group

This function sends the signal SIGQUIT to the process process. This signal is the one sent by the quit character (usually C-\) when you are not inside Emacs.

Function: stop-process &optional process current-group

This function stops the specified process. If it is a real subprocess running a program, it sends the signal SIGTSTP to that subprocess. If process represents a network, serial, or pipe connection, this function inhibits handling of the incoming data from the connection; for a network server, this means not accepting new connections. Use continue-process to resume normal execution.

Outside of Emacs, on systems with job control, the stop character (usually C-z) normally sends the SIGTSTP signal to a subprocess. When current-group is non-nil, you can think of this function as typing C-z on the terminal Emacs uses to communicate with the subprocess.

Function: continue-process &optional process current-group

This function resumes execution of the process process. If it is a real subprocess running a program, it sends the signal SIGCONT to that subprocess; this presumes that process was stopped previously. If process represents a network, serial, or pipe connection, this function resumes handling of the incoming data from the connection. For serial connections, data that arrived during the time the process was stopped might be lost.

Command: signal-process process signal &optional remote

This function sends a signal to process process. The argument signal specifies which signal to send; it should be an integer, or a symbol whose name is a signal.

The process argument can be a system process ID (an integer); that allows you to send signals to processes that are not children of Emacs. See Accessing Other Processes.

If process is a process object which contains the property remote-pid, or process is a number and remote is a remote file name, process is interpreted as process on the respective remote host, which will be the process to signal.

If process is a string, it is interpreted as process object with the respective process name, or as a number.

Sometimes, it is necessary to send a signal to a non-local asynchronous process. This is possible by writing an own interrupt-process or signal-process implementation. This function must be added then to interrupt-process-functions or signal-process-functions, respectively.

Variable: interrupt-process-functions

This variable is a list of functions to be called for interrupt-process. The arguments of the functions are the same as for interrupt-process. These functions are called in the order of the list, until one of them returns non-nil. The default function, which shall always be the last in this list, is internal-default-interrupt-process.

This is the mechanism, how Tramp implements interrupt-process.

Variable: signal-process-functions

This variable is a list of functions to be called for signal-process. The arguments of the functions are the same as for signal-process. These functions are called in the order of the list, until one of them returns non-nil. The default function, which shall always be the last in this list, is internal-default-signal-process.

This is the mechanism, how Tramp implements signal-process.


40.9 Receiving Output from Processes

The output that an asynchronous subprocess writes to its standard output stream is passed to a function called the filter function. The default filter function simply inserts the output into a buffer, which is called the associated buffer of the process (see Process Buffers). If the process has no buffer then the default filter discards the output.

If the subprocess writes to its standard error stream, by default the error output is also passed to the process filter function. Alternatively, you could use the :stderr parameter with a non-nil value in a call to make-process (see make-process) to make the destination of the error output separate from the standard output.

When a subprocess terminates, Emacs reads any pending output, then stops reading output from that subprocess. Therefore, if the subprocess has children that are still live and still producing output, Emacs won’t receive that output.

Output from a subprocess can arrive only while Emacs is waiting: when reading terminal input (see the function waiting-for-user-input-p), in sit-for and sleep-for (see Waiting for Elapsed Time or Input), in accept-process-output (see Accepting Output from Processes), and in functions which send data to processes (see Sending Input to Processes). This minimizes the problem of timing errors that usually plague parallel programming. For example, you can safely create a process and only then specify its buffer or filter function; no output can arrive before you finish, if the code in between does not call any primitive that waits.

Variable: process-adaptive-read-buffering

On some systems, when Emacs reads the output from a subprocess, the output data is read in very small blocks, potentially resulting in very poor performance. This behavior can be remedied to some extent by setting the variable process-adaptive-read-buffering to a non-nil value (the default), as it will automatically delay reading from such processes, thus allowing them to produce more output before Emacs tries to read it.


40.9.1 Process Buffers

A process can (and usually does) have an associated buffer, which is an ordinary Emacs buffer that is used for two purposes: storing the output from the process, and deciding when to kill the process. You can also use the buffer to identify a process to operate on, since in normal practice only one process is associated with any given buffer. Many applications of processes also use the buffer for editing input to be sent to the process, but this is not built into Emacs Lisp.

By default, process output is inserted in the associated buffer. (You can change this by defining a custom filter function, see Process Filter Functions.) The position to insert the output is determined by the process-mark, which is then updated to point to the end of the text just inserted. Usually, but not always, the process-mark is at the end of the buffer.

Killing the associated buffer of a process also kills the process. Emacs asks for confirmation first, if the process’s process-query-on-exit-flag is non-nil (see Querying Before Exit). This confirmation is done by the function process-kill-buffer-query-function, which is run from kill-buffer-query-functions (see Killing Buffers).

Function: process-buffer process

This function returns the associated buffer of the specified process.

(process-buffer (get-process "shell"))
     ⇒ #<buffer *shell*>
Function: process-mark process

This function returns the process marker for process, which is the marker that says where to insert output from the process.

If process does not have a buffer, process-mark returns a marker that points nowhere.

The default filter function uses this marker to decide where to insert process output, and updates it to point after the inserted text. That is why successive batches of output are inserted consecutively.

Custom filter functions normally should use this marker in the same fashion. For an example of a filter function that uses process-mark, see Process Filter Example.

When the user is expected to enter input in the process buffer for transmission to the process, the process marker separates the new input from previous output.

Function: set-process-buffer process buffer

This function sets the buffer associated with process to buffer. If buffer is nil, the process becomes associated with no buffer; if non-nil and different from the buffer associated with the process, the process mark will be set to point to the end of buffer (unless the process mark is already associated with buffer).

Function: get-buffer-process buffer-or-name

This function returns a nondeleted process associated with the buffer specified by buffer-or-name. If there are several processes associated with it, this function chooses one (currently, the one most recently created, but don’t count on that). Deletion of a process (see delete-process) makes it ineligible for this function to return.

It is usually a bad idea to have more than one process associated with the same buffer.

(get-buffer-process "*shell*")
     ⇒ #<process shell>

Killing the process’s buffer deletes the process, which kills the subprocess with a SIGHUP signal (see Sending Signals to Processes).

If the process’s buffer is displayed in a window, your Lisp program may wish to tell the process the dimensions of that window, so that the process could adapt its output to those dimensions, much as it adapts to the screen dimensions. The following functions allow communicating this kind of information to processes; however, not all systems support the underlying functionality, so it is best to provide fallbacks, e.g., via command-line arguments or environment variables.

Function: set-process-window-size process height width

Tell process that its logical window size has dimensions width by height, in character units. If this function succeeds in communicating this information to the process, it returns t; otherwise it returns nil.

When windows that display buffers associated with process change their dimensions, the affected processes should be told about these changes. By default, when the window configuration changes, Emacs will automatically call set-process-window-size on behalf of every process whose buffer is displayed in a window, passing it the smallest dimensions of all the windows displaying the process’s buffer. This works via window-configuration-change-hook (see Hooks for Window Scrolling and Changes), which is told to invoke the function that is the value of the variable window-adjust-process-window-size-function for each process whose buffer is displayed in at least one window. You can customize this behavior by setting the value of that variable.

User Option: window-adjust-process-window-size-function

The value of this variable should be a function of two arguments: a process and the list of windows displaying the process’s buffer. When the function is called, the process’s buffer is the current buffer. The function should return a cons cell (width . height) that describes the dimensions of the logical process window to be passed via a call to set-process-window-size. The function can also return nil, in which case Emacs will not call set-process-window-size for this process.

Emacs supplies two predefined values for this variable: window-adjust-process-window-size-smallest, which returns the smallest of all the dimensions of the windows that display a process’s buffer; and window-adjust-process-window-size-largest, which returns the largest dimensions. For more complex strategies, write your own function.

This variable can be buffer-local.

If the process has the adjust-window-size-function property (see Process Information), its value overrides the global and buffer-local values of window-adjust-process-window-size-function.


40.9.2 Process Filter Functions

A process filter function is a function that receives the standard output from the associated process. All output from that process is passed to the filter. The default filter simply outputs directly to the process buffer.

By default, the error output from the process, if any, is also passed to the filter function, unless the destination for the standard error stream of the process was separated from the standard output when the process was created. Emacs will only call the filter function during certain function calls. See Receiving Output from Processes. Note that if any of those functions are called by the filter, the filter may be called recursively.

A filter function must accept two arguments: the associated process and a string, which is output just received from it. The function is then free to do whatever it chooses with the output.

Quitting is normally inhibited within a filter function—otherwise, the effect of typing C-g at command level or to quit a user command would be unpredictable. If you want to permit quitting inside a filter function, bind inhibit-quit to nil. In most cases, the right way to do this is with the macro with-local-quit. See Quitting.

If an error happens during execution of a filter function, it is caught automatically, so that it doesn’t stop the execution of whatever program was running when the filter function was started. However, if debug-on-error is non-nil, errors are not caught. This makes it possible to use the Lisp debugger to debug filter functions. See The Lisp Debugger. If an error is caught, Emacs pauses for process-error-pause-time seconds so that the user sees the error. See Creating an Asynchronous Process

Many filter functions sometimes (or always) insert the output in the process’s buffer, mimicking the actions of the default filter. Such filter functions need to make sure that they save the current buffer, select the correct buffer (if different) before inserting output, and then restore the original buffer. They should also check whether the buffer is still alive, update the process marker, and in some cases update the value of point. Here is how to do these things:

(defun ordinary-insertion-filter (proc string)
  (when (buffer-live-p (process-buffer proc))
    (with-current-buffer (process-buffer proc)
      (let ((moving (= (point) (process-mark proc))))
        (save-excursion
          ;; Insert the text, advancing the process marker.
          (goto-char (process-mark proc))
          (insert string)
          (set-marker (process-mark proc) (point)))
        (if moving (goto-char (process-mark proc)))))))

To make the filter force the process buffer to be visible whenever new text arrives, you could insert a line like the following just before the with-current-buffer construct:

(display-buffer (process-buffer proc))

To force point to the end of the new output, no matter where it was previously, eliminate the variable moving from the example and call goto-char unconditionally. Note that this doesn’t necessarily move the window point. The default filter actually uses insert-before-markers which moves all markers, including the window point. This may move unrelated markers, so it’s generally better to move the window point explicitly, or set its insertion type to t (see Windows and Point).

Note that Emacs automatically saves and restores the match data while executing filter functions. See The Match Data.

The output to the filter may come in chunks of any size. A program that produces the same output twice in a row may send it as one batch of 200 characters one time, and five batches of 40 characters the next. If the filter looks for certain text strings in the subprocess output, make sure to handle the case where one of these strings is split across two or more batches of output; one way to do this is to insert the received text into a temporary buffer, which can then be searched.

Function: set-process-filter process filter

This function gives process the filter function filter. If filter is nil, it gives the process the default filter, which inserts the process output into the process buffer. If filter is t, Emacs stops accepting output from the process, unless it’s a network server process that listens for incoming connections.

Function: process-filter process

This function returns the filter function of process.

In case the process’s output needs to be passed to several filters, you can use add-function to combine an existing filter with a new one. See Advising Emacs Lisp Functions.

Here is an example of the use of a filter function:

(defun keep-output (process output)
   (setq kept (cons output kept)))
     ⇒ keep-output
(setq kept nil)
     ⇒ nil
(set-process-filter (get-process "shell") 'keep-output)
     ⇒ keep-output
(process-send-string "shell" "ls ~/other\n")
     ⇒ nil
kept
     ⇒ ("lewis@slug:$ "
"FINAL-W87-SHORT.MSS    backup.otl              kolstad.mss~
address.txt             backup.psf              kolstad.psf
backup.bib~             david.mss               resume-Dec-86.mss~
backup.err              david.psf               resume-Dec.psf
backup.mss              dland                   syllabus.mss
"
"#backups.mss#          backup.mss~             kolstad.mss
")

40.9.3 Decoding Process Output

When Emacs writes process output directly into a multibyte buffer, it decodes the output according to the process output coding system. If the coding system is raw-text or no-conversion, Emacs converts the unibyte output to multibyte using string-to-multibyte, and inserts the resulting multibyte text.

You can use set-process-coding-system to specify which coding system to use (see Process Information). Otherwise, the coding system comes from coding-system-for-read, if that is non-nil; or else from the defaulting mechanism (see Default Coding Systems). If the text output by a process contains null bytes, Emacs by default uses no-conversion for it; see inhibit-null-byte-detection, for how to control this behavior.

Warning: Coding systems such as undecided, which determine the coding system from the data, do not work entirely reliably with asynchronous subprocess output. This is because Emacs has to process asynchronous subprocess output in batches, as it arrives. Emacs must try to detect the proper coding system from one batch at a time, and this does not always work. Therefore, if at all possible, specify a coding system that determines both the character code conversion and the end of line conversion—that is, one like latin-1-unix, rather than undecided or latin-1.

When Emacs calls a process filter function, it provides the process output as a multibyte string or as a unibyte string according to the process’s filter coding system. Emacs decodes the output according to the process output coding system, which usually produces a multibyte string, except for coding systems such as binary and raw-text.


40.9.4 Accepting Output from Processes

Output from asynchronous subprocesses normally arrives only while Emacs is waiting for some sort of external event, such as elapsed time or terminal input. Occasionally it is useful in a Lisp program to explicitly permit output to arrive at a specific point, or even to wait until output arrives from a process.

Function: accept-process-output &optional process seconds millisec just-this-one

This function allows Emacs to read pending output from processes. The output is given to their filter functions. If process is non-nil then this function does not return until some output has been received from process or process has closed the connection.

The arguments seconds and millisec let you specify timeout periods. The former specifies a period measured in seconds and the latter specifies one measured in milliseconds. The two time periods thus specified are added together, and accept-process-output returns after that much time, even if there is no subprocess output.

The argument millisec is obsolete (and should not be used), because seconds can be floating point to specify waiting a fractional number of seconds. If seconds is 0, the function accepts whatever output is pending but does not wait.

If process is a process, and the argument just-this-one is non-nil, only output from that process is handled, suspending output from other processes until some output has been received from that process or the timeout expires. If just-this-one is an integer, also inhibit running timers. This feature is generally not recommended, but may be necessary for specific applications, such as speech synthesis.

The function accept-process-output returns non-nil if it got output from process, or from any process if process is nil; this can occur even after a process has exited if the corresponding connection contains buffered data. The function returns nil if the timeout expired or the connection was closed before output arrived.

If a connection from a process contains buffered data, accept-process-output can return non-nil even after the process has exited. Therefore, although the following loop:

;; This loop contains a bug.
(while (process-live-p process)
  (accept-process-output process))

will often read all output from process, it has a race condition and can miss some output if process-live-p returns nil while the connection still contains data. Better is to write the loop like this:

(while (accept-process-output process))

If you have passed a non-nil stderr to make-process, it will have a standard error process. See Creating an Asynchronous Process. In that case, waiting for process output from the main process doesn’t wait for output from the standard error process. To make sure you have received both all of standard output and all of standard error from a process, use the following code:

(while (accept-process-output process))
(while (accept-process-output stderr-process))

If you passed a buffer to the stderr argument of make-process, you still have to wait for the standard error process, like so:

(let* ((stdout (generate-new-buffer "stdout"))
       (stderr (generate-new-buffer "stderr"))
       (process (make-process :name "test"
                              :command '("my-program")
                              :buffer stdout
                              :stderr stderr))
       (stderr-process (get-buffer-process stderr)))
  (unless (and process stderr-process)
    (error "Process unexpectedly nil"))
  (while (accept-process-output process))
  (while (accept-process-output stderr-process)))

Only when both accept-process-output forms return nil, you can be sure that the process has exited and Emacs has read all its output.

Reading pending standard error from a process running on a remote host is not possible this way.


40.9.5 Processes and Threads

Because threads were a relatively late addition to Emacs Lisp, and due to the way dynamic binding was sometimes used in conjunction with accept-process-output, by default a process is locked to the thread that created it. When a process is locked to a thread, output from the process can only be accepted by that thread.

A Lisp program can specify to which thread a process is to be locked, or instruct Emacs to unlock a process, in which case its output can be processed by any thread. Only a single thread will wait for output from a given process at one time—once one thread begins waiting for output, the process is temporarily locked until accept-process-output or sit-for returns.

If the thread exits, all the processes locked to it are unlocked.

Function: process-thread process

Return the thread to which process is locked. If process is unlocked, return nil.

Function: set-process-thread process thread

Set the locking thread of process to thread. thread may be nil, in which case the process is unlocked.


40.10 Sentinels: Detecting Process Status Changes

A process sentinel is a function that is called whenever the associated process changes status for any reason, including signals (whether sent by Emacs or caused by the process’s own actions) that terminate, stop, or continue the process. The process sentinel is also called if the process exits. The sentinel receives two arguments: the process for which the event occurred, and a string describing the type of event.

If no sentinel function was specified for a process, it will use the default sentinel function, which inserts a message in the process’s buffer with the process name and the string describing the event.

The string describing the event looks like one of the following (but this is not an exhaustive list of event strings):

  • "finished\n".
  • "deleted\n".
  • "exited abnormally with code exitcode (core dumped)\n". The “core dumped” part is optional, and only appears if the process dumped core.
  • "failed with code fail-code\n".
  • "signal-description (core dumped)\n". The signal-description is a system-dependent textual description of a signal, e.g., "killed" for SIGKILL. The “core dumped” part is optional, and only appears if the process dumped core.
  • "open from host-name\n".
  • "open\n".
  • "run\n".
  • "connection broken by remote peer\n".

A sentinel runs only while Emacs is waiting (e.g., for terminal input, or for time to elapse, or for process output). This avoids the timing errors that could result from running sentinels at random places in the middle of other Lisp programs. A program can wait, so that sentinels will run, by calling sit-for or sleep-for (see Waiting for Elapsed Time or Input), or accept-process-output (see Accepting Output from Processes). Emacs also allows sentinels to run when the command loop is reading input. delete-process calls the sentinel when it terminates a running process.

Emacs does not keep a queue of multiple reasons to call the sentinel of one process; it records just the current status and the fact that there has been a change. Therefore two changes in status, coming in quick succession, can call the sentinel just once. However, process termination will always run the sentinel exactly once. This is because the process status can’t change again after termination.

Emacs explicitly checks for output from the process before running the process sentinel. Once the sentinel runs due to process termination, no further output can arrive from the process.

A sentinel that writes the output into the buffer of the process should check whether the buffer is still alive. If it tries to insert into a dead buffer, it will get an error. If the buffer is dead, (buffer-name (process-buffer process)) returns nil.

Quitting is normally inhibited within a sentinel—otherwise, the effect of typing C-g at command level or to quit a user command would be unpredictable. If you want to permit quitting inside a sentinel, bind inhibit-quit to nil. In most cases, the right way to do this is with the macro with-local-quit. See Quitting.

If an error happens during execution of a sentinel, it is caught automatically, so that it doesn’t stop the execution of whatever programs was running when the sentinel was started. However, if debug-on-error is non-nil, errors are not caught. This makes it possible to use the Lisp debugger to debug the sentinel. See The Lisp Debugger. If an error is caught, Emacs pauses for process-error-pause-time seconds so that the user sees the error. See Creating an Asynchronous Process

While a sentinel is running, the process sentinel is temporarily set to nil so that the sentinel won’t run recursively. For this reason it is not possible for a sentinel to specify a new sentinel.

Note that Emacs automatically saves and restores the match data while executing sentinels. See The Match Data.

Function: set-process-sentinel process sentinel

This function associates sentinel with process. If sentinel is nil, then the process will have the default sentinel, which inserts a message in the process’s buffer when the process status changes.

Changes in process sentinels take effect immediately—if the sentinel is slated to be run but has not been called yet, and you specify a new sentinel, the eventual call to the sentinel will use the new one.

(defun msg-me (process event)
   (princ
     (format "Process: %s had the event '%s'" process event)))
(set-process-sentinel (get-process "shell") 'msg-me)
     ⇒ msg-me
(kill-process (get-process "shell"))
     -| Process: #<process shell> had the event 'killed'
     ⇒ #<process shell>
Function: process-sentinel process

This function returns the sentinel of process.

In case a process status changes need to be passed to several sentinels, you can use add-function to combine an existing sentinel with a new one. See Advising Emacs Lisp Functions.

Function: waiting-for-user-input-p

While a sentinel or filter function is running, this function returns non-nil if Emacs was waiting for keyboard input from the user at the time the sentinel or filter function was called, or nil if it was not.


40.11 Querying Before Exit

When Emacs exits, it terminates all its subprocesses. For subprocesses that run a program, it sends them the SIGHUP signal; connections are simply closed. Because subprocesses may be doing valuable work, Emacs normally asks the user to confirm that it is ok to terminate them. Each process has a query flag, which, if non-nil, says that Emacs should ask for confirmation before exiting and thus killing that process. The default for the query flag is t, meaning do query.

Function: process-query-on-exit-flag process

This returns the query flag of process.

Function: set-process-query-on-exit-flag process flag

This function sets the query flag of process to flag. It returns flag.

Here is an example of using set-process-query-on-exit-flag on a shell process to avoid querying:

(set-process-query-on-exit-flag (get-process "shell") nil)
     ⇒ nil
User Option: confirm-kill-processes

If this user option is set to t (the default), then Emacs asks for confirmation before killing processes on exit. If it is nil, Emacs kills processes without confirmation, i.e., the query flag of all processes is ignored.


40.12 Accessing Other Processes

In addition to accessing and manipulating processes that are subprocesses of the current Emacs session, Emacs Lisp programs can also access other processes. We call these system processes, to distinguish them from Emacs subprocesses.

Emacs provides several primitives for accessing system processes. Not all platforms support these primitives; on those which don’t, these primitives return nil.

Function: list-system-processes

This function returns a list of all the processes running on the system. Each process is identified by its PID, a numerical process ID that is assigned by the OS and distinguishes the process from all the other processes running on the same machine at the same time.

If default-directory points to a remote host, processes of that host are returned.

Function: process-attributes pid

This function returns an alist of attributes for the process specified by its process ID pid. Each association in the alist is of the form (key . value), where key designates the attribute and value is the value of that attribute. The various attribute keys that this function can return are listed below. Not all platforms support all of these attributes; if an attribute is not supported, its association will not appear in the returned alist.

If default-directory points to a remote host, pid is regarded as process of that host.

euid

The effective user ID of the user who invoked the process. The corresponding value is a number. If the process was invoked by the same user who runs the current Emacs session, the value is identical to what user-uid returns (see User Identification).

user

User name corresponding to the process’s effective user ID, a string.

egid

The group ID of the effective user ID, a number.

group

Group name corresponding to the effective user’s group ID, a string.

comm

The name of the command that runs in the process. This is a string that usually specifies the name of the executable file of the process, without the leading directories. However, some special system processes can report strings that do not correspond to an executable file of a program.

state

The state code of the process. This is a short string that encodes the scheduling state of the process. Here’s a list of the most frequently seen codes:

"D"

uninterruptible sleep (usually I/O)

"R"

running

"S"

interruptible sleep (waiting for some event)

"T"

stopped, e.g., by a job control signal

"Z"

zombie: a process that terminated, but was not reaped by its parent

For the full list of the possible states, see the manual page of the ps command.

ppid

The process ID of the parent process, a number.

pgrp

The process group ID of the process, a number.

sess

The session ID of the process. This is a number that is the process ID of the process’s session leader.

ttname

A string that is the name of the process’s controlling terminal. On Unix and GNU systems, this is normally the file name of the corresponding terminal device, such as /dev/pts65.

tpgid

The numerical process group ID of the foreground process group that uses the process’s terminal.

minflt

The number of minor page faults caused by the process since its beginning. (Minor page faults are those that don’t involve reading from disk.)

majflt

The number of major page faults caused by the process since its beginning. (Major page faults require a disk to be read, and are thus more expensive than minor page faults.)

cminflt
cmajflt

Like minflt and majflt, but include the number of page faults for all the child processes of the given process.

utime

Time spent by the process in the user context, for running the application’s code. The corresponding value is a Lisp timestamp (see Time of Day).

stime

Time spent by the process in the system (kernel) context, for processing system calls. The corresponding value is a Lisp timestamp.

time

The sum of utime and stime. The corresponding value is a Lisp timestamp.

cutime
cstime
ctime

Like utime, stime, and time, but include the times of all the child processes of the given process.

pri

The numerical priority of the process.

nice

The nice value of the process, a number. (Processes with smaller nice values get scheduled more favorably.)

thcount

The number of threads in the process.

start

The time when the process was started, as a Lisp timestamp.

etime

The time elapsed since the process started, as a Lisp timestamp.

vsize

The virtual memory size of the process, measured in kilobytes.

rss

The size of the process’s resident set, the number of kilobytes occupied by the process in the machine’s physical memory.

pcpu

The percentage of the CPU time used by the process since it started. The corresponding value is a nonnegative floating-point number. Although in theory the number can exceed 100 on a multicore platform, it is usually less than 100 because Emacs is typically single-threaded.

pmem

The percentage of the total physical memory installed on the machine used by the process’s resident set. The value is a floating-point number between 0 and 100.

args

The command-line with which the process was invoked. This is a string in which individual command-line arguments are separated by blanks; whitespace characters that are embedded in the arguments are quoted as appropriate for the system’s shell: escaped by backslash characters on GNU and Unix, and enclosed in double quote characters on Windows. Thus, this command-line string can be directly used in primitives such as shell-command.


40.13 Transaction Queues

You can use a transaction queue to communicate with a subprocess using transactions. First use tq-create to create a transaction queue communicating with a specified process. Then you can call tq-enqueue to send a transaction.

Function: tq-create process

This function creates and returns a transaction queue communicating with process. The argument process should be a subprocess capable of sending and receiving streams of bytes. It may be a child process, or it may be a TCP connection to a server, possibly on another machine.

Function: tq-enqueue queue question regexp closure fn &optional delay-question

This function sends a transaction to queue queue. Specifying the queue has the effect of specifying the subprocess to talk to.

The argument question is the outgoing message that starts the transaction. The argument fn is the function to call when the corresponding answer comes back; it is called with two arguments: closure, and the answer received.

The argument regexp is a regular expression that should match text at the end of the entire answer, but nothing before; that’s how tq-enqueue determines where the answer ends.

If the argument delay-question is non-nil, delay sending this question until the process has finished replying to any previous questions. This produces more reliable results with some processes.

Function: tq-close queue

Shut down transaction queue queue, waiting for all pending transactions to complete, and then terminate the connection or child process.

Transaction queues are implemented by means of a filter function. See Process Filter Functions.


40.14 Network Connections

Emacs Lisp programs can open stream (TCP) and datagram (UDP) network connections (see Datagrams) to other processes on the same machine or other machines. A network connection is handled by Lisp much like a subprocess, and is represented by a process object. However, the process you are communicating with is not a child of the Emacs process, has no process ID, and you can’t kill it or send it signals. All you can do is send and receive data. delete-process closes the connection, but does not kill the program at the other end; that program must decide what to do about closure of the connection.

Lisp programs can listen for connections by creating network servers. A network server is also represented by a kind of process object, but unlike a network connection, the network server never transfers data itself. When it receives a connection request, it creates a new network connection to represent the connection just made. (The network connection inherits certain information, including the process plist, from the server.) The network server then goes back to listening for more connection requests.

Network connections and servers are created by calling make-network-process with an argument list consisting of keyword/argument pairs, for example :server t to create a server process, or :type 'datagram to create a datagram connection. See Low-Level Network Access, for details. You can also use the open-network-stream function described below.

To distinguish the different types of processes, the process-type function returns the symbol network for a network connection or server, serial for a serial port connection, pipe for a pipe connection, or real for a real subprocess.

The process-status function returns open, closed, connect, stop, or failed for network connections. For a network server, the status is always listen. Except for stop, none of those values is possible for a real subprocess. See Process Information.

You can stop and resume operation of a network process by calling stop-process and continue-process. For a server process, being stopped means not accepting new connections. (Up to 5 connection requests will be queued for when you resume the server; you can increase this limit, unless it is imposed by the operating system—see the :server keyword of make-network-process, make-network-process.) For a network stream connection, being stopped means not processing input (any arriving input waits until you resume the connection). For a datagram connection, some number of packets may be queued but input may be lost. You can use the function process-command to determine whether a network connection or server is stopped; a non-nil value means yes.

Emacs can create encrypted network connections, using the built-in support for the GnuTLS Transport Layer Security Library; see the GnuTLS project page. If your Emacs was compiled with GnuTLS support, the function gnutls-available-p is defined and returns non-nil. For more details, see Overview in The Emacs-GnuTLS manual. The open-network-stream function can transparently handle the details of creating encrypted connections for you, using whatever support is available.

Function: open-network-stream name buffer host service &rest parameters

This function opens a TCP connection, with optional encryption, and returns a process object that represents the connection.

The name argument specifies the name for the process object. It is modified as necessary to make it unique.

The buffer argument is the buffer to associate with the connection. Output from the connection is inserted in the buffer, unless you specify your own filter function to handle the output. If buffer is nil, it means that the connection is not associated with any buffer.

The arguments host and service specify where to connect to; host is the host name (a string), and service is the name of a defined network service (a string) or a port number (an integer like 80 or an integer string like "80").

The remaining arguments parameters are keyword/argument pairs that are mainly relevant to encrypted connections:

:nowait boolean

If non-nil, try to make an asynchronous connection.

:noquery query-flag

Initialize the process query flag to query-flag. See Querying Before Exit.

:coding coding

Use this to set the coding systems used by the network process, in preference to binding coding-system-for-read or coding-system-for-write. See make-network-process, for details.

:type type

The type of connection. Options are:

plain

An ordinary, unencrypted connection.

tls
ssl

A TLS (Transport Layer Security) connection.

nil
network

Start with a plain connection, and if parameters ‘:success’ and ‘:capability-command’ are supplied, try to upgrade to an encrypted connection via STARTTLS. If that fails, retain the unencrypted connection.

starttls

As for nil, but if STARTTLS fails, drop the connection.

shell

A shell connection.

:always-query-capabilities boolean

If non-nil, always ask for the server’s capabilities, even when doing a ‘plain’ connection.

:capability-command capability-command

Command to query the host capabilities. This can either be a string (which will then be sent verbatim to the server) or a function (called with a single parameter: the “greeting” from the server when connecting) that should return a string.

:end-of-command regexp
:end-of-capability regexp

Regular expression matching the end of a command, or the end of the command capability-command. The latter defaults to the former.

:starttls-function function

Function of one argument (the response to capability-command), which returns either nil or the command to activate STARTTLS, if supported.

:success regexp

Regular expression matching a successful STARTTLS negotiation.

:use-starttls-if-possible boolean

If non-nil, do opportunistic STARTTLS upgrades even if Emacs doesn’t have built-in TLS support.

:warn-unless-encrypted boolean

If non-nil, and :return-value is also non-nil, Emacs will warn if the connection isn’t encrypted. This is useful for protocols like IMAP and the like, where most users would expect the network traffic to be encrypted.

:client-certificate list-or-t

Either a list of the form (key-file cert-file), naming the certificate key file and certificate file itself, or t, meaning to query auth-source for this information (see auth-source in Emacs auth-source Library). Only used for TLS or STARTTLS. To enable automatic queries of auth-source when :client-certificate is not specified customize network-stream-use-client-certificates to t.

:return-list cons-or-nil

The return value of this function. If omitted or nil, return a process object. Otherwise, a cons of the form (process-object . plist), where plist can include the following keywords:

:greeting string-or-nil

If non-nil, the greeting string returned by the host.

:capabilities string-or-nil

If non-nil, the host’s capability string.

:type symbol

The connection type: ‘plain’ or ‘tls’.

:shell-command string-or-nil

If the connection type is shell, this parameter will be interpreted as a format-spec string (see Custom Format Strings) that will be executed to make the connection. The specs available are ‘%s’ for the host name and ‘%p’ for the port number. For instance, if you want to first ssh to ‘gateway’ before making a plain connection, then this parameter’s value could be something like ‘ssh gateway nc %s %p’.


40.15 Network Servers

You create a server by calling make-network-process (see make-network-process) with :server t. The server will listen for connection requests from clients. When it accepts a client connection request, that creates a new network connection, itself a process object, with the following parameters:

  • The connection’s process name is constructed by concatenating the server process’s name with a client identification string. The client identification string for an IPv4 connection looks like ‘<a.b.c.d:p>’, which represents an address and port number. Otherwise, it is a unique number in brackets, as in ‘<nnn>’. The number is unique for each connection in the Emacs session.
  • If the server has a non-default filter, the connection process does not get a separate process buffer; otherwise, Emacs creates a new buffer for the purpose. The buffer name is the server’s buffer name or process name, concatenated with the client identification string.

    The server’s process buffer value is never used directly, but the log function can retrieve it and use it to log connections by inserting text there.

  • The communication type and the process filter and sentinel are inherited from those of the server. The server never directly uses its filter and sentinel; their sole purpose is to initialize connections made to the server.
  • The connection’s process contact information is set according to the client’s addressing information (typically an IP address and a port number). This information is associated with the process-contact keywords :host, :service, :remote.
  • The connection’s local address is set up according to the port number used for the connection.
  • The client process’s plist is initialized from the server’s plist.

40.16 Datagrams

A datagram connection communicates with individual packets rather than streams of data. Each call to process-send sends one datagram packet (see Sending Input to Processes), and each datagram received results in one call to the filter function.

The datagram connection doesn’t have to talk with the same remote peer all the time. It has a remote peer address which specifies where to send datagrams to. Each time an incoming datagram is passed to the filter function, the peer address is set to the address that datagram came from; that way, if the filter function sends a datagram, it will go back to that place. You can specify the remote peer address when you create the datagram connection using the :remote keyword. You can change it later on by calling set-process-datagram-address.

Function: process-datagram-address process

If process is a datagram connection or server, this function returns its remote peer address.

Function: set-process-datagram-address process address

If process is a datagram connection or server, this function sets its remote peer address to address.


40.17 Low-Level Network Access

You can also create network connections by operating at a lower level than that of open-network-stream, using make-network-process.


40.17.1 make-network-process

The basic function for creating network connections and network servers is make-network-process. It can do either of those jobs, depending on the arguments you give it.

Function: make-network-process &rest args

This function creates a network connection or server and returns the process object that represents it. The arguments args are a list of keyword/argument pairs. Omitting a keyword is always equivalent to specifying it with value nil, except for :coding, :filter-multibyte, and :reuseaddr. Here are the meaningful keywords (those corresponding to network options are listed in the following section):

:name name

Use the string name as the process name. It is modified if necessary to make it unique.

:type type

Specify the communication type. A value of nil specifies a stream connection (the default); datagram specifies a datagram connection; seqpacket specifies a sequenced packet stream connection. Both connections and servers can be of these types.

:server server-flag

If server-flag is non-nil, create a server. Otherwise, create a connection. For a stream type server, server-flag may be an integer, which then specifies the length of the queue of pending connections to the server. The default queue length is 5.

:host host

Specify the host to connect to. host should be a host name or Internet address, as a string, or the symbol local to specify the local host. If you specify host for a server, it must specify a valid address for the local host, and only clients connecting to that address will be accepted. When using local, by default IPv4 will be used, specify a family of ipv6 to override this. To listen on all interfaces, specify an address of ‘"0.0.0.0"’ for IPv4 or ‘"::"’ for IPv6. Note that on some operating systems, listening on ‘"::"’ will also listen on IPv4, so attempting to then listen separately on IPv4 will result in EADDRINUSE errors (‘"Address already in use"’).

:service service

service specifies a port number to connect to; or, for a server, the port number to listen on. It should be a service name like ‘"https"’ that translates to a port number, or an integer like ‘443’ or an integer string like ‘"443"’ that specifies the port number directly. For a server, it can also be t, which means to let the system select an unused port number.

:family family

family specifies the address (and protocol) family for communication. nil means determine the proper address family automatically for the given host and service. local specifies a Unix socket, in which case host is ignored. ipv4 and ipv6 specify to use IPv4 and IPv6, respectively.

:use-external-socket use-external-socket

If use-external-socket is non-nil use any sockets passed to Emacs on invocation instead of allocating one. This is used by the Emacs server code to allow on-demand socket activation. If Emacs wasn’t passed a socket, this option is silently ignored.

:local local-address

For a server process, local-address is the address to listen on. It overrides family, host and service, so you might as well not specify them.

:remote remote-address

For a connection, remote-address is the address to connect to. It overrides family, host and service, so you might as well not specify them.

For a datagram server, remote-address specifies the initial setting of the remote datagram address.

The format of local-address or remote-address depends on the address family:

  • An IPv4 address is represented as a five-element vector of four 8-bit integers and one 16-bit integer [a b c d p] corresponding to numeric IPv4 address a.b.c.d and port number p.
  • An IPv6 address is represented as a nine-element vector of 16-bit integers [a b c d e f g h p] corresponding to numeric IPv6 address a:b:c:d:e:f:g:h and port number p.
  • A local address is represented as a string, which specifies the address in the local address space.
  • An unsupported-family address is represented by a cons (f . av), where f is the family number and av is a vector specifying the socket address using one element per address data byte. Do not rely on this format in portable code, as it may depend on implementation defined constants, data sizes, and data structure alignment.
:nowait bool

If bool is non-nil for a stream connection, return without waiting for the connection to complete. When the connection succeeds or fails, Emacs will call the sentinel function, with a second argument matching "open" (if successful) or "failed". The default is to block, so that make-network-process does not return until the connection has succeeded or failed.

If you’re setting up an asynchronous TLS connection, you have to also provide the :tls-parameters parameter (see below).

Depending on the capabilities of Emacs, how asynchronous :nowait is may vary. The three elements that may (or may not) be done asynchronously are domain name resolution, socket setup, and (for TLS connections) TLS negotiation.

Many functions that interact with process objects, (for instance, process-datagram-address) rely on them at least having a socket before they can return a useful value. These functions will block until the socket has achieved the desired status. The recommended way of interacting with asynchronous sockets is to place a sentinel on the process, and not try to interact with it before it has changed status to ‘"run"’. That way, none of these functions will block.

:tls-parameters

When opening a TLS connection, this should be where the first element is the TLS type (which should either be gnutls-x509pki or gnutls-anon, and the remaining elements should form a keyword list acceptable for gnutls-boot. (This keyword list can be obtained from the gnutls-boot-parameters function.) The TLS connection will then be negotiated after completing the connection to the host.

:stop stopped

If stopped is non-nil, start the network connection or server in the stopped state.

:buffer buffer

Use buffer as the process buffer.

:coding coding

Use coding as the coding system for this process. To specify different coding systems for decoding data from the connection and for encoding data sent to it, specify (decoding . encoding) for coding.

If you don’t specify this keyword at all, the default is to determine the coding systems from the data.

:noquery query-flag

Initialize the process query flag to query-flag. See Querying Before Exit.

:filter filter

Initialize the process filter to filter.

:filter-multibyte multibyte

If multibyte is non-nil, strings given to the process filter are multibyte, otherwise they are unibyte. The default is t.

:sentinel sentinel

Initialize the process sentinel to sentinel.

:log log

Initialize the log function of a server process to log. The log function is called each time the server accepts a network connection from a client. The arguments passed to the log function are server, connection, and message; where server is the server process, connection is the new process for the connection, and message is a string describing what has happened.

:plist plist

Initialize the process plist to plist.

The original argument list, modified with the actual connection information, is available via the process-contact function.


40.17.2 Network Options

The following network options can be specified when you create a network process. Except for :reuseaddr, you can also set or modify these options later, using set-network-process-option.

For a server process, the options specified with make-network-process are not inherited by the client connections, so you will need to set the necessary options for each child connection as it is created.

:bindtodevice device-name

If device-name is a non-empty string identifying a network interface name (see network-interface-list), only handle packets received on that interface. If device-name is nil (the default), handle packets received on any interface.

Using this option may require special privileges on some systems.

:broadcast broadcast-flag

If broadcast-flag is non-nil for a datagram process, the process will receive datagram packet sent to a broadcast address, and be able to send packets to a broadcast address. This is ignored for a stream connection.

:dontroute dontroute-flag

If dontroute-flag is non-nil, the process can only send to hosts on the same network as the local host.

:keepalive keepalive-flag

If keepalive-flag is non-nil for a stream connection, enable exchange of low-level keep-alive messages.

:linger linger-arg

If linger-arg is non-nil, wait for successful transmission of all queued packets on the connection before it is deleted (see delete-process). If linger-arg is an integer, it specifies the maximum time in seconds to wait for queued packets to be sent before closing the connection. The default is nil, which means to discard unsent queued packets when the process is deleted.

:oobinline oobinline-flag

If oobinline-flag is non-nil for a stream connection, receive out-of-band data in the normal data stream. Otherwise, ignore out-of-band data.

:priority priority

Set the priority for packets sent on this connection to the integer priority. The interpretation of this number is protocol specific; such as setting the TOS (type of service) field on IP packets sent on this connection. It may also have system dependent effects, such as selecting a specific output queue on the network interface.

:reuseaddr reuseaddr-flag

If reuseaddr-flag is non-nil (the default) for a stream server process, allow this server to reuse a specific port number (see :service), unless another process on this host is already listening on that port. If reuseaddr-flag is nil, there may be a period of time after the last use of that port (by any process on the host) where it is not possible to make a new server on that port.

Function: set-network-process-option process option value &optional no-error

This function sets or modifies a network option for network process process. The accepted options and values are as for make-network-process. If no-error is non-nil, this function returns nil instead of signaling an error if option is not a supported option. If the function successfully completes, it returns t.

The current setting of an option is available via the process-contact function.


40.17.3 Testing Availability of Network Features

To test for the availability of a given network feature, use featurep like this:

(featurep 'make-network-process '(keyword value))

The result of this form is t if it works to specify keyword with value value in make-network-process. Here are some of the keywordvalue pairs you can test in this way.

(:nowait t)

Non-nil if non-blocking connect is supported.

(:type datagram)

Non-nil if datagrams are supported.

(:family local)

Non-nil if local (a.k.a. “UNIX domain”) sockets are supported.

(:family ipv6)

Non-nil if IPv6 is supported.

(:service t)

Non-nil if the system can select the port for a server.

To test for the availability of a given network option, use featurep like this:

(featurep 'make-network-process 'keyword)

The accepted keyword values are :bindtodevice, etc. For the complete list, see Network Options. This form returns non-nil if that particular network option is supported by make-network-process (or set-network-process-option).


40.18 Misc Network Facilities

These additional functions are useful for creating and operating on network connections. Note that they are supported only on some systems.

Function: network-interface-list &optional full family

This function returns a list describing the network interfaces of the machine you are using. The value is an alist whose elements have the form (ifname . address). ifname is a string naming the interface, address has the same form as the local-address and remote-address arguments to make-network-process, i.e. a vector of integers. By default both IPv4 and IPv6 addresses are returned if possible.

Optional argument full non-nil means to instead return a list of one or more elements of the form (ifname addr bcast netmask). ifname is a non-unique string naming the interface. addr, bcast, and netmask are vectors of integers detailing the IP address, broadcast address, and network mask.

Optional argument family specified as symbol ipv4 or ipv6 restricts the returned information to IPv4 and IPv6 addresses respectively, independently of the value of full. Specifying ipv6 when IPv6 support is not available will result in an error being signaled.

Some examples:

(network-interface-list) ⇒
(("vmnet8" .
  [172 16 76 1 0])
 ("vmnet1" .
  [172 16 206 1 0])
 ("lo0" .
  [65152 0 0 0 0 0 0 1 0])
 ("lo0" .
  [0 0 0 0 0 0 0 1 0])
 ("lo0" .
  [127 0 0 1 0]))
(network-interface-list t) ⇒
(("vmnet8"
  [172 16 76 1 0]
  [172 16 76 255 0]
  [255 255 255 0 0])
 ("vmnet1"
  [172 16 206 1 0]
  [172 16 206 255 0]
  [255 255 255 0 0])
 ("lo0"
  [65152 0 0 0 0 0 0 1 0]
  [65152 0 0 0 65535 65535 65535 65535 0]
  [65535 65535 65535 65535 0 0 0 0 0])
 ("lo0"
  [0 0 0 0 0 0 0 1 0]
  [0 0 0 0 0 0 0 1 0]
  [65535 65535 65535 65535 65535 65535 65535 65535 0])
 ("lo0"
  [127 0 0 1 0]
  [127 255 255 255 0]
  [255 0 0 0 0]))
Function: network-interface-info ifname

This function returns information about the network interface named ifname. The value is a list of the form (addr bcast netmask hwaddr flags).

addr

The Internet protocol address.

bcast

The broadcast address.

netmask

The network mask.

hwaddr

The layer 2 address (Ethernet MAC address, for instance).

flags

The current flags of the interface.

Note that this function returns only IPv4 information.

Function: format-network-address address &optional omit-port

This function converts the Lisp representation of a network address to a string.

A five-element vector [a b c d p] represents an IPv4 address a.b.c.d and port number p. format-network-address converts that to the string "a.b.c.d:p".

A nine-element vector [a b c d e f g h p] represents an IPv6 address along with a port number. format-network-address converts that to the string "[a:b:c:d:e:f:g:h]:p".

If the vector does not include the port number, p, or if omit-port is non-nil, the result does not include the :p suffix.

Function: network-lookup-address-info name &optional family hints

This function perform hostname lookups on name, which is expected to be an ASCII-only string, otherwise it signals an error. Call puny-encode-domain on name first if you wish to lookup internationalized hostnames.

If successful, this function returns a list of Lisp representations of network addresses (see make-network-process, for a description of the format), otherwise return nil. In the latter case, it also logs an error message hopefully explaining what went wrong.

By default, this function attempts both IPv4 and IPv6 lookups. The optional argument family controls this behavior, specifying the symbol ipv4 or ipv6 restricts lookups to IPv4 and IPv6 respectively.

If optional argument hints is numeric, the function treats the name as a numerical IP address (and does not perform DNS lookups). This can be used to check whether a string is a valid numerical representation of an IP address, or to convert a numerical string to its canonical representation. e.g.

(network-lookup-address-info "127.1" 'ipv4 'numeric)
    ⇒ ([127 0 0 1 0])

(network-lookup-address-info "::1" nil 'numeric)
    ⇒ ([0 0 0 0 0 0 0 1 0])

Be warned that there are some surprising valid forms, especially for IPv4, e.g ‘0xe3010203’ and ‘0343.1.2.3’ are both valid, as are ‘0’ and ‘1’ (but they are invalid for IPv6).


40.19 Communicating with Serial Ports

Emacs can communicate with serial ports. For interactive use, M-x serial-term opens a terminal window. In a Lisp program, make-serial-process creates a process object.

The serial port can be configured at run-time, without having to close and re-open it. The function serial-process-configure lets you change the speed, bytesize, and other parameters. In a terminal window created by serial-term, you can click on the mode line for configuration.

A serial connection is represented by a process object, which can be used in a similar way to a subprocess or network process. You can send and receive data, and configure the serial port. A serial process object has no process ID, however, and you can’t send signals to it, and the status codes are different from other types of processes. delete-process on the process object or kill-buffer on the process buffer close the connection, but this does not affect the device connected to the serial port.

The function process-type returns the symbol serial for a process object representing a serial port connection.

Serial ports are available on GNU/Linux, Unix, and MS Windows systems.

Command: serial-term port speed &optional line-mode

Start a terminal-emulator for a serial port in a new buffer. port is the name of the serial port to connect to. For example, this could be /dev/ttyS0 on Unix. On MS Windows, this could be COM1, or \\.\COM10 (double the backslashes in Lisp strings).

speed is the speed of the serial port in bits per second. 9600 is a common value. The buffer is in Term mode; see Term Mode in The GNU Emacs Manual, for the commands to use in that buffer. You can change the speed and the configuration in the mode line menu. If line-mode is non-nil, term-line-mode is used; otherwise term-raw-mode is used.

Function: make-serial-process &rest args

This function creates a process and a buffer. Arguments are specified as keyword/argument pairs. Here’s the list of the meaningful keywords, with the first two (port and speed) being mandatory:

:port port

This is the name of the serial port. On Unix and GNU systems, this is a file name such as /dev/ttyS0. On Windows, this could be COM1, or \\.\COM10 for ports higher than COM9 (double the backslashes in Lisp strings).

:speed speed

The speed of the serial port in bits per second. This function calls serial-process-configure to handle the speed; see the following documentation of that function for more details.

:name name

The name of the process. If name is not given, port will serve as the process name as well.

:buffer buffer

The buffer to associate with the process. The value can be either a buffer or a string that names a buffer. Process output goes at the end of that buffer, unless you specify an output stream or filter function to handle the output. If buffer is not given, the process buffer’s name is taken from the value of the :name keyword.

:coding coding

If coding is a symbol, it specifies the coding system used for both reading and writing for this process. If coding is a cons (decoding . encoding), decoding is used for reading, and encoding is used for writing. If not specified, the default is to determine the coding systems from the data itself.

:noquery query-flag

Initialize the process query flag to query-flag. See Querying Before Exit. The flags defaults to nil if unspecified.

:stop bool

Start process in the stopped state if bool is non-nil. In the stopped state, a serial process does not accept incoming data, but you can send outgoing data. The stopped state is cleared by continue-process and set by stop-process.

:filter filter

Install filter as the process filter.

:sentinel sentinel

Install sentinel as the process sentinel.

:plist plist

Install plist as the initial plist of the process.

:bytesize
:parity
:stopbits
:flowcontrol

These are handled by serial-process-configure, which is called by make-serial-process.

The original argument list, possibly modified by later configuration, is available via the function process-contact.

Here is an example:

(make-serial-process :port "/dev/ttyS0" :speed 9600)
Function: serial-process-configure &rest args

This function configures a serial port connection. Arguments are specified as keyword/argument pairs. Attributes that are not given are re-initialized from the process’s current configuration (available via the function process-contact), or set to reasonable default values. The following arguments are defined:

:process process
:name name
:buffer buffer
:port port

Any of these arguments can be given to identify the process that is to be configured. If none of these arguments is given, the current buffer’s process is used.

:speed speed

The speed of the serial port in bits per second, a.k.a. baud rate. The value can be any number, but most serial ports work only at a few defined values between 1200 and 115200, with 9600 being the most common value. If speed is nil, the function ignores all other arguments and does not configure the port. This may be useful for special serial ports such as Bluetooth-to-serial converters, which can only be configured through ‘AT’ commands sent through the connection. The value of nil for speed is valid only for connections that were already opened by a previous call to make-serial-process or serial-term.

:bytesize bytesize

The number of bits per byte, which can be 7 or 8. If bytesize is not given or nil, it defaults to 8.

:parity parity

The value can be nil (don’t use parity), the symbol odd (use odd parity), or the symbol even (use even parity). If parity is not given, it defaults to no parity.

:stopbits stopbits

The number of stopbits used to terminate a transmission of each byte. stopbits can be 1 or 2. If stopbits is not given or nil, it defaults to 1.

:flowcontrol flowcontrol

The type of flow control to use for this connection, which is either nil (don’t use flow control), the symbol hw (use RTS/CTS hardware flow control), or the symbol sw (use XON/XOFF software flow control). If flowcontrol is not given, it defaults to no flow control.

Internally, make-serial-process calls serial-process-configure for the initial configuration of the serial port.


40.20 Packing and Unpacking Byte Arrays

This section describes how to pack and unpack arrays of bytes, usually for binary network protocols. These functions convert byte arrays to alists, and vice versa. The byte array can be represented as a unibyte string or as a vector of integers, while the alist associates symbols either with fixed-size objects or with recursive sub-alists. To use the functions referred to in this section, load the bindat library.

Conversion from byte arrays to nested alists is also known as deserializing or unpacking, while going in the opposite direction is also known as serializing or packing.


40.20.1 Describing Data Layout

To control unpacking and packing, you write a data layout specification, also called a Bindat type expression. This can be a base type or a composite type made of several fields, where the specification controls the length of each field to be processed, and how to pack or unpack it. We normally keep bindat type values in variables whose names end in -bindat-spec; that kind of name is automatically recognized as risky (see File Local Variables).

Macro: bindat-type &rest type

Creates a Bindat type value object according to the Bindat type expression type.

A field’s type describes the size (in bytes) of the object that the field represents and, in the case of multibyte fields, how the bytes are ordered within the field. The two possible orderings are big endian (also known as “network byte ordering”) and little endian. For instance, the number #x23cd (decimal 9165) in big endian would be the two bytes #x23 #xcd; and in little endian, #xcd #x23. Here are the possible type values:

u8
byte

Unsigned byte, with length 1.

uint bitlen &optional le

Unsigned integer in network byte order (big-endian), with bitlen bits. bitlen has to be a multiple of 8. If le is non-nil, then use little-endian byte order.

sint bitlen le

Signed integer in network byte order (big-endian), with bitlen bits. bitlen has to be a multiple of 8. If le is non-nil, then use little-endian byte order.

str len

Unibyte string (see Text Representations) of length len bytes. When packing, the first len bytes of the input string are copied to the packed output. If the input string is shorter than len, the remaining bytes will be null (zero) unless a pre-allocated string was provided to bindat-pack, in which case the remaining bytes are left unmodified. If the input string is multibyte with only ASCII and eight-bit characters, it is converted to unibyte before it is packed; other multibyte strings signal an error. When unpacking, any null bytes in the packed input string will appear in the unpacked output.

strz &optional len

If len is not provided, this is a variable-length null-terminated unibyte string (see Text Representations). When packing into strz, the entire input string is copied to the packed output followed by a null (zero) byte. (If pre-allocated string is provided for packing into strz, that pre-allocated string should have enough space for the additional null byte appended to the output string contents, see Functions to Unpack and Pack Bytes). The length of the packed output is the length of the input string plus one (for the null terminator). The input string must not contain any null bytes. If the input string is multibyte with only ASCII and eight-bit characters, it is converted to unibyte before it is packed; other multibyte strings signal an error. When unpacking a strz, the resulting output string will contain all bytes up to (but excluding) the null byte that terminated the input string.

If len is provided, strz behaves the same as str, but with a couple of differences:

  • When packing, a null terminator is written after the packed input string if the number of characters in the input string is less than len.
  • When unpacking, the first null byte encountered in the packed string is interpreted as the terminating byte, and it and all subsequent bytes are excluded from the result of the unpacking.

Caution: The packed output will not be null-terminated unless the input string is shorter than len bytes or it contains a null byte within the first len bytes.

vec len [type]

Vector of len elements. The type of the elements is given by type, defaulting to bytes. The type can be any Bindat type expression.

repeat len [type]

Like vec, but it unpacks to and packs from lists, whereas vec unpacks to vectors.

bits len

List of bits that are set to 1 in len bytes. The bytes are taken in big-endian order, and the bits are numbered starting with 8 * len − 1 and ending with zero. For example: bits 2 unpacks #x28 #x1c to (2 3 4 11 13) and #x1c #x28 to (3 5 10 11 12).

fill len

len bytes used as a mere filler. In packing, these bytes are left unchanged, which normally means they remain zero. When unpacking, this just returns nil.

align len

Same as fill except the number of bytes is that needed to skip to the next multiple of len bytes.

type exp

This lets you refer to a type indirectly: exp is a Lisp expression which should return a Bindat type value.

unit exp

This is a trivial type which uses up 0 bits of space. exp describes the value returned when we try to “unpack” such a field.

struct fields...

Composite type made of several fields. Every field is of the form (name type) where type can be any Bindat type expression. name can be _ when the field’s value does not deserve to be named, as is often the case for align and fill fields. When the context makes it clear that this is a Bindat type expression, the symbol struct can be omitted.

In the types above, len and bitlen are given as an integer specifying the number of bytes (or bits) in the field. When the length of a field is not fixed, it typically depends on the value of preceding fields. For this reason, the length len does not have to be a constant but can be any Lisp expression and it can refer to the value of previous fields via their name.

For example, the specification of a data layout where a leading byte gives the size of a subsequent vector of 16 bit integers could be:

(bindat-type
  (len      u8)
  (payload  vec (1+ len) uint 16))

40.20.2 Functions to Unpack and Pack Bytes

In the following documentation, type refers to a Bindat type value as returned from bindat-type, raw to a byte array, and struct to an alist representing unpacked field data.

Function: bindat-unpack type raw &optional idx

This function unpacks data from the unibyte string or byte array raw according to type. Normally, this starts unpacking at the beginning of the byte array, but if idx is non-nil, it specifies a zero-based starting position to use instead.

The value is an alist or nested alist in which each element describes one unpacked field.

Function: bindat-get-field struct &rest name

This function selects a field’s data from the nested alist struct. Usually struct was returned by bindat-unpack. If name corresponds to just one argument, that means to extract a top-level field value. Multiple name arguments specify repeated lookup of sub-structures. An integer name acts as an array index.

For example, (bindat-get-field struct a b 2 c) means to find field c in the third element of subfield b of field a. (This corresponds to struct.a.b[2].c in the C programming language syntax.)

Although packing and unpacking operations change the organization of data (in memory), they preserve the data’s total length, which is the sum of all the fields’ lengths, in bytes. This value is not generally inherent in either the specification or alist alone; instead, both pieces of information contribute to its calculation. Likewise, the length of a string or array being unpacked may be longer than the data’s total length as described by the specification.

Function: bindat-length type struct

This function returns the total length of the data in struct, according to type.

Function: bindat-pack type struct &optional raw idx

This function returns a byte array packed according to type from the data in the alist struct. It normally creates and fills a new byte array starting at the beginning. However, if raw is non-nil, it specifies a pre-allocated unibyte string or vector to pack into. If idx is non-nil, it specifies the starting offset for packing into raw.

When pre-allocating, you should make sure (length raw) meets or exceeds the total length to avoid an out-of-range error.

Function: bindat-ip-to-string ip

Convert the Internet address vector ip to a string in the usual dotted notation.

(bindat-ip-to-string [127 0 0 1])
     ⇒ "127.0.0.1"

40.20.3 Advanced data layout specifications

Bindat type expressions are not limited to the types described earlier. They can also be arbitrary Lisp forms returning Bindat type expressions. For example, the type below describes data which can either contain a 24-bit error code or a vector of bytes:

(bindat-type
  (len      u8)
  (payload  . (if (zerop len) (uint 24) (vec (1- len)))))

Furthermore, while composite types are normally unpacked to (and packed from) association lists, this can be changed via the use of the following special keyword arguments:

:unpack-val exp

When the list of fields ends with this keyword argument, then the value returned when unpacking is the value of exp instead of the standard alist. exp can refer to all the previous fields by their name.

:pack-val exp

If a field’s type is followed by this keyword argument, then the value packed into this field is returned by exp instead of being extracted from the alist.

:pack-var name

If the list of fields is preceded by this keyword argument, then all the subsequent :pack-val arguments can refer to the overall value to pack into this composite type via the variable named name.

For example, one could describe a 16-bit signed integer as follows:

(defconst sint16-bindat-spec
  (let* ((max (ash 1 15))
         (wrap (+ max max)))
    (bindat-type :pack-var v
                 (n uint 16 :pack-val (if (< v 0) (+ v wrap) v))
                 :unpack-val (if (>= n max) (- n wrap) n))))

Which would then behave as follows:

(bindat-pack sint16-bindat-spec -8)
     ⇒ "\377\370"

(bindat-unpack sint16-bindat-spec "\300\100")
     ⇒ -16320

Finally, you can define new Bindat type forms to use in Bindat type expressions with bindat-defmacro:

Macro: bindat-defmacro name args &rest body

Define a new Bindat type expression named name and taking arguments args. Its behavior follows that of defmacro, which the important difference that the new forms can only be used within Bindat type expressions.


41 Emacs Display

This chapter describes a number of features related to the display that Emacs presents to the user.


41.1 Refreshing the Screen

The function redraw-frame clears and redisplays the entire contents of a given frame (see Frames). This is useful if the screen is corrupted.

Function: redraw-frame &optional frame

This function clears and redisplays frame frame. If frame is omitted or nil, it redraws the selected frame.

Even more powerful is redraw-display:

Command: redraw-display

This function clears and redisplays all visible frames.

In Emacs, processing user input takes priority over redisplay. If you call these functions when input is available, they don’t redisplay immediately, but the requested redisplay does happen eventually—after all the input has been processed.

On text terminals, suspending and resuming Emacs normally also refreshes the screen. Some terminal emulators record separate contents for display-oriented programs such as Emacs and for ordinary sequential display. If you are using such a terminal, you might want to inhibit the redisplay on resumption.

User Option: no-redraw-on-reenter

This variable controls whether Emacs redraws the entire screen after it has been suspended and resumed. Non-nil means there is no need to redraw, nil means redrawing is needed. The default is nil.


41.2 Forcing Redisplay

Emacs normally tries to redisplay the screen whenever it waits for input. With the following function, you can request an immediate attempt to redisplay, in the middle of Lisp code, without actually waiting for input.

Function: redisplay &optional force

This function tries immediately to redisplay. The optional argument force, if non-nil, forces the redisplay to be performed, instead of being preempted if input is pending.

The function returns t if it actually tried to redisplay, and nil otherwise. A value of t does not mean that redisplay proceeded to completion; it could have been preempted by newly arriving input.

Although redisplay tries immediately to redisplay, it does not change how Emacs decides which parts of its frame(s) to redisplay. By contrast, the following function adds certain windows to the pending redisplay work (as if their contents had completely changed), but does not immediately try to perform redisplay.

Function: force-window-update &optional object

This function forces some or all windows to be updated the next time Emacs does a redisplay. If object is a window, that window is to be updated. If object is a buffer or buffer name, all windows displaying that buffer are to be updated. If object is nil (or omitted), all windows are to be updated.

This function does not do a redisplay immediately; Emacs does that as it waits for input, or when the function redisplay is called.

Variable: pre-redisplay-function

A function run just before redisplay. It is called with one argument, the set of windows to be redisplayed. The set can be nil, meaning only the selected window, or t, meaning all the windows.

Variable: pre-redisplay-functions

This hook is run just before redisplay. It is called once in each window that is about to be redisplayed, with current-buffer set to the buffer displayed in that window.


41.3 Truncation

When a line of text extends beyond the right edge of a window, Emacs can continue the line (make it wrap to the next screen line), or truncate the line (limit it to one screen line). The additional screen lines used to display a long text line are called continuation lines. Continuation is not the same as filling; continuation happens on the screen only, not in the buffer contents, and it breaks a line precisely at the right margin, not at a word boundary. See Filling.

On a graphical display, tiny arrow images in the window fringes indicate truncated and continued lines (see Fringes). On a text terminal, a ‘$’ in the rightmost column of the window indicates truncation; a ‘\’ on the rightmost column indicates a line that wraps. (The display table can specify alternate characters to use for this; see Display Tables).

Since wrapping and truncation of text contradict each other, Emacs turns off line truncation when wrapping is requested, and vice versa.

User Option: truncate-lines

If this buffer-local variable is non-nil, lines that extend beyond the right edge of the window are truncated; otherwise, they are continued. As a special exception, the variable truncate-partial-width-windows takes precedence in partial-width windows (i.e., windows that do not occupy the entire frame width).

User Option: truncate-partial-width-windows

This variable controls line truncation in partial-width windows. A partial-width window is one that does not occupy the entire frame width (see Splitting Windows). If the value is nil, line truncation is determined by the variable truncate-lines (see above). If the value is an integer n, lines are truncated if the partial-width window has fewer than n columns, regardless of the value of truncate-lines; if the partial-width window has n or more columns, line truncation is determined by truncate-lines. For any other non-nil value, lines are truncated in every partial-width window, regardless of the value of truncate-lines.

When horizontal scrolling (see Horizontal Scrolling) is in use in a window, that forces truncation.

Variable: wrap-prefix

If this buffer-local variable is non-nil, it defines a wrap prefix which Emacs displays at the start of every continuation line. (If lines are truncated, wrap-prefix is never used.) Its value may be a string or an image (see Other Display Specifications), or a stretch of whitespace such as specified by the :width or :align-to display properties (see Specified Spaces). The value is interpreted in the same way as a display text property. See The display Property.

A wrap prefix may also be specified for regions of text, using the wrap-prefix text or overlay property. This takes precedence over the wrap-prefix variable. See Properties with Special Meanings.

Variable: line-prefix

If this buffer-local variable is non-nil, it defines a line prefix which Emacs displays at the start of every non-continuation line. Its value may be a string or an image (see Other Display Specifications), or a stretch of whitespace such as specified by the :width or :align-to display properties (see Specified Spaces). The value is interpreted in the same way as a display text property. See The display Property.

A line prefix may also be specified for regions of text using the line-prefix text or overlay property. This takes precedence over the line-prefix variable. See Properties with Special Meanings.


41.4 The Echo Area

The echo area is used for displaying error messages (see Errors), for messages made with the message primitive, and for echoing keystrokes. It is not the same as the minibuffer, despite the fact that the minibuffer appears (when active) in the same place on the screen as the echo area. See The Minibuffer in The GNU Emacs Manual.

Apart from the functions documented in this section, you can print Lisp objects to the echo area by specifying t as the output stream. See Output Streams.


41.4.1 Displaying Messages in the Echo Area

This section describes the standard functions for displaying messages in the echo area.

Function: message format-string &rest arguments

This function displays a message in the echo area. format-string is a format string, and arguments are the objects for its format specifications, like in the format-message function (see Formatting Strings). The resulting formatted string is displayed in the echo area; if it contains face text properties, it is displayed with the specified faces (see Faces). The string is also added to the *Messages* buffer, but without text properties (see Logging Messages in *Messages*).

Typically grave accent and apostrophe in the format translate to matching curved quotes, e.g., "Missing `%s'" might result in "Missing ‘foo’". See Text Quoting Style, for how to influence or inhibit this translation.

In batch mode, the message is printed to the standard error stream, followed by a newline.

When inhibit-message is non-nil, no message will be displayed in the echo area, it will only be logged to ‘*Messages*’.

If format-string is nil or the empty string, message clears the echo area; if the echo area has been expanded automatically, this brings it back to its normal size. If the minibuffer is active, this brings the minibuffer contents back onto the screen immediately.

(message "Reverting `%s'..." (buffer-name))
 -| Reverting ‘subr.el’...
⇒ "Reverting ‘subr.el’..."

---------- Echo Area ----------
Reverting ‘subr.el’...
---------- Echo Area ----------

To automatically display a message in the echo area or in a pop-buffer, depending on its size, use display-message-or-buffer (see below).

Warning: If you want to use your own string as a message verbatim, don’t just write (message string). If string contains ‘%’, ‘`’, or ‘'’ it may be reformatted, with undesirable results. Instead, use (message "%s" string).

The following facilities allow users and Lisp programs to control how echo-area messages are displayed.

Variable: set-message-function

If this variable is non-nil, it should be a function of one argument, the text of a message to display in the echo area. That function will be called by message and related functions. If the function returns nil, the message is displayed in the echo area as usual. If the function returns a string, that string is displayed in the echo area instead of the original message. If the function returns any other non-nil value, that means the message was already handled, so message will not display anything in the echo area.

The default value calls set-minibuffer-message, described below.

Variable: clear-message-function

If this variable is non-nil, it should be a function of no arguments; message and related functions call it when their argument message is nil or the empty string, to clear the echo area.

Usually this function is called when the next input event arrives after displaying an echo-area message. The function is expected to clear the message displayed by its counterpart function specified by set-message-function, but doesn’t have to. If the function wants the echo area to remain uncleared, it should return the symbol dont-clear-message; any other value will result in the echo area being cleared.

The default value is the function that clears the message displayed in an active minibuffer.

User Option: set-message-functions

The value of this user option is a list of functions to be called for handling display of echo-area messages. Each function is called with one argument, the text of the message to display. If the function returns a string, that string replaces the original message, and the next function in the list is called with the new message text. If the function returns nil, the next function in the list is called with the same text; if the last function in the list returns nil, the message text is displayed in the echo area. If the function returns a non-nil value that is not a string, the message is considered to be handled, and no further functions in the list are called.

The three useful functions to be put in the list that is the value of this option are described below.

Function: set-minibuffer-message message

This function displays message in the echo-area when the minibuffer is not active, and at the end of the minibuffer when the minibuffer is active. However, if the text shown in the active minibuffer has the minibuffer-message text property (see Properties with Special Meanings) on some character, the message will be displayed before the first character having that property.

This function is by default the only member of the list in set-message-functions.

Function: inhibit-message message

If an echo-area message matches any regexp in the list that is the value of the user option inhibit-message-regexps, this function suppresses the display of that message and returns a non-nil value that is not a string. Thus, if this function is in the list set-message-functions, the rest of the functions in the list will not be called when message matches the regexps in inhibit-message-regexps. To ensure a matching message will never be displayed, make this function be the first element of the list in set-message-functions.

Function: set-multi-message message

This function accumulates several echo-area messages emitted one after another, and returns them as a single string in which individual messages are separated by newlines. Up to multi-message-max recent messages can be accumulated. The accumulated messages are discarded when more than multi-message-timeout seconds have elapsed since the time the first message was emitted.

Variable: inhibit-message

When this variable is non-nil, message and related functions will not display any messages in the Echo Area. Echo-area messages are still logged in the *Messages* buffer, though.

Macro: with-temp-message message &rest body

This construct displays a message in the echo area temporarily, during the execution of body. It displays message, executes body, then returns the value of the last body form while restoring the previous echo area contents.

Function: message-or-box format-string &rest arguments

This function displays a message like message, but may display it in a dialog box instead of the echo area. If this function is called in a command that was invoked using the mouse—more precisely, if last-nonmenu-event (see Information from the Command Loop) is either nil or a list—then it uses a dialog box or pop-up menu to display the message. Otherwise, it uses the echo area. (This is the same criterion that y-or-n-p uses to make a similar decision; see Yes-or-No Queries.)

You can force use of the mouse or of the echo area by binding last-nonmenu-event to a suitable value around the call.

Function: message-box format-string &rest arguments

This function displays a message like message, but uses a dialog box (or a pop-up menu) whenever that is possible. If it is impossible to use a dialog box or pop-up menu, because the terminal does not support them, then message-box uses the echo area, like message.

Function: display-message-or-buffer message &optional buffer-name action frame

This function displays the message message, which may be either a string or a buffer. If it is shorter than the maximum height of the echo area, as defined by max-mini-window-height, it is displayed in the echo area, using message. Otherwise, display-buffer is used to show it in a pop-up buffer.

Returns either the string shown in the echo area, or when a pop-up buffer is used, the window used to display it.

If message is a string, then the optional argument buffer-name is the name of the buffer used to display it when a pop-up buffer is used, defaulting to *Message*. In the case where message is a string and displayed in the echo area, it is not specified whether the contents are inserted into the buffer anyway.

The optional arguments action and frame are as for display-buffer, and only used if a buffer is displayed.

Function: current-message

This function returns the message currently being displayed in the echo area, or nil if there is none.


41.4.2 Reporting Operation Progress

When an operation can take a while to finish, you should inform the user about the progress it makes. This way the user can estimate remaining time and clearly see that Emacs is busy working, not hung. A convenient way to do this is to use a progress reporter.

Here is a working example that does nothing useful:

(let ((progress-reporter
       (make-progress-reporter "Collecting mana for Emacs..."
                               0  500)))
  (dotimes (k 500)
    (sit-for 0.01)
    (progress-reporter-update progress-reporter k))
  (progress-reporter-done progress-reporter))
Function: make-progress-reporter message &optional min-value max-value current-value min-change min-time

This function creates and returns a progress reporter object, which you will use as an argument for the other functions listed below. The idea is to precompute as much data as possible to make progress reporting very fast.

When this progress reporter is subsequently used, it will display message in the echo area, followed by progress percentage. message is treated as a simple string. If you need it to depend on a filename, for instance, use format-message before calling this function.

The arguments min-value and max-value should be numbers standing for the starting and final states of the operation. For instance, an operation that scans a buffer should set these to the results of point-min and point-max correspondingly. max-value should be greater than min-value.

Alternatively, you can set min-value and max-value to nil. In that case, the progress reporter does not report process percentages; it instead displays a “spinner” that rotates a notch each time you update the progress reporter.

If min-value and max-value are numbers, you can give the argument current-value a numerical value specifying the initial progress; if omitted, this defaults to min-value.

The remaining arguments control the rate of echo area updates. The progress reporter will wait for at least min-change more percents of the operation to be completed before printing next message; the default is one percent. min-time specifies the minimum time in seconds to pass between successive prints; the default is 0.2 seconds. (On some operating systems, the progress reporter may handle fractions of seconds with varying precision).

This function calls progress-reporter-update, so the first message is printed immediately.

Function: progress-reporter-update reporter &optional value suffix

This function does the main work of reporting progress of your operation. It displays the message of reporter, followed by progress percentage determined by value. If percentage is zero, or close enough according to the min-change and min-time arguments, then it is omitted from the output.

reporter must be the result of a call to make-progress-reporter. value specifies the current state of your operation and must be between min-value and max-value (inclusive) as passed to make-progress-reporter. For instance, if you scan a buffer, then value should be the result of a call to point.

Optional argument suffix is a string to be displayed after reporter’s main message and progress text. If reporter is a non-numerical reporter, then value should be nil, or a string to use instead of suffix.

This function respects min-change and min-time as passed to make-progress-reporter and so does not output new messages on every invocation. It is thus very fast and normally you should not try to reduce the number of calls to it: resulting overhead will most likely negate your effort.

Function: progress-reporter-force-update reporter &optional value new-message suffix

This function is similar to progress-reporter-update except that it prints a message in the echo area unconditionally.

reporter, value, and suffix have the same meaning as for progress-reporter-update. Optional new-message allows you to change the message of the reporter. Since this function always updates the echo area, such a change will be immediately presented to the user.

Function: progress-reporter-done reporter

This function should be called when the operation is finished. It prints the message of reporter followed by word ‘done’ in the echo area.

You should always call this function and not hope for progress-reporter-update to print ‘100%’. Firstly, it may never print it, there are many good reasons for this not to happen. Secondly, ‘done’ is more explicit.

Macro: dotimes-with-progress-reporter (var count [result]) reporter-or-message body…

This is a convenience macro that works the same way as dotimes does, but also reports loop progress using the functions described above. It allows you to save some typing. The argument reporter-or-message can be either a string or a progress reporter object.

You can rewrite the example in the beginning of this subsection using this macro as follows:

(dotimes-with-progress-reporter
    (k 500)
    "Collecting some mana for Emacs..."
  (sit-for 0.01))

Using a reporter object as the reporter-or-message argument is useful if you want to specify the optional arguments in make-progress-reporter. For instance, you can write the previous example as follows:

(dotimes-with-progress-reporter
    (k 500)
    (make-progress-reporter "Collecting some mana for Emacs..." 0 500 0 1 1.5)
  (sit-for 0.01))
Macro: dolist-with-progress-reporter (var list [result]) reporter-or-message body…

This is another convenience macro that works the same way as dolist does, but also reports loop progress using the functions described above. As in dotimes-with-progress-reporter, reporter-or-message can be a progress reporter or a string. You can rewrite the previous example with this macro as follows:

(dolist-with-progress-reporter
    (k (number-sequence 0 500))
    "Collecting some mana for Emacs..."
  (sit-for 0.01))
Macro: with-delayed-message (timeout message) body…

Sometimes it’s unclear whether an operation will take a long time to execute or not, or it can be inconvenient to implement a progress reporter. This macro can be used in those situations.

(with-delayed-message (2 (format "Gathering data for %s" entry))
  (setq data (gather-data entry)))

In this example, if the body takes more than two seconds to execute, the message will be displayed. If it takes a shorter time than that, the message won’t be displayed. In either case, the body is evaluated as normally, and the return value of the final element in the body is the return value of the macro.

The message element is evaluated before body, and is always evaluated, whether the message is displayed or not.


41.4.3 Logging Messages in *Messages*

Almost all the messages displayed in the echo area are also recorded in the *Messages* buffer so that the user can refer back to them. This includes all the messages that are output with message. By default, this buffer is read-only and uses the major mode messages-buffer-mode. Nothing prevents the user from killing the *Messages* buffer, but the next display of a message recreates it. Any Lisp code that needs to access the *Messages* buffer directly and wants to ensure that it exists should use the function messages-buffer.

Function: messages-buffer

This function returns the *Messages* buffer. If it does not exist, it creates it, and switches it to messages-buffer-mode.

User Option: message-log-max

This variable specifies how many lines to keep in the *Messages* buffer. The value t means there is no limit on how many lines to keep. The value nil disables message logging entirely. Here’s how to display a message and prevent it from being logged:

(let (message-log-max)
  (message …))
Variable: messages-buffer-name

This variable has the name of the buffer where messages should be logged to, and defaults to *Messages*. Some packages may find it useful to temporarily redirect the output to a different buffer (perhaps to write the buffer out to a log file later), and they can bind this variable to a different buffer name. (Note that this buffer (if it doesn’t exist already), will be created and put into messages-buffer-mode.)

To make *Messages* more convenient for the user, the logging facility combines successive identical messages. It also combines successive related messages for the sake of two cases: question followed by answer, and a series of progress messages.

A question followed by an answer has two messages like the ones produced by y-or-n-p: the first is ‘question’, and the second is ‘question...answer’. The first message conveys no additional information beyond what’s in the second, so logging the second message discards the first from the log.

A series of progress messages has successive messages like those produced by make-progress-reporter. They have the form ‘base...how-far’, where base is the same each time, while how-far varies. Logging each message in the series discards the previous one, provided they are consecutive.

The functions make-progress-reporter and y-or-n-p don’t have to do anything special to activate the message log combination feature. It operates whenever two consecutive messages are logged that share a common prefix ending in ‘...’.


41.4.4 Echo Area Customization

These variables control details of how the echo area works.

Variable: cursor-in-echo-area

This variable controls where the cursor appears when a message is displayed in the echo area. If it is non-nil, then the cursor appears at the end of the message. Otherwise, the cursor appears at point—not in the echo area at all.

The value is normally nil; Lisp programs bind it to t for brief periods of time.

Variable: echo-area-clear-hook

This normal hook is run whenever the echo area is cleared—either by (message nil) or for any other reason.

User Option: echo-keystrokes

This variable determines how much time should elapse before command characters echo. Its value must be a number, and specifies the number of seconds to wait before echoing. If the user types a prefix key (such as C-x) and then delays this many seconds before continuing, the prefix key is echoed in the echo area. (Once echoing begins in a key sequence, all subsequent characters in the same key sequence are echoed immediately.)

If the value is zero, then command input is not echoed.

Variable: message-truncate-lines

Normally, displaying a long message resizes the echo area to display the entire message, wrapping long line as needed. But if the variable message-truncate-lines is non-nil, long lines of echo-area message are instead truncated to fit the mini-window width.

The variable max-mini-window-height, which specifies the maximum height for resizing minibuffer windows, also applies to the echo area (which is really a special use of the minibuffer window; see Minibuffer Windows).


41.5 Reporting Warnings

Warnings are a facility for a program to inform the user of a possible problem, but continue running (as opposed to signaling an error, see Errors).


41.5.1 Warning Basics

Every warning is a textual message, which explains the problem for the user, with the associated severity level which is a symbol. Here are the supported severity levels, in order of decreasing severity, and their meanings:

:emergency

A problem that will seriously impair Emacs operation soon if the user does not attend to it promptly.

:error

A report about data or circumstances that are inherently wrong.

:warning

A report about data or circumstances that are not inherently wrong, but raise suspicion of a possible problem.

:debug

A report of information that may be useful if the user is currently debugging the Lisp program which issues the warning.

When your program encounters invalid input data, it can either signal a Lisp error by calling error or signal (see How to Signal an Error) or report a warning with severity :error. Signaling a Lisp error is the easiest thing to do, but it means the signaling program cannot continue execution. If you want to take the trouble of implementing a way to continue processing despite the invalid data, then reporting a warning of severity :error is the right way of informing the user of the problem. For instance, the Emacs Lisp byte compiler can report an error that way and continue compiling other functions. (If the program signals a Lisp error and then handles it with condition-case, the user won’t see the error message; reporting that as a warning instead avoids that problem.)

In addition to severity level, each warning has a warning type to classify it. The warning type is either a symbol or a list of symbols. If it is a symbol, it should be the custom group that you use for the program’s user options; if it is a list, the first element of the list should be that custom group. For example, byte compiler warnings use the warning type (bytecomp). If the warning type is a list, the elements of the list after the first one, which should be arbitrary symbols, represent subcategories of the warning: they will be displayed to the user to better explain the nature of the warning.

Function: display-warning type message &optional level buffer-name

This function reports a warning, using the string message as the warning text and type as the warning type. level should be the severity level, and defaults to :warning if omitted or nil.

buffer-name, if non-nil, specifies the name of the buffer for logging the warning message. By default, it is *Warnings*.

Function: lwarn type level message &rest args

This function reports a warning using the value returned by (format-message message args…) as the message text in the *Warnings* buffer. In other respects it is equivalent to display-warning.

Function: warn message &rest args

This function reports a warning using the value returned by (format-message message args…) as the message text, emacs as the warning type, and :warning as the severity level. It exists for compatibility only; we recommend not using it, because you should specify a specific warning type.


41.5.2 Warning Variables

Programs can customize how their warnings appear by binding the variables described in this section.

Variable: warning-levels

This list defines the meaning and severity order of the warning severity levels. Each element defines one severity level, and they are arranged in order of decreasing severity.

Each element has the form (level string [function]), where level is the severity level it defines. string specifies the textual description of this level. string should use ‘%s’ to specify where to put the warning type information, or it can omit the ‘%s’ so as not to include that information.

The optional function, if non-nil, is a function to call with no arguments, to get the user’s attention. A notable example is ding (see Beeping).

Normally you should not change the value of this variable.

Variable: warning-prefix-function

If non-nil, the value is a function to generate prefix text for warnings. Programs can bind the variable to a suitable function. display-warning calls this function with the warnings buffer the current buffer, and the function can insert text into it. That text becomes the beginning of the warning message.

The function is called with two arguments, the severity level and its entry in warning-levels. It should return a list to use instead of that entry (the value need not be an actual member of warning-levels, but it must have the same structure). By constructing this value, the function can change the severity of the warning, or specify different handling for a given severity level.

If the variable’s value is nil, there’s no prefix text, before the warning is displayed, starting with the string part of the entry in warning-levels corresponding to the warning’s level.

Variable: warning-series

Programs can bind this variable to t to say that the next warning should begin a series. When several warnings form a series, that means to leave point on the first warning of the series, rather than keep moving it for each warning so that it appears on the last one. The series ends when the local binding of this variable is unbound and warning-series becomes nil again.

The value can also be a symbol with a function definition. That is equivalent to t, except that the next warning will also call the function with no arguments with the warnings buffer the current buffer. The function can, for example, insert text which will serve as a header for the series of warnings.

Once a series has begun, the value of this variable is a marker which points to the buffer position in the warnings buffer of the start of the series.

The variable’s normal value is nil, which means to handle each warning separately.

Variable: warning-fill-prefix

When this variable is non-nil, it specifies a fill prefix to use for filling the text of each warning.

Variable: warning-fill-column

The column at which to fill warnings.

Variable: warning-type-format

This variable specifies the format for displaying the warning type in the warning text. The result of formatting the type this way gets included in the message under the control of the string in the entry in warning-levels. The default value is " (%s)". If you bind it to the empty string "" then the warning type won’t appear at all.


41.5.3 Warning Options

These variables are used by users to control what happens when a Lisp program reports a warning.

User Option: warning-minimum-level

This user option specifies the minimum severity level that should be shown immediately to the user, by popping the warnings buffer in some window. The default is :warning, which means to show the warning buffer for any warning severity except :debug. The warnings of lower severity levels will still be written into the warnings buffer, but the buffer will not be forced onto display.

User Option: warning-minimum-log-level

This user option specifies the minimum severity level that should be logged in the warnings buffer. Warnings of lower severity will be completely ignored: not written to the warnings buffer and not displayed. The default is :warning, which means to log warnings of any severity except :debug.

User Option: warning-suppress-types

This list specifies which warning types should not be displayed immediately when they occur. Each element of the list should be a list of symbols. If an element of this list has the same elements as the first elements in a warning type, then the warning of that type will not be shown on display by popping the warnings buffer in some window (the warning will still be logged in the warnings buffer).

For example, if the value of this variable is a list like this:

((foo) (bar subtype))

then warnings whose types are foo or (foo) or (foo something) or (bar subtype other) will not be shown to the user.

User Option: warning-suppress-log-types

This list specifies which warning types should be ignored: not logged in the warnings buffer and not shown to the user. The structure and the matching of warning types are the same as for warning-suppress-types above.

During startup, Emacs delays showing any warnings until after it loads and processes the site-wide and user’s init files (see Summary: Sequence of Actions at Startup). Let-binding (see Local Variables) the values of these options around some code in your init files which might emit a warning will therefore not work, because it will not be in effect by the time the warning is actually processed. Thus, if you want to suppress some warnings during startup, change the values of the above options in your init file early enough, or put those let-binding forms in your after-init-hook or emacs-startup-hook functions. See The Init File.


41.5.4 Delayed Warnings

Sometimes, you may wish to avoid showing a warning while a command is running, and only show it only after the end of the command. You can use the function delay-warning for this. Emacs automatically delays any warnings emitted during the early stages of startup, and shows them only after the init files are processed.

Function: delay-warning type message &optional level buffer-name

This function is the delayed counterpart to display-warning (see Warning Basics), and it is called with the same arguments. The warning message is queued into delayed-warnings-list.

Variable: delayed-warnings-list

The value of this variable is a list of warnings to be displayed after the current command has finished. Each element must be a list

(type message [level [buffer-name]])

with the same form, and the same meanings, as the argument list of display-warning. Immediately after running post-command-hook (see Command Loop Overview), the Emacs command loop displays all the warnings specified by this variable, then resets the variable to nil.

Programs which need to further customize the delayed warnings mechanism can change the variable delayed-warnings-hook:

Variable: delayed-warnings-hook

This is a normal hook which is run by the Emacs command loop, after post-command-hook, in order to process and display delayed warnings. Emacs also runs this hook during startup, after loading the site-start and user init files (see Summary: Sequence of Actions at Startup), because warnings emitted before that are automatically delayed.

Its default value is a list of two functions:

(collapse-delayed-warnings display-delayed-warnings)

The function collapse-delayed-warnings removes repeated entries from delayed-warnings-list. The function display-delayed-warnings calls display-warning on each of the entries in delayed-warnings-list, in turn, and then sets delayed-warnings-list to nil.


41.6 Invisible Text

You can make characters invisible, so that they do not appear on the screen, with the invisible property. This can be either a text property (see Text Properties) or an overlay property (see Overlays). Cursor motion also partly ignores these characters; if the command loop finds that point is inside a range of invisible text after a command, it relocates point to the other side of the text.

In the simplest case, any non-nil invisible property makes a character invisible. This is the default case—if you don’t alter the default value of buffer-invisibility-spec, this is how the invisible property works. You should normally use t as the value of the invisible property if you don’t plan to set buffer-invisibility-spec yourself.

More generally, you can use the variable buffer-invisibility-spec to control which values of the invisible property make text invisible. This permits you to classify the text into different subsets in advance, by giving them different invisible values, and subsequently make various subsets visible or invisible by changing the value of buffer-invisibility-spec.

Controlling visibility with buffer-invisibility-spec is especially useful in a program to display the list of entries in a database. It permits the implementation of convenient filtering commands to view just a part of the entries in the database. Setting this variable is very fast, much faster than scanning all the text in the buffer looking for properties to change.

Variable: buffer-invisibility-spec

This variable specifies which kinds of invisible properties actually make a character invisible. Setting this variable makes it buffer-local.

t

A character is invisible if its invisible property is non-nil. This is the default.

a list

Each element of the list specifies a criterion for invisibility; if a character’s invisible property fits any one of these criteria, the character is invisible. The list can have two kinds of elements:

atom

A character is invisible if its invisible property value is atom or if it is a list with atom as a member; comparison is done with eq.

(atom . t)

A character is invisible if its invisible property value is atom or if it is a list with atom as a member; comparison is done with eq. Moreover, a sequence of such characters displays as an ellipsis.

Two functions are specifically provided for adding elements to buffer-invisibility-spec and removing elements from it.

Function: add-to-invisibility-spec element

This function adds the element element to buffer-invisibility-spec. If buffer-invisibility-spec was t, it changes to a list, (t), so that text whose invisible property is t remains invisible.

Function: remove-from-invisibility-spec element

This removes the element element from buffer-invisibility-spec. This does nothing if element is not in the list.

A convention for use of buffer-invisibility-spec is that a major mode should use the mode’s own name as an element of buffer-invisibility-spec and as the value of the invisible property:

;; If you want to display an ellipsis:
(add-to-invisibility-spec '(my-symbol . t))
;; If you don’t want ellipsis:
(add-to-invisibility-spec 'my-symbol)

(overlay-put (make-overlay beginning end)
             'invisible 'my-symbol)

;; When done with the invisibility:
(remove-from-invisibility-spec '(my-symbol . t))
;; Or respectively:
(remove-from-invisibility-spec 'my-symbol)

You can check for invisibility using the following function:

Function: invisible-p pos-or-prop

If pos-or-prop is a marker or number, this function returns a non-nil value if the text at that position is currently invisible.

If pos-or-prop is any other kind of Lisp object, that is taken to mean a possible value of the invisible text or overlay property. In that case, this function returns a non-nil value if that value would cause text to become invisible, based on the current value of buffer-invisibility-spec.

The return value of this function is t if the text would be completely hidden on display, or a non-nil, non-t value if the text would be replaced by an ellipsis.

Ordinarily, functions that operate on text or move point do not care whether the text is invisible, they process invisible characters and visible characters alike. The user-level line motion commands, such as next-line, previous-line, ignore invisible newlines if line-move-ignore-invisible is non-nil (the default), i.e., behave like these invisible newlines didn’t exist in the buffer, but only because they are explicitly programmed to do so.

If a command ends with point inside or at the boundary of invisible text, the main editing loop relocates point to one of the two ends of the invisible text. Emacs chooses the direction of relocation so that it is the same as the overall movement direction of the command; if in doubt, it prefers a position where an inserted char would not inherit the invisible property. Additionally, if the text is not replaced by an ellipsis and the command only moved within the invisible text, then point is moved one extra character so as to try and reflect the command’s movement by a visible movement of the cursor.

Thus, if the command moved point back to an invisible range (with the usual stickiness), Emacs moves point back to the beginning of that range. If the command moved point forward into an invisible range, Emacs moves point forward to the first visible character that follows the invisible text and then forward one more character.

These adjustments of point that ended up in the middle of invisible text can be disabled by setting disable-point-adjustment to a non-nil value. See Adjusting Point After Commands.

Incremental search can make invisible overlays visible temporarily and/or permanently when a match includes invisible text. To enable this, the overlay should have a non-nil isearch-open-invisible property. The property value should be a function to be called with the overlay as an argument. This function should make the overlay visible permanently; it is used when the match overlaps the overlay on exit from the search.

During the search, such overlays are made temporarily visible by temporarily modifying their invisible and intangible properties. If you want this to be done differently for a certain overlay, give it an isearch-open-invisible-temporary property which is a function. The function is called with two arguments: the first is the overlay, and the second is nil to make the overlay visible, or t to make it invisible again.


41.7 Selective Display

Selective display refers to a pair of related features for hiding certain lines on the screen.

The first variant, explicit selective display, was designed for use in a Lisp program: it controls which lines are hidden by altering the text. This kind of hiding is now obsolete and deprecated; instead you should use the invisible property (see Invisible Text) to get the same effect.

In the second variant, the choice of lines to hide is made automatically based on indentation. This variant is designed to be a user-level feature.

The way you control explicit selective display is by replacing a newline (control-j) with a carriage return (control-m). The text that was formerly a line following that newline is now hidden. Strictly speaking, it is temporarily no longer a line at all, since only newlines can separate lines; it is now part of the previous line.

Selective display does not directly affect editing commands. For example, C-f (forward-char) moves point unhesitatingly into hidden text. However, the replacement of newline characters with carriage return characters affects some editing commands. For example, next-line skips hidden lines, since it searches only for newlines. Modes that use selective display can also define commands that take account of the newlines, or that control which parts of the text are hidden.

When you write a selectively displayed buffer into a file, all the control-m’s are output as newlines. This means that when you next read in the file, it looks OK, with nothing hidden. The selective display effect is seen only within Emacs.

Variable: selective-display

This buffer-local variable enables selective display. This means that lines, or portions of lines, may be made hidden.

  • If the value of selective-display is t, then the character control-m marks the start of hidden text; the control-m, and the rest of the line following it, are not displayed. This is explicit selective display.
  • If the value of selective-display is a positive integer, then lines that start with more than that many columns of indentation are not displayed.

When some portion of a buffer is hidden, the vertical movement commands operate as if that portion did not exist, allowing a single next-line command to skip any number of hidden lines. However, character movement commands (such as forward-char) do not skip the hidden portion, and it is possible (if tricky) to insert or delete text in a hidden portion.

In the examples below, we show the display appearance of the buffer foo, which changes with the value of selective-display. The contents of the buffer do not change.

(setq selective-display nil)
     ⇒ nil

---------- Buffer: foo ----------
1 on this column
 2on this column
  3n this column
  3n this column
 2on this column
1 on this column
---------- Buffer: foo ----------

(setq selective-display 2)
     ⇒ 2

---------- Buffer: foo ----------
1 on this column
 2on this column
 2on this column
1 on this column
---------- Buffer: foo ----------
User Option: selective-display-ellipses

If this buffer-local variable is non-nil, then Emacs displays ‘’ at the end of a line that is followed by hidden text. This example is a continuation of the previous one.

(setq selective-display-ellipses t)
     ⇒ t

---------- Buffer: foo ----------
1 on this column
 2on this column ...
 2on this column
1 on this column
---------- Buffer: foo ----------

You can use a display table to substitute other text for the ellipsis (‘’). See Display Tables.


41.8 Temporary Displays

Temporary displays are used by Lisp programs to put output into a buffer and then present it to the user for perusal rather than for editing. Many help commands use this feature.

Macro: with-output-to-temp-buffer buffer-name body…

This function executes the forms in body while arranging to insert any output they print into the buffer named buffer-name, which is first created if necessary, and put into Help mode. (See the similar form with-temp-buffer-window below.) Finally, the buffer is displayed in some window, but that window is not selected.

If the forms in body do not change the major mode in the output buffer, so that it is still Help mode at the end of their execution, then with-output-to-temp-buffer makes this buffer read-only at the end, and also scans it for function and variable names to make them into clickable cross-references. See Tips for Documentation Strings, in particular the item on hyperlinks in documentation strings, for more details.

The string buffer-name specifies the temporary buffer, which need not already exist. The argument must be a string, not a buffer. The buffer is erased initially (with no questions asked), and it is marked as unmodified after with-output-to-temp-buffer exits.

with-output-to-temp-buffer binds standard-output to the temporary buffer, then it evaluates the forms in body. Output using the Lisp output functions within body goes by default to that buffer (but screen display and messages in the echo area, although they are “output” in the general sense of the word, are not affected). See Output Functions.

Several hooks are available for customizing the behavior of this construct; they are listed below.

The value of the last form in body is returned.

---------- Buffer: foo ----------
 This is the contents of foo.
---------- Buffer: foo ----------

(with-output-to-temp-buffer "foo"
    (print 20)
    (print standard-output))
⇒ #<buffer foo>

---------- Buffer: foo ----------

20

#<buffer foo>

---------- Buffer: foo ----------
User Option: temp-buffer-show-function

If this variable is non-nil, with-output-to-temp-buffer calls it as a function to do the job of displaying a help buffer. The function gets one argument, which is the buffer it should display.

It is a good idea for this function to run temp-buffer-show-hook just as with-output-to-temp-buffer normally would, inside of save-selected-window and with the chosen window and buffer selected.

Variable: temp-buffer-setup-hook

This normal hook is run by with-output-to-temp-buffer before evaluating body. When the hook runs, the temporary buffer is current. This hook is normally set up with a function to put the buffer in Help mode.

Variable: temp-buffer-show-hook

This normal hook is run by with-output-to-temp-buffer after displaying the temporary buffer. When the hook runs, the temporary buffer is current, and the window it was displayed in is selected.

Macro: with-temp-buffer-window buffer-or-name action quit-function body…

This macro is similar to with-output-to-temp-buffer. Like that construct, it executes body while arranging to insert any output it prints into the buffer named buffer-or-name and displays that buffer in some window. Unlike with-output-to-temp-buffer, however, it does not automatically switch that buffer to Help mode.

The argument buffer-or-name specifies the temporary buffer. It can be either a buffer, which must already exist, or a string, in which case a buffer of that name is created, if necessary. The buffer is marked as unmodified and read-only when with-temp-buffer-window exits.

This macro does not call temp-buffer-show-function. Rather, it passes the action argument to display-buffer (see Choosing a Window for Displaying a Buffer) in order to display the buffer.

The value of the last form in body is returned, unless the argument quit-function is specified. In that case, it is called with two arguments: the window showing the buffer and the result of body. The final return value is then whatever quit-function returns.

This macro uses the normal hooks temp-buffer-window-setup-hook and temp-buffer-window-show-hook in place of the analogous hooks run by with-output-to-temp-buffer.

The two constructs described next are mostly identical to with-temp-buffer-window but differ from it as specified:

Macro: with-current-buffer-window buffer-or-name action quit-function &rest body

This macro is like with-temp-buffer-window but unlike that makes the buffer specified by buffer-or-name current for running body.

A window showing a temporary buffer can be fitted to the size of that buffer using the following mode:

User Option: temp-buffer-resize-mode

When this minor mode is enabled, windows showing a temporary buffer are automatically resized to fit their buffer’s contents.

A window is resized if and only if it has been specially created for the buffer. In particular, windows that have shown another buffer before are not resized. By default, this mode uses fit-window-to-buffer (see Resizing Windows) for resizing. You can specify a different function by customizing the options temp-buffer-max-height and temp-buffer-max-width below.

The effect of this option can be overridden by providing a suitable window-height, window-width or window-size action alist entry for display-buffer (see Action Alists for Buffer Display).

User Option: temp-buffer-max-height

This option specifies the maximum height (in lines) of a window displaying a temporary buffer when temp-buffer-resize-mode is enabled. It can also be a function to be called to choose the height for such a buffer. It gets one argument, the buffer, and should return a positive integer. At the time the function is called, the window to be resized is selected.

User Option: temp-buffer-max-width

This option specifies the maximum width of a window (in columns) displaying a temporary buffer when temp-buffer-resize-mode is enabled. It can also be a function to be called to choose the width for such a buffer. It gets one argument, the buffer, and should return a positive integer. At the time the function is called, the window to be resized is selected.

The following function uses the current buffer for temporary display:

Function: momentary-string-display string position &optional char message

This function momentarily displays string in the current buffer at position. It has no effect on the undo list or on the buffer’s modification status.

The momentary display remains until the next input event. If the next input event is char, momentary-string-display ignores it and returns. Otherwise, that event remains buffered for subsequent use as input. Thus, typing char will simply remove the string from the display, while typing (say) C-f will remove the string from the display and later (presumably) move point forward. The argument char is a space by default.

The return value of momentary-string-display is not meaningful.

If the string string does not contain control characters, you can do the same job in a more general way by creating (and then subsequently deleting) an overlay with a before-string property. See Overlay Properties.

If message is non-nil, it is displayed in the echo area while string is displayed in the buffer. If it is nil, a default message says to type char to continue.

In this example, point is initially located at the beginning of the second line:

---------- Buffer: foo ----------
This is the contents of foo.
∗Second line.
---------- Buffer: foo ----------

(momentary-string-display
  "**** Important Message! ****"
  (point) ?\r
  "Type RET when done reading")
⇒ t

---------- Buffer: foo ----------
This is the contents of foo.
**** Important Message! ****Second line.
---------- Buffer: foo ----------

---------- Echo Area ----------
Type RET when done reading
---------- Echo Area ----------

41.9 Overlays

You can use overlays to alter the appearance of a buffer’s text on the screen, for the sake of presentation features. An overlay is an object that belongs to a particular buffer, and has a specified beginning and end. It also has properties that you can examine and set; these affect the display of the text within the overlaid portion of the buffer.

Editing the text of the buffer adjusts the beginning and end of each overlay so that it stays with the text. When you create the overlay, you can specify whether text inserted at the beginning should be inside the overlay or outside, and likewise for the end of the overlay.


41.9.1 Managing Overlays

This section describes the functions to create, delete and move overlays, and to examine their contents. Overlay changes are not recorded in the buffer’s undo list, since the overlays are not considered part of the buffer’s contents.

Function: overlayp object

This function returns t if object is an overlay.

Function: make-overlay start end &optional buffer front-advance rear-advance

This function creates and returns an overlay that belongs to buffer and ranges from start to end. Both start and end must specify buffer positions; they may be integers or markers. If buffer is omitted, the overlay is created in the current buffer.

An overlay whose start and end specify the same buffer position is known as empty. A non-empty overlay can become empty if the text between its start and end is deleted. When that happens, the overlay is by default not deleted, but you can cause it to be deleted by giving it the ‘evaporate’ property (see evaporate property).

The arguments front-advance and rear-advance specify what happens when text is inserted at the beginning (i.e., before start) and at the end. If they are both nil, the default, then the overlay extends to include any text inserted at the beginning, but not text inserted at the end. If front-advance is non-nil, text inserted at the beginning of the overlay is excluded from the overlay. If rear-advance is non-nil, text inserted at the end of the overlay is included in the overlay.

Function: overlay-start overlay

This function returns the position at which overlay starts, as an integer.

Function: overlay-end overlay

This function returns the position at which overlay ends, as an integer.

Function: overlay-buffer overlay

This function returns the buffer that overlay belongs to. It returns nil if overlay has been deleted.

Function: delete-overlay overlay

This function deletes the specified overlay. The overlay continues to exist as a Lisp object, and its property list is unchanged, but it ceases to be attached to the buffer it belonged to, and ceases to have any effect on display.

A deleted overlay is not permanently disconnected. You can give it a position in a buffer again by calling move-overlay.

Function: move-overlay overlay start end &optional buffer

This function moves overlay to buffer, and places its bounds at start and end in that buffer. Both arguments start and end must specify buffer positions; they may be integers or markers.

If buffer is omitted, overlay stays in the same buffer it was already associated with; if overlay was previously deleted (and thus isn’t associated with any buffer), it goes into the current buffer.

The return value is overlay.

This function is the only valid way to change the endpoints of an overlay.

Function: remove-overlays &optional start end name value

This function removes all the overlays between start and end whose property name has the specified value. It can move the endpoints of the overlays in the region, or split them.

If name is omitted or nil, it means to delete all overlays in the specified region. If start and/or end are omitted or nil, that means the beginning and end of the buffer respectively. Therefore, (remove-overlays) removes all the overlays in the current buffer.

Function: copy-overlay overlay

This function returns a copy of overlay. The copy has the same endpoints and properties as overlay. However, the text insertion type for the start of the overlay and for the end of the overlay are set to their default values.

Here are some examples:

;; Create an overlay.
(setq foo (make-overlay 1 10))
     ⇒ #<overlay from 1 to 10 in display.texi>
(overlay-start foo)
     ⇒ 1
(overlay-end foo)
     ⇒ 10
(overlay-buffer foo)
     ⇒ #<buffer display.texi>
;; Give it a property we can check later.
(overlay-put foo 'happy t)
     ⇒ t
;; Verify the property is present.
(overlay-get foo 'happy)
     ⇒ t
;; Move the overlay.
(move-overlay foo 5 20)
     ⇒ #<overlay from 5 to 20 in display.texi>
(overlay-start foo)
     ⇒ 5
(overlay-end foo)
     ⇒ 20
;; Delete the overlay.
(delete-overlay foo)
     ⇒ nil
;; Verify it is deleted.
foo
     ⇒ #<overlay in no buffer>
;; A deleted overlay has no position.
(overlay-start foo)
     ⇒ nil
(overlay-end foo)
     ⇒ nil
(overlay-buffer foo)
     ⇒ nil
;; Undelete the overlay.
(move-overlay foo 1 20)
     ⇒ #<overlay from 1 to 20 in display.texi>
;; Verify the results.
(overlay-start foo)
     ⇒ 1
(overlay-end foo)
     ⇒ 20
(overlay-buffer foo)
     ⇒ #<buffer display.texi>
;; Moving and deleting the overlay does not change its properties.
(overlay-get foo 'happy)
     ⇒ t

41.9.2 Overlay Properties

Overlay properties are like text properties in that the properties that alter how a character is displayed can come from either source. But in most respects they are different. See Text Properties, for comparison.

