GNU Coreutils

Short Table of Contents

Table of Contents

Next: , Up: (dir)   [Contents][Index]

GNU Coreutils

This manual documents version 8.24 of the GNU core utilities, including the standard programs for text and file manipulation.

Copyright © 1994-2015 Free Software Foundation, Inc.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, with no Front-Cover Texts, and with no Back-Cover Texts. A copy of the license is included in the section entitled “GNU Free Documentation License”.


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

1 Introduction

This manual is a work in progress: many sections make no attempt to explain basic concepts in a way suitable for novices. Thus, if you are interested, please get involved in improving this manual. The entire GNU community will benefit.

The GNU utilities documented here are mostly compatible with the POSIX standard. Please report bugs to bug-coreutils@gnu.org">bug-coreutils@gnu.org. Remember to include the version number, machine architecture, input files, and any other information needed to reproduce the bug: your input, what you expected, what you got, and why it is wrong. Diffs are welcome, but please include a description of the problem as well, since this is sometimes difficult to infer. See Bugs in Using and Porting GNU CC.

This manual was originally derived from the Unix man pages in the distributions, which were written by David MacKenzie and updated by Jim Meyering. What you are reading now is the authoritative documentation for these utilities; the man pages are no longer being maintained. The original fmt man page was written by Ross Paterson. François Pinard did the initial conversion to Texinfo format. Karl Berry did the indexing, some reorganization, and editing of the results. Brian Youmans of the Free Software Foundation office staff combined the manuals for textutils, fileutils, and sh-utils to produce the present omnibus manual. Richard Stallman contributed his usual invaluable insights to the overall process.


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

2 Common options

Certain options are available in all of these programs. Rather than writing identical descriptions for each of the programs, they are described here. (In fact, every GNU program accepts (or should accept) these options.)

Normally options and operands can appear in any order, and programs act as if all the options appear before any operands. For example, ‘sort -r passwd -t :’ acts like ‘sort -r -t : passwd’, since ‘:’ is an option-argument of -t. However, if the POSIXLY_CORRECT environment variable is set, options must appear before operands, unless otherwise specified for a particular command.

A few programs can usefully have trailing operands with leading ‘-’. With such a program, options must precede operands even if POSIXLY_CORRECT is not set, and this fact is noted in the program description. For example, the env command’s options must appear before its operands, since in some cases the operands specify a command that itself contains options.

Most programs that accept long options recognize unambiguous abbreviations of those options. For example, ‘rmdir --ignore-fail-on-non-empty’ can be invoked as ‘rmdir --ignore-fail’ or even ‘rmdir --i’. Ambiguous options, such as ‘ls --h’, are identified as such.

Some of these programs recognize the --help and --version options only when one of them is the sole command line argument. For these programs, abbreviations of the long options are not always recognized.

--help

Print a usage message listing all available options, then exit successfully.

--version

Print the version number, then exit successfully.

--

Delimit the option list. Later arguments, if any, are treated as operands even if they begin with ‘-’. For example, ‘sort -- -r’ reads from the file named -r.

A single ‘-’ operand is not really an option, though it looks like one. It stands for a file operand, and some tools treat it as standard input, or as standard output if that is clear from the context. For example, ‘sort -’ reads from standard input, and is equivalent to plain ‘sort’. Unless otherwise specified, a ‘-’ can appear as any operand that requires a file name.


Next: , Up: Common options   [Contents][Index]

2.1 Exit status

Nearly every command invocation yields an integral exit status that can be used to change how other commands work. For the vast majority of commands, an exit status of zero indicates success. Failure is indicated by a nonzero value—typically ‘1’, though it may differ on unusual platforms as POSIX requires only that it be nonzero.

However, some of the programs documented here do produce other exit status values and a few associate different meanings with the values ‘0’ and ‘1’. Here are some of the exceptions: chroot, env, expr, nice, nohup, numfmt, printenv, sort, stdbuf, test, timeout, tty.


Next: , Previous: , Up: Common options   [Contents][Index]

2.2 Backup options

Some GNU programs (at least cp, install, ln, and mv) optionally make backups of files before writing new versions. These options control the details of these backups. The options are also briefly mentioned in the descriptions of the particular programs.

-b
--backup[=method]

Make a backup of each file that would otherwise be overwritten or removed. Without this option, the original versions are destroyed. Use method to determine the type of backups to make. When this option is used but method is not specified, then the value of the VERSION_CONTROL environment variable is used. And if VERSION_CONTROL is not set, the default backup type is ‘existing’.

Note that the short form of this option, -b does not accept any argument. Using -b is equivalent to using --backup=existing.

This option corresponds to the Emacs variable ‘version-control’; the values for method are the same as those used in Emacs. This option also accepts more descriptive names. The valid methods are (unique abbreviations are accepted):

none
off

Never make backups.

numbered
t

Always make numbered backups.

existing
nil

Make numbered backups of files that already have them, simple backups of the others.

simple
never

Always make simple backups. Please note ‘never’ is not to be confused with ‘none’.

-S suffix
--suffix=suffix

Append suffix to each backup file made with -b. If this option is not specified, the value of the SIMPLE_BACKUP_SUFFIX environment variable is used. And if SIMPLE_BACKUP_SUFFIX is not set, the default is ‘~’, just as in Emacs.


Next: , Previous: , Up: Common options   [Contents][Index]

2.3 Block size

Some GNU programs (at least df, du, and ls) display sizes in “blocks”. You can adjust the block size and method of display to make sizes easier to read. The block size used for display is independent of any file system block size. Fractional block counts are rounded up to the nearest integer.

The default block size is chosen by examining the following environment variables in turn; the first one that is set determines the block size.

DF_BLOCK_SIZE

This specifies the default block size for the df command. Similarly, DU_BLOCK_SIZE specifies the default for du and LS_BLOCK_SIZE for ls.

BLOCK_SIZE

This specifies the default block size for all three commands, if the above command-specific environment variables are not set.

BLOCKSIZE

This specifies the default block size for all values that are normally printed as blocks, if neither BLOCK_SIZE nor the above command-specific environment variables are set. Unlike the other environment variables, BLOCKSIZE does not affect values that are normally printed as byte counts, e.g., the file sizes contained in ls -l output.

POSIXLY_CORRECT

If neither command_BLOCK_SIZE, nor BLOCK_SIZE, nor BLOCKSIZE is set, but this variable is set, the block size defaults to 512.

If none of the above environment variables are set, the block size currently defaults to 1024 bytes in most contexts, but this number may change in the future. For ls file sizes, the block size defaults to 1 byte.

A block size specification can be a positive integer specifying the number of bytes per block, or it can be human-readable or si to select a human-readable format. Integers may be followed by suffixes that are upward compatible with the http://www.bipm.org/en/publications/si-brochure/chapter3.html">SI prefixes for decimal multiples and with the ISO/IEC 80000-13 (formerly IEC 60027-2) prefixes for binary multiples.

With human-readable formats, output sizes are followed by a size letter such as ‘M’ for megabytes. BLOCK_SIZE=human-readable uses powers of 1024; ‘M’ stands for 1,048,576 bytes. BLOCK_SIZE=si is similar, but uses powers of 1000 and appends ‘B’; ‘MB’ stands for 1,000,000 bytes.

A block size specification preceded by ‘'’ causes output sizes to be displayed with thousands separators. The LC_NUMERIC locale specifies the thousands separator and grouping. For example, in an American English locale, ‘--block-size="'1kB"’ would cause a size of 1234000 bytes to be displayed as ‘1,234’. In the default C locale, there is no thousands separator so a leading ‘'’ has no effect.

An integer block size can be followed by a suffix to specify a multiple of that size. A bare size letter, or one followed by ‘iB’, specifies a multiple using powers of 1024. A size letter followed by ‘B’ specifies powers of 1000 instead. For example, ‘1M’ and ‘1MiB’ are equivalent to ‘1048576’, whereas ‘1MB’ is equivalent to ‘1000000’.

A plain suffix without a preceding integer acts as if ‘1’ were prepended, except that it causes a size indication to be appended to the output. For example, ‘--block-size="kB"’ displays 3000 as ‘3kB’.

The following suffixes are defined. Large sizes like 1Y may be rejected by your computer due to limitations of its arithmetic.

kB

kilobyte: 10^3 = 1000.

k
K
KiB

kibibyte: 2^{10} = 1024. ‘K’ is special: the SI prefix is ‘k’ and the ISO/IEC 80000-13 prefix is ‘Ki’, but tradition and POSIX use ‘k’ to mean ‘KiB’.

MB

megabyte: 10^6 = 1,000,000.

M
MiB

mebibyte: 2^{20} = 1,048,576.

GB

gigabyte: 10^9 = 1,000,000,000.

G
GiB

gibibyte: 2^{30} = 1,073,741,824.

TB

terabyte: 10^{12} = 1,000,000,000,000.

T
TiB

tebibyte: 2^{40} = 1,099,511,627,776.

PB

petabyte: 10^{15} = 1,000,000,000,000,000.

P
PiB

pebibyte: 2^{50} = 1,125,899,906,842,624.

EB

exabyte: 10^{18} = 1,000,000,000,000,000,000.

E
EiB

exbibyte: 2^{60} = 1,152,921,504,606,846,976.

ZB

zettabyte: 10^{21} = 1,000,000,000,000,000,000,000

Z
ZiB

2^{70} = 1,180,591,620,717,411,303,424.

YB

yottabyte: 10^{24} = 1,000,000,000,000,000,000,000,000.

Y
YiB

2^{80} = 1,208,925,819,614,629,174,706,176.

Block size defaults can be overridden by an explicit --block-size=size option. The -k option is equivalent to --block-size=1K, which is the default unless the POSIXLY_CORRECT environment variable is set. The -h or --human-readable option is equivalent to --block-size=human-readable. The --si option is equivalent to --block-size=si. Note for ls the -k option does not control the display of the apparent file sizes, whereas the --block-size option does.


Next: , Previous: , Up: Common options   [Contents][Index]

2.4 Floating point numbers

Commands that accept or produce floating point numbers employ the floating point representation of the underlying system, and suffer from rounding error, overflow, and similar floating-point issues. Almost all modern systems use IEEE-754 floating point, and it is typically portable to assume IEEE-754 behavior these days. IEEE-754 has positive and negative infinity, distinguishes positive from negative zero, and uses special values called NaNs to represent invalid computations such as dividing zero by itself. For more information, please see David Goldberg’s paper What Every Computer Scientist Should Know About Floating-Point Arithmetic.

Commands that accept floating point numbers as options, operands or input use the standard C functions strtod and strtold to convert from text to floating point numbers. These floating point numbers therefore can use scientific notation like 1.0e-34 and -10e100. Commands that parse floating point also understand case-insensitive inf, infinity, and NaN, although whether such values are useful depends on the command in question. Modern C implementations also accept hexadecimal floating point numbers such as -0x.ep-3, which stands for -14/16 times 2^-3, which equals -0.109375. The LC_NUMERIC locale determines the decimal-point character. See Parsing of Floats in The GNU C Library Reference Manual.


Next: , Previous: , Up: Common options   [Contents][Index]

2.5 Signal specifications

A signal may be a signal name like ‘HUP’, or a signal number like ‘1’, or an exit status of a process terminated by the signal. A signal name can be given in canonical form or prefixed by ‘SIG’. The case of the letters is ignored. The following signal names and numbers are supported on all POSIX compliant systems:

HUP

1. Hangup.

INT

2. Terminal interrupt.

QUIT

3. Terminal quit.

ABRT

6. Process abort.

KILL

9. Kill (cannot be caught or ignored).

ALRM

14. Alarm Clock.

TERM

15. Termination.

Other supported signal names have system-dependent corresponding numbers. All systems conforming to POSIX 1003.1-2001 also support the following signals:

BUS

Access to an undefined portion of a memory object.

CHLD

Child process terminated, stopped, or continued.

CONT

Continue executing, if stopped.

FPE

Erroneous arithmetic operation.

ILL

Illegal Instruction.

PIPE

Write on a pipe with no one to read it.

SEGV

Invalid memory reference.

STOP

Stop executing (cannot be caught or ignored).

TSTP

Terminal stop.

TTIN

Background process attempting read.

TTOU

Background process attempting write.

URG

High bandwidth data is available at a socket.

USR1

User-defined signal 1.

USR2

User-defined signal 2.

POSIX 1003.1-2001 systems that support the XSI extension also support the following signals:

POLL

Pollable event.

PROF

Profiling timer expired.

SYS

Bad system call.

TRAP

Trace/breakpoint trap.

VTALRM

Virtual timer expired.

XCPU

CPU time limit exceeded.

XFSZ

File size limit exceeded.

POSIX 1003.1-2001 systems that support the XRT extension also support at least eight real-time signals called ‘RTMIN’, ‘RTMIN+1’, …, ‘RTMAX-1’, ‘RTMAX’.


Next: , Previous: , Up: Common options   [Contents][Index]

2.6 chown, chgrp, chroot, id: Disambiguating user names and IDs

Since the user and group arguments to these commands may be specified as names or numeric IDs, there is an apparent ambiguity. What if a user or group name is a string of digits? 1 Should the command interpret it as a user name or as an ID? POSIX requires that these commands first attempt to resolve the specified string as a name, and only once that fails, then try to interpret it as an ID. This is troublesome when you want to specify a numeric ID, say 42, and it must work even in a pathological situation where ‘42’ is a user name that maps to some other user ID, say 1000. Simply invoking chown 42 F, will set Fs owner ID to 1000—not what you intended.

GNU chown, chgrp, chroot, and id provide a way to work around this, that at the same time may result in a significant performance improvement by eliminating a database look-up. Simply precede each numeric user ID and/or group ID with a ‘+’, in order to force its interpretation as an integer:

chown +42 F
chgrp +$numeric_group_id another-file
chown +0:+0 /

The name look-up process is skipped for each ‘+’-prefixed string, because a string containing ‘+’ is never a valid user or group name. This syntax is accepted on most common Unix systems, but not on Solaris 10.


Next: , Previous: , Up: Common options   [Contents][Index]

2.7 Sources of random data

The shuf, shred, and sort commands sometimes need random data to do their work. For example, ‘sort -R’ must choose a hash function at random, and it needs random data to make this selection.

By default these commands use an internal pseudo-random generator initialized by a small amount of entropy, but can be directed to use an external source with the --random-source=file option. An error is reported if file does not contain enough bytes.

For example, the device file /dev/urandom could be used as the source of random data. Typically, this device gathers environmental noise from device drivers and other sources into an entropy pool, and uses the pool to generate random bits. If the pool is short of data, the device reuses the internal pool to produce more bits, using a cryptographically secure pseudo-random number generator. But be aware that this device is not designed for bulk random data generation and is relatively slow.

/dev/urandom suffices for most practical uses, but applications requiring high-value or long-term protection of private data may require an alternate data source like /dev/random or /dev/arandom. The set of available sources depends on your operating system.

To reproduce the results of an earlier invocation of a command, you can save some random data into a file and then use that file as the random source in earlier and later invocations of the command. Rather than depending on a file, one can generate a reproducible arbitrary amount of pseudo-random data given a seed value, using for example:

get_seeded_random()
{
  seed="$1"
  openssl enc -aes-256-ctr -pass pass:"$seed" -nosalt \
    </dev/zero 2>/dev/null
}

shuf -i1-100 --random-source=<(get_seeded_random 42)

Next: , Previous: , Up: Common options   [Contents][Index]

2.8 Target directory

The cp, install, ln, and mv commands normally treat the last operand specially when it is a directory or a symbolic link to a directory. For example, ‘cp source dest’ is equivalent to ‘cp source dest/source’ if dest is a directory. Sometimes this behavior is not exactly what is wanted, so these commands support the following options to allow more fine-grained control:

-T
--no-target-directory

Do not treat the last operand specially when it is a directory or a symbolic link to a directory. This can help avoid race conditions in programs that operate in a shared area. For example, when the command ‘mv /tmp/source /tmp/dest’ succeeds, there is no guarantee that /tmp/source was renamed to /tmp/dest: it could have been renamed to /tmp/dest/source instead, if some other process created /tmp/dest as a directory. However, if mv -T /tmp/source /tmp/dest succeeds, there is no question that /tmp/source was renamed to /tmp/dest.

In the opposite situation, where you want the last operand to be treated as a directory and want a diagnostic otherwise, you can use the --target-directory (-t) option.

-t directory
--target-directory=directory

Use directory as the directory component of each destination file name.

The interface for most programs is that after processing options and a finite (possibly zero) number of fixed-position arguments, the remaining argument list is either expected to be empty, or is a list of items (usually files) that will all be handled identically. The xargs program is designed to work well with this convention.

The commands in the mv-family are unusual in that they take a variable number of arguments with a special case at the end (namely, the target directory). This makes it nontrivial to perform some operations, e.g., “move all files from here to ../d/”, because mv * ../d/ might exhaust the argument space, and ls | xargs ... doesn’t have a clean way to specify an extra final argument for each invocation of the subject command. (It can be done by going through a shell command, but that requires more human labor and brain power than it should.)

The --target-directory (-t) option allows the cp, install, ln, and mv programs to be used conveniently with xargs. For example, you can move the files from the current directory to a sibling directory, d like this:

ls | xargs mv -t ../d --

However, this doesn’t move files whose names begin with ‘.’. If you use the GNU find program, you can move those files too, with this command:

find . -mindepth 1 -maxdepth 1 \
  | xargs mv -t ../d

But both of the above approaches fail if there are no files in the current directory, or if any file has a name containing a blank or some other special characters. The following example removes those limitations and requires both GNU find and GNU xargs:

find . -mindepth 1 -maxdepth 1 -print0 \
  | xargs --null --no-run-if-empty \
      mv -t ../d

The --target-directory (-t) and --no-target-directory (-T) options cannot be combined.


Next: , Previous: , Up: Common options   [Contents][Index]

2.9 Trailing slashes

Some GNU programs (at least cp and mv) allow you to remove any trailing slashes from each source argument before operating on it. The --strip-trailing-slashes option enables this behavior.

This is useful when a source argument may have a trailing slash and specify a symbolic link to a directory. This scenario is in fact rather common because some shells can automatically append a trailing slash when performing file name completion on such symbolic links. Without this option, mv, for example, (via the system’s rename function) must interpret a trailing slash as a request to dereference the symbolic link and so must rename the indirectly referenced directory and not the symbolic link. Although it may seem surprising that such behavior be the default, it is required by POSIX and is consistent with other parts of that standard.


Next: , Previous: , Up: Common options   [Contents][Index]

2.10 Traversing symlinks

The following options modify how chown and chgrp traverse a hierarchy when the --recursive (-R) option is also specified. If more than one of the following options is specified, only the final one takes effect. These options specify whether processing a symbolic link to a directory entails operating on just the symbolic link or on all files in the hierarchy rooted at that directory.

These options are independent of --dereference and --no-dereference (-h), which control whether to modify a symlink or its referent.

-H

If --recursive (-R) is specified and a command line argument is a symbolic link to a directory, traverse it.

-L

In a recursive traversal, traverse every symbolic link to a directory that is encountered.

-P

Do not traverse any symbolic links. This is the default if none of -H, -L, or -P is specified.


Next: , Previous: , Up: Common options   [Contents][Index]

2.11 Treating / specially

Certain commands can operate destructively on entire hierarchies. For example, if a user with appropriate privileges mistakenly runs ‘rm -rf / tmp/junk’, that may remove all files on the entire system. Since there are so few legitimate uses for such a command, GNU rm normally declines to operate on any directory that resolves to /. If you really want to try to remove all the files on your system, you can use the --no-preserve-root option, but the default behavior, specified by the --preserve-root option, is safer for most purposes.

The commands chgrp, chmod and chown can also operate destructively on entire hierarchies, so they too support these options. Although, unlike rm, they don’t actually unlink files, these commands are arguably more dangerous when operating recursively on /, since they often work much more quickly, and hence damage more files before an alert user can interrupt them. Tradition and POSIX require these commands to operate recursively on /, so they default to --no-preserve-root, but using the --preserve-root option makes them safer for most purposes. For convenience you can specify --preserve-root in an alias or in a shell function.

Note that the --preserve-root option also ensures that chgrp and chown do not modify / even when dereferencing a symlink pointing to /.


Next: , Previous: , Up: Common options   [Contents][Index]

2.12 Special built-in utilities

Some programs like nice can invoke other programs; for example, the command ‘nice cat file’ invokes the program cat by executing the command ‘cat file’. However, special built-in utilities like exit cannot be invoked this way. For example, the command ‘nice exit’ does not have a well-defined behavior: it may generate an error message instead of exiting.

Here is a list of the special built-in utilities that are standardized by POSIX 1003.1-2004.

. : break continue eval exec exit export readonly return set shift times trap unset

For example, because ‘.’, ‘:’, and ‘exec’ are special, the commands ‘nice . foo.sh’, ‘nice :’, and ‘nice exec pwd’ do not work as you might expect.

Many shells extend this list. For example, Bash has several extra special built-in utilities like history, and suspend, and with Bash the command ‘nice suspend’ generates an error message instead of suspending.


Next: , Previous: , Up: Common options   [Contents][Index]

2.13 Standards conformance

In a few cases, the GNU utilities’ default behavior is incompatible with the POSIX standard. To suppress these incompatibilities, define the POSIXLY_CORRECT environment variable. Unless you are checking for POSIX conformance, you probably do not need to define POSIXLY_CORRECT.

Newer versions of POSIX are occasionally incompatible with older versions. For example, older versions of POSIX required the command ‘sort +1’ to sort based on the second and succeeding fields in each input line, but starting with POSIX 1003.1-2001 the same command is required to sort the file named +1, and you must instead use the command ‘sort -k 2’ to get the field-based sort.

The GNU utilities normally conform to the version of POSIX that is standard for your system. To cause them to conform to a different version of POSIX, define the _POSIX2_VERSION environment variable to a value of the form yyyymm specifying the year and month the standard was adopted. Three values are currently supported for _POSIX2_VERSION: ‘199209’ stands for POSIX 1003.2-1992, ‘200112’ stands for POSIX 1003.1-2001, and ‘200809’ stands for POSIX 1003.1-2008. For example, if you have a newer system but are running software that assumes an older version of POSIX and uses ‘sort +1’ or ‘tail +10’, you can work around any compatibility problems by setting ‘_POSIX2_VERSION=199209’ in your environment.


Previous: , Up: Common options   [Contents][Index]

2.14 coreutils: Multi-call program

The coreutils command invokes an individual utility, either implicitly selected by the last component of the name used to invoke coreutils, or explicitly with the --coreutils-prog option. Synopsis:

coreutils --coreutils-prog=PROGRAM

The coreutils command is not installed by default, so portable scripts should not rely on its existence.


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

3 Output of entire files

These commands read and write entire files, possibly transforming them in some way.


Next: , Up: Output of entire files   [Contents][Index]

3.1 cat: Concatenate and write files

cat copies each file (‘-’ means standard input), or standard input if none are given, to standard output. Synopsis:

cat [option] [file]…

The program accepts the following options. Also see Common options.

-A
--show-all

Equivalent to -vET.

-b
--number-nonblank

Number all nonempty output lines, starting with 1.

-e

Equivalent to -vE.

-E
--show-ends

Display a ‘$’ after the end of each line.

-n
--number

Number all output lines, starting with 1. This option is ignored if -b is in effect.

-s
--squeeze-blank

Suppress repeated adjacent empty lines; output just one empty line instead of several.

-t

Equivalent to -vT.

-T
--show-tabs

Display TAB characters as ‘^I’.

-u

Ignored; for POSIX compatibility.

-v
--show-nonprinting

Display control characters except for LFD and TAB using ‘^’ notation and precede characters that have the high bit set with ‘M-’.

On systems like MS-DOS that distinguish between text and binary files, cat normally reads and writes in binary mode. However, cat reads in text mode if one of the options -bensAE is used or if cat is reading from standard input and standard input is a terminal. Similarly, cat writes in text mode if one of the options -bensAE is used or if standard output is a terminal.

An exit status of zero indicates success, and a nonzero value indicates failure.

Examples:

# Output f's contents, then standard input, then g's contents.
cat f - g

# Copy standard input to standard output.
cat

Next: , Previous: , Up: Output of entire files   [Contents][Index]

3.2 tac: Concatenate and write files in reverse

tac copies each file (‘-’ means standard input), or standard input if none are given, to standard output, reversing the records (lines by default) in each separately. Synopsis:

tac [option]… [file]…

Records are separated by instances of a string (newline by default). By default, this separator string is attached to the end of the record that it follows in the file.

The program accepts the following options. Also see Common options.

-b
--before

The separator is attached to the beginning of the record that it precedes in the file.

-r
--regex

Treat the separator string as a regular expression.

-s separator
--separator=separator

Use separator as the record separator, instead of newline.

On systems like MS-DOS that distinguish between text and binary files, tac reads and writes in binary mode.

An exit status of zero indicates success, and a nonzero value indicates failure.

Example:

# Reverse a file character by character.
tac -r -s 'x\|[^x]'

Next: , Previous: , Up: Output of entire files   [Contents][Index]

3.3 nl: Number lines and write files

nl writes each file (‘-’ means standard input), or standard input if none are given, to standard output, with line numbers added to some or all of the lines. Synopsis:

nl [option]… [file]…

nl decomposes its input into (logical) pages; by default, the line number is reset to 1 at the top of each logical page. nl treats all of the input files as a single document; it does not reset line numbers or logical pages between files.

A logical page consists of three sections: header, body, and footer. Any of the sections can be empty. Each can be numbered in a different style from the others.

The beginnings of the sections of logical pages are indicated in the input file by a line containing exactly one of these delimiter strings:

\:\:\:

start of header;

\:\:

start of body;

\:

start of footer.

The two characters from which these strings are made can be changed from ‘\’ and ‘:’ via options (see below), but the pattern and length of each string cannot be changed.

A section delimiter is replaced by an empty line on output. Any text that comes before the first section delimiter string in the input file is considered to be part of a body section, so nl treats a file that contains no section delimiters as a single body section.

The program accepts the following options. Also see Common options.

-b style
--body-numbering=style

Select the numbering style for lines in the body section of each logical page. When a line is not numbered, the current line number is not incremented, but the line number separator character is still prepended to the line. The styles are:

a

number all lines,

t

number only nonempty lines (default for body),

n

do not number lines (default for header and footer),

pbre

number only lines that contain a match for the basic regular expression bre. See Regular Expressions in The GNU Grep Manual.

-d cd
--section-delimiter=cd

Set the section delimiter characters to cd; default is ‘\:’. If only c is given, the second remains ‘:’. (Remember to protect ‘\’ or other metacharacters from shell expansion with quotes or extra backslashes.)

-f style
--footer-numbering=style

Analogous to --body-numbering.

-h style
--header-numbering=style

Analogous to --body-numbering.

-i number
--line-increment=number

Increment line numbers by number (default 1).

-l number
--join-blank-lines=number

Consider number (default 1) consecutive empty lines to be one logical line for numbering, and only number the last one. Where fewer than number consecutive empty lines occur, do not number them. An empty line is one that contains no characters, not even spaces or tabs.

-n format
--number-format=format

Select the line numbering format (default is rn):

ln

left justified, no leading zeros;

rn

right justified, no leading zeros;

rz

right justified, leading zeros.

-p
--no-renumber

Do not reset the line number at the start of a logical page.

-s string
--number-separator=string

Separate the line number from the text line in the output with string (default is the TAB character).

-v number
--starting-line-number=number

Set the initial line number on each logical page to number (default 1).

-w number
--number-width=number

Use number characters for line numbers (default 6).

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Output of entire files   [Contents][Index]

3.4 od: Write files in octal or other formats

od writes an unambiguous representation of each file (‘-’ means standard input), or standard input if none are given. Synopses:

od [option]… [file]…
od [-abcdfilosx]… [file] [[+]offset[.][b]]
od [option]… --traditional [file] [[+]offset[.][b] [[+]label[.][b]]]

Each line of output consists of the offset in the input, followed by groups of data from the file. By default, od prints the offset in octal, and each group of file data is a C short int’s worth of input printed as a single octal number.

If offset is given, it specifies how many input bytes to skip before formatting and writing. By default, it is interpreted as an octal number, but the optional trailing decimal point causes it to be interpreted as decimal. If no decimal is specified and the offset begins with ‘0x’ or ‘0X’ it is interpreted as a hexadecimal number. If there is a trailing ‘b’, the number of bytes skipped will be offset multiplied by 512.

If a command is of both the first and second forms, the second form is assumed if the last operand begins with ‘+’ or (if there are two operands) a digit. For example, in ‘od foo 10’ and ‘od +10’ the ‘10’ is an offset, whereas in ‘od 10’ the ‘10’ is a file name.

The program accepts the following options. Also see Common options.

-A radix
--address-radix=radix

Select the base in which file offsets are printed. radix can be one of the following:

d

decimal;

o

octal;

x

hexadecimal;

n

none (do not print offsets).

The default is octal.

--endian=order

Reorder input bytes, to handle inputs with differing byte orders, or to provide consistent output independent of the endian convention of the current system. Swapping is performed according to the specified --type size and endian order, which can be ‘little’ or ‘big’.

-j bytes
--skip-bytes=bytes

Skip bytes input bytes before formatting and writing. If bytes begins with ‘0x’ or ‘0X’, it is interpreted in hexadecimal; otherwise, if it begins with ‘0’, in octal; otherwise, in decimal. bytes may be, or may be an integer optionally followed by, one of the following multiplicative suffixes:

b’  =>            512 ("blocks")
‘KB’ =>           1000 (KiloBytes)
‘K’  =>           1024 (KibiBytes)
‘MB’ =>      1000*1000 (MegaBytes)
‘M’  =>      1024*1024 (MebiBytes)
‘GB’ => 1000*1000*1000 (GigaBytes)
‘G’  => 1024*1024*1024 (GibiBytes)

and so on for ‘T’, ‘P’, ‘E’, ‘Z’, and ‘Y’.

