GNU Texinfo 6.8

Next: , Up: (dir)   [Contents][Index]

Texinfo

This manual is for GNU Texinfo (version 6.8, 8 June 2021), a documentation system that can produce both online information and a printed manual from a single source using semantic markup.

Table of Contents

Short Table of Contents


Next: , Previous: , Up: Texinfo   [Contents][Index]

Texinfo Copying Conditions

GNU Texinfo is free software; this means that everyone is free to use it and free to redistribute it on certain conditions. Texinfo is not in the public domain; it is copyrighted and there are restrictions on its distribution, but these restrictions are designed to permit everything that a good cooperating citizen would want to do. What is not allowed is to try to prevent others from further sharing any version of Texinfo that they might get from you.

Specifically, we want to make sure that you have the right to give away copies of the programs that relate to Texinfo, that you receive source code or else can get it if you want it, that you can change these programs or use pieces of them in new free programs, and that you know you can do these things.

To make sure that everyone has such rights, we have to forbid you to deprive anyone else of these rights. For example, if you distribute copies of the Texinfo related programs, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must tell them their rights.

Also, for our own protection, we must make certain that everyone finds out that there is no warranty for the programs that relate to Texinfo. If these programs are modified by someone else and passed on, we want their recipients to know that what they have is not what we distributed, so that any problems introduced by others will not reflect on our reputation.

The precise conditions of the licenses for the programs currently being distributed that relate to Texinfo are found in the General Public Licenses that accompany them. This manual is covered by the GNU Free Documentation License (see GNU Free Documentation License).


1 Overview of Texinfo

Texinfo is a documentation system that uses a single source file to produce both online information and printed output. This means that instead of writing several different documents, one for each output format, you need only write one document.

Using Texinfo, you can create a printed document (via the TeX typesetting system) in PDF or PostScript format, including chapters, sections, cross-references, and indices. From the same Texinfo source file, you can create an HTML output file suitable for use with a web browser, you can create an Info file with special features to make browsing documentation easy, and also create a DocBook file or a transliteration to XML format.

A Texinfo source file is a plain text file containing text interspersed with @-commands (words preceded by an ‘@’) that tell the Texinfo processors what to do. Texinfo’s markup commands are almost entirely semantic; that is, they specify the intended meaning of text in the document, rather than physical formatting instructions. You can edit a Texinfo file with any text editor, but it is especially convenient to use GNU Emacs since that editor has a special mode, called Texinfo mode, that provides various Texinfo-related features. (See Using Texinfo Mode.)

Texinfo was devised specifically for the purpose of writing software documentation and manuals. If you want to write a good manual for your program, Texinfo has many features which we hope will make your job easier. However, Texinfo is not intended to be a general-purpose formatting program. It provides almost no commands for controlling the final formatting, so may be inappropriate for your needs if you if you want to lay out a newspaper, devise a glossy magazine ad, or follow the exact formatting requirements of a publishing house.

Spell “Texinfo” with a capital “T” and the other letters in lowercase. The first syllable of “Texinfo” is pronounced like “speck”, not “hex”. This odd pronunciation is derived from the pronunciation of TeX. Pronounce TeX as if the ‘X’ were the last sound in the name ‘Bach’. In the word TeX, the ‘X’ is, rather than the English letter “ex”, actually the Greek letter “chi”.

Texinfo is the official documentation format of the GNU project. More information, including manuals for GNU packages, is available at the GNU documentation web page.


1.1 Reporting Bugs

We welcome bug reports and suggestions for any aspect of the Texinfo system: programs, documentation, installation, etc. Please email them to bug-texinfo@gnu.org. You can get the latest version of Texinfo via its home page, http://www.gnu.org/software/texinfo.

For bug reports, please include enough information for the maintainers to reproduce the problem. Generally speaking, that means:

  • The version number of Texinfo and the program(s) or manual(s) involved.
  • The contents of any input files necessary to reproduce the bug.
  • Precisely how you ran any program(s) involved.
  • A description of the problem and samples of any erroneous output.
  • Hardware and operating system names and versions.
  • Anything else that you think would be helpful.

When in doubt whether something is needed or not, include it. It’s better to include too much than to leave out something important.

It is critical to send an actual input file that reproduces the problem. What’s not critical is to “narrow down” the example to the smallest possible input—the actual input with which you discovered the bug will suffice. (Of course, if you do do experiments, the smaller the input file, the better.)

Any problems with the Info reader in Emacs should be reported to the Emacs developers: see Bugs in The GNU Emacs Manual.

Patches are most welcome; if possible, please make them with ‘diff -c’ (see Comparing and Merging Files) and include ChangeLog entries (see Change Log in The GNU Emacs Manual), and follow the existing coding style.


1.2 Output Formats

Here is a brief overview of the output formats currently supported by Texinfo.

Info

(Generated via makeinfo.) Info format is mostly a plain text transliteration of the Texinfo source. It adds a few control characters to provide navigational information for cross-references, indices, and so on. The Emacs Info subsystem (see Info), and the standalone info program (see GNU Info), among others, can read these files. See Info Files, and Creating and Installing Info Files.

Plain text

(Generated via makeinfo --plaintext.) This is almost the same as Info output with the navigational control characters are omitted.

HTML

(Generated via makeinfo --html.) HTML, standing for Hyper Text Markup Language, has become the most commonly used language for writing documents on the World Wide Web. Web browsers, such as Mozilla, Lynx, and Emacs-W3, can render this language online. There are many versions of HTML, both different standards and browser-specific variations. makeinfo tries to use a subset of the language that can be interpreted by any common browser, intentionally not using many newer or less widely-supported tags. Although the native output is thus rather plain, it can be customized at various levels, if desired. For details of the HTML language and much related information, see http://www.w3.org/MarkUp/. See Generating HTML.

DVI

(Generated via texi2dvi.) The DeVIce Independent binary format is output by the TeX typesetting program (http://tug.org). This is then read by a DVI ‘driver’, which knows the actual device-specific commands that can be viewed or printed, notably Dvips for translation to PostScript (see Dvips) and Xdvi for viewing on an X display (http://sourceforge.net/projects/xdvi/). See Formatting and Printing Hardcopy. (Be aware that the Texinfo language is very different from and much stricter than TeX’s usual languages: plain TeX, LaTeX, ConTeXt, etc.)

PostScript

(Generated via texi2dvi --ps.) PostScript is a page description language that became widely used around 1985 and is still used today. http://en.wikipedia.org/wiki/PostScript gives a basic description and more preferences. By default, Texinfo uses the dvips program to convert TeX’s DVI output to PostScript. See Dvips.

PDF

(Generated via texi2dvi --pdf or texi2pdf.) This format was developed by Adobe Systems for portable document interchange, based on their previous PostScript language. It can represent the exact appearance of a document, including fonts and graphics, and supporting arbitrary scaling. It is intended to be platform-independent and easily viewable, among other design goals; http://en.wikipedia.org/wiki/Portable_Document_Format and http://tug.org/TUGboat/tb22-3/tb72beebe-pdf.pdf have some background. By default, Texinfo uses the pdftex program, an extension of TeX, to output PDF; see http://tug.org/applications/pdftex. See PDF Output.

DocBook

(Generated via makeinfo --docbook.) This is an XML-based format developed some years ago, primarily for technical documentation. It therefore bears some resemblance, in broad outline, to Texinfo. See http://www.docbook.org. Various converters from DocBook to Texinfo have also been developed; see the Texinfo web pages.

XML

(Generated via makeinfo --xml.) XML is a generic syntax specification usable for any sort of content (a reference is at http://www.w3.org/XML). The makeinfo XML output, unlike all the other output formats, is a transliteration of the Texinfo source rather than processed output. That is, it translates the Texinfo markup commands into XML syntax, for further processing by XML tools. The XML contains enough information to recreate the original content, except for syntactic constructs such as Texinfo macros and conditionals. The Texinfo source distribution includes a utility script txixml2texi to do that backward transformation.

The details of the output syntax are defined in an XML DTD as usual, which is contained in a file texinfo.dtd included in the Texinfo source distribution and available via the Texinfo web pages. Texinfo XML files, and XML files in general, cannot be viewed in typical web browsers; they won’t follow the DTD reference and as a result will simply report a (misleading) syntax error.


1.3 Info Files

As mentioned above, Info format is mostly a plain text transliteration of the Texinfo source, with the addition of a few control characters to separate nodes and provide navigational information, so that Info-reading programs can operate on it.

Info files are nearly always created by processing a Texinfo source document. makeinfo, also known as texi2any, is the principal command that converts a Texinfo file into an Info file; see texi2any: The Generic Translator for Texinfo.

Generally, you enter an Info file through a node that by convention is named ‘Top’. This node normally contains just a brief summary of the file’s purpose, and a large menu through which the rest of the file is reached. From this node, you can either traverse the file systematically by going from node to node, or you can go to a specific node listed in the main menu, or you can search the index menus and then go directly to the node that has the information you want. Alternatively, with the standalone Info program, you can specify specific menu items on the command line (see Info).

If you want to read through an Info file in sequence, as if it were a printed manual, you can hit SPC repeatedly, or you get the whole file with the advanced Info command g *. (See Advanced Info commands in Info.)

The dir file in the info directory serves as the departure point for the whole Info system. From it, you can reach the ‘Top’ nodes of each of the documents in a complete Info system.

If you wish to refer to an Info file via a URI, you can use the (unofficial) syntax exemplified by the following. This works with Emacs/W3, for example:

info:emacs#Dissociated%20Press
info:///usr/info/emacs#Dissociated%20Press
info://localhost/usr/info/emacs#Dissociated%20Press

The info program itself does not follow URIs of any kind.


1.4 Printed Books

A Texinfo file can be formatted and typeset as a printed book or manual. To do this, you need TeX, a sophisticated typesetting program written by Donald Knuth of Stanford University. It is not part of the Texinfo distribution.

Texinfo provides a file texinfo.tex that contains the definitions that TeX uses when it typesets a Texinfo file. You can get the latest version of texinfo.tex from the Texinfo home page, http://www.gnu.org/software/texinfo/.

A Texinfo-based book is similar to any other typeset, printed work: it can have a title page, copyright page, table of contents, and preface, as well as chapters, numbered or unnumbered sections and subsections, page headers, cross-references, footnotes, and indices.

TeX is very powerful and has a great many features. However, because a Texinfo file must be able to present information both on a character-only terminal in Info form and in a typeset book, the formatting commands that Texinfo supports are necessarily limited.

See Formatting and Printing Hardcopy, for more information on processing a manual with TeX.


1.5 Adding Output Formats

The output formats in the previous sections handle a wide variety of usage, but of course there is always room for more.

If you are a programmer and would like to contribute to the GNU project by implementing additional output formats for Texinfo, that would be excellent. The way to do this that would be most useful is to write a new back-end for texi2any, our reference implementation of a Texinfo parser; it creates a tree representation of the Texinfo input that you can use for the conversion. The documentation in the source file tp/Texinfo/Convert/Converter.pm is a good place to start. See texi2any: The Generic Translator for Texinfo.

Another viable approach is use the Texinfo XML output from texi2any as your input. This XML is an essentially complete representation of the input, but without the Texinfo syntax and option peculiarities, as described above.

If you still cannot resist the temptation of writing a new program that reads Texinfo source directly, let us give some more caveats: please do not underestimate the amount of work required. Texinfo is by no means a simple language to parse correctly, and remains under development, so you would be committing to an ongoing task. You are advised to check that the tests of the language that come with texi2any give correct results with your new program.

From time to time, proposals are made to generate traditional Unix man pages from Texinfo source. However, because man pages have a strict conventional format, creating a good man page requires a completely different source from that needed for the typical Texinfo applications of writing a good user tutorial and/or a good reference manual. This makes generating man pages incompatible with the Texinfo design goal of not having to document the same information in different ways for different output formats. You might as well write the man page directly.

As an alternative way to support man pages, you may find the program help2man to be useful. It generates a traditional man page from the ‘--help’ output of a program. In fact, the man pages for the programs in the Texinfo distribution are generated with this. It is GNU software written by Brendan O’Dea, available from http://www.gnu.org/software/help2man.


1.6 History

Richard M. Stallman invented the Texinfo format, wrote the initial processors, and created Edition 1.0 of this manual. Robert J. Chassell greatly revised and extended the manual, starting with Edition 1.1. Brian Fox was responsible for the standalone Texinfo distribution until version 3.8, and originally wrote the standalone makeinfo and info programs. Karl Berry continued maintenance from Texinfo 3.8 (manual edition 2.22), and Gavin Smith has continued maintenance since Texinfo 6.0.

Our thanks go out to all who helped improve this work, particularly the indefatigable Eli Zaretskii and Andreas Schwab, who have provided patches beyond counting. François Pinard and David D. Zuhn, tirelessly recorded and reported mistakes and obscurities. Zack Weinberg did the impossible by implementing the macro syntax in texinfo.tex. Thanks to Melissa Weisshaus for her frequent reviews of nearly similar editions. Dozens of others have contributed patches and suggestions, they are gratefully acknowledged in the ChangeLog file. Our mistakes are our own.

Beginnings

In the 1970’s at CMU, Brian Reid developed a program and format named Scribe to mark up documents for printing. It used the @ character to introduce commands, as Texinfo does. Much more consequentially, it strove to describe document contents rather than formatting, an idea wholeheartedly adopted by Texinfo.

Meanwhile, people at MIT developed another, not too dissimilar format called Bolio. This then was converted to using TeX as its typesetting language: BoTeX. The earliest BoTeX version seems to have been 0.02 on October 31, 1984.

BoTeX could only be used as a markup language for documents to be printed, not for online documents. Richard Stallman (RMS) worked on both Bolio and BoTeX. He also developed a nifty on-line help format called Info, and then combined BoTeX and Info to create Texinfo, a mark up language for text that is intended to be read both online and as printed hard copy.

Moving forward, the original translator to create Info was written (primarily by RMS and Bob Chassell) in Emacs Lisp, namely the texinfo-format-buffer and other functions. In the early 1990s, Brian Fox reimplemented the conversion program in C, now called makeinfo.

Reimplementing in Perl

In 2012, the C makeinfo was itself replaced by a Perl implementation generically called texi2any. This version supports the same level of output customization as texi2html, an independent program originally written by Lionel Cons, later with substantial work by many others. The many additional features needed to make texi2html a replacement for makeinfo were implemented by Patrice Dumas. The first never-released version of texi2any was based on the texi2html code.

That implementation, however, was abandoned in favor of the current program (also written by Patrice Dumas), which parses the Texinfo input into a tree for processing. It inherited the design of customization and other features from texi2html (for more on texi2html compatibility, see texi2html: Ancestor of texi2any). However, texi2any is a full reimplementation: it constructs a tree-based representation of the input document for all back-ends to work from.

The new Perl program is much slower than the old C program. The speed gap has partially closed since first release, but it may not ever be entirely comparable. So why did we switch? In short, we intend and hope that the present program will be much easier than the previous C implementation of makeinfo to extend to different output styles, back-end output formats, and all other customizations. In more detail:

  • HTML customization. Many GNU and other free software packages had been happily using the HTML customization features in texi2html for years. Thus, in effect two independent implementations of the Texinfo language had developed, and keeping them in sync was not simple. Adding the HTML customization possible in texi2html to a C program would have been an enormous effort.
  • Unicode, and multilingual support generally, especially of east Asian languages. Although of course it’s perfectly plausible to write such support in C, in the particular case of makeinfo, it would have been tantamount to rewriting the entire program. In Perl, much of that comes essentially for free.
  • Additional back-ends. The makeinfo code had become convoluted to the point where adding a new back-end was quite complex, requiring complex interactions with existing back-ends. In contrast, our Perl implementation provides a clean tree-based representation for all back-ends to work from. People have requested numerous different back-ends (LaTeX, the latest (X)HTML, …), and they will now be much more feasible to implement. Which leads to the last item:
  • Making contributions easier. In general, due to the cleaner structure, the Perl program should be considerably easier than the C for anyone to read and contribute to, with the resulting obvious benefits.

texi2any is intended to be a reference implementation that defines parts of the language not fully specified by the manual. Without such a reference, alternative implementations would be very likely to have subtle, or not-so-subtle, differences in behavior, and thus Texinfo documents would become dependent on the processor. It is also important to have consistent command-line options for all processors. Extensive tests of the language and processor were developed at the same time as texi2any; we encourage anyone thinking of writing a program to parse Texinfo input to make use of these tests.

With the release of texi2any as the reference implementation, development of both the C implementation of makeinfo and texi2html has been halted. Going forward, we ask authors of Texinfo documents to use only texi2any.


Next: , Previous: , Up: Texinfo   [Contents][Index]

2 Writing a Texinfo File

This chapter describes Texinfo syntax and what is required in a Texinfo file, and gives a short sample file.


2.1 General Syntactic Conventions

This section describes the general conventions used in all Texinfo documents.

  • All printable ASCII characters except ‘@’, ‘{’ and ‘}’ can appear in a Texinfo file and stand for themselves. ‘@’ is the escape character which introduces commands, while ‘{’ and ‘}’ are used to surround arguments to certain commands. To put one of these special characters into the document, put an ‘@’ character in front of it, like this: ‘@@’, ‘@{’, and ‘@}’.
  • In a Texinfo file, the commands you write to describe the contents of the manual are preceded by an ‘@’ character; they are called @-commands. (The ‘@’ in Texinfo has the same meaning that ‘\’ has in plain TeX.)

    Depending on what they do or what arguments they take, you need to write @-commands on lines of their own, or as part of sentences. As a general rule, a command requires braces if it mingles among other text; but it does not need braces if it is on a line of its own. For more details of Texinfo command syntax, see @-Command Syntax.

  • Whitespace following an @-command name is optional and (usually) ignored if present. The exceptions are contexts when whitespace is significant, e.g., an @example environment.
  • Texinfo supports the usual quotation marks used in English and in other languages; see Inserting Quotation Marks.
  • Use three hyphens in a row, ‘---’, to produce a long dash—like this (called an em dash), used for punctuation in sentences. Use two hyphens, ‘--’, to produce a medium dash (called an en dash), used primarily for numeric ranges, as in “June 25–26”. Use a single hyphen, ‘-’, to produce a standard hyphen used in compound words. For display on the screen, Info reduces three hyphens to two and two hyphens to one (not transitively!). Of course, any number of hyphens in the source remain as they are in literal contexts, such as @code and @example.
  • Do not use tab characters in a Texinfo file! (Except perhaps in verbatim modes.) TeX uses variable-width fonts, which means that it is impractical at best to define a tab to work in all circumstances. Consequently, TeX treats tabs like single spaces, and that is not what they look like in the source. Furthermore, makeinfo does nothing special with tabs, and thus a tab character in your input file will usually have a different appearance in the output.

    To avoid this problem, Texinfo mode in GNU Emacs inserts multiple spaces when you press the TAB key. Also, you can run untabify in Emacs to convert tabs in a region to multiple spaces, or use the unexpand command from the shell.

  • Lastly, form feed (CTRL-l) characters in the input are handled as follows:
    PDF/DVI

    In normal text, treated as ending any open paragraph; essentially ignored between paragraphs.

    Info

    Output as-is between paragraphs (their most common use); in other contexts, they may be treated as regular spaces (and thus consolidated with surrounding whitespace).

    HTML

    Written as a numeric entity except contexts where spaces are ignored; for example, in ‘@footnote{ ^L foo}’, the form feed is ignored.

    XML

    Keep them everywhere; in attributes, escaped as ‘\f’; also, ‘\’ is escaped as ‘\\’ and newline as ‘\n’.

    DocBook

    Completely removed, as they are not allowed.

    As you can see, because of these differing requirements of the output formats, it’s not possible to use form feeds completely portably.


2.2 Comments

You can write comments in a Texinfo file by using the @comment command, which may be abbreviated to @c. Such comments are for a person looking at the Texinfo source file. All the text on a line that follows either @comment or @c is a comment; the rest of the line does not appear in the visible output. (To be precise, the character after the @c or @comment must be something other than a dash or alphanumeric, or it will be taken as part of the command.)

Often, you can write the @comment or @c in the middle of a line, and only the text that follows after the @comment or @c command does not appear; but some commands, such as @settitle, work on a whole line. You cannot use @comment or @c within a line beginning with such a command.

In cases of nested command invocations, complicated macro definitions, etc., @c and @comment may provoke an error when processing with TeX. Therefore, you can also use the DEL character (ASCII 127 decimal, 0x7f hex, 0177 octal) as a true TeX comment character (catcode 14, in TeX internals). Everything on the line after the DEL will be ignored.

You can also have long stretches of text ignored by the Texinfo processors with the @ignore and @end ignore commands. Write each of these commands on a line of its own, starting each command at the beginning of the line. Text between these two commands does not appear in the processed output. You can use @ignore and @end ignore for writing comments. (For some caveats regarding nesting of such commands, see Conditional Nesting.)


2.3 What a Texinfo File Must Have

By convention, the name of a Texinfo file ends with one of the extensions .texi, .texinfo, .txi, or .tex.

In order to be made into a printed manual and other output formats, a Texinfo file must begin with lines like this:

\input texinfo
@settitle name-of-manual

The contents of the file follow this beginning, and then you must end the Texinfo source with a line like this:

@bye

The @bye line at the end of the file on a line of its own tells the formatters that the file is ended and to stop formatting. If you leave this out, you’ll be dumped at TeX’s prompt at the end of the run.

Furthermore, you will usually provide a Texinfo file with a title page, indices, and the like, all of which are explained in this manual. But the minimum, which can be useful for short documents, is just the two lines at the beginning and the one line at the end.


2.4 Short Sample

Here is a short sample Texinfo file. The makeinfo program transforms a Texinfo source file such as this into an Info file or HTML, and TeX typesets it for a printed manual.

\input texinfo
@settitle Sample Manual 1.0

@copying
This is a short example of a complete Texinfo file.

Copyright @copyright{} 2016 Free Software Foundation, Inc.
@end copying

@titlepage
@title Sample Title
@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage

@contents

@node Top
@top GNU Sample

This manual is for GNU Sample
(version @value{VERSION}, @value{UPDATED}).

@menu
* First Chapter::    The first chapter is the
                      only chapter in this sample.
* Index::            Complete index.
@end menu


@node First Chapter
@chapter First Chapter

@cindex chapter, first
This is the first chapter.
@cindex index entry, another

Here is a numbered list.

@enumerate
@item
This is the first item.

@item
This is the second item.
@end enumerate


@node First Section
@section First Section

First section of first chapter.


@node Second Section
@section Second Section

Second section of first chapter.


@node Index
@unnumbered Index

@printindex cp

@bye

2.5 Texinfo File Header

Texinfo files start with at least two lines:

\input texinfo
@settitle Sample Manual 1.0

The ‘\input texinfo’ line tells TeX to use the texinfo.tex file, which tells TeX how to translate the Texinfo @-commands into TeX typesetting commands. (Note the use of the backslash, ‘\’; this is correct for TeX.)

The @settitle line specifies a title for the page headers (or footers) of the printed manual, and the default title and document description for the ‘<head>’ in HTML. (Strictly speaking, @settitle is optional—if you don’t mind your document being titled ‘Untitled’.)

Also, you can optionally write the @settitle line between start-of-header and end-of-header lines if you want to format just part of the Texinfo file in Emacs.

It makes sense to include any command that affects document formatting as a whole in the header. @synindex (see @synindex: Combining Indices), for instance, is another command often included in the header.


2.5.1 The First Line of a Texinfo File

Every Texinfo file that is to be the top-level input to TeX must begin with a line that looks like this:

\input texinfo

When the file is processed by TeX, the ‘\input texinfo’ command tells TeX to load the macros needed for processing a Texinfo file. These are in a file called texinfo.tex, which should have been installed on your system along with either the TeX or Texinfo software. TeX uses the backslash, ‘\’, to mark the beginning of a command, exactly as Texinfo uses ‘@’. The texinfo.tex file causes the switch from ‘\’ to ‘@’; before the switch occurs, TeX requires ‘\’, which is why it appears at the beginning of the file.

You may optionally follow this line with a comment to tell GNU Emacs to use Texinfo mode when the file is edited:

\input texinfo   @c -*-texinfo-*-

This may be useful when Emacs doesn’t detect the file type from the file extension automatically.


2.5.2 Start of Header

A start-of-header line is a Texinfo comment that looks like this:

@c %**start of header

Write the start-of-header line on the second line of a Texinfo file. Follow the start-of-header line with an @settitle line and, optionally, with other commands that globally affect the document formatting, such as @synindex or @footnotestyle; and then by an end-of-header line (see End of Header).

The start- and end-of-header lines allow you to format only part of a Texinfo file for Info or printing. See The texinfo-format… Commands.

The odd string of characters, ‘%**’, is to ensure that no other comment is accidentally taken for a start-of-header line. You can change it if you wish by setting the tex-start-of-header and/or tex-end-of-header Emacs variables. See Formatting and Printing in Texinfo Mode.


2.5.3 @setfilename: Set the Output File Name

The @setfilename line specifies the name of the output file to be generated. When present, it should be the first Texinfo command (that is, after ‘\input texinfo’). Write the @setfilename command at the beginning of a line and follow it on the same line by the Info file name.

@setfilename info-file-name

The name must be different from the name of the Texinfo file. There are two conventions for choosing the name: you can either remove the extension (such as ‘.texi’) entirely from the input file name, or (recommended) replace it with the ‘.info’ extension.

When a @setfilename line is present, the Texinfo processors ignore everything written before the @setfilename line. This is why the very first line of the file (the \input line) does not show up in the output.

If there is no @setfilename line, makeinfo uses the input file name to determine the output name: first, any of the extensions .texi, .tex, .txi or .texinfo is removed from the input file name; then, the output format specific extension is added—.html when generating HTML, .info when generating Info, etc. The \input line is still ignored in this processing, as well as leading blank lines.

When producing another output format, makeinfo will replace any final extension with the output format-specific extension (‘html’ when generating HTML, for example), or add a dot followed by the extension (‘.html’ for HTML) if the given name has no extension.

@setfilename used to be required by the Texinfo processors, and some other programs may still expect it to be present; for example, Automake (see Texinfo in GNU Automake).

Although an explicit ‘.info’ extension is preferable, some operating systems cannot handle long file names. You can run into a problem even when the file name you specify is itself short enough. This occurs because the Info formatters split a long Info file into short indirect subfiles, and name them by appending ‘-1’, ‘-2’, …, ‘-10’, ‘-11’, and so on, to the original file name. (See Tag Files and Split Files.) The subfile name texinfo.info-10, for example, is too long for old systems with a 14-character limit on filenames; so the Info file name for this document is texinfo rather than texinfo.info. When makeinfo is running on operating systems such as MS-DOS which impose severe limits on file names, it may remove some characters from the original file name to leave enough space for the subfile suffix, thus producing files named texin-10, gcc.i12, etc.

See also the --output option in Invoking texi2any/makeinfo from a Shell.


2.5.4 @settitle: Set the Document Title

A Texinfo file should contain a line that looks like this:

@settitle title

Write the @settitle command at the beginning of a line and follow it on the same line by the title. Do not write anything else on the line. The @settitle command should precede everything that generates actual output. The best place for it is right after the @setfilename command (described in the previous section).

This command tells TeX the title to use in a header or footer for double-sided output, in case such headings are output. For more on headings for TeX, see Heading Generation.

In the HTML file produced by makeinfo, title serves as the document ‘<title>’. It also becomes the default document description in the ‘<head>’ part.

When the title page is used in the output, the title in the @settitle command does not affect the title as it appears on the title page. Thus, the two do not need not to match exactly. A practice we recommend is to include the version or edition number of the manual in the @settitle title; on the title page, the version number generally appears as a @subtitle so it would be omitted from the @title. See @titlepage.


2.5.5 End of Header

Follow the header lines with an end-of-header line, which is a Texinfo comment that looks like this:

@c %**end of header

See Start of Header.


2.6 Document Permissions

This segment describes the document and contains the copyright notice and copying permissions. This is done with the @copying command. A real manual includes more text here, according to the license under which it is distributed.

@copying
This is a short example of a complete Texinfo file, version 1.0.

Copyright @copyright{} 2016 Free Software Foundation, Inc.
@end copying

The copyright notice and copying permissions for a document need to appear in several places in the various Texinfo output formats. Therefore, Texinfo provides a command (@copying) to declare this text once, and another command (@insertcopying) to insert the text at appropriate points.

If the document is a software manual, the software is typically under a different license—for GNU and many other free software packages, software is usually released under the GNU GPL, and manuals are released under the GNU FDL. It is helpful to state the license of the software of the manual, but giving the complete text of the software license is not necessarily required.


2.6.1 @copying: Declare Copying Permissions

The @copying command should be given very early in the document; the recommended location is right after the header material (see Texinfo File Header). It conventionally consists of a sentence or two about what the program is, identification of the documentation itself, the legal copyright line, and the copying permissions. Here is a skeletal example:

@copying
This manual is for program (version version, updated
date), which …

Copyright @copyright{} years copyright-owner.

@quotation
Permission is granted to …
@end quotation
@end copying

The @quotation has no legal significance; it’s there to improve readability in some contexts.

The text of @copying is output as a comment at the beginning of Info, HTML, XML, and DocBook output files. It is not output implicitly in plain text or TeX; it’s up to you to use @insertcopying to emit the copying information. See the next section for details.

The @copyright{} command generates a ‘c’ inside a circle when the output format supports this glyph (print and HTML always do, for instance). When the glyph is not supported in the output, it generates the three-character sequence ‘(C)’.

The copyright notice itself has the following legally-prescribed form:

Copyright © years copyright-owner.

The word ‘Copyright’ must always be written in English, even if the document is otherwise written in another language. This is due to international law.

The list of years should include all years in which a version was completed (even if it was released in a subsequent year). It is simplest for each year to be written out individually and in full, separated by commas.

The copyright owner (or owners) is whoever holds legal copyright on the work. In the case of works assigned to the FSF, the owner is ‘Free Software Foundation, Inc.’.

The copyright ‘line’ may actually be split across multiple lines, both in the source document and in the output. This often happens for documents with a long history, having many different years of publication. If you do use several lines, do not indent any of them (or anything else in the @copying block) in the source file.

See Copyright Notices in GNU Maintainer Information, for additional information. See GNU Sample Texts, for the full text to be used in GNU manuals. See GNU Free Documentation License, for the license itself under which GNU and other free manuals are distributed.


2.6.2 @insertcopying: Include Permissions Text

The @insertcopying command is simply written on a line by itself, like this:

@insertcopying

This inserts the text previously defined by @copying. To meet legal requirements, it must be used on the copyright page in the printed manual (see Copyright Page).

The @copying command itself causes the permissions text to appear in an Info file before the first node. The text is also copied into the beginning of each split Info output file, as is legally necessary. This location implies a human reading the manual using Info does not see this text (except when using the advanced Info command g *), but this does not matter for legal purposes, because the text is present.

Similarly, the @copying text is automatically included at the beginning of each HTML output file, as an HTML comment. Again, this text is not visible (unless the reader views the HTML source).

The permissions text defined by @copying also appears automatically at the beginning of the XML and DocBook output files.


2.8 Generating a Table of Contents

The @chapter, @section, and other structuring commands (see Chapter Structuring) supply the information to make up a table of contents, but they do not cause an actual table to appear in the manual. To do this, you must use the @contents and/or @summarycontents command(s).

@contents

Generates a table of contents in a printed manual, including all chapters, sections, subsections, etc., as well as appendices and unnumbered chapters. Headings generated by @majorheading, @chapheading, and the other @…heading commands do not appear in the table of contents (see Structuring Command Types).

@shortcontents
@summarycontents

(@summarycontents is a synonym for @shortcontents.)

Generates a short or summary table of contents that lists only the chapters, appendices, and unnumbered chapters. Sections, subsections and subsubsections are omitted. Only a long manual needs a short table of contents in addition to the full table of contents.

Both contents commands should be written on a line by themselves, and placed near the beginning of the file, after the @end titlepage (see @titlepage), before any sectioning command. The contents commands automatically generate a chapter-like heading at the top of the first table of contents page, so don’t include any sectioning command such as @unnumbered before them.

Since an Info file uses menus instead of tables of contents, the Info formatting commands ignore the contents commands. But the contents are included in plain text output (generated by makeinfo --plaintext) and in other output formats, such as HTML.

When makeinfo writes a short table of contents while producing HTML output, the links in the short table of contents point to corresponding entries in the full table of contents rather than the text of the document. The links in the full table of contents point to the main text of the document.


2.9 The ‘Top’ Node and Master Menu

The ‘Top’ node is the node in which a reader enters an Info manual. As such, it should contain a very brief description of the manual (including the version number). The contents of the ‘Top’ node do not appear in printed output.

It is conventional to write a @top sectioning command line containing the title of the document immediately after the @node Top line (see The @top Sectioning Command). This command helps makeinfo determine the relationships between nodes.

We repeat the short description from the beginning of the ‘@copying’ text, but there’s no need to repeat the copyright information, so we don’t use ‘@insertcopying’ here.

The ‘Top’ node contains a top-level menu listing the chapters, and possibly a detailed menu listing all the nodes in the entire document.

@node Top
@top Short Sample

This is a short sample Texinfo file.

@menu
* First Chapter::    The first chapter is the
                       only chapter in this sample.
* Index::            Complete index.
@end menu

2.9.1 Parts of a Master Menu

A master menu is the main menu. It is customary to include a detailed menu listing all the nodes in the document in this menu. Like any other menu, a master menu is enclosed in @menu and @end menu and does not appear in the printed output.

The master menu contains entries for the major nodes in the Texinfo file: the nodes for the chapters, chapter-like sections, and the appendices, followed by nodes for the indices.

You may choose to follow these entries with a defailed menu. This lists other, lower-level nodes, often ordered by chapter. These items may be a convenience for an inquirer who can go directly to a particular node when searching for specific information, rather than going through an intermediate menu. If you use a detailed menu in your master menu, mark it with the @detailmenu … @end detailmenu environment, or makeinfo will get confused.

Each section in the menu can be introduced by a descriptive line. So long as the line does not begin with an asterisk, it will not be treated as a menu entry. (See Writing a Menu, for more information.)

For example, the master menu for this manual looks like the following (but has many more entries):

@menu
* Copying Conditions::  Your rights.
* Overview::            Texinfo in brief.
…
* Command and Variable Index::
* General Index::

@detailmenu
--- The Detailed Node Listing ---

Overview of Texinfo

* Reporting Bugs:: …
…

Beginning a Texinfo File

* Sample Beginning:: …
…
@end detailmenu
@end menu

2.10 The Body of the Document

The body segment contains all the text of the document. A manual is divided into one or more nodes (see Nodes). The example illustrates a chapter made of three nodes, one for introductory material in the chapter, and two sections. The introductory material contains an enumerated list.

@node First Chapter
@chapter First Chapter

@cindex chapter, first
This is the first chapter.
@cindex index entry, another

Here is a numbered list.

@enumerate
@item
This is the first item.

@item
This is the second item.
@end enumerate


@node First Section
@section First Section

First section of first chapter.


@node Second Section
@section Second Section

Second section of first chapter.

In the Info output, the ‘First Chapter’ node will contain a menu listing the two sections in the chapter. Similarly, when this node is output in its own HTML file, it will contain a table of contents for the chapter.

Here is what the contents of this chapter will look like:


1. First Chapter

This is the first chapter.

Here is a numbered list.

  1. This is the first item.
  2. This is the second item.

1.1 First Section

First section of first chapter.

1.2 Second Section

Second section of first chapter.

(In the Info and HTML output, the chapter would also be split into nodes.)


2.11 Ending a Texinfo File

The end of a Texinfo file should include commands to create indices (see Printing Indices and Menus), and the @bye command to mark the last line to be processed. For example:

@node Index
@unnumbered Index

@printindex cp

@bye

A @bye command terminates Texinfo processing. It should be on a line by itself. Anything following @bye is completely ignored.


3 Nodes

A node is a region of text that begins at a @node command, and continues until the next @node command. To specify a node, write a @node command at the beginning of a line, and follow it with the name of the node. Each node contains the discussion of one topic. Info readers display one node at a time, and provide commands for the user to move to related nodes. The HTML output can be similarly navigated.

Nodes are used as the targets of cross-references. Cross-references, such as the one at the end of this sentence, are made with @xref and related commands; see Cross-references. Cross-references can be sprinkled throughout the text, and provide a way to represent links that do not fit a hierarchical structure.

Normally, you put a node command immediately before each chapter structuring command—for example, an @section or @subsection line. (See Chapter Structuring.) You must do this even if you do not intend to format the file for Info. This is because TeX uses both @node names and chapter-structuring names in the output for cross-references. The only time you are likely to use the chapter structuring commands without also using nodes is if you are writing a document that contains no cross references and will only be printed, not transformed into Info, HTML, or other formats.


3.1 Writing a @node Line

Write @node at the beginning of a line followed by the name of the node, like this:

@node node-name

After you have inserted a @node line, you should immediately write an @-command for the chapter or section and insert its name.

You may optionally follow the node name argument to @node with up to three optional arguments on the rest of the same line, separating the arguments with commas. These are the names of the ‘Next’, ‘Previous’, and ‘Up’ pointers, in that order. Hence, the template for a fully-written-out node line with ‘Next’, ‘Previous’, and ‘Up’ pointers looks like this:

@node node-name, next, previous, up

The node-name argument must be present, but the others are optional. If you wish to specify some but not others, just insert commas as needed, as in: ‘@node mynode,,,uppernode’. Any spaces before or after each name on the @node line are ignored. However, if your Texinfo document is hierarchically organized, as virtually all are, we recommend leaving off all the pointers and letting makeinfo determine them.

The makeinfo program automatically determines node pointers for a hierarchically organized document. For it do to do this, each @node command should be followed immediately by a sectioning command such as @chapter or @section (except that comment lines may intervene). Finally, you must follow the ‘Top’ @node line with a line beginning with @top to mark the top-level node in the file. See The @top Sectioning Command.

Even when you explicitly specify all pointers, you cannot write the nodes in the Texinfo source file in an arbitrary order. You must write the nodes in the order you wish them to appear in the output. For Info format one can imagine that the order may not matter, but it matters for the other formats.

In most cases, you will want to take advantage of the pointer creation feature, and not redundantly specify node pointers that the programs can determine. However, Texinfo documents are not required to be organized hierarchically or in fact to contain sectioning commands at all (for example, if you never intend the document to be printed), so node pointers may still be specified explicitly, in full generality.

If you are using GNU Emacs, you can use the update node commands provided by Texinfo mode to insert the names of the pointers. (See Updating Nodes and Menus.)

Alternatively, you can insert the ‘Next’, ‘Previous’, and ‘Up’ pointers yourself. If you do this, you may find it helpful to use the Texinfo mode keyboard command C-c C-c n. This command inserts ‘@node’ and a comment line listing the names of the pointers in their proper order. The comment line helps you keep track of which arguments are for which pointers.

If you wish, you can ignore @node lines altogether in your first draft and then use the texinfo-insert-node-lines command to create @node lines for you. However, we do not recommend this practice. It is better to name the node itself at the same time that you write a segment so you can easily make cross-references. Useful cross-references are an especially important feature of a good Texinfo manual.


3.2 Choosing Node Names

The name of a node identifies the node. For all the details of node names, see @node Line Requirements).

Here are some suggestions for node names:

  • Try to pick node names that are informative but short.

    In the Info file, the file name, node name, and pointer names are all inserted on one line, which may run into the right edge of the window. (This does not cause a problem with Info, but is ugly.)

  • Try to pick node names that differ from each other near the beginnings of their names. This way, it is easy to use automatic name completion in Info.
  • Conventionally, node names are capitalized in the same way as section and chapter titles. In this manual, initial and significant words are capitalized; others are not. In other manuals, just initial words and proper nouns are capitalized. Either way is fine; we recommend just being consistent.
  • In HTML output, any characters in the node name other than plain ASCII letters, numbers or spaces will be changed in the file name. (See HTML Cross-reference Node Name Expansion.) This can make the URL’s for the pages in your manual less user-friendly; for example, in this manual the ‘@dots’ node is output as __0040dots.html.

Because node names are used in cross-references, it is not desirable to casually change them once published. Such name changes invalidate references from other manuals, from mail archives, and so on.

The pointers from a given node enable you to reach other nodes and consist simply of the names of those nodes.

Normally, a node’s ‘Up’ pointer contains the name of the node whose menu mentions that node. The node’s ‘Next’ pointer contains the name of the node that follows the present node in that menu and its ‘Previous’ pointer contains the name of the node that precedes it in that menu. When a node’s ‘Previous’ node is the same as its ‘Up’ node, both pointers name the same node.

Usually, the first node of a Texinfo file is the ‘Top’ node, and its ‘Up’ pointer points to the dir file, which contains the main menu for all of Info.


Next: , Previous: , Up: Nodes   [Contents][Index]

3.3 @node Line Requirements

Names used with @node have several requirements:

  • All the node names in a single Texinfo file must be unique.

    This means, for example, that if you end every chapter with a summary, you must name each summary node differently. You cannot just call them all “Summary”. You may, however, duplicate the titles of chapters, sections, and the like. Thus you can end each chapter with a section called “Summary”, so long as the node names for those sections are all different.

  • Node names can contain @-commands. The output is generally the natural result of the command; for example, using @TeX{} in a node name results in the TeX logo being output, as it would be in normal text. Cross-references should use @TeX{} just as the node name does.

    For Info and HTML output, especially, it is necessary to expand commands to some sequence of plain characters; for instance, @TeX{} expands to the three letters ‘TeX’ in the Info node name. However, cross-references to the node should not take the “shortcut” of using ‘TeX’; stick to the actual node name, commands and all.

    Some commands do not make sense in node names; for instance, environments (e.g., @quotation), commands that read a whole line as their argument (e.g., @sp), and plenty of others.

    For the complete list of commands that are allowed, and their expansion for HTML identifiers and file names, see HTML Cross-reference Command Expansion. The expansions for Info are generally given with main the description of the command.

    Prior to the Texinfo 5 release in 2013, this feature was supported in an ad hoc way (the --commands-in-node-names option to makeinfo). Now it is part of the language.

  • Unfortunately, you cannot reliably use periods, commas, or colons within a node name; these can confuse the Info reader. Also, a node name may not start with a left parenthesis preceding a right parenthesis, as in (not)allowed, since this syntax is used to specify an external manual. (Perhaps these limitations will be removed some day.)

    makeinfo warns about such problematic usage in node names, menu items, and cross-references. If you don’t want to see the warnings, you can set the customization variable INFO_SPECIAL_CHARS_WARNING to ‘0’ (see Other Customization Variables).

    Also, if you insist on using these characters in node names (accepting the resulting substandard Info output), in order not to confuse the Texinfo processors you must still escape those characters, by using either special insertions (see Inserting ‘,’ with @comma{}) or @asis (see @asis). For example:

    @node foo@asis{::}bar
    

    As an example of avoiding the special characters, the following is a section title in this manual:

    @section @code{@@unnumbered}, @code{@@appendix}: Chapters with...
    

    But the corresponding node name lacks the commas and the subtitle:

    @node @code{@@unnumbered @@appendix}
    
  • Case is significant in node names.
  • Spaces before and after names on the ‘@node’ line are ignored. Multiple whitespace characters “inside” a name are collapsed to a single space. For example:
    @node foo bar
    @node  foo bar,
    @node foo bar ,
    @node foo  bar,
    @node  foo  bar ,
    

    all define the same node, namely ‘foo bar’. In menu entries, this is the name that should be used: no leading or trailing spaces, and a single internal space. (For cross-references, the node name used in the Texinfo sources is automatically normalized in this way.)


3.4 The First Node

The first node of a Texinfo file is the Top node, except in an included file (see Include Files). The Top node should contain a short summary and a master menu. See The ‘Top’ Node and Master Menu for more information on the Top node contents and examples. Straight text before the Top node outside of any node should be avoided.

Here is a description of the node pointers to be used in the Top node:

  • The Top node (which must be named ‘top’ or ‘Top’) should have as its ‘Up’ node the name of a node in another file, where there is a menu that leads to this file. Specify the file name in parentheses.

    Usually, all Info files are available through a single virtual Info tree, constructed from multiple directories. In this case, use ‘(dir)’ as the parent of the Top node; this specifies the top-level node in the dir file, which contains the main menu for the Info system as a whole. (Each directory with Info files is intended to contain a file named dir.)

    That’s fine for Info, but for HTML output, one might well want the Up link from the Top node to go to some specific place. For example, for GNU the natural place would be http://www.gnu.org/manual/ (a web page collecting links to most GNU manuals), better specified as just /manual/ if the manual will be installed on www.gnu.org. This can be specified with the TOP_NODE_UP_URL customization variable (see HTML Customization Variables), as in

    $ makeinfo --html -c TOP_NODE_UP_URL=/manual/ ...
    
  • The ‘Prev’ node of the Top node is usually either omitted or also set to (dir). Either is fine.
  • The ‘Next’ node of the Top node should be the first chapter in your document.

See Installing an Info File, for more information about installing an Info file in the info directory.

It is usually best to leave the pointers off entirely and let the tools implicitly define them, with this simple result:

@node Top

3.5 The @top Sectioning Command

The @top command is a special sectioning command that you should only use after a ‘@node Top’ line at the beginning of a Texinfo file. The @top command tells the makeinfo formatter which node is to be used as the root of the node tree.

It produces the same sort of output as @unnumbered (see @unnumbered, @appendix: Chapters with Other Labeling).

@top is ignored when raising or lowering sections. That is, it is never lowered and nothing can be raised to it (see Raise/lower Sections: @raisesections and @lowersections).

It used to be conventionial to wrap the ‘Top’ node in an @ifnottex conditional so that it would not appear in TeX output (see Conditionally Visible Text). Thus, a Top node often looked like this:

@ifnottex
@node Top
@top your-manual-title

very-high-level-summary
@end ifnottex

This is no longer necessary, as the ‘Top’ node is now never output for TeX output.


3.6 Texinfo Document Structure

Nodes can contain menus, which contain the names of child nodes within the parent node; for example, a node corresponding to a chapter would have a menu of the sections in that chapter. The menus allow the user to move to the child nodes in the Info output.

In addition, nodes contain node pointers that name other nodes. The ‘Next’ and ‘Previous’ pointers link nodes at the same sectioning level into a chain. As you might imagine, the ‘Next’ pointer links to the next node, and the ‘Previous’ pointer links to the previous node. Thus, for example, all the nodes that are at the level of sections within a chapter are linked together, and the order in this chain is the same as the order of the children in the menu of the parent chapter. Each child node records the parent node name as its ‘Up’ pointer.

The Info and HTML output from makeinfo for each node includes links to the ‘Next’, ‘Previous’, and ‘Up’ nodes. The HTML also uses the accesskey attribute with the values ‘n’, ‘p’, and ‘u’ respectively. This allows people using web browsers to follow the navigation using (typically) M-letter, e.g., M-n for the ‘Next’ node, from anywhere within the node. Node pointers and menus provide structure for Info files just as chapters, sections, subsections, and the like provide structure for printed books. The two structures are theoretically distinct; in practice, however, the tree structure of printed books is essentially always used for the node and menu structure also, as this leads to a document which is easiest to follow.