Text properties are considered a part of the text; overlays and their properties are specifically considered not to be part of the text. Thus, copying text between various buffers and strings preserves text properties, but does not try to preserve overlays. Changing a buffer’s text properties marks the buffer as modified, while moving an overlay or changing its properties does not. Unlike text property changes, overlay property changes are not recorded in the buffer’s undo list.

Since more than one overlay can specify a property value for the same character, Emacs lets you specify a priority value of each overlay. The priority value is used to decide which of the overlapping overlays will “win”.

These functions read and set the properties of an overlay:

Function: overlay-get overlay prop

This function returns the value of property prop recorded in overlay, if any. If overlay does not record any value for that property, but it does have a category property which is a symbol, that symbol’s prop property is used. Otherwise, the value is nil.

Function: overlay-put overlay prop value

This function sets the value of property prop recorded in overlay to value. It returns value.

Function: overlay-properties overlay

This returns a copy of the property list of overlay.

See also the function get-char-property which checks both overlay properties and text properties for a given character. See Examining Text Properties.

Many overlay properties have special meanings; here is a table of them:

priority

This property’s value determines the priority of the overlay. If you want to specify a priority value, use either nil (or zero), or a positive integer, or a cons of two values. Any other value triggers undefined behavior.

The priority matters when two or more overlays cover the same character and both specify the same property with different values; the one whose priority value is higher overrides the other. (For the face property, the higher priority overlay’s value does not completely override the other value; instead, its individual face attributes override the corresponding face attributes of the face property whose priority is lower.) If two overlays have the same priority value, and one is “nested” in the other (i.e., covers fewer buffer or string positions), then the inner one will prevail over the outer one. If neither is nested in the other then you should not make assumptions about which overlay will prevail.

When a Lisp program puts overlays with defined priorities on text that might have overlays without priorities, this could cause undesirable results, because any overlay with a positive priority value will override all the overlays without a priority. Since most Emacs features that use overlays don’t specify priorities for their overlays, integer priorities should be used with care. Instead of using integer priorities and risk overriding other overlays, you can use priority values of the form (primary . secondary), where the primary value is used as described above, and secondary is the fallback value used when primary and the nesting considerations fail to resolve the precedence between overlays. In particular, priority value (nil . n), with n a positive integer, allows to have the overlays ordered by priority when necessary without completely overriding other overlays.

Currently, all overlays take priority over text properties.

If you need to put overlays in priority order, use the sorted argument of overlays-at. See Searching for Overlays.

window

If the window property is non-nil, then the overlay applies only on that window.

category

If an overlay has a category property, we call it the category of the overlay. It should be a symbol. The properties of the symbol serve as defaults for the properties of the overlay.

face

This property controls the appearance of the text (see Faces). The value of the property can be the following:

  • A face name (a symbol or string).
  • An anonymous face: a property list of the form (keyword value …), where each keyword is a face attribute name and value is a value for that attribute.
  • A list of faces. Each list element should be either a face name or an anonymous face. This specifies a face which is an aggregate of the attributes of each of the listed faces. Faces occurring earlier in the list have higher priority.
  • A cons cell of the form (foreground-color . color-name) or (background-color . color-name). This specifies the foreground or background color, similar to (:foreground color-name) or (:background color-name). This form is supported for backward compatibility only, and should be avoided.
mouse-face

This property is used instead of face when the mouse is within the range of the overlay. However, Emacs ignores all face attributes from this property that alter the text size (e.g., :height, :weight, and :slant); those attributes are always the same as in the unhighlighted text.

display

This property activates various features that change the way text is displayed. For example, it can make text appear taller or shorter, higher or lower, wider or narrower, or replaced with an image. See The display Property.

help-echo

If an overlay has a help-echo property, then when you move the mouse onto the text in the overlay, Emacs displays a help string in the echo area, or as a tooltip. For details see Text help-echo.

field

Consecutive characters with the same field property constitute a field. Some motion functions including forward-word and beginning-of-line stop moving at a field boundary. See Defining and Using Fields.

modification-hooks

This property’s value is a list of functions to be called if any character within the overlay is changed or if text is inserted strictly within the overlay.

The hook functions are called both before and after each change. If the functions save the information they receive, and compare notes between calls, they can determine exactly what change has been made in the buffer text.

When called before a change, each function receives four arguments: the overlay, nil, and the beginning and end of the text range to be modified.

When called after a change, each function receives five arguments: the overlay, t, the beginning and end of the text range just modified, and the length of the pre-change text replaced by that range. (For an insertion, the pre-change length is zero; for a deletion, that length is the number of characters deleted, and the post-change beginning and end are equal.)

When these functions are called, inhibit-modification-hooks is bound to non-nil. If the functions modify the buffer, you might want to bind inhibit-modification-hooks to nil, so as to cause the change hooks to run for these modifications. However, doing this may call your own change hook recursively, so be sure to prepare for that. See Change Hooks.

Text properties also support the modification-hooks property, but the details are somewhat different (see Properties with Special Meanings).

insert-in-front-hooks

This property’s value is a list of functions to be called before and after inserting text right at the beginning of the overlay. The calling conventions are the same as for the modification-hooks functions.

insert-behind-hooks

This property’s value is a list of functions to be called before and after inserting text right at the end of the overlay. The calling conventions are the same as for the modification-hooks functions.

invisible

The invisible property can make the text in the overlay invisible, which means that it does not appear on the screen. See Invisible Text, for details.

intangible

The intangible property on an overlay works just like the intangible text property. It is obsolete. See Properties with Special Meanings, for details.

isearch-open-invisible

This property tells incremental search (see Incremental Search in The GNU Emacs Manual) how to make an invisible overlay visible, permanently, if the final match overlaps it. See Invisible Text.

isearch-open-invisible-temporary

This property tells incremental search how to make an invisible overlay visible, temporarily, during the search. See Invisible Text.

before-string

This property’s value is a string to add to the display at the beginning of the overlay. The string does not appear in the buffer in any sense—only on the screen. Note that if the text at the beginning of the overlay is made invisible, the string will not be displayed.

after-string

This property’s value is a string to add to the display at the end of the overlay. The string does not appear in the buffer in any sense—only on the screen. Note that if the text at the end of the overlay is made invisible, the string will not be displayed.

line-prefix

This property specifies a display spec to prepend to each non-continuation line at display-time. See Truncation.

wrap-prefix

This property specifies a display spec to prepend to each continuation line at display-time. See Truncation.

evaporate

If this property is non-nil, the overlay is deleted automatically if it becomes empty (i.e., if its length becomes zero). If you give an empty overlay (see empty overlay) a non-nil evaporate property, that deletes it immediately. Note that, unless an overlay has this property, it will not be deleted when the text between its starting and ending positions is deleted from the buffer.

keymap

If this property is non-nil, it specifies a keymap for a portion of the text. This keymap takes precedence over most other keymaps (see Active Keymaps), and it is used when point is within the overlay, where the front- and rear-advance properties define whether the boundaries are considered as being within or not.

local-map

The local-map property is similar to keymap but replaces the buffer’s local map rather than augmenting existing keymaps. This also means it has lower precedence than minor mode keymaps.

The keymap and local-map properties do not affect a string displayed by the before-string, after-string, or display properties. This is only relevant for mouse clicks and other mouse events that fall on the string, since point is never on the string. To bind special mouse events for the string, assign it a keymap or local-map text property. See Properties with Special Meanings.


41.9.3 Searching for Overlays

Function: overlays-at pos &optional sorted

This function returns a list of all the overlays that cover the character at position pos in the current buffer. If sorted is non-nil, the list is in decreasing order of priority, otherwise it is in no particular order. An overlay contains position pos if it begins at or before pos, and ends after pos.

To illustrate usage, here is a Lisp function that returns a list of the overlays that specify property prop for the character at point:

(defun find-overlays-specifying (prop)
  (let ((overlays (overlays-at (point)))
        found)
    (while overlays
      (let ((overlay (car overlays)))
        (if (overlay-get overlay prop)
            (setq found (cons overlay found))))
      (setq overlays (cdr overlays)))
    found))
Function: overlays-in beg end

This function returns a list of the overlays that overlap the region beg through end. An overlay overlaps with a region if it contains one or more characters in the region; empty overlays (see empty overlay) overlap if they are at beg, strictly between beg and end, or at end when end denotes the position at the end of the accessible part of the buffer.

Function: next-overlay-change pos

This function returns the buffer position of the next beginning or end of an overlay, after pos. If there is none, it returns (point-max).

Function: previous-overlay-change pos

This function returns the buffer position of the previous beginning or end of an overlay, before pos. If there is none, it returns (point-min).

As an example, here’s a simplified (and inefficient) version of the primitive function next-single-char-property-change (see Text Property Search Functions). It searches forward from position pos for the next position where the value of a given property prop, as obtained from either overlays or text properties, changes.

(defun next-single-char-property-change (position prop)
  (save-excursion
    (goto-char position)
    (let ((propval (get-char-property (point) prop)))
      (while (and (not (eobp))
                  (eq (get-char-property (point) prop) propval))
        (goto-char (min (next-overlay-change (point))
                        (next-single-property-change (point) prop)))))
    (point)))

41.10 Size of Displayed Text

Since not all characters have the same width, these functions let you check the width of a character. See Indentation Primitives, and Motion by Screen Lines, for related functions.

Function: char-width char

This function returns the width in columns of the character char, if it were displayed in the current buffer (i.e., taking into account the buffer’s display table, if any; see Display Tables). The width of a tab character is usually tab-width (see Usual Display Conventions).

Function: char-uppercase-p char

Return non-nil if char is an uppercase character according to Unicode.

Function: string-width string &optional from to

This function returns the width in columns of the string string, if it were displayed in the current buffer and the selected window. Optional arguments from and to specify the substring of string to consider, and are interpreted as in substring (see Creating Strings).

The return value is an approximation: it only considers the values returned by char-width for the constituent characters, always takes a tab character as taking tab-width columns, ignores display properties and fonts, etc. For these reasons, we recommend using window-text-pixel-size or string-pixel-width, described below, instead.

Function: truncate-string-to-width string width &optional start-column padding ellipsis ellipsis-text-property

This function returns a new string that is a truncation of string which fits within width columns on display.

If string is narrower than width, the result is equal to string; otherwise excess characters are omitted from the result. If a multi-column character in string exceeds the goal width, that character is omitted from the result. Thus, the result can sometimes fall short of width, but cannot go beyond it.

The optional argument start-column specifies the starting column; it defaults to zero. If this is non-nil, then the first start-column columns of the string are omitted from the result. If one multi-column character in string extends across the column start-column, that character is omitted.

The optional argument padding, if non-nil, is a padding character added at the beginning and end of the result string, to extend it to exactly width columns. The padding character is appended at the end of the result if it falls short of width, as many times as needed to reach width. It is also prepended at the beginning of the result if a multi-column character in string extends across the column start-column.

If ellipsis is non-nil, it should be a string which will replace the end of string when it is truncated. In this case, more characters will be removed from string to free enough space for ellipsis to fit within width columns. However, if the display width of string is less than the display width of ellipsis, ellipsis will not be appended to the result. If ellipsis is non-nil and not a string, it stands for the value returned by the function truncate-string-ellipsis, described below.

The optional argument ellipsis-text-property, if non-nil, means hide the excess parts of string with a display text property (see The display Property) showing the ellipsis, instead of actually truncating the string.

(truncate-string-to-width "\tab\t" 12 4)
     ⇒ "ab"
(truncate-string-to-width "\tab\t" 12 4 ?\s)
     ⇒ "    ab  "

This function uses string-width and char-width to find the suitable truncation point when string is too wide, so it suffers from the same basic issues as string-width does. In particular, when character composition happens within string, the display width of a string could be smaller than the sum of widths of the constituent characters, and this function might return inaccurate results.

Function: truncate-string-ellipsis

This function returns the string to be used as an ellipses in truncate-string-to-width and other similar contexts. The value is that of the variable truncate-string-ellipsis, if it’s non-nil, the string with the single character U+2026 HORIZONTAL ELLIPSIS if that character can be displayed on the selected frame, and the string ‘...’ otherwise.

The following function returns the size in pixels of text as if it were displayed in a given window. This function is used by fit-window-to-buffer and fit-frame-to-buffer (see Resizing Windows) to make a window exactly as large as the text it contains.

Function: window-text-pixel-size &optional window from to x-limit y-limit mode-lines ignore-line-at-end

This function returns the size of the text of window’s buffer in pixels. window must be a live window and defaults to the selected one. The return value is a cons of the maximum pixel-width of any text line and the maximum pixel-height of all text lines. This function exists to allow Lisp programs to adjust the dimensions of window to the buffer text it needs to display, and for other similar situations.

The return value can also optionally (see below) include the buffer position of the first line whose dimensions were measured.

The optional argument from, if non-nil, specifies the first text position to consider, and defaults to the minimum accessible position of the buffer. If from is t, it stands for the minimum accessible position that is not a newline character. If from is a cons, its car specifies a buffer position, and its cdr specifies the vertical offset in pixels from that position to the first screen line whose text is to be measured. (The measurement will start from the visual beginning of that screen line.) In that case, the return value will instead be a list of the pixel-width, pixel-height, and the buffer position of the first line that was measured. The optional argument to, if non-nil, specifies the last text position to consider, and defaults to the maximum accessible position of the buffer. If to is t, it stands for the maximum accessible position that is not a newline character.

The optional argument x-limit, if non-nil, specifies the maximum X coordinate beyond which text should be ignored; it is therefore also the largest value of pixel-width that the function can return. If x-limit nil or omitted, it means to use the pixel-width of window’s body (see Window Sizes); this default means that text of truncated lines wider than the window will be ignored. This default is useful when the caller does not intend to change the width of window. Otherwise, the caller should specify here the maximum width window’s body may assume; in particular, if truncated lines are expected and their text needs to be accounted for, x-limit should be set to a large value. Since calculating the width of long lines can take some time, it’s always a good idea to make this argument as small as needed; in particular, if the buffer might contain long lines that will be truncated anyway.

The optional argument y-limit, if non-nil, specifies the maximum Y coordinate beyond which text is to be ignored; it is therefore also the maximum pixel-height that the function can return. If y-limit is nil or omitted, it means to consider all the lines of text till the buffer position specified by to. Since calculating the pixel-height of a large buffer can take some time, it makes sense to specify this argument; in particular, if the caller does not know the size of the buffer.

The optional argument mode-lines nil or omitted means to not include the height of the mode-, tab- or header-line of window in the return value. If it is either the symbol mode-line, tab-line or header-line, include only the height of that line, if present, in the return value. If it is t, include the height of all of these lines, if present, in the return value.

The optional argument ignore-line-at-end controls whether or not to count the height of text in to’s screen line as part of the returned pixel-height. This is useful if your Lisp program is only interested in the dimensions of text up to and excluding the visual beginning of to’s screen line.

window-text-pixel-size treats the text displayed in a window as a whole and does not care about the size of individual lines. The following function does.

Function: window-lines-pixel-dimensions &optional window first last body inverse left

This function calculates the pixel dimensions of each line displayed in the specified window. It does so by walking window’s current glyph matrix—a matrix storing the glyph (see Glyphs) of each buffer character currently displayed in window. If successful, it returns a list of cons pairs representing the x- and y-coordinates of the lower right corner of the last character of each line. Coordinates are measured in pixels from an origin (0, 0) at the top-left corner of window. window must be a live window and defaults to the selected one.

If the optional argument first is an integer, it denotes the index (starting with 0) of the first line of window’s glyph matrix to be returned. Note that if window has a header line, the line with index 0 is that header line. If first is nil, the first line to be considered is determined by the value of the optional argument body: If body is non-nil, this means to start with the first line of window’s body, skipping any header line, if present. Otherwise, this function will start with the first line of window’s glyph matrix, possibly the header line.

If the optional argument last is an integer, it denotes the index of the last line of window’s glyph matrix that shall be returned. If last is nil, the last line to be considered is determined by the value of body: If body is non-nil, this means to use the last line of window’s body, omitting window’s mode line, if present. Otherwise, this means to use the last line of window which may be the mode line.

The optional argument inverse, if nil, means that the y-pixel value returned for any line specifies the distance in pixels from the left edge (body edge if body is non-nil) of window to the right edge of the last glyph of that line. inverse non-nil means that the y-pixel value returned for any line specifies the distance in pixels from the right edge of the last glyph of that line to the right edge (body edge if body is non-nil) of window. This is useful for determining the amount of slack space at the end of each line.

The optional argument left, if non-nil means to return the x- and y-coordinates of the lower left corner of the leftmost character on each line. This is the value that should be used for windows that mostly display text from right to left.

If left is non-nil and inverse is nil, this means that the y-pixel value returned for any line specifies the distance in pixels from the left edge of the last (leftmost) glyph of that line to the right edge (body edge if body is non-nil) of window. If left and inverse are both non-nil, the y-pixel value returned for any line specifies the distance in pixels from the left edge (body edge if body is non-nil) of window to the left edge of the last (leftmost) glyph of that line.

This function returns nil if the current glyph matrix of window is not up-to-date which usually happens when Emacs is busy, for example, when processing a command. The value should be retrievable though when this function is run from an idle timer with a delay of zero seconds.

Function: buffer-text-pixel-size &optional buffer-or-name window from to x-limit y-limit

This is much like window-text-pixel-size, but can be used when the buffer isn’t shown in a window. (window-text-pixel-size is faster when it is, so this function shouldn’t be used in that case.)

buffer-or-name must specify a live buffer or the name of a live buffer and defaults to the current buffer. window must be a live window and defaults to the selected one; the function will compute the text dimensions as if buffer is displayed in window. The return value is a cons of the maximum pixel-width of any text line and the pixel-height of all the text lines of the buffer specified by buffer-or-name.

The optional arguments x-limit and y-limit have the same meaning as with window-text-pixel-size.

Function: string-pixel-width string

This is a convenience function that uses window-text-pixel-size to compute the width of string (in pixels).

Function: line-pixel-height

This function returns the height in pixels of the line at point in the selected window. The value includes the line spacing of the line (see Line Height).

Function: string-glyph-split string

When character compositions are in effect, sequence of characters can be composed for display to form grapheme clusters, for example to display accented characters, or ligatures, or Emoji, or when complex text shaping requires that for some scripts. When that happens, characters no longer map in a simple way to display columns, and display layout decisions with such strings, such as truncating too wide strings, can be a complex job. This function helps in performing such jobs: it splits up its argument string into a list of substrings, where each substring produces a single grapheme cluster that should be displayed as a unit. Lisp programs can then use this list to construct visually-valid substrings of string which will look correctly on display, or compute the width of any substring of string by adding the width of its constituents in the returned list, etc.

For instance, if you want to display a string without the first glyph, you can say:

(apply #'insert (cdr (string-glyph-split string))))

When a buffer is displayed with line numbers (see Display Custom in The GNU Emacs Manual), it is sometimes useful to know the width taken for displaying the line numbers. The following function is for Lisp programs which need this information for layout calculations.

Function: line-number-display-width &optional pixelwise

This function returns the width used for displaying the line numbers in the selected window. If the optional argument pixelwise is the symbol columns, the return value is a float number of the frame’s canonical columns; if pixelwise is t or any other non-nil value, the value is an integer and is measured in pixels. If pixelwise is omitted or nil, the value is the integer number of columns of the font defined for the line-number face, and doesn’t include the 2 columns used to pad the numbers on display. If line numbers are not displayed in the selected window, the value is zero regardless of the value of pixelwise. Use with-selected-window (see Selecting Windows) if you need this information about another window.


41.11 Line Height

The total height of each display line consists of the height of the contents of the line, plus optional additional vertical line spacing above or below the display line.

The height of the line contents is the maximum height of any character or image on that display line, including the final newline if there is one. (A display line that is continued doesn’t include a final newline.) That is the default line height, if you do nothing to specify a greater height. (In the most common case, this equals the height of the corresponding frame’s default font, see Frame Font.)

There are several ways to explicitly specify a larger line height, either by specifying an absolute height for the display line, or by specifying vertical space. However, no matter what you specify, the actual line height can never be less than the default.

A newline can have a line-height text or overlay property that controls the total height of the display line ending in that newline. The property value can be one of several forms:

t

If the property value is t, the newline character has no effect on the displayed height of the line—the visible contents alone determine the height. The line-spacing property of the newline, described below, is also ignored in this case. This is useful for tiling small images (or image slices) without adding blank areas between the images.

(height total)

If the property value is a list of the form shown, that adds extra space below the display line. First Emacs uses height as a height spec to control extra space above the line; then it adds enough space below the line to bring the total line height up to total. In this case, any value of line-spacing property for the newline is ignored.

Any other kind of property value is a height spec, which translates into a number—the specified line height. There are several ways to write a height spec; here’s how each of them translates into a number:

integer

If the height spec is a positive integer, the height value is that integer.

float

If the height spec is a float, float, the numeric height value is float times the frame’s default line height.

(face . ratio)

If the height spec is a cons of the format shown, the numeric height is ratio times the height of face face. ratio can be any type of number, or nil which means a ratio of 1. If face is t, it refers to the current face.

(nil . ratio)

If the height spec is a cons of the format shown, the numeric height is ratio times the height of the contents of the line.

Thus, any valid height spec determines the height in pixels, one way or another. If the line contents’ height is less than that, Emacs adds extra vertical space above the line to achieve the specified total height.

If you don’t specify the line-height property, the line’s height consists of the contents’ height plus the line spacing. There are several ways to specify the line spacing for different parts of Emacs text.

On graphical terminals, you can specify the line spacing for all lines in a frame, using the line-spacing frame parameter (see Layout Parameters). However, if the default value of line-spacing is non-nil, it overrides the frame’s line-spacing parameter. An integer specifies the number of pixels put below lines. A floating-point number specifies the spacing relative to the frame’s default line height.

You can specify the line spacing for all lines in a buffer via the buffer-local line-spacing variable. An integer specifies the number of pixels put below lines. A floating-point number specifies the spacing relative to the default frame line height. This overrides line spacings specified for the frame.

Finally, a newline can have a line-spacing text or overlay property that can enlarge the default frame line spacing and the buffer local line-spacing variable: if its value is larger than the buffer or frame defaults, that larger value is used instead, for the display line ending in that newline (unless the newline also has the line-height property whose value is one of the special values which cause line-spacing to be ignored, see above).

One way or another, these mechanisms specify a Lisp value for the spacing of each line. The value is a height spec, and it translates into a Lisp value as described above. However, in this case the numeric height value specifies the line spacing, rather than the line height.

On text terminals, the line spacing cannot be altered.


41.12 Faces

A face is a collection of graphical attributes for displaying text: font, foreground color, background color, optional underlining, etc. Faces control how Emacs displays text in buffers, as well as other parts of the frame such as the mode line.

One way to represent a face is as a property list of attributes, like (:foreground "red" :weight bold). Such a list is called an anonymous face. For example, you can assign an anonymous face as the value of the face text property, and Emacs will display the underlying text with the specified attributes. See Properties with Special Meanings.

More commonly, a face is referred to via a face name: a Lisp symbol associated with a set of face attributes29. Named faces are defined using the defface macro (see Defining Faces). Emacs comes with several standard named faces (see Basic Faces).

Some parts of Emacs require named faces (e.g., the functions documented in Face Attribute Functions). Unless otherwise stated, we will use the term face to refer only to named faces.

Function: facep object

This function returns a non-nil value if object is a named face: a Lisp symbol or string which serves as a face name. Otherwise, it returns nil.


41.12.1 Face Attributes

Face attributes determine the visual appearance of a face. The following table lists all the face attributes, their possible values, and their effects.

Apart from the values given below, each face attribute can have the value unspecified. This special value means that the face doesn’t specify that attribute directly. An unspecified attribute tells Emacs to refer instead to a parent face (see the description :inherit attribute below); or, failing that, to an underlying face (see Displaying Faces). (However, unspecified is not a valid value in defface.)

A face attribute can also have the value reset. This special value stands for the value of the corresponding attribute of the default face.

The default face must explicitly specify all attributes, and cannot use the special value reset.

Some of these attributes are meaningful only on certain kinds of displays. If your display cannot handle a certain attribute, the attribute is ignored.

:family

Font family name (a string). See Fonts in The GNU Emacs Manual, for more information about font families. The function font-family-list (see below) returns a list of available family names.

:foundry

The name of the font foundry for the font family specified by the :family attribute (a string). See Fonts in The GNU Emacs Manual.

:width

Relative character width. This should be one of the symbols ultra-condensed, extra-condensed, condensed, semi-condensed, normal, regular, medium, semi-expanded, expanded, extra-expanded, or ultra-expanded.

:height

The height of the font. In the simplest case, this is an integer in units of 1/10 point.

The value can also be floating point or a function, which specifies the height relative to an underlying face (see Displaying Faces). A floating-point value specifies the amount by which to scale the height of the underlying face. A function value is called with one argument, the height of the underlying face, and returns the height of the new face. If the function is passed an integer argument, it must return an integer.

The height of the default face must be specified using an integer; floating point and function values are not allowed.

:weight

Font weight—one of the symbols (from densest to faintest) ultra-bold, extra-bold, bold, semi-bold, normal, semi-light, light, extra-light, or ultra-light. On text terminals which support variable-brightness text, any weight greater than normal is displayed as extra bright, and any weight less than normal is displayed as half-bright.

:slant

Font slant—one of the symbols italic, oblique, normal, reverse-italic, or reverse-oblique. On text terminals that support variable-brightness text, slanted text is displayed as half-bright.

:foreground

Foreground color, a string. The value can be a system-defined color name, or a hexadecimal color specification. See Color Names. On black-and-white displays, certain shades of gray are implemented by stipple patterns.

:distant-foreground

Alternative foreground color, a string. This is like :foreground but the color is only used as a foreground when the background color is near to the foreground that would have been used. This is useful for example when marking text (i.e., the region face). If the text has a foreground that is visible with the region face, that foreground is used. If the foreground is near the region face background, :distant-foreground is used instead so the text is readable.

:background

Background color, a string. The value can be a system-defined color name, or a hexadecimal color specification. See Color Names.

:underline

Whether or not characters should be underlined, and in what way. The possible values of the :underline attribute are:

nil

Don’t underline.

t

Underline with the foreground color of the face.

color

Underline in color color, a string specifying a color.

(:color color :style style :position position)

color is either a string, or the symbol foreground-color, meaning the foreground color of the face. Omitting the attribute :color means to use the foreground color of the face. style should be a symbol line or wave, meaning to use a straight or wavy line. Omitting the attribute :style means to use a straight line. position, if non-nil, means to display the underline at the descent of the text, instead of at the baseline level. If it is a number, then it specifies the amount of pixels above the descent to display the underline.

:overline

Whether or not characters should be overlined, and in what color. If the value is t, overlining uses the foreground color of the face. If the value is a string, overlining uses that color. The value nil means do not overline.

:strike-through

Whether or not characters should be strike-through, and in what color. The value is used like that of :overline.

:box

Whether or not a box should be drawn around characters, its color, the width of the box lines, and 3D appearance. Here are the possible values of the :box attribute, and what they mean:

nil

Don’t draw a box.

t

Draw a box with lines of width 1, in the foreground color.

color

Draw a box with lines of width 1, in color color.

(:line-width (vwidth . hwidth) :color color :style style)

You can explicitly specify all aspects of the box with a plist on this form. Any element in this plist can be omitted.

The values of vwidth and hwidth specify respectively the width of the vertical and horizontal lines to draw; they default to (1 . 1). A negative horizontal or vertical width −n means to draw a line of width n that occupies the space of the underlying text, thus avoiding any increase in the character height or width. For simplification the width could be specified with only a single number n instead of a list, such case is equivalent to ((abs n) . n).

The value of color specifies the color to draw with. The default is the background color of the face for 3D boxes and flat-button, and the foreground color of the face for other boxes.

The value of style specifies whether to draw a 3D box. If it is released-button, the box looks like a 3D button that is not being pressed. If it is pressed-button, the box looks like a 3D button that is being pressed. If it is nil, flat-button or omitted, a plain 2D box is used.

:inverse-video

Whether or not characters should be displayed in inverse video. The value should be t (yes) or nil (no).

:stipple

The background stipple, a bitmap.

The value can be a string; that should be the name of a file containing external-format X bitmap data. The file is found in the directories listed in the variable x-bitmap-file-path.

Alternatively, the value can specify the bitmap directly, with a list of the form (width height data). Here, width and height specify the size in pixels, and data is a string containing the raw bits of the bitmap, row by row. Each row occupies (width + 7) / 8 consecutive bytes in the string (which should be a unibyte string for best results). This means that each row always occupies at least one whole byte.

If the value is nil, that means use no stipple pattern.

Normally you do not need to set the stipple attribute, because it is used automatically to handle certain shades of gray.

:font

The font used to display the face. Its value should be a font object or a fontset. If it is a font object, it specifies the font to be used by the face for displaying ASCII characters. See Low-Level Font Representation, for information about font objects, font specs, and font entities. See Fontsets, for information about fontsets.

When specifying this attribute using set-face-attribute or set-face-font (see Face Attribute Functions), you may also supply a font spec, a font entity, or a string. Emacs converts such values to an appropriate font object, and stores that font object as the actual attribute value. If you specify a string, the contents of the string should be a font name (see Fonts in The GNU Emacs Manual); if the font name is an XLFD containing wildcards, Emacs chooses the first font matching those wildcards. Specifying this attribute also changes the values of the :family, :foundry, :width, :height, :weight, and :slant attributes.

:inherit

The name of a face from which to inherit attributes, or a list of face names. Attributes from inherited faces are merged into the face like an underlying face would be, with higher priority than underlying faces (see Displaying Faces). If the face to inherit from is unspecified, it is treated the same as nil, since Emacs never merges :inherit attributes. If a list of faces is used, attributes from faces earlier in the list override those from later faces.

:extend

Whether or not this face will be extended beyond end of line and will affect the display of the empty space between the end of line and the edge of the window. The value should be t to display the empty space between end of line and edge of the window using this face, or nil to not use this face for the space between the end of the line and the edge of the window. When Emacs merges several faces for displaying the empty space beyond end of line, only those faces with :extend non-nil will be merged. By default, only a small number of faces, notably, region, have this attribute set. This attribute is different from the others in that when a theme doesn’t specify an explicit value for a face, the value from the original face definition by defface is inherited (see Defining Faces).

Some modes, like hl-line-mode, use a face with an :extend property to mark the entire current line. Note, however, that Emacs will always allow you to move point after the final character in a buffer, and if the buffer ends with a newline character, point can be placed on what is seemingly a line at the end of the buffer—but Emacs can’t highlight that “line”, because it doesn’t really exist.

Function: font-family-list &optional frame

This function returns a list of available font family names. The optional argument frame specifies the frame on which the text is to be displayed; if it is nil, the selected frame is used.

User Option: underline-minimum-offset

This variable specifies the minimum distance between the baseline and the underline, in pixels, when displaying underlined text.

User Option: x-bitmap-file-path

This variable specifies a list of directories for searching for bitmap files, for the :stipple attribute.

Function: bitmap-spec-p object

This returns t if object is a valid bitmap specification, suitable for use with :stipple (see above). It returns nil otherwise.


41.12.2 Defining Faces

The usual way to define a face is through the defface macro. This macro associates a face name (a symbol) with a default face spec. A face spec is a construct which specifies what attributes a face should have on any given terminal; for example, a face spec might specify one foreground color on high-color terminals, and a different foreground color on low-color terminals.

People are sometimes tempted to create a variable whose value is a face name. In the vast majority of cases, this is not necessary; the usual procedure is to define a face with defface, and then use its name directly.

Note that once you have defined a face (usually with defface), you cannot later undefine this face safely, except by restarting Emacs.

Macro: defface face spec doc [keyword value]…

This macro declares face as a named face whose default face spec is given by spec. You should not quote the symbol face, and it should not end in ‘-face’ (that would be redundant). The argument doc is a documentation string for the face. The additional keyword arguments have the same meanings as in defgroup and defcustom (see Common Item Keywords).

If face already has a default face spec, this macro does nothing.

The default face spec determines face’s appearance when no customizations are in effect (see Customization Settings). If face has already been customized (via Custom themes or via customizations read from the init file), its appearance is determined by the custom face spec(s), which override the default face spec spec. However, if the customizations are subsequently removed, the appearance of face will again be determined by its default face spec.

As an exception, if you evaluate a defface form with C-M-x (eval-defun) or with C-x C-e (eval-last-sexp) in Emacs Lisp mode, a special feature of these commands overrides any custom face specs on the face, causing the face to reflect exactly what the defface says.

The spec argument is a face spec, which states how the face should appear on different kinds of terminals. It should be an alist whose elements each have the form

(display . plist)

display specifies a class of terminals (see below). plist is a property list of face attributes and their values, specifying how the face appears on such terminals. For backward compatibility, you can also write an element as (display plist).

The display part of an element of spec determines which terminals the element matches. If more than one element of spec matches a given terminal, the first element that matches is the one used for that terminal. There are three possibilities for display:

default

This element of spec doesn’t match any terminal; instead, it specifies defaults that apply to all terminals. This element, if used, must be the first element of spec. Each of the following elements can override any or all of these defaults.

t

This element of spec matches all terminals. Therefore, any subsequent elements of spec are never used. Normally t is used in the last (or only) element of spec.

a list

If display is a list, each element should have the form (characteristic value…). Here characteristic specifies a way of classifying terminals, and the values are possible classifications which display should apply to. Here are the possible values of characteristic:

type

The kind of window system the terminal uses—either graphic (any graphics-capable display), x, pc (for the MS-DOS console), w32 (for MS Windows 9X/NT/2K/XP), haiku (for Haiku), pgtk (for pure GTK), or tty (a non-graphics-capable display). See window-system.

class

What kinds of colors the terminal supports—either color, grayscale, or mono.

background

The kind of background—either light or dark.

min-colors

An integer that represents the minimum number of colors the terminal should support. This matches a terminal if its display-color-cells value is at least the specified integer.

supports

Whether or not the terminal can display the face attributes given in value… (see Face Attributes). See Display Face Attribute Testing, for more information on exactly how this testing is done.

If an element of display specifies more than one value for a given characteristic, any of those values is acceptable. If display has more than one element, each element should specify a different characteristic; then each characteristic of the terminal must match one of the values specified for it in display.

For example, here’s the definition of the standard face highlight:

(defface highlight
  '((((class color) (min-colors 88) (background light))
     :background "darkseagreen2")
    (((class color) (min-colors 88) (background dark))
     :background "darkolivegreen")
    (((class color) (min-colors 16) (background light))
     :background "darkseagreen2")
    (((class color) (min-colors 16) (background dark))
     :background "darkolivegreen")
    (((class color) (min-colors 8))
     :background "green" :foreground "black")
    (t :inverse-video t))
  "Basic face for highlighting."
  :group 'basic-faces)

Internally, Emacs stores each face’s default spec in its face-defface-spec symbol property (see Symbol Properties). The saved-face property stores any face spec saved by the user using the customization buffer; the customized-face property stores the face spec customized for the current session, but not saved; and the theme-face property stores an alist associating the active customization settings and Custom themes with the face specs for that face. The face’s documentation string is stored in the face-documentation property.

Normally, a face is declared just once, using defface, and any further changes to its appearance are applied using the Customize framework (e.g., via the Customize user interface or via the custom-set-faces function; see Applying Customizations), or by face remapping (see Face Remapping). In the rare event that you need to change a face spec directly from Lisp, you can use the face-spec-set function.

Function: face-spec-set face spec &optional spec-type

This function applies spec as a face spec for face. spec should be a face spec, as described in the above documentation for defface.

This function also defines face as a valid face name if it is not already one, and (re)calculates its attributes on existing frames.

The optional argument spec-type determines which spec to set. If it is omitted or nil or face-override-spec, this function sets the override spec, which overrides face specs on face of all the other types mentioned below. This is useful when calling this function outside of Custom code. If spec-type is customized-face or saved-face, this function sets the customized spec or the saved custom spec, respectively. If it is face-defface-spec, this function sets the default face spec (the same one set by defface). If it is reset, this function clears out all customization specs and override specs from face (in this case, the value of spec is ignored). The effect of any other value of spec-type on the face specs is reserved for internal use, but the function will still define face itself and recalculate its attributes, as described above.


41.12.3 Face Attribute Functions

This section describes functions for directly accessing and modifying the attributes of a named face.

Function: face-attribute face attribute &optional frame inherit

This function returns the value of the attribute attribute for face on frame. See Face Attributes, for the supported attributes.

If frame is omitted or nil, that means the selected frame (see Input Focus). If frame is t, this function returns the value of the specified attribute for newly-created frames, i.e. the value of the attribute before applying the face spec in the face’s defface definition (see Defining Faces) or the spec set by face-spec-set. This default value of attribute is normally unspecified, unless you have specified some other value using set-face-attribute; see below.

If inherit is nil, only attributes directly defined by face are considered, so the return value may be unspecified, or a relative value. If inherit is non-nil, face’s definition of attribute is merged with the faces specified by its :inherit attribute; however the return value may still be unspecified or relative. If inherit is a face or a list of faces, then the result is further merged with that face (or faces), until it becomes specified and absolute.

To ensure that the return value is always specified and absolute, use a value of default for inherit; this will resolve any unspecified or relative values by merging with the default face (which is always completely specified).

For example,

(face-attribute 'bold :weight)
     ⇒ bold
Function: face-attribute-relative-p attribute value

This function returns non-nil if value, when used as the value of the face attribute attribute, is relative. This means it would modify, rather than completely override, any value that comes from a subsequent face in the face list or that is inherited from another face.

unspecified is a relative value for all attributes. For :height, floating point and function values are also relative.

For example:

(face-attribute-relative-p :height 2.0)
     ⇒ t
Function: face-all-attributes face &optional frame

This function returns an alist of attributes of face. The elements of the result are name-value pairs of the form (attr-name . attr-value). Optional argument frame specifies the frame whose definition of face to return; if omitted or nil, the returned value describes the default attributes of face for newly created frames, i.e. the values these attributes have before applying the face spec in the face’s defface definition or the spec set by face-spec-set. These default values of the attributes are normally unspecified, unless you have specified some other value using set-face-attribute; see below.

Function: merge-face-attribute attribute value1 value2

If value1 is a relative value for the face attribute attribute, returns it merged with the underlying value value2; otherwise, if value1 is an absolute value for the face attribute attribute, returns value1 unchanged.

Normally, Emacs uses the face specs of each face to automatically calculate its attributes on each frame (see Defining Faces). The function set-face-attribute can override this calculation by directly assigning attributes to a face, either on a specific frame or for all frames. This function is mostly intended for internal usage.

Function: set-face-attribute face frame &rest arguments

This function sets one or more attributes of face for frame. The attributes specified in this way override the face spec(s) belonging to face. See Face Attributes, for the supported attributes.

The extra arguments arguments specify the attributes to set, and the values for them. They should consist of alternating attribute names (such as :family or :underline) and values. Thus,

(set-face-attribute 'foo nil :weight 'bold :slant 'italic)

sets the attribute :weight to bold and the attribute :slant to italic.

If frame is t, this function sets the default attributes for newly created frames; they will effectively override the attribute values specified by defface. If frame is nil, this function sets the attributes for all existing frames, as well as for newly created frames.

To reset the value of an attribute, that is, to indicate that the face doesn’t by itself specify a value for the attribute, use the special value unspecified (not nil!) for the attribute, and set the frame argument to t, in addition to the call with frame set to nil. This is because the default attributes for newly created frames are merged with the face’s spec in defface when a new frame is created, and so having unspecified in the default attributes for new frames will be unable to override defface; the special call to this function as described above will arrange for defface to be overridden.

Note that the attribute-value pairs are evaluated in the order they are specified, with the exception of the :family and :foundry attributes, which are evaluated first. This means that if a certain attribute is specified more than once, only the last value will be used. It also means that in some cases a different order of attributes will produce different results. For example, when :weight is placed before :font, the weight value is applied to the current font of the face, and might be rounded to the closest available weight of that font, whereas when :font is placed before :weight the weight value is applied to the specified font.

The following commands and functions mostly provide compatibility with old versions of Emacs. They work by calling set-face-attribute. Values of t and nil (or omitted) for their frame argument are handled just like set-face-attribute and face-attribute. The commands read their arguments using the minibuffer, if called interactively.

Command: set-face-foreground face color &optional frame
Command: set-face-background face color &optional frame

These set the :foreground attribute (or :background attribute, respectively) of face to color.

Command: set-face-stipple face pattern &optional frame

This sets the :stipple attribute of face to pattern.

Command: set-face-font face font &optional frame

Change the font-related attributes of face to those of font (a string or a font object). See face-font-attribute, for the supported formats of the font argument. This function sets the attribute :font of the face, and indirectly also the :family, :foundry, :width, :height, :weight, and :slant attributes, as defined by the font. If frame is non-nil, only change the attributes on the specified frame.

Function: set-face-bold face bold-p &optional frame

This sets the :weight attribute of face to normal if bold-p is nil, and to bold otherwise.

Function: set-face-italic face italic-p &optional frame

This sets the :slant attribute of face to normal if italic-p is nil, and to italic otherwise.

Command: set-face-underline face underline &optional frame

This sets the :underline attribute of face to underline.

Command: set-face-inverse-video face inverse-video-p &optional frame

This sets the :inverse-video attribute of face to inverse-video-p.

Command: invert-face face &optional frame

This swaps the foreground and background colors of face face.

Command: set-face-extend face extend &optional frame

This sets the :extend attribute of face to extend.

The following functions examine the attributes of a face. They mostly provide compatibility with old versions of Emacs. If you don’t specify frame, they refer to the selected frame; t refers to the default data for new frames. They return unspecified if the face doesn’t define any value for that attribute. If inherit is nil, only an attribute directly defined by the face is returned. If inherit is non-nil, any faces specified by its :inherit attribute are considered as well, and if inherit is a face or a list of faces, then they are also considered, until a specified attribute is found. To ensure that the return value is always specified, use a value of default for inherit.

Function: face-font face &optional frame character

This function returns the name of the font used by the specified face.

If the optional argument frame is specified, it returns the name of the font of face for that frame; frame defaults to the selected frame if it is nil or omitted. If frame is t, the function reports on the font defaults for face to be used for new frames.

By default, the returned font is for displaying ASCII characters, but if frame is anything but t, and the optional third argument character is supplied, the function returns the font name used by face for that character.

Function: face-foreground face &optional frame inherit
Function: face-background face &optional frame inherit

These functions return the foreground color (or background color, respectively) of face face, as a string. If the color is unspecified, they return nil.

Function: face-stipple face &optional frame inherit

This function returns the name of the background stipple pattern of face face, or nil if it doesn’t have one.

Function: face-bold-p face &optional frame inherit

This function returns a non-nil value if the :weight attribute of face is bolder than normal (i.e., one of semi-bold, bold, extra-bold, or ultra-bold). Otherwise, it returns nil.

Function: face-italic-p face &optional frame inherit

This function returns a non-nil value if the :slant attribute of face is italic or oblique, and nil otherwise.

Function: face-underline-p face &optional frame inherit

This function returns non-nil if face face specifies a non-nil :underline attribute.

Function: face-inverse-video-p face &optional frame inherit

This function returns non-nil if face face specifies a non-nil :inverse-video attribute.

Function: face-extend-p face &optional frame inherit

This function returns non-nil if face face specifies a non-nil :extend attribute. The inherit argument is passed to face-attribute.


41.12.4 Displaying Faces

When Emacs displays a given piece of text, the visual appearance of the text may be determined by faces drawn from different sources. If these various sources together specify more than one face for a particular character, Emacs merges the attributes of the various faces. Here is the order in which Emacs merges the faces, from highest to lowest priority:

  • If the text consists of a special glyph, the glyph can specify a particular face. See Glyphs.
  • If the text lies within an active region, Emacs highlights it using the region face. See Standard Faces in The GNU Emacs Manual.
  • If the text lies within an overlay with a non-nil face property, Emacs applies the face(s) specified by that property. If the overlay has a mouse-face property and the mouse is near enough to the overlay, Emacs applies the face or face attributes specified by the mouse-face property instead. See Overlay Properties.

    When multiple overlays cover the same character, an overlay with higher priority overrides those with lower priority. See Overlays.

  • If the text contains a face or mouse-face property, Emacs applies the specified faces and face attributes. See Properties with Special Meanings. (This is how Font Lock mode faces are applied. See Font Lock Mode.)
  • If the text lies within the mode line of the selected window, Emacs applies the mode-line face. For the mode line of a non-selected window, Emacs applies the mode-line-inactive face. For a header line, Emacs applies the header-line face. For a tab line, Emacs applies the tab-line face.
  • If the text comes from an overlay string via before-string or after-string properties (see Overlay Properties), or from a display string (see Other Display Specifications), and the string doesn’t contain a face or mouse-face property, or these properties leave some face attributes undefined, but the buffer text affected by the overlay/display property does define a face or those attributes, Emacs applies the face attributes of the “underlying” buffer text. Note that this is so even if the overlay or display string is displayed in the display margins (see Displaying in the Margins).
  • If any given attribute has not been specified during the preceding steps, Emacs applies the attribute of the default face.

At each stage, if a face has a valid :inherit attribute, Emacs treats any attribute with an unspecified value as having the corresponding value drawn from the parent face(s). see Face Attributes. Note that the parent face(s) may also leave the attribute unspecified; in that case, the attribute remains unspecified at the next level of face merging.


41.12.5 Face Remapping

The variable face-remapping-alist is used for buffer-local or global changes in the appearance of a face. For instance, it is used to implement the text-scale-adjust command (see Text Scale in The GNU Emacs Manual).

Variable: face-remapping-alist

The value of this variable is an alist whose elements have the form (face . remapping). This causes Emacs to display any text having the face face with remapping, rather than the ordinary definition of face.

remapping may be any face spec suitable for a face text property: either a face (i.e., a face name or a property list of attribute/value pairs), or a list of faces. For details, see the description of the face text property in Properties with Special Meanings. remapping serves as the complete specification for the remapped face—it replaces the normal definition of face, instead of modifying it.

If face-remapping-alist is buffer-local, its local value takes effect only within that buffer. If face-remapping-alist includes faces applicable only to certain windows, by using the (:filtered (:window param valspec), that face takes effect only in windows that match the filter conditions (see Properties with Special Meanings). To turn off face filtering temporarily, bind face-filters-always-match to a non-nil value, then all face filters will match any window.

Note: face remapping is non-recursive. If remapping references the same face name face, either directly or via the :inherit attribute of some other face in remapping, that reference uses the normal definition of face. For instance, if the mode-line face is remapped using this entry in face-remapping-alist:

(mode-line italic mode-line)

then the new definition of the mode-line face inherits from the italic face, and the normal (non-remapped) definition of mode-line face.

The following functions implement a higher-level interface to face-remapping-alist. Most Lisp code should use these functions instead of setting face-remapping-alist directly, to avoid trampling on remappings applied elsewhere. These functions are intended for buffer-local remappings, so they all make face-remapping-alist buffer-local as a side-effect. They manage face-remapping-alist entries of the form

  (face relative-spec-1 relative-spec-2 ... base-spec)

where, as explained above, each of the relative-spec-N and base-spec is either a face name, or a property list of attribute/value pairs. Each of the relative remapping entries, relative-spec-N, is managed by the face-remap-add-relative and face-remap-remove-relative functions; these are intended for simple modifications like changing the text size. The base remapping entry, base-spec, has the lowest priority and is managed by the face-remap-set-base and face-remap-reset-base functions; it is intended for major modes to remap faces in the buffers they control.

Function: face-remap-add-relative face &rest specs

This function adds specs as relative remappings for face face in the current buffer. specs should be a list where each element is either a face name, or a property list of attribute/value pairs.

The return value is a Lisp object that serves as a cookie; you can pass this object as an argument to face-remap-remove-relative if you need to remove the remapping later.

;; Remap the 'escape-glyph' face into a combination
;; of the 'highlight' and 'italic' faces:
(face-remap-add-relative 'escape-glyph 'highlight 'italic)

;; Increase the size of the 'default' face by 50%:
(face-remap-add-relative 'default :height 1.5)

Note that buffer-local face remapping does not work reliably for parent faces of basic faces (see Basic Faces). (These are the faces that are used in mode lines, header lines, and other basic decorations of windows and frames.) For instance, mode-line-inactive inherits from mode-line, but remapping mode-line won’t normally have the desired effect on mode-line-inactive, especially if done locally for some buffers. Instead you have to remap mode-line-inactive directly.

Function: face-remap-remove-relative cookie

This function removes a relative remapping previously added by face-remap-add-relative. cookie should be the Lisp object returned by face-remap-add-relative when the remapping was added.

Function: face-remap-set-base face &rest specs

This function sets the base remapping of face in the current buffer to specs. If specs is empty, the default base remapping is restored, similar to calling face-remap-reset-base (see below); note that this is different from specs containing a single value nil, which has the opposite result (the global definition of face is ignored).

This overwrites the default base-spec, which inherits the global face definition, so it is up to the caller to add such inheritance if so desired.

Function: face-remap-reset-base face

This function sets the base remapping of face to its default value, which inherits from face’s global definition.


41.12.6 Functions for Working with Faces

Here are additional functions for creating and working with faces.

Function: face-list

This function returns a list of all defined face names.

Function: face-id face

This function returns the face number of face face. This is a number that uniquely identifies a face at low levels within Emacs. It is seldom necessary to refer to a face by its face number. However, functions that manipulate glyphs, such as make-glyph-code and glyph-face (see Glyphs) access the face numbers internally. Note that the face number is stored as the value of the face property of the face symbol, so we recommend not to set that property of a face to any value of your own.

Function: face-documentation face

This function returns the documentation string of face face, or nil if none was specified for it.

Function: face-equal face1 face2 &optional frame

This returns t if the faces face1 and face2 have the same attributes for display.

Function: face-differs-from-default-p face &optional frame

This returns non-nil if the face face displays differently from the default face.

A face alias provides an equivalent name for a face. You can define a face alias by giving the alias symbol the face-alias property, with a value of the target face name. The following example makes modeline an alias for the mode-line face.

(put 'modeline 'face-alias 'mode-line)
Macro: define-obsolete-face-alias obsolete-face current-face when

This macro defines obsolete-face as an alias for current-face, and also marks it as obsolete, indicating that it may be removed in future. when should be a string indicating when obsolete-face was made obsolete (usually a version number string).


41.12.7 Automatic Face Assignment

This hook is used for automatically assigning faces to text in the buffer. It is part of the implementation of Jit-Lock mode, used by Font-Lock.

Variable: fontification-functions

This variable holds a list of functions that are called by Emacs redisplay as needed, just before doing redisplay. They are called even when Font Lock Mode isn’t enabled. When Font Lock Mode is enabled, this variable usually holds just one function, jit-lock-function.

The functions are called in the order listed, with one argument, a buffer position pos. Collectively they should attempt to assign faces to the text in the current buffer starting at pos.

The functions should record the faces they assign by setting the face property. They should also add a non-nil fontified property to all the text they have assigned faces to. That property tells redisplay that faces have been assigned to that text already.

It is probably a good idea for the functions to do nothing if the character after pos already has a non-nil fontified property, but this is not required. If one function overrides the assignments made by a previous one, the properties after the last function finishes are the ones that really matter.

For efficiency, we recommend writing these functions so that they usually assign faces to around 400 to 600 characters at each call.

Note that, when the buffer text includes very long lines, these functions are called as if they were in a with-restriction form (see Narrowing), with a long-line-optimizations-in-fontification-functions label and with the buffer narrowed to a portion around pos.


41.12.8 Basic Faces

If your Emacs Lisp program needs to assign some faces to text, it is often a good idea to use certain existing faces or inherit from them, rather than defining entirely new faces. This way, if other users have customized those existing faces to give Emacs a certain look, your program will fit in without additional customization.

Some of the basic faces defined in Emacs are listed below. In addition to these, you might want to make use of the Font Lock faces for syntactic highlighting, if highlighting is not already handled by Font Lock mode, or if some Font Lock faces are not in use. See Faces for Font Lock.

default

The default face, whose attributes are all specified. All other faces implicitly inherit from it: any unspecified attribute defaults to the attribute on this face (see Face Attributes).

mode-line-active
mode-line-inactive
header-line
tab-line

Basic faces used for the mode line, header line, and tab line.

tool-bar
tab-bar
fringe
scroll-bar
window-divider
border
child-frame-border

Basic faces used for the corresponding decorations of GUI frames.

cursor

The basic face used for the text cursor.

mouse

The basic face used for displaying mouse-sensitive text when the mouse pointer is on that text.

bold
italic
bold-italic
underline
fixed-pitch
fixed-pitch-serif
variable-pitch

These have the attributes indicated by their names (e.g., bold has a bold :weight attribute), with all other attributes unspecified (and so given by default).

shadow

For dimmed-out text. For example, it is used for the ignored part of a filename in the minibuffer (see Minibuffers for File Names in The GNU Emacs Manual).

link
link-visited

For clickable text buttons that send the user to a different buffer or location.

highlight

For stretches of text that should temporarily stand out. For example, it is commonly assigned to the mouse-face property for cursor highlighting (see Properties with Special Meanings).

match
isearch
lazy-highlight

For text matching (respectively) permanent search matches, interactive search matches, and lazy highlighting other matches than the current interactive one.

error
warning
success

For text concerning errors, warnings, or successes. For example, these are used for messages in *Compilation* buffers.


41.12.9 Font Selection

Before Emacs can draw a character on a graphical display, it must select a font for that character30. See Fonts in The GNU Emacs Manual. Normally, Emacs automatically chooses a font based on the faces assigned to that character—specifically, the face attributes :family, :weight, :slant, and :width (see Face Attributes). The choice of font also depends on the character to be displayed; some fonts can only display a limited set of characters. If no available font exactly fits the requirements, Emacs looks for the closest matching font. The variables in this section control how Emacs makes this selection.

User Option: face-font-family-alternatives

If a given family is specified but does not exist, this variable specifies alternative font families to try. Each element should have this form:

(family alternate-families…)

If family is specified but not available, Emacs will try the other families given in alternate-families, one by one, until it finds a family that does exist.

User Option: face-font-selection-order

If there is no font that exactly matches all desired face attributes (:width, :height, :weight, and :slant), this variable specifies the order in which these attributes should be considered when selecting the closest matching font. The value should be a list containing those four attribute symbols, in order of decreasing importance. The default is (:width :height :weight :slant).

Font selection first finds the best available matches for the first attribute in the list; then, among the fonts which are best in that way, it searches for the best matches in the second attribute, and so on.

The attributes :weight and :width have symbolic values in a range centered around normal. Matches that are more extreme (farther from normal) are somewhat preferred to matches that are less extreme (closer to normal); this is designed to ensure that non-normal faces contrast with normal ones, whenever possible.

One example of a case where this variable makes a difference is when the default font has no italic equivalent. With the default ordering, the italic face will use a non-italic font that is similar to the default one. But if you put :slant before :height, the italic face will use an italic font, even if its height is not quite right.

User Option: face-font-registry-alternatives

This variable lets you specify alternative font registries to try, if a given registry is specified and doesn’t exist. Each element should have this form:

(registry alternate-registries…)

If registry is specified but not available, Emacs will try the other registries given in alternate-registries, one by one, until it finds a registry that does exist.

Emacs can make use of scalable fonts, but by default it does not use them.

User Option: scalable-fonts-allowed

This variable controls which scalable fonts to use. A value of nil, the default, means do not use scalable fonts. t means to use any scalable font that seems appropriate for the text.

Otherwise, the value must be a list of regular expressions. Then a scalable font is enabled for use if its name matches any regular expression in the list. For example,

(setq scalable-fonts-allowed '("iso10646-1$"))

allows the use of scalable fonts with registry iso10646-1.

Variable: face-font-rescale-alist

This variable specifies scaling for certain faces. Its value should be a list of elements of the form

(fontname-regexp . scale-factor)

If fontname-regexp matches the font name that is about to be used, this says to choose a larger similar font according to the factor scale-factor. You would use this feature to normalize the font size if certain fonts are bigger or smaller than their nominal heights and widths would suggest.


41.12.10 Looking Up Fonts

Function: x-list-fonts name &optional reference-face frame maximum width

This function returns a list of available font names that match name. name should be a string containing a font name in either the Fontconfig, GTK+, or XLFD format (see Fonts in The GNU Emacs Manual). Within an XLFD string, wildcard characters may be used: the ‘*’ character matches any substring, and the ‘?’ character matches any single character. Case is ignored when matching font names.

If the optional arguments reference-face and frame are specified, the returned list includes only fonts that are the same size as reference-face (a face name) currently is on the frame frame.

The optional argument maximum sets a limit on how many fonts to return. If it is non-nil, then the return value is truncated after the first maximum matching fonts. Specifying a small value for maximum can make this function much faster, in cases where many fonts match the pattern.

The optional argument width specifies a desired font width. If it is non-nil, the function only returns those fonts whose characters are (on average) width times as wide as reference-face.

Function: x-family-fonts &optional family frame

This function returns a list describing the available fonts for family family on frame. If family is omitted or nil, this list applies to all families, and therefore, it contains all available fonts. Otherwise, family must be a string; it may contain the wildcards ‘?’ and ‘*’.

The list describes the display that frame is on; if frame is omitted or nil, it applies to the selected frame’s display (see Input Focus).

Each element in the list is a vector of the following form:

[family width point-size weight slant
 fixed-p full registry-and-encoding]

The first five elements correspond to face attributes; if you specify these attributes for a face, it will use this font.

The last three elements give additional information about the font. fixed-p is non-nil if the font is fixed-pitch. full is the full name of the font, and registry-and-encoding is a string giving the registry and encoding of the font.


41.12.11 Fontsets

A fontset is a list of fonts, each assigned to a range of character codes. An individual font cannot display the whole range of characters that Emacs supports, but a fontset can. Fontsets have names, just as fonts do, and you can use a fontset name in place of a font name when you specify the font for a frame or a face. Here is information about defining a fontset under Lisp program control.

Function: create-fontset-from-fontset-spec fontset-spec &optional style-variant-p noerror

This function defines a new fontset according to the specification string fontset-spec. The string should have this format:

fontpattern, [charset:font]…

Whitespace characters before and after the commas are ignored.

The first part of the string, fontpattern, should have the form of a standard X font name, except that the last two fields should be ‘fontset-alias’.

The new fontset has two names, one long and one short. The long name is fontpattern in its entirety. The short name is ‘fontset-alias’. You can refer to the fontset by either name. If a fontset with the same name already exists, an error is signaled, unless noerror is non-nil, in which case this function does nothing.

If optional argument style-variant-p is non-nil, that says to create bold, italic and bold-italic variants of the fontset as well. These variant fontsets do not have a short name, only a long one, which is made by altering fontpattern to indicate the bold and/or italic status.

The specification string also says which fonts to use in the fontset. See below for the details.

The construct ‘charset:font’ specifies which font to use (in this fontset) for one particular character set. Here, charset is the name of a character set, and font is the font to use for that character set. You can use this construct any number of times in the specification string.

For the remaining character sets, those that you don’t specify explicitly, Emacs chooses a font based on fontpattern: it replaces ‘fontset-alias’ with a value that names one character set. For the ASCII character set, ‘fontset-alias’ is replaced with ‘ISO8859-1’.

In addition, when several consecutive fields are wildcards, Emacs collapses them into a single wildcard. This is to prevent use of auto-scaled fonts. Fonts made by scaling larger fonts are not usable for editing, and scaling a smaller font is not useful because it is better to use the smaller font in its own size, which Emacs does.

Thus if fontpattern is this,

-*-fixed-medium-r-normal-*-24-*-*-*-*-*-fontset-24

the font specification for ASCII characters would be this:

-*-fixed-medium-r-normal-*-24-*-ISO8859-1

and the font specification for Chinese GB2312 characters would be this:

-*-fixed-medium-r-normal-*-24-*-gb2312*-*

You may not have any Chinese font matching the above font specification. Most X distributions include only Chinese fonts that have ‘song ti’ or ‘fangsong ti’ in the family field. In such a case, ‘Fontset-n’ can be specified as below:

Emacs.Fontset-0: -*-fixed-medium-r-normal-*-24-*-*-*-*-*-fontset-24,\
        chinese-gb2312:-*-*-medium-r-normal-*-24-*-gb2312*-*

Then, the font specifications for all but Chinese GB2312 characters have ‘fixed’ in the family field, and the font specification for Chinese GB2312 characters has a wild card ‘*’ in the family field.

Function: set-fontset-font fontset characters font-spec &optional frame add

This function modifies the existing fontset to use the font specified by font-spec for displaying the specified characters.

If fontset is nil, this function modifies the fontset of the selected frame or that of frame if frame is not nil.

If fontset is t, this function modifies the default fontset, whose short name as a string is ‘fontset-default’.

The characters argument can be a single character which should be displayed using font-spec. It can also be a cons cell (from . to), where from and to are characters. In that case, use font-spec for all the characters in the range from and to (inclusive).

characters may be a charset symbol (see Character Sets). In that case, use font-spec for all the characters in the charset.

characters may be a script symbol (see char-script-table). In that case, use font-spec for all the characters belonging to the script.

characters may be nil, which means to use font-spec for any character in fontset for which no font-spec is specified.

font-spec may be a font-spec object created by the function font-spec (see Low-Level Font Representation).

font-spec may be a cons cell (family . registry), where family is a family name of a font (possibly including a foundry name at the head), and registry is a registry name of a font (possibly including an encoding name at the tail).

font-spec may be a font name, a string.

font-spec may be nil, which explicitly specifies that there’s no font for the specified characters. This is useful, for example, to avoid expensive system-wide search for fonts for characters that have no glyphs, like those from the Unicode Private Use Area (PUA).

The optional argument add, if non-nil, specifies how to add font-spec to the font specifications previously set for characters. If it is prepend, font-spec is prepended to the existing specs. If it is append, font-spec is appended. By default, font-spec overwrites the previously set font specs.

For instance, this changes the default fontset to use a font whose family name is ‘Kochi Gothic’ for all characters belonging to the charset japanese-jisx0208:

(set-fontset-font t 'japanese-jisx0208
                  (font-spec :family "Kochi Gothic"))
Function: char-displayable-p char

This function returns non-nil if Emacs ought to be able to display char. Or more precisely, if the selected frame’s fontset has a font to display the character set that char belongs to.

Fontsets can specify a font on a per-character basis; when the fontset does that, this function’s value may not be accurate.

This function may return non-nil even when there is no font available, since it also checks whether the coding system for the text terminal can encode the character (see Terminal I/O Encoding).


41.12.12 Low-Level Font Representation

Normally, it is not necessary to manipulate fonts directly. In case you need to do so, this section explains how.

In Emacs Lisp, fonts are represented using three different Lisp object types: font objects, font specs, and font entities.

Function: fontp object &optional type

Return t if object is a font object, font spec, or font entity. Otherwise, return nil.

The optional argument type, if non-nil, determines the exact type of Lisp object to check for. In that case, type should be one of font-object, font-spec, or font-entity.

A font object is a Lisp object that represents a font that Emacs has opened. Font objects cannot be modified in Lisp, but they can be inspected.

Function: font-at position &optional window string

Return the font object that is being used to display the character at position position in the window window. If window is nil, it defaults to the selected window. If string is nil, position specifies a position in the current buffer; otherwise, string should be a string, and position specifies a position in that string.

A font spec is a Lisp object that contains a set of specifications that can be used to find a font. More than one font may match the specifications in a font spec.

Function: font-spec &rest arguments

Return a new font spec using the specifications in arguments, which should come in property-value pairs. The possible specifications are as follows:

:name

The font name (a string), in either XLFD, Fontconfig, or GTK+ format. See Fonts in The GNU Emacs Manual.

:family
:foundry
:weight
:slant
:width

These have the same meanings as the face attributes of the same name. See Face Attributes. :family and :foundry are strings, while the other three are symbols. As example values, :slant may be italic, :weight may be bold and :width may be normal.

:size

The font size—either a non-negative integer that specifies the pixel size, or a floating-point number that specifies the point size.

:adstyle

Additional typographic style information for the font, such as ‘sans’. The value should be a string or a symbol.

:registry

The charset registry and encoding of the font, such as ‘iso8859-1’. The value should be a string or a symbol.

:dpi

The resolution in dots per inch for which the font is designed. The value must be a non-negative number.

:spacing

The spacing of the font: proportional, dual, mono, or charcell. The value should be either an integer (0 for proportional, 90 for dual, 100 for mono, 110 for charcell) or a one-letter symbol (one of P, D, M, or C).

:avgwidth

The average width of the font in 1/10 pixel units. The value should be a non-negative number.

:script

The script that the font must support (a symbol).

:lang

The language that the font should support. The value should be a symbol whose name is a two-letter ISO-639 language name. On X, the value is matched against the “Additional Style” field of the XLFD name of a font, if it is non-empty. On MS-Windows, fonts matching the spec are required to support codepages needed for the language. Currently, only a small set of CJK languages is supported with this property: ‘ja’, ‘ko’, and ‘zh’.

:otf

The font must be an OpenType font that supports these OpenType features, provided Emacs is compiled with a library, such as ‘libotf’ on GNU/Linux, that supports complex text layout for scripts which need that. The value must be a list of the form

(script-tag langsys-tag gsub gpos)

where script-tag is the OpenType script tag symbol; langsys-tag is the OpenType language system tag symbol, or nil to use the default language system; gsub is a list of OpenType GSUB feature tag symbols, or nil if none is required; and gpos is a list of OpenType GPOS feature tag symbols, or nil if none is required. If gsub or gpos is a list, a nil element in that list means that the font must not match any of the remaining tag symbols. The gpos element may be omitted. For the list of OpenType script, language, and feature tags, see the list of registered OTF tags.

:type

The symbol that specifies the font backend used to draw the characters. The possible values depend on the platform and on how Emacs was configured at build time. Typical values include ftcrhb and xfthb on X, harfbuzz on MS-Windows, ns on GNUstep, etc. It can also be nil if left unspecified, typically in a font-spec.

Function: font-put font-spec property value

Set the font property property in the font-spec font-spec to value. The property can any of the ones described above.

A font entity is a reference to a font that need not be open. Its properties are intermediate between a font object and a font spec: like a font object, and unlike a font spec, it refers to a single, specific font. Unlike a font object, creating a font entity does not load the contents of that font into computer memory. Emacs may open multiple font objects of different sizes from a single font entity referring to a scalable font.

Function: find-font font-spec &optional frame

This function returns a font entity that best matches the font spec font-spec on frame frame. If frame is nil, it defaults to the selected frame.

Function: list-fonts font-spec &optional frame num prefer

This function returns a list of all font entities that match the font spec font-spec.

The optional argument frame, if non-nil, specifies the frame on which the fonts are to be displayed. The optional argument num, if non-nil, should be an integer that specifies the maximum length of the returned list. The optional argument prefer, if non-nil, should be another font spec, which is used to control the order of the returned list; the returned font entities are sorted in order of decreasing closeness to that font spec.

If you call set-face-attribute and pass a font spec, font entity, or font name string as the value of the :font attribute, Emacs opens the best matching font that is available for display. It then stores the corresponding font object as the actual value of the :font attribute for that face.

The following functions can be used to obtain information about a font. For these functions, the font argument can be a font object, a font entity, or a font spec.

Function: font-get font property

This function returns the value of the font property property for font. The property can any of the ones that font-spec supports.

If font is a font spec and the font spec does not specify property, the return value is nil. If font is a font object or font entity, the value for the :script property may be a list of scripts supported by the font, and the value of the :otf property is a cons of the form (gsub . gpos), where gsub and gpos are lists representing OpenType features supported by the font, of the form

((script-tag (langsys-tag feature…) …) …)

where script-tag, langsys-tag, and feature are symbols representing OpenType layout tags.

If font is a font object, the special property :combining-capability is non-nil if the font backend of font supports rendering of combining characters for non-OpenType fonts.

Function: font-face-attributes font &optional frame

This function returns a list of face attributes corresponding to font. The optional argument frame specifies the frame on which the font is to be displayed. If it is nil, the selected frame is used. The return value has the form

(:family family :height height :weight weight
   :slant slant :width width)

where the values of family, height, weight, slant, and width are face attribute values. Some of these key-attribute pairs may be omitted from the list if they are not specified by font.

Function: font-xlfd-name font &optional fold-wildcards

This function returns the XLFD (X Logical Font Descriptor), a string, matching font. See Fonts in The GNU Emacs Manual, for information about XLFDs. If the name is too long for an XLFD (which can contain at most 255 characters), the function returns nil.

If the optional argument fold-wildcards is non-nil, consecutive wildcards in the XLFD are folded into one.

The following two functions return important information about a font.

Function: font-info name &optional frame

This function returns information about a font specified by its name, a string, as it is used on frame. If frame is omitted or nil, it defaults to the selected frame.

The value returned by the function is a vector of the form [opened-name full-name size height baseline-offset relative-compose default-ascent max-width ascent descent space-width average-width filename capability]. Here’s the description of each components of this vector:

opened-name

The name used to open the font, a string.

full-name

The full name of the font, a string.

size

The pixel size of the font.

height

The height of the font in pixels.

baseline-offset

The offset in pixels from the ASCII baseline, positive upward.

relative-compose
default-ascent

Numbers controlling how to compose characters.

max-width

The maximum advance width of the font.

ascent
descent

The ascent and descent of this font. The sum of these two numbers should be equal to the value of height above.

space-width

The width, in pixels, of the font’s space character.

average-width

The average width of the font characters. If this is zero, Emacs uses the value of space-width instead, when it calculates text layout on display.

filename

The file name of the font as a string. This can be nil if the font back-end does not provide a way to find out the font’s file name.

capability

A list whose first element is a symbol representing the font type, one of x, opentype, truetype, type1, pcf, or bdf. For OpenType fonts, the list includes 2 additional elements describing the GSUB and GPOS features supported by the font. Each of these elements is a list of the form ((script (langsys feature …) …) …), where script is a symbol representing an OpenType script tag, langsys is a symbol representing an OpenType langsys tag (or nil, which stands for the default langsys), and each feature is a symbol representing an OpenType feature tag.

Function: query-font font-object

This function returns information about a font-object. (This is in contrast to font-info, which takes the font name, a string, as its argument.)

The value returned by the function is a vector of the form [name filename pixel-size max-width ascent descent space-width average-width capability]. Here’s the description of each components of this vector:

name

The font name, a string.

filename

The file name of the font as a string. This can be nil if the font back-end does not provide a way to find out the font’s file name.

pixel-size

The pixel size of the font used to open the font.

max-width

The maximum advance width of the font.

ascent
descent

The ascent and descent of this font. The sum of these two numbers gives the font height.

space-width

The width, in pixels, of the font’s space character.

average-width

The average width of the font characters. If this is zero, Emacs uses the value of space-width instead, when it calculates text layout on display.

capability

A list whose first element is a symbol representing the font type, one of x, opentype, truetype, type1, pcf, or bdf. For OpenType fonts, the list includes 2 additional elements describing the GSUB and GPOS features supported by the font. Each of these elements is a list of the form ((script (langsys feature …) …) …), where script is a symbol representing an OpenType script tag, langsys is a symbol representing an OpenType langsys tag (or nil, which stands for the default langsys), and each feature is a symbol representing an OpenType feature tag.

The following four functions return size information about fonts used by various faces, allowing various layout considerations in Lisp programs. These functions take face remapping into consideration, returning information about the remapped face, if the face in question was remapped. See Face Remapping.

Function: default-font-width

This function returns the average width in pixels of the font used by the current buffer’s default face, as that face is defined for the selected frame.

Function: default-font-height

This function returns the height in pixels of the font used by the current buffer’s default face, as that face is defined for the selected frame.

Function: window-font-width &optional window face

This function returns the average width in pixels for the font used by face in window. The specified window must be a live window. If nil or omitted, window defaults to the selected window, and face defaults to the default face in window.

Function: window-font-height &optional window face

This function returns the height in pixels for the font used by face in window. The specified window must be a live window. If nil or omitted, window defaults to the selected window, and face defaults to the default face in window.


41.13 Fringes

On graphical displays, Emacs draws fringes next to each window: thin vertical strips down the sides which can display bitmaps indicating truncation, continuation, horizontal scrolling, and so on.


41.13.1 Fringe Size and Position

The following buffer-local variables control the position and width of fringes in windows showing that buffer.

Variable: fringes-outside-margins

The fringes normally appear between the display margins and the window text. If the value is non-nil, they appear outside the display margins. See Displaying in the Margins.

Variable: left-fringe-width

This variable, if non-nil, specifies the width of the left fringe in pixels. A value of nil means to use the left fringe width from the window’s frame.

Variable: right-fringe-width

This variable, if non-nil, specifies the width of the right fringe in pixels. A value of nil means to use the right fringe width from the window’s frame.

Any buffer which does not specify values for these variables uses the values specified by the left-fringe and right-fringe frame parameters (see Layout Parameters).

The above variables actually take effect via the function set-window-buffer (see Buffers and Windows), which calls set-window-fringes as a subroutine. If you change one of these variables, the fringe display is not updated in existing windows showing the buffer, unless you call set-window-buffer again in each affected window. You can also use set-window-fringes to control the fringe display in individual windows.

Function: set-window-fringes window left &optional right outside-margins persistent

This function sets the fringe widths of window window. If window is nil, the selected window is used.

The argument left specifies the width in pixels of the left fringe, and likewise right for the right fringe. A value of nil for either one stands for the default width. If outside-margins is non-nil, that specifies that fringes should appear outside of the display margins.

If window is not large enough to accommodate fringes of the desired width, this leaves the fringes of window unchanged.

The values specified here may be later overridden by invoking set-window-buffer (see Buffers and Windows) on window with its keep-margins argument nil or omitted. However, if the optional fifth argument persistent is non-nil and the other arguments are processed successfully, the values specified here unconditionally survive subsequent invocations of set-window-buffer. This can be used to permanently turn off fringes in the minibuffer window, consult the description of set-window-scroll-bars for an example (see Scroll Bars).

Function: window-fringes &optional window

This function returns information about the fringes of a window window. If window is omitted or nil, the selected window is used. The value has the form (left-width right-width outside-margins persistent).


41.13.2 Fringe Indicators

Fringe indicators are tiny icons displayed in the window fringe to indicate truncated or continued lines, buffer boundaries, etc.

User Option: indicate-empty-lines

When this is non-nil, Emacs displays a special glyph in the fringe of each empty line at the end of the buffer, on graphical displays. See Fringes. This variable is automatically buffer-local in every buffer.

User Option: indicate-buffer-boundaries

This buffer-local variable controls how the buffer boundaries and window scrolling are indicated in the window fringes.

Emacs can indicate the buffer boundaries—that is, the first and last line in the buffer—with angle icons when they appear on the screen. In addition, Emacs can display an up-arrow in the fringe to show that there is text above the screen, and a down-arrow to show there is text below the screen.

There are three kinds of basic values:

nil

Don’t display any of these fringe icons.

left

Display the angle icons and arrows in the left fringe.

right

Display the angle icons and arrows in the right fringe.

any non-alist

Display the angle icons in the left fringe and don’t display the arrows.

Otherwise the value should be an alist that specifies which fringe indicators to display and where. Each element of the alist should have the form (indicator . position). Here, indicator is one of top, bottom, up, down, and t (which covers all the icons not yet specified), while position is one of left, right and nil.

For example, ((top . left) (t . right)) places the top angle bitmap in left fringe, and the bottom angle bitmap as well as both arrow bitmaps in right fringe. To show the angle bitmaps in the left fringe, and no arrow bitmaps, use ((top . left) (bottom . left)).

Variable: fringe-indicator-alist

This buffer-local variable specifies the mapping from logical fringe indicators to the actual bitmaps displayed in the window fringes. The value is an alist of elements (indicator . bitmaps), where indicator specifies a logical indicator type and bitmaps specifies the fringe bitmaps to use for that indicator.

Each indicator should be one of the following symbols:

truncation, continuation.

Used for truncation and continuation lines.

up, down, top, bottom, top-bottom

Used when indicate-buffer-boundaries is non-nil: up and down indicate a buffer boundary lying above or below the window edge; top and bottom indicate the topmost and bottommost buffer text line; and top-bottom indicates where there is just one line of text in the buffer.

empty-line

Used to indicate empty lines after the buffer end when indicate-empty-lines is non-nil.

overlay-arrow

Used for overlay arrows (see The Overlay Arrow).

Each bitmaps value may be a list of symbols (left right [left1 right1]). The left and right symbols specify the bitmaps shown in the left and/or right fringe, for the specific indicator. left1 and right1 are specific to the bottom and top-bottom indicators, and are used to indicate that the last text line has no final newline. Alternatively, bitmaps may be a single symbol which is used in both left and right fringes.

See Fringe Bitmaps, for a list of standard bitmap symbols and how to define your own. In addition, nil represents the empty bitmap (i.e., an indicator that is not shown).

When fringe-indicator-alist has a buffer-local value, and there is no bitmap defined for a logical indicator, or the bitmap is t, the corresponding value from the default value of fringe-indicator-alist is used.


41.13.3 Fringe Cursors

When a line is exactly as wide as the window, Emacs displays the cursor in the right fringe instead of using two lines. Different bitmaps are used to represent the cursor in the fringe depending on the current buffer’s cursor type.

User Option: overflow-newline-into-fringe

If this is non-nil, lines exactly as wide as the window (not counting the final newline character) are not continued. Instead, when point is at the end of the line, the cursor appears in the right fringe.

Variable: fringe-cursor-alist

This variable specifies the mapping from logical cursor type to the actual fringe bitmaps displayed in the right fringe. The value is an alist where each element has the form (cursor-type . bitmap), which means to use the fringe bitmap bitmap to display cursors of type cursor-type.

Each cursor-type should be one of box, hollow, bar, hbar, or hollow-small. The first four have the same meanings as in the cursor-type frame parameter (see Cursor Parameters). The hollow-small type is used instead of hollow when the normal hollow-rectangle bitmap is too tall to fit on a specific display line.

Each bitmap should be a symbol specifying the fringe bitmap to be displayed for that logical cursor type. See Fringe Bitmaps.

When fringe-cursor-alist has a buffer-local value, and there is no bitmap defined for a cursor type, the corresponding value from the default value of fringes-indicator-alist is used.


41.13.4 Fringe Bitmaps

The fringe bitmaps are the actual bitmaps which represent the logical fringe indicators for truncated or continued lines, buffer boundaries, overlay arrows, etc. Each bitmap is represented by a symbol. These symbols are referred to by the variable fringe-indicator-alist, which maps fringe indicators to bitmaps (see Fringe Indicators), and the variable fringe-cursor-alist, which maps fringe cursors to bitmaps (see Fringe Cursors).

Lisp programs can also directly display a bitmap in the left or right fringe, by using a display property for one of the characters appearing in the line (see Other Display Specifications). Such a display specification has the form

(fringe bitmap [face])

fringe is either the symbol left-fringe or right-fringe. bitmap is a symbol identifying the bitmap to display. The optional face names a face whose foreground and background colors are to be used to display the bitmap, using the attributes of the fringe face for colors that face didn’t specify. If face is omitted, that means to use the attributes of the default face for the colors which the fringe face didn’t specify. For predictable results that don’t depend on the attributes of the default and fringe faces, we recommend you never omit face, but always provide a specific face. In particular, if you want the bitmap to be always displayed in the fringe face, use fringe as face.

For instance, to display an arrow in the left fringe, using the warning face, you could say something like:

(overlay-put
 (make-overlay (point) (point))
 'before-string (propertize
                 "x" 'display
                 `(left-fringe right-arrow warning)))

Here is a list of the standard fringe bitmaps defined in Emacs, and how they are currently used in Emacs (via fringe-indicator-alist and fringe-cursor-alist):

left-arrow, right-arrow

Used to indicate truncated lines.

left-curly-arrow, right-curly-arrow

Used to indicate continued lines.

right-triangle, left-triangle

The former is used by overlay arrows. The latter is unused.

up-arrow, down-arrow
bottom-left-angle, bottom-right-angle
top-left-angle, top-right-angle
left-bracket, right-bracket
empty-line

Used to indicate buffer boundaries.

filled-rectangle, hollow-rectangle
filled-square, hollow-square
vertical-bar, horizontal-bar

Used for different types of fringe cursors.

exclamation-mark, question-mark
large-circle

Not used by core Emacs features.

The next subsection describes how to define your own fringe bitmaps.

Function: fringe-bitmaps-at-pos &optional pos window

This function returns the fringe bitmaps of the display line containing position pos in window window. The return value has the form (left right ov), where left is the symbol for the fringe bitmap in the left fringe (or nil if no bitmap), right is similar for the right fringe, and ov is non-nil if there is an overlay arrow in the left fringe.

The value is nil if pos is not visible in window. If window is nil, that stands for the selected window. If pos is nil, that stands for the value of point in window.


41.13.5 Customizing Fringe Bitmaps

Function: define-fringe-bitmap bitmap bits &optional height width align

This function defines the symbol bitmap as a new fringe bitmap, or replaces an existing bitmap with that name.

The argument bits specifies the image to use. It should be either a string or a vector of integers, where each element (an integer) corresponds to one row of the bitmap. Each bit of an integer corresponds to one pixel of the bitmap, where the low bit corresponds to the rightmost pixel of the bitmap. (Note that this order of bits is opposite of the order in XBM images; see XBM Images.)

The height is normally the length of bits. However, you can specify a different height with non-nil height. The width is normally 8, but you can specify a different width with non-nil width. The width must be an integer between 1 and 16.

The argument align specifies the positioning of the bitmap relative to the range of rows where it is used; the default is to center the bitmap. The allowed values are top, center, or bottom.

The align argument may also be a list (align periodic) where align is interpreted as described above. If periodic is non-nil, it specifies that the rows in bits should be repeated enough times to reach the specified height.

Function: destroy-fringe-bitmap bitmap

This function destroys the fringe bitmap identified by bitmap. If bitmap identifies a standard fringe bitmap, it actually restores the standard definition of that bitmap, instead of eliminating it entirely.

Function: set-fringe-bitmap-face bitmap &optional face

This sets the face for the fringe bitmap bitmap to face. If face is nil, it selects the fringe face. The bitmap’s face controls the color to draw it in.

face is merged with the fringe face, so normally face should specify only the foreground color.


41.13.6 The Overlay Arrow

The overlay arrow is useful for directing the user’s attention to a particular line in a buffer. For example, in the modes used for interface to debuggers, the overlay arrow indicates the line of code about to be executed. This feature has nothing to do with overlays (see Overlays).

Variable: overlay-arrow-string

This variable holds the string to display to call attention to a particular line, or nil if the arrow feature is not in use. On a graphical display the contents of the string are ignored if the left fringe is shown; instead a glyph is displayed in the fringe area to the left of the display area.

Variable: overlay-arrow-position

This variable holds a marker that indicates where to display the overlay arrow. It should point at the beginning of a line. On a non-graphical display, or when the left fringe is not shown, the arrow text appears at the beginning of that line, overlaying any text that would otherwise appear. Since the arrow is usually short, and the line usually begins with indentation, normally nothing significant is overwritten.

The overlay-arrow string is displayed in any given buffer if the value of overlay-arrow-position in that buffer points into that buffer. Thus, it is possible to display multiple overlay arrow strings by creating buffer-local bindings of overlay-arrow-position. However, it is usually cleaner to use overlay-arrow-variable-list to achieve this result.

You can do a similar job by creating an overlay with a before-string property. See Overlay Properties.

You can define multiple overlay arrows via the variable overlay-arrow-variable-list.

Variable: overlay-arrow-variable-list

This variable’s value is a list of variables, each of which specifies the position of an overlay arrow. The variable overlay-arrow-position has its normal meaning because it is on this list.

Each variable on this list can have properties overlay-arrow-string and overlay-arrow-bitmap that specify an overlay arrow string (for text terminals or graphical terminals without the left fringe shown) or fringe bitmap (for graphical terminals with a left fringe) to display at the corresponding overlay arrow position. If either property is not set, the default overlay-arrow-string or overlay-arrow fringe indicator is used.


41.14 Scroll Bars

Normally the frame parameter vertical-scroll-bars controls whether the windows in the frame have vertical scroll bars, and whether they are on the left or right. The frame parameter scroll-bar-width specifies how wide they are (nil meaning the default).

The frame parameter horizontal-scroll-bars controls whether the windows in the frame have horizontal scroll bars. The frame parameter scroll-bar-height specifies how high they are (nil meaning the default). See Layout Parameters.

Horizontal scroll bars are not available on all platforms. The function horizontal-scroll-bars-available-p which takes no argument returns non-nil if they are available on your system.

The following three functions take as argument a live frame which defaults to the selected one.

Function: frame-current-scroll-bars &optional frame

This function reports the scroll bar types for frame frame. The value is a cons cell (vertical-type . horizontal-type), where vertical-type is either left, right, or nil (which means no vertical scroll bar.) horizontal-type is either bottom or nil (which means no horizontal scroll bar).

Function: frame-scroll-bar-width &optional frame

This function returns the width of vertical scroll bars of frame in pixels.

Function: frame-scroll-bar-height &optional frame

This function returns the height of horizontal scroll bars of frame in pixels.

You can override the frame specific settings for individual windows by using the following function:

Function: set-window-scroll-bars window &optional width vertical-type height horizontal-type persistent

This function sets the width and/or height and the types of scroll bars for window window. If window is nil, the selected window is used.

width specifies the width of the vertical scroll bar in pixels (nil means use the width specified for the frame). vertical-type specifies whether to have a vertical scroll bar and, if so, where. The possible values are left, right, t, which means to use the frame’s default, and nil for no vertical scroll bar.

height specifies the height of the horizontal scroll bar in pixels (nil means use the height specified for the frame). horizontal-type specifies whether to have a horizontal scroll bar. The possible values are bottom, t, which means to use the frame’s default, and nil for no horizontal scroll bar. Note that for a mini window the value t has the same meaning as nil, namely to not show a horizontal scroll bar. You have to explicitly specify bottom in order to show a horizontal scroll bar in a mini window.

If window is not large enough to accommodate a scroll bar of the desired dimension, this leaves the corresponding scroll bar unchanged.

The values specified here may be later overridden by invoking set-window-buffer (see Buffers and Windows) on window with its keep-margins argument nil or omitted. However, if the optional fifth argument persistent is non-nil and the other arguments are processed successfully, the values specified here unconditionally survive subsequent invocations of set-window-buffer.

Using the persistent argument of set-window-scroll-bars and set-window-fringes (see Fringe Size and Position) you can reliably and permanently turn off scroll bars and/or fringes in any minibuffer window by adding the following snippet to your early init file (see The Init File).

(add-hook 'after-make-frame-functions
          (lambda (frame)
            (set-window-scroll-bars
             (minibuffer-window frame) 0 nil 0 nil t)
            (set-window-fringes
             (minibuffer-window frame) 0 0 nil t)))

The following four functions take as argument a live window which defaults to the selected one.

Function: window-scroll-bars &optional window

This function returns a list of the form (width columns vertical-type height lines horizontal-type persistent).

The value width is the value that was specified for the width of the vertical scroll bar (which may be nil); columns is the (possibly rounded) number of columns that the vertical scroll bar actually occupies.

The value height is the value that was specified for the height of the horizontal scroll bar (which may be nil); lines is the (possibly rounded) number of lines that the horizontally scroll bar actually occupies.

The value of persistent is the value specified for window with the last successful invocation of set-window-scroll-bars, nil if there never was one.

Function: window-current-scroll-bars &optional window

This function reports the scroll bar type for window window. The value is a cons cell (vertical-type . horizontal-type). Unlike window-scroll-bars, this reports the scroll bar type actually used, once frame defaults and scroll-bar-mode are taken into account.

Function: window-scroll-bar-width &optional window

This function returns the width in pixels of window’s vertical scrollbar.

Function: window-scroll-bar-height &optional window

This function returns the height in pixels of window’s horizontal scrollbar.

If you do not specify a window’s scroll bar settings via set-window-scroll-bars, the buffer-local variables vertical-scroll-bar, horizontal-scroll-bar, scroll-bar-width and scroll-bar-height in the buffer being displayed control the window’s scroll bars. The function set-window-buffer examines these variables. If you change them in a buffer that is already visible in a window, you can make the window take note of the new values by calling set-window-buffer specifying the same buffer that is already displayed.

You can control the appearance of scroll bars for a particular buffer by setting the following variables which automatically become buffer-local when set.

Variable: vertical-scroll-bar

This variable specifies the location of the vertical scroll bar. The possible values are left, right, t, which means to use the frame’s default, and nil for no scroll bar.

Variable: horizontal-scroll-bar

This variable specifies the location of the horizontal scroll bar. The possible values are bottom, t, which means to use the frame’s default, and nil for no scroll bar.

Variable: scroll-bar-width

This variable specifies the width of the buffer’s vertical scroll bars, measured in pixels. A value of nil means to use the value specified by the frame.

Variable: scroll-bar-height

This variable specifies the height of the buffer’s horizontal scroll bar, measured in pixels. A value of nil means to use the value specified by the frame.

Finally you can toggle the display of scroll bars on all frames by customizing the variables scroll-bar-mode and horizontal-scroll-bar-mode.

User Option: scroll-bar-mode

This variable controls whether and where to put vertical scroll bars in all frames. The possible values are nil for no scroll bars, left to put scroll bars on the left and right to put scroll bars on the right.

User Option: horizontal-scroll-bar-mode

This variable controls whether to display horizontal scroll bars on all frames.


41.15 Window Dividers

Window dividers are bars drawn between a frame’s windows. A right divider is drawn between a window and any adjacent windows on the right. Its width (thickness) is specified by the frame parameter right-divider-width. A bottom divider is drawn between a window and adjacent windows on the bottom or the echo area. Its width is specified by the frame parameter bottom-divider-width. In either case, specifying a width of zero means to not draw such dividers. See Layout Parameters.

Technically, a right divider belongs to the window on its left, which means that its width contributes to the total width of that window. A bottom divider belongs to the window above it, which means that its width contributes to the total height of that window. See Window Sizes. When a window has both, a right and a bottom divider, the bottom divider prevails. This means that a bottom divider is drawn over the full total width of its window while the right divider ends above the bottom divider.

Dividers can be dragged with the mouse and are therefore useful for adjusting the sizes of adjacent windows with the mouse. They also serve to visually set apart adjacent windows when no scroll bars or mode lines are present. The following three faces allow the customization of the appearance of dividers:

window-divider

When a divider is less than three pixels wide, it is drawn solidly with the foreground of this face. For larger dividers this face is used for the inner part only, excluding the first and last pixel.

window-divider-first-pixel

This is the face used for drawing the first pixel of a divider that is at least three pixels wide. To obtain a solid appearance, set this to the same value used for the window-divider face.

window-divider-last-pixel

This is the face used for drawing the last pixel of a divider that is at least three pixels wide. To obtain a solid appearance, set this to the same value used for the window-divider face.

You can get the sizes of the dividers of a specific window with the following two functions.

Function: window-right-divider-width &optional window

Return the width (thickness) in pixels of window’s right divider. window must be a live window and defaults to the selected one. The return value is always zero for a rightmost window.

Function: window-bottom-divider-width &optional window

Return the width (thickness) in pixels of window’s bottom divider. window must be a live window and defaults to the selected one. The return value is zero for the minibuffer window or a bottommost window on a minibuffer-less frame.


41.16 The display Property

The display text property (or overlay property) is used to insert images into text, and to control other aspects of how text displays. Display specifications in the same display property value generally apply in parallel to the text they cover.

If several sources (overlays and/or a text property) specify values for the display property, only one of the values takes effect, following the rules of get-char-property. See Examining Text Properties.

The value of the display property should be a display specification, or a list or vector containing several display specifications.

Function: get-display-property position prop &optional object properties

This convenience function can be used to get a specific display property, no matter whether the display property is a vector, a list or a simple property. This is like get-text-property (see Examining Text Properties), but works on the display property only.

position is the position in the buffer or string to examine, and prop is the display property to return. The optional object argument should be either a string or a buffer, and defaults to the current buffer. If the optional properties argument is non-nil, it should be a display property, and in that case, position and object are ignored. (This can be useful if you’ve already gotten the display property with get-char-property, for instance (see Examining Text Properties).

Function: add-display-text-property start end prop value &optional object

Add display property prop of value to the text from start to end.

If any text in the region has a non-nil display property, those properties are retained. For instance:

(add-display-text-property 4 8 'height 2.0)
(add-display-text-property 2 12 'raise 0.5)

After doing this, the region from 2 to 4 will have the raise display property, the region from 4 to 8 will have both the raise and height display properties, and finally the region from 8 to 12 will only have the raise display property.

If object is non-nil, it should be a string or a buffer. If nil, this defaults to the current buffer.

Some of the display specifications allow inclusion of Lisp forms, which are evaluated at display time. This could be unsafe in certain situations, e.g., when the display specification was generated by some external program/agent. Wrapping a display specification in a list that begins with the special symbol disable-eval, as in (disable-eval spec), will disable evaluation of any Lisp in spec, while still supporting all the other display property features.

The rest of this section describes several kinds of display specifications and what they mean.


41.16.1 Display Specs That Replace The Text

Some kinds of display specifications specify something to display instead of the text that has the property. These are called replacing display specifications. Emacs does not allow the user to interactively move point into the middle of buffer text that is replaced in this way.

If a list of display specifications includes more than one replacing display specification, the first overrides the rest. Replacing display specifications make most other display specifications irrelevant, since those don’t apply to the replacement.

For replacing display specifications, the text that has the property means all the consecutive characters that have the same Lisp object as their display property; these characters are replaced as a single unit. If two characters have different Lisp objects as their display properties (i.e., objects which are not eq), they are handled separately.

Here is an example which illustrates this point. A string serves as a replacing display specification, which replaces the text that has the property with the specified string (see Other Display Specifications). Consider the following function:

(defun foo ()
  (dotimes (i 5)
    (let ((string (concat "A"))
          (start (+ i i (point-min))))
      (put-text-property start (1+ start) 'display string)
      (put-text-property start (+ 2 start) 'display string))))

This function gives each of the first ten characters in the buffer a display property which is a string "A", but they don’t all get the same string object. The first two characters get the same string object, so they are replaced with one ‘A’; the fact that the display property was assigned in two separate calls to put-text-property is irrelevant. Similarly, the next two characters get a second string (concat creates a new string object), so they are replaced with one ‘A’; and so on. Thus, the ten characters appear as five A’s.


41.16.2 Specified Spaces

To display a space of specified width and/or height, use a display specification of the form (space . props), where props is a property list (a list of alternating properties and values). You can put this property on one or more consecutive characters; a space of the specified height and width is displayed in place of all of those characters. These are the properties you can use in props to specify the weight of the space:

:width width

If width is a number, it specifies that the space width should be width times the normal character width. width can also be a pixel width specification (see Pixel Specification for Spaces).

:relative-width factor

Specifies that the width of the stretch should be computed from the first character in the group of consecutive characters that have the same display property. The space width is the pixel width of that character, multiplied by factor. (On text-mode terminals, the “pixel width” of a character is usually 1, but it could be more for TABs and double-width CJK characters.)

:align-to hpos

Specifies that the space should be wide enough to reach the column hpos. If hpos is a number, it is a column number, and is measured in units of the canonical character width (see Frame Font). hpos can also be a pixel width specification (see Pixel Specification for Spaces). When the current line is wider than the window, and is either displayed by one or more continuation lines, or is truncated and possibly scrolled horizontally (see Horizontal Scrolling), hpos is measured from the beginning of the logical line, not from the visual beginning of the screen line. This way, alignment produced by :align-to is consistent with functions that count columns, such as current-column and move-to-column (see Counting Columns).

You should use one and only one of the above properties. You can also specify the height of the space, with these properties:

:height height

Specifies the height of the space. If height is a number, it specifies that the space height should be height times the normal character height. The height may also be a pixel height specification (see Pixel Specification for Spaces).

:relative-height factor

Specifies the height of the space, multiplying the ordinary height of the text having this display specification by factor.

:ascent ascent

If the value of ascent is a non-negative number no greater than 100, it specifies that ascent percent of the height of the space should be considered as the ascent of the space—that is, the part above the baseline. The ascent may also be specified in pixel units with a pixel ascent specification (see Pixel Specification for Spaces).

Don’t use both :height and :relative-height together.

The :width and :align-to properties are supported on non-graphic terminals, but the other space properties in this section are not.

Note that space properties are treated as paragraph separators for the purposes of reordering bidirectional text for display. See Bidirectional Display, for the details.


41.16.3 Pixel Specification for Spaces

The value of the :width, :align-to, :height, and :ascent properties can be a special kind of expression that is evaluated during redisplay. The result of the evaluation is used as an absolute number of pixels.

The following expressions are supported:

  expr ::= num | (num) | unit | elem | pos | image | xwidget | form
  num  ::= integer | float | symbol
  unit ::= in | mm | cm | width | height
  elem ::= left-fringe | right-fringe | left-margin | right-margin
        |  scroll-bar | text
  pos  ::= left | center | right
  form ::= (num . expr) | (op expr ...)
  op   ::= + | -

The form num specifies a fraction of the default frame font height or width. The form (num) specifies an absolute number of pixels. If num is a symbol, symbol, its buffer-local variable binding is used; that binding can be either a number or a cons cell of the forms shown above (including yet another cons cell whose car is a symbol that has a buffer-local binding).

The in, mm, and cm units specify the number of pixels per inch, millimeter, and centimeter, respectively. The width and height units correspond to the default width and height of the current face. An image specification of the form (image . props) (see Image Descriptors) corresponds to the width or height of the specified image. Similarly, an xwidget specification of the form (xwidget . props) stands for the width or height of the specified xwidget. See Embedded Native Widgets.

The elements left-fringe, right-fringe, left-margin, right-margin, scroll-bar, and text specify the width of the corresponding area of the window. When the window displays line numbers (see Size of Displayed Text), the width of the text area is decreased by the screen space taken by the line-number display.

The left, center, and right positions can be used with :align-to to specify a position relative to the left edge, center, or right edge of the text area. When the window displays line numbers, and :align-to is used in display properties of buffer text (as opposed to header line, see below), the left and the center positions are offset to account for the screen space taken by the line-number display.

Any of the above window elements (except text) can also be used with :align-to to specify that the position is relative to the left edge of the given area. Once the base offset for a relative position has been set (by the first occurrence of one of these symbols), further occurrences of these symbols are interpreted as the width of the specified area. For example, to align to the center of the left-margin, use

:align-to (+ left-margin (0.5 . left-margin))

If no specific base offset is set for alignment, it is always relative to the left edge of the text area. For example, ‘:align-to 0’ aligns with the first text column in the text area. When the window displays line numbers, the text is considered to start where the space used for line-number display ends.

A value of the form (num . expr) stands for the product of the values of num and expr. For example, (2 . in) specifies a width of 2 inches, while (0.5 . image) specifies half the width (or height) of the specified image (which should be given by its image spec).

The form (+ expr ...) adds up the value of the expressions. The form (- expr ...) negates or subtracts the value of the expressions.

Text shown in the header line that uses :align-to display specifications is not automatically realigned when display-line-numbers-mode is turned on and off, or when the width of line numbers on display changes. To arrange for the header-line text alignment to be updated, thus keeping the header-line text aligned with the buffer text, turn on the header-line-indent-mode in the buffer and use its two variables, header-line-indent and header-line-indent-width, in the display specification. See Window Header Lines. Here’s a simple example:

(setq header-line-format
      (concat (propertize " "
                          'display
                          '(space :align-to
                                  (+ header-line-indent-width 10)))
              "Column"))

This will keep the text ‘Column’ on the header line aligned with column 10 of buffer text, regardless of whether display-line-numbers-mode is on or off, and also when line-number display changes its width.


41.16.4 Other Display Specifications

Here are the other sorts of display specifications that you can use in the display text property.

string

Display string instead of the text that has this property.

Recursive display specifications are not supported—string’s display properties, if any, are not used.

(image . image-props)

This kind of display specification is an image descriptor (see Image Descriptors). When used as a display specification, it means to display the image instead of the text that has the display specification.

(slice x y width height)

This specification together with image specifies a slice (a partial area) of the image to display. The elements y and x specify the top left corner of the slice, within the image; width and height specify the width and height of the slice. Integers are numbers of pixels. A floating-point number in the range 0.0–1.0 stands for that fraction of the width or height of the entire image.

((margin nil) string)

A display specification of this form means to display string instead of the text that has the display specification, at the same position as that text. It is equivalent to using just string, but it is done as a special case of marginal display (see Displaying in the Margins).

(left-fringe bitmap [face])
(right-fringe bitmap [face])

This display specification on any character of a line of text causes the specified bitmap be displayed in the left or right fringes for that line, instead of the characters that have the display specification. The optional face specifies the face whose colors are to be used for the bitmap display. See Fringe Bitmaps, for the details.

(space-width factor)

This display specification affects all the space characters within the text that has the specification. It displays all of these spaces factor times as wide as normal. The element factor should be an integer or float. Characters other than spaces are not affected at all; in particular, this has no effect on tab characters.

(min-width (width))

This display specification ensures the text that has it takes at least width space on display, by adding a stretch of white space to the end of the text if the text is shorter than width. The text is partitioned using the identity of the parameter, which is why the parameter is a list with one element. For instance:

(insert (propertize "foo" 'display '(min-width (6.0))))

This will add padding after ‘foo’ bringing the total width up to the width of six normal characters. Note that the affected characters are identified by the (6.0) list in the display property, compared with eq. The element width can be either an integer or a float specifying the required minimum width of the text (see Pixel Specification for Spaces).

(height height)

This display specification makes the text taller or shorter. Here are the possibilities for height:

(+ n)

This means to use a font that is n steps larger. A step is defined by the set of available fonts—specifically, those that match what was otherwise specified for this text, in all attributes except height. Each size for which a suitable font is available counts as another step. n should be an integer.

(- n)

This means to use a font that is n steps smaller.

a number, factor

A number, factor, means to use a font that is factor times as tall as the default font.

a symbol, function

A symbol is a function to compute the height. It is called with the current height as argument, and should return the new height to use.

anything else, form

If the height value doesn’t fit the previous possibilities, it is a form. Emacs evaluates it to get the new height, with the symbol height bound to the current specified font height.

(raise factor)

This kind of display specification raises or lowers the text it applies to, relative to the baseline of the line. It is mainly meant to support display of subscripts and superscripts.

The factor must be a number, which is interpreted as a multiple of the height of the affected text. If it is positive, that means to display the characters raised. If it is negative, that means to display them lower down.

Note that if the text also has a height display specification, which was specified before (i.e. to the left of) raise, the latter will affect the amount of raising or lowering in pixels, because that is based on the height of the text being raised. Therefore, if you want to display a sub- or superscript that is smaller than the normal text height, consider specifying raise before height.

You can make any display specification conditional. To do that, package it in another list of the form (when condition . spec). Then the specification spec applies only when condition evaluates to a non-nil value. During the evaluation, object is bound to the string or buffer having the conditional display property. position and buffer-position are bound to the position within object and the buffer position where the display property was found, respectively. Both positions can be different when object is a string.

Note that condition will only be evaluated when redisplay examines the text where this display spec is located, so this feature is best suited for conditions that are relatively stable, i.e. yield, for each particular buffer position, the same results on every evaluation. If the results change for the same text location, e.g., if the result depends on the position of point, then the conditional specification might not do what you want, because redisplay examines only those parts of buffer text where it has reasons to assume that something changed since the last display cycle.


41.16.5 Displaying in the Margins

A buffer can have blank areas called display margins on the left and on the right. Ordinary text never appears in these areas, but you can put things into the display margins using the display property. There is currently no way to make text or images in the margin mouse-sensitive.