-N bytes
--read-bytes=bytes

Output at most bytes bytes of the input. Prefixes and suffixes on bytes are interpreted as for the -j option.

-S bytes
--strings[=bytes]

Instead of the normal output, output only string constants: at least bytes consecutive ASCII graphic characters, followed by a zero byte (ASCII NUL). Prefixes and suffixes on bytes are interpreted as for the -j option.

If bytes is omitted with --strings, the default is 3.

-t type
--format=type

Select the format in which to output the file data. type is a string of one or more of the below type indicator characters. If you include more than one type indicator character in a single type string, or use this option more than once, od writes one copy of each output line using each of the data types that you specified, in the order that you specified.

Adding a trailing “z” to any type specification appends a display of the single byte character representation of the printable characters to the output line generated by the type specification.

a

named character, ignoring high-order bit

c

printable single byte character, C backslash escape or a 3 digit octal sequence

d

signed decimal

f

floating point (see Floating point)

o

octal

u

unsigned decimal

x

hexadecimal

The type a outputs things like ‘sp’ for space, ‘nl’ for newline, and ‘nul’ for a zero byte. Only the least significant seven bits of each byte is used; the high-order bit is ignored. Type c outputs ‘ ’, ‘\n’, and \0, respectively.

Except for types ‘a’ and ‘c’, you can specify the number of bytes to use in interpreting each number in the given data type by following the type indicator character with a decimal integer. Alternately, you can specify the size of one of the C compiler’s built-in data types by following the type indicator character with one of the following characters. For integers (‘d’, ‘o’, ‘u’, ‘x’):

C

char

S

short

I

int

L

long

For floating point (f):

F

float

D

double

L

long double

-v
--output-duplicates

Output consecutive lines that are identical. By default, when two or more consecutive output lines would be identical, od outputs only the first line, and puts just an asterisk on the following line to indicate the elision.

-w[n]
--width[=n]

Dump n input bytes per output line. This must be a multiple of the least common multiple of the sizes associated with the specified output types.

If this option is not given at all, the default is 16. If n is omitted, the default is 32.

The next several options are shorthands for format specifications. GNU od accepts any combination of shorthands and format specification options. These options accumulate.

-a

Output as named characters. Equivalent to ‘-t a’.

-b

Output as octal bytes. Equivalent to ‘-t o1’.

-c

Output as printable single byte characters, C backslash escapes or 3 digit octal sequences. Equivalent to ‘-t c’.

-d

Output as unsigned decimal two-byte units. Equivalent to ‘-t u2’.

-f

Output as floats. Equivalent to ‘-t fF’.

-i

Output as decimal ints. Equivalent to ‘-t dI’.

-l

Output as decimal long ints. Equivalent to ‘-t dL’.

-o

Output as octal two-byte units. Equivalent to -t o2.

-s

Output as decimal two-byte units. Equivalent to -t d2.

-x

Output as hexadecimal two-byte units. Equivalent to ‘-t x2’.

--traditional

Recognize the non-option label argument that traditional od accepted. The following syntax:

od --traditional [file] [[+]offset[.][b] [[+]label[.][b]]]

can be used to specify at most one file and optional arguments specifying an offset and a pseudo-start address, label. The label argument is interpreted just like offset, but it specifies an initial pseudo-address. The pseudo-addresses are displayed in parentheses following any normal address.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Output of entire files   [Contents][Index]

3.5 base64: Transform data into printable data

base64 transforms data read from a file, or standard input, into (or from) base64 encoded form. The base64 encoded form uses printable ASCII characters to represent binary data. Synopses:

base64 [option]… [file]
base64 --decode [option]… [file]

The base64 encoding expands data to roughly 133% of the original. The format conforms to ftp://ftp.rfc-editor.org/in-notes/rfc4648.txt">RFC 4648.

The program accepts the following options. Also see Common options.

-w cols
--wrap=cols

During encoding, wrap lines after cols characters. This must be a positive number.

The default is to wrap after 76 characters. Use the value 0 to disable line wrapping altogether.

-d
--decode

Change the mode of operation, from the default of encoding data, to decoding data. Input is expected to be base64 encoded data, and the output will be the original data.

-i
--ignore-garbage

When decoding, newlines are always accepted. During decoding, ignore unrecognized bytes, to permit distorted data to be decoded.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

4 Formatting file contents

These commands reformat the contents of files.


Next: , Up: Formatting file contents   [Contents][Index]

4.1 fmt: Reformat paragraph text

fmt fills and joins lines to produce output lines of (at most) a given number of characters (75 by default). Synopsis:

fmt [option]… [file]…

fmt reads from the specified file arguments (or standard input if none are given), and writes to standard output.

By default, blank lines, spaces between words, and indentation are preserved in the output; successive input lines with different indentation are not joined; tabs are expanded on input and introduced on output.

fmt prefers breaking lines at the end of a sentence, and tries to avoid line breaks after the first word of a sentence or before the last word of a sentence. A sentence break is defined as either the end of a paragraph or a word ending in any of ‘.?!’, followed by two spaces or end of line, ignoring any intervening parentheses or quotes. Like TeX, fmt reads entire “paragraphs” before choosing line breaks; the algorithm is a variant of that given by Donald E. Knuth and Michael F. Plass in “Breaking Paragraphs Into Lines”, Software—Practice & Experience 11, 11 (November 1981), 1119–1184.

The program accepts the following options. Also see Common options.

-c
--crown-margin

Crown margin mode: preserve the indentation of the first two lines within a paragraph, and align the left margin of each subsequent line with that of the second line.

-t
--tagged-paragraph

Tagged paragraph mode: like crown margin mode, except that if indentation of the first line of a paragraph is the same as the indentation of the second, the first line is treated as a one-line paragraph.

-s
--split-only

Split lines only. Do not join short lines to form longer ones. This prevents sample lines of code, and other such “formatted” text from being unduly combined.

-u
--uniform-spacing

Uniform spacing. Reduce spacing between words to one space, and spacing between sentences to two spaces.

-width
-w width
--width=width

Fill output lines up to width characters (default 75 or goal plus 10, if goal is provided).

-g goal
--goal=goal

fmt initially tries to make lines goal characters wide. By default, this is 7% shorter than width.

-p prefix
--prefix=prefix

Only lines beginning with prefix (possibly preceded by whitespace) are subject to formatting. The prefix and any preceding whitespace are stripped for the formatting and then re-attached to each formatted output line. One use is to format certain kinds of program comments, while leaving the code unchanged.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Formatting file contents   [Contents][Index]

4.2 pr: Paginate or columnate files for printing

pr writes each file (‘-’ means standard input), or standard input if none are given, to standard output, paginating and optionally outputting in multicolumn format; optionally merges all files, printing all in parallel, one per column. Synopsis:

pr [option]… [file]…

By default, a 5-line header is printed at each page: two blank lines; a line with the date, the file name, and the page count; and two more blank lines. A footer of five blank lines is also printed. The default page_length is 66 lines. The default number of text lines is therefore 56. The text line of the header takes the form ‘date string page’, with spaces inserted around string so that the line takes up the full page_width. Here, date is the date (see the -D or --date-format option for details), string is the centered header string, and page identifies the page number. The LC_MESSAGES locale category affects the spelling of page; in the default C locale, it is ‘Page number’ where number is the decimal page number.

Form feeds in the input cause page breaks in the output. Multiple form feeds produce empty pages.

Columns are of equal width, separated by an optional string (default is ‘space’). For multicolumn output, lines will always be truncated to page_width (default 72), unless you use the -J option. For single column output no line truncation occurs by default. Use -W option to truncate lines in that case.

The program accepts the following options. Also see Common options.

+first_page[:last_page]
--pages=first_page[:last_page]

Begin printing with page first_page and stop with last_page. Missing ‘:last_page’ implies end of file. While estimating the number of skipped pages each form feed in the input file results in a new page. Page counting with and without ‘+first_page’ is identical. By default, counting starts with the first page of input file (not first page printed). Line numbering may be altered by -N option.

-column
--columns=column

With each single file, produce column columns of output (default is 1) and print columns down, unless -a is used. The column width is automatically decreased as column increases; unless you use the -W/-w option to increase page_width as well. This option might well cause some lines to be truncated. The number of lines in the columns on each page are balanced. The options -e and -i are on for multiple text-column output. Together with -J option column alignment and line truncation is turned off. Lines of full length are joined in a free field format and -S option may set field separators. -column may not be used with -m option.

-a
--across

With each single file, print columns across rather than down. The -column option must be given with column greater than one. If a line is too long to fit in a column, it is truncated.

-c
--show-control-chars

Print control characters using hat notation (e.g., ‘^G’); print other nonprinting characters in octal backslash notation. By default, nonprinting characters are not changed.

-d
--double-space

Double space the output.

-D format
--date-format=format

Format header dates using format, using the same conventions as for the command ‘date +format’. See date invocation. Except for directives, which start with ‘%’, characters in format are printed unchanged. You can use this option to specify an arbitrary string in place of the header date, e.g., --date-format="Monday morning".

The default date format is ‘%Y-%m-%d %H:%M’ (for example, ‘2001-12-04 23:59’); but if the POSIXLY_CORRECT environment variable is set and the LC_TIME locale category specifies the POSIX locale, the default is ‘%b %e %H:%M %Y’ (for example, ‘Dec  4 23:59 2001’.

Time stamps are listed according to the time zone rules specified by the TZ environment variable, or by the system default rules if TZ is not set. See Specifying the Time Zone with TZ in The GNU C Library Reference Manual.

-e[in-tabchar[in-tabwidth]]
--expand-tabs[=in-tabchar[in-tabwidth]]

Expand tabs to spaces on input. Optional argument in-tabchar is the input tab character (default is the TAB character). Second optional argument in-tabwidth is the input tab character’s width (default is 8).

-f
-F
--form-feed

Use a form feed instead of newlines to separate output pages. This does not alter the default page length of 66 lines.

-h header
--header=header

Replace the file name in the header with the centered string header. When using the shell, header should be quoted and should be separated from -h by a space.

-i[out-tabchar[out-tabwidth]]
--output-tabs[=out-tabchar[out-tabwidth]]

Replace spaces with tabs on output. Optional argument out-tabchar is the output tab character (default is the TAB character). Second optional argument out-tabwidth is the output tab character’s width (default is 8).

-J
--join-lines

Merge lines of full length. Used together with the column options -column, -a -column or -m. Turns off -W/-w line truncation; no column alignment used; may be used with --sep-string[=string]. -J has been introduced (together with -W and --sep-string) to disentangle the old (POSIX-compliant) options -w and -s along with the three column options.

-l page_length
--length=page_length

Set the page length to page_length (default 66) lines, including the lines of the header [and the footer]. If page_length is less than or equal to 10, the header and footer are omitted, as if the -t option had been given.

-m
--merge

Merge and print all files in parallel, one in each column. If a line is too long to fit in a column, it is truncated, unless the -J option is used. --sep-string[=string] may be used. Empty pages in some files (form feeds set) produce empty columns, still marked by string. The result is a continuous line numbering and column marking throughout the whole merged file. Completely empty merged pages show no separators or line numbers. The default header becomes ‘date page’ with spaces inserted in the middle; this may be used with the -h or --header option to fill up the middle blank part.

-n[number-separator[digits]]
--number-lines[=number-separator[digits]]

Provide digits digit line numbering (default for digits is 5). With multicolumn output the number occupies the first digits column positions of each text column or only each line of -m output. With single column output the number precedes each line just as -m does. Default counting of the line numbers starts with the first line of the input file (not the first line printed, compare the --page option and -N option). Optional argument number-separator is the character appended to the line number to separate it from the text followed. The default separator is the TAB character. In a strict sense a TAB is always printed with single column output only. The TAB width varies with the TAB position, e.g., with the left margin specified by -o option. With multicolumn output priority is given to ‘equal width of output columns’ (a POSIX specification). The TAB width is fixed to the value of the first column and does not change with different values of left margin. That means a fixed number of spaces is always printed in the place of the number-separator TAB. The tabification depends upon the output position.

-N line_number
--first-line-number=line_number

Start line counting with the number line_number at first line of first page printed (in most cases not the first line of the input file).

-o margin
--indent=margin

Indent each line with a margin margin spaces wide (default is zero). The total page width is the size of the margin plus the page_width set with the -W/-w option. A limited overflow may occur with numbered single column output (compare -n option).

-r
--no-file-warnings

Do not print a warning message when an argument file cannot be opened. (The exit status will still be nonzero, however.)

-s[char]
--separator[=char]

Separate columns by a single character char. The default for char is the TAB character without -w and ‘no character’ with -w. Without -s the default separator ‘space’ is set. -s[char] turns off line truncation of all three column options (-COLUMN|-a -COLUMN|-m) unless -w is set. This is a POSIX-compliant formulation.

-S[string]
--sep-string[=string]

Use string to separate output columns. The -S option doesn’t affect the -W/-w option, unlike the -s option which does. It does not affect line truncation or column alignment. Without -S, and with -J, pr uses the default output separator, TAB. Without -S or -J, pr uses a ‘space’ (same as -S" "). If no ‘string’ argument is specified, ‘""’ is assumed.

-t
--omit-header

Do not print the usual header [and footer] on each page, and do not fill out the bottom of pages (with blank lines or a form feed). No page structure is produced, but form feeds set in the input files are retained. The predefined pagination is not changed. -t or -T may be useful together with other options; e.g.: -t -e4, expand TAB characters in the input file to 4 spaces but don’t make any other changes. Use of -t overrides -h.

-T
--omit-pagination

Do not print header [and footer]. In addition eliminate all form feeds set in the input files.

-v
--show-nonprinting

Print nonprinting characters in octal backslash notation.

-w page_width
--width=page_width

Set page width to page_width characters for multiple text-column output only (default for page_width is 72). The specified page_width is rounded down so that columns have equal width. -s[CHAR] turns off the default page width and any line truncation and column alignment. Lines of full length are merged, regardless of the column options set. No page_width setting is possible with single column output. A POSIX-compliant formulation.

-W page_width
--page_width=page_width

Set the page width to page_width characters, honored with and without a column option. With a column option, the specified page_width is rounded down so that columns have equal width. Text lines are truncated, unless -J is used. Together with one of the three column options (-column, -a -column or -m) column alignment is always used. The separator options -S or -s don’t disable the -W option. Default is 72 characters. Without -W page_width and without any of the column options NO line truncation is used (defined to keep downward compatibility and to meet most frequent tasks). That’s equivalent to -W 72 -J. The header line is never truncated.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Formatting file contents   [Contents][Index]

4.3 fold: Wrap input lines to fit in specified width

fold writes each file (- means standard input), or standard input if none are given, to standard output, breaking long lines. Synopsis:

fold [option]… [file]…

By default, fold breaks lines wider than 80 columns. The output is split into as many lines as necessary.

fold counts screen columns by default; thus, a tab may count more than one column, backspace decreases the column count, and carriage return sets the column to zero.

The program accepts the following options. Also see Common options.

-b
--bytes

Count bytes rather than columns, so that tabs, backspaces, and carriage returns are each counted as taking up one column, just like other characters.

-s
--spaces

Break at word boundaries: the line is broken after the last blank before the maximum line length. If the line contains no such blanks, the line is broken at the maximum line length as usual.

-w width
--width=width

Use a maximum line length of width columns instead of 80.

For compatibility fold supports an obsolete option syntax -width. New scripts should use -w width instead.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

5 Output of parts of files

These commands output pieces of the input.


Next: , Up: Output of parts of files   [Contents][Index]

5.1 head: Output the first part of files

head prints the first part (10 lines by default) of each file; it reads from standard input if no files are given or when given a file of -. Synopsis:

head [option]… [file]…

If more than one file is specified, head prints a one-line header consisting of:

==> file name <==

before the output for each file.

The program accepts the following options. Also see Common options.

-c k
--bytes=k

Print the first k bytes, instead of initial lines. However, if k starts with a ‘-’, print all but the last k bytes of each file. k may be, or may be an integer optionally followed by, one of the following multiplicative suffixes:

b’  =>            512 ("blocks")
‘KB’ =>           1000 (KiloBytes)
‘K’  =>           1024 (KibiBytes)
‘MB’ =>      1000*1000 (MegaBytes)
‘M’  =>      1024*1024 (MebiBytes)
‘GB’ => 1000*1000*1000 (GigaBytes)
‘G’  => 1024*1024*1024 (GibiBytes)

and so on for ‘T’, ‘P’, ‘E’, ‘Z’, and ‘Y’.

-n k
--lines=k

Output the first k lines. However, if k starts with a ‘-’, print all but the last k lines of each file. Size multiplier suffixes are the same as with the -c option.

-q
--quiet
--silent

Never print file name headers.

-v
--verbose

Always print file name headers.

For compatibility head also supports an obsolete option syntax -countoptions, which is recognized only if it is specified first. count is a decimal number optionally followed by a size letter (‘b’, ‘k’, ‘m’) as in -c, or ‘l’ to mean count by lines, or other option letters (‘cqv’). Scripts intended for standard hosts should use -c count or -n count instead. If your script must also run on hosts that support only the obsolete syntax, it is usually simpler to avoid head, e.g., by using ‘sed 5q’ instead of ‘head -5’.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Output of parts of files   [Contents][Index]

5.2 tail: Output the last part of files

tail prints the last part (10 lines by default) of each file; it reads from standard input if no files are given or when given a file of ‘-’. Synopsis:

tail [option]… [file]…

If more than one file is specified, tail prints a one-line header consisting of:

==> file name <==

before the output for each file.

GNU tail can output any amount of data (some other versions of tail cannot). It also has no -r option (print in reverse), since reversing a file is really a different job from printing the end of a file; BSD tail (which is the one with -r) can only reverse files that are at most as large as its buffer, which is typically 32 KiB. A more reliable and versatile way to reverse files is the GNU tac command.

The program accepts the following options. Also see Common options.

-c k
--bytes=k

Output the last k bytes, instead of final lines. However, if k starts with a ‘+’, start printing with the kth byte from the start of each file, instead of from the end. k may be, or may be an integer optionally followed by, one of the following multiplicative suffixes:

b’  =>            512 ("blocks")
‘KB’ =>           1000 (KiloBytes)
‘K’  =>           1024 (KibiBytes)
‘MB’ =>      1000*1000 (MegaBytes)
‘M’  =>      1024*1024 (MebiBytes)
‘GB’ => 1000*1000*1000 (GigaBytes)
‘G’  => 1024*1024*1024 (GibiBytes)

and so on for ‘T’, ‘P’, ‘E’, ‘Z’, and ‘Y’.

-f
--follow[=how]

Loop forever trying to read more characters at the end of the file, presumably because the file is growing. If more than one file is given, tail prints a header whenever it gets output from a different file, to indicate which file that output is from.

There are two ways to specify how you’d like to track files with this option, but that difference is noticeable only when a followed file is removed or renamed. If you’d like to continue to track the end of a growing file even after it has been unlinked, use --follow=descriptor. This is the default behavior, but it is not useful if you’re tracking a log file that may be rotated (removed or renamed, then reopened). In that case, use --follow=name to track the named file, perhaps by reopening it periodically to see if it has been removed and recreated by some other program. Note that the inotify-based implementation handles this case without the need for any periodic reopening.

No matter which method you use, if the tracked file is determined to have shrunk, tail prints a message saying the file has been truncated and resumes tracking the end of the file from the newly-determined endpoint.

When a file is removed, tail’s behavior depends on whether it is following the name or the descriptor. When following by name, tail can detect that a file has been removed and gives a message to that effect, and if --retry has been specified it will continue checking periodically to see if the file reappears. When following a descriptor, tail does not detect that the file has been unlinked or renamed and issues no message; even though the file may no longer be accessible via its original name, it may still be growing.

The option values ‘descriptor’ and ‘name’ may be specified only with the long form of the option, not with -f.

The -f option is ignored if no file operand is specified and standard input is a FIFO or a pipe. Likewise, the -f option has no effect for any operand specified as ‘-’, when standard input is a FIFO or a pipe.

With kernel inotify support, output is triggered by file changes and is generally very prompt. Otherwise, tail sleeps for one second between checks— use --sleep-interval=n to change that default—which can make the output appear slightly less responsive or bursty. When using tail without inotify support, you can make it more responsive by using a sub-second sleep interval, e.g., via an alias like this:

alias tail='tail -s.1'
-F

This option is the same as --follow=name --retry. That is, tail will attempt to reopen a file when it is removed. Should this fail, tail will keep trying until it becomes accessible again.

--retry

Indefinitely try to open the specified file. This option is useful mainly when following (and otherwise issues a warning).

When following by file descriptor (i.e., with --follow=descriptor), this option only affects the initial open of the file, as after a successful open, tail will start following the file descriptor.

When following by name (i.e., with --follow=name), tail infinitely retries to re-open the given files until killed.

Without this option, when tail encounters a file that doesn’t exist or is otherwise inaccessible, it reports that fact and never checks it again.

--sleep-interval=number

Change the number of seconds to wait between iterations (the default is 1.0). During one iteration, every specified file is checked to see if it has changed size. Historical implementations of tail have required that number be an integer. However, GNU tail accepts an arbitrary floating point number. See Floating point. When tail uses inotify, this polling-related option is usually ignored. However, if you also specify --pid=p, tail checks whether process p is alive at least every number seconds.

--pid=pid

When following by name or by descriptor, you may specify the process ID, pid, of the sole writer of all file arguments. Then, shortly after that process terminates, tail will also terminate. This will work properly only if the writer and the tailing process are running on the same machine. For example, to save the output of a build in a file and to watch the file grow, if you invoke make and tail like this then the tail process will stop when your build completes. Without this option, you would have had to kill the tail -f process yourself.

$ make >& makerr & tail --pid=$! -f makerr

If you specify a pid that is not in use or that does not correspond to the process that is writing to the tailed files, then tail may terminate long before any files stop growing or it may not terminate until long after the real writer has terminated. Note that --pid cannot be supported on some systems; tail will print a warning if this is the case.

--max-unchanged-stats=n

When tailing a file by name, if there have been n (default n=5) consecutive iterations for which the file has not changed, then open/fstat the file to determine if that file name is still associated with the same device/inode-number pair as before. When following a log file that is rotated, this is approximately the number of seconds between when tail prints the last pre-rotation lines and when it prints the lines that have accumulated in the new log file. This option is meaningful only when polling (i.e., without inotify) and when following by name.

-n k
--lines=k

Output the last k lines. However, if k starts with a ‘+’, start printing with the kth line from the start of each file, instead of from the end. Size multiplier suffixes are the same as with the -c option.

-q
--quiet
--silent

Never print file name headers.

-v
--verbose

Always print file name headers.

For compatibility tail also supports an obsolete usage ‘tail -[count][bcl][f] [file]’, which is recognized only if it does not conflict with the usage described above. This obsolete form uses exactly one option and at most one file. In the option, count is an optional decimal number optionally followed by a size letter (‘b’, ‘c’, ‘l’) to mean count by 512-byte blocks, bytes, or lines, optionally followed by ‘f’ which has the same meaning as -f.

On older systems, the leading ‘-’ can be replaced by ‘+’ in the obsolete option syntax with the same meaning as in counts, and obsolete usage overrides normal usage when the two conflict. This obsolete behavior can be enabled or disabled with the _POSIX2_VERSION environment variable (see Standards conformance).

Scripts intended for use on standard hosts should avoid obsolete syntax and should use -c count[b], -n count, and/or -f instead. If your script must also run on hosts that support only the obsolete syntax, you can often rewrite it to avoid problematic usages, e.g., by using ‘sed -n '$p'’ rather than ‘tail -1’. If that’s not possible, the script can use a test like ‘if tail -c +1 </dev/null >/dev/null 2>&1; then …’ to decide which syntax to use.

Even if your script assumes the standard behavior, you should still beware usages whose behaviors differ depending on the POSIX version. For example, avoid ‘tail - main.c’, since it might be interpreted as either ‘tail main.c’ or as ‘tail -- - main.c’; avoid ‘tail -c 4’, since it might mean either ‘tail -c4’ or ‘tail -c 10 4’; and avoid ‘tail +4’, since it might mean either ‘tail ./+4’ or ‘tail -n +4’.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Output of parts of files   [Contents][Index]

5.3 split: Split a file into pieces.

split creates output files containing consecutive or interleaved sections of input (standard input if none is given or input is ‘-’). Synopsis:

split [option] [input [prefix]]

By default, split puts 1000 lines of input (or whatever is left over for the last section), into each output file.

The output files’ names consist of prefix (‘x’ by default) followed by a group of characters (‘aa’, ‘ab’, … by default), such that concatenating the output files in traditional sorted order by file name produces the original input file (except -nr/n). By default split will initially create files with two generated suffix characters, and will increase this width by two when the next most significant position reaches the last character. (‘yz’, ‘zaaa’, ‘zaab’, …). In this way an arbitrary number of output files are supported, which sort as described above, even in the presence of an --additional-suffix option. If the -a option is specified and the output file names are exhausted, split reports an error without deleting the output files that it did create.

The program accepts the following options. Also see Common options.

-l lines
--lines=lines

Put lines lines of input into each output file. If --separator is specified, then lines determines the number of records.

For compatibility split also supports an obsolete option syntax -lines. New scripts should use -l lines instead.

-b size
--bytes=size

Put size bytes of input into each output file. size may be, or may be an integer optionally followed by, one of the following multiplicative suffixes:

b’  =>            512 ("blocks")
‘KB’ =>           1000 (KiloBytes)
‘K’  =>           1024 (KibiBytes)
‘MB’ =>      1000*1000 (MegaBytes)
‘M’  =>      1024*1024 (MebiBytes)
‘GB’ => 1000*1000*1000 (GigaBytes)
‘G’  => 1024*1024*1024 (GibiBytes)

and so on for ‘T’, ‘P’, ‘E’, ‘Z’, and ‘Y’.

-C size
--line-bytes=size

Put into each output file as many complete lines of input as possible without exceeding size bytes. Individual lines or records longer than size bytes are broken into multiple files. size has the same format as for the --bytes option. If --separator is specified, then lines determines the number of records.

--filter=command

With this option, rather than simply writing to each output file, write through a pipe to the specified shell command for each output file. command should use the $FILE environment variable, which is set to a different output file name for each invocation of the command. For example, imagine that you have a 1TiB compressed file that, if uncompressed, would be too large to reside on disk, yet you must split it into individually-compressed pieces of a more manageable size. To do that, you might run this command:

xz -dc BIG.xz | split -b200G --filter='xz > $FILE.xz' - big-

Assuming a 10:1 compression ratio, that would create about fifty 20GiB files with names big-aa.xz, big-ab.xz, big-ac.xz, etc.

-n chunks
--number=chunks

Split input to chunks output files where chunks may be:

n      generate n files based on current size of input
k/n    only output kth of n to stdout
l/n    generate n files without splitting lines or records
l/k/n  likewise but only output kth of n to stdout
r/n    like ‘l’ but use round robin distribution
r/k/n  likewise but only output kth of n to stdout

Any excess bytes remaining after dividing the input into n chunks, are assigned to the last chunk. Any excess bytes appearing after the initial calculation are discarded (except when using ‘r’ mode).

All n files are created even if there are fewer than n lines, or the input is truncated.

For ‘l’ mode, chunks are approximately input size / n. The input is partitioned into n equal sized portions, with the last assigned any excess. If a line starts within a partition it is written completely to the corresponding file. Since lines or records are not split even if they overlap a partition, the files written can be larger or smaller than the partition size, and even empty if a line/record is so long as to completely overlap the partition.

For ‘r’ mode, the size of input is irrelevant, and so can be a pipe for example.

-a length
--suffix-length=length

Use suffixes of length length. If a length of 0 is specified, this is the same as if (any previous) -a was not specified, and thus enables the default behavior, which starts the suffix length at 2, and unless -n or --numeric-suffixes=from is specified, will auto increase the length by 2 as required.

-d
--numeric-suffixes[=from]

Use digits in suffixes rather than lower-case letters. The numerical suffix counts from from if specified, 0 otherwise.

from is used to either set the initial suffix for a single run, or to set the suffix offset for independently split inputs, and consequently the auto suffix length expansion described above is disabled. Therefore you may also want to use option -a to allow suffixes beyond ‘99’. Note if option --number is specified and the number of files is less than from, a single run is assumed and the minimum suffix length required is automatically determined.

--additional-suffix=suffix

Append an additional suffix to output file names. suffix must not contain slash.

-e
--elide-empty-files

Suppress the generation of zero-length output files. This can happen with the --number option if a file is (truncated to be) shorter than the number requested, or if a line is so long as to completely span a chunk. The output file sequence numbers, always run consecutively even when this option is specified.

-t separator
--separator=separator

Use character separator as the record separator instead of the default newline character (ASCII LF). To specify ASCII NUL as the separator, use the two-character string ‘\0’, e.g., ‘split -t '\0'’.

-u
--unbuffered

Immediately copy input to output in --number r/… mode, which is a much slower mode of operation.

--verbose

Write a diagnostic just before each output file is opened.

An exit status of zero indicates success, and a nonzero value indicates failure.