Typically, the sectioning structure and the node structure are completely parallel, with one node for each chapter, section, etc., and with the nodes following the same hierarchical arrangement as the sectioning. Thus, if a node is at the logical level of a chapter, its child nodes are at the level of sections; similarly, the child nodes of sections are at the level of subsections.

Although it is technically possible to create Texinfo documents with only one structure or the other, or for the two structures not to be parallel, or for either the sectioning or node structure to be abnormally formed, etc., this is not at all recommended. To the best of our knowledge, all the Texinfo manuals currently in general use do follow the conventional parallel structure.


Next: , Previous: , Up: Nodes   [Contents][Index]

3.7 Node and Menu Illustration

Here is a diagram that illustrates a Texinfo file with three chapters, each of which contains two sections.

The “root” is at the top of the diagram and the “leaves” are at the bottom. This is how such a diagram is drawn conventionally; it illustrates an upside-down tree. For this reason, the root node is called the ‘Top’ node, and ‘Up’ node pointers carry you closer to the root.

                         Top
                          |
        -------------------------------------
       |                  |                  |
    Chapter 1          Chapter 2          Chapter 3
       |                  |                  |
    --------           --------           --------
   |        |         |        |         |        |
Section  Section   Section  Section   Section  Section
  1.1      1.2       2.1      2.2       3.1      3.2

Using explicit pointers (not recommended, but shown for purposes of the example), the fully-written command to start Chapter 2 would be this:

@node     Chapter 2,  Chapter 3, Chapter 1, Top
@comment  node-name,  next,      previous,  up

This @node line says that the name of this node is “Chapter 2”, the name of the ‘Next’ node is “Chapter 3”, the name of the ‘Previous’ node is “Chapter 1”, and the name of the ‘Up’ node is “Top”. You can (and should) omit writing out these node names if your document is hierarchically organized , but the pointer relationships still obtain.

Note: ‘Next’ and ‘Previous’ refer to nodes at the same hierarchical level in the manual, not necessarily to the next node within the Texinfo file. In the Texinfo file, the subsequent node may be at a lower level—a section-level node most often follows a chapter-level node, for example. (The ‘Top’ node contains the exception to this rule. Since the ‘Top’ node is the only node at that level, ‘Next’ refers to the first following node, which is almost always a chapter or chapter-level node.)

To go to Sections 2.1 and 2.2 using Info, you need a menu inside Chapter 2. (See Menus.) You would write the menu just before the beginning of Section 2.1, like this:

   @menu
   * Sect. 2.1::    Description of this section.
   * Sect. 2.2::    Description.
   @end menu

Using explicit pointers, the node for Sect. 2.1 is written like this:

@node     Sect. 2.1, Sect. 2.2, Chapter 2, Chapter 2
@comment  node-name, next,      previous,  up

In Info format, the ‘Next’ and ‘Previous’ pointers of a node usually lead to other nodes at the same level—from chapter to chapter or from section to section (sometimes, as shown, the ‘Previous’ pointer points up); an ‘Up’ pointer usually leads to a node at the level above (closer to the ‘Top’ node); and a ‘Menu’ leads to nodes at a level below (closer to ‘leaves’). (A cross-reference can point to a node at any level; see Cross-references.)

A @node command and a chapter structuring command are conventionally used together, in that order, often followed by indexing commands. (As shown in the example above, you may follow the @node line with a comment line, e.g., to show which pointer is which if explicit pointers are used.) The Texinfo processors use this construct to determine the relationships between nodes and sectioning commands.

Here is the beginning of the chapter in this manual called “Ending a Texinfo File”. This shows a @node line followed by a @chapter line, and then by indexing lines.

@node Ending a File
@chapter Ending a Texinfo File
@cindex Ending a Texinfo file
@cindex Texinfo file ending
@cindex File ending

Next: , Previous: , Up: Texinfo   [Contents][Index]

4 Chapter Structuring

Texinfo’s chapter structuring commands divide a document into a hierarchy of chapters, sections, subsections, and subsubsections. These commands generate large headings in the text, like the one above. They also provide information for generating the table of contents (see Generating a Table of Contents).

Normally you put a @node command immediately before each chapter structuring command. See Nodes.


4.1 Tree Structure of Sections

A Texinfo file is usually structured like a book with chapters, sections, subsections, and the like. This structure can be visualized as a tree (or rather as an upside-down tree) with the root at the top and the levels corresponding to chapters, sections, subsection, and subsubsections.

Here is a diagram that shows a Texinfo file with three chapters, each with two sections.

                         Top
                          |
        -------------------------------------
       |                  |                  |
    Chapter 1          Chapter 2          Chapter 3
       |                  |                  |
    --------           --------           --------
   |        |         |        |         |        |
Section  Section   Section  Section   Section  Section
  1.1      1.2       2.1      2.2       3.1      3.2

In a Texinfo file that has this structure, the beginning of Chapter 2 would be written like this:

@node    Chapter 2
@chapter Chapter 2

For purposes of example, here is how it would be written with explicit node pointers:

@node    Chapter 2,  Chapter 3, Chapter 1, Top
@chapter Chapter 2

The chapter structuring commands are described in the sections that follow; the @node command is described in the previous chapter (see Nodes).


4.2 Structuring Command Types

The chapter structuring commands fall into four groups or series, each of which contains structuring commands corresponding to the hierarchical levels of chapters, sections, subsections, and subsubsections.

The four groups of commands are the @chapter series, the @unnumbered series, the @appendix series, and the @heading series. Each command produces a title with a different appearance in the body of the document. Some of the commands list their titles in the tables of contents, while others do not. Here are the details:

  • The @chapter and @appendix series of commands produce numbered or lettered entries both in the body of a document and in its table of contents.
  • The @unnumbered series of commands produce unnumbered entries both in the body of a document and in its table of contents. The @top command, which has a special use, is a member of this series (see The @top Sectioning Command). An @unnumbered section is a normal part of the document structure.
  • The @heading series of commands produce simple unnumbered headings that do not appear in a table of contents, are not associated with nodes, and cannot be cross-referenced. These heading commands never start a new page.

When a @setchapternewpage command says to do so, the @chapter, @unnumbered, and @appendix commands start new pages in the printed manual; the @heading commands do not. See @setchapternewpage: Blank Pages Before Chapters.

Here is a summary:

No new page
NumberedUnnumberedLettered/numberedUnnumbered
In contentsIn contentsIn contentsNot in contents
@top@majorheading
@chapter@unnumbered@appendix@chapheading
@section@unnumberedsec@appendixsec@heading
@subsection@unnumberedsubsec@appendixsubsec@subheading
@subsubsection@unnumberedsubsubsec@appendixsubsubsec@subsubheading

4.3 @chapter: Chapter Structuring

@chapter identifies a chapter in the document–the highest level of the normal document structuring hierarchy. Write the command at the beginning of a line and follow it on the same line by the title of the chapter. The chapter is numbered automatically, starting from 1.

For example, the present chapter in this manual is entitled “@chapter: Chapter Structuring”; the @chapter line looks like this:

@chapter @code{@@chapter}: Chapter Structuring

In TeX, the @chapter command produces a chapter heading in the document.

In Info and plain text output, the @chapter command causes the title to appear on a line by itself, with a line of asterisks inserted underneath. So, the above example produces the following output:

5 Chapter Structuring
*********************

In HTML, the @chapter command produces an <h2>-level header by default (controlled by the CHAPTER_HEADER_LEVEL customization variable, see Other Customization Variables).

In the XML and DocBook output, a <chapter> element is produced that includes all the following sections, up to the next chapter.


4.4 @unnumbered, @appendix: Chapters with Other Labeling

Use the @unnumbered command to start a chapter-level element that appears without chapter numbers of any kind. Use the @appendix command to start an appendix that is labeled by letter (‘A’, ‘B’, …) instead of by number; appendices are also at the chapter level of structuring.

Write an @appendix or @unnumbered command at the beginning of a line and follow it on the same line by the title, just as with @chapter.

Texinfo also provides a command @centerchap, which is analogous to @unnumbered, but centers its argument in the printed and HTML outputs. This kind of stylistic choice is not usually offered by Texinfo. You are recommended not to use this command, as it may be removed in future releases of Texinfo.

With @unnumbered, if the name of the associated node is one of these English words (case-insensitive):

Acknowledgements  Colophon  Dedication  Preface

then the DocBook output uses corresponding special tags (<preface>, etc.) instead of the default <chapter>. The argument to @unnumbered itself can be anything, and is output as the following <title> text as usual.


4.5 @majorheading, @chapheading: Chapter-level Headings

The @majorheading and @chapheading commands produce chapter-like headings in the body of a document.

However, neither command produces an entry in the table of contents, and neither command causes TeX to start a new page in a printed manual.

In TeX, a @majorheading command generates a larger vertical whitespace before the heading than a @chapheading command but is otherwise the same.

In Info and plain text, the @majorheading and @chapheading commands produce the same output as @chapter: the title is printed on a line by itself with a line of asterisks underneath. Similarly for HTML. The only difference is the lack of numbering and the lack of any association with nodes. See @chapter: Chapter Structuring.


4.6 @section: Sections Below Chapters

An @section command identifies a section within a chapter unit, whether created with @chapter, @unnumbered, or @appendix, following the numbering scheme of the chapter-level command. Thus, within a @chapter chapter numbered ‘1’, the sections are numbered ‘1.1’, ‘1.2’, etc.; within an @appendix “chapter” labeled ‘A’, the sections are numbered ‘A.1’, ‘A.2’, etc.; within an @unnumbered chapter, the section gets no number. The output is underlined with ‘=’ in Info and plain text.

To make a section, write the @section command at the beginning of a line and follow it on the same line by the section title. For example,

@section This is a section

might produce the following in Info:

5.7 This is a section
=====================

Section titles are listed in the table of contents.

The TeX, HTML, DocBook, and XML output is all analogous to the chapter-level output, just “one level down”; see @chapter: Chapter Structuring.


4.7 @unnumberedsec, @appendixsec, @heading

The @unnumberedsec, @appendixsec, and @heading commands are, respectively, the unnumbered, appendix-like, and heading-like equivalents of the @section command (see the previous section).

@unnumberedsec and @appendixsec do not need to be used in ordinary circumstances, because @section may also be used within @unnumbered and @appendix chapters; again, see the previous section.

@unnumberedsec

The @unnumberedsec command may be used within an unnumbered chapter or within a regular chapter or appendix to produce an unnumbered section.

@appendixsec
@appendixsection

@appendixsection is a longer spelling of the @appendixsec command; the two are synonymous.

Conventionally, the @appendixsec or @appendixsection command is used only within appendices.

@heading

You may use the @heading command (almost) anywhere for a section-style heading that will not appear in the table of contents. The @heading-series commands can appear inside most environments, for example, though pathological and useless locations such as inside @titlepage, as an argument to another command, etc., are not allowed.


4.8 @subsection: Subsections Below Sections

Subsections are to sections as sections are to chapters; see @section: Sections Below Chapters. In Info and plain text, subsection titles are underlined with ‘-’. For example,

@subsection This is a subsection

might produce

1.2.3 This is a subsection
--------------------------

Subsection titles are listed in the table of contents.

The TeX, HTML, DocBook, and XML output is all analogous to the chapter-level output, just “two levels down”; see @chapter: Chapter Structuring.


4.9 The @subsection-like Commands

The @unnumberedsubsec, @appendixsubsec, and @subheading commands are, respectively, the unnumbered, appendix-like, and heading-like equivalents of the @subsection command. (See @subsection: Subsections Below Sections.)

@unnumberedsubsec and @appendixsubsec do not need to be used in ordinary circumstances, because @subsection may also be used within sections of @unnumbered and @appendix chapters (see @section: Sections Below Chapters).

An @subheading command produces a heading like that of a subsection except that it is not numbered and does not appear in the table of contents. Similarly, an @unnumberedsubsec command produces an unnumbered heading like that of a subsection and an @appendixsubsec command produces a subsection-like heading labeled with a letter and numbers; both of these commands produce headings that appear in the table of contents. In Info and plain text, the @subsection-like commands generate a title underlined with hyphens.


4.10 @subsection and Other Subsub Commands

The fourth and lowest level sectioning commands in Texinfo are the ‘subsub’ commands. They are:

@subsubsection

Subsubsections are to subsections as subsections are to sections. (See @subsection: Subsections Below Sections.) Subsubsection titles appear in the table of contents.

@unnumberedsubsubsec

Unnumbered subsubsection titles appear in the table of contents, but lack numbers. Otherwise, unnumbered subsubsections are the same as subsubsections.

@appendixsubsubsec

Conventionally, appendix commands are used only for appendices and are lettered and numbered appropriately. They also appear in the table of contents.

@subsubheading

The @subsubheading command may be used anywhere that you want a small heading that will not appear in the table of contents.

As with subsections, @unnumberedsubsubsec and @appendixsubsubsec do not need to be used in ordinary circumstances, because @subsubsection may also be used within subsections of @unnumbered and @appendix chapters (see @section: Sections Below Chapters).

In Info, ‘subsub’ titles are underlined with periods. For example,

@subsubsection This is a subsubsection

might produce

1.2.3.4 This is a subsubsection
...............................

The TeX, HTML, DocBook, and XML output is all analogous to the chapter-level output, just “three levels down”; see @chapter: Chapter Structuring.


4.11 @part: Groups of Chapters

The final sectioning command is @part, to mark a part of a manual, that is, a group of chapters or (rarely) appendices. This behaves quite differently from the other sectioning commands, to fit with the way such “parts” are conventionally used in books.

No @node command is associated with @part. Just write the command on a line by itself, including the part title, at the place in the document you want to mark off as starting that part. For example:

@part Part I:@* The beginning

As can be inferred from this example, no automatic numbering or labeling of the @part text is done. The text is taken as-is.

Because parts are not associated with nodes, no general text can follow the @part line. To produce the intended output, it must be followed by a chapter-level command (including its node). Thus, to continue the example:

@part Part I:@* The beginning

@node Introduction
@chapter Introduction
...

In the TeX output, the @part text is included in both the normal and short tables of contents (see Generating a Table of Contents), without a page number (since that is the normal convention). In addition, a “part page” is output in the body of the document, with just the @part text. In the example above, the @* causes a line break on the part page (but is replaced with a space in the tables of contents). This part page is always forced to be on an odd (right-hand) page, regardless of the chapter pagination (see @setchapternewpage: Blank Pages Before Chapters).

In the HTML output, the @part text is similarly included in the tables of contents, and a heading is included in the main document text, as part of the following chapter or appendix node.

In the XML and DocBook output, the <part> element includes all the following chapters, up to the next <part>. A <part> containing chapters is also closed at an appendix.

In the Info and plain text output, @part has no effect.

@part is ignored when raising or lowering sections (see next section). That is, it is never lowered and nothing can be raised to it.


4.12 Raise/lower Sections: @raisesections and @lowersections

The @raisesections and @lowersections commands implicitly raise and lower the hierarchical level of following chapters, sections and the other sectioning commands (excluding parts).

That is, the @raisesections command changes sections to chapters, subsections to sections, and so on. Conversely, the @lowersections command changes chapters to sections, sections to subsections, and so on. Thus, a @lowersections command cancels a @raisesections command, and vice versa.

You can use @lowersections to include text written as an outer or standalone Texinfo file in another Texinfo file as an inner, included file (see Include Files). Typical usage looks like this:

@lowersections
@include somefile.texi
@raisesections

(Without the @raisesections, all the subsequent sections in the main file would also be lowered.)

If the included file being lowered has a @top node, you’ll need to conditionalize its inclusion with a flag (see @set and @value).

As a practical matter, you generally only want to raise or lower large chunks, usually in external files as shown above. The final result has to have menus that take the raising and lowering into account, so you cannot just arbitrarily sprinkle @raisesections and @lowersections commands throughout the document.

Repeated use of the commands continues to raise or lower the hierarchical level a step at a time. An attempt to raise above ‘chapter’ reproduces chapter commands; an attempt to lower below ‘subsubsection’ reproduces subsubsection commands. Also, lowered subsubsections and raised chapters will not work with makeinfo’s feature of implicitly determining node pointers, since the menu structure cannot be represented correctly.

Write each @raisesections and @lowersections command on a line of its own.


5 Cross-references

Cross-references are used to refer the reader to other parts of the same or different Texinfo files.


5.1 What References Are For

Often, but not always, a printed document should be designed so that it can be read sequentially. People tire of flipping back and forth to find information that should be presented to them as they need it.

However, in any document, some information will be too detailed for the current context, or incidental to it; use cross-references to provide access to such information. Also, an online help system or a reference manual is not like a novel; few read such documents in sequence from beginning to end. Instead, people look up what they need. For this reason, such creations should contain many cross-references to help readers find other information that they may not have read.

In a printed manual, a cross-reference results in a page reference, unless it is to another manual altogether, in which case the cross-reference names that manual. In Info, a cross-reference results in an entry that you can follow using the Info ‘f’ command. (See Following cross-references in Info.) In HTML, a cross-reference results in an hyperlink.

The various cross-reference commands use nodes (or anchors, see @anchor: Defining Arbitrary Cross-reference Targets) to define cross-reference locations. TeX needs nodes to define cross-reference locations. When TeX generates a DVI file, it records each node’s page number and uses the page numbers in making references. Thus, even if you are writing a manual that will only be printed, and not used online, you must nonetheless write @node lines in order to name the places to which you make cross-references.


5.2 Different Cross-reference Commands

There are three different cross-reference commands:

@xref

Used to start a sentence in the printed manual and in HTML with ‘See …’ or an Info cross-reference saying ‘*Note name: node.’.

@ref

Used within or, more often, at the end of a sentence; produces just the reference in the printed manual and in HTML without the preceding ‘See’ (same as @xref for Info).

@pxref

Used within parentheses, at the end of a sentence, or otherwise before punctuation, to make a reference. Its output starts with a lowercase ‘see’ in the printed manual and in HTML, and a lowercase ‘*note’ in Info. (‘p’ is for ‘parenthesis’.)

Additionally, there are commands to produce references to documents outside the Texinfo system. The @cite command is used to make references to books and manuals. @url produces a URL, for example a reference to a page on the World Wide Web.


5.3 Parts of a Cross-reference

A cross-reference command requires only one argument, which is the name of the node to which it refers. Here is a simple example:

@xref{Node name}.

In Info output, this produces

*Note Node name::.

In a printed manual, the output is

See Section nnn [Node name], page ppp.

A cross-reference command may contain up to four additional arguments. By using these arguments, you can provide a cross-reference name, a topic description or section title for the printed output, the name of a different manual file, and the name of a different printed manual. To refer to another manual as a whole, the manual file and/or the name of the printed manual are the only required arguments (see Referring to a Manual as a Whole).

Here is an example of a full five-part cross-reference:

@xref{Node name, Online Label, Printed Label,
info-file-name, A Printed Manual}, for details.

which produces

*Note Online Label: (info-file-name)Node name,
for details.

in Info and

See section “Printed Label” in A Printed Manual, for details.

in a printed book.

The five possible arguments for a cross-reference are:

  1. The node or anchor name (required, except for reference to whole manuals). This is the location to which the cross-reference takes you. In a printed document, the location of the node provides the page reference only for references within the same document. Use @node to define the node (see Writing a @node Line), or @anchor (see @anchor: Defining Arbitrary Cross-reference Targets).

    Write a node name in a cross-reference in exactly the same way as in the @node line, including the same capitalization; otherwise, the formatters may not find the reference.

  2. A label for online output. It is usually omitted; then the topic description (third argument) is used if it was specified; if that was omitted as well, the node name is used.
  3. A label for printed output. Often, this is the title or topic of the section. This is used as the name of the reference in the printed manual. If omitted, the node name is used.
  4. The name of the manual file in which the reference is located, if it is different from the current file. This name is used both for Info and HTML.
  5. The name of a printed manual from a different Texinfo file.

The template for a full five argument cross-reference looks like this:

@xref{node-name, online-label, printed-label,
info-file-name, printed-manual-title}

Whitespace before and after the commas separating these arguments is ignored. To include a comma in one of the arguments, use @comma{} (see Inserting ‘,’ with @comma{}).

When processing with TeX, a comma is automatically inserted after the page number for cross-references to within the same manual, unless the closing brace of the argument is followed by non-whitespace (such as a comma or period). This gives you the choice of whether to have a comma there in Info or HTML output. For example,

@xref{Another Section} for more information

produces ‘See Another Section, page ppp, for more information’ in the printed output, and ‘*Note Another Section:: for more information in the Info output.

If an unwanted comma is added, follow the argument with a command such as ‘@:’. For example, ‘@xref{Hurricanes}@: --- for the details produces

See Hurricanes, page ppp — for the details

instead of ‘See Hurricanes, page ppp, — for the details’.

Cross-references with one, two, three, four, and five arguments are described separately following the description of @xref.

makeinfo warns when the text of a cross-reference (and node names and menu items) contains a problematic construct that will interfere with its parsing in Info. If you don’t want to see the warnings, you can set the customization variable INFO_SPECIAL_CHARS_WARNING to ‘0’ (see Other Customization Variables).


5.4 @xref

The @xref command generates a cross-reference for the beginning of a sentence.


5.4.1 @xref with One Argument

The simplest form of @xref takes one argument, the name of another node in the same Texinfo file.

For example,

@xref{Tropical Storms}.

produces

*Note Tropical Storms::.

in Info and

See Section 3.1 [Tropical Storms], page 24.

in a printed manual.


5.4.2 @xref with Two Arguments

With two arguments, the second is used as a label for the online output.

The template is like this:

@xref{node-name, online-label}.

For example,

@xref{Electrical Effects, Lightning}.

produces:

*Note Lightning: Electrical Effects.

in Info and

See Section 5.2 [Electrical Effects], page 57.

in a printed manual, where the node name is printed.

The second argument to cross-references must observe some of the restrictions for node names (see @node Line Requirements). The most common issue is that colons cannot be used, since that interferes with the parsing of the Info file.


5.4.3 @xref with Three Arguments

A third argument replaces the node name in the TeX output. The third argument should be the name of the section in the printed output, or else state the topic discussed by that section.

The template is like this:

@xref{node-name, online-label, printed-label}.

For example,

@xref{Electrical Effects, Lightning, Thunder and Lightning},
for details.

produces

*Note Lightning: Electrical Effects, for details.

in Info and

See Section 5.2 [Thunder and Lightning], page 57, for details.

in a printed manual.

If a third argument is given and the second one is empty, then the third argument serves for both. (Note how two commas, side by side, mark the empty second argument.)

@xref{Electrical Effects, , Thunder and Lightning},
for details.

produces

*Note Thunder and Lightning: Electrical Effects, for details.

in Info and

See Section 5.2 [Thunder and Lightning], page 57, for details.

in a printed manual.

The third argument to cross-references must observe some of the restrictions for node names (see @node Line Requirements). The most common issue is that colons cannot be used, since that interferes with the parsing of the Info file.

As a practical matter, it is often best to write cross-references with just the first argument if the node name and the section title are the same (or nearly so), and with the first and third arguments only if the node name and title are different.

Texinfo offers a setting to use the section title instead of node names by default in cross-references (an explicitly specified third argument still takes precedence):

@xrefautomaticsectiontitle on

Typically this line would be given near the beginning of the document and used for the whole manual. But you can turn it off if you want (@xrefautomaticsectiontitle off), for example, if you’re including some other sub-document that doesn’t have suitable section names. This setting also applies to node headers in HTML, if @xrefautomaticsectiontitle is on, the sections names are used in node headers instead of the node names when possible.


5.4.4 @xref with Four and Five Arguments

In a cross-reference, a fourth argument specifies the name of another Info file, different from the file in which the reference appears, and a fifth argument specifies its title as a printed manual.

The full template is:

@xref{node-name, online-label, printed-label,
info-file-name, printed-manual-title}.

For example,

@xref{Electrical Effects, Lightning, Thunder and Lightning,
weather, An Introduction to Meteorology}.

produces this output in Info:

*Note Lightning: (weather)Electrical Effects.

As you can see, the name of the Info file is enclosed in parentheses and precedes the name of the node.

In a printed manual, the reference looks like this:

See section “Thunder and Lightning” in An Introduction to Meteorology.

The title of the printed manual is typeset like @cite; and the reference lacks a page number since TeX cannot know to which page a reference refers when that reference is to another manual.

Next case: often, you will leave out the second argument when you use the long version of @xref. In this case, the third argument, the topic description, will be used as the cross-reference name in Info. For example,

@xref{Electrical Effects, , Thunder and Lightning,
weather, An Introduction to Meteorology}.

produces

*Note Thunder and Lightning: (weather)Electrical Effects.

in Info and

See section “Thunder and Lightning” in An Introduction to Meteorology.

in a printed manual.

Next case: If the node name and the section title are the same in the other manual, you may also leave out the section title. In this case, the node name is used in both instances. For example,

@xref{Electrical Effects,,,
weather, An Introduction to Meteorology}.

produces

*Note (weather)Electrical Effects::.

in Info and

See section “Electrical Effects” in An Introduction to Meteorology.

in a printed manual.

A very unusual case: you may want to refer to another manual file that is within a single printed manual—when multiple Texinfo files are incorporated into the same TeX run but can create separate Info or HTML output files. In this case, you need to specify only the fourth argument, and not the fifth.

Finally, it’s also allowed to leave out all the arguments except the fourth and fifth, to refer to another manual as a whole. See the next section.


Next: , Previous: , Up: Cross-references   [Contents][Index]

5.5 Referring to a Manual as a Whole

Ordinarily, you must always name a node in a cross-reference. However, it’s not unusual to want to refer to another manual as a whole, rather than a particular section within it. In this case, giving any section name is an unnecessary distraction.

So, with cross-references to other manuals (see @xref with Four and Five Arguments), if the first argument is either ‘Top’ (capitalized just that way) or omitted entirely, and the third argument is omitted, the printed output includes no node or section name. (The Info output includes ‘Top’ if it was given.) For example,

@xref{Top,,, make, The GNU Make Manual}.

produces

*Note (make)Top::.

and

See The GNU Make Manual.

Info readers will go to the Top node of the manual whether or not the ‘Top’ node is explicitly specified.

It’s also possible (and is historical practice) to refer to a whole manual by specifying the ‘Top’ node and an appropriate entry for the third argument to the @xref command. Using this idiom, to make a cross-reference to The GNU Make Manual, you would write:

@xref{Top,, Overview, make, The GNU Make Manual}.

which produces

*Note Overview: (make)Top.

in Info and

See section “Overview” in The GNU Make Manual.

in a printed manual.

In this example, ‘Top’ is the name of the first node, and ‘Overview’ is the name of the first section of the manual. There is no widely-used convention for naming the first section in a printed manual, this is just what the Make manual happens to use. This arbitrariness of the first name is a principal reason why omitting the third argument in whole-manual cross-references is preferable.


5.6 @ref

@ref is nearly the same as @xref except that it does not generate a ‘See’ in the printed output, just the reference itself. This makes it useful as the last part of a sentence.

For example,

For more information, @pxref{This}, and @ref{That}.

produces in Info:

For more information, *note This::, and *note That::.

and in printed output:

For more information, see Section 1.1 [This], page 1, and Section 1.2 [That], page 2.

The @ref command can tempt writers to express themselves in a manner that is suitable for a printed manual but looks awkward in the Info format. Bear in mind that your audience could be using both the printed and the Info format. For example:

Sea surges are described in @ref{Hurricanes}.

looks ok in the printed output:

Sea surges are described in Section 6.7 [Hurricanes], page 72.

but is awkward to read in Info, “note” being a verb:

Sea surges are described in *note Hurricanes::.

5.7 @pxref

The parenthetical reference command, @pxref, is nearly the same as @xref, but it is best used within parentheses. The command differs from @xref in that TeX typesets the reference for the printed manual with a lowercase ‘see’ rather than an uppercase ‘See’.

With one argument, a parenthetical cross-reference looks like this:

… storms cause flooding (@pxref{Hurricanes}) …

which produces

… storms cause flooding (*note Hurricanes::) …

in Info and

… storms cause flooding (see Section 6.7 [Hurricanes], page 72) …

in a printed manual.

With two arguments, a parenthetical cross-reference has this template:

… (@pxref{node-name, cross-reference-name}) …

which produces

… (*note cross-reference-name: node-name.) …

in Info and

… (see Section nnn [node-name], page ppp) …

in a printed manual.

@pxref can be used with up to five arguments, just like @xref (see @xref).

In past versions of Texinfo, it was not allowed to write punctuation after a @pxref, so it could be used only before a right parenthesis. This is no longer the case. The effect of ‘@pxref{node-name}’ is similar to that of ‘see @ref{node-name}’. However, in many circumstance the latter is preferrable, as this makes it clear in the Info output that the word “see” should be present.


5.8 @anchor: Defining Arbitrary Cross-reference Targets

An anchor is a position in your document, labelled so that cross-references can refer to it, just as they can to nodes. You create an anchor with the @anchor command, and give the label as a normal brace-delimited argument. For example:

This marks the @anchor{x-spot}spot.
…
@xref{x-spot,,the spot}.

produces:

This marks the spot.
…
See [the spot], page 1.

As you can see, the @anchor command itself produces no output. This example defines an anchor ‘x-spot’ just before the word ‘spot’. You can refer to it later with an @xref or other cross reference command, as shown (see Cross-references).

It is best to put @anchor commands just before the position you wish to refer to; that way, the reader’s eye is led on to the correct text when they jump to the anchor. You can put the @anchor command on a line by itself if that helps readability of the source. Whitespace (including newlines) is ignored after @anchor.

Anchor names and node names may not conflict. Anchors and nodes are given similar treatment in some ways; for example, the goto-node command takes either an anchor name or a node name as an argument. (See Go to node in Info.)

Also like node names, anchor names cannot include some characters (see @node Line Requirements).

Because of this duality, when you delete or rename a node, it is usually a good idea to define an @anchor with the old name. That way, any links to the old node, whether from other Texinfo manuals or general web pages, keep working.


5.9 @inforef: Cross-references to Info-only Material

@inforef is used for making cross-references to Info documents—even from a printed manual. This was originally used for Info files that were not generated from any Texinfo source. The command is now obsolete and should not be used.

The command takes either two or three arguments, in the following order:

  1. The node name.
  2. The cross-reference name (optional).
  3. The Info file name.

The template is:

@inforef{node-name, cross-reference-name, info-file-name}

For example,

@inforef{Advanced, Advanced Info commands, info},
for more information.

produces (in Info):

*Note Advanced Info commands: (info)Advanced,
for more information.

and (in the printed output):

See Info file info, node ‘Advanced’, for more information.

(This particular example is not realistic, since the Info manual is written in Texinfo, so all formats are available. In fact, we don’t know of any extant Info-only manuals.)

The converse of @inforef is @cite, which is used to refer to printed works for which no Info form exists. See @cite{reference}.


5.10 @url, @uref{url[, text][, replacement]}

@url produces a reference to a uniform resource locator (url). It takes one mandatory argument, the url, and two optional arguments which control the text that is displayed. In HTML and PDF output, @url produces a link you can follow. (To merely indicate a url without creating a link people can follow, use @indicateurl, see @indicateurl{uniform-resource-locator}.)

@uref is a synonym for @url. (Originally, @url had the meaning of @indicateurl and @uref was required to produce a working link, but in practice @url was almost always misused. So we’ve changed the meaning.)

The second argument, if specified, is the text to display (the default is the url itself); in Info, DVI, and PDF output, but not in HTML output, the url is output in addition to this text.

The third argument, if specified, is the text to display, but in this case the url is not output in any format. This is useful when the text is already sufficiently referential, as in a man page. Also, if the third argument is given, the second argument is ignored.


5.10.1 @url Examples

First, here is an example of the simplest form of @url, with just one argument. The given url is both the target and the visible text of the link:

The official GNU ftp site is @url{http://ftp.gnu.org/gnu}.

produces:

The official GNU ftp site is http://ftp.gnu.org/gnu.

Two-argument form of @url

Here is an example of the two-argument form:

The official @url{http://ftp.gnu.org/gnu, GNU ftp site}
holds programs and texts.

which produces:

The official GNU ftp site
holds programs and texts.

that is, the Info (and TeX, etc.) output is this:

The official GNU ftp site (http://ftp.gnu.org/gnu)
holds programs and texts.

while the HTML output is this:

The official <a href="http://ftp.gnu.org/gnu">GNU ftp site</a>
holds programs and texts.

Three-argument form of @url

Finally, an example of the three-argument form:

The @url{/man.cgi/1/ls,,ls} program …

which, except for HTML, produces:

The ls program …

but with HTML:

The <a href="/man.cgi/1/ls">ls</a> program …

By the way, some people prefer to display urls in the unambiguous format:

<URL:http://host/path>

You can use this form in the input file if you wish. We feel it’s not necessary to include the ‘<URL:’ and ‘>’ in the output, since to be useful any software that tries to detect urls in text already has to detect them without the ‘<URL:’.


5.10.2 URL Line Breaking

TeX allows line breaking within urls at only a few characters (which are special in urls): ‘&’, ‘.’, ‘#’, ‘?’, and ‘/’ (but not between two ‘/’ characters). A tiny amount of stretchable space is also inserted around these characters to help with line breaking.

For HTML output, modern browsers will also do line breaking within displayed urls. If you need to allow breaks at other characters you can insert @/ as needed (see @* and @/: Generate and Allow Line Breaks).

By default, in TeX any such breaks at special characters will occur after the character. Some people prefer such breaks to happen before the special character. This can be controlled with the @urefbreakstyle command (this command has effect only in TeX):

@urefbreakstyle how

where the argument how is one of these words:

after

(the default) Potentially break after the special characters.

before

Potentially break before the special characters.

none

Do not consider breaking at the special characters at all; any potential breaks must be manually inserted.


5.10.3 @url PDF Output Format

If the ultimate purpose of a PDF is only to be viewed online, perhaps similar to HTML in some inchoate way, you may not want the urls to be included in the visible text (just as urls are not visible to readers of web pages). Texinfo provides a PDF-specific option for this, which must be used inside @tex:

@tex
\global\urefurlonlylinktrue
@end tex

The result is that @url{http://www.gnu.org, GNU} has the visible output of just ‘GNU’, with a link target of http://www.gnu.org. Ordinarily, the visible output would include both the label and the url: ‘GNU (http://www.gnu.org)’.

This option only has effect when the PDF output is produced with the pdfTeX program, not with other ways of getting from Texinfo to PDF (e.g., TeX to DVI to PDF). Consequently, it is ok to specify this option unconditionally within @tex, as shown above. It is ignored when DVI is being produced.


5.10.4 PDF Colors

By default, urls and cross-reference links are printed in black in PDF output. Very occasionally, however, you may want to highlight such “live” links with a different color, as is commonly done on web pages. Texinfo provides a PDF-specific option for specifying these colors, which must be used inside @tex:

@tex
\global\def\linkcolor{1 0 0}  % red
\global\def\urlcolor{0 1 0}   % green
@end tex

\urlcolor changes the color of @url output (both the actual url and any textual label), while \linkcolor changes the color for cross-references to nodes, etc. They are independent.

The three given values must be numbers between 0 and 1, specifying the amount of red, green, and blue respectively.

These definitions only have an effect when the PDF output is produced with the pdfTeX program, not with other ways of getting from Texinfo to PDF (e.g., TeX to DVI to PDF). Consequently, it is ok to specify this option unconditionally within @tex, as shown above. It is ignored when DVI is being produced.

We do not recommend colorizing just for fun; unless you have a specific reason to use colors, best to skip it.


5.11 @cite{reference}

Use the @cite command for the name of a book that lacks a companion Info file. The command produces italics in the printed manual, and quotation marks in the Info file.

If a book is written in Texinfo, it is better to use a cross-reference command since a reader can easily follow such a reference in Info. See @xref.


6 Marking Text, Words and Phrases

In Texinfo, you can mark words and phrases in a variety of ways. The Texinfo formatters use this information to determine how to highlight the text. You can specify, for example, whether a word or phrase is a defining occurrence, a metasyntactic variable, or a symbol used in a program. Also, you can emphasize text, in several different ways.


6.1 Indicating Definitions, Commands, etc.

Texinfo has commands for indicating just what kind of object a piece of text refers to. For example, email addresses are marked by @email; that way, the result can be a live link to send email when the output format supports it. If the email address was simply marked as “print in a typewriter font”, that would not be possible.


6.1.1 Highlighting Commands are Useful

The commands serve a variety of purposes:

@code{sample-code}

Indicate text that is a literal example of a piece of a program. See @code{sample-code}.

@kbd{keyboard-characters}

Indicate keyboard input. See @kbd{keyboard-characters}.

@key{key-name}

Indicate the conventional name for a key on a keyboard. See @key{key-name}.

@samp{text}

Indicate text that is a literal example of a sequence of characters. See @samp{text}.

@verb{text}

Write a verbatim sequence of characters. See @verb{chartextchar}.

@var{metasyntactic-variable}

Indicate a metasyntactic variable. See @var{metasyntactic-variable}.

@env{environment-variable}

Indicate an environment variable. See @env{environment-variable}.

@file{file-name}

Indicate the name of a file. See @file{file-name}.

@command{command-name}

Indicate the name of a command. See @command{command-name}.

@option{option}

Indicate a command-line option. See @option{option-name}.

@dfn{term}

Indicate the introductory or defining use of a term. See @dfn{term}.

@cite{reference}

Indicate the name of a book. See @cite{reference}.

@abbr{abbreviation}

Indicate an abbreviation, such as ‘Comput.’.

@acronym{acronym}

Indicate an acronym. See @acronym{acronym[, meaning]}.

@indicateurl{uniform-resource-locator}

Indicate an example (that is, nonfunctional) uniform resource locator. See @indicateurl{uniform-resource-locator}. (Use @url (see @url, @uref{url[, text][, replacement]}) for live urls.)

@email{email-address[, displayed-text]}

Indicate an electronic mail address. See @email{email-address[, displayed-text]}.


6.1.2 @code{sample-code}

Use the @code command to indicate text that is a piece of a program and which consists of entire syntactic tokens. Enclose the text in braces.

Thus, you should use @code for an expression in a program, for the name of a variable or function used in a program, or for a keyword in a programming language.

Use @code for command names in languages that resemble programming languages, such as Texinfo. For example, @code and @samp are produced by writing ‘@code{@@code}’ and ‘@code{@@samp}’ in the Texinfo source, respectively.

It is incorrect to alter the case of a word inside a @code command when it appears at the beginning of a sentence. Most computer languages are case sensitive. In C, for example, Printf is different from the identifier printf, and most likely is a misspelling of it. Even in languages which are not case sensitive, it is confusing to a human reader to see identifiers spelled in different ways. Pick one spelling and always use that. If you do not want to start a sentence with a command name written all in lowercase, you should rearrange the sentence.

In the Info output, @code results in single quotation marks around the text. In other formats, @code argument is typeset in a typewriter (monospace) font. For example,

The function returns @code{nil}.

produces this:

The function returns nil.

Here are some cases for which it is preferable not to use @code:

  • For shell command names, such as ls (use @command).
  • For environment variables, such as TEXINPUTS (use @env).
  • For shell options, such as ‘-c’, when such options stand alone (use @option).
  • An entire shell command often looks better if written using @samp rather than @code. In this case, the rule is to choose the more pleasing format.
  • For a string of characters shorter than a syntactic token. For example, if you are writing about ‘goto-ch’, which is just a part of the name for the goto-char Emacs Lisp function, you should use @samp.
  • In general, when writing about the characters used in a token; for example, do not use @code when you are explaining what letters or printable symbols can be used in the names of functions. (Use @samp.) Also, you should not use @code to mark text that is considered input to programs unless the input is written in a language that is like a programming language. For example, you should not use @code for the keystroke commands of GNU Emacs (use @kbd instead) although you may use @code for the names of the Emacs Lisp functions that the keystroke commands invoke.

By default, TeX will consider breaking lines at ‘-’ and ‘_’ characters within @code and related commands. This can be controlled with @allowcodebreaks (see @allowcodebreaks: Control Line Breaks in @code). The HTML output attempts to respect this for ‘-’, but ultimately it is up to the browser’s behavior. For Info, it seems better never to make such breaks.

For Info, the quotes are omitted in the output of the @code command and related commands (e.g., @kbd, @command), in typewriter-like contexts such as the @example environment (see @example: Example Text) and @code itself, etc.

To control which quoting characters are implicitly inserted by Texinfo processors in the output of ‘@code’, etc., see the OPEN_QUOTE_SYMBOL and CLOSE_QUOTE_SYMBOL customization variables (see Other Customization Variables). This is separate from how actual quotation characters in the input document are handled (see Inserting Quote Characters).


6.1.3 @kbd{keyboard-characters}

Use the @kbd command for characters of input to be typed by users. For example, to refer to the characters M-a, write:

@kbd{M-a}

and to refer to the characters M-x shell, write:

@kbd{M-x shell}

By default, the @kbd command produces a different font (slanted typewriter instead of normal typewriter), so users can distinguish the characters that they are supposed to type from those that the computer outputs.

Since the usage of @kbd varies from manual to manual, you can control the font switching with the @kbdinputstyle command. This command has no effect on Info output. Write this command at the beginning of a line with a single word as an argument, one of the following:

code

Always use the same font for @kbd as @code.

example

Use the distinguishing font for @kbd only in @example and similar environments.

distinct

(the default) Always use the distinguishing font for @kbd.

You can embed another @-command inside the braces of a @kbd command. Here, for example, is the way to describe a command that would be described more verbosely as “press the ‘r’ key and then press the RETURN key”:

@kbd{r @key{RET}}

This produces: r RET. (The present manual uses the default for @kbdinputstyle.)

You also use the @kbd command if you are spelling out the letters you type; for example:

To give the @code{logout} command,
type the characters @kbd{l o g o u t @key{RET}}.

This produces:

To give the logout command, type the characters l o g o u t RET.

(Also, this example shows that you can add spaces for clarity. If you explicitly want to mention a space character as one of the characters of input, write @key{SPC} for it.)


6.1.4 @key{key-name}

Use the @key command for the conventional name for a key on a keyboard, as in:

@key{RET}

You can use the @key command within the argument of an @kbd command when the sequence of characters to be typed includes one or more keys that are described by name.

For example, to produce C-x ESC and M-TAB you would type:

@kbd{C-x @key{ESC}}
@kbd{M-@key{TAB}}

Here is a list of the recommended names for keys:

SPC

Space

RET

Return

LFD

Linefeed (however, since most keyboards nowadays do not have a Linefeed key, it might be better to call this character C-j)

TAB

Tab

BS

Backspace

ESC

Escape

DELETE

Delete

SHIFT

Shift

CTRL

Control

META

Meta

There are subtleties to handling words like ‘meta’ or ‘ctrl’ that are names of modifier keys. When mentioning a character in which the modifier key is used, such as Meta-a, use the @kbd command alone; do not use the @key command; but when you are referring to the modifier key in isolation, use the @key command. For example, write ‘@kbd{Meta-a}’ to produce Meta-a and ‘@key{META}’ to produce META.

As a convention in GNU manuals, @key should not be used in index entries.


6.1.5 @samp{text}

Use the @samp command to indicate text that is a literal example or ‘sample’ of a sequence of characters in a file, string, pattern, etc. Enclose the text in braces. The argument appears within single quotation marks in both the Info file and the printed manual; in addition, it is printed in a fixed-width font.

To match @samp{foo} at the end of the line,
use the regexp @samp{foo$}.

produces

To match ‘foo’ at the end of the line, use the regexp ‘foo$’.

Any time you are referring to single characters, you should use @samp unless @kbd or @key is more appropriate. Also, you may use @samp for entire statements in C and for entire shell commands—in this case, @samp often looks better than @code. Basically, @samp is a catchall for whatever is not covered by @code, @kbd, @key, @command, etc.

Only include punctuation marks within braces if they are part of the string you are specifying. Write punctuation marks outside the braces if those punctuation marks are part of the English text that surrounds the string. In the following sentence, for example, the commas and period are outside of the braces:

In English, the vowels are @samp{a}, @samp{e},
@samp{i}, @samp{o}, @samp{u}, and sometimes
@samp{y}.

This produces:

In English, the vowels are ‘a’, ‘e’, ‘i’, ‘o’, ‘u’, and sometimes ‘y’.


6.1.6 @verb{chartextchar}

Use the @verb command to print a verbatim sequence of characters.

Like LaTeX’s \verb command, the verbatim text can be quoted using any unique delimiter character. Enclose the verbatim text, including the delimiters, in braces. Text is printed in a fixed-width font:

How many @verb{|@|}-escapes does one need to print this
@verb{.@a @b.@c.} string or @verb{+@'e?`{}!`\+} this?

produces

How many @-escapes does one need to print this
@a @b.@c string or @'e?`{}!`\ this?

This is in contrast to @samp (see the previous section), @code, and similar commands; in those cases, the argument is normal Texinfo text, where the three characters @{} are special, as usual. With @verb, nothing is special except the delimiter character you choose.

The delimiter character itself may appear inside the verbatim text, as shown above. As another example, ‘@verb{...}’ prints a single (fixed-width) period.

It is not reliable to use @verb inside other Texinfo constructs. In particular, it does not work to use @verb in anything related to cross-referencing, such as section titles or figure captions.


6.1.7 @var{metasyntactic-variable}

Use the @var command to indicate metasyntactic variables. A metasyntactic variable is something that stands for another piece of text. For example, you should use a metasyntactic variable in the documentation of a function to describe the arguments that are passed to that function.

Do not use @var for the names of normal variables in computer programs. These are specific names, so @code is correct for them (@code). For example, the Emacs Lisp variable texinfo-tex-command is not a metasyntactic variable; it is properly formatted using @code.

Do not use @var for environment variables either; @env is correct for them (see the next section).

The effect of @var in the Info file is to change the case of the argument to all uppercase. In the printed manual and HTML output, the argument is output in slanted type.

For example,

To delete file @var{filename},
type @samp{rm @var{filename}}.

produces

To delete file filename, type ‘rm filename’.

(Note that @var may appear inside @code, @samp, @file, etc.)

Write a metasyntactic variable all in lowercase without spaces, and use hyphens to make it more readable. Thus, the Texinfo source for the illustration of how to begin a Texinfo manual looks like this:

\input texinfo
@@settitle @var{name-of-manual}

This produces:

\input texinfo
@settitle name-of-manual

In some documentation styles, metasyntactic variables are shown with angle brackets, for example:

…, type rm <filename>

However, that is not the style that Texinfo uses.


6.1.8 @env{environment-variable}

Use the @env command to indicate environment variables, as used by many operating systems, including GNU. Do not use it for metasyntactic variables; use @var for those (see the previous section).

@env is equivalent to @code in its effects. For example:

The @env{PATH} environment variable …

produces

The PATH environment variable …


6.1.9 @file{file-name}

Use the @file command to indicate text that is the name of a file, buffer, or directory, or is the name of a node in Info. You can also use the command for file name suffixes. Do not use @file for symbols in a programming language; use @code.

@file is equivalent to code in its effects. For example,

The @file{.el} files are in
the @file{/usr/local/emacs/lisp} directory.

produces

The .el files are in the /usr/local/emacs/lisp directory.


6.1.10 @command{command-name}

Use the @command command to indicate command names, such as ls or cc.