The way to display something in the margins is to specify it in a margin display specification in the display property of some text. This is a replacing display specification, meaning that the text you put it on does not get displayed; the margin display appears, but that text does not.

A margin display specification looks like ((margin right-margin) spec) or ((margin left-margin) spec). Here, spec is another display specification that says what to display in the margin. Typically it is a string of text to display, or an image descriptor.

To display something in the margin in association with certain buffer text, without altering or preventing the display of that text, put on that text an overlay with a before-string property, and put the margin display specification on the contents of the before-string.

Note that if the string to be displayed in the margin doesn’t specify a face, its face is determined using the same rules and priorities as it is for strings displayed in the text area (see Displaying Faces). If this results in undesirable “leaking” of faces into the margin, make sure the string has an explicit face specified for it.

Before the display margins can display anything, you must give them a nonzero width. The usual way to do that is to set these variables:

Variable: left-margin-width

This variable specifies the width of the left margin, in character cell (a.k.a. “column”) units. It is buffer-local in all buffers. A value of nil means no left marginal area.

Variable: right-margin-width

This variable specifies the width of the right margin, in character cell units. It is buffer-local in all buffers. A value of nil means no right marginal area.

Setting these variables does not immediately affect the window. These variables are checked when a new buffer is displayed in the window. Thus, you can make changes take effect by calling set-window-buffer. Do not use these variables to try to determine the current width of the left or right margin. Instead, use the function window-margins.

You can also set the margin widths immediately.

Function: set-window-margins window left &optional right

This function specifies the margin widths for window window, in character cell units. The argument left controls the left margin, and right controls the right margin (default 0).

If window is not large enough to accommodate margins of the desired width, this leaves the margins of window unchanged.

The values specified here may be later overridden by invoking set-window-buffer (see Buffers and Windows) on window with its keep-margins argument nil or omitted.

Function: window-margins &optional window

This function returns the width of the left and right margins of window as a cons cell of the form (left . right). If one of the two marginal areas does not exist, its width is returned as nil; if neither of the two margins exist, the function returns (nil). If window is nil, the selected window is used.


41.17 Images

To display an image in an Emacs buffer, you must first create an image descriptor, then use it as a display specifier in the display property of text that is displayed (see The display Property).

Emacs is usually able to display images when it is run on a graphical terminal. Images cannot be displayed in a text terminal, on certain graphical terminals that lack the support for this, or if Emacs is compiled without image support. You can use the function display-images-p to determine if images can in principle be displayed (see Display Feature Testing).


41.17.1 Image Formats

Emacs can display a number of different image formats. Some of these image formats are supported only if particular support libraries are installed. On some platforms, Emacs can load support libraries on demand; if so, the variable dynamic-library-alist can be used to modify the set of known names for these dynamic libraries. See Dynamically Loaded Libraries.

Supported image formats (and the required support libraries) include PBM and XBM (which do not depend on support libraries and are always available), XPM (libXpm), GIF (libgif or libungif), JPEG (libjpeg), TIFF (libtiff), PNG (libpng), SVG (librsvg), and WebP (libwebp).

Each of these image formats is associated with an image type symbol. The symbols for the above formats are, respectively, pbm, xbm, xpm, gif, jpeg, tiff, png, svg, and webp.

On some platforms, the built-in image support that doesn’t require any optional libraries includes BMP images.31

Furthermore, if you build Emacs with ImageMagick (libMagickWand) support, Emacs can display any image format that ImageMagick can. See ImageMagick Images. All images displayed via ImageMagick have type symbol imagemagick.

Variable: image-types

This variable contains a list of type symbols for image formats which are potentially supported in the current configuration.

“Potentially” means that Emacs knows about the image types, not necessarily that they can be used (for example, they could depend on unavailable dynamic libraries). To know which image types are really available, use image-type-available-p.

Function: image-type-available-p type

This function returns non-nil if images of type type can be loaded and displayed. type must be an image type symbol.

For image types whose support libraries are statically linked, this function always returns t. For image types whose support libraries are dynamically loaded, it returns t if the library could be loaded and nil otherwise.


41.17.2 Image Descriptors

An image descriptor is a list which specifies the underlying data for an image, and how to display it. It is typically used as the value of a display overlay or text property (see Other Display Specifications); but See Showing Images, for convenient helper functions to insert images into buffers.

Each image descriptor has the form (image . props), where props is a property list of alternating keyword symbols and values, including at least the pair :type type that specifies the image type.

Image descriptors which define image dimensions, :width, :height, :max-width and :max-height, may take either an integer, which represents the dimension in pixels, or a pair (value . em), where value is the dimension’s length in ems32. One em is equivalent to the height of the font and value may be an integer or a float.

The following is a list of properties that are meaningful for all image types (there are also properties which are meaningful only for certain image types, as documented in the following subsections):

:type type

The image type. See Image Formats. Every image descriptor must include this property.

:file file

This says to load the image from file file. If file is not an absolute file name, it is expanded relative to each of the directories mentioned by image-load-path (see Defining Images).

:data data

This specifies the raw image data. Each image descriptor must have either :data or :file, but not both.

For most image types, the value of a :data property should be a string containing the image data. Some image types do not support :data; for some others, :data alone is not enough, so you need to use other image properties along with :data. See the following subsections for details.

:margin margin

This specifies how many pixels to add as an extra margin around the image. The value, margin, must be a non-negative number, or a pair (x . y) of such numbers. If it is a pair, x specifies how many pixels to add horizontally, and y specifies how many pixels to add vertically. If :margin is not specified, the default is zero.

:ascent ascent

This specifies the amount of the image’s height to use for its ascent—that is, the part above the baseline. The value, ascent, must be a number in the range 0 to 100, or the symbol center.

If ascent is a number, that percentage of the image’s height is used for its ascent.

If ascent is center, the image is vertically centered around a centerline which would be the vertical centerline of text drawn at the position of the image, in the manner specified by the text properties and overlays that apply to the image.

If this property is omitted, it defaults to 50.

:relief relief

This adds a shadow rectangle around the image. The value, relief, specifies the width of the shadow lines, in pixels. If relief is negative, shadows are drawn so that the image appears as a pressed button; otherwise, it appears as an unpressed button.

:width width, :height height

The :width and :height keywords are used for scaling the image. If only one of them is specified, the other one will be calculated so as to preserve the aspect ratio. If both are specified, aspect ratio may not be preserved.

:max-width max-width, :max-height max-height

The :max-width and :max-height keywords are used for scaling if the size of the image exceeds these values. If :width is set, it will have precedence over max-width, and if :height is set, it will have precedence over max-height, but you can otherwise mix these keywords as you wish.

If both :max-width and :height are specified, but :width is not, preserving the aspect ratio might require that width exceeds :max-width. If this happens, scaling will use a smaller value for the height so as to preserve the aspect ratio while not exceeding :max-width. Similarly when both :max-height and :width are specified, but :height is not. For example, if you have a 200x100 image and specify that :width should be 400 and :max-height should be 150, you’ll end up with an image that is 300x150: Preserving the aspect ratio and not exceeding the “max” setting. This combination of parameters is a useful way of saying “display this image as large as possible, but no larger than the available display area”.

:scale scale

This should be a number, where values higher than 1 means to increase the size, and lower means to decrease the size, by multiplying both the width and height. For instance, a value of 0.25 will make the image a quarter size of what it originally was. If the scaling makes the image larger than specified by :max-width or :max-height, the resulting size will not exceed those two values. If both :scale and :height/:width are specified, the height/width will be adjusted by the specified scaling factor.

:rotation angle

Specifies a rotation angle in degrees. Only multiples of 90 degrees are supported, unless the image type is imagemagick. Positive values rotate clockwise, negative values counter-clockwise. Rotation is performed after scaling and cropping.

:flip flip

If this is t, the image will be horizontally flipped. Currently it has no effect if the image type is imagemagick. Vertical flipping can be achieved by rotating the image 180 degrees and toggling this value.

:transform-smoothing smooth

If this is t, any image transform will have smoothing applied; if nil, no smoothing will be applied. The exact algorithm used is platform dependent, but should be equivalent to bilinear filtering. Disabling smoothing will use the nearest neighbor algorithm.

If this property is not specified, create-image will use the image-transform-smoothing user option to say whether scaling should be done or not. This option can be nil (no smoothing), t (use smoothing) or a predicate function that’s called with the image object as the only parameter, and should return either nil or t. The default is for down-scaling to apply smoothing, and for large up-scaling to not apply smoothing.

:index frame

See Multi-Frame Images.

:conversion algorithm

This specifies a conversion algorithm that should be applied to the image before it is displayed; the value, algorithm, specifies which algorithm.

laplace
emboss

Specifies the Laplace edge detection algorithm, which blurs out small differences in color while highlighting larger differences. People sometimes consider this useful for displaying the image for a disabled button.

(edge-detection :matrix matrix :color-adjust adjust)

Specifies a general edge-detection algorithm. matrix must be either a nine-element list or a nine-element vector of numbers. A pixel at position x/y in the transformed image is computed from original pixels around that position. matrix specifies, for each pixel in the neighborhood of x/y, a factor with which that pixel will influence the transformed pixel; element 0 specifies the factor for the pixel at x-1/y-1, element 1 the factor for the pixel at x/y-1 etc., as shown below:

  (x-1/y-1  x/y-1  x+1/y-1
   x-1/y    x/y    x+1/y
   x-1/y+1  x/y+1  x+1/y+1)

The resulting pixel is computed from the color intensity of the color resulting from summing up the RGB values of surrounding pixels, multiplied by the specified factors, and dividing that sum by the sum of the factors’ absolute values.

Laplace edge-detection currently uses a matrix of

  (1  0  0
   0  0  0
   0  0 -1)

Emboss edge-detection uses a matrix of

  ( 2 -1  0
   -1  0  1
    0  1 -2)
disabled

Specifies transforming the image so that it looks disabled.

:mask mask

If mask is heuristic or (heuristic bg), build a clipping mask for the image, so that the background of a frame is visible behind the image. If bg is not specified, or if bg is t, determine the background color of the image by looking at the four corners of the image, assuming the most frequently occurring color from the corners is the background color of the image. Otherwise, bg must be a list (red green blue) specifying the color to assume for the background of the image.

If mask is nil, remove a mask from the image, if it has one. Images in some formats include a mask which can be removed by specifying :mask nil.

:pointer shape

This specifies the pointer shape when the mouse pointer is over this image. See Pointer Shape, for available pointer shapes.

:map map

This associates an image map of hot spots with this image.

An image map is an alist where each element has the format (area id plist). An area is specified as either a rectangle, a circle, or a polygon.

A rectangle is a cons (rect . ((x0 . y0) . (x1 . y1))) which specifies the pixel coordinates of the upper left and bottom right corners of the rectangle area.

A circle is a cons (circle . ((x0 . y0) . r)) which specifies the center and the radius of the circle; r may be a float or integer.

A polygon is a cons (poly . [x0 y0 x1 y1 ...]) where each pair in the vector describes one corner in the polygon.

When the mouse pointer lies on a hot-spot area of an image, the plist of that hot-spot is consulted; if it contains a help-echo property, that defines a tool-tip for the hot-spot, and if it contains a pointer property, that defines the shape of the mouse cursor when it is on the hot-spot. See Pointer Shape, for available pointer shapes.

When you click the mouse when the mouse pointer is over a hot-spot, an event is composed by combining the id of the hot-spot with the mouse event; for instance, [area4 mouse-1] if the hot-spot’s id is area4.

Note that the map’s coordinates should reflect the displayed image after all transforms have been done (rotation, scaling and so on), and also note that Emacs (by default) performs auto-scaling of images, so to make things match up, you should either specify :scale 1.0 when creating the image, or use the result of image-compute-scaling-factor to compute the elements of the map.

Function: image-mask-p spec &optional frame

This function returns t if image spec has a mask bitmap. frame is the frame on which the image will be displayed. frame nil or omitted means to use the selected frame (see Input Focus).

Function: image-transforms-p &optional frame

This function returns non-nil if frame supports image scaling and rotation. frame nil or omitted means to use the selected frame (see Input Focus). The returned list includes symbols that indicate which image transform operations are supported:

scale

Image scaling is supported by frame via the :scale, :width, :height, :max-width, and :max-height properties.

rotate90

Image rotation is supported by frame if the rotation angle is an integral multiple of 90 degrees.

If image transforms are not supported, :rotation, :crop, :width, :height, :scale, :max-width and :max-height will only be usable through ImageMagick, if available (see ImageMagick Images).


41.17.3 XBM Images

To use XBM format, specify xbm as the image type. This image format doesn’t require an external library, so images of this type are always supported.

Additional image properties supported for the xbm image type are:

:foreground foreground

The value, foreground, should be a string specifying the image foreground color, or nil for the default color. This color is used for each pixel in the XBM that is 1. The default is the frame’s foreground color.

:background background

The value, background, should be a string specifying the image background color, or nil for the default color. This color is used for each pixel in the XBM that is 0. The default is the frame’s background color.

If you specify an XBM image using data within Emacs instead of an external file, use the following three properties:

:data data

The value, data, specifies the contents of the image. There are three formats you can use for data:

  • A vector of strings or bool-vectors, each specifying one line of the image. Do specify :data-height and :data-width.
  • A string containing the same byte sequence as an XBM file would contain.
  • A string or a bool-vector containing the bits of the image (plus perhaps some extra bits at the end that will not be used). It should contain at least stride * height bits, where stride is the smallest multiple of 8 greater than or equal to the width of the image. In this case, you should specify :data-height, :data-width and :stride, both to indicate that the string contains just the bits rather than a whole XBM file, and to specify the size of the image.
:stride stride

The number of bool vector entries stored for each row; the smallest multiple of 8 greater than or equal to width.


41.17.4 XPM Images

To use XPM format, specify xpm as the image type. The additional image property :color-symbols is also meaningful with the xpm image type:

:color-symbols symbols

The value, symbols, should be an alist whose elements have the form (name . color). In each element, name is the name of a color as it appears in the image file, and color specifies the actual color to use for displaying that name.


41.17.5 ImageMagick Images

If your Emacs build has ImageMagick support, you can use the ImageMagick library to load many image formats (see File Conveniences in The GNU Emacs Manual). The image type symbol for images loaded via ImageMagick is imagemagick, regardless of the actual underlying image format.

To check for ImageMagick support, use the following:

(image-type-available-p 'imagemagick)
Function: imagemagick-types

This function returns a list of image file extensions supported by the current ImageMagick installation. Each list element is a symbol representing an internal ImageMagick name for an image type, such as BMP for .bmp images.

User Option: imagemagick-enabled-types

The value of this variable is a list of ImageMagick image types which Emacs may attempt to render using ImageMagick. Each list element should be one of the symbols in the list returned by imagemagick-types, or an equivalent string. Alternatively, a value of t enables ImageMagick for all possible image types. Regardless of the value of this variable, imagemagick-types-inhibit (see below) takes precedence.

User Option: imagemagick-types-inhibit

The value of this variable lists the ImageMagick image types which should never be rendered using ImageMagick, regardless of the value of imagemagick-enabled-types. A value of t disables ImageMagick entirely.

Variable: image-format-suffixes

This variable is an alist mapping image types to file name extensions. Emacs uses this in conjunction with the :format image property (see below) to give a hint to the ImageMagick library as to the type of an image. Each element has the form (type extension), where type is a symbol specifying an image content-type, and extension is a string that specifies the associated file name extension.

Images loaded with ImageMagick support the following additional image descriptor properties:

:background background

background, if non-nil, should be a string specifying a color, which is used as the image’s background color if the image supports transparency. If the value is nil, it defaults to the frame’s background color.

:format type

The value, type, should be a symbol specifying the type of the image data, as found in image-format-suffixes. This is used when the image does not have an associated file name, to provide a hint to ImageMagick to help it detect the image type.

:crop geometry

The value of geometry should be a list of the form (width height x y). width and height specify the width and height of the cropped image. If x is a positive number it specifies the offset of the cropped area from the left of the original image, and if negative the offset from the right. If y is a positive number it specifies the offset from the top of the original image, and if negative from the bottom. If x or y are nil or unspecified the crop area will be centered on the original image.

If the crop area is outside or overlaps the edge of the image it will be reduced to exclude any areas outside of the image. This means it is not possible to use :crop to increase the size of the image by entering large width or height values.

Cropping is performed after scaling but before rotation.


41.17.6 SVG Images

SVG (Scalable Vector Graphics) is an XML format for specifying images. SVG images support the following additional image descriptor properties:

:foreground foreground

foreground, if non-nil, should be a string specifying a color, which is used as the image’s foreground color. If the value is nil, it defaults to the current face’s foreground color.

:background background

background, if non-nil, should be a string specifying a color, which is used as the image’s background color if the image supports transparency. If the value is nil, it defaults to the current face’s background color.

:css css

css, if non-nil, should be a string specifying the CSS to override the default CSS used when generating the image.

SVG library

If your Emacs build has SVG support, you can create and manipulate these images with the following functions from the svg.el library.

Function: svg-create width height &rest args

Create a new, empty SVG image with the specified dimensions. args is an argument plist with you can specify following:

:stroke-width

The default width (in pixels) of any lines created.

:stroke

The default stroke color on any lines created.

This function returns an SVG object, a Lisp data structure that specifies an SVG image, and all the following functions work on that structure. The argument svg in the following functions specifies such an SVG object.

Function: svg-gradient svg id type stops

Create a gradient in svg with identifier id. type specifies the gradient type, and can be either linear or radial. stops is a list of percentage/color pairs.

The following will create a linear gradient that goes from red at the start, to green 25% of the way, to blue at the end:

(svg-gradient svg "gradient1" 'linear
              '((0 . "red") (25 . "green") (100 . "blue")))

The gradient created (and inserted into the SVG object) can later be used by all functions that create shapes.

All the following functions take an optional list of keyword parameters that alter the various attributes from their default values. Valid attributes include:

:stroke-width

The width (in pixels) of lines drawn, and outlines around solid shapes.

:stroke-color

The color of lines drawn, and outlines around solid shapes.

:fill-color

The color used for solid shapes.

:id

The identified of the shape.

:gradient

If given, this should be the identifier of a previously defined gradient object.

:clip-path

Identifier of a clip path.

Function: svg-rectangle svg x y width height &rest args

Add to svg a rectangle whose upper left corner is at position x/y and whose size is width/height.

(svg-rectangle svg 100 100 500 500 :gradient "gradient1")
Function: svg-circle svg x y radius &rest args

Add to svg a circle whose center is at x/y and whose radius is radius.

Function: svg-ellipse svg x y x-radius y-radius &rest args

Add to svg an ellipse whose center is at x/y, and whose horizontal radius is x-radius and the vertical radius is y-radius.

Function: svg-line svg x1 y1 x2 y2 &rest args

Add to svg a line that starts at x1/y1 and extends to x2/y2.

Function: svg-polyline svg points &rest args

Add to svg a multiple-segment line (a.k.a. “polyline”) that goes through points, which is a list of X/Y position pairs.

(svg-polyline svg '((200 . 100) (500 . 450) (80 . 100))
              :stroke-color "green")
Function: svg-polygon svg points &rest args

Add a polygon to svg where points is a list of X/Y pairs that describe the outer circumference of the polygon.

(svg-polygon svg '((100 . 100) (200 . 150) (150 . 90))
             :stroke-color "blue" :fill-color "red")
Function: svg-path svg commands &rest args

Add the outline of a shape to svg according to commands, see SVG Path Commands.

Coordinates by default are absolute. To use coordinates relative to the last position, or – initially – to the origin, set the attribute :relative to t. This attribute can be specified for the function or for individual commands. If specified for the function, then all commands use relative coordinates by default. To make an individual command use absolute coordinates, set :relative to nil.

(svg-path svg
	  '((moveto ((100 . 100)))
	    (lineto ((200 . 0) (0 . 200) (-200 . 0)))
	    (lineto ((100 . 100)) :relative nil))
	  :stroke-color "blue"
	  :fill-color "lightblue"
	  :relative t)
Function: svg-text svg text &rest args

Add the specified text to svg.

(svg-text
 svg "This is a text"
 :font-size "40"
 :font-weight "bold"
 :stroke "black"
 :fill "white"
 :font-family "impact"
 :letter-spacing "4pt"
 :x 300
 :y 400
 :stroke-width 1)
Function: svg-embed svg image image-type datap &rest args

Add an embedded (raster) image to svg. If datap is nil, image should be a file name; otherwise it should be a string containing the image data as raw bytes. image-type should be a MIME image type, for instance "image/jpeg".

(svg-embed svg "~/rms.jpg" "image/jpeg" nil
           :width "100px" :height "100px"
           :x "50px" :y "75px")
Function: svg-embed-base-uri-image svg relative-filename &rest args

To svg add an embedded (raster) image placed at relative-filename. relative-filename is searched inside file-name-directory of the :base-uri svg image property. :base-uri specifies a (possibly non-existing) file name of the svg image to be created, thus all the embedded files are searched relatively to the :base-uri filename’s directory. If :base-uri is omitted, then filename from where svg image is loaded is used. Using :base-uri improves the performance of embedding large images, comparing to svg-embed, because all the work is done directly by librsvg.

;; Embedding /tmp/subdir/rms.jpg and /tmp/another/rms.jpg
(svg-embed-base-uri-image svg "subdir/rms.jpg"
           :width "100px" :height "100px"
           :x "50px" :y "75px")
(svg-embed-base-uri-image svg "another/rms.jpg"
           :width "100px" :height "100px"
           :x "75px" :y "50px")
(svg-image svg :scale 1.0
           :base-uri "/tmp/dummy"
           :width 175 :height 175)
Function: svg-clip-path svg &rest args

Add a clipping path to svg. If applied to a shape via the :clip-path property, parts of that shape which lie outside of the clipping path are not drawn.

(let ((clip-path (svg-clip-path svg :id "foo")))
  (svg-circle clip-path 200 200 175))
(svg-rectangle svg 50 50 300 300
               :fill-color "red"
               :clip-path "url(#foo)")
Function: svg-node svg tag &rest args

Add the custom node tag to svg.

(svg-node svg
          'rect
          :width 300 :height 200 :x 50 :y 100 :fill-color "green")
Function: svg-remove svg id

Remove the element with identifier id from the svg.

Function: svg-image svg

Finally, the svg-image takes an SVG object as its argument and returns an image object suitable for use in functions like insert-image.

Here’s a complete example that creates and inserts an image with a circle:

(let ((svg (svg-create 400 400 :stroke-width 10)))
  (svg-gradient svg "gradient1" 'linear '((0 . "red") (100 . "blue")))
  (svg-circle svg 200 200 100 :gradient "gradient1"
                  :stroke-color "green")
  (insert-image (svg-image svg)))

SVG Path Commands

SVG paths allow creation of complex images by combining lines, curves, arcs, and other basic shapes. The functions described below allow invoking SVG path commands from a Lisp program.

Command: moveto points

Move the pen to the first point in points. Additional points are connected with lines. points is a list of X/Y coordinate pairs. Subsequent moveto commands represent the start of a new subpath.

(svg-path svg '((moveto ((200 . 100) (100 . 200) (0 . 100))))
          :fill "white" :stroke "black")
Command: closepath

End the current subpath by connecting it back to its initial point. A line is drawn along the connection.

(svg-path svg '((moveto ((200 . 100) (100 . 200) (0 . 100)))
                (closepath)
                (moveto ((75 . 125) (100 . 150) (125 . 125)))
                (closepath))
          :fill "red" :stroke "black")
Command: lineto points

Draw a line from the current point to the first element in points, a list of X/Y position pairs. If more than one point is specified, draw a polyline.

(svg-path svg '((moveto ((200 . 100)))
                (lineto ((100 . 200) (0 . 100))))
          :fill "yellow" :stroke "red")
Command: horizontal-lineto x-coordinates

Draw a horizontal line from the current point to the first element in x-coordinates. Specifying multiple coordinates is possible, although this usually doesn’t make sense.

(svg-path svg '((moveto ((100 . 200)))
                (horizontal-lineto (300)))
          :stroke "green")
Command: vertical-lineto y-coordinates

Draw vertical lines.

(svg-path svg '((moveto ((200 . 100)))
                (vertical-lineto (300)))
          :stroke "green")
Command: curveto coordinate-sets

Using the first element in coordinate-sets, draw a cubic Bézier curve from the current point. If there are multiple coordinate sets, draw a polybezier. Each coordinate set is a list of the form (x1 y1 x2 y2 x y), where (xy) is the curve’s end point. (x1y1) and (x2y2) are control points at the beginning and at the end, respectively.

(svg-path svg '((moveto ((100 . 100)))
                (curveto ((200 100 100 200 200 200)
                          (300 200 0 100 100 100))))
          :fill "transparent" :stroke "red")
Command: smooth-curveto coordinate-sets

Using the first element in coordinate-sets, draw a cubic Bézier curve from the current point. If there are multiple coordinate sets, draw a polybezier. Each coordinate set is a list of the form (x2 y2 x y), where (xy) is the curve’s end point and (x2y2) is the corresponding control point. The first control point is the reflection of the second control point of the previous command relative to the current point, if that command was curveto or smooth-curveto. Otherwise the first control point coincides with the current point.

(svg-path svg '((moveto ((100 . 100)))
                (curveto ((200 100 100 200 200 200)))
                (smooth-curveto ((0 100 100 100))))
          :fill "transparent" :stroke "blue")
Command: quadratic-bezier-curveto coordinate-sets

Using the first element in coordinate-sets, draw a quadratic Bézier curve from the current point. If there are multiple coordinate sets, draw a polybezier. Each coordinate set is a list of the form (x1 y1 x y), where (xy) is the curve’s end point and (x1y1) is the control point.

(svg-path svg '((moveto ((200 . 100)))
                (quadratic-bezier-curveto ((300 100 300 200)))
                (quadratic-bezier-curveto ((300 300 200 300)))
                (quadratic-bezier-curveto ((100 300 100 200)))
                (quadratic-bezier-curveto ((100 100 200 100))))
          :fill "transparent" :stroke "pink")
Command: smooth-quadratic-bezier-curveto coordinate-sets

Using the first element in coordinate-sets, draw a quadratic Bézier curve from the current point. If there are multiple coordinate sets, draw a polybezier. Each coordinate set is a list of the form (x y), where (xy) is the curve’s end point. The control point is the reflection of the control point of the previous command relative to the current point, if that command was quadratic-bezier-curveto or smooth-quadratic-bezier-curveto. Otherwise the control point coincides with the current point.

(svg-path svg '((moveto ((200 . 100)))
                (quadratic-bezier-curveto ((300 100 300 200)))
                (smooth-quadratic-bezier-curveto ((200 300)))
                (smooth-quadratic-bezier-curveto ((100 200)))
                (smooth-quadratic-bezier-curveto ((200 100))))
          :fill "transparent" :stroke "lightblue")
Command: elliptical-arc coordinate-sets

Using the first element in coordinate-sets, draw an elliptical arc from the current point. If there are multiple coordinate sets, draw a sequence of elliptical arcs. Each coordinate set is a list of the form (rx ry x y), where (xy) is the end point of the ellipse, and (rxry) are its radii. Attributes may be appended to the list:

:x-axis-rotation

The angle in degrees by which the x-axis of the ellipse is rotated relative to the x-axis of the current coordinate system.

:large-arc

If set to t, draw an arc sweep greater than or equal to 180 degrees. Otherwise, draw an arc sweep smaller than or equal to 180 degrees.

:sweep

If set to t, draw an arc in positive angle direction. Otherwise, draw it in negative angle direction.

(svg-path svg '((moveto ((200 . 250)))
                (elliptical-arc ((75 75 200 350))))
          :fill "transparent" :stroke "red")
(svg-path svg '((moveto ((200 . 250)))
                (elliptical-arc ((75 75 200 350 :large-arc t))))
          :fill "transparent" :stroke "green")
(svg-path svg '((moveto ((200 . 250)))
                (elliptical-arc ((75 75 200 350 :sweep t))))
          :fill "transparent" :stroke "blue")
(svg-path svg '((moveto ((200 . 250)))
                (elliptical-arc ((75 75 200 350 :large-arc t
                                     :sweep t))))
          :fill "transparent" :stroke "gray")
(svg-path svg '((moveto ((160 . 100)))
                (elliptical-arc ((40 100 80 0)))
                (elliptical-arc ((40 100 -40 -70
                                     :x-axis-rotation -120)))
                (elliptical-arc ((40 100 -40 70
                                     :x-axis-rotation -240))))
          :stroke "pink" :fill "lightblue"
          :relative t)

41.17.7 Other Image Types

For PBM images, specify image type pbm. Color, gray-scale and monochromatic images are supported. For mono PBM images, two additional image properties are supported.

:foreground foreground

The value, foreground, should be a string specifying the image foreground color, or nil for the default color. This color is used for each pixel in the PBM that is 1. The default is the frame’s foreground color.

:background background

The value, background, should be a string specifying the image background color, or nil for the default color. This color is used for each pixel in the PBM that is 0. The default is the frame’s background color.

The remaining image types that Emacs can support are:

GIF

Image type gif. Supports the :index property. See Multi-Frame Images.

JPEG

Image type jpeg.

PNG

Image type png.

TIFF

Image type tiff. Supports the :index property. See Multi-Frame Images.

WebP

Image type webp.


41.17.8 Defining Images

The functions create-image, defimage and find-image provide convenient ways to create image descriptors.

Function: create-image file-or-data &optional type data-p &rest props

This function creates and returns an image descriptor which uses the data in file-or-data. file-or-data can be a file name or a string containing the image data; data-p should be nil for the former case, non-nil for the latter case. If file-or-data is a relative file name, the function will search for it in directories mentioned in image-load-path.

The optional argument type is a symbol specifying the image type. If type is omitted or nil, create-image tries to determine the image type from the file’s first few bytes, or else from the file’s name.

The remaining arguments, props, specify additional image properties—for example,

(create-image "foo.xpm" 'xpm nil :heuristic-mask t)

The function returns nil if images of this type are not supported. Otherwise it returns an image descriptor.

Macro: defimage symbol specs &optional doc

This macro defines symbol as an image name. The arguments specs is a list which specifies how to display the image. The third argument, doc, is an optional documentation string.

Each argument in specs has the form of a property list, and each one should specify at least the :type property and either the :file or the :data property. The value of :type should be a symbol specifying the image type, the value of :file is the file to load the image from, and the value of :data is a string containing the actual image data. Here is an example:

(defimage test-image
  ((:type xpm :file "~/test1.xpm")
   (:type xbm :file "~/test1.xbm")))

defimage tests each argument, one by one, to see if it is usable—that is, if the type is supported and the file exists. The first usable argument is used to make an image descriptor which is stored in symbol.

If none of the alternatives will work, then symbol is defined as nil.

Function: image-property image property

Return the value of property in image. Properties can be set by using setf. Setting a property to nil will remove the property from the image.

Function: find-image specs

This function provides a convenient way to find an image satisfying one of a list of image specifications specs.

Each specification in specs is a property list with contents depending on image type. All specifications must at least contain the properties :type type and either :file file or :data data, where type is a symbol specifying the image type, e.g., xbm, file is the file to load the image from, and data is a string containing the actual image data. The first specification in the list whose type is supported, and file exists, is used to construct the image specification to be returned. If no specification is satisfied, nil is returned.

The image is looked for in image-load-path.

User Option: image-load-path

This variable’s value is a list of locations in which to search for image files. If an element is a string or a variable symbol whose value is a string, the string is taken to be the name of a directory to search. If an element is a variable symbol whose value is a list, that is taken to be a list of directories to search.

The default is to search in the images subdirectory of the directory specified by data-directory, then the directory specified by data-directory, and finally in the directories in load-path. Subdirectories are not automatically included in the search, so if you put an image file in a subdirectory, you have to supply the subdirectory explicitly. For example, to find the image images/foo/bar.xpm within data-directory, you should specify the image as follows:

(defimage foo-image '((:type xpm :file "foo/bar.xpm")))
Function: image-load-path-for-library library image &optional path no-error

This function returns a suitable search path for images used by the Lisp package library.

The function searches for image first using image-load-path, excluding data-directory/images, and then in load-path, followed by a path suitable for library, which includes ../../etc/images and ../etc/images relative to the library file itself, and finally in data-directory/images.

Then this function returns a list of directories which contains first the directory in which image was found, followed by the value of load-path. If path is given, it is used instead of load-path.

If no-error is non-nil and a suitable path can’t be found, don’t signal an error. Instead, return a list of directories as before, except that nil appears in place of the image directory.

Here is an example of using image-load-path-for-library:

(defvar image-load-path) ; shush compiler
(let* ((load-path (image-load-path-for-library
                    "mh-e" "mh-logo.xpm"))
       (image-load-path (cons (car load-path)
                              image-load-path)))
  (mh-tool-bar-folder-buttons-init))

Images are automatically scaled when created based on the image-scaling-factor variable. The value is either a floating point number (where numbers higher than 1 means to increase the size and lower means to shrink the size), or the symbol auto, which will compute a scaling factor based on the font pixel size.


41.17.9 Showing Images

You can use an image descriptor by setting up the display property yourself, but it is easier to use the functions in this section.

Function: insert-image image &optional string area slice inhibit-isearch

This function inserts image in the current buffer at point. The value image should be an image descriptor; it could be a value returned by create-image, or the value of a symbol defined with defimage. The argument string specifies the text to put in the buffer to hold the image. If it is omitted or nil, insert-image uses " " by default.

The argument area specifies whether to put the image in a margin. If it is left-margin, the image appears in the left margin; right-margin specifies the right margin. If area is nil or omitted, the image is displayed at point within the buffer’s text.

The argument slice specifies a slice of the image to insert. If slice is nil or omitted the whole image is inserted. (However, note that images are chopped on display at the window’s right edge, because wrapping images is not supported.) Otherwise, slice is a list (x y width height) which specifies the x and y positions and width and height of the image area to insert. Integer values are in units of pixels. A floating-point number in the range 0.0–1.0 stands for that fraction of the width or height of the entire image.

Internally, this function inserts string in the buffer, and gives it a display property which specifies image. See The display Property. By default, doing interactive searches in the buffer will consider string when searching. If inhibit-isearch is non-nil, this is inhibited.

Function: insert-sliced-image image &optional string area rows cols

This function inserts image in the current buffer at point, like insert-image, but splits the image into rowsxcols equally sized slices.

Emacs displays each slice as a separate image, and allows more intuitive scrolling up/down, instead of jumping up/down the entire image when paging through a buffer that displays (large) images.

Function: put-image image pos &optional string area

This function puts image image in front of pos in the current buffer. The argument pos should be an integer or a marker. It specifies the buffer position where the image should appear. The argument string specifies the text that should hold the image as an alternative to the default ‘x’.

The argument image must be an image descriptor, perhaps returned by create-image or stored by defimage.

The argument area specifies whether to put the image in a margin. If it is left-margin, the image appears in the left margin; right-margin specifies the right margin. If area is nil or omitted, the image is displayed at point within the buffer’s text.

Internally, this function creates an overlay, and gives it a before-string property containing text that has a display property whose value is the image. (Whew! that was a mouthful…)

Function: remove-images start end &optional buffer

This function removes images in buffer between positions start and end. If buffer is omitted or nil, images are removed from the current buffer.

This removes only images that were put into buffer the way put-image does it, not images that were inserted with insert-image or in other ways.

Function: image-size spec &optional pixels frame

This function returns the size of an image as a pair (width . height). spec is an image specification. pixels non-nil means return sizes measured in pixels, otherwise return sizes measured in the default character size of frame (see Frame Font). frame is the frame on which the image will be displayed. frame nil or omitted means use the selected frame (see Input Focus).

Variable: max-image-size

This variable is used to define the maximum size of image that Emacs will load. Emacs will refuse to load (and display) any image that is larger than this limit.

If the value is an integer, it directly specifies the maximum image height and width, measured in pixels. If it is floating point, it specifies the maximum image height and width as a ratio to the frame height and width. If the value is non-numeric, there is no explicit limit on the size of images.

The purpose of this variable is to prevent unreasonably large images from accidentally being loaded into Emacs. It only takes effect the first time an image is loaded. Once an image is placed in the image cache, it can always be displayed, even if the value of max-image-size is subsequently changed (see Image Cache).

Function: image-at-point-p

This function returns t if point is on an image, and nil otherwise.

Images inserted with the insertion functions above also get a local keymap installed in the text properties (or overlays) that span the displayed image. This keymap defines the following commands:

i +

Increase the image size (image-increase-size)

i -

Decrease the image size (image-decrease-size).

i r

Rotate the image (image-rotate).

i h

Flip the image horizontally (image-flip-horizontally).

i v

Flip the image vertically (image-flip-vertically).

i o

Save the image to a file (image-save).

i c

Interactively crop the image (image-crop).

i x

Interactively cut a rectangle from the image (image-cut).

See Image Mode in The GNU Emacs Manual, for more details about these image-specific key bindings.


41.17.10 Multi-Frame Images

Some image files can contain more than one image. We say that there are multiple “frames” in the image. At present, Emacs supports multiple frames for GIF, TIFF, and certain ImageMagick formats such as DJVM.

The frames can be used either to represent multiple pages (this is usually the case with multi-frame TIFF files, for example), or to create animation (usually the case with multi-frame GIF files).

A multi-frame image has a property :index, whose value is an integer (counting from 0) that specifies which frame is being displayed.

Function: image-multi-frame-p image

This function returns non-nil if image contains more than one frame. The actual return value is a cons (nimages . delay), where nimages is the number of frames and delay is the delay in seconds between them, or nil if the image does not specify a delay. Images that are intended to be animated usually specify a frame delay, whereas ones that are intended to be treated as multiple pages do not.

Function: image-current-frame image

This function returns the index of the current frame number for image, counting from 0.

Function: image-show-frame image n &optional nocheck

This function switches image to frame number n. It replaces a frame number outside the valid range with that of the end of the range, unless nocheck is non-nil. If image does not contain a frame with the specified number, the image displays as a hollow box.

Function: image-animate image &optional index limit

This function animates image. The optional integer index specifies the frame from which to start (default 0). The optional argument limit controls the length of the animation. If omitted or nil, the image animates once only; if t it loops forever; if a number animation stops after that many seconds.

Animation operates by means of a timer. Note that Emacs imposes a minimum frame delay of 0.01 (image-minimum-frame-delay) seconds. If the image itself does not specify a delay, Emacs uses image-default-frame-delay.

Function: image-animate-timer image

This function returns the timer responsible for animating image, if there is one.


41.17.11 Image Cache

Emacs caches images so that it can display them again more efficiently. When Emacs displays an image, it searches the image cache for an existing image specification equal to the desired specification. If a match is found, the image is displayed from the cache. Otherwise, Emacs loads the image normally.

Function: image-flush spec &optional frame

This function removes the image with specification spec from the image cache of frame frame. Image specifications are compared using equal. If frame is nil, it defaults to the selected frame. If frame is t, the image is flushed on all existing frames.

In Emacs’s current implementation, each graphical terminal possesses an image cache, which is shared by all the frames on that terminal (see Multiple Terminals). Thus, refreshing an image in one frame also refreshes it in all other frames on the same terminal.

One use for image-flush is to tell Emacs about a change in an image file. If an image specification contains a :file property, the image is cached based on the file’s contents when the image is first displayed. Even if the file subsequently changes, Emacs continues displaying the old version of the image. Calling image-flush flushes the image from the cache, forcing Emacs to re-read the file the next time it needs to display that image.

Another use for image-flush is for memory conservation. If your Lisp program creates a large number of temporary images over a period much shorter than image-cache-eviction-delay (see below), you can opt to flush unused images yourself, instead of waiting for Emacs to do it automatically.

Function: clear-image-cache &optional filter

This function clears an image cache, removing all the images stored in it. If filter is omitted or nil, it clears the cache for the selected frame. If filter is a frame, it clears the cache for that frame. If filter is t, all image caches are cleared. Otherwise, filter is taken to be a file name, and all images associated with that file name are removed from all image caches.

If an image in the image cache has not been displayed for a specified period of time, Emacs removes it from the cache and frees the associated memory.

Variable: image-cache-eviction-delay

This variable specifies the number of seconds an image can remain in the cache without being displayed. When an image is not displayed for this length of time, Emacs removes it from the image cache.

Under some circumstances, if the number of images in the cache grows too large, the actual eviction delay may be shorter than this.

If the value is nil, Emacs does not remove images from the cache except when you explicitly clear it. This mode can be useful for debugging.

Function: image-cache-size

This function returns the total size of the current image cache, in bytes. An image of size 200x100 with 24 bits per color will have a cache size of 60000 bytes, for instance.


41.18 Icons

Emacs sometimes uses buttons (for clicking on) or small graphics (to illustrate something). Since Emacs is available on a wide variety of systems with different capabilities, and users have different preferences, Emacs provides a facility to handle this in a convenient way, allowing customization, graceful degradation, accessibility, as well as themability: Icons.

The central macro here is define-icon, and here’s a simple example:

(define-icon outline-open button
  '((image "right.svg" "open.xpm" "open.pbm" :height line)
    (emoji "▶️")
    (symbol "▶" "➤")
    (text "open" :face icon-button))
  "Icon used for buttons for opening a section in outline buffers."
  :version "29.1"
  :help-echo "Open this section")

Which alternative will actually be displayed depends on the value of the user option icon-preference (see Icons in The GNU Emacs Manual) and on the results of run-time checks for what the current frame’s terminal can actually display.

The macro in the example above defines outline-open as an icon, and inherits properties from the icon called button (so this is meant as a clickable button to be inserted in a buffer). It is followed by a list of icon types along with the actual icon shapes themselves. In addition, there’s a doc string and various keywords that contain additional information and properties.

To instantiate an icon, you use icon-string, which will consult the current Customize theming, and the icon-preference user option, and finally what the Emacs is able to actually display. If icon-preference is (image emoji symbol text) (i.e., allowing all of these forms of icons), in this case, icon-string will first check that Emacs is able to display images at all, and then whether it has support for each of those different image formats. If that fails, Emacs will check whether Emacs can display emojis (in the current frame). If that fails, it’ll check whether it can display the symbol in question. If that fails, it’ll use the plain text version.

For instance, if icon-preference doesn’t contain image or emoji, it’ll skip those entries.

Code can confidently call icon-string in all circumstances and be sure that something readable will appear on the screen, no matter whether the user is on a graphical terminal or a text terminal, and no matter which features Emacs was built with.

Macro: define-icon name parent specs doc &rest keywords

Define an icon name, a symbol, with the display alternatives in spec, that can be later instantiated using icon-string. The name is the name of the resulting keyword.

The resulting icon will inherit specs from parent, and from their parent’s parents, and so on, and the lowest descendent element wins.

specs is a list of icon specifications. The first element of each specification is the type, and the rest is something that can be used as an icon of that type, and then optionally followed by a keyword list. The following icon types are available:

image

In this case, there may be many images listed as candidates. Emacs will choose the first one that the current Emacs instance can show. If an image is listed is an absolute file name, it’s used as is, but it’s otherwise looked up in the list image-load-path (see Defining Images).

emoji

This should be a (possibly colorful) emoji.

symbol

This should be a (monochrome) symbol character.

text

Icons should also have a textual fallback. This can also be used for the visually impaired: if icon-preference is just (text), all icons will be replaced by text.

Various keywords may follow the list of icon specifications. For instance:

(symbol "▶" "➤" :face icon-button)

Unknown keywords are ignored. The following keywords are allowed:

:face

The face to be used for the icon.

:height

This is only valid for image icons, and can be either a number (which specifies the height in pixels), or the symbol line, which will use the default line height in the currently selected window.

:width

This is only valid for image icons, and can be either a number (which specifies the width in pixels), or the symbol font, which will use the width in pixels of the current buffer’s default face font.

doc should be a doc string.

keywords is a list of keyword/value pairs. The following keywords are allowed:

:version

The (approximate) Emacs version this button first appeared. (This keyword is mandatory.)

:group

The customization group this icon belongs in. If not present, it is inferred.

:help-echo

The help string shown when hovering over the icon with the mouse pointer.

Function: icon-string icon

This function returns a string suitable for display in the current buffer for icon.

Function: icon-elements icon

Alternatively, you can get a “deconstructed” version of icon with this function. It returns a plist (see Property Lists) where the keys are string, face and image. (The latter is only present if the icon is represented by an image.) This can be useful if the icon isn’t to be inserted directly in the buffer, but needs some sort of pre-processing first.

Icons can be customized with M-x customize-icon. Themes can specify changes to icons with, for instance:

(custom-theme-set-icons
  'my-theme
  '(outline-open ((image :height 100)
                  (text " OPEN ")))
  '(outline-close ((image :height 100)
                   (text " CLOSE " :face warning))))

41.19 Embedded Native Widgets

Emacs is able to display native widgets, such as GTK+ WebKit widgets, in Emacs buffers when it was built with the necessary support libraries and is running on a graphical terminal. To test whether Emacs supports display of embedded widgets, check that the xwidget-internal feature is available (see Features).

To display an embedded widget in a buffer, you must first create an xwidget object, and then use that object as the display specifier in a display text or overlay property (see The display Property).

Embedded widgets can send events notifying Lisp code about changes occurring within them. (see Xwidget events).

Function: make-xwidget type title width height arguments &optional buffer related

This creates and returns an xwidget object. If buffer is omitted or nil, it defaults to the current buffer. If buffer names a buffer that doesn’t exist, it will be created. The type identifies the type of the xwidget component, it can be one of the following:

webkit

The WebKit component.

The width and height arguments specify the widget size in pixels, and title, a string, specifies its title. related is used internally by the WebKit widget, and specifies another WebKit widget that the newly created widget should share settings and subprocesses with.

The xwidget that is returned will be killed alongside its buffer (see Killing Buffers). You can also kill it using kill-xwidget. Once it is killed, the xwidget may continue to exist as a Lisp object and act as a display property until all references to it are gone, but most actions that can be performed on live xwidgets will no longer be available.

Function: xwidgetp object

This function returns t if object is an xwidget, nil otherwise.

Function: xwidget-live-p object

This function returns t if object is an xwidget that hasn’t been killed, and nil otherwise.

Function: kill-xwidget xwidget

This function kills xwidget, by removing it from its buffer and releasing window system resources it holds.

Function: xwidget-plist xwidget

This function returns the property list of xwidget.

Function: set-xwidget-plist xwidget plist

This function replaces the property list of xwidget with a new property list given by plist.

Function: xwidget-buffer xwidget

This function returns the buffer of xwidget. If xwidget has been killed, it returns nil.

Function: set-xwidget-buffer xwidget buffer

This function sets the buffer of xwidget to buffer.

Function: get-buffer-xwidgets buffer

This function returns a list of xwidget objects associated with the buffer, which can be specified as a buffer object or a name of an existing buffer, a string. The value is nil if buffer contains no xwidgets.

Function: xwidget-webkit-goto-uri xwidget uri

This function browses the specified uri in the given xwidget. The uri is a string that specifies the name of a file or a URL.

Function: xwidget-webkit-execute-script xwidget script

This function causes the browser widget specified by xwidget to execute the specified JavaScript script.

Function: xwidget-webkit-execute-script-rv xwidget script &optional default

This function executes the specified script like xwidget-webkit-execute-script does, but it also returns the script’s return value as a string. If script doesn’t return a value, this function returns default, or nil if default was omitted.

Function: xwidget-webkit-get-title xwidget

This function returns the title of xwidget as a string.

Function: xwidget-resize xwidget width height

This function resizes the specified xwidget to the size widthxheight pixels.

Function: xwidget-size-request xwidget

This function returns the desired size of xwidget as a list of the form (width height). The dimensions are in pixels.

Function: xwidget-info xwidget

This function returns the attributes of xwidget as a vector of the form [type title width height]. The attributes are usually determined by make-xwidget when the xwidget is created.

Function: set-xwidget-query-on-exit-flag xwidget flag

This function allows you to arrange that Emacs will ask the user for confirmation before exiting or before killing a buffer that has xwidget associated with it. If flag is non-nil, Emacs will query the user, otherwise it will not.

Function: xwidget-query-on-exit-flag xwidget

This function returns the current setting of xwidgets query-on-exit flag, either t or nil.

Function: xwidget-perform-lispy-event xwidget event frame

Send an input event event to xwidget. The precise action performed is platform-specific. See Input Events.

You can optionally pass the frame on which the event was generated via frame. On X11, modifier keys in key events will not be considered if frame is nil, and the selected frame is not an X-Windows frame.

On GTK, only keyboard and function key events are supported. Mouse, motion, and click events are dispatched to the xwidget without going through Lisp code, and as such shouldn’t require this function to be called.

Function: xwidget-webkit-search query xwidget &optional case-insensitive backwards wrap-around

Start an incremental search on the WebKit widget xwidget with the string query as the query. case-insensitive denotes whether or not the search is case-insensitive, backwards determines if the search is performed backwards towards the start of the document, and wrap-around determines whether or not the search terminates at the end of the document.

If the function is called while a search query is already present, then the query specified here will replace the existing query.

To stop a search query, use xwidget-webkit-finish-search.

Function: xwidget-webkit-next-result xwidget

Display the next search result in xwidget. This function will signal an error if a search query has not been already started in xwidget through xwidget-webkit-search.

If wrap-around was non-nil when xwidget-webkit-search was called, then the search will restart from the beginning of the document when its end is reached.

Function: xwidget-webkit-previous-result xwidget

Display the previous search result in xwidget. This function signals an error if a search query has not been already started in xwidget through xwidget-webkit-search.

If wrap-around was non-nil when xwidget-webkit-search was called, then the search will restart from the end of the document when its beginning is reached.

Function: xwidget-webkit-finish-search xwidget

Finish a search operation started with xwidget-webkit-search in xwidget. If there is no query currently ongoing, this function signals an error.

Function: xwidget-webkit-load-html xwidget text &optional base-uri

Load text, a string, into xwidget, which should be a WebKit xwidget. Any HTML markup in text will be processed by xwidget while rendering the text.

Optional argument base-uri, which should be a string, specifies the absolute location of the web resources referenced by text, to be used for resolving relative links in text.

Function: xwidget-webkit-goto-history xwidget rel-pos

Make xwidget, a WebKit widget, load the rel-posth element in its navigation history.

If rel-pos is zero, the current page will be reloaded instead.

Function: xwidget-webkit-back-forward-list xwidget &optional limit

Return the navigation history of xwidget, up to limit items in each direction. If not specified, limit defaults to 50.

The returned value is a list of the form (back here forward), where here is the current navigation item, while back is a list of items containing the items recorded by WebKit before the current navigation item, and forward is a list of items recorded after the current navigation item. back, here and forward can all be nil.

When here is nil, it means that no items have been recorded yet; if back or forward are nil, it means that there is no history recorded before or after the current item respectively.

Navigation items are themselves lists of the form (idx title uri). In these lists, idx is an index that can be passed to xwidget-webkit-goto-history, title is the human-readable title of the item, and uri is the URI of the item. The user should normally have no reason to load uri manually to reach a specific history item. Instead, idx should be passed as an index to xwidget-webkit-goto-history.

Function: xwidget-webkit-estimated-load-progress xwidget

Return an estimate of how much data is remaining to be transferred before the page displayed by the WebKit widget xwidget is fully loaded.

The value returned is a float ranging between 0.0 and 1.0.

Function: xwidget-webkit-set-cookie-storage-file xwidget file

Make the WebKit widget xwidget store cookies in file.

file must be an absolute file name. The new setting will also affect any xwidget that was created with xwidget as the related argument to make-xwidget, and widgets related to those as well.

If this function is not called at least once on xwidget or a related widget, xwidget will not store cookies on disk at all.

Function: xwidget-webkit-stop-loading xwidget

Terminate any data transfer still in progress in the WebKit widget xwidget as part of a page-loading operation. If a page is not being loaded, this function does nothing.


41.20 Buttons

The Button package defines functions for inserting and manipulating buttons that can be activated with the mouse or via keyboard commands. These buttons are typically used for various kinds of hyperlinks.

A button is essentially a set of text or overlay properties, attached to a stretch of text in a buffer. These properties are called button properties. One of these properties, the action property, specifies a function which is called when the user invokes the button using the keyboard or the mouse. The action function may examine the button and use its other properties as desired.

In some ways, the Button package duplicates the functionality in the Widget package. See Introduction in The Emacs Widget Library. The advantage of the Button package is that it is faster, smaller, and simpler to program. From the point of view of the user, the interfaces produced by the two packages are very similar.


41.20.1 Button Properties

Each button has an associated list of properties defining its appearance and behavior, and other arbitrary properties may be used for application specific purposes. The following properties have special meaning to the Button package:

action

The function to call when the user invokes the button, which is passed the single argument button. By default this is ignore, which does nothing.

mouse-action

This is similar to action, and when present, will be used instead of action for button invocations resulting from mouse-clicks (instead of the user hitting RET). If not present, mouse-clicks use action instead.

face

This is an Emacs face controlling how buttons of this type are displayed; by default this is the button face.

mouse-face

This is an additional face which controls appearance during mouse-overs (merged with the usual button face); by default this is the usual Emacs highlight face.

keymap

The button’s keymap, defining bindings active within the button region. By default this is the usual button region keymap, stored in the variable button-map, which defines RET and mouse-2 to invoke the button.

type

The button type. See Button Types.

help-echo

A string displayed by the Emacs tooltip help system; by default, "mouse-2, RET: Push this button". Alternatively, a function that returns, or a form that evaluates to, a string to be displayed or nil. For details see Text help-echo.

The function is called with three arguments, window, object, and pos. The second argument, object, is either the overlay that had the property (for overlay buttons), or the buffer containing the button (for text property buttons). The other arguments have the same meaning as for the special text property help-echo.

The follow-link property, defining how a mouse-1 click behaves on this button, See Defining Clickable Text.

button

All buttons have a non-nil button property, which may be useful in finding regions of text that comprise buttons (which is what the standard button functions do).

There are other properties defined for the regions of text in a button, but these are not generally interesting for typical uses.


41.20.2 Button Types

Every button has a button type, which defines default values for the button’s properties. Button types are arranged in a hierarchy, with specialized types inheriting from more general types, so that it’s easy to define special-purpose types of buttons for specific tasks.

Function: define-button-type name &rest properties

Define a button type called name (a symbol). The remaining arguments form a sequence of property value pairs, specifying default property values for buttons with this type (a button’s type may be set by giving it a type property when creating the button, using the :type keyword argument).

In addition, the keyword argument :supertype may be used to specify a button-type from which name inherits its default property values. Note that this inheritance happens only when name is defined; subsequent changes to a supertype are not reflected in its subtypes.

Using define-button-type to define default properties for buttons is not necessary—buttons without any specified type use the built-in button-type button—but it is encouraged, since doing so usually makes the resulting code clearer and more efficient.


41.20.3 Making Buttons

Buttons are associated with a region of text, using an overlay or text properties to hold button-specific information, all of which are initialized from the button’s type (which defaults to the built-in button type button). Like all Emacs text, the appearance of the button is governed by the face property; by default (via the face property inherited from the button button-type) this is a simple underline, like a typical web-page link.

For convenience, there are two sorts of button-creation functions, those that add button properties to an existing region of a buffer, called make-...button, and those that also insert the button text, called insert-...button.

The button-creation functions all take the &rest argument properties, which should be a sequence of property value pairs, specifying properties to add to the button; see Button Properties. In addition, the keyword argument :type may be used to specify a button-type from which to inherit other properties; see Button Types. Any properties not explicitly specified during creation will be inherited from the button’s type (if the type defines such a property).

The following functions add a button using an overlay (see Overlays) to hold the button properties:

Function: make-button beg end &rest properties

This makes a button from beg to end in the current buffer, and returns it.

Function: insert-button label &rest properties

This inserts a button with the label label at point, and returns it.

The following functions are similar, but using text properties (see Text Properties) to hold the button properties. Such buttons do not add markers to the buffer, so editing in the buffer does not slow down if there is an extremely large numbers of buttons. However, if there is an existing face text property on the text (e.g., a face assigned by Font Lock mode), the button face may not be visible. Both of these functions return the starting position of the new button.

Function: make-text-button beg end &rest properties

This makes a button from beg to end in the current buffer, using text properties.

Function: insert-text-button label &rest properties

This inserts a button with the label label at point, using text properties.

Function: buttonize string callback &optional data

Sometimes it’s more convenient to make a string into a button without inserting it into a buffer immediately, for instance when creating data structures that may then, later, be inserted into a buffer. This function makes string into such a string, and callback will be called when the user clicks on the button. The optional data parameter will be used as the parameter when callback is called. If nil, the button is used as the parameter instead.


41.20.4 Manipulating Buttons

These are functions for getting and setting properties of buttons. Often these are used by a button’s invocation function to determine what to do.

Where a button parameter is specified, it means an object referring to a specific button, either an overlay (for overlay buttons), or a buffer-position or marker (for text property buttons). Such an object is passed as the first argument to a button’s invocation function when it is invoked.

Function: button-start button

Return the position at which button starts.

Function: button-end button

Return the position at which button ends.

Function: button-get button prop

Get the property of button button named prop.

Function: button-put button prop val

Set button’s prop property to val.

Function: button-activate button &optional use-mouse-action

Call button’s action property (i.e., invoke the function that is the value of that property, passing it the single argument button). If use-mouse-action is non-nil, try to invoke the button’s mouse-action property instead of action; if the button has no mouse-action property, use action as normal. If the button-data property is present in button, use that as the argument for the action function instead of button.

Function: button-label button

Return button’s text label.

Function: button-type button

Return button’s button-type.

Function: button-has-type-p button type

Return t if button has button-type type, or one of type’s subtypes.

Function: button-at pos

Return the button at position pos in the current buffer, or nil. If the button at pos is a text property button, the return value is a marker pointing to pos.

Function: button-type-put type prop val

Set the button-type type’s prop property to val.

Function: button-type-get type prop

Get the property of button-type type named prop.

Function: button-type-subtype-p type supertype

Return t if button-type type is a subtype of supertype.


41.20.5 Button Buffer Commands

These are commands and functions for locating and operating on buttons in an Emacs buffer.

push-button is the command that a user uses to actually push a button, and is bound by default in the button itself to RET and to mouse-2 using a local keymap in the button’s overlay or text properties. Commands that are useful outside the buttons itself, such as forward-button and backward-button are additionally available in the keymap stored in button-buffer-map; a mode which uses buttons may want to use button-buffer-map as a parent keymap for its keymap. Alternatively, the button-mode can be switched on for much the same effect: It’s a minor mode that does nothing else than install button-buffer-map as a minor mode keymap.

If the button has a non-nil follow-link property, and mouse-1-click-follows-link is set, a quick mouse-1 click will also activate the push-button command. See Defining Clickable Text.

Command: push-button &optional pos use-mouse-action

Perform the action specified by a button at location pos. pos may be either a buffer position or a mouse-event. If use-mouse-action is non-nil, or pos is a mouse-event (see Mouse Events), try to invoke the button’s mouse-action property instead of action; if the button has no mouse-action property, use action as normal. pos defaults to point, except when push-button is invoked interactively as the result of a mouse-event, in which case, the mouse event’s position is used. If there’s no button at pos, do nothing and return nil, otherwise return t.

Command: forward-button n &optional wrap display-message no-error

Move to the nth next button, or nth previous button if n is negative. If n is zero, move to the start of any button at point. If wrap is non-nil, moving past either end of the buffer continues from the other end. If display-message is non-nil, the button’s help-echo string is displayed. Any button with a non-nil skip property is skipped over. Returns the button found, and signals an error if no buttons can be found. If no-error is non-nil, return nil instead of signaling the error.

Command: backward-button n &optional wrap display-message no-error

Move to the nth previous button, or nth next button if n is negative. If n is zero, move to the start of any button at point. If wrap is non-nil, moving past either end of the buffer continues from the other end. If display-message is non-nil, the button’s help-echo string is displayed. Any button with a non-nil skip property is skipped over. Returns the button found, and signals an error if no buttons can be found. If no-error is non-nil, return nil instead of signaling the error.

Function: next-button pos &optional count-current
Function: previous-button pos &optional count-current

Return the next button after (for next-button) or before (for previous-button) position pos in the current buffer. If count-current is non-nil, count any button at pos in the search, instead of starting at the next button.


41.21 Abstract Display

The Ewoc package constructs buffer text that represents a structure of Lisp objects, and updates the text to follow changes in that structure. This is like the “view” component in the “model–view–controller” design paradigm. Ewoc means “Emacs’s Widget for Object Collections”.

An ewoc is a structure that organizes information required to construct buffer text that represents certain Lisp data. The buffer text of the ewoc has three parts, in order: first, fixed header text; next, textual descriptions of a series of data elements (Lisp objects that you specify); and last, fixed footer text. Specifically, an ewoc contains information on:

  • The buffer which its text is generated in.
  • The text’s start position in the buffer.
  • The header and footer strings.
  • A doubly-linked chain of nodes, each of which contains:
    • A data element, a single Lisp object.
    • Links to the preceding and following nodes in the chain.
  • A pretty-printer function which is responsible for inserting the textual representation of a data element value into the current buffer.

Typically, you define an ewoc with ewoc-create, and then pass the resulting ewoc structure to other functions in the Ewoc package to build nodes within it, and display it in the buffer. Once it is displayed in the buffer, other functions determine the correspondence between buffer positions and nodes, move point from one node’s textual representation to another, and so forth. See Abstract Display Functions.

A node encapsulates a data element much the way a variable holds a value. Normally, encapsulation occurs as a part of adding a node to the ewoc. You can retrieve the data element value and place a new value in its place, like so:

(ewoc-data node)
⇒ value

(ewoc-set-data node new-value)
⇒ new-value

You can also use, as the data element value, a Lisp object (list or vector) that is a container for the real value, or an index into some other structure. The example (see Abstract Display Example) uses the latter approach.

When the data changes, you will want to update the text in the buffer. You can update all nodes by calling ewoc-refresh, or just specific nodes using ewoc-invalidate, or all nodes satisfying a predicate using ewoc-map. Alternatively, you can delete invalid nodes using ewoc-delete or ewoc-filter, and add new nodes in their place. Deleting a node from an ewoc deletes its associated textual description from buffer, as well.


41.21.1 Abstract Display Functions

In this subsection, ewoc and node stand for the structures described above (see Abstract Display), while data stands for an arbitrary Lisp object used as a data element.

Function: ewoc-create pretty-printer &optional header footer nosep

This constructs and returns a new ewoc, with no nodes (and thus no data elements). pretty-printer should be a function that takes one argument, a data element of the sort you plan to use in this ewoc, and inserts its textual description at point using insert (and never insert-before-markers, because that would interfere with the Ewoc package’s internal mechanisms).

Normally, a newline is automatically inserted after the header, the footer and every node’s textual description. If nosep is non-nil, no newline is inserted. This may be useful for displaying an entire ewoc on a single line, for example, or for making nodes invisible by arranging for pretty-printer to do nothing for those nodes.

An ewoc maintains its text in the buffer that is current when you create it, so switch to the intended buffer before calling ewoc-create.

Function: ewoc-buffer ewoc

This returns the buffer where ewoc maintains its text.

Function: ewoc-get-hf ewoc

This returns a cons cell (header . footer) made from ewoc’s header and footer.

Function: ewoc-set-hf ewoc header footer

This sets the header and footer of ewoc to the strings header and footer, respectively.

Function: ewoc-enter-first ewoc data
Function: ewoc-enter-last ewoc data

These add a new node encapsulating data, putting it, respectively, at the beginning or end of ewoc’s chain of nodes.

Function: ewoc-enter-before ewoc node data
Function: ewoc-enter-after ewoc node data

These add a new node encapsulating data, adding it to ewoc before or after node, respectively.

Function: ewoc-prev ewoc node
Function: ewoc-next ewoc node

These return, respectively, the previous node and the next node of node in ewoc.

Function: ewoc-nth ewoc n

This returns the node in ewoc found at zero-based index n. A negative n means count from the end. ewoc-nth returns nil if n is out of range.

Function: ewoc-data node

This extracts the data encapsulated by node and returns it.

Function: ewoc-set-data node data

This sets the data encapsulated by node to data.

Function: ewoc-locate ewoc &optional pos guess

This determines the node in ewoc which contains point (or pos if specified), and returns that node. If ewoc has no nodes, it returns nil. If pos is before the first node, it returns the first node; if pos is after the last node, it returns the last node. The optional third arg guess should be a node that is likely to be near pos; this doesn’t alter the result, but makes the function run faster.

Function: ewoc-location node

This returns the start position of node.

Function: ewoc-goto-prev ewoc arg
Function: ewoc-goto-next ewoc arg

These move point to the previous or next, respectively, argth node in ewoc. ewoc-goto-prev does not move if it is already at the first node or if ewoc is empty, whereas ewoc-goto-next moves past the last node, returning nil. Excepting this special case, these functions return the node moved to.

Function: ewoc-goto-node ewoc node

This moves point to the start of node in ewoc.

Function: ewoc-refresh ewoc

This function regenerates the text of ewoc. It works by deleting the text between the header and the footer, i.e., all the data elements’ representations, and then calling the pretty-printer function for each node, one by one, in order.

Function: ewoc-invalidate ewoc &rest nodes

This is similar to ewoc-refresh, except that only nodes in ewoc are updated instead of the entire set.

Function: ewoc-delete ewoc &rest nodes

This deletes each node in nodes from ewoc.

Function: ewoc-filter ewoc predicate &rest args

This calls predicate for each data element in ewoc and deletes those nodes for which predicate returns nil. Any args are passed to predicate.

Function: ewoc-collect ewoc predicate &rest args

This calls predicate for each data element in ewoc and returns a list of those elements for which predicate returns non-nil. The elements in the list are ordered as in the buffer. Any args are passed to predicate.

Function: ewoc-map map-function ewoc &rest args

This calls map-function for each data element in ewoc and updates those nodes for which map-function returns non-nil. Any args are passed to map-function.


41.21.2 Abstract Display Example

Here is a simple example using functions of the ewoc package to implement a color components display, an area in a buffer that represents a vector of three integers (itself representing a 24-bit RGB value) in various ways.

(setq colorcomp-ewoc nil
      colorcomp-data nil
      colorcomp-mode-map nil
      colorcomp-labels ["Red" "Green" "Blue"])

(defun colorcomp-pp (data)
  (if data
      (let ((comp (aref colorcomp-data data)))
        (insert (aref colorcomp-labels data) "\t: #x"
                (format "%02X" comp) " "
                (make-string (ash comp -2) ?#) "\n"))
    (let ((cstr (format "#%02X%02X%02X"
                        (aref colorcomp-data 0)
                        (aref colorcomp-data 1)
                        (aref colorcomp-data 2)))
          (samp " (sample text) "))
      (insert "Color\t: "
              (propertize samp 'face
                          `(foreground-color . ,cstr))
              (propertize samp 'face
                          `(background-color . ,cstr))
              "\n"))))

(defun colorcomp (color)
  "Allow fiddling with COLOR in a new buffer.
The buffer is in Color Components mode."
  (interactive "sColor (name or #RGB or #RRGGBB): ")
  (when (string= "" color)
    (setq color "green"))
  (unless (color-values color)
    (error "No such color: %S" color))
  (switch-to-buffer
   (generate-new-buffer (format "originally: %s" color)))
  (kill-all-local-variables)
  (setq major-mode 'colorcomp-mode
        mode-name "Color Components")
  (use-local-map colorcomp-mode-map)
  (erase-buffer)
  (buffer-disable-undo)
  (let ((data (apply 'vector (mapcar (lambda (n) (ash n -8))
                                     (color-values color))))
        (ewoc (ewoc-create 'colorcomp-pp
                           "\nColor Components\n\n"
                           (substitute-command-keys
                            "\n\\{colorcomp-mode-map}"))))
    (set (make-local-variable 'colorcomp-data) data)
    (set (make-local-variable 'colorcomp-ewoc) ewoc)
    (ewoc-enter-last ewoc 0)
    (ewoc-enter-last ewoc 1)
    (ewoc-enter-last ewoc 2)
    (ewoc-enter-last ewoc nil)))

This example can be extended to be a color selection widget (in other words, the “controller” part of the “model–view–controller” design paradigm) by defining commands to modify colorcomp-data and to finish the selection process, and a keymap to tie it all together conveniently.

(defun colorcomp-mod (index limit delta)
  (let ((cur (aref colorcomp-data index)))
    (unless (= limit cur)
      (aset colorcomp-data index (+ cur delta)))
    (ewoc-invalidate
     colorcomp-ewoc
     (ewoc-nth colorcomp-ewoc index)
     (ewoc-nth colorcomp-ewoc -1))))

(defun colorcomp-R-more () (interactive) (colorcomp-mod 0 255 1))
(defun colorcomp-G-more () (interactive) (colorcomp-mod 1 255 1))
(defun colorcomp-B-more () (interactive) (colorcomp-mod 2 255 1))
(defun colorcomp-R-less () (interactive) (colorcomp-mod 0 0 -1))
(defun colorcomp-G-less () (interactive) (colorcomp-mod 1 0 -1))
(defun colorcomp-B-less () (interactive) (colorcomp-mod 2 0 -1))

(defun colorcomp-copy-as-kill-and-exit ()
  "Copy the color components into the kill ring and kill the buffer.
The string is formatted #RRGGBB (hash followed by six hex digits)."
  (interactive)
  (kill-new (format "#%02X%02X%02X"
                    (aref colorcomp-data 0)
                    (aref colorcomp-data 1)
                    (aref colorcomp-data 2)))
  (kill-buffer nil))

(setq colorcomp-mode-map
      (define-keymap :suppress t
        "i" 'colorcomp-R-less
        "o" 'colorcomp-R-more
        "k" 'colorcomp-G-less
        "l" 'colorcomp-G-more
        "," 'colorcomp-B-less
        "." 'colorcomp-B-more
        "SPC" 'colorcomp-copy-as-kill-and-exit))

Note that we never modify the data in each node, which is fixed when the ewoc is created to be either nil or an index into the vector colorcomp-data, the actual color components.


41.22 Blinking Parentheses

This section describes the mechanism by which Emacs shows a matching open parenthesis when the user inserts a close parenthesis.

The value of this variable should be a function (of no arguments) to be called whenever a character with close parenthesis syntax is inserted. The value of blink-paren-function may be nil, in which case nothing is done.

If this variable is nil, then blink-matching-open does nothing.

This variable specifies the maximum distance to scan for a matching parenthesis before giving up.

This variable specifies the number of seconds to keep indicating the matching parenthesis. A fraction of a second often gives good results, but the default is 1, which works on all systems.

This function is the default value of blink-paren-function. It assumes that point follows a character with close parenthesis syntax and applies the appropriate effect momentarily to the matching opening character. If that character is not already on the screen, it displays the character’s context in the echo area. To avoid long delays, this function does not search farther than blink-matching-paren-distance characters.

Here is an example of calling this function explicitly.

(defun interactive-blink-matching-open ()
  "Indicate momentarily the start of parenthesized sexp before point."
  (interactive)
  (let ((blink-matching-paren-distance
         (buffer-size))
        (blink-matching-paren t))
    (blink-matching-open)))

41.23 Character Display

This section describes how characters are actually displayed by Emacs. Typically, a character is displayed as a glyph (a graphical symbol which occupies one character position on the screen), whose appearance corresponds to the character itself. For example, the character ‘a’ (character code 97) is displayed as ‘a’. Some characters, however, are displayed specially. For example, the formfeed character (character code 12) is usually displayed as a sequence of two glyphs, ‘^L’, while the newline character (character code 10) starts a new screen line.

You can modify how each character is displayed by defining a display table, which maps each character code into a sequence of glyphs. See Display Tables.


41.23.1 Usual Display Conventions

Here are the conventions for displaying each character code (in the absence of a display table, which can override these conventions; see Display Tables).

  • The printable ASCII characters, character codes 32 through 126 (consisting of numerals, English letters, and symbols like ‘#’) are displayed literally.
  • The tab character (character code 9) displays as whitespace stretching up to the next tab stop column. See Text Display in The GNU Emacs Manual. The variable tab-width controls the number of spaces per tab stop (see below).
  • The newline character (character code 10) has a special effect: it ends the preceding line and starts a new line.
  • The non-printable ASCII control characters—character codes 0 through 31, as well as the DEL character (character code 127)—display in one of two ways according to the variable ctl-arrow. If this variable is non-nil (the default), these characters are displayed as sequences of two glyphs, where the first glyph is ‘^’ (a display table can specify a glyph to use instead of ‘^’); e.g., the DEL character is displayed as ‘^?’.

    If ctl-arrow is nil, these characters are displayed as octal escapes (see below).

    This rule also applies to carriage return (character code 13), if that character appears in the buffer. But carriage returns usually do not appear in buffer text; they are eliminated as part of end-of-line conversion (see Basic Concepts of Coding Systems).

  • Raw bytes are non-ASCII characters with codes 128 through 255 (see Text Representations). These characters display as octal escapes: sequences of four glyphs, where the first glyph is the ASCII code for ‘\’, and the others are digit characters representing the character code in octal. (A display table can specify a glyph to use instead of ‘\’.)
  • Each non-ASCII character with code above 255 is displayed literally, if the terminal supports it. If the terminal does not support it, the character is said to be glyphless, and it is usually displayed using a placeholder glyph. For example, if a graphical terminal has no font for a character, Emacs usually displays a box containing the character code in hexadecimal. See Glyphless Character Display.

The above display conventions apply even when there is a display table, for any character whose entry in the active display table is nil. Thus, when you set up a display table, you need only specify the characters for which you want special display behavior.

The following variables affect how certain characters are displayed on the screen. Since they change the number of columns the characters occupy, they also affect the indentation functions. They also affect how the mode line is displayed; if you want to force redisplay of the mode line using the new values, call the function force-mode-line-update (see Mode Line Format).

User Option: ctl-arrow

This buffer-local variable controls how control characters are displayed. If it is non-nil, they are displayed as a caret followed by the character: ‘^A’. If it is nil, they are displayed as octal escapes: a backslash followed by three octal digits, as in ‘\001’.

User Option: tab-width

The value of this buffer-local variable is the spacing between tab stops used for displaying tab characters in Emacs buffers. The value is in units of columns, and the default is 8. Note that this feature is completely independent of the user-settable tab stops used by the command tab-to-tab-stop. See Adjustable Tab Stops.


41.23.2 Display Tables

A display table is a special-purpose char-table (see Char-Tables), with display-table as its subtype, which is used to override the usual character display conventions. This section describes how to make, inspect, and assign elements to a display table object. The next section (see Active Display Table) describes the various standard display tables and their precedence.

Function: make-display-table

This creates and returns a display table. The table initially has nil in all elements.

The ordinary elements of the display table are indexed by character codes; the element at index c says how to display the character code c. The value should be nil (which means to display the character c according to the usual display conventions; see Usual Display Conventions), or a vector of glyph codes (which means to display the character c as those glyphs; see Glyphs).

Warning: if you use the display table to change the display of newline characters, the whole buffer will be displayed as one long line.

The display table also has six extra slots which serve special purposes. Here is a table of their meanings; nil in any slot means to use the default for that slot, as stated below.

0

The glyph for the end of a truncated screen line (the default for this is ‘$’). See Glyphs. On graphical terminals, Emacs by default uses arrows in the fringes to indicate truncation, so the display table has no effect, unless you disable the fringes (see Window Fringes in the GNU Emacs Manual).

1

The glyph for the end of a continued line (the default is ‘\’). On graphical terminals, Emacs by default uses curved arrows in the fringes to indicate continuation, so the display table has no effect, unless you disable the fringes.

2

The glyph for indicating a character displayed as an octal character code (the default is ‘\’).

3

The glyph for indicating a control character (the default is ‘^’).

4

A vector of glyphs for indicating the presence of invisible lines (the default is ‘...’). See Selective Display.

5

The glyph used to draw the border between side-by-side windows (the default is ‘|’). See Splitting Windows. This currently has effect only on text terminals; on graphical terminals, if vertical scroll bars are supported and in use, a scroll bar separates the two windows, and if there are no vertical scroll bars and no dividers (see Window Dividers), Emacs uses a thin line to indicate the border.

For example, here is how to construct a display table that mimics the effect of setting ctl-arrow to a non-nil value (see Glyphs, for the function make-glyph-code):

(setq disptab (make-display-table))
(dotimes (i 32)
  (or (= i ?\t)
      (= i ?\n)
      (aset disptab i
            (vector (make-glyph-code ?^ 'escape-glyph)
                    (make-glyph-code (+ i 64) 'escape-glyph)))))
(aset disptab 127
      (vector (make-glyph-code ?^ 'escape-glyph)
              (make-glyph-code ?? 'escape-glyph)))
Function: display-table-slot display-table slot

This function returns the value of the extra slot slot of display-table. The argument slot may be a number from 0 to 5 inclusive, or a slot name (symbol). Valid symbols are truncation, wrap, escape, control, selective-display, and vertical-border.

Function: set-display-table-slot display-table slot value

This function stores value in the extra slot slot of display-table. The argument slot may be a number from 0 to 5 inclusive, or a slot name (symbol). Valid symbols are truncation, wrap, escape, control, selective-display, and vertical-border.

Function: describe-display-table display-table

This function displays a description of the display table display-table in a help buffer.

Command: describe-current-display-table

This command displays a description of the current display table in a help buffer.


41.23.3 Active Display Table

Each window can specify a display table, and so can each buffer. The window’s display table, if there is one, takes precedence over the buffer’s display table. If neither exists, Emacs tries to use the standard display table; if that is nil, Emacs uses the usual character display conventions (see Usual Display Conventions). (Emacs does not “merge” display tables: For instance, if the window has a display table, the buffer’s display table and the standard display table are completely ignored.)

Note that display tables affect how the mode line is displayed, so if you want to force redisplay of the mode line using a new display table, call force-mode-line-update (see Mode Line Format).

Function: window-display-table &optional window

This function returns window’s display table, or nil if there is none. The default for window is the selected window.

Function: set-window-display-table window table

This function sets the display table of window to table. The argument table should be either a display table or nil.

Variable: buffer-display-table

This variable is automatically buffer-local in all buffers; its value specifies the buffer’s display table. If it is nil, there is no buffer display table.

Variable: standard-display-table

The value of this variable is the standard display table, which is used when Emacs is displaying a buffer in a window with neither a window display table nor a buffer display table defined, or when Emacs is outputting text to the standard output or error streams. Although its default is typically nil, in an interactive session if the terminal cannot display curved quotes, its default maps curved quotes to ASCII approximations. See Text Quoting Style.

The disp-table library defines several functions for changing the standard display table.


41.23.4 Glyphs

A glyph is a graphical symbol which occupies a single character position on the screen. Each glyph is represented in Lisp as a glyph code, which specifies a character and optionally a face to display it in (see Faces). The main use of glyph codes is as the entries of display tables (see Display Tables). The following functions are used to manipulate glyph codes:

Function: make-glyph-code char &optional face

This function returns a glyph code representing char char with face face. If face is omitted or nil, the glyph uses the default face; in that case, the glyph code is an integer. If face is non-nil, the glyph code is not necessarily an integer object.

Function: glyph-char glyph

This function returns the character of glyph code glyph.

Function: glyph-face glyph

This function returns face of glyph code glyph, or nil if glyph uses the default face.

You can set up a glyph table to change how glyph codes are actually displayed on text terminals. This feature is semi-obsolete; use glyphless-char-display instead (see Glyphless Character Display).

Variable: glyph-table

The value of this variable, if non-nil, is the current glyph table. It takes effect only on character terminals; on graphical displays, all glyphs are displayed literally. The glyph table should be a vector whose gth element specifies how to display glyph code g, where g is the glyph code for a glyph whose face is unspecified. Each element should be one of the following:

nil

Display this glyph literally.

a string

Display this glyph by sending the specified string to the terminal.

a glyph code

Display the specified glyph code instead.

Any integer glyph code greater than or equal to the length of the glyph table is displayed literally.


41.23.5 Glyphless Character Display

Glyphless characters are characters which are displayed in a special way, e.g., as a box containing a hexadecimal code, instead of being displayed literally. These include characters which are explicitly defined to be glyphless, as well as characters for which there is no available font (on a graphical display), and characters which cannot be encoded by the terminal’s coding system (on a text terminal).

The glyphless-display-mode minor mode can be used to toggle displaying glyphless characters in a convenient manner in the current buffer. If this mode is enabled, all the glyphless characters are displayed as boxes that display acronyms of their character names.

Variable: glyphless-char-display

For more fine-grained (and global) control, this variable can be used. The value of this variable is a char-table which defines glyphless characters and how they are displayed. Each entry must be one of the following display methods:

nil

Display the character in the usual way.

zero-width

Don’t display the character.

thin-space

Display a thin space, 1-pixel wide on graphical displays, or 1-character wide on text terminals.

empty-box

Display an empty box.

hex-code

Display a box containing the Unicode codepoint of the character, in hexadecimal notation.

an ASCII string

Display a box containing that string. The string should contain at most 6 ASCII characters. As an exception, if the string includes just one character, on text-mode terminals that character will be displayed without a box; this allows to handle such “acronyms” as a replacement character for characters that cannot be displayed by the terminal.

a cons cell (graphical . text)

Display with graphical on graphical displays, and with text on text terminals. Both graphical and text must be one of the display methods described above.

The thin-space, empty-box, hex-code, and ASCII string display methods are drawn with the glyphless-char face. On text terminals, a box is emulated by square brackets, ‘[]’.

The char-table has one extra slot, which determines how to display any character that cannot be displayed with any available font, or cannot be encoded by the terminal’s coding system. Its value should be one of the above display methods, except zero-width.

If a character has a non-nil entry in an active display table, the display table takes effect; in this case, Emacs does not consult glyphless-char-display at all.

User Option: glyphless-char-display-control

This user option provides a convenient way to set glyphless-char-display for groups of similar characters. Do not set its value directly from Lisp code; the value takes effect only via a custom :set function (see Defining Customization Variables), which updates glyphless-char-display.

Its value should be an alist of elements (group . method), where group is a symbol specifying a group of characters, and method is a symbol specifying how to display them.

group should be one of the following:

c0-control

ASCII control characters U+0000 to U+001F, excluding the newline and tab characters (normally displayed as escape sequences like ‘^A’; see How Text Is Displayed in The GNU Emacs Manual).

c1-control

Non-ASCII, non-printing characters U+0080 to U+009F (normally displayed as octal escape sequences like ‘\230’).

format-control

Characters of Unicode General Category [Cf], such as U+200E LEFT-TO-RIGHT MARK, but excluding characters that have graphic images, such as U+00AD SOFT HYPHEN.

bidi-control

This is a subset of format-control, but only includes characters that are related to bidirectional formatting control, like U+2069 POP DIRECTIONAL ISOLATE and U+202A LEFT-TO-RIGHT EMBEDDING. See Bidirectional Display.

Characters of Unicode General Category [Cf], such as U+200E LEFT-TO-RIGHT MARK, but excluding characters that have graphic images, such as U+00AD SOFT HYPHEN.

variation-selectors

Unicode VS-1 through VS-256 (U+FE00 through U+FE0F and U+E0100 through U+E01EF), which are used to select between different glyphs for the same codepoints (typically emojis).

no-font

Characters for which there is no suitable font, or which cannot be encoded by the terminal’s coding system, or those for which the text-mode terminal has no glyphs.

The method symbol should be one of zero-width, thin-space, empty-box, or hex-code. These have the same meanings as in glyphless-char-display, above.


41.24 Beeping

This section describes how to make Emacs ring the bell (or blink the screen) to attract the user’s attention. Be conservative about how often you do this; frequent bells can become irritating. Also be careful not to use just beeping when signaling an error is more appropriate (see Errors).

Function: ding &optional do-not-terminate

This function beeps, or flashes the screen (see visible-bell below). It also terminates any keyboard macro currently executing unless do-not-terminate is non-nil.

Function: beep &optional do-not-terminate

This is a synonym for ding.

User Option: visible-bell

This variable determines whether Emacs should flash the screen to represent a bell. Non-nil means yes, nil means no. This is effective on graphical displays, and on text terminals provided the terminal’s Termcap entry defines the visible bell capability (‘vb’).

User Option: ring-bell-function

If this is non-nil, it specifies how Emacs should ring the bell. Its value should be a function of no arguments. If this is non-nil, it takes precedence over the visible-bell variable.


41.25 Window Systems

Emacs works with several window systems, most notably the X Window System. Both Emacs and X use the term “window”, but use it differently. An Emacs frame is a single window as far as X is concerned; the individual Emacs windows are not known to X at all.

Variable: window-system

This terminal-local variable tells Lisp programs what window system Emacs is using for displaying the frame. The possible values are

x

Emacs is displaying the frame using X.

w32

Emacs is displaying the frame using native MS-Windows GUI.

ns

Emacs is displaying the frame using the Nextstep interface (used on GNUstep and macOS).

pc

Emacs is displaying the frame using MS-DOS direct screen writes.

haiku

Emacs is displaying the frame using the Application Kit on Haiku.

pgtk

Emacs is displaying the frame using pure GTK facilities.

nil

Emacs is displaying the frame on a character-based terminal.

Variable: initial-window-system

This variable holds the value of window-system used for the first frame created by Emacs during startup. (When Emacs is invoked as a daemon, it does not create any initial frames, so initial-window-system is nil, except on MS-Windows, where it is still w32. See daemon in The GNU Emacs Manual.)

Function: window-system &optional frame

This function returns a symbol whose name tells what window system is used for displaying frame (which defaults to the currently selected frame). The list of possible symbols it returns is the same one documented for the variable window-system above.

Do not use window-system and initial-window-system as predicates or boolean flag variables, if you want to write code that works differently on text terminals and graphic displays. That is because window-system is not a good indicator of Emacs capabilities on a given display type. Instead, use display-graphic-p or any of the other display-*-p predicates described in Display Feature Testing.


41.26 Tooltips

Tooltips are special frames (see Frames) that are used to display helpful hints (a.k.a. “tips”) related to the current position of the mouse pointer. Emacs uses tooltips to display help strings about active portions of text (see Properties with Special Meanings) and about various UI elements, such as menu items (see Extended Menu Items) and tool-bar buttons (see Tool bars).

Function: tooltip-mode

Tooltip Mode is a minor mode that enables display of tooltips. Turning off this mode causes the tooltips be displayed in the echo area. On text-mode (a.k.a. “TTY”) frames, tooltips are always displayed in the echo area.

When Emacs is built with the GTK+ toolkit or Haiku windowing support, it by default displays tooltips using toolkit functions, and the appearance of the tooltips is then controlled by the toolkit’s settings. Toolkit-provided tooltips can be disabled by changing the value of the variable use-system-tooltips to nil. The rest of this subsection describes how to control non-toolkit tooltips, which are presented by Emacs itself.

Tooltips are displayed in special frames called tooltip frames, which have their own frame parameters (see Frame Parameters). Unlike other frames, the default parameters for tooltip frames are stored in a special variable.

User Option: tooltip-frame-parameters

This customizable option holds the default frame parameters used for displaying tooltips. Any font and color parameters are ignored, and the corresponding attributes of the tooltip face are used instead. If left or top parameters are included, they are used as absolute frame-relative coordinates where the tooltip should be shown. (Mouse-relative position of the tooltip can be customized using the variables described in Tooltips in The GNU Emacs Manual.) Note that the left and top parameters, if present, override the values of mouse-relative offsets.

The tooltip face determines the appearance of text shown in tooltips. It should generally use a variable-pitch font of size that is preferably smaller than the default frame font.

Variable: tooltip-functions

This abnormal hook is a list of functions to call when Emacs needs to display a tooltip. Each function is called with a single argument event which is a copy of the last mouse movement event. If a function on this list actually displays the tooltip, it should return non-nil, and then the rest of the functions will not be called. The default value of this variable is a single function tooltip-help-tips.

If you write your own function to be put on the tooltip-functions list, you may need to know the buffer of the mouse event that triggered the tooltip display. The following function provides that information.

Function: tooltip-event-buffer event

This function returns the buffer over which event occurred. Call it with the argument of the function from tooltip-functions to obtain the buffer whose text triggered the tooltip. Note that the event might occur not over a buffer (e.g., over the tool bar), in which case this function will return nil.

Other aspects of tooltip display are controlled by several customizable settings; see Tooltips in The GNU Emacs Manual.


41.27 Bidirectional Display

Emacs can display text written in scripts, such as Arabic, Farsi, and Hebrew, whose natural ordering for horizontal text display runs from right to left. Furthermore, segments of Latin script and digits embedded in right-to-left text are displayed left-to-right, while segments of right-to-left script embedded in left-to-right text (e.g., Arabic or Hebrew text in comments or strings in a program source file) are appropriately displayed right-to-left. We call such mixtures of left-to-right and right-to-left text bidirectional text. This section describes the facilities and options for editing and displaying bidirectional text.

Text is stored in Emacs buffers and strings in logical (or reading) order, i.e., the order in which a human would read each character. In right-to-left and bidirectional text, the order in which characters are displayed on the screen (called visual order) is not the same as logical order; the characters’ screen positions do not increase monotonically with string or buffer position. In performing this bidirectional reordering, Emacs follows the Unicode Bidirectional Algorithm (a.k.a. UBA), which is described in Annex #9 of the Unicode standard (https://www.unicode.org/reports/tr9/). Emacs provides a “Full Bidirectionality” class implementation of the UBA, consistent with the requirements of the Unicode Standard v9.0. Note, however, that the way Emacs displays continuation lines when text direction is opposite to the base paragraph direction deviates from the UBA, which requires to perform line wrapping before reordering text for display.

Variable: bidi-display-reordering

If the value of this buffer-local variable is non-nil (the default), Emacs performs bidirectional reordering for display. The reordering affects buffer text, as well as display strings and overlay strings from text and overlay properties in the buffer (see Overlay Properties, and see The display Property). If the value is nil, Emacs does not perform bidirectional reordering in the buffer.

The default value of bidi-display-reordering controls the reordering of strings which are not directly supplied by a buffer, including the text displayed in mode lines (see Mode Line Format) and header lines (see Window Header Lines).

Emacs never reorders the text of a unibyte buffer, even if bidi-display-reordering is non-nil in the buffer. This is because unibyte buffers contain raw bytes, not characters, and thus lack the directionality properties required for reordering. Therefore, to test whether text in a buffer will be reordered for display, it is not enough to test the value of bidi-display-reordering alone. The correct test is this:

 (if (and enable-multibyte-characters
          bidi-display-reordering)
     ;; Buffer is being reordered for display
   )

However, unibyte display and overlay strings are reordered if their parent buffer is reordered. This is because plain-ASCII strings are stored by Emacs as unibyte strings. If a unibyte display or overlay string includes non-ASCII characters, these characters are assumed to have left-to-right direction.

Text covered by display text properties, by overlays with display properties whose value is a string, and by any other properties that replace buffer text, is treated as a single unit when it is reordered for display. That is, the entire chunk of text covered by these properties is reordered together. Moreover, the bidirectional properties of the characters in such a chunk of text are ignored, and Emacs reorders them as if they were replaced with a single character U+FFFC, known as the Object Replacement Character. This means that placing a display property over a portion of text may change the way that the surrounding text is reordered for display. To prevent this unexpected effect, always place such properties on text whose directionality is identical with text that surrounds it.

Each paragraph of bidirectional text has a base direction, either right-to-left or left-to-right. Left-to-right paragraphs are displayed beginning at the left margin of the window, and are truncated or continued when the text reaches the right margin. Right-to-left paragraphs are displayed beginning at the right margin, and are continued or truncated at the left margin.

Where exactly paragraphs start and end, for the purpose of the Emacs UBA implementation, is determined by the following two buffer-local variables (note that paragraph-start and paragraph-separate have no influence on this). By default both of these variables are nil, and paragraphs are bounded by empty lines, i.e., lines that consist entirely of zero or more whitespace characters followed by a newline.

Variable: bidi-paragraph-start-re

If non-nil, this variable’s value should be a regular expression matching a line that starts or separates two paragraphs. The regular expression is always matched after a newline, so it is best to anchor it, i.e., begin it with a "^".

Variable: bidi-paragraph-separate-re

If non-nil, this variable’s value should be a regular expression matching a line separates two paragraphs. The regular expression is always matched after a newline, so it is best to anchor it, i.e., begin it with a "^".

If you modify any of these two variables, you should normally modify both, to make sure they describe paragraphs consistently. For example, to have each new line start a new paragraph for bidi-reordering purposes, set both variables to "^".

By default, Emacs determines the base direction of each paragraph by looking at the text at its beginning. The precise method of determining the base direction is specified by the UBA; in a nutshell, the first character in a paragraph that has an explicit directionality determines the base direction of the paragraph. However, sometimes a buffer may need to force a certain base direction for its paragraphs. For example, buffers containing program source code should force all paragraphs to be displayed left-to-right. You can use following variable to do this:

User Option: bidi-paragraph-direction

If the value of this buffer-local variable is the symbol right-to-left or left-to-right, all paragraphs in the buffer are assumed to have that specified direction. Any other value is equivalent to nil (the default), which means to determine the base direction of each paragraph from its contents.

Modes for program source code should set this to left-to-right. Prog mode does this by default, so modes derived from Prog mode do not need to set this explicitly (see Basic Major Modes).

Function: current-bidi-paragraph-direction &optional buffer

This function returns the paragraph direction at point in the named buffer. The returned value is a symbol, either left-to-right or right-to-left. If buffer is omitted or nil, it defaults to the current buffer. If the buffer-local value of the variable bidi-paragraph-direction is non-nil, the returned value will be identical to that value; otherwise, the returned value reflects the paragraph direction determined dynamically by Emacs. For buffers whose value of bidi-display-reordering is nil as well as unibyte buffers, this function always returns left-to-right.

Sometimes there’s a need to move point in strict visual order, either to the left or to the right of its current screen position. Emacs provides a primitive to do that.

Function: move-point-visually direction

This function moves point of the currently selected window to the buffer position that appears immediately to the right or to the left of point on the screen. If direction is positive, point will move one screen position to the right, otherwise it will move one screen position to the left. Note that, depending on the surrounding bidirectional context, this could potentially move point many buffer positions away. If invoked at the end of a screen line, the function moves point to the rightmost or leftmost screen position of the next or previous screen line, as appropriate for the value of direction.

The function returns the new buffer position as its value.

Bidirectional reordering can have surprising and unpleasant effects when two strings with bidirectional content are juxtaposed in a buffer, or otherwise programmatically concatenated into a string of text. A typical problematic case is when a buffer consists of sequences of text fields separated by whitespace or punctuation characters, like Buffer Menu mode or Rmail Summary Mode. Because the punctuation characters used as separators have weak directionality, they take on the directionality of surrounding text. As result, a numeric field that follows a field with bidirectional content can be displayed to the left of the preceding field, messing up the expected layout. There are several ways to avoid this problem:

  • Append the special character U+200E LEFT-TO-RIGHT MARK, or LRM, to the end of each field that may have bidirectional content, or prepend it to the beginning of the following field. The function bidi-string-mark-left-to-right, described below, comes in handy for this purpose. (In a right-to-left paragraph, use U+200F RIGHT-TO-LEFT MARK, or RLM, instead.) This is one of the solutions recommended by the UBA.
  • Include the tab character in the field separator. The tab character plays the role of segment separator in bidirectional reordering, causing the text on either side to be reordered separately.
  • Separate fields with a display property or overlay with a property value of the form (space . PROPS) (see Specified Spaces). Emacs treats this display specification as a paragraph separator, and reorders the text on either side separately.
Function: bidi-string-mark-left-to-right string

This function returns its argument string, possibly modified, such that the result can be safely concatenated with another string, or juxtaposed with another string in a buffer, without disrupting the relative layout of this string and the next one on display. If the string returned by this function is displayed as part of a left-to-right paragraph, it will always appear on display to the left of the text that follows it. The function works by examining the characters of its argument, and if any of those characters could cause reordering on display, the function appends the LRM character to the string. The appended LRM character is made invisible by giving it an invisible text property of t (see Invisible Text).

The reordering algorithm uses the bidirectional properties of the characters stored as their bidi-class property (see Character Properties). Lisp programs can change these properties by calling the put-char-code-property function. However, doing this requires a thorough understanding of the UBA, and is therefore not recommended. Any changes to the bidirectional properties of a character have global effect: they affect all Emacs frames and windows.

Similarly, the mirroring property is used to display the appropriate mirrored character in the reordered text. Lisp programs can affect the mirrored display by changing this property. Again, any such changes affect all of Emacs display.

The bidirectional properties of characters can be overridden by inserting into the text special directional control characters, LEFT-TO-RIGHT OVERRIDE (LRO) and RIGHT-TO-LEFT OVERRIDE (RLO). Any characters between a RLO and the following newline or POP DIRECTIONAL FORMATTING (PDF) control character, whichever comes first, will be displayed as if they were strong right-to-left characters, i.e. they will be reversed on display. Similarly, any characters between LRO and PDF or newline will display as if they were strong left-to-right, and will not be reversed even if they are strong right-to-left characters.

These overrides are useful when you want to make some text unaffected by the reordering algorithm, and instead directly control the display order. But they can also be used for malicious purposes, known as phishing. Specifically, a URL on a Web page or a link in an email message can be manipulated to make its visual appearance unrecognizable, or similar to some popular benign location, while the real location, interpreted by a browser in the logical order, is very different.

Emacs provides a primitive that applications can use to detect instances of text whose bidirectional properties were overridden so as to make a left-to-right character display as if it were a right-to-left character, or vice versa.

Function: bidi-find-overridden-directionality from to &optional object

This function looks at the text of the specified object between positions from (inclusive) and to (exclusive), and returns the first position where it finds a strong left-to-right character whose directional properties were forced to display the character as right-to-left, or for a strong right-to-left character that was forced to display as left-to-right. If it finds no such characters in the specified region of text, it returns nil.

The optional argument object specifies which text to search, and defaults to the current buffer. If object is non-nil, it can be some other buffer, or it can be a string or a window. If it is a string, the function searches that string. If it is a window, the function searches the buffer displayed in that window. If a buffer whose text you want to examine is displayed in some window, we recommend to specify it by that window, rather than pass the buffer to the function. This is because telling the function about the window allows it to correctly account for window-specific overlays, which might change the result of the function if some text in the buffer is covered by overlays.

When text that includes mixed right-to-left and left-to-right characters and bidirectional controls is copied into a different location, it can change its visual appearance, and also can affect the visual appearance of the surrounding text at destination. This is because reordering of bidirectional text specified by the UBA has non-trivial context-dependent effects both on the copied text and on the text at copy destination that will surround it.

Sometimes, a Lisp program may need to preserve the exact visual appearance of the copied text at destination, and of the text that surrounds the copy. Lisp programs can use the following function to achieve that effect.

Function: buffer-substring-with-bidi-context start end &optional no-properties

This function works similar to buffer-substring (see Examining Buffer Contents), but it prepends and appends to the copied text bidi directional control characters necessary to preserve the visual appearance of the text when it is inserted at another place. Optional argument no-properties, if non-nil, means remove the text properties from the copy of the text.


42 Operating System Interface

This chapter is about starting and getting out of Emacs, access to values in the operating system environment, and terminal input, output.

See Building Emacs, for related information. See Emacs Display, for additional operating system status information pertaining to the terminal and the screen.


42.1 Starting Up Emacs

This section describes what Emacs does when it is started, and how you can customize these actions.


42.1.1 Summary: Sequence of Actions at Startup

When Emacs is started up, it performs the following operations (see normal-top-level in startup.el):

  1. It adds subdirectories to load-path, by running the file named subdirs.el in each directory in the list. Normally, this file adds the directory’s subdirectories to the list, and those are scanned in their turn. The files subdirs.el are normally generated automatically when Emacs is installed.
  2. It loads any leim-list.el that it finds in the load-path directories. This file is intended for registering input methods. The search is only for any personal leim-list.el files that you may have created; it skips the directories containing the standard Emacs libraries (these should contain only a single leim-list.el file, which is compiled into the Emacs executable).
  3. It sets the variable before-init-time to the value of current-time (see Time of Day). It also sets after-init-time to nil, which signals to Lisp programs that Emacs is being initialized.
  4. It sets the language environment and the terminal coding system, if requested by environment variables such as LANG.
  5. It does some basic parsing of the command-line arguments.
  6. It loads your early init file (see Early Init File in The GNU Emacs Manual). This is not done if the options ‘-q’, ‘-Q’, or ‘--batch’ were specified. If the ‘-u’ option was specified, Emacs looks for the init file in that user’s home directory instead.
  7. It calls the function package-activate-all to activate any optional Emacs Lisp package that has been installed. See Packaging Basics. However, Emacs doesn’t activate the packages when package-enable-at-startup is nil or when it’s started with one of the options ‘-q’, ‘-Q’, or ‘--batch’. To activate the packages in the latter case, package-activate-all should be called explicitly (e.g., via the ‘--funcall’ option).
  8. If not running in batch mode, it initializes the window system that the variable initial-window-system specifies (see initial-window-system). The initialization function, window-system-initialization, is a generic function (see Generic Functions) whose actual implementation is different for each supported window system. If the value of initial-window-system is windowsystem, then the appropriate implementation of the initialization function is defined in the file term/windowsystem-win.el. This file should have been compiled into the Emacs executable when it was built.
  9. It runs the normal hook before-init-hook.
  10. If appropriate, it creates a graphical frame. As part of creating the graphical frame, it initializes the window system specified by initial-frame-alist and default-frame-alist (see Initial Frame Parameters) for the graphical frame, by calling the window-system-initialization function for that window system. This is not done in batch (noninteractive) or daemon mode.
  11. It initializes the initial frame’s faces, and sets up the menu bar and tool bar if needed. If graphical frames are supported, it sets up the tool bar even if the current frame is not a graphical one, since a graphical frame may be created later on.
  12. It use custom-reevaluate-setting to re-initialize the members of the list custom-delayed-init-variables. These are any pre-loaded user options whose default value depends on the run-time, rather than build-time, context. See custom-initialize-delay.
  13. It loads the library site-start, if it exists. This is not done if the options ‘-Q’ or ‘--no-site-file’ were specified.
  14. It loads your init file (see The Init File). This is not done if the options ‘-q’, ‘-Q’, or ‘--batch’ were specified. If the ‘-u’ option was specified, Emacs looks for the init file in that user’s home directory instead.
  15. It loads the library default, if it exists. This is not done if inhibit-default-init is non-nil, nor if the options ‘-q’, ‘-Q’, or ‘--batch’ were specified.
  16. It loads your abbrevs from the file specified by abbrev-file-name, if that file exists and can be read (see abbrev-file-name). This is not done if the option ‘--batch’ was specified.
  17. It sets the variable after-init-time to the value of current-time. This variable was set to nil earlier; setting it to the current time signals that the initialization phase is over, and, together with before-init-time, provides the measurement of how long it took.
  18. It runs the normal hooks after-init-hook and delayed-warnings-hook. The latter shows any warnings emitted during previous stages of startup, which are automatically delayed.
  19. If the buffer *scratch* exists and is still in Fundamental mode (as it should be by default), it sets its major mode according to initial-major-mode.
  20. If started on a text terminal, it loads the terminal-specific Lisp library (see Terminal-Specific Initialization), and runs the hook tty-setup-hook. This is not done in --batch mode, nor if term-file-prefix is nil.
  21. It displays the initial echo area message, unless you have suppressed that with inhibit-startup-echo-area-message.
  22. It processes any command-line options that were not handled earlier.
  23. It now exits if the option --batch was specified.
  24. If the *scratch* buffer exists and is empty, it inserts (substitute-command-keys initial-scratch-message) into that buffer.
  25. If initial-buffer-choice is a string, it visits the file (or directory) with that name. If it is a function, it calls the function with no arguments and selects the buffer that it returns. If one file is given as a command line argument, that file is visited and its buffer displayed alongside initial-buffer-choice. If more than one file is given, all of the files are visited and the *Buffer List* buffer is displayed alongside initial-buffer-choice.
  26. It runs emacs-startup-hook.
  27. It calls frame-notice-user-settings, which modifies the parameters of the selected frame according to whatever the init files specify.
  28. It runs window-setup-hook. The only difference between this hook and emacs-startup-hook is that this one runs after the previously mentioned modifications to the frame parameters.
  29. It displays the startup screen, which is a special buffer that contains information about copyleft and basic Emacs usage. This is not done if inhibit-startup-screen or initial-buffer-choice are non-nil, or if the ‘--no-splash’ or ‘-Q’ command-line options were specified.
  30. If a daemon was requested, it calls server-start. (On POSIX systems, if a background daemon was requested, it then detaches from the controlling terminal.) See Emacs Server in The GNU Emacs Manual.
  31. If started by the X session manager, it calls emacs-session-restore passing it as argument the ID of the previous session. See Session Management.