Here are a few examples to illustrate how the --number (-n) option works:

Notice how, by default, one line may be split onto two or more:

$ seq -w 6 10 > k; split -n3 k; head xa?
==> xaa <==
06
07
==> xab <==

08
0
==> xac <==
9
10

Use the "l/" modifier to suppress that:

$ seq -w 6 10 > k; split -nl/3 k; head xa?
==> xaa <==
06
07

==> xab <==
08
09

==> xac <==
10

Use the "r/" modifier to distribute lines in a round-robin fashion:

$ seq -w 6 10 > k; split -nr/3 k; head xa?
==> xaa <==
06
09

==> xab <==
07
10

==> xac <==
08

You can also extract just the Kth chunk. This extracts and prints just the 7th "chunk" of 33:

$ seq 100 > k; split -nl/7/33 k
20
21
22

Previous: , Up: Output of parts of files   [Contents][Index]

5.4 csplit: Split a file into context-determined pieces

csplit creates zero or more output files containing sections of input (standard input if input is ‘-’). Synopsis:

csplit [option]… input pattern

The contents of the output files are determined by the pattern arguments, as detailed below. An error occurs if a pattern argument refers to a nonexistent line of the input file (e.g., if no remaining line matches a given regular expression). After every pattern has been matched, any remaining input is copied into one last output file.

By default, csplit prints the number of bytes written to each output file after it has been created.

The types of pattern arguments are:

n

Create an output file containing the input up to but not including line n (a positive integer). If followed by a repeat count, also create an output file containing the next n lines of the input file once for each repeat.

/regexp/[offset]

Create an output file containing the current line up to (but not including) the next line of the input file that contains a match for regexp. The optional offset is an integer. If it is given, the input up to (but not including) the matching line plus or minus offset is put into the output file, and the line after that begins the next section of input.

%regexp%[offset]

Like the previous type, except that it does not create an output file, so that section of the input file is effectively ignored.

{repeat-count}

Repeat the previous pattern repeat-count additional times. The repeat-count can either be a positive integer or an asterisk, meaning repeat as many times as necessary until the input is exhausted.

The output files’ names consist of a prefix (‘xx’ by default) followed by a suffix. By default, the suffix is an ascending sequence of two-digit decimal numbers from ‘00’ to ‘99’. In any case, concatenating the output files in sorted order by file name produces the original input file.

By default, if csplit encounters an error or receives a hangup, interrupt, quit, or terminate signal, it removes any output files that it has created so far before it exits.

The program accepts the following options. Also see Common options.

-f prefix
--prefix=prefix

Use prefix as the output file name prefix.

-b suffix
--suffix=suffix

Use suffix as the output file name suffix. When this option is specified, the suffix string must include exactly one printf(3)-style conversion specification, possibly including format specification flags, a field width, a precision specifications, or all of these kinds of modifiers. The format letter must convert a binary unsigned integer argument to readable form. The format letters ‘d’ and ‘i’ are aliases for ‘u’, and the ‘u’, ‘o’, ‘x’, and ‘X’ conversions are allowed. The entire suffix is given (with the current output file number) to sprintf(3) to form the file name suffixes for each of the individual output files in turn. If this option is used, the --digits option is ignored.

-n digits
--digits=digits

Use output file names containing numbers that are digits digits long instead of the default 2.

-k
--keep-files

Do not remove output files when errors are encountered.

--suppress-matched

Do not output lines matching the specified pattern. I.e., suppress the boundary line from the start of the second and subsequent splits.

-z
--elide-empty-files

Suppress the generation of zero-length output files. (In cases where the section delimiters of the input file are supposed to mark the first lines of each of the sections, the first output file will generally be a zero-length file unless you use this option.) The output file sequence numbers always run consecutively starting from 0, even when this option is specified.

-s
-q
--silent
--quiet

Do not print counts of output file sizes.

An exit status of zero indicates success, and a nonzero value indicates failure.

Here is an example of its usage. First, create an empty directory for the exercise, and cd into it:

$ mkdir d && cd d

Now, split the sequence of 1..14 on lines that end with 0 or 5:

$ seq 14 | csplit - '/[05]$/' '{*}'
8
10
15

Each number printed above is the size of an output file that csplit has just created. List the names of those output files:

$ ls
xx00  xx01  xx02

Use head to show their contents:

$ head xx*
==> xx00 <==
1
2
3
4

==> xx01 <==
5
6
7
8
9

==> xx02 <==
10
11
12
13
14

Example of splitting input by empty lines:

$ csplit --suppress-matched input.txt '/^$/' '{*}'

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

6 Summarizing files

These commands generate just a few numbers representing entire contents of files.


Next: , Up: Summarizing files   [Contents][Index]

6.1 wc: Print newline, word, and byte counts

wc counts the number of bytes, characters, whitespace-separated words, and newlines in each given file, or standard input if none are given or for a file of ‘-’. Synopsis:

wc [option]… [file]…

wc prints one line of counts for each file, and if the file was given as an argument, it prints the file name following the counts. If more than one file is given, wc prints a final line containing the cumulative counts, with the file name total. The counts are printed in this order: newlines, words, characters, bytes, maximum line length. Each count is printed right-justified in a field with at least one space between fields so that the numbers and file names normally line up nicely in columns. The width of the count fields varies depending on the inputs, so you should not depend on a particular field width. However, as a GNU extension, if only one count is printed, it is guaranteed to be printed without leading spaces.

By default, wc prints three counts: the newline, words, and byte counts. Options can specify that only certain counts be printed. Options do not undo others previously given, so

wc --bytes --words

prints both the byte counts and the word counts.

With the --max-line-length option, wc prints the length of the longest line per file, and if there is more than one file it prints the maximum (not the sum) of those lengths. The line lengths here are measured in screen columns, according to the current locale and assuming tab positions in every 8th column.

The program accepts the following options. Also see Common options.

-c
--bytes

Print only the byte counts.

-m
--chars

Print only the character counts.

-w
--words

Print only the word counts.

-l
--lines

Print only the newline counts.

-L
--max-line-length

Print only the maximum display widths. Tabs are set at every 8th column. Display widths of wide characters are considered. Non-printable characters are given 0 width.

--files0-from=file

Disallow processing files named on the command line, and instead process those named in file file; each name being terminated by a zero byte (ASCII NUL). This is useful when the list of file names is so long that it may exceed a command line length limitation. In such cases, running wc via xargs is undesirable because it splits the list into pieces and makes wc print a total for each sublist rather than for the entire list. One way to produce a list of ASCII NUL terminated file names is with GNU find, using its -print0 predicate. If file is ‘-’ then the ASCII NUL terminated file names are read from standard input.

For example, to find the length of the longest line in any .c or .h file in the current hierarchy, do this:

find . -name '*.[ch]' -print0 |
  wc -L --files0-from=- | tail -n1

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Summarizing files   [Contents][Index]

6.2 sum: Print checksum and block counts

sum computes a 16-bit checksum for each given file, or standard input if none are given or for a file of ‘-’. Synopsis:

sum [option]… [file]…

sum prints the checksum for each file followed by the number of blocks in the file (rounded up). If more than one file is given, file names are also printed (by default). (With the --sysv option, corresponding file names are printed when there is at least one file argument.)

By default, GNU sum computes checksums using an algorithm compatible with BSD sum and prints file sizes in units of 1024-byte blocks.

The program accepts the following options. Also see Common options.

-r

Use the default (BSD compatible) algorithm. This option is included for compatibility with the System V sum. Unless -s was also given, it has no effect.

-s
--sysv

Compute checksums using an algorithm compatible with System V sum’s default, and print file sizes in units of 512-byte blocks.

sum is provided for compatibility; the cksum program (see next section) is preferable in new applications.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Summarizing files   [Contents][Index]

6.3 cksum: Print CRC checksum and byte counts

cksum computes a cyclic redundancy check (CRC) checksum for each given file, or standard input if none are given or for a file of ‘-’. Synopsis:

cksum [option]… [file]…

cksum prints the CRC checksum for each file along with the number of bytes in the file, and the file name unless no arguments were given.

cksum is typically used to ensure that files transferred by unreliable means (e.g., netnews) have not been corrupted, by comparing the cksum output for the received files with the cksum output for the original files (typically given in the distribution).

The CRC algorithm is specified by the POSIX standard. It is not compatible with the BSD or System V sum algorithms (see the previous section); it is more robust.

The only options are --help and --version. See Common options.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Summarizing files   [Contents][Index]

6.4 md5sum: Print or check MD5 digests

md5sum computes a 128-bit checksum (or fingerprint or message-digest) for each specified file.

Note: The MD5 digest is more reliable than a simple CRC (provided by the cksum command) for detecting accidental file corruption, as the chances of accidentally having two files with identical MD5 are vanishingly small. However, it should not be considered secure against malicious tampering: although finding a file with a given MD5 fingerprint is considered infeasible at the moment, it is known how to modify certain files, including digital certificates, so that they appear valid when signed with an MD5 digest. For more secure hashes, consider using SHA-2. See sha2 utilities.

If a file is specified as ‘-’ or if no files are given md5sum computes the checksum for the standard input. md5sum can also determine whether a file and checksum are consistent. Synopsis:

md5sum [option]… [file]…

For each file, ‘md5sum’ outputs by default, the MD5 checksum, a space, a flag indicating binary or text input mode, and the file name. Binary mode is indicated with ‘*’, text mode with ‘ ’ (space). Binary mode is the default on systems where it’s significant, otherwise text mode is the default. If file contains a backslash or newline, the line is started with a backslash, and each problematic character in the file name is escaped with a backslash, making the output unambiguous even in the presence of arbitrary file names. If file is omitted or specified as ‘-’, standard input is read.

The program accepts the following options. Also see Common options.

-b
--binary

Treat each input file as binary, by reading it in binary mode and outputting a ‘*’ flag. This is the inverse of --text. On systems like GNU that do not distinguish between binary and text files, this option merely flags each input mode as binary: the MD5 checksum is unaffected. This option is the default on systems like MS-DOS that distinguish between binary and text files, except for reading standard input when standard input is a terminal.

-c
--check

Read file names and checksum information (not data) from each file (or from stdin if no file was specified) and report whether the checksums match the contents of the named files. The input to this mode of md5sum is usually the output of a prior, checksum-generating run of ‘md5sum’. Three input formats are supported. Either the default output format described above, the --tag output format, or the BSD reversed mode format which is similar to the default mode, but doesn’t use a character to distinguish binary and text modes.


For each such line, md5sum reads the named file and computes its MD5 checksum. Then, if the computed message digest does not match the one on the line with the file name, the file is noted as having failed the test. Otherwise, the file passes the test. By default, for each valid line, one line is written to standard output indicating whether the named file passed the test. After all checks have been performed, if there were any failures, a warning is issued to standard error. Use the --status option to inhibit that output. If any listed file cannot be opened or read, if any valid line has an MD5 checksum inconsistent with the associated file, or if no valid line is found, md5sum exits with nonzero status. Otherwise, it exits successfully.

--quiet

This option is useful only when verifying checksums. When verifying checksums, don’t generate an ’OK’ message per successfully checked file. Files that fail the verification are reported in the default one-line-per-file format. If there is any checksum mismatch, print a warning summarizing the failures to standard error.

--status

This option is useful only when verifying checksums. When verifying checksums, don’t generate the default one-line-per-file diagnostic and don’t output the warning summarizing any failures. Failures to open or read a file still evoke individual diagnostics to standard error. If all listed files are readable and are consistent with the associated MD5 checksums, exit successfully. Otherwise exit with a status code indicating there was a failure.

--tag

Output BSD style checksums, which indicate the checksum algorithm used. As a GNU extension, file names with problematic characters are escaped as described above, with the same escaping indicator of ‘\’ at the start of the line, being used. The --tag option implies binary mode, and is disallowed with --text mode as supporting that would unnecessarily complicate the output format, while providing little benefit.

-t
--text

Treat each input file as text, by reading it in text mode and outputting a ‘ ’ flag. This is the inverse of --binary. This option is the default on systems like GNU that do not distinguish between binary and text files. On other systems, it is the default for reading standard input when standard input is a terminal. This mode is never defaulted to if --tag is used.

-w
--warn

When verifying checksums, warn about improperly formatted MD5 checksum lines. This option is useful only if all but a few lines in the checked input are valid.

--strict

When verifying checksums, if one or more input line is invalid, exit nonzero after all warnings have been issued.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Summarizing files   [Contents][Index]

6.5 sha1sum: Print or check SHA-1 digests

sha1sum computes a 160-bit checksum for each specified file. The usage and options of this command are precisely the same as for md5sum. See md5sum invocation.

Note: The SHA-1 digest is more secure than MD5, and no collisions of it are known (different files having the same fingerprint). However, it is known that they can be produced with considerable, but not unreasonable, resources. For this reason, it is generally considered that SHA-1 should be gradually phased out in favor of the more secure SHA-2 hash algorithms. See sha2 utilities.


Previous: , Up: Summarizing files   [Contents][Index]

6.6 sha2 utilities: Print or check SHA-2 digests

The commands sha224sum, sha256sum, sha384sum and sha512sum compute checksums of various lengths (respectively 224, 256, 384 and 512 bits), collectively known as the SHA-2 hashes. The usage and options of these commands are precisely the same as for md5sum and sha1sum. See md5sum invocation.

Note: The SHA384 and SHA512 digests are considerably slower to compute, especially on 32-bit computers, than SHA224 or SHA256.


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

7 Operating on sorted files

These commands work with (or produce) sorted files.


Next: , Up: Operating on sorted files   [Contents][Index]

7.1 sort: Sort text files

sort sorts, merges, or compares all the lines from the given files, or standard input if none are given or for a file of ‘-’. By default, sort writes the results to standard output. Synopsis:

sort [option]… [file]…

sort has three modes of operation: sort (the default), merge, and check for sortedness. The following options change the operation mode:

-c
--check
--check=diagnose-first

Check whether the given file is already sorted: if it is not all sorted, print a diagnostic containing the first out-of-order line and exit with a status of 1. Otherwise, exit successfully. At most one input file can be given.

-C
--check=quiet
--check=silent

Exit successfully if the given file is already sorted, and exit with status 1 otherwise. At most one input file can be given. This is like -c, except it does not print a diagnostic.

-m
--merge

Merge the given files by sorting them as a group. Each input file must always be individually sorted. It always works to sort instead of merge; merging is provided because it is faster, in the case where it works.

A pair of lines is compared as follows: sort compares each pair of fields, in the order specified on the command line, according to the associated ordering options, until a difference is found or no fields are left. If no key fields are specified, sort uses a default key of the entire line. Finally, as a last resort when all keys compare equal, sort compares entire lines as if no ordering options other than --reverse (-r) were specified. The --stable (-s) option disables this last-resort comparison so that lines in which all fields compare equal are left in their original relative order. The --unique (-u) option also disables the last-resort comparison.

Unless otherwise specified, all comparisons use the character collating sequence specified by the LC_COLLATE locale.2

GNU sort (as specified for all GNU utilities) has no limit on input line length or restrictions on bytes allowed within lines. In addition, if the final byte of an input file is not a newline, GNU sort silently supplies one. A line’s trailing newline is not part of the line for comparison purposes.

Exit status:

0 if no error occurred
1 if invoked with -c or -C and the input is not sorted
2 if an error occurred

If the environment variable TMPDIR is set, sort uses its value as the directory for temporary files instead of /tmp. The --temporary-directory (-T) option in turn overrides the environment variable.

The following options affect the ordering of output lines. They may be specified globally or as part of a specific key field. If no key fields are specified, global options apply to comparison of entire lines; otherwise the global options are inherited by key fields that do not specify any special options of their own. In pre-POSIX versions of sort, global options affect only later key fields, so portable shell scripts should specify global options first.

-b
--ignore-leading-blanks

Ignore leading blanks when finding sort keys in each line. By default a blank is a space or a tab, but the LC_CTYPE locale can change this. Note blanks may be ignored by your locale’s collating rules, but without this option they will be significant for character positions specified in keys with the -k option.

-d
--dictionary-order

Sort in phone directory order: ignore all characters except letters, digits and blanks when sorting. By default letters and digits are those of ASCII and a blank is a space or a tab, but the LC_CTYPE locale can change this.

-f
--ignore-case

Fold lowercase characters into the equivalent uppercase characters when comparing so that, for example, ‘b’ and ‘B’ sort as equal. The LC_CTYPE locale determines character types. When used with --unique those lower case equivalent lines are thrown away. (There is currently no way to throw away the upper case equivalent instead. (Any --reverse given would only affect the final result, after the throwing away.))

-g
--general-numeric-sort
--sort=general-numeric

Sort numerically, converting a prefix of each line to a long double-precision floating point number. See Floating point. Do not report overflow, underflow, or conversion errors. Use the following collating sequence:

Use this option only if there is no alternative; it is much slower than --numeric-sort (-n) and it can lose information when converting to floating point.

-h
--human-numeric-sort
--sort=human-numeric

Sort numerically, first by numeric sign (negative, zero, or positive); then by SI suffix (either empty, or ‘k’ or ‘K’, or one of ‘MGTPEZY’, in that order; see Block size); and finally by numeric value. For example, ‘1023M’ sorts before ‘1G’ because ‘M’ (mega) precedes ‘G’ (giga) as an SI suffix. This option sorts values that are consistently scaled to the nearest suffix, regardless of whether suffixes denote powers of 1000 or 1024, and it therefore sorts the output of any single invocation of the df, du, or ls commands that are invoked with their --human-readable or --si options. The syntax for numbers is the same as for the --numeric-sort option; the SI suffix must immediately follow the number. Note also the numfmt command, which can be used to reformat numbers to human format after the sort, thus often allowing sort to operate on more accurate numbers.

-i
--ignore-nonprinting

Ignore nonprinting characters. The LC_CTYPE locale determines character types. This option has no effect if the stronger --dictionary-order (-d) option is also given.

-M
--month-sort
--sort=month

An initial string, consisting of any amount of blanks, followed by a month name abbreviation, is folded to UPPER case and compared in the order ‘JAN’ < ‘FEB’ < … < ‘DEC’. Invalid names compare low to valid names. The LC_TIME locale category determines the month spellings. By default a blank is a space or a tab, but the LC_CTYPE locale can change this.

-n
--numeric-sort
--sort=numeric

Sort numerically. The number begins each line and consists of optional blanks, an optional ‘-’ sign, and zero or more digits possibly separated by thousands separators, optionally followed by a decimal-point character and zero or more digits. An empty number is treated as ‘0’. The LC_NUMERIC locale specifies the decimal-point character and thousands separator. By default a blank is a space or a tab, but the LC_CTYPE locale can change this.

Comparison is exact; there is no rounding error.

Neither a leading ‘+’ nor exponential notation is recognized. To compare such strings numerically, use the --general-numeric-sort (-g) option.

-V
--version-sort

Sort by version name and number. It behaves like a standard sort, except that each sequence of decimal digits is treated numerically as an index/version number. (See Details about version sort.)

-r
--reverse

Reverse the result of comparison, so that lines with greater key values appear earlier in the output instead of later.

-R
--random-sort
--sort=random

Sort by hashing the input keys and then sorting the hash values. Choose the hash function at random, ensuring that it is free of collisions so that differing keys have differing hash values. This is like a random permutation of the inputs (see shuf invocation), except that keys with the same value sort together.

If multiple random sort fields are specified, the same random hash function is used for all fields. To use different random hash functions for different fields, you can invoke sort more than once.

The choice of hash function is affected by the --random-source option.

Other options are:

--compress-program=prog

Compress any temporary files with the program prog.

With no arguments, prog must compress standard input to standard output, and when given the -d option it must decompress standard input to standard output.

Terminate with an error if prog exits with nonzero status.

White space and the backslash character should not appear in prog; they are reserved for future use.

--files0-from=file

Disallow processing files named on the command line, and instead process those named in file file; each name being terminated by a zero byte (ASCII NUL). This is useful when the list of file names is so long that it may exceed a command line length limitation. In such cases, running sort via xargs is undesirable because it splits the list into pieces and makes sort print sorted output for each sublist rather than for the entire list. One way to produce a list of ASCII NUL terminated file names is with GNU find, using its -print0 predicate. If file is ‘-’ then the ASCII NUL terminated file names are read from standard input.

-k pos1[,pos2]
--key=pos1[,pos2]

Specify a sort field that consists of the part of the line between pos1 and pos2 (or the end of the line, if pos2 is omitted), inclusive.

Each pos has the form ‘f[.c][opts]’, where f is the number of the field to use, and c is the number of the first character from the beginning of the field. Fields and character positions are numbered starting with 1; a character position of zero in pos2 indicates the field’s last character. If ‘.c’ is omitted from pos1, it defaults to 1 (the beginning of the field); if omitted from pos2, it defaults to 0 (the end of the field). opts are ordering options, allowing individual keys to be sorted according to different rules; see below for details. Keys can span multiple fields.

Example: To sort on the second field, use --key=2,2 (-k 2,2). See below for more notes on keys and more examples. See also the --debug option to help determine the part of the line being used in the sort.

--debug

Highlight the portion of each line used for sorting. Also issue warnings about questionable usage to stderr.

--batch-size=nmerge

Merge at most nmerge inputs at once.

When sort has to merge more than nmerge inputs, it merges them in groups of nmerge, saving the result in a temporary file, which is then used as an input in a subsequent merge.

A large value of nmerge may improve merge performance and decrease temporary storage utilization at the expense of increased memory usage and I/O. Conversely a small value of nmerge may reduce memory requirements and I/O at the expense of temporary storage consumption and merge performance.

The value of nmerge must be at least 2. The default value is currently 16, but this is implementation-dependent and may change in the future.

The value of nmerge may be bounded by a resource limit for open file descriptors. The commands ‘ulimit -n’ or ‘getconf OPEN_MAX’ may display limits for your systems; these limits may be modified further if your program already has some files open, or if the operating system has other limits on the number of open files. If the value of nmerge exceeds the resource limit, sort silently uses a smaller value.

-o output-file
--output=output-file

Write output to output-file instead of standard output. Normally, sort reads all input before opening output-file, so you can safely sort a file in place by using commands like sort -o F F and cat F | sort -o F. However, sort with --merge (-m) can open the output file before reading all input, so a command like cat F | sort -m -o F - G is not safe as sort might start writing F before cat is done reading it.

On newer systems, -o cannot appear after an input file if POSIXLY_CORRECT is set, e.g., ‘sort F -o F’. Portable scripts should specify -o output-file before any input files.

--random-source=file

Use file as a source of random data used to determine which random hash function to use with the -R option. See Random sources.

-s
--stable

Make sort stable by disabling its last-resort comparison. This option has no effect if no fields or global ordering options other than --reverse (-r) are specified.

-S size
--buffer-size=size

Use a main-memory sort buffer of the given size. By default, size is in units of 1024 bytes. Appending ‘%’ causes size to be interpreted as a percentage of physical memory. Appending ‘K’ multiplies size by 1024 (the default), ‘M’ by 1,048,576, ‘G’ by 1,073,741,824, and so on for ‘T’, ‘P’, ‘E’, ‘Z’, and ‘Y’. Appending ‘b’ causes size to be interpreted as a byte count, with no multiplication.

This option can improve the performance of sort by causing it to start with a larger or smaller sort buffer than the default. However, this option affects only the initial buffer size. The buffer grows beyond size if sort encounters input lines larger than size.

-t separator
--field-separator=separator

Use character separator as the field separator when finding the sort keys in each line. By default, fields are separated by the empty string between a non-blank character and a blank character. By default a blank is a space or a tab, but the LC_CTYPE locale can change this.

That is, given the input line ‘ foo bar, sort breaks it into fields ‘ foo and ‘ bar. The field separator is not considered to be part of either the field preceding or the field following, so with ‘sort -t " "’ the same input line has three fields: an empty field, ‘foo’, and ‘bar’. However, fields that extend to the end of the line, as -k 2, or fields consisting of a range, as -k 2,3, retain the field separators present between the endpoints of the range.

To specify ASCII NUL as the field separator, use the two-character string ‘\0’, e.g., ‘sort -t '\0'’.

-T tempdir
--temporary-directory=tempdir

Use directory tempdir to store temporary files, overriding the TMPDIR environment variable. If this option is given more than once, temporary files are stored in all the directories given. If you have a large sort or merge that is I/O-bound, you can often improve performance by using this option to specify directories on different disks and controllers.

--parallel=n

Set the number of sorts run in parallel to n. By default, n is set to the number of available processors, but limited to 8, as there are diminishing performance gains after that. Note also that using n threads increases the memory usage by a factor of log n. Also see nproc invocation.

-u
--unique

Normally, output only the first of a sequence of lines that compare equal. For the --check (-c or -C) option, check that no pair of consecutive lines compares equal.

This option also disables the default last-resort comparison.

The commands sort -u and sort | uniq are equivalent, but this equivalence does not extend to arbitrary sort options. For example, sort -n -u inspects only the value of the initial numeric string when checking for uniqueness, whereas sort -n | uniq inspects the entire line. See uniq invocation.

-z
--zero-terminated

Delimit items with a zero byte rather than a newline (ASCII LF). I.e., treat input as items separated by ASCII NUL and terminate output items with ASCII NUL. This option can be useful in conjunction with ‘perl -0’ or ‘find -print0’ and ‘xargs -0’ which do the same in order to reliably handle arbitrary file names (even those containing blanks or other special characters).

Historical (BSD and System V) implementations of sort have differed in their interpretation of some options, particularly -b, -f, and -n. GNU sort follows the POSIX behavior, which is usually (but not always!) like the System V behavior. According to POSIX, -n no longer implies -b. For consistency, -M has been changed in the same way. This may affect the meaning of character positions in field specifications in obscure cases. The only fix is to add an explicit -b.

A position in a sort field specified with -k may have any of the option letters ‘MbdfghinRrV’ appended to it, in which case no global ordering options are inherited by that particular field. The -b option may be independently attached to either or both of the start and end positions of a field specification, and if it is inherited from the global options it will be attached to both. If input lines can contain leading or adjacent blanks and -t is not used, then -k is typically combined with -b or an option that implicitly ignores leading blanks (‘Mghn’) as otherwise the varying numbers of leading blanks in fields can cause confusing results.

If the start position in a sort field specifier falls after the end of the line or after the end field, the field is empty. If the -b option was specified, the ‘.c’ part of a field specification is counted from the first nonblank character of the field.

On older systems, sort supports an obsolete origin-zero syntax ‘+pos1 [-pos2]’ for specifying sort keys. The obsolete sequence ‘sort +a.x -b.y’ is equivalent to ‘sort -k a+1.x+1,b’ if y is ‘0’ or absent, otherwise it is equivalent to ‘sort -k a+1.x+1,b+1.y’.

This obsolete behavior can be enabled or disabled with the _POSIX2_VERSION environment variable (see Standards conformance); it can also be enabled when POSIXLY_CORRECT is not set by using the obsolete syntax with ‘-pos2’ present.

Scripts intended for use on standard hosts should avoid obsolete syntax and should use -k instead. For example, avoid ‘sort +2’, since it might be interpreted as either ‘sort ./+2’ or ‘sort -k 3’. If your script must also run on hosts that support only the obsolete syntax, it can use a test like ‘if sort -k 1 </dev/null >/dev/null 2>&1; then …’ to decide which syntax to use.

Here are some examples to illustrate various combinations of options.


Next: , Previous: , Up: Operating on sorted files   [Contents][Index]

7.2 shuf: Shuffling text

shuf shuffles its input by outputting a random permutation of its input lines. Each output permutation is equally likely. Synopses:

shuf [option]… [file]
shuf -e [option]… [arg]…
shuf -i lo-hi [option]…

shuf has three modes of operation that affect where it obtains its input lines. By default, it reads lines from standard input. The following options change the operation mode:

-e
--echo

Treat each command-line operand as an input line.

-i lo-hi
--input-range=lo-hi

Act as if input came from a file containing the range of unsigned decimal integers lohi, one per line.

shuf’s other options can affect its behavior in all operation modes:

-n lines
--head-count=count

Output at most count lines. By default, all input lines are output.

-o output-file
--output=output-file

Write output to output-file instead of standard output. shuf reads all input before opening output-file, so you can safely shuffle a file in place by using commands like shuf -o F <F and cat F | shuf -o F.

--random-source=file

Use file as a source of random data used to determine which permutation to generate. See Random sources.

-r
--repeat

Repeat output values, that is, select with replacement. With this option the output is not a permutation of the input; instead, each output line is randomly chosen from all the inputs. This option is typically combined with --head-count; if --head-count is not given, shuf repeats indefinitely.

-z
--zero-terminated

Delimit items with a zero byte rather than a newline (ASCII LF). I.e., treat input as items separated by ASCII NUL and terminate output items with ASCII NUL. This option can be useful in conjunction with ‘perl -0’ or ‘find -print0’ and ‘xargs -0’ which do the same in order to reliably handle arbitrary file names (even those containing blanks or other special characters).

For example:

shuf <<EOF
A man,
a plan,
a canal:
Panama!
EOF

might produce the output

Panama!
A man,
a canal:
a plan,

Similarly, the command:

shuf -e clubs hearts diamonds spades

might output:

clubs
diamonds
spades
hearts

and the command ‘shuf -i 1-4’ might output:

4
2
1
3

The above examples all have four input lines, so shuf might produce any of the twenty-four possible permutations of the input. In general, if there are n input lines, there are n! (i.e., n factorial, or n * (n - 1) * … * 1) possible output permutations.

To output 50 random numbers each in the range 0 through 9, use:

shuf -r -n 50 -i 0-9

To simulate 100 coin flips, use:

shuf -r -n 100 -e Head Tail

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Operating on sorted files   [Contents][Index]

7.3 uniq: Uniquify files

uniq writes the unique lines in the given input, or standard input if nothing is given or for an input name of ‘-’. Synopsis:

uniq [option]… [input [output]]

By default, uniq prints its input lines, except that it discards all but the first of adjacent repeated lines, so that no output lines are repeated. Optionally, it can instead discard lines that are not repeated, or all repeated lines.

The input need not be sorted, but repeated input lines are detected only if they are adjacent. If you want to discard non-adjacent duplicate lines, perhaps you want to use sort -u. See sort invocation.