@command is equivalent to @code in its effects. For example:

The command @command{ls} lists directory contents.

produces

The command ls lists directory contents.

You should write the name of a program in the ordinary text font, rather than using @command, if you regard it as a new English word, such as ‘Emacs’ or ‘Bison’.

When writing an entire shell command invocation, as in ‘ls -l’, you should use either @samp or @code at your discretion.


6.1.11 @option{option-name}

Use the @option command to indicate a command-line option; for example, -l or --version or --output=filename.

@option is equivalent to @code in its effects. For example:

The option @option{-l} produces a long listing.

produces

The option -l produces a long listing.


6.1.12 @dfn{term}

Use the @dfn command to identify the introductory or defining use of a technical term. Use the command only in passages whose purpose is to introduce a term which will be used again or which the reader ought to know. Mere passing mention of a term for the first time does not deserve @dfn. The command generates italics in the printed manual, and double quotation marks in the Info file. For example:

Getting rid of a file is called @dfn{deleting} it.

produces

Getting rid of a file is called deleting it.

As a general rule, a sentence containing the defining occurrence of a term should be a definition of the term. The sentence does not need to say explicitly that it is a definition, but it should contain the information of a definition—it should make the meaning clear.


6.1.13 @abbr{abbreviation[, meaning]}

You can use the @abbr command for general abbreviations. The abbreviation is given as the single argument in braces, as in ‘@abbr{Comput.}’. As a matter of style, or for particular abbreviations, you may prefer to omit periods, as in ‘@abbr{Mr} Stallman’.

@abbr accepts an optional second argument, intended to be used for the meaning of the abbreviation.

If the abbreviation ends with a lowercase letter and a period, and is not at the end of a sentence, and has no second argument, remember to use the @. command (see Ending a Sentence) to get the correct spacing. However, you do not have to use @. within the abbreviation itself; Texinfo automatically assumes periods within the abbreviation do not end a sentence.

In TeX and in the Info output, the first argument is printed as-is; if the second argument is present, it is printed in parentheses after the abbreviation. In HTML the <abbr> tag is used; in DocBook, the <abbrev> tag is used. For instance:

@abbr{Comput. J., Computer Journal}

produces:

Comput. J. (Computer Journal)

For abbreviations consisting of all capital letters, you may prefer to use the @acronym command instead. See the next section for more on the usage of these two commands.


6.1.14 @acronym{acronym[, meaning]}

You can use the @acronym command for abbreviations written in all capital letters, such as ‘NASA’. The abbreviation is given as the single argument in braces, as in ‘@acronym{NASA}’. As a matter of style, or for particular acronyms, you may prefer to use periods, as in ‘@acronym{N.A.S.A.}’.

@acronym accepts an optional second argument, intended to be used for the meaning of the acronym.

If the acronym is at the end of a sentence, and if there is no second argument, remember to use the @. or similar command (see Ending a Sentence) to get the correct spacing.

In TeX, the acronym is printed in slightly smaller font. In the Info output, the argument is printed as-is. In either format, if the second argument is present, it is printed in parentheses after the acronym. In HTML and DocBook the <acronym> tag is used.

For instance (since GNU is a recursive acronym, we use @acronym recursively):

@acronym{GNU, @acronym{GNU}'s Not Unix}

produces:

GNU (GNU’s Not Unix)

In some circumstances, it is conventional to print family names in all capitals. Don’t use @acronym for this, since a name is not an acronym. Use @sc instead (see @sc{text}: The Small Caps Font).

@abbr and @acronym are closely related commands: they both signal to the reader that a shortened form is being used, and possibly give a meaning. When choosing whether to use these two commands, please bear the following in mind.

  • - In common English usage, acronyms are a subset of abbreviations: they include pronounceable words like ‘NATO’, ‘radar’, and ‘snafu’; some sources also include syllable acronyms like ‘Usenet’, hybrids like ‘SIGGRAPH’, and unpronounceable initialisms like ‘FBI’.
  • - In Texinfo, an acronym (but not an abbreviation) should consist only of capital letters and periods, no lowercase.
  • - In TeX, an acronym (but not an abbreviation) is printed in a slightly smaller font.
  • - Some browsers place a dotted bottom border under abbreviations but not acronyms.
  • - It usually turns out to be quite difficult and/or time-consuming to consistently use @acronym for all sequences of uppercase letters. Furthermore, it looks strange for some acronyms to be in the normal font size and others to be smaller. Thus, a simpler approach you may wish to consider is to avoid @acronym and just typeset everything as normal text in all capitals: ‘GNU’, producing the output ‘GNU’.
  • - In general, it’s not essential to use either of these commands for all abbreviations; use your judgment. Text is perfectly readable without them.

6.1.15 @indicateurl{uniform-resource-locator}

Use the @indicateurl command to indicate a uniform resource locator on the World Wide Web. This is purely for markup purposes and does not produce a link you can follow (use the @url or @uref command for that, see @url, @uref{url[, text][, replacement]}). @indicateurl is useful for urls which do not actually exist. For example:

For example, the url might be @indicateurl{http://example.org/path}.

which produces:

For example, the url might be ‘http://example.org/path’.

The output from @indicateurl is more or less like that of @samp (see @samp{text}).


6.1.16 @email{email-address[, displayed-text]}

Use the @email command to indicate an electronic mail address. It takes one mandatory argument, the address, and one optional argument, the text to display (the default is the address itself).

In Info, the address is shown in angle brackets, preceded by the text to display if any. In TeX, the angle brackets are omitted. In HTML output, @email produces a ‘mailto’ link that usually brings up a mail composition window. For example:

Send bug reports to @email{bug-texinfo@@gnu.org},
suggestions to the @email{bug-texinfo@@gnu.org, same place}.

produces

Send bug reports to bug-texinfo@gnu.org,
suggestions to the same place.

6.2 Emphasizing Text

Usually, Texinfo changes the font to mark words in the text according to the category the words belong to; an example is the @code command. Most often, this is the best way to mark words. However, sometimes you will want to emphasize text without indicating a category. Texinfo has two commands to do this. Also, Texinfo has several commands that specify the font in which text will be output. These commands have no effect in Info and only one of them, the @r command, has any regular use.


6.2.1 @emph{text} and @strong{text}

The @emph and @strong commands are for emphasis; @strong is stronger. In printed output, @emph produces italics and @strong produces bold. In the Info output, @emph surrounds the text with underscores (‘_’), and @strong puts asterisks around the text.

For example,

@strong{Caution:} @samp{rm * .[^.]*}
removes @emph{all} files in the directory.

produces the following:

Caution: ‘rm * .[^.]*’ removes all files in the directory.

The @strong command is seldom used except to mark what is, in effect, a typographical element, such as the word ‘Caution’ in the preceding example.

Caution: Do not use @strong with the word ‘Note’ followed by a space; Info will mistake the combination for a cross-reference. Use a phrase such as Please notice or Caution instead, or the optional argument to @quotation—‘Note’ is allowable there.


6.2.2 @sc{text}: The Small Caps Font

Use the ‘@sc’ command to set text in A SMALL CAPS FONT (where possible). Write the text you want to be in small caps between braces in lowercase, like this:

Richard @sc{Stallman} a commencé le projet GNU.

This produces:

Richard STALLMAN a commencé le projet GNU.

As shown here, we recommend reserving @sc for special cases where you want typographic small caps; family names are one such, especially in languages other than English, though there are no hard-and-fast rules about such things.

TeX typesets any uppercase letters between the braces of an @sc command in full-size capitals; only lowercase letters are printed in the small caps font. In the Info output, the argument to @sc is printed in all uppercase. In HTML, the argument is uppercased and the output marked with the <small> tag to reduce the font size, since HTML cannot easily represent true small caps.

Overall, we recommend using standard upper- and lowercase letters wherever possible.


6.2.3 Fonts for Printing

Texinfo provides one command to change the size of the main body font in the TeX output for a document: @fonttextsize. It has no effect in other output. It takes a single argument on the remainder of the line, which must be either ‘10’ or ‘11’. For example:

@fonttextsize 10

The effect is to reduce the body font to a 10pt size (the default is 11pt). Fonts for other elements, such as sections and chapters, are reduced accordingly. This should only be used in conjunction with @smallbook (see @smallbook: Printing “Small” Books) or similar, since 10pt fonts on standard paper (8.5x11 or A4) are too small. One reason to use this command is to save pages, and hence printing cost, for physical books.

Texinfo does not at present have commands to switch the font family to use, or more general size-changing commands.

Texinfo also provides a number of font commands that specify font changes in the printed manual and (where possible) in the HTML output. They have no effect in Info. All the commands apply to a following argument surrounded by braces.

@b

selects bold face;

@i

selects an italic font;

@r

selects a roman font, which is the usual font in which text is printed. It may or may not be seriffed.

@sansserif

selects a sans serif font;

@slanted

selects a slanted font;

@t

selects the fixed-width, typewriter-style font used by @code;

(The commands with longer names were invented much later than the others, at which time it did not seem desirable to use very short names for such infrequently needed features.)

The @r command can be useful in example-like environments, to write comments in the standard roman font instead of the fixed-width font. This looks better in printed output, and produces a <lineannotation> tag in DocBook output.

For example,

@lisp
(+ 2 2)    ; @r{Add two plus two.}
@end lisp

produces

(+ 2 2)    ; Add two plus two.

The @t command can occasionally be useful to produce output in a typewriter font where that is supported (e.g., HTML and PDF), but no distinction is needed in Info or plain text: @t{foo} produces foo, cf. @code{foo} producing foo.

In general, the other font commands are unlikely to be useful; they exist primarily to make it possible to document the functionality of specific font effects, such as in TeX and related packages.


7 Quotations and Examples

Quotations and examples are blocks of text consisting of one or more whole paragraphs that are set off from the bulk of the text and treated differently. They are usually indented in the output.

In Texinfo, you always begin a quotation or example by writing an @-command at the beginning of a line by itself, and end it by writing an @end command that is also at the beginning of a line by itself. For instance, you begin an example by writing @example by itself at the beginning of a line and end the example by writing @end example on a line by itself, at the beginning of that line, and with only one space between the @end and the example.


7.1 Block Enclosing Commands

Here is a summary of commands that enclose blocks of text, also known as environments. They’re explained further in the following sections.

@quotation

Indicate text that is quoted. The text is filled, indented (from both margins), and printed in a roman font by default.

@indentedblock

Like @quotation, but the text is indented only on the left.

@example

Illustrate code, commands, and the like. The text is printed in a fixed-width font, and indented but not filled.

@lisp

Like @example, but specifically for illustrating Lisp code. The text is printed in a fixed-width font, and indented but not filled.

@verbatim

Mark a piece of text that is to be printed verbatim; no character substitutions are made and all commands are ignored, until the next @end verbatim. The text is printed in a fixed-width font, and not indented or filled. Extra spaces and blank lines are significant, and tabs are expanded.

@display

Display illustrative text. The text is indented but not filled, and no font is selected (so, by default, the font is roman).

@format

Like @display (the text is not filled and no font is selected), but the text is not indented.

@smallquotation
@smallindentedblock
@smallexample
@smalllisp
@smalldisplay
@smallformat

These @small... commands are just like their non-small counterparts, except that they output text in a smaller font size, where possible.

@flushleft
@flushright

Text is not filled, but is set flush with the left or right margin, respectively.

@raggedright

Text is filled, but only justified on the left, leaving the right margin ragged.

@cartouche

Highlight text, often an example or quotation, by drawing a box with rounded corners around it.

The @exdent command is used within the above constructs to undo the indentation of a line.

The @noindent command may be used after one of the above constructs (or at the beginning of any paragraph) to prevent the following text from being indented as a new paragraph.


7.2 @quotation: Block Quotations

The text of a quotation is processed like normal text (regular font, text is filled) except that:

  • both the left and right margins are closer to the center of the page, so the whole of the quotation is indented;
  • the first lines of paragraphs are indented no more than other lines; and
  • an @author command may be given to specify the author of the quotation.

This is an example of text written between a @quotation command and an @end quotation command. A @quotation command is most often used to indicate text that is excerpted from another (real or hypothetical) printed work.

Write a @quotation command as text on a line by itself. This line will disappear from the output. Mark the end of the quotation with a line beginning with and containing only @end quotation. The @end quotation line will likewise disappear from the output.

@quotation takes one optional argument, given on the remainder of the line. This text, if present, is included at the beginning of the quotation in bold or otherwise emphasized, and followed with a ‘:’. For example:

@quotation Note
This is
a foo.
@end quotation

produces

Note: This is a foo.

If the @quotation argument is one of these English words (case-insensitive):

Caution  Important  Note  Tip  Warning

then the DocBook output uses corresponding special tags (<note>, etc.) instead of the default <blockquote>. HTML output always uses <blockquote>.

If the author of the quotation is specified in the @quotation block with the @author command, a line with the author name is displayed after the quotation:

@quotation
People sometimes ask me if it is a sin in the Church of Emacs to use
vi.  Using a free version of vi is not a sin; it is a penance.  So happy
hacking.

@author Richard Stallman
@end quotation

produces

People sometimes ask me if it is a sin in the Church of Emacs to use vi. Using a free version of vi is not a sin; it is a penance. So happy hacking.

Richard Stallman

7.3 @indentedblock: Indented text blocks

The @indentedblock environment is similar to @quotation, except that text is only indented on the left (and there is no optional argument for an author). Thus, the text font remains unchanged, and text is gathered and filled as usual, but the left margin is increased. For example:

This is an example of text written between an @indentedblock command and an @end indentedblock command. The @indentedblock environment can contain any text or other commands desired.

This is written in the Texinfo source as:

@indentedblock
This is an example ...
@end indentedblock

7.4 @example: Example Text

The @example environment is used to indicate computer input or output that is not part of the running text. If you want to embed code fragments within sentences, use the @code command or its relatives instead (see @code{sample-code}).

Write an @example command at the beginning of a line by itself. Mark the end of the block with @end example. For example,

@example
cp foo @var{dest1}; \
 cp foo @var{dest2}
@end example

produces

cp foo dest1; \
 cp foo dest2

The output uses a fixed-width font and is indented. Each line in the input file is a line in the output; that is, the source text is not filled. Extra spaces and blank lines are significant. Texinfo commands are expanded; if you want the output to be the input verbatim, use the @verbatim environment instead (see @verbatim: Literal Text).

Examples are often, logically speaking, “in the middle” of a paragraph, and the text that continues afterwards should not be indented, as in the example above. The @noindent command prevents a piece of text from being indented as if it were a new paragraph (see @noindent: Omitting Indentation).

If you wish to use the normal roman font for a code comment, you can use the @r command (see Fonts for Printing).

You may optionally give arguments to the @example command, separated by commas if there is more than one. In the HTML output, any such arguments are output as class names. This may be useful for adding syntax highlighting to manuals for code samples. We recommend that when you give multiple arguments to @example, you use the first argument to specify the language of the code (e.g. ‘C’, ‘lisp’, ‘Cplusplus’). You may find uses for other arguments, such as providing a formatting hint or marking code samples for extraction and further processing, but for now nothing definitive is recommended. Perhaps this will change in future Texinfo releases.

Caution: Do not use tabs in the lines of an example! (Or anywhere else in Texinfo, except in verbatim environments.) TeX treats tabs as single spaces, and that is not what they look like.


7.5 @verbatim: Literal Text

Use the @verbatim environment for printing of text that may contain special characters or commands that should not be interpreted, such as computer input or output (@example interprets its text as regular Texinfo commands). This is especially useful for including automatically generated files in a Texinfo manual.

In general, the output will be just the same as the input. No character substitutions are made, e.g., all spaces and blank lines are significant, including tabs. In the printed manual, the text is typeset in a fixed-width font, and not indented or filled.

Write a @verbatim command at the beginning of a line by itself. This line will disappear from the output. Mark the end of the verbatim block with an @end verbatim command, also written at the beginning of a line by itself. The @end verbatim will also disappear from the output.

For example:

@verbatim
{
TAB@command with strange characters: @'e
expandTABme
}
@end verbatim

This produces:

{
        @command with strange characters: @'e
expand	me
}

Since the lines containing @verbatim and @end verbatim produce no output, typically you should put a blank line before the @verbatim and another blank line after the @end verbatim. Blank lines between the beginning @verbatim and the ending @end verbatim will appear in the output.

You can get a “small” verbatim by enclosing the @verbatim in an @smallformat environment, as shown here:

@smallformat
@verbatim
... still verbatim, but in a smaller font ...
@end verbatim
@end smallformat

Finally, a word of warning: it is not reliable to use @verbatim inside other Texinfo constructs.

See also @verbatiminclude file: Include a File Verbatim.


7.6 @lisp: Marking a Lisp Example

The @lisp command is used for Lisp code:

@lisp
Example lisp code
@end lisp

This is synonymous with the following:

@example lisp
Example lisp code
@end example

Use @lisp to preserve information regarding the nature of the example. This is useful, for example, if you write a function that evaluates only and all the Lisp code in a Texinfo file. Then you can use the Texinfo file as a Lisp library.


7.7 @display: Examples Using the Text Font

The @display command begins another kind of environment, where the font is left unchanged, not switched to typewriter as with @example. Each line of input still produces a line of output, and the output is still indented.

This is an example of text written between a @display command
and an @end display command.  The @display command
indents the text, but does not fill it.

7.8 @format: Examples Using the Full Line Width

The @format command is similar to @display, except it leaves the text unindented. Like @display, it does not select the fixed-width font.

This is an example of text written between a @format command
and an @end format command.  As you can see
from this example,
the @format command does not fill the text.

7.9 @exdent: Undoing a Line’s Indentation

The @exdent command removes any indentation a line might have. The command is written at the beginning of a line and applies only to the text that follows the command that is on the same line. Do not use braces around the text. In a printed manual, the text on an @exdent line is printed in the roman font.

@exdent is usually used within examples. Thus,

@example
This line follows an @@example command.
@exdent This line is exdented.
This line follows the exdented line.
The @@end example comes on the next line.
@end example

produces

This line follows an @example command.
This line is exdented.
This line follows the exdented line.
The @end example comes on the next line.

In practice, the @exdent command is rarely used. Usually, you un-indent text by ending the example and returning the page to its normal width.

@exdent has no effect in HTML output.


7.10 @flushleft and @flushright

The @flushleft and @flushright commands line up the ends of lines on the left and right margins of a page, but do not fill the text. The commands are written on lines of their own, without braces. The @flushleft and @flushright commands are ended by @end flushleft and @end flushright commands on lines of their own.

For example,

@flushleft
This text is
written flushleft.
@end flushleft

produces

This text is written flushleft.

@flushright produces the type of indentation often used in the return address of letters. For example,

@flushright
Here is an example of text written
flushright.  The @code{@flushright} command
right justifies every line but leaves the
left end ragged.
@end flushright

produces

Here is an example of text written flushright. The @flushright command right justifies every line but leaves the left end ragged.


7.11 @raggedright: Ragged Right Text

The @raggedright fills text as usual, but the text is only justified on the left; the right margin is ragged. The command is written on a line of its own, without braces. The @raggedright command is ended by @end raggedright on a line of its own. This command has no effect in Info and HTML output, where text is always set ragged right.

The @raggedright command can be useful with paragraphs containing lists of commands with long names, when it is known in advance that justifying the text on both margins will make the paragraph look bad.

An example (from elsewhere in this manual):

@raggedright
Commands for double and single angle quotation marks:
@code{@@guillemetleft@{@}}, @code{@@guillemetright@{@}},
@code{@@guillemotleft@{@}}, @code{@@guillemotright@{@}},
@code{@@guilsinglleft@{@}}, @code{@@guilsinglright@{@}}.
@end raggedright

produces

Commands for double and single angle quotation marks: @guillemetleft{}, @guillemetright{}, @guillemotleft{}, @guillemotright{}, @guilsinglleft{}, @guilsinglright{}.


7.12 @noindent: Omitting Indentation

An example or other inclusion can break a paragraph into segments. Ordinarily, the formatters indent text that follows an example as a new paragraph. You can prevent this on a case-by-case basis by writing @noindent at the beginning of a line, preceding the continuation text. You can also disable indentation for all paragraphs globally with @paragraphindent (see @paragraphindent: Controlling Paragraph Indentation).

Here is an example showing how to eliminate the normal indentation of the text after an @example, a common situation:

@example
This is an example
@end example

@noindent
This line is not indented.  As you can see, the
beginning of the line is fully flush left with the
line that follows after it.

produces:

This is an example

This line is not indented.  As you can see, the
beginning of the line is fully flush left with the
line that follows after it.

The standard usage of @noindent is just as above: at the beginning of what would otherwise be a paragraph, to eliminate the indentation that normally happens there. It can either be followed by text or be on a line by itself. There is no reason to use it in other contexts, such as in the middle of a paragraph or inside an environment (see Quotations and Examples).

You can control the number of blank lines in the Info file output by adjusting the input as desired: a line containing just @noindent does not generate a blank line, and neither does an @end line for an environment.

Do not put braces after a @noindent command; they are not used, since @noindent is a command used outside of paragraphs (see @-Command Syntax).


7.13 @indent: Forcing Indentation

To complement the @noindent command (see the previous section), Texinfo provides the @indent command to force a paragraph to be indented. For instance, this paragraph (the first in this section) is indented using an @indent command.

And indeed, the first paragraph of a section is the most likely place to use @indent, to override the normal behavior of no indentation there (see @paragraphindent: Controlling Paragraph Indentation). It can either be followed by text or be on a line by itself.

As a special case, when @indent is used in an environment where text is not filled, it produces a paragraph indentation space in the TeX output. (These environments are where a line of input produces a line of output, such as @example and @display; for a summary of all environments, see Block Enclosing Commands.)

Do not put braces after an @indent command; they are not used, since @indent is a command used outside of paragraphs (see @-Command Syntax).


7.14 @cartouche: Rounded Rectangles

In a printed manual, the @cartouche command draws a box with rounded corners around its contents. In HTML, a normal rectangle is drawn. @cartouche has no effect in Info output.

You can use this command to further highlight an example or quotation. For instance, you could write a manual in which one type of example is surrounded by a cartouche for emphasis.

For example,

@cartouche
@example
% pwd
/usr/local/share/emacs
@end example
@end cartouche

surrounds the two-line example with a box with rounded corners, in the printed manual.

The output from the example looks like this (if you’re reading this in Info, you’ll see the @cartouche had no effect):

% pwd
/usr/local/share/emacs

@cartouche also implies @group (see @group: Prevent Page Breaks).


7.15 @small… Block Commands

In addition to the regular @example and similar commands, Texinfo has “small” example-style commands. These are @smallquotation, @smallindentedblock, @smalldisplay, @smallexample, @smallformat, and @smalllisp.

In Info and HTML output, the @small… commands are equivalent to their non-small companion commands.

In TeX, however, the @small… commands typeset text in a smaller font than the non-small example commands. Thus, for instance, code examples can contain longer lines and still fit on a page without needing to be rewritten.

A smaller font size is also retained in the Texinfo XML transliteration.

Mark the end of a @small… block with a corresponding @end small…. For example, pair @smallexample with @end smallexample.

Here is an example of the font used by the @smallexample command (in Info, the output will be the same as usual):

… to make sure that you have the freedom to
distribute copies of free software (and charge for
this service 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.

The @small… commands use the same font style as their normal counterparts: @smallexample and @smalllisp use a fixed-width font, and everything else uses the regular font. They also have the same behavior in other respects—whether filling is done and whether margins are narrowed.

As a general rule, it’s better to just use the regular commands (such as @example instead of @smallexample), unless you have a good reason for it.


8 Lists and Tables

Texinfo has several ways of making lists and tables. Lists can be bulleted or numbered; two-column tables can highlight the items in the first column; multi-column tables are also supported.


8.1 Introducing Lists

Texinfo automatically indents the text in lists or tables, and numbers an enumerated list. This last feature is useful if you modify the list, since you do not need to renumber it yourself.

Numbered lists and tables begin with the appropriate @-command at the beginning of a line, and end with the corresponding @end command on a line by itself. The table and itemized-list commands also require that you write formatting information on the same line as the beginning @-command.

Begin an enumerated list, for example, with an @enumerate command and end the list with an @end enumerate command. Begin an itemized list with an @itemize command, followed on the same line by a formatting command such as @bullet, and end the list with an @end itemize command.

Precede each element of a list with an @item or @itemx command.


Here is an itemized list of the different kinds of table and lists:

  • Itemized lists with and without bullets.
  • Enumerated lists, using numbers or letters.
  • Two-column tables with highlighting.

Here is an enumerated list with the same items:

  1. Itemized lists with and without bullets.
  2. Enumerated lists, using numbers or letters.
  3. Two-column tables with highlighting.

And here is a two-column table with the same items and their @-commands:

@itemize

Itemized lists with and without bullets.

@enumerate

Enumerated lists, using numbers or letters.

@table
@ftable
@vtable

Two-column tables, optionally with indexing.


8.2 @itemize: Making an Itemized List

The @itemize command produces a sequence of “items”, each starting with a bullet or other mark inside the left margin, and generally indented.

Begin an itemized list by writing @itemize at the beginning of a line. Follow the command, on the same line, with a character or a Texinfo command that generates a mark. Usually, you will use @bullet after @itemize, but you can use @minus, or any command or character that results in a single character in the Info file. (When you write the mark command such as @bullet after an @itemize command, you may omit the ‘{}’.) If you don’t specify a mark command, the default is @bullet. If you don’t want any mark at all, but still want logical items, use @w{} (in this case the braces are required).

After the @itemize, write your items, each starting with @item. Text can follow on the same line as the @item. The text of an item can continue for more than one paragraph.

There should be at least one @item inside the @itemize environment. If none are present, makeinfo gives a warning. If you just want indented text and not a list of items, use @indentedblock; see @indentedblock: Indented text blocks.

Index entries and comments that are given before an @item including the first, are automatically moved (internally) to after the @item, so the output is as expected. Historically this has been a common practice.

Usually, you should put a blank line between items. This puts a blank line in the Info file. (TeX inserts the proper vertical space in any case.) Except when the entries are very brief, these blank lines make the list look better.

Here is an example of the use of @itemize, followed by the output it produces. @bullet produces an ‘*’ in Info and a round dot in other output formats.

@itemize @bullet
@item
Some text for foo.

@item
Some text
for bar.
@end itemize

This produces:

  • Some text for foo.
  • Some text for bar.

Itemized lists may be embedded within other itemized lists. Here is a list marked with dashes embedded in a list marked with bullets:

@itemize @bullet
@item
First item.

@itemize @minus
@item
Inner item.

@item
Second inner item.
@end itemize

@item
Second outer item.
@end itemize

This produces:

  • First item.
    • - Inner item.
    • - Second inner item.
  • Second outer item.

8.3 @enumerate: Making a Numbered or Lettered List

@enumerate is like @itemize (see @itemize: Making an Itemized List), except that the labels on the items are successive integers or letters instead of bullets.

Write the @enumerate command at the beginning of a line. The command does not require an argument, but accepts either a number or a letter as an option. Without an argument, @enumerate starts the list with the number ‘1’. With a numeric argument, such as ‘3’, the command starts the list with that number. With an upper- or lowercase letter, such as ‘a’ or ‘A’, the command starts the list with that letter.

Write the text of the enumerated list in the same way as an itemized list: write a line starting with @item at the beginning of each item in the enumeration. It is ok to have text following the @item, and the text for an item can continue for several paragraphs.

You should put a blank line between entries in the list. This generally makes it easier to read the Info file.

Here is an example of @enumerate without an argument:

@enumerate
@item
Underlying causes.

@item
Proximate causes.
@end enumerate

This produces:

  1. Underlying causes.
  2. Proximate causes.

Here is an example with an argument of 3:


@enumerate 3
@item
Predisposing causes.

@item
Precipitating causes.

@item
Perpetuating causes.
@end enumerate

This produces:

  1. Predisposing causes.
  2. Precipitating causes.
  3. Perpetuating causes.

Here is a brief summary of the alternatives. The summary is constructed using @enumerate with an argument of a.


  1. @enumerate

    Without an argument, produce a numbered list, with the first item numbered 1.

  2. @enumerate unsigned-integer

    With an (unsigned) numeric argument, start a numbered list with that number. You can use this to continue a list that you interrupted with other text.

  3. @enumerate upper-case-letter

    With an uppercase letter as argument, start a list in which each item is marked by a letter, beginning with that uppercase letter.

  4. @enumerate lower-case-letter

    With a lowercase letter as argument, start a list in which each item is marked by a letter, beginning with that lowercase letter.

You can also nest enumerated lists, as in an outline.


8.4 Making a Two-column Table

@table is similar to @itemize (see @itemize: Making an Itemized List), but allows you to specify a name or heading line for each item. The @table command is used to produce two-column tables, and is especially useful for glossaries, explanatory exhibits, and command-line option summaries.


8.4.1 Using the @table Command

Use the @table command to produce a two-column table. This command is typically used when you have a list of items and a brief text with each one, such as a list of definitions.

Write the @table command at the beginning of a line, after a blank line, and follow it on the same line with an argument that is an ‘indicating’ command, such as @code, @samp, @var, @option, or @kbd (see Indicating Definitions, Commands, etc.). This command will be applied to the text in the first column. For example, @table @code will cause the text in the first column to be output as if it had been the argument to a @code command.

You may use the @asis command as an argument to @table. @asis is a command that does nothing: if you use this command after @table, the first column entries are output without added highlighting (“as is”).

The @table command works with other commands besides those explicitly mentioned here. However, you can only use predefined Texinfo commands that take an argument in braces. You cannot reliably use a new command defined with @macro, although an @alias (for a suitable predefined command) is acceptable. See Defining New Texinfo Commands.

Begin each table entry with an @item command at the beginning of a line. Write the text for the first column on the same line as the @item command. Write the text for the second column on the line following the @item line and on subsequent lines. You may write as many lines of supporting text as you wish, even several paragraphs. But only the text on the same line as the @item will be placed in the first column (including any footnotes). You do not need to type anything for an empty second column.

Normally, you should put a blank line before an @item line (except the first one). This puts a blank line in the Info file. Except when the entries are very brief, a blank line looks better. End the table with a line consisting of @end table, followed by a blank line. TeX will always start a new paragraph after the table, so the blank line is needed for the Info output to be analogous.

For example, the following table highlights the text in the first column with the @samp command:

@table @samp
@item foo
This is the text for
@samp{foo}.

@item bar
Text for @samp{bar}.
@end table

This produces:

foo

This is the text for ‘foo’.

bar

Text for ‘bar’.

If you want to list two or more named items with a single block of text, use the @itemx command. (See @itemx: Second and Subsequent Items.)

The @table command (see Using the @table Command) is not supported inside @display. Since @display is line-oriented, it doesn’t make sense to use them together. If you want to indent a table, try @quotation (see @quotation: Block Quotations) or @indentedblock (see @indentedblock: Indented text blocks).


8.4.2 @ftable and @vtable

The @ftable and @vtable commands are the same as the @table command except that @ftable automatically enters each of the items in the first column of the table into the index of functions and @vtable automatically enters each of the items in the first column of the table into the index of variables. This simplifies the task of creating indices. Only the items on the same line as the @item or @itemx commands are indexed, and they are indexed in exactly the form that they appear on that line. See Indices, for more information about indices.

Begin a two-column table using @ftable or @vtable by writing the @-command at the beginning of a line, followed on the same line by an argument that is a Texinfo command such as @code, exactly as you would for a @table command; and end the table with an @end ftable or @end vtable command on a line by itself.

See the example for @table in the previous section.


8.4.3 @itemx: Second and Subsequent Items

Use the @itemx command inside a table when you have two or more first column entries for the same item, each of which should appear on a line of its own.

Use @item for the first entry, and @itemx for all subsequent entries; @itemx must always follow an @item command, with no blank line intervening.

The @itemx command works exactly like @item except that it does not generate extra vertical space above the first column text. If you have multiple consecutive @itemx commands, do not insert any blank lines between them.

For example,

@table @code
@item upcase
@itemx downcase
These two functions accept a character or a string as
argument, and return the corresponding uppercase (lowercase)
character or string.
@end table

This produces:

upcase
downcase

These two functions accept a character or a string as argument, and return the corresponding uppercase (lowercase) character or string.

(Note also that this example illustrates multi-line supporting text in a two-column table.)


8.5 @multitable: Multi-column Tables

@multitable allows you to construct tables with any number of columns, with each column having any width you like.

You define the column widths on the @multitable line itself, and write each row of the actual table following an @item command, with columns separated by a @tab command. Finally, @end multitable completes the table. Details in the sections below.


8.5.1 Multitable Column Widths

You can define the column widths for a multitable in two ways: as fractions of the line length; or with a prototype row. Mixing the two methods is not supported. In either case, the widths are defined entirely on the same line as the @multitable command.

  1. To specify column widths as fractions of the line length, write @columnfractions and the decimal numbers (presumably less than 1; a leading zero is allowed and ignored) after the @multitable command, as in:
    @multitable @columnfractions .33 .33 .33
    

    The fractions need not add up exactly to 1.0, as these do not. This allows you to produce tables that do not need the full line length.

    When using @columnfractions, the leftmost column may appear slightly wider than you might expect, relative to the other columns. This is due to spacing between columns being included in the width of the other columns.

  2. To specify a prototype row, write the longest entry for each column enclosed in braces after the @multitable command. For example:
    @multitable {some text for column one} {for column two}
    

    The first column will then have the width of the typeset ‘some text for column one’, and the second column the width of ‘for column two’.

    The prototype entries need not appear in the table itself.

    Although we used simple text in this example, the prototype entries can contain Texinfo commands; markup commands such as @code are particularly likely to be useful.


8.5.2 Multitable Rows

After the @multitable command defining the column widths (see the previous section), you begin each row in the body of a multitable with @item, and separate the column entries with @tab. Line breaks are not special within the table body, and you may break input lines in your source file as necessary.

You can also use @headitem instead of @item to produce a heading row. The TeX output for such a row is in bold, and the HTML and DocBook output uses the <thead> tag. In Info, the heading row is followed by a separator line made of dashes (‘-’ characters).

The command @headitemfont can be used in templates when the entries in a @headitem row need to be used in a template. It is a synonym for @b, but using @headitemfont avoids any dependency on that particular font style, in case we provide a way to change it in the future.

Here is a complete example of a multi-column table (the text is from The GNU Emacs Manual, see Splitting Windows in The GNU Emacs Manual):

@multitable @columnfractions .15 .45 .4
@headitem Key @tab Command @tab Description
@item C-x 2
@tab @code{split-window-vertically}
@tab Split the selected window into two windows,
with one above the other.
@item C-x 3
@tab @code{split-window-horizontally}
@tab Split the selected window into two windows
positioned side by side.
@item C-Mouse-2
@tab
@tab In the mode line or scroll bar of a window,
split that window.
@end multitable

produces:

KeyCommandDescription
C-x 2split-window-verticallySplit the selected window into two windows, with one above the other.
C-x 3split-window-horizontallySplit the selected window into two windows positioned side by side.
C-Mouse-2In the mode line or scroll bar of a window, split that window.

Next: , Previous: , Up: Texinfo   [Contents][Index]

9 Special Displays

The commands in this chapter allow you to write text that is specially displayed (output format permitting), outside of the normal document flow.

One set of such commands is for creating “floats”, that is, figures, tables, and the like, set off from the main text, possibly numbered, captioned, and/or referred to from elsewhere in the document. Images are often included in these displays.

Another group of commands is for creating footnotes in Texinfo.


9.1 Floats

A float is a display which is set off from the main text. It is typically labeled as being a “Figure”, “Table”, “Example”, or some similar type.

A float is so-named because, in principle, it can be moved to the bottom or top of the current page, or to a following page, in the printed output. (Floating does not make sense in other output formats.) In the present version of Texinfo, however, this floating is unfortunately not yet implemented. Instead, the floating material is simply output at the current location, more or less as if it were an @group (see @group: Prevent Page Breaks).


9.1.1 @float [type][,label]: Floating Material

To produce floating material, enclose the material you want to be displayed separate between @float and @end float commands, on lines by themselves.

Floating material often uses @image to display an already-existing graphic (see Inserting Images), or @multitable to display a table (see @multitable: Multi-column Tables). However, the contents of the float can be anything. Here’s an example with simple text:

@float Figure,fig:ex1
This is an example float.
@end float

And the output:

This is an example float.

Figure 9.1

As shown in the example, @float takes two arguments (separated by a comma), type and label. Both are optional.

type

Specifies the sort of float this is; typically a word such as “Figure”, “Table”, etc. If this is not given, and label is, any cross-referencing will simply use a bare number.

label

Specifies a cross-reference label for this float. If given, this float is automatically given a number, and will appear in any @listoffloats output (see @listoffloats: Tables of Contents for Floats). Cross references to label are allowed.

On the other hand, if label is not given, then the float will not be numbered and consequently will not appear in the @listoffloats output or be cross-referenceable.

Ordinarily, you specify both type and label, to get a labeled and numbered float.

In Texinfo, all floats are numbered in the same way: with the chapter number (or appendix letter), a period, and the float number, which simply counts 1, 2, 3, …, and is reset at each chapter. Each float type is counted independently.

Floats within an @unnumbered, or outside of any chapter, are simply numbered consecutively from 1.

These numbering conventions are not, at present, changeable.


9.1.2 @caption & @shortcaption

You may write a @caption anywhere within a @float environment, to define a caption for the float. It is not allowed in any other context. @caption takes a single argument, enclosed in braces. Here’s an example:

@float
An example float, with caption.
@caption{Caption for example float.}
@end float

The output is:

An example float, with caption.

Caption for example float.

@caption can appear anywhere within the float; it is not processed until the @end float. The caption text is usually a sentence or two, but may consist of several paragraphs if necessary.

In the output, the caption always appears below the float; this is not currently changeable. It is preceded by the float type and/or number, as specified to the @float command (see the previous section).

The @shortcaption command likewise may be used only within @float, and takes a single argument in braces. The short caption text is used instead of the caption text in a list of floats (see the next section). Thus, you can write a long caption for the main document, and a short title to appear in the list of floats. For example:

@float
... as above ...
@shortcaption{Text for list of floats.}
@end float

The text for @shortcaption may not contain comments (@c), verbatim text (@verb), environments such as @example, footnotes (@footnote) or other complex constructs. The same constraints apply to @caption unless there is a @shortcaption.


9.1.3 @listoffloats: Tables of Contents for Floats

You can write a @listoffloats command to generate a list of floats for a given float type (see @float [type][,label]: Floating Material), analogous to the document’s overall table of contents. Typically, it is written in its own @unnumbered node to provide a heading and structure, rather like @printindex (see Printing Indices and Menus).

@listoffloats takes one optional argument, the float type. Here’s an example:

@node List of Figures
@unnumbered List of Figures
@listoffloats Figure

And here’s what the output from @listoffloats looks like, given the example figure earlier in this chapter (the Info output is formatted as a menu):

Without any argument, @listoffloats generates a list of floats for which no float type was specified, i.e., no first argument to the @float command (see @float [type][,label]: Floating Material).

Each line in the list of floats contains the float type (if any), the float number, and the caption, if any—the @shortcaption argument, if it was specified, else the @caption argument. In Info, the result is a menu where each float can be selected. In HTML, each line is a link to the float. In printed output, the page number is included.

Unnumbered floats (those without cross-reference labels) are omitted from the list of floats.


Next: , Previous: , Up: Special Displays   [Contents][Index]

9.2 Inserting Images

You can insert an image given in an external file with the @image command. Although images can be used anywhere, including the middle of a paragraph, we describe them in this chapter since they are most often part of a displayed figure or example.


9.2.1 Image Syntax

Here is the synopsis of the @image command:

@image{filename[, width[, height[, alttext[, extension]]]]}

The filename argument is mandatory, and must not have an extension, because the different processors support different formats:

  • TeX (DVI output) reads the file filename.eps (Encapsulated PostScript format).
  • pdfTeX reads filename.pdf, filename.png, filename.jpg, or filename.jpeg (in that order). It also tries uppercase versions of the extensions. The PDF format does not support EPS images, so such must be converted first.
  • For Info, makeinfo includes filename.txt verbatim (more or less as if it were in @verbatim). The Info output may also include a reference to filename.png or filename.jpg. (See below.)
  • For HTML, makeinfo outputs a reference to filename.png, filename.jpg, filename.jpeg or filename.gif (in that order). If none of those exist, it gives an error, and outputs a reference to filename.jpg anyway.
  • For DocBook, makeinfo outputs references to filename.eps, filename.gif filename.jpeg, filename.jpg, filename.pdf, filename.png and filename.svg, for every file found. Also, filename.txt is included verbatim, if present. (The subsequent DocBook processor is supposed to choose the appropriate one.)
  • For Info and HTML output, makeinfo uses the optional fifth argument extension to @image for the file extension, if it is specified and the file is found. Any leading period should be included in extension. For example:
    @image{foo,,,,.xpm}
    

If you want to install image files for use by Info readers too, we recommend putting them in a subdirectory like ‘foo-figures’ for a package foo. Copying the files into $(infodir)/foo-figures/ should be done in your Makefile.

The width and height arguments are described in the next section.

If an image is the first thing in a paragraph and followed by more text, then you should precede the @image command with @indent or @noindent to indicate the beginning of paragraph formatting. This is especially important for TeX output to get correct paragraph indentation.

Use @center to center an image (see @titlefont, @center, and @sp).

For HTML output, makeinfo sets the alt attribute for inline images to the optional alttext (fourth) argument to @image, if supplied. If not supplied, makeinfo uses the full file name of the image being displayed. The alttext is processed as Texinfo text, so special characters such as ‘"’ and ‘<’ and ‘&’ are escaped in the HTML output; also, you can get an empty alt string with @- (a command that produces no output; see @- and @hyphenation: Helping TeX Hyphenate).

For Info output, the alt string is also processed as Texinfo text and output. In this case, ‘\’ is escaped as ‘\\’ and ‘"’ as ‘\"’; no other escapes are done.

In Info output, makeinfo writes a reference to the binary image file (trying filename suffixed with extension, .extension, .png, or .jpg, in that order) if one exists. It also literally includes the .txt file if one exists. This way, Info readers which can display images (such as the Emacs Info browser, running under X) can do so, whereas Info readers which can only use text (such as the standalone Info reader) can display the textual version.

The implementation of this is to put the following construct into the Info output:

^@^H[image src="binaryfile" text="txtfile"
           alt="alttext ... ^@^H]

where ‘^@’ and ‘^H’ stand for the actual null and backspace control characters. If one of the files is not present, the corresponding argument is omitted.

The reason for mentioning this here is that older Info browsers (this feature was introduced in Texinfo version 4.6) will display the above literally, which, although not pretty, should not be harmful.


9.2.2 Image Scaling

The optional width and height arguments to the @image command (see the previous section) specify the size to which to scale the image. They are only taken into account in TeX. If neither is specified, the image is presented in its natural size (given in the file); if only one is specified, the other is scaled proportionately; and if both are specified, both are respected, thus likely distorting the original image by changing its aspect ratio.

The width and height may be specified using any valid TeX dimension, namely:

pt

point (72.27pt = 1in)

pc

pica (1pc = 12pt)

bp

big point (72bp = 1in)

in

inch

cm

centimeter (2.54cm = 1in)

mm

millimeter (10mm = 1cm)

dd

didôt point (1157dd = 1238pt)

cc

cicero (1cc = 12dd)

sp

scaled point (65536sp = 1pt)

For example, the following will scale a file ridt.eps to one inch vertically, with the width scaled proportionately:

@image{ridt,,1in}

For @image to work with TeX, the file epsf.tex must be installed somewhere that TeX can find it. (The standard location is texmf/tex/generic/dvips/epsf.tex, where texmf is a root of your TeX directory tree.) This file is included in the Texinfo distribution and is also available from ftp://tug.org/tex/epsf.tex, among other places.

@image can be used within a line as well as for displayed figures. Therefore, if you intend it to be displayed, be sure to leave a blank line before the command, or the output will run into the preceding text.

Image scaling is presently implemented only in TeX, not in HTML or any other sort of output.


9.3 Footnotes

A footnote is for a reference that documents or elucidates the primary text.2

Footnotes are distracting; use them sparingly at most, and it is best to avoid them completely. Standard bibliographical references are usually better placed in a bibliography at the end of a document instead of in footnotes throughout.


9.3.1 Footnote Commands

In Texinfo, footnotes are created with the @footnote command. This command is followed immediately by a left brace, then by the text of the footnote, and then by a terminating right brace. Footnotes may be of any length (they will be broken across pages if necessary), but are usually short. The template is:

ordinary text@footnote{text of footnote}

As shown here, the @footnote command should come right after the text being footnoted, with no intervening space; otherwise, the footnote marker might end up starting a line.

For example, this clause is followed by a sample footnote3; in the Texinfo source, it looks like this:

…a sample footnote@footnote{Here is the sample
footnote.}; in the Texinfo source…

As you can see, this source includes two punctuation marks next to each other; in this case, ‘.};’ is the sequence. This is normal (the first ends the footnote and the second belongs to the sentence being footnoted), so don’t worry that it looks odd. (Another style, perfectly acceptable, is to put the footnote after punctuation belonging to the sentence, as in ‘;@footnote{...’.)

In a printed manual or book, the reference mark for a footnote is a small, superscripted number; the text of the footnote appears at the bottom of the page, below a horizontal line.

In Info, the reference mark for a footnote is a pair of parentheses with the footnote number between them, like this: ‘(1)’. The reference mark is followed by a cross-reference link to the footnote text if footnotes are put in separate nodes (see Footnote Styles).

In the HTML output, footnote references are generally marked with a small, superscripted number which is rendered as a hypertext link to the footnote text.

Footnotes cannot be nested, and cannot appear in section headings of any kind or other “unusual” places.

A final tip: footnotes in the argument of an @item command for an @table must be entirely on the same line as the @item (as usual). See Making a Two-column Table.


Previous: , Up: Footnotes   [Contents][Index]

9.3.2 Footnote Styles

Info has two footnote styles, which determine where the text of the footnote is located:

  • In the ‘End’ node style, all the footnotes for a single node are placed at the end of that node. The footnotes are separated from the rest of the node by a line of dashes with the word ‘Footnotes’ within it. Each footnote begins with an ‘(n)’ reference mark.

    Here is an example of the Info output for a single footnote in the end-of-node style:

    --------- Footnotes ---------
    
    (1)  Here is a sample footnote.
    
  • In the ‘Separate’ node style, all the footnotes for a single node are placed in an automatically constructed node of their own. In this style, a “footnote reference” follows each ‘(n)’ reference mark in the body of the node. The footnote reference is actually a cross-reference which you use to reach the footnote node.

    The name of the node with the footnotes is constructed by appending ‘-Footnotes to the name of the node that contains the footnotes. (Consequently, the footnotes’ node for the Footnotes node is Footnotes-Footnotes!) The footnotes’ node has an ‘Up’ node pointer that leads back to its parent node.

    Here is how the first footnote in this manual looks after being formatted for Info in the separate node style:

    File: texinfo.info  Node: Overview-Footnotes, Up: Overview
    
    (1) The first syllable of "Texinfo" is pronounced like "speck", not
    "hex". …
    

Unless your document has long and important footnotes (as in, say, Gibbon’s Decline and Fall …), we recommend the ‘end’ style, as it is simpler for readers to follow.