The following options affect some aspects of the startup sequence.

User Option: inhibit-startup-screen

This variable, if non-nil, inhibits the startup screen. In that case, Emacs typically displays the *scratch* buffer; but see initial-buffer-choice, below.

Do not set this variable in the init file of a new user, or in a way that affects more than one user, as that would prevent new users from receiving information about copyleft and basic Emacs usage.

inhibit-startup-message and inhibit-splash-screen are aliases for this variable.

User Option: initial-buffer-choice

If non-nil, this variable is a string that specifies a file or directory for Emacs to display after starting up, instead of the startup screen. If its value is a function, Emacs calls that function which must return a buffer which is then displayed. If its value is t, Emacs displays the *scratch* buffer.

User Option: inhibit-startup-echo-area-message

This variable controls the display of the startup echo area message. You can suppress the startup echo area message by adding text with this form to your init file:

(setq inhibit-startup-echo-area-message
      "your-login-name")

Emacs explicitly checks for an expression as shown above in your init file; your login name must appear in the expression as a Lisp string constant. You can also use the Customize interface. Other methods of setting inhibit-startup-echo-area-message to the same value do not inhibit the startup message. This way, you can easily inhibit the message for yourself if you wish, but thoughtless copying of your init file will not inhibit the message for someone else.

User Option: initial-scratch-message

This variable, if non-nil, should be a string, which is treated as documentation to be inserted into the *scratch* buffer when Emacs starts up or when that buffer is recreated. If it is nil, the *scratch* buffer is empty.

The following command-line options affect some aspects of the startup sequence. See Initial Options in The GNU Emacs Manual.

--no-splash

Do not display a splash screen.

--batch

Run without an interactive terminal. See Batch Mode.

--daemon
--bg-daemon
--fg-daemon

Do not initialize any display; just start a server. (A “background” daemon automatically runs in the background.)

--no-init-file
-q

Do not load either the init file, or the default library.

--no-site-file

Do not load the site-start library.

--quick
-Q

Equivalent to ‘-q --no-site-file --no-splash’.

--init-directory

Specify the directory to use when finding the Emacs init files.


42.1.2 The Init File

When you start Emacs, it normally attempts to load your init file. This is either a file named .emacs or .emacs.el in your home directory, or a file named init.el in a subdirectory named .emacs.d in your home directory.

The command-line switches ‘-q’, ‘-Q’, and ‘-u’ control whether and where to find the init file; ‘-q’ (and the stronger ‘-Q’) says not to load an init file, while ‘-u user’ says to load user’s init file instead of yours. See Entering Emacs in The GNU Emacs Manual. If neither option is specified, Emacs uses the LOGNAME environment variable, or the USER (most systems) or USERNAME (MS systems) variable, to find your home directory and thus your init file; this way, even if you have su’d, Emacs still loads your own init file. If those environment variables are absent, though, Emacs uses your user-id to find your home directory.

Emacs also attempts to load a second init file, called the early init file, if it exists. This is a file named early-init.el in your ~/.emacs.d directory. The difference between the early init file and the regular init file is that the early init file is loaded much earlier during the startup process, so you can use it to customize some things that are initialized before loading the regular init file. For example, you can customize the process of initializing the package system, by setting variables such as package-load-list or package-enable-at-startup. See Package Installation in The GNU Emacs Manual.

An Emacs installation may have a default init file, which is a Lisp library named default.el. Emacs finds this file through the standard search path for libraries (see How Programs Do Loading). The Emacs distribution does not come with this file; it is intended for local customizations. If the default init file exists, it is loaded whenever you start Emacs. But your own personal init file, if any, is loaded first; if it sets inhibit-default-init to a non-nil value, then Emacs does not subsequently load the default.el file. In batch mode, or if you specify ‘-q’ (or ‘-Q’), Emacs loads neither your personal init file nor the default init file.

Another file for site-customization is site-start.el. Emacs loads this before the user’s init file. You can inhibit the loading of this file with the option ‘--no-site-file’.

User Option: site-run-file

This variable specifies the site-customization file to load before the user’s init file. Its normal value is "site-start". The only way you can change it with real effect is to do so before dumping Emacs.

See Init File Examples in The GNU Emacs Manual, for examples of how to make various commonly desired customizations in your .emacs file.

User Option: inhibit-default-init

If this variable is non-nil, it prevents Emacs from loading the default initialization library file. The default value is nil.

Variable: before-init-hook

This normal hook is run, once, just before loading all the init files (site-start.el, your init file, and default.el). (The only way to change it with real effect is before dumping Emacs.)

Variable: after-init-hook

This normal hook is run, once, just after loading all the init files (site-start.el, your init file, and default.el), before loading the terminal-specific library (if started on a text terminal) and processing the command-line action arguments.

Variable: emacs-startup-hook

This normal hook is run, once, just after handling the command line arguments. In batch mode, Emacs does not run this hook.

Variable: window-setup-hook

This normal hook is very similar to emacs-startup-hook. The only difference is that it runs slightly later, after setting of the frame parameters. See window-setup-hook.

Variable: user-init-file

This variable holds the absolute file name of the user’s init file. If the actual init file loaded is a compiled file, such as .emacs.elc, the value refers to the corresponding source file.

Variable: user-emacs-directory

This variable holds the name of the Emacs default directory. It defaults to ${XDG_CONFIG_HOME-'~/.config'}/emacs/ if that directory exists and ~/.emacs.d/ and ~/.emacs do not exist, otherwise to ~/.emacs.d/ on all platforms but MS-DOS. Here, ${XDG_CONFIG_HOME-'~/.config'} stands for the value of the environment variable XDG_CONFIG_HOME if that variable is set, and for ~/.config otherwise. See How Emacs Finds Your Init File in The GNU Emacs Manual.


42.1.3 Terminal-Specific Initialization

Each terminal type can have its own Lisp library that Emacs loads when run on that type of terminal. The library’s name is constructed by concatenating the value of the variable term-file-prefix and the terminal type (specified by the environment variable TERM). Normally, term-file-prefix has the value "term/"; changing this is not recommended. If there is an entry matching TERM in the term-file-aliases association list, Emacs uses the associated value in place of TERM. Emacs finds the file in the normal manner, by searching the load-path directories, and trying the ‘.elc’ and ‘.el’ suffixes.

The usual role of a terminal-specific library is to enable special keys to send sequences that Emacs can recognize. It may also need to set or add to input-decode-map if the Termcap or Terminfo entry does not specify all the terminal’s function keys. See Terminal Input.

When the name of the terminal type contains a hyphen or underscore, and no library is found whose name is identical to the terminal’s name, Emacs strips from the terminal’s name the last hyphen or underscore and everything that follows it, and tries again. This process is repeated until Emacs finds a matching library, or until there are no more hyphens or underscores in the name (i.e., there is no terminal-specific library). For example, if the terminal name is ‘xterm-256color’ and there is no term/xterm-256color.el library, Emacs tries to load term/xterm.el. If necessary, the terminal library can evaluate (getenv "TERM") to find the full name of the terminal type.

Your init file can prevent the loading of the terminal-specific library by setting the variable term-file-prefix to nil.

You can also arrange to override some of the actions of the terminal-specific library by using tty-setup-hook. This is a normal hook that Emacs runs after initializing a new text terminal. You could use this hook to define initializations for terminals that do not have their own libraries. See Hooks.

User Option: term-file-prefix

If the value of this variable is non-nil, Emacs loads a terminal-specific initialization file as follows:

(load (concat term-file-prefix (getenv "TERM")))

You may set the term-file-prefix variable to nil in your init file if you do not wish to load the terminal-initialization file.

On MS-DOS, Emacs sets the TERM environment variable to ‘internal’.

User Option: term-file-aliases

This variable is an association list mapping terminal types to their aliases. For example, an element of the form ("vt102" . "vt100") means to treat a terminal of type ‘vt102’ like one of type ‘vt100’.

Variable: tty-setup-hook

This variable is a normal hook that Emacs runs after initializing a new text terminal. (This applies when Emacs starts up in non-windowed mode, and when making a tty emacsclient connection.) The hook runs after loading your init file (if applicable) and the terminal-specific Lisp file, so you can use it to adjust the definitions made by that file.

For a related feature, see window-setup-hook.


42.1.4 Command-Line Arguments

You can use command-line arguments to request various actions when you start Emacs. Note that the recommended way of using Emacs is to start it just once, after logging in, and then do all editing in the same Emacs session (see Entering Emacs in The GNU Emacs Manual). For this reason, you might not use command-line arguments very often; nonetheless, they can be useful when invoking Emacs from session scripts or debugging Emacs. This section describes how Emacs processes command-line arguments.

Function: command-line

This function parses the command line that Emacs was called with, processes it, and (amongst other things) loads the user’s init file and displays the startup messages.

Variable: command-line-processed

The value of this variable is t once the command line has been processed.

If you redump Emacs by calling dump-emacs (see Building Emacs), you may wish to set this variable to nil first in order to cause the new dumped Emacs to process its new command-line arguments.

Variable: command-switch-alist

This variable is an alist of user-defined command-line options and associated handler functions. By default it is empty, but you can add elements if you wish.

A command-line option is an argument on the command line, which has the form:

-option

The elements of the command-switch-alist look like this:

(option . handler-function)

The CAR, option, is a string, the name of a command-line option (including the initial hyphen). The handler-function is called to handle option, and receives the option name as its sole argument.

In some cases, the option is followed in the command line by an argument. In these cases, the handler-function can find all the remaining command-line arguments in the variable command-line-args-left (see below). (The entire list of command-line arguments is in command-line-args.)

Note that the handling of command-switch-alist doesn’t treat equals signs in option specially. That is, if there’s an option like --name=value on the command line, then only a command-switch-alist member whose car is literally --name=value will match this option. If you want to parse such options, you need to use command-line-functions instead (see below).

The command-line arguments are parsed by the command-line-1 function in the startup.el file. See also Command Line Arguments for Emacs Invocation in The GNU Emacs Manual.

Variable: command-line-args

The value of this variable is the list of command-line arguments passed to Emacs.

Variable: command-line-args-left

The value of this variable is the list of command-line arguments that have not yet been processed.

Variable: command-line-functions

This variable’s value is a list of functions for handling an unrecognized command-line argument. Each time the next argument to be processed has no special meaning, the functions in this list are called, in order of appearance, until one of them returns a non-nil value.

These functions are called with no arguments. They can access the command-line argument under consideration through the variable argi, which is bound temporarily at this point. The remaining arguments (not including the current one) are in the variable command-line-args-left.

When a function recognizes and processes the argument in argi, it should return a non-nil value to say it has dealt with that argument. If it has also dealt with some of the following arguments, it can indicate that by deleting them from command-line-args-left.

If all of these functions return nil, then the argument is treated as a file name to visit.


42.2 Getting Out of Emacs

There are two ways to get out of Emacs: you can kill the Emacs job, which exits permanently, or you can suspend it, which permits you to reenter the Emacs process later. (In a graphical environment, you can of course simply switch to another application without doing anything special to Emacs, then switch back to Emacs when you want.)


42.2.1 Killing Emacs

Killing Emacs means ending the execution of the Emacs process. If you started Emacs from a terminal, the parent process normally resumes control. The low-level primitive for killing Emacs is kill-emacs.

Command: kill-emacs &optional exit-data restart

This command calls the hook kill-emacs-hook, then exits the Emacs process and kills it.

If exit-data is an integer, that is used as the exit status of the Emacs process. (This is useful primarily in batch operation; see Batch Mode.)

If exit-data is a string, its contents are stuffed into the terminal input buffer so that the shell (or whatever program next reads input) can read them.

If exit-data is neither an integer nor a string, or is omitted, that means to use the (system-specific) exit status which indicates successful program termination.

If restart is non-nil, instead of just exiting at the end, start a new Emacs process, using the same command line arguments as the currently running Emacs process.

The kill-emacs function is normally called via the higher-level command C-x C-c (save-buffers-kill-terminal). See Exiting in The GNU Emacs Manual. It is also called automatically if Emacs receives a SIGTERM or SIGHUP operating system signal (e.g., when the controlling terminal is disconnected), or if it receives a SIGINT signal while running in batch mode (see Batch Mode).

Variable: kill-emacs-hook

This normal hook is run by kill-emacs, before it kills Emacs.

Because kill-emacs can be called in situations where user interaction is impossible (e.g., when the terminal is disconnected), functions on this hook should not attempt to interact with the user. If you want to interact with the user when Emacs is shutting down, use kill-emacs-query-functions, described below.

When Emacs is killed, all the information in the Emacs process, aside from files that have been saved, is lost. Because killing Emacs inadvertently can lose a lot of work, the save-buffers-kill-terminal command queries for confirmation if you have buffers that need saving or subprocesses that are running. It also runs the abnormal hook kill-emacs-query-functions:

User Option: kill-emacs-query-functions

When save-buffers-kill-terminal is killing Emacs, it calls the functions in this hook, after asking the standard questions and before calling kill-emacs. The functions are called in order of appearance, with no arguments. Each function can ask for additional confirmation from the user. If any of them returns nil, save-buffers-kill-emacs does not kill Emacs, and does not run the remaining functions in this hook. Calling kill-emacs directly does not run this hook.

Command: restart-emacs

This command does the same as save-buffers-kill-emacs, but instead of just killing the current Emacs process at the end, it’ll restart a new Emacs process, using the same command line arguments as the currently running Emacs process.


42.2.2 Suspending Emacs

On text terminals, it is possible to suspend Emacs, which means stopping Emacs temporarily and returning control to its superior process, which is usually the shell. This allows you to resume editing later in the same Emacs process, with the same buffers, the same kill ring, the same undo history, and so on. To resume Emacs, use the appropriate command in the parent shell—most likely fg.

Suspending works only on a terminal device from which the Emacs session was started. We call that device the controlling terminal of the session. Suspending is not allowed if the controlling terminal is a graphical terminal. Suspending is usually not relevant in graphical environments, since you can simply switch to another application without doing anything special to Emacs.

Some operating systems (those without SIGTSTP, or MS-DOS) do not support suspension of jobs; on these systems, suspension actually creates a new shell temporarily as a subprocess of Emacs. Then you would exit the shell to return to Emacs.

Command: suspend-emacs &optional string

This function stops Emacs and returns control to the superior process. If and when the superior process resumes Emacs, suspend-emacs returns nil to its caller in Lisp.

This function works only on the controlling terminal of the Emacs session; to relinquish control of other tty devices, use suspend-tty (see below). If the Emacs session uses more than one terminal, you must delete the frames on all the other terminals before suspending Emacs, or this function signals an error. See Multiple Terminals.

If string is non-nil, its characters are sent to Emacs’s superior shell, to be read as terminal input. The characters in string are not echoed by the superior shell; only the results appear.

Before suspending, suspend-emacs runs the normal hook suspend-hook. After the user resumes Emacs, suspend-emacs runs the normal hook suspend-resume-hook. See Hooks.

The next redisplay after resumption will redraw the entire screen, unless the variable no-redraw-on-reenter is non-nil. See Refreshing the Screen.

Here is an example of how you could use these hooks:

(add-hook 'suspend-hook
          (lambda () (or (y-or-n-p "Really suspend?")
                         (error "Suspend canceled"))))
(add-hook 'suspend-resume-hook (lambda () (message "Resumed!")
                                 (sit-for 2)))

Here is what you would see upon evaluating (suspend-emacs "pwd"):

---------- Buffer: Minibuffer ----------
Really suspend? y
---------- Buffer: Minibuffer ----------

---------- Parent Shell ----------
bash$ /home/username
bash$ fg

---------- Echo Area ----------
Resumed!

Note that ‘pwd’ is not echoed after Emacs is suspended. But it is read and executed by the shell.

Variable: suspend-hook

This variable is a normal hook that Emacs runs before suspending.

Variable: suspend-resume-hook

This variable is a normal hook that Emacs runs on resuming after a suspension.

Function: suspend-tty &optional tty

If tty specifies a terminal device used by Emacs, this function relinquishes the device and restores it to its prior state. Frames that used the device continue to exist, but are not updated and Emacs doesn’t read input from them. tty can be a terminal object, a frame (meaning the terminal for that frame), or nil (meaning the terminal for the selected frame). See Multiple Terminals.

If tty is already suspended, this function does nothing.

This function runs the hook suspend-tty-functions, passing the terminal object as an argument to each function.

Function: resume-tty &optional tty

This function resumes the previously suspended terminal device tty; where tty has the same possible values as it does for suspend-tty.

This function reopens the terminal device, re-initializes it, and redraws it with that terminal’s selected frame. It then runs the hook resume-tty-functions, passing the terminal object as an argument to each function.

If the same device is already used by another Emacs terminal, this function signals an error. If tty is not suspended, this function does nothing.

Function: controlling-tty-p &optional tty

This function returns non-nil if tty is the controlling terminal of the Emacs session; tty can be a terminal object, a frame (meaning the terminal for that frame), or nil (meaning the terminal for the selected frame).

Command: suspend-frame

This command suspends a frame. For GUI frames, it calls iconify-frame (see Visibility of Frames); for frames on text terminals, it calls either suspend-emacs or suspend-tty, depending on whether the frame is displayed on the controlling terminal device or not.


42.3 Operating System Environment

Emacs provides access to variables in the operating system environment through various functions. These variables include the name of the system, the user’s UID, and so on.

Variable: system-configuration

This variable holds the standard GNU configuration name for the hardware/software configuration of your system, as a string. For example, a typical value for a 64-bit GNU/Linux system is ‘"x86_64-unknown-linux-gnu"’.

Variable: system-type

The value of this variable is a symbol indicating the type of operating system Emacs is running on. The possible values are:

aix

IBM’s AIX.

berkeley-unix

Berkeley BSD and its variants.

cygwin

Cygwin, a POSIX layer on top of MS-Windows.

darwin

Darwin (macOS).

gnu

The GNU system (using the GNU kernel, which consists of the HURD and Mach).

gnu/linux

A GNU/Linux system—that is, a variant GNU system, using the Linux kernel. (These systems are the ones people often call “Linux”, but actually Linux is just the kernel, not the whole system.)

gnu/kfreebsd

A GNU (glibc-based) system with a FreeBSD kernel.

haiku

The Haiku operating system, a derivative of the Be Operating System.

hpux

Hewlett-Packard HPUX operating system.

nacl

Google Native Client (NaCl) sandboxing system.

ms-dos

Microsoft’s DOS. Emacs compiled with DJGPP for MS-DOS binds system-type to ms-dos even when you run it on MS-Windows.

usg-unix-v

AT&T Unix System V.

windows-nt

Microsoft Windows NT, 9X and later. The value of system-type is always windows-nt, e.g., even on Windows 10.

We do not wish to add new symbols to make finer distinctions unless it is absolutely necessary! In fact, we hope to eliminate some of these alternatives in the future. If you need to make a finer distinction than system-type allows for, you can test system-configuration, e.g., against a regexp.

Function: system-name

This function returns the name of the machine you are running on, as a string.

User Option: mail-host-address

If this variable is non-nil, it is used instead of system-name for purposes of generating email addresses. For example, it is used when constructing the default value of user-mail-address. See User Identification.

Command: getenv var &optional frame

This function returns the value of the environment variable var, as a string. var should be a string. If var is undefined in the environment, getenv returns nil. It returns ‘""’ if var is set but null. Within Emacs, a list of environment variables and their values is kept in the variable process-environment.

(getenv "USER")
     ⇒ "lewis"

The shell command printenv prints all or part of the environment:

bash$ printenv
PATH=/usr/local/bin:/usr/bin:/bin
USER=lewis
TERM=xterm
SHELL=/bin/bash
HOME=/home/lewis
Command: setenv variable &optional value substitute

This command sets the value of the environment variable named variable to value. variable should be a string. Internally, Emacs Lisp can handle any string. However, normally variable should be a valid shell identifier, that is, a sequence of letters, digits and underscores, starting with a letter or underscore. Otherwise, errors may occur if subprocesses of Emacs try to access the value of variable. If value is omitted or nil (or, interactively, with a prefix argument), setenv removes variable from the environment. Otherwise, value should be a string.

If the optional argument substitute is non-nil, Emacs calls the function substitute-env-vars to expand any environment variables in value.

setenv works by modifying process-environment; binding that variable with let is also reasonable practice.

setenv returns the new value of variable, or nil if it removed variable from the environment.

Macro: with-environment-variables variables body…

This macro sets the environment variables according to variables temporarily when executing body. The previous values are restored when the form finishes. The argument variables should be a list of pairs of strings of the form (var value), where var is the name of the environment variable and value is that variable’s value.

(with-environment-variables (("LANG" "C")
                             ("LANGUAGE" "en_US:en"))
  (call-process "ls" nil t))
Variable: process-environment

This variable is a list of strings, each describing one environment variable. The functions getenv and setenv work by means of this variable.

process-environment
⇒ ("PATH=/usr/local/bin:/usr/bin:/bin"
    "USER=lewis"
    "TERM=xterm"
    "SHELL=/bin/bash"
    "HOME=/home/lewis"
    …)

If process-environment contains multiple elements that specify the same environment variable, the first of these elements specifies the variable, and the others are ignored.

Variable: initial-environment

This variable holds the list of environment variables Emacs inherited from its parent process when Emacs started.

Variable: path-separator

This variable holds a string that says which character separates directories in a search path (as found in an environment variable). Its value is ":" for Unix and GNU systems, and ";" for MS systems.

Function: path-separator

This function returns the connection-local value of variable path-separator. That is ";" for MS systems and a local default-directory, and ":" for Unix and GNU systems, or a remote default-directory.

Function: parse-colon-path path

This function takes a search path string such as the value of the PATH environment variable, and splits it at the separators, returning a list of directories. nil in this list means the current directory. Although the function’s name says “colon”, it actually uses the value of variable path-separator.

(parse-colon-path ":/foo:/bar")
     ⇒ (nil "/foo/" "/bar/")
Variable: invocation-name

This variable holds the program name under which Emacs was invoked. The value is a string, and does not include a directory name.

Variable: invocation-directory

This variable holds the directory in which the Emacs executable was located when it was run, or nil if that directory cannot be determined.

Variable: installation-directory

If non-nil, this is a directory within which to look for the lib-src and etc subdirectories. In an installed Emacs, it is normally nil. It is non-nil when Emacs can’t find those directories in their standard installed locations, but can find them in a directory related somehow to the one containing the Emacs executable (i.e., invocation-directory).

Function: load-average &optional use-float

This function returns the current 1-minute, 5-minute, and 15-minute system load averages, in a list. The load average indicates the number of processes trying to run on the system.

By default, the values are integers that are 100 times the system load averages, but if use-float is non-nil, then they are returned as floating-point numbers without multiplying by 100.

If it is impossible to obtain the load average, this function signals an error. On some platforms, access to load averages requires installing Emacs as setuid or setgid so that it can read kernel information, and that usually isn’t advisable.

If the 1-minute load average is available, but the 5- or 15-minute averages are not, this function returns a shortened list containing the available averages.

(load-average)
     ⇒ (169 48 36)
(load-average t)
     ⇒ (1.69 0.48 0.36)

The shell command uptime returns similar information.

Function: emacs-pid

This function returns the process ID of the Emacs process, as an integer.

Variable: tty-erase-char

This variable holds the erase character that was selected in the system’s terminal driver, before Emacs was started.

Variable: null-device

This variable holds the system null device. Its value is "/dev/null" for Unix and GNU systems, and "NUL" for MS systems.

Function: null-device

This function returns the connection-local value of variable null-device. That is "NUL" for MS systems and a local default-directory, and "/dev/null" for Unix and GNU systems, or a remote default-directory.


42.4 User Identification

Variable: init-file-user

This variable says which user’s init files should be used by Emacs—or nil if none. "" stands for the user who originally logged in. The value reflects command-line options such as ‘-q’ or ‘-u user’.

Lisp packages that load files of customizations, or any other sort of user profile, should obey this variable in deciding where to find it. They should load the profile of the user name found in this variable. If init-file-user is nil, meaning that the ‘-q’, ‘-Q’, or ‘-batch’ option was used, then Lisp packages should not load any customization files or user profile.

User Option: user-mail-address

This holds the email address of the user who is using Emacs.

Function: user-login-name &optional uid

This function returns the name under which the user is logged in. It uses the environment variables LOGNAME or USER if either is set. Otherwise, the value is based on the effective UID, not the real UID.

If you specify uid (a number), the result is the user name that corresponds to uid, or nil if there is no such user.

Function: user-real-login-name

This function returns the user name corresponding to Emacs’s real UID. This ignores the effective UID, and the environment variables LOGNAME and USER.

Function: user-full-name &optional uid

This function returns the full name of the logged-in user—or the value of the environment variable NAME, if that is set.

If the Emacs process’s user-id does not correspond to any known user (and provided NAME is not set), the result is "unknown".

If uid is non-nil, then it should be a number (a user-id) or a string (a login name). Then user-full-name returns the full name corresponding to that user-id or login name. If you specify a user-id or login name that isn’t defined, it returns nil.

The symbols user-login-name, user-real-login-name and user-full-name are variables as well as functions. The functions return the same values that the variables hold. These variables allow you to fake out Emacs by telling the functions what to return. The variables are also useful for constructing frame titles (see Frame Titles).

Function: user-real-uid

This function returns the real UID of the user.

Function: user-uid

This function returns the effective UID of the user.

Function: group-gid

This function returns the effective GID of the Emacs process.

Function: group-real-gid

This function returns the real GID of the Emacs process.

Function: system-users

This function returns a list of strings, listing the user names on the system. If Emacs cannot retrieve this information, the return value is a list containing just the value of user-real-login-name.

Function: system-groups

This function returns a list of strings, listing the names of user groups on the system. If Emacs cannot retrieve this information, the return value is nil.

Function: group-name gid

This function returns the group name that corresponds to the numeric group ID gid, or nil if there is no such group.


42.5 Time of Day

This section explains how to determine the current time and time zone.

Many functions like current-time and file-attributes return Lisp timestamp values that count seconds, and that can represent absolute time by counting seconds since the epoch of 1970-01-01 00:00:00 UTC (Coordinated Universal Time). Typically these counts ignore leap seconds; however, GNU and some other operating systems can be configured to count leap seconds.

Although traditionally Lisp timestamps were integer pairs, their form has evolved and programs ordinarily should not depend on the current default form. If your program needs a particular timestamp form, you can use the time-convert function to convert it to the needed form. See Time Conversion.

There are currently three forms of Lisp timestamps, each of which represents a number of seconds:

  • An integer. Although this is the simplest form, it cannot represent subsecond timestamps.
  • A pair of integers (ticks . hz), where hz is positive. This represents ticks/hz seconds, which is the same time as plain ticks if hz is 1. A common value for hz is 1000000000, for a nanosecond-resolution clock.
  • A list of four integers (high low micro pico), where 0≤low<65536, 0≤micro<1000000, and 0≤pico<1000000. This represents the number of seconds using the formula: high * 2**16 + low + micro * 10**−6 + pico * 10**−12. If current-time-list is t, some functions may default to returning two- or three-element lists, with omitted micro and pico components defaulting to zero. On all current machines pico is a multiple of 1000, but this may change as higher-resolution clocks become available.

Function arguments, e.g., the time argument to format-time-string, accept a more-general time value format, which can be a Lisp timestamp, nil for the current time, a finite floating-point number for seconds, or a list (high low micro) or (high low) that is a truncated list timestamp with missing elements taken to be zero.

Time values can be converted to and from calendrical and other forms. Some of these conversions rely on operating system functions that limit the range of possible time values, and signal an error such as ‘"Specified time is not representable"’ if the limits are exceeded. For instance, a system might not support timestamps before the epoch, or years far in the future. You can convert a time value into a human-readable string using format-time-string, into a Lisp timestamp using time-convert, and into other forms using decode-time and float-time. These functions are described in the following sections.

Function: current-time-string &optional time zone

This function returns the current time and date as a human-readable string. The format does not vary for the initial part of the string, which contains the day of week, month, day of month, and time of day in that order: the number of characters used for these fields is always the same, although (unless you require English weekday or month abbreviations regardless of locale) it is typically more convenient to use format-time-string than to extract fields from the output of current-time-string, as the year might not have exactly four digits, and additional information may some day be added at the end.

The argument time, if given, specifies a time to format, instead of the current time. The optional argument zone defaults to the current time zone rule. See Time Zone Rules. The operating system limits the range of time and zone values.

(current-time-string)
     ⇒ "Fri Nov  1 15:59:49 2019"
Variable: current-time-list

This boolean variable is a transition aid. If t, current-time and related functions return timestamps in list form, typically (high low micro pico); otherwise, they use (ticks . hz) form. Currently this variable defaults to t, for behavior compatible with previous Emacs versions. Developers are encouraged to test timestamp-related code with this variable set to nil, as it will default to nil in a future Emacs version, and will be removed in some version after that.

Function: current-time

This function returns the current time as a Lisp timestamp. If current-time-list is nil, the timestamp has the form (ticks . hz) where ticks counts clock ticks and hz is the clock ticks per second. Otherwise, the timestamp has the list form (high low usec psec). You can use (time-convert nil t) or (time-convert nil 'list) to obtain a particular form regardless of the value of current-time-list. See Time Conversion.

Function: float-time &optional time

This function returns the current time as a floating-point number of seconds since the epoch. The optional argument time, if given, specifies a time to convert instead of the current time.

Warning: Since the result is floating point, it may not be exact. Do not use this function if precise time stamps are required. For example, on typical systems (float-time '(1 . 10)) displays as ‘0.1’ but is slightly greater than 1/10.

time-to-seconds is an alias for this function.

Function: current-cpu-time

Return the current CPU time along with its resolution. The return value is a pair (CPU-TICKS . TICKS-PER-SEC). The CPU-TICKS counter can wrap around, so values cannot be meaningfully compared if too much time has passed between them.


42.6 Time Zone Rules

The default time zone is determined by the TZ environment variable. See Operating System Environment. For example, you can tell Emacs to default to Universal Time with (setenv "TZ" "UTC0"). If TZ is not in the environment, Emacs uses system wall clock time, which is a platform-dependent default time zone.

The set of supported TZ strings is system-dependent. GNU and many other systems support TZDB timezones, e.g., ‘"America/New_York"’ specifies the time zone and daylight saving time history for locations near New York City. GNU and most other systems support POSIX-style TZ strings, e.g., ‘"EST5EDT,M4.1.0,M10.5.0"’ specifies the rules used in New York from 1987 through 2006. All systems support the string ‘"UTC0"’ meaning Universal Time.

Functions that convert to and from local time accept an optional time zone rule argument, which specifies the conversion’s time zone and daylight saving time history. If the time zone rule is omitted or nil, the conversion uses Emacs’s default time zone. If it is t, the conversion uses Universal Time. If it is wall, the conversion uses the system wall clock time. If it is a string, the conversion uses the time zone rule equivalent to setting TZ to that string. If it is a list (offset abbr), where offset is an integer number of seconds east of Universal Time and abbr is a string, the conversion uses a fixed time zone with the given offset and abbreviation. An integer offset is treated as if it were (offset abbr), where abbr is a numeric abbreviation on POSIX-compatible platforms and is unspecified on MS-Windows.

Function: current-time-zone &optional time zone

This function returns a list describing the time zone that the user is in.

The value has the form (offset abbr). Here offset is an integer giving the number of seconds ahead of Universal Time (east of Greenwich). A negative value means west of Greenwich. The second element, abbr, is a string giving an abbreviation for the time zone, e.g., ‘"CST"’ for China Standard Time or for U.S. Central Standard Time. Both elements can change when daylight saving time begins or ends; if the user has specified a time zone that does not use a seasonal time adjustment, then the value is constant through time.

If the operating system doesn’t supply all the information necessary to compute the value, the unknown elements of the list are nil.

The argument time, if given, specifies a time value to analyze instead of the current time. The optional argument zone defaults to the current time zone rule. The operating system limits the range of time and zone values.


42.7 Time Conversion

These functions convert time values (see Time of Day) to Lisp timestamps, or into calendrical information and vice versa.

Many operating systems use 64-bit signed integers to count seconds, and can represent times far in the past or future. However, some are more limited. For example, old-fashioned operating systems that use 32-bit signed integers typically handle only times from 1901-12-13 20:45:52 through 2038-01-19 03:14:07 Universal Time.

Calendrical conversion functions use the Gregorian calendar even for dates before the Gregorian calendar was introduced, and for dates in the far distant past or future for which the Gregorian calendar is wildly inaccurate and disagrees with common practice in scientific fields like astronomy and paleontology, which use Julian-calendar year lengths. Year numbers count since the year 1 BCE, and do not skip zero as traditional Gregorian years do; for example, the year number −37 represents the Gregorian year 38 BCE.

Function: time-convert time form

This function converts a time value into a Lisp timestamp.

The form argument specifies the timestamp form to be returned. If form is the symbol integer, this function returns an integer count of seconds. If form is a positive integer, it specifies a clock frequency and this function returns an integer-pair timestamp (ticks . form). If form is t, this function treats it as a positive integer suitable for representing the timestamp; for example, it is treated as 1000000000 if time is nil and the platform timestamp has nanosecond resolution. If form is list, this function returns an integer list (high low micro pico). Although a nil form currently acts like list, this is planned to change in a future Emacs version, so callers requiring list timestamps should pass list explicitly.

If time is not a time value, this function signals an error. Otherwise, if time cannot be represented exactly, conversion truncates it toward minus infinity. When form is t, conversion is always exact so no truncation occurs, and the returned clock resolution is no less than that of time. By way of contrast, although float-time can also convert any time value without signaling an error, the result might not be exact. See Time of Day.

For efficiency this function might return a value that is eq to time, or that otherwise shares structure with time.

Although (time-convert nil nil) is equivalent to (current-time), the latter may be a bit faster.

(setq a (time-convert nil t))
⇒ (1564826753904873156 . 1000000000)
(time-convert a 100000)
⇒ (156482675390487 . 100000)
(time-convert a 'integer)
⇒ 1564826753
(time-convert a 'list)
⇒ (23877 23681 904873 156000)
Function: decode-time &optional time zone form

This function converts a time value into calendrical information. If you don’t specify time, it decodes the current time, and similarly zone defaults to the current time zone rule. See Time Zone Rules. The operating system limits the range of time and zone values.

The form argument controls the form of the returned seconds element, as described below. The return value is a list of nine elements, as follows:

(seconds minutes hour day month year dow dst utcoff)

Here is what the elements mean:

seconds

The number of seconds past the minute, with form described below.

minutes

The number of minutes past the hour, as an integer between 0 and 59.

hour

The hour of the day, as an integer between 0 and 23.

day

The day of the month, as an integer between 1 and 31.

month

The month of the year, as an integer between 1 and 12.

year

The year, an integer typically greater than 1900.

dow

The day of week, as an integer between 0 and 6, where 0 stands for Sunday.

dst

t if daylight saving time is effect, nil if it is not in effect, and −1 if this information is not available.

utcoff

An integer indicating the Universal Time offset in seconds, i.e., the number of seconds east of Greenwich.

The seconds element is a Lisp timestamp that is nonnegative and less than 61; it is less than 60 except during positive leap seconds (assuming the operating system supports leap seconds). If the optional form argument is t, seconds uses the same precision as time; if form is integer, seconds is truncated to an integer. For example, if time is the timestamp (1566009571321 . 1000), which represents 2019-08-17 02:39:31.321 UTC on typical systems that lack leap seconds, then (decode-time time t t) returns ((31321 . 1000) 39 2 17 8 2019 6 nil 0), whereas (decode-time time t 'integer) returns (31 39 2 17 8 2019 6 nil 0). If form is omitted or nil, it currently defaults to integer but this default may change in future Emacs releases, so callers requiring a particular form should specify form.

Common Lisp Note: Common Lisp has different meanings for dow, dst and utcoff, and its second is an integer between 0 and 59 inclusive.

To access (or alter) the elements in the calendrical information, the decoded-time-second, decoded-time-minute, decoded-time-hour, decoded-time-day, decoded-time-month, decoded-time-year, decoded-time-weekday, decoded-time-dst and decoded-time-zone accessors can be used.

Function: encode-time time &rest obsolescent-arguments

This function converts time to a Lisp timestamp. It can act as the inverse of decode-time.

Ordinarily the first argument is a list (second minute hour day month year ignored dst zone) that specifies a decoded time in the style of decode-time. For the meanings of these list elements, see the table under decode-time. In particular, dst says how to interpret timestamps during a daylight saving fallback when timestamps are repeated. If dst is −1, the DST value is guessed; if it is t or nil the timestamp with that DST value is returned, with an error signaled if no such timestamp exists. Unfortunately a dst value of t or nil does not disambiguate timestamps duplicated when a TZDB-based timezone moves further west of Greenwich, such as disambiguating the two standard-time timestamps 2020-12-27 01:30 when zone is ‘"Europe/Volgograd"’, which at 02:00 that day changed standard time from 4 to 3 hours east of Greenwich; if you need to handle situations like this you can use a numeric zone to disambiguate instead.

The first argument can also be a list (second minute hour day month year), which is treated like the list (second minute hour day month year nil -1 nil).

As an obsolescent calling convention, this function can be given six or more arguments. The first six arguments second, minute, hour, day, month, and year specify most of the components of a decoded time. If there are more than six arguments the last argument is used as zone and any other extra arguments are ignored, so that (apply #'encode-time (decode-time ...)) works. In this obsolescent convention, dst is −1 and zone defaults to the current time zone rule (see Time Zone Rules). When modernizing an obsolescent caller, ensure that the more-modern list equivalent contains 9 elements with a dst element that is −1, not nil.

Year numbers less than 100 are not treated specially. If you want them to stand for years above 1900, or years above 2000, you must alter them yourself before you call encode-time. The operating system limits the range of time and zone values. However, timestamps ranging from the epoch to the near future are always supported.

The encode-time function acts as a rough inverse to decode-time. For example, you can pass the output of the latter to the former as follows:

(encode-time (decode-time …))

You can perform simple date arithmetic by using out-of-range values for seconds, minutes, hour, day, and month; for example, day 0 means the day preceding the given month. Take care when doing so, as it is common for this to fail in some cases. For example:

;; Try to compute the time one month from now.
;; Watch out; this might not work as expected.
(let ((time (decode-time)))
  (setf (decoded-time-month time)
        (+ (decoded-time-month time) 1))
  time)

Unfortunately, this code might not work as expected if the resulting time is invalid due to month length differences, daylight saving transitions, time zone changes, or missing leap days or leap seconds. For example, if executed on January 30 this code yields a nonexistent date February 30, which encode-time would adjust to early March. Similarly, adding four years to February 29, 2096 would yield the nonexistent date February 29, 2100; and adding one hour to 01:30 on March 13, 2022 in New York would yield a timestamp 02:30 that does not exist because clocks sprang forward from 02:00 to 03:00 that day. To avoid some (though not all) of the problem, you can base calculations on the middle of the affected unit, e.g., start at the 15th of the month when adding months. Alternatively, you can use the calendar and time-date libraries.


42.8 Parsing and Formatting Times

These functions convert time values to text in a string, and vice versa. Time values are either represented as a Lisp timestamp (see Time of Day) or a decoded time structure (see Time Conversion).

Function: date-to-time string

This function parses the time-string string and returns the corresponding Lisp timestamp. The argument string should represent a date-time, and should be in one of the forms recognized by parse-time-string (see below). This function assumes Universal Time if string lacks explicit time zone information, and assumes earliest values if string lacks month, day, or time. The operating system limits the range of time and zone values.

Function: parse-time-string string

This function parses the time-string string into a decoded time structure (see Time Conversion). The argument string should resemble an RFC 822 (or later) or ISO 8601 string, like “Fri, 25 Mar 2016 16:24:56 +0100” or “1998-09-12T12:21:54-0200”, but this function will attempt to parse less well-formed time strings as well.

Function: iso8601-parse string

For a more strict function (that will error out upon invalid input), this function can be used instead. It can parse all variants of the ISO 8601 standard, so in addition to the formats mentioned above, it also parses things like “1998W45-3” (week number) and “1998-245” (ordinal day number). To parse durations, there’s iso8601-parse-duration, and to parse intervals, there’s iso8601-parse-interval. All these functions return decoded time structures, except the final one, which returns three of them (the start, the end, and the duration).

Function: format-time-string format-string &optional time zone

This function converts time (which should be a Lisp timestamp, and defaults to the current time if time is omitted or nil) to a string according to format-string. The conversion uses the time zone rule zone, which defaults to the current time zone rule. See Time Zone Rules. The argument format-string may contain ‘%’-sequences which say to substitute parts of the time. Here is a table of what the ‘%’-sequences mean:

%a

This stands for the abbreviated name of the day of week.

%A

This stands for the full name of the day of week.

%b

This stands for the abbreviated name of the month.

%B

This stands for the full name of the month.

%c

This is a synonym for ‘%x %X’.

%C

This stands for the century, that is, the year divided by 100, truncated toward zero. The default field width is 2.

%d

This stands for the day of month, zero-padded.

%D

This is a synonym for ‘%m/%d/%y’.

%e

This stands for the day of month, blank-padded.

%F

This stands for the ISO 8601 date format, which is like ‘%+4Y-%m-%d’ except that any flags or field width override the ‘+’ and (after subtracting 6) the ‘4’.

%g

This stands for the year without century (00–99) corresponding to the current ISO week number. ISO weeks start on Monday and end on Sunday. If an ISO week begins in one year and ends in another, the rules regarding which year ‘%g’ will produce are complex and will not be described here; however, in general, if most of the week’s days are in the ending year, ‘%g’ will produce that year.

%G

This stands for the year with century corresponding to the current ISO week number.

%h

This is a synonym for ‘%b’.

%H

This stands for the hour (00–23).

%I

This stands for the hour (01–12).

%j

This stands for the day of the year (001–366).

%k

This stands for the hour (0–23), blank padded.

%l

This stands for the hour (1–12), blank padded.

%m

This stands for the month (01–12).

%M

This stands for the minute (00–59).

%n

This stands for a newline.

%N

This stands for the nanoseconds (000000000–999999999). To ask for fewer digits, use ‘%3N’ for milliseconds, ‘%6N’ for microseconds, etc. Any excess digits are discarded, without rounding.

%p

This stands for ‘AM’ or ‘PM’, as appropriate.

%q

This stands for the calendar quarter (1–4).

%r

This is a synonym for ‘%I:%M:%S %p’.

%R

This is a synonym for ‘%H:%M’.

%s

This stands for the integer number of seconds since the epoch.

%S

This stands for the second (00–59, or 00–60 on platforms that support leap seconds).

%t

This stands for a tab character.

%T

This is a synonym for ‘%H:%M:%S’.

%u

This stands for the numeric day of week (1–7). Monday is day 1.

%U

This stands for the week of the year (01–52), assuming that weeks start on Sunday.

%V

This stands for the week of the year according to ISO 8601.

%w

This stands for the numeric day of week (0–6). Sunday is day 0.

%W

This stands for the week of the year (01–52), assuming that weeks start on Monday.

%x

This has a locale-specific meaning. In the default locale (named ‘C’), it is equivalent to ‘%D’.

%X

This has a locale-specific meaning. In the default locale (named ‘C’), it is equivalent to ‘%T’.

%y

This stands for the year without century (00–99).

%Y

This stands for the year with century.

%Z

This stands for the time zone abbreviation (e.g., ‘EST’).

%z

This stands for the time zone numerical offset. The ‘z’ can be preceded by one, two, or three colons; if plain ‘%z’ stands for ‘-0500’, then ‘%:z’ stands for ‘-05:00’, ‘%::z’ stands for ‘-05:00:00’, and ‘%:::z’ is like ‘%::z’ except it suppresses trailing instances of ‘:00’ so it stands for ‘-05’ in the same example.

%%

This stands for a single ‘%’.

One or more flag characters can appear immediately after the ‘%’. ‘0’ pads with zeros, ‘+’ pads with zeros and also puts ‘+’ before nonnegative year numbers with more than four digits, ‘_’ pads with blanks, ‘-’ suppresses padding, ‘^’ upper-cases letters, and ‘#’ reverses the case of letters.

You can also specify the field width and type of padding for any of these ‘%’-sequences. This works as in printf: you write the field width as digits in a ‘%’-sequence, after any flags. For example, ‘%S’ specifies the number of seconds since the minute; ‘%03S’ means to pad this with zeros to 3 positions, ‘%_3S’ to pad with spaces to 3 positions. Plain ‘%3S’ pads with zeros, because that is how ‘%S’ normally pads to two positions.

The characters ‘E’ and ‘O’ act as modifiers when used after any flags and field widths in a ‘%’-sequence. ‘E’ specifies using the current locale’s alternative version of the date and time. In a Japanese locale, for example, %Ex might yield a date format based on the Japanese Emperors’ reigns. ‘E’ is allowed in ‘%Ec’, ‘%EC’, ‘%Ex’, ‘%EX’, ‘%Ey’, and ‘%EY’.

O’ means to use the current locale’s alternative representation of numbers, instead of the ordinary decimal digits. This is allowed with most letters, all the ones that output numbers.

To help debug programs, unrecognized ‘%’-sequences stand for themselves and are output as-is. Programs should not rely on this behavior, as future versions of Emacs may recognize new ‘%’-sequences as extensions.

This function uses the C library function strftime (see Formatting Calendar Time in The GNU C Library Reference Manual) to do most of the work. In order to communicate with that function, it first converts time and zone to internal form; the operating system limits the range of time and zone values. This function also encodes format-string using the coding system specified by locale-coding-system (see Locales); after strftime returns the resulting string, this function decodes the string using that same coding system.

Function: format-seconds format-string seconds

This function converts its argument seconds into a string of years, days, hours, etc., according to format-string. The argument format-string may contain ‘%’-sequences which control the conversion. Here is a table of what the ‘%’-sequences mean:

%y
%Y

The integer number of 365-day years.

%d
%D

The integer number of days.

%h
%H

The integer number of hours.

%m
%M

The integer number of minutes.

%s
%S

The number of seconds. If the optional ‘,’ parameter is used, it’s a floating point number, and the number after the ‘,’ specifies how many decimals to be used. ‘%,2s’ means “use two decimals”.

%z

Non-printing control flag. When it is used, other specifiers must be given in the order of decreasing size, i.e., years before days, hours before minutes, etc. Nothing will be produced in the result string to the left of ‘%z’ until the first non-zero conversion is encountered. For example, the default format used by emacs-uptime (see emacs-uptime) "%Y, %D, %H, %M, %z%S" means that the number of seconds will always be produced, but years, days, hours, and minutes will only be shown if they are non-zero.

%x

Non-printing control flag that works along the same lines as ‘%z’, but instead suppresses printing of trailing zero-value time elements.

%%

Produces a literal ‘%’.

Upper-case format sequences produce the units in addition to the numbers, lower-case formats produce only the numbers.

You can also specify the field width by following the ‘%’ with a number; shorter numbers will be padded with blanks. An optional period before the width requests zero-padding instead. For example, "%.3Y" might produce "004 years".


42.9 Processor Run time

Emacs provides several functions and primitives that return time, both elapsed and processor time, used by the Emacs process.

Command: emacs-uptime &optional format

This function returns a string representing the Emacs uptime—the elapsed wall-clock time this instance of Emacs is running. The string is formatted by format-seconds according to the optional argument format. For the available format descriptors, see format-seconds. If format is nil or omitted, it defaults to "%Y, %D, %H, %M, %z%S".

When called interactively, it prints the uptime in the echo area.

Function: get-internal-run-time

This function returns the processor run time used by Emacs, as a Lisp timestamp (see Time of Day).

Note that the time returned by this function excludes the time Emacs was not using the processor, and if the Emacs process has several threads, the returned value is the sum of the processor times used up by all Emacs threads.

If the system doesn’t provide a way to determine the processor run time, get-internal-run-time returns the same time as current-time.

Command: emacs-init-time

This function returns the duration of the Emacs initialization (see Summary: Sequence of Actions at Startup) in seconds, as a string. When called interactively, it prints the duration in the echo area.


42.10 Time Calculations

These functions perform calendrical computations using time values (see Time of Day). As with any time value, a value of nil for any of their time-value arguments stands for the current system time, and a finite number stands for the number of seconds since the epoch.

Function: time-less-p t1 t2

This returns t if the time value t1 is less than the time value t2.

Function: time-equal-p t1 t2

This returns t if the two time values t1 and t2 are equal. The result is nil if either argument is a NaN. For the purpose of comparison, a nil argument represents the current time with infinite resolution, so this function returns nil if one argument is nil and the other is not, and callers can therefore use nil to represent an unknown time value that does not equal any timestamp.

Function: time-subtract t1 t2

This returns the time difference t1t2 between two time values, as a Lisp timestamp. The result is exact and its clock resolution is no worse than the worse of its two arguments’ resolutions. If you need the difference in units of elapsed seconds, you can convert it with time-convert or float-time. See Time Conversion.

Function: time-add t1 t2

This returns the sum of two time values, using the same conversion rules as time-subtract. One argument should represent a time difference rather than a point in time, as a time value that is often just a single number of elapsed seconds. Here is how to add a number of seconds to a time value:

(time-add time seconds)
Function: time-to-days time-value

This function returns the number of days between the beginning of year 1 and time-value, assuming the default time zone. The operating system limits the range of time and zone values.

Function: days-to-time days

This is not quite the inverse of the time-to-days function, as it uses the Emacs epoch (instead of the year 1) for historical reasons. To get the inverse, subtract (time-to-days 0) from days, in which case days-to-time may return nil if days is negative.

Function: time-to-day-in-year time-value

This returns the day number within the year corresponding to time-value, assuming the default time zone. The operating system limits the range of time and zone values.

Function: date-leap-year-p year

This function returns t if year is a leap year.

Function: date-days-in-month year month

Return the number of days in month in year. For instance, February 2020 has 29 days.

Function: date-ordinal-to-time year ordinal

Return the date of ordinal in year as a decoded time structure. For instance, the 120th day in 2004 is April 29th.


42.11 Timers for Delayed Execution

You can set up a timer to call a function at a specified future time or after a certain length of idleness. A timer is a special object that stores the information about the next invocation times and the function to invoke.

Function: timerp object

This predicate function returns non-nil if object is a timer.

Emacs cannot run timers at any arbitrary point in a Lisp program; it can run them only when Emacs could accept output from a subprocess: namely, while waiting or inside certain primitive functions such as sit-for or read-event which can wait. Therefore, a timer’s execution may be delayed if Emacs is busy. However, the time of execution is very precise if Emacs is idle.

Emacs binds inhibit-quit to t before calling the timer function, because quitting out of many timer functions can leave things in an inconsistent state. This is normally unproblematical because most timer functions don’t do a lot of work. Indeed, for a timer to call a function that takes substantial time to run is likely to be annoying. If a timer function needs to allow quitting, it should use with-local-quit (see Quitting). For example, if a timer function calls accept-process-output to receive output from an external process, that call should be wrapped inside with-local-quit, to ensure that C-g works if the external process hangs.

It is usually a bad idea for timer functions to alter buffer contents. When they do, they usually should call undo-boundary both before and after changing the buffer, to separate the timer’s changes from user commands’ changes and prevent a single undo entry from growing to be quite large.

Timer functions should also avoid calling functions that cause Emacs to wait, such as sit-for (see Waiting for Elapsed Time or Input). This can lead to unpredictable effects, since other timers (or even the same timer) can run while waiting. If a timer function needs to perform an action after a certain time has elapsed, it can do this by scheduling a new timer.

If a timer function performs a remote file operation, it can be in conflict with an already running remote file operation of the same connection. Such conflicts are detected, and they result in a remote-file-error error (see Standard Errors). This should be protected by wrapping the timer function body with

(ignore-error 'remote-file-error
  …)

If a timer function calls functions that can change the match data, it should save and restore the match data. See Saving and Restoring the Match Data.

Command: run-at-time time repeat function &rest args

This sets up a timer that calls the function function with arguments args at time time. If repeat is a number (integer or floating point), the timer is scheduled to run again every repeat seconds after time. If repeat is nil, the timer runs only once.

time may specify an absolute or a relative time.

Absolute times may be specified using a string with a limited variety of formats, and are taken to be times today, even if already in the past. The recognized forms are ‘xxxx’, ‘x:xx’, or ‘xx:xx’ (military time), and ‘xxam’, ‘xxAM’, ‘xxpm’, ‘xxPM’, ‘xx:xxam’, ‘xx:xxAM’, ‘xx:xxpm’, or ‘xx:xxPM’. A period can be used instead of a colon to separate the hour and minute parts.

To specify a relative time as a string, use numbers followed by units. For example:

1 min

denotes 1 minute from now.

1 min 5 sec

denotes 65 seconds from now.

1 min 2 sec 3 hour 4 day 5 week 6 fortnight 7 month 8 year

denotes exactly 103 months, 123 days, and 10862 seconds from now.

For relative time values, Emacs considers a month to be exactly thirty days, and a year to be exactly 365.25 days.

Not all convenient formats are strings. If time is a number (integer or floating point), that specifies a relative time measured in seconds. The result of encode-time can also be used to specify an absolute value for time.

In most cases, repeat has no effect on when first call takes place—time alone specifies that. There is one exception: if time is t, then the timer runs whenever the time is a multiple of repeat seconds after the epoch. This is useful for functions like display-time. For instance, the following will make function run at every “whole” minute (e.g., ‘11:03:00’, ‘11:04:00’, etc):

(run-at-time t 60 function)

If Emacs didn’t get any CPU time when the timer would have run (for example if the system was busy running another process or if the computer was sleeping or in a suspended state), the timer will run as soon as Emacs resumes and is idle.

The function run-at-time returns a timer value that identifies the particular scheduled future action. You can use this value to call cancel-timer (see below).

Command: run-with-timer secs repeat function &rest args

This is exactly the same as run-at-time (so see that definition for an explanation of the parameters; secs is passed as time to that function), but is meant to be used when the delay is specified in seconds.

A repeating timer nominally ought to run every repeat seconds, but remember that any invocation of a timer can be late. Lateness of one repetition has no effect on the scheduled time of the next repetition. For instance, if Emacs is busy computing for long enough to cover three scheduled repetitions of the timer, and then starts to wait, it will immediately call the timer function three times in immediate succession (presuming no other timers trigger before or between them). If you want a timer to run again no less than n seconds after the last invocation, don’t use the repeat argument. Instead, the timer function should explicitly reschedule the timer.

User Option: timer-max-repeats

This variable’s value specifies the maximum number of times to repeat calling a timer function in a row, when many previously scheduled calls were unavoidably delayed.

Macro: with-timeout (seconds timeout-forms…) body…

Execute body, but give up after seconds seconds. If body finishes before the time is up, with-timeout returns the value of the last form in body. If, however, the execution of body is cut short by the timeout, then with-timeout executes all the timeout-forms and returns the value of the last of them.

This macro works by setting a timer to run after seconds seconds. If body finishes before that time, it cancels the timer. If the timer actually runs, it terminates execution of body, then executes timeout-forms.

Since timers can run within a Lisp program only when the program calls a primitive that can wait, with-timeout cannot stop executing body while it is in the midst of a computation—only when it calls one of those primitives. So use with-timeout only with a body that waits for input, not one that does a long computation.

The function y-or-n-p-with-timeout provides a simple way to use a timer to avoid waiting too long for an answer. See Yes-or-No Queries.

Function: cancel-timer timer

This cancels the requested action for timer, which should be a timer—usually, one previously returned by run-at-time or run-with-idle-timer. This cancels the effect of that call to one of these functions; the arrival of the specified time will not cause anything special to happen.

The list-timers command lists all the currently active timers. The command c (timer-list-cancel) will cancel the timer on the line under point. You can sort the list by column using the command S (tabulated-list-sort).


42.12 Idle Timers

Here is how to set up a timer that runs when Emacs is idle for a certain length of time. Aside from how to set them up, idle timers work just like ordinary timers.

Command: run-with-idle-timer secs repeat function &rest args

Set up a timer which runs the next time Emacs is idle for secs seconds. The value of secs may be a number or a value of the type returned by current-idle-time.

If repeat is nil, the timer runs just once, the first time Emacs remains idle for a long enough time. More often repeat is non-nil, which means to run the timer each time Emacs remains idle for secs seconds.

The function run-with-idle-timer returns a timer value which you can use in calling cancel-timer (see Timers for Delayed Execution).

Emacs becomes idle when it starts waiting for user input (unless it waits for input with a timeout, see Reading One Event), and it remains idle until the user provides some input. If a timer is set for five seconds of idleness, it runs approximately five seconds after Emacs first becomes idle. Even if repeat is non-nil, this timer will not run again as long as Emacs remains idle, because the duration of idleness will continue to increase and will not go down to five seconds again.

Emacs can do various things while idle: garbage collect, autosave or handle data from a subprocess. But these interludes during idleness do not interfere with idle timers, because they do not reset the clock of idleness to zero. An idle timer set for 600 seconds will run when ten minutes have elapsed since the last user command was finished, even if subprocess output has been accepted thousands of times within those ten minutes, and even if there have been garbage collections and autosaves.

When the user supplies input, Emacs becomes non-idle while executing the input. Then it becomes idle again, and all the idle timers that are set up to repeat will subsequently run another time, one by one.

Do not write an idle timer function containing a loop which does a certain amount of processing each time around, and exits when (input-pending-p) is non-nil. This approach seems very natural but has two problems:

  • It blocks out all process output (since Emacs accepts process output only while waiting).
  • It blocks out any idle timers that ought to run during that time.

Similarly, do not write an idle timer function that sets up another idle timer (including the same idle timer) with secs argument less than or equal to the current idleness time. Such a timer will run almost immediately, and continue running again and again, instead of waiting for the next time Emacs becomes idle. The correct approach is to reschedule with an appropriate increment of the current value of the idleness time, as described below.

Function: current-idle-time

If Emacs is idle, this function returns the length of time Emacs has been idle, using the same format as current-time (see Time of Day).

When Emacs is not idle, current-idle-time returns nil. This is a convenient way to test whether Emacs is idle.

The main use of current-idle-time is when an idle timer function wants to “take a break” for a while. It can set up another idle timer to call the same function again, after a few seconds more idleness. Here’s an example:

(defvar my-resume-timer nil
  "Timer for `my-timer-function' to reschedule itself, or nil.")

(defun my-timer-function ()
  ;; If the user types a command while my-resume-timer
  ;; is active, the next time this function is called from
  ;; its main idle timer, deactivate my-resume-timer.
  (when my-resume-timer
    (cancel-timer my-resume-timer))
  ...do the work for a while...
  (when taking-a-break
    (setq my-resume-timer
          (run-with-idle-timer
            ;; Compute an idle time break-length
            ;; more than the current value.
            (time-add (current-idle-time) break-length)
            nil
            'my-timer-function))))

42.13 Terminal Input

This section describes functions and variables for recording or manipulating terminal input. See Emacs Display, for related functions.


42.13.1 Input Modes

Function: set-input-mode interrupt flow meta &optional quit-char

This function sets the mode for reading keyboard input. If interrupt is non-nil, then Emacs uses input interrupts. If it is nil, then it uses CBREAK mode. The default setting is system-dependent. Some systems always use CBREAK mode regardless of what is specified.

When Emacs communicates directly with X, it ignores this argument and uses interrupts if that is the way it knows how to communicate.

If flow is non-nil, then Emacs uses XON/XOFF (C-q, C-s) flow control for output to the terminal. This has no effect except in CBREAK mode.

The argument meta controls support for input character codes above 127. If meta is t, Emacs converts characters with the 8th bit set into Meta characters, before it decodes them as needed (see Terminal I/O Encoding). If meta is nil, Emacs disregards the 8th bit; this is necessary when the terminal uses it as a parity bit. If meta is the symbol encoded, Emacs first decodes the characters using all the 8 bits of each byte, and then converts the decoded single-byte characters into Meta characters if they have their eighth bit set. Finally, if meta is neither t nor nil nor encoded, Emacs uses all 8 bits of input unchanged, both before and after decoding them. This is good for terminals that use 8-bit character sets and don’t encode the Meta modifier as the eighth bit.

If quit-char is non-nil, it specifies the character to use for quitting. Normally this character is C-g. See Quitting.

The current-input-mode function returns the input mode settings Emacs is currently using.

Function: current-input-mode

This function returns the current mode for reading keyboard input. It returns a list, corresponding to the arguments of set-input-mode, of the form (interrupt flow meta quit) in which:

interrupt

is non-nil when Emacs is using interrupt-driven input. If nil, Emacs is using CBREAK mode.

flow

is non-nil if Emacs uses XON/XOFF (C-q, C-s) flow control for output to the terminal. This value is meaningful only when interrupt is nil.

meta

is t if Emacs treats the eighth bit of input characters as the Meta bit before decoding input; encoded if Emacs treats the eighth bit of the decoded single-byte characters as the Meta bit; nil if Emacs clears the eighth bit of every input character; any other value means Emacs uses all eight bits as the basic character code.

quit

is the character Emacs currently uses for quitting, usually C-g.


42.13.2 Recording Input

Function: recent-keys &optional include-cmds

This function returns a vector containing the last 300 input events from the keyboard or mouse. All input events are included, whether or not they were used as parts of key sequences. Thus, you always get the last 300 input events, not counting events generated by keyboard macros. (These are excluded because they are less interesting for debugging; it should be enough to see the events that invoked the macros.)

If include-cmds is non-nil, complete key sequences in the result vector are interleaved with pseudo-events of the form (nil . COMMAND), where COMMAND is the binding of the key sequence (see Command Loop Overview).

A call to clear-this-command-keys (see Information from the Command Loop) causes this function to return an empty vector immediately afterward.

Command: open-dribble-file filename

This function opens a dribble file named filename. When a dribble file is open, each input event from the keyboard or mouse (but not those from keyboard macros) is written in that file. A non-character event is expressed using its printed representation surrounded by ‘<…>’. Be aware that sensitive information (such as passwords) may end up recorded in the dribble file.

You close the dribble file by calling this function with an argument of nil.

See also the open-termscript function (see Terminal Output).


42.14 Terminal Output

The terminal output functions send output to a text terminal, or keep track of output sent to the terminal. The variable baud-rate tells you what Emacs thinks is the output speed of the terminal.

User Option: baud-rate

This variable’s value is the output speed of the terminal, as far as Emacs knows. Setting this variable does not change the speed of actual data transmission, but the value is used for calculations such as padding.

It also affects decisions about whether to scroll part of the screen or repaint on text terminals. See Forcing Redisplay, for the corresponding functionality on graphical terminals.

The value is measured in baud.

If you are running across a network, and different parts of the network work at different baud rates, the value returned by Emacs may be different from the value used by your local terminal. Some network protocols communicate the local terminal speed to the remote machine, so that Emacs and other programs can get the proper value, but others do not. If Emacs has the wrong value, it makes decisions that are less than optimal. To fix the problem, set baud-rate.

Function: send-string-to-terminal string &optional terminal

This function sends string to terminal without alteration. Control characters in string have terminal-dependent effects. (If you need to display non-ASCII text on the terminal, encode it using one of the functions described in Explicit Encoding and Decoding.) This function operates only on text terminals. terminal may be a terminal object, a frame, or nil for the selected frame’s terminal. In batch mode, string is sent to stdout when terminal is nil.

One use of this function is to define function keys on terminals that have downloadable function key definitions. For example, this is how (on certain terminals) to define function key 4 to move forward four characters (by transmitting the characters C-u C-f to the computer):

(send-string-to-terminal "\eF4\^U\^F")
     ⇒ nil
Command: open-termscript filename

This function is used to open a termscript file that will record all the characters sent by Emacs to the terminal. It returns nil. Termscript files are useful for investigating problems where Emacs garbles the screen, problems that are due to incorrect Termcap entries or to undesirable settings of terminal options more often than to actual Emacs bugs. Once you are certain which characters were actually output, you can determine reliably whether they correspond to the Termcap specifications in use.

(open-termscript "../junk/termscript")
     ⇒ nil

You close the termscript file by calling this function with an argument of nil.

See also open-dribble-file in Recording Input.


42.15 Sound Output

To play sound using Emacs, use the function play-sound. Only certain systems are supported; if you call play-sound on a system which cannot really do the job, it gives an error.

The sound must be stored as a file in RIFF-WAVE format (‘.wav’) or Sun Audio format (‘.au’).

Function: play-sound sound

This function plays a specified sound. The argument, sound, has the form (sound properties...), where the properties consist of alternating keywords (particular symbols recognized specially) and values corresponding to them.

Here is a table of the keywords that are currently meaningful in sound, and their meanings:

:file file

This specifies the file containing the sound to play. If the file name is not absolute, it is expanded against the directory data-directory.

:data data

This specifies the sound to play without need to refer to a file. The value, data, should be a string containing the same bytes as a sound file. We recommend using a unibyte string.

:volume volume

This specifies how loud to play the sound. It should be a number in the range of 0 to 1. The default is to use whatever volume has been specified before.

:device device

This specifies the system device on which to play the sound, as a string. The default device is system-dependent.

Before actually playing the sound, play-sound calls the functions in the list play-sound-functions. Each function is called with one argument, sound.

Command: play-sound-file file &optional volume device

This function is an alternative interface to playing a sound file specifying an optional volume and device.

Variable: play-sound-functions

A list of functions to be called before playing a sound. Each function is called with one argument, a property list that describes the sound.


42.16 Operating on X11 Keysyms

To define system-specific X11 keysyms, set the variable system-key-alist.

Variable: system-key-alist

This variable’s value should be an alist with one element for each system-specific keysym. Each element has the form (code . symbol), where code is the numeric keysym code (not including the vendor-specific bit, −2**28), and symbol is the name for the function key.

For example (168 . mute-acute) defines a system-specific key (used by HP X servers) whose numeric code is −2**28 + 168.

It is not crucial to exclude from the alist the keysyms of other X servers; those do no harm, as long as they don’t conflict with the ones used by the X server actually in use.

The variable is always local to the current terminal, and cannot be buffer-local. See Multiple Terminals.

You can specify which keysyms Emacs should use for the Control, Meta, Alt, Hyper, and Super modifiers by setting these variables:

Variable: x-ctrl-keysym
Variable: x-alt-keysym
Variable: x-meta-keysym
Variable: x-hyper-keysym
Variable: x-super-keysym

The name of the keysym that should stand for the Control modifier (respectively, for Alt, Meta, Hyper, and Super). For example, here is how to swap the Meta and Alt modifiers within Emacs:

(setq x-alt-keysym 'meta)
(setq x-meta-keysym 'alt)

42.17 Batch Mode

The command-line option ‘-batch’ causes Emacs to run noninteractively. In this mode, Emacs does not read commands from the terminal, it does not alter the terminal modes, and it does not expect to be outputting to an erasable screen. The idea is that you specify Lisp programs to run; when they are finished, Emacs should exit. The way to specify the programs to run is with ‘-l file’, which loads the library named file, or ‘-f function’, which calls function with no arguments, or ‘--eval=form’.

Variable: noninteractive

This variable is non-nil when Emacs is running in batch mode.

If the specified Lisp program signals an unhandled error in batch mode, Emacs exits with a non-zero exit status after invoking the Lisp debugger which shows the Lisp backtrace (see Invoking the Debugger) on the standard error stream:

$ emacs -Q --batch --eval '(error "foo")'; echo $?

Error: error ("foo")
mapbacktrace(#f(compiled-function (evald func args flags) #<bytecode -0x4f85c5
7c45e2f81>))
debug-early-backtrace()
debug-early(error (error "foo"))
signal(error ("foo"))
error("foo")
eval((error "foo") t)
command-line-1(("--eval" "(error \"foo\")"))
command-line()
normal-top-level()
foo
255

Any Lisp program output that would normally go to the echo area, either using message, or using prin1, etc., with t as the stream (see Output Streams), goes instead to Emacs’s standard descriptors when in batch mode: message writes to the standard error descriptor, while prin1 and other print functions write to the standard output. Similarly, input that would normally come from the minibuffer is read from the standard input descriptor. Thus, Emacs behaves much like a noninteractive application program. (The echo area output that Emacs itself normally generates, such as command echoing, is suppressed entirely.)

Non-ASCII text written to the standard output or error descriptors is by default encoded using locale-coding-system (see Locales) if it is non-nil; this can be overridden by binding coding-system-for-write to a coding system of you choice (see Explicit Encoding and Decoding).

In batch mode, Emacs will enlarge the value of the gc-cons-percentage variable from the default of ‘0.1’ up to ‘1.0’. Batch jobs that are supposed to run for a long time should adjust the limit back down again, because this means that less garbage collection will be performed by default (and more memory consumed).


42.18 Session Management

Emacs supports the X Session Management Protocol, which is used to suspend and restart applications. In the X Window System, a program called the session manager is responsible for keeping track of the applications that are running. When the X server shuts down, the session manager asks applications to save their state, and delays the actual shutdown until they respond. An application can also cancel the shutdown.

When the session manager restarts a suspended session, it directs these applications to individually reload their saved state. It does this by specifying a special command-line argument that says what saved session to restore. For Emacs, this argument is ‘--smid session’.

Variable: emacs-save-session-functions

Emacs supports saving state via a hook called emacs-save-session-functions. Emacs runs this hook when the session manager tells it that the window system is shutting down. The functions are called with no arguments, and with the current buffer set to a temporary buffer. Each function can use insert to add Lisp code to this buffer. At the end, Emacs saves the buffer in a file, called the session file.

Subsequently, when the session manager restarts Emacs, it loads the session file automatically (see Loading). This is performed by a function named emacs-session-restore, which is called during startup. See Summary: Sequence of Actions at Startup.

If a function in emacs-save-session-functions returns non-nil, Emacs tells the session manager to cancel the shutdown.

Here is an example that just inserts some text into *scratch* when Emacs is restarted by the session manager.

(add-hook 'emacs-save-session-functions 'save-yourself-test)

(defun save-yourself-test ()
  (insert
   (format "%S" '(with-current-buffer "*scratch*"
                   (insert "I am restored"))))
  nil)

42.19 Desktop Notifications

Emacs is able to send notifications on systems that support the freedesktop.org Desktop Notifications Specification and on MS-Windows. In order to use this functionality on POSIX hosts, Emacs must have been compiled with D-Bus support, and the notifications library must be loaded. See D-Bus in D-Bus integration in Emacs. The following function is supported when D-Bus support is available:

Function: notifications-notify &rest params

This function sends a notification to the desktop via D-Bus, consisting of the parameters specified by the params arguments. These arguments should consist of alternating keyword and value pairs. The supported keywords and values are as follows:

:bus bus

The D-Bus bus. This argument is needed only if a bus other than :session shall be used.

:title title

The notification title.

:body text

The notification body text. Depending on the implementation of the notification server, the text could contain HTML markups, like ‘"<b>bold text</b>"’, hyperlinks, or images. Special HTML characters must be encoded, as ‘"Contact &lt;postmaster@localhost&gt;!"’.

:app-name name

The name of the application sending the notification. The default is notifications-application-name.

:replaces-id id

The notification id that this notification replaces. id must be the result of a previous notifications-notify call.

:app-icon icon-file

The file name of the notification icon. If set to nil, no icon is displayed. The default is notifications-application-icon.

:actions (key title key title ...)

A list of actions to be applied. key and title are both strings. The default action (usually invoked by clicking the notification) should have a key named ‘"default"’. The title can be anything, though implementations are free not to display it.

:timeout timeout

The timeout time in milliseconds since the display of the notification at which the notification should automatically close. If −1, the notification’s expiration time is dependent on the notification server’s settings, and may vary for the type of notification. If 0, the notification never expires. Default value is −1.

:urgency urgency

The urgency level. It can be low, normal, or critical.

:action-items

When this keyword is given, the title string of the actions is interpreted as icon name.

:category category

The type of notification this is, a string. See the Desktop Notifications Specification for a list of standard categories.

:desktop-entry filename

This specifies the name of the desktop filename representing the calling program, like ‘"emacs"’.

:image-data (width height rowstride has-alpha bits channels data)

This is a raw data image format that describes the width, height, rowstride, whether there is an alpha channel, bits per sample, channels and image data, respectively.

:image-path path

This is represented either as a URI (‘file://’ is the only URI schema supported right now) or a name in a freedesktop.org-compliant icon theme from ‘$XDG_DATA_DIRS/icons’.

:sound-file filename

The path to a sound file to play when the notification pops up.

:sound-name name

A themable named sound from the freedesktop.org sound naming specification from ‘$XDG_DATA_DIRS/sounds’, to play when the notification pops up. Similar to the icon name, only for sounds. An example would be ‘"message-new-instant"’.

:suppress-sound

Causes the server to suppress playing any sounds, if it has that ability.

:resident

When set the server will not automatically remove the notification when an action has been invoked. The notification will remain resident in the server until it is explicitly removed by the user or by the sender. This hint is likely only useful when the server has the :persistence capability.

:transient

When set the server will treat the notification as transient and by-pass the server’s persistence capability, if it should exist.

:x position
:y position

Specifies the X, Y location on the screen that the notification should point to. Both arguments must be used together.

:on-action function

Function to call when an action is invoked. The notification id and the key of the action are passed as arguments to the function.

:on-close function

Function to call when the notification has been closed by timeout or by the user. The function receive the notification id and the closing reason as arguments:

  • expired if the notification has expired
  • dismissed if the notification was dismissed by the user
  • close-notification if the notification was closed by a call to notifications-close-notification
  • undefined if the notification server hasn’t provided a reason

Which parameters are accepted by the notification server can be checked via notifications-get-capabilities.

This function returns a notification id, an integer, which can be used to manipulate the notification item with notifications-close-notification or the :replaces-id argument of another notifications-notify call. For example:

(defun my-on-action-function (id key)
  (message "Message %d, key \"%s\" pressed" id key))
     ⇒ my-on-action-function

(defun my-on-close-function (id reason)
  (message "Message %d, closed due to \"%s\"" id reason))
     ⇒ my-on-close-function

(notifications-notify
 :title "Title"
 :body "This is <b>important</b>."
 :actions '("Confirm" "I agree" "Refuse" "I disagree")
 :on-action 'my-on-action-function
 :on-close 'my-on-close-function)
     ⇒ 22

A message window opens on the desktop.  Press ``I agree''.
     ⇒ Message 22, key "Confirm" pressed
        Message 22, closed due to "dismissed"
Function: notifications-close-notification id &optional bus

This function closes a notification with identifier id. bus can be a string denoting a D-Bus connection, the default is :session.

Function: notifications-get-capabilities &optional bus

Returns the capabilities of the notification server, a list of symbols. bus can be a string denoting a D-Bus connection, the default is :session. The following capabilities can be expected:

:actions

The server will provide the specified actions to the user.

:body

Supports body text.

:body-hyperlinks

The server supports hyperlinks in the notifications.

:body-images

The server supports images in the notifications.

:body-markup

Supports markup in the body text.

:icon-multi

The server will render an animation of all the frames in a given image array.

:icon-static

Supports display of exactly 1 frame of any given image array. This value is mutually exclusive with :icon-multi.

:persistence

The server supports persistence of notifications.

:sound

The server supports sounds on notifications.

Further vendor-specific caps start with :x-vendor, like :x-gnome-foo-cap.

Function: notifications-get-server-information &optional bus

Return information on the notification server, a list of strings. bus can be a string denoting a D-Bus connection, the default is :session. The returned list is (name vendor version spec-version).

name

The product name of the server.

vendor

The vendor name. For example, ‘"KDE"’, ‘"GNOME"’.

version

The server’s version number.

spec-version

The specification version the server is compliant with.

If spec_version is nil, the server supports a specification prior to ‘"1.0"’.

When Emacs runs on MS-Windows as a GUI session, it supports a small subset of the D-Bus notifications functionality via a native primitive:

Function: w32-notification-notify &rest params

This function displays an MS-Windows tray notification as specified by params. MS-Windows tray notifications are displayed in a balloon from an icon in the notification area of the taskbar.

Value is the integer unique ID of the notification that can be used to remove the notification using w32-notification-close, described below. If the function fails, the return value is nil.

The arguments params are specified as keyword/value pairs. All the parameters are optional, but if no parameters are specified, the function will do nothing and return nil.

The following parameters are supported:

:icon icon

Display icon in the system tray. If icon is a string, it should specify a file name from which to load the icon; the specified file should be a .ico Windows icon file. If icon is not a string, or if this parameter is not specified, the standard Emacs icon will be used.

:tip tip

Use tip as the tooltip for the notification. If tip is a string, this is the text of a tooltip that will be shown when the mouse pointer hovers over the tray icon added by the notification. If tip is not a string, or if this parameter is not specified, the default tooltip text is ‘Emacs notification’. The tooltip text can be up to 127 characters long (63 on Windows versions before W2K). Longer strings will be truncated.

:level level

Notification severity level, one of info, warning, or error. If given, the value determines the icon displayed to the left of the notification title, but only if the :title parameter (see below) is also specified and is a string.

:title title

The title of the notification. If title is a string, it is displayed in a larger font immediately above the body text. The title text can be up to 63 characters long; longer text will be truncated.

:body body

The body of the notification. If body is a string, it specifies the text of the notification message. Use embedded newlines to control how the text is broken into lines. The body text can be up to 255 characters long, and will be truncated if it’s longer. Unlike with D-Bus, the body text should be plain text, with no markup.

Note that versions of Windows before W2K support only :icon and :tip. The other parameters can be passed, but they will be ignored on those old systems.

There can be at most one active notification at any given time. An active notification must be removed by calling w32-notification-close before a new one can be shown.

To remove the notification and its icon from the taskbar, use the following function:

Function: w32-notification-close id

This function removes the tray notification given by its unique id.


42.20 Notifications on File Changes

Several operating systems support watching of filesystems for changes to files or their attributes. If configured properly, Emacs links a respective library like inotify, kqueue, gfilenotify, or w32notify statically. These libraries enable watching of filesystems on the local machine.

It is also possible to watch filesystems on remote machines, see Remote Files in The GNU Emacs Manual. This does not depend on one of the libraries linked to Emacs.

Since all these libraries emit different events upon notified file changes, Emacs provides a special library filenotify which presents a unified interface to applications. Lisp programs that want to receive file notifications should always use this library in preference to the native ones. This section documents the filenotify library functions and variables.

Function: file-notify-add-watch file flags callback

Add a watch for filesystem events pertaining to file. This arranges for filesystem events pertaining to file to be reported to Emacs.

The returned value is a descriptor for the added watch. Its type depends on the underlying library, and in general cannot be assumed to be an integer as in the example below. It should be used for comparison by equal only.

If the file cannot be watched for some reason, this function signals a file-notify-error error.

Sometimes, mounted filesystems cannot be watched for file changes. This is not detected by this function, and so a non-nil return value does not guarantee that changes on file will be actually notified.

If file is a symlink, it doesn’t follow that link. Just file itself will be watched.

flags is a list of conditions to set what will be watched for. It can include the following symbols:

change

watch for changes in file’s contents

attribute-change

watch for changes in file attributes, like permissions or modification time

If file is a directory, change watches for file creation and deletion in that directory. Some of the native file notification libraries also report file changes in that case. This does not work recursively.

When any event happens, Emacs will call the callback function passing it a single argument event, which is of the form

(descriptor action file [file1])

descriptor is the same object as the one returned by this function. action is the description of the event. It could be any one of the following symbols:

created

file was created

deleted

file was deleted

changed

file’s contents has changed; with w32notify library, reports attribute changes as well

renamed

file has been renamed to file1

attribute-changed

a file attribute was changed

stopped

watching file has stopped

Note that the w32notify library does not report attribute-changed events. When some file’s attribute, like permissions or modification time, has changed, this library reports a changed event. Likewise, the kqueue library does not reliably report file attribute changes when watching a directory.

The stopped event means that watching the file has been discontinued. This could be because file-notify-rm-watch was called (see below), or because the file being watched was deleted, or due to another error reported from the underlying library which makes further watching impossible.

file and file1 are the name of the file(s) whose event is being reported. For example:

(require 'filenotify)
     ⇒ filenotify

(defun my-notify-callback (event)
  (message "Event %S" event))
     ⇒ my-notify-callback

(file-notify-add-watch
  "/tmp" '(change attribute-change) 'my-notify-callback)
     ⇒ 35025468

(write-region "foo" nil "/tmp/foo")
     ⇒ Event (35025468 created "/tmp/.#foo")
        Event (35025468 created "/tmp/foo")
        Event (35025468 changed "/tmp/foo")
        Event (35025468 deleted "/tmp/.#foo")

(write-region "bla" nil "/tmp/foo")
     ⇒ Event (35025468 created "/tmp/.#foo")
        Event (35025468 changed "/tmp/foo")
        Event (35025468 deleted "/tmp/.#foo")

(set-file-modes "/tmp/foo" (default-file-modes) 'nofollow)
     ⇒ Event (35025468 attribute-changed "/tmp/foo")

Whether the action renamed is returned depends on the used watch library. Otherwise, the actions deleted and created could be returned in a random order.

(rename-file "/tmp/foo" "/tmp/bla")
     ⇒ Event (35025468 renamed "/tmp/foo" "/tmp/bla")

(delete-file "/tmp/bla")
     ⇒ Event (35025468 deleted "/tmp/bla")
Function: file-notify-rm-watch descriptor

Removes an existing file watch specified by its descriptor. descriptor should be an object returned by file-notify-add-watch.

Command: file-notify-rm-all-watches

Removes all existing file notification watches from Emacs.

Use this command with caution, because it could have unexpected side effects on packages relying on file watches. It is intended mainly for debugging purposes, or when Emacs has been stalled.

Function: file-notify-valid-p descriptor

Checks a watch specified by its descriptor for validity. descriptor should be an object returned by file-notify-add-watch.

A watch can become invalid if the file or directory it watches is deleted, or if the watcher thread exits abnormally for any other reason. Removing the watch by calling file-notify-rm-watch also makes it invalid.

(make-directory "/tmp/foo")
     ⇒ Event (35025468 created "/tmp/foo")

(setq desc
      (file-notify-add-watch
        "/tmp/foo" '(change) 'my-notify-callback))
     ⇒ 11359632

(file-notify-valid-p desc)
     ⇒ t

(write-region "bla" nil "/tmp/foo/bla")
     ⇒ Event (11359632 created "/tmp/foo/.#bla")
        Event (11359632 created "/tmp/foo/bla")
        Event (11359632 changed "/tmp/foo/bla")
        Event (11359632 deleted "/tmp/foo/.#bla")

;; Deleting a file in the directory doesn't invalidate the watch.
(delete-file "/tmp/foo/bla")
     ⇒ Event (11359632 deleted "/tmp/foo/bla")

(write-region "bla" nil "/tmp/foo/bla")
     ⇒ Event (11359632 created "/tmp/foo/.#bla")
        Event (11359632 created "/tmp/foo/bla")
        Event (11359632 changed "/tmp/foo/bla")
        Event (11359632 deleted "/tmp/foo/.#bla")

;; Deleting the directory invalidates the watch.
;; Events arrive for different watch descriptors.
(delete-directory "/tmp/foo" 'recursive)
     ⇒ Event (35025468 deleted "/tmp/foo")
        Event (11359632 deleted "/tmp/foo/bla")
        Event (11359632 deleted "/tmp/foo")
        Event (11359632 stopped "/tmp/foo")

(file-notify-valid-p desc)
     ⇒ nil

42.21 Dynamically Loaded Libraries

A dynamically loaded library is a library that is loaded on demand, when its facilities are first needed. Emacs supports such on-demand loading of support libraries for some of its features.

Variable: dynamic-library-alist

This is an alist of dynamic libraries and external library files implementing them.

Each element is a list of the form (library files…), where the car is a symbol representing a supported external library, and the rest are strings giving alternate filenames for that library.

Emacs tries to load the library from the files in the order they appear in the list; if none is found, the Emacs session won’t have access to that library, and the features it provides will be unavailable.

Image support on some platforms uses this facility. Here’s an example of setting this variable for supporting images on MS-Windows:

(setq dynamic-library-alist
      '((xpm "libxpm.dll" "xpm4.dll" "libXpm-nox4.dll")
        (png "libpng12d.dll" "libpng12.dll" "libpng.dll"
             "libpng13d.dll" "libpng13.dll")
        (jpeg "jpeg62.dll" "libjpeg.dll" "jpeg-62.dll"
              "jpeg.dll")
        (tiff "libtiff3.dll" "libtiff.dll")
        (gif "giflib4.dll" "libungif4.dll" "libungif.dll")
        (svg "librsvg-2-2.dll")
        (gdk-pixbuf "libgdk_pixbuf-2.0-0.dll")
        (glib "libglib-2.0-0.dll")
        (gobject "libgobject-2.0-0.dll")))

Note that image types pbm and xbm do not need entries in this variable because they do not depend on external libraries and are always available in Emacs.

Also note that this variable is not meant to be a generic facility for accessing external libraries; only those already known by Emacs can be loaded through it.

This variable is ignored if the given library is statically linked into Emacs.


42.22 Security Considerations

Like any application, Emacs can be run in a secure environment, where the operating system enforces rules about access and the like. With some care, Emacs-based applications can also be part of a security perimeter that checks such rules. Although the default settings for Emacs work well for a typical software development environment, they may require adjustment in environments containing untrusted users that may include attackers. Here is a compendium of security issues that may be helpful if you are developing such applications. It is by no means complete; it is intended to give you an idea of the security issues involved, rather than to be a security checklist.

File local variables

A file that Emacs visits can contain variable settings that affect the buffer visiting that file; See File Local Variables. Similarly, a directory can specify local variable values common to all files in that directory; see Directory Local Variables. Although Emacs takes some effort to protect against misuse of these variables, a security hole can be created merely by a package setting safe-local-variable too optimistically, a problem that is all too common. To disable this feature for both files and directories, set enable-local-variables to nil.

Access control

Although Emacs normally respects access permissions of the underlying operating system, in some cases it handles accesses specially. For example, file names can have handlers that treat the files specially, with their own access checking. See Making Certain File Names “Magic”. Also, a buffer can be read-only even if the corresponding file is writable, and vice versa, which can result in messages such as ‘File passwd is write-protected; try to save anyway? (yes or no)’. See Read-Only Buffers.

Authentication

Emacs has several functions that deal with passwords, e.g., read-passwd. See Reading a Password. Although these functions do not attempt to broadcast passwords to the world, their implementations are not proof against determined attackers with access to Emacs internals. For example, even if Elisp code uses clear-string to scrub a password from its memory after using it, remnants of the password may still reside in the garbage-collected free list. See Modifying Strings.

Code injection

Emacs can send commands to many other applications, and applications should take care that strings sent as operands of these commands are not misinterpreted as directives. For example, when using a shell command to rename a file a to b, do not simply use the string mv a b, because either file name might start with ‘-’, or might contain shell metacharacters like ‘;’. Although functions like shell-quote-argument can help avoid this sort of problem, they are not panaceas; for example, on a POSIX platform shell-quote-argument quotes shell metacharacters but not leading ‘-’. On MS-Windows, quoting for ‘%’ assumes none of the environment variables have ‘^’ in their name. See Shell Arguments. Typically it is safer to use call-process than a subshell. See Creating a Synchronous Process. And it is safer yet to use builtin Emacs functions; for example, use (rename-file "a" "b" t) instead of invoking mv. See Changing File Names and Attributes.

Coding systems

Emacs attempts to infer the coding systems of the files and network connections it accesses. See Coding Systems. If Emacs infers incorrectly, or if the other parties to the network connection disagree with Emacs’s inferences, the resulting system could be unreliable. Also, even when it infers correctly, Emacs often can use bytes that other programs cannot. For example, although to Emacs the null byte is just a character like any other, many other applications treat it as a string terminator and mishandle strings or files containing null bytes.

Environment and configuration variables

POSIX specifies several environment variables that can affect how Emacs behaves. Any environment variable whose name consists entirely of uppercase ASCII letters, digits, and the underscore may affect the internal behavior of Emacs. Emacs uses several such variables, e.g., EMACSLOADPATH. See Library Search. On some platforms some environment variables (e.g., PATH, POSIXLY_CORRECT, SHELL, TMPDIR) need to have properly-configured values in order to get standard behavior for any utility Emacs might invoke. Even seemingly-benign variables like TZ may have security implications. See Operating System Environment.

Emacs has customization and other variables with similar considerations. For example, if the variable shell-file-name specifies a shell with nonstandard behavior, an Emacs-based application may misbehave.

Installation

When Emacs is installed, if the installation directory hierarchy can be modified by untrusted users, the application cannot be trusted. This applies also to the directory hierarchies of the programs that Emacs uses, and of the files that Emacs reads and writes.

Network access

Emacs often accesses the network, and you may want to configure it to avoid network accesses that it would normally do. For example, unless you set tramp-mode to nil, file names using a certain syntax are interpreted as being network files, and are retrieved across the network. See The Tramp Manual in The Tramp Manual.

Race conditions

Emacs applications have the same sort of race-condition issues that other applications do. For example, even when (file-readable-p "foo.txt") returns t, it could be that foo.txt is unreadable because some other program changed the file’s permissions between the call to file-readable-p and now. See Testing Accessibility.

Resource limits

When Emacs exhausts memory or other operating system resources, its behavior can be less reliable, in that computations that ordinarily run to completion may abort back to the top level. This may cause Emacs to neglect operations that it normally would have done.


43 Preparing Lisp code for distribution

Emacs provides a standard way to distribute Emacs Lisp code to users. A package is a collection of one or more files, formatted and bundled in such a way that users can easily download, install, uninstall, and upgrade it.

The following sections describe how to create a package, and how to put it in a package archive for others to download. See Packages in The GNU Emacs Manual, for a description of user-level features of the packaging system.

These sections are mostly directed towards package archive maintainers—much of this information is not relevant for package authors (i.e., people who write code that will be distributed via these archives).


43.1 Packaging Basics

A package is either a simple package or a multi-file package. A simple package is stored in a package archive as a single Emacs Lisp file, while a multi-file package is stored as a tar file (containing multiple Lisp files, and possibly non-Lisp files such as a manual).

In ordinary usage, the difference between simple packages and multi-file packages is relatively unimportant; the Package Menu interface makes no distinction between them. However, the procedure for creating them differs, as explained in the following sections.

Each package (whether simple or multi-file) has certain attributes:

Name

A short word (e.g., ‘auctex’). This is usually also the symbol prefix used in the program (see Emacs Lisp Coding Conventions).

Version

A version number, in a form that the function version-to-list understands (e.g., ‘11.86’). Each release of a package should be accompanied by an increase in the version number so that it will be recognized as an upgrade by users querying the package archive.

Brief description

This is shown when the package is listed in the Package Menu. It should occupy a single line, ideally in 36 characters or less.

Long description

This is shown in the buffer created by C-h P (describe-package), following the package’s brief description and installation status. It normally spans multiple lines, and should fully describe the package’s capabilities and how to begin using it once it is installed.

Dependencies

A list of other packages (possibly including minimal acceptable version numbers) on which this package depends. The list may be empty, meaning this package has no dependencies. Otherwise, installing this package also automatically installs its dependencies, recursively; if any dependency cannot be found, the package cannot be installed.

Installing a package, either via the command package-install-file, or via the Package Menu, creates a subdirectory of package-user-dir named name-version, where name is the package’s name and version its version (e.g., ~/.emacs.d/elpa/auctex-11.86/). We call this the package’s content directory. It is where Emacs puts the package’s contents (the single Lisp file for a simple package, or the files extracted from a multi-file package).

Emacs then searches every Lisp file in the content directory for autoload magic comments (see Autoload). These autoload definitions are saved to a file named name-autoloads.el in the content directory. They are typically used to autoload the principal user commands defined in the package, but they can also perform other tasks, such as adding an element to auto-mode-alist (see How Emacs Chooses a Major Mode). Note that a package typically does not autoload every function and variable defined within it—only the handful of commands typically called to begin using the package. Emacs then byte-compiles every Lisp file in the package.

After installation, the installed package is loaded: Emacs adds the package’s content directory to load-path, and evaluates the autoload definitions in name-autoloads.el.

Whenever Emacs starts up, it automatically calls the function package-activate-all to make installed packages available to the current session. This is done after loading the early init file, but before loading the regular init file (see Summary: Sequence of Actions at Startup). Packages are not automatically made available if the user option package-enable-at-startup is set to nil in the early init file.

Function: package-activate-all

This function makes the packages available to the current session. The user option package-load-list specifies which packages to make available; by default, all installed packages are made available. See Package Installation in The GNU Emacs Manual.

In most cases, you should not need to call package-activate-all, as this is done automatically during startup. Simply make sure to put any code that should run before package-activate-all in the early init file, and any code that should run after it in the primary init file (see Init File in The GNU Emacs Manual).

Command: package-initialize &optional no-activate

This function initializes Emacs’ internal record of which packages are installed, and then calls package-activate-all.

The optional argument no-activate, if non-nil, causes Emacs to update its record of installed packages without actually making them available.


43.2 Simple Packages

A simple package consists of a single Emacs Lisp source file. The file must conform to the Emacs Lisp library header conventions (see Conventional Headers for Emacs Libraries). The package’s attributes are taken from the various headers, as illustrated by the following example:

;;; superfrobnicator.el --- Frobnicate and bifurcate flanges  -*- lexical-binding:t -*-

;; Copyright (C) 2022 Free Software Foundation, Inc.

;; Author: J. R. Hacker <jrh@example.com>
;; Version: 1.3
;; Package-Requires: ((flange "1.0"))
;; Keywords: multimedia, hypermedia
;; URL: https://example.com/jrhacker/superfrobnicate

…

;;; Commentary:

;; This package provides a minor mode to frobnicate and/or
;; bifurcate any flanges you desire.  To activate it, just type
…

;;;###autoload
(define-minor-mode superfrobnicator-mode
…

The name of the package is the same as the base name of the file, as written on the first line. Here, it is ‘superfrobnicator’.

The brief description is also taken from the first line. Here, it is ‘Frobnicate and bifurcate flanges’.

The version number comes from the ‘Package-Version’ header, if it exists, or from the ‘Version’ header otherwise. One or the other must be present. Here, the version number is 1.3.

If the file has a ‘;;; Commentary:’ section, this section is used as the long description. (When displaying the description, Emacs omits the ‘;;; Commentary:’ line, as well as the leading comment characters in the commentary itself.)

If the file has a ‘Package-Requires’ header, that is used as the package dependencies. In the above example, the package depends on the ‘flange’ package, version 1.0 or higher. See Conventional Headers for Emacs Libraries, for a description of the ‘Package-Requires’ header. If the header is omitted, the package has no dependencies.

The ‘Keywords’ and ‘URL’ headers are optional, but recommended. The command describe-package uses these to add links to its output. The ‘Keywords’ header should contain at least one standard keyword from the finder-known-keywords list.

The file ought to also contain one or more autoload magic comments, as explained in Packaging Basics. In the above example, a magic comment autoloads superfrobnicator-mode.

See Creating and Maintaining Package Archives, for an explanation of how to add a single-file package to a package archive.


43.3 Multi-file Packages

A multi-file package is less convenient to create than a single-file package, but it offers more features: it can include multiple Emacs Lisp files, an Info manual, and other file types (such as images).

Prior to installation, a multi-file package is stored in a package archive as a tar file. The tar file must be named name-version.tar, where name is the package name and version is the version number. Its contents, once extracted, must all appear in a directory named name-version, the content directory (see Packaging Basics). Files may also extract into subdirectories of the content directory.

One of the files in the content directory must be named name-pkg.el. It must contain a single Lisp form, consisting of a call to the function define-package, described below. This defines the package’s attributes: version, brief description, and requirements.

For example, if we distribute version 1.3 of the superfrobnicator as a multi-file package, the tar file would be superfrobnicator-1.3.tar. Its contents would extract into the directory superfrobnicator-1.3, and one of these would be the file superfrobnicator-pkg.el.

Function: define-package name version &optional docstring requirements

This function defines a package. name is the package name, a string. version is the version, as a string of a form that can be understood by the function version-to-list. docstring is the brief description.

requirements is a list of required packages and their versions. Each element in this list should have the form (dep-name dep-version), where dep-name is a symbol whose name is the dependency’s package name, and dep-version is the dependency’s version (a string).

If the content directory contains a file named README, this file is used as the long description (overriding any ‘;;; Commentary:’ section).

If the content directory contains a file named dir, this is assumed to be an Info directory file made with install-info. See Invoking install-info in Texinfo. The relevant Info files should also be present in the content directory. In this case, Emacs will automatically add the content directory to Info-directory-list when the package is activated.

Do not include any .elc files in the package. Those are created when the package is installed. Note that there is no way to control the order in which files are byte-compiled.

Do not include any file named name-autoloads.el. This file is reserved for the package’s autoload definitions (see Packaging Basics). It is created automatically when the package is installed, by searching all the Lisp files in the package for autoload magic comments.

If the multi-file package contains auxiliary data files (such as images), the package’s Lisp code can refer to these files via the variable load-file-name (see Loading). Here is an example:

(defconst superfrobnicator-base (file-name-directory load-file-name))

(defun superfrobnicator-fetch-image (file)
  (expand-file-name file superfrobnicator-base))

43.4 Creating and Maintaining Package Archives

Via the Package Menu, users may download packages from package archives. Such archives are specified by the variable package-archives, whose default value lists the archives hosted on GNU ELPA and non-GNU ELPA. This section describes how to set up and maintain a package archive.

User Option: package-archives

The value of this variable is an alist of package archives recognized by the Emacs package manager.

Each alist element corresponds to one archive, and should have the form (id . location), where id is the name of the archive (a string) and location is its base location (a string).

If the base location starts with ‘http:’ or ‘https:’, it is treated as an HTTP(S) URL, and packages are downloaded from this archive via HTTP(S) (as is the case for the default GNU archive).

Otherwise, the base location should be a directory name. In this case, Emacs retrieves packages from this archive via ordinary file access. Such local archives are mainly useful for testing.

A package archive is simply a directory in which the package files, and associated files, are stored. If you want the archive to be reachable via HTTP, this directory must be accessible to a web server; See Interfacing to an archive web server.

A convenient way to set up and update a package archive is via the package-x library. This is included with Emacs, but not loaded by default; type M-x load-library RET package-x RET to load it, or add (require 'package-x) to your init file. See Lisp Libraries in The GNU Emacs Manual.

After you create an archive, remember that it is not accessible in the Package Menu interface unless it is in package-archives.

Maintaining a public package archive entails a degree of responsibility. When Emacs users install packages from your archive, those packages can cause Emacs to run arbitrary code with the permissions of the installing user. (This is true for Emacs code in general, not just for packages.) So you should ensure that your archive is well-maintained and keep the hosting system secure.

One way to increase the security of your packages is to sign them using a cryptographic key. If you have generated a private/public gpg key pair, you can use gpg to sign the package like this:

gpg -ba -o file.sig file

For a single-file package, file is the package Lisp file; for a multi-file package, it is the package tar file. You can also sign the archive’s contents file in the same way. Make the .sig files available in the same location as the packages. You should also make your public key available for people to download; e.g., by uploading it to a key server such as https://pgp.mit.edu/. When people install packages from your archive, they can use your public key to verify the signatures.

A full explanation of these matters is outside the scope of this manual. For more information on cryptographic keys and signing, see GnuPG in The GNU Privacy Guard Manual. Emacs comes with an interface to GNU Privacy Guard, see EasyPG in Emacs EasyPG Assistant Manual.


43.5 Interfacing to an archive web server

A web server providing access to a package archive must support the following queries:

archive-contents

Return a lisp form describing the archive contents. The form is a list of ’package-desc’ structures (see package.el), except the first element of the list is the archive version.

<package name>-readme.txt

Return the long description of the package.

<file name>.sig

Return the signature for the file.

<file name>

Return the file. This will be the tarball for a multi-file package, or the single file for a simple package.


Appendix A Emacs 28 Antinews

For those users who live backwards in time, here is information about downgrading to Emacs version 28.2. We hope you will enjoy the greater simplicity that results from the absence of many Emacs 29.2 features.


Appendix B GNU Free Documentation License

Version 1.3, 3 November 2008
Copyright © 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc.
https://fsf.org/

Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
  1. PREAMBLE

    The purpose of this License is to make a manual, textbook, or other functional and useful document free in the sense of freedom: to assure everyone the effective freedom to copy and redistribute it, with or without modifying it, either commercially or noncommercially. Secondarily, this License preserves for the author and publisher a way to get credit for their work, while not being considered responsible for modifications made by others.

    This License is a kind of “copyleft”, which means that derivative works of the document must themselves be free in the same sense. It complements the GNU General Public License, which is a copyleft license designed for free software.

    We have designed this License in order to use it for manuals for free software, because free software needs free documentation: a free program should come with manuals providing the same freedoms that the software does. But this License is not limited to software manuals; it can be used for any textual work, regardless of subject matter or whether it is published as a printed book. We recommend this License principally for works whose purpose is instruction or reference.

  2. APPLICABILITY AND DEFINITIONS

    This License applies to any manual or other work, in any medium, that contains a notice placed by the copyright holder saying it can be distributed under the terms of this License. Such a notice grants a world-wide, royalty-free license, unlimited in duration, to use that work under the conditions stated herein. The “Document”, below, refers to any such manual or work. Any member of the public is a licensee, and is addressed as “you”. You accept the license if you copy, modify or distribute the work in a way requiring permission under copyright law.

    A “Modified Version” of the Document means any work containing the Document or a portion of it, either copied verbatim, or with modifications and/or translated into another language.

    A “Secondary Section” is a named appendix or a front-matter section of the Document that deals exclusively with the relationship of the publishers or authors of the Document to the Document’s overall subject (or to related matters) and contains nothing that could fall directly within that overall subject. (Thus, if the Document is in part a textbook of mathematics, a Secondary Section may not explain any mathematics.) The relationship could be a matter of historical connection with the subject or with related matters, or of legal, commercial, philosophical, ethical or political position regarding them.

    The “Invariant Sections” are certain Secondary Sections whose titles are designated, as being those of Invariant Sections, in the notice that says that the Document is released under this License. If a section does not fit the above definition of Secondary then it is not allowed to be designated as Invariant. The Document may contain zero Invariant Sections. If the Document does not identify any Invariant Sections then there are none.

    The “Cover Texts” are certain short passages of text that are listed, as Front-Cover Texts or Back-Cover Texts, in the notice that says that the Document is released under this License. A Front-Cover Text may be at most 5 words, and a Back-Cover Text may be at most 25 words.

    A “Transparent” copy of the Document means a machine-readable copy, represented in a format whose specification is available to the general public, that is suitable for revising the document straightforwardly with generic text editors or (for images composed of pixels) generic paint programs or (for drawings) some widely available drawing editor, and that is suitable for input to text formatters or for automatic translation to a variety of formats suitable for input to text formatters. A copy made in an otherwise Transparent file format whose markup, or absence of markup, has been arranged to thwart or discourage subsequent modification by readers is not Transparent. An image format is not Transparent if used for any substantial amount of text. A copy that is not “Transparent” is called “Opaque”.

    Examples of suitable formats for Transparent copies include plain ASCII without markup, Texinfo input format, LaTeX input format, SGML or XML using a publicly available DTD, and standard-conforming simple HTML, PostScript or PDF designed for human modification. Examples of transparent image formats include PNG, XCF and JPG. Opaque formats include proprietary formats that can be read and edited only by proprietary word processors, SGML or XML for which the DTD and/or processing tools are not generally available, and the machine-generated HTML, PostScript or PDF produced by some word processors for output purposes only.

    The “Title Page” means, for a printed book, the title page itself, plus such following pages as are needed to hold, legibly, the material this License requires to appear in the title page. For works in formats which do not have any title page as such, “Title Page” means the text near the most prominent appearance of the work’s title, preceding the beginning of the body of the text.

    The “publisher” means any person or entity that distributes copies of the Document to the public.