Comparisons honor the rules specified by the LC_COLLATE locale category.

If no output file is specified, uniq writes to standard output.

The program accepts the following options. Also see Common options.

-f n
--skip-fields=n

Skip n fields on each line before checking for uniqueness. Use a null string for comparison if a line has fewer than n fields. Fields are sequences of non-space non-tab characters that are separated from each other by at least one space or tab.

For compatibility uniq supports an obsolete option syntax -n. New scripts should use -f n instead.

-s n
--skip-chars=n

Skip n characters before checking for uniqueness. Use a null string for comparison if a line has fewer than n characters. If you use both the field and character skipping options, fields are skipped over first.

On older systems, uniq supports an obsolete option syntax +n. This obsolete behavior can be enabled or disabled with the _POSIX2_VERSION environment variable (see Standards conformance), but portable scripts should avoid commands whose behavior depends on this variable. For example, use ‘uniq ./+10’ or ‘uniq -s 10’ rather than the ambiguous ‘uniq +10’.

-c
--count

Print the number of times each line occurred along with the line.

-i
--ignore-case

Ignore differences in case when comparing lines.

-d
--repeated

Discard lines that are not repeated. When used by itself, this option causes uniq to print the first copy of each repeated line, and nothing else.

-D
--all-repeated[=delimit-method]

Do not discard the second and subsequent repeated input lines, but discard lines that are not repeated. This option is useful mainly in conjunction with other options e.g., to ignore case or to compare only selected fields. The optional delimit-method specifies how to delimit groups of repeated lines, and must be one of the following:

none

Do not delimit groups of repeated lines. This is equivalent to --all-repeated (-D).

prepend

Output a newline before each group of repeated lines. With --zero-terminated (-z), use a zero byte (ASCII NUL) instead of a newline as the delimiter.

separate

Separate groups of repeated lines with a single newline. This is the same as using ‘prepend’, except that no delimiter is inserted before the first group, and hence may be better suited for output direct to users. With --zero-terminated (-z), use a zero byte (ASCII NUL) instead of a newline as the delimiter.

Note that when groups are delimited and the input stream contains two or more consecutive blank lines, then the output is ambiguous. To avoid that, filter the input through ‘tr -s '\n'’ to replace each sequence of consecutive newlines with a single newline.

This is a GNU extension.

--group[=delimit-method]

Output all lines, and delimit each unique group. With --zero-terminated (-z), use a zero byte (ASCII NUL) instead of a newline as the delimiter. The optional delimit-method specifies how to delimit groups, and must be one of the following:

separate

Separate unique groups with a single delimiter. This is the default delimiting method if none is specified, and better suited for output direct to users.

prepend

Output a delimiter before each group of unique items.

append

Output a delimiter after each group of unique items.

both

Output a delimiter around each group of unique items.

Note that when groups are delimited and the input stream contains two or more consecutive blank lines, then the output is ambiguous. To avoid that, filter the input through ‘tr -s '\n'’ to replace each sequence of consecutive newlines with a single newline.

This is a GNU extension.

-u
--unique

Discard the last line that would be output for a repeated input group. When used by itself, this option causes uniq to print unique lines, and nothing else.

-w n
--check-chars=n

Compare at most n characters on each line (after skipping any specified fields and characters). By default the entire rest of the lines are compared.

-z
--zero-terminated

Delimit items with a zero byte rather than a newline (ASCII LF). I.e., treat input as items separated by ASCII NUL and terminate output items with ASCII NUL. This option can be useful in conjunction with ‘perl -0’ or ‘find -print0’ and ‘xargs -0’ which do the same in order to reliably handle arbitrary file names (even those containing blanks or other special characters).

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Operating on sorted files   [Contents][Index]

7.4 comm: Compare two sorted files line by line

comm writes to standard output lines that are common, and lines that are unique, to two input files; a file name of ‘-’ means standard input. Synopsis:

comm [option]… file1 file2

Before comm can be used, the input files must be sorted using the collating sequence specified by the LC_COLLATE locale. If an input file ends in a non-newline character, a newline is silently appended. The sort command with no options always outputs a file that is suitable input to comm.

With no options, comm produces three-column output. Column one contains lines unique to file1, column two contains lines unique to file2, and column three contains lines common to both files. Columns are separated by a single TAB character.

The options -1, -2, and -3 suppress printing of the corresponding columns (and separators). Also see Common options.

Unlike some other comparison utilities, comm has an exit status that does not depend on the result of the comparison. Upon normal completion comm produces an exit code of zero. If there is an error it exits with nonzero status.

If the --check-order option is given, unsorted inputs will cause a fatal error message. If the option --nocheck-order is given, unsorted inputs will never cause an error message. If neither of these options is given, wrongly sorted inputs are diagnosed only if an input file is found to contain unpairable lines. If an input file is diagnosed as being unsorted, the comm command will exit with a nonzero status (and the output should not be used).

Forcing comm to process wrongly sorted input files containing unpairable lines by specifying --nocheck-order is not guaranteed to produce any particular output. The output will probably not correspond with whatever you hoped it would be.

--check-order

Fail with an error message if either input file is wrongly ordered.

--nocheck-order

Do not check that both input files are in sorted order.

Other options are:

--output-delimiter=str

Print str between adjacent output columns, rather than the default of a single TAB character.

The delimiter str may not be empty.


Next: , Previous: , Up: Operating on sorted files   [Contents][Index]

7.5 ptx: Produce permuted indexes

ptx reads a text file and essentially produces a permuted index, with each keyword in its context. The calling sketch is either one of:

ptx [option …] [file …]
ptx -G [option …] [input [output]]

The -G (or its equivalent: --traditional) option disables all GNU extensions and reverts to traditional mode, thus introducing some limitations and changing several of the program’s default option values. When -G is not specified, GNU extensions are always enabled. GNU extensions to ptx are documented wherever appropriate in this document. See Compatibility in ptx, for the full list.

Individual options are explained in the following sections.

When GNU extensions are enabled, there may be zero, one or several files after the options. If there is no file, the program reads the standard input. If there is one or several files, they give the name of input files which are all read in turn, as if all the input files were concatenated. However, there is a full contextual break between each file and, when automatic referencing is requested, file names and line numbers refer to individual text input files. In all cases, the program outputs the permuted index to the standard output.

When GNU extensions are not enabled, that is, when the program operates in traditional mode, there may be zero, one or two parameters besides the options. If there are no parameters, the program reads the standard input and outputs the permuted index to the standard output. If there is only one parameter, it names the text input to be read instead of the standard input. If two parameters are given, they give respectively the name of the input file to read and the name of the output file to produce. Be very careful to note that, in this case, the contents of file given by the second parameter is destroyed. This behavior is dictated by System V ptx compatibility; GNU Standards normally discourage output parameters not introduced by an option.

Note that for any file named as the value of an option or as an input text file, a single dash ‘-’ may be used, in which case standard input is assumed. However, it would not make sense to use this convention more than once per program invocation.


Next: , Up: ptx invocation   [Contents][Index]

7.5.1 General options

-G
--traditional

As already explained, this option disables all GNU extensions to ptx and switches to traditional mode.

--help

Print a short help on standard output, then exit without further processing.

--version

Print the program version on standard output, then exit without further processing.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: ptx invocation   [Contents][Index]

7.5.2 Charset selection

As it is set up now, the program assumes that the input file is coded using 8-bit ISO 8859-1 code, also known as Latin-1 character set, unless it is compiled for MS-DOS, in which case it uses the character set of the IBM-PC. (GNU ptx is not known to work on smaller MS-DOS machines anymore.) Compared to 7-bit ASCII, the set of characters which are letters is different; this alters the behavior of regular expression matching. Thus, the default regular expression for a keyword allows foreign or diacriticized letters. Keyword sorting, however, is still crude; it obeys the underlying character set ordering quite blindly.

-f
--ignore-case

Fold lower case letters to upper case for sorting.


Next: , Previous: , Up: ptx invocation   [Contents][Index]

7.5.3 Word selection and input processing

-b file
--break-file=file

This option provides an alternative (to -W) method of describing which characters make up words. It introduces the name of a file which contains a list of characters which cannot be part of one word; this file is called the Break file. Any character which is not part of the Break file is a word constituent. If both options -b and -W are specified, then -W has precedence and -b is ignored.

When GNU extensions are enabled, the only way to avoid newline as a break character is to write all the break characters in the file with no newline at all, not even at the end of the file. When GNU extensions are disabled, spaces, tabs and newlines are always considered as break characters even if not included in the Break file.

-i file
--ignore-file=file

The file associated with this option contains a list of words which will never be taken as keywords in concordance output. It is called the Ignore file. The file contains exactly one word in each line; the end of line separation of words is not subject to the value of the -S option.

-o file
--only-file=file

The file associated with this option contains a list of words which will be retained in concordance output; any word not mentioned in this file is ignored. The file is called the Only file. The file contains exactly one word in each line; the end of line separation of words is not subject to the value of the -S option.

There is no default for the Only file. When both an Only file and an Ignore file are specified, a word is considered a keyword only if it is listed in the Only file and not in the Ignore file.

-r
--references

On each input line, the leading sequence of non-white space characters will be taken to be a reference that has the purpose of identifying this input line in the resulting permuted index. See Output formatting in ptx, for more information about reference production. Using this option changes the default value for option -S.

Using this option, the program does not try very hard to remove references from contexts in output, but it succeeds in doing so when the context ends exactly at the newline. If option -r is used with -S default value, or when GNU extensions are disabled, this condition is always met and references are completely excluded from the output contexts.

-S regexp
--sentence-regexp=regexp

This option selects which regular expression will describe the end of a line or the end of a sentence. In fact, this regular expression is not the only distinction between end of lines or end of sentences, and input line boundaries have no special significance outside this option. By default, when GNU extensions are enabled and if -r option is not used, end of sentences are used. In this case, this regex is imported from GNU Emacs:

[.?!][]\"')}]*\\($\\|\t\\|  \\)[ \t\n]*

Whenever GNU extensions are disabled or if -r option is used, end of lines are used; in this case, the default regexp is just:

\n

Using an empty regexp is equivalent to completely disabling end of line or end of sentence recognition. In this case, the whole file is considered to be a single big line or sentence. The user might want to disallow all truncation flag generation as well, through option -F "". See Syntax of Regular Expressions in The GNU Emacs Manual.

When the keywords happen to be near the beginning of the input line or sentence, this often creates an unused area at the beginning of the output context line; when the keywords happen to be near the end of the input line or sentence, this often creates an unused area at the end of the output context line. The program tries to fill those unused areas by wrapping around context in them; the tail of the input line or sentence is used to fill the unused area on the left of the output line; the head of the input line or sentence is used to fill the unused area on the right of the output line.

As a matter of convenience to the user, many usual backslashed escape sequences from the C language are recognized and converted to the corresponding characters by ptx itself.

-W regexp
--word-regexp=regexp

This option selects which regular expression will describe each keyword. By default, if GNU extensions are enabled, a word is a sequence of letters; the regexp used is ‘\w+’. When GNU extensions are disabled, a word is by default anything which ends with a space, a tab or a newline; the regexp used is ‘[^ \t\n]+’.

An empty regexp is equivalent to not using this option. See Syntax of Regular Expressions in The GNU Emacs Manual.

As a matter of convenience to the user, many usual backslashed escape sequences, as found in the C language, are recognized and converted to the corresponding characters by ptx itself.


Next: , Previous: , Up: ptx invocation   [Contents][Index]

7.5.4 Output formatting

Output format is mainly controlled by the -O and -T options described in the table below. When neither -O nor -T are selected, and if GNU extensions are enabled, the program chooses an output format suitable for a dumb terminal. Each keyword occurrence is output to the center of one line, surrounded by its left and right contexts. Each field is properly justified, so the concordance output can be readily observed. As a special feature, if automatic references are selected by option -A and are output before the left context, that is, if option -R is not selected, then a colon is added after the reference; this nicely interfaces with GNU Emacs next-error processing. In this default output format, each white space character, like newline and tab, is merely changed to exactly one space, with no special attempt to compress consecutive spaces. This might change in the future. Except for those white space characters, every other character of the underlying set of 256 characters is transmitted verbatim.

Output format is further controlled by the following options.

-g number
--gap-size=number

Select the size of the minimum white space gap between the fields on the output line.

-w number
--width=number

Select the maximum output width of each final line. If references are used, they are included or excluded from the maximum output width depending on the value of option -R. If this option is not selected, that is, when references are output before the left context, the maximum output width takes into account the maximum length of all references. If this option is selected, that is, when references are output after the right context, the maximum output width does not take into account the space taken by references, nor the gap that precedes them.

-A
--auto-reference

Select automatic references. Each input line will have an automatic reference made up of the file name and the line ordinal, with a single colon between them. However, the file name will be empty when standard input is being read. If both -A and -r are selected, then the input reference is still read and skipped, but the automatic reference is used at output time, overriding the input reference.

-R
--right-side-refs

In the default output format, when option -R is not used, any references produced by the effect of options -r or -A are placed to the far right of output lines, after the right context. With default output format, when the -R option is specified, references are rather placed at the beginning of each output line, before the left context. For any other output format, option -R is ignored, with one exception: with -R the width of references is not taken into account in total output width given by -w.

This option is automatically selected whenever GNU extensions are disabled.

-F string
--flac-truncation=string

This option will request that any truncation in the output be reported using the string string. Most output fields theoretically extend towards the beginning or the end of the current line, or current sentence, as selected with option -S. But there is a maximum allowed output line width, changeable through option -w, which is further divided into space for various output fields. When a field has to be truncated because it cannot extend beyond the beginning or the end of the current line to fit in, then a truncation occurs. By default, the string used is a single slash, as in -F /.

string may have more than one character, as in -F. Also, in the particular case when string is empty (-F ""), truncation flagging is disabled, and no truncation marks are appended in this case.

As a matter of convenience to the user, many usual backslashed escape sequences, as found in the C language, are recognized and converted to the corresponding characters by ptx itself.

-M string
--macro-name=string

Select another string to be used instead of ‘xx’, while generating output suitable for nroff, troff or TeX.

-O
--format=roff

Choose an output format suitable for nroff or troff processing. Each output line will look like:

.xx "tail" "before" "keyword_and_after" "head" "ref"

so it will be possible to write a ‘.xx’ roff macro to take care of the output typesetting. This is the default output format when GNU extensions are disabled. Option -M can be used to change ‘xx’ to another macro name.

In this output format, each non-graphical character, like newline and tab, is merely changed to exactly one space, with no special attempt to compress consecutive spaces. Each quote character ‘"’ is doubled so it will be correctly processed by nroff or troff.

-T
--format=tex

Choose an output format suitable for TeX processing. Each output line will look like:

\xx {tail}{before}{keyword}{after}{head}{ref}

so it will be possible to write a \xx definition to take care of the output typesetting. Note that when references are not being produced, that is, neither option -A nor option -r is selected, the last parameter of each \xx call is inhibited. Option -M can be used to change ‘xx’ to another macro name.

In this output format, some special characters, like ‘$’, ‘%’, ‘&’, ‘#’ and ‘_’ are automatically protected with a backslash. Curly brackets ‘{’, ‘}’ are protected with a backslash and a pair of dollar signs (to force mathematical mode). The backslash itself produces the sequence \backslash{}. Circumflex and tilde diacritical marks produce the sequence ^\{ } and ~\{ } respectively. Other diacriticized characters of the underlying character set produce an appropriate TeX sequence as far as possible. The other non-graphical characters, like newline and tab, and all other characters which are not part of ASCII, are merely changed to exactly one space, with no special attempt to compress consecutive spaces. Let me know how to improve this special character processing for TeX.


Previous: , Up: ptx invocation   [Contents][Index]

7.5.5 The GNU extensions to ptx

This version of ptx contains a few features which do not exist in System V ptx. These extra features are suppressed by using the -G command line option, unless overridden by other command line options. Some GNU extensions cannot be recovered by overriding, so the simple rule is to avoid -G if you care about GNU extensions. Here are the differences between this program and System V ptx.


Previous: , Up: Operating on sorted files   [Contents][Index]

7.6 tsort: Topological sort

tsort performs a topological sort on the given file, or standard input if no input file is given or for a file of ‘-’. For more details and some history, see tsort background. Synopsis:

tsort [option] [file]

tsort reads its input as pairs of strings, separated by blanks, indicating a partial ordering. The output is a total ordering that corresponds to the given partial ordering.

For example

tsort <<EOF
a b c
d
e f
b c d e
EOF

will produce the output

a
b
c
d
e
f

Consider a more realistic example. You have a large set of functions all in one file, and they may all be declared static except one. Currently that one (say main) is the first function defined in the file, and the ones it calls directly follow it, followed by those they call, etc. Let’s say that you are determined to take advantage of prototypes, so you have to choose between declaring all of those functions (which means duplicating a lot of information from the definitions) and rearranging the functions so that as many as possible are defined before they are used. One way to automate the latter process is to get a list for each function of the functions it calls directly. Many programs can generate such lists. They describe a call graph. Consider the following list, in which a given line indicates that the function on the left calls the one on the right directly.

main parse_options
main tail_file
main tail_forever
tail_file pretty_name
tail_file write_header
tail_file tail
tail_forever recheck
tail_forever pretty_name
tail_forever write_header
tail_forever dump_remainder
tail tail_lines
tail tail_bytes
tail_lines start_lines
tail_lines dump_remainder
tail_lines file_lines
tail_lines pipe_lines
tail_bytes xlseek
tail_bytes start_bytes
tail_bytes dump_remainder
tail_bytes pipe_bytes
file_lines dump_remainder
recheck pretty_name

then you can use tsort to produce an ordering of those functions that satisfies your requirement.

example$ tsort call-graph | tac
dump_remainder
start_lines
file_lines
pipe_lines
xlseek
start_bytes
pipe_bytes
tail_lines
tail_bytes
pretty_name
write_header
tail
recheck
parse_options
tail_file
tail_forever
main

tsort detects any cycles in the input and writes the first cycle encountered to standard error.

Note that for a given partial ordering, generally there is no unique total ordering. In the context of the call graph above, the function parse_options may be placed anywhere in the list as long as it precedes main.

The only options are --help and --version. See Common options.

An exit status of zero indicates success, and a nonzero value indicates failure.


Up: tsort invocation   [Contents][Index]

7.6.1 tsort: Background

tsort exists because very early versions of the Unix linker processed an archive file exactly once, and in order. As ld read each object in the archive, it decided whether it was needed in the program based on whether it defined any symbols which were undefined at that point in the link.

This meant that dependencies within the archive had to be handled specially. For example, scanf probably calls read. That means that in a single pass through an archive, it was important for scanf.o to appear before read.o, because otherwise a program which calls scanf but not read might end up with an unexpected unresolved reference to read.

The way to address this problem was to first generate a set of dependencies of one object file on another. This was done by a shell script called lorder. The GNU tools don’t provide a version of lorder, as far as I know, but you can still find it in BSD distributions.

Then you ran tsort over the lorder output, and you used the resulting sort to define the order in which you added objects to the archive.

This whole procedure has been obsolete since about 1980, because Unix archives now contain a symbol table (traditionally built by ranlib, now generally built by ar itself), and the Unix linker uses the symbol table to effectively make multiple passes over an archive file.

Anyhow, that’s where tsort came from. To solve an old problem with the way the linker handled archive files, which has since been solved in different ways.


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

8 Operating on fields


Next: , Up: Operating on fields   [Contents][Index]

8.1 cut: Print selected parts of lines

cut writes to standard output selected parts of each line of each input file, or standard input if no files are given or for a file name of ‘-’. Synopsis:

cut option… [file]…

In the table which follows, the byte-list, character-list, and field-list are one or more numbers or ranges (two numbers separated by a dash) separated by commas. Bytes, characters, and fields are numbered starting at 1. Incomplete ranges may be given: -m means ‘1-m’; ‘n-’ means ‘n’ through end of line or last field. The list elements can be repeated, can overlap, and can be specified in any order; but the selected input is written in the same order that it is read, and is written exactly once.

The program accepts the following options. Also see Common options.

-b byte-list
--bytes=byte-list

Select for printing only the bytes in positions listed in byte-list. Tabs and backspaces are treated like any other character; they take up 1 byte. If an output delimiter is specified, (see the description of --output-delimiter), then output that string between ranges of selected bytes.

-c character-list
--characters=character-list

Select for printing only the characters in positions listed in character-list. The same as -b for now, but internationalization will change that. Tabs and backspaces are treated like any other character; they take up 1 character. If an output delimiter is specified, (see the description of --output-delimiter), then output that string between ranges of selected bytes.

-f field-list
--fields=field-list

Select for printing only the fields listed in field-list. Fields are separated by a TAB character by default. Also print any line that contains no delimiter character, unless the --only-delimited (-s) option is specified.

Note awk supports more sophisticated field processing, and by default will use (and discard) runs of blank characters to separate fields, and ignore leading and trailing blanks.

awk '{print $2}'    # print the second field
awk '{print $NF-1}' # print the penultimate field
awk '{print $2,$1}' # reorder the first two fields

In the unlikely event that awk is unavailable, one can use the join command, to process blank characters as awk does above.

join -a1 -o 1.2     - /dev/null # print the second field
join -a1 -o 1.2,1.1 - /dev/null # reorder the first two fields
-d input_delim_byte
--delimiter=input_delim_byte

With -f, use the first byte of input_delim_byte as the input fields separator (default is TAB).

-n

Do not split multi-byte characters (no-op for now).

-s
--only-delimited

For -f, do not print lines that do not contain the field separator character. Normally, any line without a field separator is printed verbatim.

--output-delimiter=output_delim_string

With -f, output fields are separated by output_delim_string. The default with -f is to use the input delimiter. When using -b or -c to select ranges of byte or character offsets (as opposed to ranges of fields), output output_delim_string between non-overlapping ranges of selected bytes.

--complement

This option is a GNU extension. Select for printing the complement of the bytes, characters or fields selected with the -b, -c or -f options. In other words, do not print the bytes, characters or fields specified via those options. This option is useful when you have many fields and want to print all but a few of them.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Operating on fields   [Contents][Index]

8.2 paste: Merge lines of files

paste writes to standard output lines consisting of sequentially corresponding lines of each given file, separated by a TAB character. Standard input is used for a file name of ‘-’ or if no input files are given.

Synopsis:

paste [option]… [file]…

For example, with:

$ cat num2
1
2
$ cat let3
a
b
c

Take lines sequentially from each file:

$ paste num2 let3
1       a
2       b
        c

Duplicate lines from a file:

$ paste num2 let3 num2
1       a      1
2       b      2
        c

Intermix lines from stdin:

$ paste - let3 - < num2
1       a      2
        b
        c

Join consecutive lines with a space:

$ seq 4 | paste -d ' ' - -
1 2
3 4

The program accepts the following options. Also see Common options.

-s
--serial

Paste the lines of one file at a time rather than one line from each file. Using the above example data:

$ paste -s num2 let3
1       2
a       b       c
-d delim-list
--delimiters=delim-list

Consecutively use the characters in delim-list instead of TAB to separate merged lines. When delim-list is exhausted, start again at its beginning. Using the above example data:

$ paste -d '%_' num2 let3 num2
1%a_1
2%b_2
%c_

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Operating on fields   [Contents][Index]

8.3 join: Join lines on a common field

join writes to standard output a line for each pair of input lines that have identical join fields. Synopsis:

join [option]… file1 file2

Either file1 or file2 (but not both) can be ‘-’, meaning standard input. file1 and file2 should be sorted on the join fields.

Normally, the sort order is that of the collating sequence specified by the LC_COLLATE locale. Unless the -t option is given, the sort comparison ignores blanks at the start of the join field, as in sort -b. If the --ignore-case option is given, the sort comparison ignores the case of characters in the join field, as in sort -f.

The sort and join commands should use consistent locales and options if the output of sort is fed to join. You can use a command like ‘sort -k 1b,1’ to sort a file on its default join field, but if you select a non-default locale, join field, separator, or comparison options, then you should do so consistently between join and sort. If ‘join -t ''’ is specified then the whole line is considered which matches the default operation of sort.

If the input has no unpairable lines, a GNU extension is available; the sort order can be any order that considers two fields to be equal if and only if the sort comparison described above considers them to be equal. For example:

$ cat file1
a a1
c c1
b b1
$ cat file2
a a2
c c2
b b2
$ join file1 file2
a a1 a2
c c1 c2
b b1 b2

If the --check-order option is given, unsorted inputs will cause a fatal error message. If the option --nocheck-order is given, unsorted inputs will never cause an error message. If neither of these options is given, wrongly sorted inputs are diagnosed only if an input file is found to contain unpairable lines, and when both input files are non empty. If an input file is diagnosed as being unsorted, the join command will exit with a nonzero status (and the output should not be used).

Forcing join to process wrongly sorted input files containing unpairable lines by specifying --nocheck-order is not guaranteed to produce any particular output. The output will probably not correspond with whatever you hoped it would be.

The defaults are:

The program accepts the following options. Also see Common options.

-a file-number

Print a line for each unpairable line in file file-number (either ‘1’ or ‘2’), in addition to the normal output.

--check-order

Fail with an error message if either input file is wrongly ordered.

--nocheck-order

Do not check that both input files are in sorted order. This is the default.

-e string

Replace those output fields that are missing in the input with string. I.e., missing fields specified with the -12jo options.

--header

Treat the first line of each input file as a header line. The header lines will be joined and printed as the first output line. If -o is used to specify output format, the header line will be printed according to the specified format. The header lines will not be checked for ordering even if --check-order is specified. Also if the header lines from each file do not match, the heading fields from the first file will be used.

-i
--ignore-case

Ignore differences in case when comparing keys. With this option, the lines of the input files must be ordered in the same way. Use ‘sort -f’ to produce this ordering.

-1 field

Join on field field (a positive integer) of file 1.

-2 field

Join on field field (a positive integer) of file 2.

-j field

Equivalent to -1 field -2 field.

-o field-list
-o auto

If the keyword ‘auto’ is specified, infer the output format from the first line in each file. This is the same as the default output format but also ensures the same number of fields are output for each line. Missing fields are replaced with the -e option and extra fields are discarded.

Otherwise, construct each output line according to the format in field-list. Each element in field-list is either the single character ‘0’ or has the form m.n where the file number, m, is ‘1’ or ‘2’ and n is a positive field number.

A field specification of ‘0’ denotes the join field. In most cases, the functionality of the ‘0’ field spec may be reproduced using the explicit m.n that corresponds to the join field. However, when printing unpairable lines (using either of the -a or -v options), there is no way to specify the join field using m.n in field-list if there are unpairable lines in both files. To give join that functionality, POSIX invented the ‘0’ field specification notation.

The elements in field-list are separated by commas or blanks. Blank separators typically need to be quoted for the shell. For example, the commands ‘join -o 1.2,2.2’ and ‘join -o '1.2 2.2'’ are equivalent.

All output lines—including those printed because of any -a or -v option—are subject to the specified field-list.

-t char

Use character char as the input and output field separator. Treat as significant each occurrence of char in the input file. Use ‘sort -t char’, without the -b option of ‘sort’, to produce this ordering. If ‘join -t ''’ is specified, the whole line is considered, matching the default operation of sort. If ‘-t '\0'’ is specified then the ASCII NUL character is used to delimit the fields.

-v file-number

Print a line for each unpairable line in file file-number (either ‘1’ or ‘2’), instead of the normal output.

-z
--zero-terminated

Delimit items with a zero byte rather than a newline (ASCII LF). I.e., treat input as items separated by ASCII NUL and terminate output items with ASCII NUL. This option can be useful in conjunction with ‘perl -0’ or ‘find -print0’ and ‘xargs -0’ which do the same in order to reliably handle arbitrary file names (even those containing blanks or other special characters).

An exit status of zero indicates success, and a nonzero value indicates failure.


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

9 Operating on characters

These commands operate on individual characters.


Next: , Up: Operating on characters   [Contents][Index]

9.1 tr: Translate, squeeze, and/or delete characters

Synopsis:

tr [option]… set1 [set2]

tr copies standard input to standard output, performing one of the following operations:

The set1 and (if given) set2 arguments define ordered sets of characters, referred to below as set1 and set2. These sets are the characters of the input that tr operates on. The --complement (-c, -C) option replaces set1 with its complement (all of the characters that are not in set1).

Currently tr fully supports only single-byte characters. Eventually it will support multibyte characters; when it does, the -C option will cause it to complement the set of characters, whereas -c will cause it to complement the set of values. This distinction will matter only when some values are not characters, and this is possible only in locales using multibyte encodings when the input contains encoding errors.

The program accepts the --help and --version options. See Common options. Options must precede operands.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Up: tr invocation   [Contents][Index]

9.1.1 Specifying sets of characters

The format of the set1 and set2 arguments resembles the format of regular expressions; however, they are not regular expressions, only lists of characters. Most characters simply represent themselves in these strings, but the strings can contain the shorthands listed below, for convenience. Some of them can be used only in set1 or set2, as noted below.

Backslash escapes

The following backslash escape sequences are recognized:

\a

Control-G.

\b

Control-H.

\f

Control-L.

\n

Control-J.

\r

Control-M.

\t

Control-I.

\v

Control-K.

\ooo

The 8-bit character with the value given by ooo, which is 1 to 3 octal digits. Note that ‘\400’ is interpreted as the two-byte sequence, ‘\040’ ‘0’.

\\

A backslash.

While a backslash followed by a character not listed above is interpreted as that character, the backslash also effectively removes any special significance, so it is useful to escape ‘[’, ‘]’, ‘*’, and ‘-’.

Ranges

The notation ‘m-n’ expands to all of the characters from m through n, in ascending order. m should collate before n; if it doesn’t, an error results. As an example, ‘0-9’ is the same as ‘0123456789’.