Use the @footnotestyle command to specify an Info file’s footnote style. Write this command at the beginning of a line followed by an argument, either ‘end’ for the end node style or ‘separate’ for the separate node style.

For example,

@footnotestyle end

or

@footnotestyle separate

Write a @footnotestyle command before or shortly after the end-of-header line at the beginning of a Texinfo file. (You should include any @footnotestyle command between the start-of-header and end-of-header lines, so the region formatting commands will format footnotes as specified.)

In HTML, when the footnote style is ‘end’, or if the output is not split, footnotes are put at the end of the output. If set to ‘separate’, and the output is split, they are placed in a separate file.


10 Indices

Using Texinfo, you can generate indices without having to sort and collate entries manually. In an index, the entries are listed in alphabetical order, together with information on how to find the discussion of each entry. In a printed manual, this information consists of page numbers. In an Info file, this information is a menu entry leading to the first node referenced.

Texinfo provides several predefined kinds of indices: an index for functions, an index for variables, an index for concepts, and so on. You can combine indices or use them for other than their canonical purpose. Lastly, you can define your own new indices.


10.1 Predefined Indices

Texinfo provides six predefined indices. Here are their nominal meanings, abbreviations, and the corresponding index entry commands:

cp

(@cindex) Concept index, for general concepts.

fn

(@findex) Function index, for function and function-like names (such as entry points of libraries).

ky

(@kindex) Keystroke index, for keyboard commands.

pg

(@pindex) Program index, for names of programs.

tp

(@tindex) Data type index, for type names (such as structures defined in header files).

vr

(@vindex) Variable index, for variable names (such as library global variables).

Not every manual needs all of these, and most manuals use only two or three at most. The present manual, for example, has two indices: a concept index and an @-command index. (The latter is actually the function index but is called a command index in the chapter heading.)

You are not required to use the predefined indices strictly for their canonical purposes. For example, suppose you wish to index some C preprocessor macros. You could put them in the function index along with actual functions, just by writing @findex commands for them; then, when you print the “Function Index” as an unnumbered chapter, you could give it the title ‘Function and Macro Index’ and all will be consistent for the reader.

On the other hand, it is best not to stray too far from the meaning of the predefined indices. Otherwise, in the event that your text is combined with other text from other manuals, the index entries will not match up. Instead, define your own new index (see Defining New Indices).

We recommend having a single index in the final document whenever possible, however many source indices you use, since then readers have only one place to look. Two or more source indices can be combined into one output index by using the @synindex or @syncodeindex commands (see Combining Indices).


10.2 Defining the Entries of an Index

The data to make an index come from many individual indexing commands scattered throughout the Texinfo source file. Each command says to add one entry to a particular index; after formatting, the index will give the current page number or node name as the reference.

An index entry consists of an indexing command at the beginning of a line followed, on the rest of the line, by the entry.

For example, this section begins with the following five entries for the concept index:

@cindex Defining indexing entries
@cindex Index entries, defining
@cindex Entries for an index
@cindex Specifying index entries
@cindex Creating index entries

Each predefined index has its own indexing command—@cindex for the concept index, @findex for the function index, and so on, as listed in the previous section.

Index entries should precede the visible material that is being indexed. For instance:

@cindex hello
Hello, there!

Among other reasons, that way following indexing links (in whatever context) ends up before the material, where readers want to be, instead of after.

Try to avoid using a colon in index entries, as this may confuse some Info readers. See The Parts of a Menu for more information about the structure of a menu entry.

By default, entries for a concept index are printed in a small roman font and entries for the other indices are printed in a small @code font. You may change the way part of an entry is printed with the usual Texinfo commands, such as @file for file names (see Marking Text, Words and Phrases), and @r for the normal roman font (see Fonts for Printing).

For the printed output, you may specify an explicit sort key for an index entry using @sortas following either the index command or the text of the entry. For example: ‘@findex @sortas{\} \ @r{(literal \ in @code{@@math})’ sorts the index entry this produces under backslash.

To reduce the quantity of sort keys you need to provide explicitly, you may choose to ignore certain characters in index entries for the purposes of sorting. The characters that you can currently choose to ignore are ‘\’, ‘-’, ‘<’ and ‘@’, which are ignored by giving as an argument to the @set command, respectively, txiindexbackslashignore, txiindexhyphenignore, txiindexlessthanignore and txiindexatsignignore. For example, specifying ‘@set txiindexbackslashignore’ causes the ‘\mathopsup’ entry in the index for this manual to be sorted as if it were ‘mathopsup’, so that it appears among the other entries beginning with ‘M’.


10.3 Advanced Indexing Commands

Texinfo provides several commands for doing advanced indexing, similar to the indices you may see in professionally published books.

First, you can create multilevel index entries, allowing you to group many related subtopics under the same higher-level topic. You do this by separating the parts of such an entry with the @subentry command. Such commands might look like this:

@cindex Superhumans @subentry villains
@cindex Superhumans @subentry heroes

You may have up to three levels in an entry:

@cindex coffee makers @subentry electric @subentry pink
@cindex coffee makers @subentry electric @subentry blue

You can use the @sortas command mentioned earlier with any or all of the three parts of an entry to cause them to sort differently than they would by default.

Second, you may provide an index entry that points to another, using the @seeentry (“see entry”) command. For example:

@cindex Indexes @seeentry{Indices}

Such an entry should be unique in your document; the idea is to redirect the reader to the other entry where they will find all the information they are looking for.

Finally, you may provide a “see also” entry using the @seealso command. These entries go along with regular entries, and are grouped together with them in the final printed index. For example:

@cindex Coffee
@cindex Coffee @subentry With milk and sugar
@cindex Coffee @subentry With doughnuts
@cindex Coffee @subentry Decaffeinated
@cindex Coffee @seealso{Tea}

When using all three of these advanced commands, do not place a comma between the different parts of the index text. The texindex program, which sorts the index entries and generates the indexing formatting commands, takes care of placing commas in the correct places for you.

These features are most useful with printed documents created with TeX, and when translating Texinfo to DocBook.


10.4 Making Index Entries

Concept index entries consist of text. The best way to write an index is to devise entries which are terse yet clear. If you can do this, the index usually looks better if the entries are written just as they would appear in the middle of a sentence, that is, capitalizing only proper names and acronyms that always call for uppercase letters. This is the case convention we use in most GNU manuals’ indices.

If you don’t see how to make an entry terse yet clear, make it longer and clear—not terse and confusing. If many of the entries are several words long, the index may look better if you use a different convention: capitalize the first word of each entry. Whichever case convention you use, use it consistently.

In any event, do not ever capitalize a case-sensitive name such as a C or Lisp function name or a shell command; that would be a spelling error. Entries in indices other than the concept index are symbol names in programming languages, or program names; these names are usually case-sensitive, so likewise use upper- and lowercase as required.

It is a good idea to make index entries unique wherever feasible. That way, people using the printed output or online completion of index entries don’t see undifferentiated lists. Consider this an opportunity to make otherwise-identical index entries be more specific, so readers can more easily find the exact place they are looking for. The advanced indexing features described in Advanced Indexing Commands can help with this, as well.

When you are making index entries, it is good practice to think of the different ways people may look for something. Different people do not think of the same words when they look something up. A helpful index will have items indexed under all the different words that people may use. For example, one reader may think it obvious that the two-letter names for indices should be listed under “Indices, two-letter names”, since “Indices” are the general concept. But another reader may remember the specific concept of two-letter names and search for the entry listed as “Two letter names for indices”. A good index will have both entries and will help both readers.

Like typesetting, the construction of an index is a skilled art, the subtleties of which may not be appreciated until you need to do it yourself.


10.5 Printing Indices and Menus

To print an index means to include it as part of a manual or Info file. This does not happen automatically just because you use @cindex or other index-entry generating commands in the Texinfo file; those just cause the raw data for the index to be accumulated. To generate an index, you must include the @printindex command at the place in the document where you want the index to appear. Also, as part of the process of creating a printed manual, you must run a program called texindex (see Formatting and Printing Hardcopy) to sort the raw data to produce a sorted index file. The sorted index file is what is actually used to print the index.

Texinfo offers six separate types of predefined index, which suffice in most cases. See the other parts of this chapter for information on this, as well as advanced indexing commands, defining your own new indices, combining indices, and, most importantly, advice on writing the actual index entries. This section focuses on printing indices, which is done with the @printindex command.

@printindex takes one argument, a two-letter index abbreviation. It reads the corresponding sorted index file (for printed output), and formats it appropriately into an index.

The @printindex command does not generate a chapter heading for the index, since different manuals have different needs. Consequently, you should precede the @printindex command with a suitable section or chapter command (usually @appendix or @unnumbered) to supply the chapter heading and put the index into the table of contents. Precede the chapter heading with an @node line as usual.

For example:

@node Variable Index
@unnumbered Variable Index

@printindex vr

@node Concept Index
@unnumbered Concept Index

@printindex cp

If you have more than one index, we recommend placing the concept index last.

  • In printed output, @printindex produces a traditional two-column index, with dot leaders between the index terms and page numbers.
  • In Info output, @printindex produces a special menu containing the line number of the entry, relative to the start of the node. Info readers can use this to go to the exact line of an entry, not just the containing node. (Older Info readers will just go to the node.) Here’s an example:
    * First index entry:   Top.   (line  7)
    

    The actual number of spaces is variable, to right-justify the line number; it’s been reduced here to make the line fit in the printed manual.

  • In plain text output, @printindex produces the same menu, but the line numbers are relative to the start of the file, since that’s more convenient for that format.
  • In HTML output, @printindex produces links to the index entries.
  • In XML and DocBook output, it simply records the index to be printed.

10.6 Combining Indices

Sometimes you will want to combine two disparate indices such as functions and concepts, perhaps because you have few enough entries that a separate index would look silly.

You could put functions into the concept index by writing @cindex commands for them instead of @findex commands, and produce a consistent manual by printing the concept index with the title ‘Function and Concept Index’ and not printing the ‘Function Index’ at all; but this is not a robust procedure. It works only if your document is never included as part of another document that is designed to have a separate function index; if your document were to be included with such a document, the functions from your document and those from the other would not end up together. Also, to make your function names appear in the right font in the concept index, you would need to enclose every one of them between the braces of @code.


10.6.1 @syncodeindex: Combining Indices Using @code

When you want to combine functions and concepts into one index, you should index the functions with @findex and index the concepts with @cindex, and use the @syncodeindex command to redirect the function index entries into the concept index.

The @syncodeindex command takes two arguments; they are the name of the index to redirect, and the name of the index to redirect it to. The template looks like this:

@syncodeindex from to

For this purpose, the indices are given two-letter names:

cp

Concept index

fn

Function index

ky

Key index

pg

Program index

tp

Data type index

vr

Variable index

Write a @syncodeindex command before or shortly after the end-of-header line at the beginning of a Texinfo file. For example, to merge a function index with a concept index, write the following:

@syncodeindex fn cp

This causes all entries designated for the function index to merge in with the concept index instead.

To merge both a variable index and a function index into a concept index, write the following:

@syncodeindex vr cp
@syncodeindex fn cp

The @syncodeindex command puts all the entries from the ‘from’ index (the redirected index) into the @code font, overriding whatever default font is used by the index to which the entries are now directed. This way, if you direct function names from a function index into a concept index, all the function names are printed in the @code font as you would expect.


10.6.2 @synindex: Combining Indices

The @synindex command is nearly the same as the @syncodeindex command, except that it does not put the ‘from’ index entries into the @code font; rather it puts them in the roman font. Thus, you use @synindex when you merge a concept index into a function index.

See Printing Indices and Menus, for information about printing an index at the end of a book or creating an index menu in an Info file.


Previous: , Up: Indices   [Contents][Index]

10.7 Defining New Indices

In addition to the predefined indices (see Predefined Indices), you may use the @defindex and @defcodeindex commands to define new indices. These commands create new indexing @-commands with which you mark index entries. The @defindex command is used like this:

@defindex name

New index names are usually two-letter words, such as ‘au’. For example:

@defindex au

This defines a new index, called the ‘au’ index. At the same time, it creates a new indexing command, @auindex, that you can use to make index entries. Use this new indexing command just as you would use a predefined indexing command.

For example, here is a section heading followed by a concept index entry and two ‘au’ index entries.

@section Cognitive Semantics
@cindex kinesthetic image schemas
@auindex Johnson, Mark
@auindex Lakoff, George

(Evidently, ‘au’ serves here as an abbreviation for “author”.)

Texinfo constructs the new indexing command by concatenating the name of the index with ‘index’; thus, defining an ‘xy’ index leads to the automatic creation of an @xyindex command.

Use the @printindex command to print the index, as you do with the predefined indices. For example:

@node Author Index
@unnumbered Author Index

@printindex au

The @defcodeindex command is like the @defindex command, except that, in the printed output, it prints entries in an @code font by default instead of in a roman font.

You should define new indices before the end-of-header line of a Texinfo file, and (of course) before any @synindex or @syncodeindex commands (see Texinfo File Header).

As mentioned earlier (see Predefined Indices), we recommend having a single index in the final document whenever possible (no matter how many source indices you use), since then readers have only one place to look.

When creating an index, TeX creates a file whose extension is the name of the index (see Names of index files). Therefore you should avoid using index names that collide with extensions used for other purposes, such as ‘.aux’ or ‘.xml’. makeinfo already reports an error if a new index conflicts well-known extension name.


11 Special Insertions

Texinfo provides several commands for inserting characters that have special meaning in Texinfo, such as braces, and for other graphic elements that do not correspond to simple characters you can type.


11.1 Special Characters: Inserting @ {} , \ # &

@’ and curly braces are the basic special characters in Texinfo. To insert these characters so they appear in text, you must put an ‘@’ in front of these characters to prevent Texinfo from misinterpreting them. Alphabetic commands are also provided.

The other characters (comma, backslash, hash, ampersand) are special only in restricted contexts, as explained in the respective sections.


11.1.1 Inserting ‘@’ with @@ and @atchar{}

@@ produces a single ‘@’ character in the output. Do not put braces after an @@ command.

@atchar{} also produces a single ‘@’ character in the output. It does need following braces, as usual for alphabetic commands. In inline conditionals (see Inline Conditionals: @inline, @inlineifelse, @inlineraw), it can be necessary to avoid using the literal ‘@’ character in the source (and may be clearer in other contexts).


11.1.2 Inserting ‘{ ‘}’ with @{ @} and @l rbracechar{}

@{ produces a single ‘{’ in the output, and @} produces a single ‘}’. Do not put braces after either an @{ or an @} command.

@lbracechar{} and @rbracechar{} also produce single ‘{’ and ‘}’ characters in the output. They do need following braces, as usual for alphabetic commands. In inline conditionals (see Inline Conditionals: @inline, @inlineifelse, @inlineraw), it can be necessary to avoid using literal brace characters in the source (and may be clearer in other contexts).


11.1.3 Inserting ‘,’ with @comma{}

Ordinarily, a comma ‘,’ is a normal character that can be simply typed in your input where you need it.

However, Texinfo uses the comma as a special character only in one context: to separate arguments to those Texinfo commands, such as @acronym (see @acronym{acronym[, meaning]}) and @xref (see Cross-references), as well as user-defined macros (see Defining Macros), which take more than one argument.

Since a comma character would confuse Texinfo’s parsing for these commands, you must use the command ‘@comma{}’ instead if you want to pass an actual comma. Here are some examples:

@acronym{ABC, A Bizarre @comma{}}
@xref{Comma,, The @comma{} symbol}
@mymac{One argument@comma{} containing a comma}

Although ‘@comma{}’ can be used nearly anywhere, there is no need for it anywhere except in this unusual case.

(Incidentally, the name ‘@comma’ lacks the ‘char’ suffix used in its companion commands only for historical reasons. It didn’t seem important enough to define a synonym.)


11.1.4 Inserting ‘\’ with @backslashchar{}

Ordinarily, a backslash ‘\’ is a normal character in Texinfo that can be simply typed in your input where you need it. The result is to typeset the backslash from the typewriter font.

However, Texinfo uses the backslash as a special character in one restricted context: to delimit formal arguments in the bodies of user-defined macros (see Defining Macros).

Due to the vagaries of macro argument parsing, it is more reliable to pass an alphabetic command that produces a backslash instead of using a literal \. Hence @backslashchar{}. Here is an example macro call:

@mymac{One argument@backslashchar{} with a backslash}

Texinfo documents may also use \ as a command character inside @math (see @math and @displaymath: Formatting Mathematics). In this case, @\ or \backslash produces a “math” backslash (from the math symbol font), while @backslashchar{} produces a typewriter backslash as usual.

Although ‘@backslashchar{}’ can be used nearly anywhere, there is no need for it except in these unusual cases.


11.1.5 Inserting ‘#’ with @hashchar{}

Ordinarily, a hash ‘#’ is a normal character in Texinfo that can be simply typed in your input where you need it. The result is to typeset the hash character from the current font.

This character has many other names, varying by locale, such as “number sign”, “pound”, and “octothorp”. It is also sometimes called “sharp” or “sharp sign” since it vaguely resembles the musical symbol by that name. In situations where Texinfo is used, “hash” is the most common in our experience.

However, Texinfo uses the hash character as a special character in one restricted context: to introduce the so-called #line directive and variants (see External Macro Processors: Line Directives).

So, in order to typeset an actual hash character in such a place (for example, in a program that needs documentation about #line), it’s necessary to use @hashchar{} or some other construct. Here’s an example:

@hashchar{} 10 "example.c"

Although ‘@hashchar{}’ can be used nearly anywhere, there is no need for it anywhere except this unusual case.


11.1.6 Inserting ‘&’ with @& and @ampchar{}

Ordinarily, an ampersand ‘&’ is a normal character in Texinfo that can be simply typed in your input where you need it. The result is to typeset the ampersand character.

However, the ampersand character has a special meaning in Texinfo in just one restricted context. In the argument to a definition command (see Definition Commands), Emacs Lisp keywords beginning with ampersands are recognized and typeset specially. See A Sample Function Description in GNU Emacs Lisp Reference Manual. For example:

@defun foo integer1 &optional integer2 &rest integers
@code{foo} described here.
@end defun

leads to the output

Function: foo integer1 &optional integer2 &rest integers

foo described here.

So, in order to typeset an ampersand in such a context (for example, in documentation of a C++ function taking a reference as a parameter), it’s necessary to use @& or some other construct. Here’s an example:

@deftypefn Function int foo (@code{const std::vector<int>@&} bar)
Documentation of @code{foo}.
@end deftypefn

This gives the output

Function: int foo (const std::vector<int>& bar)

Documentation of foo.

Although ‘@&’ and ‘@ampchar{}’ can be used nearly anywhere, there is no need for them anywhere except this unusual case.


11.2 Inserting Quote Characters

As explained in the early section on general Texinfo input conventions (see General Syntactic Conventions), Texinfo source files use the ASCII character ` (96 decimal) to produce a left quote (‘), and ASCII ' (39 decimal) to produce a right quote (’). Doubling these input characters (`` and '') produces double quotes (“ and ”). These are the conventions used by TeX.

This works all right for text. However, in examples of computer code, readers are especially likely to cut and paste the text verbatim—and, unfortunately, some document viewers will mangle these characters. (The free PDF reader xpdf works fine, but other PDF readers, both free and nonfree, have problems.)

If this is a concern for you, Texinfo provides these two commands:

@codequoteundirected on-off

causes the output for the ' character in code environments to be the undirected single quote, like this: '.

@codequotebacktick on-off

causes the output for the ` character in code environments to be the backtick character (standalone grave accent), like this: `.

If you want these settings for only part of the document, @codequote... off will restore the normal behavior, as in @codequoteundirected off.

These settings affect @code, @example, @kbd, @samp, @verb, and @verbatim. See Highlighting Commands are Useful.

This feature can also be controlled by using @set to change the values of the corresponding variables txicodequoteundirected and txicodequotebacktick.


11.3 Inserting Space

The following sections describe commands that control spacing of various kinds within and after sentences.


11.3.1 Multiple Spaces

Ordinarily, multiple whitespace characters (space, tab, and newline) are collapsed into a single space.

Occasionally, you may want to produce several consecutive spaces, either for purposes of example (e.g., what your program does with multiple spaces as input), or merely for purposes of appearance in headings or lists. Texinfo supports three commands: @SPACE, @TAB, and @NL, all of which insert a single space into the output. (Here, @SPACE represents an ‘@’ character followed by a space, i.e., ‘@ ’, TAB represents an actual tab character, and @NL represents an ‘@’ character and end-of-line, i.e., when ‘@’ is the last character on a line.)

For example,

Spacey@ @ @ @
example.

produces

Spacey    example.

Other possible uses of @SPACE have been subsumed by @multitable (see @multitable: Multi-column Tables).

Do not follow any of these commands with braces.

To produce a non-breakable space, see @tie{}: Inserting an Unbreakable Space.


11.3.2 Not Ending a Sentence

When a period, exclamation point or question mark is at the end of a sentence, slightly more space is inserted after it in a typeset manual.

Usually, Texinfo can determine automatically when a period ends a sentence. However, special commands are needed in some circumstances. Use the @: command after a period, question mark, exclamation mark or colon that should not be followed by extra space. This is necessary in the following situations:

  1. After a period that ends a lowercase abbreviation which is not at the end of a sentence.
  2. When a parenthetical remark in the middle of a sentence (like this one!) ends with a period, exclamation point or question mark, @: should be used after the right parenthesis. Similarly for right brackets and right quotes (both single and double).

For example:

foo vs.@: bar (or?)@: baz’,

The first line below shows the output, and for comparison, the second line shows the spacing when the ‘@:’ commands were not used.

foo vs. bar (or?) baz
foo vs. bar (or?) baz

It may help you to remember what @: does by imagining that it stands for an invisible lower-case character that stops a word ending in a period.

A few Texinfo commands force normal interword spacing, so that you don’t have to insert @: where you otherwise would. These are the code-like highlighting commands, @var, @abbr, and @acronym (see Highlighting Commands are Useful). For example, in ‘@code{foo. bar}’ the period is not considered to be the end of a sentence, and no extra space is inserted.

@: has no effect on the HTML or DocBook output.


11.3.3 Ending a Sentence

As mentioned above, Texinfo normally inserts additional space after the end of a sentence. It uses the same heuristic for this as TeX: a sentence ends with a period, exclamation point, or question mark, either preceded or followed by optional closing punctuation, and then whitespace, and not preceded by a capital letter.

Use @. instead of a period, @! instead of an exclamation point, and @? instead of a question mark at the end of a sentence that does end with a capital letter. Do not put braces after any of these commands. For example:

Give it to M.I.B. and to M.E.W@.  Also, give it to R.J.C@.
Give it to M.I.B. and to M.E.W.  Also, give it to R.J.C.

The output follows. In printed output and Info, you can see the desired extra whitespace after the ‘W’ in the first line.

Give it to M.I.B. and to M.E.W. Also, give it to R.J.C.
Give it to M.I.B. and to M.E.W. Also, give it to R.J.C.

In the HTML output, @. is equivalent to a simple ‘.’; likewise for @! and @?.

The “closing punctuation” mentioned above is defined as a right parenthesis (‘)’, right bracket (‘]’), or right quote, either single or double (‘'’ and ‘''’; the many possible additional Unicode right quotes are not included). These characters can be thought of as invisible with respect to whether a given period ends a sentence. (This is the same rule as TeX.) For instance, the periods in ‘foo.) Bar’ and ‘foo.'' Bar’ do end sentences.

The meanings of @: and @., etc. in Texinfo are designed to work well with the Emacs sentence motion commands (see Sentences in The GNU Emacs Manual). It may help to imagine that the ‘@’ in ‘@.’, etc., is an invisible lower-case letter ‘a’ which makes an upper-case letter before it immaterial for the purposes of deciding whether the period ends the sentence.

A few Texinfo commands are not considered as being an abbreviation, even though they may end with a capital letter when expanded, so that you don’t have to insert @. and companions. Notably, this is the case for code-like highlighting commands, @var arguments ending with a capital letter, @LaTeX, and @TeX. For example, that sentence ended with ‘... @code{@@TeX}.’; @. was not needed. Similarly, in ... @var{VARNAME}. Text the period after VARNAME ends the sentence; there is no need to use @..


11.3.4 @frenchspacing val: Control Sentence Spacing

In American typography, it is traditional and correct to put extra space at the end of a sentence. This is the default in Texinfo (implemented in Info and printed output; for HTML, we don’t try to override the browser). In French typography (and others), this extra space is wrong; all spaces are uniform.

Therefore Texinfo provides the @frenchspacing command to control the spacing after punctuation. It reads the rest of the line as its argument, which must be the single word ‘on’ or ‘off’ (always these words, regardless of the language of the document). Here is an example:

@frenchspacing on
This is text. Two sentences. Three sentences. French spacing.

@frenchspacing off
This is text. Two sentences. Three sentences. Non-French spacing.

produces:

This is text. Two sentences. Three sentences. French spacing.

This is text. Two sentences. Three sentences. Non-French spacing.

@frenchspacing also affects the output after @., @!, and @? (see Ending a Sentence).

@frenchspacing has no effect on the HTML or DocBook output; for XML, it outputs a transliteration of itself (see Output Formats).


11.3.5 @dmn{dimension}: Format a Dimension

You can use the @dmn command to format a dimension with a little extra space in the printed output. That is, on seeing @dmn, TeX inserts just enough space for proper typesetting; in other output formats, the formatting commands insert no space at all.

To use the @dmn command, write the number and then follow it immediately, with no intervening space, by @dmn, and then by the dimension within braces. For example,

A4 paper is 8.27@dmn{in} wide.

produces

A4 paper is 8.27in wide.

Not everyone uses this style. Some people prefer ‘8.27 in.’ or ‘8.27 inches’. In these cases, however, you need to use @tie (see @tie{}: Inserting an Unbreakable Space) or @w (see @w{text}: Prevent Line Breaks) so that no line break can occur between the number and the dimension. Also, if you write a period after an abbreviation within a sentence (as with the ‘in.’ above), you should write ‘@:’ after the period to prevent TeX from inserting extra whitespace, as shown here. See Not Ending a Sentence.


11.4 Inserting Accents

Here is a table with the commands Texinfo provides for inserting floating accents. They all need an argument, the character to accent, which can either be given in braces as usual (@'{e}), or, as a special case, the braces can be omitted, in which case the argument is the next character (@'e). This is to make the source as convenient as possible to type and read, since accented characters are very common in some languages.

If the command is alphabetic, such as @dotaccent, then there must be a space between the command name and argument if braces are not used. If the command is non-alphabetic, such as @', then there must not be a space; the argument is the very next character.

Exception: the argument to @tieaccent must be enclosed in braces (since it is two characters instead of one).

To get the true accented characters output in Info, not just the ASCII transliterations, it is necessary to specify @documentencoding with an encoding which supports the required characters (see @documentencoding enc: Set Input Encoding). In this case, you can also use non-ASCII (e.g., pre-accented) characters in the source file.

CommandOutputWhat
@"oöumlaut accent
@'oóacute accent
@,{c}çcedilla accent
@=oōmacron/overbar accent
@^oôcircumflex accent
@`oògrave accent
@~oõtilde accent
@dotaccent{o}ȯoverdot accent
@H{o}őlong Hungarian umlaut
@ogonek{a}ąogonek
@ringaccent{o}o*ring accent
@tieaccent{oo}oo[tie-after accent
@u{o}ŏbreve accent
@ubaraccent{o}o_underbar accent
@udotaccent{o}underdot accent
@v{o}ǒcaron/hacek/check accent

This table lists the Texinfo commands for inserting other characters commonly used in languages other than English.

@exclamdown{}¡upside-down !
@questiondown{}¿upside-down ?
@aa{} @AA{}å Åa,A with circle
@ae{} @AE{}æ Æae,AE ligatures
@dh{} @DH{}ð ÐIcelandic eth
@dotless{i}idotless i
@dotless{j}jdotless j
@l{} @L{}ł Łsuppressed-L,l
@o{} @O{}ø ØO,o with slash
@oe{} @OE{}œ Œoe,OE ligatures
@ordf{} @ordm{}ª ºSpanish ordinals
@ss{}ßes-zet or sharp S
@th{} @TH{}þ ÞIcelandic thorn

11.5 Inserting Quotation Marks

Use doubled single-quote characters to begin and end quotations: ``…''. TeX converts two single quotes to left- and right-hand doubled quotation marks, and Info converts doubled single-quote characters to ASCII double quotes: ``…'' becomes "…".

You may occasionally need to produce two consecutive single quotes; for example, in documenting a computer language such as Maxima where '' is a valid command. You can do this with the input '@w{}'; the empty @w command stops the combination into the double-quote characters.

The left quote character (`, ASCII code 96) used in Texinfo is a grave accent in ANSI and ISO character set standards. We use it as a quote character because that is how TeX is set up, by default.

Texinfo supports several other quotation marks used in languages other than English. Below is a table with the commands Texinfo provides for inserting quotation marks.

In order to get the symbols for the quotation marks in encoded Info output, it is necessary to specify @documentencoding UTF-8. (See @documentencoding enc: Set Input Encoding.) Double guillemets are also present in ISO 8859-1 (aka Latin 1) and ISO 8859-15 (aka Latin 9).

The standard TeX fonts support the usual quotation marks used in English (the ones produced with single and doubled ASCII single quotes). For the other quotation marks, TeX uses European Computer Modern (EC) fonts (ecrm1000 and other variants). These fonts are freely available, of course; you can download them from http://ctan.org/pkg/ec, among other places.

The free EC fonts are bitmap fonts created with Metafont. Especially for on-line viewing, Type 1 (vector) versions of the fonts are preferable; these are available in the CM-Super font package (http://ctan.org/pkg/cm-super).

Both distributions include installation instructions.

CommandGlyphUnicode name (point)
@quotedblleft{} ``Left double quotation mark (U+201C)
@quotedblright{} ''Right double quotation mark (U+201D)
@quoteleft{} `Left single quotation mark (U+2018)
@quoteright{} 'Right single quotation mark (U+2019)
@quotedblbase{}Double low-9 quotation mark (U+201E)
@quotesinglbase{}Single low-9 quotation mark (U+201A)
@guillemetleft{}«Left-pointing double angle quotation mark (U+00AB)
@guillemetright{}»Right-pointing double angle quotation mark (U+00BB)
@guilsinglleft{}Single left-pointing angle quotation mark (U+2039)
@guilsinglright{}Single right-pointing angle quotation mark (U+203A)

For the double angle quotation marks, Adobe and LaTeX glyph names are also supported: @guillemotleft and @guillemotright. These names are incorrect; a “guillemot” is a bird species (a type of auk).

Traditions for quotation mark usage vary to a great extent between languages (http://en.wikipedia.org/wiki/Quotation_mark). Texinfo does not provide commands or configurations for typesetting quotation marks according to the numerous traditions. Therefore, you have to choose the commands appropriate for the language of your manual. Sometimes aliases (see @alias new=existing) can simplify the usage and make the source code more readable. For example, in German, @quotedblbase is used for the left double quote, and the right double quote is the glyph produced by @quotedblleft, which is counterintuitive. Thus, in this case the following aliases would be convenient:

@alias lgqq = quotedblbase
@alias rgqq = quotedblleft

11.6 @sub and @sup: Inserting Subscripts and Superscripts

You can insert subscripts and superscripts with the @sub and @sup commands. For example:

here@sub{below}@sup{above}

produces:

herebelowabove

In Info and plain text, @sub{text} is output as ‘_{text}’ and @sup{text} as ‘^{text}’, including the literal braces (to mark the beginning and end of the “script” text to the reader).

When the output format (and display program) permit (TeX, HTML), the superscript is set above the subscript when both commands are given consecutively.

For subscripts and superscripts in mathematical expressions, it is better to use TeX’s ‘_’ and ‘^’ characters. See the next section.


11.7 @math and @displaymath: Formatting Mathematics

You can write a short mathematical expression with the @math command. Write the mathematical expression between braces, like this:

@math{\sin 2\pi \equiv \cos 3\pi}

which looks like this:

\(\sin 2\pi \equiv \cos 3\pi\)

The @math command has no special effect on the Info output or (by default) the HTML output, merely outputting the contents verbatim.

For the TeX output, @math switches into math mode. This allows you to use all the plain TeX math control sequences for symbols, functions, and so on, starting with ‘\’. It’s best to use ‘\’ instead of ‘@’ for any such mathematical commands; otherwise, texi2any will complain.

By default, the HTML output is only enclosed by <em>. texi2any provides three options for displaying properly formatted mathematics for HTML. You can select these with the HTML_MATH variable (see HTML Customization Variables). With HTML_MATH set to ‘l2h’, texi2any attempts to use the latex2html program to produce image files for mathematical material. With the ‘t4h’ setting, texi2any attempts to use the tex4ht program. With the ‘mathjax’ setting, texi2any inserts references in the output files to MathJax scripts to format the material. The MathJax option requires JavaScript to be enabled in the browser to work. See also MathJax Customization Variables and latex2html Customization Variables.

For displayed equations, you can use the @displaymath command. Example:

@displaymath
f(x) = {1\over\sigma\sqrt{2\pi}}
e^{-{1\over2}\left({x-\mu\over\sigma}\right)^2}
@end displaymath

which produces:

\[f(x) = {1\over\sigma\sqrt{2\pi}} e^{-{1\over2}\left({x-\mu\over\sigma}\right)^2} \]

Although @sub and @sup may work inside math mode in some contexts, it is better to use TeX’s ‘_’ and ‘^’ characters to denote subscripts and superscripts within mathematical expressions. In general, the contents of @math or @displaymath should be plain TeX only, with no interspersed Texinfo commands.

Due to the conflict with Texinfo’s @sup command, you can access the plain TeX command \sup as \mathopsup instead, in the unlikely occurrence that you want to do this (but only when processing with TeX, not with any of the HTML_MATH options).


11.8 Glyphs for Text

Texinfo has support for a few additional glyphs that are commonly used in printed text but not available in ASCII. Of course, there are many thousands more. It is possible to use Unicode characters as-is as far as makeinfo is concerned, but TeX is not so lucky.


11.8.1 @TeX{} (TeX) and @LaTeX{} (LaTeX)

Use the @TeX{} command to generate ‘TeX’. In a printed manual, this is a special logo that is different from three ordinary letters. In Info, it just looks like ‘TeX’.

Similarly, use the @LaTeX{} command to generate ‘LaTeX’, which is even more special in printed manuals (and different from the incorrect La@TeX{}. In Info, the result is just ‘LaTeX’. (LaTeX is another macro package built on top of TeX, very loosely analogous to Texinfo in that it emphasizes logical structure, but much (much) larger.)

The spelling of these commands is unusual for Texinfo, in that they use both uppercase and lowercase letters.


11.8.2 @copyright{} (©)

Use the @copyright{} command to generate the copyright symbol, ‘©’. Where possible, this is a ‘c’ inside a circle; in Info, this is ‘(C)’.

Legally, it’s not necessary to use the copyright symbol; the English word ‘Copyright’ suffices, according to international treaty.


11.8.3 @registeredsymbol{} (®)

Use the @registeredsymbol{} command to generate the registered symbol, ‘®’. Where possible, this is an ‘R’ inside a circle; in Info, this is ‘(R)’.


11.8.4 @dots (…) and @enddots (...)

An ellipsis (a sequence of dots) would be spaced wrong when typeset as a string of periods, so a special command is used in Texinfo: use the @dots{} command to generate a normal ellipsis, which is three dots in a row, appropriately spaced … like so. To emphasize: do not simply write three periods in the input file; that would work for the Info file output, but would produce the wrong amount of space between the periods in the printed manual.

The @enddots{} command generates an end-of-sentence ellipsis, which also has three dots, but with different spacing afterwards, ... Look closely to see the difference.

Here is an ellipsis: … Here are three periods in a row: ...

In printed (and usually HTML) output, the three periods in a row are much closer together than the dots in the ellipsis.


11.8.5 @bullet (•)

Use the @bullet{} command to generate a large round dot, or the closest possible thing to one. In Info, an asterisk is used. Here is a bullet: •

When you use @bullet in @itemize, you do not need to type the braces, because @itemize supplies them. (see @itemize: Making an Itemized List).


11.8.6 @euro (€): Euro Currency Symbol

Use the @euro{} command to generate ‘€’. Where possible, this is the symbol for the Euro currency. Otherwise, the word ‘Euro’ is used.

Texinfo cannot magically synthesize support for the Euro symbol where the underlying system (fonts, software, whatever) does not support it. Therefore, you may find it preferable to use the word “Euro”. (In banking contexts, the abbreviation for the Euro is EUR.)

In order to get the Euro symbol in encoded Info output, for example, it is necessary to specify @documentencoding ISO-8859-15 or @documentencoding UTF-8 (See @documentencoding enc: Set Input Encoding.) The Euro symbol is in ISO 8859-15 (aka Latin 9), and is not in the more widely-used ISO 8859-1 (Latin 1).

The Euro symbol does not exist in the standard TeX fonts (which were designed before the Euro was legislated into existence). Therefore, TeX uses an additional font, named feymr10 (along with other variables). It is freely available, of course; you can download it from http://ctan.org/pkg/eurosym, among other places. The distribution includes installation instructions.


11.8.7 @pounds (£): Pounds Sterling

Use the @pounds{} command to generate ‘£’. Where possible, this is the symbol for the pounds sterling British currency. Otherwise, it is ‘#’.


11.8.8 @textdegree (°): Degrees Symbol

Use the @textdegree{} command to generate ‘°’. Where possible, this is the normal symbol for degrees. Otherwise, it is an ‘o’.


11.8.9 @minus (-): Inserting a Minus Sign

Use the @minus{} command to generate a minus sign. In a fixed-width font, this is a single hyphen, but in a proportional font, the symbol is the customary length for a minus sign—a little longer than a hyphen, shorter than an em-dash:

-’ is a minus sign generated with ‘@minus{}’,

‘-’ is a hyphen generated with the character ‘-’,

‘—’ is an em-dash for text.

In the fixed-width font used by Info, @minus{} is the same as a hyphen.

You should not use @minus{} inside @code or @example because the width distinction is not made in the fixed-width font they use.

When you use @minus to specify the mark beginning each entry in an itemized list, you do not need to type the braces (see @itemize: Making an Itemized List).

If you actually want to typeset some math that does a subtraction, it is better to use @math. Then the regular ‘-’ character produces a minus sign, as in @math{a-b} (see @math and @displaymath: Formatting Mathematics).


11.8.10 @geq (≥) and @leq (≤): Inserting Relations

Use the @geq{} and @leq{} commands to generate greater-than-or-equal and less-than-equal-signs, ‘≥’ and ‘≤’. When those symbols are not available, the ASCII sequences ‘>=’ and ‘<=’ are output.


11.9 Glyphs for Programming

In Texinfo, code is often illustrated in examples that are delimited by @example and @end example, or by @lisp and @end lisp. In such examples, you can indicate the results of evaluation or an expansion using ‘’ or ‘’. Likewise, there are commands to insert glyphs to indicate printed output, error messages, equivalence of expressions, the location of point in an editor, and GUI operation sequences.

The glyph-insertion commands do not need to be used within an example, but most often they are. All glyph-insertion commands are followed by empty braces.


11.9.1 Glyphs Summary

Here is a summary of the glyph commands:

@result{} indicates the result of an expression.

@expansion{} indicates the results of a macro expansion.

-|

@print{} indicates printed output.

error→

@error{} indicates the following text is an error message.

@equiv{} indicates the exact equivalence of two forms.

@point{} shows the location of point.

A → B

@clicksequence{A @click{} B} indicates a GUI operation sequence: first A, then clicking B, or choosing B from a menu, or otherwise selecting it.


11.9.2 @result{} (⇒): Result of an Expression

Use the @result{} command to indicate the result of evaluating an expression.

The @result{} command is displayed as ‘’, either a double stemmed arrow or (when that is not available) the ASCII sequence ‘=>’.

Thus, the following,

(cdr '(1 2 3))
    ⇒ (2 3)

may be read as “(cdr '(1 2 3)) evaluates to (2 3)”.


11.9.3 @expansion{} (→): Indicating an Expansion

When an expression is a macro call, it expands into a new expression. You can indicate the result of the expansion with the @expansion{} command.

The @expansion{} command is displayed as ‘’, either a long arrow with a flat base or (when that is not available) the ASCII sequence ‘==>’.

For example, the following

@lisp
(third '(a b c))
    @expansion{} (car (cdr (cdr '(a b c))))
    @result{} c
@end lisp

produces

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

which may be read as:

(third '(a b c)) expands to (car (cdr (cdr '(a b c)))); the result of evaluating the expression is c.

Often, as in this case, an example looks better if the @expansion{} and @result{} commands are indented.


11.9.4 @print{} (-|): Indicating Generated Output

Sometimes an expression will generate output during its execution. You can indicate such displayed output with the @print{} command.

The @print{} command is displayed as ‘-|’, either a horizontal dash butting against a vertical bar or (when that is not available) the ASCII sequence ‘-|’.

In the following example, the printed text is indicated with ‘-|’, and the value of the expression follows on the last line.

(progn (print 'foo) (print 'bar))
    -| foo
    -| bar
    ⇒ bar

In a Texinfo source file, this example is written as follows:

@lisp
(progn (print 'foo) (print 'bar))
    @print{} foo
    @print{} bar
    @result{} bar
@end lisp

11.9.5 @error{} (error→): Indicating an Error Message

A piece of code may cause an error when you evaluate it. You can designate the error message with the @error{} command.

The @error{} command is displayed as ‘error→’, either the word ‘error’ in a box in the printed output, the word error followed by an arrow in other formats or (when no arrow is available) ‘error-->’.

Thus,

@lisp
(+ 23 'x)
@error{} Wrong type argument: integer-or-marker-p, x
@end lisp

produces

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

This indicates that the following error message is printed when you evaluate the expression:

Wrong type argument: integer-or-marker-p, x

The word ‘error→’ itself is not part of the error message.


11.9.6 @equiv{} (≡): Indicating Equivalence

Sometimes two expressions produce identical results. You can indicate the exact equivalence of two forms with the @equiv{} command. The @equiv{} command is displayed as ‘’, either a standard mathematical equivalence sign (three parallel horizontal lines) or (when that is not available) as the ASCII sequence ‘==’.

Thus,

@lisp
(make-sparse-keymap) @equiv{} (list 'keymap)
@end lisp

produces

(make-sparse-keymap) ≡ (list 'keymap)

This indicates that evaluating (make-sparse-keymap) produces identical results to evaluating (list 'keymap).


11.9.7 @point{} (∗): Indicating Point in a Buffer

Sometimes you need to show an example of text in an Emacs buffer. In such examples, the convention is to include the entire contents of the buffer in question between two lines of dashes containing the buffer name.

You can use the ‘@point{}’ command to show the location of point in the text in the buffer. (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 located.)

The @point{} command is displayed as ‘’, either a pointed star or (when that is not available) the ASCII sequence ‘-!-’.

The following example shows the contents of buffer foo before and after evaluating a Lisp command to insert the word changed.

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

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

In a Texinfo source file, the example is written like this:

@example
---------- Buffer: foo ----------
This is the @point{}contents of foo.
---------- Buffer: foo ----------

(insert "changed ")
    @result{} nil
---------- Buffer: foo ----------
This is the changed @point{}contents of foo.
---------- Buffer: foo ----------
@end example

11.9.8 Click Sequences

When documenting graphical interfaces, it is necessary to describe sequences such as ‘Click on ‘File’, then choose ‘Open’, then …’. Texinfo offers commands @clicksequence and click to represent this, typically used like this:

… @clicksequence{File @click{} Open} …

which produces:

… File → Open …

The @click command produces a right arrow by default; this glyph is also available independently via the command @arrow{}.

You can change the glyph produced by @click with the command @clickstyle, which takes a command name as its single argument on the rest of the line, much like @itemize and friends (see @itemize: Making an Itemized List). The command should produce a glyph, and the usual empty braces ‘{}’ are omitted. Here’s an example:

@clickstyle @result
… @clicksequence{File @click{} Open} …

now produces:

… File ⇒ Open …

11.10 Inserting Unicode: @U

The command @U{hex} inserts a representation of the Unicode character U+hex. For example, @U{0132} inserts the Dutch ‘IJ’ ligature (‘IJ’).

The hex value should be at least four hex digits; leading zeros are not added. In general, hex must specify a valid normal Unicode character; e.g., U+10FFFF (the very last code point) is invalid by definition, and thus cannot be inserted this way.

@U is useful for inserting occasional glyphs for which Texinfo has no dedicated command, while allowing the Texinfo source to remain purely 7-bit ASCII for maximum portability.

This command has many limitations—the same limitations as inserting Unicode characters in UTF-8 or another binary form. First and most importantly, TeX knows nothing about most of Unicode. Supporting specific additional glyphs upon request is possible, but it’s not viable for texinfo.tex to support whole additional scripts (Japanese, Urdu, …). The @U command does nothing to change this. If the specified character is not supported in TeX, an error is given. (See @documentencoding enc: Set Input Encoding.)

In HTML, XML, and DocBook, the output from @U is always an entity reference of the form ‘&#xhex;’, as in ‘&#x0132;’ for the example above. This should work even when an HTML document uses some other encoding (say, Latin 1) and the given character is not supported in that encoding.

In Info and plain text, if the output encoding is not UTF-8, the output is the ASCII sequence ‘U+hex’, as in the six ASCII characters ‘U+0132’ for the example above.


12 Forcing and Preventing Breaks

Line and page breaks can sometimes occur in the ‘wrong’ place in one or another form of output. It’s up to you to ensure that text looks right in all the output formats.

For example, in a printed manual, page breaks may occur awkwardly in the middle of an example; to prevent this, you can hold text together using a grouping command that keeps the text from being split across two pages. Conversely, you may want to force a page break where none would normally occur.

You can use the break, break prevention, or pagination commands to fix problematic line and page breaks.


12.1 Break Commands

The break commands create or allow line and paragraph breaks:

@*

Force a line break.

@sp n

Skip n blank lines.

@-

Insert a discretionary hyphen.

@hyphenation{hy-phen-a-ted words}

Define hyphen points in hy-phen-a-ted words.

These commands hold text together on a single line:

@w{text}

Prevent text from being split and hyphenated across two lines.

@tie{}

Insert a normal interword space at which a line break may not occur.

The pagination commands apply only to printed output, since other output formats do not have pages.

@page

Start a new page.

@group

Hold text together that must appear on one page.

@need mils

Start a new page if not enough space on this one.


12.2 @* and @/: Generate and Allow Line Breaks

The @* command forces a line break in all output formats. The @/ command allows a line break (printed manual only).

Here is an example with @*:

This sentence is broken @*into two lines.

produces

This sentence is broken
into two lines.

The @/ command can be useful within long urls or other identifiers where TeX can’t find a good place to break. TeX will automatically break urls at the natural places (see URL Line Breaking), so only use @/ if you need it. @/ has no effect on the other output format.