    A section “Entitled XYZ” means a named subunit of the Document whose title either is precisely XYZ or contains XYZ in parentheses following text that translates XYZ in another language. (Here XYZ stands for a specific section name mentioned below, such as “Acknowledgements”, “Dedications”, “Endorsements”, or “History”.) To “Preserve the Title” of such a section when you modify the Document means that it remains a section “Entitled XYZ” according to this definition.

    The Document may include Warranty Disclaimers next to the notice which states that this License applies to the Document. These Warranty Disclaimers are considered to be included by reference in this License, but only as regards disclaiming warranties: any other implication that these Warranty Disclaimers may have is void and has no effect on the meaning of this License.

  3. VERBATIM COPYING

    You may copy and distribute the Document in any medium, either commercially or noncommercially, provided that this License, the copyright notices, and the license notice saying this License applies to the Document are reproduced in all copies, and that you add no other conditions whatsoever to those of this License. You may not use technical measures to obstruct or control the reading or further copying of the copies you make or distribute. However, you may accept compensation in exchange for copies. If you distribute a large enough number of copies you must also follow the conditions in section 3.

    You may also lend copies, under the same conditions stated above, and you may publicly display copies.

  4. COPYING IN QUANTITY

    If you publish printed copies (or copies in media that commonly have printed covers) of the Document, numbering more than 100, and the Document’s license notice requires Cover Texts, you must enclose the copies in covers that carry, clearly and legibly, all these Cover Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on the back cover. Both covers must also clearly and legibly identify you as the publisher of these copies. The front cover must present the full title with all words of the title equally prominent and visible. You may add other material on the covers in addition. Copying with changes limited to the covers, as long as they preserve the title of the Document and satisfy these conditions, can be treated as verbatim copying in other respects.

    If the required texts for either cover are too voluminous to fit legibly, you should put the first ones listed (as many as fit reasonably) on the actual cover, and continue the rest onto adjacent pages.

    If you publish or distribute Opaque copies of the Document numbering more than 100, you must either include a machine-readable Transparent copy along with each Opaque copy, or state in or with each Opaque copy a computer-network location from which the general network-using public has access to download using public-standard network protocols a complete Transparent copy of the Document, free of added material. If you use the latter option, you must take reasonably prudent steps, when you begin distribution of Opaque copies in quantity, to ensure that this Transparent copy will remain thus accessible at the stated location until at least one year after the last time you distribute an Opaque copy (directly or through your agents or retailers) of that edition to the public.

    It is requested, but not required, that you contact the authors of the Document well before redistributing any large number of copies, to give them a chance to provide you with an updated version of the Document.

  5. MODIFICATIONS

    You may copy and distribute a Modified Version of the Document under the conditions of sections 2 and 3 above, provided that you release the Modified Version under precisely this License, with the Modified Version filling the role of the Document, thus licensing distribution and modification of the Modified Version to whoever possesses a copy of it. In addition, you must do these things in the Modified Version:

    1. Use in the Title Page (and on the covers, if any) a title distinct from that of the Document, and from those of previous versions (which should, if there were any, be listed in the History section of the Document). You may use the same title as a previous version if the original publisher of that version gives permission.
    2. List on the Title Page, as authors, one or more persons or entities responsible for authorship of the modifications in the Modified Version, together with at least five of the principal authors of the Document (all of its principal authors, if it has fewer than five), unless they release you from this requirement.
    3. State on the Title page the name of the publisher of the Modified Version, as the publisher.
    4. Preserve all the copyright notices of the Document.
    5. Add an appropriate copyright notice for your modifications adjacent to the other copyright notices.
    6. Include, immediately after the copyright notices, a license notice giving the public permission to use the Modified Version under the terms of this License, in the form shown in the Addendum below.
    7. Preserve in that license notice the full lists of Invariant Sections and required Cover Texts given in the Document’s license notice.
    8. Include an unaltered copy of this License.
    9. Preserve the section Entitled “History”, Preserve its Title, and add to it an item stating at least the title, year, new authors, and publisher of the Modified Version as given on the Title Page. If there is no section Entitled “History” in the Document, create one stating the title, year, authors, and publisher of the Document as given on its Title Page, then add an item describing the Modified Version as stated in the previous sentence.
    10. Preserve the network location, if any, given in the Document for public access to a Transparent copy of the Document, and likewise the network locations given in the Document for previous versions it was based on. These may be placed in the “History” section. You may omit a network location for a work that was published at least four years before the Document itself, or if the original publisher of the version it refers to gives permission.
    11. For any section Entitled “Acknowledgements” or “Dedications”, Preserve the Title of the section, and preserve in the section all the substance and tone of each of the contributor acknowledgements and/or dedications given therein.
    12. Preserve all the Invariant Sections of the Document, unaltered in their text and in their titles. Section numbers or the equivalent are not considered part of the section titles.
    13. Delete any section Entitled “Endorsements”. Such a section may not be included in the Modified Version.
    14. Do not retitle any existing section to be Entitled “Endorsements” or to conflict in title with any Invariant Section.
    15. Preserve any Warranty Disclaimers.

    If the Modified Version includes new front-matter sections or appendices that qualify as Secondary Sections and contain no material copied from the Document, you may at your option designate some or all of these sections as invariant. To do this, add their titles to the list of Invariant Sections in the Modified Version’s license notice. These titles must be distinct from any other section titles.

    You may add a section Entitled “Endorsements”, provided it contains nothing but endorsements of your Modified Version by various parties—for example, statements of peer review or that the text has been approved by an organization as the authoritative definition of a standard.

    You may add a passage of up to five words as a Front-Cover Text, and a passage of up to 25 words as a Back-Cover Text, to the end of the list of Cover Texts in the Modified Version. Only one passage of Front-Cover Text and one of Back-Cover Text may be added by (or through arrangements made by) any one entity. If the Document already includes a cover text for the same cover, previously added by you or by arrangement made by the same entity you are acting on behalf of, you may not add another; but you may replace the old one, on explicit permission from the previous publisher that added the old one.

    The author(s) and publisher(s) of the Document do not by this License give permission to use their names for publicity for or to assert or imply endorsement of any Modified Version.

  6. COMBINING DOCUMENTS

    You may combine the Document with other documents released under this License, under the terms defined in section 4 above for modified versions, provided that you include in the combination all of the Invariant Sections of all of the original documents, unmodified, and list them all as Invariant Sections of your combined work in its license notice, and that you preserve all their Warranty Disclaimers.

    The combined work need only contain one copy of this License, and multiple identical Invariant Sections may be replaced with a single copy. If there are multiple Invariant Sections with the same name but different contents, make the title of each such section unique by adding at the end of it, in parentheses, the name of the original author or publisher of that section if known, or else a unique number. Make the same adjustment to the section titles in the list of Invariant Sections in the license notice of the combined work.

    In the combination, you must combine any sections Entitled “History” in the various original documents, forming one section Entitled “History”; likewise combine any sections Entitled “Acknowledgements”, and any sections Entitled “Dedications”. You must delete all sections Entitled “Endorsements.”

  7. COLLECTIONS OF DOCUMENTS

    You may make a collection consisting of the Document and other documents released under this License, and replace the individual copies of this License in the various documents with a single copy that is included in the collection, provided that you follow the rules of this License for verbatim copying of each of the documents in all other respects.

    You may extract a single document from such a collection, and distribute it individually under this License, provided you insert a copy of this License into the extracted document, and follow this License in all other respects regarding verbatim copying of that document.

  8. AGGREGATION WITH INDEPENDENT WORKS

    A compilation of the Document or its derivatives with other separate and independent documents or works, in or on a volume of a storage or distribution medium, is called an “aggregate” if the copyright resulting from the compilation is not used to limit the legal rights of the compilation’s users beyond what the individual works permit. When the Document is included in an aggregate, this License does not apply to the other works in the aggregate which are not themselves derivative works of the Document.

    If the Cover Text requirement of section 3 is applicable to these copies of the Document, then if the Document is less than one half of the entire aggregate, the Document’s Cover Texts may be placed on covers that bracket the Document within the aggregate, or the electronic equivalent of covers if the Document is in electronic form. Otherwise they must appear on printed covers that bracket the whole aggregate.

  9. TRANSLATION

    Translation is considered a kind of modification, so you may distribute translations of the Document under the terms of section 4. Replacing Invariant Sections with translations requires special permission from their copyright holders, but you may include translations of some or all Invariant Sections in addition to the original versions of these Invariant Sections. You may include a translation of this License, and all the license notices in the Document, and any Warranty Disclaimers, provided that you also include the original English version of this License and the original versions of those notices and disclaimers. In case of a disagreement between the translation and the original version of this License or a notice or disclaimer, the original version will prevail.

    If a section in the Document is Entitled “Acknowledgements”, “Dedications”, or “History”, the requirement (section 4) to Preserve its Title (section 1) will typically require changing the actual title.

  10. TERMINATION

    You may not copy, modify, sublicense, or distribute the Document except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense, or distribute it is void, and will automatically terminate your rights under this License.

    However, if you cease all violation of this License, then your license from a particular copyright holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and finally terminates your license, and (b) permanently, if the copyright holder fails to notify you of the violation by some reasonable means prior to 60 days after the cessation.

    Moreover, your license from a particular copyright holder is reinstated permanently if the copyright holder notifies you of the violation by some reasonable means, this is the first time you have received notice of violation of this License (for any work) from that copyright holder, and you cure the violation prior to 30 days after your receipt of the notice.

    Termination of your rights under this section does not terminate the licenses of parties who have received copies or rights from you under this License. If your rights have been terminated and not permanently reinstated, receipt of a copy of some or all of the same material does not give you any rights to use it.

  11. FUTURE REVISIONS OF THIS LICENSE

    The Free Software Foundation may publish new, revised versions of the GNU Free Documentation License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. See https://www.gnu.org/licenses/.

    Each version of the License is given a distinguishing version number. If the Document specifies that a particular numbered version of this License “or any later version” applies to it, you have the option of following the terms and conditions either of that specified version or of any later version that has been published (not as a draft) by the Free Software Foundation. If the Document does not specify a version number of this License, you may choose any version ever published (not as a draft) by the Free Software Foundation. If the Document specifies that a proxy can decide which future versions of this License can be used, that proxy’s public statement of acceptance of a version permanently authorizes you to choose that version for the Document.

  12. RELICENSING

    “Massive Multiauthor Collaboration Site” (or “MMC Site”) means any World Wide Web server that publishes copyrightable works and also provides prominent facilities for anybody to edit those works. A public wiki that anybody can edit is an example of such a server. A “Massive Multiauthor Collaboration” (or “MMC”) contained in the site means any set of copyrightable works thus published on the MMC site.

    “CC-BY-SA” means the Creative Commons Attribution-Share Alike 3.0 license published by Creative Commons Corporation, a not-for-profit corporation with a principal place of business in San Francisco, California, as well as future copyleft versions of that license published by that same organization.

    “Incorporate” means to publish or republish a Document, in whole or in part, as part of another Document.

    An MMC is “eligible for relicensing” if it is licensed under this License, and if all works that were first published under this License somewhere other than this MMC, and subsequently incorporated in whole or in part into the MMC, (1) had no cover texts or invariant sections, and (2) were thus incorporated prior to November 1, 2008.

    The operator of an MMC Site may republish an MMC contained in the site under CC-BY-SA on the same site at any time before August 1, 2009, provided the MMC is eligible for relicensing.

ADDENDUM: How to use this License for your documents

To use this License in a document you have written, include a copy of the License in the document and put the following copyright and license notices just after the title page:

  Copyright (C)  year  your name.
  Permission is granted to copy, distribute and/or modify this document
  under the terms of the GNU Free Documentation License, Version 1.3
  or any later version published by the Free Software Foundation;
  with no Invariant Sections, no Front-Cover Texts, and no Back-Cover
  Texts.  A copy of the license is included in the section entitled ``GNU
  Free Documentation License''.

If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts, replace the “with…Texts.” line with this:

    with the Invariant Sections being list their titles, with
    the Front-Cover Texts being list, and with the Back-Cover Texts
    being list.

If you have Invariant Sections without Cover Texts, or some other combination of the three, merge those two alternatives to suit the situation.

If your document contains nontrivial examples of program code, we recommend releasing these examples in parallel under your choice of free software license, such as the GNU General Public License, to permit their use in free software.


Appendix C GNU General Public License

Version 3, 29 June 2007
Copyright © 2007 Free Software Foundation, Inc. https://fsf.org/

Everyone is permitted to copy and distribute verbatim copies of this
license document, but changing it is not allowed.

Preamble

The GNU General Public License is a free, copyleft license for software and other kinds of works.

The licenses for most software and other practical works are designed to take away your freedom to share and change the works. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change all versions of a program—to make sure it remains free software for all its users. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work released this way by its authors. You can apply it to your programs, too.

When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for them if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs, and that you know you can do these things.

To protect your rights, we need to prevent others from denying you these rights or asking you to surrender the rights. Therefore, you have certain responsibilities if you distribute copies of the software, or if you modify it: responsibilities to respect the freedom of others.

For example, if you distribute copies of such a program, whether gratis or for a fee, you must pass on to the recipients the same freedoms that you received. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights.

Developers that use the GNU GPL protect your rights with two steps: (1) assert copyright on the software, and (2) offer you this License giving you legal permission to copy, distribute and/or modify it.

For the developers’ and authors’ protection, the GPL clearly explains that there is no warranty for this free software. For both users’ and authors’ sake, the GPL requires that modified versions be marked as changed, so that their problems will not be attributed erroneously to authors of previous versions.

Some devices are designed to deny users access to install or run modified versions of the software inside them, although the manufacturer can do so. This is fundamentally incompatible with the aim of protecting users’ freedom to change the software. The systematic pattern of such abuse occurs in the area of products for individuals to use, which is precisely where it is most unacceptable. Therefore, we have designed this version of the GPL to prohibit the practice for those products. If such problems arise substantially in other domains, we stand ready to extend this provision to those domains in future versions of the GPL, as needed to protect the freedom of users.

Finally, every program is threatened constantly by software patents. States should not allow patents to restrict development and use of software on general-purpose computers, but in those that do, we wish to avoid the special danger that patents applied to a free program could make it effectively proprietary. To prevent this, the GPL assures that patents cannot be used to render the program non-free.

The precise terms and conditions for copying, distribution and modification follow.

TERMS AND CONDITIONS

  1. Definitions.

    “This License” refers to version 3 of the GNU General Public License.

    “Copyright” also means copyright-like laws that apply to other kinds of works, such as semiconductor masks.

    “The Program” refers to any copyrightable work licensed under this License. Each licensee is addressed as “you”. “Licensees” and “recipients” may be individuals or organizations.

    To “modify” a work means to copy from or adapt all or part of the work in a fashion requiring copyright permission, other than the making of an exact copy. The resulting work is called a “modified version” of the earlier work or a work “based on” the earlier work.

    A “covered work” means either the unmodified Program or a work based on the Program.

    To “propagate” a work means to do anything with it that, without permission, would make you directly or secondarily liable for infringement under applicable copyright law, except executing it on a computer or modifying a private copy. Propagation includes copying, distribution (with or without modification), making available to the public, and in some countries other activities as well.

    To “convey” a work means any kind of propagation that enables other parties to make or receive copies. Mere interaction with a user through a computer network, with no transfer of a copy, is not conveying.

    An interactive user interface displays “Appropriate Legal Notices” to the extent that it includes a convenient and prominently visible feature that (1) displays an appropriate copyright notice, and (2) tells the user that there is no warranty for the work (except to the extent that warranties are provided), that licensees may convey the work under this License, and how to view a copy of this License. If the interface presents a list of user commands or options, such as a menu, a prominent item in the list meets this criterion.

  2. Source Code.

    The “source code” for a work means the preferred form of the work for making modifications to it. “Object code” means any non-source form of a work.

    A “Standard Interface” means an interface that either is an official standard defined by a recognized standards body, or, in the case of interfaces specified for a particular programming language, one that is widely used among developers working in that language.

    The “System Libraries” of an executable work include anything, other than the work as a whole, that (a) is included in the normal form of packaging a Major Component, but which is not part of that Major Component, and (b) serves only to enable use of the work with that Major Component, or to implement a Standard Interface for which an implementation is available to the public in source code form. A “Major Component”, in this context, means a major essential component (kernel, window system, and so on) of the specific operating system (if any) on which the executable work runs, or a compiler used to produce the work, or an object code interpreter used to run it.

    The “Corresponding Source” for a work in object code form means all the source code needed to generate, install, and (for an executable work) run the object code and to modify the work, including scripts to control those activities. However, it does not include the work’s System Libraries, or general-purpose tools or generally available free programs which are used unmodified in performing those activities but which are not part of the work. For example, Corresponding Source includes interface definition files associated with source files for the work, and the source code for shared libraries and dynamically linked subprograms that the work is specifically designed to require, such as by intimate data communication or control flow between those subprograms and other parts of the work.

    The Corresponding Source need not include anything that users can regenerate automatically from other parts of the Corresponding Source.

    The Corresponding Source for a work in source code form is that same work.

  3. Basic Permissions.

    All rights granted under this License are granted for the term of copyright on the Program, and are irrevocable provided the stated conditions are met. This License explicitly affirms your unlimited permission to run the unmodified Program. The output from running a covered work is covered by this License only if the output, given its content, constitutes a covered work. This License acknowledges your rights of fair use or other equivalent, as provided by copyright law.

    You may make, run and propagate covered works that you do not convey, without conditions so long as your license otherwise remains in force. You may convey covered works to others for the sole purpose of having them make modifications exclusively for you, or provide you with facilities for running those works, provided that you comply with the terms of this License in conveying all material for which you do not control copyright. Those thus making or running the covered works for you must do so exclusively on your behalf, under your direction and control, on terms that prohibit them from making any copies of your copyrighted material outside their relationship with you.

    Conveying under any other circumstances is permitted solely under the conditions stated below. Sublicensing is not allowed; section 10 makes it unnecessary.

  4. Protecting Users’ Legal Rights From Anti-Circumvention Law.

    No covered work shall be deemed part of an effective technological measure under any applicable law fulfilling obligations under article 11 of the WIPO copyright treaty adopted on 20 December 1996, or similar laws prohibiting or restricting circumvention of such measures.

    When you convey a covered work, you waive any legal power to forbid circumvention of technological measures to the extent such circumvention is effected by exercising rights under this License with respect to the covered work, and you disclaim any intention to limit operation or modification of the work as a means of enforcing, against the work’s users, your or third parties’ legal rights to forbid circumvention of technological measures.

  5. Conveying Verbatim Copies.

    You may convey verbatim copies of the Program’s source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice; keep intact all notices stating that this License and any non-permissive terms added in accord with section 7 apply to the code; keep intact all notices of the absence of any warranty; and give all recipients a copy of this License along with the Program.

    You may charge any price or no price for each copy that you convey, and you may offer support or warranty protection for a fee.

  6. Conveying Modified Source Versions.

    You may convey a work based on the Program, or the modifications to produce it from the Program, in the form of source code under the terms of section 4, provided that you also meet all of these conditions:

    1. The work must carry prominent notices stating that you modified it, and giving a relevant date.
    2. The work must carry prominent notices stating that it is released under this License and any conditions added under section 7. This requirement modifies the requirement in section 4 to “keep intact all notices”.
    3. You must license the entire work, as a whole, under this License to anyone who comes into possession of a copy. This License will therefore apply, along with any applicable section 7 additional terms, to the whole of the work, and all its parts, regardless of how they are packaged. This License gives no permission to license the work in any other way, but it does not invalidate such permission if you have separately received it.
    4. If the work has interactive user interfaces, each must display Appropriate Legal Notices; however, if the Program has interactive interfaces that do not display Appropriate Legal Notices, your work need not make them do so.

    A compilation of a covered work with other separate and independent works, which are not by their nature extensions of the covered work, and which are not combined with it such as to form a larger program, in or on a volume of a storage or distribution medium, is called an “aggregate” if the compilation and its resulting copyright are not used to limit the access or legal rights of the compilation’s users beyond what the individual works permit. Inclusion of a covered work in an aggregate does not cause this License to apply to the other parts of the aggregate.

  7. Conveying Non-Source Forms.

    You may convey a covered work in object code form under the terms of sections 4 and 5, provided that you also convey the machine-readable Corresponding Source under the terms of this License, in one of these ways:

    1. Convey the object code in, or embodied in, a physical product (including a physical distribution medium), accompanied by the Corresponding Source fixed on a durable physical medium customarily used for software interchange.
    2. Convey the object code in, or embodied in, a physical product (including a physical distribution medium), accompanied by a written offer, valid for at least three years and valid for as long as you offer spare parts or customer support for that product model, to give anyone who possesses the object code either (1) a copy of the Corresponding Source for all the software in the product that is covered by this License, on a durable physical medium customarily used for software interchange, for a price no more than your reasonable cost of physically performing this conveying of source, or (2) access to copy the Corresponding Source from a network server at no charge.
    3. Convey individual copies of the object code with a copy of the written offer to provide the Corresponding Source. This alternative is allowed only occasionally and noncommercially, and only if you received the object code with such an offer, in accord with subsection 6b.
    4. Convey the object code by offering access from a designated place (gratis or for a charge), and offer equivalent access to the Corresponding Source in the same way through the same place at no further charge. You need not require recipients to copy the Corresponding Source along with the object code. If the place to copy the object code is a network server, the Corresponding Source may be on a different server (operated by you or a third party) that supports equivalent copying facilities, provided you maintain clear directions next to the object code saying where to find the Corresponding Source. Regardless of what server hosts the Corresponding Source, you remain obligated to ensure that it is available for as long as needed to satisfy these requirements.
    5. Convey the object code using peer-to-peer transmission, provided you inform other peers where the object code and Corresponding Source of the work are being offered to the general public at no charge under subsection 6d.

    A separable portion of the object code, whose source code is excluded from the Corresponding Source as a System Library, need not be included in conveying the object code work.

    A “User Product” is either (1) a “consumer product”, which means any tangible personal property which is normally used for personal, family, or household purposes, or (2) anything designed or sold for incorporation into a dwelling. In determining whether a product is a consumer product, doubtful cases shall be resolved in favor of coverage. For a particular product received by a particular user, “normally used” refers to a typical or common use of that class of product, regardless of the status of the particular user or of the way in which the particular user actually uses, or expects or is expected to use, the product. A product is a consumer product regardless of whether the product has substantial commercial, industrial or non-consumer uses, unless such uses represent the only significant mode of use of the product.

    “Installation Information” for a User Product means any methods, procedures, authorization keys, or other information required to install and execute modified versions of a covered work in that User Product from a modified version of its Corresponding Source. The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with solely because modification has been made.

    If you convey an object code work under this section in, or with, or specifically for use in, a User Product, and the conveying occurs as part of a transaction in which the right of possession and use of the User Product is transferred to the recipient in perpetuity or for a fixed term (regardless of how the transaction is characterized), the Corresponding Source conveyed under this section must be accompanied by the Installation Information. But this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM).

    The requirement to provide Installation Information does not include a requirement to continue to provide support service, warranty, or updates for a work that has been modified or installed by the recipient, or for the User Product in which it has been modified or installed. Access to a network may be denied when the modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network.

    Corresponding Source conveyed, and Installation Information provided, in accord with this section must be in a format that is publicly documented (and with an implementation available to the public in source code form), and must require no special password or key for unpacking, reading or copying.

  8. Additional Terms.

    “Additional permissions” are terms that supplement the terms of this License by making exceptions from one or more of its conditions. Additional permissions that are applicable to the entire Program shall be treated as though they were included in this License, to the extent that they are valid under applicable law. If additional permissions apply only to part of the Program, that part may be used separately under those permissions, but the entire Program remains governed by this License without regard to the additional permissions.

    When you convey a copy of a covered work, you may at your option remove any additional permissions from that copy, or from any part of it. (Additional permissions may be written to require their own removal in certain cases when you modify the work.) You may place additional permissions on material, added by you to a covered work, for which you have or can give appropriate copyright permission.

    Notwithstanding any other provision of this License, for material you add to a covered work, you may (if authorized by the copyright holders of that material) supplement the terms of this License with terms:

    1. Disclaiming warranty or limiting liability differently from the terms of sections 15 and 16 of this License; or
    2. Requiring preservation of specified reasonable legal notices or author attributions in that material or in the Appropriate Legal Notices displayed by works containing it; or
    3. Prohibiting misrepresentation of the origin of that material, or requiring that modified versions of such material be marked in reasonable ways as different from the original version; or
    4. Limiting the use for publicity purposes of names of licensors or authors of the material; or
    5. Declining to grant rights under trademark law for use of some trade names, trademarks, or service marks; or
    6. Requiring indemnification of licensors and authors of that material by anyone who conveys the material (or modified versions of it) with contractual assumptions of liability to the recipient, for any liability that these contractual assumptions directly impose on those licensors and authors.

    All other non-permissive additional terms are considered “further restrictions” within the meaning of section 10. If the Program as you received it, or any part of it, contains a notice stating that it is governed by this License along with a term that is a further restriction, you may remove that term. If a license document contains a further restriction but permits relicensing or conveying under this License, you may add to a covered work material governed by the terms of that license document, provided that the further restriction does not survive such relicensing or conveying.

    If you add terms to a covered work in accord with this section, you must place, in the relevant source files, a statement of the additional terms that apply to those files, or a notice indicating where to find the applicable terms.

    Additional terms, permissive or non-permissive, may be stated in the form of a separately written license, or stated as exceptions; the above requirements apply either way.

  9. Termination.

    You may not propagate or modify a covered work except as expressly provided under this License. Any attempt otherwise to propagate or modify it is void, and will automatically terminate your rights under this License (including any patent licenses granted under the third paragraph of section 11).

    However, if you cease all violation of this License, then your license from a particular copyright holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and finally terminates your license, and (b) permanently, if the copyright holder fails to notify you of the violation by some reasonable means prior to 60 days after the cessation.

    Moreover, your license from a particular copyright holder is reinstated permanently if the copyright holder notifies you of the violation by some reasonable means, this is the first time you have received notice of violation of this License (for any work) from that copyright holder, and you cure the violation prior to 30 days after your receipt of the notice.

    Termination of your rights under this section does not terminate the licenses of parties who have received copies or rights from you under this License. If your rights have been terminated and not permanently reinstated, you do not qualify to receive new licenses for the same material under section 10.

  10. Acceptance Not Required for Having Copies.

    You are not required to accept this License in order to receive or run a copy of the Program. Ancillary propagation of a covered work occurring solely as a consequence of using peer-to-peer transmission to receive a copy likewise does not require acceptance. However, nothing other than this License grants you permission to propagate or modify any covered work. These actions infringe copyright if you do not accept this License. Therefore, by modifying or propagating a covered work, you indicate your acceptance of this License to do so.

  11. Automatic Licensing of Downstream Recipients.

    Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. You are not responsible for enforcing compliance by third parties with this License.

    An “entity transaction” is a transaction transferring control of an organization, or substantially all assets of one, or subdividing an organization, or merging organizations. If propagation of a covered work results from an entity transaction, each party to that transaction who receives a copy of the work also receives whatever licenses to the work the party’s predecessor in interest had or could give under the previous paragraph, plus a right to possession of the Corresponding Source of the work from the predecessor in interest, if the predecessor has it or can get it with reasonable efforts.

    You may not impose any further restrictions on the exercise of the rights granted or affirmed under this License. For example, you may not impose a license fee, royalty, or other charge for exercise of rights granted under this License, and you may not initiate litigation (including a cross-claim or counterclaim in a lawsuit) alleging that any patent claim is infringed by making, using, selling, offering for sale, or importing the Program or any portion of it.

  12. Patents.

    A “contributor” is a copyright holder who authorizes use under this License of the Program or a work on which the Program is based. The work thus licensed is called the contributor’s “contributor version”.

    A contributor’s “essential patent claims” are all patent claims owned or controlled by the contributor, whether already acquired or hereafter acquired, that would be infringed by some manner, permitted by this License, of making, using, or selling its contributor version, but do not include claims that would be infringed only as a consequence of further modification of the contributor version. For purposes of this definition, “control” includes the right to grant patent sublicenses in a manner consistent with the requirements of this License.

    Each contributor grants you a non-exclusive, worldwide, royalty-free patent license under the contributor’s essential patent claims, to make, use, sell, offer for sale, import and otherwise run, modify and propagate the contents of its contributor version.

    In the following three paragraphs, a “patent license” is any express agreement or commitment, however denominated, not to enforce a patent (such as an express permission to practice a patent or covenant not to sue for patent infringement). To “grant” such a patent license to a party means to make such an agreement or commitment not to enforce a patent against the party.

    If you convey a covered work, knowingly relying on a patent license, and the Corresponding Source of the work is not available for anyone to copy, free of charge and under the terms of this License, through a publicly available network server or other readily accessible means, then you must either (1) cause the Corresponding Source to be so available, or (2) arrange to deprive yourself of the benefit of the patent license for this particular work, or (3) arrange, in a manner consistent with the requirements of this License, to extend the patent license to downstream recipients. “Knowingly relying” means you have actual knowledge that, but for the patent license, your conveying the covered work in a country, or your recipient’s use of the covered work in a country, would infringe one or more identifiable patents in that country that you have reason to believe are valid.

    If, pursuant to or in connection with a single transaction or arrangement, you convey, or propagate by procuring conveyance of, a covered work, and grant a patent license to some of the parties receiving the covered work authorizing them to use, propagate, modify or convey a specific copy of the covered work, then the patent license you grant is automatically extended to all recipients of the covered work and works based on it.

    A patent license is “discriminatory” if it does not include within the scope of its coverage, prohibits the exercise of, or is conditioned on the non-exercise of one or more of the rights that are specifically granted under this License. You may not convey a covered work if you are a party to an arrangement with a third party that is in the business of distributing software, under which you make payment to the third party based on the extent of your activity of conveying the work, and under which the third party grants, to any of the parties who would receive the covered work from you, a discriminatory patent license (a) in connection with copies of the covered work conveyed by you (or copies made from those copies), or (b) primarily for and in connection with specific products or compilations that contain the covered work, unless you entered into that arrangement, or that patent license was granted, prior to 28 March 2007.

    Nothing in this License shall be construed as excluding or limiting any implied license or other defenses to infringement that may otherwise be available to you under applicable patent law.

  13. No Surrender of Others’ Freedom.

    If conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot convey a covered work so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not convey it at all. For example, if you agree to terms that obligate you to collect a royalty for further conveying from those to whom you convey the Program, the only way you could satisfy both those terms and this License would be to refrain entirely from conveying the Program.

  14. Use with the GNU Affero General Public License.

    Notwithstanding any other provision of this License, you have permission to link or combine any covered work with a work licensed under version 3 of the GNU Affero General Public License into a single combined work, and to convey the resulting work. The terms of this License will continue to apply to the part which is the covered work, but the special requirements of the GNU Affero General Public License, section 13, concerning interaction through a network will apply to the combination as such.

  15. Revised Versions of this License.

    The Free Software Foundation may publish revised and/or new versions of the GNU General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns.

    Each version is given a distinguishing version number. If the Program specifies that a certain numbered version of the GNU General Public License “or any later version” applies to it, you have the option of following the terms and conditions either of that numbered version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of the GNU General Public License, you may choose any version ever published by the Free Software Foundation.

    If the Program specifies that a proxy can decide which future versions of the GNU General Public License can be used, that proxy’s public statement of acceptance of a version permanently authorizes you to choose that version for the Program.

    Later license versions may give you additional or different permissions. However, no additional obligations are imposed on any author or copyright holder as a result of your choosing to follow a later version.

  16. Disclaimer of Warranty.

    THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

  17. Limitation of Liability.

    IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.

  18. Interpretation of Sections 15 and 16.

    If the disclaimer of warranty and limitation of liability provided above cannot be given local legal effect according to their terms, reviewing courts shall apply local law that most closely approximates an absolute waiver of all civil liability in connection with the Program, unless a warranty or assumption of liability accompanies a copy of the Program in return for a fee.

END OF TERMS AND CONDITIONS

How to Apply These Terms to Your New Programs

If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms.

To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively state the exclusion of warranty; and each file should have at least the “copyright” line and a pointer to where the full notice is found.

one line to give the program's name and a brief idea of what it does.
Copyright (C) year name of author

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or (at
your option) any later version.

This program is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see https://www.gnu.org/licenses/.

Also add information on how to contact you by electronic and paper mail.

If the program does terminal interaction, make it output a short notice like this when it starts in an interactive mode:

program Copyright (C) year name of author
This program comes with ABSOLUTELY NO WARRANTY; for details type ‘show w’.
This is free software, and you are welcome to redistribute it
under certain conditions; type ‘show c’ for details.

The hypothetical commands ‘show w’ and ‘show c’ should show the appropriate parts of the General Public License. Of course, your program’s commands might be different; for a GUI interface, you would use an “about box”.

You should also get your employer (if you work as a programmer) or school, if any, to sign a “copyright disclaimer” for the program, if necessary. For more information on this, and how to apply and follow the GNU GPL, see https://www.gnu.org/licenses/.

The GNU General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Lesser General Public License instead of this License. But first, please read https://www.gnu.org/licenses/why-not-lgpl.html.


Appendix D Tips and Conventions

This chapter describes no additional features of Emacs Lisp. Instead it gives advice on making effective use of the features described in the previous chapters, and describes conventions Emacs Lisp programmers should follow.

You can automatically check some of the conventions described below by running the command M-x checkdoc RET when visiting a Lisp file. It cannot check all of the conventions, and not all the warnings it gives necessarily correspond to problems, but it is worth examining them all. Alternatively, use the command M-x checkdoc-current-buffer RET to check the conventions in the current buffer, or checkdoc-file when you want to check a file in batch mode, e.g., with a command run by M-x compile RET.


D.1 Emacs Lisp Coding Conventions

Here are conventions that you should follow when writing Emacs Lisp code intended for widespread use:

  • Simply loading a package should not change Emacs’s editing behavior. Include a command or commands to enable and disable the feature, or to invoke it.

    This convention is mandatory for any file that includes custom definitions. If fixing such a file to follow this convention requires an incompatible change, go ahead and make the incompatible change; don’t postpone it.

  • You should choose a short word to distinguish your program from other Lisp programs. The names of all global symbols in your program, that is the names of variables, constants, and functions, should begin with that chosen prefix. Separate the prefix from the rest of the name with a hyphen, ‘-’. This practice helps avoid name conflicts, since all global variables in Emacs Lisp share the same name space, and all functions share another name space33. Use two hyphens to separate prefix and name if the symbol is not meant to be used by other packages.

    Occasionally, for a command name intended for users to use, it is more convenient if some words come before the package’s name prefix. For example, it is our convention to have commands that list objects named as ‘list-something’, e.g., a package called ‘frob’ could have a command ‘list-frobs’, when its other global symbols begin with ‘frob-’. Also, constructs that define functions, variables, etc., work better if they start with ‘define-’, so put the name prefix later on in the name.

    This recommendation applies even to names for traditional Lisp primitives that are not primitives in Emacs Lisp—such as copy-list. Believe it or not, there is more than one plausible way to define copy-list. Play it safe; append your name prefix to produce a name like foo-copy-list or mylib-copy-list instead.

    If you write a function that you think ought to be added to Emacs under a certain name, such as twiddle-files, don’t call it by that name in your program. Call it mylib-twiddle-files in your program, and send mail to ‘bug-gnu-emacs@gnu.org’ suggesting we add it to Emacs. If and when we do, we can change the name easily enough.

    If one prefix is insufficient, your package can use two or three alternative common prefixes, so long as they make sense.

  • We recommend enabling lexical-binding in new code, and converting existing Emacs Lisp code to enable lexical-binding if it doesn’t already. See Using Lexical Binding.
  • Put a call to provide at the end of each separate Lisp file. See Features.
  • If a file requires certain other Lisp programs to be loaded beforehand, then the comments at the beginning of the file should say so. Also, use require to make sure they are loaded. See Features.
  • If a file foo uses a macro defined in another file bar, but does not use any functions or variables defined in bar, then foo should contain the following expression:
    (eval-when-compile (require 'bar))
    

    This tells Emacs to load bar just before byte-compiling foo, so that the macro definition is available during compilation. Using eval-when-compile avoids loading bar when the compiled version of foo is used. It should be called before the first use of the macro in the file. See Macros and Byte Compilation.

  • Avoid loading additional libraries at run time unless they are really needed. If your file simply cannot work without some other library, then just require that library at the top-level and be done with it. But if your file contains several independent features, and only one or two require the extra library, then consider putting require statements inside the relevant functions rather than at the top-level. Or use autoload statements to load the extra library when needed. This way people who don’t use those aspects of your file do not need to load the extra library.
  • If you need Common Lisp extensions, use the cl-lib library rather than the old cl library. The latter library is deprecated and will be removed in a future version of Emacs.
  • When defining a major mode, please follow the major mode conventions. See Major Mode Conventions.
  • When defining a minor mode, please follow the minor mode conventions. See Conventions for Writing Minor Modes.
  • If the purpose of a function is to tell you whether a certain condition is true or false, give the function a name that ends in ‘p’ (which stands for “predicate”). If the name is one word, add just ‘p’; if the name is multiple words, add ‘-p’. Examples are framep and frame-live-p. We recommend to avoid using this -p suffix in boolean variable names, unless the variable is bound to a predicate function; instead, use a -flag suffix or names like is-foo.
  • If the purpose of a variable is to store a single function, give it a name that ends in ‘-function’. If the purpose of a variable is to store a list of functions (i.e., the variable is a hook), please follow the naming conventions for hooks. See Hooks.
  • Using unload-feature will undo the changes usually done by loading a feature (like adding functions to hooks). However, if loading feature does something unusual and more complex, you can define a function named feature-unload-function, and make it undo any such special changes. unload-feature will then automatically run this function if it exists. See Unloading.
  • It is a bad idea to define aliases for the Emacs primitives. Normally you should use the standard names instead. The case where an alias may be useful is where it facilitates backwards compatibility or portability.
  • If a package needs to define an alias or a new function for compatibility with some other version of Emacs, name it with the package prefix, not with the raw name with which it occurs in the other version. Here is an example from Gnus, which provides many examples of such compatibility issues.
    (defalias 'gnus-point-at-bol
      (if (fboundp 'point-at-bol)
          'point-at-bol
        'line-beginning-position))
    
  • Redefining or advising an Emacs primitive is a bad idea. It may do the right thing for a particular program, but there is no telling what other programs might break as a result.
  • It is likewise a bad idea for one Lisp package to advise a function in another Lisp package (see Advising Emacs Lisp Functions).
  • Avoid using eval-after-load and with-eval-after-load in libraries and packages (see Hooks for Loading). This feature is meant for personal customizations; using it in a Lisp program is unclean, because it modifies the behavior of another Lisp file in a way that’s not visible in that file. This is an obstacle for debugging, much like advising a function in the other package.
  • If a file does replace any of the standard functions or library programs of Emacs, prominent comments at the beginning of the file should say which functions are replaced, and how the behavior of the replacements differs from that of the originals.
  • Constructs that define a function or variable should be macros, not functions, and their names should start with ‘define-’. The macro should receive the name to be defined as the first argument. That will help various tools find the definition automatically. Avoid constructing the names in the macro itself, since that would confuse these tools.
  • In some other systems there is a convention of choosing variable names that begin and end with ‘*’. We don’t use that convention in Emacs Lisp, so please don’t use it in your programs. (Emacs uses such names only for special-purpose buffers.) People will find Emacs more coherent if all libraries use the same conventions.
  • The default file coding system for Emacs Lisp source files is UTF-8 (see Text Representations). In the rare event that your program contains characters which are not in UTF-8, you should specify an appropriate coding system in the source file’s ‘-*-’ line or local variables list. See Local Variables in Files in The GNU Emacs Manual.
  • Indent the file using the default indentation parameters.
  • Don’t make a habit of putting close-parentheses on lines by themselves; Lisp programmers find this disconcerting.
  • Please put a copyright notice and copying permission notice on the file if you distribute copies. See Conventional Headers for Emacs Libraries.
  • For variables holding (or functions returning) a file or directory name, avoid using path in its name, preferring file, file-name, or directory instead, since Emacs follows the GNU convention to use the term path only for search paths, which are lists of directory names.

D.2 Key Binding Conventions

  • Many special major modes, like Dired, Info, Compilation, and Occur, are designed to handle read-only text that contains hyper-links. Such a major mode should redefine mouse-2 and RET to follow the links. It should also set up a follow-link condition, so that the link obeys mouse-1-click-follows-link. See Defining Clickable Text. See Buttons, for an easy method of implementing such clickable links.
  • Don’t define C-c letter as a key in Lisp programs. Sequences consisting of C-c and a letter (either upper or lower case; ASCII or non-ASCII) are reserved for users; they are the only sequences reserved for users, so do not block them.

    Changing all the Emacs major modes to respect this convention was a lot of work; abandoning this convention would make that work go to waste, and inconvenience users. Please comply with it.

  • Function keys F5 through F9 without modifier keys are also reserved for users to define.
  • Sequences consisting of C-c followed by a control character or a digit are reserved for major modes.
  • Sequences consisting of C-c followed by {, }, <, >, : or ; are also reserved for major modes.
  • Sequences consisting of C-c followed by any other ASCII punctuation or symbol character are allocated for minor modes. Using them in a major mode is not absolutely prohibited, but if you do that, the major mode binding may be shadowed from time to time by minor modes.
  • Don’t bind C-h following any prefix character (including C-c). If you don’t bind C-h, it is automatically available as a help character for listing the subcommands of the prefix character.
  • Don’t bind a key sequence ending in ESC except following another ESC. (That is, it is OK to bind a sequence ending in ESC ESC.)

    The reason for this rule is that a non-prefix binding for ESC in any context prevents recognition of escape sequences as function keys in that context.

  • Similarly, don’t bind a key sequence ending in C-g, since that is commonly used to cancel a key sequence.
  • Anything that acts like a temporary mode or state that the user can enter and leave should define ESC ESC or ESC ESC ESC as a way to escape.

    For a state that accepts ordinary Emacs commands, or more generally any kind of state in which ESC followed by a function key or arrow key is potentially meaningful, then you must not define ESC ESC, since that would preclude recognizing an escape sequence after ESC. In these states, you should define ESC ESC ESC as the way to escape. Otherwise, define ESC ESC instead.


D.3 Emacs Programming Tips

Following these conventions will make your program fit better into Emacs when it runs.

  • Don’t use next-line or previous-line in programs; nearly always, forward-line is more convenient as well as more predictable and robust. See Motion by Text Lines.
  • Don’t call functions that set the mark, unless setting the mark is one of the intended features of your program. The mark is a user-level feature, so it is incorrect to change the mark except to supply a value for the user’s benefit. See The Mark.

    In particular, don’t use any of these functions:

    • beginning-of-buffer, end-of-buffer
    • replace-string, replace-regexp
    • insert-file, insert-buffer

    If you just want to move point, or replace a certain string, or insert a file or buffer’s contents, without any of the other features intended for interactive users, you can replace these functions with one or two lines of simple Lisp code.

  • Use lists rather than vectors, except when there is a particular reason to use a vector. Lisp has more facilities for manipulating lists than for vectors, and working with lists is usually more convenient.

    Vectors are advantageous for tables that are substantial in size and are accessed in random order (not searched front to back), provided there is no need to insert or delete elements (only lists allow that).

  • The recommended way to show a message in the echo area is with the message function, not princ. See The Echo Area.
  • When you encounter an error condition, call the function error (or signal). The function error does not return. See How to Signal an Error.

    Don’t use message, throw, sleep-for, or beep to report errors.

  • An error message should start with a capital letter but should not end with a period or other punctuation.

    It is occasionally useful to tell the user where an error originated, even if debug-on-error is nil. In such cases, a lower-case Lisp symbol can be prepended to the error message. For example, the error message “Invalid input” could be extended to say “some-function: Invalid input”.

  • A question asked in the minibuffer with yes-or-no-p or y-or-n-p should start with a capital letter and end with ‘?’.
  • When you mention a default value in a minibuffer prompt, put it and the word ‘default’ inside parentheses. It should look like this:
    Enter the answer (default 42):
    
  • In interactive, if you use a Lisp expression to produce a list of arguments, don’t try to provide the correct default values for region or position arguments. Instead, provide nil for those arguments if they were not specified, and have the function body compute the default value when the argument is nil. For instance, write this:
    (defun foo (pos)
      (interactive
       (list (if specified specified-pos)))
      (unless pos (setq pos default-pos))
      ...)
    

    rather than this:

    (defun foo (pos)
      (interactive
       (list (if specified specified-pos
                 default-pos)))
      ...)
    

    This is so that repetition of the command will recompute these defaults based on the current circumstances.

    You do not need to take such precautions when you use interactive specs ‘d’, ‘m’ and ‘r’, because they make special arrangements to recompute the argument values on repetition of the command.

  • Many commands that take a long time to execute display a message that says something like ‘Operating...’ when they start, and change it to ‘Operating...done’ when they finish. Please keep the style of these messages uniform: no space around the ellipsis, and no period after ‘done’. See Reporting Operation Progress, for an easy way to generate such messages.
  • Try to avoid using recursive edits. Instead, do what the Rmail e command does: use a new local keymap that contains a command defined to switch back to the old local keymap. Or simply switch to another buffer and let the user switch back at will. See Recursive Editing.

D.4 Tips for Making Compiled Code Fast

Here are ways of improving the execution speed of byte-compiled Lisp programs.

  • Profile your program, to find out where the time is being spent. See Profiling.
  • Use iteration rather than recursion whenever possible. Function calls are slow in Emacs Lisp even when a compiled function is calling another compiled function.
  • Using the primitive list-searching functions memq, member, assq, or assoc is even faster than explicit iteration. It can be worth rearranging a data structure so that one of these primitive search functions can be used.
  • Certain built-in functions are handled specially in byte-compiled code, avoiding the need for an ordinary function call. It is a good idea to use these functions rather than alternatives. To see whether a function is handled specially by the compiler, examine its byte-compile property. If the property is non-nil, then the function is handled specially.

    For example, the following input will show you that aref is compiled specially (see Functions that Operate on Arrays):

    (get 'aref 'byte-compile)
         ⇒ byte-compile-two-args
    

    Note that in this case (and many others), you must first load the bytecomp library, which defines the byte-compile property.

  • If calling a small function accounts for a substantial part of your program’s running time, make the function inline. This eliminates the function call overhead. Since making a function inline reduces the flexibility of changing the program, don’t do it unless it gives a noticeable speedup in something slow enough that users care about the speed. See Inline Functions.

D.5 Tips for Avoiding Compiler Warnings

  • Try to avoid compiler warnings about undefined free variables, by adding dummy defvar definitions for these variables, like this:
    (defvar foo)
    

    Such a definition has no effect except to tell the compiler not to warn about uses of the variable foo in this file.

  • Similarly, to avoid a compiler warning about an undefined function that you know will be defined, use a declare-function statement (see Telling the Compiler that a Function is Defined).
  • If you use many functions, macros, and variables from a certain file, you can add a require (see require) for that package to avoid compilation warnings for them, like this:
    (require 'foo)
    

    If you need only macros from some file, you can require it only at compile time (see Evaluation During Compilation). For instance,

    (eval-when-compile
      (require 'foo))
    
  • If you bind a variable in one function, and use it or set it in another function, the compiler warns about the latter function unless the variable has a definition. But adding a definition would be unclean if the variable has a short name, since Lisp packages should not define short variable names. The right thing to do is to rename this variable to start with the name prefix used for the other functions and variables in your package.
  • The last resort for avoiding a warning, when you want to do something that is usually a mistake but you know is not a mistake in your usage, is to put it inside with-no-warnings. See Compiler Errors.

D.6 Tips for Documentation Strings

Here are some tips and conventions for the writing of documentation strings. You can check many of these conventions by running the command M-x checkdoc-minor-mode.

  • Every command, function, or variable intended for users to know about should have a documentation string.
  • An internal variable or subroutine of a Lisp program might as well have a documentation string. Documentation strings take up very little space in a running Emacs.
  • Format the documentation string so that it fits in an Emacs window on an 80-column screen. It is a good idea for most lines to be no wider than 60 characters. The first line should not be wider than 67 characters or it will look bad in the output of apropos.

    You can fill the text if that looks good. Emacs Lisp mode fills documentation strings to the width specified by emacs-lisp-docstring-fill-column. However, you can sometimes make a documentation string much more readable by adjusting its line breaks with care. Use blank lines between sections if the documentation string is long.

  • The first line of the documentation string should consist of one or two complete sentences that stand on their own as a summary. M-x apropos displays just the first line, and if that line’s contents don’t stand on their own, the result looks bad. In particular, start the first line with a capital letter and end it with a period.

    For a function, the first line should briefly answer the question, “What does this function do?” For a variable, the first line should briefly answer the question, “What does this value mean?” Prefer to answer these questions in a way that will make sense to users and callers of the function or the variable. In particular, do not tell what the function does by enumerating the actions of its code; instead, describe the role of these actions and the function’s contract.

    Don’t limit the documentation string to one line; use as many lines as you need to explain the details of how to use the function or variable. Please use complete sentences for the rest of the text too.

  • When the user tries to use a disabled command, Emacs displays just the first paragraph of its documentation string—everything through the first blank line. If you wish, you can choose which information to include before the first blank line so as to make this display useful.
  • The first line should mention all the important arguments of the function (in particular, the mandatory arguments), and should mention them in the order that they are written in a function call. If the function has many arguments, then it is not feasible to mention them all in the first line; in that case, the first line should mention the first few arguments, including the most important arguments.
  • When a function’s documentation string mentions the value of an argument of the function, use the argument name in capital letters as if it were a name for that value. Thus, the documentation string of the function eval refers to its first argument as ‘FORM’, because the actual argument name is form:
    Evaluate FORM and return its value.
    

    Also write metasyntactic variables in capital letters, such as when you show the decomposition of a list or vector into subunits, some of which may vary. ‘KEY’ and ‘VALUE’ in the following example illustrate this practice:

    The argument TABLE should be an alist whose elements
    have the form (KEY . VALUE).  Here, KEY is ...
    
  • Never change the case of a Lisp symbol when you mention it in a doc string. If the symbol’s name is foo, write “foo”, not “Foo” (which is a different symbol).

    This might appear to contradict the policy of writing function argument values, but there is no real contradiction; the argument value is not the same thing as the symbol that the function uses to hold the value.

    If this puts a lower-case letter at the beginning of a sentence and that annoys you, rewrite the sentence so that the symbol is not at the start of it.

  • Do not start or end a documentation string with whitespace.
  • Do not indent subsequent lines of a documentation string so that the text is lined up in the source code with the text of the first line. This looks nice in the source code, but looks bizarre when users view the documentation. Remember that the indentation before the starting double-quote is not part of the string!
  • When documentation should display an ASCII apostrophe or grave accent, use ‘\\='’ or ‘\\=`’ in the documentation string literal so that the character is displayed as-is.
  • In documentation strings, do not quote expressions that are not Lisp symbols, as these expressions can stand for themselves. For example, write ‘Return the list (NAME TYPE RANGE) ...’ instead of ‘Return the list `(NAME TYPE RANGE)' ...’ or ‘Return the list \\='(NAME TYPE RANGE) ...’.
  • When a documentation string refers to a Lisp symbol, write it as it would be printed (which usually means in lower case), with a grave accent ‘`’ before and apostrophe ‘'’ after it. There are two exceptions: write t and nil without surrounding punctuation. For example:
    CODE can be `lambda', nil, or t.
    

    Note that when Emacs displays these doc strings, Emacs will usually display ‘`’ (grave accent) as ‘’ (left single quotation mark) and ‘'’ (apostrophe) as ‘’ (right single quotation mark), if the display supports displaying these characters. See Substituting Key Bindings in Documentation. (Some previous versions of this section recommended using the non-ASCII single quotation marks directly in doc strings, but this is now discouraged, since that leads to broken help string displays on terminals that don’t support displaying those characters.)

    Help mode automatically creates a hyperlink when a documentation string uses a single-quoted symbol name, if the symbol has either a function or a variable definition. You do not need to do anything special to make use of this feature. However, when a symbol has both a function definition and a variable definition, and you want to refer to just one of them, you can specify which one by writing one of the words ‘variable’, ‘option’, ‘function’, or ‘command’, immediately before the symbol name. (Case makes no difference in recognizing these indicator words.) For example, if you write

    This function sets the variable `buffer-file-name'.
    

    then the hyperlink will refer only to the variable documentation of buffer-file-name, and not to its function documentation.

    If a symbol has a function definition and/or a variable definition, but those are irrelevant to the use of the symbol that you are documenting, you can write the words ‘symbol’ or ‘program’ before the symbol name to prevent making any hyperlink. For example,

    If the argument KIND-OF-RESULT is the symbol `list',
    this function returns a list of all the objects
    that satisfy the criterion.
    

    does not make a hyperlink to the documentation, irrelevant here, of the function list.

    Normally, no hyperlink is made for a variable without variable documentation. You can force a hyperlink for such variables by preceding them with one of the words ‘variable’ or ‘option’.

    Hyperlinks for faces are only made if the face name is preceded or followed by the word ‘face’. In that case, only the face documentation will be shown, even if the symbol is also defined as a variable or as a function.

    To make a hyperlink to Info documentation, write the single-quoted name of the Info node (or anchor), preceded by ‘info node’, ‘Info node’, ‘info anchor’ or ‘Info anchor’. The Info file name defaults to ‘emacs’. For example,

    See Info node `Font Lock' and Info node `(elisp)Font Lock Basics'.
    

    To make a hyperlink to a man page, write the single-quoted name of the man page, preceded by ‘Man page’, ‘man page’, or ‘man page for’. For example,

    See the man page `chmod(1)' for details.
    

    The Info documentation is always preferable to man pages, so be sure to link to an Info manual where available. For example, chmod is documented in the GNU Coreutils manual, so it is better to link to that instead of the man page.

    To link to a customization group, write the single-quoted name of the group, preceded by ‘customization group’ (the first character in each word is case-insensitive). For example,

    See the customization group `whitespace' for details.
    

    Finally, to create a hyperlink to URLs, write the single-quoted URL, preceded by ‘URL’. For example,

    The GNU project website has more information (see URL
    `https://www.gnu.org/').
    
  • Don’t write key sequences directly in documentation strings. Instead, use the ‘\\[…]’ construct to stand for them. For example, instead of writing ‘C-f’, write the construct ‘\\[forward-char]’. When Emacs displays the documentation string, it substitutes whatever key is currently bound to forward-char. (This is normally ‘C-f’, but it may be some other character if the user has moved key bindings.) See Substituting Key Bindings in Documentation.
  • In documentation strings for a major mode, you will want to refer to the key bindings of that mode’s local map, rather than global ones. Therefore, use the construct ‘\\<…>’ once in the documentation string to specify which key map to use. Do this before the first use of ‘\\[…]’. The text inside the ‘\\<…>’ should be the name of the variable containing the local keymap for the major mode.

    Each use of ‘\\[…]’ slows the display of the documentation string by a tiny amount. If you use a lot of them, these tiny slowdowns will add up, and might become tangible, especially on slow systems. So our recommendation is not to over-use them; e.g., try to avoid using more than one reference to the same command in the same doc string.

  • For consistency, phrase the verb in the first sentence of a function’s documentation string as an imperative—for instance, use “Return the cons of A and B.” in preference to “Returns the cons of A and B.” Usually it looks good to do likewise for the rest of the first paragraph. Subsequent paragraphs usually look better if each sentence is indicative and has a proper subject.
  • The documentation string for a function that is a yes-or-no predicate should start with words such as “Return t if”, to indicate explicitly what constitutes truth. The word “return” avoids starting the sentence with lower-case “t”, which could be somewhat distracting.
  • Write documentation strings in the active voice, not the passive, and in the present tense, not the future. For instance, use “Return a list containing A and B.” instead of “A list containing A and B will be returned.”
  • Avoid using the word “cause” (or its equivalents) unnecessarily. Instead of, “Cause Emacs to display text in boldface”, write just “Display text in boldface”.
  • Avoid using “iff” (a mathematics term meaning “if and only if”), since many people are unfamiliar with it and mistake it for a typo. In most cases, the meaning is clear with just “if”. Otherwise, try to find an alternate phrasing that conveys the meaning.
  • Try to avoid using abbreviations such as “e.g.” (for “for example”), “i.e.” (for “that is”), “no.” (for “number”), “cf.” (for “compare”/“see also”) and “w.r.t.” (for “with respect to”) as much as possible. It is almost always clearer and easier to read the expanded version.34
  • When a command is meaningful only in a certain mode or situation, do mention that in the documentation string. For example, the documentation of dired-find-file is:
    In Dired, visit the file or directory named on this line.
    
  • When you define a variable that represents an option users might want to set, use defcustom. See Defining Global Variables.
  • The documentation string for a variable that is a yes-or-no flag should start with words such as “Non-nil means”, to make it clear that all non-nil values are equivalent and indicate explicitly what nil and non-nil mean.
  • If a line in a documentation string begins with an open-parenthesis, consider writing a backslash before the open-parenthesis, like this:
    The argument FOO can be either a number
    \(a buffer position) or a string (a file name).
    

    This avoids a bug in Emacs versions older than 27.1, where the ‘(’ was treated as the start of a defun (see Defuns in The GNU Emacs Manual). If you do not anticipate anyone editing your code with older Emacs versions, there is no need for this work-around.


D.7 Tips on Writing Comments

We recommend these conventions for comments:

;

Comments that start with a single semicolon, ‘;’, should all be aligned to the same column on the right of the source code. Such comments usually explain how the code on that line does its job. For example:

(setq base-version-list                 ; There was a base
      (assoc (substring fn 0 start-vn)  ; version to which
             file-version-assoc-list))  ; this looks like
                                        ; a subversion.
;;

Comments that start with two semicolons, ‘;;’, should be aligned to the same level of indentation as the code. Such comments usually describe the purpose of the following lines or the state of the program at that point. For example:

(prog1 (setq auto-fill-function
             …
             …
  ;; Update mode line.
  (force-mode-line-update)))

We also normally use two semicolons for comments outside functions.

;; This Lisp code is run in Emacs when it is to operate as
;; a server for other processes.

If a function has no documentation string, it should instead have a two-semicolon comment right before the function, explaining what the function does and how to call it properly. Explain precisely what each argument means and how the function interprets its possible values. It is much better to convert such comments to documentation strings, though.

;;;

Comments that start with three (or more) semicolons, ‘;;;’, should start at the left margin. We use them for comments that should be considered a heading by Outline minor mode. By default, comments starting with at least three semicolons (followed by a single space and a non-whitespace character) are considered section headings, comments starting with two or fewer are not.

(Historically, triple-semicolon comments have also been used for commenting out lines within a function, but this use is discouraged in favor of using just two semicolons. This also applies when commenting out entire functions; when doing that use two semicolons as well.)

Three semicolons are used for top-level sections, four for sub-sections, five for sub-sub-sections and so on.

Typically libraries have at least four top-level sections. For example when the bodies of all of these sections are hidden:

;;; backquote.el --- implement the ` Lisp construct...
;;; Commentary:...
;;; Code:...
;;; backquote.el ends here

(In a sense the last line is not a section heading as it must never be followed by any text; after all it marks the end of the file.)

For longer libraries it is advisable to split the code into multiple sections. This can be done by splitting the ‘Code:’ section into multiple sub-sections. Even though that was the only recommended approach for a long time, many people have chosen to use multiple top-level code sections instead. You may chose either style.

Using multiple top-level code sections has the advantage that it avoids introducing an additional nesting level but it also means that the section named ‘Code’ does not contain all the code, which is awkward. To avoid that, you should put no code at all inside that section; that way it can be considered a separator instead of a section heading.

Finally, we recommend that you don’t end headings with a colon or any other punctuation for that matter. For historic reasons the ‘Code:’ and ‘Commentary:’ headings end with a colon, but we recommend that you don’t do the same for other headings anyway.

Generally speaking, the M-; (comment-dwim) command automatically starts a comment of the appropriate type; or indents an existing comment to the right place, depending on the number of semicolons. See Manipulating Comments in The GNU Emacs Manual.


D.8 Conventional Headers for Emacs Libraries

Emacs has conventions for using special comments in Lisp libraries to divide them into sections and give information such as who wrote them. Using a standard format for these items makes it easier for tools (and people) to extract the relevant information. This section explains these conventions, starting with an example:

;;; foo.el --- Support for the Foo programming language  -*- lexical-binding: t; -*-

;; Copyright (C) 2010-2021 Your Name

;; Author: Your Name <yourname@example.com>
;; Maintainer: Someone Else <someone@example.com>
;; Created: 14 Jul 2010
;; Keywords: languages
;; URL: https://example.com/foo

;; This file is not part of GNU Emacs.

;; This file is free software…
…
;; along with this file.  If not, see <https://www.gnu.org/licenses/>.

The very first line should have this format:

;;; filename --- description  -*- lexical-binding: t; -*-

The description should be contained in one line. If the file needs to set more variables in the ‘-*-’ specification, add it after lexical-binding. If this would make the first line too long, use a Local Variables section at the end of the file.

The copyright notice usually lists your name (if you wrote the file). If you have an employer who claims copyright on your work, you might need to list them instead. Do not say that the copyright holder is the Free Software Foundation (or that the file is part of GNU Emacs) unless your file has been accepted into the Emacs distribution. For more information on the form of copyright and license notices, see the guide on the GNU website.

After the copyright notice come several header comment lines, each beginning with ‘;; header-name:’. Here is a table of the conventional possibilities for header-name:

Author

This header states the name and email address of at least the principal author of the library. If there are multiple authors, list them on continuation lines led by ;; and a tab or at least two spaces. We recommend including a contact email address, of the form ‘<…>’. For example:

;; Author: Your Name <yourname@example.com>
;;      Someone Else <someone@example.com>
;;      Another Person <another@example.com>
Maintainer

This header has the same format as the Author header. It lists the person(s) who currently maintain(s) the file (respond to bug reports, etc.).

If there is no Maintainer header, the person(s) in the Author header is/are presumed to be the maintainer(s). Some files in Emacs use ‘emacs-devel@gnu.org’ for the maintainer, which means the author is no longer responsible for the file, and that it is maintained as part of Emacs.

Created

This optional line gives the original creation date of the file, and is for historical interest only.

Version

If you wish to record version numbers for the individual Lisp program, put them in this line. Lisp files distributed with Emacs generally do not have a ‘Version’ header, since the version number of Emacs itself serves the same purpose. If you are distributing a collection of multiple files, we recommend not writing the version in every file, but only the main one.

Keywords

This line lists keywords for the finder-by-keyword help command. Please use that command to see a list of the meaningful keywords. The command M-x checkdoc-package-keywords RET will find and display any keywords that are not in finder-known-keywords. If you set the variable checkdoc-package-keywords-flag non-nil, checkdoc commands will include the keyword verification in its checks.

This field is how people will find your package when they’re looking for things by topic. To separate the keywords, you can use spaces, commas, or both.

The name of this field is unfortunate, since people often assume it is the place to write arbitrary keywords that describe their package, rather than just the relevant Finder keywords.

URL
Homepage

These lines state the website of the library.

Package-Version

If ‘Version’ is not suitable for use by the package manager, then a package can define ‘Package-Version’; it will be used instead. This is handy if ‘Version’ is an RCS id or something else that cannot be parsed by version-to-list. See Packaging Basics.

Package-Requires

If this exists, it names packages on which the current package depends for proper operation. See Packaging Basics. This is used by the package manager both at download time (to ensure that a complete set of packages is downloaded) and at activation time (to ensure that a package is only activated if all its dependencies have been).

Its format is a list of lists on a single line. The car of each sub-list is the name of a package, as a symbol. The cadr of each sub-list is the minimum acceptable version number, as a string that can be parsed by version-to-list. An entry that lacks a version (i.e., an entry which is just a symbol, or a sub-list of one element) is equivalent to entry with version "0". For instance:

;; Package-Requires: ((gnus "1.0") (bubbles "2.7.2") cl-lib (seq))

Packages that don’t need to support Emacs versions older than Emacs 27 can have the ‘Package-Requires’ header split across multiple lines, like this:

;; Package-Requires: ((emacs "27.1")
;;                    (compat "29.1.4.1"))

Note that with this format, you still need to start the list on the same line as ‘Package-Requires’.

The package code automatically defines a package named ‘emacs’ with the version number of the currently running Emacs. This can be used to require a minimal version of Emacs for a package.

Just about every Lisp library ought to have the ‘Author’ and ‘Keywords’ header comment lines. Use the others if they are appropriate. You can also put in header lines with other header names—they have no standard meanings, so they can’t do any harm.

We use additional stylized comments to subdivide the contents of the library file. These should be separated from anything else by blank lines. Here is a table of them:

;;; Commentary:

This begins introductory comments that explain how the library works. It should come right after the copying permissions, terminated by a ‘Change Log’, ‘History’ or ‘Code’ comment line. This text is used by the Finder package, so it should make sense in that context.

;;; Change Log:

This begins an optional log of changes to the file over time. Don’t put too much information in this section—it is better to keep the detailed logs in a version control system (as Emacs does) or in a separate ChangeLog file. ‘History’ is an alternative to ‘Change Log’.

;;; Code:

This begins the actual code of the program.

;;; filename ends here

This is the footer line; it appears at the very end of the file. Its purpose is to enable people to detect truncated versions of the file from the lack of a footer line.


Appendix E GNU Emacs Internals

This chapter describes how the runnable Emacs executable is dumped with the preloaded Lisp libraries in it, how storage is allocated, and some internal aspects of GNU Emacs that may be of interest to C programmers.


E.1 Building Emacs

This section explains the steps involved in building the Emacs executable. You don’t have to know this material to build and install Emacs, since the makefiles do all these things automatically. This information is pertinent to Emacs developers.

Building Emacs requires GNU Make version 3.81 or later.

Compilation of the C source files in the src directory produces an executable file called temacs, also called a bare impure Emacs. It contains the Emacs Lisp interpreter and I/O routines, but not the editing commands.

The command temacs -l loadup would run temacs and direct it to load loadup.el. The loadup library loads additional Lisp libraries, which set up the normal Emacs editing environment. After this step, the Emacs executable is no longer bare.

Because it takes some time to load the standard Lisp files, the temacs executable usually isn’t run directly by users. Instead, one of the last steps of building Emacs runs the command ‘temacs -batch -l loadup --temacs=dump-method. The special option --temacs tells temacs how to record all the standard preloaded Lisp functions and variables, so that when you subsequently run Emacs, it will start much faster. The --temacs option requires an argument dump-method, which can be one of the following:

pdump

Record the preloaded Lisp data in a dump file. This method produces an additional data file which Emacs will load at startup. The produced dump file is usually called emacs.pdmp, and is installed in the Emacs exec-directory (see Help Functions). This method is the most preferred one, as it does not require Emacs to employ any special techniques of memory allocation, which might get in the way of various memory-layout techniques used by modern systems to enhance security and privacy.

pbootstrap

Like ‘pdump’, but used while bootstrapping Emacs, when no previous Emacs binary and no *.elc byte-compiled Lisp files are available. The produced dump file is usually named bootstrap-emacs.pdmp in this case.

dump

This method causes temacs to dump out an executable program, called emacs, which has all the standard Lisp files already preloaded into it. (The ‘-batch’ argument prevents temacs from trying to initialize any of its data on the terminal, so that the tables of terminal information are empty in the dumped Emacs.) This method is also known as unexec, because it produces a program file from a running process, and thus is in some sense the opposite of executing a program to start a process. Although this method was the way that Emacs traditionally saved its state, it is now deprecated.

bootstrap

Like ‘dump’, but used when bootstrapping Emacs with the unexec method.

The dumped emacs executable (also called a pure Emacs) is the one which is installed. If the portable dumper was used to build Emacs, the emacs executable is actually an exact copy of temacs, and the corresponding emacs.pdmp file is installed as well. The variable preloaded-file-list stores a list of the preloaded Lisp files recorded in the dump file or in the dumped Emacs executable. If you port Emacs to a new operating system, and are not able to implement dumping of any kind, then Emacs must load loadup.el each time it starts.

By default the dumped emacs executable records details such as the build time and host name. Use the --disable-build-details option of configure to suppress these details, so that building and installing Emacs twice from the same sources is more likely to result in identical copies of Emacs.

You can specify additional files to preload by writing a library named site-load.el that loads them. You may need to rebuild Emacs with an added definition

#define SITELOAD_PURESIZE_EXTRA n

to make n added bytes of pure space to hold the additional files; see src/puresize.h. (Try adding increments of 20000 until it is big enough.) However, the advantage of preloading additional files decreases as machines get faster. On modern machines, it is usually not advisable.

After loadup.el reads site-load.el, it finds the documentation strings for primitive and preloaded functions (and variables) in the file etc/DOC where they are stored, by calling Snarf-documentation (see Accessing Documentation).

You can specify other Lisp expressions to execute just before dumping by putting them in a library named site-init.el. This file is executed after the documentation strings are found.

If you want to preload function or variable definitions, there are three ways you can do this and make their documentation strings accessible when you subsequently run Emacs:

  • Arrange to scan these files when producing the etc/DOC file, and load them with site-load.el.
  • Load the files with site-init.el, then copy the files into the installation directory for Lisp files when you install Emacs.
  • Specify a nil value for byte-compile-dynamic-docstrings as a local variable in each of these files, and load them with either site-load.el or site-init.el. (This method has the drawback that the documentation strings take up space in Emacs all the time.)

It is not advisable to put anything in site-load.el or site-init.el that would alter any of the features that users expect in an ordinary unmodified Emacs. If you feel you must override normal features for your site, do it with default.el, so that users can override your changes if they wish. See Summary: Sequence of Actions at Startup. Note that if either site-load.el or site-init.el changes load-path, the changes will be lost after dumping. See Library Search. To make a permanent change to load-path, use the --enable-locallisppath option of configure.

In a package that can be preloaded, it is sometimes necessary (or useful) to delay certain evaluations until Emacs subsequently starts up. The vast majority of such cases relate to the values of customizable variables. For example, tutorial-directory is a variable defined in startup.el, which is preloaded. The default value is set based on data-directory. The variable needs to access the value of data-directory when Emacs starts, not when it is dumped, because the Emacs executable has probably been installed in a different location since it was dumped.

Function: custom-initialize-delay symbol value

This function delays the initialization of symbol to the next Emacs start. You normally use this function by specifying it as the :initialize property of a customizable variable. (The argument value is unused, and is provided only for compatibility with the form Custom expects.)

In the unlikely event that you need a more general functionality than custom-initialize-delay provides, you can use before-init-hook (see Summary: Sequence of Actions at Startup).

Function: dump-emacs-portable to-file &optional track-referrers

This function dumps the current state of Emacs into a dump file to-file, using the pdump method. Normally, the dump file is called emacs-name.dmp, where emacs-name is the name of the Emacs executable file. The optional argument track-referrers, if non-nil, causes the portable dumper to keep additional information to help track down the provenance of object types that are not yet supported by the pdump method.

Although the portable dumper code can run on many platforms, the dump files that it produces are not portable—they can be loaded only by the Emacs executable that dumped them.

If you want to use this function in an Emacs that was already dumped, you must run Emacs with the ‘-batch’ option.

If you’re including ‘.el’ files in the dumped Emacs and that ‘.el’ file has code that is normally run at load time, that code won’t be run when Emacs starts after dumping. To help work around that problem, you can put functions on the after-pdump-load-hook hook. This hook is run when starting Emacs.

Function: dump-emacs to-file from-file

This function dumps the current state of Emacs into an executable file to-file, using the unexec method. It takes symbols from from-file (this is normally the executable file temacs).

This function cannot be used in an Emacs that was already dumped. This function is deprecated, and by default Emacs is built without unexec support so this function is not available.

Function: pdumper-stats

If the current Emacs session restored its state from a dump file, this function returns information about the dump file and the time it took to restore the Emacs state. The value is an alist ((dumped-with-pdumper . t) (load-time . time(dump-file-name . file)), where file is the name of the dump file, and time is the time in seconds it took to restore the state from the dump file. If the current session was not restored from a dump file, the value is nil.


E.2 Pure Storage

Emacs Lisp uses two kinds of storage for user-created Lisp objects: normal storage and pure storage. Normal storage is where all the new data created during an Emacs session are kept (see Garbage Collection). Pure storage is used for certain data in the preloaded standard Lisp files—data that should never change during actual use of Emacs.

Pure storage is allocated only while temacs is loading the standard preloaded Lisp libraries. In the file emacs, it is marked as read-only (on operating systems that permit this), so that the memory space can be shared by all the Emacs jobs running on the machine at once. Pure storage is not expandable; a fixed amount is allocated when Emacs is compiled, and if that is not sufficient for the preloaded libraries, temacs allocates dynamic memory for the part that didn’t fit. If Emacs will be dumped using the pdump method (see Building Emacs), the pure-space overflow is of no special importance (it just means some of the preloaded stuff cannot be shared with other Emacs jobs). However, if Emacs will be dumped using the now obsolete unexec method, the resulting image will work, but garbage collection (see Garbage Collection) is disabled in this situation, causing a memory leak. Such an overflow normally won’t happen unless you try to preload additional libraries or add features to the standard ones. Emacs will display a warning about the overflow when it starts, if it was dumped using unexec. If this happens, you should increase the compilation parameter SYSTEM_PURESIZE_EXTRA in the file src/puresize.h and rebuild Emacs.

Function: purecopy object

This function makes a copy in pure storage of object, and returns it. It copies a string by simply making a new string with the same characters, but without text properties, in pure storage. It recursively copies the contents of vectors and cons cells. It does not make copies of other objects such as symbols, but just returns them unchanged. It signals an error if asked to copy markers.

This function is a no-op except while Emacs is being built and dumped; it is usually called only in preloaded Lisp files.

Variable: pure-bytes-used

The value of this variable is the number of bytes of pure storage allocated so far. Typically, in a dumped Emacs, this number is very close to the total amount of pure storage available—if it were not, we would preallocate less.

Variable: purify-flag

This variable determines whether defun should make a copy of the function definition in pure storage. If it is non-nil, then the function definition is copied into pure storage.

This flag is t while loading all of the basic functions for building Emacs initially (allowing those functions to be shareable and non-collectible). Dumping Emacs as an executable always writes nil in this variable, regardless of the value it actually has before and after dumping.

You should not change this flag in a running Emacs.


E.3 Garbage Collection

When a program creates a list or the user defines a new function (such as by loading a library), that data is placed in normal storage. If normal storage runs low, then Emacs asks the operating system to allocate more memory. Different types of Lisp objects, such as symbols, cons cells, small vectors, markers, etc., are segregated in distinct blocks in memory. (Large vectors, long strings, buffers and certain other editing types, which are fairly large, are allocated in individual blocks, one per object; small strings are packed into blocks of 8k bytes, and small vectors are packed into blocks of 4k bytes).

Beyond the basic vector, a lot of objects like markers, overlays and buffers are managed as if they were vectors. The corresponding C data structures include the union vectorlike_header field whose size member contains the subtype enumerated by enum pvec_type and an information about how many Lisp_Object fields this structure contains and what the size of the rest data is. This information is needed to calculate the memory footprint of an object, and used by the vector allocation code while iterating over the vector blocks.

It is quite common to use some storage for a while, then release it by (for example) killing a buffer or deleting the last pointer to an object. Emacs provides a garbage collector to reclaim this abandoned storage. The garbage collector operates, in essence, by finding and marking all Lisp objects that are still accessible to Lisp programs. To begin with, it assumes all the symbols, their values and associated function definitions, and any data presently on the stack, are accessible. Any objects that can be reached indirectly through other accessible objects are also accessible, but this calculation is done “conservatively”, so it may slightly overestimate how many objects that are accessible.

When marking is finished, all objects still unmarked are garbage. No matter what the Lisp program or the user does, it is impossible to refer to them, since there is no longer a way to reach them. Their space might as well be reused, since no one will miss them. The second (sweep) phase of the garbage collector arranges to reuse them. (But since the marking was done “conservatively”, not all unused objects are guaranteed to be garbage-collected by any one sweep.)

The sweep phase puts unused cons cells onto a free list for future allocation; likewise for symbols and markers. It compacts the accessible strings so they occupy fewer 8k blocks; then it frees the other 8k blocks. Unreachable vectors from vector blocks are coalesced to create largest possible free areas; if a free area spans a complete 4k block, that block is freed. Otherwise, the free area is recorded in a free list array, where each entry corresponds to a free list of areas of the same size. Large vectors, buffers, and other large objects are allocated and freed individually.

Common Lisp note: Unlike other Lisps, GNU Emacs Lisp does not call the garbage collector when the free list is empty. Instead, it simply requests the operating system to allocate more storage, and processing continues until gc-cons-threshold bytes have been used.

This means that you can make sure that the garbage collector will not run during a certain portion of a Lisp program by calling the garbage collector explicitly just before it (provided that portion of the program does not use so much space as to force a second garbage collection).

Command: garbage-collect

This command runs a garbage collection, and returns information on the amount of space in use. (Garbage collection can also occur spontaneously if you use more than gc-cons-threshold bytes of Lisp data since the previous garbage collection.)

garbage-collect returns a list with information on amount of space in use, where each entry has the form ‘(name size used)’ or ‘(name size used free)’. In the entry, name is a symbol describing the kind of objects this entry represents, size is the number of bytes used by each one, used is the number of those objects that were found live in the heap, and optional free is the number of those objects that are not live but that Emacs keeps around for future allocations. So an overall result is:

((conses cons-size used-conses free-conses)
 (symbols symbol-size used-symbols free-symbols)
 (strings string-size used-strings free-strings)
 (string-bytes byte-size used-bytes)
 (vectors vector-size used-vectors)
 (vector-slots slot-size used-slots free-slots)
 (floats float-size used-floats free-floats)
 (intervals interval-size used-intervals free-intervals)
 (buffers buffer-size used-buffers)
 (heap unit-size total-size free-size))

Here is an example:

(garbage-collect)
      ⇒ ((conses 16 49126 8058) (symbols 48 14607 0)
                 (strings 32 2942 2607)
                 (string-bytes 1 78607) (vectors 16 7247)
                 (vector-slots 8 341609 29474) (floats 8 71 102)
                 (intervals 56 27 26) (buffers 944 8)
                 (heap 1024 11715 2678))

Below is a table explaining each element. Note that last heap entry is optional and present only if an underlying malloc implementation provides mallinfo function.

cons-size

Internal size of a cons cell, i.e., sizeof (struct Lisp_Cons).

used-conses

The number of cons cells in use.

free-conses

The number of cons cells for which space has been obtained from the operating system, but that are not currently being used.

symbol-size

Internal size of a symbol, i.e., sizeof (struct Lisp_Symbol).

used-symbols

The number of symbols in use.

free-symbols

The number of symbols for which space has been obtained from the operating system, but that are not currently being used.

string-size

Internal size of a string header, i.e., sizeof (struct Lisp_String).

used-strings

The number of string headers in use.

free-strings

The number of string headers for which space has been obtained from the operating system, but that are not currently being used.

byte-size

This is used for convenience and equals to sizeof (char).

used-bytes

The total size of all string data in bytes.

vector-size

Size in bytes of a vector of length 1, including its header.

used-vectors

The number of vector headers allocated from the vector blocks.

slot-size

Internal size of a vector slot, always equal to sizeof (Lisp_Object).

used-slots

The number of slots in all used vectors. Slot counts might include some or all overhead from vector headers, depending on the platform.

free-slots

The number of free slots in all vector blocks.

float-size

Internal size of a float object, i.e., sizeof (struct Lisp_Float). (Do not confuse it with the native platform float or double.)

used-floats

The number of floats in use.

free-floats

The number of floats for which space has been obtained from the operating system, but that are not currently being used.

interval-size

Internal size of an interval object, i.e., sizeof (struct interval).

used-intervals

The number of intervals in use.

free-intervals

The number of intervals for which space has been obtained from the operating system, but that are not currently being used.

buffer-size

Internal size of a buffer, i.e., sizeof (struct buffer). (Do not confuse with the value returned by buffer-size function.)

used-buffers

The number of buffer objects in use. This includes killed buffers invisible to users, i.e., all buffers in all_buffers list.

unit-size

The unit of heap space measurement, always equal to 1024 bytes.

total-size

Total heap size, in unit-size units.

free-size

Heap space which is not currently used, in unit-size units.

If there was overflow in pure space (see Pure Storage), and Emacs was dumped using the (now obsolete) unexec method (see Building Emacs), then garbage-collect returns nil, because a real garbage collection cannot be done in that case.

User Option: garbage-collection-messages

If this variable is non-nil, Emacs displays a message at the beginning and end of garbage collection. The default value is nil.

Variable: post-gc-hook

This is a normal hook that is run at the end of garbage collection. Garbage collection is inhibited while the hook functions run, so be careful writing them.

User Option: gc-cons-threshold

The value of this variable is the number of bytes of storage that must be allocated for Lisp objects after one garbage collection in order to trigger another garbage collection. You can use the result returned by garbage-collect to get an information about size of the particular object type; space allocated to the contents of buffers does not count.

The initial threshold value is GC_DEFAULT_THRESHOLD, defined in alloc.c. Since it’s defined in word_size units, the value is 400,000 for the default 32-bit configuration and 800,000 for the 64-bit one. If you specify a larger value, garbage collection will happen less often. This reduces the amount of time spent garbage collecting (so Lisp programs will run faster between cycles of garbage collection that happen more rarely), but increases total memory use. You may want to do this when running a program that creates lots of Lisp data, especially if you need it to run faster. However, we recommend against increasing the threshold for prolonged periods of time, and advise that you never set it higher than needed for the program to run in reasonable time. Using thresholds higher than necessary could potentially cause higher system-wide memory pressure, and also make each garbage-collection cycle take much more time, and should therefore be avoided.

You can make collections more frequent by specifying a smaller value, down to 1/10th of GC_DEFAULT_THRESHOLD. A value less than this minimum will remain in effect only until the subsequent garbage collection, at which time garbage-collect will set the threshold back to the minimum.

User Option: gc-cons-percentage

The value of this variable specifies the amount of consing before a garbage collection occurs, as a fraction of the current heap size. This criterion and gc-cons-threshold apply in parallel, and garbage collection occurs only when both criteria are satisfied.

As the heap size increases, the time to perform a garbage collection increases. Thus, it can be desirable to do them less frequently in proportion.

As with gc-cons-threshold, do not enlarge this more than necessary, and never for prolonged periods of time.

Control over the garbage collector via gc-cons-threshold and gc-cons-percentage is only approximate. Although Emacs checks for threshold exhaustion regularly, for efficiency reasons it does not do so immediately after every change to the heap or to gc-cons-threshold or gc-cons-percentage, so exhausting the threshold does not immediately trigger garbage collection. Also, for efficiency in threshold calculations Emacs approximates the heap size, which counts the bytes used by currently-accessible objects in the heap.

The value returned by garbage-collect describes the amount of memory used by Lisp data, broken down by data type. By contrast, the function memory-limit provides information on the total amount of memory Emacs is currently using.

Function: memory-limit

This function returns an estimate of the total amount of bytes of virtual memory that Emacs is currently using, divided by 1024. You can use this to get a general idea of how your actions affect the memory usage.

Variable: memory-full

This variable is t if Emacs is nearly out of memory for Lisp objects, and nil otherwise.

Function: memory-use-counts

This returns a list of numbers that count the number of objects created in this Emacs session. Each of these counters increments for a certain kind of object. See the documentation string for details.

Function: memory-info

This functions returns an amount of total system memory and how much of it is free. On an unsupported system, the value may be nil.

If default-directory points to a remote host, memory information of that host is returned.

Variable: gcs-done

This variable contains the total number of garbage collections done so far in this Emacs session.

Variable: gc-elapsed

This variable contains the total number of seconds of elapsed time during garbage collection so far in this Emacs session, as a floating-point number.

Function: memory-report

It can sometimes be useful to see where Emacs is using memory (in various variables, buffers, and caches). This command will open a new buffer (called ‘"*Memory Report*"’) that will give an overview, in addition to listing the “largest” buffers and variables.

All the data here is approximate, because there’s really no consistent way to compute the size of a variable. For instance, two variables may share parts of a data structure, and this will be counted twice, but this command may still give a useful high-level overview of which parts of Emacs are using memory.


E.4 Stack-allocated Objects

The garbage collector described above is used to manage data visible from Lisp programs, as well as most of the data internally used by the Lisp interpreter. Sometimes it may be useful to allocate temporary internal objects using the C stack of the interpreter. This can help performance, as stack allocation is typically faster than using heap memory to allocate and the garbage collector to free. The downside is that using such objects after they are freed results in undefined behavior, so uses should be well thought out and carefully debugged by using the GC_CHECK_MARKED_OBJECTS feature (see src/alloc.c). In particular, stack-allocated objects should never be made visible to user Lisp code.

Currently, cons cells and strings can be allocated this way. This is implemented by C macros like AUTO_CONS and AUTO_STRING that define a named Lisp_Object with block lifetime. These objects are not freed by the garbage collector; instead, they have automatic storage duration, i.e., they are allocated like local variables and are automatically freed at the end of execution of the C block that defined the object.

For performance reasons, stack-allocated strings are limited to ASCII characters, and many of these strings are immutable, i.e., calling ASET on them produces undefined behavior.


E.5 Memory Usage

These functions and variables give information about the total amount of memory allocation that Emacs has done, broken down by data type. Note the difference between these and the values returned by garbage-collect; those count objects that currently exist, but these count the number or size of all allocations, including those for objects that have since been freed.

Variable: cons-cells-consed

The total number of cons cells that have been allocated so far in this Emacs session.

Variable: floats-consed

The total number of floats that have been allocated so far in this Emacs session.

Variable: vector-cells-consed

The total number of vector cells that have been allocated so far in this Emacs session. This includes vector-like objects such as markers and overlays, plus certain objects not visible to users.

Variable: symbols-consed

The total number of symbols that have been allocated so far in this Emacs session.

Variable: string-chars-consed

The total number of string characters that have been allocated so far in this session.

Variable: intervals-consed

The total number of intervals that have been allocated so far in this Emacs session.

Variable: strings-consed

The total number of strings that have been allocated so far in this Emacs session.


E.6 C Dialect

The C part of Emacs is portable to C99 or later: C11-specific features such as ‘<stdalign.h>’ and ‘_Noreturn’ are not used without a check, typically at configuration time, and the Emacs build procedure provides a substitute implementation if necessary. Some C11 features, such as anonymous structures and unions, are too difficult to emulate, so they are avoided entirely.

At some point in the future the base C dialect will no doubt change to C11.


E.7 Writing Emacs Primitives

Lisp primitives are Lisp functions implemented in C. The details of interfacing the C function so that Lisp can call it are handled by a few C macros. The only way to really understand how to write new C code is to read the source, but we can explain some things here.

An example of a special form is the definition of or, from eval.c. (An ordinary function would have the same general appearance.)

DEFUN ("or", For, Sor, 0, UNEVALLED, 0,
       doc: /* Eval args until one of them yields non-nil,
then return that value.
The remaining args are not evalled at all.
If all args return nil, return nil.
usage: (or CONDITIONS...)  */)
  (Lisp_Object args)
{
  Lisp_Object val = Qnil;

  while (CONSP (args))
    {
      val = eval_sub (XCAR (args));
      if (!NILP (val))
        break;
      args = XCDR (args);
      maybe_quit ();
    }

  return val;
}

Let’s start with a precise explanation of the arguments to the DEFUN macro. Here is a template for them:

DEFUN (lname, fname, sname, min, max, interactive, doc)
lname

This is the name of the Lisp symbol to define as the function name; in the example above, it is or.

fname

This is the C function name for this function. This is the name that is used in C code for calling the function. The name is, by convention, ‘F’ prepended to the Lisp name, with all dashes (‘-’) in the Lisp name changed to underscores. Thus, to call this function from C code, call For.

sname

This is a C variable name to use for a structure that holds the data for the subr object that represents the function in Lisp. This structure conveys the Lisp symbol name to the initialization routine that will create the symbol and store the subr object as its definition. By convention, this name is always fname with ‘F’ replaced with ‘S’.

min

This is the minimum number of arguments that the function requires. The function or allows a minimum of zero arguments.

max

This is the maximum number of arguments that the function accepts, if there is a fixed maximum. Alternatively, it can be UNEVALLED, indicating a special form that receives unevaluated arguments, or MANY, indicating an unlimited number of evaluated arguments (the equivalent of &rest). Both UNEVALLED and MANY are macros. If max is a number, it must be more than min but less than 8.

interactive

This is an interactive specification, a string such as might be used as the argument of interactive in a Lisp function (see Using interactive). In the case of or, it is 0 (a null pointer), indicating that or cannot be called interactively. A value of "" indicates a function that should receive no arguments when called interactively. If the value begins with a ‘"(’, the string is evaluated as a Lisp form. For example:

DEFUN ("foo", Ffoo, Sfoo, 0, 3,
       "(list (read-char-by-name \"Insert character: \")\
              (prefix-numeric-value current-prefix-arg)\
              t)",
       doc: /* … */)
doc

This is the documentation string. It uses C comment syntax rather than C string syntax because comment syntax requires nothing special to include multiple lines. The ‘doc:’ identifies the comment that follows as the documentation string. The ‘/*’ and ‘*/’ delimiters that begin and end the comment are not part of the documentation string.

If the last line of the documentation string begins with the keyword ‘usage:’, the rest of the line is treated as the argument list for documentation purposes. This way, you can use different argument names in the documentation string from the ones used in the C code. ‘usage:’ is required if the function has an unlimited number of arguments.

Some primitives have multiple definitions, one per platform (e.g., x-create-frame). In such cases, rather than writing the same documentation string in each definition, only one definition has the actual documentation. The others have placeholders beginning with ‘SKIP’, which are ignored by the function that parses the DOC file.

All the usual rules for documentation strings in Lisp code (see Tips for Documentation Strings) apply to C code documentation strings too.

The documentation string can be followed by a list of C function attributes for the C function that implements the primitive, like this:

DEFUN ("bar", Fbar, Sbar, 0, UNEVALLED, 0
       doc: /* … */
       attributes: attr1 attr2 …)

You can specify more than a single attribute, one after the other. Currently, only the following attributes are recognized:

noreturn

Declares the C function as one that never returns. This corresponds to the C11 keyword _Noreturn and to __attribute__ ((__noreturn__)) attribute of GCC (see Function Attributes in Using the GNU Compiler Collection).

const

Declares that the function does not examine any values except its arguments, and has no effects except the return value. This corresponds to __attribute__ ((__const__)) attribute of GCC.

noinline

This corresponds to __attribute__ ((__noinline__)) attribute of GCC, which prevents the function from being considered for inlining. This might be needed, e.g., to countermand effects of link-time optimizations on stack-based variables.

After the call to the DEFUN macro, you must write the argument list for the C function, including the types for the arguments. If the primitive accepts a fixed maximum number of Lisp arguments, there must be one C argument for each Lisp argument, and each argument must be of type Lisp_Object. (Various macros and functions for creating values of type Lisp_Object are declared in the file lisp.h.) If the primitive is a special form, it must accept a Lisp list containing its unevaluated Lisp arguments as a single argument of type Lisp_Object. If the primitive has no upper limit on the number of evaluated Lisp arguments, it must have exactly two C arguments: the first is the number of Lisp arguments, and the second is the address of a block containing their values. These have types ptrdiff_t and Lisp_Object *, respectively. Since Lisp_Object can hold any Lisp object of any data type, you can determine the actual data type only at run time; so if you want a primitive to accept only a certain type of argument, you must check the type explicitly using a suitable predicate (see Type Predicates).

Within the function For itself, the local variable args refers to objects controlled by Emacs’s stack-marking garbage collector. Although the garbage collector does not reclaim objects reachable from C Lisp_Object stack variables, it may move some of the components of an object, such as the contents of a string or the text of a buffer. Therefore, functions that access these components must take care to refetch their addresses after performing Lisp evaluation. This means that instead of keeping C pointers to string contents or buffer text, the code should keep the buffer or string position, and recompute the C pointer from the position after performing Lisp evaluation. Lisp evaluation can occur via calls to eval_sub or Feval, either directly or indirectly.

Note the call to maybe_quit inside the loop: this function checks whether the user pressed C-g, and if so, aborts the processing. You should do that in any loop that can potentially require a large number of iterations; in this case, the list of arguments could be very long. This increases Emacs responsiveness and improves user experience.

You must not use C initializers for static or global variables unless the variables are never written once Emacs is dumped. These variables with initializers are allocated in an area of memory that becomes read-only (on certain operating systems) as a result of dumping Emacs. See Pure Storage.

Defining the C function is not enough to make a Lisp primitive available; you must also create the Lisp symbol for the primitive and store a suitable subr object in its function cell. The code looks like this:

defsubr (&sname);

Here sname is the name you used as the third argument to DEFUN.

If you add a new primitive to a file that already has Lisp primitives defined in it, find the function (near the end of the file) named syms_of_something, and add the call to defsubr there. If the file doesn’t have this function, or if you create a new file, add to it a syms_of_filename (e.g., syms_of_myfile). Then find the spot in emacs.c where all of these functions are called, and add a call to syms_of_filename there.

The function syms_of_filename is also the place to define any C variables that are to be visible as Lisp variables. DEFVAR_LISP makes a C variable of type Lisp_Object visible in Lisp. DEFVAR_INT makes a C variable of type int visible in Lisp with a value that is always an integer. DEFVAR_BOOL makes a C variable of type int visible in Lisp with a value that is either t or nil. Note that variables defined with DEFVAR_BOOL are automatically added to the list byte-boolean-vars used by the byte compiler.

These macros all expect three arguments:

lname

The name of the variable to be used by Lisp programs.

vname

The name of the variable in the C sources.

doc

The documentation for the variable, as a C comment. See Documentation Basics, for more details.

By convention, when defining variables of a “native” type (int and bool), the name of the C variable is the name of the Lisp variable with - replaced by _. When the variable has type Lisp_Object, the convention is to also prefix the C variable name with V. i.e.

DEFVAR_INT ("my-int-variable", my_int_variable,
           doc: /* An integer variable.  */);

DEFVAR_LISP ("my-lisp-variable", Vmy_lisp_variable,
           doc: /* A Lisp variable.  */);

There are situations in Lisp where you need to refer to the symbol itself rather than the value of that symbol. One such case is when temporarily overriding the value of a variable, which in Lisp is done with let. In C sources, this is done by defining a corresponding, constant symbol, and using specbind. By convention, Qmy_lisp_variable corresponds to Vmy_lisp_variable; to define it, use the DEFSYM macro. i.e.

DEFSYM (Qmy_lisp_variable, "my-lisp-variable");

To perform the actual binding:

specbind (Qmy_lisp_variable, Qt);

In Lisp symbols sometimes need to be quoted, to achieve the same effect in C you again use the corresponding constant symbol Qmy_lisp_variable. For example, when creating a buffer-local variable (see Buffer-Local Variables) in Lisp you would write:

(make-variable-buffer-local 'my-lisp-variable)

In C the corresponding code uses Fmake_variable_buffer_local in combination with DEFSYM, i.e.

DEFSYM (Qmy_lisp_variable, "my-lisp-variable");
Fmake_variable_buffer_local (Qmy_lisp_variable);

If you want to make a Lisp variable that is defined in C behave like one declared with defcustom, add an appropriate entry to cus-start.el. See Defining Customization Variables, for a description of the format to use.

If you directly define a file-scope C variable of type Lisp_Object, you must protect it from garbage-collection by calling staticpro in syms_of_filename, like this:

staticpro (&variable);

Here is another example function, with more complicated arguments. This comes from the code in window.c, and it demonstrates the use of macros and functions to manipulate Lisp objects.

DEFUN ("coordinates-in-window-p", Fcoordinates_in_window_p,
       Scoordinates_in_window_p, 2, 2, 0,
       doc: /* Return non-nil if COORDINATES are in WINDOW.
  …
  or `right-margin' is returned.  */)
  (register Lisp_Object coordinates, Lisp_Object window)
{
  struct window *w;
  struct frame *f;
  int x, y;
  Lisp_Object lx, ly;

  w = decode_live_window (window);
  f = XFRAME (w->frame);
  CHECK_CONS (coordinates);
  lx = Fcar (coordinates);
  ly = Fcdr (coordinates);
  CHECK_NUMBER (lx);
  CHECK_NUMBER (ly);
  x = FRAME_PIXEL_X_FROM_CANON_X (f, lx) + FRAME_INTERNAL_BORDER_WIDTH (f);
  y = FRAME_PIXEL_Y_FROM_CANON_Y (f, ly) + FRAME_INTERNAL_BORDER_WIDTH (f);

  switch (coordinates_in_window (w, x, y))
    {
    case ON_NOTHING:            /* NOT in window at all.  */
      return Qnil;
    case ON_MODE_LINE:          /* In mode line of window.  */
      return Qmode_line;
    case ON_SCROLL_BAR:         /* On scroll-bar of window.  */
      /* Historically we are supposed to return nil in this case.  */
      return Qnil;

    default:
      emacs_abort ();
    }
}

Note that C code cannot call functions by name unless they are defined in C. The way to call a function written in Lisp is to use Ffuncall, which embodies the Lisp function funcall. Since the Lisp function funcall accepts an unlimited number of arguments, in C it takes two: the number of Lisp-level arguments, and a one-dimensional array containing their values. The first Lisp-level argument is the Lisp function to call, and the rest are the arguments to pass to it.

The C functions call0, call1, call2, and so on, provide handy ways to call a Lisp function conveniently with a fixed number of arguments. They work by calling Ffuncall.

eval.c is a very good file to look through for examples; lisp.h contains the definitions for some important macros and functions.

If you define a function which is side-effect free or pure, give it a non-nil side-effect-free or pure property, respectively (see Standard Symbol Properties).


E.8 Writing Dynamically-Loaded Modules

This section describes the Emacs module API and how to use it as part of writing extension modules for Emacs. The module API is defined in the C programming language, therefore the description and the examples in this section assume the module is written in C. For other programming languages, you will need to use the appropriate bindings, interfaces and facilities for calling C code. Emacs C code requires a C99 or later compiler (see C Dialect), and so the code examples in this section also follow that standard.

Writing a module and integrating it into Emacs comprises the following tasks:

  • Writing initialization code for the module.
  • Writing one or more module functions.
  • Communicating values and objects between Emacs and your module functions.
  • Handling of error conditions and nonlocal exits.

The following subsections describe these tasks and the API itself in more detail.

Once your module is written, compile it to produce a shared library, according to the conventions of the underlying platform. Then place the shared library in a directory mentioned in load-path (see Library Search), where Emacs will find it.

If you wish to verify the conformance of a module to the Emacs dynamic module API, invoke Emacs with the --module-assertions option. See Initial Options in The GNU Emacs Manual.


E.8.1 Module Initialization Code

Begin your module by including the header file emacs-module.h and defining the GPL compatibility symbol:

#include <emacs-module.h>

int plugin_is_GPL_compatible;

The emacs-module.h file is installed into your system’s include tree as part of the Emacs installation. Alternatively, you can find it in the Emacs source tree.

Next, write an initialization function for the module.

Function: int emacs_module_init (struct emacs_runtime *runtime)

Emacs calls this function when it loads a module. If a module does not export a function named emacs_module_init, trying to load the module will signal an error. The initialization function should return zero if the initialization succeeds, non-zero otherwise. In the latter case, Emacs will signal an error, and the loading of the module will fail. If the user presses C-g during the initialization, Emacs ignores the return value of the initialization function and quits (see Quitting). (If needed, you can catch user quitting inside the initialization function, see should_quit.)

The argument runtime is a pointer to a C struct that includes 2 public fields: size, which provides the size of the structure in bytes; and get_environment, which provides a pointer to a function that allows the module initialization function access to the Emacs environment object and its interfaces.

The initialization function should perform whatever initialization is required for the module. In addition, it can perform the following tasks:

Compatibility verification

A module can verify that the Emacs executable which loads the module is compatible with the module, by comparing the size member of the runtime structure with the value compiled into the module:

int
emacs_module_init (struct emacs_runtime *runtime)
{
  if (runtime->size < sizeof (*runtime))
    return 1;
}

If the size of the runtime object passed to the module is smaller than what it expects, it means the module was compiled for an Emacs version newer (later) than the one which attempts to load it, i.e. the module might be incompatible with the Emacs binary.

In addition, a module can verify the compatibility of the module API with what the module expects. The following sample code assumes it is part of the emacs_module_init function shown above:

  emacs_env *env = runtime->get_environment (runtime);
  if (env->size < sizeof (*env))
    return 2;

This calls the get_environment function using the pointer provided in the runtime structure to retrieve a pointer to the API’s environment, a C struct which also has a size field holding the size of the structure in bytes.

Finally, you can write a module that will work with older versions of Emacs, by comparing the size of the environment passed by Emacs with known sizes, like this:

  emacs_env *env = runtime->get_environment (runtime);
  if (env->size >= sizeof (struct emacs_env_26))
    emacs_version = 26;  /* Emacs 26 or later.  */
  else if (env->size >= sizeof (struct emacs_env_25))
    emacs_version = 25;
  else
    return 2; /* Unknown or unsupported version.  */

This works because later Emacs versions always add members to the environment, never remove any members, so the size can only grow with new Emacs releases. Given the version of Emacs, the module can use only the parts of the module API that existed in that version, since those parts are identical in later versions.

emacs-module.h defines a preprocessor macro EMACS_MAJOR_VERSION. It expands to an integer literal which is the latest major version of Emacs supported by the header. See Version Information. Note that the value of EMACS_MAJOR_VERSION is a compile-time constant and does not represent the version of Emacs that is currently running and has loaded your module. If you want your module to be compatible with various versions of emacs-module.h as well as various versions of Emacs, you can use conditional compilation based on EMACS_MAJOR_VERSION.

We recommend that modules always perform the compatibility verification, unless they do their job entirely in the initialization function, and don’t access any Lisp objects or use any Emacs functions accessible through the environment structure.

Binding module functions to Lisp symbols

This gives the module functions names so that Lisp code could call it by that name. We describe how to do this in Writing Module Functions below.


E.8.2 Writing Module Functions

The main reason for writing an Emacs module is to make additional functions available to Lisp programs that load the module. This subsection describes how to write such module functions.

A module function has the following general form and signature:

Function: emacs_value emacs_function (emacs_env *env, ptrdiff_t nargs, emacs_value *args, void *data)

The env argument provides a pointer to the API environment, needed to access Emacs objects and functions. The nargs argument is the required number of arguments, which can be zero (see make_function below for more flexible specification of the argument number), and args is a pointer to the array of the function arguments. The argument data points to additional data required by the function, which was arranged when make_function (see below) was called to create an Emacs function from emacs_function.

Module functions use the type emacs_value to communicate Lisp objects between Emacs and the module (see Conversion Between Lisp and Module Values). The API, described below and in the following subsections, provides facilities for conversion between basic C data types and the corresponding emacs_value objects.

In the module function’s body, do not attempt to access elements of the args array beyond the index nargs-1: memory for the args array is allocated exactly to accommodate nargs values, and accessing beyond that will most probably crash your module. In particular, if the value of nargs passed to the function at run time is zero, it must not access args at all, as no memory will have been allocated for it in that case.

A module function always returns a value. If the function returns normally, the Lisp code which called it will see the Lisp object corresponding to the emacs_value value the function returned. However, if the user typed C-g, or if the module function or its callees signaled an error or exited nonlocally (see Nonlocal Exits in Modules), Emacs will ignore the returned value and quit or throw as it does when Lisp code encounters the same situations.

The header emacs-module.h provides the type emacs_function as an alias type for a function pointer to a module function.

After writing your C code for a module function, you should make a Lisp function object from it using the make_function function, whose pointer is provided in the environment (recall that the pointer to the environment is returned by get_environment). This is normally done in the module initialization function (see module initialization function), after verifying the API compatibility.

Function: emacs_value make_function (emacs_env *env, ptrdiff_t min_arity, ptrdiff_t max_arity, emacs_function func, const char *docstring, void *data)

This returns an Emacs function created from the C function func, whose signature is as described for emacs_function above. The arguments min_arity and max_arity specify the minimum and maximum number of arguments that func can accept. The max_arity argument can have the special value emacs_variadic_function, which makes the function accept an unlimited number of arguments, like the &rest keyword in Lisp (see Features of Argument Lists).