GNU tr does not support the System V syntax that uses square brackets to enclose ranges. Translations specified in that format sometimes work as expected, since the brackets are often transliterated to themselves. However, they should be avoided because they sometimes behave unexpectedly. For example, ‘tr -d '[0-9]'’ deletes brackets as well as digits.

Many historically common and even accepted uses of ranges are not portable. For example, on EBCDIC hosts using the ‘A-Z’ range will not do what most would expect because ‘A’ through ‘Z’ are not contiguous as they are in ASCII. If you can rely on a POSIX compliant version of tr, then the best way to work around this is to use character classes (see below). Otherwise, it is most portable (and most ugly) to enumerate the members of the ranges.

Repeated characters

The notation ‘[c*n]’ in set2 expands to n copies of character c. Thus, ‘[y*6]’ is the same as ‘yyyyyy’. The notation ‘[c*]’ in string2 expands to as many copies of c as are needed to make set2 as long as set1. If n begins with ‘0’, it is interpreted in octal, otherwise in decimal.

Character classes

The notation ‘[:class:]’ expands to all of the characters in the (predefined) class class. The characters expand in no particular order, except for the upper and lower classes, which expand in ascending order. When the --delete (-d) and --squeeze-repeats (-s) options are both given, any character class can be used in set2. Otherwise, only the character classes lower and upper are accepted in set2, and then only if the corresponding character class (upper and lower, respectively) is specified in the same relative position in set1. Doing this specifies case conversion. The class names are given below; an error results when an invalid class name is given.

alnum

Letters and digits.

alpha

Letters.

blank

Horizontal whitespace.

cntrl

Control characters.

digit

Digits.

graph

Printable characters, not including space.

lower

Lowercase letters.

print

Printable characters, including space.

punct

Punctuation characters.

space

Horizontal or vertical whitespace.

upper

Uppercase letters.

xdigit

Hexadecimal digits.

Equivalence classes

The syntax ‘[=c=]’ expands to all of the characters that are equivalent to c, in no particular order. Equivalence classes are a relatively recent invention intended to support non-English alphabets. But there seems to be no standard way to define them or determine their contents. Therefore, they are not fully implemented in GNU tr; each character’s equivalence class consists only of that character, which is of no particular use.


Next: , Previous: , Up: tr invocation   [Contents][Index]

9.1.2 Translating

tr performs translation when set1 and set2 are both given and the --delete (-d) option is not given. tr translates each character of its input that is in set1 to the corresponding character in set2. Characters not in set1 are passed through unchanged. When a character appears more than once in set1 and the corresponding characters in set2 are not all the same, only the final one is used. For example, these two commands are equivalent:

tr aaa xyz
tr a z

A common use of tr is to convert lowercase characters to uppercase. This can be done in many ways. Here are three of them:

tr abcdefghijklmnopqrstuvwxyz ABCDEFGHIJKLMNOPQRSTUVWXYZ
tr a-z A-Z
tr '[:lower:]' '[:upper:]'

But note that using ranges like a-z above is not portable.

When tr is performing translation, set1 and set2 typically have the same length. If set1 is shorter than set2, the extra characters at the end of set2 are ignored.

On the other hand, making set1 longer than set2 is not portable; POSIX says that the result is undefined. In this situation, BSD tr pads set2 to the length of set1 by repeating the last character of set2 as many times as necessary. System V tr truncates set1 to the length of set2.

By default, GNU tr handles this case like BSD tr. When the --truncate-set1 (-t) option is given, GNU tr handles this case like the System V tr instead. This option is ignored for operations other than translation.

Acting like System V tr in this case breaks the relatively common BSD idiom:

tr -cs A-Za-z0-9 '\012'

because it converts only zero bytes (the first element in the complement of set1), rather than all non-alphanumerics, to newlines.

By the way, the above idiom is not portable because it uses ranges, and it assumes that the octal code for newline is 012. Assuming a POSIX compliant tr, here is a better way to write it:

tr -cs '[:alnum:]' '[\n*]'

Previous: , Up: tr invocation   [Contents][Index]

9.1.3 Squeezing repeats and deleting

When given just the --delete (-d) option, tr removes any input characters that are in set1.

When given just the --squeeze-repeats (-s) option, tr replaces each input sequence of a repeated character that is in set1 with a single occurrence of that character.

When given both --delete and --squeeze-repeats, tr first performs any deletions using set1, then squeezes repeats from any remaining characters using set2.

The --squeeze-repeats option may also be used when translating, in which case tr first performs translation, then squeezes repeats from any remaining characters using set2.

Here are some examples to illustrate various combinations of options:


Next: , Previous: , Up: Operating on characters   [Contents][Index]

9.2 expand: Convert tabs to spaces

expand writes the contents of each given file, or standard input if none are given or for a file of ‘-’, to standard output, with tab characters converted to the appropriate number of spaces. Synopsis:

expand [option]… [file]…

By default, expand converts all tabs to spaces. It preserves backspace characters in the output; they decrement the column count for tab calculations. The default action is equivalent to -t 8 (set tabs every 8 columns).

The program accepts the following options. Also see Common options.

-t tab1[,tab2]…
--tabs=tab1[,tab2]…

If only one tab stop is given, set the tabs tab1 spaces apart (default is 8). Otherwise, set the tabs at columns tab1, tab2, … (numbered from 0), and replace any tabs beyond the last tab stop given with single spaces. Tab stops can be separated by blanks as well as by commas.

For compatibility, GNU expand also accepts the obsolete option syntax, -t1[,t2]…. New scripts should use -t t1[,t2]… instead.

-i
--initial

Only convert initial tabs (those that precede all non-space or non-tab characters) on each line to spaces.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Operating on characters   [Contents][Index]

9.3 unexpand: Convert spaces to tabs

unexpand writes the contents of each given file, or standard input if none are given or for a file of ‘-’, to standard output, converting blanks at the beginning of each line into as many tab characters as needed. In the default POSIX locale, a blank is a space or a tab; other locales may specify additional blank characters. Synopsis:

unexpand [option]… [file]…

By default, unexpand converts only initial blanks (those that precede all non-blank characters) on each line. It preserves backspace characters in the output; they decrement the column count for tab calculations. By default, tabs are set at every 8th column.

The program accepts the following options. Also see Common options.

-t tab1[,tab2]…
--tabs=tab1[,tab2]…

If only one tab stop is given, set the tabs tab1 columns apart instead of the default 8. Otherwise, set the tabs at columns tab1, tab2, … (numbered from 0), and leave blanks beyond the tab stops given unchanged. Tab stops can be separated by blanks as well as by commas. This option implies the -a option.

For compatibility, GNU unexpand supports the obsolete option syntax, -tab1[,tab2]…, where tab stops must be separated by commas. (Unlike -t, this obsolete option does not imply -a.) New scripts should use --first-only -t tab1[,tab2]… instead.

-a
--all

Also convert all sequences of two or more blanks just before a tab stop, even if they occur after non-blank characters in a line.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

10 Directory listing

This chapter describes the ls command and its variants dir and vdir, which list information about files.


Next: , Up: Directory listing   [Contents][Index]

10.1 ls: List directory contents

The ls program lists information about files (of any type, including directories). Options and file arguments can be intermixed arbitrarily, as usual.

For non-option command-line arguments that are directories, by default ls lists the contents of directories, not recursively, and omitting files with names beginning with ‘.’. For other non-option arguments, by default ls lists just the file name. If no non-option argument is specified, ls operates on the current directory, acting as if it had been invoked with a single argument of ‘.’.

By default, the output is sorted alphabetically, according to the locale settings in effect.3 If standard output is a terminal, the output is in columns (sorted vertically) and control characters are output as question marks; otherwise, the output is listed one per line and control characters are output as-is.

Because ls is such a fundamental program, it has accumulated many options over the years. They are described in the subsections below; within each section, options are listed alphabetically (ignoring case). The division of options into the subsections is not absolute, since some options affect more than one aspect of ls’s operation.

Exit status:

0 success
1 minor problems  (e.g., failure to access a file or directory not
  specified as a command line argument.  This happens when listing a
  directory in which entries are actively being removed or renamed.)
2 serious trouble (e.g., memory exhausted, invalid option, failure
  to access a file or directory specified as a command line argument
  or a directory loop)

Also see Common options.


Next: , Up: ls invocation   [Contents][Index]

10.1.1 Which files are listed

These options determine which files ls lists information for. By default, ls lists files and the contents of any directories on the command line, except that in directories it ignores files whose names start with ‘.’.

-a
--all

In directories, do not ignore file names that start with ‘.’.

-A
--almost-all

In directories, do not ignore all file names that start with ‘.’; ignore only . and ... The --all (-a) option overrides this option.

-B
--ignore-backups

In directories, ignore files that end with ‘~’. This option is equivalent to ‘--ignore='*~' --ignore='.*~'’.

-d
--directory

List just the names of directories, as with other types of files, rather than listing their contents. Do not follow symbolic links listed on the command line unless the --dereference-command-line (-H), --dereference (-L), or --dereference-command-line-symlink-to-dir options are specified.

-H
--dereference-command-line

If a command line argument specifies a symbolic link, show information for the file the link references rather than for the link itself.

--dereference-command-line-symlink-to-dir

Do not dereference symbolic links, with one exception: if a command line argument specifies a symbolic link that refers to a directory, show information for that directory rather than for the link itself. This is the default behavior when no other dereferencing-related option has been specified (--classify (-F), --directory (-d), (-l), --dereference (-L), or --dereference-command-line (-H)).

--group-directories-first

Group all the directories before the files and then sort the directories and the files separately using the selected sort key (see –sort option). That is, this option specifies a primary sort key, and the –sort option specifies a secondary key. However, any use of --sort=none (-U) disables this option altogether.

--hide=PATTERN

In directories, ignore files whose names match the shell pattern pattern, unless the --all (-a) or --almost-all (-A) is also given. This option acts like --ignore=pattern except that it has no effect if --all (-a) or --almost-all (-A) is also given.

This option can be useful in shell aliases. For example, if lx is an alias for ‘ls --hide='*~'’ and ly is an alias for ‘ls --ignore='*~'’, then the command ‘lx -A’ lists the file README~ even though ‘ly -A’ would not.

-I pattern
--ignore=pattern

In directories, ignore files whose names match the shell pattern (not regular expression) pattern. As in the shell, an initial ‘.’ in a file name does not match a wildcard at the start of pattern. Sometimes it is useful to give this option several times. For example,

$ ls --ignore='.??*' --ignore='.[^.]' --ignore='#*'

The first option ignores names of length 3 or more that start with ‘.’, the second ignores all two-character names that start with ‘.’ except ‘..’, and the third ignores names that start with ‘#’.

-L
--dereference

When showing file information for a symbolic link, show information for the file the link references rather than the link itself. However, even with this option, ls still prints the name of the link itself, not the name of the file that the link points to.

-R
--recursive

List the contents of all directories recursively.


Next: , Previous: , Up: ls invocation   [Contents][Index]

10.1.2 What information is listed

These options affect the information that ls displays. By default, only file names are shown.

--author

List each file’s author when producing long format directory listings. In GNU/Hurd, file authors can differ from their owners, but in other operating systems the two are the same.

-D
--dired

With the long listing (-l) format, print an additional line after the main output:

//DIRED// beg1 end1 beg2 end2

The begn and endn are unsigned integers that record the byte position of the beginning and end of each file name in the output. This makes it easy for Emacs to find the names, even when they contain unusual characters such as space or newline, without fancy searching.

If directories are being listed recursively (-R), output a similar line with offsets for each subdirectory name:

//SUBDIRED// beg1 end1

Finally, output a line of the form:

//DIRED-OPTIONS// --quoting-style=word

where word is the quoting style (see Formatting the file names).

Here is an actual example:

$ mkdir -p a/sub/deeper a/sub2
$ touch a/f1 a/f2
$ touch a/sub/deeper/file
$ ls -gloRF --dired a
  a:
  total 8
  -rw-r--r-- 1    0 Jun 10 12:27 f1
  -rw-r--r-- 1    0 Jun 10 12:27 f2
  drwxr-xr-x 3 4096 Jun 10 12:27 sub/
  drwxr-xr-x 2 4096 Jun 10 12:27 sub2/

  a/sub:
  total 4
  drwxr-xr-x 2 4096 Jun 10 12:27 deeper/

  a/sub/deeper:
  total 0
  -rw-r--r-- 1 0 Jun 10 12:27 file

  a/sub2:
  total 0
//DIRED// 48 50 84 86 120 123 158 162 217 223 282 286
//SUBDIRED// 2 3 167 172 228 240 290 296
//DIRED-OPTIONS// --quoting-style=literal

Note that the pairs of offsets on the ‘//DIRED//’ line above delimit these names: f1, f2, sub, sub2, deeper, file. The offsets on the ‘//SUBDIRED//’ line delimit the following directory names: a, a/sub, a/sub/deeper, a/sub2.

Here is an example of how to extract the fifth entry name, ‘deeper’, corresponding to the pair of offsets, 222 and 228:

$ ls -gloRF --dired a > out
$ dd bs=1 skip=222 count=6 < out 2>/dev/null; echo
deeper

Note that although the listing above includes a trailing slash for the ‘deeper’ entry, the offsets select the name without the trailing slash. However, if you invoke ls with --dired along with an option like --escape (aka -b) and operate on a file whose name contains special characters, notice that the backslash is included:

$ touch 'a b'
$ ls -blog --dired 'a b'
  -rw-r--r-- 1 0 Jun 10 12:28 a\ b
//DIRED// 30 34
//DIRED-OPTIONS// --quoting-style=escape

If you use a quoting style that adds quote marks (e.g., --quoting-style=c), then the offsets include the quote marks. So beware that the user may select the quoting style via the environment variable QUOTING_STYLE. Hence, applications using --dired should either specify an explicit --quoting-style=literal option (aka -N or --literal) on the command line, or else be prepared to parse the escaped names.

--full-time

Produce long format directory listings, and list times in full. It is equivalent to using --format=long with --time-style=full-iso (see Formatting file timestamps).

-g

Produce long format directory listings, but don’t display owner information.

-G
--no-group

Inhibit display of group information in a long format directory listing. (This is the default in some non-GNU versions of ls, so we provide this option for compatibility.)

-h
--human-readable

Append a size letter to each size, such as ‘M’ for mebibytes. Powers of 1024 are used, not 1000; ‘M’ stands for 1,048,576 bytes. This option is equivalent to --block-size=human-readable. Use the --si option if you prefer powers of 1000.

-i
--inode

Print the inode number (also called the file serial number and index number) of each file to the left of the file name. (This number uniquely identifies each file within a particular file system.)

-l
--format=long
--format=verbose

In addition to the name of each file, print the file type, file mode bits, number of hard links, owner name, group name, size, and timestamp (see Formatting file timestamps), normally the modification time. Print question marks for information that cannot be determined.

Normally the size is printed as a byte count without punctuation, but this can be overridden (see Block size). For example, -h prints an abbreviated, human-readable count, and ‘--block-size="'1"’ prints a byte count with the thousands separator of the current locale.

For each directory that is listed, preface the files with a line ‘total blocks’, where blocks is the total disk allocation for all files in that directory. The block size currently defaults to 1024 bytes, but this can be overridden (see Block size). The blocks computed counts each hard link separately; this is arguably a deficiency.

The file type is one of the following characters:

-

regular file

b

block special file

c

character special file

C

high performance (“contiguous data”) file

d

directory

D

door (Solaris 2.5 and up)

l

symbolic link

M

off-line (“migrated”) file (Cray DMF)

n

network special file (HP-UX)

p

FIFO (named pipe)

P

port (Solaris 10 and up)

s

socket

?

some other file type

The file mode bits listed are similar to symbolic mode specifications (see Symbolic Modes). But ls combines multiple bits into the third character of each set of permissions as follows:

s

If the set-user-ID or set-group-ID bit and the corresponding executable bit are both set.

S

If the set-user-ID or set-group-ID bit is set but the corresponding executable bit is not set.

t

If the restricted deletion flag or sticky bit, and the other-executable bit, are both set. The restricted deletion flag is another name for the sticky bit. See Mode Structure.

T

If the restricted deletion flag or sticky bit is set but the other-executable bit is not set.

x

If the executable bit is set and none of the above apply.

-

Otherwise.

Following the file mode bits is a single character that specifies whether an alternate access method such as an access control list applies to the file. When the character following the file mode bits is a space, there is no alternate access method. When it is a printing character, then there is such a method.

GNU ls uses a ‘.’ character to indicate a file with a security context, but no other alternate access method.

A file with any other combination of alternate access methods is marked with a ‘+’ character.

-n
--numeric-uid-gid

Produce long format directory listings, but display numeric user and group IDs instead of the owner and group names.

-o

Produce long format directory listings, but don’t display group information. It is equivalent to using --format=long with --no-group .

-s
--size

Print the disk allocation of each file to the left of the file name. This is the amount of disk space used by the file, which is usually a bit more than the file’s size, but it can be less if the file has holes.

Normally the disk allocation is printed in units of 1024 bytes, but this can be overridden (see Block size).

For files that are NFS-mounted from an HP-UX system to a BSD system, this option reports sizes that are half the correct values. On HP-UX systems, it reports sizes that are twice the correct values for files that are NFS-mounted from BSD systems. This is due to a flaw in HP-UX; it also affects the HP-UX ls program.

--si

Append an SI-style abbreviation to each size, such as ‘M’ for megabytes. Powers of 1000 are used, not 1024; ‘M’ stands for 1,000,000 bytes. This option is equivalent to --block-size=si. Use the -h or --human-readable option if you prefer powers of 1024.

-Z
--context

Display the SELinux security context or ‘?’ if none is found. When used with the -l option, print the security context to the left of the size column.


Next: , Previous: , Up: ls invocation   [Contents][Index]

10.1.3 Sorting the output

These options change the order in which ls sorts the information it outputs. By default, sorting is done by character code (e.g., ASCII order).

-c
--time=ctime
--time=status

If the long listing format (e.g., -l, -o) is being used, print the status change time (the ‘ctime’ in the inode) instead of the modification time. When explicitly sorting by time (--sort=time or -t) or when not using a long listing format, sort according to the status change time.

-f

Primarily, like -U—do not sort; list the files in whatever order they are stored in the directory. But also enable -a (list all files) and disable -l, --color, and -s (if they were specified before the -f).

-r
--reverse

Reverse whatever the sorting method is—e.g., list files in reverse alphabetical order, youngest first, smallest first, or whatever.

-S
--sort=size

Sort by file size, largest first.

-t
--sort=time

Sort by modification time (the ‘mtime’ in the inode), newest first.

-u
--time=atime
--time=access
--time=use

If the long listing format (e.g., --format=long) is being used, print the last access time (the ‘atime’ in the inode). When explicitly sorting by time (--sort=time or -t) or when not using a long listing format, sort according to the access time.

-U
--sort=none

Do not sort; list the files in whatever order they are stored in the directory. (Do not do any of the other unrelated things that -f does.) This is especially useful when listing very large directories, since not doing any sorting can be noticeably faster.

-v
--sort=version

Sort by version name and number, lowest first. It behaves like a default sort, except that each sequence of decimal digits is treated numerically as an index/version number. (See Details about version sort.)

-X
--sort=extension

Sort directory contents alphabetically by file extension (characters after the last ‘.’); files with no extension are sorted first.


Next: , Previous: , Up: ls invocation   [Contents][Index]

10.1.4 Details about version sort

Version sorting handles the fact that file names frequently include indices or version numbers. Standard sorting usually does not produce the order that one expects because comparisons are made on a character-by-character basis. Version sorting is especially useful when browsing directories that contain many files with indices/version numbers in their names:

$ ls -1            $ ls -1v
abc.zml-1.gz       abc.zml-1.gz
abc.zml-12.gz      abc.zml-2.gz
abc.zml-2.gz       abc.zml-12.gz

Version-sorted strings are compared such that if ver1 and ver2 are version numbers and prefix and suffix (suffix matching the regular expression ‘(\.[A-Za-z~][A-Za-z0-9~]*)*’) are strings then ver1 < ver2 implies that the name composed of “prefix ver1 suffix” sorts before “prefix ver2 suffix”.

Note also that leading zeros of numeric parts are ignored:

$ ls -1            $ ls -1v
abc-1.007.tgz      abc-1.01a.tgz
abc-1.012b.tgz     abc-1.007.tgz
abc-1.01a.tgz      abc-1.012b.tgz

This functionality is implemented using gnulib’s filevercmp function, which has some caveats worth noting.


Next: , Previous: , Up: ls invocation   [Contents][Index]

10.1.5 General output formatting

These options affect the appearance of the overall output.

-1
--format=single-column

List one file per line. This is the default for ls when standard output is not a terminal. See also the -b and -q options to suppress direct output of newline characters within a file name.

-C
--format=vertical

List files in columns, sorted vertically. This is the default for ls if standard output is a terminal. It is always the default for the dir program. GNU ls uses variable width columns to display as many files as possible in the fewest lines.

--color [=when]

Specify whether to use color for distinguishing file types. when may be omitted, or one of:

Specifying --color and no when is equivalent to --color=always. Piping a colorized listing through a pager like more or less usually produces unreadable results. However, using more -f does seem to work.

Note that using the --color option may incur a noticeable performance penalty when run in a directory with very many entries, because the default settings require that ls stat every single file it lists. However, if you would like most of the file-type coloring but can live without the other coloring options (e.g., executable, orphan, sticky, other-writable, capability), use dircolors to set the LS_COLORS environment variable like this,

eval $(dircolors -p | perl -pe \
  's/^((CAP|S[ET]|O[TR]|M|E)\w+).*/$1 00/' | dircolors -)

and on a dirent.d_type-capable file system, ls will perform only one stat call per command line argument.

-F
--classify
--indicator-style=classify

Append a character to each file name indicating the file type. Also, for regular files that are executable, append ‘*’. The file type indicators are ‘/’ for directories, ‘@’ for symbolic links, ‘|’ for FIFOs, ‘=’ for sockets, ‘>’ for doors, and nothing for regular files. Do not follow symbolic links listed on the command line unless the --dereference-command-line (-H), --dereference (-L), or --dereference-command-line-symlink-to-dir options are specified.

--file-type
--indicator-style=file-type

Append a character to each file name indicating the file type. This is like -F, except that executables are not marked.

--indicator-style=word

Append a character indicator with style word to entry names, as follows:

none

Do not append any character indicator; this is the default.

slash

Append ‘/’ for directories. This is the same as the -p option.

file-type

Append ‘/’ for directories, ‘@’ for symbolic links, ‘|’ for FIFOs, ‘=’ for sockets, and nothing for regular files. This is the same as the --file-type option.

classify

Append ‘*’ for executable regular files, otherwise behave as for ‘file-type’. This is the same as the -F or --classify option.

-k
--kibibytes

Set the default block size to its normal value of 1024 bytes, overriding any contrary specification in environment variables (see Block size). This option is in turn overridden by the --block-size, -h or --human-readable, and --si options.

The -k or --kibibytes option affects the per-directory block count written by the -l and similar options, and the size written by the -s or --size option. It does not affect the file size written by -l.

-m
--format=commas

List files horizontally, with as many as will fit on each line, separated by ‘, ’ (a comma and a space).

-p
--indicator-style=slash

Append a ‘/’ to directory names.

-x
--format=across
--format=horizontal

List the files in columns, sorted horizontally.

-T cols
--tabsize=cols

Assume that each tab stop is cols columns wide. The default is 8. ls uses tabs where possible in the output, for efficiency. If cols is zero, do not use tabs at all.

Some terminal emulators might not properly align columns to the right of a TAB following a non-ASCII byte. You can avoid that issue by using the -T0 option or put TABSIZE=0 in your environment, to tell ls to align using spaces, not tabs.

-w
--width=cols

Assume the screen is cols columns wide. The default is taken from the terminal settings if possible; otherwise the environment variable COLUMNS is used if it is set; otherwise the default is 80.


Next: , Previous: , Up: ls invocation   [Contents][Index]

10.1.6 Formatting file timestamps

By default, file timestamps are listed in abbreviated form, using a date like ‘Mar 30  2002’ for non-recent timestamps, and a date-without-year and time like ‘Mar 30 23:45’ for recent timestamps. This format can change depending on the current locale as detailed below.

A timestamp is considered to be recent if it is less than six months old, and is not dated in the future. If a timestamp dated today is not listed in recent form, the timestamp is in the future, which means you probably have clock skew problems which may break programs like make that rely on file timestamps.

Time stamps are listed according to the time zone rules specified by the TZ environment variable, or by the system default rules if TZ is not set. See Specifying the Time Zone with TZ in The GNU C Library Reference Manual.

The following option changes how file timestamps are printed.

--time-style=style

List timestamps in style style. The style should be one of the following:

+format

List timestamps using format, where format is interpreted like the format argument of date (see date invocation). For example, --time-style="+%Y-%m-%d %H:%M:%S" causes ls to list timestamps like ‘2002-03-30 23:45:56’. As with date, format’s interpretation is affected by the LC_TIME locale category.

If format contains two format strings separated by a newline, the former is used for non-recent files and the latter for recent files; if you want output columns to line up, you may need to insert spaces in one of the two formats.

full-iso

List timestamps in full using ISO 8601 date, time, and time zone format with nanosecond precision, e.g., ‘2002-03-30 23:45:56.477817180 -0700’. This style is equivalent to ‘+%Y-%m-%d %H:%M:%S.%N %z’.

This is useful because the time output includes all the information that is available from the operating system. For example, this can help explain make’s behavior, since GNU make uses the full timestamp to determine whether a file is out of date.

long-iso

List ISO 8601 date and time in minutes, e.g., ‘2002-03-30 23:45’. These timestamps are shorter than ‘full-iso’ timestamps, and are usually good enough for everyday work. This style is equivalent to ‘+%Y-%m-%d %H:%M’.

iso

List ISO 8601 dates for non-recent timestamps (e.g., ‘2002-03-30 ’), and ISO 8601 month, day, hour, and minute for recent timestamps (e.g., ‘03-30 23:45’). These timestamps are uglier than ‘long-iso’ timestamps, but they carry nearly the same information in a smaller space and their brevity helps ls output fit within traditional 80-column output lines. The following two ls invocations are equivalent:

newline='
'
ls -l --time-style="+%Y-%m-%d $newline%m-%d %H:%M"
ls -l --time-style="iso"
locale

List timestamps in a locale-dependent form. For example, a Finnish locale might list non-recent timestamps like ‘maalis 30  2002’ and recent timestamps like ‘maalis 30 23:45’. Locale-dependent timestamps typically consume more space than ‘iso’ timestamps and are harder for programs to parse because locale conventions vary so widely, but they are easier for many people to read.

The LC_TIME locale category specifies the timestamp format. The default POSIX locale uses timestamps like ‘Mar 30  2002’ and ‘Mar 30 23:45’; in this locale, the following two ls invocations are equivalent:

newline='
'
ls -l --time-style="+%b %e  %Y$newline%b %e %H:%M"
ls -l --time-style="locale"

Other locales behave differently. For example, in a German locale, --time-style="locale" might be equivalent to --time-style="+%e. %b %Y $newline%e. %b %H:%M" and might generate timestamps like ‘30. Mär 2002 ’ and ‘30. Mär 23:45’.

posix-style

List POSIX-locale timestamps if the LC_TIME locale category is POSIX, style timestamps otherwise. For example, the ‘posix-long-iso’ style lists timestamps like ‘Mar 30  2002’ and ‘Mar 30 23:45’ when in the POSIX locale, and like ‘2002-03-30 23:45’ otherwise.

You can specify the default value of the --time-style option with the environment variable TIME_STYLE; if TIME_STYLE is not set the default style is ‘locale’. GNU Emacs 21.3 and later use the --dired option and therefore can parse any date format, but if you are using Emacs 21.1 or 21.2 and specify a non-POSIX locale you may need to set ‘TIME_STYLE="posix-long-iso"’.

To avoid certain denial-of-service attacks, timestamps that would be longer than 1000 bytes may be treated as errors.


Previous: , Up: ls invocation   [Contents][Index]

10.1.7 Formatting the file names

These options change how file names themselves are printed.

-b
--escape
--quoting-style=escape

Quote nongraphic characters in file names using alphabetic and octal backslash sequences like those used in C.

-N
--literal
--quoting-style=literal

Do not quote file names. However, with ls nongraphic characters are still printed as question marks if the output is a terminal and you do not specify the --show-control-chars option.

-q
--hide-control-chars

Print question marks instead of nongraphic characters in file names. This is the default if the output is a terminal and the program is ls.

-Q
--quote-name
--quoting-style=c

Enclose file names in double quotes and quote nongraphic characters as in C.

--quoting-style=word

Use style word to quote file names and other strings that may contain arbitrary characters. The word should be one of the following:

literal

Output strings as-is; this is the same as the -N or --literal option.

shell

Quote strings for the shell if they contain shell metacharacters or would cause ambiguous output. The quoting is suitable for POSIX-compatible shells like bash, but it does not always work for incompatible shells like csh.

shell-always

Quote strings for the shell, even if they would normally not require quoting.

c

Quote strings as for C character string literals, including the surrounding double-quote characters; this is the same as the -Q or --quote-name option.

escape

Quote strings as for C character string literals, except omit the surrounding double-quote characters; this is the same as the -b or --escape option.

clocale

Quote strings as for C character string literals, except use surrounding quotation marks appropriate for the locale.

locale

Quote strings as for C character string literals, except use surrounding quotation marks appropriate for the locale, and quote 'like this' instead of "like this" in the default C locale. This looks nicer on many displays.

You can specify the default value of the --quoting-style option with the environment variable QUOTING_STYLE. If that environment variable is not set, the default value is ‘literal’, but this default may change to ‘shell’ in a future version of this package.