12.3 @- and @hyphenation: Helping TeX Hyphenate

Although TeX’s hyphenation algorithm is generally pretty good, it does miss useful hyphenation points from time to time. (Or, far more rarely, insert an incorrect hyphenation.) So, for documents with an unusual vocabulary or when fine-tuning for a printed edition, you may wish to help TeX out. Texinfo supports two commands for this:

@-

Insert a discretionary hyphen, i.e., a place where TeX can (but does not have to) hyphenate. This is especially useful when you notice an overfull hbox is due to TeX missing a hyphenation (see Overfull “hboxes”). TeX will not insert any hyphenation points itself into a word containing @-.

@hyphenation{hy-phen-a-ted words}

Tell TeX how to hyphenate hy-phen-a-ted words. As shown, you put a ‘-’ at each hyphenation point. For example:

@hyphenation{man-u-script man-u-scripts}

TeX only uses the specified hyphenation points when the words match exactly, so give all necessary variants, such as plurals.

Info, HTML, and other non-TeX output is not hyphenated, so none of these commands have any effect there.


12.4 @allowcodebreaks: Control Line Breaks in @code

Ordinarily, TeX considers breaking lines at ‘-’ and ‘_’ characters within @code and related commands (see @code{sample-code}), more or less as if they were “empty” hyphenation points.

This is necessary since many manuals, especially for Lisp-family languages, must document very long identifiers. On the other hand, some manuals don’t have this problem, and you may not wish to allow a line break at the underscore in, for example, SIZE_MAX, or even worse, after any of the four underscores in __typeof__.

So Texinfo provides this command:

@allowcodebreaks false

to prevent from breaking at ‘-’ or ‘_’ within @code. You can go back to allowing such breaks with @allowcodebreaks true. Write these commands on lines by themselves.

These commands can be given anywhere in the document. For example, you may have just one problematic paragraph where you need to turn off the breaks, but want them in general, or vice versa.

This command has no effect except in HTML and TeX output.


12.5 @w{text}: Prevent Line Breaks

@w{text} outputs text, while prohibiting line breaks within text.

Thus, you can use @w to produce a non-breakable space, fixed at the width of a normal interword space:

@w{ } @w{ } @w{ } indentation.

produces:

      indentation.

The space from @w{ }, as well as being non-breakable, also will not stretch or shrink. Sometimes that is what you want, for instance if you’re doing manual indenting. However, usually you want a normal interword space that does stretch and shrink (in the printed output); for that, see the @tie command in the next section.

You can also use the @w command to prevent TeX from automatically hyphenating a long name or phrase that happens to fall near the end of a line. makeinfo does not ever hyphenate words.

You can also use @w to avoid unwanted keyword expansion in source control systems. For example, to literally write $Id$ in your document, use @w{$}Id$. This trick isn’t effective in Info or plain text output, though.


12.6 @tie{}: Inserting an Unbreakable Space

The @tie{} command produces a normal interword space at which a line break may not occur. Always write it with following (empty) braces, as usual for commands used within a paragraph. Here’s an example:

@TeX{} was written by Donald E.@tie{}Knuth.

produces:

TeX was written by Donald E. Knuth.

There are two important differences between @tie{} and @w{ }:

  • The space produced by @tie{} will stretch and shrink slightly along with the normal interword spaces in the paragraph; the space produced by @w{ } will not vary.
  • @tie{} allows hyphenation of the surrounding words, while @w{ } inhibits hyphenation of those words (for TeXnical reasons, namely that it produces an ‘\hbox’).

12.7 @sp n: Insert Blank Lines

A line beginning with and containing only @sp n generates n blank lines of space in both the printed manual and the Info file. @sp also forces a paragraph break. For example,

@sp 2

generates two blank lines.

The @sp command is most often used in the title page.


12.8 @page: Start a New Page

A line containing only @page starts a new page in a printed manual. In other formats, without the concept of pages, it starts a new paragraph. A @page command is often used in the @titlepage section of a Texinfo file to start the copyright page.


12.9 @group: Prevent Page Breaks

The @group command (on a line by itself) is used inside an @example or similar construct to begin an unsplittable vertical group, which will appear entirely on one page in the printed output. The group is terminated by a line containing only @end group. These two lines produce no output of their own, and in the Info file output they have no effect at all.

Although @group would make sense conceptually in a wide variety of contexts, its current implementation works reliably only within @example and variants, and within @display, @format, @flushleft and @flushright. See Quotations and Examples. (What all these commands have in common is that each line of input produces a line of output.) In other contexts, @group can cause anomalous vertical spacing.

This formatting requirement means that you should write:

@example
@group
…
@end group
@end example

with the @group and @end group commands inside the @example and @end example commands.

The @group command is most often used to hold an example together on one page. In this Texinfo manual, more than 100 examples contain text that is enclosed between @group and @end group.

If you forget to end a group, you may get strange and unfathomable error messages when you run TeX. This is because TeX keeps trying to put the rest of the Texinfo file onto the one page and does not start to generate error messages until it has processed considerable text. It is a good rule of thumb to look for a missing @end group if you get incomprehensible error messages in TeX.


12.10 @need mils: Prevent Page Breaks

A line containing only @need n starts a new page in a printed manual if fewer than n mils (thousandths of an inch) remain on the current page. Do not use braces around the argument n. The @need command has no effect on other output formats since they are not paginated.

This paragraph is preceded by a @need command that tells TeX to start a new page if fewer than 800 mils (eight-tenths inch) remain on the page. It looks like this:

@need 800
This paragraph is preceded by …

The @need command is useful for preventing orphans: single lines at the bottoms of printed pages.


13 Definition Commands

The @deffn command and the other definition commands enable you to describe functions, variables, macros, commands, user options, special forms and other such artifacts in a uniform format.

In the Info file, a definition causes the entity category—‘Function’, ‘Variable’, or whatever—to appear at the beginning of the first line of the definition, followed by the entity’s name and arguments. In the printed manual, the command causes TeX to print the entity’s name and its arguments on the left margin and print the category next to the right margin. In both output formats, the body of the definition is indented. Also, the name of the entity is entered into the appropriate index: @deffn enters the name into the index of functions, @defvr enters it into the index of variables, and so on (see Predefined Indices).

A manual need not and should not contain more than one definition for a given name. An appendix containing a summary should use @table rather than the definition commands.


13.1 The Template for a Definition

The @deffn command is used for definitions of entities that resemble functions. To write a definition using the @deffn command, write the @deffn command at the beginning of a line and follow it on the same line by the category of the entity, the name of the entity itself, and its arguments (if any). Then write the body of the definition on succeeding lines. (You may embed examples in the body.) Finally, end the definition with an @end deffn command written on a line of its own.

The other definition commands follow the same format: a line with the @def… command and whatever arguments are appropriate for that command; the body of the definition; and a corresponding @end line.

The template for a definition looks like this:

@deffn category name argumentsbody-of-definition
@end deffn

For example,

@deffn Command forward-word count
This command moves point forward @var{count} words
(or backward if @var{count} is negative). …
@end deffn

produces

Command: forward-word count

This command moves point forward count words (or backward if count is negative). …

Capitalize the category name like a title. If the name of the category contains spaces, as in the phrase ‘Interactive Command’, enclose it in braces. For example:

@deffn {Interactive Command} isearch-forward
…
@end deffn

Otherwise, the second word will be mistaken for the name of the entity. As a general rule, when any of the arguments in the heading line except the last one are more than one word, you need to enclose them in braces. This may also be necessary if the text contains commands, for example, ‘{declaraci@'on}’ if you are writing in Spanish.

Some of the definition commands are more general than others. The @deffn command, for example, is the general definition command for functions and the like—for entities that may take arguments. When you use this command, you specify the category to which the entity belongs. Three predefined, specialized variations (@defun, @defmac, and @defspec) specify the category for you: “Function”, “Macro”, and “Special Form” respectively. (In Lisp, a special form is an entity much like a function.) Similarly, the general @defvr command is accompanied by several specialized variations for describing particular kinds of variables.

See A Sample Function Definition, for a detailed example of a function definition, including the use of @example inside the definition.


13.2 Definition Command Continuation Lines

The heading line of a definition command can get very long. Therefore, Texinfo has a special syntax allowing them to be continued over multiple lines of the source file: a lone ‘@’ at the end of each line to be continued. Here’s an example:

@defun fn-name @
  arg1 arg2 arg3
This is the basic continued defun.
@end defun

produces:

Function: fn-name arg1 arg2 arg3

This is the basic continued defun.

As you can see, the continued lines are combined, as if they had been typed on one source line.

Although this example only shows a one-line continuation, continuations may extend over any number of lines, in the same way; put an @ at the end of each line to be continued.

In general, any number of spaces or tabs before the @ continuation character are collapsed into a single space. There is one exception: the Texinfo processors will not fully collapse whitespace around a continuation inside braces. For example:

@deffn {Category @
  Name} …

The output (not shown) has excess space between ‘Category’ and ‘Name’. To avoid this, elide the unwanted whitespace in your input, or put the continuation @ outside braces.

@ does not function as a continuation character in any other context. Ordinarily, ‘@’ followed by a whitespace character (space, tab, newline) produces a normal interword space (see Multiple Spaces).


13.3 Optional and Repeated Arguments

Some entities take optional or repeated arguments, conventionally specified by using square brackets and ellipses: an argument enclosed within square brackets is optional, and an argument followed by an ellipsis is optional and may be repeated more than once.

Thus, [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 in Lisp.

Here is the @defspec line of an example of an imaginary (complicated) special form:

Special Form: foobar (var [from to [inc]]) body…

In this example, 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.

In a Texinfo source file, this @defspec line is written like this:

@defspec foobar (var [from to [inc]]) body@dots{}

The function is listed in the Command and Variable Index under ‘foobar’.


13.4 @deffnx, et al.: Two or More ‘First’ Lines

To create two or more ‘first’ or header lines for a definition, follow the first @deffn line by a line beginning with @deffnx. The @deffnx command works exactly like @deffn except that it does not generate extra vertical white space between it and the preceding line.

For example,

@deffn {Interactive Command} isearch-forward
@deffnx {Interactive Command} isearch-backward
These two search commands are similar except …
@end deffn

produces

Interactive Command: isearch-forward
Interactive Command: isearch-backward

These two search commands are similar except …

Each definition command has an ‘x’ form: @defunx, @defvrx, @deftypefunx, etc.

The ‘x’ forms work similarly to @itemx (see @itemx: Second and Subsequent Items).


13.5 The Definition Commands

Texinfo provides more than a dozen definition commands, all of which are described in this section.

The definition commands automatically enter the name of the entity in the appropriate index: for example, @deffn, @defun, and @defmac enter function names in the index of functions; @defvr and @defvar enter variable names in the index of variables.

Although the examples that follow mostly illustrate Lisp, the commands can be used for other programming languages.


13.5.1 Functions and Similar Entities

This section describes the commands for describing functions and similar entities:

@deffn category name arguments

The @deffn command is the general definition command for functions, interactive commands, and similar entities that may take arguments. You must choose a term to describe the category of entity being defined; for example, “Function” could be used if the entity is a function. The @deffn command is written at the beginning of a line and is followed on the same line by the category of entity being described, the name of this particular entity, and its arguments, if any. Terminate the definition with @end deffn on a line of its own.

For example, here is a definition:

@deffn Command forward-char nchars
Move point forward @var{nchars} characters.
@end deffn

This shows a rather terse definition for a “command” named forward-char with one argument, nchars.

@deffn prints argument names such as nchars in slanted type in the printed output, because we think of these names as metasyntactic variables—they stand for the actual argument values. Within the text of the description, however, write an argument name explicitly with @var to refer to the value of the argument. In the example above, we used ‘@var{nchars}’ in this way.

In the extremely unusual case when an argument name contains ‘--’, or another character sequence which is treated specially (see General Syntactic Conventions), use @code around the special characters. This avoids the conversion to typographic en-dashes and em-dashes.

The template for @deffn is:

@deffn category name argumentsbody-of-definition
@end deffn
@defun name arguments

The @defun command is the definition command for functions. @defun is equivalent to ‘@deffn Function …’. Terminate the definition with @end defun on a line of its own. Thus, the template is:

@defun function-name argumentsbody-of-definition
@end defun
@defmac name arguments

The @defmac command is the definition command for macros. @defmac is equivalent to ‘@deffn Macro …’ and works like @defun.

@defspec name arguments

The @defspec command is the definition command for special forms. (In Lisp, a special form is an entity much like a function; see Special Forms in GNU Emacs Lisp Reference Manual.) @defspec is equivalent to ‘@deffn {Special Form} …’ and works like @defun.

All these commands create entries in the index of functions.


13.5.2 Variables and Similar Entities

Here are the commands for defining variables and similar entities:

@defvr category name

The @defvr command is a general definition command for something like a variable—an entity that records a value. You must choose a term to describe the category of entity being defined; for example, “Variable” could be used if the entity is a variable. Write the @defvr command at the beginning of a line and follow it on the same line by the category of the entity and the name of the entity.

We recommend capitalizing the category name like a title. If the name of the category contains spaces, as in the name “User Option”, enclose it in braces. Otherwise, the second word will be mistaken for the name of the entity. For example,

@defvr {User Option} fill-column
This buffer-local variable specifies
the maximum width of filled lines.
…
@end defvr

Terminate the definition with @end defvr on a line of its own.

The template is:

@defvr category name
body-of-definition
@end defvr

@defvr creates an entry in the index of variables for name.

@defvar name

The @defvar command is the definition command for variables. @defvar is equivalent to ‘@defvr Variable …’.

For example:

@defvar kill-ring
…
@end defvar

The template is:

@defvar name
body-of-definition
@end defvar

@defvar creates an entry in the index of variables for name.

@defopt name

The @defopt command is the definition command for user options, i.e., variables intended for users to change according to taste; Emacs has many such (see Variables in The GNU Emacs Manual). @defopt is equivalent to ‘@defvr {User Option} …’ and works like @defvar. It creates an entry in the index of variables.


13.5.3 Functions in Typed Languages

The @deftypefn command and its variations are for describing functions in languages in which you must declare types of variables and functions, such as C and C++.

@deftypefn category data-type name arguments

The @deftypefn command is the general definition command for functions and similar entities that may take arguments and that are typed. The @deftypefn command is written at the beginning of a line and is followed on the same line by the category of entity being described, the type of the returned value, the name of this particular entity, and its arguments, if any.

For example,

@deftypefn {Library Function} int foobar @
  (int @var{foo}, float @var{bar})
…
@end deftypefn

produces:

Library Function: int foobar (int foo, float bar)

This means that foobar is a “library function” that returns an int, and its arguments are foo (an int) and bar (a float).

Since in typed languages, the actual names of the arguments are typically scattered among data type names and keywords, Texinfo cannot find them without help. You can either (a) write everything as straight text, and it will be printed in slanted type; (b) use @var for the variable names, which will uppercase the variable names in Info and use the slanted typewriter font in printed output; (c) use @var for the variable names and @code for the type names and keywords, which will be dutifully obeyed.

The template for @deftypefn is:

@deftypefn category data-type name argumentsbody-of-description
@end deftypefn

Note that if the category or data type is more than one word then it must be enclosed in braces to make it a single argument.

If you are describing a procedure in a language that has packages, such as Ada, you might consider using @deftypefn in a manner somewhat contrary to the convention described in the preceding paragraphs. For example:

@deftypefn stacks private push @
       (@var{s}:in out stack; @
       @var{n}:in integer)
…
@end deftypefn

(In these examples the @deftypefn arguments are shown using continuations (see Definition Command Continuation Lines), but could be on a single line.)

In this instance, the procedure is classified as belonging to the package stacks rather than classified as a ‘procedure’ and its data type is described as private. (The name of the procedure is push, and its arguments are s and n.)

@deftypefn creates an entry in the index of functions for name.

@deftypefun data-type name arguments

The @deftypefun command is the specialized definition command for functions in typed languages. The command is equivalent to ‘@deftypefn Function …’. The template is:

@deftypefun type name argumentsbody-of-description
@end deftypefun

@deftypefun creates an entry in the index of functions for name.

Ordinarily, the return type is printed on the same line as the function name and arguments, as shown above. In source code, GNU style is to put the return type on a line by itself. So Texinfo provides an option to do that: @deftypefnnewline on.

This affects typed functions only—not untyped functions, not typed variables, etc. Specifically, it affects the commands in this section, and the analogous commands for object-oriented languages, namely @deftypeop and @deftypemethod (see Object-Oriented Methods).

Specifying @deftypefnnewline off reverts to the default.


13.5.4 Variables in Typed Languages

Variables in typed languages are handled in a manner similar to functions in typed languages. See Functions in Typed Languages. The general definition command @deftypevr corresponds to @deftypefn and the specialized definition command @deftypevar corresponds to @deftypefun.

@deftypevr category data-type name

The @deftypevr command is the general definition command for something like a variable in a typed language—an entity that records a value. You must choose a term to describe the category of the entity being defined; for example, “Variable” could be used if the entity is a variable.

The @deftypevr command is written at the beginning of a line and is followed on the same line by the category of the entity being described, the data type, and the name of this particular entity.

For example:

@deftypevr {Global Flag} int enable
…
@end deftypevr

produces the following:

Global Flag: int enable

The template is:

@deftypevr category data-type name
body-of-description
@end deftypevr
@deftypevar data-type name

The @deftypevar command is the specialized definition command for variables in typed languages. @deftypevar is equivalent to ‘@deftypevr Variable …’. The template is:

@deftypevar data-type name
body-of-description
@end deftypevar

These commands create entries in the index of variables.


13.5.5 Data Types

Here is the command for data types:

@deftp category name attributes

The @deftp command is the generic definition command for data types. The command is written at the beginning of a line and is followed on the same line by the category, by the name of the type (which is a word like int or float), and then by names of attributes of objects of that type. Thus, you could use this command for describing int or float, in which case you could use data type as the category. (A data type is a category of certain objects for purposes of deciding which operations can be performed on them.)

In Lisp, for example, pair names a particular data type, and an object of that type has two slots called the CAR and the CDR. Here is how you would write the first line of a definition of pair.

@deftp {Data type} pair car cdr
…
@end deftp

The template is:

@deftp category name-of-type attributesbody-of-definition
@end deftp

@deftp creates an entry in the index of data types.


13.5.6 Object-Oriented Programming

Here are the commands for formatting descriptions about abstract objects, such as are used in object-oriented programming. A class is a defined type of abstract object. An instance of a class is a particular object that has the type of the class. An instance variable is a variable that belongs to the class but for which each instance has its own value.


13.5.6.1 Object-Oriented Variables

These commands allow you to define different sorts of variables in object-oriented programming languages.

@defcv category class name

The @defcv command is the general definition command for variables associated with classes in object-oriented programming. The @defcv command is followed by three arguments: the category of thing being defined, the class to which it belongs, and its name. For instance:

@defcv {Class Option} Window border-pattern
…
@end defcv

produces:

Class Option of Window: border-pattern

@defcv creates an entry in the index of variables.

@deftypecv category class data-type name

The @deftypecv command is the definition command for typed class variables in object-oriented programming. It is analogous to @defcv with the addition of the data-type parameter to specify the type of the instance variable. Ordinarily, the data type is a programming language construct that should be marked with @code. For instance:

@deftypecv {Class Option} Window @code{int} border-pattern
…
@end deftypecv

produces:

Class Option of Window: int border-pattern

@deftypecv creates an entry in the index of variables.

@defivar class name

The @defivar command is the definition command for instance variables in object-oriented programming. @defivar is equivalent to ‘@defcv {Instance Variable} …’. For instance:

@defivar Window border-pattern
…
@end defivar

produces:

Instance Variable of Window: border-pattern

@defivar creates an entry in the index of variables.

@deftypeivar class data-type name

The @deftypeivar command is the definition command for typed instance variables in object-oriented programming. It is analogous to @defivar with the addition of the data-type parameter to specify the type of the instance variable. Ordinarily, the data type is a programming language construct that should be marked with @code. For instance:

@deftypeivar Window @code{int} border-pattern
…
@end deftypeivar

produces:

Instance Variable of Window: int border-pattern

@deftypeivar creates an entry in the index of variables.


13.5.6.2 Object-Oriented Methods

These commands allow you to define different sorts of function-like entities resembling methods in object-oriented programming languages. These entities take arguments, as functions do, but are associated with particular classes of objects.

@defop category class name arguments

The @defop command is the general definition command for these method-like entities.

For example, some systems have constructs called wrappers that are associated with classes as methods are, but that act more like macros than like functions. You could use @defop Wrapper to describe one of these.

Sometimes it is useful to distinguish methods and operations. You can think of an operation as the specification for a method. Thus, a window system might specify that all window classes have a method named expose; we would say that this window system defines an expose operation on windows in general. Typically, the operation has a name and also specifies the pattern of arguments; all methods that implement the operation must accept the same arguments, since applications that use the operation do so without knowing which method will implement it.

Often it makes more sense to document operations than methods. For example, window application developers need to know about the expose operation, but need not be concerned with whether a given class of windows has its own method to implement this operation. To describe this operation, you would write:

@defop Operation windows expose

The @defop command is written at the beginning of a line and is followed on the same line by the overall name of the category of operation, the name of the class of the operation, the name of the operation, and its arguments, if any.

The template is:

@defop category class name argumentsbody-of-definition
@end defop

@defop creates an entry, such as ‘expose on windows’, in the index of functions.

@deftypeop category class data-type name arguments

The @deftypeop command is the definition command for typed operations in object-oriented programming. It is similar to @defop with the addition of the data-type parameter to specify the return type of the method. @deftypeop creates an entry in the index of functions.

@defmethod class name arguments

The @defmethod command is the definition command for methods in object-oriented programming. A method is a kind of function that implements an operation for a particular class of objects and its subclasses.

@defmethod is equivalent to ‘@defop Method …’. The command is written at the beginning of a line and is followed by the name of the class of the method, the name of the method, and its arguments, if any.

For example:

@defmethod bar-class bar-method argument
…
@end defmethod

illustrates the definition for a method called bar-method of the class bar-class. The method takes an argument.

@defmethod creates an entry in the index of functions.

@deftypemethod class data-type name arguments

The @deftypemethod command is the definition command for methods in object-oriented typed languages, such as C++ and Java. It is similar to the @defmethod command with the addition of the data-type parameter to specify the return type of the method. @deftypemethod creates an entry in the index of functions.

The typed commands are affected by the @deftypefnnewline option (see Functions in Typed Languages).


13.6 Conventions for Writing Definitions

When you write a definition using @deffn, @defun, or one of the other definition commands, please take care to use arguments that indicate the meaning, as with the count argument to the forward-word function. Also, if the name of an argument contains the name of a type, such as integer, take care that the argument actually is of that type.


13.7 A Sample Function Definition

A function definition uses the @defun and @end defun commands. The name of the function follows immediately after the @defun command and it is followed, on the same line, by the parameter list.

Here is a definition from Calling Functions in The GNU Emacs Lisp Reference Manual.

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 arguments to give to function, rather than a single argument. We also say that this list is appended to the other arguments.

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)

An interesting example of using apply is found in the description of mapcar.

In the Texinfo source file, this example looks like this:

@defun apply function &rest arguments
@code{apply} calls @var{function} with
@var{arguments}, just like @code{funcall} but with one
difference: the last of @var{arguments} is a list of
arguments to give to @var{function}, rather than a single
argument.  We also say that this list is @dfn{appended}
to the other arguments.

@code{apply} returns the result of calling
@var{function}.  As with @code{funcall},
@var{function} must either be a Lisp function or a
primitive function; special forms and macros do not make
sense in @code{apply}.

@example
(setq f 'list)
    @result{} list
(apply f 'x 'y 'z)
@error{} Wrong type argument: listp, z
(apply '+ 1 2 '(3 4))
    @result{} 10
(apply '+ '(1 2 3 4))
    @result{} 10

(apply 'append '((a b c) nil (x y z) nil))
    @result{} (a b c x y z)
@end example

An interesting example of using @code{apply} is found
in the description of @code{mapcar}.
@end defun

In this manual, this function is listed in the Command and Variable Index under apply.

Ordinary variables and user options are described using a format like that for functions except that variables do not take arguments.


14 Internationalization

Texinfo has some support for writing in languages other than English, although this area still needs considerable work. (If you are the one helping to translate the fixed strings written to documents, see Internationalization of Document Strings.)

For a list of the various accented and special characters Texinfo supports, see Inserting Accents.


14.1 @documentlanguage ll[_cc]: Set the Document Language

The @documentlanguage command declares the current document locale. Write it on a line by itself, near the beginning of the file.

@documentlanguage ll[_cc]

Include a two-letter ISO 639-2 language code (ll) following the command name, optionally followed by an underscore and two-letter ISO 3166 two-letter country code (cc). If you have a multilingual document, the intent is to be able to use this command multiple times, to declare each language change. If the command is not used at all, the default is en_US for US English.

As with GNU Gettext (see Gettext), if the country code is omitted, the main dialect is assumed where possible. For example, de is equivalent to de_DE (German as spoken in Germany).

For Info and other online output, this command changes the translation of various document strings such as “see” in cross-references (see Cross-references), “Function” in defuns (see Definition Commands), and so on. Some strings, such as “Node:”, “Next:”, “Menu:”, etc., are keywords in Info output, so are not translated there; they are translated in other output formats.

For TeX, this command causes a file txi-locale.tex to be read (if it exists). If @documentlanguage argument contains the optional ‘_cc’ suffix, this is tried first. For example, with @documentlanguage de_DE, TeX first looks for txi-de_DE.tex, then txi-de.tex.

Such a txi-* file is intended to redefine the various English words used in TeX output, such as ‘Chapter’, ‘See’, and so on. We are aware that individual words like these cannot always be translated in isolation, and that a very different strategy would be required for ideographic (among other) scripts. Help in improving Texinfo’s language support is welcome.

@documentlanguage also changes TeX’s current hyphenation patterns, if the TeX program being run has the necessary support included. This will generally not be the case for tex itself, but will usually be the case for up-to-date distributions of the extended TeX programs etex (DVI output) and pdftex (PDF output). texi2dvi will use the extended TeXs if they are available (see Format with texi2dvi).

In September 2006, the W3C Internationalization Activity released a new recommendation for specifying languages: http://www.rfc-editor.org/rfc/bcp/bcp47.txt. When Gettext supports this new scheme, Texinfo will too.

Since the lists of language codes and country codes are updated relatively frequently, we don’t attempt to list them here. The valid language codes are on the official home page for ISO 639, http://www.loc.gov/standards/iso639-2/. The country codes and the official web site for ISO 3166 can be found via http://en.wikipedia.org/wiki/ISO_3166.


14.2 @documentencoding enc: Set Input Encoding

The @documentencoding command declares the input document encoding, and can also affect the encoding of the output. Write it on a line by itself, with a valid encoding specification following, near the beginning of the file.

@documentencoding enc

Texinfo supports these encodings:

US-ASCII

This has no particular effect, but it’s included for completeness.

UTF-8

The vast global character encoding, expressed in 8-bit bytes.

ISO-8859-1
ISO-8859-15
ISO-8859-2

These specify the standard encodings for Western European (the first two) and Eastern European languages (the third), respectively. ISO 8859-15 replaces some little-used characters from 8859-1 (e.g., precomposed fractions) with more commonly needed ones, such as the Euro symbol (€).

A full description of the encodings is beyond our scope here; one useful reference is http://czyborra.com/charsets/iso8859.html.

koi8-r

This is the commonly used encoding for the Russian language.

koi8-u

This is the commonly used encoding for the Ukrainian language.

Specifying an encoding enc has the following effects:

In Info output, a so-called ‘Local Variables’ section (see File Variables in The GNU Emacs Manual) is output including enc. This allows Info readers to set the encoding appropriately. It looks like this:

Local Variables:
coding: enc
End:

Also, in Info and plain text output, unless the option --disable-encoding is given to makeinfo, accent constructs and special characters, such as @'e, are output as the actual 8-bit or UTF-8 character in the given encoding where possible.

In HTML output, a ‘<meta>’ tag is output, in the ‘<head>’ section of the HTML, that specifies enc. Web servers and browsers cooperate to use this information so the correct encoding is used to display the page, if supported by the system. That looks like this:

<meta http-equiv="Content-Type" content="text/html;
     charset=enc">

In XML and DocBook output, UTF-8 is always used for the output, according to the conventions of those formats.

In TeX output, the characters which are supported in the standard Computer Modern fonts are output accordingly. For example, this means using constructed accents rather than precomposed glyphs. Using a missing character generates a warning message, as does specifying an unimplemented encoding.

Although modern TeX systems support nearly every script in use in the world, this wide-ranging support is not available in texinfo.tex, and it’s not feasible to duplicate or incorporate all that effort. (Our plan to support other scripts is to create a LaTeX back-end to texi2any, where the support is already present.)

For maximum portability of Texinfo documents across the many different user environments in the world, we recommend sticking to 7-bit ASCII in the input unless your particular manual needs a substantial amount of non-ASCII, e.g., it’s written in German. You can use the @U command to insert an occasional needed character (see Inserting Unicode: @U).


15 Conditionally Visible Text

The conditional commands allow you to use different text for different output formats, or for general conditions that you define. For example, you can use them to specify different text for the printed manual and the Info output.

The conditional commands comprise the following categories.


15.1 Conditional Commands

Texinfo has an @ifformat environment for each output format, to allow conditional inclusion of text for a particular output format.

@ifinfo begins segments of text that should be ignored by TeX when it typesets the printed manual, and by makeinfo when not producing Info output. The segment of text appears only in the Info file and, for historical compatibility, the plain text output.

The environments for the other formats are analogous:

@ifdocbook … @end ifdocbook

Text to appear only in the DocBook output.

@ifhtml … @end ifhtml

Text to appear only in the HTML output.

@ifplaintext … @end ifplaintext

Text to appear only in the plain text output.

@iftex … @end iftex

Text to appear only in the printed manual.

@ifxml … @end ifxml

Text to appear only in the XML output.

The @if… and @end if… commands must appear on lines by themselves in your source file. The newlines following the commands are (more or less) treated as whitespace, so that the conditional text is flowed normally into a surrounding paragraph.

The @if… constructs are intended to conditionalize normal Texinfo source; see Raw Formatter Commands, for using underlying format commands directly.

Here is an example showing all these conditionals:

@iftex
This text will appear only in the printed manual.
@end iftex
@ifinfo
However, this text will appear only in Info and plain text.
@end ifinfo
@ifhtml
And this text will only appear in HTML.
@end ifhtml
@ifplaintext
Whereas this text will only appear in plain text.
@end ifplaintext
@ifxml
Notwithstanding that this will only appear in XML.
@end ifxml
@ifdocbook
Nevertheless, this will only appear in DocBook.
@end ifdocbook

The preceding example produces the following line:

And this text will only appear in HTML.

Notice that you only see one of the input lines, depending on which version of the manual you are reading.

In complex documents, you may want Texinfo to issue an error message in some conditionals that should not ever be processed. The @errormsg{text} command will do this; it takes one argument, the text of the error message.

We mention @errormsg{} here even though it is not strictly related to conditionals, since in practice it is most likely to be useful in that context. Technically, it can be used anywhere. See External Macro Processors: Line Directives, for a caveat regarding the line numbers which @errormsg emits in TeX.


15.2 Conditional Not Commands

You can specify text to be included in any output format other than a given one with the @ifnot… environments:

@ifnotdocbook … @end ifnotdocbook
@ifnothtml … @end ifnothtml
@ifnotinfo … @end ifnotinfo
@ifnotplaintext … @end ifnotplaintext
@ifnottex … @end ifnottex
@ifnotxml … @end ifnotxml

The @ifnot… command and the @end command must appear on lines by themselves in your actual source file.

If the output file is being made in the given format, the region is ignored. Otherwise, it is included.

There is one exception (for historical compatibility): @ifnotinfo text is omitted for both Info and plain text output, not just Info. To specify text which appears only in Info and not in plain text, use @ifnotplaintext, like this:

@ifinfo
@ifnotplaintext
This will be in Info, but not plain text.
@end ifnotplaintext
@end ifinfo

The regions delimited by these commands are ordinary Texinfo source as with @iftex, not raw formatter source as with @tex (see Raw Formatter Commands).


15.3 Raw Formatter Commands

The @if… conditionals just described must be used only with normal Texinfo source. For instance, most features of plain TeX will not work within @iftex. The purpose of @if… is to provide conditional processing for Texinfo source, not provide access to underlying formatting features. For that, Texinfo provides so-called raw formatter commands. They should only be used when truly required (most documents do not need them).

The first raw formatter command is @tex. You can enter plain TeX completely, and use ‘\’ in the TeX commands, by delineating a region with the @tex and @end tex commands. All plain TeX commands and category codes are restored within a @tex region. The sole exception is that the @ character still introduces a command, so that @end tex can be recognized. Texinfo processors will not output material in such a region unless TeX output is being produced.

In complex cases, you may wish to define new TeX macros within @tex. You must use \gdef to do this, not \def, because @tex regions are processed in a TeX group. If you need to make several definitions, you may wish to set \globaldefs=1 (its value will be restored to zero as usual when the group ends at @end tex, so it won’t cause problems with the rest of the document).

As an example, here is a displayed equation written in plain TeX:

@tex
$$ \chi^2 = \sum_{i=1}^N
         \left (y_i - (a + b x_i)
         \over \sigma_i\right)^2 $$
@end tex

The output of this example will appear only in a printed manual. If you are reading this in a format not generated by TeX, you will not see the equation that appears in the printed manual.

Analogously, you can use @ifhtml … @end ifhtml to delimit Texinfo source to be included in HTML output only, and @html … @end html for a region of raw HTML.

Likewise, you can use @ifxml … @end ifxml to delimit Texinfo source to be included in XML output only, and @xml … @end xml for a region of raw XML. Regions of raw text in other formats will also be present in the XML output, but with protection of XML characters and within corresponding elements. For example, the raw HTML text:

@html
<br />
@end html

will be included in the XML output as:

<html>
&lt;br /&gt;
</html>

Again likewise, you can use @ifdocbook … @end ifdocbook to delimit Texinfo source to be included in DocBook output only, and @docbook … @end docbook for a region of raw DocBook.

The behavior of newlines in raw regions is unspecified.

In all cases, in raw processing, @ retains the same meaning as in the remainder of the document. Thus, the Texinfo processors must recognize and even execute, to some extent, the contents of the raw regions, regardless of the final output format. Therefore, specifying changes that globally affect the document inside a raw region leads to unpredictable and generally undesirable behavior. For example, using the @kbdinputstyle command inside a raw region is undefined.

The remedy is simple: don’t do that. Use the raw formatter commands for their intended purpose, of providing material directly in the underlying format. When you simply want to give different Texinfo specifications for different output formats, use the @if… conditionals and stay in Texinfo syntax.


15.4 Inline Conditionals: @inline, @inlineifelse, @inlineraw

Texinfo provides a set of conditional commands with arguments given within braces:

@inlinefmt{format, text}

Process the Texinfo text if format output is being generated.

@inlinefmtifelse{format, then-text, else-text}

Process the Texinfo then-text if format output is being generated; otherwise, process else-text.

@inlineraw{format, text}

Similar, but for raw text (see Raw Formatter Commands).

The supported format names are:

docbook  html  info  plaintext  tex  xml

For example,

@inlinefmt{html, @emph{HTML-only text}}

is nearly equivalent to

@ifhtml
@emph{HTML-only text}
@end ifhtml

except that no whitespace is added, as happens in the latter (environment) case.

In these commands, whitespace is ignored after the comma separating the arguments, as usual, but is not ignored at the end of text.

To insert a literal at sign, left brace, or right brace in one of the arguments, you must use the alphabetic commands @atchar{} (see Inserting ‘@’ with @@ and @atchar{}), and @lbracechar{} or @rbracechar{} (see Inserting ‘{ ‘}’ with @{ @} and @l rbracechar{}), or the parsing will become confused.

With @inlinefmtifelse, it is also necessary to use @comma{} to avoid mistaking a ‘,’ in the text for the delimiter. With @inlinefmt and @inlineraw, @comma{} is not required (though it’s fine to use it), since these commands always have exactly two arguments.

For TeX, the processed text cannot contain newline-delimited commands. Text to be ignored (i.e., for non-TeX) can, though.

Two other @inline... conditionals complement the @ifset and @ifclear commands; see the next section.


15.5 Flags: @set, @clear, conditionals, and @value

You can direct the Texinfo formatting commands to format or ignore parts of a Texinfo file with the @set, @clear, @ifset, and @ifclear commands.

Here are brief descriptions of these commands, see the following sections for more details:

@set flag [value]

Set the variable flag, to the optional value if specified.

@clear flag

Undefine the variable flag, whether or not it was previously defined.

@ifset flag

If flag is set, text through the next @end ifset command is formatted. If flag is clear, text through the following @end ifset command is ignored.

@inlineifset{flag, text}

Brace-delimited version of @ifset.

@ifclear flag

If flag is set, text through the next @end ifclear command is ignored. If flag is clear, text through the following @end ifclear command is formatted.

@inlineifclear{flag, text}

Brace-delimited version of @ifclear.


15.5.1 @set and @value

You use the @set command to specify a value for a flag, which is later expanded by the @value command.

A flag (aka variable) name is an identifier starting with an alphanumeric, ‘-’, or ‘_’. Subsequent characters, if any, may not be whitespace, ‘@’, braces, angle brackets, or any of ‘~`^+|’; other characters, such as ‘%’, may work. However, it is best to use only letters and numerals in a flag name, not ‘-’ or ‘_’ or others—they will work in some contexts, but not all, due to limitations in TeX.

The value is the remainder of the input line, and can contain anything. However, unlike most other commands which take the rest of the line as a value, @set need not appear at the beginning of a line.

Write the @set command like this:

@set foo This is a string.

This sets the value of the flag foo to “This is a string.”.

The Texinfo formatters then replace a @value{flag} command with the string to which flag is set. Thus, when foo is set as shown above, the Texinfo formatters convert this:

@value{foo}
to this:
This is a string.

You can write a @value command within a paragraph; but you must write a @set command on a line of its own.

If you write the @set command like this:

@set foo

without specifying a string, the value of foo is the empty string.

If you clear a previously set flag with @clear flag, a subsequent @value{flag} command will report an error.

For example, if you set foo as follows:

@set howmuch very, very, very

then the formatters transform

It is a @value{howmuch} wet day.
into
It is a very, very, very wet day.

If you write

@clear howmuch

then the formatters transform

It is a @value{howmuch} wet day.
into
It is a {No value for "howmuch"} wet day.

@value cannot be reliably used as the argument to an accent command (see Inserting Accents). For example, this fails:

@set myletter a
@'@value{myletter}    

15.5.2 @ifset and @ifclear

When a flag is set, the Texinfo formatting commands format text between subsequent pairs of @ifset flag and @end ifset commands. When the flag is cleared, the Texinfo formatting commands do not format the text. @ifclear operates analogously.

Write the conditionally formatted text between @ifset flag and @end ifset commands, like this:

@ifset flag
conditional-text
@end ifset

For example, you can create one document that has two variants, such as a manual for a ‘large’ and ‘small’ model:

You can use this machine to dig up shrubs
without hurting them.

@set large

@ifset large
It can also dig up fully grown trees.
@end ifset

Remember to replant promptly …

In the example, the formatting commands will format the text between @ifset large and @end ifset because the large flag is set.

When flag is cleared, the Texinfo formatting commands do not format the text between @ifset flag and @end ifset; that text is ignored and does not appear in either printed or Info output.

For example, if you clear the flag of the preceding example by writing an @clear large command after the @set large command (but before the conditional text), then the Texinfo formatting commands ignore the text between the @ifset large and @end ifset commands. In the formatted output, that text does not appear; in both printed and Info output, you see only the lines that say, “You can use this machine to dig up shrubs without hurting them. Remember to replant promptly …”.

If a flag is cleared with a @clear flag command, then the formatting commands format text between subsequent pairs of @ifclear and @end ifclear commands. But if the flag is set with @set flag, then the formatting commands do not format text between an @ifclear and an @end ifclear command; rather, they ignore that text. An @ifclear command looks like this:

@ifclear flag

15.5.3 @inlineifset and @inlineifclear

@inlineifset and @inlineifclear provide brace-delimited alternatives to the @ifset and @ifclear forms, similar to the other @inline... Commands (see Inline Conditionals: @inline, @inlineifelse, @inlineraw). The same caveats about argument parsing given there apply here too.

@inlineifset{var, text}

Process the Texinfo text if the flag var is defined.

@inlineifclear{var, text}

Process the Texinfo text if the flag var is not defined.

Except for the syntax, their general behavior and purposes is the same as with @ifset and @ifclear, described in the previous section.


15.5.4 @value Example

You can use the @value command to minimize the number of places you need to change when you record an update to a manual. See GNU Sample Texts, for the full text of an example of using this to work with Automake distributions.

This example is adapted from The GNU Make Manual.

  1. Set the flags:
    @set EDITION 0.35 Beta
    @set VERSION 3.63 Beta
    @set UPDATED 14 August 1992
    @set UPDATE-MONTH August 1992
    
  2. Write text for the @copying section (see @copying: Declare Copying Permissions):
    @copying
    This is Edition @value{EDITION},
    last updated @value{UPDATED},
    of @cite{The GNU Make Manual},
    for @code{make}, version @value{VERSION}.
    
    Copyright …
    
    Permission is granted …
    @end copying
    
  3. Write text for the title page, for people reading the printed manual:
    @titlepage
    @title GNU Make
    @subtitle A Program for Directing Recompilation
    @subtitle Edition @value{EDITION}, …
    @subtitle @value{UPDATE-MONTH}
    @page
    @insertcopying
    …
    @end titlepage
    

    (On a printed cover, a date listing the month and the year looks less fussy than a date listing the day as well as the month and year.)

  4. Write text for the Top node, for people reading the Info file:
    @ifnottex
    @node Top
    @top Make
    
    This is Edition @value{EDITION},
    last updated @value{UPDATED},
    of @cite{The GNU Make Manual},
    for @code{make}, version @value{VERSION}.
    @end ifnottex
    

    After you format the manual, the @value constructs have been expanded, so the output contains text like this:

    This is Edition 0.35 Beta, last updated 14 August 1992,
    of `The GNU Make Manual', for `make', Version 3.63 Beta.
    

When you update the manual, you change only the values of the flags; you do not need to edit the three sections.


15.6 Testing for Texinfo Commands: @ifcommanddefined, @ifcommandnotdefined

Occasionally, you may want to arrange for your manual to test if a given Texinfo command is available and (presumably) do some sort of fallback formatting if not. There are conditionals @ifcommanddefined and @ifcommandnotdefined to do this. For example:

@ifcommanddefined node
Good, @samp{@@node} is defined.
@end ifcommanddefined

will output the expected ‘Good, ‘@node’ is defined.’.

This conditional will also consider any new commands defined by the document via @macro, @alias, @definfoenclose, and @def(code)index (see Defining New Texinfo Commands) to be true. Caveat: the TeX implementation reports internal TeX commands, in addition to all the Texinfo commands, as being “defined”; the makeinfo implementation is reliable in this regard, however.

You can check the NEWS file in the Texinfo source distribution and linked from the Texinfo home page (http://www.gnu.org/software/texinfo) to see when a particular command was added.

These command-checking conditionals themselves were added in Texinfo 5.0, released in 2013—decades after Texinfo’s inception. In order to test if they themselves are available, the predefined flag txicommandconditionals can be tested, like this:

@ifset txicommandconditionals
@ifcommandnotdefined foobarnode
(Good, @samp{@@foobarnode} is not defined.)
@end ifcommandnotdefined
@end ifset

Since flags (see the previous section) were added early in the existence of Texinfo, there is no problem with assuming they are available.

We recommend avoiding these tests whenever possible—which is usually the case. For many software packages, it is reasonable for all developers to have a given version of Texinfo (or newer) installed, and thus no reason to worry about older versions. (It is straightforward for anyone to download and install the Texinfo source; it does not have any problematic dependencies.)

The issue of Texinfo versions does not generally arise for end users. With properly distributed packages, users need not process the Texinfo manual simply to build and install the package; they can use preformatted Info (or other) output files. This is desirable in general, to avoid unnecessary dependencies between packages (see Releases in GNU Coding Standards).


15.7 Conditional Nesting

Conditionals can be nested; however, the details are a little tricky. The difficulty comes with failing conditionals, such as @ifhtml when HTML is not being produced, where the included text is to be ignored. However, it is not to be completely ignored, since it is useful to have one @ifset inside another, for example—that is a way to include text only if two conditions are met. Here’s an example:

@ifset somevar
@ifset anothervar
Both somevar and anothervar are set.
@end ifset
@ifclear anothervar
Somevar is set, anothervar is not.
@end ifclear
@end ifset

Technically, Texinfo requires that for a failing conditional, the ignored text must be properly nested with respect to that failing conditional. Unfortunately, it’s not always feasible to check that all conditionals are properly nested, because then the processors could have to fully interpret the ignored text, which defeats the purpose of the command. Here’s an example illustrating these rules:

@ifset a
@ifset b
@ifclear ok  - ok, ignored
@end junky   - ok, ignored
@end ifset
@c WRONG - missing @end ifset.

Finally, as mentioned above, all conditional commands must be on lines by themselves, with no text (even spaces) before or after. Otherwise, the processors cannot reliably determine which commands to consider for nesting purposes.


16 Defining New Texinfo Commands

Texinfo provides several ways to define new commands (in all cases, it’s not recommended to try redefining existing commands):

Most generally of all (not just for defining new commands), it is possible to invoke any external macro processor and have Texinfo recognize so-called #line directives for error reporting.

If you want to do simple text substitution, @set and @value is the simplest approach (see Flags: @set, @clear, conditionals, and @value).


16.1 Defining Macros

You use the Texinfo @macro command to define a macro, like this:

@macro macroname{param1, param2, …}
text … \param1\ …
@end macro

The parameters param1, param2, … correspond to arguments supplied when the macro is subsequently used in the document (described in the next section).

For a macro to work consistently with TeX, macroname must consist entirely of letters: no digits, hyphens, underscores, or other special characters. So, we recommend using only letters. However, makeinfo will accept anything consisting of alphanumerics, and (except as the first character) ‘-’. The ‘_’ character is excluded so that macros can be called inside @math without a following space (see @math and @displaymath: Formatting Mathematics).

If a macro needs no parameters, you can define it either with an empty list (‘@macro foo {}’) or with no braces at all (‘@macro foo’).

The definition or body of the macro can contain most Texinfo commands, including macro invocations. However, a macro definition that defines another macro does not work in TeX due to limitations in the design of @macro.

In the macro body, instances of a parameter name surrounded by backslashes, as in ‘\param1\’ in the example above, are replaced by the corresponding argument from the macro invocation. You can use parameter names any number of times in the body, including zero.

To get a single ‘\’ in the macro expansion, use ‘\\’. Any other use of ‘\’ in the body yields a warning.

The newline characters after the @macro line and before the @end macro line are ignored, that is, not included in the macro body. All other whitespace is treated according to the usual Texinfo rules.

To allow a macro to be used recursively, that is, in an argument to a call to itself, you must define it with ‘@rmacro’, like this:

@rmacro rmac {arg}
a\arg\b
@end rmacro
…
@rmac{1@rmac{text}2}

This produces the output ‘a1atextb2b’. With ‘@macro’ instead of ‘@rmacro’, an error message is given.