The argument data is a way to arrange for arbitrary additional data to be passed to func when it is called. Whatever pointer is passed to make_function will be passed unaltered to func.

The argument docstring specifies the documentation string for the function. It should be either an ASCII string, or a UTF-8 encoded non-ASCII string, or a NULL pointer; in the latter case the function will have no documentation. The documentation string can end with a line that specifies the advertised calling convention, see Documentation Strings of Functions.

Since every module function must accept the pointer to the environment as its first argument, the call to make_function could be made from any module function, but you will normally want to do that from the module initialization function, so that all the module functions are known to Emacs once the module is loaded.

Finally, you should bind the Lisp function to a symbol, so that Lisp code could call your function by name. For that, use the module API function intern (see intern) whose pointer is also provided in the environment that module functions can access.

Combining the above steps, code that arranges for a C function module_func to be callable as module-func from Lisp will look like this, as part of the module initialization function:

 emacs_env *env = runtime->get_environment (runtime);
 emacs_value func = env->make_function (env, min_arity, max_arity,
                                        module_func, docstring, data);
 emacs_value symbol = env->intern (env, "module-func");
 emacs_value args[] = {symbol, func};
 env->funcall (env, env->intern (env, "defalias"), 2, args);

This makes the symbol module-func known to Emacs by calling env->intern, then invokes defalias from Emacs to bind the function to that symbol. Note that it is possible to use fset instead of defalias; the differences are described in defalias.

Module functions including the emacs_module_init function (see module initialization function) may only interact with Emacs by calling environment functions from some live emacs_env pointer while being called directly or indirectly from Emacs. In other words, if a module function wants to call Lisp functions or Emacs primitives, convert emacs_value objects to and from C datatypes (see Conversion Between Lisp and Module Values), or interact with Emacs in any other way, some call from Emacs to emacs_module_init or to a module function must be in the call stack. Module functions may not interact with Emacs while garbage collection is running; see Garbage Collection. They may only interact with Emacs from Lisp interpreter threads (including the main thread) created by Emacs; see Threads. The --module-assertions command-line option can detect some violations of the above requirements. See Initial Options in The GNU Emacs Manual.

Using the module API, it is possible to define more complex function and data types: inline functions, macros, etc. However, the resulting C code will be cumbersome and hard to read. Therefore, we recommend that you limit the module code which creates functions and data structures to the absolute minimum, and leave the rest for a Lisp package that will accompany your module, because doing these additional tasks in Lisp is much easier, and will produce a much more readable code. For example, given a module function module-func defined as above, one way of making a macro module-macro based on it is with the following simple Lisp wrapper:

(defmacro module-macro (&rest args)
  "Documentation string for the macro."
  (module-func args))

The Lisp package which goes with your module could then load the module using the load primitive (see Emacs Dynamic Modules) when the package is loaded into Emacs.

By default, module functions created by make_function are not interactive. To make them interactive, you can use the following function.

Function: void make_interactive (emacs_env *env, emacs_value function, emacs_value spec)

This function, which is available since Emacs 28, makes the function function interactive using the interactive specification spec. Emacs interprets spec like the argument to the interactive form. Using interactive, and see Code Characters for interactive. function must be an Emacs module function returned by make_function.

Note that there is no native module support for retrieving the interactive specification of a module function. Use the function interactive-form for that. Using interactive. It is not possible to make a module function non-interactive once you have made it interactive using make_interactive.

If you want to run some code when a module function object (i.e., an object returned by make_function) is garbage-collected, you can install a function finalizer. Function finalizers are available since Emacs 28. For example, if you have passed some heap-allocated structure to the data argument of make_function, you can use the finalizer to deallocate the structure. See (libc)Basic Allocation, and see (libc)Freeing after Malloc. The finalizer function has the following signature:

void finalizer (void *data)

Here, data receives the value passed to data when calling make_function. Note that the finalizer can’t interact with Emacs in any way.

Directly after calling make_function, the newly-created function doesn’t have a finalizer. Use set_function_finalizer to add one, if desired.

Function: void emacs_finalizer (void *ptr)

The header emacs-module.h provides the type emacs_finalizer as a type alias for an Emacs finalizer function.

Function: emacs_finalizer get_function_finalizer (emacs_env *env, emacs_value arg)

This function, which is available since Emacs 28, returns the function finalizer associated with the module function represented by arg. arg must refer to a module function, that is, an object returned by make_function. If no finalizer is associated with the function, NULL is returned.

Function: void set_function_finalizer (emacs_env *env, emacs_value arg, emacs_finalizer fin)

This function, which is available since Emacs 28, sets the function finalizer associated with the module function represented by arg to fin. arg must refer to a module function, that is, an object returned by make_function. fin can either be NULL to clear arg’s function finalizer, or a pointer to a function to be called when the object represented by arg is garbage-collected. At most one function finalizer can be set per function; if arg already has a finalizer, it is replaced by fin.


E.8.3 Conversion Between Lisp and Module Values

With very few exceptions, most modules need to exchange data with Lisp programs that call them: accept arguments to module functions and return values from module functions. For this purpose, the module API provides the emacs_value type, which represents Emacs Lisp objects communicated via the API; it is the functional equivalent of the Lisp_Object type used in Emacs C primitives (see Writing Emacs Primitives). This section describes the parts of the module API that allow to create emacs_value objects corresponding to basic Lisp data types, and how to access from C data in emacs_value objects that correspond to Lisp objects.

All of the functions described below are actually function pointers provided via the pointer to the environment which every module function accepts. Therefore, module code should call these functions through the environment pointer, like this:

emacs_env *env;  /* the environment pointer */
env->some_function (arguments…);

The emacs_env pointer will usually come from the first argument to the module function, or from the call to get_environment if you need the environment in the module initialization function.

Most of the functions described below became available in Emacs 25, the first Emacs release that supported dynamic modules. For the few functions that became available in later Emacs releases, we mention the first Emacs version that supported them.

The following API functions extract values of various C data types from emacs_value objects. They all raise the wrong-type-argument error condition (see Type Predicates) if the argument emacs_value object is not of the type expected by the function. See Nonlocal Exits in Modules, for details of how signaling errors works in Emacs modules, and how to catch error conditions inside the module before they are reported to Emacs. The API function type_of (see type_of) can be used to obtain the type of a emacs_value object.

Function: intmax_t extract_integer (emacs_env *env, emacs_value arg)

This function returns the value of a Lisp integer specified by arg. The C data type of the return value, intmax_t, is the widest integer data type supported by the C compiler, typically long long. If the value of arg doesn’t fit into an intmax_t, the function signals an error using the error symbol overflow-error.

Function: bool extract_big_integer (emacs_env *env, emacs_value arg, int *sign, ptrdiff_t *count, emacs_limb_t *magnitude)

This function, which is available since Emacs 27, extracts the integer value of arg. The value of arg must be an integer (fixnum or bignum). If sign is not NULL, it stores the sign of arg (-1, 0, or +1) into *sign. The magnitude is stored into magnitude as follows. If count and magnitude are both non-NULL, then magnitude must point to an array of at least *count unsigned long elements. If magnitude is large enough to hold the magnitude of arg, then this function writes the magnitude into the magnitude array in little-endian form, stores the number of array elements written into *count, and returns true. If magnitude is not large enough, it stores the required array size into *count, signals an error, and returns false. If count is not NULL and magnitude is NULL, then the function stores the required array size into *count and returns true.

Emacs guarantees that the maximum required value of *count never exceeds min (PTRDIFF_MAX, SIZE_MAX) / sizeof (emacs_limb_t), so you can use malloc (*count * sizeof *magnitude) to allocate the magnitude array without worrying about integer overflow in the size calculation.

Type alias: emacs_limb_t

This is an unsigned integer type, used as the element type for the magnitude arrays for the big integer conversion functions. The type is guaranteed to have unique object representations, i.e., no padding bits.

Macro: EMACS_LIMB_MAX

This macro expands to a constant expression specifying the maximum possible value for an emacs_limb_t object. The expression is suitable for use in #if.

Function: double extract_float (emacs_env *env, emacs_value arg)

This function returns the value of a Lisp float specified by arg, as a C double value.

Function: struct timespec extract_time (emacs_env *env, emacs_value arg)

This function, which is available since Emacs 27, interprets arg as an Emacs Lisp time value and returns the corresponding struct timespec. See Time of Day. struct timespec represents a timestamp with nanosecond precision. It has the following members:

time_t tv_sec

Whole number of seconds.

long tv_nsec

Fractional seconds as a number of nanoseconds. For timestamps returned by extract_time, this is always nonnegative and less than one billion. (Although POSIX requires the type of tv_nsec to be long, the type is long long on some nonstandard platforms.)

See (libc)Elapsed Time.

If time has higher precision than nanoseconds, then this function truncates it to nanosecond precision towards negative infinity. This function signals an error if time (truncated to nanoseconds) cannot be represented by struct timespec. For example, if time_t is a 32-bit integer type, then a time value of ten billion seconds would signal an error, but a time value of 600 picoseconds would get truncated to zero.

If you need to deal with time values that are not representable by struct timespec, or if you want higher precision, call the Lisp function encode-time and work with its return value. See Time Conversion.

Function: bool copy_string_contents (emacs_env *env, emacs_value arg, char *buf, ptrdiff_t *len)

This function stores the UTF-8 encoded text of a Lisp string specified by arg in the array of char pointed by buf, which should have enough space to hold at least *len bytes, including the terminating null byte. The argument len must not be a NULL pointer, and, when the function is called, it should point to a value that specifies the size of buf in bytes.

If the buffer size specified by *len is large enough to hold the string’s text, the function stores in *len the actual number of bytes copied to buf, including the terminating null byte, and returns true. If the buffer is too small, the function raises the args-out-of-range error condition, stores the required number of bytes in *len, and returns false. See Nonlocal Exits in Modules, for how to handle pending error conditions.

The argument buf can be a NULL pointer, in which case the function stores in *len the number of bytes required for storing the contents of arg, and returns true. This is how you can determine the size of buf needed to store a particular string: first call copy_string_contents with NULL as buf, then allocate enough memory to hold the number of bytes stored by the function in *len, and call the function again with non-NULL buf to actually perform the text copying.

Function: emacs_value vec_get (emacs_env *env, emacs_value vector, ptrdiff_t index)

This function returns the element of vector at index. The index of the first vector element is zero. The function raises the args-out-of-range error condition if the value of index is invalid. To extract C data from the value the function returns, use the other extraction functions described here, as appropriate for the Lisp data type stored in that element of the vector.

Function: ptrdiff_t vec_size (emacs_env *env, emacs_value vector)

This function returns the number of elements in vector.

Function: void vec_set (emacs_env *env, emacs_value vector, ptrdiff_t index, emacs_value value)

This function stores value in the element of vector whose index is index. It raises the args-out-of-range error condition if the value of index is invalid.

The following API functions create emacs_value objects from basic C data types. They all return the created emacs_value object.

Function: emacs_value make_integer (emacs_env *env, intmax_t n)

This function takes an integer argument n and returns the corresponding emacs_value object. It returns either a fixnum or a bignum depending on whether the value of n is inside the limits set by most-negative-fixnum and most-positive-fixnum (see Integer Basics).

Function: emacs_value make_big_integer (emacs_env *env, int sign, ptrdiff_t count, const emacs_limb_t *magnitude)

This function, which is available since Emacs 27, takes an arbitrary-sized integer argument and returns a corresponding emacs_value object. The sign argument gives the sign of the return value. If sign is nonzero, then magnitude must point to an array of at least count elements specifying the little-endian magnitude of the return value.

The following example uses the GNU Multiprecision Library (GMP) to calculate the next probable prime after a given integer. See (gmp)Top, for a general overview of GMP, and see (gmp)Integer Import and Export for how to convert the magnitude array to and from GMP mpz_t values.

#include <emacs-module.h>
int plugin_is_GPL_compatible;

#include <assert.h>
#include <limits.h>
#include <stdint.h>
#include <stdlib.h>
#include <string.h>

#include <gmp.h>

static void
memory_full (emacs_env *env)
{
  static const char message[] = "Memory exhausted";
  emacs_value data = env->make_string (env, message,
                                       strlen (message));
  env->non_local_exit_signal
    (env, env->intern (env, "error"),
     env->funcall (env, env->intern (env, "list"), 1, &data));
}

enum
{
  order = -1, endian = 0, nails = 0,
  limb_size = sizeof (emacs_limb_t),
  max_nlimbs = ((SIZE_MAX < PTRDIFF_MAX ? SIZE_MAX : PTRDIFF_MAX)
                / limb_size)
};

static bool
extract_big_integer (emacs_env *env, emacs_value arg, mpz_t result)
{
  ptrdiff_t nlimbs;
  bool ok = env->extract_big_integer (env, arg, NULL, &nlimbs, NULL);
  if (!ok)
    return false;
  assert (0 < nlimbs && nlimbs <= max_nlimbs);
  emacs_limb_t *magnitude = malloc (nlimbs * limb_size);
  if (magnitude == NULL)
    {
      memory_full (env);
      return false;
    }
  int sign;
  ok = env->extract_big_integer (env, arg, &sign, &nlimbs, magnitude);
  assert (ok);
  mpz_import (result, nlimbs, order, limb_size, endian, nails, magnitude);
  free (magnitude);
  if (sign < 0)
    mpz_neg (result, result);
  return true;
}

static emacs_value
make_big_integer (emacs_env *env, const mpz_t value)
{
  size_t nbits = mpz_sizeinbase (value, 2);
  int bitsperlimb = CHAR_BIT * limb_size - nails;
  size_t nlimbs = nbits / bitsperlimb + (nbits % bitsperlimb != 0);
  emacs_limb_t *magnitude
    = nlimbs <= max_nlimbs ? malloc (nlimbs * limb_size) : NULL;
  if (magnitude == NULL)
    {
      memory_full (env);
      return NULL;
    }
  size_t written;
  mpz_export (magnitude, &written, order, limb_size, endian, nails, value);
  assert (written == nlimbs);
  assert (nlimbs <= PTRDIFF_MAX);
  emacs_value result = env->make_big_integer (env, mpz_sgn (value),
                                              nlimbs, magnitude);
  free (magnitude);
  return result;
}

static emacs_value
next_prime (emacs_env *env, ptrdiff_t nargs, emacs_value *args,
            void *data)
{
  assert (nargs == 1);
  mpz_t p;
  mpz_init (p);
  extract_big_integer (env, args[0], p);
  mpz_nextprime (p, p);
  emacs_value result = make_big_integer (env, p);
  mpz_clear (p);
  return result;
}

int
emacs_module_init (struct emacs_runtime *runtime)
{
  emacs_env *env = runtime->get_environment (runtime);
  emacs_value symbol = env->intern (env, "next-prime");
  emacs_value func
    = env->make_function (env, 1, 1, next_prime, NULL, NULL);
  emacs_value args[] = {symbol, func};
  env->funcall (env, env->intern (env, "defalias"), 2, args);
  return 0;
}
Function: emacs_value make_float (emacs_env *env, double d)

This function takes a double argument d and returns the corresponding Emacs floating-point value.

Function: emacs_value make_time (emacs_env *env, struct timespec time)

This function, which is available since Emacs 27, takes a struct timespec argument time and returns the corresponding Emacs timestamp as a pair (ticks . hz). See Time of Day. The return value represents exactly the same timestamp as time: all input values are representable, and there is never a loss of precision. time.tv_sec and time.tv_nsec can be arbitrary values. In particular, there’s no requirement that time be normalized. This means that time.tv_nsec can be negative or larger than 999,999,999.

Function: emacs_value make_string (emacs_env *env, const char *str, ptrdiff_t len)

This function creates an Emacs string from C text string pointed by str whose length in bytes, not including the terminating null byte, is len. The original string in str can be either an ASCII string or a UTF-8 encoded non-ASCII string; it can include embedded null bytes, and doesn’t have to end in a terminating null byte at str[len]. The function raises the overflow-error error condition if len is negative or exceeds the maximum length of an Emacs string. If len is zero, then str can be NULL, otherwise it must point to valid memory. For nonzero len, make_string returns unique mutable string objects.

Function: emacs_value make_unibyte_string (emacs_env *env, const char *str, ptrdiff_t len)

This function is like make_string, but has no restrictions on the values of the bytes in the C string, and can be used to pass binary data to Emacs in the form of a unibyte string.

The API does not provide functions to manipulate Lisp data structures, for example, create lists with cons and list (see Building Cons Cells and Lists), extract list members with car and cdr (see Accessing Elements of Lists), create vectors with vector (see Functions for Vectors), etc. For these, use intern and funcall, described in the next subsection, to call the corresponding Lisp functions.

Normally, emacs_value objects have a rather short lifetime: it ends when the emacs_env pointer used for their creation goes out of scope. Occasionally, you may need to create global references: emacs_value objects that live as long as you wish. Use the following two functions to manage such objects.

Function: emacs_value make_global_ref (emacs_env *env, emacs_value value)

This function returns a global reference for value.

Function: void free_global_ref (emacs_env *env, emacs_value global_value)

This function frees the global_value previously created by make_global_ref. The global_value is no longer valid after the call. Your module code should pair each call to make_global_ref with the corresponding free_global_ref.

An alternative to keeping around C data structures that need to be passed to module functions later is to create user pointer objects. A user pointer, or user-ptr, object is a Lisp object that encapsulates a C pointer and can have an associated finalizer function, which is called when the object is garbage-collected (see Garbage Collection). The module API provides functions to create and access user-ptr objects. These functions raise the wrong-type-argument error condition if they are called on emacs_value that doesn’t represent a user-ptr object.

Function: emacs_value make_user_ptr (emacs_env *env, emacs_finalizer fin, void *ptr)

This function creates and returns a user-ptr object which wraps the C pointer ptr. The finalizer function fin can be a NULL pointer (meaning no finalizer), or it can be a function of the following signature:

typedef void (*emacs_finalizer) (void *ptr);

If fin is not a NULL pointer, it will be called with the ptr as the argument when the user-ptr object is garbage-collected. Don’t run any expensive code in a finalizer, because GC must finish quickly to keep Emacs responsive.

Function: void * get_user_ptr (emacs_env *env, emacs_value arg)

This function extracts the C pointer from the Lisp object represented by arg.

Function: void set_user_ptr (emacs_env *env, emacs_value arg, void *ptr)

This function sets the C pointer embedded in the user-ptr object represented by arg to ptr.

Function: emacs_finalizer get_user_finalizer (emacs_env *env, emacs_value arg)

This function returns the finalizer of the user-ptr object represented by arg, or NULL if it doesn’t have a finalizer.

Function: void set_user_finalizer (emacs_env *env, emacs_value arg, emacs_finalizer fin)

This function changes the finalizer of the user-ptr object represented by arg to be fin. If fin is a NULL pointer, the user-ptr object will have no finalizer.

Note that the emacs_finalizer type works for both user pointer an module function finalizers. See Module Function Finalizers.


E.8.4 Miscellaneous Convenience Functions for Modules

This subsection describes a few convenience functions provided by the module API. Like the functions described in previous subsections, all of them are actually function pointers, and need to be called via the emacs_env pointer. Description of functions that were introduced after Emacs 25 calls out the first version where they became available.

Function: bool eq (emacs_env *env, emacs_value a, emacs_value b)

This function returns true if the Lisp objects represented by a and b are identical, false otherwise. This is the same as the Lisp function eq (see Equality Predicates), but avoids the need to intern the objects represented by the arguments.

There are no API functions for other equality predicates, so you will need to use intern and funcall, described below, to perform more complex equality tests.

Function: bool is_not_nil (emacs_env *env, emacs_value arg)

This function tests whether the Lisp object represented by arg is non-nil; it returns true or false accordingly.

Note that you could implement an equivalent test by using intern to get an emacs_value representing nil, then use eq, described above, to test for equality. But using this function is more convenient.

Function: emacs_value type_of (emacs_env *env, emacs_value arg)

This function returns the type of arg as a value that represents a symbol: string for a string, integer for an integer, process for a process, etc. See Type Predicates. You can use intern and eq to compare against known type symbols, if your code needs to depend on the object type.

Function: emacs_value intern (emacs_env *env, const char *name)

This function returns an interned Emacs symbol whose name is name, which should be an ASCII null-terminated string. It creates a new symbol if one does not already exist.

Together with funcall, described below, this function provides a means for invoking any Lisp-callable Emacs function, provided that its name is a pure ASCII string. For example, here’s how to intern a symbol whose name name_str is non-ASCII, by calling the more powerful Emacs intern function (see Creating and Interning Symbols):

emacs_value fintern = env->intern (env, "intern");
emacs_value sym_name =
  env->make_string (env, name_str, strlen (name_str));
emacs_value symbol = env->funcall (env, fintern, 1, &sym_name);
Function: emacs_value funcall (emacs_env *env, emacs_value func, ptrdiff_t nargs, emacs_value *args)

This function calls the specified func passing it nargs arguments from the array pointed to by args. The argument func can be a function symbol (e.g., returned by intern described above), a module function returned by make_function (see Writing Module Functions), a subroutine written in C, etc. If nargs is zero, args can be a NULL pointer.

The function returns the value that func returned.

If your module includes potentially long-running code, it is a good idea to check from time to time in that code whether the user wants to quit, e.g., by typing C-g (see Quitting). The following function, which is available since Emacs 26.1, is provided for that purpose.

Function: bool should_quit (emacs_env *env)

This function returns true if the user wants to quit. In that case, we recommend that your module function aborts any on-going processing and returns as soon as possible. In most cases, use process_input instead.

To process input events in addition to checking whether the user wants to quit, use the following function, which is available since Emacs 27.1.

Function: enum emacs_process_input_result process_input (emacs_env *env)

This function processes pending input events. It returns emacs_process_input_quit if the user wants to quit or an error occurred while processing signals. In that case, we recommend that your module function aborts any on-going processing and returns as soon as possible. If the module code may continue running, process_input returns emacs_process_input_continue. The return value is emacs_process_input_continue if and only if there is no pending nonlocal exit in env. If the module continues after calling process_input, global state such as variable values and buffer content may have been modified in arbitrary ways.

Function: int open_channel (emacs_env *env, emacs_value pipe_process)

This function, which is available since Emacs 28, opens a channel to an existing pipe process. pipe_process must refer to an existing pipe process created by make-pipe-process. Pipe Processes. If successful, the return value will be a new file descriptor that you can use to write to the pipe. Unlike all other module functions, you can use the returned file descriptor from arbitrary threads, even if no module environment is active. You can use the write function to write to the file descriptor. Once done, close the file descriptor using close. (libc)Low-Level I/O.


E.8.5 Nonlocal Exits in Modules

Emacs Lisp supports nonlocal exits, whereby program control is transferred from one point in a program to another remote point. See Nonlocal Exits. Thus, Lisp functions called by your module might exit nonlocally by calling signal or throw, and your module functions must handle such nonlocal exits properly. Such handling is needed because C programs will not automatically release resources and perform other cleanups in these cases; your module code must itself do it. The module API provides facilities for that, described in this subsection. They are generally available since Emacs 25; those of them that became available in later releases explicitly call out the first Emacs version where they became part of the API.

When some Lisp code called by a module function signals an error or throws, the nonlocal exit is trapped, and the pending exit and its associated data are stored in the environment. Whenever a nonlocal exit is pending in the environment, any module API function called with a pointer to that environment will return immediately without any processing (the functions non_local_exit_check, non_local_exit_get, and non_local_exit_clear are exceptions from this rule). If your module function then does nothing and returns to Emacs, a pending nonlocal exit will cause Emacs to act on it: signal an error or throw to the corresponding catch.

So the simplest “handling” of nonlocal exits in module functions is to do nothing special and let the rest of your code to run as if nothing happened. However, this can cause two classes of problems:

  • Your module function might use uninitialized or undefined values, since API functions return immediately without producing the expected results.
  • Your module might leak resources, because it might not have the opportunity to release them.

Therefore, we recommend that your module functions check for nonlocal exit conditions and recover from them, using the functions described below.

Function: enum emacs_funcall_exit non_local_exit_check (emacs_env *env)

This function returns the kind of nonlocal exit condition stored in env. The possible values are:

emacs_funcall_exit_return

The last API function exited normally.

emacs_funcall_exit_signal

The last API function signaled an error.

emacs_funcall_exit_throw

The last API function exited via throw.

Function: enum emacs_funcall_exit non_local_exit_get (emacs_env *env, emacs_value *symbol, emacs_value *data)

This function returns the kind of nonlocal exit condition stored in env, like non_local_exit_check does, but it also returns the full information about the nonlocal exit, if any. If the return value is emacs_funcall_exit_signal, the function stores the error symbol in *symbol and the error data in *data (see How to Signal an Error). If the return value is emacs_funcall_exit_throw, the function stores the catch tag symbol in *symbol and the throw value in *data. The function doesn’t store anything in memory pointed by these arguments when the return value is emacs_funcall_exit_return.

You should check nonlocal exit conditions where it matters: before you allocated some resource or after you allocated a resource that might need freeing, or where a failure means further processing is impossible or infeasible.

Once your module function detected that a nonlocal exit is pending, it can either return to Emacs (after performing the necessary local cleanup), or it can attempt to recover from the nonlocal exit. The following API functions will help with these tasks.

Function: void non_local_exit_clear (emacs_env *env)

This function clears the pending nonlocal exit conditions and data from env. After calling it, the module API functions will work normally. Use this function if your module function can recover from nonlocal exits of the Lisp functions it calls and continue, and also before calling any of the following two functions (or any other API functions, if you want them to perform their intended processing when a nonlocal exit is pending).

Function: void non_local_exit_throw (emacs_env *env, emacs_value tag, emacs_value value)

This function throws to the Lisp catch symbol represented by tag, passing it value as the value to return. Your module function should in general return soon after calling this function. One use of this function is when you want to re-throw a non-local exit from one of the called API or Lisp functions.

Function: void non_local_exit_signal (emacs_env *env, emacs_value symbol, emacs_value data)

This function signals the error represented by the error symbol symbol with the specified error data data. The module function should return soon after calling this function. This function could be useful, e.g., for signaling errors from module functions to Emacs.


E.9 Object Internals

Emacs Lisp provides a rich set of the data types. Some of them, like cons cells, integers and strings, are common to nearly all Lisp dialects. Some others, like markers and buffers, are quite special and needed to provide the basic support to write editor commands in Lisp. To implement such a variety of object types and provide an efficient way to pass objects between the subsystems of an interpreter, there is a set of C data structures and a special type to represent the pointers to all of them, which is known as tagged pointer.

In C, the tagged pointer is an object of type Lisp_Object. Any initialized variable of such a type always holds the value of one of the following basic data types: integer, symbol, string, cons cell, float, or vectorlike object. Each of these data types has the corresponding tag value. All tags are enumerated by enum Lisp_Type and placed into a 3-bit bitfield of the Lisp_Object. The rest of the bits is the value itself. Integers are immediate, i.e., directly represented by those value bits, and all other objects are represented by the C pointers to a corresponding object allocated from the heap. Width of the Lisp_Object is platform- and configuration-dependent: usually it’s equal to the width of an underlying platform pointer (i.e., 32-bit on a 32-bit machine and 64-bit on a 64-bit one), but also there is a special configuration where Lisp_Object is 64-bit but all pointers are 32-bit. The latter trick was designed to overcome the limited range of values for Lisp integers on a 32-bit system by using 64-bit long long type for Lisp_Object.

The following C data structures are defined in lisp.h to represent the basic data types beyond integers:

struct Lisp_Cons

Cons cell, an object used to construct lists.

struct Lisp_String

String, the basic object to represent a sequence of characters.

struct Lisp_Vector

Array, a fixed-size set of Lisp objects which may be accessed by an index.

struct Lisp_Symbol

Symbol, the unique-named entity commonly used as an identifier.

struct Lisp_Float

Floating-point value.

These types are the first-class citizens of an internal type system. Since the tag space is limited, all other types are the subtypes of Lisp_Vectorlike. Vector subtypes are enumerated by enum pvec_type, and nearly all complex objects like windows, buffers, frames, and processes fall into this category.

Below there is a description of a few subtypes of Lisp_Vectorlike. Buffer object represents the text to display and edit. Window is the part of display structure which shows the buffer or is used as a container to recursively place other windows on the same frame. (Do not confuse Emacs Lisp window object with the window as an entity managed by the user interface system like X; in Emacs terminology, the latter is called frame.) Finally, process object is used to manage the subprocesses.


E.9.1 Buffer Internals

Two structures (see buffer.h) are used to represent buffers in C. The buffer_text structure contains fields describing the text of a buffer; the buffer structure holds other fields. In the case of indirect buffers, two or more buffer structures reference the same buffer_text structure.

Here are some of the fields in struct buffer_text:

beg

The address of the buffer contents. The buffer contents is a linear C array of char, with the gap somewhere in its midst.

gpt
gpt_byte

The character and byte positions of the buffer gap. See The Buffer Gap.

z
z_byte

The character and byte positions of the end of the buffer text.

gap_size

The size of buffer’s gap. See The Buffer Gap.

modiff
save_modiff
chars_modiff
overlay_modiff

These fields count the number of buffer-modification events performed in this buffer. modiff is incremented after each buffer-modification event, and is never otherwise changed; save_modiff contains the value of modiff the last time the buffer was visited or saved; chars_modiff counts only modifications to the characters in the buffer, ignoring all other kinds of changes (such as text properties); and overlay_modiff counts only modifications to the buffer’s overlays.

beg_unchanged
end_unchanged

The number of characters at the start and end of the text that are known to be unchanged since the last complete redisplay.

unchanged_modified
overlay_unchanged_modified

The values of modiff and overlay_modiff, respectively, after the last complete redisplay. If their current values match modiff or overlay_modiff, that means beg_unchanged and end_unchanged contain no useful information.

markers

The markers that refer to this buffer. This is actually a single marker, and successive elements in its marker chain (a linked list) are the other markers referring to this buffer text.

intervals

The interval tree which records the text properties of this buffer.

Some of the fields of struct buffer are:

header

A header of type union vectorlike_header is common to all vectorlike objects.

own_text

A struct buffer_text structure that ordinarily holds the buffer contents. In indirect buffers, this field is not used.

text

A pointer to the buffer_text structure for this buffer. In an ordinary buffer, this is the own_text field above. In an indirect buffer, this is the own_text field of the base buffer.

next

A pointer to the next buffer, in the chain of all buffers, including killed buffers. This chain is used only for allocation and garbage collection, in order to collect killed buffers properly.

pt
pt_byte

The character and byte positions of point in a buffer.

begv
begv_byte

The character and byte positions of the beginning of the accessible range of text in the buffer.

zv
zv_byte

The character and byte positions of the end of the accessible range of text in the buffer.

base_buffer

In an indirect buffer, this points to the base buffer. In an ordinary buffer, it is null.

local_flags

This field contains flags indicating that certain variables are local in this buffer. Such variables are declared in the C code using DEFVAR_PER_BUFFER, and their buffer-local bindings are stored in fields in the buffer structure itself. (Some of these fields are described in this table.)

modtime

The modification time of the visited file. It is set when the file is written or read. Before writing the buffer into a file, this field is compared to the modification time of the file to see if the file has changed on disk. See Buffer Modification.

auto_save_modified

The time when the buffer was last auto-saved.

last_window_start

The window-start position in the buffer as of the last time the buffer was displayed in a window.

clip_changed

This flag indicates that narrowing has changed in the buffer. See Narrowing.

prevent_redisplay_optimizations_p

This flag indicates that redisplay optimizations should not be used to display this buffer.

inhibit_buffer_hooks

This flag indicates that the buffer should not run the hooks kill-buffer-hook, kill-buffer-query-functions (see Killing Buffers), and buffer-list-update-hook (see The Buffer List). It is set at buffer creation (see Creating Buffers), and avoids slowing down internal or temporary buffers, such as those created by with-temp-buffer (see Current Buffer).

name

A Lisp string that names the buffer. It is guaranteed to be unique. See Buffer Names. This and the following fields have their names in the C struct definition end in a _ to indicate that they should not be accessed directly, but via the BVAR macro, like this:

  Lisp_Object buf_name = BVAR (buffer, name);
save_length

The length of the file this buffer is visiting, when last read or saved. It can have 2 special values: −1 means auto-saving was turned off in this buffer, and −2 means don’t turn off auto-saving if buffer text shrinks a lot. This and other fields concerned with saving are not kept in the buffer_text structure because indirect buffers are never saved.

directory

The directory for expanding relative file names. This is the value of the buffer-local variable default-directory (see Functions that Expand Filenames).

filename

The name of the file visited in this buffer, or nil. This is the value of the buffer-local variable buffer-file-name (see Buffer File Name).

undo_list
backed_up
auto_save_file_name
auto_save_file_format
read_only
file_format
file_truename
invisibility_spec
display_count
display_time

These fields store the values of Lisp variables that are automatically buffer-local (see Buffer-Local Variables), whose corresponding variable names have the additional prefix buffer- and have underscores replaced with dashes. For instance, undo_list stores the value of buffer-undo-list.

mark

The mark for the buffer. The mark is a marker, hence it is also included on the list markers. See The Mark.

local_var_alist

The association list describing the buffer-local variable bindings of this buffer, not including the built-in buffer-local bindings that have special slots in the buffer object. (Those slots are omitted from this table.) See Buffer-Local Variables.

major_mode

Symbol naming the major mode of this buffer, e.g., lisp-mode.

mode_name

Pretty name of the major mode, e.g., "Lisp".

keymap
abbrev_table
syntax_table
category_table
display_table

These fields store the buffer’s local keymap (see Keymaps), abbrev table (see Abbrev Tables), syntax table (see Syntax Tables), category table (see Categories), and display table (see Display Tables).

downcase_table
upcase_table
case_canon_table

These fields store the conversion tables for converting text to lower case, upper case, and for canonicalizing text for case-fold search. See The Case Table.

minor_modes

An alist of the minor modes of this buffer.

pt_marker
begv_marker
zv_marker

These fields are only used in an indirect buffer, or in a buffer that is the base of an indirect buffer. Each holds a marker that records pt, begv, and zv respectively, for this buffer when the buffer is not current.

mode_line_format
header_line_format
case_fold_search
tab_width
fill_column
left_margin
auto_fill_function
truncate_lines
word_wrap
ctl_arrow
bidi_display_reordering
bidi_paragraph_direction
selective_display
selective_display_ellipses
overwrite_mode
abbrev_mode
mark_active
enable_multibyte_characters
buffer_file_coding_system
cache_long_line_scans
point_before_scroll
left_fringe_width
right_fringe_width
fringes_outside_margins
scroll_bar_width
indicate_empty_lines
indicate_buffer_boundaries
fringe_indicator_alist
fringe_cursor_alist
scroll_up_aggressively
scroll_down_aggressively
cursor_type
cursor_in_non_selected_windows

These fields store the values of Lisp variables that are automatically buffer-local (see Buffer-Local Variables), whose corresponding variable names have underscores replaced with dashes. For instance, mode_line_format stores the value of mode-line-format.

overlays

The interval tree containing this buffer’s overlays.

last_selected_window

This is the last window that was selected with this buffer in it, or nil if that window no longer displays this buffer.


E.9.2 Window Internals

The fields of a window (for a complete list, see the definition of struct window in window.h) include:

frame

The frame that this window is on, as a Lisp object.

mini

Non-zero if this window is a minibuffer window, a window showing the minibuffer or the echo area.

pseudo_window_p

Non-zero if this window is a pseudo window. A pseudo window is either a window used to display the menu bar or the tool bar (when Emacs uses toolkits that don’t display their own menu bar and tool bar) or the tab bar or a window showing a tooltip on a tooltip frame. Pseudo windows are in general not accessible from Lisp code.

parent

Internally, Emacs arranges windows in a tree; each group of siblings has a parent window whose area includes all the siblings. This field points to the window’s parent in that tree, as a Lisp object. For the root window of the tree and a minibuffer window this is always nil.

Parent windows do not display buffers, and play little role in display except to shape their child windows. Emacs Lisp programs cannot directly manipulate parent windows; they operate on the windows at the leaves of the tree, which actually display buffers.

contents

For a leaf window and windows showing a tooltip, this is the buffer, as a Lisp object, that the window is displaying. For an internal (“parent”) window, this is its first child window. For a pseudo window showing a menu or tool bar this is nil. It is also nil for a window that has been deleted.

next
prev

The next and previous sibling of this window as Lisp objects. next is nil if the window is the right-most or bottom-most in its group; prev is nil if it is the left-most or top-most in its group. Whether the sibling is left/right or up/down is determined by the horizontal field of the sibling’s parent: if it’s non-zero, the siblings are arranged horizontally.

As a special case, next of a frame’s root window points to the frame’s minibuffer window, provided this is not a minibuffer-only or minibuffer-less frame. On such frames prev of the minibuffer window points to that frame’s root window. In any other case, the root window’s next and the minibuffer window’s (if present) prev fields are nil.

left_col

The left-hand edge of the window, measured in columns, relative to the leftmost column (column 0) of the window’s native frame.

top_line

The top edge of the window, measured in lines, relative to the topmost line (line 0) of the window’s native frame.

pixel_left
pixel_top

The left-hand and top edges of this window, measured in pixels, relative to the top-left corner (0, 0) of the window’s native frame.

total_cols
total_lines

The total width and height of the window, measured in columns and lines respectively. The values include scroll bars and fringes, dividers and/or the separator line on the right of the window (if any).

pixel_width;
pixel_height;

The total width and height of the window measured in pixels.

start

A marker pointing to the position in the buffer that is the first character (in the logical order, see Bidirectional Display) displayed in the window.

pointm

This is the value of point in the current buffer when this window is selected; when it is not selected, it retains its previous value.

old_pointm

The value of pointm at the last redisplay time.

force_start

If this flag is non-nil, it says that the window has been scrolled explicitly by the Lisp program, and the value of the window’s start was set for redisplay to honor. This affects what the next redisplay does if point is off the screen: instead of scrolling the window to show the text around point, it moves point to a location that is on the screen.

optional_new_start

This is similar to force_start, but the next redisplay will only obey it if point stays visible.

start_at_line_beg

Non-nil means current value of start was the beginning of a line when it was chosen.

use_time

This is the last time that the window was selected. The function get-lru-window uses this field.

sequence_number

A unique number assigned to this window when it was created.

last_modified

The modiff field of the window’s buffer, as of the last time a redisplay completed in this window.

last_overlay_modified

The overlay_modiff field of the window’s buffer, as of the last time a redisplay completed in this window.

last_point

The buffer’s value of point, as of the last time a redisplay completed in this window.

last_had_star

A non-zero value means the window’s buffer was modified when the window was last updated.

vertical_scroll_bar_type
horizontal_scroll_bar_type

The types of this window’s vertical and horizontal scroll bars.

scroll_bar_width
scroll_bar_height

The width of this window’s vertical scroll bar and the height of this window’s horizontal scroll bar, in pixels.

left_margin_cols
right_margin_cols

The widths of the left and right margins in this window. A value of zero means no margin.

left_fringe_width
right_fringe_width

The pixel widths of the left and right fringes in this window. A value of −1 means use the values of the frame.

fringes_outside_margins

A non-zero value means the fringes outside the display margins; othersize they are between the margin and the text.

window_end_pos

This is computed as z minus the buffer position of the last glyph in the current matrix of the window. The value is only valid if window_end_valid is non-zero.

window_end_bytepos

The byte position corresponding to window_end_pos.

window_end_vpos

The window-relative vertical position of the line containing window_end_pos.

window_end_valid

This field is set to a non-zero value if window_end_pos and window_end_vpos are truly valid. This is zero if nontrivial redisplay is pre-empted, since in that case the display that window_end_pos was computed for did not get onto the screen.

cursor

A structure describing where the cursor is in this window.

last_cursor_vpos

The window-relative vertical position of the line showing the cursor as of the last redisplay that finished.

phys_cursor

A structure describing where the cursor of this window physically is.

phys_cursor_type
phys_cursor_height
phys_cursor_width

The type, height, and width of the cursor that was last displayed on this window.

phys_cursor_on_p

This field is non-zero if the cursor is physically on.

cursor_off_p

Non-zero means the cursor in this window is logically off. This is used for blinking the cursor.

last_cursor_off_p

This field contains the value of cursor_off_p as of the time of the last redisplay.

must_be_updated_p

This is set to 1 during redisplay when this window must be updated.

hscroll

This is the number of columns that the display in the window is scrolled horizontally to the left. Normally, this is 0. When only the current line is hscrolled, this describes how much the current line is scrolled.

min_hscroll

Minimum value of hscroll, set by the user via set-window-hscroll (see Horizontal Scrolling). When only the current line is hscrolled, this describes the horizontal scrolling of lines other than the current one.

vscroll

Vertical scroll amount, in pixels. Normally, this is 0.

dedicated

Non-nil if this window is dedicated to its buffer.

combination_limit

This window’s combination limit, meaningful only for a parent window. If this is t, then it is not allowed to delete this window and recombine its child windows with other siblings of this window.

window_parameters

The alist of this window’s parameters.

display_table

The window’s display table, or nil if none is specified for it.

update_mode_line

Non-zero means this window’s mode line needs to be updated.

mode_line_height
header_line_height

The height in pixels of the mode line and the header line, or −1 if not known.

base_line_number

The line number of a certain position in the buffer, or zero. This is used for displaying the line number of point in the mode line.

base_line_pos

The position in the buffer for which the line number is known, or zero meaning none is known. If it is −1, don’t display the line number as long as the window shows that buffer.

column_number_displayed

The column number currently displayed in this window’s mode line, or −1 if column numbers are not being displayed.

current_matrix
desired_matrix

Glyph matrices describing the current and desired display of this window.


E.9.3 Process Internals

The fields of a process (for a complete list, see the definition of struct Lisp_Process in process.h) include:

name

A Lisp string, the name of the process.

command

A list containing the command arguments that were used to start this process. For a network or serial process, it is nil if the process is running or t if the process is stopped.

filter

A Lisp function used to accept output from the process.

sentinel

A Lisp function called whenever the state of the process changes.

buffer

The associated buffer of the process.

pid

An integer, the operating system’s process ID. Pseudo-processes such as network or serial connections use a value of 0.

childp

A flag, t if this is really a child process. For a network or serial connection, it is a plist based on the arguments to make-network-process or make-serial-process.

mark

A marker indicating the position of the end of the last output from this process inserted into the buffer. This is often but not always the end of the buffer.

kill_without_query

If this is non-zero, killing Emacs while this process is still running does not ask for confirmation about killing the process.

raw_status

The raw process status, as returned by the wait system call.

status

The process status, as process-status should return it. This is a Lisp symbol, a cons cell, or a list.

tick
update_tick

If these two fields are not equal, a change in the status of the process needs to be reported, either by running the sentinel or by inserting a message in the process buffer.

pty_flag

Non-zero if communication with the subprocess uses a pty; zero if it uses a pipe.

infd

The file descriptor for input from the process.

outfd

The file descriptor for output to the process.

tty_name

The name of the terminal that the subprocess is using, or nil if it is using pipes.

decode_coding_system

Coding-system for decoding the input from this process.

decoding_buf

A working buffer for decoding.

decoding_carryover

Size of carryover in decoding.

encode_coding_system

Coding-system for encoding the output to this process.

encoding_buf

A working buffer for encoding.

inherit_coding_system_flag

Flag to set coding-system of the process buffer from the coding system used to decode process output.

type

Symbol indicating the type of process: real, network, serial.


E.10 C Integer Types

Here are some guidelines for use of integer types in the Emacs C source code. These guidelines sometimes give competing advice; common sense is advised.

  • Avoid arbitrary limits. For example, avoid int len = strlen (s); unless the length of s is required for other reasons to fit in int range.
  • Do not assume that signed integer arithmetic wraps around on overflow. This is no longer true of Emacs porting targets: signed integer overflow has undefined behavior in practice, and can dump core or even cause earlier or later code to behave illogically. Unsigned overflow does wrap around reliably, modulo a power of two.
  • Prefer signed types to unsigned, as code gets confusing when signed and unsigned types are combined. Many other guidelines assume that types are signed; in the rarer cases where unsigned types are needed, similar advice may apply to the unsigned counterparts (e.g., size_t instead of ptrdiff_t, or uintptr_t instead of intptr_t).
  • Prefer int for Emacs character codes, in the range 0 .. 0x3FFFFF. More generally, prefer int for integers known to be in int range, e.g., screen column counts.
  • Prefer ptrdiff_t for sizes, i.e., for integers bounded by the maximum size of any individual C object or by the maximum number of elements in any C array. This is part of Emacs’s general preference for signed types. Using ptrdiff_t limits objects to PTRDIFF_MAX bytes, but larger objects would cause trouble anyway since they would break pointer subtraction, so this does not impose an arbitrary limit.
  • Avoid ssize_t except when communicating to low-level APIs that have ssize_t-related limitations. Although it’s equivalent to ptrdiff_t on typical platforms, ssize_t is occasionally narrower, so using it for size-related calculations could overflow. Also, ptrdiff_t is more ubiquitous and better-standardized, has standard printf formats, and is the basis for Emacs’s internal size-overflow checking. When using ssize_t, please note that POSIX requires support only for values in the range −1 .. SSIZE_MAX.
  • Normally, prefer intptr_t for internal representations of pointers, or for integers bounded only by the number of objects that can exist at any given time or by the total number of bytes that can be allocated. However, prefer uintptr_t to represent pointer arithmetic that could cross page boundaries. For example, on a machine with a 32-bit address space an array could cross the 0x7fffffff/0x80000000 boundary, which would cause an integer overflow when adding 1 to (intptr_t) 0x7fffffff.
  • Prefer the Emacs-defined type EMACS_INT for representing values converted to or from Emacs Lisp fixnums, as fixnum arithmetic is based on EMACS_INT.
  • When representing a system value (such as a file size or a count of seconds since the Epoch), prefer the corresponding system type (e.g., off_t, time_t). Do not assume that a system type is signed, unless this assumption is known to be safe. For example, although off_t is always signed, time_t need not be.
  • Prefer intmax_t for representing values that might be any signed integer value. A printf-family function can print such a value via a format like "%"PRIdMAX.
  • Prefer bool, false and true for booleans. Using bool can make programs easier to read and a bit faster than using int. Although it is also OK to use int, 0 and 1, this older style is gradually being phased out. When using bool, respect the limitations of the replacement implementation of bool. In particular, boolean bitfields should be of type bool_bf, not bool, so that they work correctly even when compiling Objective C with standard GCC.
  • In bitfields, prefer unsigned int or signed int to int, as int is less portable: it might be signed, and might not be. Single-bit bit fields should be unsigned int or bool_bf so that their values are 0 or 1.

Appendix F Standard Errors

Here is a list of the more important error symbols in standard Emacs, grouped by concept. The list includes each symbol’s message and a cross reference to a description of how the error can occur.

Each error symbol has a set of parent error conditions that is a list of symbols. Normally this list includes the error symbol itself and the symbol error. Occasionally it includes additional symbols, which are intermediate classifications, narrower than error but broader than a single error symbol. For example, all the errors in accessing files have the condition file-error. If we do not say here that a certain error symbol has additional error conditions, that means it has none.

As a special exception, the error symbols quit and minibuffer-quit don’t have the condition error, because quitting is not considered an error.

Most of these error symbols are defined in C (mainly data.c), but some are defined in Lisp. For example, the file userlock.el defines the file-locked and file-supersession errors. Several of the specialized Lisp libraries distributed with Emacs define their own error symbols. We do not attempt to list of all those here.

See Errors, for an explanation of how errors are generated and handled.

error

The message is ‘error’. See Errors.

quit

The message is ‘Quit’. See Quitting.

minibuffer-quit

The message is ‘Quit’. This is a subcategory of quit. See Quitting.

args-out-of-range

The message is ‘Args out of range’. This happens when trying to access an element beyond the range of a sequence, buffer, or other container-like object. See Sequences, Arrays, and Vectors, and see Text.

arith-error

The message is ‘Arithmetic error’. This occurs when trying to perform integer division by zero. See Numeric Conversions, and see Arithmetic Operations.

beginning-of-buffer

The message is ‘Beginning of buffer’. See Motion by Characters.

buffer-read-only

The message is ‘Buffer is read-only’. See Read-Only Buffers.

circular-list

The message is ‘List contains a loop’. This happens when a circular structure is encountered. See Read Syntax for Circular Objects.

cl-assertion-failed

The message is ‘Assertion failed’. This happens when the cl-assert macro fails a test. See Assertions in Common Lisp Extensions.

coding-system-error

The message is ‘Invalid coding system’. See Coding Systems in Lisp.

cyclic-function-indirection

The message is ‘Symbol's chain of function indirections contains a loop’. See Symbol Function Indirection.

cyclic-variable-indirection

The message is ‘Symbol's chain of variable indirections contains a loop’. See Variable Aliases.

dbus-error

The message is ‘D-Bus error’. See Errors and Events in D-Bus integration in Emacs.

end-of-buffer

The message is ‘End of buffer’. See Motion by Characters.

end-of-file

The message is ‘End of file during parsing’. Note that this is not a subcategory of file-error, because it pertains to the Lisp reader, not to file I/O. See Input Functions.

file-already-exists

This is a subcategory of file-error. See Writing to Files.

permission-denied

This is a subcategory of file-error, which occurs when the OS doesn’t allow Emacs to access a file or a directory for some reason.

file-date-error

This is a subcategory of file-error. It occurs when copy-file tries and fails to set the last-modification time of the output file. See Changing File Names and Attributes.

file-error

We do not list the error-strings of this error and its subcategories, because the error message is normally constructed from the data items alone when the error condition file-error is present. Thus, the error-strings are not very relevant. However, these error symbols do have error-message properties, and if no data is provided, the error-message property is used. See Files.

file-missing

This is a subcategory of file-error. It occurs when an operation attempts to act on a file that is missing. See Changing File Names and Attributes.

compression-error

This is a subcategory of file-error, which results from problems handling a compressed file. See How Programs Do Loading.

file-locked

This is a subcategory of file-error. See File Locks.

file-supersession

This is a subcategory of file-error. See Buffer Modification Time.

file-notify-error

This is a subcategory of file-error. It happens, when a file could not be watched for changes. See Notifications on File Changes.

remote-file-error

This is a subcategory of file-error, which results from problems in accessing a remote file. See Remote Files in The GNU Emacs Manual. Often, this error appears when timers, process filters, process sentinels or special events in general try to access a remote file, and collide with another remote file operation. In general it is a good idea to write a bug report. See Bugs in The GNU Emacs Manual.

ftp-error

This is a subcategory of remote-file-error, which results from problems in accessing a remote file using ftp. See Remote Files in The GNU Emacs Manual.

invalid-function

The message is ‘Invalid function’. See Symbol Function Indirection.

invalid-read-syntax

The message is usually ‘Invalid read syntax’. See Printed Representation and Read Syntax. This error can also be raised by commands like eval-expression when there’s text following an expression. In that case, the message is ‘Trailing garbage following expression’.

invalid-regexp

The message is ‘Invalid regexp’. See Regular Expressions.

mark-inactive

The message is ‘The mark is not active now’. See The Mark.

no-catch

The message is ‘No catch for tag’. See Explicit Nonlocal Exits: catch and throw.

range-error

The message is Arithmetic range error.

overflow-error

The message is ‘Arithmetic overflow error’. This is a subcategory of range-error. This can happen with integers exceeding the integer-width limit. See Integer Basics.

scan-error

The message is ‘Scan error’. This happens when certain syntax-parsing functions find invalid syntax or mismatched parentheses. Conventionally raised with three argument: a human-readable error message, the start of the obstacle that cannot be moved over, and the end of the obstacle. See Moving over Balanced Expressions, and see Parsing Expressions.

search-failed

The message is ‘Search failed’. See Searching and Matching.

setting-constant

The message is ‘Attempt to set a constant symbol’. This happens when attempting to assign values to nil, t, most-positive-fixnum, most-negative-fixnum, and keyword symbols. It also happens when attempting to assign values to enable-multibyte-characters and some other symbols whose direct assignment is not allowed for some reason. See Variables that Never Change.

text-read-only

The message is ‘Text is read-only’. This is a subcategory of buffer-read-only. See Properties with Special Meanings.

undefined-color

The message is ‘Undefined color’. See Color Names.

user-error

The message is the empty string. See How to Signal an Error.

user-search-failed

This is like ‘search-failed’, but doesn’t trigger the debugger, like ‘user-error’. See How to Signal an Error, and see Searching and Matching. This is used for searching in Info files, see Search Text in Info.

void-function

The message is ‘Symbol's function definition is void’. See Accessing Function Cell Contents.

void-variable

The message is ‘Symbol's value as variable is void’. See Accessing Variable Values.

wrong-number-of-arguments

The message is ‘Wrong number of arguments’. See Features of Argument Lists.

wrong-type-argument

The message is ‘Wrong type argument’. See Type Predicates.

unknown-image-type

The message is ‘Cannot determine image type’. See Images.

inhibited-interaction

The message is ‘User interaction while inhibited’. This error is signaled when inhibit-interaction is non-nil and a user interaction function (like read-from-minibuffer) is called.


Appendix G Standard Keymaps

In this section we list some of the more general keymaps. Many of these exist when Emacs is first started, but some are loaded only when the respective feature is accessed.

There are many other, more specialized, maps than these; in particular those associated with major and minor modes. The minibuffer uses several keymaps (see Minibuffer Commands that Do Completion). For more details on keymaps, see Keymaps.

2C-mode-map

A sparse keymap for subcommands of the prefix C-x 6.
See Two-Column Editing in The GNU Emacs Manual.

abbrev-map

A sparse keymap for subcommands of the prefix C-x a.
See Defining Abbrevs in The GNU Emacs Manual.

button-buffer-map

A sparse keymap useful for buffers containing buffers.
You may want to use this as a parent keymap. See Buttons.

button-map

A sparse keymap used by buttons.

ctl-x-4-map

A sparse keymap for subcommands of the prefix C-x 4.

ctl-x-5-map

A sparse keymap for subcommands of the prefix C-x 5.

ctl-x-map

A full keymap for C-x commands.

ctl-x-r-map

A sparse keymap for subcommands of the prefix C-x r.
See Registers in The GNU Emacs Manual.

esc-map

A full keymap for ESC (or Meta) commands.

function-key-map

The parent keymap of all local-function-key-map (q.v.) instances.

global-map

The full keymap containing default global key bindings.
Modes should not modify the Global map.

goto-map

A sparse keymap used for the M-g prefix key.

help-map

A sparse keymap for the keys following the help character C-h.
See Help Functions.

Helper-help-map

A full keymap used by the help utility package.
It has the same keymap in its value cell and in its function cell.

input-decode-map

The keymap for translating keypad and function keys.
If there are none, then it contains an empty sparse keymap. See Keymaps for Translating Sequences of Events.

key-translation-map

A keymap for translating keys. This one overrides ordinary key bindings, unlike local-function-key-map. See Keymaps for Translating Sequences of Events.

kmacro-keymap

A sparse keymap for keys that follows the C-x C-k prefix search.
See Keyboard Macros in The GNU Emacs Manual.

local-function-key-map

The keymap for translating key sequences to preferred alternatives.
If there are none, then it contains an empty sparse keymap. See Keymaps for Translating Sequences of Events.

menu-bar-file-menu
menu-bar-edit-menu
menu-bar-options-menu
global-buffers-menu-map
menu-bar-tools-menu
menu-bar-help-menu

These keymaps display the main, top-level menus in the menu bar.
Some of them contain sub-menus. For example, the Edit menu contains menu-bar-search-menu, etc. See The Menu Bar.

minibuffer-inactive-mode-map

A full keymap used in the minibuffer when it is not active.
See Editing in the Minibuffer in The GNU Emacs Manual.

mode-line-coding-system-map
mode-line-input-method-map
mode-line-column-line-number-mode-map

These keymaps control various areas of the mode line.
See Mode Line Format.

mode-specific-map

The keymap for characters following C-c. Note, this is in the global map. This map is not actually mode-specific: its name was chosen to be informative in C-h b (display-bindings), where it describes the main use of the C-c prefix key.

mouse-appearance-menu-map

A sparse keymap used for the S-mouse-1 key.

mule-keymap

The global keymap used for the C-x RET prefix key.

narrow-map

A sparse keymap for subcommands of the prefix C-x n.

prog-mode-map

The keymap used by Prog mode.
See Basic Major Modes.

query-replace-map
multi-query-replace-map

A sparse keymap used for responses in query-replace and related commands; also for y-or-n-p and map-y-or-n-p. The functions that use this map do not support prefix keys; they look up one event at a time. multi-query-replace-map extends query-replace-map for multi-buffer replacements. See query-replace-map.

search-map

A sparse keymap that provides global bindings for search-related commands.

special-mode-map

The keymap used by Special mode.
See Basic Major Modes.

tab-prefix-map

The global keymap used for the C-x t prefix key for tab-bar related commands.
See Tab Bars in The GNU Emacs Manual.

tab-bar-map

The keymap defining the contents of the tab bar.
See Tab Bars in The GNU Emacs Manual.

tool-bar-map

The keymap defining the contents of the tool bar.
See Tool bars.

universal-argument-map

A sparse keymap used while processing C-u.
See Prefix Command Arguments.

vc-prefix-map

The global keymap used for the C-x v prefix key.

x-alternatives-map

A sparse keymap used to map certain keys under graphical frames.
The function x-setup-function-keys uses this.


Appendix H Standard Hooks

The following is a list of some hook variables that let you provide functions to be called from within Emacs on suitable occasions.

Most of these variables have names ending with ‘-hook’. They are normal hooks, run by means of run-hooks. The value of such a hook is a list of functions; the functions are called with no arguments and their values are completely ignored. The recommended way to put a new function on such a hook is to call add-hook. See Hooks, for more information about using hooks.

The variables whose names end in ‘-functions’ are usually abnormal hooks (some old code may also use the deprecated ‘-hooks’ suffix). Their values are lists of functions, but these functions are called in a special way: they are either passed arguments, or their return values are used in some way. The variables whose names end in ‘-function’ have single functions as their values.

This is not an exhaustive list, it only covers the more general hooks. For example, every major mode defines a hook named ‘modename-mode-hook’. The major mode command runs this normal hook with run-mode-hooks as the very last thing it does. See Mode Hooks. Most minor modes have mode hooks too.

A special feature allows you to specify expressions to evaluate if and when a file is loaded (see Hooks for Loading). That feature is not exactly a hook, but does a similar job.

activate-mark-hook
deactivate-mark-hook

See The Mark.

after-change-functions
before-change-functions
first-change-hook

See Change Hooks.

after-change-major-mode-hook
change-major-mode-after-body-hook

See Mode Hooks.

after-init-hook
before-init-hook
emacs-startup-hook
window-setup-hook

See The Init File.

after-insert-file-functions
write-region-annotate-functions
write-region-post-annotation-function

See File Format Conversion.

after-make-frame-functions
before-make-frame-hook
server-after-make-frame-hook

See Creating Frames.

after-save-hook
before-save-hook
write-contents-functions
write-file-functions

See Saving Buffers.

after-setting-font-hook

Hook run after a frame’s font changes.

auto-save-hook

See Auto-Saving.

before-hack-local-variables-hook
hack-local-variables-hook

See File Local Variables.

buffer-access-fontify-functions

See Lazy Computation of Text Properties.

buffer-list-update-hook

Hook run when the buffer list changes (see The Buffer List).

buffer-quit-function

Function to call to quit the current buffer.

change-major-mode-hook

See Creating and Deleting Buffer-Local Bindings.

comint-password-function

This abnormal hook permits a derived mode to supply a password for the underlying command interpreter without prompting the user.

command-line-functions

See Command-Line Arguments.

delayed-warnings-hook

The command loop runs this soon after post-command-hook (q.v.).

focus-in-hook
focus-out-hook

See Input Focus.

delete-frame-functions
after-delete-frame-functions

See Deleting Frames.

delete-terminal-functions

See Multiple Terminals.

pop-up-frame-function
split-window-preferred-function

See Additional Options for Displaying Buffers.

echo-area-clear-hook

See Echo Area Customization.

find-file-hook
find-file-not-found-functions

See Functions for Visiting Files.

font-lock-extend-after-change-region-function

See Region to Fontify after a Buffer Change.

font-lock-extend-region-functions

See Multiline Font Lock Constructs.

font-lock-fontify-buffer-function
font-lock-fontify-region-function
font-lock-mark-block-function
font-lock-unfontify-buffer-function
font-lock-unfontify-region-function

See Other Font Lock Variables.

fontification-functions

See Automatic Face Assignment.

frame-auto-hide-function

See Quitting Windows.

quit-window-hook

See Quitting Windows.

kill-buffer-hook
kill-buffer-query-functions

See Killing Buffers.

kill-emacs-hook
kill-emacs-query-functions

See Killing Emacs.

menu-bar-update-hook

See The Menu Bar.

minibuffer-setup-hook
minibuffer-exit-hook

See Minibuffer Miscellany.

mouse-leave-buffer-hook

Hook run when the user mouse-clicks in a window.

mouse-position-function

See Mouse Position.

prefix-command-echo-keystrokes-functions

An abnormal hook run by prefix commands (such as C-u) which should return a string describing the current prefix state. For example, C-u produces ‘C-u-’ and ‘C-u 1 2 3-’. Each hook function is called with no arguments and should return a string describing the current prefix state, or nil if there’s no prefix state. See Prefix Command Arguments.

prefix-command-preserve-state-hook

Hook run when a prefix command needs to preserve the prefix by passing the current prefix command state to the next command. For example, C-u needs to pass the state to the next command when the user types C-u - or follows C-u with a digit.

pre-redisplay-functions

Hook run in each window just before redisplaying it. See Forcing Redisplay.

post-command-hook
pre-command-hook

See Command Loop Overview.

post-gc-hook

See Garbage Collection.

post-self-insert-hook

See Keymaps and Minor Modes.

suspend-hook
suspend-resume-hook
suspend-tty-functions
resume-tty-functions

See Suspending Emacs.

syntax-begin-function
syntax-propertize-extend-region-functions
syntax-propertize-function
font-lock-syntactic-face-function

See Syntactic Font Lock. See Syntax Properties.

temp-buffer-setup-hook
temp-buffer-show-function
temp-buffer-show-hook

See Temporary Displays.

tty-setup-hook

See Terminal-Specific Initialization.

window-configuration-change-hook
window-scroll-functions
window-size-change-functions

See Hooks for Window Scrolling and Changes.


Index

Jump to:   "   #   $   %   &   '   (   )   *   +   ,   -   .   /   0   1   2   3   :   ;   <   =   >   ?   @   [   \   ]   ^   `   |  
A   B   C   D   E   F   G   H   I   J   K   L   M   N   O   P   Q   R   S   T   U   V   W   X   Y   Z  
Index Entry  Section

"
"’ in printing: Output Functions
"’ in strings: Syntax for Strings

#
##’ read syntax: Symbol Type
#$: Docs and Compilation
#'’ syntax: Anonymous Functions
#(’ read syntax: Text Props and Strings
#:’ read syntax: Symbol Type
#@count: Docs and Compilation
#n#’ read syntax: Circular Objects
#n=’ read syntax: Circular Objects
#^’ read syntax: Char-Table Type

$
$’ in display: Truncation
$’ in regexp: Regexp Special

%
%: Arithmetic Operations
%’ in format: Formatting Strings

&
&’ in replacement: Replacing Match
&optional: Argument List
&rest: Argument List

'
'’ for quoting: Quoting

(
(’ in regexp: Regexp Backslash
(…)’ in lists: Cons Cell Type
(?:’ in regexp: Regexp Backslash

)
)’ in regexp: Regexp Backslash

*
*: Arithmetic Operations
*’ in interactive: Using Interactive
*’ in regexp: Regexp Special
* in rx: Rx Constructs
** in rx: Rx Constructs
*? in rx: Rx Constructs
*scratch*: Auto Major Mode

+
+: Arithmetic Operations
+’ in regexp: Regexp Special
+ in rx: Rx Constructs
+? in rx: Rx Constructs

,
, (with backquote): Backquote
,@ (with backquote): Backquote

-
-: Arithmetic Operations
--disable-build-details option to configure: Building Emacs
--enable-locallisppath option to configure: Building Emacs
--temacs option, and dumping method: Building Emacs
–enable-profiling option of configure: Profiling

.
.’ in lists: Dotted Pair Notation
.’ in regexp: Regexp Special
.#, lock file names: File Locks
.emacs: Init File

/
/: Arithmetic Operations
/=: Comparison of Numbers
/dev/tty: Serial Ports

0
0+ in rx: Rx Constructs

1
1+: Arithmetic Operations
1+ in rx: Rx Constructs
1-: Arithmetic Operations
1value: Test Coverage

2
2C-mode-map: Prefix Keys
2D box: Face Attributes

3
3D box: Face Attributes

:
: in rx: Rx Constructs
:deferred, JSONRPC keyword: JSONRPC deferred requests
:documentation: Function Documentation
:equal: Pattern Matching
:match: Pattern Matching
:notification-dispatcher: JSONRPC Overview
:pred: Pattern Matching
:repeat: Creating Keymaps
:request-dispatcher: JSONRPC Overview

;
;’ for commenting: Comments

<
<: Comparison of Numbers
<=: Comparison of Numbers

=
=: Comparison of Numbers
= in rx: Rx Constructs

>
>: Comparison of Numbers
>=: Comparison of Numbers
>= in rx: Rx Constructs

?
?’ in character constant: Basic Char Syntax
? in minibuffer: Text from Minibuffer
?’ in regexp: Regexp Special
? in rx: Rx Constructs
?? in rx: Rx Constructs

@
@’ in interactive: Using Interactive

[
[’ in regexp: Regexp Special
[…] (Edebug): Specification List

\
\’ in character constant: General Escape Syntax
\’ in display: Truncation
\’ in printing: Output Functions
\’ in regexp: Regexp Special
\’ in replacement: Replacing Match
\’ in strings: Syntax for Strings
\’ in symbols: Symbol Type
\'’ in regexp: Regexp Backslash
\( in strings: List Motion
\<’ in regexp: Regexp Backslash
\=’ in regexp: Regexp Backslash
\>’ in regexp: Regexp Backslash
\a: Basic Char Syntax
\b: Basic Char Syntax
\b’ in regexp: Regexp Backslash
\B’ in regexp: Regexp Backslash
\e: Basic Char Syntax
\f: Basic Char Syntax
\n: Basic Char Syntax
\n’ in print: Output Variables
\n’ in replacement: Replacing Match
\r: Basic Char Syntax
\s: Basic Char Syntax
\s’ in regexp: Regexp Backslash
\S’ in regexp: Regexp Backslash
\t: Basic Char Syntax
\v: Basic Char Syntax
\w’ in regexp: Regexp Backslash
\W’ in regexp: Regexp Backslash
\_<’ in regexp: Regexp Backslash
\_>’ in regexp: Regexp Backslash
\`’ in regexp: Regexp Backslash

]
]’ in regexp: Regexp Special

^
^’ in interactive: Using Interactive
^’ in regexp: Regexp Special

`
`: Backquote
‘ (list substitution): Backquote

|
|’ in regexp: Regexp Backslash
| in rx: Rx Constructs

A
abbrev: Abbrevs
abbrev properties: Abbrev Properties
abbrev table properties: Abbrev Table Properties
abbrev tables: Abbrev Tables
abbrev tables in modes: Major Mode Conventions
abbrev-all-caps: Abbrev Expansion
abbrev-expand-function: Abbrev Expansion
abbrev-expansion: Abbrev Expansion
abbrev-file-name: Abbrev Files
abbrev-get: Abbrev Properties
abbrev-insert: Abbrev Expansion
abbrev-map: Standard Keymaps
abbrev-minor-mode-table-alist: Standard Abbrev Tables
abbrev-prefix-mark: Abbrev Expansion
abbrev-put: Abbrev Properties
abbrev-start-location: Abbrev Expansion
abbrev-start-location-buffer: Abbrev Expansion
abbrev-symbol: Abbrev Expansion
abbrev-table-get: Abbrev Table Properties
abbrev-table-name-list: Abbrev Tables
abbrev-table-p: Abbrev Tables
abbrev-table-put: Abbrev Table Properties
abbreviate-file-name: Directory Names
abbreviated file names: Directory Names
abbrevs, looking up and expanding: Abbrev Expansion
abbrevs-changed: Abbrev Files
abnormal hook: Hooks
abort-recursive-edit: Recursive Editing
aborting: Recursive Editing
abs: Comparison of Numbers
absolute edges: Frame Layout
absolute file name: Relative File Names
absolute frame edges: Frame Layout
absolute frame position: Frame Layout
absolute position: Frame Layout
accept input from processes: Accepting Output
accept-change-group: Atomic Changes
accept-process-output: Accepting Output
access control list: Extended Attributes
access minibuffer contents: Minibuffer Contents
access-file: Testing Accessibility
accessibility of a file: Testing Accessibility
accessible portion (of a buffer): Narrowing
accessible-keymaps: Scanning Keymaps
accessing documentation strings: Accessing Documentation
accessing hash tables: Hash Access
accessing plist properties: Plist Access
ACL entries: Extended Attributes
acos: Math Functions
action (button property): Button Properties
action alist for buffer display: Buffer Display Action Alists
action function, for buffer display: Buffer Display Action Functions
action, customization keyword: Type Keywords
activate-change-group: Atomic Changes
activate-mark-hook: The Mark
active display table: Active Display Table
active keymap: Active Keymaps
active keymap, controlling: Controlling Active Maps
active minibuffer: Intro to Minibuffers
active-minibuffer-window: Minibuffer Windows
ad-activate: Porting Old Advice
adaptive-fill-first-line-regexp: Adaptive Fill
adaptive-fill-function: Adaptive Fill
adaptive-fill-mode: Adaptive Fill
adaptive-fill-regexp: Adaptive Fill
add-display-text-property: Display Property
add-face-text-property: Changing Properties
add-function: Core Advising Primitives
add-hook: Setting Hooks
add-name-to-file: Changing Files
add-text-properties: Changing Properties
add-to-history: Minibuffer History
add-to-invisibility-spec: Invisible Text
add-to-list: List Variables
add-to-ordered-list: List Variables
add-variable-watcher: Watching Variables
address field of register: Cons Cell Type
adjust-window-trailing-edge: Resizing Windows
adjusting point: Adjusting Point
advertised binding: Keys in Documentation
advertised-calling-convention (declare spec): Declare Form
advice, add and remove: Core Advising Primitives
advice-add: Advising Named Functions
advice-eval-interactive-spec: Core Advising Primitives
advice-function-mapc: Core Advising Primitives
advice-function-member-p: Core Advising Primitives
advice-mapc: Advising Named Functions
advice-member-p: Advising Named Functions
advice-remove: Advising Named Functions
advices, porting from defadvice: Porting Old Advice
advising functions: Advising Functions
advising named functions: Advising Named Functions
AEAD cipher: GnuTLS Cryptography
affixation-function, in completion: Programmed Completion
after-change notifier, for tree-sitter parse-tree: Using Parser
after-change-functions: Change Hooks
after-change-major-mode-hook: Mode Hooks
after-delete-frame-functions: Deleting Frames
after-find-file: Subroutines of Visiting
after-focus-change-function: Input Focus
after-init-hook: Init File
after-init-time: Startup Summary
after-insert-file-functions: Format Conversion Piecemeal
after-load-functions: Hooks for Loading
after-make-frame-functions: Creating Frames
after-pdump-load-hook: Building Emacs
after-revert-hook: Reverting
after-save-hook: Saving Buffers
after-setting-font-hook: Standard Hooks
after-string (overlay property): Overlay Properties
alias, for coding systems: Coding System Basics
alias, for faces: Face Functions
alias, for functions: Defining Functions
alias, for variables: Variable Aliases
aligning header line, when line numbers are displayed: Pixel Specification
alist: Association Lists
alist vs. plist: Plists and Alists
alist-get: Association Lists
all-completions: Basic Completion
all-threads: Basic Thread Functions
allow-no-window, a buffer display action alist entry: Buffer Display Action Alists
alnum character class, regexp: Char Classes
alpha character class, regexp: Char Classes
alpha, a frame parameter: Font and Color Parameters
alpha-background, a frame parameter: Font and Color Parameters
alt characters: Other Char Bits
alternative commands, defining: Generic Commands
always: Calling Functions
amalgamating commands, and undo: Undo
amalgamating-undo-limit: Undo
ancestor directory of file: Contents of Directories
and: Combining Conditions
and in rx: Rx Constructs
animation: Multi-Frame Images
annotation-function, in completion: Programmed Completion
anonymous face: Faces
anonymous function: Anonymous Functions
anonymous node, tree-sitter: Language Grammar
any in rx: Rx Constructs
anychar in rx: Rx Constructs
anything in rx: Rx Constructs
apostrophe for quoting: Quoting
apostrophe, quoting in documentation strings: Documentation Tips
append: Building Lists
append-to-file: Writing to Files
apply: Calling Functions
apply, and debugging: Internals of Debugger
apply-partially: Calling Functions
applying customizations: Applying Customizations
apropos: Help Functions
archive web server: Archive Web Server
aref: Array Functions
args, customization keyword: Composite Types
argument: What Is a Function
argument binding: Argument List
argument lists, features: Argument List
arguments for shell commands: Shell Arguments
arguments, interactive entry: Using Interactive
arguments, reading: Minibuffers
argv: Command-Line Arguments
arith-error example: Handling Errors
arith-error in division: Arithmetic Operations
arithmetic operations: Arithmetic Operations
arithmetic shift: Bitwise Operations
array: Arrays
array elements: Array Functions
arrayp: Array Functions
ascii character class, regexp: Char Classes
ASCII character codes: Character Type
ASCII control characters: Usual Display
ascii-case-table: Case Tables
aset: Array Functions
ash: Bitwise Operations
asin: Math Functions
ask-user-about-lock: File Locks
ask-user-about-supersession-threat: Modification Time
asking the user questions: Yes-or-No Queries
assoc: Association Lists
assoc-default: Association Lists
assoc-delete-all: Association Lists
assoc-string: Text Comparison
association list: Association Lists
assq: Association Lists
assq-delete-all: Association Lists
asynchronous native compilation, disable: Native-Compilation Variables
asynchronous subprocess: Asynchronous Processes
atan: Math Functions
atom: List-related Predicates
atomic changes: Atomic Changes
atomic windows: Atomic Windows
atoms: Cons Cell Type
attributes of text: Text Properties
Auto Fill mode: Auto Filling
auto-coding-alist: Default Coding Systems
auto-coding-functions: Default Coding Systems
auto-coding-regexp-alist: Default Coding Systems
auto-fill-chars: Auto Filling
auto-fill-function: Auto Filling
auto-hide-function, a frame parameter: Frame Interaction Parameters
auto-hscroll-mode: Horizontal Scrolling
auto-lower, a frame parameter: Management Parameters
auto-mode-alist: Auto Major Mode
auto-raise, a frame parameter: Management Parameters
auto-raise-tool-bar-buttons: Tool Bar
auto-resize-tool-bars: Tool Bar
auto-save-default: Auto-Saving
auto-save-file-name-p: Auto-Saving
auto-save-file-name-transforms: Auto-Saving
auto-save-hook: Auto-Saving
auto-save-interval: Auto-Saving
auto-save-list-file-name: Auto-Saving
auto-save-list-file-prefix: Auto-Saving
auto-save-mode: Auto-Saving
auto-save-timeout: Auto-Saving
auto-save-visited-file-name: Auto-Saving
auto-selection of window: Mouse Window Auto-selection
auto-window-vscroll: Vertical Scrolling
autoload: Autoload
autoload: Autoload
autoload by prefix: Autoload by Prefix
autoload cookie: Autoload
autoload cookie, and safe values of variable: File Local Variables
autoload errors: Autoload
autoload object: What Is a Function
autoload, when to use: When to Autoload
autoload-compute-prefixes: Autoload by Prefix
autoload-do-load: Autoload
autoloadp: Autoload
automatic face assignment: Auto Faces
automatically buffer-local: Intro to Buffer-Local

B
back-to-indentation: Motion by Indent
background-color, a frame parameter: Font and Color Parameters
background-mode, a frame parameter: Font and Color Parameters
backing store: Display Feature Testing
backquote (list substitution): Backquote
backquote-style patterns: Backquote Patterns
backref in rx: Rx Constructs
backslash in character constants: General Escape Syntax
backslash in regular expressions: Regexp Backslash
backslash in strings: Syntax for Strings
backslash in symbols: Symbol Type
backspace: Basic Char Syntax
backtrace: Internals of Debugger
backtrace buffer: Backtraces
backtrace from emacsclient’s --eval: Error Debugging
backtrace of thread: The Thread List
backtrace-debug: Internals of Debugger
backtrace-frame: Internals of Debugger
backtrace-on-error-noninteractive: Invoking the Debugger
backtrace-on-redisplay-error: Debugging Redisplay
backtracking: Backtracking
backtracking and POSIX regular expressions: POSIX Regexps
backtracking and regular expressions: Regexp Special
backup file: Backup Files
backup files, rename or copy: Rename or Copy
backup-buffer: Making Backups
backup-by-copying: Rename or Copy
backup-by-copying-when-linked: Rename or Copy
backup-by-copying-when-mismatch: Rename or Copy
backup-by-copying-when-privileged-mismatch: Rename or Copy
backup-directory-alist: Making Backups
backup-enable-predicate: Making Backups
backup-file-name-p: Backup Names
backup-inhibited: Making Backups
backups and auto-saving: Backups and Auto-Saving
backward-button: Button Buffer Commands
backward-char: Character Motion
backward-delete-char-untabify: Deletion
backward-delete-char-untabify-method: Deletion
backward-list: List Motion
backward-prefix-chars: Motion and Syntax
backward-sexp: List Motion
backward-to-indentation: Motion by Indent
backward-up-list: List Motion
backward-word: Word Motion
backward-word-strictly: Word Motion
balance-windows: Resizing Windows
balance-windows-area: Resizing Windows
balanced parenthesis motion: List Motion
balancing parentheses: Blinking
balancing window sizes: Resizing Windows
bare symbol: Symbols with Position
bare-symbol: Symbols with Position
barf-if-buffer-read-only: Read Only Buffers
base 64 encoding: Base 64
base buffer: Indirect Buffers
base coding system: Coding System Basics
base direction of a paragraph: Bidirectional Display
base for reading an integer: Integer Basics
base location, package archive: Package Archives
base remapping, faces: Face Remapping
base type, in bindat specification: Bindat Types
base64-decode-region: Base 64
base64-decode-string: Base 64
base64-encode-region: Base 64
base64-encode-string: Base 64
base64url-encode-region: Base 64
base64url-encode-string: Base 64
basic code (of input character): Keyboard Events
basic faces: Basic Faces
batch mode: Batch Mode
batch-byte-compile: Compilation Functions
batch-native-compile: Native-Compilation Functions
baud, in serial connections: Serial Ports
baud-rate: Terminal Output
beep: Beeping
before point, insertion: Insertion
before-change-functions: Change Hooks
before-hack-local-variables-hook: File Local Variables
before-init-hook: Init File
before-init-time: Startup Summary
before-make-frame-hook: Creating Frames
before-revert-hook: Reverting
before-save-hook: Saving Buffers
before-string (overlay property): Overlay Properties
beginning of line: Text Lines
beginning of line in regexp: Regexp Special
beginning-of-buffer: Buffer End Motion
beginning-of-defun: List Motion
beginning-of-defun-function: List Motion
beginning-of-line: Text Lines
bell: Beeping
bell character: Basic Char Syntax
benchmark.el: Profiling
benchmarking: Profiling
best practices: Tips
bidi-display-reordering: Bidirectional Display
bidi-find-overridden-directionality: Bidirectional Display
bidi-paragraph-direction: Bidirectional Display
bidi-paragraph-separate-re: Bidirectional Display
bidi-paragraph-start-re: Bidirectional Display
bidi-string-mark-left-to-right: Bidirectional Display
bidirectional class of characters: Character Properties
bidirectional display: Bidirectional Display
bidirectional reordering: Bidirectional Display
big endian, in bindat specification: Bindat Types
bignum range: Integer Basics
bignump: Predicates on Numbers
binary coding system: Coding System Basics
binary I/O in batch mode: Input Functions
bindat computed types: Bindat Computed Types
bindat functions: Bindat Functions
bindat packing and unpacking into arbitrary types: Bindat Computed Types
bindat type expression: Bindat Types
bindat types: Bindat Types
bindat, define new type forms: Bindat Computed Types
bindat-defmacro: Bindat Computed Types
bindat-get-field: Bindat Functions
bindat-ip-to-string: Bindat Functions
bindat-length: Bindat Functions
bindat-pack: Bindat Functions
bindat-type: Bindat Types
bindat-unpack: Bindat Functions
binding arguments: Argument List
binding local variables: Local Variables
binding of a key: Keymap Basics
bitmap-spec-p: Face Attributes
bitmaps, fringe: Fringe Bitmaps
bitwise arithmetic: Bitwise Operations
blink-cursor-alist: Cursor Parameters
blink-matching-delay: Blinking
blink-matching-open: Blinking
blink-matching-paren: Blinking
blink-matching-paren-distance: Blinking
blink-paren-function: Blinking
blinking parentheses: Blinking
bobp: Near Point
body height of a window: Window Sizes
body of a window: Basic Windows
body of function: Lambda Components
body size of a window: Window Sizes
body width of a window: Window Sizes
body-function, a buffer display action alist entry: Buffer Display Action Alists
bol in rx: Rx Constructs
bolp: Near Point
bool-vector: Bool-Vectors
bool-vector length: Sequence Functions
bool-vector-count-consecutive: Bool-Vectors
bool-vector-count-population: Bool-Vectors
bool-vector-exclusive-or: Bool-Vectors
bool-vector-intersection: Bool-Vectors
bool-vector-not: Bool-Vectors
bool-vector-p: Bool-Vectors
bool-vector-set-difference: Bool-Vectors
bool-vector-subsetp: Bool-Vectors
bool-vector-union: Bool-Vectors
Bool-vectors: Bool-Vectors
boolean: nil and t
booleanp: nil and t
bootstrapping Emacs: Building Emacs
border-color, a frame parameter: Font and Color Parameters
border-width, a frame parameter: Layout Parameters
bos in rx: Rx Constructs
bot in rx: Rx Constructs
bottom dividers: Window Dividers
bottom-divider, prefix key: Key Sequence Input
bottom-divider-width, a frame parameter: Layout Parameters
bottom-visible, a frame parameter: Mouse Dragging Parameters
boundp: Void Variables
bow in rx: Rx Constructs
box diagrams, for lists: Box Diagrams
break: Debugger
breakpoints (Edebug): Breakpoints
bucket (in obarray): Creating Symbols
buffer: Buffers
buffer boundaries, indicating: Fringe Indicators
buffer contents: Text
buffer display: Displaying Buffers
buffer display action alist: Buffer Display Action Alists
buffer display action function: Buffer Display Action Functions
buffer display action functions, precedence: Precedence of Action Functions
buffer display conventions: The Zen of Buffer Display
buffer display display action: Choosing Window
buffer file name: Buffer File Name
buffer gap: Buffer Gap
buffer input stream: Input Streams
buffer internals: Buffer Internals
buffer list: Buffer List
buffer modification: Buffer Modification
buffer names: Buffer Names
buffer output stream: Output Streams
buffer point changed by Edebug: Edebug Display Update
buffer portion as string: Buffer Contents
buffer position: Positions
buffer text notation: Buffer Text Notation
buffer, read-only: Read Only Buffers
buffer-access-fontified-property: Lazy Properties
buffer-access-fontify-functions: Lazy Properties
buffer-auto-save-file-format: Format Conversion Round-Trip
buffer-auto-save-file-name: Auto-Saving
buffer-backed-up: Making Backups
buffer-base-buffer: Indirect Buffers
buffer-button-map: Button Buffer Commands
buffer-chars-modified-tick: Buffer Modification
buffer-disable-undo: Maintaining Undo
buffer-display-count: Buffers and Windows
buffer-display-table: Active Display Table
buffer-display-time: Buffers and Windows
buffer-enable-undo: Maintaining Undo
buffer-end: Point
buffer-end in rx: Rx Constructs
buffer-file-coding-system: Encoding and I/O
buffer-file-format: Format Conversion Round-Trip
buffer-file-name: Buffer File Name
buffer-file-name: Buffer File Name
buffer-file-number: Buffer File Name
buffer-file-truename: Buffer File Name
buffer-hash: Checksum/Hash
buffer-invisibility-spec: Invisible Text
buffer-list: Buffer List
buffer-list, a frame parameter: Buffer Parameters
buffer-list-update-hook: Buffer List
buffer-list-update-hook: Standard Hooks
buffer-list-update-hook in temporary buffers: Current Buffer
buffer-live-p: Killing Buffers
buffer-local variables: Buffer-Local Variables
buffer-local variables in modes: Major Mode Conventions
buffer-local-boundp: Creating Buffer-Local
buffer-local-restore-state: Defining Minor Modes
buffer-local-set-state: Defining Minor Modes
buffer-local-value: Creating Buffer-Local
buffer-local-variables: Creating Buffer-Local
buffer-match-p: Buffer List
buffer-modified-p: Buffer Modification
buffer-modified-tick: Buffer Modification
buffer-name: Buffer Names
buffer-name-history: Minibuffer History
buffer-narrowed-p: Narrowing
buffer-offer-save: Killing Buffers
buffer-predicate, a frame parameter: Buffer Parameters
buffer-quit-function: Standard Hooks
buffer-read-only: Read Only Buffers
buffer-save-without-query: Killing Buffers
buffer-saved-size: Auto-Saving
buffer-size: Point
buffer-stale-function: Reverting
buffer-start in rx: Rx Constructs
buffer-string: Buffer Contents
buffer-substring: Buffer Contents
buffer-substring-no-properties: Buffer Contents
buffer-substring-with-bidi-context: Bidirectional Display
buffer-swap-text: Swapping Text
buffer-text-pixel-size: Size of Displayed Text
buffer-undo-list: Undo
bufferp: Buffer Basics
bufferpos-to-filepos: Text Representations
buffers to display on frame: Buffer Parameters
buffers without undo information: Buffer Names
buffers, controlled in windows: Buffers and Windows
buffers, creating: Creating Buffers
buffers, killing: Killing Buffers
bugs: Caveats
bugs in this manual: Caveats
build details: Building Emacs
building Emacs: Building Emacs
building lists: Building Lists
built-in function: What Is a Function
bump-use-time, a buffer display action alist entry: Buffer Display Action Alists
bury-buffer: Buffer List
butlast: List Elements
button (button property): Button Properties
button buffer commands: Button Buffer Commands
button properties: Button Properties
button types: Button Types
button-activate: Manipulating Buttons
button-at: Manipulating Buttons
button-down event: Button-Down Events
button-end: Manipulating Buttons
button-face, customization keyword: Type Keywords
button-get: Manipulating Buttons
button-has-type-p: Manipulating Buttons
button-label: Manipulating Buttons
button-map: Button Properties
button-prefix, customization keyword: Type Keywords
button-put: Manipulating Buttons
button-start: Manipulating Buttons
button-suffix, customization keyword: Type Keywords
button-type: Manipulating Buttons
button-type-get: Manipulating Buttons
button-type-put: Manipulating Buttons
button-type-subtype-p: Manipulating Buttons
buttonize: Making Buttons
buttons in buffers: Buttons
byte compilation: Byte Compilation
byte compiler warnings, how to avoid: Warning Tips
byte packing and unpacking: Byte Packing
byte to string: Converting Representations
byte-boolean-vars: Variables with Restricted Values
byte-boolean-vars: Writing Emacs Primitives
byte-code: Byte Compilation
byte-code function: Byte-Code Objects
byte-code object: Byte-Code Objects
byte-code-function-p: What Is a Function
byte-compile: Compilation Functions
byte-compile and byte-optimize, advising: Advice and Byte Code
byte-compile-debug: Compilation Functions
byte-compile-dynamic: Dynamic Loading
byte-compile-dynamic-docstrings: Docs and Compilation
byte-compile-error-on-warn: Compiler Errors
byte-compile-file: Compilation Functions
byte-compiler errors: Compiler Errors
byte-compiler warnings: Compiler Errors
byte-compiling macros: Compiling Macros
byte-compiling require: Named Features
byte-recompile-directory: Compilation Functions
byte-to-position: Text Representations
byte-to-string: Converting Representations
bytes: Strings and Characters
bytesize, in serial connections: Serial Ports

C
C programming language: C Dialect
C-c: Prefix Keys
C-g: Quitting
C-h: Prefix Keys
C-M-x: Instrumenting
C-x: Prefix Keys
C-x 4: Prefix Keys
C-x 5: Prefix Keys
C-x 6: Prefix Keys
C-x C-a C-m: Edebug Execution Modes
C-x RET: Prefix Keys
C-x t: Prefix Keys
C-x v: Prefix Keys
C-x X =: Coverage Testing
caaaar: List Elements
caaadr: List Elements
caaar: List Elements
caadar: List Elements
caaddr: List Elements
caadr: List Elements
caar: List Elements
cadaar: List Elements
cadadr: List Elements
cadar: List Elements
caddar: List Elements
cadddr: List Elements
caddr: List Elements
cadr: List Elements
calendrical computations: Time Calculations
calendrical information: Time Conversion
call stack: Internals of Debugger
call-interactively: Interactive Call
call-process: Synchronous Processes
call-process, command-line arguments from minibuffer: Shell Arguments
call-process-region: Synchronous Processes
call-process-shell-command: Synchronous Processes
call-shell-region: Synchronous Processes
called-interactively-p: Distinguish Interactive
calling a function: Calling Functions
cancel-change-group: Atomic Changes
cancel-debug-on-entry: Function Debugging
cancel-debug-on-variable-change: Variable Debugging
cancel-timer: Timers
canonical character height: Frame Font
canonical character width: Frame Font
capitalization: Case Conversion
capitalize: Case Conversion
capitalize-region: Case Changes
capitalize-word: Case Changes
caption bar: Frame Layout
capturing, tree-sitter node: Pattern Matching
car: List Elements
car-safe: List Elements
case conversion in buffers: Case Changes
case conversion in Lisp: Case Conversion
case in replacements: Replacing Match
case-fold, text property: Text from Minibuffer
case-fold-search: Searching and Case
case-replace: Searching and Case
case-table-p: Case Tables
catch: Catch and Throw
catch-all: Parser-based Indentation
categories of characters: Categories
category (overlay property): Overlay Properties
category (text property): Special Properties
category in rx: Rx Constructs
category set: Categories
category table: Categories
category, in completion: Programmed Completion
category, regexp search for: Regexp Backslash
category-docstring: Categories
category-set-mnemonics: Categories
category-table: Categories
category-table-p: Categories
cdaaar: List Elements
cdaadr: List Elements
cdaar: List Elements
cdadar: List Elements
cdaddr: List Elements
cdadr: List Elements
cdar: List Elements
cddaar: List Elements
cddadr: List Elements
cddar: List Elements
cdddar: List Elements
cddddr: List Elements
cdddr: List Elements
cddr: List Elements
cdr: List Elements
cdr-safe: List Elements
ceiling: Numeric Conversions
centering point: Textual Scrolling
change hooks: Change Hooks
change hooks for a character: Special Properties
change load-path at configure time: Building Emacs
change-major-mode-after-body-hook: Mode Hooks
change-major-mode-hook: Creating Buffer-Local
changing key bindings: Changing Key Bindings
changing text properties: Changing Properties
changing to another buffer: Current Buffer
changing window size: Resizing Windows
char in rx: Rx Constructs
char-after: Near Point
char-before: Near Point
char-category-set: Categories
char-charset: Character Sets
char-code-property-description: Character Properties
char-displayable-p: Fontsets
char-equal: Text Comparison
char-from-name: Character Codes
char-or-string-p: Predicates for Strings
char-property-alias-alist: Examining Properties
char-script-table: Character Properties
char-syntax: Syntax Table Functions
char-table length: Sequence Functions
char-table-extra-slot: Char-Tables
char-table-p: Char-Tables
char-table-parent: Char-Tables
char-table-range: Char-Tables
char-table-subtype: Char-Tables
char-tables: Char-Tables
char-to-string: String Conversion
char-uppercase-p: Size of Displayed Text
char-width: Size of Displayed Text
char-width-table: Character Properties
character alternative (in regexp): Regexp Special
character arrays: Strings and Characters
character case: Case Conversion
character categories: Categories
character class in rx: Rx Constructs
character class in rx: Rx Constructs
character classes in regexp: Char Classes
character code conversion: Coding System Basics
character codepoint: Text Representations
character codes: Character Codes
character event: Keyboard Events
character insertion: Commands for Insertion
character printing: Describing Characters
character properties: Character Properties
character set, searching: Scanning Charsets
character sets: Character Sets
character to string: String Conversion
character translation tables: Translation of Characters
character width on display: Size of Displayed Text
characterp: Character Codes
characters: Strings and Characters
characters for interactive codes: Interactive Codes
characters, multi-byte: Non-ASCII Characters
characters, representation in buffers and strings: Text Representations
charset: Character Sets
charset, coding systems to encode: Lisp and Coding Systems
charset, text property on buffer text: Explicit Encoding
charset, text property on strings: Explicit Encoding
charset-after: Scanning Charsets
charset-list: Character Sets
charset-plist: Character Sets
charset-priority-list: Character Sets
charsetp: Character Sets
charsets supported by a coding system: Lisp and Coding Systems
check-coding-system: Lisp and Coding Systems
check-coding-systems-region: Lisp and Coding Systems
check-declare-directory: Declaring Functions
check-declare-file: Declaring Functions
checkdoc: Tips
checkdoc-current-buffer: Tips
checkdoc-file: Tips
checkdoc-minor-mode: Documentation Tips
checkdoc-package-keywords: Library Headers
checkdoc-package-keywords-flag: Library Headers
child frames: Child Frames
child process: Processes
child window: Windows and Frames
child-frame-border-width, a frame parameter: Layout Parameters
child-frame-parameters, a buffer display action alist entry: Buffer Display Action Alists
choice, customization types: Splicing into Lists
cipher, AEAD: GnuTLS Cryptography
cipher, symmetric: GnuTLS Cryptography
circular list: Cons Cells
circular structure, read syntax: Circular Objects
cl: Lisp History
CL note—allocate more storage: Garbage Collection
CL note—case of letters: Symbol Type
CL note—default optional arg: Argument List
CL note—interning existing symbol: Creating Symbols
CL note—lack union, intersection: Sets And Lists
CL note—no continuable errors: Signaling Errors
CL note—no setf functions: Adding Generalized Variables
CL note—only throw in Emacs: Catch and Throw
CL note—rplaca vs setcar: Modifying Lists
CL note—special forms compared: Special Forms
CL note—symbol in obarrays: Creating Symbols
cl-call-next-method: Generic Functions
cl-defgeneric: Generic Functions
cl-defmethod: Generic Functions
cl-next-method-p: Generic Functions
cl-old-struct-compat-mode: Backward Compatibility
classification of file types: Kinds of Files
classifying events: Classifying Events
clean-mode: Major Modes
cleanup forms: Cleanups
clear-abbrev-table: Abbrev Tables
clear-image-cache: Image Cache
clear-message-function: Displaying Messages
clear-string: Modifying Strings
clear-this-command-keys: Command Loop Info
clear-visited-file-modtime: Modification Time
click event: Click Events
clickable buttons in buffers: Buttons
clickable text: Clickable Text
clipboard: Window System Selections
clipboard support (for MS-Windows): Window System Selections
clone-indirect-buffer: Indirect Buffers
clone-of, a window parameter: Window Parameters
CLOS: Generic Functions
closepath: SVG Images
closure: Closures
closures, example of using: Lexical Binding
clrhash: Hash Access
coded character set: Character Sets
codepoint, largest value: Character Codes
codes, interactive, description of: Interactive Codes
codespace: Text Representations
coding conventions in Emacs Lisp: Coding Conventions
coding standards: Tips
coding system: Coding Systems
coding system for operation: Specifying Coding Systems
coding system, automatically determined: Default Coding Systems
coding system, validity check: Lisp and Coding Systems
coding systems for encoding a string: Lisp and Coding Systems
coding systems for encoding region: Lisp and Coding Systems
coding systems, priority: Specifying Coding Systems
coding-system-aliases: Coding System Basics
coding-system-change-eol-conversion: Lisp and Coding Systems
coding-system-change-text-conversion: Lisp and Coding Systems
coding-system-charset-list: Lisp and Coding Systems
coding-system-eol-type: Lisp and Coding Systems
coding-system-for-read: Specifying Coding Systems
coding-system-for-write: Specifying Coding Systems
coding-system-get: Coding System Basics
coding-system-list: Lisp and Coding Systems
coding-system-p: Lisp and Coding Systems
coding-system-priority-list: Specifying Coding Systems
coding-system-require-warning: Specifying Coding Systems
collapse-delayed-warnings: Delayed Warnings
color names: Color Names
color-dark-p: Color Names
color-defined-p: Color Names
color-gray-p: Color Names
color-name-to-rgb: Color Names
color-supported-p: Color Names
color-values: Color Names
colors on text terminals: Text Terminal Colors
column width: Frame Font
column-0: Parser-based Indentation
columns: Columns
COM1: Serial Ports
combine-after-change-calls: Change Hooks
combine-and-quote-strings: Shell Arguments
combine-change-calls: Change Hooks
combining conditions: Combining Conditions
command: What Is a Function
command descriptions: A Sample Function Description
command history: Command History
command in keymap: Key Lookup
command loop: Command Loop
command loop variables: Command Loop Info
command loop, recursive: Recursive Editing
command-completion-default-include-p: Interactive Call
command-debug-status: Internals of Debugger
command-error-function: Processing of Errors
command-execute: Interactive Call
command-history: Command History
command-line: Command-Line Arguments
command-line arguments: Command-Line Arguments
command-line options: Command-Line Arguments
command-line-args: Command-Line Arguments
command-line-args-left: Command-Line Arguments
command-line-functions: Command-Line Arguments
command-line-processed: Command-Line Arguments
command-query: Disabling Commands
command-remapping: Remapping Commands
command-switch-alist: Command-Line Arguments
commandp: Interactive Call
commandp example: High-Level Completion
commands, defining: Defining Commands
commands, mode-specific: Command Modes
commands, specify as mode-specific: Command Modes
comment style: Syntax Flags
comment syntax: Syntax Class Table
comment-auto-fill-only-comments: Auto Filling
comment-end: Parser-based Indentation
comment-end-can-be-escaped: Control Parsing
comment-start: Parser-based Indentation
commentary, in a Lisp library: Library Headers
comments: Comments
comments, Lisp convention for: Comment Tips
Common Lisp: Lisp History
comp-native-version-dir: Library Search
compare tree-sitter syntax nodes: Accessing Node Information
compare-buffer-substrings: Comparing Text
compare-strings: Text Comparison
comparing buffer text: Comparing Text
comparing file modification time: Modification Time
comparing numbers: Comparison of Numbers
comparing time values: Time Calculations
compatibility, between modules and Emacs: Module Initialization
compilation (Emacs Lisp): Byte Compilation
compilation functions: Compilation Functions
compilation to native code (Emacs Lisp): Native Compilation
compile-defun: Compilation Functions
compile-time constant: Eval During Compile
compiled function: Byte-Code Objects
compiled-function-p: What Is a Function
compiler errors: Compiler Errors
compiler macro: Declare Form
compiler macros, advising: Advice and Byte Code
compiling tree-sitter queries: Pattern Matching
complete key: Keymap Basics
completing-read: Minibuffer Completion
completing-read-function: Minibuffer Completion
completion: Completion
completion category: Programmed Completion
completion metadata: Programmed Completion
completion styles: Completion Variables
completion table: Basic Completion
completion table, modifying: Basic Completion
completion tables, combining: Basic Completion
completion, file name: File Name Completion
completion-at-point: Completion in Buffers
completion-at-point-functions: Completion in Buffers
completion-auto-help: Completion Commands
completion-boundaries: Basic Completion
completion-category-overrides: Completion Variables
completion-extra-properties: Completion Variables
completion-ignore-case: Basic Completion
completion-ignored-extensions: File Name Completion
completion-in-region: Completion in Buffers
completion-regexp-list: Basic Completion
completion-styles: Completion Variables
completion-styles-alist: Completion Variables
completion-table-case-fold: Basic Completion
completion-table-dynamic: Programmed Completion
completion-table-in-turn: Basic Completion
completion-table-merge: Basic Completion
completion-table-subvert: Basic Completion
completion-table-with-cache: Programmed Completion
completion-table-with-predicate: Basic Completion
completion-table-with-quoting: Basic Completion
completion-table-with-terminator: Basic Completion
complex arguments: Minibuffers
complex command: Command History
composite type, in bindat specification: Bindat Types
composite types (customization): Composite Types
composition (text property): Special Properties
composition property, and point display: Adjusting Point
compute-motion: Screen Lines
computed documentation string: Function Documentation
concat: Creating Strings
concatenating bidirectional strings: Bidirectional Display
concatenating lists: Rearrangement
concatenating strings: Creating Strings
concurrency: Threads
cond: Conditionals
condition name: Error Symbols
condition-case: Handling Errors
condition-case-unless-debug: Handling Errors
condition-mutex: Condition Variables
condition-name: Condition Variables
condition-notify: Condition Variables
condition-variable-p: Condition Variables
condition-wait: Condition Variables
conditional evaluation: Conditionals
conditional selection of windows: Cyclic Window Ordering
confirm-kill-processes: Query Before Exit
connection local profiles: Connection Local Profiles
connection local variables: Connection Local Variables
connection local variables, applying: Applying Connection Local Variables
connection-local-criteria-alist: Connection Local Profiles
connection-local-default-application: Applying Connection Local Variables
connection-local-get-profile-variables: Connection Local Profiles
connection-local-profile-alist: Connection Local Profiles
connection-local-profile-name-for-setq: Applying Connection Local Variables
connection-local-set-profile-variables: Connection Local Profiles
connection-local-set-profiles: Connection Local Profiles
cons: Building Lists
cons cells: Building Lists
cons-cells-consed: Memory Usage
consing: Building Lists
consp: List-related Predicates
constant variables: Constant Variables
constant variables: Defining Variables
constrain-to-field: Fields
content directory, package: Packaging Basics
context menus, for a major mode: Major Mode Conventions
context-menu-functions: Major Mode Conventions
continuation lines: Truncation
continue-process: Signals to Processes
control character key constants: Low-Level Key Binding
control character printing: Describing Characters
control characters: Ctl-Char Syntax
control characters in display: Usual Display
control characters, reading: Quoted Character Input
control structures: Control Structures
Control-X-prefix: Prefix Keys
controller part, model/view/controller: Abstract Display Example
controlling terminal: Suspending Emacs
controlling-tty-p: Suspending Emacs
conventions for documentation strings: Documentation Tips
conventions for Emacs Lisp programs: Coding Conventions
conventions for Emacs programming: Programming Tips
conventions for key bindings: Key Binding Conventions
conventions for library header comments: Library Headers
conventions for Lisp comments: Comment Tips
conventions for writing major modes: Major Mode Conventions
conventions for writing minor modes: Minor Mode Conventions
conversion of strings: String Conversion
convert buffer position to file byte: Text Representations
convert file byte to buffer position: Text Representations
convert sequence to another type: Sequence Functions
convert-standard-filename: Standard File Names
converting file names from/to MS-Windows syntax: File Names
converting numbers: Numeric Conversions
coordinate, relative to frame: Coordinates and Windows
Coordinated Universal Time: Time of Day
coordinates-in-window-p: Coordinates and Windows
copy-abbrev-table: Abbrev Tables
copy-alist: Association Lists
copy-category-table: Categories
copy-directory: Create/Delete Dirs
copy-file: Changing Files
copy-hash-table: Other Hash
copy-keymap: Creating Keymaps
copy-marker: Creating Markers
copy-overlay: Managing Overlays
copy-region-as-kill: Kill Functions
copy-sequence: Sequence Functions
copy-syntax-table: Syntax Table Functions
copy-tree: Building Lists
copying alists: Association Lists
copying bidirectional text, preserve visual order: Bidirectional Display
copying files: Changing Files
copying lists: Building Lists
copying sequences: Sequence Functions
copying strings: Creating Strings
copying vectors: Vector Functions
copysign: Float Basics
copy_string_contents: Module Values
cos: Math Functions
count-lines: Text Lines
count-loop: A Sample Function Description
count-screen-lines: Screen Lines
count-words: Text Lines
counting columns: Columns
counting set bits: Bitwise Operations
coverage testing: Test Coverage
coverage testing (Edebug): Coverage Testing
create subprocess: Subprocess Creation
create-file-buffer: Subroutines of Visiting
create-fontset-from-fontset-spec: Fontsets
create-image: Defining Images
create-lockfiles: File Locks
creating buffers: Creating Buffers
creating hash tables: Creating Hash
creating keymaps: Creating Keymaps
creating markers: Creating Markers
creating strings: Creating Strings
creating tree-sitter parsers: Using Parser
creating, copying and deleting directories: Create/Delete Dirs
cryptographic hash: Checksum/Hash
cryptographic hash: GnuTLS Cryptography
ctl-arrow: Usual Display
ctl-x-4-map: Prefix Keys
ctl-x-5-map: Prefix Keys
ctl-x-map: Prefix Keys
ctl-x-r-map: Standard Keymaps
curly quotes: Text Quoting Style
curly quotes: Documentation Tips
curly quotes, in formatted messages: Formatting Strings
current binding: Local Variables
current buffer: Current Buffer
current buffer mark: The Mark
current buffer point and mark (Edebug): Edebug Display Update
current buffer position: Point
current command: Command Loop Info
current stack frame: Backtraces
current-active-maps: Active Keymaps
current-bidi-paragraph-direction: Bidirectional Display
current-buffer: Current Buffer
current-case-table: Case Tables
current-column: Columns
current-cpu-time: Time of Day
current-fill-column: Margins
current-frame-configuration: Frame Configurations
current-global-map: Controlling Active Maps
current-idle-time: Idle Timers
current-indentation: Primitive Indent
current-input-method: Input Methods
current-input-mode: Input Modes
current-justification: Filling
current-kill: Low-Level Kill Ring
current-left-margin: Margins
current-local-map: Controlling Active Maps
current-message: Displaying Messages
current-minibuffer-command: Command Loop Info
current-minor-mode-maps: Controlling Active Maps
current-prefix-arg: Prefix Command Arguments
current-thread: Basic Thread Functions
current-time: Time of Day
current-time-list: Time of Day
current-time-string: Time of Day
current-time-zone: Time Zone Rules
current-window-configuration: Window Configurations
current-word: Buffer Contents
currying: Calling Functions
cursor: Window Point
cursor (text property): Special Properties
cursor position for display properties and overlays: Special Properties
cursor, and frame parameters: Cursor Parameters
cursor, fringe: Fringe Cursors
cursor-color, a frame parameter: Font and Color Parameters
cursor-face (text property): Special Properties
cursor-face-highlight-mode: Special Properties
cursor-face-highlight-nonselected-window: Special Properties
cursor-in-echo-area: Echo Area Customization
cursor-in-non-selected-windows: Cursor Parameters
cursor-intangible (text property): Special Properties
cursor-intangible-mode: Special Properties
cursor-sensor-functions (text property): Special Properties
cursor-sensor-inhibit: Special Properties
cursor-sensor-mode: Special Properties
cursor-type: Cursor Parameters
cursor-type: Cursor Parameters
cursor-type, a frame parameter: Cursor Parameters
curved quotes: Text Quoting Style
curved quotes: Documentation Tips
curved quotes, in formatted messages: Formatting Strings
curveto: SVG Images
custom ‘%’-sequence in format: Custom Format Strings
custom format string: Custom Format Strings
custom themes: Custom Themes
custom-add-frequent-value: Variable Definitions
custom-group property: Group Definitions
custom-initialize-delay: Building Emacs
custom-known-themes: Custom Themes
custom-reevaluate-setting: Variable Definitions
custom-set-faces: Applying Customizations
custom-set-variables: Applying Customizations
custom-theme-p: Custom Themes
custom-theme-set-faces: Custom Themes
custom-theme-set-variables: Custom Themes
custom-unlispify-remove-prefixes: Group Definitions
custom-variable-history: Minibuffer History
custom-variable-p: Variable Definitions
customizable variables, how to define: Variable Definitions
customization groups, defining: Group Definitions
customization item: Customization
customization keywords: Common Keywords
customization types: Customization Types
customization types, define new: Defining New Types
customize-package-emacs-version-alist: Common Keywords
customized-face (face symbol property): Defining Faces
cycle-sort-function, in completion: Programmed Completion
cyclic ordering of windows: Cyclic Window Ordering
cygwin-convert-file-name-from-windows: File Names
cygwin-convert-file-name-to-windows: File Names