--show-control-chars

Print nongraphic characters as-is in file names. This is the default unless the output is a terminal and the program is ls.


Next: , Previous: , Up: Directory listing   [Contents][Index]

10.2 dir: Briefly list directory contents

dir is equivalent to ls -C -b; that is, by default files are listed in columns, sorted vertically, and special characters are represented by backslash escape sequences.

See ls.


Next: , Previous: , Up: Directory listing   [Contents][Index]

10.3 vdir: Verbosely list directory contents

vdir is equivalent to ls -l -b; that is, by default files are listed in long format and special characters are represented by backslash escape sequences.

See ls.


Previous: , Up: Directory listing   [Contents][Index]

10.4 dircolors: Color setup for ls

dircolors outputs a sequence of shell commands to set up the terminal for color output from ls (and dir, etc.). Typical usage:

eval "$(dircolors [option]… [file])"

If file is specified, dircolors reads it to determine which colors to use for which file types and extensions. Otherwise, a precompiled database is used. For details on the format of these files, run ‘dircolors --print-database’.

To make dircolors read a ~/.dircolors file if it exists, you can put the following lines in your ~/.bashrc (or adapt them to your favorite shell):

d=.dircolors
test -r $d && eval "$(dircolors $d)"

The output is a shell command to set the LS_COLORS environment variable. You can specify the shell syntax to use on the command line, or dircolors will guess it from the value of the SHELL environment variable.

The program accepts the following options. Also see Common options.

-b
--sh
--bourne-shell

Output Bourne shell commands. This is the default if the SHELL environment variable is set and does not end with ‘csh’ or ‘tcsh’.

-c
--csh
--c-shell

Output C shell commands. This is the default if SHELL ends with csh or tcsh.

-p
--print-database

Print the (compiled-in) default color configuration database. This output is itself a valid configuration file, and is fairly descriptive of the possibilities.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

11 Basic operations

This chapter describes the commands for basic file manipulation: copying, moving (renaming), and deleting (removing).


Next: , Up: Basic operations   [Contents][Index]

11.1 cp: Copy files and directories

cp copies files (or, optionally, directories). The copy is completely independent of the original. You can either copy one file to another, or copy arbitrarily many files to a destination directory. Synopses:

cp [option]… [-T] source dest
cp [option]… sourcedirectory
cp [option]… -t directory source

Generally, files are written just as they are read. For exceptions, see the --sparse option below.

By default, cp does not copy directories. However, the -R, -a, and -r options cause cp to copy recursively by descending into source directories and copying files to corresponding destination directories.

When copying from a symbolic link, cp normally follows the link only when not copying recursively or when --link (-l) is used. This default can be overridden with the --archive (-a), -d, --dereference (-L), --no-dereference (-P), and -H options. If more than one of these options is specified, the last one silently overrides the others.

When copying to a symbolic link, cp follows the link only when it refers to an existing regular file. However, when copying to a dangling symbolic link, cp refuses by default, and fails with a diagnostic, since the operation is inherently dangerous. This behavior is contrary to historical practice and to POSIX. Set POSIXLY_CORRECT to make cp attempt to create the target of a dangling destination symlink, in spite of the possible risk. Also, when an option like --backup or --link acts to rename or remove the destination before copying, cp renames or removes the symbolic link rather than the file it points to.

By default, cp copies the contents of special files only when not copying recursively. This default can be overridden with the --copy-contents option.

cp generally refuses to copy a file onto itself, with the following exception: if --force --backup is specified with source and dest identical, and referring to a regular file, cp will make a backup file, either regular or numbered, as specified in the usual ways (see Backup options). This is useful when you simply want to make a backup of an existing file before changing it.

The program accepts the following options. Also see Common options.

-a
--archive

Preserve as much as possible of the structure and attributes of the original files in the copy (but do not attempt to preserve internal directory structure; i.e., ‘ls -U’ may list the entries in a copied directory in a different order). Try to preserve SELinux security context and extended attributes (xattr), but ignore any failure to do that and print no corresponding diagnostic. Equivalent to -dR --preserve=all with the reduced diagnostics.

--attributes-only

Copy only the specified attributes of the source file to the destination. If the destination already exists, do not alter its contents. See the --preserve option for controlling which attributes to copy.

-b
--backup[=method]

See Backup options. Make a backup of each file that would otherwise be overwritten or removed. As a special case, cp makes a backup of source when the force and backup options are given and source and dest are the same name for an existing, regular file. One useful application of this combination of options is this tiny Bourne shell script:

#!/bin/sh
# Usage: backup FILE...
# Create a GNU-style backup of each listed FILE.
fail=0
for i; do
  cp --backup --force --preserve=all -- "$i" "$i" || fail=1
done
exit $fail
--copy-contents

If copying recursively, copy the contents of any special files (e.g., FIFOs and device files) as if they were regular files. This means trying to read the data in each source file and writing it to the destination. It is usually a mistake to use this option, as it normally has undesirable effects on special files like FIFOs and the ones typically found in the /dev directory. In most cases, cp -R --copy-contents will hang indefinitely trying to read from FIFOs and special files like /dev/console, and it will fill up your destination disk if you use it to copy /dev/zero. This option has no effect unless copying recursively, and it does not affect the copying of symbolic links.

-d

Copy symbolic links as symbolic links rather than copying the files that they point to, and preserve hard links between source files in the copies. Equivalent to --no-dereference --preserve=links.

-f
--force

When copying without this option and an existing destination file cannot be opened for writing, the copy fails. However, with --force, when a destination file cannot be opened, cp then removes it and tries to open it again. Contrast this behavior with that enabled by --link and --symbolic-link, whereby the destination file is never opened but rather is removed unconditionally. Also see the description of --remove-destination.

This option is independent of the --interactive or -i option: neither cancels the effect of the other.

This option is ignored when the --no-clobber or -n option is also used.

-H

If a command line argument specifies a symbolic link, then copy the file it points to rather than the symbolic link itself. However, copy (preserving its nature) any symbolic link that is encountered via recursive traversal.

-i
--interactive

When copying a file other than a directory, prompt whether to overwrite an existing destination file. The -i option overrides a previous -n option.

-l
--link

Make hard links instead of copies of non-directories.

-L
--dereference

Follow symbolic links when copying from them. With this option, cp cannot create a symbolic link. For example, a symlink (to regular file) in the source tree will be copied to a regular file in the destination tree.

-n
--no-clobber

Do not overwrite an existing file. The -n option overrides a previous -i option. This option is mutually exclusive with -b or --backup option.

-P
--no-dereference

Copy symbolic links as symbolic links rather than copying the files that they point to. This option affects only symbolic links in the source; symbolic links in the destination are always followed if possible.

-p
--preserve[=attribute_list]

Preserve the specified attributes of the original files. If specified, the attribute_list must be a comma-separated list of one or more of the following strings:

mode

Preserve the file mode bits and access control lists.

ownership

Preserve the owner and group. On most modern systems, only users with appropriate privileges may change the owner of a file, and ordinary users may preserve the group ownership of a file only if they happen to be a member of the desired group.

timestamps

Preserve the times of last access and last modification, when possible. On older systems, it is not possible to preserve these attributes when the affected file is a symbolic link. However, many systems now provide the utimensat function, which makes it possible even for symbolic links.

links

Preserve in the destination files any links between corresponding source files. Note that with -L or -H, this option can convert symbolic links to hard links. For example,

$ mkdir c; : > a; ln -s a b; cp -aH a b c; ls -i1 c
74161745 a
74161745 b

Note the inputs: b is a symlink to regular file a, yet the files in destination directory, c/, are hard-linked. Since -a implies --no-dereference it would copy the symlink, but the later -H tells cp to dereference the command line arguments where it then sees two files with the same inode number. Then the --preserve=links option also implied by -a will preserve the perceived hard link.

Here is a similar example that exercises cp’s -L option:

$ mkdir b c; (cd b; : > a; ln -s a b); cp -aL b c; ls -i1 c/b
74163295 a
74163295 b
context

Preserve SELinux security context of the file, or fail with full diagnostics.

xattr

Preserve extended attributes of the file, or fail with full diagnostics. If cp is built without xattr support, ignore this option. If SELinux context, ACLs or Capabilities are implemented using xattrs, they are preserved implicitly by this option as well, i.e., even without specifying --preserve=mode or --preserve=context.

all

Preserve all file attributes. Equivalent to specifying all of the above, but with the difference that failure to preserve SELinux security context or extended attributes does not change cp’s exit status. In contrast to -a, all but ‘Operation not supported’ warnings are output.

Using --preserve with no attribute_list is equivalent to --preserve=mode,ownership,timestamps.

In the absence of this option, the permissions of existing destination files are unchanged. Each new file is created with the mode of the corresponding source file minus the set-user-ID, set-group-ID, and sticky bits as the create mode; the operating system then applies either the umask or a default ACL, possibly resulting in a more restrictive file mode. See File permissions.

--no-preserve=attribute_list

Do not preserve the specified attributes. The attribute_list has the same form as for --preserve.

--parents

Form the name of each destination file by appending to the target directory a slash and the specified name of the source file. The last argument given to cp must be the name of an existing directory. For example, the command:

cp --parents a/b/c existing_dir

copies the file a/b/c to existing_dir/a/b/c, creating any missing intermediate directories.

-R
-r
--recursive

Copy directories recursively. By default, do not follow symbolic links in the source unless used together with the --link (-l) option; see the --archive (-a), -d, --dereference (-L), --no-dereference (-P), and -H options. Special files are copied by creating a destination file of the same type as the source; see the --copy-contents option. It is not portable to use -r to copy symbolic links or special files. On some non-GNU systems, -r implies the equivalent of -L and --copy-contents for historical reasons. Also, it is not portable to use -R to copy symbolic links unless you also specify -P, as POSIX allows implementations that dereference symbolic links by default.

--reflink[=when]

Perform a lightweight, copy-on-write (COW) copy, if supported by the file system. Once it has succeeded, beware that the source and destination files share the same disk data blocks as long as they remain unmodified. Thus, if a disk I/O error affects data blocks of one of the files, the other suffers the same fate.

The when value can be one of the following:

always

The default behavior: if the copy-on-write operation is not supported then report the failure for each file and exit with a failure status.

auto

If the copy-on-write operation is not supported then fall back to the standard copy behavior.

This option is overridden by the --link, --symbolic-link and --attributes-only options, thus allowing it to be used to configure the default data copying behavior for cp. For example, with the following alias, cp will use the minimum amount of space supported by the file system.

alias cp='cp --reflink=auto --sparse=always'
--remove-destination

Remove each existing destination file before attempting to open it (contrast with -f above).

--sparse=when

A sparse file contains holes—a sequence of zero bytes that does not occupy any physical disk blocks; the ‘read’ system call reads these as zeros. This can both save considerable disk space and increase speed, since many binary files contain lots of consecutive zero bytes. By default, cp detects holes in input source files via a crude heuristic and makes the corresponding output file sparse as well. Only regular files may be sparse.

The when value can be one of the following:

auto

The default behavior: if the input file is sparse, attempt to make the output file sparse, too. However, if an output file exists but refers to a non-regular file, then do not attempt to make it sparse.

always

For each sufficiently long sequence of zero bytes in the input file, attempt to create a corresponding hole in the output file, even if the input file does not appear to be sparse. This is useful when the input file resides on a file system that does not support sparse files (for example, ‘efs’ file systems in SGI IRIX 5.3 and earlier), but the output file is on a type of file system that does support them. Holes may be created only in regular files, so if the destination file is of some other type, cp does not even try to make it sparse.

never

Never make the output file sparse. This is useful in creating a file for use with the mkswap command, since such a file must not have any holes.

--strip-trailing-slashes

Remove any trailing slashes from each source argument. See Trailing slashes.

-s
--symbolic-link

Make symbolic links instead of copies of non-directories. All source file names must be absolute (starting with ‘/’) unless the destination files are in the current directory. This option merely results in an error message on systems that do not support symbolic links.

-S suffix
--suffix=suffix

Append suffix to each backup file made with -b. See Backup options.

-t directory
--target-directory=directory

Specify the destination directory. See Target directory.

-T
--no-target-directory

Do not treat the last operand specially when it is a directory or a symbolic link to a directory. See Target directory.

-u
--update

Do not copy a non-directory that has an existing destination with the same or newer modification time. If time stamps are being preserved, the comparison is to the source time stamp truncated to the resolutions of the destination file system and of the system calls used to update time stamps; this avoids duplicate work if several ‘cp -pu’ commands are executed with the same source and destination. If --preserve=links is also specified (like with ‘cp -au’ for example), that will take precedence. Consequently, depending on the order that files are processed from the source, newer files in the destination may be replaced, to mirror hard links in the source.

-v
--verbose

Print the name of each file before copying it.

-x
--one-file-system

Skip subdirectories that are on different file systems from the one that the copy started on. However, mount point directories are copied.

-Z
--context[=context]

Without a specified context, adjust the SELinux security context according to the system default type for destination files, similarly to the restorecon command. The long form of this option with a specific context specified, will set the context for newly created files only. With a specified context, if both SELinux and SMACK are disabled, a warning is issued. This option is mutually exclusive with the --preserve=context option, and overrides the --preserve=all and -a options.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Basic operations   [Contents][Index]

11.2 dd: Convert and copy a file

dd copies a file (from standard input to standard output, by default) with a changeable I/O block size, while optionally performing conversions on it. Synopses:

dd [operand]…
dd option

The only options are --help and --version. See Common options. dd accepts the following operands, whose syntax was inspired by the DD (data definition) statement of OS/360 JCL.

if=file

Read from file instead of standard input.

of=file

Write to file instead of standard output. Unless ‘conv=notrunc’ is given, dd truncates file to zero bytes (or the size specified with ‘seek=’).

ibs=bytes

Set the input block size to bytes. This makes dd read bytes per block. The default is 512 bytes.

obs=bytes

Set the output block size to bytes. This makes dd write bytes per block. The default is 512 bytes.

bs=bytes

Set both input and output block sizes to bytes. This makes dd read and write bytes per block, overriding any ‘ibs’ and ‘obs’ settings. In addition, if no data-transforming conv option is specified, input is copied to the output as soon as it’s read, even if it is smaller than the block size.

cbs=bytes

Set the conversion block size to bytes. When converting variable-length records to fixed-length ones (conv=block) or the reverse (conv=unblock), use bytes as the fixed record length.

skip=n

Skip nibs’-byte blocks in the input file before copying. If ‘iflag=skip_bytes’ is specified, n is interpreted as a byte count rather than a block count.

seek=n

Skip nobs’-byte blocks in the output file before copying. if ‘oflag=seek_bytes’ is specified, n is interpreted as a byte count rather than a block count.

count=n

Copy nibs’-byte blocks from the input file, instead of everything until the end of the file. if ‘iflag=count_bytes’ is specified, n is interpreted as a byte count rather than a block count. Note if the input may return short reads as could be the case when reading from a pipe for example, ‘iflag=fullblock’ will ensure that ‘count=’ corresponds to complete input blocks rather than the traditional POSIX specified behavior of counting input read operations.

status=level

Transfer information is normally output to stderr upon receipt of the ‘INFO’ signal or when dd exits. Specifying level will adjust the amount of information printed, with the last level specified taking precedence.

none

Do not print any informational or warning messages to stderr. Error messages are output as normal.

noxfer

Do not print the final transfer rate and volume statistics that normally make up the last status line.

progress

Print the transfer rate and volume statistics on stderr, when processing each input block. Statistics are output on a single line at most once every second, but updates can be delayed when waiting on I/O.

conv=conversion[,conversion]…

Convert the file as specified by the conversion argument(s). (No spaces around any comma(s).)

Conversions:

ascii

Convert EBCDIC to ASCII, using the conversion table specified by POSIX. This provides a 1:1 translation for all 256 bytes. This option implies ‘conv=unblock’; input is converted to ASCII before trailing spaces are deleted.

ebcdic

Convert ASCII to EBCDIC. This is the inverse of the ‘ascii’ conversion. This option implies ‘conv=block’; trailing spaces are added before being converted to EBCDIC.

ibm

This acts like ‘conv=ebcdic’, except it uses the alternate conversion table specified by POSIX. This is not a 1:1 translation, but reflects common historical practice for ‘~’, ‘[’, and ‘]’.

The ‘ascii’, ‘ebcdic’, and ‘ibm’ conversions are mutually exclusive. If you use any of these options, you should also use the ‘cbs=’ option.

block

For each line in the input, output ‘cbs’ bytes, replacing the input newline with a space and padding with spaces as necessary.

unblock

Remove any trailing spaces in each ‘cbs’-sized input block, and append a newline.

The ‘block’ and ‘unblock’ conversions are mutually exclusive.

lcase

Change uppercase letters to lowercase.

ucase

Change lowercase letters to uppercase.

The ‘lcase’ and ‘ucase’ conversions are mutually exclusive.

sparse

Try to seek rather than write NUL output blocks. On a file system that supports sparse files, this will create sparse output when extending the output file. Be careful when using this option in conjunction with ‘conv=notrunc’ or ‘oflag=append’. With ‘conv=notrunc’, existing data in the output file corresponding to NUL blocks from the input, will be untouched. With ‘oflag=append’ the seeks performed will be ineffective. Similarly, when the output is a device rather than a file, NUL input blocks are not copied, and therefore this option is most useful with virtual or pre zeroed devices.

swab

Swap every pair of input bytes. GNU dd, unlike others, works when an odd number of bytes are read—the last byte is simply copied (since there is nothing to swap it with).

sync

Pad every input block to size of ‘ibs’ with trailing zero bytes. When used with ‘block’ or ‘unblock’, pad with spaces instead of zero bytes.

The following “conversions” are really file flags and don’t affect internal processing:

excl

Fail if the output file already exists; dd must create the output file itself.

nocreat

Do not create the output file; the output file must already exist.

The ‘excl’ and ‘nocreat’ conversions are mutually exclusive.

notrunc

Do not truncate the output file.

noerror

Continue after read errors.

fdatasync

Synchronize output data just before finishing. This forces a physical write of output data.

fsync

Synchronize output data and metadata just before finishing. This forces a physical write of output data and metadata.

iflag=flag[,flag]…

Access the input file using the flags specified by the flag argument(s). (No spaces around any comma(s).)

oflag=flag[,flag]…

Access the output file using the flags specified by the flag argument(s). (No spaces around any comma(s).)

Here are the flags. Not every flag is supported on every operating system.

append

Write in append mode, so that even if some other process is writing to this file, every dd write will append to the current contents of the file. This flag makes sense only for output. If you combine this flag with the ‘of=file’ operand, you should also specify ‘conv=notrunc’ unless you want the output file to be truncated before being appended to.

cio

Use concurrent I/O mode for data. This mode performs direct I/O and drops the POSIX requirement to serialize all I/O to the same file. A file cannot be opened in CIO mode and with a standard open at the same time.

direct

Use direct I/O for data, avoiding the buffer cache. Note that the kernel may impose restrictions on read or write buffer sizes. For example, with an ext4 destination file system and a Linux-based kernel, using ‘oflag=direct’ will cause writes to fail with EINVAL if the output buffer size is not a multiple of 512.

directory

Fail unless the file is a directory. Most operating systems do not allow I/O to a directory, so this flag has limited utility.

dsync

Use synchronized I/O for data. For the output file, this forces a physical write of output data on each write. For the input file, this flag can matter when reading from a remote file that has been written to synchronously by some other process. Metadata (e.g., last-access and last-modified time) is not necessarily synchronized.

sync

Use synchronized I/O for both data and metadata.

nocache

Discard the data cache for a file. When count=0 all cache is discarded, otherwise the cache is dropped for the processed portion of the file. Also when count=0 failure to discard the cache is diagnosed and reflected in the exit status. Here as some usage examples:

# Advise to drop cache for whole file
dd if=ifile iflag=nocache count=0

# Ensure drop cache for the whole file
dd of=ofile oflag=nocache conv=notrunc,fdatasync count=0

# Drop cache for part of file
dd if=ifile iflag=nocache skip=10 count=10 of=/dev/null

# Stream data using just the read-ahead cache
dd if=ifile of=ofile iflag=nocache oflag=nocache
nonblock

Use non-blocking I/O.

noatime

Do not update the file’s access time. Some older file systems silently ignore this flag, so it is a good idea to test it on your files before relying on it.

noctty

Do not assign the file to be a controlling terminal for dd. This has no effect when the file is not a terminal. On many hosts (e.g., GNU/Linux hosts), this option has no effect at all.

nofollow

Do not follow symbolic links.

nolinks

Fail if the file has multiple hard links.

binary

Use binary I/O. This option has an effect only on nonstandard platforms that distinguish binary from text I/O.

text

Use text I/O. Like ‘binary’, this option has no effect on standard platforms.

fullblock

Accumulate full blocks from input. The read system call may return early if a full block is not available. When that happens, continue calling read to fill the remainder of the block. This flag can be used only with iflag. This flag is useful with pipes for example as they may return short reads. In that case, this flag is needed to ensure that a ‘count=’ argument is interpreted as a block count rather than a count of read operations.

count_bytes

Interpret the ‘count=’ operand as a byte count, rather than a block count, which allows specifying a length that is not a multiple of the I/O block size. This flag can be used only with iflag.

skip_bytes

Interpret the ‘skip=’ operand as a byte count, rather than a block count, which allows specifying an offset that is not a multiple of the I/O block size. This flag can be used only with iflag.

seek_bytes

Interpret the ‘seek=’ operand as a byte count, rather than a block count, which allows specifying an offset that is not a multiple of the I/O block size. This flag can be used only with oflag.

These flags are not supported on all systems, and ‘dd’ rejects attempts to use them when they are not supported. When reading from standard input or writing to standard output, the ‘nofollow’ and ‘noctty’ flags should not be specified, and the other flags (e.g., ‘nonblock’) can affect how other processes behave with the affected file descriptors, even after dd exits.

The numeric-valued strings above (n and bytes) can be followed by a multiplier: ‘b’=512, ‘c’=1, ‘w’=2, ‘xm’=m, or any of the standard block size suffixes like ‘k’=1024 (see Block size).

Any block size you specify via ‘bs=’, ‘ibs=’, ‘obs=’, ‘cbs=’ should not be too large—values larger than a few megabytes are generally wasteful or (as in the gigabyte..exabyte case) downright counterproductive or error-inducing.

To process data that is at an offset or size that is not a multiple of the I/O block size, you can use the ‘skip_bytes’, ‘seek_bytes’ and ‘count_bytes’ flags. Alternatively the traditional method of separate dd invocations can be used. For example, the following shell commands copy data in 512 KiB blocks between a disk and a tape, but do not save or restore a 4 KiB label at the start of the disk:

disk=/dev/rdsk/c0t1d0s2
tape=/dev/rmt/0

# Copy all but the label from disk to tape.
(dd bs=4k skip=1 count=0 && dd bs=512k) <$disk >$tape

# Copy from tape back to disk, but leave the disk label alone.
(dd bs=4k seek=1 count=0 && dd bs=512k) <$tape >$disk

For failing disks, other tools come with a great variety of extra functionality to ease the saving of as much data as possible before the disk finally dies, e.g. GNU ddrescue. However, in some cases such a tool is not available or the administrator feels more comfortable with the handling of dd. As a simple rescue method, call dd as shown in the following example: the options ‘conv=noerror,sync’ are used to continue after read errors and to pad out bad reads with NULs, while ‘iflag=fullblock’ caters for short reads (which traditionally never occur on disk based devices):

# Rescue data from an (unmounted!) partition of a failing disk.
dd conv=noerror,sync iflag=fullblock </dev/sda1 > /mnt/rescue.img

Sending an ‘INFO’ signal (or ‘USR1’ signal where that is unavailable) to a running dd process makes it print I/O statistics to standard error and then resume copying. In the example below, dd is run in the background to copy 5GB of data. The kill command makes it output intermediate I/O statistics, and when dd completes normally or is killed by the SIGINT signal, it outputs the final statistics.

# Ignore the signal so we never inadvertently terminate the dd child.
# Note this is not needed when SIGINFO is available.
trap '' USR1

# Run dd with the fullblock iflag to avoid short reads
# which can be triggered by reception of signals.
dd iflag=fullblock if=/dev/zero of=/dev/null count=5000000 bs=1000 & pid=$!

# Output stats every half second
until ! kill -s USR1 $pid 2>/dev/null; do sleep .5; done

The above script will output in the following format

859+0 records in
859+0 records out
4295000000 bytes (4.3 GB) copied, 0.539934 s, 8.0 GB/s
1000+0 records in
1000+0 records out
5000000000 bytes (5.0 GB) copied, 0.630785 s, 7.9 GB/s

Note also the ‘status=progress’ option which periodically updates the last line of the transfer statistics above.

On systems lacking the ‘INFO’ signal dd responds to the ‘USR1’ signal instead, unless the POSIXLY_CORRECT environment variable is set.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Basic operations   [Contents][Index]

11.3 install: Copy files and set attributes

install copies files while setting their file mode bits and, if possible, their owner and group. Synopses:

install [option]… [-T] source dest
install [option]… sourcedirectory
install [option]… -t directory source…
install [option]… -d directory

install is similar to cp, but allows you to control the attributes of destination files. It is typically used in Makefiles to copy programs into their destination directories. It refuses to copy files onto themselves.

install never preserves extended attributes (xattr).

The program accepts the following options. Also see Common options.

-b
--backup[=method]

See Backup options. Make a backup of each file that would otherwise be overwritten or removed.

-C
--compare

Compare each pair of source and destination files, and if the destination has identical content and any specified owner, group, permissions, and possibly SELinux context, then do not modify the destination at all. Note this option is best used in conjunction with --user, --group and --mode options, lest install incorrectly determines the default attributes that installed files would have (as it doesn’t consider setgid directories and POSIX default ACLs for example). This could result in redundant copies or attributes that are not reset to the correct defaults.

-c

Ignored; for compatibility with old Unix versions of install.

-D

Create any missing parent directories of dest, then copy source to dest. Explicitly specifying the --target-directory=dir will similarly ensure the presence of that hierarchy before copying source arguments.

-d
--directory

Create any missing parent directories, giving them the default attributes. Then create each given directory, setting their owner, group and mode as given on the command line or to the defaults.

-g group
--group=group

Set the group ownership of installed files or directories to group. The default is the process’s current group. group may be either a group name or a numeric group ID.

-m mode
--mode=mode

Set the file mode bits for the installed file or directory to mode, which can be either an octal number, or a symbolic mode as in chmod, with ‘a=’ (no access allowed to anyone) as the point of departure (see File permissions). The default mode is ‘u=rwx,go=rx,a-s’—read, write, and execute for the owner, read and execute for group and other, and with set-user-ID and set-group-ID disabled. This default is not quite the same as ‘755’, since it disables instead of preserving set-user-ID and set-group-ID on directories. See Directory Setuid and Setgid.

-o owner
--owner=owner

If install has appropriate privileges (is run as root), set the ownership of installed files or directories to owner. The default is root. owner may be either a user name or a numeric user ID.

--preserve-context

Preserve the SELinux security context of files and directories. Failure to preserve the context in all of the files or directories will result in an exit status of 1. If SELinux is disabled then print a warning and ignore the option.

-p
--preserve-timestamps

Set the time of last access and the time of last modification of each installed file to match those of each corresponding original file. When a file is installed without this option, its last access and last modification times are both set to the time of installation. This option is useful if you want to use the last modification times of installed files to keep track of when they were last built as opposed to when they were last installed.

-s
--strip

Strip the symbol tables from installed binary executables.

--strip-program=program

Program used to strip binaries.

-S suffix
--suffix=suffix

Append suffix to each backup file made with -b. See Backup options.

-t directory
--target-directory=directory

Specify the destination directory. See Target directory.

-T
--no-target-directory

Do not treat the last operand specially when it is a directory or a symbolic link to a directory. See Target directory.

-v
--verbose

Print the name of each file before copying it.

-Z
--context[=context]

Without a specified context, adjust the SELinux security context according to the system default type for destination files, similarly to the restorecon command. The long form of this option with a specific context specified, will set the context for newly created files only. With a specified context, if both SELinux and SMACK are disabled, a warning is issued. This option is mutually exclusive with the --preserve-context option.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Basic operations   [Contents][Index]

11.4 mv: Move (rename) files

mv moves or renames files (or directories). Synopses:

mv [option]… [-T] source dest
mv [option]… sourcedirectory
mv [option]… -t directory source

mv can move any type of file from one file system to another. Prior to version 4.0 of the fileutils, mv could move only regular files between file systems. For example, now mv can move an entire directory hierarchy including special device files from one partition to another. It first uses some of the same code that’s used by cp -a to copy the requested directories and files, then (assuming the copy succeeded) it removes the originals. If the copy fails, then the part that was copied to the destination partition is removed. If you were to copy three directories from one partition to another and the copy of the first directory succeeded, but the second didn’t, the first would be left on the destination partition and the second and third would be left on the original partition.

mv always tries to copy extended attributes (xattr), which may include SELinux context, ACLs or Capabilities. Upon failure all but ‘Operation not supported’ warnings are output.

If a destination file exists but is normally unwritable, standard input is a terminal, and the -f or --force option is not given, mv prompts the user for whether to replace the file. (You might own the file, or have write permission on its directory.) If the response is not affirmative, the file is skipped.

Warning: Avoid specifying a source name with a trailing slash, when it might be a symlink to a directory. Otherwise, mv may do something very surprising, since its behavior depends on the underlying rename system call. On a system with a modern Linux-based kernel, it fails with errno=ENOTDIR. However, on other systems (at least FreeBSD 6.1 and Solaris 10) it silently renames not the symlink but rather the directory referenced by the symlink. See Trailing slashes.