You can undefine a macro foo with @unmacro foo. It is not an error to undefine a macro that is already undefined. For example:

@unmacro foo

16.2 Invoking Macros

After a macro is defined (see the previous section), you can invoke (use) it in your document like this:

@macroname {arg1, arg2, …}

and the result will be more or less as if you typed the body of macroname at that spot. For example:

@macro foo {p, q}
Together: \p\ & \q\.
@end macro
@foo{a, b}

produces:

Together: a & b.

Thus, the arguments and parameters are separated by commas and delimited by braces; any whitespace after (but not before) a comma is ignored. The braces are required in the invocation even when the macro takes no arguments, consistent with other Texinfo commands. For example:

@macro argless {}
No arguments here.
@end macro
@argless{}

produces:

No arguments here.

Passing macro arguments containing commas requires care, since commas also separate the arguments. To include a comma character in an argument, the most reliable method is to use the @comma{} command. For makeinfo, you can also prepend a backslash character, as in ‘\,’, but this does not work with TeX.

It’s not always necessary to worry about commas. To facilitate use of macros, makeinfo implements two rules for automatic quoting in some circumstances:

  1. If a macro takes only one argument, all commas in its invocation are quoted by default. For example:
    @macro TRYME{text}
    @strong{TRYME: \text\}
    @end macro
    
    @TRYME{A nice feature, though it can be dangerous.}
    

    will produce the following output

    TRYME: A nice feature, though it can be dangerous.
    

    And indeed, it can. Namely, makeinfo does not control the number of arguments passed to one-argument macros, so be careful when you invoke them.

  2. If a macro invocation includes another command (including a recursive invocation of itself), any commas in the nested command invocation(s) are quoted by default. For example, in
    @say{@strong{Yes, I do}, person one}
    

    the comma after ‘Yes’ is implicitly quoted. Here’s another example, with a recursive macro:

    @rmacro cat{a,b}
    \a\\b\
    @end rmacro
    
    @cat{@cat{foo, bar}, baz}
    

    will produce the string ‘foobarbaz’.

  3. Otherwise, a comma should be explicitly quoted, as above, for it to be treated as a part of an argument.

The backslash itself can be quoted in macro arguments with another backslash. For example:

@macname {\\bleh}

will pass the argument ‘\bleh’ to macname.

makeinfo also recognizes ‘\{’ and ‘\}’ sequences for curly braces, but these are not recognized by the implementation in TeX. There should, however, rarely be a need for these, as they are only needed when a macro argument contains unbalanced braces.

If a macro is defined to take exactly one argument, it can be invoked without any braces, taking all of the line after the macro name as the argument. For example:

@macro bar {p}
Twice: \p\ & \p\.
@end macro
@bar aah

produces:

Twice: aah & aah.

In these arguments, there is no escaping of special characters, so each ‘\’ stands for itself.

If a macro is defined to take more than one argument, but is called with only one (in braces), the remaining arguments are set to the empty string, and no error is given. For example:

@macro addtwo {p, q}
Both: \p\\q\.
@end macro
@addtwo{a}

produces simply:

Both: a.

16.3 Macro Details and Caveats

By design, macro expansion does not happen in the following contexts in makeinfo:

  • @macro and @unmacro lines;
  • @if... lines, including @ifset and similar;
  • @set, @clear, @value;
  • @clickstyle lines;
  • @end lines.

Unfortunately, TeX may do some expansion in these situations, possibly yielding errors.

Also, quite a few macro-related constructs cause problems with TeX; some of the caveats are listed below. Thus, if you get macro-related errors when producing the printed version of a manual, you might try expanding the macros with makeinfo by invoking texi2dvi with the ‘-E’ option (see Format with texi2dvi). Or, more reliably, eschew Texinfo macros altogether and use a language designed for macro processing, such as M4 (see External Macro Processors: Line Directives).

  • As mentioned earlier, macro names must consist entirely of letters.
  • It is not advisable to redefine any TeX primitive, plain, or Texinfo command name as a macro. Unfortunately, this is a large and open-ended set of names, and the possible resulting errors are unpredictable.
  • Arguments to macros taking more than one argument cannot cross lines.
  • Macros containing a command which must be on a line by itself, such as a conditional, cannot be invoked in the middle of a line. Similarly, macros containing line-oriented commands or text, such as @example environments, may behave unpredictably in TeX.
  • If you have problems using conditionals within a macro, an alternative is to use separate macro definitions inside conditional blocks. For example, instead of
    @macro Mac
    @iftex
    text for TeX output
    @end iftex
    @ifnottex
    text for not TeX output
    @end ifnottex
    @end macro
    

    you can do the following instead:

    @iftex
    @macro Mac
    text for TeX output
    @end macro
    @end iftex
    
    @ifnottex
    @macro Mac
    text for not TeX output
    @end macro
    @end ifnottex
    
  • Texinfo commands in the expansion of a macro in the text of an index entry may end up being typeset as literal text (including an “@” sign), instead of being interpreted with their intended meaning.
  • White space is ignored at the beginnings of lines.
  • Macros can’t be reliably used in the argument to accent commands (see Inserting Accents).
  • The backslash escape for commas in macro arguments does not work; @comma{} must be used.
  • Likewise, if you want to pass an argument with the Texinfo command @, (to produce a cedilla, see Inserting Accents), you have to use @value or another workaround. Otherwise, the comma may be taken as separating the arguments. For example,
    @macro mactwo{argfirst, argsecond}
    \argfirst\+\argsecond\.
    @end macro
    @set fc Fran@,cois
    @mactwo{@value{fc},}
    

    produces:

    François+.
    
  • Ending a macro body with ‘@c’ may cause text following the macro invocation to be ignored as a comment in makeinfo. This is not the case when processing with TeX. This was often done to “comment out” an unwanted newline at the end of a macro body, but this is not necessary any more, as the final newline before ‘@end macro’ is not included in the macro body anyway.
  • In general, you can’t arbitrarily substitute a macro (or @value) call for Texinfo command arguments, even when the text is the same. Texinfo is not M4 (or even plain TeX). It might work with some commands, it fails with others. Best not to do it at all. For instance, this fails:
    @macro offmacro
    off
    @end macro
    @headings @offmacro
    

    This looks equivalent to @headings off, but for TeXnical reasons, it fails with a mysterious error message (namely, ‘Paragraph ended before @headings was complete’).

  • Macros cannot define macros in the natural way. To do this, you must use conditionals and raw TeX. For example:
    @ifnottex
    @macro ctor {name, arg}
    @macro \name\
    something involving \arg\ somehow
    @end macro
    @end macro
    @end ifnottex
    @tex
    \gdef\ctor#1{\ctorx#1,}
    \gdef\ctorx#1,#2,{\def#1{something involving #2 somehow}}
    @end tex
    

The makeinfo implementation also has the following limitations (by design):

  • @verbatim and macros do not mix; for instance, you can’t start a verbatim block inside a macro and end it outside (see @verbatim: Literal Text). Starting any environment inside a macro and ending it outside may or may not work, for that matter.
  • Macros that completely define macros are ok, but it’s not possible to have incompletely nested macro definitions. That is, @macro and @end macro (likewise for @rmacro) must be correctly paired. For example, you cannot start a macro definition within a macro, and then end that nested definition outside the macro.

In the makeinfo implementation before Texinfo 5.0, ends of lines from expansion of a @macro definition did not end an @-command line-delimited argument (@chapter, @center, etc.). This is no longer the case. For example:

@macro twolines{}
aaa
bbb
@end macro
@center @twolines{}

In the current makeinfo, this is equivalent to:

@center aaa
bbb

with just ‘aaa’ as the argument to @center. In the earlier implementation, it would have been parsed as this:

@center aaa bbb

16.4 ‘@alias new=existing

The ‘@alias’ command defines a new command to be just like an existing one. This is useful for defining additional markup names, thus preserving additional semantic information in the input even though the output result may be the same.

Write the ‘@alias’ command on a line by itself, followed by the new command name, an equals sign, and the existing command name. Whitespace around the equals sign is optional and ignored if present. Thus:

@alias new = existing

For example, if your document contains citations for both books and some other media (movies, for example), you might like to define a macro @moviecite{} that does the same thing as an ordinary @cite{} but conveys the extra semantic information as well. You’d do this as follows:

@alias moviecite = cite

Macros do not always have the same effect as aliases, due to vagaries of argument parsing. Also, aliases are much simpler to define than macros. So the command is not redundant.

Unfortunately, it’s not possible to alias Texinfo environments; for example, @alias lang=example is an error.

Aliases must not be recursive, directly or indirectly.

It is not advisable to redefine any TeX primitive, plain TeX, or Texinfo command name as an alias. Unfortunately this is a very large set of names, and the possible resulting errors from TeX are unpredictable.

makeinfo will accept the same identifiers for aliases as it does for macro names, that is, alphanumerics and (except as the first character) ‘-’.


16.5 @definfoenclose: Customized Highlighting

An @definfoenclose command may be used to define a highlighting command for all the non-TeX output formats. A command defined using @definfoenclose marks text by enclosing it in strings that precede and follow the text.

In practice, there is little use for this command, and we do not recommend you use it. Support for @definfoenclose may be removed in future releases of Texinfo.

Write a @definfoenclose command at the beginning of a line followed by three comma-separated arguments. The first argument to @definfoenclose is the @-command name (without the @); the second argument is the start delimiter string; and the third argument is the end delimiter string. The latter two arguments enclose the highlighted text in the output.

A delimiter string may contain spaces. Neither the start nor end delimiter is required. If you do not want a start delimiter but do want an end delimiter, you must follow the command name with two commas in a row; otherwise, the end delimiter string you intended will naturally be (mis)interpreted as the start delimiter string.

An enclosure command defined this way takes one argument in braces, since it is intended for new markup commands (see Marking Text, Words and Phrases).

For example, you can write:

@definfoenclose phoo,//,\\

near the beginning of a Texinfo file to define @phoo as an Info formatting command that inserts ‘//’ before and ‘\\’ after the argument to @phoo. You can then write @phoo{bar} wherever you want ‘//bar\\’ highlighted in Info.

For TeX formatting, you could write

@iftex
@global@let@phoo=@i
@end iftex

to define @phoo as a command that causes TeX to typeset the argument to @phoo in italics.

Each definition applies to its own formatter: one for TeX, the other for everything else. The raw TeX commands need to be in ‘@iftex’. @definfoenclose command need not be within ‘@ifinfo’ unless you want to use different definitions for different output formats.

Here is another example: write

@definfoenclose headword, , :

near the beginning of the file, to define @headword as an Info formatting command that inserts nothing before and a colon after the argument to @headword.

@definfoenclose’ definitions must not be recursive, directly or indirectly.


16.6 External Macro Processors: Line Directives

Texinfo macros (and its other text substitution facilities) work fine in straightforward cases. If your document needs unusually complex processing, however, their fragility and limitations can be a problem. In this case, you may want to use a different macro processor altogether, such as M4 (see M4) or CPP (see The C Preprocessor).

With one exception, Texinfo does not need to know whether its input is “original” source or preprocessed from some other source file. Therefore, you can arrange your build system to invoke whatever programs you like to handle macro expansion or other preprocessing needs. Texinfo does not offer built-in support for any particular preprocessor, since no one program seemed likely to suffice for the requirements of all documents.

The one exception is line numbers in error messages. In that case, the line number should refer to the original source file, whatever it may be. There’s a well-known mechanism for this: the so-called ‘#line’ directive. Texinfo supports this.


16.6.1 ‘#line’ Directive

An input line such as this:

#line 100 "foo.ptexi"

indicates that the next line was line 100 of the file foo.ptexi, and so that’s what an error message should refer to. Both M4 (see Preprocessor features in GNU M4) and CPP (see Line Control in The C Preprocessor, and Preprocessor Output in The C Preprocessor) can generate such lines.

The makeinfo program recognizes these lines by default, except within @verbatim blocks (see @verbatim: Literal Text). Their recognition can be turned off completely with CPP_LINE_DIRECTIVES (see Other Customization Variables), though there is normally no reason to do so.

For those few programs (M4, CPP, Texinfo) which need to document ‘#line’ directives and therefore have examples which would otherwise match the pattern, the command @hashchar{} can be used (see Inserting ‘#’ with @hashchar{}). The example line above looks like this in the source for this manual:

@hashchar{}line 100 "foo.ptexi"

The @hashchar command was added to Texinfo in 2013. If you don’t want to rely on it, you can also use @set and @value to insert the literal ‘#’:

@set hash #
@value{hash}line 1 "example.c"

Or, if suitable, a @verbatim environment can be used instead of @example. As mentioned above, #line-recognition is disabled inside verbatim blocks.


16.6.2 ‘#line’ and TeX

As mentioned, makeinfo recognizes the ‘#line’ directives described in the previous section. However, texinfo.tex does not and cannot. Therefore, such a line will be incorrectly typeset verbatim if TeX sees it. The solution is to use makeinfo’s macro expansion options before running TeX. There are three approaches:

  • If you run texi2dvi or its variants (see Format with texi2dvi), you can pass -E and texi2dvi will run makeinfo first to expand macros and eliminate ‘#line’.
  • If you run makeinfo or its variants (see texi2any: The Generic Translator for Texinfo), you can specify --no-ifinfo --iftex -E somefile.out, and then give somefile.out to texi2dvi in a separate command.
  • Or you can run makeinfo --dvi --Xopt -E. (Or --pdf instead of --dvi.) makeinfo will then call texi2dvi -E.

One last caveat regarding use with TeX: since the #line directives are not recognized, the line numbers emitted by the @errormsg{} command (see Conditional Commands), or by TeX itself, are the (incorrect) line numbers from the derived file which TeX is reading, rather than the preprocessor-specified line numbers. This is another example of why we recommend running makeinfo for the best diagnostics (see makeinfo Advantages).


16.6.3 ‘#line’ Syntax Details

Syntax details for the ‘#line’ directive: the ‘#’ character can be preceded or followed by whitespace, the word ‘line’ is optional, and the file name can be followed by a whitespace-separated list of integers (these are so-called “flags” output by CPP in some cases). For those who like to know the gory details, the actual (Perl) regular expression which is matched is this:

/^\s*#\s*(line)? (\d+)(( "([^"]+)")(\s+\d+)*)?\s*$/

As far as we’ve been able to tell, the trailing integer flags only occur in conjunction with a file name, so that is reflected in the regular expression.

As an example, the following is a syntactically valid ‘#line’ directive, meaning line 1 of /usr/include/stdio.h:

# 1 "/usr/include/stdio.h" 2 3 4

Unfortunately, the quoted file name (‘"..."’) has to be optional, because M4 (especially) can often generate ‘#line’ directives within a single file. Since the ‘line’ is also optional, the result is that lines might match which you wouldn’t expect, e.g.,

# 1

The possible solutions are described above (see #line’ Directive).


17 Include Files

When a Texinfo processor sees an @include command in a Texinfo file, it processes the contents of the file named by the @include and incorporates them into the output files being created. Include files thus let you keep a single large document as a collection of conveniently small parts.


17.1 How to Use Include Files

To include another file within a Texinfo file, write the @include command at the beginning of a line and follow it on the same line by the name of a file to be included. For example:

@include buffers.texi

@-commands are expanded in file names. The one most likely to be useful is @value (see @set and @value), and even then only in complicated situations.

An included file should simply be a segment of text that you expect to be included as is into the overall or outer Texinfo file; it should not contain the standard beginning and end parts of a Texinfo file. In particular, you should not start an included file with a line saying ‘\input texinfo’; if you do, that text is inserted into the output file literally. Likewise, you should not end an included file with a @bye command; nothing after @bye is formatted.

In the long-ago past, you were required to write an @setfilename line at the beginning of an included file, but no longer. Now, it does not matter whether you write such a line. If an @setfilename line exists in an included file, it is ignored.


17.2 texinfo-multiple-files-update

GNU Emacs Texinfo mode provides the texinfo-multiple-files-update command. This command creates or updates ‘Next’, ‘Previous’, and ‘Up’ pointers of included files as well as those in the outer or overall Texinfo file, and it creates or updates a main menu in the outer file. Depending on whether you call it with optional arguments, the command updates only the pointers in the first @node line of the included files or all of them:

M-x texinfo-multiple-files-update

Called without any arguments:

  • - Create or update the ‘Next’, ‘Previous’, and ‘Up’ pointers of the first @node line in each file included in an outer or overall Texinfo file.
  • - Create or update the ‘Top’ level node pointers of the outer or overall file.
  • - Create or update a main menu in the outer file.
C-u M-x texinfo-multiple-files-update

Called with C-u as a prefix argument:

  • - Create or update pointers in the first @node line in each included file.
  • - Create or update the ‘Top’ level node pointers of the outer file.
  • - Create and insert a master menu in the outer file. The master menu is made from all the menus in all the included files.
C-u 8 M-x texinfo-multiple-files-update

Called with a numeric prefix argument, such as C-u 8:

  • - Create or update all the ‘Next’, ‘Previous’, and ‘Up’ pointers of all the included files.
  • - Create or update all the menus of all the included files.
  • - Create or update the ‘Top’ level node pointers of the outer or overall file.
  • - And then create a master menu in the outer file. This is similar to invoking texinfo-master-menu with an argument when you are working with just one file.

Note the use of the prefix argument in interactive use: with a regular prefix argument, just C-u, the texinfo-multiple-files-update command inserts a master menu; with a numeric prefix argument, such as C-u 8, the command updates every pointer and menu in all the files and then inserts a master menu.


17.3 Include Files Requirements

If you plan to use the texinfo-multiple-files-update command, the outer Texinfo file that lists included files within it should contain nothing but the beginning and end parts of a Texinfo file, and a number of @include commands listing the included files. It should not even include indices, which should be listed in an included file of their own.

Moreover, each of the included files must contain exactly one highest level node (conventionally, @chapter or equivalent), and this node must be the first node in the included file. Furthermore, each of these highest level nodes in each included file must be at the same hierarchical level in the file structure. Usually, each is a @chapter, an @appendix, or an @unnumbered node. Thus, normally, each included file contains one, and only one, chapter or equivalent-level node.

The outer file should contain only one node, the ‘Top’ node. It should not contain any nodes besides the single ‘Top’ node. The texinfo-multiple-files-update command will not process them.


17.4 Sample File with @include

Here is an example of an outer Texinfo file with @include files within it before running texinfo-multiple-files-update, which would insert a main or master menu:

\input texinfo @c -*-texinfo-*-
@settitle Include Example

... See Sample Texinfo Files, for
examples of the rest of the frontmatter ...

@ifnottex
@node Top
@top Include Example
@end ifnottex

@include foo.texinfo
@include bar.texinfo
@include concept-index.texinfo
@bye

An included file, such as foo.texinfo, might look like this:

@node First
@chapter First Chapter

Contents of first chapter …

The full contents of concept-index.texinfo might be as simple as this:

@node Concept Index
@unnumbered Concept Index

@printindex cp

The outer Texinfo source file for The GNU Emacs Lisp Reference Manual is named elisp.texi. This outer file contains a master menu with 417 entries and a list of 41 @include files.


17.5 @verbatiminclude file: Include a File Verbatim

You can include the exact contents of a file in the document with the @verbatiminclude command:

@verbatiminclude filename

The contents of filename is printed in a verbatim environment (see @verbatim: Literal Text). Generally, the file is printed exactly as it is, with all special characters and white space retained. No indentation is added; if you want indentation, enclose the @verbatiminclude within @example (see @example: Example Text).

The name of the file is taken literally, with a single exception: @value{var} references are expanded. This makes it possible to include files in other directories within a distribution, for instance:

@verbatiminclude @value{top_srcdir}/NEWS

(You still have to get top_srcdir defined in the first place.)

For a method on printing the file contents in a smaller font size, see the end of the section on @verbatim.


18 Formatting and Printing Hardcopy

Running the texi2dvi or texi2pdf command is the simplest way to create printable output. These commands are installed as part of the Texinfo package.

In more detail, three major shell commands are used to print formatted output from a Texinfo manual: one converts the Texinfo source into something printable, a second sorts indices, and a third actually prints the formatted document. When you use the shell commands, you can either work directly in the operating system shell or work within a shell inside GNU Emacs (or some other computing environment).

If you are using GNU Emacs, you can use commands provided by Texinfo mode instead of shell commands. In addition to the three commands to format a file, sort the indices, and print the result, Texinfo mode offers key bindings for commands to recenter the output buffer, show the print queue, and delete a job from the print queue.

In the United States, documents are most often printed on 8.5 inch by 11 inch pages (216mm by 280mm); this is the default size. But you can also print for 7 inch by 9.25 inch pages (178mm by 235mm, the @smallbook size; or on A4 or A5 size paper (@afourpaper, @afivepaper). See @smallbook: Printing “Small” Books, and Printing on A4 Paper.

Details are in the following sections.


18.1 Use TeX

The typesetting program called TeX is used to format a Texinfo document for printable output. TeX is a very powerful typesetting program and, when used correctly, does an exceptionally good job. It is not included in the Texinfo package, being a vast suite of software in itself.

TeX is a document formatter that is used by the FSF for its documentation. It is the easiest way to get printed output (e.g., PDF and PostScript) for Texinfo manuals. TeX is freely redistributable, and you can get it over the Internet or on physical media. See http://tug.org/texlive.


18.2 Format with texi2dvi

The texi2dvi program takes care of all the steps for producing a TeX DVI file from a Texinfo document. Similarly, texi2pdf produces a PDF file.

To run texi2dvi or texi2pdf on an input file foo.texi, do this (where ‘prompt$ ’ is your shell prompt):

prompt$ texi2dvi foo.texi
prompt$ texi2pdf foo.texi

As shown in this example, the file names given to texi2dvi and texi2pdf must include any extension, such as ‘.texi’.

For a list of all the options, run ‘texi2dvi --help’. Some of the options are discussed below.

With the --pdf option, texi2dvi produces PDF output instead of DVI (see PDF Output), by running pdftex instead of tex. Alternatively, the command texi2pdf is an abbreviation for running ‘texi2dvi --pdf’. The command pdftexi2dvi is also provided as a convenience for AUC-TeX (see AUC-TeX), as it prefers to merely prepend ‘pdf’ to DVI producing tools to have PDF producing tools.

With the --dvipdf option, texi2dvi produces PDF output by running TeX and then a DVI-to-PDF program: if the DVIPDF environment variable is set, that value is used, else the first program extant among dvipdfmx, dvipdfm, dvipdf, dvi2pdf, dvitopdf. This method generally supports CJK typesetting better than pdftex.

With the --ps option, texi2dvi produces PostScript instead of DVI, by running tex and then dvips (see Dvips). (Or the value of the DVIPS environment variable, if set.)

texi2dvi can also be used to process LaTeX files. Normally texi2dvi is able to guess the input file language by its contents and file name extension; however, if it guesses wrong you can explicitly specify the input language using --language=lang command line option, where lang is either ‘latex’ or ‘texinfo’.

One useful option to texi2dvi is ‘--command=cmd’. This inserts cmd on a line by itself at the start of the file in a temporary copy of the input file, before running TeX. With this, you can specify different printing formats, such as @smallbook (see @smallbook: Printing “Small” Books), @afourpaper (see Printing on A4 Paper), or @pagesizes (see @pagesizes [width][, height]: Custom Page Sizes), without actually changing the document source. (You can also do this on a site-wide basis with texinfo.cnf; see Preparing for TeX).

The option -E (equivalently, -e and --expand) does Texinfo macro expansion using makeinfo instead of the TeX implementation (see Macro Details and Caveats). Each implementation has its own limitations and advantages. If this option is used, no line in the source file may begin with the string @c _texi2dvi or the string @c (_texi2dvi).

texi2dvi takes the --build=mode option to specify where the TeX compilation takes place, and, as a consequence, how auxiliary files are treated. The build mode can also be set using the environment variable TEXI2DVI_BUILD_MODE. The valid values for mode are:

local

Compile in the current directory, leaving all the auxiliary files around. This is the traditional TeX use.

tidy

Compile in a local *.t2d directory, where the auxiliary files are left. Output files are copied back to the original file.

Using the ‘tidy’ mode brings several advantages:

  • - the current directory is not cluttered with plethora of temporary files.
  • - clutter can be even further reduced using --build-dir=dir: all the *.t2d directories are stored there.
  • - clutter can be reduced to zero using, e.g., --build-dir=/tmp/\$USER.t2d or --build-dir=\$HOME/.t2d.
  • - the output file is updated after every successful TeX run, for sake of concurrent visualization of the output. In a ‘local’ build the viewer stops during the whole TeX run.
  • - if the compilation fails, the previous state of the output file is preserved.
  • - PDF and DVI compilation are kept in separate subdirectories preventing any possibility of auxiliary file incompatibility.

On the other hand, because ‘tidy’ compilation takes place in another directory, occasionally TeX won’t be able to find some files (e.g., when using \graphicspath): in that case, use -I to specify the additional directories to consider.

clean

Same as ‘tidy’, but remove the auxiliary directory afterwards. Every compilation therefore requires the full cycle.

texi2dvi will use etex (or pdfetex) if it is available, because it runs faster in some cases, and provides additional tracing information when debugging texinfo.tex. Nevertheless, this extended version of TeX is not required, and the DVI output is identical. (These days, pdftex and pdfetex are exactly the same, but we still run pdfetex to cater to ancient TeX installations.)

texi2dvi attempts to detect auxiliary files output by TeX, either by using the -recorder option, or by scanning for ‘\openout’ in the log file that a run of TeX produces. You may control how texi2dvi does this with the TEXI2DVI_USE_RECORDER environment variable. Valid values are:

yes

use the -recorder option, no checks.

no

scan for ‘\openout’ in the log file, no checks.

yesmaybe

check whether -recorder option is supported, and if yes use it, otherwise check for tracing ‘\openout’ in the log file is supported, and if yes use it, else it is an error.

nomaybe

same as ‘yesmaybe’, except that the ‘\openout’ trace in log file is checked first.

The default is ‘nomaybe’. This environment variable is provided for troubleshooting purposes, and may change or disappear in the future.


18.3 Format with tex/texindex

You can do the basic formatting of a Texinfo file with the shell command tex followed by the name of the Texinfo file. For example:

tex foo.texi

TeX will produce a DVI file as well as several auxiliary files containing information for indices, cross-references, etc. The DVI file (for DeVice Independent file) can be printed on virtually any device, perhaps after a further conversion (see the previous section).

The tex formatting command itself does not sort the indices; it writes an output file of unsorted index data. To generate a printed index after running the tex command, you first need a sorted index to work from. The texindex command sorts indices. (texi2dvi, described in the previous section, runs tex and texindex as necessary.)

tex outputs unsorted index files under names following a standard convention: the name of your main input file with any ‘.texi’ or similar extension replaced by the two letter index name. For example, the raw index output files for the input file foo.texi would be, by default, foo.cp, foo.vr, foo.fn, foo.tp, foo.pg and foo.ky. Those are exactly the arguments to give to texindex.

Instead of specifying all the unsorted index file names explicitly, it’s typical to use ‘??’ as shell wildcards and give the command in this form:

texindex foo.??

This command will run texindex on all the unsorted index files, including any two letter indices that you have defined yourself using @defindex or @defcodeindex. You can safely run ‘texindex foo.??’ even if there are files with two letter extensions that are not index files, such as ‘foo.el’. The texindex command reports but otherwise ignores such files.

For each file specified, texindex generates a sorted index file whose name is made by appending ‘s’ to the input file name; for example, foo.cps is made from foo.cp. The @printindex command looks for a file with that name (see Printing Indices and Menus). TeX does not read the raw index output file, and texindex does not alter it.

After you have sorted the indices, you need to rerun tex on the Texinfo file. This regenerates the output file, this time with up-to-date index entries.

Finally, you may need to run tex one more time, to get the page numbers in the cross-references correct.

To summarize, this is a five-step process. (Alternatively, it’s a one-step process: run texi2dvi; see the previous section.)

  1. Run tex on your Texinfo file. This generates a DVI file (with undefined cross-references and no indices), and the raw index files (with two letter extensions).
  2. Run texindex on the raw index files. This creates the corresponding sorted index files (with three letter extensions).
  3. Run tex again on your Texinfo file. This regenerates the DVI file, this time with indices and defined cross-references, but with page numbers for the cross-references from the previous run, generally incorrect.
  4. Sort the indices again, with texindex.
  5. Run tex one last time. This time the correct page numbers are written for the cross-references.

18.3.1 Formatting Partial Documents

Sometimes you may wish to print a document while you know it is incomplete, or to print just one chapter of a document. In such a case, the usual auxiliary files that TeX creates and warnings TeX gives about undefined cross-references are just nuisances. You can avoid them with the @novalidate command, which you must give before any sectioning or cross-reference commands.

Thus, the beginning of your file would look approximately like this:

\input texinfo
@novalidate
…

@novalidate also turns off validation in makeinfo, just like its --no-validate option (see Invoking texi2any/makeinfo from a Shell).

Furthermore, you need not run texindex each time after you run tex. The tex formatting command simply uses whatever sorted index files happen to exist from a previous use of texindex. If those are out of date, that is usually ok while you are creating or debugging a document.


18.3.2 Details of texindex

In Texinfo version 6, released in 2015, the texindex program was completely reimplemented. The principal functional difference is that index entries beginning with a left brace or right brace (‘{’ resp. ‘}’) can work properly. For example, these simple index entries are processed correctly, including the “index initial” shown in the index:

@cindex @{
@cindex @}
...
@printindex cp

Although not a matter of functionality, readers may be interested to know that the new texindex is a literate program (http://en.wikipedia.org/wiki/Literate_programming) using Texinfo for documentation and (portable) awk for code. A single source file, texindex/ti.twjr in this case, produces the runnable program, a printable document, and an online document.

The system is called TexiWeb Jr. and was created by Arnold Robbins, who also wrote the new texindex. Not coincidentally, he is also the long-time maintainer of gawk (GNU Awk, see The GNU Awk User’s Guide). The file texindex/Makefile.am shows example usage of the system.


18.5 Printing From an Emacs Shell

You can give formatting and printing commands from a shell within GNU Emacs, just like any other shell command. To create a shell within Emacs, type M-x shell (see Shell in The GNU Emacs Manual). In this shell, you can format and print the document. See Format and Print Hardcopy, for details.

You can switch to and from the shell buffer while tex is running and do other editing. If you are formatting a long document on a slow machine, this can be very convenient.

For example, you can use texi2dvi from an Emacs shell. Here is one way to use texi2pdf to format and print Using and Porting GNU CC from a shell within Emacs:

texi2pdf gcc.texi
lpr gcc.pdf

See the next section for more information about formatting and printing in Texinfo mode.


18.6 Formatting and Printing in Texinfo Mode

Texinfo mode provides several predefined key commands for TeX formatting and printing. These include commands for sorting indices, looking at the printer queue, killing the formatting job, and recentering the display of the buffer in which the operations occur.

C-c C-t C-b
M-x texinfo-tex-buffer

Run texi2dvi on the current buffer.

C-c C-t C-r
M-x texinfo-tex-region

Run TeX on the current region.

C-c C-t C-i
M-x texinfo-texindex

Sort the indices of a Texinfo file formatted with texinfo-tex-region.

C-c C-t C-p
M-x texinfo-tex-print

Print a DVI file that was made with texinfo-tex-region or texinfo-tex-buffer.

C-c C-t C-q
M-x tex-show-print-queue

Show the print queue.

C-c C-t C-d
M-x texinfo-delete-from-print-queue

Delete a job from the print queue; you will be prompted for the job number shown by a preceding C-c C-t C-q command (texinfo-show-tex-print-queue).

C-c C-t C-k
M-x tex-kill-job

Kill the currently running TeX job started by either texinfo-tex-region or texinfo-tex-buffer, or any other process running in the Texinfo shell buffer.

C-c C-t C-x
M-x texinfo-quit-job

Quit a TeX formatting job that has stopped because of an error by sending an x to it. When you do this, TeX preserves a record of what it did in a .log file.

C-c C-t C-l
M-x tex-recenter-output-buffer

Redisplay the shell buffer in which the TeX printing and formatting commands are run to show its most recent output.

Thus, the usual sequence of commands for formatting a buffer is as follows (with comments to the right):

C-c C-t C-b             Run texi2dvi on the buffer.
C-c C-t C-p             Print the DVI file.
C-c C-t C-q             Display the printer queue.

The Texinfo mode TeX formatting commands start a subshell in Emacs called the *tex-shell*. The texinfo-tex-command, texinfo-texindex-command, and tex-dvi-print-command commands are all run in this shell.

You can watch the commands operate in the ‘*tex-shell*’ buffer, and you can switch to and from and use the ‘*tex-shell*’ buffer as you would any other shell buffer.

The formatting and print commands depend on the values of several variables. The default values are:

    Variable                              Default value

texinfo-texi2dvi-command                  "texi2dvi"
texinfo-tex-command                       "tex"
texinfo-texindex-command                  "texindex"
texinfo-delete-from-print-queue-command   "lprm"
texinfo-tex-trailer                       "@bye"
tex-start-of-header                       "%**start"
tex-end-of-header                         "%**end"
tex-dvi-print-command                     "lpr -d"
tex-show-queue-command                    "lpq"

You can change the values of these variables with the M-x set-variable command (see Examining and Setting Variables in The GNU Emacs Manual), or with your .emacs initialization file (see Init File in The GNU Emacs Manual).

Beginning with version 20, GNU Emacs offers a user-friendly interface, called Customize, for changing values of user-definable variables. See Easy Customization Interface in The GNU Emacs Manual, for more details about this. The Texinfo variables can be found in the ‘Development/Docs/Texinfo’ group, once you invoke the M-x customize command.


18.7 Using the Local Variables List

Yet another way to apply the TeX formatting command to a Texinfo file is to put that command in a local variables list at the end of the Texinfo file. You can then specify the tex or texi2dvi commands as a compile-command and have Emacs run it by typing M-x compile. This creates a special shell called the *compilation* buffer in which Emacs runs the compile command. For example, at the end of the gdb.texi file, after the @bye, you could put the following:

Local Variables:
compile-command: "texi2dvi gdb.texi"
End:

This technique is most often used by programmers who also compile programs this way; see Compilation in The GNU Emacs Manual.


18.8 TeX Formatting Requirements Summary

Every Texinfo file that is to be input to TeX must begin with a \input command:

\input texinfo

This instructs TeX to load the macros it needs to process a Texinfo file.

Every Texinfo file must end with a line that terminates TeX’s processing and forces out unfinished pages:

@bye

Strictly speaking, these two lines are all a Texinfo file needs to be processed successfully by TeX.

Usually, however, the beginning includes a @settitle command to define the title of the printed manual, a title page, a copyright page, permissions, and a table of contents. Besides @bye, the end of a file usually includes indices. (Not to mention that most manuals contain a body of text as well.)

For more information, see:


18.9 Preparing for TeX

TeX needs to find the texinfo.tex file that the ‘\input texinfo’ command on the first line reads. The texinfo.tex file tells TeX how to handle @-commands; it is included in all standard GNU distributions. The latest version released for general use is available from the usual GNU servers and mirrors:

The latest development version is available from the Texinfo source repository:

texinfo.tex is essentially a standalone file, so, if you need or want to try a newer version than came with your system, it nearly always suffices to download it and put it anywhere that TeX will find it. You can replace any existing texinfo.tex with a newer version (of course saving the original in case of disaster).

Also, you should install epsf.tex, if it is not already installed from another distribution. More details are at the end of the description of the @image command (see Inserting Images).

To use quotation marks other than those used in English, you’ll need to have the European Computer Modern fonts (e.g., ecrm1000) and (for PDF output) CM-Super fonts (see Inserting Quotation Marks).

To use the @euro command, you’ll need the ‘feym*’ fonts (e.g., feymr10). See @euro (€): Euro Currency Symbol.

All of the above files should be installed by default in a reasonable TeX installation.

Optionally, you may create a file texinfo.cnf for site configuration. When processing a Texinfo file, TeX looks for this file in its search path, which includes the current directory and standard installation directories. You can use this file for local conventions. For example, if texinfo.cnf contains the line ‘@afourpaper’ (see Printing on A4 Paper), then all Texinfo documents will be processed with that page size in effect. If you have nothing to put in texinfo.cnf, you do not need to create it.

You can set the TEXINPUTS environment variable to allow TeX to find texinfo.cnf. (This also works for texinfo.tex and any other file TeX might read). For example, if you are using a Bourne shell-compatible shell (sh, bash, ksh, …), your .profile file could contain the lines:

TEXINPUTS=.:/home/me/mylib:
export TEXINPUTS

These settings would cause TeX first to look for an \input file in the current directory, indicated by the ‘.’, then in a hypothetical user ‘me’’s mylib directory, and finally in the system directories. (A leading, trailing, or doubled ‘:’ indicates searching the system directories at that point.)


18.10 Overfull “hboxes”

TeX is sometimes unable to typeset a line within the normal margins. This most often occurs when TeX comes upon what it interprets as a long word that it cannot hyphenate, such as an electronic mail network address or a very long identifier. When this happens, TeX prints an error message like this:

Overfull @hbox (20.76302pt too wide)

(In TeX, lines are in “horizontal boxes”, hence the term, “hbox”. ‘@hbox’ is a TeX primitive not used in the Texinfo language.)

TeX also provides the line number in the Texinfo source file and the text of the offending line, which is marked at all the places that TeX considered hyphenation. See Debugging with TeX, for more information about typesetting errors.

If the Texinfo file has an overfull hbox, you can rewrite the sentence so the overfull hbox does not occur, or you can decide to leave it. A small excursion into the right margin often does not matter and may not even be noticeable.

If you have many overfull boxes and/or an antipathy to rewriting, you can coerce TeX into greatly increasing the allowable interword spacing, thus (if you’re lucky) avoiding many of the bad line breaks, like this:

@tex
\global\emergencystretch = .9\hsize
@end tex

(You should adjust the fraction as needed.) This huge value for \emergencystretch cannot be the default, since then the typeset output would generally be of noticeably lower quality; its default value is ‘.15\hsize’. \hsize is the TeX dimension containing the current line width.

For any overfull boxes you do have, TeX will print a large, ugly, black rectangle beside the line that contains the overfull hbox unless told otherwise. This is so you will notice the location of the problem if you are correcting a draft.

To prevent such a monstrosity from marring your final printout, write the following in the beginning of the Texinfo file on a line of its own, before the @titlepage command:

@finalout

18.11 PDF Output

The simplest way to generate PDF output from Texinfo source is to run the convenience script texi2pdf (or pdftexi2dvi); this executes the texi2dvi script with the --pdf option (see Format with texi2dvi). If for some reason you want to process the document by hand, you can run the pdftex program instead of plain tex. That is, run ‘pdftex foo.texi’ instead of ‘tex foo.texi’.

PDF stands for ‘Portable Document Format’. It was invented by Adobe Systems some years ago for document interchange, based on their PostScript language. Related links:

At present, Texinfo does not provide ‘@ifpdf’ or ‘@pdf’ commands as for the other output formats, since PDF documents contain many internal low-level offsets and cross-references that would be hard or impossible to specify at the Texinfo source level.

PDF files require dedicated software to be displayed, unlike the plain ASCII formats (Info, HTML) that Texinfo supports. They also tend to be much larger than the DVI files output by TeX by default. Nevertheless, a PDF file does define an actual typeset document in a self-contained file, notably including all the fonts that are used, so it has its place.


19 texi2any: The Generic Translator for Texinfo

texi2any is the generic translator for Texinfo that can produce different output formats and is highly customizable. It supports these formats:

Info (by default, or with --info),
HTML (with --html),
plain text (with --plaintext),
DocBook (with --docbook),
Texinfo XML (with --xml).

makeinfo is an alias for texi2any. By default, both texi2any and makeinfo generate Info output; indeed, there are no differences in behavior based on the name.

Beside these default formats, command line options to texi2any can change many aspects of the output. Beyond that, initialization files provide even more control over the final output—nearly anything not specified in the Texinfo input file. Initialization files are written in Perl, like the main program, and anything which can be specified on the command line can also be specified within a initialization file.


19.1 Invoking texi2any/makeinfo from a Shell

To process a Texinfo file, invoke texi2any or makeinfo (the two names are synonyms for the same program; we’ll use the names interchangeably) followed by the name of the Texinfo file. Also select the format you want to output with the appropriate command line option (default is Info). Thus, to create the Info file for Bison, type the following to the shell:

texi2any --info bison.texinfo

You can specify more than one input file name; each is processed in turn. If an input file name is ‘-’, standard input is read.

The texi2any program accepts many options. Perhaps the most basic are those that change the output format. By default, texi2any outputs Info.

Each command line option is either a long name preceded by ‘--’ or a single letter preceded by ‘-’. You can use abbreviations for the long option names as long as they are unique.

For example, you could use the following shell command to create an Info file for bison.texinfo in which lines are filled to only 68 columns:

texi2any --fill-column=68 bison.texinfo

You can write two or more options in sequence, like this:

texi2any --no-split --fill-column=70 …

(This would keep the Info file together as one possibly very long file and would also set the fill column to 70.)

The options are (approximately in alphabetical order):

--commands-in-node-names

This option now does nothing, but remains for compatibility. (It used to ensure that @-commands in node names were expanded throughout the document, especially @value. This is now done by default.)

--conf-dir=dir

Prepend dir to the directory search list for finding customization files that may be loaded with --init-file (see below). The dir value can be a single directory, or a list of several directories separated by the usual path separator character (‘:’ on Unix-like systems, ‘;’ on Windows).

--css-include=file

When producing HTML, literally include the contents of file, which should contain W3C cascading style sheets specifications, in the ‘<style>’ block of the HTML output. If file is ‘-’, read standard input. See HTML CSS.

--css-ref=url

When producing HTML, add a ‘<link>’ tag to the output which references a cascading style sheet at url. This allows using standalone style sheets.

-D var
-D 'var value'

Cause the Texinfo variable var to be defined. This is equivalent to @set var in the Texinfo file (see Flags: @set, @clear, conditionals, and @value).

The argument to the option is always one word to the shell; if it contains internal whitespace, the first word is taken as the variable name and the remainder as the value. For example, -D 'myvar someval' is equivalent to @set myvar someval.

--disable-encoding
--enable-encoding

By default, or with --enable-encoding, output accented and special characters in Info and plain text output based on ‘@documentencoding’. With --disable-encoding, 7-bit ASCII transliterations are output. See @documentencoding enc: Set Input Encoding, and Inserting Accents.

--docbook

Generate DocBook output (rather than Info).

--document-language=lang

Use lang to translate Texinfo keywords which end up in the output document. The default is the locale specified by the @documentlanguage command if there is one, otherwise English (see @documentlanguage ll[_cc]: Set the Document Language).

--dvi

Generate a TeX DVI file using texi2dvi, rather than Info (see texi2any Printed Output).

--dvipdf

Generate a PDF file using texi2dvi --dvipdf, rather than Info (see texi2any Printed Output).

--error-limit=limit
-e limit

Report limit errors before aborting (on the assumption that continuing would be useless); default 100.

--fill-column=width
-f width

Specify the maximum number of columns in a line; this is the right-hand edge of a line. Paragraphs that are filled will be filled to this width. (Filling is the process of breaking up and connecting lines so that lines are the same length as or shorter than the number specified as the fill column. Lines are broken between words.) The default value is 72.

--footnote-style=style
-s style

Set the footnote style to style: either ‘end’ for the end node style (the default) or ‘separate’ for the separate node style. The value set by this option overrides the value set in a Texinfo file by a @footnotestyle command (see Footnote Styles).

When the footnote style is ‘separate’, makeinfo makes a new node containing the footnotes found in the current node. When the footnote style is ‘end’, makeinfo places the footnote references at the end of the current node.

In HTML, when the footnote style is ‘end’, or if the output is not split, footnotes are put at the end of the output. If set to ‘separate’, and the output is split, they are placed in a separate file.

--force
-F

Ordinarily, if the input file has errors, the output files are not created. With this option, they are preserved.

--help
-h

Print a message with available options and basic usage, then exit successfully.

--html

Generate HTML output (rather than Info). By default, the HTML output is split into one output file per Texinfo source node, and the split output is written into a subdirectory based on the name of the top-level Info file. See Generating HTML.

-I dir

Append dir to the directory search list for finding files that are included using the @include command. By default, texi2any searches only the current directory. If dir is not given, the current directory is appended. The dir value can be a single directory or a list of several directories separated by the usual path separator character (‘:’ on Unix-like systems, ‘;’ on Windows).

--ifdocbook
--ifhtml
--ifinfo
--ifplaintext
--iftex
--ifxml

For the given format, process ‘@ifformat’ and ‘@format’ commands, and do not process ‘@ifnotformat’, regardless of the format being output. For instance, if --iftex is given, then ‘@iftex’ and ‘@tex’ blocks will be read, and ‘@ifnottex’ blocks will be ignored.

--info

Generate Info output. By default, if the output file contains more than about 300,000 bytes, it is split into shorter subfiles of about that size. The name of the output file and any subfiles is determined by @setfilename (see @setfilename: Set the Output File Name). See Tag Files and Split Files.

--init-file=file

Load file as code to modify the behavior and output of the generated manual. It is customary to use the .pm or the .init extensions for these customization files, but that is not enforced; the file name can be anything. The --conf-dir option (see above) can be used to add to the list of directories in which these customization files are searched for.

In HTML mode, output a tab-separated file containing three columns: the internal link to an indexed item or item in the table of contents, the name of the index (or table of contents) in which it occurs, and the term which was indexed or entered. The items are in the natural sorting order for the given element. This dump can be useful for post-processors.

--macro-expand=file
-E file

Output the Texinfo source, with all Texinfo macros expanded, to file. Normally, the result of macro expansion is used internally by makeinfo and then discarded.

--no-headers

Do not include menus or node separator lines in the output.

When generating Info, this is the same as using --plaintext, resulting in a simple plain text file. Furthermore, @setfilename is ignored, and output is to standard output unless overridden with -o. (This behavior is for backward compatibility.)

When generating HTML, and output is split, also output navigation links only at the beginning of each file. If output is not split, do not include navigation links at the top of each node at all. See Generating HTML.

--no-ifdocbook
--no-ifhtml
--no-ifinfo
--no-ifplaintext
--no-iftex
--no-ifxml

For the given format, do not process ‘@ifformat’ and ‘@format’ commands, and do process ‘@ifnotformat’, regardless of the format being output. For instance, if --no-ifhtml is given, then ‘@ifhtml’ and ‘@html’ blocks will not be read, and ‘@ifnothtml’ blocks will be.

--no-node-files
--node-files

When generating HTML, create redirection files for anchors and any nodes not already output with the file name corresponding to the node name (see HTML Cross-reference Node Name Expansion). This makes it possible for section- and chapter-level cross-manual references to succeed (see HTML Cross-reference Configuration: htmlxref.cnf).

If the output is split, this is enabled by default. If the output is not split, --node-files enables the creation of the redirection files, in addition to the monolithic main output file. --no-node-files suppresses the creation of redirection files in any case. This option has no effect with any output format other than HTML. See Generating HTML.

--no-number-footnotes
--number-footnotes

Suppress automatic footnote numbering. By default, footnotes are numbered sequentially within a node, i.e., the current footnote number is reset to 1 at the start of each node.