D
dangling symlinks, testing for existence: Testing Accessibility
data layout specification: Bindat Types
data type: Lisp Data Types
data-directory: Help Functions
database access, SQLite: Database
database object: Database
datagrams: Datagrams
date-days-in-month: Time Calculations
date-leap-year-p: Time Calculations
date-ordinal-to-time: Time Calculations
date-to-time: Time Parsing
days-to-time: Time Calculations
deactivate-mark: The Mark
deactivate-mark: The Mark
deactivate-mark-hook: The Mark
debug: Invoking the Debugger
debug-allow-recursive-debug: Error Debugging
debug-allow-recursive-debug: Debugger Commands
debug-ignored-errors: Error Debugging
debug-on-entry: Function Debugging
debug-on-error: Error Debugging
debug-on-error use: Processing of Errors
debug-on-event: Error Debugging
debug-on-message: Error Debugging
debug-on-next-call: Internals of Debugger
debug-on-quit: Infinite Loops
debug-on-signal: Error Debugging
debug-on-variable-change: Variable Debugging
debugger: Internals of Debugger
debugger command list: Debugger Commands
debugger for Emacs Lisp: Debugger
debugger, explicit entry: Explicit Debug
debugger-bury-or-kill: Using Debugger
debugger-stack-frame-as-list: Internals of Debugger
debugging changes to variables: Variable Debugging
debugging errors: Error Debugging
debugging font-lock: Other Font Lock Variables
debugging invalid Lisp syntax: Syntax Errors
debugging lisp programs: Debugging
debugging redisplay errors: Debugging Redisplay
debugging specific functions: Function Debugging
declare: Declare Form
declare: Declare Form
declare-function: Declaring Functions
declare-function: Declaring Functions
declaring functions: Declaring Functions
decode process output: Decoding Output
decode-char: Character Sets
decode-coding-inserted-region: Explicit Encoding
decode-coding-region: Explicit Encoding
decode-coding-string: Explicit Encoding
decode-time: Time Conversion
decoding file formats: Format Conversion
decoding in coding systems: Explicit Encoding
decrement field of register: Cons Cell Type
dedicated window: Dedicated Windows
dedicated, a buffer display action alist entry: Buffer Display Action Alists
def-edebug-elem-spec: Specification List
def-edebug-spec: Instrumenting Macro Calls
defadvice: Porting Old Advice
defalias: Defining Functions
defalias-fset-function property: Defining Functions
default argument string: Interactive Codes
default character height: Frame Font
default character size: Frame Font
default character width: Frame Font
default coding system: Default Coding Systems
default coding system, functions to determine: Default Coding Systems
default filter function of a process: Filter Functions
default font: Frame Font
default height of character: Frame Font
default init file: Init File
default key binding: Format of Keymaps
default sentinel function of a process: Sentinels
default value: Default Value
default value of char-table: Char-Tables
default width of character: Frame Font
default-boundp: Default Value
default-directory: File Name Expansion
default-file-modes: Changing Files
default-font-height: Low-Level Font
default-font-width: Low-Level Font
default-frame-alist: Initial Parameters
default-input-method: Input Methods
default-justification: Filling
default-minibuffer-frame: Minibuffers and Frames
default-process-coding-system: Default Coding Systems
default-text-properties: Examining Properties
default-toplevel-value: Default Value
default-value: Default Value
default.el: Startup Summary
defconst: Defining Variables
defcustom: Variable Definitions
deferred evaluation: Deferred Eval
defface: Defining Faces
defgroup: Group Definitions
defimage: Defining Images
define customization group: Group Definitions
define customization options: Variable Definitions
define hash comparisons: Defining Hash
define image: Defining Images
define new bindat type forms: Bindat Computed Types
define new customization types: Defining New Types
define-abbrev: Defining Abbrevs
define-abbrev-table: Abbrev Tables
define-advice: Advising Named Functions
define-alternatives: Generic Commands
define-button-type: Button Types
define-category: Categories
define-derived-mode: Derived Modes
define-error: Error Symbols
define-error: Error Symbols
define-fringe-bitmap: Customizing Bitmaps
define-generic-mode: Generic Modes
define-globalized-minor-mode: Defining Minor Modes
define-hash-table-test: Defining Hash
define-icon: Icons
define-inline: Inline Functions
define-key: Low-Level Key Binding
define-key-after: Low-Level Key Binding
define-keymap: Creating Keymaps
define-minor-mode: Defining Minor Modes
define-multisession-variable: Multisession Variables
define-obsolete-face-alias: Face Functions
define-obsolete-function-alias: Obsolete Functions
define-obsolete-variable-alias: Variable Aliases
define-package: Multi-file Packages
define-prefix-command: Prefix Keys
define-short-documentation-group: Documentation Groups
defined-colors: Color Names
defining a function: Defining Functions
defining abbrevs: Defining Abbrevs
defining commands: Defining Commands
defining customization variables in C: Writing Emacs Primitives
defining faces: Defining Faces
defining functions dynamically: Defining Functions
defining Lisp variables in C: Writing Emacs Primitives
defining macros: Defining Macros
defining menus: Defining Menus
defining tokens, SMIE: SMIE Lexer
defining-kbd-macro: Keyboard Macros
definition-prefixes: Autoload by Prefix
definitions of symbols: Definitions
defmacro: Defining Macros
defsubr, Lisp symbol for a primitive: Writing Emacs Primitives
defsubst: Inline Functions
deftheme: Custom Themes
defun: Defining Functions
DEFUN, C macro to define Lisp primitives: Writing Emacs Primitives
defun-prompt-regexp: List Motion
defvar: Defining Variables
defvar-keymap: Creating Keymaps
defvar-local: Creating Buffer-Local
defvaralias: Variable Aliases
DEFVAR_INT, DEFVAR_LISP, DEFVAR_BOOL, DEFSYM: Writing Emacs Primitives
delay-mode-hooks: Mode Hooks
delay-warning: Delayed Warnings
delayed warnings: Delayed Warnings
delayed-warnings-hook: Delayed Warnings
delayed-warnings-hook: Standard Hooks
delayed-warnings-list: Delayed Warnings
delete: Sets And Lists
delete-and-extract-region: Deletion
delete-auto-save-file-if-necessary: Auto-Saving
delete-auto-save-files: Auto-Saving
delete-backward-char: Deletion
delete-before, a frame parameter: Frame Interaction Parameters
delete-blank-lines: User-Level Deletion
delete-by-moving-to-trash: Changing Files
delete-by-moving-to-trash: Create/Delete Dirs
delete-char: Deletion
delete-directory: Create/Delete Dirs
delete-dups: Sets And Lists
delete-exited-processes: Deleting Processes
delete-field: Fields
delete-file: Changing Files
delete-frame: Deleting Frames
delete-frame event: Misc Events
delete-frame-functions: Deleting Frames
delete-horizontal-space: User-Level Deletion
delete-indentation: User-Level Deletion
delete-minibuffer-contents: Minibuffer Contents
delete-old-versions: Numbered Backups
delete-other-frames: Deleting Frames
delete-other-windows: Deleting Windows
delete-other-windows, a window parameter: Window Parameters
delete-overlay: Managing Overlays
delete-process: Deleting Processes
delete-region: Deletion
delete-selection, symbol property: The Mark
delete-selection-helper: The Mark
delete-selection-pre-hook: The Mark
delete-terminal: Multiple Terminals
delete-terminal-functions: Multiple Terminals
delete-to-left-margin: Margins
delete-trailing-whitespace: User-Level Deletion
delete-window: Deleting Windows
delete-window, a window parameter: Window Parameters
delete-window-choose-selected: Deleting Windows
delete-windows-on: Deleting Windows
deleting files: Changing Files
deleting frames: Deleting Frames
deleting list elements: Sets And Lists
deleting previous char: Deletion
deleting processes: Deleting Processes
deleting text vs killing: Deletion
deleting whitespace: User-Level Deletion
deleting windows: Deleting Windows
delq: Sets And Lists
dependencies: Packaging Basics
derived mode: Derived Modes
derived-mode-p: Derived Modes
describe characters and events: Describing Characters
describe-bindings: Scanning Keymaps
describe-buffer-case-table: Case Tables
describe-categories: Categories
describe-current-display-table: Display Tables
describe-display-table: Display Tables
describe-mode: Mode Help
describe-prefix-bindings: Help Functions
describe-syntax: Syntax Table Functions
description for interactive codes: Interactive Codes
description format: Format of Descriptions
deserializing: Byte Packing
desktop notifications: Desktop Notifications
desktop save mode: Desktop Save Mode
desktop-buffer-mode-handlers: Desktop Save Mode
desktop-save-buffer: Desktop Save Mode
destroy-fringe-bitmap: Customizing Bitmaps
destructive list operations: Modifying Lists
destructuring with pcase patterns: Destructuring with pcase Patterns
detect-coding-region: Lisp and Coding Systems
detect-coding-string: Lisp and Coding Systems
deterministic build: Building Emacs
device names: Command Loop Info
device-class: Command Loop Info
diagrams, boxed, for lists: Box Diagrams
dialog boxes: Dialog Boxes
digit-argument: Prefix Command Arguments
ding: Beeping
dir-locals-class-alist: Directory Local Variables
dir-locals-directory-cache: Directory Local Variables
dir-locals-file: Directory Local Variables
dir-locals-set-class-variables: Directory Local Variables
dir-locals-set-directory-class: Directory Local Variables
direct save protocol: Drag and Drop
direction, a buffer display action alist entry: Buffer Display Action Alists
directional overrides: Bidirectional Display
directory file name: Directory Names
directory local variables: Directory Local Variables
directory name: Directory Names
directory part (of file name): File Name Components
directory-abbrev-alist: Directory Names
directory-empty-p: Contents of Directories
directory-file-name: Directory Names
directory-files: Contents of Directories
directory-files-and-attributes: Contents of Directories
directory-files-no-dot-files-regexp: Contents of Directories
directory-files-recursively: Contents of Directories
directory-name-p: Directory Names
directory-oriented functions: Contents of Directories
dired-kept-versions: Numbered Backups
disable asynchronous native compilation: Native-Compilation Variables
disable-command: Disabling Commands
disable-point-adjustment: Adjusting Point
disable-theme: Custom Themes
disabled: Disabling Commands
disabled command: Disabling Commands
disabled-command-function: Disabling Commands
disabling multibyte: Disabling Multibyte
disabling undo: Maintaining Undo
disassemble: Disassembly
disassembled byte-code: Disassembly
discard-input: Event Input Misc
discarding input: Event Input Misc
dispatch of methods for generic function: Generic Functions
display (overlay property): Overlay Properties
display (text property): Display Property
display action: Choosing Window
display area: Frame Layout
display feature testing: Display Feature Testing
display margins: Display Margins
display message in echo area: Displaying Messages
display name on X: Multiple Terminals
display origin: Frame Layout
display properties, and bidi reordering of text: Bidirectional Display
display property: Display Property
display property, and point display: Adjusting Point
display property, unsafe evaluation: Display Property
display specification: Display Property
display table: Display Tables
display, a frame parameter: Basic Parameters
display, abstract: Abstract Display
display, arbitrary objects: Abstract Display
display-backing-store: Display Feature Testing
display-buffer: Choosing Window
display-buffer-alist: Choosing Window
display-buffer-at-bottom: Buffer Display Action Functions
display-buffer-base-action: Choosing Window
display-buffer-below-selected: Buffer Display Action Functions
display-buffer-fallback-action: Choosing Window
display-buffer-full-frame: Buffer Display Action Functions
display-buffer-in-atom-window: Atomic Windows
display-buffer-in-child-frame: Buffer Display Action Functions
display-buffer-in-direction: Buffer Display Action Functions
display-buffer-in-previous-window: Buffer Display Action Functions
display-buffer-in-side-window: Displaying Buffers in Side Windows
display-buffer-no-window: Buffer Display Action Functions
display-buffer-overriding-action: Choosing Window
display-buffer-pop-up-frame: Buffer Display Action Functions
display-buffer-pop-up-window: Buffer Display Action Functions
display-buffer-reuse-mode-window: Buffer Display Action Functions
display-buffer-reuse-window: Buffer Display Action Functions
display-buffer-same-window: Buffer Display Action Functions
display-buffer-use-least-recent-window: Buffer Display Action Functions
display-buffer-use-some-frame: Buffer Display Action Functions
display-buffer-use-some-window: Buffer Display Action Functions
display-color-cells: Display Feature Testing
display-color-p: Display Feature Testing
display-completion-list: Completion Commands
display-delayed-warnings: Delayed Warnings
display-graphic-p: Display Feature Testing
display-grayscale-p: Display Feature Testing
display-images-p: Display Feature Testing
display-message-or-buffer: Displaying Messages
display-mm-dimensions-alist: Display Feature Testing
display-mm-height: Display Feature Testing
display-mm-width: Display Feature Testing
display-monitor-attributes-list: Multiple Terminals
display-monitors-changed-functions: Multiple Terminals
display-mouse-p: Display Feature Testing
display-pixel-height: Display Feature Testing
display-pixel-width: Display Feature Testing
display-planes: Display Feature Testing
display-popup-menus-p: Display Feature Testing
display-save-under: Display Feature Testing
display-screens: Display Feature Testing
display-selections-p: Display Feature Testing
display-sort-function, in completion: Programmed Completion
display-start position: Window Start and End
display-supports-face-attributes-p: Display Feature Testing
display-table-slot: Display Tables
display-type, a frame parameter: Basic Parameters
display-visual-class: Display Feature Testing
display-warning: Warning Basics
displaying a buffer: Displaying Buffers
displaying faces: Displaying Faces
displays, multiple: Multiple Terminals
distance between strings: Text Comparison
distinguish interactive calls: Distinguish Interactive
dlet: Local Variables
dnd-begin-drag-files: Drag and Drop
dnd-begin-file-drag: Drag and Drop
dnd-begin-text-drag: Drag and Drop
dnd-direct-save: Drag and Drop
dnd-protocol-alist: Drag and Drop
do-auto-save: Auto-Saving
DOC (documentation) file: Documentation Basics
doc, customization keyword: Type Keywords
doc-directory: Accessing Documentation
Document Object Model: Document Object Model
documentation: Accessing Documentation
documentation conventions: Documentation Basics
documentation for major mode: Mode Help
documentation groups: Documentation Groups
documentation notation: Evaluation Notation
documentation string of function: Function Documentation
documentation strings: Documentation
documentation strings, conventions and tips: Documentation Tips
documentation, keys in: Keys in Documentation
documentation-property: Accessing Documentation
dolist: Iteration
dolist-with-progress-reporter: Progress
DOM: Document Object Model
dom-node: Document Object Model
dotimes: Iteration
dotimes-with-progress-reporter: Progress
dotted list: Cons Cells
dotted lists (Edebug): Specification List
dotted pair notation: Dotted Pair Notation
double-click events: Repeat Events
double-click-fuzz: Repeat Events
double-click-time: Repeat Events
double-quote in strings: Syntax for Strings
down-list: List Motion
downcase: Case Conversion
downcase-region: Case Changes
downcase-word: Case Changes
downcasing in lookup-key: Key Sequence Input
download-callback xwidget events: Xwidget Events
download-started xwidget events: Xwidget Events
drag and drop: Drag and Drop
drag and drop protocols, X: Drag and Drop
drag and drop, other formats: Drag and Drop
drag and drop, X: Drag and Drop
drag event: Drag Events
drag-internal-border, a frame parameter: Mouse Dragging Parameters
drag-n-drop event: Misc Events
drag-with-header-line, a frame parameter: Mouse Dragging Parameters
drag-with-mode-line, a frame parameter: Mouse Dragging Parameters
drag-with-tab-line, a frame parameter: Mouse Dragging Parameters
dribble file: Recording Input
drop target, in drag-and-drop operations: Drag and Drop
dump file: Building Emacs
dump-emacs: Building Emacs
dump-emacs-portable: Building Emacs
dumping Emacs: Building Emacs
dynamic binding: Variable Scoping
dynamic binding, temporarily: Local Variables
dynamic extent: Variable Scoping
dynamic let-binding: Local Variables
dynamic libraries: Dynamic Libraries
dynamic loading of documentation: Docs and Compilation
dynamic loading of functions: Dynamic Loading
dynamic modules: Dynamic Modules
dynamic modules, writing: Writing Dynamic Modules
dynamic scope: Variable Scoping
dynamic-library-alist: Dynamic Libraries

E
eager macro expansion: How Programs Do Loading
early init file: Init File
early-init.el: Init File
easy-menu-define: Easy Menu
easy-mmode-define-minor-mode: Defining Minor Modes
echo area: The Echo Area
echo area customization: Echo Area Customization
echo-area-clear-hook: Echo Area Customization
echo-keystrokes: Echo Area Customization
edebug: Source Breakpoints
Edebug debugging facility: Edebug
Edebug execution modes: Edebug Execution Modes
edebug overwrites buffer point position: Edebug Display Update
Edebug specification list: Specification List
edebug, failure to instrument: Instrumenting
edebug-after-instrumentation-function: Edebug Options
edebug-all-defs: Edebug Options
edebug-all-forms: Edebug Options
edebug-backtrace-hide-instrumentation: Edebug Misc
edebug-backtrace-show-instrumentation: Edebug Misc
edebug-behavior-alist: Edebug Options
edebug-continue-kbd-macro: Edebug Options
edebug-defun: Instrumenting
edebug-display-freq-count: Coverage Testing
edebug-eval-macro-args: Instrumenting Macro Calls
edebug-eval-macro-args: Edebug Options
edebug-eval-top-level-form: Instrumenting
edebug-global-break-condition: Edebug Options
edebug-initial-mode: Edebug Options
edebug-max-depth: Checking Whether to Stop
edebug-new-definition-function: Edebug Options
edebug-on-error: Edebug Options
edebug-on-quit: Edebug Options
edebug-print-circle: Printing in Edebug
edebug-print-circle: Edebug Options
edebug-print-length: Printing in Edebug
edebug-print-length: Edebug Options
edebug-print-level: Printing in Edebug
edebug-print-level: Edebug Options
edebug-print-trace-after: Trace Buffer
edebug-print-trace-before: Trace Buffer
edebug-remove-instrumentation: Instrumenting
edebug-save-displayed-buffer-points: Edebug Options
edebug-save-windows: Edebug Options
edebug-set-global-break-condition: Global Break Condition
edebug-set-initial-mode: Edebug Execution Modes
edebug-setup-hook: Edebug Options
edebug-sit-for-seconds: Edebug Execution Modes
edebug-sit-for-seconds: Edebug Options
edebug-sit-on-break: Edebug Options
edebug-temp-display-freq-count: Coverage Testing
edebug-test-coverage: Edebug Options
edebug-trace: Edebug Options
edebug-trace: Trace Buffer
edebug-tracing: Trace Buffer
edebug-unwrap-results: Edebug Options
edge detection, images: Image Descriptors
edit distance between strings: Text Comparison
edit-and-eval-command: Object from Minibuffer
editing types: Editing Types
editor command loop: Command Loop
eight-bit, a charset: Character Sets
electric-future-map: A Sample Variable Description
element (of list): Lists
elements of sequences: Sequence Functions
elliptical-arc: SVG Images
elp.el: Profiling
elt: Sequence Functions
Emacs event standard notation: Describing Characters
Emacs process run time: Processor Run Time
emacs, a charset: Character Sets
emacs-build-number: Version Info
emacs-build-time: Version Info
emacs-init-time: Processor Run Time
emacs-internal coding system: Coding System Basics
emacs-lisp-docstring-fill-column: Documentation Tips
emacs-lisp-native-compile: Native-Compilation Functions
emacs-lisp-native-compile-and-load: Native-Compilation Functions
emacs-major-version: Version Info
emacs-minor-version: Version Info
emacs-pid: System Environment
emacs-repository-branch: Version Info
emacs-repository-version: Version Info
emacs-save-session-functions: Session Management
emacs-session-restore: Session Management
emacs-startup-hook: Init File
emacs-uptime: Processor Run Time
emacs-version: Version Info
emacs-version: Version Info
emacsclient, getting a backtrace: Error Debugging
EMACSLOADPATH environment variable: Library Search
emacs_finalizer: Module Functions
emacs_funcall_exit, enumeration: Module Nonlocal
emacs_funcall_exit_return: Module Nonlocal
emacs_funcall_exit_signal: Module Nonlocal
emacs_funcall_exit_throw: Module Nonlocal
emacs_function: Module Functions
EMACS_LIMB_MAX: Module Values
emacs_module_init: Dynamic Modules
emacs_module_init: Module Initialization
emacs_value data type: Module Values
emacs_variadic_function: Module Functions
embedded language, tree-sitter: Multiple Languages
embedded widgets: Xwidgets
empty file name: Relative File Names
empty file name, and file-exists-p: Testing Accessibility
empty file names, and expand-file-name: File Name Expansion
empty lines, indicating: Fringe Indicators
empty list: Box Diagrams
empty overlay: Managing Overlays
empty region: The Region
emulation-mode-map-alists: Controlling Active Maps
enable-command: Disabling Commands
enable-connection-local-variables: Applying Connection Local Variables
enable-dir-local-variables: Directory Local Variables
enable-local-eval: File Local Variables
enable-local-variables: File Local Variables
enable-multibyte-characters: Text Representations
enable-multibyte-characters: Disabling Multibyte
enable-recursive-minibuffers: Recursive Mini
enable-theme: Custom Themes
encapsulation, ewoc: Abstract Display
encode-char: Character Sets
encode-coding-region: Explicit Encoding
encode-coding-string: Explicit Encoding
encode-time: Time Conversion
encoding file formats: Format Conversion
encoding in coding systems: Explicit Encoding
encrypted network connections: Network
end of line in regexp: Regexp Special
end-of-buffer: Buffer End Motion
end-of-defun: List Motion
end-of-defun-function: List Motion
end-of-file: Input Functions
end-of-line: Text Lines
end-of-line conversion: Coding System Basics
end-session event: Misc Events
endianness, in bindat specification: Bindat Types
ensure-empty-lines: Commands for Insertion
ensure-list: Building Lists
environment: Intro Eval
environment variable access: System Environment
environment variables, subprocesses: Subprocess Creation
eobp: Near Point
EOL conversion: Coding System Basics
eol conversion of coding system: Lisp and Coding Systems
eol in rx: Rx Constructs
eol type of coding system: Lisp and Coding Systems
eolp: Near Point
eos in rx: Rx Constructs
eot in rx: Rx Constructs
eow in rx: Rx Constructs
epoch: Time of Day
eq: Equality Predicates
eq: Module Misc
eql: Comparison of Numbers
equal: Equality Predicates
equal-including-properties: Equality Predicates
equality: Equality Predicates
erase-buffer: Deletion
error: Signaling Errors
error cleanup: Cleanups
error debugging: Error Debugging
error description: Handling Errors
error display: The Echo Area
error handler: Handling Errors
error in debug: Invoking the Debugger
error message notation: Error Messages
error name: Error Symbols
error symbol: Error Symbols
error-conditions: Error Symbols
error-message-string: Handling Errors
errors: Errors
ESC: Functions for Key Lookup
esc-map: Prefix Keys
ESC-prefix: Prefix Keys
escape (ASCII character): Basic Char Syntax
escape characters: Output Variables
escape characters in printing: Output Functions
escape sequence: Basic Char Syntax
escaping apostrophe and grave accent in doc strings: Documentation Tips
eval: Eval
eval during compilation: Eval During Compile
eval in rx: Rx Constructs
eval, and debugging: Internals of Debugger
eval-and-compile: Eval During Compile
eval-buffer: Eval
eval-buffer (Edebug): Instrumenting
eval-defun (Edebug): Instrumenting
eval-defun, and defcustom forms: Variable Definitions
eval-defun, and defface forms: Defining Faces
eval-defun, and defvar forms: Defining Variables
eval-defun, and explicit entry to debugger: Explicit Debug
eval-expression (Edebug): Instrumenting
eval-expression, and lexical-binding: Using Lexical Binding
eval-expression-debug-on-error: Error Debugging
eval-expression-print-length: Output Variables
eval-expression-print-level: Output Variables
eval-last-sexp, and defface forms: Defining Faces
eval-last-sexp, and defvar forms: Defining Variables
eval-minibuffer: Object from Minibuffer
eval-region: Eval
eval-region (Edebug): Instrumenting
eval-when-compile: Eval During Compile
evaluated expression argument: Interactive Codes
evaluation: Evaluation
evaluation list group: Eval List
evaluation notation: Evaluation Notation
evaluation of buffer contents: Eval
evaluation of special forms: Special Forms
evaporate (overlay property): Overlay Properties
even-window-sizes: Choosing Window Options
event printing: Describing Characters
event translation: Event Mod
event type: Classifying Events
event, reading only one: Reading One Event
event-basic-type: Classifying Events
event-click-count: Repeat Events
event-convert-list: Low-Level Key Binding
event-end: Accessing Mouse
event-modifiers: Classifying Events
event-start: Accessing Mouse
eventp: Input Events
events: Input Events
ewoc: Abstract Display
ewoc-buffer: Abstract Display Functions
ewoc-collect: Abstract Display Functions
ewoc-create: Abstract Display Functions
ewoc-data: Abstract Display Functions
ewoc-delete: Abstract Display Functions
ewoc-enter-after: Abstract Display Functions
ewoc-enter-before: Abstract Display Functions
ewoc-enter-first: Abstract Display Functions
ewoc-enter-last: Abstract Display Functions
ewoc-filter: Abstract Display Functions
ewoc-get-hf: Abstract Display Functions
ewoc-goto-next: Abstract Display Functions
ewoc-goto-node: Abstract Display Functions
ewoc-goto-prev: Abstract Display Functions
ewoc-invalidate: Abstract Display Functions
ewoc-locate: Abstract Display Functions
ewoc-location: Abstract Display Functions
ewoc-map: Abstract Display Functions
ewoc-next: Abstract Display Functions
ewoc-nth: Abstract Display Functions
ewoc-prev: Abstract Display Functions
ewoc-refresh: Abstract Display Functions
ewoc-set-data: Abstract Display Functions
ewoc-set-hf: Abstract Display Functions
examining text properties: Examining Properties
examining the interactive form: Using Interactive
examining windows: Buffers and Windows
examples of using interactive: Interactive Examples
excess close parentheses: Excess Close
excess open parentheses: Excess Open
excursion: Excursions
exec-directory: Subprocess Creation
exec-path: Subprocess Creation
exec-path: Subprocess Creation
exec-suffixes: Subprocess Creation
executable-find: Locating Files
execute program: Subprocess Creation
execute with prefix argument: Interactive Call
execute-extended-command: Interactive Call
execute-extended-command-for-buffer: Interactive Call
execute-kbd-macro: Keyboard Macros
executing-kbd-macro: Keyboard Macros
execution order of buffer display action functions: Precedence of Action Functions
execution speed: Compilation Tips
exit: Recursive Editing
exit recursive editing: Recursive Editing
exit-minibuffer: Minibuffer Commands
exit-recursive-edit: Recursive Editing
exiting Emacs: Getting Out
exp: Math Functions
expand-abbrev: Abbrev Expansion
expand-file-name: File Name Expansion
expanding abbrevs: Abbrev Expansion
expansion of file names: File Name Expansion
expansion of macros: Expansion
explicit selective display: Selective Display
explicit-name, a frame parameter: Basic Parameters
explore tree-sitter syntax tree: Language Grammar
expression: Intro Eval
expt: Math Functions
extended file attributes: Extended Attributes
extended menu item: Extended Menu Items
extended-command-history: Minibuffer History
extent: Variable Scoping
external border: Frame Layout
external menu bar: Frame Layout
external tool bar: Frame Layout
external-debugging-output: Output Streams
extra node, tree-sitter: Accessing Node Information
extra slots of char-table: Char-Tables
extra-keyboard-modifiers: Event Mod
extract_big_integer: Module Values
extract_float: Module Values
extract_integer: Module Values
extract_time: Module Values

F
face (button property): Button Properties
face (non-removability of): Defining Faces
face (overlay property): Overlay Properties
face (text property): Special Properties
face alias: Face Functions
face attributes: Face Attributes
face attributes, access and modification: Attribute Functions
face codes of text: Special Properties
face merging: Displaying Faces
face name: Faces
face number: Face Functions
face property of face symbols: Face Functions
face remapping: Face Remapping
face spec: Defining Faces
face-all-attributes: Attribute Functions
face-attribute: Attribute Functions
face-attribute-relative-p: Attribute Functions
face-background: Attribute Functions
face-bold-p: Attribute Functions
face-defface-spec (face symbol property): Defining Faces
face-differs-from-default-p: Face Functions
face-documentation: Accessing Documentation
face-documentation: Face Functions
face-documentation (face symbol property): Defining Faces
face-equal: Face Functions
face-extend-p: Attribute Functions
face-filters-always-match: Special Properties
face-font: Attribute Functions
face-font-family-alternatives: Font Selection
face-font-registry-alternatives: Font Selection
face-font-rescale-alist: Font Selection
face-font-selection-order: Font Selection
face-foreground: Attribute Functions
face-id: Face Functions
face-inverse-video-p: Attribute Functions
face-italic-p: Attribute Functions
face-list: Face Functions
face-name-history: Minibuffer History
face-remap-add-relative: Face Remapping
face-remap-remove-relative: Face Remapping
face-remap-reset-base: Face Remapping
face-remap-set-base: Face Remapping
face-remapping-alist: Face Remapping
face-spec-set: Defining Faces
face-stipple: Attribute Functions
face-underline-p: Attribute Functions
facep: Faces
faces: Faces
faces for font lock: Faces for Font Lock
faces, automatic choice: Auto Faces
false: nil and t
fboundp: Function Cells
fceiling: Rounding Operations
feature-unload-function: Unloading
featurep: Named Features
features: Named Features
features: Named Features
fetch-bytecode: Dynamic Loading
ffloor: Rounding Operations
field (overlay property): Overlay Properties
field (text property): Special Properties
field name, tree-sitter: Language Grammar
field numbers in format spec: Formatting Strings
field width: Formatting Strings
field-beginning: Fields
field-end: Fields
field-is: Parser-based Indentation
field-string: Fields
field-string-no-properties: Fields
fields: Fields
fifo data structure: Rings
file accessibility: Testing Accessibility
file age: File Attributes
file attributes: File Attributes
file classification: Kinds of Files
file contents, and default coding system: Default Coding Systems
file format conversion: Format Conversion
file hard link: Changing Files
file local variables: File Local Variables
file local variables: Security Considerations
file locks: File Locks
file mode specification error: Auto Major Mode
file modes: Testing Accessibility
file modes and MS-DOS: Testing Accessibility
file modes, setting: Changing Files
file modification time: File Attributes
file name abbreviations: Directory Names
file name completion subroutines: File Name Completion
file name handler: Magic File Names
file name of buffer: Buffer File Name
file name of directory: Directory Names
file name, and default coding system: Default Coding Systems
file names: File Names
file names in directory: Contents of Directories
file names, trailing whitespace: Information about Files
file notifications: File Notifications
file open error: Subroutines of Visiting
file permissions: Testing Accessibility
file permissions, setting: Changing Files
file with multiple names: Changing Files
file, ancestor directory of: Contents of Directories
file, information about: Information about Files
file-accessible-directory-p: Testing Accessibility
file-acl: Extended Attributes
file-already-exists: Changing Files
file-attributes: File Attributes
file-backup-file-names: Backup Names
file-chase-links: Truenames
file-coding-system-alist: Default Coding Systems
file-directory-p: Kinds of Files
file-equal-p: Truenames
file-error: How Programs Do Loading
file-executable-p: Testing Accessibility
file-exists-p: Testing Accessibility
file-expand-wildcards: Contents of Directories
file-extended-attributes: Extended Attributes
file-has-changed-p: File Attributes
file-in-directory-p: Contents of Directories
file-local-copy: Magic File Names
file-local-name: Magic File Names
file-local-variables-alist: File Local Variables
file-locked: File Locks
file-locked-p: File Locks
file-modes: Testing Accessibility
file-modes-number-to-symbolic: Changing Files
file-modes-symbolic-to-number: Changing Files
file-name encoding, MS-Windows: Encoding and I/O
file-name-absolute-p: Relative File Names
file-name-all-completions: File Name Completion
file-name-as-directory: Directory Names
file-name-base: File Name Components
file-name-case-insensitive-p: Truenames
file-name-coding-system: Encoding and I/O
file-name-completion: File Name Completion
file-name-concat: Directory Names
file-name-directory: File Name Components
file-name-extension: File Name Components
file-name-handler-alist: Magic File Names
file-name-history: Minibuffer History
file-name-nondirectory: File Name Components
file-name-parent-directory: Directory Names
file-name-quote: File Name Expansion
file-name-quoted-p: File Name Expansion
file-name-sans-extension: File Name Components
file-name-sans-versions: File Name Components
file-name-split: File Name Components
file-name-unquote: File Name Expansion
file-name-with-extension: File Name Components
file-newer-than-file-p: File Attributes
file-newest-backup: Backup Names
file-nlinks: File Attributes
file-notify-add-watch: File Notifications
file-notify-rm-all-watches: File Notifications
file-notify-rm-watch: File Notifications
file-notify-valid-p: File Notifications
file-ownership-preserved-p: Testing Accessibility
file-precious-flag: Saving Buffers
file-readable-p: Testing Accessibility
file-regular-p: Kinds of Files
file-relative-name: Relative File Names
file-remote-p: Magic File Names
file-selinux-context: Extended Attributes
file-supersession: Modification Time
file-symlink-p: Kinds of Files
file-truename: Truenames
file-writable-p: Testing Accessibility
filepos-to-bufferpos: Text Representations
fill-column: Margins
fill-context-prefix: Adaptive Fill
fill-forward-paragraph-function: Filling
fill-individual-paragraphs: Filling
fill-individual-varying-indent: Filling
fill-nobreak-predicate: Margins
fill-paragraph: Filling
fill-paragraph-function: Filling
fill-prefix: Margins
fill-region: Filling
fill-region-as-paragraph: Filling
fill-separate-heterogeneous-words-with-space: Filling
fillarray: Array Functions
filling text: Filling
filling, automatic: Auto Filling
filter function: Filter Functions
filter multibyte flag, of process: Decoding Output
filter-buffer-substring: Buffer Contents
filter-buffer-substring-function: Buffer Contents
filter-buffer-substring-functions: Buffer Contents
filtering killed text: Kill Functions
filtering sequences: Sequence Functions
find file in path: Locating Files
find library: Library Search
find-auto-coding: Default Coding Systems
find-backup-file-name: Backup Names
find-buffer-visiting: Buffer File Name
find-charset-region: Scanning Charsets
find-charset-string: Scanning Charsets
find-coding-systems-for-charsets: Lisp and Coding Systems
find-coding-systems-region: Lisp and Coding Systems
find-coding-systems-string: Lisp and Coding Systems
find-file: Visiting Functions
find-file-hook: Visiting Functions
find-file-literally: Visiting Functions
find-file-literally: Visiting Functions
find-file-name-handler: Magic File Names
find-file-noselect: Visiting Functions
find-file-not-found-functions: Visiting Functions
find-file-other-window: Visiting Functions
find-file-read-only: Visiting Functions
find-file-wildcards: Visiting Functions
find-font: Low-Level Font
find-image: Defining Images
find-operation-coding-system: Default Coding Systems
find-word-boundary-function-table: Word Motion
finding files: Visiting Files
finding windows: Cyclic Window Ordering
first-change-hook: Change Hooks
first-sibling: Parser-based Indentation
fit-frame-to-buffer: Resizing Windows
fit-frame-to-buffer: Resizing Windows
fit-frame-to-buffer-margins: Resizing Windows
fit-frame-to-buffer-margins, a frame parameter: Size Parameters
fit-frame-to-buffer-sizes: Resizing Windows
fit-frame-to-buffer-sizes, a frame parameter: Size Parameters
fit-window-to-buffer: Resizing Windows
fit-window-to-buffer-horizontally: Resizing Windows
fixed-size window: Window Sizes
fixnump: Predicates on Numbers
fixup-whitespace: User-Level Deletion
flags in format specifications: Formatting Strings
flatten-tree: Building Lists
float: Numeric Conversions
float-e: Math Functions
float-output-format: Output Variables
float-pi: Math Functions
float-time: Time of Day
floating-point functions: Math Functions
floatp: Predicates on Numbers
floats-consed: Memory Usage
floor: Numeric Conversions
flowcontrol, in serial connections: Serial Ports
flush-standard-output: Output Functions
flushing input: Event Input Misc
fmakunbound: Function Cells
fn in function’s documentation string: Autoload
focus event: Focus Events
focus-follows-mouse: Input Focus
focus-in-hook: Standard Hooks
focus-out-hook: Standard Hooks
follow links: Clickable Text
follow-link (button property): Button Properties
follow-link (text or overlay property): Clickable Text
following-char: Near Point
font and color, frame parameters: Font and Color Parameters
font backend: Low-Level Font
font entity: Low-Level Font
font information for layout: Low-Level Font
font lock faces: Faces for Font Lock
Font Lock mode: Font Lock Mode
font lookup: Font Lookup
font object: Low-Level Font
font property: Low-Level Font
font registry: Low-Level Font
font selection: Font Selection
font spec: Low-Level Font
font, a frame parameter: Font and Color Parameters
font-at: Low-Level Font
font-backend, a frame parameter: Font and Color Parameters
font-face-attributes: Low-Level Font
font-family-list: Face Attributes
font-get: Low-Level Font
font-info: Low-Level Font
font-lock-add-keywords: Customizing Keywords
font-lock-bracket-face: Faces for Font Lock
font-lock-builtin-face: Faces for Font Lock
font-lock-comment-delimiter-face: Faces for Font Lock
font-lock-comment-face: Faces for Font Lock
font-lock-constant-face: Faces for Font Lock
font-lock-debug-fontify: Font Lock Basics
font-lock-defaults: Font Lock Basics
font-lock-delimiter-face: Faces for Font Lock
font-lock-doc-face: Faces for Font Lock
font-lock-doc-markup-face: Faces for Font Lock
font-lock-ensure &optional beg end: Font Lock Basics
font-lock-ensure-function: Other Font Lock Variables
font-lock-escape-face: Faces for Font Lock
font-lock-extend-after-change-region-function: Region to Refontify
font-lock-extra-managed-props: Other Font Lock Variables
font-lock-face (text property): Special Properties
font-lock-flush &optional beg end: Font Lock Basics
font-lock-flush-function: Other Font Lock Variables
font-lock-fontify-buffer: Font Lock Basics
font-lock-fontify-buffer-function: Other Font Lock Variables
font-lock-fontify-region beg end &optional loudly: Font Lock Basics
font-lock-fontify-region-function: Other Font Lock Variables
font-lock-function-call-face: Faces for Font Lock
font-lock-function-name-face: Faces for Font Lock
font-lock-ignore: Customizing Keywords
font-lock-keyword-face: Faces for Font Lock
font-lock-keywords: Search-based Fontification
font-lock-keywords-case-fold-search: Search-based Fontification
font-lock-keywords-only: Syntactic Font Lock
font-lock-mark-block-function: Other Font Lock Variables
font-lock-misc-punctuation-face: Faces for Font Lock
font-lock-multiline: Font Lock Multiline
font-lock-negation-char-face: Faces for Font Lock
font-lock-number-face: Faces for Font Lock
font-lock-operator-face: Faces for Font Lock
font-lock-preprocessor-face: Faces for Font Lock
font-lock-property-name-face: Faces for Font Lock
font-lock-property-use-face: Faces for Font Lock
font-lock-punctuation-face: Faces for Font Lock
font-lock-remove-keywords: Customizing Keywords
font-lock-string-face: Faces for Font Lock
font-lock-syntactic-face-function: Syntactic Font Lock
font-lock-syntax-table: Syntactic Font Lock
font-lock-type-face: Faces for Font Lock
font-lock-unfontify-buffer: Font Lock Basics
font-lock-unfontify-buffer-function: Other Font Lock Variables
font-lock-unfontify-region beg end: Font Lock Basics
font-lock-unfontify-region-function: Other Font Lock Variables
font-lock-variable-name-face: Faces for Font Lock
font-lock-variable-use-face: Faces for Font Lock
font-lock-warning-face: Faces for Font Lock
font-put: Low-Level Font
font-spec: Low-Level Font
font-xlfd-name: Low-Level Font
fontification-functions: Auto Faces
fontifications with tree-sitter, overview: Parser-based Font Lock
fontified (text property): Special Properties
fontp: Low-Level Font
fontset: Fontsets
foo: A Sample Function Description
for: Argument Evaluation
force coding system for operation: Specifying Coding Systems
force entry to debugger: Explicit Debug
force-mode-line-update: Mode Line Basics
force-window-update: Forcing Redisplay
forcing redisplay: Forcing Redisplay
foreground-color, a frame parameter: Font and Color Parameters
form: Intro Eval
form, self-evaluating: Self-Evaluating Forms
format: Formatting Strings
format definition: Format Conversion Round-Trip
format of gnutls cryptography inputs: Format of GnuTLS Cryptography Inputs
format of keymaps: Format of Keymaps
format specification: Formatting Strings
format, customization keyword: Type Keywords
format-alist: Format Conversion Round-Trip
format-find-file: Format Conversion Round-Trip
format-insert-file: Format Conversion Round-Trip
format-message: Formatting Strings
format-mode-line: Emulating Mode Line
format-network-address: Misc Network
format-prompt: Text from Minibuffer
format-seconds: Time Parsing
format-spec: Custom Format Strings
format-time-string: Time Parsing
format-write-file: Format Conversion Round-Trip
formatting numbers for rereading later: Formatting Strings
formatting strings: Formatting Strings
formatting time values: Time Parsing
formfeed: Basic Char Syntax
forms for cleanup: Cleanups
forms for control structures: Control Structures
forms for handling errors: Handling Errors
forms for nonlocal exits: Catch and Throw
forms for sequential execution: Sequencing
forms, backquote: Backquote
forms, conditional: Conditionals
forms, function call: Function Forms
forms, iteration: Iteration
forms, list: Classifying Lists
forms, macro call: Macro Forms
forms, quote: Quoting
forms, special: Special Forms
forms, symbol: Symbol Forms
forward-button: Button Buffer Commands
forward-char: Character Motion
forward-comment: Motion via Parsing
forward-line: Text Lines
forward-list: List Motion
forward-sexp: List Motion
forward-to-indentation: Motion by Indent
forward-word: Word Motion
forward-word-strictly: Word Motion
frame: Frames
frame configuration: Frame Configurations
frame creation: Creating Frames
frame geometry: Frame Geometry
frame height ratio: Size Parameters
frame interaction parameters: Frame Interaction Parameters
frame layout: Frame Layout
frame layout parameters: Layout Parameters
frame parameters: Frame Parameters
frame parameters for windowed displays: Window Frame Parameters
frame position: Frame Geometry
frame position: Frame Position
frame position: Position Parameters
frame position changes, a hook: Frame Position
frame size: Frame Geometry
frame size: Frame Size
frame stacking order: Raising and Lowering
frame title: Frame Titles
frame visibility: Visibility of Frames
frame width ratio: Size Parameters
frame without a minibuffer: Minibuffers and Frames
frame Z-order: Raising and Lowering
frame, which buffers to display: Buffer Parameters
frame-alpha-lower-limit: Font and Color Parameters
frame-ancestor-p: Child Frames
frame-auto-hide-function: Quitting Windows
frame-char-height: Frame Font
frame-char-width: Frame Font
frame-current-scroll-bars: Scroll Bars
frame-edges: Frame Layout
frame-first-window: Windows and Frames
frame-focus-state: Input Focus
frame-geometry: Frame Layout
frame-height: Frame Size
frame-inherited-parameters: Creating Frames
frame-inhibit-implied-resize: Implied Frame Resizing
frame-inner-height: Frame Size
frame-inner-width: Frame Size
frame-list: Finding All Frames
frame-list-z-order: Finding All Frames
frame-live-p: Deleting Frames
frame-monitor-attributes: Multiple Terminals
frame-native-height: Frame Size
frame-native-width: Frame Size
frame-old-selected-window: Window Hooks
frame-outer-height: Frame Size
frame-outer-width: Frame Size
frame-parameter: Parameter Access
frame-parameters: Parameter Access
frame-parent: Child Frames
frame-pointer-visible-p: Mouse Position
frame-position: Frame Position
frame-predicate, a buffer display action alist entry: Buffer Display Action Alists
frame-relative coordinate: Coordinates and Windows
frame-resize-pixelwise: Frame Size
frame-restack: Raising and Lowering
frame-root-window: Windows and Frames
frame-scroll-bar-height: Scroll Bars
frame-scroll-bar-width: Scroll Bars
frame-selected-window: Selecting Windows
frame-size-changed-p: Frame Size
frame-terminal: Frames
frame-text-height: Frame Size
frame-text-width: Frame Size
frame-title-format: Frame Titles
frame-visible-p: Visibility of Frames
frame-width: Frame Size
frame-window-state-change: Window Hooks
framep: Frames
frames, scanning all: Finding All Frames
frameset: Registers
free list: Garbage Collection
free variable: Converting to Lexical Binding
free variable, byte-compiler warning: Compiler Errors
free_global_ref: Module Values
frequency counts: Coverage Testing
frexp: Float Basics
fringe bitmaps: Fringe Bitmaps
fringe bitmaps, customizing: Customizing Bitmaps
fringe cursors: Fringe Cursors
fringe indicators: Fringe Indicators
fringe-bitmaps-at-pos: Fringe Bitmaps
fringe-cursor-alist: Fringe Cursors
fringe-indicator-alist: Fringe Indicators
fringes: Fringes
fringes, and empty line indication: Fringe Indicators
fringes-outside-margins: Fringe Size/Pos
front-sticky text property: Sticky Properties
fround: Rounding Operations
fset: Function Cells
ftp-login: Cleanups
ftruncate: Rounding Operations
full keymap: Format of Keymaps
full-height window: Window Sizes
full-width window: Window Sizes
fullboth frames: Size Parameters
fullheight frames: Size Parameters
fullscreen, a frame parameter: Size Parameters
fullscreen-restore, a frame parameter: Size Parameters
fullwidth frames: Size Parameters
func-arity: What Is a Function
funcall: Calling Functions
funcall: Module Misc
funcall, and debugging: Internals of Debugger
funcall-interactively: Interactive Call
function: Anonymous Functions
function aliases: Defining Functions
function call: Function Forms
function call debugging: Function Debugging
function cell: Symbol Components
function cell in autoload: Autoload
function declaration: Declaring Functions
function definition: Function Names
function descriptions: A Sample Function Description
function form evaluation: Function Forms
function groups: Documentation Groups
function input stream: Input Streams
function invocation: Calling Functions
function keys: Function Keys
function name: Function Names
function not known to be defined, compilation warning: Compiler Errors
function output stream: Output Streams
function quoting: Anonymous Functions
function safety: Function Safety
function’s documentation string: Function Documentation
function-alias-p: Defining Functions
function-documentation: Accessing Documentation
function-documentation property: Documentation Basics
function-get: Symbol Plists
function-history (function symbol property): Where Defined
function-key-map: Translation Keymaps
function-put: Symbol Plists
functionals: Calling Functions
functionp: What Is a Function
functions in modes: Major Mode Conventions
functions, making them interactive: Defining Commands
fundamental-mode: Major Modes
fundamental-mode-abbrev-table: Standard Abbrev Tables
future history in minibuffer input: Text from Minibuffer

G
gamma correction: Font and Color Parameters
gap-position: Buffer Gap
gap-size: Buffer Gap
garbage collection: Garbage Collection
garbage collection protection: Writing Emacs Primitives
garbage-collect: Garbage Collection
garbage-collection-messages: Garbage Collection
gc-cons-percentage: Garbage Collection
gc-cons-percentage, in batch mode: Batch Mode
gc-cons-threshold: Garbage Collection
gc-elapsed: Garbage Collection
gcs-done: Garbage Collection
generalized variable: Generalized Variables
generate-new-buffer: Creating Buffers
generate-new-buffer-name: Buffer Names
generated-autoload-file: Autoload
generators: Generators
generic commands: Generic Commands
generic functions: Generic Functions
generic mode: Generic Modes
gensym: Creating Symbols
geometry specification: Geometry
get: Symbol Plists
get node, tree-sitter: Retrieving Nodes
get, defcustom keyword: Variable Definitions
get-buffer: Buffer Names
get-buffer-create: Creating Buffers
get-buffer-process: Process Buffers
get-buffer-window: Buffers and Windows
get-buffer-window-list: Buffers and Windows
get-buffer-xwidgets: Xwidgets
get-byte: Character Codes
get-char-code-property: Character Properties
get-char-property: Examining Properties
get-char-property-and-overlay: Examining Properties
get-charset-property: Character Sets
get-device-terminal: Multiple Terminals
get-display-property: Display Property
get-file-buffer: Buffer File Name
get-internal-run-time: Processor Run Time
get-largest-window: Cyclic Window Ordering
get-load-suffixes: Load Suffixes
get-lru-window: Cyclic Window Ordering
get-mru-window: Cyclic Window Ordering
get-pos-property: Examining Properties
get-process: Process Information
get-register: Registers
get-text-property: Examining Properties
get-unused-category: Categories
get-variable-watchers: Watching Variables
get-window-with-predicate: Cyclic Window Ordering
getenv: System Environment
gethash: Hash Access
get_function_finalizer: Module Functions
get_user_finalizer: Module Values
get_user_ptr: Module Values
GID: User Identification
global binding: Local Variables
global break condition: Global Break Condition
global keymap: Active Keymaps
global variable: Global Variables
global-abbrev-table: Standard Abbrev Tables
global-buffers-menu-map: Standard Keymaps
global-disable-point-adjustment: Adjusting Point
global-key-binding: Low-Level Key Binding
global-map: Controlling Active Maps
global-minor-modes: Minor Modes
global-mode-string: Mode Line Variables
global-set-key: Low-Level Key Binding
global-unset-key: Low-Level Key Binding
glyph: Glyphs
glyph code: Glyphs
glyph-char: Glyphs
glyph-face: Glyphs
glyph-table: Glyphs
glyphless characters: Glyphless Chars
glyphless-char face: Glyphless Chars
glyphless-char-display: Glyphless Chars
glyphless-char-display-control: Glyphless Chars
glyphless-display-mode: Glyphless Chars
GNU ELPA: Package Archives
gnutls cryptographic functions: GnuTLS Cryptographic Functions
gnutls cryptography inputs format: Format of GnuTLS Cryptography Inputs
gnutls-ciphers: GnuTLS Cryptographic Functions
gnutls-digests: GnuTLS Cryptographic Functions
gnutls-hash-digest: GnuTLS Cryptographic Functions
gnutls-hash-mac: GnuTLS Cryptographic Functions
gnutls-macs: GnuTLS Cryptographic Functions
gnutls-symmetric-decrypt: GnuTLS Cryptographic Functions
gnutls-symmetric-encrypt: GnuTLS Cryptographic Functions
goto-char: Character Motion
goto-history-element: Minibuffer Commands
goto-line-history: Minibuffer History
goto-map: Prefix Keys
grammar, SMIE: SMIE Grammar
grand-parent: Parser-based Indentation
grapheme cluster: Size of Displayed Text
graphical display: Frames
graphical terminal: Frames
grave accent, quoting in documentation strings: Documentation Tips
great-grand-parent: Parser-based Indentation
group in rx: Rx Constructs
group, customization keyword: Common Keywords
group-function, in completion: Programmed Completion
group-gid: User Identification
group-n in rx: Rx Constructs
group-name: User Identification
group-real-gid: User Identification
groups of functions: Documentation Groups
gui-get-selection: Window System Selections
gui-set-selection: Window System Selections
guidelines for buffer display: The Zen of Buffer Display
gv-define-expander: Adding Generalized Variables
gv-define-setter: Adding Generalized Variables
gv-define-simple-setter: Adding Generalized Variables
gv-letplace: Adding Generalized Variables

H
hack-connection-local-variables: Applying Connection Local Variables
hack-connection-local-variables-apply: Applying Connection Local Variables
hack-dir-local-variables: Directory Local Variables
hack-dir-local-variables-non-file-buffer: Directory Local Variables
hack-local-variables: File Local Variables
hack-local-variables-hook: File Local Variables
Hamming weight: Bitwise Operations
handle Lisp errors: Handling Errors
handle-focus-in: Input Focus
handle-shift-selection: The Mark
handle-switch-frame: Input Focus
handling errors: Handling Errors
hardening: Security Considerations
has error, tree-sitter node: Accessing Node Information
hash code: Defining Hash
hash notation: Printed Representation
hash table access: Hash Access
hash tables: Hash Tables
hash, cryptographic: Checksum/Hash
hash, cryptographic: GnuTLS Cryptography
hash-table-count: Other Hash
hash-table-p: Other Hash
hash-table-rehash-size: Other Hash
hash-table-rehash-threshold: Other Hash
hash-table-size: Other Hash
hash-table-test: Other Hash
hash-table-weakness: Other Hash
hashing: Creating Symbols
header comments: Library Headers
header line (of a window): Header Lines
header line alignment when line numbers are displayed: Pixel Specification
header-line, prefix key: Key Sequence Input
header-line-format: Header Lines
header-line-format, a window parameter: Window Parameters
header-line-format, and :align-to: Pixel Specification
header-line-indent: Header Lines
header-line-indent-mode: Header Lines
header-line-indent-width: Header Lines
height of a line: Line Height
height of a window: Window Sizes
height spec: Line Height
height, a frame parameter: Size Parameters
help for major mode: Mode Help
help functions: Help Functions
help-buffer: Help Functions
help-char: Help Functions
help-command: Help Functions
help-echo (button property): Button Properties
help-echo (overlay property): Overlay Properties
help-echo (text property): Special Properties
help-echo event: Misc Events
help-echo text, avoid command-key substitution: Special Properties
help-echo, customization keyword: Type Keywords
help-echo-inhibit-substitution (text property): Special Properties
help-event-list: Help Functions
help-form: Help Functions
help-key-binding (face): Keys in Documentation
help-map: Help Functions
help-setup-xref: Help Functions
help-window-select: Help Functions
Helper-describe-bindings: Help Functions
Helper-help: Help Functions
Helper-help-map: Help Functions
hex numbers: Integer Basics
hidden buffers: Buffer Names
history list: Minibuffer History
history of commands: Command History
history-add-new-input: Minibuffer History
history-delete-duplicates: Minibuffer History
history-length: Minibuffer History
HOME environment variable: Subprocess Creation
hook variables, list of: Standard Hooks
hooks: Hooks
hooks for changing a character: Special Properties
hooks for loading: Hooks for Loading
hooks for motion of point: Special Properties
hooks for text changes: Change Hooks
hooks for window operations: Window Hooks
horizontal combination: Windows and Frames
horizontal position: Columns
horizontal scrolling: Horizontal Scrolling
horizontal-lineto: SVG Images
horizontal-scroll-bar: Scroll Bars
horizontal-scroll-bar, prefix key: Key Sequence Input
horizontal-scroll-bar-mode: Scroll Bars
horizontal-scroll-bars, a frame parameter: Layout Parameters
horizontal-scroll-bars-available-p: Scroll Bars
host language, tree-sitter: Multiple Languages
how to visit files: Visiting Functions
HTML DOM: Document Object Model
hyper characters: Other Char Bits
hyperlinks in documentation strings: Documentation Tips

I
icon keywords: Icons
icon types: Icons
icon-elements: Icons
icon-left, a frame parameter: Position Parameters
icon-name, a frame parameter: Management Parameters
icon-string: Icons
icon-title-format: Frame Titles
icon-top, a frame parameter: Position Parameters
icon-type, a frame parameter: Management Parameters
iconified frame: Visibility of Frames
iconify-child-frame: Child Frames
iconify-frame: Visibility of Frames
iconify-frame event: Misc Events
identical-contents objects, and byte-compiler: Equality Predicates
identity: Calling Functions
idle timers: Idle Timers
idleness: Idle Timers
IEEE floating point: Float Basics
if: Conditionals
if-let: Conditionals
ignore: Calling Functions
ignore-error: Handling Errors
ignore-errors: Handling Errors
ignore-window-parameters: Window Parameters
ignored-local-variable-values: File Local Variables
ignored-local-variables: File Local Variables
image animation: Multi-Frame Images
image cache: Image Cache
image descriptor: Image Descriptors
image formats: Image Formats
image frames: Multi-Frame Images
image maps: Image Descriptors
image slice: Showing Images
image types: Image Formats
image-animate: Multi-Frame Images
image-animate-timer: Multi-Frame Images
image-at-point-p: Showing Images
image-cache-eviction-delay: Image Cache
image-cache-size: Image Cache
image-crop: Showing Images
image-current-frame: Multi-Frame Images
image-cut: Showing Images
image-decrease-size: Showing Images
image-default-frame-delay: Multi-Frame Images
image-flip-horizontally: Showing Images
image-flip-vertically: Showing Images
image-flush: Image Cache
image-format-suffixes: ImageMagick Images
image-increase-size: Showing Images
image-load-path: Defining Images
image-load-path-for-library: Defining Images
image-mask-p: Image Descriptors
image-minimum-frame-delay: Multi-Frame Images
image-multi-frame-p: Multi-Frame Images
image-property: Defining Images
image-rotate: Showing Images
image-save: Showing Images
image-scaling-factor: Defining Images
image-show-frame: Multi-Frame Images
image-size: Showing Images
image-transforms-p: Image Descriptors
image-type-available-p: Image Formats
image-types: Image Formats
ImageMagick images: ImageMagick Images
imagemagick-enabled-types: ImageMagick Images
imagemagick-types: ImageMagick Images
imagemagick-types-inhibit: ImageMagick Images
images in buffers: Images
images, support for more formats: ImageMagick Images
Imenu: Imenu
imenu-add-to-menubar: Imenu
imenu-case-fold-search: Imenu
imenu-create-index-function: Imenu
imenu-extract-index-name-function: Imenu
imenu-generic-expression: Imenu
imenu-prev-index-position-function: Imenu
imenu-syntax-alist: Imenu
implicit progn: Sequencing
implied frame resizing: Implied Frame Resizing
implied resizing of frame: Implied Frame Resizing
in in rx: Rx Constructs
inactive minibuffer: Intro to Minibuffers
inc: Simple Macro
indefinite extent: Variable Scoping
indent-according-to-mode: Mode-Specific Indent
indent-code-rigidly: Region Indent
indent-for-tab-command: Mode-Specific Indent
indent-line-function: Mode-Specific Indent
indent-region: Region Indent
indent-region-function: Region Indent
indent-relative: Relative Indent
indent-relative-first-indent-point: Relative Indent
indent-rigidly: Region Indent
indent-tabs-mode: Primitive Indent
indent-to: Primitive Indent
indent-to-left-margin: Margins
indentation: Indentation
indentation rules, for parser-based indentation: Parser-based Indentation
indentation rules, SMIE: SMIE Indentation
indicate-buffer-boundaries: Fringe Indicators
indicate-empty-lines: Fringe Indicators
indicators, fringe: Fringe Indicators
indirect buffers: Indirect Buffers
indirect specifications: Specification List
indirect-function: Function Indirection
indirect-variable: Variable Aliases
indirection for functions: Function Indirection
infinite loops: Infinite Loops
infinity: Float Basics
information of node, syntax trees: Accessing Node Information
inheritance, for faces: Face Attributes
inheritance, keymap: Inheritance and Keymaps
inheritance, syntax table: Syntax Basics
inheritance, text property: Sticky Properties
inhibit asynchronous native compilation: Native-Compilation Variables
inhibit-default-init: Init File
inhibit-double-buffering, a frame parameter: Management Parameters
inhibit-eol-conversion: Specifying Coding Systems
inhibit-field-text-motion: Word Motion
inhibit-file-name-handlers: Magic File Names
inhibit-file-name-operation: Magic File Names
inhibit-interaction: Inhibiting Interaction
inhibit-isearch (text property): Special Properties
inhibit-iso-escape-detection: Lisp and Coding Systems
inhibit-local-variables-regexps: File Local Variables
inhibit-local-variables-regexps: Auto Major Mode
inhibit-message: Displaying Messages
inhibit-message: Displaying Messages
inhibit-message-regexps: Displaying Messages
inhibit-modification-hooks: Change Hooks
inhibit-null-byte-detection: Lisp and Coding Systems
inhibit-point-motion-hooks: Special Properties
inhibit-quit: Quitting
inhibit-read-only: Read Only Buffers
inhibit-read-only (text property): Special Properties
inhibit-same-window, a buffer display action alist entry: Buffer Display Action Alists
inhibit-splash-screen: Startup Summary
inhibit-startup-echo-area-message: Startup Summary
inhibit-startup-message: Startup Summary
inhibit-startup-screen: Startup Summary
inhibit-switch-frame, a buffer display action alist entry: Buffer Display Action Alists
inhibit-x-resources: Resources
init file: Init File
init-file-user: User Identification
init.el: Init File
initial-buffer-choice: Startup Summary
initial-environment: System Environment
initial-frame-alist: Initial Parameters
initial-major-mode: Auto Major Mode
initial-scratch-message: Startup Summary
initial-window-system: Window Systems
initial-window-system, and startup: Startup Summary
initialization of Emacs: Startup Summary
initialize, defcustom keyword: Variable Definitions
initiating drag-and-drop: Drag and Drop
initiating drag-and-drop, low-level: Drag and Drop
inline completion: Completion in Buffers
inline functions: Inline Functions
inline-const-p: Inline Functions
inline-const-val: Inline Functions
inline-error: Inline Functions
inline-letevals: Inline Functions
inline-quote: Inline Functions
inner edges: Frame Layout
inner frame: Frame Layout
inner height: Frame Layout
inner size: Frame Layout
inner width: Frame Layout
innermost containing parentheses: Parser State
input devices: Command Loop Info
input events: Input Events
input events, prevent recording: Event Input Misc
input focus: Input Focus
input methods: Input Methods
input modes: Input Modes
input stream: Input Streams
input-decode-map: Translation Keymaps
input-method-alist: Input Methods
input-method-function: Invoking the Input Method
input-pending-p: Event Input Misc
insecure text: Suspicious Text
insert: Insertion
insert-abbrev-table-description: Abbrev Tables
insert-and-inherit: Sticky Properties
insert-before-markers: Insertion
insert-before-markers-and-inherit: Sticky Properties
insert-behind-hooks (overlay property): Overlay Properties
insert-behind-hooks (text property): Special Properties
insert-buffer: Commands for Insertion
insert-buffer-substring: Insertion
insert-buffer-substring-as-yank: Yanking
insert-buffer-substring-no-properties: Insertion
insert-button: Making Buttons
insert-char: Insertion
insert-default-directory: Reading File Names
insert-directory: Contents of Directories
insert-directory-program: Contents of Directories
insert-file-contents: Reading from Files
insert-file-contents-literally: Reading from Files
insert-for-yank: Yanking
insert-image: Showing Images
insert-in-front-hooks (overlay property): Overlay Properties
insert-in-front-hooks (text property): Special Properties
insert-into-buffer: Insertion
insert-register: Registers
insert-sliced-image: Showing Images
insert-text-button: Making Buttons
inserting killed text: Yank Commands
insertion before point: Insertion
insertion of text: Insertion
insertion type of a marker: Marker Insertion Types
inside comment: Parser State
inside string: Parser State
inspection of tree-sitter parse tree nodes: Language Grammar
installation-directory: System Environment
instrumenting for Edebug: Instrumenting
int-to-string: String Conversion
intangible (overlay property): Overlay Properties
intangible (text property): Special Properties
integer range: Integer Basics
integer to decimal: String Conversion
integer to hexadecimal: Formatting Strings
integer to octal: Formatting Strings
integer to string: String Conversion
integer types (C programming language): C Integer Types
integer-or-marker-p: Predicates on Markers
integer-width: Integer Basics
integerp: Predicates on Numbers
integers: Numbers
integers in specific radix: Integer Basics
interaction parameters between frames: Frame Interaction Parameters
interactive: Using Interactive
interactive call: Interactive Call
interactive code description: Interactive Codes
interactive completion: Interactive Codes
interactive function: Defining Commands
interactive spec, using: Using Interactive
interactive specification in primitives: Writing Emacs Primitives
interactive, examples of using: Interactive Examples
interactive-form: Using Interactive
interactive-form property: Defining Commands
interactive-form, symbol property: Using Interactive
interactive-only property: Defining Commands
intern: Creating Symbols
intern: Module Misc
intern-soft: Creating Symbols
internal menu bar: Frame Layout
internal representation of characters: Text Representations
internal tool bar: Frame Layout
internal windows: Basic Windows
internal-border-width, a frame parameter: Layout Parameters
internals, of buffer: Buffer Internals
internals, of process: Process Internals
internals, of window: Window Internals
interning: Creating Symbols
interpreter: Evaluation
interpreter: Evaluation
interpreter-mode-alist: Auto Major Mode
interprogram-cut-function: Low-Level Kill Ring
interprogram-paste-function: Low-Level Kill Ring
interrupt Lisp functions: Quitting
interrupt-process: Signals to Processes
interrupt-process-functions: Signals to Processes
intersection in rx: Rx Constructs
intersection of sequences: Sequence Functions
intervals: Not Intervals
intervals-consed: Memory Usage
invalid prefix key error: Changing Key Bindings
invalid-function: Function Indirection
invalid-read-syntax: Printed Representation
invalid-regexp: Regexp Backslash
invert-face: Attribute Functions
invisible (overlay property): Overlay Properties
invisible (text property): Special Properties
invisible frame: Visibility of Frames
invisible text: Invisible Text
invisible-p: Invisible Text
invisible/intangible text, and point: Adjusting Point
invocation-directory: System Environment
invocation-name: System Environment
invoking input method: Invoking the Input Method
invoking lisp debugger: Invoking the Debugger
is this call interactive: Distinguish Interactive
isnan: Float Basics
ISO 8601 date/time strings: Time Parsing
ISO week, in time formatting: Time Parsing
iso8601-parse: Time Parsing
is_not_nil: Module Misc
italic text: Face Attributes
iter-close: Generators
iter-defun: Generators
iter-do: Generators
iter-lambda: Generators
iter-next: Generators
iter-yield: Generators
iter-yield-from: Generators
iteration: Iteration
iteration over vector or string: Sequence Functions

J
JavaScript Object Notation: Parsing JSON
javascript-callback xwidget events: Xwidget Events
jit-lock functions, debugging: Other Font Lock Variables
jit-lock-debug-mode: Other Font Lock Variables
jit-lock-register: Other Font Lock Variables
jit-lock-unregister: Other Font Lock Variables
joining lists: Rearrangement
JSON: Parsing JSON
JSON remote procedure call protocol: JSONRPC
json-available-p: Parsing JSON
json-insert: Parsing JSON
json-parse-buffer: Parsing JSON
json-parse-string: Parsing JSON
json-serialize: Parsing JSON
JSONRPC: JSONRPC
JSONRPC application interfaces: JSONRPC Overview
JSONRPC connection initargs: Process-based JSONRPC connections
JSONRPC deferred requests: JSONRPC deferred requests
JSONRPC object format: JSONRPC JSON object format
JSONRPC process-based connections: Process-based JSONRPC connections
jsonrpc-async-request: JSONRPC Overview
jsonrpc-connection: JSONRPC Overview
jsonrpc-connection-ready-p: JSONRPC deferred requests
jsonrpc-connection-receive: JSONRPC Overview
jsonrpc-connection-send: JSONRPC Overview
jsonrpc-error: JSONRPC Overview
jsonrpc-lambda: JSONRPC JSON object format
jsonrpc-notify: JSONRPC Overview
jsonrpc-process-connection: Process-based JSONRPC connections
jsonrpc-request: JSONRPC Overview
jsonrpc-running-p: JSONRPC Overview
jsonrpc-shutdown: JSONRPC Overview
jumbled display of bidirectional text: Bidirectional Display
just-one-space: User-Level Deletion
justify-current-line: Filling

K
kbd: Key Sequences
kbd-macro-termination-hook: Keyboard Macros
keep-ratio, a frame parameter: Frame Interaction Parameters
kept-new-versions: Numbered Backups
kept-old-versions: Numbered Backups
key: Key Sequences
key binding: Keymap Basics
key binding, conventions for: Key Binding Conventions
key lookup: Key Lookup
key sequence: Key Sequences
key sequence error: Changing Key Bindings
key sequence input: Key Sequence Input
key substitution sequence: Keys in Documentation
key translation function: Translation Keymaps
key-binding: Low-Level Key Binding
key-description: Describing Characters
key-translate: Event Mod
key-translation-map: Translation Keymaps
key-valid-p: Key Sequences
keyboard events: Keyboard Events
keyboard events in strings: Strings of Events
keyboard events, data in: Accessing Mouse
keyboard input: Reading Input
keyboard input decoding on X: Locales
keyboard macro execution: Interactive Call
keyboard macro termination: Beeping
keyboard macro, terminating: Event Input Misc
keyboard macros: Keyboard Macros
keyboard macros (Edebug): Edebug Execution Modes
keyboard-coding-system: Terminal I/O Encoding
keyboard-quit: Quitting
keyboard-translate: Low-Level Key Binding
keyboard-translate-table: Event Mod
keymap: Keymaps
keymap (button property): Button Properties
keymap (overlay property): Overlay Properties
keymap (text property): Special Properties
keymap entry: Key Lookup
keymap format: Format of Keymaps
keymap in keymap: Key Lookup
keymap inheritance: Inheritance and Keymaps
keymap inheritance from multiple maps: Inheritance and Keymaps
keymap of character: Special Properties
keymap of character (and overlays): Overlay Properties
keymap prompt string: Format of Keymaps
keymap-global-lookup: Functions for Key Lookup
keymap-global-set: Key Binding Commands
keymap-global-unset: Key Binding Commands
keymap-local-lookup: Functions for Key Lookup
keymap-local-set: Key Binding Commands
keymap-local-unset: Key Binding Commands
keymap-lookup: Functions for Key Lookup
keymap-parent: Inheritance and Keymaps
keymap-prompt: Defining Menus
keymap-set: Changing Key Bindings
keymap-set: Changing Key Bindings
keymap-set-after: Modifying Menus
keymap-unset: Changing Key Bindings
keymap-unset: Changing Key Bindings
keymapp: Format of Keymaps
keymaps for translating events: Translation Keymaps
keymaps in modes: Major Mode Conventions
keymaps, scanning: Scanning Keymaps
keymaps, standard: Standard Keymaps
keys in documentation strings: Keys in Documentation
keys, reserved: Key Binding Conventions
keystroke: Key Sequences
keyword symbol: Constant Variables
keywordp: Constant Variables
kill command repetition: Command Loop Info
kill ring: The Kill Ring
kill-all-local-variables: Creating Buffer-Local
kill-append: Low-Level Kill Ring
kill-buffer: Killing Buffers
kill-buffer-hook: Killing Buffers
kill-buffer-hook in temporary buffers: Current Buffer
kill-buffer-query-functions: Killing Buffers
kill-buffer-query-functions in temporary buffers: Current Buffer
kill-emacs: Killing Emacs
kill-emacs-hook: Killing Emacs
kill-emacs-query-functions: Killing Emacs
kill-local-variable: Creating Buffer-Local
kill-new: Low-Level Kill Ring
kill-process: Signals to Processes
kill-read-only-ok: Kill Functions
kill-region: Kill Functions
kill-ring: Internals of Kill Ring
kill-ring-max: Internals of Kill Ring
kill-ring-yank-pointer: Internals of Kill Ring
kill-xwidget: Xwidgets
killing buffers: Killing Buffers
killing Emacs: Killing Emacs
kinship, syntax tree nodes: Retrieving Nodes
kmacro-keymap: Standard Keymaps

L
labeled narrowing: Narrowing
labeled restriction: Narrowing
lambda: Anonymous Functions
lambda expression: Lambda Expressions
lambda in debug: Invoking the Debugger
lambda in keymap: Key Lookup
lambda list: Lambda Components
lambda-list (Edebug): Specification List
language argument, for tree-sitter: Language Grammar
language grammar version, compatibility: Language Grammar
language grammar, for tree-sitter: Language Grammar
language-change event: Misc Events
largest fixnum: Integer Basics
largest window: Cyclic Window Ordering
last: List Elements
last visible position in a window: Window Start and End
last-abbrev: Abbrev Expansion
last-abbrev-location: Abbrev Expansion
last-abbrev-text: Abbrev Expansion
last-buffer: Buffer List
last-coding-system-used: Encoding and I/O
last-command: Command Loop Info
last-command-event: Command Loop Info
last-event-device: Command Loop Info
last-event-frame: Command Loop Info
last-input-event: Event Input Misc
last-kbd-macro: Keyboard Macros
last-nonmenu-event: Command Loop Info
last-prefix-arg: Prefix Command Arguments
last-repeatable-command: Command Loop Info
lax-plist-get: Plist Access
lax-plist-put: Plist Access
layout of frame: Frame Layout
layout on display, and bidirectional text: Bidirectional Display
layout parameters of frames: Layout Parameters
lazy evaluation: Deferred Eval
lazy loading: Dynamic Loading
lazy-completion-table: Basic Completion
ldexp: Float Basics
leaf node, of tree-sitter parse tree: Retrieving Nodes
leap seconds: Time of Day
least recently used window: Cyclic Window Ordering
left and right window decorations: Basic Windows
left position ratio: Position Parameters
left, a frame parameter: Position Parameters
left-fringe, a frame parameter: Layout Parameters
left-fringe, prefix key: Key Sequence Input
left-fringe-width: Fringe Size/Pos
left-margin: Margins
left-margin, prefix key: Key Sequence Input
left-margin-width: Display Margins
length: Sequence Functions
length<: Sequence Functions
length=: Sequence Functions
length>: Sequence Functions
let: Local Variables
let*: Local Variables
let-alist: Association Lists
letrec: Local Variables
Levenshtein distance: Text Comparison
lexical binding: Variable Scoping
lexical binding (Edebug): Edebug Eval
lexical comparison of strings: Text Comparison
lexical environment: Lexical Binding
lexical scope: Variable Scoping
lexical-binding: Using Lexical Binding
library: Loading
library compilation: Compilation Functions
library header comments: Library Headers
library search: Library Search
libxml-available-p: Parsing HTML/XML
libxml-parse-html-region: Parsing HTML/XML
libxml-parse-xml-region: Parsing HTML/XML
line end conversion: Coding System Basics
line height: Frame Font
line height: Line Height
line number: Text Lines
line truncation: Truncation
line wrapping: Truncation
line-beginning-position: Text Lines
line-end in rx: Rx Constructs
line-end-position: Text Lines
line-height (text property): Special Properties
line-height (text property): Line Height
line-move-ignore-invisible: Invisible Text
line-number-at-pos: Text Lines
line-number-display-width: Size of Displayed Text
line-pixel-height: Size of Displayed Text
line-prefix: Truncation
line-spacing: Line Height
line-spacing (text property): Special Properties
line-spacing (text property): Line Height
line-spacing, a frame parameter: Layout Parameters
line-start in rx: Rx Constructs
lines: Text Lines
lines in region: Text Lines
lineto: SVG Images
link, customization keyword: Common Keywords
linked list: Cons Cell Type
linking files: Changing Files
Lisp debugger: Debugger
Lisp expression motion: List Motion
Lisp history: Lisp History
Lisp library: Loading
Lisp nesting error: Eval
Lisp object: Lisp Data Types
Lisp objects, stack-allocated: Stack-allocated Objects
Lisp package: Packaging
Lisp printer: Output Functions
Lisp reader: Streams Intro
Lisp timestamp: Time of Day
lisp variables defined in C, restrictions: Variables with Restricted Values
lisp-directory: Library Search
lisp-indent-function property: Indenting Macros
lisp-mode-abbrev-table: Standard Abbrev Tables
lisp-mode-autoload-regexp: Autoload
lisp-mode.el: Example Major Modes
list: Building Lists
list all coding systems: Lisp and Coding Systems
list elements: List Elements
list form evaluation: Classifying Lists
list in keymap: Key Lookup
list length: Sequence Functions
list modification: List Variables
list motion: List Motion
list of threads: The Thread List
list predicates: List-related Predicates
list reverse: Sequence Functions
list structure: Cons Cell Type
list structure: Cons Cells
list to vector: Sequence Functions
list, replace element: Setcar
list-buffers-directory: Buffer File Name
list-charset-chars: Character Sets
list-fonts: Low-Level Font
list-load-path-shadows: Library Search
list-multisession-values: Multisession Variables
list-processes: Process Information
list-system-processes: System Processes
list-threads: The Thread List
list-timers: Timers
listify-key-sequence: Event Input Misc
listing all buffers: Buffer List
listp: List-related Predicates
lists: Lists
lists and cons cells: Cons Cells
lists as sets: Sets And Lists
literal evaluation: Self-Evaluating Forms
literal in rx: Rx Constructs
literate programming: Mode-Specific Indent
little endian, in bindat specification: Bindat Types
live buffer: Killing Buffers
live node, tree-sitter: Accessing Node Information
live windows: Basic Windows
ln: Changing Files
load: How Programs Do Loading
load error with require: Named Features
load errors: How Programs Do Loading
load, customization keyword: Common Keywords
load-average: System Environment
load-changed xwidget event: Xwidget Events
load-committed’ in xwidgets: Xwidget Events
load-file: How Programs Do Loading
load-file-name: How Programs Do Loading
load-file-rep-suffixes: Load Suffixes
load-finished’ in xwidgets: Xwidget Events
load-history: Where Defined
load-in-progress: How Programs Do Loading
load-library: How Programs Do Loading
load-path: Library Search
load-prefer-newer: Load Suffixes
load-read-function: How Programs Do Loading
load-redirected’ in xwidgets: Xwidget Events
load-started’ in xwidgets: Xwidget Events
load-suffixes: Load Suffixes
load-theme: Custom Themes
loaddefs-generate: Autoload
loading: Loading
loading and configuring features: Named Features
loading hooks: Hooks for Loading
loading language grammar for tree-sitter: Language Grammar
loading, and non-ASCII characters: Loading Non-ASCII
loadup.el: Building Emacs
local binding: Local Variables
local keymap: Active Keymaps
local part of remote file name: Magic File Names
local variables: Local Variables
local variables, killed by major mode: Creating Buffer-Local
local, defcustom keyword: Variable Definitions
local-abbrev-table: Standard Abbrev Tables
local-function-key-map: Translation Keymaps
local-key-binding: Low-Level Key Binding
local-map (overlay property): Overlay Properties
local-map (text property): Special Properties
local-minor-modes: Minor Modes
local-set-key: Low-Level Key Binding
local-unset-key: Low-Level Key Binding
local-variable-if-set-p: Creating Buffer-Local
local-variable-p: Creating Buffer-Local
locale: Locales
locale-coding-system: Locales
locale-dependent string comparison: Text Comparison
locale-dependent string equivalence: Text Comparison
locale-info: Locales
locate file in path: Locating Files
locate-dominating-file: Contents of Directories
locate-file: Locating Files
locate-library: Library Search
locate-user-emacs-file: Standard File Names
lock file: File Locks
lock-buffer: File Locks
lock-file-mode: File Locks
lock-file-name-transforms: File Locks
log: Math Functions
logand: Bitwise Operations
logb: Float Basics
logcount: Bitwise Operations
logging echo-area messages: Logging Messages
logical arithmetic: Bitwise Operations
logical lines, moving by: Text Lines
logical order: Bidirectional Display
logical shift: Bitwise Operations
logior: Bitwise Operations
lognot: Bitwise Operations
logxor: Bitwise Operations
looking up abbrevs: Abbrev Expansion
looking up fonts: Font Lookup
looking-at: Regexp Search
looking-at-p: Regexp Search
looking-back: Regexp Search
lookup tables: Hash Tables
lookup-key: Low-Level Key Binding
loops, infinite: Infinite Loops
low-level key bindings: Low-Level Key Binding
lower case: Case Conversion
lower-frame: Raising and Lowering
lowering a frame: Raising and Lowering
LRO: Bidirectional Display
lru-frames, a buffer display action alist entry: Buffer Display Action Alists
lru-time, a buffer display action alist entry: Buffer Display Action Alists
lsh: Bitwise Operations
lwarn: Warning Basics

M
M-g: Prefix Keys
M-s: Prefix Keys
M-S-x: Interactive Call
M-X: Interactive Call
M-x: Interactive Call
Maclisp: Lisp History
macro: What Is a Function
macro argument evaluation: Argument Evaluation
macro call: Expansion
macro call evaluation: Macro Forms
macro caveats: Problems with Macros
macro compilation: Compilation Functions
macro descriptions: A Sample Function Description
macro expansion: Expansion
macro, how to define: Defining Macros
macroexpand: Expansion
macroexpand-1: Expansion
macroexpand-all: Expansion
macrop: Simple Macro
macros: Macros
macros, at compile time: Eval During Compile
magic autoload comment: Autoload
magic file names: Magic File Names
magic-fallback-mode-alist: Auto Major Mode
magic-mode-alist: Auto Major Mode
mail-host-address: System Environment
main window: Side Windows
main window of a frame: Side Windows
main-thread: Basic Thread Functions
major mode: Major Modes
major mode command: Major Modes
major mode conventions: Major Mode Conventions
major mode hook: Major Mode Conventions
major mode keymap: Active Keymaps
major mode, automatic selection: Auto Major Mode
major mode, developing with tree-sitter: Tree-sitter Major Modes
major-mode: Major Modes
major-mode-restore: Major Modes
major-mode-suspend: Major Modes
make-abbrev-table: Abbrev Tables
make-auto-save-file-name: Auto-Saving
make-backup-file-name: Backup Names
make-backup-file-name-function: Making Backups
make-backup-files: Making Backups
make-bool-vector: Bool-Vectors
make-button: Making Buttons
make-byte-code: Byte-Code Objects
make-category-set: Categories
make-category-table: Categories
make-char-table: Char-Tables
make-composed-keymap: Inheritance and Keymaps
make-condition-variable: Condition Variables
make-directory: Create/Delete Dirs
make-display-table: Display Tables
make-empty-file: Create/Delete Dirs
make-finalizer: Finalizer Type
make-frame: Creating Frames
make-frame-invisible: Visibility of Frames
make-frame-on-display: Multiple Terminals
make-frame-on-monitor: Multiple Terminals
make-frame-visible: Visibility of Frames
make-frame-visible event: Misc Events
make-glyph-code: Glyphs
make-hash-table: Creating Hash
make-help-screen: Help Functions
make-indirect-buffer: Indirect Buffers
make-keymap: Creating Keymaps
make-list: Building Lists
make-local-variable: Creating Buffer-Local
make-marker: Creating Markers
make-multisession: Multisession Variables
make-mutex: Mutexes
make-nearby-temp-file: Unique File Names
make-network-process: Network Processes
make-obsolete: Obsolete Functions
make-obsolete-generalized-variable: Adding Generalized Variables
make-obsolete-variable: Variable Aliases
make-overlay: Managing Overlays
make-pipe-process: Asynchronous Processes
make-process: Asynchronous Processes
make-progress-reporter: Progress
make-record: Record Functions
make-ring: Rings
make-serial-process: Serial Ports
make-sparse-keymap: Creating Keymaps
make-string: Creating Strings
make-symbol: Creating Symbols
make-symbolic-link: Changing Files
make-syntax-table: Syntax Table Functions
make-temp-file: Unique File Names
make-temp-name: Unique File Names
make-text-button: Making Buttons
make-thread: Basic Thread Functions
make-translation-table: Translation of Characters
make-translation-table-from-alist: Translation of Characters
make-translation-table-from-vector: Translation of Characters
make-variable-buffer-local: Creating Buffer-Local
make-vector: Vector Functions
make-vtable: Tabulated List Mode
make-xwidget: Xwidgets
make_big_integer: Module Values
make_float: Module Values
make_function: Module Functions
make_global_ref: Module Values
make_integer: Module Values
make_interactive: Module Functions
make_string: Module Values
make_time: Module Values
make_unibyte_string: Module Values
make_user_ptr: Module Values
making backup files: Making Backups
making buttons: Making Buttons
makunbound: Void Variables
malicious use of directional overrides: Bidirectional Display
managing overlays: Managing Overlays
manipulating buttons: Manipulating Buttons
map-char-table: Char-Tables
map-charset-chars: Character Sets
map-keymap: Scanning Keymaps
map-y-or-n-p: Multiple Queries
mapatoms: Creating Symbols
mapbacktrace: Internals of Debugger
mapc: Mapping Functions
mapcan: Mapping Functions
mapcar: Mapping Functions
mapconcat: Mapping Functions
maphash: Hash Access
mapped frame: Visibility of Frames
mapping functions: Mapping Functions
margins, display: Display Margins
margins, filling: Margins
mark: The Mark
mark excursion: Excursions
mark ring: The Mark
mark, the: The Mark
mark-active: The Mark
mark-even-if-inactive: The Mark
mark-marker: The Mark
mark-ring: The Mark
mark-ring-max: The Mark
marker argument: Interactive Codes
marker creation: Creating Markers
marker garbage collection: Overview of Markers
marker information: Information from Markers
marker input stream: Input Streams
marker output stream: Output Streams
marker relocation: Overview of Markers
marker, how to move position: Moving Markers
marker-buffer: Information from Markers
marker-insertion-type: Marker Insertion Types
marker-position: Information from Markers
markerp: Predicates on Markers
markers: Markers
markers as numbers: Overview of Markers
markers, predicates for: Predicates on Markers
match: Parser-based Indentation
match data: Match Data
match, customization keyword: Type Keywords
match-alternatives, customization keyword: Composite Types
match-beginning: Simple Match Data
match-buffers: Buffer List
match-data: Entire Match Data
match-end: Simple Match Data
match-inline, customization keyword: Type Keywords
match-string: Simple Match Data
match-string-no-properties: Simple Match Data
match-substitute-replacement: Replacing Match
matching, structural: Backquote Patterns
mathematical functions: Math Functions
max: Comparison of Numbers
max-char: Character Codes
max-image-size: Showing Images
max-lisp-eval-depth: Eval
max-mini-window-height: Minibuffer Windows
maximal-match in rx: Rx Constructs
maximize-window: Resizing Windows
maximized frames: Size Parameters
maximizing windows: Resizing Windows
maximum fixnum: Integer Basics
maximum value of character codepoint: Character Codes
maximum value of sequence: Sequence Functions
maximum-scroll-margin: Textual Scrolling
maybe_quit, use in Lisp primitives: Writing Emacs Primitives
md5: Checksum/Hash
MD5 checksum: Checksum/Hash
MD5 checksum: GnuTLS Cryptography
measuring resource usage: Profiling
member: Sets And Lists
member-ignore-case: Sets And Lists
membership in a list: Sets And Lists
memory allocation: Garbage Collection
memory usage: Profiling
memory usage: Memory Usage
memory-full: Garbage Collection
memory-info: Garbage Collection
memory-limit: Garbage Collection
memory-report: Garbage Collection
memory-use-counts: Garbage Collection
memq: Sets And Lists
memql: Sets And Lists
menu bar: Menu Bar
menu bar keymaps: Standard Keymaps
menu definition example: Menu Example
menu item: Defining Menus
menu keymaps: Menu Keymaps
menu modification: Modifying Menus
menu prompt string: Defining Menus
menu separators: Menu Separators
menu-bar, prefix key: Key Sequence Input
menu-bar-file-menu: Standard Keymaps
menu-bar-final-items: Menu Bar
menu-bar-help-menu: Standard Keymaps
menu-bar-lines, a frame parameter: Layout Parameters
menu-bar-options-menu: Standard Keymaps
menu-bar-tools-menu: Standard Keymaps
menu-bar-update-hook: Menu Bar
menu-item: Extended Menu Items
menu-prompt-more-char: Keyboard Menus
menus, popup: Pop-Up Menus
merge-face-attribute: Attribute Functions
message: Displaying Messages
message digest: Checksum/Hash
message, finding what causes a particular message: Error Debugging
message-box: Displaying Messages
message-log-max: Logging Messages
message-or-box: Displaying Messages
message-truncate-lines: Echo Area Customization
messages-buffer: Logging Messages
messages-buffer-name: Logging Messages
meta character key constants: Low-Level Key Binding
meta character printing: Describing Characters
meta characters: Meta-Char Syntax
meta characters lookup: Format of Keymaps
meta-prefix-char: Functions for Key Lookup
min: Comparison of Numbers
min-height, a frame parameter: Size Parameters
min-margins, a window parameter: Window Parameters
min-width, a frame parameter: Size Parameters
minibuffer: Minibuffers
minibuffer completion: Minibuffer Completion
minibuffer contents, accessing: Minibuffer Contents
minibuffer history: Minibuffer History
minibuffer input: Recursive Editing
minibuffer input, and command-line arguments: Shell Arguments
minibuffer input, reading lisp objects: Object from Minibuffer
minibuffer input, reading text strings: Text from Minibuffer
minibuffer window: Basic Windows
minibuffer window, and next-window: Cyclic Window Ordering
minibuffer windows: Minibuffer Windows
minibuffer, a frame parameter: Buffer Parameters
minibuffer-allow-text-properties: Text from Minibuffer
minibuffer-auto-raise: Raising and Lowering
minibuffer-beginning-of-buffer-movement: Completion Commands
minibuffer-complete: Completion Commands
minibuffer-complete-and-exit: Completion Commands
minibuffer-complete-word: Completion Commands
minibuffer-completion-confirm: Completion Commands
minibuffer-completion-help: Completion Commands
minibuffer-completion-predicate: Completion Commands
minibuffer-completion-table: Completion Commands
minibuffer-confirm-exit-commands: Completion Commands
minibuffer-contents: Minibuffer Contents
minibuffer-contents-no-properties: Minibuffer Contents
minibuffer-default-add-function: Minibuffer History
minibuffer-default-prompt-format: Text from Minibuffer
minibuffer-depth: Recursive Mini
minibuffer-exit, a frame parameter: Frame Interaction Parameters
minibuffer-exit-hook: Minibuffer Misc
minibuffer-frame-alist: Initial Parameters
minibuffer-help-form: Minibuffer Misc
minibuffer-history: Minibuffer History
minibuffer-inactive-mode: Minibuffer Misc
minibuffer-inactive-mode-map: Standard Keymaps
minibuffer-less frame: Frame Layout
minibuffer-local-completion-map: Completion Commands
minibuffer-local-filename-completion-map: Completion Commands
minibuffer-local-map: Text from Minibuffer
minibuffer-local-must-match-map: Completion Commands
minibuffer-local-ns-map: Text from Minibuffer
minibuffer-local-shell-command-map: Reading File Names
minibuffer-message: Minibuffer Misc
minibuffer-message (text property): Special Properties
minibuffer-message-timeout: Minibuffer Misc
minibuffer-mode-map: Text from Minibuffer
minibuffer-only frame: Frame Layout
minibuffer-only frame: Initial Parameters
minibuffer-prompt: Minibuffer Contents
minibuffer-prompt-end: Minibuffer Contents
minibuffer-prompt-properties: Text from Minibuffer
minibuffer-prompt-width: Minibuffer Contents
minibuffer-scroll-window: Minibuffer Misc
minibuffer-selected-window: Minibuffer Misc
minibuffer-setup-hook: Minibuffer Misc
minibuffer-window: Minibuffer Windows
minibuffer-window-active-p: Minibuffer Windows
minibuffer-with-setup-hook: Minibuffer Misc
minibufferp: Minibuffer Misc
minimal-match in rx: Rx Constructs
minimize-window: Resizing Windows
minimized frame: Visibility of Frames
minimizing windows: Resizing Windows
minimum fixnum: Integer Basics
minimum value of sequence: Sequence Functions
minor mode: Minor Modes
minor mode conventions: Minor Mode Conventions
minor-mode-alist: Mode Line Variables
minor-mode-key-binding: Functions for Key Lookup
minor-mode-list: Minor Modes
minor-mode-map-alist: Controlling Active Maps
minor-mode-overriding-map-alist: Controlling Active Maps
mirroring of characters: Character Properties
missing node, tree-sitter: Accessing Node Information
mkdir: Create/Delete Dirs
mod: Arithmetic Operations
mode: Modes
mode bits: Testing Accessibility
mode help: Mode Help
mode hook: Major Mode Conventions
mode line: Mode Line Format
mode line construct: Mode Line Data
mode loading: Major Mode Conventions
mode variable: Minor Mode Conventions
mode, a buffer display action alist entry: Buffer Display Action Alists
mode-class (property): Major Mode Conventions
mode-line, prefix key: Key Sequence Input
mode-line-buffer-identification: Mode Line Variables
mode-line-client: Mode Line Variables
mode-line-coding-system-map: Standard Keymaps
mode-line-column-line-number-mode-map: Standard Keymaps
mode-line-compact: Mode Line Basics
mode-line-end-spaces: Mode Line Variables
mode-line-format: Mode Line Top
mode-line-format, a window parameter: Window Parameters
mode-line-frame-identification: Mode Line Variables
mode-line-front-space: Mode Line Variables
mode-line-input-method-map: Standard Keymaps
mode-line-misc-info: Mode Line Variables
mode-line-modes: Mode Line Variables
mode-line-modified: Mode Line Variables
mode-line-mule-info: Mode Line Variables
mode-line-percent-position: Mode Line Variables
mode-line-position: Mode Line Variables
mode-line-position-column-format: Mode Line Variables
mode-line-position-column-line-format: Mode Line Variables
mode-line-position-line-format: Mode Line Variables
mode-line-process: Mode Line Variables
mode-line-remote: Mode Line Variables
mode-line-window-selected-p: Mode Line Basics
mode-name: Mode Line Variables
mode-specific commands: Command Modes
mode-specific-map: Prefix Keys
model/view/controller: Abstract Display
modification flag (of buffer): Buffer Modification
modification of lists: Rearrangement
modification time of buffer: Modification Time
modification time of file: File Attributes
modification-hooks (overlay property): Overlay Properties
modification-hooks (text property): Special Properties
modifier bits (of input character): Keyboard Events
modifiers of events: Event Mod
modify a list: List Variables
modify-all-frames-parameters: Parameter Access
modify-category-entry: Categories
modify-frame-parameters: Parameter Access
modify-syntax-entry: Syntax Table Functions
modifying strings: Modifying Strings
module API: Writing Dynamic Modules
module functions: Module Functions
module initialization: Module Initialization
module runtime environment: Module Initialization
module values, conversion: Module Values
module-file-suffix: Dynamic Modules
module-load: Dynamic Modules
modulus: Arithmetic Operations
momentary-string-display: Temporary Displays
monitor change functions: Multiple Terminals
most recently selected windows: Selecting Windows
most recently used window: Cyclic Window Ordering
most-negative-fixnum: Integer Basics
most-positive-fixnum: Integer Basics
motion based on parsing: Motion via Parsing
motion by chars, words, lines, lists: Motion
motion event: Motion Events
mouse click event: Click Events
mouse drag event: Drag Events
mouse dragging parameters: Mouse Dragging Parameters
mouse events, data in: Accessing Mouse
mouse events, in special parts of window or frame: Key Sequence Input
mouse events, repeated: Repeat Events
mouse motion events: Motion Events
mouse pointer shape: Pointer Shape
mouse position: Mouse Position
mouse position list: Click Events
mouse position list, accessing: Accessing Mouse
mouse tracking: Mouse Tracking
mouse wheel event: Click Events
mouse, a face: Font and Color Parameters
mouse, availability: Display Feature Testing
mouse-1: Clickable Text
mouse-1-click-follows-link: Clickable Text
mouse-2: Key Binding Conventions
mouse-absolute-pixel-position: Mouse Position
mouse-action (button property): Button Properties
mouse-appearance-menu-map: Standard Keymaps
mouse-autoselect-window: Mouse Window Auto-selection
mouse-color, a frame parameter: Font and Color Parameters
mouse-face (button property): Button Properties
mouse-face (overlay property): Overlay Properties
mouse-face (text property): Special Properties
mouse-fine-grained-tracking: Motion Events
mouse-leave-buffer-hook: Standard Hooks
mouse-movement-p: Classifying Events
mouse-on-link-p: Clickable Text
mouse-pixel-position: Mouse Position
mouse-position: Mouse Position
mouse-position-function: Mouse Position
mouse-wheel-down-event: Misc Events
mouse-wheel-frame, a frame parameter: Frame Interaction Parameters
mouse-wheel-left-event: Misc Events
mouse-wheel-right-event: Misc Events
mouse-wheel-up-event: Misc Events
move to beginning or end of buffer: Buffer End Motion
move-frame-functions: Frame Position
move-marker: Moving Markers
move-overlay: Managing Overlays
move-point-visually: Bidirectional Display
move-to-column: Columns
move-to-left-margin: Margins
move-to-window-group-line: Screen Lines
move-to-window-group-line-function: Screen Lines
move-to-window-line: Screen Lines
movemail: Subprocess Creation
moveto: SVG Images
moving across syntax classes: Motion and Syntax
moving markers: Moving Markers
MS-DOS and file modes: Testing Accessibility
MS-Windows file-name syntax: File Names
mule-keymap: Prefix Keys
multi-file package: Multi-file Packages
multi-frame images: Multi-Frame Images
multi-message-max: Displaying Messages
multi-message-timeout: Displaying Messages
multi-mode indentation: Mode-Specific Indent
multi-monitor: Multiple Terminals
multi-query-replace-map: Search and Replace
multi-tty: Multiple Terminals
multibyte characters: Non-ASCII Characters
multibyte text: Text Representations
multibyte-char-to-unibyte: Converting Representations
multibyte-string-p: Text Representations
multibyte-syntax-as-symbol: Control Parsing
multiline font lock: Multiline Font Lock
multiple languages, parsing with tree-sitter: Multiple Languages
multiple terminals: Multiple Terminals
multiple windows: Basic Windows
multiple X displays: Multiple Terminals
multiple yes-or-no questions: Multiple Queries
multiple-dispatch methods: Generic Functions
multiple-frames: Frame Titles
multisession variable: Multisession Variables
multisession-delete: Multisession Variables
multisession-directory: Multisession Variables
multisession-edit-mode: Multisession Variables
multisession-storage: Multisession Variables
multisession-value: Multisession Variables
mutable lists: Modifying Lists
mutable objects: Mutability
mutex-lock: Mutexes
mutex-name: Mutexes
mutex-unlock: Mutexes
mutexp: Mutexes
mwheel-coalesce-scroll-events: Misc Events