The program accepts the following options. Also see Common options.

-b
--backup[=method]

See Backup options. Make a backup of each file that would otherwise be overwritten or removed.

-f
--force

Do not prompt the user before removing a destination file. If you specify more than one of the -i, -f, -n options, only the final one takes effect.

-i
--interactive

Prompt whether to overwrite each existing destination file, regardless of its permissions. If the response is not affirmative, the file is skipped. If you specify more than one of the -i, -f, -n options, only the final one takes effect.

-n
--no-clobber

Do not overwrite an existing file. If you specify more than one of the -i, -f, -n options, only the final one takes effect. This option is mutually exclusive with -b or --backup option.

-u
--update

Do not move a non-directory that has an existing destination with the same or newer modification time. If the move is across file system boundaries, the comparison is to the source time stamp truncated to the resolutions of the destination file system and of the system calls used to update time stamps; this avoids duplicate work if several ‘mv -u’ commands are executed with the same source and destination.

-v
--verbose

Print the name of each file before moving it.

--strip-trailing-slashes

Remove any trailing slashes from each source argument. See Trailing slashes.

-S suffix
--suffix=suffix

Append suffix to each backup file made with -b. See Backup options.

-t directory
--target-directory=directory

Specify the destination directory. See Target directory.

-T
--no-target-directory

Do not treat the last operand specially when it is a directory or a symbolic link to a directory. See Target directory.

-Z
--context

This option functions similarly to the restorecon command, by adjusting the SELinux security context according to the system default type for destination files.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Basic operations   [Contents][Index]

11.5 rm: Remove files or directories

rm removes each given file. By default, it does not remove directories. Synopsis:

rm [option]… [file]…

If the -I or --interactive=once option is given, and there are more than three files or the -r, -R, or --recursive are given, then rm prompts the user for whether to proceed with the entire operation. If the response is not affirmative, the entire command is aborted.

Otherwise, if a file is unwritable, standard input is a terminal, and the -f or --force option is not given, or the -i or --interactive=always option is given, rm prompts the user for whether to remove the file. If the response is not affirmative, the file is skipped.

Any attempt to remove a file whose last file name component is . or .. is rejected without any prompting, as mandated by POSIX.

Warning: If you use rm to remove a file, it is usually possible to recover the contents of that file. If you want more assurance that the contents are truly unrecoverable, consider using shred.

The program accepts the following options. Also see Common options.

-d
--dir

Remove the listed directories if they are empty.

-f
--force

Ignore nonexistent files and missing operands, and never prompt the user. Ignore any previous --interactive (-i) option.

-i

Prompt whether to remove each file. If the response is not affirmative, the file is skipped. Ignore any previous --force (-f) option. Equivalent to --interactive=always.

-I

Prompt once whether to proceed with the command, if more than three files are named or if a recursive removal is requested. Ignore any previous --force (-f) option. Equivalent to --interactive=once.

--interactive [=when]

Specify when to issue an interactive prompt. when may be omitted, or one of:

--interactive with no when is equivalent to --interactive=always.

--one-file-system

When removing a hierarchy recursively, skip any directory that is on a file system different from that of the corresponding command line argument.

This option is useful when removing a build “chroot” hierarchy, which normally contains no valuable data. However, it is not uncommon to bind-mount /home into such a hierarchy, to make it easier to use one’s start-up file. The catch is that it’s easy to forget to unmount /home. Then, when you use rm -rf to remove your normally throw-away chroot, that command will remove everything under /home, too. Use the --one-file-system option, and it will warn about and skip directories on other file systems. Of course, this will not save your /home if it and your chroot happen to be on the same file system.

--preserve-root

Fail upon any attempt to remove the root directory, /, when used with the --recursive option. This is the default behavior. See Treating / specially.

--no-preserve-root

Do not treat / specially when removing recursively. This option is not recommended unless you really want to remove all the files on your computer. See Treating / specially.

-r
-R
--recursive

Remove the listed directories and their contents recursively.

-v
--verbose

Print the name of each file before removing it.

One common question is how to remove files whose names begin with a ‘-’. GNU rm, like every program that uses the getopt function to parse its arguments, lets you use the ‘--’ option to indicate that all following arguments are non-options. To remove a file called -f in the current directory, you could type either:

rm -- -f

or:

rm ./-f

The Unix rm program’s use of a single ‘-’ for this purpose predates the development of the getopt standard syntax.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Basic operations   [Contents][Index]

11.6 shred: Remove files more securely

shred overwrites devices or files, to help prevent even very expensive hardware from recovering the data.

Ordinarily when you remove a file (see rm invocation), the data is not actually destroyed. Only the index listing where the file is stored is destroyed, and the storage is made available for reuse. There are undelete utilities that will attempt to reconstruct the index and can bring the file back if the parts were not reused.

On a busy system with a nearly-full drive, space can get reused in a few seconds. But there is no way to know for sure. If you have sensitive data, you may want to be sure that recovery is not possible by actually overwriting the file with non-sensitive data.

However, even after doing that, it is possible to take the disk back to a laboratory and use a lot of sensitive (and expensive) equipment to look for the faint “echoes” of the original data underneath the overwritten data. If the data has only been overwritten once, it’s not even that hard.

The best way to remove something irretrievably is to destroy the media it’s on with acid, melt it down, or the like. For cheap removable media like floppy disks, this is the preferred method. However, hard drives are expensive and hard to melt, so the shred utility tries to achieve a similar effect non-destructively.

This uses many overwrite passes, with the data patterns chosen to maximize the damage they do to the old data. While this will work on floppies, the patterns are designed for best effect on hard drives. For more details, see the source code and Peter Gutmann’s paper http://www.cs.auckland.ac.nz/~pgut001/pubs/secure_del.html">Secure Deletion of Data from Magnetic and Solid-State Memory, from the proceedings of the Sixth USENIX Security Symposium (San Jose, California, July 22–25, 1996).

Please note that shred relies on a very important assumption: that the file system overwrites data in place. This is the traditional way to do things, but many modern file system designs do not satisfy this assumption. Exceptions include:

In the particular case of ext3 file systems, the above disclaimer applies (and shred is thus of limited effectiveness) only in data=journal mode, which journals file data in addition to just metadata. In both the data=ordered (default) and data=writeback modes, shred works as usual. Ext3 journaling modes can be changed by adding the data=something option to the mount options for a particular file system in the /etc/fstab file, as documented in the mount man page (man mount).

If you are not sure how your file system operates, then you should assume that it does not overwrite data in place, which means that shred cannot reliably operate on regular files in your file system.

Generally speaking, it is more reliable to shred a device than a file, since this bypasses the problem of file system design mentioned above. However, even shredding devices is not always completely reliable. For example, most disks map out bad sectors invisibly to the application; if the bad sectors contain sensitive data, shred won’t be able to destroy it.

shred makes no attempt to detect or report this problem, just as it makes no attempt to do anything about backups. However, since it is more reliable to shred devices than files, shred by default does not truncate or remove the output file. This default is more suitable for devices, which typically cannot be truncated and should not be removed.

Finally, consider the risk of backups and mirrors. File system backups and remote mirrors may contain copies of the file that cannot be removed, and that will allow a shredded file to be recovered later. So if you keep any data you may later want to destroy using shred, be sure that it is not backed up or mirrored.

shred [option]… file[…]

The program accepts the following options. Also see Common options.

-f
--force

Override file permissions if necessary to allow overwriting.

-number
-n number
--iterations=number

By default, shred uses 3 passes of overwrite. You can reduce this to save time, or increase it if you think it’s appropriate. After 25 passes all of the internal overwrite patterns will have been used at least once.

--random-source=file

Use file as a source of random data used to overwrite and to choose pass ordering. See Random sources.

-s bytes
--size=bytes

Shred the first bytes bytes of the file. The default is to shred the whole file. bytes can be followed by a size specification like ‘K’, ‘M’, or ‘G’ to specify a multiple. See Block size.

-u
--remove[=how]

After shredding a file, truncate it (if possible) and then remove it. If a file has multiple links, only the named links will be removed. Often the file name is less sensitive than the file data, in which case the optional how parameter gives control of how to more efficiently remove each directory entry. The ‘unlink’ parameter will just use a standard unlink call, ‘wipe’ will also first obfuscate bytes in the name, and ‘wipesync’ will also sync each obfuscated byte in the name to disk. Note ‘wipesync’ is the default method, but can be expensive, requiring a sync for every character in every file. This can become significant with many files, or is redundant if your file system provides synchronous metadata updates.

-v
--verbose

Display to standard error all status updates as sterilization proceeds.

-x
--exact

By default, shred rounds the size of a regular file up to the next multiple of the file system block size to fully erase the slack space in the last block of the file. This space may contain portions of the current system memory on some systems for example. Use --exact to suppress that behavior. Thus, by default if you shred a 10-byte regular file on a system with 512-byte blocks, the resulting file will be 512 bytes long. With this option, shred does not increase the apparent size of the file.

-z
--zero

Normally, the last pass that shred writes is made up of random data. If this would be conspicuous on your hard drive (for example, because it looks like encrypted data), or you just think it’s tidier, the --zero option adds an additional overwrite pass with all zero bits. This is in addition to the number of passes specified by the --iterations option.

You might use the following command to erase all trace of the file system you’d created on the floppy disk in your first drive. That command takes about 20 minutes to erase a “1.44MB” (actually 1440 KiB) floppy.

shred --verbose /dev/fd0

Similarly, to erase all data on a selected partition of your hard disk, you could give a command like this:

shred --verbose /dev/sda5

On modern disks, a single pass should be adequate, and it will take one third the time of the default three-pass approach.

# 1 pass, write pseudo-random data; 3x faster than the default
shred --verbose -n1 /dev/sda5

To be on the safe side, use at least one pass that overwrites using pseudo-random data. I.e., don’t be tempted to use ‘-n0 --zero’, in case some disk controller optimizes the process of writing blocks of all zeros, and thereby does not clear all bytes in a block. Some SSDs may do just that.

A file of ‘-’ denotes standard output. The intended use of this is to shred a removed temporary file. For example:

i=$(mktemp)
exec 3<>"$i"
rm -- "$i"
echo "Hello, world" >&3
shred - >&3
exec 3>-

However, the command ‘shred - >file’ does not shred the contents of file, since the shell truncates file before invoking shred. Use the command ‘shred file’ or (if using a Bourne-compatible shell) the command ‘shred - 1<>file’ instead.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

12 Special file types

This chapter describes commands which create special types of files (and rmdir, which removes directories, one special file type).

Although Unix-like operating systems have markedly fewer special file types than others, not everything can be treated only as the undifferentiated byte stream of normal files. For example, when a file is created or removed, the system must record this information, which it does in a directory—a special type of file. Although you can read directories as normal files, if you’re curious, in order for the system to do its job it must impose a structure, a certain order, on the bytes of the file. Thus it is a “special” type of file.

Besides directories, other special file types include named pipes (FIFOs), symbolic links, sockets, and so-called special files.


Next: , Up: Special file types   [Contents][Index]

12.1 link: Make a hard link via the link syscall

link creates a single hard link at a time. It is a minimalist interface to the system-provided link function. See Hard Links in The GNU C Library Reference Manual. It avoids the bells and whistles of the more commonly-used ln command (see ln invocation). Synopsis:

link filename linkname

filename must specify an existing file, and linkname must specify a nonexistent entry in an existing directory. link simply calls link (filename, linkname) to create the link.

On a GNU system, this command acts like ‘ln --directory --no-target-directory filename linkname’. However, the --directory and --no-target-directory options are not specified by POSIX, and the link command is more portable in practice.

If filename is a symbolic link, it is unspecified whether linkname will be a hard link to the symbolic link or to the target of the symbolic link. Use ln -P or ln -L to specify which behavior is desired.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Special file types   [Contents][Index]

12.2 ln: Make links between files

ln makes links between files. By default, it makes hard links; with the -s option, it makes symbolic (or soft) links. Synopses:

ln [option]… [-T] target linkname
ln [option]… target
ln [option]… targetdirectory
ln [option]… -t directory target

Normally ln does not remove existing files. Use the --force (-f) option to remove them unconditionally, the --interactive (-i) option to remove them conditionally, and the --backup (-b) option to rename them.

A hard link is another name for an existing file; the link and the original are indistinguishable. Technically speaking, they share the same inode, and the inode contains all the information about a file—indeed, it is not incorrect to say that the inode is the file. Most systems prohibit making a hard link to a directory; on those where it is allowed, only the super-user can do so (and with caution, since creating a cycle will cause problems to many other utilities). Hard links cannot cross file system boundaries. (These restrictions are not mandated by POSIX, however.)

Symbolic links (symlinks for short), on the other hand, are a special file type (which not all kernels support: System V release 3 (and older) systems lack symlinks) in which the link file actually refers to a different file, by name. When most operations (opening, reading, writing, and so on) are passed the symbolic link file, the kernel automatically dereferences the link and operates on the target of the link. But some operations (e.g., removing) work on the link file itself, rather than on its target. The owner and group of a symlink are not significant to file access performed through the link, but do have implications on deleting a symbolic link from a directory with the restricted deletion bit set. On the GNU system, the mode of a symlink has no significance and cannot be changed, but on some BSD systems, the mode can be changed and will affect whether the symlink will be traversed in file name resolution. See Symbolic Links in The GNU C Library Reference Manual.

Symbolic links can contain arbitrary strings; a dangling symlink occurs when the string in the symlink does not resolve to a file. There are no restrictions against creating dangling symbolic links. There are trade-offs to using absolute or relative symlinks. An absolute symlink always points to the same file, even if the directory containing the link is moved. However, if the symlink is visible from more than one machine (such as on a networked file system), the file pointed to might not always be the same. A relative symbolic link is resolved in relation to the directory that contains the link, and is often useful in referring to files on the same device without regards to what name that device is mounted on when accessed via networked machines.

When creating a relative symlink in a different location than the current directory, the resolution of the symlink will be different than the resolution of the same string from the current directory. Therefore, many users prefer to first change directories to the location where the relative symlink will be created, so that tab-completion or other file resolution will find the same target as what will be placed in the symlink.

The program accepts the following options. Also see Common options.

-b
--backup[=method]

See Backup options. Make a backup of each file that would otherwise be overwritten or removed.

-d
-F
--directory

Allow users with appropriate privileges to attempt to make hard links to directories. However, note that this will probably fail due to system restrictions, even for the super-user.

-f
--force

Remove existing destination files.

-i
--interactive

Prompt whether to remove existing destination files.

-L
--logical

If -s is not in effect, and the source file is a symbolic link, create the hard link to the file referred to by the symbolic link, rather than the symbolic link itself.

-n
--no-dereference

Do not treat the last operand specially when it is a symbolic link to a directory. Instead, treat it as if it were a normal file.

When the destination is an actual directory (not a symlink to one), there is no ambiguity. The link is created in that directory. But when the specified destination is a symlink to a directory, there are two ways to treat the user’s request. ln can treat the destination just as it would a normal directory and create the link in it. On the other hand, the destination can be viewed as a non-directory—as the symlink itself. In that case, ln must delete or backup that symlink before creating the new link. The default is to treat a destination that is a symlink to a directory just like a directory.

This option is weaker than the --no-target-directory (-T) option, so it has no effect if both options are given.

-P
--physical

If -s is not in effect, and the source file is a symbolic link, create the hard link to the symbolic link itself. On platforms where this is not supported by the kernel, this option creates a symbolic link with identical contents; since symbolic link contents cannot be edited, any file name resolution performed through either link will be the same as if a hard link had been created.

-r
--relative

Make symbolic links relative to the link location.

Example:

ln -srv /a/file /tmp
'/tmp/file' -> '../a/file'

Relative symbolic links are generated based on their canonicalized containing directory, and canonicalized targets. I.e., all symbolic links in these file names will be resolved. See realpath invocation, which gives greater control over relative file name generation, as demonstrated in the following example:

ln--relative() {
  test "$1" = --no-symlinks && { nosym=$1; shift; }
  target="$1";
  test -d "$2" && link="$2/." || link="$2"
  rtarget="$(realpath $nosym -m "$target" \
              --relative-to "$(dirname "$link")")"
  ln -s -v "$rtarget" "$link"
}
-s
--symbolic

Make symbolic links instead of hard links. This option merely produces an error message on systems that do not support symbolic links.

-S suffix
--suffix=suffix

Append suffix to each backup file made with -b. See Backup options.

-t directory
--target-directory=directory

Specify the destination directory. See Target directory.

-T
--no-target-directory

Do not treat the last operand specially when it is a directory or a symbolic link to a directory. See Target directory.

-v
--verbose

Print the name of each file after linking it successfully.

If -L and -P are both given, the last one takes precedence. If -s is also given, -L and -P are silently ignored. If neither option is given, then this implementation defaults to -P if the system link supports hard links to symbolic links (such as the GNU system), and -L if link follows symbolic links (such as on BSD).

An exit status of zero indicates success, and a nonzero value indicates failure.

Examples:

Bad Example:

# Create link ../a pointing to a in that directory.
# Not really useful because it points to itself.
ln -s a ..

Better Example:

# Change to the target before creating symlinks to avoid being confused.
cd ..
ln -s adir/a .

Bad Example:

# Hard coded file names don't move well.
ln -s $(pwd)/a /some/dir/

Better Example:

# Relative file names survive directory moves and also
# work across networked file systems.
ln -s afile anotherfile
ln -s ../adir/afile yetanotherfile

Next: , Previous: , Up: Special file types   [Contents][Index]

12.3 mkdir: Make directories

mkdir creates directories with the specified names. Synopsis:

mkdir [option]… name

mkdir creates each directory name in the order given. It reports an error if name already exists, unless the -p option is given and name is a directory.

The program accepts the following options. Also see Common options.

-m mode
--mode=mode

Set the file permission bits of created directories to mode, which uses the same syntax as in chmod and uses ‘a=rwx’ (read, write and execute allowed for everyone) for the point of the departure. See File permissions.

Normally the directory has the desired file mode bits at the moment it is created. As a GNU extension, mode may also mention special mode bits, but in this case there may be a temporary window during which the directory exists but its special mode bits are incorrect. See Directory Setuid and Setgid, for how the set-user-ID and set-group-ID bits of directories are inherited unless overridden in this way.

-p
--parents

Make any missing parent directories for each argument, setting their file permission bits to the umask modified by ‘u+wx’. Ignore existing parent directories, and do not change their file permission bits.

To set the file permission bits of any newly-created parent directories to a value that includes ‘u+wx’, you can set the umask before invoking mkdir. For example, if the shell command ‘(umask u=rwx,go=rx; mkdir -p P/Q)’ creates the parent P it sets the parent’s permission bits to ‘u=rwx,go=rx’. To set a parent’s special mode bits as well, you can invoke chmod after mkdir. See Directory Setuid and Setgid, for how the set-user-ID and set-group-ID bits of newly-created parent directories are inherited.

-v
--verbose

Print a message for each created directory. This is most useful with --parents.

-Z
--context[=context]

Without a specified context, adjust the SELinux security context according to the system default type for destination files, similarly to the restorecon command. The long form of this option with a specific context specified, will set the context for newly created files only. With a specified context, if both SELinux and SMACK are disabled, a warning is issued.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Special file types   [Contents][Index]

12.4 mkfifo: Make FIFOs (named pipes)

mkfifo creates FIFOs (also called named pipes) with the specified names. Synopsis:

mkfifo [option] name

A FIFO is a special file type that permits independent processes to communicate. One process opens the FIFO file for writing, and another for reading, after which data can flow as with the usual anonymous pipe in shells or elsewhere.

The program accepts the following options. Also see Common options.

-m mode
--mode=mode

Set the mode of created FIFOs to mode, which is symbolic as in chmod and uses ‘a=rw’ (read and write allowed for everyone) for the point of departure. mode should specify only file permission bits. See File permissions.

-Z
--context[=context]

Without a specified context, adjust the SELinux security context according to the system default type for destination files, similarly to the restorecon command. The long form of this option with a specific context specified, will set the context for newly created files only. With a specified context, if both SELinux and SMACK are disabled, a warning is issued.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Special file types   [Contents][Index]

12.5 mknod: Make block or character special files

mknod creates a FIFO, character special file, or block special file with the specified name. Synopsis:

mknod [option]… name type [major minor]

Unlike the phrase “special file type” above, the term special file has a technical meaning on Unix: something that can generate or receive data. Usually this corresponds to a physical piece of hardware, e.g., a printer or a disk. (These files are typically created at system-configuration time.) The mknod command is what creates files of this type. Such devices can be read either a character at a time or a “block” (many characters) at a time, hence we say there are block special files and character special files.

Due to shell aliases and built-in mknod functions, using an unadorned mknod interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env mknod …) to avoid interference from the shell.

The arguments after name specify the type of file to make:

p

for a FIFO

b

for a block special file

c

for a character special file

When making a block or character special file, the major and minor device numbers must be given after the file type. If a major or minor device number begins with ‘0x’ or ‘0X’, it is interpreted as hexadecimal; otherwise, if it begins with ‘0’, as octal; otherwise, as decimal.

The program accepts the following options. Also see Common options.

-m mode
--mode=mode

Set the mode of created files to mode, which is symbolic as in chmod and uses ‘a=rw’ as the point of departure. mode should specify only file permission bits. See File permissions.

-Z
--context[=context]

Without a specified context, adjust the SELinux security context according to the system default type for destination files, similarly to the restorecon command. The long form of this option with a specific context specified, will set the context for newly created files only. With a specified context, if both SELinux and SMACK are disabled, a warning is issued.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Special file types   [Contents][Index]

12.6 readlink: Print value of a symlink or canonical file name

readlink may work in one of two supported modes:

Readlink mode

readlink outputs the value of the given symbolic links. If readlink is invoked with an argument other than the name of a symbolic link, it produces no output and exits with a nonzero exit code.

Canonicalize mode

readlink outputs the absolute name of the given files which contain no ., .. components nor any repeated separators (/) or symbolic links.

readlink [option]… file

By default, readlink operates in readlink mode.

The program accepts the following options. Also see Common options.

-f
--canonicalize

Activate canonicalize mode. If any component of the file name except the last one is missing or unavailable, readlink produces no output and exits with a nonzero exit code. A trailing slash is ignored.

-e
--canonicalize-existing

Activate canonicalize mode. If any component is missing or unavailable, readlink produces no output and exits with a nonzero exit code. A trailing slash requires that the name resolve to a directory.

-m
--canonicalize-missing

Activate canonicalize mode. If any component is missing or unavailable, readlink treats it as a directory.

-n
--no-newline

Do not print the output delimiter, when a single file is specified. Print a warning if specified along with multiple files.

-s
-q
--silent
--quiet

Suppress most error messages.

-v
--verbose

Report error messages.

-z
--zero

Output a zero byte (ASCII NUL) at the end of each line, rather than a newline. This option enables other programs to parse the output even when that output would contain data with embedded newlines.

The readlink utility first appeared in OpenBSD 2.1.

The realpath command without options, operates like readlink in canonicalize mode.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Special file types   [Contents][Index]

12.7 rmdir: Remove empty directories

rmdir removes empty directories. Synopsis:

rmdir [option]… directory

If any directory argument does not refer to an existing empty directory, it is an error.

The program accepts the following options. Also see Common options.

--ignore-fail-on-non-empty

Ignore each failure to remove a directory that is solely because the directory is non-empty.

-p
--parents

Remove directory, then try to remove each component of directory. So, for example, ‘rmdir -p a/b/c’ is similar to ‘rmdir a/b/c a/b a’. As such, it fails if any of those directories turns out not to be empty. Use the --ignore-fail-on-non-empty option to make it so such a failure does not evoke a diagnostic and does not cause rmdir to exit unsuccessfully.

-v
--verbose

Give a diagnostic for each successful removal. directory is removed.

See rm invocation, for how to remove non-empty directories (recursively).

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Special file types   [Contents][Index]

12.8 unlink: Remove files via the unlink syscall

unlink deletes a single specified file name. It is a minimalist interface to the system-provided unlink function. See Deleting Files in The GNU C Library Reference Manual. Synopsis: It avoids the bells and whistles of the more commonly-used rm command (see rm invocation).

unlink filename

On some systems unlink can be used to delete the name of a directory. On others, it can be used that way only by a privileged user. In the GNU system unlink can never delete the name of a directory.

The unlink command honors the --help and --version options. To remove a file whose name begins with ‘-’, prefix the name with ‘./’, e.g., ‘unlink ./--help’.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

13 Changing file attributes

A file is not merely its contents, a name, and a file type (see Special file types). A file also has an owner (a user ID), a group (a group ID), permissions (what the owner can do with the file, what people in the group can do, and what everyone else can do), various timestamps, and other information. Collectively, we call these a file’s attributes.

These commands change file attributes.


Next: , Up: Changing file attributes   [Contents][Index]

13.1 chown: Change file owner and group

chown changes the user and/or group ownership of each given file to new-owner or to the user and group of an existing reference file. Synopsis:

chown [option]… {new-owner | --reference=ref_file} file

If used, new-owner specifies the new owner and/or group as follows (with no embedded white space):

[owner] [ : [group] ]

Specifically:

owner

If only an owner (a user name or numeric user ID) is given, that user is made the owner of each given file, and the files’ group is not changed.

owner‘:’group

If the owner is followed by a colon and a group (a group name or numeric group ID), with no spaces between them, the group ownership of the files is changed as well (to group).

owner‘:

If a colon but no group name follows owner, that user is made the owner of the files and the group of the files is changed to owner’s login group.

:’group

If the colon and following group are given, but the owner is omitted, only the group of the files is changed; in this case, chown performs the same function as chgrp.

:

If only a colon is given, or if new-owner is empty, neither the owner nor the group is changed.

If owner or group is intended to represent a numeric user or group ID, then you may specify it with a leading ‘+’. See Disambiguating names and IDs.

Some older scripts may still use ‘.’ in place of the ‘:’ separator. POSIX 1003.1-2001 (see Standards conformance) does not require support for that, but for backward compatibility GNU chown supports ‘.’ so long as no ambiguity results. New scripts should avoid the use of ‘.’ because it is not portable, and because it has undesirable results if the entire owner‘.’group happens to identify a user whose name contains ‘.’.

It is system dependent whether a user can change the group to an arbitrary one, or the more portable behavior of being restricted to setting a group of which the user is a member.

The chown command sometimes clears the set-user-ID or set-group-ID permission bits. This behavior depends on the policy and functionality of the underlying chown system call, which may make system-dependent file mode modifications outside the control of the chown command. For example, the chown command might not affect those bits when invoked by a user with appropriate privileges, or when the bits signify some function other than executable permission (e.g., mandatory locking). When in doubt, check the underlying system behavior.

The program accepts the following options. Also see Common options.

-c
--changes

Verbosely describe the action for each file whose ownership actually changes.

-f
--silent
--quiet

Do not print error messages about files whose ownership cannot be changed.

--from=old-owner

Change a file’s ownership only if it has current attributes specified by old-owner. old-owner has the same form as new-owner described above. This option is useful primarily from a security standpoint in that it narrows considerably the window of potential abuse. For example, to reflect a user ID numbering change for one user’s files without an option like this, root might run

find / -owner OLDUSER -print0 | xargs -0 chown -h NEWUSER

But that is dangerous because the interval between when the find tests the existing file’s owner and when the chown is actually run may be quite large. One way to narrow the gap would be to invoke chown for each file as it is found:

find / -owner OLDUSER -exec chown -h NEWUSER {} \;

But that is very slow if there are many affected files. With this option, it is safer (the gap is narrower still) though still not perfect:

chown -h -R --from=OLDUSER NEWUSER /
--dereference

Do not act on symbolic links themselves but rather on what they point to. This is the default.

-h
--no-dereference

Act on symbolic links themselves instead of what they point to. This mode relies on the lchown system call. On systems that do not provide the lchown system call, chown fails when a file specified on the command line is a symbolic link. By default, no diagnostic is issued for symbolic links encountered during a recursive traversal, but see --verbose.

--preserve-root

Fail upon any attempt to recursively change the root directory, /. Without --recursive, this option has no effect. See Treating / specially.

--no-preserve-root

Cancel the effect of any preceding --preserve-root option. See Treating / specially.

--reference=ref_file

Change the user and group of each file to be the same as those of ref_file. If ref_file is a symbolic link, do not use the user and group of the symbolic link, but rather those of the file it refers to.

-v
--verbose

Output a diagnostic for every file processed. If a symbolic link is encountered during a recursive traversal on a system without the lchown system call, and --no-dereference is in effect, then issue a diagnostic saying neither the symbolic link nor its referent is being changed.

-R
--recursive

Recursively change ownership of directories and their contents.

-H

If --recursive (-R) is specified and a command line argument is a symbolic link to a directory, traverse it. See Traversing symlinks.

-L

In a recursive traversal, traverse every symbolic link to a directory that is encountered. See Traversing symlinks.

-P

Do not traverse any symbolic links. This is the default if none of -H, -L, or -P is specified. See Traversing symlinks.

An exit status of zero indicates success, and a nonzero value indicates failure.

Examples:

# Change the owner of /u to "root".
chown root /u

# Likewise, but also change its group to "staff".
chown root:staff /u

# Change the owner of /u and subfiles to "root".
chown -hR root /u

Next: , Previous: , Up: Changing file attributes   [Contents][Index]

13.2 chgrp: Change group ownership

chgrp changes the group ownership of each given file to group (which can be either a group name or a numeric group ID) or to the group of an existing reference file. See chown invocation. Synopsis:

chgrp [option]… {group | --reference=ref_file} file

If group is intended to represent a numeric group ID, then you may specify it with a leading ‘+’. See Disambiguating names and IDs.

It is system dependent whether a user can change the group to an arbitrary one, or the more portable behavior of being restricted to setting a group of which the user is a member.