--no-number-sections
--number-sections

With --number_sections (the default), output chapter, section, and appendix numbers as in printed manuals. This works only with hierarchically structured manuals. You should specify --no-number-sections if your manual is not normally structured.

--no-validate
--no-pointer-validate

Suppress the pointer-validation phase of makeinfo—a dangerous thing to do. This can also be done with the @novalidate command (see Formatting Partial Documents).

If you do not suppress pointer validation, makeinfo will check the validity of cross-references and menu entries in the Texinfo file, as well as node pointers if they are given explicitly.

--no-warn

Suppress warning messages (but not error messages).

--output=file
-o file

Specify that the output should be directed to file. This overrides any file name specified in a @setfilename command found in the Texinfo source. If neither @setfilename nor this option are specified, the input file name is used to determine the output name. See @setfilename: Set the Output File Name.

If file is ‘-’, output goes to standard output and ‘--no-split’ is implied.

If file is a directory or ends with a ‘/’ the usual rules are used to determine the output file name (namely, use @setfilename or the input file name) but the files are written to the file directory. For example, ‘makeinfo -o bar/ foo.texi’, with or without --no-split, will write bar/foo.info, and possibly other files, under bar/.

When generating HTML and output is split, file is used as the name for the directory into which all files are written. For example, ‘makeinfo -o bar --html foo.texi’ will write bar/index.html, among other files.

--output-indent=val

This option now does nothing, but remains for compatibility. (It used to alter indentation in XML/DocBook output.)

-P path

Prepend path to the directory search list for @include. If path is not given, the current directory is prepended. See ‘-I’ above.

--paragraph-indent=indent
-p indent

Set the paragraph indentation style to indent. The value set by this option overrides the value set in a Texinfo file by an @paragraphindent command (see @paragraphindent: Controlling Paragraph Indentation). The value of indent is interpreted as follows:

asis

Preserve any existing indentation (or lack thereof) at the beginnings of paragraphs.

0’ or ‘none

Delete any existing indentation.

num

Indent each paragraph by num spaces.

The default is to indent by two spaces, except for paragraphs following a section heading, which are not indented.

--pdf

Generate a PDF file using texi2dvi --pdf, rather than Info (see texi2any Printed Output).

--plaintext

Output a plain text file (rather than Info): do not include menus or node separator lines in the output. This results in a straightforward plain text file that you can (for example) send in email without complications, or include in a distribution (for example, an INSTALL file).

With this option, @setfilename is ignored and the output goes to standard output by default; this can be overridden with -o.

--ps

Generate a PostScript file using texi2dvi --ps, rather than Info (see texi2any Printed Output).

--set-customization-variable var=value
-c var=value

Set the customization variable var to value. The = is optional, but both var and value must be quoted to the shell as necessary so the result is a single word. Many aspects of texi2any behavior and output may be controlled by customization variables, beyond what can be set in the document by @-commands and with other command line switches. See Customization Variables.

--split=how
--no-split

When generating Info, by default large output files are split into smaller subfiles, of approximately 300k bytes. When generating HTML, by default each output file contains one node (see Generating HTML). --no-split suppresses this splitting of the output.

Alternatively, --split=how may be used to specify at which level the HTML output should be split. The possible values for how are:

chapter

The output is split at @chapter and other sectioning @-commands at this level (@appendix, etc.).

section

The output is split at @section and similar.

node

The output is split at every node. This is the default.

Plain text output can be split similarly to HTML. This may be useful for extracting sections from a Texinfo document and making them available as separate files.

--split-size=num

Keep Info files to at most num characters if possible; default is 300,000. (However, a single node will never be split across Info files.)

--transliterate-file-names

Enable transliteration of 8-bit characters in node names for the purpose of file name creation. See HTML Cross-reference 8-bit Character Expansion.

-U var

Cause var to be undefined. This is equivalent to @clear var in the Texinfo file (see Flags: @set, @clear, conditionals, and @value).

--verbose

Cause makeinfo to display messages saying what it is doing. Normally, makeinfo only outputs messages if there are errors or warnings.

--version
-V

Print the version number, then exit successfully.

--Xopt str

Pass str (a single shell word) to texi2dvi; may be repeated (see texi2any Printed Output).

--xml

Generate Texinfo XML output (rather than Info).


19.2 Environment Variables Recognized by texi2any

makeinfo also reads the environment variable TEXINFO_OUTPUT_FORMAT to determine the output format, if not overridden by a command line option. The value should be one of:

docbook  dvi  dvipdf  html  info  pdf  plaintext  ps  xml

If not set or otherwise specified, Info output is the default.

The customization variable of the same name is also read; if set, that overrides an environment variable setting, but not a command-line option. See Customization Variables and Options.

You can control texi2any’s use of Perl extension modules by setting the TEXINFO_XS environment variable. These modules are compiled native code that the interpreted Perl code can use. Ideally, these extension modules should just work, and the only noticeable difference they should make is that texi2any finishes running sooner. However, you can use this environment variable for the purposes of troubleshooting: for example, if you have problems with the output of texi2any varying depending on whether the extension modules are in use.

The following values of TEXINFO_XS are recognized by texi2any:

default

The default behavior. Try to load extension modules, and silently fall back to the interpreted Perl implementations if this fails.

warn

Try to load extension modules, and if this fails, give a warning message before falling back to the interpreted Perl implementations.

debug

Try to load extension modules, printing many messages while doing so.

omit

Do not use extension modules.

Set TEXINFO_XS_PARSER to ‘0’ to disable the use of the native code implementation of the parser module. This is the part of texi2any that converts Texinfo input into an internal tree format used for further processing into output formats. This may be useful for working around bugs or incompatibilities between the native code implementation and the implementation in pure Perl code.


19.3 texi2any Printed Output

To justify the name Texinfo-to-any, texi2any has basic support for creating printed output in the various formats: TeX DVI, PDF, and PostScript. This is done via the simple method of executing the texi2dvi program when those output formats are requested, after checking the validity of the input to give users the benefit of texi2any’s error checking. If you don’t want such error checking, perhaps because your manual plays advanced TeX tricks together with texinfo.tex, just invoke texi2dvi directly.

The output format options for this are --dvi, --dvipdf, --pdf, and --ps. See Format with texi2dvi, for more details on these options and general texi2dvi operation. In addition, the --verbose, --silent, and --quiet options are passed on if specified; the -I and -o options are likewise passed on with their arguments, and --debug without its argument.

The only option remaining that is related to the texi2dvi invocation is --Xopt. Here, just the argument is passed on and multiple --Xopt options accumulate. This provides a way to construct an arbitrary command line for texi2dvi. For example, running

texi2any --Xopt -t --Xopt @a4paper --pdf foo.texi

is equivalent to running

texi2dvi -t @a4paper --pdf foo.texi

except for the validity check.

Although one might wish that other options to texi2any would take effect, they don’t. For example, running ‘texi2any --no-number-sections --dvi foo.texi’ still results in a DVI file with numbered sections. (Perhaps this could be improved in the future, if requests are received.)

The actual name of the command that is invoked is specified by the TEXI2DVI customization variable (see Other Customization Variables). As you might guess, the default is ‘texi2dvi’.

texi2any itself does not generate any normal output when it invokes texi2dvi, only diagnostic messages.


19.4 Customization Variables

Warning: These customization variable names and meanings may change in any Texinfo release. We always try to avoid incompatible changes, but we cannot absolutely promise, since needs change over time.

Many aspects of the behavior and output of texi2any may be modified by modifying so-called customization variables. These fall into a few general categories:

  • Those associated with @-commands; for example, @documentlanguage.
  • Those associated with command-line options; for example, the customization variable SPLIT is associated with the --split command-line option, and TEXINFO_OUTPUT_FORMAT allows specifying the output format.
  • Those associated with customizing the HTML output.
  • Other ad hoc variables.

Customization variables may set on the command line using --set-customization-variable 'var value' (quoting the variable/value pair to the shell) or --set-customization-variable var=value (using =). A special value is ‘undef’, which sets the variable to this special “undefined” Perl value.

The sections below give the details for each of these.


19.4.1 Customization Variables for @-Commands

Each of the following @-commands has an associated customization variable with the same name (minus the leading @):

@allowcodebreaks       @clickstyle        @codequotebacktick
@codequoteundirected   @deftypefnnewline  @documentdescription
@documentencoding      @documentlanguage  @exampleindent
@firstparagraphindent  @footnotestyle     @frenchspacing
@kbdinputstyle         @novalidate        @paragraphindent
@setfilename           @urefbreakstyle    @xrefautomaticsectiontitle

Setting such a customization variable to a value ‘foo’ is similar to executing @cmd foo. It is not exactly the same, though, since any side effects of parsing the Texinfo source are not redone. Also, some variables do not take Texinfo code when generating particular formats, but an argument that is already formatted. This is the case, for example, for HTML for documentdescription.

Note that if texi2any is invoked to process the file with TeX (e.g., with the --pdf option), then these customization variables may not be passed on to TeX.


19.4.2 Customization Variables and Options

The following table gives the customization variables associated with some command line options. See Invoking texi2any/makeinfo from a Shell, for the meaning of the options.

OptionVariable
--enable-encodingENABLE_ENCODING
--document-languagedocumentlanguage
--error-limitERROR_LIMIT
--fill-columnFILLCOLUMN
--footnote-stylefootnotestyle
--forceFORCE
--internal-linksINTERNAL_LINKS
--macro-expandMACRO_EXPAND
--headersHEADERS, FORMAT_MENU
--no-warnNO_WARN
--no-validatenovalidate
--number-footnotesNUMBER_FOOTNOTES
--number-sectionsNUMBER_SECTIONS
--node-filesNODE_FILES
--outputOUTFILE, SUBDIR
--paragraph-indentparagraphindent
--silentSILENT
--splitSPLIT
--split-sizeSPLIT_SIZE
--transliterate-file-namesTRANSLITERATE_FILE_NAMES
--verboseVERBOSE

Setting such a customization variable to a value ‘foo’ is essentially the same as specifying the --opt=foo if the option takes an argument, or --opt if not.

In addition, the customization variable TEXINFO_OUTPUT_FORMAT allows specifying what makeinfo outputs, either one of the usual output formats that can be specified with options, or various other forms:

docbook
dvi
dvipdf
html
info
pdf
plaintext
ps
xml

These correspond to the command-line options (and TEXINFO_OUTPUT_FORMAT environment variable values) of the same name. See Invoking texi2any/makeinfo from a Shell.

debugtree

Instead of generating a regular output format, output a text representation of the tree obtained by parsing the input texinfo document.

parse

Do only Texinfo source parsing; there is no output.

plaintexinfo

Output the Texinfo source with all the macros, @include and @value{} expanded. This is similar to setting --macro-expand, but instead of being output in addition to the normal conversion, output of Texinfo is the main output.

rawtext

Output raw text, with minimal formatting. For example, footnotes are ignored and there is no paragraph filling. This is used by the parser for file names and copyright text in HTML comments, for example.

structure

Do only Texinfo source parsing and determination of the document structure; there is no output.

texinfosxml

Output the document in TexinfoSXML representation, a syntax for writing XML data using Lisp S-expressions.

textcontent

Output the text content only, stripped of commands; this is useful for spell checking or word counting, for example. The trivial detexinfo script setting this is in the util directory of the Texinfo source as an example. It’s one line:

exec texi2any -c TEXINPUT_OUTPUT_FORMAT=textcontent "$@"

19.4.3 HTML Customization Variables

This table gives the customization variables which apply to HTML output only. A few other customization variables apply to both HTML and other output formats; see Other Customization Variables.

AVOID_MENU_REDUNDANCY

If set, and the menu entry and menu description are the same, then do not print the menu description; default false.

AFTER_BODY_OPEN

If set, the corresponding text will appear at the beginning of each HTML file; default unset.

AFTER_ABOUT

For HTML, when an About-element is output. If set, the corresponding text will appear at the end of the About element; default unset.

AFTER_OVERVIEW
AFTER_TOC_LINES

If set, the corresponding text is output after the short table of contents for AFTER_OVERVIEW and after the table of contents for AFTER_TOC_LINES; otherwise, a default string is used. At the time of writing, a </div> element is closed.

In general, you should set BEFORE_OVERVIEW if AFTER_OVERVIEW is set, and you should set BEFORE_TOC_LINES if AFTER_TOC_LINES is set.

BASEFILENAME_LENGTH

The maximum length of a base file name; default 245. Changing this would make cross-manual references to such long node names invalid (see HTML Cross-reference Link Basics).

BEFORE_OVERVIEW
BEFORE_TOC_LINES

If set, the corresponding text is output before the short table of contents for BEFORE_OVERVIEW and before the table of contents for BEFORE_TOC_LINES, otherwise a default string is used. At the time of writing, a <div ...> element is opened.

In general you should set AFTER_OVERVIEW if BEFORE_OVERVIEW is set, and you should set AFTER_TOC_LINES if BEFORE_TOC_LINES is set.

BIG_RULE

Rule used after and before the top element and before special elements, but not for footers and headers; default <hr>.

BODYTEXT

The text appearing in <body>. By default, sets the HTML lang attribute to the document language (see @documentlanguage ll[_cc]: Set the Document Language).

CASE_INSENSITIVE_FILENAMES

Construct output file names as if the filesystem were case insensitive (see HTML Splitting); default false.

CHAPTER_HEADER_LEVEL

Header formatting level used for chapter level sectioning commands; default ‘2’.

CHECK_HTMLXREF

Check that manuals which are the target of external cross-references (see @xref with Four and Five Arguments) are present in htmlxref.cnf (see HTML Cross-reference Configuration: htmlxref.cnf); default false.

COMPLEX_FORMAT_IN_TABLE

If set, use tables for indentation of complex formats; default false.

CONTENTS_OUTPUT_LOCATION

If set to ‘after_top’, output the contents at the end of the @top section. If set to ‘inline’, output the contents where the @contents and similar @-commands are located. If set to ‘separate_element’ output the contents in separate elements, either at the end of the document if not split, or in a separate file. If set to ‘after_title’ the tables of contents are output after the title; default ‘after_top’.

COPIABLE_ANCHORS

If set, output copiable anchors for the definition commands (see Definition Commands) and table commands (see Making a Two-column Table) where an index entry is defined. An anchor appears as a ‘¶’ sign that appears when you hover the mouse pointer over the heading text.

CSS_LINES

CSS output, automatically determined by default (see HTML CSS).

DATE_IN_HEADER

Put the document generation date in the header; off by default.

DEF_TABLE

If set, a <table> construction for @deffn and similar @-commands is used (looking more like the TeX output), instead of definition lists; default false.

DEFAULT_RULE

Rule used between element, except before and after the top element, and before special elements, and for footers and headers; default <hr>.

DO_ABOUT

If set to 0 never do an About special element; if set to 1 always do an About special element; default 0.

EXTERNAL_DIR

Base directory for external manuals; default none. It is better to use the general external cross-reference mechanism (see HTML Cross-reference Configuration: htmlxref.cnf) than this variable.

EXTRA_HEAD

Additional text appearing within <head>; default unset.

FOOTNOTE_END_HEADER_LEVEL

Header formatting level used for the footnotes header with the ‘end’ footnotestyle; default ‘4’. See Footnote Styles.

FOOTNOTE_SEPARATE_HEADER_LEVEL

Header formatting level used for the footnotes header with the ‘separate’ footnotestyle; default ‘4’. See Footnote Styles.

FRAMES

If set, a file describing the frame layout is generated, together with a file with the short table of contents; default false.

FRAMESET_DOCTYPE

Same as DOCTYPE, but for the file containing the frame description.

HEADER_IN_TABLE

Use tables for header formatting rather than a simple <div> element; default false.

HTML_MATH

Method to use to render @math. This can be unset, set to ‘mathjax’ (see MathJax Customization Variables), set to ‘l2h’, which has the same effect as setting L2H (see latex2html Customization Variables), or set to ‘t2h’, which uses tex4ht.

ICONS

Use icons for the navigation panel; default false.

If set, the associated value is prepended to the image file links; default unset.

INDEX_ENTRY_COLON

Symbol used between the index entry and the associated node or section; default ‘:’.

INFO_JS_DIR

(Experimental.) Add a JavaScript browsing interface to the manual. The value of the variable is the directory to place the code for this interface, so you would run the program as e.g. ‘texi2any --html -c INFO_JS_DIR=js manual.texi’ to place files in a ‘js’ directory under the output. This provides some of the functionality of the Info browsers in a web browser, such as keyboard navigation and index lookup. This only works with non-split HTML output.

The interface should provide an acceptable fallback in functionality if JavaScript or web browser features are not available. However, please be cautious when using this option, in case you do make your documentation harder to access for some of your users.

INLINE_CSS_STYLE

Put CSS directly in HTML elements rather than at the beginning of the output; default false.

JS_WEBLABELS
JS_WEBLABELS_FILE

Specify how to use a JavaScript license web labels page to give licensing information and source code for any JavaScript used in the HTML files for the manual. (See https://www.gnu.org/licenses/javascript-labels.html).

With the value ‘generate’ (the default), generate a labels page at JS_WEBLABELS_FILE, and link to it in the HTML output files. Only do this if actually referencing JavaScript files (either with HTML_MATH set to ‘mathjax’, or when using the experimental JS browsing interface when INFO_JS_DIR is set). With this setting, JS_WEBLABELS_FILE must be a relative file name.

With the value ‘reference’, link to the labels file given by JS_WEBLABELS_FILE in the output, and do not generate a labels file. This setting is useful if you separately maintain a single labels file for a larger website that includes your manual.

With ‘omit’, neither generate nor link to a labels file.

KEEP_TOP_EXTERNAL_REF

If set, do not ignore ‘Top’ as the first argument for an external ref to a manual, as is done by default. See Referring to a Manual as a Whole.

MAX_HEADER_LEVEL

Maximum header formatting level used (higher header formatting level numbers correspond to lower sectioning levels); default ‘4’.

MENU_ENTRY_COLON

Symbol used between the menu entry and the description; default ‘:’.

MENU_SYMBOL

Symbol used in front of menu entries when node names are used for menu entries formatting; default ‘&bull;’.

MONOLITHIC

Output only one file including the table of contents. Set by default, but only relevant when the output is not split.

NO_CSS

Do not use CSS; default false. See HTML CSS.

PRE_ABOUT

Used when an About element is output. If set to a text string, this text will appear at the beginning of the About element. If set to a reference on a subroutine, the result of the subroutine call will appear at the beginning of the About element. If not set (the default), default text is used.

PRE_BODY_CLOSE

If set, the given text will appear at the footer of each HTML file; default unset.

PROGRAM_NAME_IN_FOOTER

If set, output the program name and miscellaneous related information in the page footers; default false.

SECTION_NAME_IN_TITLE

If set, when output is split, use the argument of the chapter structuring command (e.g., @chapter or @section) in the <title> instead of the argument to @node.

SHOW_TITLE

If set, output the title at the beginning of the document; default true.

SIMPLE_MENU

If set, use a simple preformatted style for the menu, instead of breaking down the different parts of the menu; default false. See The Parts of a Menu.

If set, links from headings to toc entries are created; default false.

TOP_FILE

This file name may be used for the top-level file. The extension is set appropriately, if necessary. This is used to override the default, and is, in general, only taken into account when output is split, and for HTML.

TOP_NODE_FILE_TARGET

File name used for the Top node in cross-references; default is index.html.

TOP_NODE_UP_URL

A url used for Top node up references; the default is undef, in that case no Top node Up reference is generated. For more about the Top node pointers, see The First Node. For overriding the Up pointer name in case TOP_NODE_UP_URL is set and for other formats, see TOP_NODE_UP in Other Customization Variables.

USE_ACCESSKEY

Use accesskey in cross-references; default true.

USE_ISO

Use entities for doubled single-quote characters (see Inserting Quotation Marks), and ‘---’ and ‘--’ (see General Syntactic Conventions); default true.

Generate <link> elements in the HTML <head> output; default true.

USE_NODE_DIRECTIONS

If true, use nodes to determine where next, up and prev link to in node headers. If false, use sections. If undefined, use USE_NODES value. Default is undefined. Note that this setting does not determine the link string only where the links points to, see xrefautomaticsectiontitle for the link string customization. If nodes and sections are systematically associated, this customization has no practical effect.

USE_REL_REV

Use rel in cross-references; default true.

VERTICAL_HEAD_NAVIGATION

If set, a vertical navigation panel is used; default false.

WORDS_IN_PAGE

When output is split by nodes, specifies the approximate minimum page length at which a navigation panel is placed at the bottom of a page. To avoid ever having the navigation buttons at the bottom of a page, set this to a sufficiently large number. The default is 300.

XREF_USE_FLOAT_LABEL

If set, for the float name in cross-references, use the float label instead of the type followed by the float number (see @float [type][,label]: Floating Material). The default is off.

XREF_USE_NODE_NAME_ARG

Only relevant for cross-reference commands with no cross reference name (second argument). If set to 1, use the node name (first) argument in cross-reference @-commands for the text displayed as the hyperlink. If set to 0, use the node name if USE_NODES is set, otherwise the section name. If set to ‘undef’, use the first argument in preformatted environments, otherwise use the node name or section name depending on USE_NODES. The default is ‘undef’.


19.4.4 MathJax Customization Variables

This table lists the customization variables which can be used when MathJax is being used, which will be the case when HTML_MATH is set to ‘mathjax’.

MATHJAX_SCRIPT

URL of the MathJax component file (e.g. tex-svg.js) you are using. texi2any provides a default value for this variable, but you are encouraged to host this file yourself on your website so that you are not dependent on others’ hosting.

MATHJAX_SOURCE

A URL of the full source code in its preferred form for modification, or instructions for obtaining such source code, for the component file named by MATHJAX_SCRIPT. ‘Preferred form for modification’ means that this should not be in a ‘minified’ form. Used in the license labels page (see HTML Customization Variables, under JS_WEBLABELS).

Again, texi2any provides a default value for this variable, but you are encouraged to host the source code for MathJax and its dependencies yourself. This is in order to make the source code available reliably, and to reduce you and your users’ dependence on others’ distribution systems.


19.4.5 latex2html Customization Variables

This table lists the customization variables which can be used when latex2html is being used.

L2H

For HTML. If set, latex2html is used to convert @math and @tex sections; default false. Best used with --iftex.

L2H_CLEAN

(Relevant only if L2H is set.) If set, the intermediate files generated in relation with latex2html are removed; default true.

L2H_FILE

(Relevant only if L2H is set.) If set, the given file is used as latex2html’s init file; default unset.

L2H_HTML_VERSION

(Relevant only if L2H is set.) The HTML version used in the latex2html call; default unset.

L2H_L2H

(Relevant only if L2H is set.) The program invoked as latex2html; default is latex2html.

L2H_SKIP

(Relevant only if L2H is set.) If set to a true value, the actual call to latex2html is skipped; previously generated content is reused instead. If set to 0, the cache is not used at all. If set to ‘undef’, the cache is used for as many TeX fragments as possible and for any remaining the command is run. The default is ‘undef’.

L2H_TMP

(Relevant only if L2H is set.) Set the directory used for temporary files. None of the file name components in this directory name may start with ‘.’; otherwise, latex2html will fail (because of dvips). The default is the empty string, which means the current directory.


19.4.6 Other Customization Variables

This table gives the remaining customization variables, which apply to multiple formats, or affect global behavior, or otherwise don’t fit into the categories of the previous sections.

CHECK_NORMAL_MENU_STRUCTURE

Warn if the nodes pointers (either explicitly or automatically set) are not consistent with the order of node menu entries.

CLOSE_QUOTE_SYMBOL

When a closing quote is needed, use this character; default &rsquo; in HTML, &#8217; in DocBook. The default for Info is the same as OPEN_QUOTE_SYMBOL (see below).

CPP_LINE_DIRECTIVES

Recognize #line directives in a “preprocessing” pass (see External Macro Processors: Line Directives); on by default.

DEBUG

If set, debugging output is generated; default is off (zero).

DOCTYPE

For DocBook, HTML, XML. Specifies the SystemLiteral, the entity’s system identifier. This is a URI which may be used to retrieve the entity, and identifies the canonical DTD for the document. The default value is different for each of HTML, DocBook and Texinfo XML.

DUMP_TEXI

For debugging. If set, no conversion is done, only parsing and macro expansion. If the option --macro-expand is set, the Texinfo source is also expanded to the corresponding file. Default false.

DUMP_TREE

For debugging. If set, the tree constructed upon parsing a Texinfo document is output to standard error; default false.

ENABLE_ENCODING_USE_ENTITY

For HTML, XML. If --enable-encoding is set, and there is an entity corresponding with the letter or the symbol being output, prefer the entity. Set by default for HTML, but not XML.

EXTERNAL_CROSSREF_SPLIT

For cross-references to other manuals, this determines if the other manual is considered to be split or monolithic. By default, it is set based on the value of SPLIT. See HTML Cross-references, and see HTML Cross-reference Configuration: htmlxref.cnf.

EXTENSION

The extension added to the output file name. The default is different for each output format.

FORMAT_MENU

If set to ‘menu’, output Texinfo menus. This is the default for Info. ‘sectiontoc’ is the default setting for HTML, where instead of the contents of @menu blocks being output, a list of subordinate sections is output in each node. If set to ‘nomenu’, do not output menus.

This variable is set to ‘nomenu’ when generating DocBook, or when --no-headers is specified.

IGNORE_BEFORE_SETFILENAME

If set, begin outputting at @setfilename, if @setfilename is present; default true.

IGNORE_SPACE_AFTER_BRACED_COMMAND_NAME

If set, spaces are ignored after an @-command that takes braces. Default true, matching the TeX behavior.

INDEX_SPECIAL_CHARS_WARNING

If set, warn about ‘:’ in index entry, as it leads to invalid entries in index menus in output Info files. For Info and plaintext only.

INFO_SPECIAL_CHARS_QUOTE

If set, whenever there are problematic characters for Info output in places such as node names or menu items, surround the part of the construct where they appear with quoting characters, as described in Info Format Specification. See @node Line Requirements.

INFO_SPECIAL_CHARS_WARNING

If set, warn about problematic constructs for Info output (such as the string ‘::’) in node names, menu items, and cross-references; default true. Do not warn about index entries, since parsing problems there don’t prevent navigation; readers can still relatively easily find their way to the node in question.

MAX_MACRO_CALL_NESTING

The maximal number of recursive calls of @-commands defined through @rmacro; default 100000. The purpose of this variable is to avoid infinite recursions.

NO_USE_SETFILENAME

If set, do not use @setfilename to set the document name; instead, base the output document name only on the input file name. The default is false.

NODE_NAME_IN_INDEX

If set, use node names in index entries, otherwise prefer section names; default true.

NODE_NAME_IN_MENU

If set, use node names in menu entries, otherwise prefer section names; default true.

OPEN_QUOTE_SYMBOL

When an opening quote is needed, e.g., for ‘@samp’ output, use the specified character; default &lsquo; for HTML, &#8216; for DocBook. For Info, the default depends on the enabled document encoding (see @documentencoding enc: Set Input Encoding); if no document encoding is set, or the encoding is US-ASCII, etc., ‘'’ is used. This character usually appears as an undirected single quote on modern systems. If the document encoding is Unicode, the Info output uses a Unicode left quote.

OUTPUT_ENCODING_NAME

Normalized encoding name used for output files. Should be a usable charset name in HTML, typically one of the preferred IANA encoding names. By default, if an input encoding is set (typically through @documentencoding), this information is used to set the output encoding name. If no input encoding is specified, the default output encoding name may be set by the output format. In particular, the XML-based formats use utf-8 for OUTPUT_ENCODING_NAME if the encoding is not otherwise specified. See @documentencoding enc: Set Input Encoding.

If set, the cross-references in the Overview link to the corresponding Table of Contents entries; default true.

PACKAGE
PACKAGE_VERSION
PACKAGE_AND_VERSION
PACKAGE_URL
PACKAGE_NAME

The implementation’s short package name, package version, package name and version concatenated, package url, and full package name, respectively. By default, these variables are all set through Autoconf, Automake, and configure.

PREFIX

The output file prefix, which is prepended to some output file names. By default it is set by @setfilename or from the input file (see @setfilename: Set the Output File Name). How this value is used depends on the value of other customization variables or command line options, such as whether the output is split. The default is unset.

PROGRAM

Name of the program used. By default, it is set to the name of the program launched, with a trailing ‘.pl’ removed.

SORT_ELEMENT_COUNT

If set, the name of a file to which a list of elements (nodes or sections, depending on the output format) is dumped, sorted by the number of lines they contain after removal of @-commands; default unset. This is used by the program texi-elements-by-size in the util/ directory of the Texinfo source distribution (see texi-elements-by-size).

SORT_ELEMENT_COUNT_WORDS

When dumping the elements-by-size file (see preceding item), use word counts instead of line counts; default false.

TEST

If set to true, some variables which are normally dynamically generated anew for each run (date, program name, version) are set to fixed and given values. This is useful to compare the output to a reference file, as is done for the tests. The default is false.

TEXI2DVI

Name of the command used to produce PostScript, PDF, and DVI; default ‘texi2dvi’. See texi2any Printed Output.

TEXI2HTML

Generate HTML and try to be as compatible as possible with texi2html; default false.

TEXINFO_DTD_VERSION

For XML. Version of the DTD used in the XML output preamble. The default is set based on a variable in configure.ac.

TEXTCONTENT_COMMENT

For stripped text content output (i.e., when TEXINFO_OUTPUT_FORMAT is set to textcontent). If set, also output comments. Default false.

TOP_NODE_UP

Up node for the Top node; default ‘(dir)’. This node name is supposed to be already formatted for the output format. In HTML can be used in attribute, so should not contain any element. Used for HTML output only if TOP_NODE_UP_URL is set to override the url, see TOP_NODE_UP_URL in HTML Customization Variables.

TREE_TRANSFORMATIONS

The associated value is a comma separated list of transformations that can be applied to the Texinfo tree prior to outputting the result. If more than one is specified, the ordering is irrelevant; each is always applied at the necessary point during processing.

The only one executed by default is ‘move_index_entries_after_items’ for HTML and DocBook output. Here’s an example of updating the master menu in a document:

makeinfo \
  -c TREE_TRANSFORMATIONS=regenerate_master_menu \
  -c TEXINFO_OUTPUT_FORMAT=plaintexinfo \
  mydoc.texi \
  -o /tmp/out

(Caveat: Since ‘plaintexinfo’ output expands Texinfo macros and conditionals, it’s necessary to remove any such differences before installing the updates in the original document. This may be remedied in a future release.)

The following transformations are currently supported (many are used in the pod2texi utility distributed with Texinfo; see Invoking pod2texi: Convert POD to Texinfo):

complete_tree_nodes_menus

Add menu entries or whole menus for nodes associated with sections of any level, based on the sectioning tree.

complete_tree_nodes_missing_menu

Add whole menus for nodes associated with sections without menu. The menus are based on the sectioning tree.

fill_gaps_in_sectioning

Adds empty @unnumbered... sections in a tree to fill gaps in sectioning. For example, an @unnumberedsec will be inserted if a @chapter is followed by a @subsection.

insert_nodes_for_sectioning_commands

Insert nodes for sectioning commands lacking a corresponding node.

move_index_entries_after_items

In @enumerate and @itemize, move index entries appearing just before an @item to just after the @item. Comment lines between index entries are moved too. As mentioned, this is always done for HTML and DocBook output.

regenerate_master_menu

Update the Top node master menu, either replacing the (first) @detailmenu in the Top node menu, or creating it at the end of the Top node menu.

simple_menu

Mostly the same as SIMPLE_MENU: use a simple preformatted style for the menu. It differs from setting SIMPLE_MENU in that SIMPLE_MENU only has an effect in HTML output.

USE_NODES

Preferentially use nodes to decide where elements are separated. If set to false, preferentially use sectioning to decide where elements are separated. The default is true.

USE_NUMERIC_ENTITY

For HTML and XML. If set, use numeric entities instead of ASCII characters when there is no named entity. By default, set to true for HTML.

USE_UP_NODE_FOR_ELEMENT_UP

Fill in up sectioning direction with node direction when there is no sectioning up direction. In practice this can only happen when there is no @top section. Not set by default.

USE_SETFILENAME_EXTENSION

Default is on for Info, off for other output. If set, use exactly what @setfilename gives for the output file name, including the extension. You should not need to explicitly set this variable.

USE_TITLEPAGE_FOR_TITLE

Use the full @titlepage as the title, not a simple title string; default false.

USE_UNIDECODE

If set to false, do not use the Text::Unidecode Perl module to transliterate more characters; default true.


19.5 Internationalization of Document Strings

texi2any writes fixed strings into the output document at various places: cross-references, page footers, the help page, alternate text for images, and so on. The string chosen depends on the value of the documentlanguage at the time of the string being output (see @documentlanguage ll[_cc]: Set the Document Language, for the Texinfo command interface).

The Gettext framework is used for those strings (see Gettext). The libintl-perl package is used as the gettext implementation; more specifically, the pure Perl implementation is used, so Texinfo can support consistent behavior across all platforms and installations, which would not otherwise be possible. libintl-perl is included in the Texinfo distribution and always installed, to ensure that it is available if needed. It is also possible to use the system gettext (the choice can be made at build-time).

The Gettext domain ‘texinfo_document’ is used for the strings. Translated strings are written as Texinfo, and may include @-commands. In translated strings, the varying parts of the string are not usually denoted by %s and the like, but by ‘{arg_name}’. (This convention is common for gettext in Perl and is fully supported in GNU Gettext; see Perl Format Strings in GNU Gettext.) For example, in the following, ‘{section}’ will be replaced by the section name:

see {section}

These Perl-style brace format strings are used for two reasons: first, changing the order of printf arguments is only available since Perl 5.8.0; second, and more importantly, the order of arguments is unpredictable, since @-command expansion may lead to different orders depending on the output format.

The expansion of a translation string is done like this:

  1. First, the string is translated. The locale is @documentlanguage.@documentencoding.

    If the @documentlanguage has the form ‘ll_CC’, that is tried first, and then just ‘ll’. If that does not exist, and the encoding is not us-ascii, then us-ascii is tried.

    The idea is that if there is a us-ascii encoding, it means that all the characters in the charset may be expressed as @-commands. For example, there is a fr.us-ascii locale that can accommodate any encoding, since all the Latin 1 characters have associated @-commands. On the other hand, Japanese has only a translation ja.utf-8, since there are no @-commands for Japanese characters.

  2. Next, the string is expanded as Texinfo, and converted. The arguments are substituted; for example, ‘{arg_name}’ is replaced by the corresponding actual argument.

In the following example, ‘{date}’, ‘{program_homepage}’ and ‘{program}’ are the arguments of the string. Since they are used in @uref, their order is not predictable. ‘{date}’, ‘{program_homepage}’ and ‘{program}’ are substituted after the expansion:

Generated on @emph{{date}} using
@uref{{program_homepage}, @emph{{program}}}.

This approach is admittedly a bit complicated. Its usefulness is that it supports having translations available in different encodings for encodings which can be covered by @-commands, and also specifying how the formatting for some commands is done, independently of the output format—yet still be language-dependent. For example, the ‘@pxref’ translation string can be like this:

see {node_file_href} section `{section}\' in @cite{{book}}

which allows for specifying a string independently of the output format, while nevertheless with rich formatting it may be translated appropriately in many languages.


19.6 Invoking pod2texi: Convert POD to Texinfo

The pod2texi program translates Perl pod documentation file(s) to Texinfo. There are two basic modes of operation: generating a standalone manual from each input pod, or (if --base-level=1 or higher is given) generating Texinfo subfiles suitable for use with @include.

Although ordinarily this documentation in the Texinfo manual would be the best place to look, in this case we have documented all the options and examples in the pod2texi program itself, since it may be useful outside of the rest of Texinfo. Thus, please see the output of pod2texi --help, the version on the web at http://www.gnu.org/software/texinfo/manual/pod2texi.html, etc.

For an example of using pod2texi to make Texinfo out of the Perl documentation itself, see contrib/perldoc-all in the Texinfo source distribution.


19.7 texi2html: Ancestor of texi2any

Conceptually, the texi2html program is the parent of today’s texi2any program. texi2html was developed independently, originally by Lionel Cons in 1998; at the time, makeinfo could not generate HTML. Many other people contributed to texi2html over the years.

The present texi2any uses little of the actual code of texi2html, and has quite a different basic approach to the implementation (namely, parsing the Texinfo document into a tree), but still, there is a family resemblance.

By design, texi2any supports nearly all the features of texi2html in some way. However, we did not attempt to maintain strict compatibility, so no texi2html executable is installed by the Texinfo package. An approximation can be run with an invocation like this (available as util/texi2html in the Texinfo source):

texi2any --set-customization-variable TEXI2HTML=1 ...

but, to emphasize, this is not a drop-in replacement for the previous texi2html. Here are the biggest differences:

  • Most blatantly, the command line options of texi2html are now customization variables, for the most part. A table of approximate equivalents is given below.
  • The program-level customization API is very different in texi2any.
  • Indices cannot be split.
  • Translated strings cannot be customized; we hope to introduce this feature in texi2any in the future.

Aside from the last, we do not intend to reimplement these differences. Therefore, the route forward for authors is alter manuals and build processes as necessary to use the new features and methods of texi2any. The texi2html maintainers (one of whom is the principal author of texi2any) do not intend to make further releases.

Here is the table showing texi2html options and corresponding texi2any customization variables.

--toc-linksTOC_LINKS
--short-extSHORTEXTN
--prefixPREFIX
--short-refSHORT_REF
--idx-sumIDX_SUMMARY
--def-tableDEF_TABLE
--ignore-preamble-textIGNORE_PREAMBLE_TEXT
--html-xref-prefixEXTERNAL_DIR
--l2hL2H
--l2h-l2hL2H_L2H
--l2h-skipL2H_SKIP
--l2h-tmpL2H_TMP
--l2h-fileL2H_FILE
--l2h-cleanL2H_CLEAN
--use-nodesUSE_NODES
--monolithicMONOLITHIC
--top-fileTOP_FILE
--framesFRAMES
--menuFORMAT_MENU
--debugDEBUG
--doctypeDOCTYPE
--frameset-doctypeFRAMESET_DOCTYPE
--testTEST

Finally, any texi2html users seeking more detailed information can check the draft file doc/texi2oldapi.texi in the Texinfo source repository. It consists mainly of very rough notes, but may still be useful to some.


20 Creating and Installing Info Files

This chapter describes how to create and install Info files. See Info Files, for general information about the file format itself.


20.1 Creating an Info File

makeinfo is a program that converts a Texinfo file into an Info file, HTML file, or plain text. texinfo-format-region and texinfo-format-buffer are GNU Emacs functions that convert Texinfo to Info.

For information on installing the Info file in the Info system, see Installing an Info File.


20.1.1 makeinfo Advantages

The makeinfo utility creates an Info file from a Texinfo source providing better error messages than either of the Emacs formatting commands. We recommend it. The makeinfo program is independent of Emacs. You can run makeinfo in any of three ways: from an operating system shell, from a shell inside Emacs, or by typing the C-c C-m C-r or the C-c C-m C-b command in Texinfo mode in Emacs.

The texinfo-format-region and the texinfo-format-buffer commands may be useful if you cannot run makeinfo.


20.1.2 Running makeinfo Within Emacs

You can run makeinfo in GNU Emacs Texinfo mode by using either the makeinfo-region or the makeinfo-buffer commands. In Texinfo mode, the commands are bound to C-c C-m C-r and C-c C-m C-b by default.

C-c C-m C-r
M-x makeinfo-region

Format the current region for Info.

C-c C-m C-b
M-x makeinfo-buffer

Format the current buffer for Info.

When you invoke makeinfo-region the output goes to a temporary buffer. When you invoke makeinfo-buffer output goes to the file set with @setfilename (see @setfilename: Set the Output File Name).

The Emacs makeinfo-region and makeinfo-buffer commands run the makeinfo program in a temporary shell buffer. If makeinfo finds any errors, Emacs displays the error messages in the temporary buffer.

You can parse the error messages by typing C-x ` (next-error). This causes Emacs to go to and position the cursor on the line in the Texinfo source that makeinfo thinks caused the error. See Running make or Compilers Generally in The GNU Emacs Manual, for more information about using the next-error command.

In addition, you can kill the shell in which the makeinfo command is running or make the shell buffer display its most recent output.

C-c C-m C-k
M-x makeinfo-kill-job

Kill the current running makeinfo job (from makeinfo-region or makeinfo-buffer).

C-c C-m C-l
M-x makeinfo-recenter-output-buffer

Redisplay the makeinfo shell buffer to display its most recent output.

(Note that the parallel commands for killing and recentering a TeX job are C-c C-t C-k and C-c C-t C-l. See Formatting and Printing in Texinfo Mode.)

You can specify options for makeinfo by setting the makeinfo-options variable with either the M-x customize or the M-x set-variable command, or by setting the variable in your .emacs initialization file.

For example, you could write the following in your .emacs file:

(setq makeinfo-options
     "--paragraph-indent=0 --no-split
      --fill-column=70 --verbose")

For more information, see
Easy Customization Interface in The GNU Emacs Manual,
Examining and Setting Variables in The GNU Emacs Manual,
Init File in The GNU Emacs Manual, and
makeinfo Options.


20.1.3 The texinfo-format… Commands

In GNU Emacs in Texinfo mode, you can format part or all of a Texinfo file with the texinfo-format-region command. This formats the current region and displays the formatted text in a temporary buffer called ‘*Info Region*’.

Similarly, you can format a buffer with the texinfo-format-buffer command. This command creates a new buffer and generates the Info file in it. Typing C-x C-s will save the Info file under the name specified by the @setfilename line which must be near the beginning of the Texinfo file.

C-c C-e C-r
texinfo-format-region

Format the current region for Info.

C-c C-e C-b
texinfo-format-buffer

Format the current buffer for Info.

The texinfo-format-region and texinfo-format-buffer commands provide you with some error checking, and other functions can provide you with further help in finding formatting errors. These procedures are described in an appendix; see Catching Mistakes. However, the makeinfo program provides better error checking (see Running makeinfo Within Emacs).

A peculiarity of the texinfo-format-buffer and texinfo-format-region commands is that they do not indent (nor fill) paragraphs that contain @w or @* commands.


20.1.4 Batch Formatting

You can format Texinfo files for Info using batch-texinfo-format and Emacs batch mode. You can run Emacs in batch mode from any shell, including a shell inside of Emacs. (See Initial Options in The GNU Emacs Manual.)

Here is a shell command to format all the files that end in .texinfo in the current directory:

emacs -batch -funcall batch-texinfo-format *.texinfo

Emacs processes all the files listed on the command line, even if an error occurs while attempting to format some of them.

Run batch-texinfo-format only with Emacs in batch mode as shown; it is not interactive. It kills the batch mode Emacs on completion.

batch-texinfo-format is convenient if you lack makeinfo and want to format several Texinfo files at once. When you use Batch mode, you create a new Emacs process. This frees your current Emacs, so you can continue working in it. (When you run texinfo-format-region or texinfo-format-buffer, you cannot use that Emacs for anything else until the command finishes.)


20.1.5 Tag Files and Split Files

If a Texinfo file has more than 30,000 bytes, texinfo-format-buffer automatically creates a tag table for its Info file; makeinfo always creates a tag table. With a tag table, Info can jump to new nodes more quickly than it can otherwise.

In addition, if the Texinfo file contains more than about 300,000 bytes, texinfo-format-buffer and makeinfo split the large Info file into shorter indirect subfiles of about 300,000 bytes each. Big files are split into smaller files so that Emacs does not need to make a large buffer to hold the whole of a large Info file; instead, Emacs allocates just enough memory for the small, split-off file that is needed at the time. This way, Emacs avoids wasting memory when you run Info. (Before splitting was implemented, Info files were always kept short and include files were designed as a way to create a single, large printed manual out of the smaller Info files. See Include Files, for more information. Include files are still used for very large documents, such as The Emacs Lisp Reference Manual, in which each chapter is a separate file.)

When a file is split, Info itself makes use of a shortened version of the original file that contains just the tag table and references to the files that were split off. The split-off files are called indirect files.

The split-off files have names that are created by appending ‘-1, ‘-2, ‘-3 and so on to the file name specified by the @setfilename command. The shortened version of the original file continues to have the name specified by @setfilename.

At one stage in writing this document, for example, the Info file was saved as the file test-texinfo and that file looked like this:

Info file: test-texinfo,    -*-Text-*-
produced by texinfo-format-buffer
from file: new-texinfo-manual.texinfo

^_
Indirect:
test-texinfo-1: 102
test-texinfo-2: 50422
test-texinfo-3: 101300
^_^L
Tag table:
(Indirect)
Node: overview^?104
Node: info file^?1271
Node: printed manual^?4853
Node: conventions^?6855
…

(But test-texinfo had far more nodes than are shown here.) Each of the split-off, indirect files, test-texinfo-1, test-texinfo-2, and test-texinfo-3, is listed in this file after the line that says ‘Indirect:’. The tag table is listed after the line that says ‘Tag table:’.

In the list of indirect files, the number following the file name records the cumulative number of bytes in the preceding indirect files, not counting the file list itself, the tag table, or any permissions text in the first file. In the tag table, the number following the node name records the location of the beginning of the node, in bytes from the beginning of the (unsplit) output.

If you are using texinfo-format-buffer to create Info files, you may want to run the Info-validate command. (The makeinfo command does such a good job on its own, you do not need Info-validate.) However, you cannot run the M-x Info-validate node-checking command on indirect files. For information on how to prevent files from being split and how to validate the structure of the nodes, see Using Info-validate.


20.2 Installing an Info File

Info files are usually kept in the info directory. You can read Info files using the standalone Info program or the Info reader built into Emacs. (See Info, for an introduction to Info.)


20.2.1 The Directory File dir

For Info to work, the info directory must contain a file that serves as a top-level directory for the Info system. By convention, this file is called dir. (You can find the location of this file within Emacs by typing C-h i to enter Info and then typing C-x C-f to see the location of the info directory.)

The dir file is itself an Info file. It contains the top-level menu for all the Info files in the system. The menu looks like this:

* Menu:
* Info:    (info).     Documentation browsing system.
* Emacs:   (emacs).    The extensible, self-documenting
                      text editor.
* Texinfo: (texinfo).  With one source file, make
                      either a printed manual using
                      @TeX{} or an Info file.
…

Each of these menu entries points to the ‘Top’ node of the Info file that is named in parentheses. (The menu entry does not need to specify the ‘Top’ node, since Info goes to the ‘Top’ node if no node name is mentioned. See Nodes in Other Info Files.)

Thus, the ‘Info’ entry points to the ‘Top’ node of the info file and the ‘Emacs’ entry points to the ‘Top’ node of the emacs file.

In each of the Info files, the ‘Up’ pointer of the ‘Top’ node refers back to the dir file. For example, the line for the ‘Top’ node of the Emacs manual looks like this in Info:

File: emacs  Node: Top, Up: (DIR), Next: Distrib

In this case, the dir file name is written in uppercase letters—it can be written in either upper- or lowercase. This is not true in general, it is a special case for dir.


20.2.2 Listing a New Info File

To add a new Info file to your system, you must write a menu entry to add to the menu in the dir file in the info directory. For example, if you were adding documentation for GDB, you would write the following new entry:

* GDB: (gdb).           The source-level C debugger.