N
n-p-gp: Parser-based Indentation
name, a frame parameter: Basic Parameters
named function: Function Names
named node, tree-sitter: Language Grammar
named-let: Local Variables
namespace etiquette: Shorthands
namespaces: Shorthands
namespacing: Shorthands
naming backup files: Backup Names
NaN: Float Basics
narrow-map: Standard Keymaps
narrow-to-page: Narrowing
narrow-to-region: Narrowing
narrowing: Narrowing
native code: Native Compilation
native compilation: Native Compilation
native compilation, prevent writing *.eln files: Native Compilation
native edges: Frame Layout
native frame: Frame Layout
native height: Frame Layout
native position: Frame Layout
native size: Frame Layout
native width: Frame Layout
native-comp-async-jobs-number: Native-Compilation Variables
native-comp-async-query-on-exit: Native-Compilation Variables
native-comp-async-report-warnings-errors: Native-Compilation Variables
native-comp-available-p: Native-Compilation Functions
native-comp-debug: Native-Compilation Variables
native-comp-eln-load-path: Library Search
native-comp-enable-subr-trampolines: Native-Compilation Variables
native-comp-jit-compilation: Native-Compilation Variables
native-comp-limple-mode: Native-Compilation Functions
native-comp-speed: Native-Compilation Variables
native-comp-verbose: Native-Compilation Functions
native-comp-verbose: Native-Compilation Variables
native-compilation functions: Native-Compilation Functions
native-compilation variables: Native-Compilation Variables
native-compile: Native-Compilation Functions
native-compile, a Lisp feature: Native Compilation
native-compile-async: Native-Compilation Functions
natnump: Predicates on Numbers
natural numbers: Predicates on Numbers
nbutlast: List Elements
nconc: Rearrangement
negative infinity: Float Basics
negative-argument: Prefix Command Arguments
nest frame: Child Frames
network byte ordering, in Bindat specification: Bindat Types
network connection: Network
network connection, encrypted: Network
network servers: Network Servers
network service name, and default coding system: Default Coding Systems
network-coding-system-alist: Default Coding Systems
network-interface-info: Misc Network
network-interface-list: Misc Network
network-lookup-address-info: Misc Network
network-stream-use-client-certificates: Network
new file message: Subroutines of Visiting
newline: Basic Char Syntax
newline: Commands for Insertion
newline and Auto Fill mode: Commands for Insertion
newline in print: Output Functions
newline in strings: Syntax for Strings
newline-and-indent: Mode-Specific Indent
next input: Event Input Misc
next-button: Button Buffer Commands
next-char-property-change: Property Search
next-complete-history-element: Minibuffer Commands
next-frame: Finding All Frames
next-history-element: Minibuffer Commands
next-matching-history-element: Minibuffer Commands
next-overlay-change: Finding Overlays
next-property-change: Property Search
next-screen-context-lines: Textual Scrolling
next-single-char-property-change: Property Search
next-single-property-change: Property Search
next-window: Cyclic Window Ordering
nil: nil and t
nil as a list: Box Diagrams
nil in keymap: Key Lookup
nil input stream: Input Streams
nil output stream: Output Streams
nlistp: List-related Predicates
no-accept-focus, a frame parameter: Management Parameters
no-byte-compile: Byte Compilation
no-catch: Catch and Throw
no-conversion coding system: Coding System Basics
no-delete-other-windows, a window parameter: Window Parameters
no-focus-on-map, a frame parameter: Management Parameters
no-indent: Parser-based Indentation
no-native-compile: Native Compilation
no-node: Parser-based Indentation
no-other-frame, a frame parameter: Frame Interaction Parameters
no-other-window, a window parameter: Window Parameters
no-redraw-on-reenter: Refresh Screen
no-self-insert property: Defining Abbrevs
no-special-glyphs, a frame parameter: Layout Parameters
node types, in a syntax tree: Language Grammar
node, ewoc: Abstract Display
node-is: Parser-based Indentation
nodes, by field name: Retrieving Nodes
nodes, by kinship: Retrieving Nodes
nodes, by position: Retrieving Nodes
non-ASCII characters: Non-ASCII Characters
non-ASCII characters in loaded files: Loading Non-ASCII
non-ASCII text in key bindings: Key Binding Commands
non-capturing group: Regexp Backslash
non-GNU ELPA: Package Archives
non-greedy repetition characters in regexp: Regexp Special
nondirectory part (of file name): File Name Components
noninteractive: Batch Mode
nonl in rx: Rx Constructs
nonlocal exits: Nonlocal Exits
nonlocal exits, cleaning up: Cleanups
nonlocal exits, in modules: Module Nonlocal
nonprinting characters, reading: Quoted Character Input
non_local_exit_check: Module Nonlocal
non_local_exit_clear: Module Nonlocal
non_local_exit_get: Module Nonlocal
non_local_exit_signal: Module Nonlocal
non_local_exit_throw: Module Nonlocal
noreturn: Test Coverage
normal hook: Hooks
normal-auto-fill-function: Auto Filling
normal-backup-enable-predicate: Making Backups
normal-mode: Auto Major Mode
not: Combining Conditions
not in rx: Rx Constructs
not recording input events: Event Input Misc
not-modified: Buffer Modification
not-newline in rx: Rx Constructs
not-word-boundary in rx: Rx Constructs
notation: Evaluation Notation
notifications, on desktop: Desktop Notifications
notifications-close-notification: Desktop Notifications
notifications-get-capabilities: Desktop Notifications
notifications-get-server-information: Desktop Notifications
notifications-notify: Desktop Notifications
notifiers, tree-sitter: Using Parser
nreverse: Sequence Functions
ns-appearance, a frame parameter: Management Parameters
ns-transparent-titlebar, a frame parameter: Management Parameters
ntake: List Elements
nth: List Elements
nth-sibling: Parser-based Indentation
nthcdr: List Elements
null: List-related Predicates
null bytes, and decoding text: Lisp and Coding Systems
null-device: System Environment
null-device: System Environment
num-input-keys: Key Sequence Input
num-nonmacro-input-events: Reading One Event
num-processors: Process Information
number comparison: Comparison of Numbers
number conversions: Numeric Conversions
number of bignum bits, limit on: Integer Basics
number-or-marker-p: Predicates on Markers
number-sequence: Building Lists
number-to-string: String Conversion
numbered backups: Numbered Backups
numberp: Predicates on Numbers
numbers: Numbers
numeric prefix argument: Prefix Command Arguments
numeric prefix argument usage: Interactive Codes
numerical RGB color specification: Color Names

O
obarray: Creating Symbols
obarray: Creating Symbols
obarray in completion: Basic Completion
object: Lisp Data Types
object internals: Object Internals
object to string: Output Functions
object-intervals: Examining Properties
objects with identical contents, and byte-compiler: Equality Predicates
obsolete (declare spec): Declare Form
obsolete functions: Obsolete Functions
oclosure-define: OClosures
oclosure-interactive-form: Using Interactive
oclosure-lambda: OClosures
oclosure-type: OClosures
oclosures: OClosures
octal character code: General Escape Syntax
octal character input: Quoted Character Input
octal escapes: Usual Display
octal numbers: Integer Basics
old advices, porting: Porting Old Advice
old-selected-frame: Window Hooks
old-selected-window: Window Hooks
one-or-more in rx: Rx Constructs
one-window-p: Cyclic Window Ordering
only-global-abbrevs: Defining Abbrevs
opacity, frame: Font and Color Parameters
opacity, frame: Font and Color Parameters
open closures: OClosures
open-dribble-file: Recording Input
open-network-stream: Network
open-paren-in-column-0-is-defun-start: List Motion
open-termscript: Terminal Output
OpenType font: Low-Level Font
open_channel: Module Misc
operating system environment: System Environment
operating system signal: Killing Emacs
operations (property): Magic File Names
operations on images: Showing Images
opt in rx: Rx Constructs
optimize regexp: Regexp Functions
option descriptions: A Sample Variable Description
optional arguments: Argument List
optional in rx: Rx Constructs
options on command line: Command-Line Arguments
options, defcustom keyword: Variable Definitions
or: Combining Conditions
or in rx: Rx Constructs
ordering of windows, cyclic: Cyclic Window Ordering
origin of display: Frame Layout
other-buffer: Buffer List
other-window: Cyclic Window Ordering
other-window, a window parameter: Window Parameters
other-window-scroll-buffer: Textual Scrolling
outdated node, tree-sitter: Accessing Node Information
outer border: Frame Layout
outer edges: Frame Layout
outer frame: Frame Layout
outer height: Frame Layout
outer position: Frame Layout
outer size: Frame Layout
outer width: Frame Layout
outer-window-id, a frame parameter: Management Parameters
output from processes: Output from Processes
output stream: Output Streams
output variables, overriding: Output Overrides
output-controlling variables: Output Variables
overall prompt string: Format of Keymaps
overflow-newline-into-fringe: Fringe Cursors
overlay properties: Overlay Properties
overlay, empty: Managing Overlays
overlay-arrow-position: Overlay Arrow
overlay-arrow-string: Overlay Arrow
overlay-arrow-variable-list: Overlay Arrow
overlay-buffer: Managing Overlays
overlay-end: Managing Overlays
overlay-get: Overlay Properties
overlay-properties: Overlay Properties
overlay-put: Overlay Properties
overlay-start: Managing Overlays
overlayp: Managing Overlays
overlays: Overlays
overlays, managing: Managing Overlays
overlays, searching for: Finding Overlays
overlays-at: Finding Overlays
overlays-in: Finding Overlays
overlined text: Face Attributes
override existing functions: Defining Functions
override redirect frames: Management Parameters
override spec (for a face): Defining Faces
override-redirect, a frame parameter: Management Parameters
overrides, in output functions: Output Overrides
overriding bidirectional properties: Bidirectional Display
overriding-local-map: Controlling Active Maps
overriding-local-map-menu-flag: Controlling Active Maps
overriding-terminal-local-map: Controlling Active Maps
overwrite-mode: Commands for Insertion

P
package: Packaging
package archive: Package Archives
package archive security: Package Archives
package attributes: Packaging Basics
package autoloads: Packaging Basics
package dependencies: Packaging Basics
package name: Packaging Basics
package signing: Package Archives
package version: Packaging Basics
package-activate-all: Packaging Basics
package-archives: Package Archives
package-initialize: Packaging Basics
package-version, customization keyword: Common Keywords
packing: Byte Packing
padding: Formatting Strings
page-delimiter: Standard Regexps
paragraph-separate: Standard Regexps
paragraph-separate, and bidirectional display: Bidirectional Display
paragraph-start: Standard Regexps
paragraph-start, and bidirectional display: Bidirectional Display
parameters for moving frames with the mouse: Mouse Dragging Parameters
parameters for resizing frames with the mouse: Mouse Dragging Parameters
parameters of initial frame: Initial Parameters
parent: Parser-based Indentation
parent directory of file: Contents of Directories
parent frames: Child Frames
parent of char-table: Char-Tables
parent process: Processes
parent window: Windows and Frames
parent window: Windows and Frames
parent-bol: Parser-based Indentation
parent-frame, a buffer display action alist entry: Buffer Display Action Alists
parent-frame, a frame parameter: Frame Interaction Parameters
parent-is: Parser-based Indentation
parenthesis: Cons Cell Type
parenthesis depth: Low-Level Parsing
parenthesis matching: Blinking
parenthesis mismatch, debugging: Syntax Errors
parity, in serial connections: Serial Ports
parse state for a position: Position Parse
parse string, tree-sitter: Using Parser
parse-colon-path: System Environment
parse-partial-sexp: Low-Level Parsing
parse-sexp-ignore-comments: Control Parsing
parse-sexp-lookup-properties: Syntax Properties
parse-time-string: Time Parsing
parser state: Parser State
parser-based font-lock: Parser-based Font Lock
parser-based indentation: Parser-based Indentation
parsing buffer text: Syntax Tables
parsing expressions: Parsing Expressions
parsing html: Parsing HTML/XML
parsing multiple languages with tree-sitter: Multiple Languages
parsing xml: Parsing HTML/XML
parsing, control parameters: Control Parsing
partial application of functions: Calling Functions
partial-width windows: Truncation
passwords, reading: Reading a Password
PATH environment variable: Subprocess Creation
path-separator: System Environment
path-separator: System Environment
pattern matching with tree-sitter nodes: Pattern Matching
pattern matching, programming style: Pattern-Matching Conditional
pattern syntax, tree-sitter query: Pattern Matching
patterns, tree-sitter, in string form: Pattern Matching
PBM: Other Image Types
pcase: pcase Macro
pcase: Pattern-Matching Conditional
pcase pattern: pcase Macro
pcase, defining new kinds of patterns: Extending pcase
pcase-defmacro: Extending pcase
pcase-dolist: Destructuring with pcase Patterns
pcase-lambda: Destructuring with pcase Patterns
pcase-let: Destructuring with pcase Patterns
pcase-let*: Destructuring with pcase Patterns
pcase-setq: Destructuring with pcase Patterns
pdumper-stats: Building Emacs
peculiar error: Error Symbols
peeking at input: Event Input Misc
percent symbol in mode line: Mode Line Data
perform-replace: Search and Replace
performance analysis: Profiling
performance analysis (Edebug): Coverage Testing
permanent local variable: Creating Buffer-Local
permanently-enabled-local-variables: File Local Variables
permissions, file: Testing Accessibility
permissions, file: Changing Files
persistent window parameters: Window Parameters
phishing using directional overrides: Bidirectional Display
physical lines, moving by: Text Lines
piece of advice: Advising Functions
pinch event: Misc Events
pipe, when to use for subprocess communications: Asynchronous Processes
pixel height of a window: Window Sizes
pixel width of a window: Window Sizes
pixel-fill-region: Filling
pixel-fill-width: Filling
pixel-resolution wheel events: Misc Events
pixelwise, resizing windows: Resizing Windows
place form: Generalized Variables
play-sound: Sound Output
play-sound-file: Sound Output
play-sound-functions: Sound Output
plist: Property Lists
plist access: Plist Access
plist vs. alist: Plists and Alists
plist-get: Plist Access
plist-member: Plist Access
plist-put: Plist Access
plistp: Property Lists
plugin_is_GPL_compatible: Dynamic Modules
point: Point
point: Point
point excursion: Excursions
point excursion: Excursions
point in rx: Rx Constructs
point in window: Window Point
point with narrowing: Point
point-entered (text property): Special Properties
point-left (text property): Special Properties
point-marker: Creating Markers
point-max: Point
point-max-marker: Creating Markers
point-min: Point
point-min-marker: Creating Markers
pointer (text property): Special Properties
pointer shape: Pointer Shape
pointers: Cons Cell Type
polymorphism: Generic Functions
pop: List Elements
pop-mark: The Mark
pop-to-buffer: Switching Buffers
pop-up-frame-alist: Choosing Window Options
pop-up-frame-function: Choosing Window Options
pop-up-frame-parameters, a buffer display action alist entry: Buffer Display Action Alists
pop-up-frames: Choosing Window Options
pop-up-frames, replacement for: Choosing Window Options
pop-up-windows: Choosing Window Options
pop-up-windows, replacement for: Choosing Window Options
popcount: Bitwise Operations
port number, and default coding system: Default Coding Systems
pos-bol: Text Lines
pos-eol: Text Lines
pos-visible-in-window-group-p: Window Start and End
pos-visible-in-window-group-p-function: Window Start and End
pos-visible-in-window-p: Window Start and End
position (in buffer): Positions
position argument: Interactive Codes
position in window: Window Point
position of frame: Frame Geometry
position of frame: Frame Position
position of mouse: Mouse Position
position-bytes: Text Representations
position-symbol: Symbols with Position
positive infinity: Float Basics
posix-looking-at: POSIX Regexps
posix-search-backward: POSIX Regexps
posix-search-forward: POSIX Regexps
posix-string-match: POSIX Regexps
posn-actual-col-row: Accessing Mouse
posn-area: Accessing Mouse
posn-at-point: Accessing Mouse
posn-at-x-y: Accessing Mouse
posn-col-row: Accessing Mouse
posn-image: Accessing Mouse
posn-object: Accessing Mouse
posn-object-width-height: Accessing Mouse
posn-object-x-y: Accessing Mouse
posn-point: Accessing Mouse
posn-string: Accessing Mouse
posn-timestamp: Accessing Mouse
posn-window: Accessing Mouse
posn-x-y: Accessing Mouse
posnp: Accessing Mouse
post-command-hook: Command Overview
post-gc-hook: Garbage Collection
post-self-insert-hook: Commands for Insertion
pp: Output Functions
pre-command-hook: Command Overview
pre-redisplay-function: Forcing Redisplay
pre-redisplay-functions: Forcing Redisplay
precedence of buffer display action functions: Precedence of Action Functions
preceding-char: Near Point
precision in format specifications: Formatting Strings
predicates for lists: List-related Predicates
predicates for markers: Predicates on Markers
predicates for numbers: Predicates on Numbers
predicates for strings: Predicates for Strings
predicates for syntax tree nodes: Accessing Node Information
preedit-text event: Misc Events
prefer-utf-8 coding system: Coding System Basics
prefix argument: Prefix Command Arguments
prefix argument unreading: Event Input Misc
prefix command: Prefix Keys
prefix key: Prefix Keys
prefix, defgroup keyword: Group Definitions
prefix-arg: Prefix Command Arguments
prefix-command-echo-keystrokes-functions: Standard Hooks
prefix-command-preserve-state-hook: Standard Hooks
prefix-help-command: Help Functions
prefix-numeric-value: Prefix Command Arguments
preloaded Lisp files: Building Emacs
preloaded-file-list: Building Emacs
preloading additional functions and variables: Building Emacs
prepare-change-group: Atomic Changes
preserve-size, a buffer display action alist entry: Buffer Display Action Alists
preserving window sizes: Preserving Window Sizes
pretty-printer: Output Functions
prev-adaptive-prefix: Parser-based Indentation
prev-line: Parser-based Indentation
prev-sibling: Parser-based Indentation
prevent warnings in init files: Warning Options
preventing backtracking: Specification List
preventing prefix key: Key Lookup
preventing quitting: Quitting
previous complete subexpression: Parser State
previous-button: Button Buffer Commands
previous-char-property-change: Property Search
previous-complete-history-element: Minibuffer Commands
previous-frame: Finding All Frames
previous-history-element: Minibuffer Commands
previous-matching-history-element: Minibuffer Commands
previous-overlay-change: Finding Overlays
previous-property-change: Property Search
previous-single-char-property-change: Property Search
previous-single-property-change: Property Search
previous-window: Cyclic Window Ordering
previous-window, a buffer display action alist entry: Buffer Display Action Alists
primary selection: Window System Selections
primitive: What Is a Function
primitive function: Primitive Function Type
primitive function internals: Writing Emacs Primitives
primitive type: Lisp Data Types
primitive-undo: Undo
prin1: Output Functions
prin1-to-string: Output Functions
princ: Output Functions
print: Output Functions
print example: Output Streams
print name cell: Symbol Components
print-charset-text-property: Output Variables
print-circle: Output Variables
print-continuous-numbering: Output Variables
print-escape-control-characters: Output Variables
print-escape-multibyte: Output Variables
print-escape-newlines: Output Variables
print-escape-nonascii: Output Variables
print-gensym: Output Variables
print-integers-as-characters: Output Variables
print-length: Output Variables
print-level: Output Variables
print-number-table: Output Variables
print-quoted: Output Variables
print-symbols-bare: Symbols with Position
print-unreadable-function: Output Variables
printable ASCII characters: Usual Display
printable-chars: Character Properties
printed representation: Printed Representation
printed representation for characters: Basic Char Syntax
printing: Streams Intro
printing (Edebug): Printing in Edebug
printing circular structures: Printing in Edebug
printing limits: Output Variables
printing notation: Printing Notation
priority (overlay property): Overlay Properties
priority order of coding systems: Specifying Coding Systems
process: Processes
process creation: Subprocess Creation
process filter: Filter Functions
process filter multibyte flag: Decoding Output
process information: Process Information
process input: Input to Processes
process internals: Process Internals
process output: Output from Processes
process sentinel: Sentinels
process signals: Signals to Processes
process-adaptive-read-buffering: Output from Processes
process-attributes: System Processes
process-buffer: Process Buffers
process-coding-system: Process Information
process-coding-system-alist: Default Coding Systems
process-command: Process Information
process-connection-type: Asynchronous Processes
process-contact: Process Information
process-datagram-address: Datagrams
process-environment: System Environment
process-error-pause-time: Asynchronous Processes
process-exit-status: Process Information
process-file: Synchronous Processes
process-file-return-signal-string: Synchronous Processes
process-file-shell-command: Synchronous Processes
process-file-side-effects: Synchronous Processes
process-filter: Filter Functions
process-get: Process Information
process-id: Process Information
process-kill-buffer-query-function: Process Buffers
process-lines: Synchronous Processes
process-lines-ignore-status: Synchronous Processes
process-list: Process Information
process-live-p: Process Information
process-mark: Process Buffers
process-name: Process Information
process-plist: Process Information
process-put: Process Information
process-query-on-exit-flag: Query Before Exit
process-running-child-p: Input to Processes
process-send-eof: Input to Processes
process-send-region: Input to Processes
process-send-string: Input to Processes
process-sentinel: Sentinels
process-status: Process Information
process-thread: Processes and Threads
process-tty-name: Process Information
process-type: Process Information
processes, threads: Processes and Threads
processing of errors: Processing of Errors
processor run time: Processor Run Time
processp: Processes
process_input: Module Misc
profile: Profiling
profiler-report: Profiling
profiler-start: Profiling
profiler-stop: Profiling
profiling: Profiling
prog-first-column: Mode-Specific Indent
prog-indentation-context: Mode-Specific Indent
prog-mode: Basic Major Modes
prog-mode, and bidi-paragraph-direction: Bidirectional Display
prog-mode-hook: Basic Major Modes
prog-mode-map: Standard Keymaps
prog1: Sequencing
prog2: Sequencing
progn: Sequencing
program arguments: Subprocess Creation
program directories: Subprocess Creation
program name, and default coding system: Default Coding Systems
programmed completion: Programmed Completion
programming conventions: Programming Tips
programming types: Programming Types
progress reporting: Progress
progress-reporter-done: Progress
progress-reporter-force-update: Progress
progress-reporter-update: Progress
prompt for file name: Reading File Names
prompt string (of menu): Defining Menus
prompt string of keymap: Format of Keymaps
proper list: Cons Cells
proper-list-p: List-related Predicates
properties of text: Text Properties
propertize: Changing Properties
property category of text character: Special Properties
property list: Property Lists
property list cell: Symbol Components
property lists vs association lists: Plists and Alists
protect C variables from garbage collection: Writing Emacs Primitives
protected forms: Cleanups
provide: Named Features
provide-theme: Custom Themes
providing features: Named Features
pseudo window: Window Internals
pty, when to use for subprocess communications: Asynchronous Processes
pure function: What Is a Function
pure property: Standard Properties
pure storage: Pure Storage
pure-bytes-used: Pure Storage
purecopy: Pure Storage
purify-flag: Pure Storage
push: List Variables
push-button: Button Buffer Commands
push-mark: The Mark
put: Symbol Plists
put-char-code-property: Character Properties
put-charset-property: Character Sets
put-image: Showing Images
put-text-property: Changing Properties
puthash: Hash Access

Q
quadratic-bezier-curveto: SVG Images
quantify node, tree-sitter: Pattern Matching
queries, compiling: Pattern Matching
query: Parser-based Indentation
query functions, tree-sitter: Pattern Matching
query, tree-sitter: Pattern Matching
query-font: Low-Level Font
query-replace-history: Minibuffer History
query-replace-map: Search and Replace
querying the user: Yes-or-No Queries
question mark in character constant: Basic Char Syntax
quietly-read-abbrev-file: Abbrev Files
quit-flag: Quitting
quit-process: Signals to Processes
quit-restore, a window parameter: Window Parameters
quit-restore-window: Quitting Windows
quit-window: Quitting Windows
quit-window-hook: Quitting Windows
quitting: Quitting
quitting from infinite loop: Infinite Loops
quitting windows: Quitting Windows
quote: Quoting
quote character: Parser State
quote special characters in regexp: Regexp Functions
quoted character input: Quoted Character Input
quoted-insert suppression: Changing Key Bindings
quoting and unquoting command-line arguments: Shell Arguments
quoting apostrophe and grave accent in doc strings: Documentation Tips
quoting characters in printing: Output Functions
quoting using apostrophe: Quoting

R
radio, customization types: Composite Types
radix for reading an integer: Integer Basics
raise-frame: Raising and Lowering
raising a frame: Raising and Lowering
random: Random Numbers
random numbers: Random Numbers
rassoc: Association Lists
rassq: Association Lists
rassq-delete-all: Association Lists
raw prefix argument: Prefix Command Arguments
raw prefix argument usage: Interactive Codes
raw syntax descriptor: Syntax Table Internals
raw-text coding system: Coding System Basics
re-builder: Regular Expressions
re-search-backward: Regexp Search
re-search-forward: Regexp Search
read: Input Functions
read command name: Interactive Call
read file names: Reading File Names
read input: Reading Input
read syntax: Printed Representation
read syntax for characters: Basic Char Syntax
read-answer: Multiple Queries
read-answer-short: Multiple Queries
read-buffer: High-Level Completion
read-buffer-completion-ignore-case: High-Level Completion
read-buffer-function: High-Level Completion
read-char: Reading One Event
read-char-choice: Reading One Event
read-char-choice-use-read-key: Reading One Event
read-char-exclusive: Reading One Event
read-char-from-minibuffer: Multiple Queries
read-circle: Input Functions
read-coding-system: User-Chosen Coding Systems
read-color: High-Level Completion
read-command: High-Level Completion
read-directory-name: Reading File Names
read-event: Reading One Event
read-expression-history: Minibuffer History
read-extended-command-predicate: Command Modes
read-extended-command-predicate: Interactive Call
read-file-modes: Changing Files
read-file-name: Reading File Names
read-file-name-completion-ignore-case: Reading File Names
read-file-name-function: Reading File Names
read-from-minibuffer: Text from Minibuffer
read-from-string: Input Functions
read-hide-char: Reading a Password
read-input-method-name: Input Methods
read-kbd-macro: Describing Characters
read-key: Reading One Event
read-key-sequence: Key Sequence Input
read-key-sequence-vector: Key Sequence Input
read-minibuffer: Object from Minibuffer
read-minibuffer-restore-windows: Text from Minibuffer
read-multiple-choice: Reading One Event
read-no-blanks-input: Text from Minibuffer
read-non-nil-coding-system: User-Chosen Coding Systems
read-number-history: Minibuffer History
read-only (text property): Special Properties
read-only buffer: Read Only Buffers
read-only buffers in interactive: Using Interactive
read-only character: Special Properties
read-only variables: Constant Variables
read-only-mode: Read Only Buffers
read-passwd: Reading a Password
read-positioning-symbols: Input Functions
read-quoted-char: Quoted Character Input
read-quoted-char quitting: Quitting
read-regexp: Text from Minibuffer
read-regexp-case-fold-search: Text from Minibuffer
read-regexp-defaults-function: Text from Minibuffer
read-shell-command: Reading File Names
read-string: Text from Minibuffer
read-string-from-buffer: Text from Minibuffer
read-symbol-shorthands: Shorthands
read-variable: High-Level Completion
read-variable, history list: Minibuffer History
readable syntax: Input Functions
readablep: Input Functions
reading: Streams Intro
reading a single event: Reading One Event
reading from files: Reading from Files
reading from minibuffer with completion: Minibuffer Completion
reading grammar definition, tree-sitter: Language Grammar
reading interactive arguments: Interactive Codes
reading numbers in hex, octal, and binary: Integer Basics
reading order: Bidirectional Display
reading symbols: Creating Symbols
real-last-command: Command Loop Info
rear-nonsticky text property: Sticky Properties
rear-nonsticky, and cursor-intangible property: Special Properties
rearrangement of lists: Rearrangement
rebinding: Changing Key Bindings
recent-auto-save-p: Auto-Saving
recent-keys: Recording Input
recenter: Textual Scrolling
recenter-positions: Textual Scrolling
recenter-redisplay: Textual Scrolling
recenter-top-bottom: Textual Scrolling
recenter-window-group: Textual Scrolling
recenter-window-group-function: Textual Scrolling
recombining windows: Recombining Windows
record: Record Functions
record command history: Interactive Call
recording input: Recording Input
recordp: Record Functions
records: Records
rectangle, as contents of a register: Registers
recursion: Iteration
recursion-depth: Recursive Editing
recursive command loop: Recursive Editing
recursive editing level: Recursive Editing
recursive evaluation: Intro Eval
recursive minibuffers: Recursive Mini
recursive traverse of directory tree: Contents of Directories
recursive-edit: Recursive Editing
redefine existing functions: Defining Functions
redirect-frame-focus: Input Focus
redisplay: Forcing Redisplay
redisplay errors: Debugging Redisplay
redo: Undo
redraw-display: Refresh Screen
redraw-frame: Refresh Screen
reducing sequences: Sequence Functions
reference to free variable, compilation warning: Compiler Errors
references, following: Key Binding Conventions
refresh the screen: Refresh Screen
regex in rx: Rx Constructs
regexp: Regular Expressions
regexp alternative: Regexp Backslash
regexp grouping: Regexp Backslash
regexp in rx: Rx Constructs
regexp searching: Regexp Search
regexp stack overflow: Regexp Problems
regexp syntax: Syntax of Regexps
regexp syntax: Rx Notation
regexp, special characters in: Regexp Special
regexp-history: Minibuffer History
regexp-opt: Regexp Functions
regexp-opt-charset: Regexp Functions
regexp-opt-depth: Regexp Functions
regexp-quote: Regexp Functions
regexp-unmatchable: Regexp Functions
regexps used standardly in editing: Standard Regexps
region: The Region
region argument: Interactive Codes
region-beginning: The Region
region-end: The Region
register preview: Registers
register-alist: Registers
register-definition-prefixes: Autoload by Prefix
register-read-with-preview: Registers
registers: Registers
regular expression: Regular Expressions
regular expression problems: Regexp Problems
regular expression searching: Regexp Search
regular expressions, developing: Regular Expressions
reindent-then-newline-and-indent: Mode-Specific Indent
relative file name: Relative File Names
relative remapping, faces: Face Remapping
remainder: Arithmetic Operations
remapping commands: Remapping Commands
remhash: Hash Access
remote-file-name-inhibit-cache: Magic File Names
remote-file-name-inhibit-locks: File Locks
remove: Sets And Lists
remove-from-invisibility-spec: Invisible Text
remove-function: Core Advising Primitives
remove-hook: Setting Hooks
remove-images: Showing Images
remove-list-of-text-properties: Changing Properties
remove-overlays: Managing Overlays
remove-text-properties: Changing Properties
remove-variable-watcher: Watching Variables
removing from sequences: Sequence Functions
removing from sequences: Sequence Functions
remq: Sets And Lists
rename-auto-save-file: Auto-Saving
rename-buffer: Buffer Names
rename-file: Changing Files
rendering html: Parsing HTML/XML
reordering, of bidirectional text: Bidirectional Display
reordering, of elements in lists: Rearrangement
reparent frame: Child Frames
repeat events: Repeat Events
repeat in rx: Rx Constructs
repeat-mode: Creating Keymaps
repeatable key bindings: Creating Keymaps
repeated loading: Repeated Loading
replace bindings: Changing Key Bindings
replace characters: Substitution
replace characters in region: Substitution
replace characters in string: Substitution
replace list element: Setcar
replace matched text: Replacing Match
replace part of list: Setcdr
replace-buffer-contents: Replacing
replace-buffer-in-windows: Buffers and Windows
replace-match: Replacing Match
replace-re-search-function: Search and Replace
replace-regexp-in-region: Search and Replace
replace-regexp-in-string: Search and Replace
replace-region-contents: Replacing
replace-search-function: Search and Replace
replace-string-in-region: Search and Replace
replacement after search: Search and Replace
replacing display specs: Replacing Specs
require: Named Features
require, customization keyword: Common Keywords
require-final-newline: Saving Buffers
require-theme: Custom Themes
requiring features: Named Features
reserved keys: Key Binding Conventions
reset, face attribute value: Face Attributes
resize window: Resizing Windows
resize-mini-frames: Minibuffer Windows
resize-mini-windows: Minibuffer Windows
rest arguments: Argument List
restacking a frame: Raising and Lowering
restart-emacs: Killing Emacs
restore-buffer-modified-p: Buffer Modification
restricted-sexp, customization types: Composite Types
restriction (in a buffer): Narrowing
resume (cf. no-redraw-on-reenter): Refresh Screen
resume-tty: Suspending Emacs
resume-tty-functions: Suspending Emacs
rethrow a signal: Handling Errors
retrieve node, tree-sitter: Retrieving Nodes
retrieving tree-sitter nodes: Retrieving Nodes
return (ASCII character): Basic Char Syntax
return value: What Is a Function
reusable-frames, a buffer display action alist entry: Buffer Display Action Alists
reverse: Sequence Functions
reversing a list: Sequence Functions
reversing a string: Sequence Functions
reversing a vector: Sequence Functions
revert-buffer: Reverting
revert-buffer-function: Reverting
revert-buffer-in-progress-p: Reverting
revert-buffer-insert-file-contents-function: Reverting
revert-without-query: Reverting
reverting buffers: Reverting
rgb value: Color Names
right dividers: Window Dividers
right-divider, prefix key: Key Sequence Input
right-divider-width, a frame parameter: Layout Parameters
right-fringe, a frame parameter: Layout Parameters
right-fringe, prefix key: Key Sequence Input
right-fringe-width: Fringe Size/Pos
right-margin, prefix key: Key Sequence Input
right-margin-width: Display Margins
right-to-left text: Bidirectional Display
ring data structure: Rings
ring-bell-function: Beeping
ring-copy: Rings
ring-elements: Rings
ring-empty-p: Rings
ring-insert: Rings
ring-insert-at-beginning: Rings
ring-length: Rings
ring-p: Rings
ring-ref: Rings
ring-remove: Rings
ring-resize: Rings
ring-size: Rings
risky, defcustom keyword: Variable Definitions
risky-local-variable-p: File Local Variables
RLO: Bidirectional Display
rm: Changing Files
root window: Windows and Frames
root window of atomic window: Atomic Windows
round: Numeric Conversions
rounding in conversions: Numeric Conversions
rounding without conversion: Rounding Operations
rplaca: Modifying Lists
rplacd: Modifying Lists
run time stack: Internals of Debugger
run-at-time: Timers
run-hook-with-args: Running Hooks
run-hook-with-args-until-failure: Running Hooks
run-hook-with-args-until-success: Running Hooks
run-hooks: Running Hooks
run-mode-hooks: Mode Hooks
run-with-idle-timer: Idle Timers
run-with-timer: Timers
running a hook when a window gets selected: Selecting Windows
rx: Rx Functions
rx: Rx Notation
rx-define: Extending Rx
rx-let: Extending Rx
rx-let-eval: Extending Rx
rx-to-string: Rx Functions

S
S-expression: Intro Eval
safe local variable: File Local Variables
safe, defcustom keyword: Variable Definitions
safe-length: List Elements
safe-local-eval-forms: File Local Variables
safe-local-variable, property of variable: File Local Variables
safe-local-variable-p: File Local Variables
safe-local-variable-values: File Local Variables
safe-magic (property): Magic File Names
safely encode a string: Lisp and Coding Systems
safely encode characters in a charset: Lisp and Coding Systems
safely encode region: Lisp and Coding Systems
safety of functions: Function Safety
same-window-buffer-names, replacement for: Choosing Window Options
same-window-regexps, replacement for: Choosing Window Options
save abbrevs in files: Abbrev Files
save-abbrevs: Abbrev Files
save-buffer: Saving Buffers
save-buffer-coding-system: Encoding and I/O
save-current-buffer: Current Buffer
save-excursion: Excursions
save-mark-and-excursion: Excursions
save-match-data: Saving Match Data
save-restriction: Narrowing
save-selected-window: Selecting Windows
save-some-buffers: Saving Buffers
save-some-buffers-default-predicate: Saving Buffers
save-window-excursion: Window Configurations
saved-face (face symbol property): Defining Faces
SaveUnder feature: Display Feature Testing
saving buffers: Saving Buffers
saving text properties: Format Conversion
saving window information: Window Configurations
scalable fonts: Font Selection
scalable-fonts-allowed: Font Selection
scan-lists: Motion via Parsing
scan-sexps: Motion via Parsing
scanning expressions: Parsing Expressions
scanning for character sets: Scanning Charsets
scanning keymaps: Scanning Keymaps
scope: Variable Scoping
scoping rule: Variable Scoping
screen layout: Frame Configuration Type
screen lines, moving by: Screen Lines
screen of terminal: Basic Windows
screen refresh: Refresh Screen
screen-gamma, a frame parameter: Font and Color Parameters
script symbols: Character Properties
scroll bar events, data in: Accessing Scroll
scroll bars: Scroll Bars
scroll-bar-background, a frame parameter: Font and Color Parameters
scroll-bar-event-ratio: Accessing Scroll
scroll-bar-foreground, a frame parameter: Font and Color Parameters
scroll-bar-height: Scroll Bars
scroll-bar-height, a frame parameter: Layout Parameters
scroll-bar-mode: Scroll Bars
scroll-bar-scale: Accessing Scroll
scroll-bar-width: Scroll Bars
scroll-bar-width, a frame parameter: Layout Parameters
scroll-command property: Textual Scrolling
scroll-conservatively: Textual Scrolling
scroll-down: Textual Scrolling
scroll-down-aggressively: Textual Scrolling
scroll-down-command: Textual Scrolling
scroll-error-top-bottom: Textual Scrolling
scroll-left: Horizontal Scrolling
scroll-margin: Textual Scrolling
scroll-other-window: Textual Scrolling
scroll-other-window-down: Textual Scrolling
scroll-preserve-screen-position: Textual Scrolling
scroll-right: Horizontal Scrolling
scroll-step: Textual Scrolling
scroll-up: Textual Scrolling
scroll-up-aggressively: Textual Scrolling
scroll-up-command: Textual Scrolling
scrolling textually: Textual Scrolling
search-backward: String Search
search-failed: String Search
search-forward: String Search
search-map: Prefix Keys
search-spaces-regexp: Regexp Search
searching: Searching and Matching
searching active keymaps for keys: Searching Keymaps
searching and case: Searching and Case
searching and replacing: Search and Replace
searching for overlays: Finding Overlays
searching for regexp: Regexp Search
searching text properties: Property Search
secondary selection: Window System Selections
secondary storage: Files and Storage
secure-hash: Checksum/Hash
secure-hash-algorithms: Checksum/Hash
security: Security Considerations
security vulnerabilities in text: Suspicious Text
security, and display specifications: Display Property
seed, for random number generation: Random Numbers
select safe coding system: User-Chosen Coding Systems
select window hooks: Selecting Windows
select-frame: Input Focus
select-frame-set-input-focus: Input Focus
select-safe-coding-system: User-Chosen Coding Systems
select-safe-coding-system-accept-default-p: User-Chosen Coding Systems
select-safe-coding-system-function: User-Chosen Coding Systems
select-window: Selecting Windows
selected frame: Input Focus
selected window: Selecting Windows
selected-frame: Input Focus
selected-window: Selecting Windows
selected-window-group: Selecting Windows
selected-window-group-function: Selecting Windows
selecting a buffer: Current Buffer
selecting a font: Font Selection
selecting a window: Selecting Windows
selection (for window systems): Window System Selections
selection-coding-system: Window System Selections
selective-display: Selective Display
selective-display-ellipses: Selective Display
selectively disabling font-lock fontifications: Customizing Keywords
self-evaluating form: Self-Evaluating Forms
self-insert-and-exit: Minibuffer Commands
self-insert-command: Commands for Insertion
self-insert-command override: Changing Key Bindings
self-insert-command, minor modes: Keymaps and Minor Modes
self-insert-uses-region-functions: Commands for Insertion
self-insertion: Commands for Insertion
SELinux context: Extended Attributes
send-string-to-terminal: Terminal Output
sending signals: Signals to Processes
sentence-end: Standard Regexps
sentence-end: Standard Regexps
sentence-end-double-space: Filling
sentence-end-without-period: Filling
sentence-end-without-space: Filling
sentinel (of process): Sentinels
seq in rx: Rx Constructs
seq library: Sequence Functions
seq-concatenate: Sequence Functions
seq-contains-p: Sequence Functions
seq-count: Sequence Functions
seq-difference: Sequence Functions
seq-do: Sequence Functions
seq-doseq: Sequence Functions
seq-drop: Sequence Functions
seq-drop-while: Sequence Functions
seq-elt: Sequence Functions
seq-empty-p: Sequence Functions
seq-every-p: Sequence Functions
seq-filter: Sequence Functions
seq-find: Sequence Functions
seq-group-by: Sequence Functions
seq-intersection: Sequence Functions
seq-into: Sequence Functions
seq-keep: Sequence Functions
seq-length: Sequence Functions
seq-let: Sequence Functions
seq-map: Sequence Functions
seq-map-indexed: Sequence Functions
seq-mapcat: Sequence Functions
seq-mapn: Sequence Functions
seq-max: Sequence Functions
seq-min: Sequence Functions
seq-partition: Sequence Functions
seq-position: Sequence Functions
seq-positions: Sequence Functions
seq-random-elt: Sequence Functions
seq-reduce: Sequence Functions
seq-remove: Sequence Functions
seq-remove-at-position: Sequence Functions
seq-set-equal-p: Sequence Functions
seq-setq: Sequence Functions
seq-some: Sequence Functions
seq-sort: Sequence Functions
seq-sort-by: Sequence Functions
seq-split: Sequence Functions
seq-subseq: Sequence Functions
seq-take: Sequence Functions
seq-take-while: Sequence Functions
seq-union: Sequence Functions
seq-uniq: Sequence Functions
seqp: Sequence Functions
sequence: Sequences Arrays Vectors
sequence destructuring: Sequence Functions
sequence destructuring: Sequence Functions
sequence functions in seq: Sequence Functions
sequence in rx: Rx Constructs
sequence iteration: Sequence Functions
sequence length: Sequence Functions
sequence maximum: Sequence Functions
sequence minimum: Sequence Functions
sequence reverse: Sequence Functions
sequencep: Sequence Functions
sequences, generalized: Sequence Functions
sequences, intersection of: Sequence Functions
sequences, union of: Sequence Functions
sequencing: Sequencing
sequencing pattern: pcase Macro
sequential execution: Sequencing
serial connections: Serial Ports
serial-process-configure: Serial Ports
serial-term: Serial Ports
serializing: Byte Packing
server-after-make-frame-hook: Creating Frames
session file: Session Management
session manager: Session Management
set: Setting Variables
set, defcustom keyword: Variable Definitions
set-advertised-calling-convention: Obsolete Functions
set-after, defcustom keyword: Variable Definitions
set-auto-coding: Default Coding Systems
set-auto-mode: Auto Major Mode
set-binary-mode: Input Functions
set-buffer: Current Buffer
set-buffer-auto-saved: Auto-Saving
set-buffer-major-mode: Auto Major Mode
set-buffer-modified-p: Buffer Modification
set-buffer-multibyte: Selecting a Representation
set-case-syntax: Case Tables
set-case-syntax-delims: Case Tables
set-case-syntax-pair: Case Tables
set-case-table: Case Tables
set-category-table: Categories
set-char-table-extra-slot: Char-Tables
set-char-table-parent: Char-Tables
set-char-table-range: Char-Tables
set-charset-priority: Character Sets
set-coding-system-priority: Specifying Coding Systems
set-default: Default Value
set-default-file-modes: Changing Files
set-default-toplevel-value: Default Value
set-display-table-slot: Display Tables
set-face-attribute: Attribute Functions
set-face-background: Attribute Functions
set-face-bold: Attribute Functions
set-face-extend: Attribute Functions
set-face-font: Attribute Functions
set-face-foreground: Attribute Functions
set-face-inverse-video: Attribute Functions
set-face-italic: Attribute Functions
set-face-stipple: Attribute Functions
set-face-underline: Attribute Functions
set-file-acl: Changing Files
set-file-extended-attributes: Changing Files
set-file-modes: Changing Files
set-file-selinux-context: Changing Files
set-file-times: Changing Files
set-fontset-font: Fontsets
set-frame-configuration: Frame Configurations
set-frame-font: Frame Font
set-frame-height: Frame Size
set-frame-parameter: Parameter Access
set-frame-position: Frame Position
set-frame-selected-window: Selecting Windows
set-frame-size: Frame Size
set-frame-width: Frame Size
set-frame-window-state-change: Window Hooks
set-fringe-bitmap-face: Customizing Bitmaps
set-input-method: Input Methods
set-input-mode: Input Modes
set-keyboard-coding-system: Terminal I/O Encoding
set-keymap-parent: Inheritance and Keymaps
set-left-margin: Margins
set-mark: The Mark
set-marker: Moving Markers
set-marker-insertion-type: Marker Insertion Types
set-match-data: Entire Match Data
set-message-function: Displaying Messages
set-message-functions: Displaying Messages
set-minibuffer-message: Displaying Messages
set-minibuffer-window: Minibuffer Windows
set-mouse-absolute-pixel-position: Mouse Position
set-mouse-pixel-position: Mouse Position
set-mouse-position: Mouse Position
set-multi-message: Displaying Messages
set-network-process-option: Network Options
set-process-buffer: Process Buffers
set-process-coding-system: Process Information
set-process-datagram-address: Datagrams
set-process-filter: Filter Functions
set-process-plist: Process Information
set-process-query-on-exit-flag: Query Before Exit
set-process-sentinel: Sentinels
set-process-thread: Processes and Threads
set-process-window-size: Process Buffers
set-register: Registers
set-right-margin: Margins
set-standard-case-table: Case Tables
set-syntax-table: Syntax Table Functions
set-terminal-coding-system: Terminal I/O Encoding
set-terminal-parameter: Terminal Parameters
set-text-properties: Changing Properties
set-transient-map: Controlling Active Maps
set-transient-map-timeout: Controlling Active Maps
set-visited-file-modtime: Modification Time
set-visited-file-name: Buffer File Name
set-window-buffer: Buffers and Windows
set-window-combination-limit: Recombining Windows
set-window-configuration: Window Configurations
set-window-dedicated-p: Dedicated Windows
set-window-display-table: Active Display Table
set-window-fringes: Fringe Size/Pos
set-window-group-start: Window Start and End
set-window-group-start-function: Window Start and End
set-window-hscroll: Horizontal Scrolling
set-window-margins: Display Margins
set-window-next-buffers: Window History
set-window-parameter: Window Parameters
set-window-point: Window Point
set-window-prev-buffers: Window History
set-window-scroll-bars: Scroll Bars
set-window-start: Window Start and End
set-window-vscroll: Vertical Scrolling
set-xwidget-buffer: Xwidgets
set-xwidget-plist: Xwidgets
set-xwidget-query-on-exit-flag: Xwidgets
setcar: Setcar
setcdr: Setcdr
setenv: System Environment
setf: Setting Generalized Variables
setopt: Setting Variables
setplist: Symbol Plists
setq: Setting Variables
setq-connection-local: Applying Connection Local Variables
setq-default: Default Value
setq-local: Creating Buffer-Local
sets: Sets And Lists
setting modes of files: Changing Files
setting-constant error: Constant Variables
set_function_finalizer: Module Functions
set_user_finalizer: Module Values
set_user_ptr: Module Values
severity level: Warning Basics
sexp: Intro Eval
sexp motion: List Motion
SHA hash: Checksum/Hash
SHA hash: GnuTLS Cryptography
sha1: Checksum/Hash
shaded, a frame parameter: Management Parameters
shadowed Lisp files: Library Search
shadowing of variables: Local Variables
shared structure, read syntax: Circular Objects
shell command arguments: Shell Arguments
shell-command-history: Minibuffer History
shell-command-to-string: Synchronous Processes
shell-quote-argument: Shell Arguments
shift-selection, and interactive spec: Using Interactive
shift-translation: Key Sequence Input
shortdoc-add-function: Documentation Groups
shorthands: Shorthands
should_quit: Module Misc
show image: Showing Images
show-help-function: Special Properties
shr-insert-document: Parsing HTML/XML
shrink-window-if-larger-than-buffer: Resizing Windows
shy groups: Regexp Backslash
sibling window: Windows and Frames
side effect: Intro Eval
side windows: Side Windows
side, a buffer display action alist entry: Buffer Display Action Alists
side-effect-free property: Standard Properties
SIGHUP: Killing Emacs
SIGINT: Killing Emacs
signal: Signaling Errors
signal-process: Signals to Processes
signal-process-functions: Signals to Processes
signaling errors: Signaling Errors
signals: Signals to Processes
SIGTERM: Killing Emacs
SIGTSTP: Suspending Emacs
sigusr1 event: Misc Events
sigusr2 event: Misc Events
simple package: Simple Packages
sin: Math Functions
single file package: Simple Packages
single-function hook: Hooks
single-key-description: Describing Characters
sit-for: Waiting
site-init.el: Building Emacs
site-lisp directories: Library Search
site-load.el: Building Emacs
site-run-file: Init File
site-start.el: Startup Summary
size of frame: Frame Geometry
size of image: Showing Images
size of text on display: Size of Displayed Text
size of window: Window Sizes
skip-chars-backward: Skipping Characters
skip-chars-forward: Skipping Characters
skip-syntax-backward: Motion and Syntax
skip-syntax-forward: Motion and Syntax
skip-taskbar, a frame parameter: Management Parameters
skipping characters: Skipping Characters
skipping characters of certain syntax: Motion and Syntax
skipping comments: Control Parsing
sleep-for: Waiting
slice, image: Showing Images
slot, a buffer display action alist entry: Buffer Display Action Alists
small-temporary-file-directory: Unique File Names
smallest fixnum: Integer Basics
SMIE: SMIE
SMIE grammar: SMIE Grammar
SMIE lexer: SMIE Lexer
smie-bnf->prec2: Operator Precedence Grammars
smie-close-block: SMIE setup
smie-config: SMIE Customization
smie-config-guess: SMIE Customization
smie-config-local: SMIE Customization
smie-config-save: SMIE Customization
smie-config-set-indent: SMIE Customization
smie-config-show-indent: SMIE Customization
smie-down-list: SMIE setup
smie-merge-prec2s: Operator Precedence Grammars
smie-prec2->grammar: Operator Precedence Grammars
smie-precs->prec2: Operator Precedence Grammars
smie-rule-bolp: SMIE Indentation Helpers
smie-rule-hanging-p: SMIE Indentation Helpers
smie-rule-next-p: SMIE Indentation Helpers
smie-rule-parent: SMIE Indentation Helpers
smie-rule-parent-p: SMIE Indentation Helpers
smie-rule-prev-p: SMIE Indentation Helpers
smie-rule-separator: SMIE Indentation Helpers
smie-rule-sibling-p: SMIE Indentation Helpers
smie-setup: SMIE setup
smooth-curveto: SVG Images
smooth-quadratic-bezier-curveto: SVG Images
snap-width, a frame parameter: Mouse Dragging Parameters
Snarf-documentation: Accessing Documentation
sort: Sequence Functions
sort-columns: Sorting
sort-fields: Sorting
sort-fold-case: Sorting
sort-lines: Sorting
sort-numeric-base: Sorting
sort-numeric-fields: Sorting
sort-pages: Sorting
sort-paragraphs: Sorting
sort-regexp-fields: Sorting
sort-subr: Sorting
sorting lists: Sequence Functions
sorting sequences: Sequence Functions
sorting text: Sorting
sorting vectors: Sequence Functions
sound: Sound Output
source breakpoints: Source Breakpoints
space (ASCII character): Basic Char Syntax
space display spec, and bidirectional text: Bidirectional Display
spaces, pixel specification: Pixel Specification
spaces, specified height or width: Specified Space
sparse keymap: Format of Keymaps
SPC in minibuffer: Text from Minibuffer
special events: Special Events
special form descriptions: A Sample Function Description
special forms: Special Forms
special forms for control structures: Control Structures
special modes: Major Mode Conventions
special read syntax: Special Read Syntax
special variables: Using Lexical Binding
special-event-map: Controlling Active Maps
special-form-p: Special Forms
special-mode: Basic Major Modes
special-mode-map: Standard Keymaps
special-variable-p: Using Lexical Binding
Specified time is not representable: Time of Day
specify coding system: Specifying Coding Systems
specify color: Color Names
speedups: Compilation Tips
splicing (with backquote): Backquote
split-height-threshold: Choosing Window Options
split-root-window-below: Splitting Windows
split-root-window-right: Splitting Windows
split-string: Creating Strings
split-string-and-unquote: Shell Arguments
split-string-default-separators: Creating Strings
split-string-shell-command: Shell Arguments
split-width-threshold: Choosing Window Options
split-window: Splitting Windows
split-window, a window parameter: Window Parameters
split-window-below: Splitting Windows
split-window-keep-point: Splitting Windows
split-window-preferred-function: Choosing Window Options
split-window-right: Splitting Windows
split-window-sensibly: Choosing Window Options
splitting windows: Splitting Windows
sqlite-available-p: Database
sqlite-close: Database
sqlite-columns: Database
sqlite-commit: Database
sqlite-execute: Database
sqlite-finalize: Database
sqlite-load-extension: Database
sqlite-mode-open-file: Database
sqlite-more-p: Database
sqlite-next: Database
sqlite-open: Database
sqlite-pragma: Database
sqlite-rollback: Database
sqlite-select: Database
sqlite-transaction: Database
sqlite-version: Database
sqlitep: Database
sqrt: Math Functions
stable sort: Sequence Functions
stack allocated Lisp objects: Stack-allocated Objects
stack frame: Backtraces
stack overflow in regexp: Regexp Problems
standalone-parent: Parser-based Indentation
standard abbrev tables: Standard Abbrev Tables
standard colors for character terminals: Font and Color Parameters
standard error process: Asynchronous Processes
standard errors: Standard Errors
standard hooks: Standard Hooks
standard regexps used in editing: Standard Regexps
standard syntax table: Syntax Basics
standard-case-table: Case Tables
standard-category-table: Categories
standard-display-table: Active Display Table
standard-input: Input Functions
standard-output: Output Variables
standard-syntax-table: Syntax Basics
standard-translation-table-for-decode: Translation of Characters
standard-translation-table-for-encode: Translation of Characters
standards of coding style: Tips
start-file-process: Asynchronous Processes
start-file-process-shell-command: Asynchronous Processes
start-process: Asynchronous Processes
start-process, command-line arguments from minibuffer: Shell Arguments
start-process-shell-command: Asynchronous Processes
STARTTLS network connections: Network
startup of Emacs: Startup Summary
startup screen: Startup Summary
startup-redirect-eln-cache: Native-Compilation Functions
startup.el: Startup Summary
statement object: Database
staticpro, protection from GC: Writing Emacs Primitives
stderr stream, use for debugging: Output Streams
sticky text properties: Sticky Properties
sticky, a frame parameter: Management Parameters
stop points: Using Edebug
stop-process: Signals to Processes
stopbits, in serial connections: Serial Ports
stopping an infinite loop: Infinite Loops
stopping on events: Global Break Condition
storage of vector-like Lisp objects: Garbage Collection
store-match-data: Entire Match Data
store-substring: Modifying Strings
stream (for printing): Output Streams
stream (for reading): Input Streams
strike-through text: Face Attributes
string: Creating Strings
string creation: Creating Strings
string equality: Text Comparison
string in keymap: Key Lookup
string input stream: Input Streams
string length: Sequence Functions
string modification: Modifying Strings
string predicates: Predicates for Strings
string reverse: Sequence Functions
string search: String Search
string to number: String Conversion
string to object: Input Functions
string to vector: Sequence Functions
string, number of bytes: Text Representations
string, writing a doc string: Documentation Basics
string-as-multibyte: Selecting a Representation
string-as-unibyte: Selecting a Representation
string-bytes: Text Representations
string-chars-consed: Memory Usage
string-chop-newline: Creating Strings
string-clean-whitespace: Creating Strings
string-collate-equalp: Text Comparison
string-collate-lessp: Text Comparison
string-distance: Text Comparison
string-end in rx: Rx Constructs
string-equal: Text Comparison
string-equal-ignore-case: Text Comparison
string-fill: Creating Strings
string-glyph-split: Size of Displayed Text
string-greaterp: Text Comparison
string-lessp: Text Comparison
string-limit: Creating Strings
string-lines: Creating Strings
string-match: Regexp Search
string-match-p: Regexp Search
string-or-null-p: Predicates for Strings
string-pad: Creating Strings
string-pixel-width: Size of Displayed Text
string-prefix-p: Text Comparison
string-replace: Search and Replace
string-search: Text Comparison
string-start in rx: Rx Constructs
string-suffix-p: Text Comparison
string-to-char: String Conversion
string-to-int: String Conversion
string-to-multibyte: Converting Representations
string-to-number: String Conversion
string-to-syntax: Syntax Table Internals
string-to-unibyte: Converting Representations
string-trim: Creating Strings
string-trim-left: Creating Strings
string-trim-right: Creating Strings
string-version-lessp: Text Comparison
string-width: Size of Displayed Text
string<: Text Comparison
string=: Text Comparison
stringp: Predicates for Strings
strings: Strings and Characters
strings with keyboard events: Strings of Events
strings, formatting them: Formatting Strings
strings-consed: Memory Usage
structural matching: Backquote Patterns
sub-sequence: Sequence Functions
submatch in rx: Rx Constructs
submatch-n in rx: Rx Constructs
submenu: Mouse Menus
subprocess: Processes
subr: What Is a Function
subr-arity: What Is a Function
subrp: What Is a Function
subst-char-in-region: Substitution
subst-char-in-string: Substitution
substitute characters: Substitution
substitute-command-keys: Keys in Documentation
substitute-in-file-name: File Name Expansion
substitute-key-definition: Changing Key Bindings
substitute-key-definition: Low-Level Key Binding
substitute-quotes: Keys in Documentation
substituting keys in documentation: Keys in Documentation
substring: Creating Strings
substring-no-properties: Creating Strings
subtype of char-table: Char-Tables
success handler: Handling Errors
suggestions: Caveats
super characters: Other Char Bits
support for touchscreens: Touchscreen Events
suppress-keymap: Changing Key Bindings
surrogate minibuffer frame: Minibuffers and Frames
suspend (cf. no-redraw-on-reenter): Refresh Screen
suspend evaluation: Recursive Editing
suspend major mode temporarily: Major Modes
suspend-emacs: Suspending Emacs
suspend-frame: Suspending Emacs
suspend-hook: Suspending Emacs
suspend-resume-hook: Suspending Emacs
suspend-tty: Suspending Emacs
suspend-tty-functions: Suspending Emacs
suspending Emacs: Suspending Emacs
suspicious text: Suspicious Text
suspicious text strings: Suspicious Text
SVG images: SVG Images
SVG object: SVG Images
svg path commands: SVG Images
svg-circle: SVG Images
svg-clip-path: SVG Images
svg-create: SVG Images
svg-ellipse: SVG Images
svg-embed: SVG Images
svg-embed-base-uri-image: SVG Images
svg-gradient: SVG Images
svg-image: SVG Images
svg-line: SVG Images
svg-node: SVG Images
svg-path: SVG Images
svg-polygon: SVG Images
svg-polyline: SVG Images
svg-rectangle: SVG Images
svg-remove: SVG Images
svg-text: SVG Images
swap text between buffers: Swapping Text
switch-to-buffer: Switching Buffers
switch-to-buffer-in-dedicated-window: Switching Buffers
switch-to-buffer-obey-display-actions: Switching Buffers
switch-to-buffer-other-frame: Switching Buffers
switch-to-buffer-other-window: Switching Buffers
switch-to-buffer-preserve-window-point: Switching Buffers
switch-to-next-buffer: Window History
switch-to-prev-buffer: Window History
switch-to-prev-buffer-skip: Window History
switch-to-prev-buffer-skip-regexp: Window History
switches on command line: Command-Line Arguments
switching to a buffer: Switching Buffers
sxhash-eq: Defining Hash
sxhash-eql: Defining Hash
sxhash-equal: Defining Hash
symbol: Symbols
symbol components: Symbol Components
symbol equality: Creating Symbols
symbol evaluation: Symbol Forms
symbol forms: Symbol Forms
symbol function indirection: Function Indirection
symbol in keymap: Key Lookup
symbol name hashing: Creating Symbols
symbol property: Symbol Properties
symbol that evaluates to itself: Constant Variables
symbol with constant value: Constant Variables
symbol with position: Symbols with Position
symbol, where defined: Where Defined
symbol-end in rx: Rx Constructs
symbol-file: Where Defined
symbol-function: Function Cells
symbol-name: Creating Symbols
symbol-plist: Symbol Plists
symbol-start in rx: Rx Constructs
symbol-value: Accessing Variables
symbol-with-pos-p: Symbols with Position
symbol-with-pos-pos: Symbols with Position
symbolic links: Kinds of Files
symbolic links: Kinds of Files
symbolic shorthands: Shorthands
symbolp: Symbols
symbols-consed: Memory Usage
symbols-with-pos-enabled: Symbols with Position
symmetric cipher: GnuTLS Cryptography
synchronized multisession variables: Multisession Variables
synchronous subprocess: Synchronous Processes
syntactic font lock: Syntactic Font Lock
syntax class: Syntax Descriptors
syntax class table: Syntax Class Table
syntax code: Syntax Table Internals
syntax descriptor: Syntax Descriptors
syntax entry, setting: Syntax Table Functions
syntax error (Edebug): Backtracking
syntax flags: Syntax Flags
syntax for characters: Basic Char Syntax
syntax highlighting and coloring: Font Lock Mode
syntax in rx: Rx Constructs
syntax of regular expressions: Syntax of Regexps
syntax table: Syntax Tables
syntax table example: Example Major Modes
syntax table internals: Syntax Table Internals
syntax tables (accessing elements of): Syntax Table Internals
syntax tables in modes: Major Mode Conventions
syntax tree nodes, by field name: Retrieving Nodes
syntax tree nodes, by kinship: Retrieving Nodes
syntax tree nodes, by position: Retrieving Nodes
syntax tree nodes, retrieving from other nodes: Retrieving Nodes
syntax tree, concrete: Language Grammar
syntax tree, from parsing program source: Parsing Program Source
syntax tree, retrieving nodes: Retrieving Nodes
syntax trees, node information: Accessing Node Information
syntax-after: Syntax Table Internals
syntax-class: Syntax Table Internals
syntax-class-to-char: Syntax Table Internals
syntax-ppss: Position Parse
syntax-ppss-context: Parser State
syntax-ppss-flush-cache: Position Parse
syntax-ppss-toplevel-pos: Parser State
syntax-propertize-extend-region-functions: Syntax Properties
syntax-propertize-function: Syntax Properties
syntax-table: Syntax Table Functions
syntax-table (text property): Syntax Properties
syntax-table-p: Syntax Basics
system abbrev: Abbrevs
system processes: System Processes
system tooltips: Tooltips
system type and name: System Environment
system-configuration: System Environment
system-groups: User Identification
system-key-alist: X11 Keysyms
system-messages-locale: Locales
system-name: System Environment
system-time-locale: Locales
system-type: System Environment
system-users: User Identification

T
t: nil and t
t input stream: Input Streams
t output stream: Output Streams
tab (ASCII character): Basic Char Syntax
tab bar: Frame Layout
tab deletion: Deletion
TAB in minibuffer: Text from Minibuffer
tab-always-indent: Mode-Specific Indent
tab-bar mouse events: Click Events
tab-bar, prefix key: Key Sequence Input
tab-bar-lines, a frame parameter: Layout Parameters
tab-bar-map: Standard Keymaps
tab-first-completion: Mode-Specific Indent
tab-line, prefix key: Key Sequence Input
tab-line-format, a window parameter: Window Parameters
tab-prefix-map: Prefix Keys
tab-stop-list: Indent Tabs
tab-to-tab-stop: Indent Tabs
tab-width: Usual Display
tabs stops for indentation: Indent Tabs
Tabulated List mode: Tabulated List Mode
tabulated-list-clear-all-tags: Tabulated List Mode
tabulated-list-delete-entry: Tabulated List Mode
tabulated-list-entries: Tabulated List Mode
tabulated-list-format: Tabulated List Mode
tabulated-list-get-entry: Tabulated List Mode
tabulated-list-get-id: Tabulated List Mode
tabulated-list-gui-sort-indicator-asc: Tabulated List Mode
tabulated-list-gui-sort-indicator-desc: Tabulated List Mode
tabulated-list-header-overlay-p: Tabulated List Mode
tabulated-list-init-header: Tabulated List Mode
tabulated-list-mode: Tabulated List Mode
tabulated-list-padding: Tabulated List Mode
tabulated-list-print: Tabulated List Mode
tabulated-list-printer: Tabulated List Mode
tabulated-list-put-tag: Tabulated List Mode
tabulated-list-revert-hook: Tabulated List Mode
tabulated-list-set-col: Tabulated List Mode
tabulated-list-sort-key: Tabulated List Mode
tabulated-list-tty-sort-indicator-asc: Tabulated List Mode
tabulated-list-tty-sort-indicator-desc: Tabulated List Mode
tabulated-list-use-header-line: Tabulated List Mode
tag on run time stack: Catch and Throw
tag, customization keyword: Common Keywords
take: List Elements
tan: Math Functions
TCP: Network
temacs: Building Emacs
TEMP environment variable: Unique File Names
temp-buffer-max-height: Temporary Displays
temp-buffer-max-width: Temporary Displays
temp-buffer-resize-mode: Temporary Displays
temp-buffer-setup-hook: Temporary Displays
temp-buffer-show-function: Temporary Displays
temp-buffer-show-hook: Temporary Displays
temp-buffer-window-setup-hook: Temporary Displays
temp-buffer-window-show-hook: Temporary Displays
temporary buffer display: Temporary Displays
temporary display: Temporary Displays
temporary file on a remote host: Unique File Names
temporary files: Unique File Names
temporary-file-directory: Unique File Names
temporary-file-directory: Unique File Names
TERM environment variable: Terminal-Specific
term-file-aliases: Terminal-Specific
term-file-prefix: Terminal-Specific
Termcap: Terminal-Specific
terminal: Frames
terminal input: Terminal Input
terminal input modes: Input Modes
terminal output: Terminal Output
terminal parameters: Terminal Parameters
terminal screen: Basic Windows
terminal type: Terminal Type
terminal-coding-system: Terminal I/O Encoding
terminal-list: Multiple Terminals
terminal-live-p: Frames
terminal-local variables: Multiple Terminals
terminal-name: Multiple Terminals
terminal-parameter: Terminal Parameters
terminal-parameters: Terminal Parameters
terminal-specific initialization: Terminal-Specific
terminology, for tree-sitter functions: Retrieving Nodes
termscript file: Terminal Output
terpri: Output Functions
test-completion: Basic Completion
testcover-mark-all: Test Coverage
testcover-next-mark: Test Coverage
testcover-start: Test Coverage
testing types: Type Predicates
text: Text
text area: Frame Layout
text area of a window: Basic Windows
text comparison: Text Comparison
text conversion of coding system: Lisp and Coding Systems
text deletion: Deletion
text height of a frame: Frame Size
text insertion: Insertion
text near point: Near Point
text parsing: Syntax Tables
text properties: Text Properties
text properties in files: Format Conversion
text properties in the mode line: Properties in Mode
text properties, changing: Changing Properties
text properties, examining: Examining Properties
text properties, read syntax: Text Props and Strings
text properties, searching: Property Search
text representation: Text Representations
text size of a frame: Frame Size
text terminal: Frames
text width of a frame: Frame Size
text-char-description: Describing Characters
text-mode: Basic Major Modes
text-mode-abbrev-table: Standard Abbrev Tables
text-properties-at: Examining Properties
text-property-any: Property Search
text-property-default-nonsticky: Sticky Properties
text-property-not-all: Property Search
text-property-search-backward: Property Search
text-property-search-forward: Property Search
text-quoting-style: Text Quoting Style
text-quoting-style: Text Quoting Style
text-terminal focus notification: Input Focus
textsec-check: Suspicious Text
textsec-suspicious (face): Suspicious Text
textsec-suspicious-p: Suspicious Text
textual order: Control Structures
textual scrolling: Textual Scrolling
theme-face (face symbol property): Defining Faces
thing-at-point: Buffer Contents
thing-at-point-provider-alist: Buffer Contents
this-command: Command Loop Info
this-command-keys: Command Loop Info
this-command-keys-shift-translated: Key Sequence Input
this-command-keys-vector: Command Loop Info
this-original-command: Command Loop Info
thread backtrace: The Thread List
thread list: The Thread List
thread--blocker: Basic Thread Functions
thread-join: Basic Thread Functions
thread-last-error: Basic Thread Functions
thread-list-refresh-seconds: The Thread List
thread-live-p: Basic Thread Functions
thread-name: Basic Thread Functions
thread-signal: Basic Thread Functions
thread-yield: Basic Thread Functions
threadp: Basic Thread Functions
threads: Threads
three-step-help: Help Functions
throw: Catch and Throw
throw example: Recursive Editing
thunk: Deferred Eval
thunk-delay: Deferred Eval
thunk-force: Deferred Eval
thunk-let: Deferred Eval
thunk-let*: Deferred Eval
tiled windows: Basic Windows
time calculations: Time Calculations
time conversion: Time Conversion
time formatting: Time Parsing
time of day: Time of Day
time parsing: Time Parsing
time value: Time of Day
time zone rule: Time Zone Rules
time zone rules: Time Zone Rules
time zone, current: Time Zone Rules
time-add: Time Calculations
time-convert: Time Conversion
time-equal-p: Time Calculations
time-less-p: Time Calculations
time-subtract: Time Calculations
time-to-day-in-year: Time Calculations
time-to-days: Time Calculations
timer-max-repeats: Timers
timerp: Timers
timers: Timers
timestamp of a mouse event: Accessing Mouse
timestamp, Lisp: Time of Day
timing programs: Profiling
tips for documentation strings: Documentation Tips
tips for faster Lisp code: Compilation Tips
tips for writing Lisp: Tips
title bar: Frame Layout
title, a frame parameter: Basic Parameters
TLS network connections: Network
TMP environment variable: Unique File Names
TMPDIR environment variable: Unique File Names
toggle-enable-multibyte-characters: Disabling Multibyte
tool bar: Tool Bar
tool-bar mouse events: Click Events
tool-bar-add-item: Tool Bar
tool-bar-add-item-from-menu: Tool Bar
tool-bar-border: Tool Bar
tool-bar-button-margin: Tool Bar
tool-bar-button-relief: Tool Bar
tool-bar-lines, a frame parameter: Layout Parameters
tool-bar-local-item-from-menu: Tool Bar
tool-bar-map: Tool Bar
tool-bar-position, a frame parameter: Layout Parameters
tooltip face: Tooltips
tooltip for help strings: Special Properties
tooltip frames: Tooltips
tooltip window: Basic Windows
tooltip-event-buffer: Tooltips
tooltip-frame-parameters: Tooltips
tooltip-functions: Tooltips
tooltip-help-tips: Tooltips
tooltip-mode: Tooltips
tooltips: Tooltips
top and bottom window decorations: Basic Windows
top frame: Raising and Lowering
top position ratio: Position Parameters
top, a frame parameter: Position Parameters
top-level: Recursive Editing
top-level default value: Default Value
top-level form: Loading
top-level frame: Frames
top-visible, a frame parameter: Mouse Dragging Parameters
total height of a window: Window Sizes
total pixel height of a window: Window Sizes
total pixel width of a window: Window Sizes
total width of a window: Window Sizes
touch point, in touchscreen events: Touchscreen Events
touch-end event: Misc Events
touchscreen events: Touchscreen Events
touchscreen-begin event: Touchscreen Events
touchscreen-end event: Touchscreen Events
touchscreen-update event: Touchscreen Events
tq-close: Transaction Queues
tq-create: Transaction Queues
tq-enqueue: Transaction Queues
trace buffer: Trace Buffer
tracing Lisp programs: Debugging
track-mouse: Mouse Tracking
tracking frame size changes: Frame Size
trailing blanks in file names: Information about Files
trampolines, in native compilation: Native-Compilation Variables
transaction queue: Transaction Queues
transcendental functions: Math Functions
transient keymap: Controlling Active Maps
transient-mark-mode: The Mark
translate-region: Substitution
translate-upper-case-key-bindings: Key Sequence Input
translating input events: Event Mod
translation keymap: Translation Keymaps
translation tables: Translation of Characters
translation-table-for-input: Translation of Characters
transparency, frame: Font and Color Parameters
transparency, frame: Font and Color Parameters
transpose-regions: Transposition
trash: Changing Files
trash: Create/Delete Dirs
tray notifications, MS-Windows: Desktop Notifications
tree-sitter extra node: Accessing Node Information
tree-sitter fontifications, overview: Parser-based Font Lock
tree-sitter host and embedded languages: Multiple Languages
tree-sitter missing node: Accessing Node Information
tree-sitter narrowing: Using Parser
tree-sitter node field name: Language Grammar
tree-sitter node that has error: Accessing Node Information
tree-sitter nodes, comparing: Accessing Node Information
tree-sitter outdated node: Accessing Node Information
tree-sitter parse string: Using Parser
tree-sitter parse tree, leaf node: Retrieving Nodes
tree-sitter parse-tree, update and after-change callback: Using Parser
tree-sitter parser, creating: Using Parser
tree-sitter parser, using: Using Parser
tree-sitter patterns as strings: Pattern Matching
tree-sitter query pattern syntax: Pattern Matching
tree-sitter, find node: Retrieving Nodes
tree-sitter, live parsing node: Accessing Node Information
treesit-available-p: Parsing Program Source
treesit-beginning-of-defun: List Motion
treesit-buffer-root-node: Retrieving Nodes
treesit-buffer-too-large: Using Parser
treesit-check-indent: Parser-based Indentation
treesit-defun-at-point: Tree-sitter Major Modes
treesit-defun-name: Tree-sitter Major Modes
treesit-defun-name-function: Tree-sitter Major Modes
treesit-defun-tactic: List Motion
treesit-defun-type-regexp: List Motion
treesit-end-of-defun: List Motion
treesit-explore-mode: Language Grammar
treesit-extra-load-path: Language Grammar
treesit-filter-child: Retrieving Nodes
treesit-font-lock-feature-list: Parser-based Font Lock
treesit-font-lock-recompute-features: Parser-based Font Lock
treesit-font-lock-rules: Parser-based Font Lock
treesit-font-lock-settings: Parser-based Font Lock
treesit-fontify-with-override: Parser-based Font Lock
treesit-indent-function: Parser-based Indentation
treesit-induce-sparse-tree: Retrieving Nodes
treesit-inspect-mode: Language Grammar
treesit-language-abi-version: Language Grammar
treesit-language-at: Multiple Languages
treesit-language-at-point-function: Multiple Languages
treesit-language-available-p: Language Grammar
treesit-library-abi-version: Language Grammar
treesit-load-language-error: Language Grammar
treesit-load-name-override-list: Language Grammar
treesit-major-mode-setup: Tree-sitter Major Modes
treesit-max-buffer-size: Using Parser
treesit-node-at: Retrieving Nodes
treesit-node-buffer: Accessing Node Information
treesit-node-check: Accessing Node Information
treesit-node-child: Retrieving Nodes
treesit-node-child-by-field-name: Retrieving Nodes
treesit-node-child-count: Accessing Node Information
treesit-node-children: Retrieving Nodes
treesit-node-descendant-for-range: Retrieving Nodes
treesit-node-end: Accessing Node Information
treesit-node-eq: Accessing Node Information
treesit-node-field-name: Accessing Node Information
treesit-node-field-name-for-child: Accessing Node Information
treesit-node-first-child-for-pos: Retrieving Nodes
treesit-node-index: Accessing Node Information
treesit-node-language: Accessing Node Information
treesit-node-next-sibling: Retrieving Nodes
treesit-node-on: Retrieving Nodes
treesit-node-outdated: Retrieving Nodes
treesit-node-p: Accessing Node Information
treesit-node-parent: Retrieving Nodes
treesit-node-parser: Accessing Node Information
treesit-node-prev-sibling: Retrieving Nodes
treesit-node-start: Accessing Node Information
treesit-node-text: Accessing Node Information
treesit-node-top-level: Retrieving Nodes
treesit-node-type: Accessing Node Information
treesit-parent-until: Retrieving Nodes
treesit-parent-while: Retrieving Nodes
treesit-parse-string: Using Parser
treesit-parser-add-notifier: Using Parser
treesit-parser-buffer: Using Parser
treesit-parser-create: Using Parser
treesit-parser-delete: Using Parser
treesit-parser-included-ranges: Multiple Languages
treesit-parser-language: Using Parser
treesit-parser-list: Using Parser
treesit-parser-notifiers: Using Parser
treesit-parser-p: Using Parser
treesit-parser-remove-notifier: Using Parser
treesit-parser-root-node: Retrieving Nodes
treesit-parser-set-included-ranges: Multiple Languages
treesit-pattern-expand: Pattern Matching
treesit-query-capture: Pattern Matching
treesit-query-compile: Pattern Matching
treesit-query-error: Pattern Matching
treesit-query-expand: Pattern Matching
treesit-query-language: Pattern Matching
treesit-query-range: Multiple Languages
treesit-query-string: Pattern Matching
treesit-query-validate: Pattern Matching
treesit-range-invalid: Multiple Languages
treesit-range-rules: Multiple Languages
treesit-range-settings: Multiple Languages
treesit-ready-p: Tree-sitter Major Modes
treesit-search-forward: Retrieving Nodes
treesit-search-forward-goto: Retrieving Nodes
treesit-search-subtree: Retrieving Nodes
treesit-simple-imenu-settings: Imenu
treesit-simple-indent-presets: Parser-based Indentation
treesit-simple-indent-rules: Parser-based Indentation
treesit-update-ranges: Multiple Languages
triple-click events: Repeat Events
true: nil and t
true list: Cons Cells
truename (of file): Truenames
truncate: Numeric Conversions
truncate-lines: Truncation
truncate-partial-width-windows: Truncation
truncate-string-ellipsis: Size of Displayed Text
truncate-string-to-width: Size of Displayed Text
truth value: nil and t
try-completion: Basic Completion
tty-color-alist: Text Terminal Colors
tty-color-approximate: Text Terminal Colors
tty-color-clear: Text Terminal Colors
tty-color-define: Text Terminal Colors
tty-color-mode, a frame parameter: Font and Color Parameters
tty-color-translate: Text Terminal Colors
tty-erase-char: System Environment
tty-menu-calls-mouse-position-function: Mouse Position
tty-setup-hook: Terminal-Specific
tty-top-frame: Raising and Lowering
turn multibyte support on or off: Disabling Multibyte
two’s complement: Integer Basics
type: Lisp Data Types
type (button property): Button Properties
type checking: Type Predicates
type checking internals: Writing Emacs Primitives
type of node, tree-sitter: Language Grammar
type predicates: Type Predicates
type, defcustom keyword: Customization Types
type-error, customization keyword: Type Keywords
type-of: Type Predicates
type_of: Module Misc
typographic conventions: Some Terms
TZ, environment variable: Time Zone Rules

U
UBA: Bidirectional Display
UDP: Network
UID: User Identification
umask: Changing Files
unassigned character codepoints: Character Properties
unbalanced parentheses: Syntax Errors
unbinding keys: Key Binding Commands
unbury-buffer: Buffer List
undecided coding system: Coding System Basics
undecided coding-system, when decoding: Explicit Encoding
undecided coding-system, when encoding: Explicit Encoding
undecorated, a frame parameter: Management Parameters
undefined: Functions for Key Lookup
undefined in keymap: Key Lookup
undefined key: Keymap Basics
underline-minimum-offset: Face Attributes
underlined text: Face Attributes
undo avoidance: Substitution
undo in temporary buffers: Current Buffer
undo-amalgamate-change-group: Atomic Changes
undo-ask-before-discard: Maintaining Undo
undo-auto-amalgamate: Undo
undo-auto-current-boundary-timer: Undo
undo-boundary: Undo
undo-in-progress: Undo
undo-limit: Maintaining Undo
undo-outer-limit: Maintaining Undo
undo-strong-limit: Maintaining Undo
unexec: Building Emacs
unexec: Building Emacs
unhandled-file-name-directory: Magic File Names
unibyte buffers, and bidi reordering: Bidirectional Display
unibyte text: Text Representations
unibyte-char-to-multibyte: Converting Representations
unibyte-string: Text Representations
Unicode: Text Representations
unicode bidirectional algorithm: Bidirectional Display
unicode character escape: General Escape Syntax
unicode general category: Character Properties
unicode, a charset: Character Sets
unicode-category-table: Character Properties
unintern: Creating Symbols
uninterned symbol: Creating Symbols
uninterned symbol, and generating Lisp code: Creating Symbols
union of sequences: Sequence Functions
unique file names: Unique File Names
Universal Time: Time of Day
universal-argument: Prefix Command Arguments
universal-argument-map: Standard Keymaps
unless: Conditionals
unload-feature: Unloading
unload-feature-special-hooks: Unloading
unloading packages: Unloading
unloading packages, preparing for: Coding Conventions
unlock-buffer: File Locks
unmapped frame: Visibility of Frames
unmatchable in rx: Rx Constructs
unnumbered group: Regexp Backslash
unpacking: Byte Packing
unread-command-events: Event Input Misc
unsafep: Function Safety
unspecified, face attribute value: Face Attributes
unsplittable, a frame parameter: Buffer Parameters
unused lexical variable: Converting to Lexical Binding
unwind-protect: Cleanups
unwinding: Cleanups
up-list: List Motion
upcase: Case Conversion
upcase-initials: Case Conversion
upcase-region: Case Changes
upcase-word: Case Changes
update callback, for tree-sitter parse-tree: Using Parser
upper case: Case Conversion
upper case key sequence: Key Sequence Input
uptime of Emacs: Processor Run Time
use time of window: Selecting Windows
use-empty-active-region: The Region
use-frame-synchronization, a frame parameter: Management Parameters
use-global-map: Controlling Active Maps
use-hard-newlines: Filling
use-local-map: Controlling Active Maps
use-package: Named Features
use-region-p: The Region
use-system-tooltips: Tooltips
user errors, signaling: Signaling Errors
user groups: User Identification
user identification: User Identification
user options, how to define: Variable Definitions
user pointer object: Dynamic Modules
user pointer, using in module functions: Module Values
user signals: Misc Events
user-defined error: Error Symbols
user-emacs-directory: Init File
user-error: Signaling Errors
user-full-name: User Identification
user-full-name: User Identification
user-init-file: Init File
user-login-name: User Identification
user-login-name: User Identification
user-mail-address: User Identification
user-position, a frame parameter: Position Parameters
user-ptr object: Dynamic Modules
user-ptrp: Dynamic Modules
user-real-login-name: User Identification
user-real-login-name: User Identification
user-real-uid: User Identification
user-size, a frame parameter: Size Parameters
user-uid: User Identification
UTC: Time of Day
utf-8-emacs coding system: Coding System Basics
utility functions for parser-based indentation: Parser-based Indentation

V
valid windows: Basic Windows
validity of coding system: Lisp and Coding Systems
value cell: Symbol Components
value of expression: Evaluation
value of function: What Is a Function
values: Eval
variable: Variables
variable aliases: Variable Aliases
variable definition: Defining Variables
variable descriptions: A Sample Variable Description
variable pitch tables: Tabulated List Mode
variable watchpoints: Watching Variables
variable with constant value: Constant Variables
variable write debugging: Variable Debugging
variable, buffer-local: Buffer-Local Variables
variable-documentation property: Documentation Basics
variable-width spaces: Specified Space
variant coding system: Coding System Basics
vc-mode: Mode Line Variables
vc-prefix-map: Prefix Keys
vc-responsible-backend: Truenames
vconcat: Vector Functions
vector: Vector Functions
vector (type): Vectors
vector evaluation: Self-Evaluating Forms
vector length: Sequence Functions
vector reverse: Sequence Functions
vector to list: Sequence Functions
vector-cells-consed: Memory Usage
vector-like objects, storage: Garbage Collection
vectorp: Vector Functions
vec_get: Module Values
vec_set: Module Values
vec_size: Module Values
verify-visited-file-modtime: Modification Time
version number (in file name): File Name Components
version, customization keyword: Common Keywords
version-control: Numbered Backups
vertical combination: Windows and Frames
vertical fractional scrolling: Vertical Scrolling
vertical scroll position: Vertical Scrolling
vertical tab: Basic Char Syntax
vertical-line, prefix key: Key Sequence Input
vertical-lineto: SVG Images
vertical-motion: Screen Lines
vertical-scroll-bar: Scroll Bars
vertical-scroll-bar, prefix key: Key Sequence Input
vertical-scroll-bars, a frame parameter: Layout Parameters
view part, model/view/controller: Abstract Display
view-register: Registers
virtual buffers: Swapping Text
visibility, a frame parameter: Management Parameters
visible frame: Visibility of Frames
visible-bell: Beeping
visible-frame-list: Finding All Frames
visited file: Buffer File Name
visited file mode: Auto Major Mode
visited-file-modtime: Modification Time
visiting files: Visiting Files
visiting files, functions for: Visiting Functions
visual lines, moving by: Screen Lines
visual order: Bidirectional Display
visual order, preserve when copying bidirectional text: Bidirectional Display
visual-order cursor motion: Bidirectional Display
void function: Function Indirection
void function cell: Function Cells
void variable: Void Variables
void-function: Function Cells
void-text-area-pointer: Pointer Shape
void-variable error: Void Variables

W
w32-collate-ignore-punctuation: Text Comparison
w32-notification-close: Desktop Notifications
w32-notification-notify: Desktop Notifications
wait-for-wm, a frame parameter: Management Parameters
waiting: Waiting
waiting for command key input: Event Input Misc
waiting-for-user-input-p: Sentinels
walk-windows: Cyclic Window Ordering
warn: Warning Basics
warning options: Warning Options
warning type: Warning Basics
warning variables: Warning Variables
warning-fill-column: Warning Variables
warning-fill-prefix: Warning Variables
warning-levels: Warning Variables
warning-minimum-level: Warning Options
warning-minimum-log-level: Warning Options
warning-prefix-function: Warning Variables
warning-series: Warning Variables
warning-suppress-log-types: Warning Options
warning-suppress-types: Warning Options
warning-type-format: Warning Variables
warnings: Warnings
warnings from byte compiler: Warning Tips
warnings, delayed: Delayed Warnings
warnings, suppressing during startup: Warning Options
watch, for filesystem events: File Notifications
watchpoints for Lisp variables: Watching Variables
webkit browser widget: Xwidgets
wheel-down event: Misc Events
wheel-up event: Misc Events
when: Conditionals
when-let: Conditionals
where was a symbol defined: Where Defined
where-is-internal: Scanning Keymaps
while: Iteration
while-let: Conditionals
while-no-input: Event Input Misc
while-no-input-ignore-events: Event Input Misc
whitespace: Basic Char Syntax
wholenump: Predicates on Numbers
widen: Narrowing
widening: Narrowing
width of a window: Window Sizes
width, a frame parameter: Size Parameters
window: Basic Windows
window (overlay property): Overlay Properties
window auto-selection: Mouse Window Auto-selection
window body: Basic Windows
window body height: Window Sizes
window body size: Window Sizes
window body width: Window Sizes
window buffer change: Window Hooks
window change functions: Window Hooks
window combination: Windows and Frames
window combination limit: Recombining Windows
window configuration (Edebug): Edebug Display Update
window configuration change: Window Hooks
window configurations: Window Configurations
window decorations: Basic Windows
window dividers: Window Dividers
window end position: Window Start and End
window excursions: Excursions
window header line: Header Lines
window height: Window Sizes
window history: Window History
window in direction: Windows and Frames
window internals: Window Internals
window layout in a frame: Window Configuration Type
window layout, all frames: Frame Configuration Type
window manager interaction, and frame parameters: Management Parameters
window order by time of last use: Selecting Windows
window ordering, cyclic: Cyclic Window Ordering
window parameters: Window Parameters
window pixel height: Window Sizes
window pixel width: Window Sizes
window point: Window Point
window point internals: Window Internals
window position: Window Point
window position: Coordinates and Windows
window position on display: Position Parameters
window positions and window managers: Position Parameters
window resizing: Resizing Windows
window selected within a frame: Selecting Windows
window selection change: Window Hooks
window size: Window Sizes
window size change: Window Hooks
window size on display: Size Parameters
window size, changing: Resizing Windows
window splitting: Splitting Windows
window start position: Window Start and End
window state: Window Configurations
window state change: Window Hooks
window state change flag: Window Hooks
window that satisfies a predicate: Cyclic Window Ordering
window top line: Window Start and End
window tree: Windows and Frames
window use time: Selecting Windows
window width: Window Sizes
window, a buffer display action alist entry: Buffer Display Action Alists
window-absolute-body-pixel-edges: Coordinates and Windows
window-absolute-pixel-edges: Coordinates and Windows
window-absolute-pixel-position: Coordinates and Windows
window-adjust-process-window-size-function: Process Buffers
window-at: Coordinates and Windows
window-at-side-p: Windows and Frames
window-atom, a window parameter: Window Parameters
window-atom-root: Atomic Windows
window-body-edges: Coordinates and Windows
window-body-height: Window Sizes
window-body-pixel-edges: Coordinates and Windows
window-body-size: Window Sizes
window-body-width: Window Sizes
window-bottom-divider-width: Window Dividers
window-buffer: Buffers and Windows
window-buffer-change-functions: Window Hooks
window-bump-use-time: Selecting Windows
window-child: Windows and Frames
window-combination-limit: Recombining Windows
window-combination-limit: Recombining Windows
window-combination-resize: Recombining Windows
window-combined-p: Windows and Frames
window-configuration-change-hook: Window Hooks
window-configuration-equal-p: Window Configurations
window-configuration-frame: Window Configurations
window-configuration-p: Window Configurations
window-current-scroll-bars: Scroll Bars
window-dedicated-p: Dedicated Windows
window-display-table: Active Display Table
window-edges: Coordinates and Windows
window-end: Window Start and End
window-font-height: Low-Level Font
window-font-width: Low-Level Font
window-frame: Windows and Frames
window-fringes: Fringe Size/Pos
window-full-height-p: Window Sizes
window-full-width-p: Window Sizes
window-group-end: Window Start and End
window-group-end-function: Window Start and End
window-group-start: Window Start and End
window-group-start-function: Window Start and End
window-header-line-height: Header Lines
window-header-line-height: Window Sizes
window-height, a buffer display action alist entry: Buffer Display Action Alists
window-hscroll: Horizontal Scrolling
window-id, a frame parameter: Management Parameters
window-in-direction: Windows and Frames
window-largest-empty-rectangle: Coordinates and Windows
window-left-child: Windows and Frames
window-line-height: Window Start and End
window-lines-pixel-dimensions: Size of Displayed Text
window-list: Windows and Frames
window-live-p: Basic Windows
window-main-window: Side Window Options and Functions
window-make-atom: Atomic Windows
window-margins: Display Margins
window-max-chars-per-line: Window Sizes
window-min-height: Window Sizes
window-min-height: Window Sizes
window-min-height, a buffer display action alist entry: Buffer Display Action Alists
window-min-size: Window Sizes
window-min-width: Window Sizes
window-min-width: Window Sizes
window-min-width, a buffer display action alist entry: Buffer Display Action Alists
window-minibuffer-p: Minibuffer Windows
window-mode-line-height: Window Sizes
window-next-buffers: Window History
window-next-sibling: Windows and Frames
window-old-body-pixel-height: Window Hooks
window-old-body-pixel-width: Window Hooks
window-old-buffer: Window Hooks
window-old-pixel-height: Window Hooks
window-old-pixel-width: Window Hooks
window-parameter: Window Parameters
window-parameters: Window Parameters
window-parameters, a buffer display action alist entry: Buffer Display Action Alists
window-parent: Windows and Frames
window-persistent-parameters: Window Parameters
window-pixel-edges: Coordinates and Windows
window-pixel-height: Window Sizes
window-pixel-width: Window Sizes
window-point: Window Point
window-point-insertion-type: Window Point
window-preserve-size: Preserving Window Sizes
window-preserved-size: Preserving Window Sizes
window-preserved-size, a window parameter: Window Parameters
window-prev-buffers: Window History
window-prev-sibling: Windows and Frames
window-resizable: Resizing Windows
window-resize: Resizing Windows
window-resize-pixelwise: Resizing Windows
window-right-divider-width: Window Dividers
window-scroll-bar-height: Scroll Bars
window-scroll-bar-width: Scroll Bars
window-scroll-bars: Scroll Bars
window-scroll-functions: Window Hooks
window-selection-change-functions: Window Hooks
window-setup-hook: Init File
window-side, a window parameter: Window Parameters
window-sides-reversed: Side Window Options and Functions
window-sides-slots: Side Window Options and Functions
window-sides-vertical: Side Window Options and Functions
window-size, a buffer display action alist entry: Buffer Display Action Alists
window-size-change-functions: Window Hooks
window-size-fixed: Preserving Window Sizes
window-slot, a window parameter: Window Parameters
window-start: Window Start and End
window-state-change-functions: Window Hooks
window-state-change-hook: Window Hooks
window-state-get: Window Configurations
window-state-put: Window Configurations
window-swap-states: Window Configurations
window-system: Window Systems
window-system: Window Systems
window-system window: Basic Windows
window-system-initialization: Startup Summary
window-tab-line-height: Window Sizes
window-text-pixel-size: Size of Displayed Text
window-toggle-side-windows: Side Window Options and Functions
window-top-child: Windows and Frames
window-total-height: Window Sizes
window-total-size: Window Sizes
window-total-width: Window Sizes
window-tree: Windows and Frames
window-use-time: Selecting Windows
window-valid-p: Basic Windows
window-vscroll: Vertical Scrolling
window-width, a buffer display action alist entry: Buffer Display Action Alists
windowp: Basic Windows
windows, controlling precisely: Buffers and Windows
windows, recombining: Recombining Windows
with-case-table: Case Tables
with-coding-priority: Specifying Coding Systems
with-connection-local-application-variables: Applying Connection Local Variables
with-connection-local-variables: Applying Connection Local Variables
with-current-buffer: Current Buffer
with-current-buffer-window: Temporary Displays
with-delayed-message: Progress
with-demoted-errors: Handling Errors
with-environment-variables: System Environment
with-eval-after-load: Hooks for Loading
with-existing-directory: Testing Accessibility
with-file-modes: Changing Files
with-help-window: Help Functions
with-local-quit: Quitting
with-mutex: Mutexes
with-no-warnings: Compiler Errors
with-output-to-string: Output Functions
with-output-to-temp-buffer: Temporary Displays
with-restriction: Narrowing
with-selected-frame: Selecting Windows
with-selected-window: Selecting Windows
with-silent-modifications: Buffer Modification
with-silent-modifications, and changes in text properties: Changing Properties
with-sqlite-transaction: Database
with-suppressed-warnings: Compiler Errors
with-syntax-table: Syntax Table Functions
with-temp-buffer: Current Buffer
with-temp-buffer-window: Temporary Displays
with-temp-file: Writing to Files
with-temp-message: Displaying Messages
with-timeout: Timers
with-undo-amalgamate: Undo
without-restriction: Narrowing
word-boundary in rx: Rx Constructs
word-end in rx: Rx Constructs
word-search-backward: String Search
word-search-backward-lax: String Search
word-search-forward: String Search
word-search-forward-lax: String Search
word-search-regexp: String Search
word-start in rx: Rx Constructs
words in region: Text Lines
words-include-escapes: Word Motion
wrap-prefix: Truncation
write-abbrev-file: Abbrev Files
write-char: Output Functions
write-contents-functions: Saving Buffers
write-file: Saving Buffers
write-file-functions: Saving Buffers
write-region: Writing to Files
write-region: Files and Storage
write-region-annotate-functions: Format Conversion Piecemeal
write-region-inhibit-fsync: Writing to Files
write-region-post-annotation-function: Format Conversion Piecemeal
writing a documentation string: Documentation Basics
writing buffer display actions: The Zen of Buffer Display
writing emacs modules: Writing Dynamic Modules
writing Emacs primitives: Writing Emacs Primitives
writing module functions: Module Functions
writing to files: Writing to Files
wrong-number-of-arguments: Argument List
wrong-type-argument: Type Predicates

X
X display names: Multiple Terminals
X Window System: Window Systems
x-alt-keysym: X11 Keysyms
x-alternatives-map: Standard Keymaps
x-begin-drag: Drag and Drop
x-bitmap-file-path: Face Attributes
x-close-connection: Multiple Terminals
x-color-defined-p: Color Names
x-color-values: Color Names
x-ctrl-keysym: X11 Keysyms
x-defined-colors: Color Names
x-display-color-p: Display Feature Testing
x-display-list: Multiple Terminals
x-dnd-direct-save-function: Drag and Drop
x-dnd-disable-motif-protocol: Drag and Drop
x-dnd-known-types: Drag and Drop
x-dnd-save-direct: Drag and Drop
x-dnd-save-direct-immediately: Drag and Drop
x-dnd-test-function: Drag and Drop
x-dnd-types-alist: Drag and Drop
x-dnd-use-offix-drop: Drag and Drop
x-dnd-use-unsupported-drop: Drag and Drop
x-double-buffered-p: Visibility of Frames
x-family-fonts: Font Lookup
x-focus-frame: Input Focus
x-get-resource: Resources
x-hyper-keysym: X11 Keysyms
x-list-fonts: Font Lookup
x-meta-keysym: X11 Keysyms
x-open-connection: Multiple Terminals
x-parse-geometry: Geometry
x-pointer-shape: Pointer Shape
x-popup-dialog: Dialog Boxes
x-popup-menu: Pop-Up Menus
x-pre-popup-menu-hook: Pop-Up Menus
x-resource-class: Resources
x-resource-name: Resources
x-sensitive-text-pointer-shape: Pointer Shape
x-server-vendor: Display Feature Testing
x-server-version: Display Feature Testing
x-setup-function-keys: Standard Keymaps
x-stretch-cursor: Cursor Parameters
x-super-keysym: X11 Keysyms
X11 keysyms: X11 Keysyms
XBM: XBM Images
xdigit character class, regexp: Char Classes
XDS: Drag and Drop
XML DOM: Document Object Model
xor: Combining Conditions
XPM: XPM Images
xwidget: Xwidgets
xwidget callbacks: Xwidget Events
xwidget property list: Xwidgets
xwidget type: Xwidget Type
xwidget-buffer: Xwidgets
xwidget-display-event event: Xwidget Events
xwidget-event event: Xwidget Events
xwidget-info: Xwidgets
xwidget-live-p: Xwidgets
xwidget-perform-lispy-event: Xwidgets
xwidget-plist: Xwidgets
xwidget-query-on-exit-flag: Xwidgets
xwidget-resize: Xwidgets
xwidget-size-request: Xwidgets
xwidget-view type: Xwidget Type
xwidget-webkit-back-forward-list: Xwidgets
xwidget-webkit-estimated-load-progress: Xwidgets
xwidget-webkit-execute-script: Xwidgets
xwidget-webkit-execute-script-rv: Xwidgets
xwidget-webkit-finish-search: Xwidgets
xwidget-webkit-get-title: Xwidgets
xwidget-webkit-goto-history: Xwidgets
xwidget-webkit-goto-uri: Xwidgets
xwidget-webkit-load-html: Xwidgets
xwidget-webkit-next-result: Xwidgets
xwidget-webkit-previous-result: Xwidgets
xwidget-webkit-search: Xwidgets
xwidget-webkit-set-cookie-storage-file: Xwidgets
xwidget-webkit-stop-loading: Xwidgets
xwidgetp: Xwidgets

Y
y-or-n-p: Yes-or-No Queries
y-or-n-p-with-timeout: Yes-or-No Queries
yank: Yank Commands
yank suppression: Changing Key Bindings
yank-excluded-properties: Yanking
yank-handled-properties: Yanking
yank-media-handler: Yanking Media
yank-pop: Yank Commands
yank-transform-functions: Yanking
yank-undo-function: Yank Commands
yanking and text properties: Yanking
yes-or-no questions: Yes-or-No Queries
yes-or-no-p: Yes-or-No Queries

Z
z-group, a frame parameter: Position Parameters
Z-order: Raising and Lowering
zero-or-more in rx: Rx Constructs
zero-or-one in rx: Rx Constructs
zerop: Predicates on Numbers
zlib-available-p: Decompression
zlib-decompress-region: Decompression


Footnotes

(1)

You may also encounter ‘#^^’, used for sub-char-tables.

(2)

This is the behavior specified for languages like Common Lisp and C for constants, and this differs from languages like JavaScript and Python where an interpreter is required to signal an error if a program attempts to change an immutable object. Ideally the Emacs Lisp interpreter will evolve in latter direction.

(3)

For more information about collation rules and their locale dependencies, see The Unicode Collation Algorithm. Some Standard C libraries, such as the GNU C Library (a.k.a. glibc) implement large portions of the Unicode Collation Algorithm and use the associated locale data, Common Locale Data Repository, or CLDR.

(4)

It is sometimes also referred to as a true list, but we generally do not use this terminology in this manual.

(5)

There is no strictly equivalent way to add an element to the end of a list. You can use (append listname (list newelt)), which creates a whole new list by copying listname and adding newelt to its end. Or you can use (nconc listname (list newelt)), which modifies listname by following all the CDRs and then replacing the terminating nil. Compare this to adding an element to the beginning of a list with cons, which neither copies nor modifies the list.

(6)

This usage of “key” is not related to the term “key sequence”; it means a value used to look up an item in a table. In this case, the table is the alist, and the alist associations are the items.

(7)

It is sometimes also referred to as an S-expression or sexp, but we generally do not use this terminology in this manual.

(8)

This definition of “environment” is specifically not intended to include all the data that can affect the result of a program.

(9)

To be precise, under the default dynamic scoping rule, the value cell always holds the variable’s current value, but this is not the case under the lexical scoping rule. See Scoping Rules for Variable Bindings, for details.

(10)

With some exceptions; for instance, a lexical binding can also be accessed from the Lisp debugger.

(11)

The MS-DOS version of Emacs uses _dir-locals.el instead, due to limitations of the DOS filesystems.

(12)

This only works in code using lexical-binding.

(13)

This is related to, but different from currying, which transforms a function that takes multiple arguments in such a way that it can be called as a chain of functions, each one with a single argument.

(14)

If the number of arguments that func can accept is unlimited, then the new function will also accept an unlimited number of arguments, so in that case apply-partially doesn’t reduce the number of arguments that the new function could accept.

(15)

Note that unlike the built-in function, this version accepts any number of arguments.

(16)

Some elements actually supply two arguments.

(17)

Button-down is the conservative antithesis of drag.

(18)

It is required for menus which do not use a toolkit, e.g., on a text terminal.

(19)

In MS-Windows versions of Emacs compiled for the Cygwin environment, you can use the functions cygwin-convert-file-name-to-windows and cygwin-convert-file-name-from-windows to convert between the two file-name syntaxes.

(20)

Emacs follows the GNU convention to use the term file name instead of the term pathname. We use the term path only for search paths, which are lists of directory names.

(21)

On PGTK frames, setting the values fullheight and fullwidth has no effect.

(22)

On Haiku, child frames are only visible when a parent frame is active, owing to a limitation of the Haiku windowing system. Owing to the same limitation, child frames are only guaranteed to appear above their top-level parent; that is to say, the top-most frame in the hierarchy, which does not have a parent frame.

(23)

An RFC, an acronym for Request for Comments, is a numbered Internet informational document describing a standard. RFCs are usually written by technical experts acting on their own initiative, and are traditionally written in a pragmatic, experience-driven manner.

(24)

This internal representation is based on one of the encodings defined by the Unicode Standard, called UTF-8, for representing any Unicode codepoint, but Emacs extends UTF-8 to represent the additional codepoints it uses for raw 8-bit bytes and characters not unified with Unicode.

(25)

The Unicode specification writes these tag names inside ‘<..>’ brackets, but the tag names in Emacs do not include the brackets; e.g., Unicode specifies ‘<small>’ where Emacs uses ‘small’.

(26)

It could be written much simpler with non-greedy operators (how?), but that would make the example less interesting.

(27)

Note that regexp-opt does not guarantee that its result is absolutely the most efficient form possible. A hand-tuned regular expression can sometimes be slightly more efficient, but is almost never worth the effort.

(28)

On other systems, Emacs uses a Lisp emulation of ls; see Contents of Directories.

(29)

For backward compatibility, you can also use a string to specify a face name; that is equivalent to a Lisp symbol with the same name.

(30)

In this context, the term font has nothing to do with Font Lock (see Font Lock Mode).

(31)

On MS-Windows, this requires w32-use-native-image-API to be set non-nil.

(32)

In typography an em is a distance equivalent to the height of the type. For example when using 12 point type 1 em is equal to 12 points. Its use ensures distances and type remain proportional.

(33)

The benefits of a Common Lisp-style package system are considered not to outweigh the costs.

(34)

We do use these occasionally, but try not to overdo it.