The first part of the menu entry is the menu entry name, followed by a colon. The second part is the name of the Info file, in parentheses, followed by a period. The third part is the description.

The name of an Info file often has a .info extension. Thus, the Info file for GDB might be called either gdb or gdb.info. The Info reader programs automatically try the file name both with and without .info4; so it is better to avoid clutter and not to write ‘.info’ explicitly in the menu entry. For example, the GDB menu entry should use just ‘gdb’ for the file name, not ‘gdb.info’.


20.2.3 Info Files in Other Directories

If an Info file is not in the info directory, there are three ways to specify its location:

  1. Write the file name in the dir file as the second part of the menu.
  2. Specify the Info directory name in the INFOPATH environment variable in your .profile or .cshrc initialization file. (Only you and others who set this environment variable will be able to find Info files whose location is specified this way.)
  3. If you are using Emacs, list the name of the file in a second dir file, in its directory; and then add the name of that directory to the Info-directory-list variable in your personal or site initialization file.

    This variable tells Emacs where to look for dir files (the files must be named dir). Emacs merges the files named dir from each of the listed directories. (In Emacs version 18, you can set the Info-directory variable to the name of only one directory.)

For example, to reach a test file in the /home/bob/info directory, you could add an entry like this to the menu in the standard dir file:

* Test: (/home/bob/info/info-test).  Bob's own test file.

In this case, the absolute file name of the info-test file is written as the second part of the menu entry.

If you don’t want to edit the system dir file, you can tell Info where to look by setting the INFOPATH environment variable in your shell startup file. This works with both the Emacs and standalone Info readers.

Emacs uses the INFOPATH environment variable to initialize the value of Emacs’s own Info-directory-list variable. The standalone Info reader merges any files named dir in any directory listed in the INFOPATH variable into a single menu presented to you in the node called ‘(dir)Top’.

However you set INFOPATH, if its last character is a colon (on MS-DOS/MS-Windows systems, use a semicolon instead), this is replaced by the default (compiled-in) path. This gives you a way to augment the default path with new directories without having to list all the standard places. For example (using sh syntax):

INFOPATH=/home/bob/info:
export INFOPATH

will search /home/bob/info first, then the standard directories. Leading or doubled colons are not treated specially.

When you create your own dir file for use with Info-directory-list or INFOPATH, it’s easiest to start by copying an existing dir file and replace all the text after the ‘* Menu:’ with your desired entries. That way, the punctuation and special CTRL-_ characters that Info needs will be present.

As one final alternative, which works only with Emacs Info, you can change the Info-directory-list variable. For example:

(add-hook 'Info-mode-hook '(lambda ()
	     (add-to-list 'Info-directory-list
			  (expand-file-name "~/info"))))

20.2.4 Installing Info Directory Files

When you install an Info file onto your system, you can use the program install-info to update the Info directory file dir. Normally the makefile for the package runs install-info, just after copying the Info file into its proper installed location.

In order for the Info file to work with install-info, you include the commands @dircategory and @direntry@end direntry in the Texinfo source file. Use @direntry to specify the menu entries to add to the Info directory file, and use @dircategory to specify which part of the Info directory to put it in. Here is how these commands are used in this manual:

@dircategory Texinfo documentation system
@direntry
* Texinfo: (texinfo).           The GNU documentation format.
* install-info: (texinfo)Invoking install-info. …
…
@end direntry

Here’s what this produces in the Info file:

INFO-DIR-SECTION Texinfo documentation system
START-INFO-DIR-ENTRY
* Texinfo: (texinfo).           The GNU documentation format.
* install-info: (texinfo)Invoking install-info. …
…
END-INFO-DIR-ENTRY

The install-info program sees these lines in the Info file, and that is how it knows what to do.

Always use the @direntry and @dircategory commands near the beginning of the Texinfo input, before the first @node command. If you use them later on in the input, install-info will not notice them.

install-info will automatically reformat the description of the menu entries it is adding. As a matter of convention, the description of the main entry (above, ‘The GNU documentation format’) should start at column 32, starting at zero (as in what-cursor-position in Emacs). This will make it align with most others. Description for individual utilities best start in column 48, where possible. For more information about formatting see the ‘--calign’, ‘--align’, and ‘--max-width’ options in Invoking install-info.

If you use @dircategory more than once in the Texinfo source, each usage specifies the ‘current’ category; any subsequent @direntry commands will add to that category.

When choosing a category name for the @dircategory command, we recommend consulting the Free Software Directory. If your program is not listed there, or listed incorrectly or incompletely, please report the situation to the directory maintainers (http://directory.fsf.org) so that the category names can be kept in sync.

Here are a few examples (see the util/dir-example file in the Texinfo distribution for large sample dir file):

Emacs
Localization
Printing
Software development
Software libraries
Text creation and manipulation

Each ‘Invoking’ node for every program installed should have a corresponding @direntry. This lets users easily find the documentation for the different programs they can run, as with the traditional man system.


20.2.5 Invoking install-info

install-info inserts menu entries from an Info file into the top-level dir file in the Info system (see the previous sections for an explanation of how the dir file works). install-info also removes menu entries from the dir file. It’s most often run as part of software installation, or when constructing a dir file for all manuals on a system. Synopsis:

install-info [option…] [info-file [dir-file]]

If info-file or dir-file are not specified, the options (described below) that define them must be. There are no compile-time defaults, and standard input is never used. install-info can read only one Info file and write only one dir file per invocation.

If dir-file (however specified) does not exist, install-info creates it if possible (with no entries).

If any input file is compressed with gzip (see Gzip), install-info automatically uncompresses it for reading. And if dir-file is compressed, install-info also automatically leaves it compressed after writing any changes. If dir-file itself does not exist, install-info tries to open dir-file.gz, dir-file.xz, dir-file.bz2, dir-file.lz, and dir-file.lzma, in that order.

Options:

--add-once

Specifies that the entry or entries will only be put into a single section.

--align=column

Specifies the column that the second and subsequent lines of menu entry’s description will be formatted to begin at. The default for this option is ‘35’. It is used in conjunction with the ‘--max-width’ option. column starts counting at 1.

--append-new-sections

Instead of alphabetizing new sections, place them at the end of the DIR file.

--calign=column

Specifies the column that the first line of menu entry’s description will be formatted to begin at. The default for this option is ‘33’. It is used in conjunction with the ‘--max-width’ option. When the name of the menu entry exceeds this column, entry’s description will start on the following line. column starts counting at 1.

--debug

Report what is being done.

--delete

Delete the entries in info-file from dir-file. The file name in the entry in dir-file must be info-file (except for an optional ‘.info’ in either one). Don’t insert any new entries. Any empty sections that result from the removal are also removed.

--description=text

Specify the explanatory portion of the menu entry. If you don’t specify a description (either via ‘--entry’, ‘--item’ or this option), the description is taken from the Info file itself.

--dir-file=name

Specify file name of the Info directory file. This is equivalent to using the dir-file argument.

--dry-run

Same as ‘--test’.

--entry=text

Insert text as an Info directory entry; text should have the form of an Info menu item line plus zero or more extra lines starting with whitespace. If you specify more than one entry, they are all added. If you don’t specify any entries, they are determined from information in the Info file itself.

--help

Display a usage message with basic usage and all available options, then exit successfully.

--info-file=file

Specify Info file to install in the directory. This is equivalent to using the info-file argument.

--info-dir=dir

Specify the directory where the directory file dir resides. Equivalent to ‘--dir-file=dir/dir’.

--infodir=dir

Same as ‘--info-dir’.

--item=text

Same as ‘--entry=text’. An Info directory entry is actually a menu item.

--keep-old

Do not replace pre-existing menu entries. When ‘--remove’ is specified, this option means that empty sections are not removed.

--max-width=column

Specifies the column that the menu entry’s description will be word-wrapped at. column starts counting at 1.

--maxwidth=column

Same as ‘--max-width’.

--menuentry=text

Same as ‘--name’.

--name=text

Specify the name portion of the menu entry. If the text does not start with an asterisk ‘*’, it is presumed to be the text after the ‘*’ and before the parentheses that specify the Info file. Otherwise text is taken verbatim, and is taken as defining the text up to and including the first period (a space is appended if necessary). If you don’t specify the name (either via ‘--entry’, ‘--item’ or this option), it is taken from the Info file itself. If the Info does not contain the name, the basename of the Info file is used.

--no-indent

Suppress formatting of new entries into the dir file.

--quiet
--silent

Suppress warnings, etc., for silent operation.

--remove

Same as ‘--delete’.

--remove-exactly

Also like ‘--delete’, but only entries if the Info file name matches exactly; .info and/or .gz suffixes are not ignored.

--section=sec

Put this file’s entries in section sec of the directory. If you specify more than one section, all the entries are added in each of the sections. If you don’t specify any sections, they are determined from information in the Info file itself. If the Info file doesn’t specify a section, the menu entries are put into the Miscellaneous section.

--section regex sec

Same as ‘--regex=regex --section=sec --add-once’.

install-info tries to detect when this alternate syntax is used, but does not always guess correctly. Here is the heuristic that install-info uses:

  1. If the second argument to --section starts with a hyphen, the original syntax is presumed.
  2. If the second argument to --section is a file that can be opened, the original syntax is presumed.
  3. Otherwise the alternate syntax is used.

When the heuristic fails because your section title starts with a hyphen, or it happens to be a file that can be opened, the syntax should be changed to ‘--regex=regex --section=sec --add-once’.

--regex=regex

Put this file’s entries into any section that matches regex. If more than one section matches, all of the entries are added in each of the sections. Specify regex using basic regular expression syntax, more or less as used with grep, for example.

--test

Suppress updating of the directory file.

--version

Display version information and exit successfully.


20.3 Info Format FAQ

Here are some questions that have been frequently asked on the project mailing lists and elsewhere.

Why when I run ‘info foo’ do I get the same output as ‘man foo’?

Check that the Info manuals are installed. Not all GNU/Linux distributions install all the Info manuals by default. This is regrettable, as often the Info manual provides more information than the provided man page.

Why not use HTML instead of Info?

Manuals are rarely written in the Info format itself, but are generated from Texinfo source by the texi2any program. texi2any can generate HTML as well as Info from Texinfo source. You can also access and download HTML manuals from the GNU website (https://www.gnu.org/manual/manual.html). Additionally, some GNU/Linux distributions provide packages for the installation of HTML manuals.

Info still has some advantages over HTML for locally installed documentation. The Info browsers support index lookup and support for links between locally installed manuals in multiple locations. It’s important to have documentation installed locally on your machine rather than relying on the Internet; this makes accessing documentation more reliable, and ensures it matches installed versions of packages. It’s hoped that support for browsing locally installed Texinfo documentation in some form of HTML can be improved in the future.

Why provide the command-line info program when it is not very good?

info has had improvements over time, so perhaps give it another go. It can be configured to change the default key bindings, as well as change the color of cross-references and search results, and to enable mouse scrolling support. You should not need to be experienced with the Emacs editor to use info. (Some recommend another program called pinfo, but this lacks in important features like index lookup.)

Some prefer to be able to scroll through the entire manual at once (as happens with man pages), rather than being limited to a single ‘node’ of content at once. Of course, there is no accounting for taste, but a single, unstructured block of text becomes more awkward as a manual becomes longer and more complicated. (However, if you really want to, you can view an info manual all at once in the less pager with ‘info foo | less’.)

Why do I have ‘see’ appearing in front of all of my links?

By default, Emacs Info mode either changes the marker ‘*note’ for cross-references to ‘see’, or hides it completely. Unfortunately, there is no way to do this reliably, as both the @pxref and @ref commands in Texinfo output this marker in the Info output, and the ‘see’ text is only appropriate for @pxref.

Yes, but how do I get a plain link, with no extra markup?

You can’t. Info is a plain text format that is displayed mostly as-is in the viewers, and without the ‘*note’ text there would be nothing to mark text as a link. Additionally, in printed output there is no such thing as a plain link, as the page number of the target would have to be printed somewhere.

If you really want a plain link in HTML output without affecting other output formats, you could create a macro with conditional definitions for each output format.

Why do lines containing links appear mysteriously short?

This due to Emacs (or info with hide-note-references set to ‘On’) hiding ‘*note’ strings, as mentioned above.

Can the Info format be extended to support fonts, colors or reflowable text?

Any extension would be incompatible with existing Info-viewing programs. Extra markup added to Info files would be displayed to the user, making files ugly and unreadable.

When Texinfo files are processed into Info files, information about which Texinfo commands were used to markup text is lost. Moreover, it is not possible to reliably detect whether text can be reflowed (e.g. a paragraph of prose), or whether line breaks should be kept (e.g. a code sample, or poem).

Info’s core purpose is to display documentation on text terminals. If you want more, you are recommended to use the HTML output from texi2any instead.


21 Generating HTML

makeinfo generates Info output by default, but given the --html option, it will generate HTML, for web browsers and other programs. This chapter gives some details on such HTML output.

makeinfo has many user-definable customization variables with which you can influence the HTML output. See Customization Variables.

makeinfo can also produce output in XML and DocBook formats, but we do not as yet describe these in detail. See Output Formats, for a brief overview of all the output formats.


21.1 HTML Translation

First, the HTML generated by makeinfo is standard HTML 4. When first written, it also tried to be compatible with earlier standards (e.g., HTML 2.0, RFC-1866).

Please report output from an error-free run of makeinfo which has practical browser portability problems as a bug (see Reporting Bugs).

Some known exceptions to HTML 3.2 (using ‘--init-file=html32.pm’ produced strict HTML 3.2 output, but this has not been tested lately; see Invoking texi2any/makeinfo from a Shell):

  1. HTML 3.2 tables are generated for the @multitable command (see @multitable: Multi-column Tables), but they should degrade reasonably in browsers without table support.
  2. The HTML 4 ‘id’ attribute is used.
  3. The HTML 4 ‘lang’ attribute on the ‘<html>’ attribute is used.
  4. Entities that are not in the HTML 3.2 standard are also used.
  5. CSS is used (see HTML CSS).
  6. Some HTML 4 elements are used: span, thead, abbr, acronym.

To achieve maximum portability and accessibility among browsers (both graphical and text-based), systems, and users, the HTML output is intentionally quite plain and generic. It has always been our goal for users to be able to customize the output to their wishes via CSS (see HTML CSS) or other means (see Customization Variables). If you cannot accomplish a reasonable customization, feel free to report that.

However, we do not wish to depart from our basic goal of widest readability for the core output. For example, using fancy CSS may make it possible for the HTML output to more closely resemble the TeX output in some details, but this result is not even close to being worth the ensuing difficulties.

It is also intentionally not our goal, and not even possible, to pass through every conceivable validation test without any diagnostics. Different validation tests have different goals, often about pedantic enforcement of some standard or another. Our overriding goal is to help users, not blindly comply with standards.

To repeat what was said at the top: please report output from an error-free run of makeinfo which has practical browser portability problems as a bug (see Reporting Bugs).

A few other general points about the HTML output follow.

Navigation bar: By default, a navigation bar is inserted at the start of each node, analogous to Info output. If the ‘--no-headers’ option is used, the navigation bar is only inserted at the beginning of split files. Header <link> elements in split output can support Info-like navigation with browsers like Lynx and Emacs W3 which implement this HTML 1.0 feature.

Footnotes: for HTML, when the footnote style is ‘end’, or if the output is not split, footnotes are put at the end of the output. If the footnote style is set to ‘separate’, and the output is split, they are placed in a separate file. See Footnote Styles.

Raw HTML: makeinfo will include segments of Texinfo source between @ifhtml and @end ifhtml in the HTML output (but not any of the other conditionals, by default). Source between @html and @end html is passed without change to the output (i.e., suppressing the normal escaping of input ‘<’, ‘>’ and ‘&’ characters which have special significance in HTML). See Conditional Commands.


21.2 HTML Splitting

When splitting output at nodes (which is the default), makeinfo writes HTML output into (basically) one output file per Texinfo source @node.

Each output file name is the node name with spaces replaced by ‘-’’s and special characters changed to ‘_’ followed by their code point in hex (see HTML Cross-references). This is to make it portable and easy to use as a file name. In the unusual case of two different nodes having the same name after this treatment, they are written consecutively to the same file, with HTML anchors so each can be referred to independently.

If makeinfo is run on a system which does not distinguish case in file names, nodes which are the same except for case (e.g., ‘index’ and ‘Index’) will also be folded into the same output file with anchors. You can also pretend to be on a case insensitive filesystem by setting the customization variable CASE_INSENSITIVE_FILENAMES.

It is also possible to split at chapters or sections with --split (see Invoking texi2any/makeinfo from a Shell). In that case, the file names are constructed after the name of the node associated with the relevant sectioning command. Also, unless --no-node-files is specified, a redirection file is output for every node in order to more reliably support cross-references to that manual (see HTML Cross-references).

When splitting, the HTML output files are written into a subdirectory, with the name chosen as follows:

  1. makeinfo first tries the subdirectory with the base name from @setfilename (that is, any extension is removed). For example, HTML output for @setfilename gcc.info would be written into a subdirectory named ‘gcc/’.
  2. If that directory cannot be created for any reason, then makeinfo tries appending ‘.html’ to the directory name. For example, output for @setfilename texinfo would be written to ‘texinfo.html/’.
  3. If the ‘name.html’ directory can’t be created either, makeinfo gives up.

In any case, the top-level output file within the directory is always named ‘index.html’.

Monolithic output (--no-split) is named according to @setfilename (with any ‘.info’ extension is replaced with ‘.html’), --output (the argument is used literally), or based on the input file name as a last resort (see @setfilename: Set the Output File Name).


21.3 HTML CSS

Cascading Style Sheets (CSS for short) is an Internet standard for influencing the display of HTML documents: see http://www.w3.org/Style/CSS/.

By default, makeinfo includes a few simple CSS commands to better implement the appearance of some Texinfo environments. Here is one of them, as an example:

pre.display { font-family:inherit }

A full explanation of CSS is (far) beyond this manual; please see the reference above. In brief, however, the above tells the web browser to use the same font as the main document for @display. By default, the HTML ‘<pre>’ command uses a monospaced font.

You can influence the CSS in the HTML output with two makeinfo options: --css-include=file and --css-ref=url.

The option --css-ref=url adds to each output HTML file a ‘<link>’ tag referencing a CSS at the given url. This allows using external style sheets. You may find the file texi2html/examples/texinfo-bright-colors.css useful for visualizing the CSS elements in Texinfo output.

The option --css-include=file includes the contents file in the HTML output, as you might expect. However, the details are somewhat tricky, as described in the following, to provide maximum flexibility.

The CSS file may begin with so-called ‘@import’ directives, which link to external CSS specifications for browsers to use when interpreting the document. Again, a full description is beyond our scope here, but we’ll describe how they work syntactically, so we can explain how makeinfo handles them.

There can be more than one ‘@import’, but they have to come first in the file, with only whitespace and comments interspersed, no normal definitions. (Technical exception: a ‘@charset’ directive may precede the ‘@import’’s. This does not alter makeinfo’s behavior, it just copies the ‘@charset’ if present.) Comments in CSS files are delimited by ‘/* ... */’, as in C. An ‘@import’ directive must be in one of these two forms:

@import url(http://example.org/foo.css);
@import "http://example.net/bar.css";

As far as makeinfo is concerned, the crucial characters are the ‘@’ at the beginning and the semicolon terminating the directive. When reading the CSS file, it simply copies any such ‘@’-directive into the output, as follows:

  • If file contains only normal CSS declarations, it is included after makeinfo’s default CSS, thus overriding it.
  • If file begins with ‘@import’ specifications (see below), then the ‘import’’s are included first (they have to come first, according to the standard), and then makeinfo’s default CSS is included. If you need to override makeinfo’s defaults from an ‘@import’, you can do so with the ‘! important’ CSS construct, as in:
    pre.example { font-size: inherit ! important }
    
  • If file contains both ‘@import’ and inline CSS specifications, the ‘@import’’s are included first, then makeinfo’s defaults, and lastly the inline CSS from file.
  • Any @-directive other than ‘@import’ and ‘@charset’ is treated as a CSS declaration, meaning makeinfo includes its default CSS and then the rest of the file.

If the CSS file is malformed or erroneous, makeinfo’s output is unspecified. makeinfo does not try to interpret the meaning of the CSS file in any way; it just looks for the special ‘@’ and ‘;’ characters and blindly copies the text into the output. Comments in the CSS file may or may not be included in the output.

In addition to the possibilities offered by CSS, makeinfo has many user-definable customization variables with which you can influence the HTML output. See Customization Variables.


21.4 HTML Cross-references

Cross-references between Texinfo manuals in HTML format become, in the end, a standard HTML <a> link, but the details are unfortunately complex. This section describes the algorithm used in detail, so that Texinfo can cooperate with other programs, such as texi2html, by writing mutually compatible HTML files.

This algorithm may or may not be used for links within HTML output for a Texinfo file. Since no issues of compatibility arise in such cases, we do not need to specify this.

We try to support references to such “external” manuals in both monolithic and split forms. A monolithic (mono) manual is entirely contained in one file, and a split manual has a file for each node. (See HTML Splitting.)

The algorithm was primarily devised by Patrice Dumas in 2003–04.


21.4.2 HTML Cross-reference Node Name Expansion

As mentioned in the previous section, the key part of the HTML cross reference algorithm is the conversion of node names in the Texinfo source into strings suitable for XHTML identifiers and file names. The restrictions are similar for each: plain ASCII letters, numbers, and the ‘-’ and ‘_’ characters are all that can be used. (Although HTML anchors can contain most characters, XHTML is more restrictive.)

Cross-references in Texinfo can refer either to nodes or anchors (see @anchor: Defining Arbitrary Cross-reference Targets). However, anchors are treated identically to nodes in this context, so we’ll continue to say “node” names for simplicity.

A special exception: the Top node (see The ‘Top’ Node and Master Menu) is always mapped to the file index.html, to match web server software. However, the HTML target is ‘Top’. Thus (in the split case):

@xref{Top,,, emacs, The GNU Emacs Manual}.
⇒ <a href="emacs/index.html#Top">
  1. The standard ASCII letters (a-z and A-Z) are not modified. All other characters may be changed as specified below.
  2. The standard ASCII numbers (0-9) are not modified except when a number is the first character of the node name. In that case, see below.
  3. Multiple consecutive space, tab and newline characters are transformed into just one space. (It’s not possible to have newlines in node names with the current implementation, but we specify it anyway, just in case.)
  4. Leading and trailing spaces are removed.
  5. After the above has been applied, each remaining space character is converted into a ‘-’ character.
  6. Other ASCII 7-bit characters are transformed into ‘_00xx’, where xx is the ASCII character code in (lowercase) hexadecimal. This includes ‘_’, which is mapped to ‘_005f’.
  7. If the node name does not begin with a letter, the literal string ‘g_t’ is prefixed to the result. (Due to the rules above, that string can never occur otherwise; it is an arbitrary choice, standing for “GNU Texinfo”.) This is necessary because XHTML requires that identifiers begin with a letter.

For example:

@node A  node --- with _'%
⇒ A-node-_002d_002d_002d-with-_005f_0027_0025

Example translations of common characters:

  • _’ ⇒ ‘_005f
  • -’ ⇒ ‘_002d
  • A node’ ⇒ ‘A-node

On case-folding computer systems, nodes differing only by case will be mapped to the same file. In particular, as mentioned above, Top always maps to the file index.html. Thus, on a case-folding system, Top and a node named ‘Index’ will both be written to index.html. Fortunately, the targets serve to distinguish these cases, since HTML target names are always case-sensitive, independent of operating system.


21.4.3 HTML Cross-reference Command Expansion

Node names may contain @-commands (see @node Line Requirements). This section describes how they are handled.

First, comments are removed.

Next, any @value commands (see @set and @value) and macro invocations (see Invoking Macros) are fully expanded.

Then, for the following commands, the command name and braces are removed, and the text of the argument is recursively transformed:

@asis @b @cite @code @command @dfn @dmn @dotless
@emph @env @file @i @indicateurl @kbd @key
@samp @sansserif @sc @slanted @strong @sub @sup
@t @U @var @verb @w

For @sc, any letters are capitalized.

In addition, the following commands are replaced by constant text, as shown below. If any of these commands have non-empty arguments, as in @TeX{bad}, it is an error, and the result is unspecified. In this table, ‘(space)’ means a space character and ‘(nothing)’ means the empty string. The notation ‘U+hhhh’ means Unicode code point hhhh (in hex, as usual).

There are further transformations of many of these expansions to yield the final file or other target name, such as space characters to ‘-’, etc., according to the other rules.

@(newline)(space)
@(space)(space)
@(tab)(space)
@!!
@*(space)
@-(nothing)
@..
@:(nothing)
@??
@@@
@{{
@}}
@LaTeXLaTeX
@TeXTeX
@arrowU+2192
@bulletU+2022
@comma,
@copyrightU+00A9
@dotsU+2026
@enddots...
@equivU+2261
@errorerror-->
@euroU+20AC
@exclamdownU+00A1
@expansionU+21A6
@geqU+2265
@leqU+2264
@minusU+2212
@ordfU+00AA
@ordmU+00BA
@pointU+2605
@poundsU+00A3
@printU+22A3
@questiondownU+00BF
@registeredsymbolU+00AE
@resultU+21D2
@textdegreeU+00B0
@tie(space)

Quotation mark @-commands (@quotedblright{} and the like), are likewise replaced by their Unicode values. Normal quotation characters (e.g., ASCII ‘ and ’) are not altered. See Inserting Quotation Marks.

Any @acronym, @abbr, @email, and @image commands are replaced by their first argument. (For these commands, all subsequent arguments are optional, and ignored here.) See @acronym{acronym[, meaning]}, and @email{email-address[, displayed-text]}, and Inserting Images.

Accents are handled according to the next section.

Any other command is an error, and the result is unspecified.


21.4.4 HTML Cross-reference 8-bit Character Expansion

Usually, characters other than plain 7-bit ASCII are transformed into the corresponding Unicode code point(s) in Normalization Form C, which uses precomposed characters where available. (This is the normalization form recommended by the W3C and other bodies.) This holds when that code point is 0xffff or less, as it almost always is.

These will then be further transformed by the rules above into the string ‘_hhhh’, where hhhh is the code point in hex.

For example, combining this rule and the previous section:

@node @b{A} @TeX{} @u{B} @point{}@enddots{}
⇒ A-TeX-B_0306-_2605_002e_002e_002e

Notice: 1) @enddots expands to three periods which in turn expands to three ‘_002e’’s; 2) @u{B} is a ‘B’ with a breve accent, which does not exist as a pre-accented Unicode character, therefore expands to ‘B_0306’ (B with combining breve).

When the Unicode code point is above 0xffff, the transformation is ‘__xxxxxx’, that is, two leading underscores followed by six hex digits. Since Unicode has declared that their highest code point is 0x10ffff, this is sufficient. (We felt it was better to define this extra escape than to always use six hex digits, since the first two would nearly always be zeros.)

This method works fine if the node name consists mostly of ASCII characters and contains only few 8-bit ones. But if the document is written in a language whose script is not based on the Latin alphabet (for example, Ukrainian), it will create file names consisting almost entirely of ‘_xxxx’ notations, which is inconvenient and all but unreadable. To handle such cases, makeinfo offers the --transliterate-file-names command line option. This option enables transliteration of node names into ASCII characters for the purposes of file name creation and referencing. The transliteration is based on phonetic principles, which makes the generated file names more easily understandable.

For the definition of Unicode Normalization Form C, see Unicode report UAX#15, http://www.unicode.org/reports/tr15/. Many related documents and implementations are available elsewhere on the web.


21.4.5 HTML Cross-reference Mismatch

As mentioned earlier (see HTML Cross-reference Link Basics), the generating software may need to guess whether a given manual being cross-referenced is available in split or monolithic form—and, inevitably, it might guess wrong. However, when the referent manual is generated, it is possible to handle at least some mismatches.

In the case where we assume the referent is split, but it is actually available in mono, the only recourse would be to generate a manual/ subdirectory full of HTML files which redirect back to the monolithic manual.html. Since this is essentially the same as a split manual in the first place, it’s not very appealing.

On the other hand, in the case where we assume the referent is mono, but it is actually available in split, it is possible to use JavaScript to redirect from the putatively monolithic manual.html to the different manual/node.html files. Here’s an example:

function redirect() {
  switch (location.hash) {
    case "#Node1":
      location.replace("manual/Node1.html#Node1"); break;
    case "#Node2" :
      location.replace("manual/Node2.html#Node2"); break;
    …
    default:;
  }
}

Then, in the <body> tag of manual.html:

<body onLoad="redirect();">

Once again, this is something the software which generated the referent manual has to do in advance, it’s not something the software generating the cross-reference in the present manual can control.


21.4.6 HTML Cross-reference Configuration: htmlxref.cnf

makeinfo reads a file named htmlxref.cnf to gather information for cross-references to other manuals in HTML output. It is looked for in the following directories:

./

(the current directory)

./.texinfo/

(under the current directory)

~/.texinfo/

(where ~ is the current user’s home directory)

sysconfdir/texinfo/

(where sysconfdir is the system configuration directory specified at compile-time, e.g., /usr/local/etc)

datadir/texinfo/

(likewise specified at compile time, e.g., /usr/local/share)

All files found are used, with earlier entries overriding later ones. The Texinfo distribution includes a default file which handles many GNU manuals; it is installed in the last of the above directories, i.e., datadir/texinfo/htmlxref.cnf.

The file is line-oriented. Lines consisting only of whitespace are ignored. Comments are indicated with a ‘#’ at the beginning of a line, optionally preceded by whitespace. Since ‘#’ can occur in urls (like almost any character), it does not otherwise start a comment.

Each non-blank non-comment line must be either a variable assignment or manual information.

A variable assignment line looks like this:

varname = varvalue

Whitespace around the ‘=’ is optional and ignored. The varname should consist of letters; case is significant. The varvalue is an arbitrary string, continuing to the end of the line. Variables are then referenced with ‘${varname}’; variable references can occur in the varvalue.

A manual information line looks like this:

manual keyword urlprefix

with manual the short identifier for a manual, keyword being one of: mono, node, section, chapter, and urlprefix described below. Variable references can occur only in the urlprefix. For example (used in the canonical htmlxref.cnf):

G = http://www.gnu.org
GS = ${G}/software
hello mono    ${GS}/hello/manual/hello.html
hello chapter ${GS}/hello/manual/html_chapter/
hello section ${GS}/hello/manual/html_section/
hello node    ${GS}/hello/manual/html_node/

If the keyword is mono, urlprefix gives the host, directory, and file name for manual as one monolithic file.

If the keyword is node, section, or chapter, urlprefix gives the host and directory for manual split into nodes, sections, or chapters, respectively.

When available, makeinfo will use the “corresponding” value for cross-references between manuals. That is, when generating monolithic output (--no-split), the mono url will be used, when generating output that is split by node, the node url will be used, etc. However, if a manual is not available in that form, anything that is available can be used. Here is the search order for each style:

node    ⇒ node,    section, chapter, mono
section ⇒ section, chapter, node,    mono
chapter ⇒ chapter, section, node,    mono
mono    ⇒ mono,    chapter, section, node

These section- and chapter-level cross-manual references can succeed only when the target manual was created using --node-files; this is the default for split output.

If you have additions or corrections to the htmlxref.cnf distributed with Texinfo, please email bug-texinfo@gnu.org as usual. You can get the latest version from http://ftpmirror.gnu.org/texinfo/htmlxref.cnf.


21.5 @documentdescription: Summary Text

When producing HTML output for a document, makeinfo writes a ‘<meta>’ element in the ‘<head>’ to give some idea of the content of the document. By default, this description is the title of the document, taken from the @settitle command (see @settitle: Set the Document Title). To change this, use the @documentdescription environment, as in:

@documentdescription
descriptive text.
@end documentdescription

This will produce the following output in the ‘<head>’ of the HTML:

<meta name=description content="descriptive text.">

@documentdescription must be specified before the first node of the document.


Next: , Previous: , Up: Texinfo   [Contents][Index]

Appendix A @-Command Details

Here are the details of @-commands: information about their syntax, a list of commands, and information about where commands can appear.


A.1 @-Command Syntax

Texinfo has the following types of @-command:

1. Brace commands

These commands start with @ followed by a letter or a word, followed by an argument within braces. For example, the command @dfn indicates the introductory or defining use of a term; it is used as follows: ‘In Texinfo, @@-commands are @dfn{mark-up} commands.

2. Line commands

These commands occupy an entire line. The line starts with @, followed by the name of the command (a word); for example, @center or @cindex. If no argument is needed, the word is followed by the end of the line. If there is an argument, it is separated from the command name by a space. Braces are not used.

3. Block commands

These commands are written at the start of a line, with general text on following lines, terminated by a matching @end command on a line of its own. For example, @example, then the lines of a coding example, then @end example. Some of these block commands take arguments as line commands do; for example, @enumerate A opening an environment terminated by @end enumerate. Here ‘A’ is the argument.

4. Symbol insertion commands with no arguments

These commands start with @ followed by a word followed by a left and right- brace. These commands insert special symbols in the document; they do not take arguments. Some examples: @dots{} ⇒ ‘’, @equiv{} ⇒ ‘’, @TeX{} ⇒ ‘TeX’, and @bullet{} ⇒ ‘’.

5. Non-alphabetic commands

The names of commands in all of the above categories consist of alphabetic characters, almost entirely in lower-case. Unlike those, the non-alphabetic commands consist of an @ followed by a punctuation mark or other character that is not part of the Latin alphabet. Non-alphabetic commands are almost always part of text within a paragraph. The non-alphabetic commands include @@, @{, @}, @., @SPACE, and most of the accent commands.

6. Miscellaneous commands

There are a handful of commands that don’t fit into any of the above categories; for example, the obsolete command @refill, which is always used at the end of a paragraph immediately following the final period or other punctuation character. @refill takes no argument and does not require braces. Likewise, @tab used in a @multitable block does not take arguments, and is not followed by braces.

Thus, the alphabetic commands fall into classes that have different argument syntaxes. You cannot tell to which class a command belongs by the appearance of its name, but you can tell by the command’s meaning: if the command stands for a glyph, it is in class 4 and does not require an argument; if it makes sense to use the command among other text as part of a paragraph, the command is in class 1 and must be followed by an argument in braces. The non-alphabetic commands, such as @:, are exceptions to the rule; they do not need braces.

The purpose of having different syntax for commands is to make Texinfo files easier to read, and also to help the GNU Emacs paragraph and filling commands work properly.


A.2 @-Command List

Here is an alphabetical list of the @-commands in Texinfo. Square brackets, [ ], indicate optional arguments; an ellipsis, ‘’, indicates repeated text.

@whitespace

An @ followed by a space, tab, or newline produces a normal, stretchable, interword space. See Multiple Spaces.

@!

Produce an exclamation point that ends a sentence (usually after an end-of-sentence capital letter). See Ending a Sentence.

@"
@'

Generate an umlaut or acute accent, respectively, over the next character, as in ö and ó. See Inserting Accents.

@&
@ampchar{}

Generate an ampersand. See Inserting ‘&’ with @& and @ampchar{}.

@*

Force a line break. See @* and @/: Generate and Allow Line Breaks.

@,{c}

Generate a cedilla accent under c, as in ç. See Inserting Accents.

@-

Insert a discretionary hyphenation point. See @- and @hyphenation: Helping TeX Hyphenate.

@.

Produce a period that ends a sentence (usually after an end-of-sentence capital letter). See Ending a Sentence.

@/

Produces no output, but allows a line break. See @* and @/: Generate and Allow Line Breaks.

@:

Tell TeX to refrain from inserting extra whitespace after an immediately preceding period, question mark, exclamation mark, or colon, as TeX normally would. See Not Ending a Sentence.

@=

Generate a macron (bar) accent over the next character, as in ō. See Inserting Accents.

@?

Produce a question mark that ends a sentence (usually after an end-of-sentence capital letter). See Ending a Sentence.

@@
@atchar{}

Insert an at sign, ‘@’. See Inserting ‘@’ with @@ and @atchar{}.

@\
@backslashchar{}

Insert a backslash, ‘\’; @backslashchar{} works anywhere, while @\ works only inside @math. See Inserting ‘\’ with @backslashchar{}, and @math and @displaymath: Formatting Mathematics.

@^
@`

Generate a circumflex (hat) or grave accent, respectively, over the next character, as in ô and è. See Inserting Accents.

@{
@lbracechar{}

Insert a left brace, ‘{’. See Inserting ‘{ ‘}’ with @{ @} and @l rbracechar{}.

@}
@rbracechar{}

Insert a right brace, ‘}’. See Inserting ‘{ ‘}’ with @{ @} and @l rbracechar{}.

@~

Generate a tilde accent over the next character, as in Ñ. See Inserting Accents.

@AA{}
@aa{}

Generate the uppercase and lowercase Scandinavian A-ring letters, respectively: Å, å. See Inserting Accents.

@abbr{abbreviation}

Indicate a general abbreviation, such as ‘Comput.’. See @abbr{abbreviation[, meaning]}.

@acronym{acronym}

Indicate an acronym in all capital letters, such as ‘NASA’. See @acronym{acronym[, meaning]}.

@AE{}
@ae{}

Generate the uppercase and lowercase AE ligatures, respectively: Æ, æ. See Inserting Accents.

@afivepaper

Change page dimensions for the A5 paper size. See Printing on A4 Paper.

@afourlatex
@afourpaper
@afourwide

Change page dimensions for the A4 paper size. See Printing on A4 Paper.

@alias new=existing

Make the command ‘@new’ a synonym for the existing command ‘@existing’. See @alias new=existing.

@allowcodebreaks true-false

Control breaking at ‘-’ and ‘_’ in TeX. See @allowcodebreaks: Control Line Breaks in @code.

@anchor{name}

Define name as the current location for use as a cross-reference target. See @anchor: Defining Arbitrary Cross-reference Targets.

@appendix title

Begin an appendix. The title appears in the table of contents. In Info, the title is underlined with asterisks. See @unnumbered, @appendix: Chapters with Other Labeling.

@appendixsec title
@appendixsection title

Begin an appendix section within an appendix. The section title appears in the table of contents. In Info, the title is underlined with equal signs. @appendixsection is a longer spelling of the @appendixsec command. See @unnumberedsec, @appendixsec, @heading.

@appendixsubsec title

Begin an appendix subsection. The title appears in the table of contents. In Info, the title is underlined with hyphens. See The @subsection-like Commands.

@appendixsubsubsec title

Begin an appendix subsubsection. The title appears in the table of contents. In Info, the title is underlined with periods. See @subsection and Other Subsub Commands.

@arrow{}

Generate a right arrow glyph: ‘’. Used by default for @click. See Click Sequences.

@asis

Used following @table, @ftable, and @vtable to print the table’s first column without highlighting (“as is”). See @asis.

@author author

Typeset author flushleft and underline it. See @title, @subtitle, and @author.

@b{text}

Set text in a bold font. No effect in Info. See Fonts for Printing.

@bullet{}

Generate a large round dot, • (‘*’ in Info). Often used with @table. See @bullet (•).

@bsixpaper

Change page dimensions for the B6 paper size. See Printing on A4 Paper.

@bye

Stop formatting a file. The formatters do not see anything in the input file following @bye. See Ending a Texinfo File.

@c comment

Begin a comment in Texinfo. The rest of the line does not appear in any output. A synonym for @comment. DEL also starts a comment. See Comments.

@caption

Define the full caption for a @float. See @caption & @shortcaption.

@cartouche

Highlight an example or quotation by drawing a box with rounded corners around it. Pair with @end cartouche. No effect in Info. See @cartouche: Rounded Rectangles.

@center line-of-text

Center the line of text following the command. See @titlefont, @center, and @sp.

@centerchap line-of-text

Like @chapter, but centers the chapter title. See @chapter: Chapter Structuring.

@chapheading title

Print an unnumbered chapter-like heading, but omit from the table of contents. In Info, the title is underlined with asterisks. See @majorheading, @chapheading: Chapter-level Headings.

@chapter title

Begin a numbered chapter. The chapter title appears in the table of contents. In Info, the title is underlined with asterisks. See @chapter: Chapter Structuring.

@cindex entry

Add entry to the index of concepts. See Defining the Entries of an Index.

@cite{reference}

Highlight the name of a book or other reference that has no companion Info file. See @cite{reference}.

@clear flag

Unset flag, preventing the Texinfo formatting commands from formatting text between subsequent pairs of @ifset flag and @end ifset commands, and preventing @value{flag} from expanding to the value to which flag is set. See Flags: @set, @clear, conditionals, and @value.

@click{}

Represent a single “click” in a GUI. Used within @clicksequence. See Click Sequences.

@clicksequence{action @click{} action}

Represent a sequence of clicks in a GUI. See Click Sequences.

@clickstyle @cmd

Execute @cmd for each @click; the default is @arrow. The usual following empty braces on @cmd are omitted. See Click Sequences.

@code{sample-code}

Indicate an expression, a syntactically complete token of a program, or a program name. Unquoted in Info output. See @code{sample-code}.

@codequotebacktick on-off
@codequoteundirected on-off

Control output of ` and ' in code examples. See Inserting Quote Characters.

@comma{}

Insert a comma ‘,’ character; only needed when a literal comma would be taken as an argument separator. See Inserting ‘,’ with @comma{}.

@command{command-name}

Indicate a command name, such as ls. See @command{command-name}.

@comment comment

Begin a comment in Texinfo. The rest of the line does not appear in any output. A synonym for @c. See Comments.

@contents

Print a complete table of contents. Has no effect in Info, which uses menus instead. See Generating a Table of Contents.

@copying

Specify copyright holders and copying conditions for the document. Pair with @end copying. See @copying: Declare Copying Permissions.

@copyright{}

Generate the copyright symbol ©. See @copyright{} (©).

@defcodeindex index-name

Define a new index and its indexing command. Print entries in an @code font. See Defining New Indices.

@defcv category class name
@defcvx category class name

Format a description for a variable associated with a class in object-oriented programming. Takes three arguments: the category of thing being defined, the class to which it belongs, and its name. See Definition Commands.

@deffn category name arguments
@deffnx category name arguments

Format a description for a function, interactive command, or similar entity that may take arguments. @deffn takes as arguments the category of entity being described, the name of this particular entity, and its arguments, if any. See Definition Commands.

@defindex index-name

Define a new index and its indexing command. Print entries in a roman font. See Defining New Indices.

@definfoenclose newcmd, before, after

Must be used within @ifinfo; create a new command @newcmd for Info that marks text by enclosing it in strings that precede and follow the text. See @definfoenclose: Customized Highlighting.

@defivar class instance-variable-name
@defivarx class instance-variable-name

Format a description for an instance variable in object-oriented programming. The command is equivalent to ‘@defcv {Instance Variable} …’. See Definition Commands.

@defmac macroname arguments
@defmacx macroname arguments

Format a description for a macro; equivalent to ‘@deffn Macro …’. See Definition Commands.

@defmethod class method-name arguments
@defmethodx class method-name arguments

Format a description for a method in object-oriented programming; equivalent to ‘@defop Method …’. See Definition Commands.

@defop category class name arguments
@defopx category class name arguments

Format a description for an operation in object-oriented programming. @defop takes as arguments the name of the category of operation, the name of the operation’s class, the name of the operation, and its arguments, if any. See Definition Commands, and Object-Oriented Programming.

@defopt option-name
@defoptx option-name

Format a description for a user option; equivalent to ‘@defvr {User Option} …’. See Definition Commands.

@defspec special-form-name arguments
@defspecx special-form-name arguments

Format a description for a special form; equivalent to ‘@deffn {Special Form} …’. See Definition Commands.

@deftp category name-of-type attributes
@deftpx category name-of-type attributes

Format a description for a data type; its arguments are the category, the name of the type (e.g., ‘int’) , and then the names of attributes of objects of that type. See Definition Commands, and Data Types.

@deftypecv category class data-type name
@deftypecvx category class data-type name

Format a description for a typed class variable in object-oriented programming. See Definition Commands, and Object-Oriented Programming.

@deftypefn category data-type name arguments
@deftypefnx category data-type name arguments

Format a description for a function or similar entity that may take arguments and that is typed. @deftypefn takes as arguments the category of entity being described, the type, the name of the entity, and its arguments, if any. See Definition Commands.

@deftypefnnewline on-off

Specifies whether return types for @deftypefn and similar are printed on lines by themselves; default is off. See Functions in Typed Languages.

@deftypefun data-type function-name arguments
@deftypefunx data-type function-name arguments

Format a description for a function in a typed language. The command is equivalent to ‘@deftypefn Function …’. See Definition Commands.

@deftypeivar class data-type variable-name
@deftypeivarx class data-type variable-name

Format a description for a typed instance variable in object-oriented programming. See Definition Commands, and Object-Oriented Programming.

@deftypemethod class data-type method-name arguments
@deftypemethodx class data-type method-name arguments

Format a description for a typed method in object-oriented programming. See Definition Commands.

@deftypeop category class data-type name arguments
@deftypeopx category class data-type name arguments

Format a description for a typed operation in object-oriented programming. See Definition Commands, and Object-Oriented Programming.

@deftypevar data-type variable-name
@deftypevarx data-type variable-name

Format a description for a variable in a typed language. The command is equivalent to ‘@deftypevr Variable …’. See Definition Commands.

@deftypevr category data-type name
@deftypevrx category data-type name

Format a description for something like a variable in a typed language—an entity that records a value. Takes as arguments the category of entity being described, the type, and the name of the entity. See Definition Commands.

@defun function-name arguments
@defunx function-name arguments

Format a description for a function; equivalent to ‘@deffn Function …’. See Definition Commands.

@defvar variable-name
@defvarx variable-name

Format a description for a variable; equivalent to ‘@defvr Variable …’. See Definition Commands.

@defvr category name
@defvrx category name

Format a description for any kind of variable. @defvr takes as arguments the category of the entity and the name of the entity. See Definition Commands.

@detailmenu

Mark the (optional) detailed node listing in a master menu. See Parts of a Master Menu.

@dfn{term}

Indicate the introductory or defining use of a term. See @dfn{term}.

@DH{}
@dh{}

Generate the uppercase and lowercase Icelandic letter eth, respectively: Ð, ð. See Inserting Accents.

@dircategory dirpart

Specify a part of the Info directory menu where this file’s entry should go. See Installing Info Directory Files.

@direntry

Begin the Info directory menu entry for this file. Pair with @end direntry. See Installing Info Directory Files.

@display

Begin a kind of example. Like @example (indent text, do not fill), but do not select a new font. Pair with @end display. See @display: Examples Using the Text Font.

@displaymath

Format a block of math in “display” format. See @math and @displaymath: Formatting Mathematics.

@dmn{dimension}

Format a unit of measure, as in 12pt. Causes TeX to insert a thin space before dimension. No effect in Info. See @dmn{dimension}: Format a Dimension.

@docbook

Enter DocBook completely. Pair with @end docbook. See Raw Formatter Commands.

@documentdescription

Set the document description text, included in the HTML output. Pair with @end documentdescription. See @documentdescription: Summary Text.

@documentencoding enc

Declare the input encoding to be enc. See @documentencoding enc: Set Input Encoding.

@documentlanguage