The program accepts the following options. Also see Common options.

-c
--changes

Verbosely describe the action for each file whose group actually changes.

-f
--silent
--quiet

Do not print error messages about files whose group cannot be changed.

--dereference

Do not act on symbolic links themselves but rather on what they point to. This is the default.

-h
--no-dereference

Act on symbolic links themselves instead of what they point to. This mode relies on the lchown system call. On systems that do not provide the lchown system call, chgrp fails when a file specified on the command line is a symbolic link. By default, no diagnostic is issued for symbolic links encountered during a recursive traversal, but see --verbose.

--preserve-root

Fail upon any attempt to recursively change the root directory, /. Without --recursive, this option has no effect. See Treating / specially.

--no-preserve-root

Cancel the effect of any preceding --preserve-root option. See Treating / specially.

--reference=ref_file

Change the group of each file to be the same as that of ref_file. If ref_file is a symbolic link, do not use the group of the symbolic link, but rather that of the file it refers to.

-v
--verbose

Output a diagnostic for every file processed. If a symbolic link is encountered during a recursive traversal on a system without the lchown system call, and --no-dereference is in effect, then issue a diagnostic saying neither the symbolic link nor its referent is being changed.

-R
--recursive

Recursively change the group ownership of directories and their contents.

-H

If --recursive (-R) is specified and a command line argument is a symbolic link to a directory, traverse it. See Traversing symlinks.

-L

In a recursive traversal, traverse every symbolic link to a directory that is encountered. See Traversing symlinks.

-P

Do not traverse any symbolic links. This is the default if none of -H, -L, or -P is specified. See Traversing symlinks.

An exit status of zero indicates success, and a nonzero value indicates failure.

Examples:

# Change the group of /u to "staff".
chgrp staff /u

# Change the group of /u and subfiles to "staff".
chgrp -hR staff /u

Next: , Previous: , Up: Changing file attributes   [Contents][Index]

13.3 chmod: Change access permissions

chmod changes the access permissions of the named files. Synopsis:

chmod [option]… {mode | --reference=ref_file} file

chmod never changes the permissions of symbolic links, since the chmod system call cannot change their permissions. This is not a problem since the permissions of symbolic links are never used. However, for each symbolic link listed on the command line, chmod changes the permissions of the pointed-to file. In contrast, chmod ignores symbolic links encountered during recursive directory traversals.

A successful use of chmod clears the set-group-ID bit of a regular file if the file’s group ID does not match the user’s effective group ID or one of the user’s supplementary group IDs, unless the user has appropriate privileges. Additional restrictions may cause the set-user-ID and set-group-ID bits of mode or ref_file to be ignored. This behavior depends on the policy and functionality of the underlying chmod system call. When in doubt, check the underlying system behavior.

If used, mode specifies the new file mode bits. For details, see the section on File permissions. If you really want mode to have a leading ‘-’, you should use -- first, e.g., ‘chmod -- -w file’. Typically, though, ‘chmod a-w file’ is preferable, and chmod -w file (without the --) complains if it behaves differently from what ‘chmod a-w file’ would do.

The program accepts the following options. Also see Common options.

-c
--changes

Verbosely describe the action for each file whose permissions actually changes.

-f
--silent
--quiet

Do not print error messages about files whose permissions cannot be changed.

--preserve-root

Fail upon any attempt to recursively change the root directory, /. Without --recursive, this option has no effect. See Treating / specially.

--no-preserve-root

Cancel the effect of any preceding --preserve-root option. See Treating / specially.

-v
--verbose

Verbosely describe the action or non-action taken for every file.

--reference=ref_file

Change the mode of each file to be the same as that of ref_file. See File permissions. If ref_file is a symbolic link, do not use the mode of the symbolic link, but rather that of the file it refers to.

-R
--recursive

Recursively change permissions of directories and their contents.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Changing file attributes   [Contents][Index]

13.4 touch: Change file timestamps

touch changes the access and/or modification times of the specified files. Synopsis:

touch [option]… file

Any file argument that does not exist is created empty, unless option --no-create (-c) or --no-dereference (-h) was in effect.

A file argument string of ‘-’ is handled specially and causes touch to change the times of the file associated with standard output.

By default, touch sets file timestamps to the current time. Because touch acts on its operands left to right, the resulting timestamps of earlier and later operands may disagree. Also, the determination of what time is “current” depends on the platform. Platforms with network file systems often use different clocks for the operating system and for file systems; because touch typically uses file systems’ clocks by default, clock skew can cause the resulting file timestamps to appear to be in a program’s “future” or “past”.

The touch command sets the file’s timestamp to the greatest representable value that is not greater than the requested time. This can differ from the requested time for several reasons. First, the requested time may have a higher resolution than supported. Second, a file system may use different resolutions for different types of times. Third, file timestamps may use a different resolution than operating system timestamps. Fourth, the operating system primitives used to update timestamps may employ yet a different resolution. For example, in theory a file system might use 10-microsecond resolution for access time and 100-nanosecond resolution for modification time, and the operating system might use nanosecond resolution for the current time and microsecond resolution for the primitive that touch uses to set a file’s timestamp to an arbitrary value.

When setting file timestamps to the current time, touch can change the timestamps for files that the user does not own but has write permission for. Otherwise, the user must own the files. Some older systems have a further restriction: the user must own the files unless both the access and modification times are being set to the current time.

Although touch provides options for changing two of the times—the times of last access and modification—of a file, there is actually a standard third one as well: the inode change time. This is often referred to as a file’s ctime. The inode change time represents the time when the file’s meta-information last changed. One common example of this is when the permissions of a file change. Changing the permissions doesn’t access the file, so the atime doesn’t change, nor does it modify the file, so the mtime doesn’t change. Yet, something about the file itself has changed, and this must be noted somewhere. This is the job of the ctime field. This is necessary, so that, for example, a backup program can make a fresh copy of the file, including the new permissions value. Another operation that modifies a file’s ctime without affecting the others is renaming. In any case, it is not possible, in normal operations, for a user to change the ctime field to a user-specified value. Some operating systems and file systems support a fourth time: the birth time, when the file was first created; by definition, this timestamp never changes.

Time stamps assume the time zone rules specified by the TZ environment variable, or by the system default rules if TZ is not set. See Specifying the Time Zone with TZ in The GNU C Library Reference Manual. You can avoid ambiguities during daylight saving transitions by using UTC time stamps.

The program accepts the following options. Also see Common options.

-a
--time=atime
--time=access
--time=use

Change the access time only.

-c
--no-create

Do not warn about or create files that do not exist.

-d
--date=time

Use time instead of the current time. It can contain month names, time zones, ‘am’ and ‘pm’, ‘yesterday’, etc. For example, --date="2004-02-27 14:19:13.489392193 +0530" specifies the instant of time that is 489,392,193 nanoseconds after February 27, 2004 at 2:19:13 PM in a time zone that is 5 hours and 30 minutes east of UTC. See Date input formats. File systems that do not support high-resolution time stamps silently ignore any excess precision here.

-f

Ignored; for compatibility with BSD versions of touch.

-h
--no-dereference

Attempt to change the timestamps of a symbolic link, rather than what the link refers to. When using this option, empty files are not created, but option -c must also be used to avoid warning about files that do not exist. Not all systems support changing the timestamps of symlinks, since underlying system support for this action was not required until POSIX 2008. Also, on some systems, the mere act of examining a symbolic link changes the access time, such that only changes to the modification time will persist long enough to be observable. When coupled with option -r, a reference timestamp is taken from a symbolic link rather than the file it refers to.

-m
--time=mtime
--time=modify

Change the modification time only.

-r file
--reference=file

Use the times of the reference file instead of the current time. If this option is combined with the --date=time (-d time) option, the reference file’s time is the origin for any relative times given, but is otherwise ignored. For example, ‘-r foo -d '-5 seconds'’ specifies a time stamp equal to five seconds before the corresponding time stamp for foo. If file is a symbolic link, the reference timestamp is taken from the target of the symlink, unless -h was also in effect.

-t [[cc]yy]mmddhhmm[.ss]

Use the argument (optional four-digit or two-digit years, months, days, hours, minutes, optional seconds) instead of the current time. If the year is specified with only two digits, then cc is 20 for years in the range 0 … 68, and 19 for years in 69 … 99. If no digits of the year are specified, the argument is interpreted as a date in the current year. On the atypical systems that support leap seconds, ss may be ‘60’.

On older systems, touch supports an obsolete syntax, as follows. If no timestamp is given with any of the -d, -r, or -t options, and if there are two or more files and the first file is of the form ‘mmddhhmm[yy]’ and this would be a valid argument to the -t option (if the yy, if any, were moved to the front), and if the represented year is in the range 1969–1999, that argument is interpreted as the time for the other files instead of as a file name. This obsolete behavior can be enabled or disabled with the _POSIX2_VERSION environment variable (see Standards conformance), but portable scripts should avoid commands whose behavior depends on this variable. For example, use ‘touch ./12312359 main.c’ or ‘touch -t 12312359 main.c’ rather than the ambiguous ‘touch 12312359 main.c’.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

14 Disk usage

No disk can hold an infinite amount of data. These commands report how much disk storage is in use or available, report other file and file status information, and write buffers to disk.


Next: , Up: Disk usage   [Contents][Index]

14.1 df: Report file system disk space usage

df reports the amount of disk space used and available on file systems. Synopsis:

df [option]… [file]…

With no arguments, df reports the space used and available on all currently mounted file systems (of all types). Otherwise, df reports on the file system containing each argument file.

Normally the disk space is printed in units of 1024 bytes, but this can be overridden (see Block size). Non-integer quantities are rounded up to the next higher unit.

For bind mounts and without arguments, df only outputs the statistics for that device with the shortest mount point name in the list of file systems (mtab), i.e., it hides duplicate entries, unless the -a option is specified.

With the same logic, df elides a mount entry of a dummy pseudo device if there is another mount entry of a real block device for that mount point with the same device number, e.g. the early-boot pseudo file system ‘rootfs’ is not shown per default when already the real root device has been mounted.

If an argument file resolves to a special file containing a mounted file system, df shows the space available on that file system rather than on the file system containing the device node. GNU df does not attempt to determine the disk usage on unmounted file systems, because on most kinds of systems doing so requires extremely nonportable intimate knowledge of file system structures.

The program accepts the following options. Also see Common options.

-a
--all

Include in the listing dummy, duplicate, or inaccessible file systems, which are omitted by default. Dummy file systems are typically special purpose pseudo file systems such as ‘/proc’, with no associated storage. Duplicate file systems are local or remote file systems that are mounted at separate locations in the local file hierarchy, or bind mounted locations. Inaccessible file systems are those which are mounted but subsequently over-mounted by another file system at that point, or otherwise inaccessible due to permissions of the mount point etc.

-B size
--block-size=size

Scale sizes by size before printing them (see Block size). For example, -BG prints sizes in units of 1,073,741,824 bytes.

-h
--human-readable

Append a size letter to each size, such as ‘M’ for mebibytes. Powers of 1024 are used, not 1000; ‘M’ stands for 1,048,576 bytes. This option is equivalent to --block-size=human-readable. Use the --si option if you prefer powers of 1000.

-H

Equivalent to --si.

-i
--inodes

List inode usage information instead of block usage. An inode (short for index node) contains information about a file such as its owner, permissions, timestamps, and location on the disk.

-k

Print sizes in 1024-byte blocks, overriding the default block size (see Block size). This option is equivalent to --block-size=1K.

-l
--local

Limit the listing to local file systems. By default, remote file systems are also listed.

--no-sync

Do not invoke the sync system call before getting any usage data. This may make df run significantly faster on systems with many disks, but on some systems (notably SunOS) the results may be slightly out of date. This is the default.

--output
--output[=field_list]

Use the output format defined by field_list, or print all fields if field_list is omitted. In the latter case, the order of the columns conforms to the order of the field descriptions below.

The use of the --output together with each of the options -i, -P, and -T is mutually exclusive.

FIELD_LIST is a comma-separated list of columns to be included in df’s output and therefore effectively controls the order of output columns. Each field can thus be used at the place of choice, but yet must only be used once.

Valid field names in the field_list are:

source

The source of the mount point, usually a device.

fstype

File system type.

itotal

Total number of inodes.

iused

Number of used inodes.

iavail

Number of available inodes.

ipcent

Percentage of iused divided by itotal.

size

Total number of blocks.

used

Number of used blocks.

avail

Number of available blocks.

pcent

Percentage of used divided by size.

file

The file name if specified on the command line.

target

The mount point.

The fields for block and inodes statistics are affected by the scaling options like -h as usual.

The definition of the field_list can even be split among several --output uses.

#!/bin/sh
# Print the TARGET (i.e., the mount point) along with their percentage
# statistic regarding the blocks and the inodes.
df --out=target --output=pcent,ipcent

# Print all available fields.
df --o
-P
--portability

Use the POSIX output format. This is like the default format except for the following:

  1. The information about each file system is always printed on exactly one line; a mount device is never put on a line by itself. This means that if the mount device name is more than 20 characters long (e.g., for some network mounts), the columns are misaligned.
  2. The labels in the header output line are changed to conform to POSIX.
  3. The default block size and output format are unaffected by the DF_BLOCK_SIZE, BLOCK_SIZE and BLOCKSIZE environment variables. However, the default block size is still affected by POSIXLY_CORRECT: it is 512 if POSIXLY_CORRECT is set, 1024 otherwise. See Block size.
--si

Append an SI-style abbreviation to each size, such as ‘M’ for megabytes. Powers of 1000 are used, not 1024; ‘M’ stands for 1,000,000 bytes. This option is equivalent to --block-size=si. Use the -h or --human-readable option if you prefer powers of 1024.

--sync

Invoke the sync system call before getting any usage data. On some systems (notably SunOS), doing this yields more up to date results, but in general this option makes df much slower, especially when there are many or very busy file systems.

--total

Print a grand total of all arguments after all arguments have been processed. This can be used to find out the total disk size, usage and available space of all listed devices. If no arguments are specified df will try harder to elide file systems insignificant to the total available space, by suppressing duplicate remote file systems.

For the grand total line, df prints ‘"total"’ into the source column, and ‘"-"’ into the target column. If there is no source column (see --output), then df prints ‘"total"’ into the target column, if present.

-t fstype
--type=fstype

Limit the listing to file systems of type fstype. Multiple file system types can be specified by giving multiple -t options. By default, nothing is omitted.

-T
--print-type

Print each file system’s type. The types printed here are the same ones you can include or exclude with -t and -x. The particular types printed are whatever is supported by the system. Here are some of the common names (this list is certainly not exhaustive):

nfs

An NFS file system, i.e., one mounted over a network from another machine. This is the one type name which seems to be used uniformly by all systems.

ext2, ext3, ext4, xfs, btrfs…

A file system on a locally-mounted hard disk. (The system might even support more than one type here; Linux does.)

iso9660, cdfs

A file system on a CD or DVD drive. HP-UX uses ‘cdfs’, most other systems use ‘iso9660’.

ntfs,fat

File systems used by MS-Windows / MS-DOS.

-x fstype
--exclude-type=fstype

Limit the listing to file systems not of type fstype. Multiple file system types can be eliminated by giving multiple -x options. By default, no file system types are omitted.

-v

Ignored; for compatibility with System V versions of df.

df is installed only on systems that have usable mount tables, so portable scripts should not rely on its existence.

An exit status of zero indicates success, and a nonzero value indicates failure. Failure includes the case where no output is generated, so you can inspect the exit status of a command like ‘df -t ext3 -t reiserfs dir’ to test whether dir is on a file system of type ‘ext3’ or ‘reiserfs’.

Since the list of file systems (mtab) is needed to determine the file system type, failure includes the cases when that list cannot be read and one or more of the options -a, -l, -t or -x is used together with a file name argument.


Next: , Previous: , Up: Disk usage   [Contents][Index]

14.2 du: Estimate file space usage

du reports the amount of disk space used by the set of specified files and for each subdirectory (of directory arguments). Synopsis:

du [option]… [file]…

With no arguments, du reports the disk space for the current directory. Normally the disk space is printed in units of 1024 bytes, but this can be overridden (see Block size). Non-integer quantities are rounded up to the next higher unit.

If two or more hard links point to the same file, only one of the hard links is counted. The file argument order affects which links are counted, and changing the argument order may change the numbers and entries that du outputs.

The program accepts the following options. Also see Common options.

-0
--null

Output a zero byte (ASCII NUL) at the end of each line, rather than a newline. This option enables other programs to parse the output even when that output would contain data with embedded newlines.

-a
--all

Show counts for all files, not just directories.

--apparent-size

Print apparent sizes, rather than disk usage. The apparent size of a file is the number of bytes reported by wc -c on regular files, or more generally, ls -l --block-size=1 or stat --format=%s. For example, a file containing the word ‘zoo’ with no newline would, of course, have an apparent size of 3. Such a small file may require anywhere from 0 to 16 KiB or more of disk space, depending on the type and configuration of the file system on which the file resides. However, a sparse file created with this command:

dd bs=1 seek=2GiB if=/dev/null of=big

has an apparent size of 2 GiB, yet on most modern systems, it actually uses almost no disk space.

-B size
--block-size=size

Scale sizes by size before printing them (see Block size). For example, -BG prints sizes in units of 1,073,741,824 bytes.

-b
--bytes

Equivalent to --apparent-size --block-size=1.

-c
--total

Print a grand total of all arguments after all arguments have been processed. This can be used to find out the total disk usage of a given set of files or directories.

-D
--dereference-args

Dereference symbolic links that are command line arguments. Does not affect other symbolic links. This is helpful for finding out the disk usage of directories, such as /usr/tmp, which are often symbolic links.

-d depth
--max-depth=depth

Show the total for each directory (and file if –all) that is at most MAX_DEPTH levels down from the root of the hierarchy. The root is at level 0, so du --max-depth=0 is equivalent to du -s.

--files0-from=file

Disallow processing files named on the command line, and instead process those named in file file; each name being terminated by a zero byte (ASCII NUL). This is useful when the list of file names is so long that it may exceed a command line length limitation. In such cases, running du via xargs is undesirable because it splits the list into pieces and makes du print with the --total (-c) option for each sublist rather than for the entire list. One way to produce a list of ASCII NUL terminated file names is with GNU find, using its -print0 predicate. If file is ‘-’ then the ASCII NUL terminated file names are read from standard input.

-H

Equivalent to --dereference-args (-D).

-h
--human-readable

Append a size letter to each size, such as ‘M’ for mebibytes. Powers of 1024 are used, not 1000; ‘M’ stands for 1,048,576 bytes. This option is equivalent to --block-size=human-readable. Use the --si option if you prefer powers of 1000.

--inodes

List inode usage information instead of block usage. This option is useful for finding directories which contain many files, and therefore eat up most of the inodes space of a file system (see df, option --inodes). It can well be combined with the options -a, -c, -h, -l, -s, -S, -t and -x; however, passing other options regarding the block size, for example -b, -m and --apparent-size, is ignored.

-k

Print sizes in 1024-byte blocks, overriding the default block size (see Block size). This option is equivalent to --block-size=1K.

-L
--dereference

Dereference symbolic links (show the disk space used by the file or directory that the link points to instead of the space used by the link).

-l
--count-links

Count the size of all files, even if they have appeared already (as a hard link).

-m

Print sizes in 1,048,576-byte blocks, overriding the default block size (see Block size). This option is equivalent to --block-size=1M.

-P
--no-dereference

For each symbolic links encountered by du, consider the disk space used by the symbolic link.

-S
--separate-dirs

Normally, in the output of du (when not using --summarize), the size listed next to a directory name, d, represents the sum of sizes of all entries beneath d as well as the size of d itself. With --separate-dirs, the size reported for a directory name, d, will exclude the size of any subdirectories.

--si

Append an SI-style abbreviation to each size, such as ‘M’ for megabytes. Powers of 1000 are used, not 1024; ‘M’ stands for 1,000,000 bytes. This option is equivalent to --block-size=si. Use the -h or --human-readable option if you prefer powers of 1024.

-s
--summarize

Display only a total for each argument.

-t size
--threshold=size

Exclude entries based on a given size. The size refers to used blocks in normal mode (see Block size), or inodes count in conjunction with the --inodes option.

If size is positive, then du will only print entries with a size greater than or equal to that.

If size is negative, then du will only print entries with a size smaller than or equal to that.

Although GNU find can be used to find files of a certain size, du’s --threshold option can be used to also filter directories based on a given size.

Please note that the --threshold option can be combined with the --apparent-size option, and in this case would elide entries based on its apparent size.

Please note that the --threshold option can be combined with the --inodes option, and in this case would elide entries based on its inodes count.

Here’s how you would use --threshold to find directories with a size greater than or equal to 200 megabytes:

du --threshold=200MB

Here’s how you would use --threshold to find directories and files - note the -a - with an apparent size smaller than or equal to 500 bytes:

du -a -t -500 --apparent-size

Here’s how you would use --threshold to find directories on the root file system with more than 20000 inodes used in the directory tree below:

du --inodes -x --threshold=20000 /
--time

Show time of the most recent modification of any file in the directory, or any of its subdirectories.

--time=ctime
--time=status
--time=use

Show the most recent status change time (the ‘ctime’ in the inode) of any file in the directory, instead of the modification time.

--time=atime
--time=access

Show the most recent access time (the ‘atime’ in the inode) of any file in the directory, instead of the modification time.

--time-style=style

List timestamps in style style. This option has an effect only if the --time option is also specified. The style should be one of the following:

+format

List timestamps using format, where format is interpreted like the format argument of date (see date invocation). For example, --time-style="+%Y-%m-%d %H:%M:%S" causes du to list timestamps like ‘2002-03-30 23:45:56’. As with date, format’s interpretation is affected by the LC_TIME locale category.

full-iso

List timestamps in full using ISO 8601 date, time, and time zone format with nanosecond precision, e.g., ‘2002-03-30 23:45:56.477817180 -0700’. This style is equivalent to ‘+%Y-%m-%d %H:%M:%S.%N %z’.

long-iso

List ISO 8601 date and time in minutes, e.g., ‘2002-03-30 23:45’. These timestamps are shorter than ‘full-iso’ timestamps, and are usually good enough for everyday work. This style is equivalent to ‘+%Y-%m-%d %H:%M’.

iso

List ISO 8601 dates for timestamps, e.g., ‘2002-03-30’. This style is equivalent to ‘+%Y-%m-%d’.

You can specify the default value of the --time-style option with the environment variable TIME_STYLE; if TIME_STYLE is not set the default style is ‘long-iso’. For compatibility with ls, if TIME_STYLE begins with ‘+’ and contains a newline, the newline and any later characters are ignored; if TIME_STYLE begins with ‘posix-’ the ‘posix-’ is ignored; and if TIME_STYLE is ‘locale’ it is ignored.

-X file
--exclude-from=file

Like --exclude, except take the patterns to exclude from file, one per line. If file is ‘-’, take the patterns from standard input.

--exclude=pattern

When recursing, skip subdirectories or files matching pattern. For example, du --exclude='*.o' excludes files whose names end in ‘.o’.

-x
--one-file-system

Skip directories that are on different file systems from the one that the argument being processed is on.

On BSD systems, du reports sizes that are half the correct values for files that are NFS-mounted from HP-UX systems. On HP-UX systems, it reports sizes that are twice the correct values for files that are NFS-mounted from BSD systems. This is due to a flaw in HP-UX; it also affects the HP-UX du program.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Disk usage   [Contents][Index]

14.3 stat: Report file or file system status

stat displays information about the specified file(s). Synopsis:

stat [option]… [file]…

With no option, stat reports all information about the given files. But it also can be used to report the information of the file systems the given files are located on. If the files are links, stat can also give information about the files the links point to.

Due to shell aliases and built-in stat functions, using an unadorned stat interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env stat …) to avoid interference from the shell.

-L
--dereference

Change how stat treats symbolic links. With this option, stat acts on the file referenced by each symbolic link argument. Without it, stat acts on any symbolic link argument directly.

-f
--file-system

Report information about the file systems where the given files are located instead of information about the files themselves. This option implies the -L option.

-c
--format=format

Use format rather than the default format. format is automatically newline-terminated, so running a command like the following with two or more file operands produces a line of output for each operand:

$ stat --format=%d:%i / /usr
2050:2
2057:2
--printf=format

Use format rather than the default format. Like --format, but interpret backslash escapes, and do not output a mandatory trailing newline. If you want a newline, include ‘\n’ in the format. Here’s how you would use --printf to print the device and inode numbers of / and /usr:

$ stat --printf='%d:%i\n' / /usr
2050:2
2057:2
-t
--terse

Print the information in terse form, suitable for parsing by other programs.

The output of the following commands are identical and the --format also identifies the items printed (in fuller form) in the default format. Note the format string would include another ‘%C’ at the end with an active SELinux security context.

$ stat --format="%n %s %b %f %u %g %D %i %h %t %T %X %Y %Z %W %o" ...
$ stat --terse ...

The same illustrating terse output in --file-system mode:

$ stat -f --format="%n %i %l %t %s %S %b %f %a %c %d" ...
$ stat -f --terse ...

The valid format directives for files with --format and --printf are:

The ‘%t’ and ‘%T’ formats operate on the st_rdev member of the stat(2) structure, and are only defined for character and block special files. On some systems or file types, st_rdev may be used to represent other quantities.

The ‘%W’, ‘%X’, ‘%Y’, and ‘%Z’ formats accept a precision preceded by a period to specify the number of digits to print after the decimal point. For example, ‘%.3X’ outputs the last access time to millisecond precision. If a period is given but no precision, stat uses 9 digits, so ‘%.X’ is equivalent to ‘%.9X’. When discarding excess precision, time stamps are truncated toward minus infinity.

zero pad:
  $ stat -c '[%015Y]' /usr
  [000001288929712]
space align:
  $ stat -c '[%15Y]' /usr
  [     1288929712]
  $ stat -c '[%-15Y]' /usr
  [1288929712     ]
precision:
  $ stat -c '[%.3Y]' /usr
  [1288929712.114]
  $ stat -c '[%.Y]' /usr
  [1288929712.114951834]

The mount point printed by ‘%m’ is similar to that output by df, except that:

When listing file system information (--file-system (-f)), you must use a different set of format directives:

Time stamps are listed according to the time zone rules specified by the TZ environment variable, or by the system default rules if TZ is not set. See Specifying the Time Zone with TZ in The GNU C Library Reference Manual.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Disk usage   [Contents][Index]

14.4 sync: Synchronize cached writes to persistent storage

sync synchronizes in memory files or file systems to persistent storage. Synopsis:

sync [option] [file]…

sync writes any data buffered in memory out to disk. This can include (but is not limited to) modified superblocks, modified inodes, and delayed reads and writes. This must be implemented by the kernel; The sync program does nothing but exercise the sync, syncfs, fsync, and fdatasync system calls.

The kernel keeps data in memory to avoid doing (relatively slow) disk reads and writes. This improves performance, but if the computer crashes, data may be lost or the file system corrupted as a result. The sync command instructs the kernel to write data in memory to persistent storage.

If any argument is specified then only those files will be synchronized using the fsync(2) syscall by default.

If at least one file is specified, it is possible to change the synchronization method with the following options. Also see Common options.

-d
--data

Use fdatasync(2) to sync only the data for the file, and any metadata required to maintain file system consistency.

-f
--file-system

Synchronize all the I/O waiting for the file systems that contain the file, using the syscall syncfs(2). Note you would usually not specify this option if passing a device node like ‘/dev/sda’ for example, as that would sync the containing file system rather than the referenced one. Note also that depending on the system, passing individual device nodes or files may have different sync characteristics than using no arguments. I.e., arguments passed to fsync(2) may provide greater guarantees through write barriers, than a global sync(2) used when no arguments are provided.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Disk usage   [Contents][Index]

14.5 truncate: Shrink or extend the size of a file

truncate shrinks or extends the size of each file to the specified size. Synopsis:

truncate optionfile

Any file that does not exist is created.

If a file is larger than the specified size, the extra data is lost. If a file is shorter, it is extended and the extended part (or hole) reads as zero bytes.

The program accepts the following options. Also see Common options.

-c
--no-create

Do not create files that do not exist.

-o
--io-blocks

Treat size as number of I/O blocks of the file rather than bytes.

-r rfile
--reference=rfile

Base the size of each file on the size of rfile.

-s size
--size=size

Set or adjust the size of each file according to size. size is in bytes unless --io-blocks is specified. size may be, or may be an integer optionally followed by, one of the following multiplicative suffixes:

KB’ =>           1000 (KiloBytes)
‘K’  =>           1024 (KibiBytes)
‘MB’ =>      1000*1000 (MegaBytes)
‘M’  =>      1024*1024 (MebiBytes)
‘GB’ => 1000*1000*1000 (GigaBytes)
‘G’  => 1024*1024*1024 (GibiBytes)

and so on for ‘T’, ‘P’, ‘E’, ‘Z’, and ‘Y’.

size may also be prefixed by one of the following to adjust the size of each file based on its current size:

+’  => extend by
‘-’  => reduce by
‘<’  => at most
‘>’  => at least
‘/’  => round down to multiple of
‘%’  => round up to multiple of

An exit status of zero indicates success, and a nonzero value indicates failure.


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

15 Printing text

This section describes commands that display text strings.


Next: , Up: Printing text   [Contents][Index]

15.1 echo: Print a line of text

echo writes each given string to standard output, with a space between each and a newline after the last one. Synopsis:

echo [option]… [string]…

Due to shell aliases and built-in echo functions, using an unadorned echo interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env echo …) to avoid interference from the shell.

The program accepts the following options. Also see Common options. Options must precede operands, and the normally-special argument ‘--’ has no special meaning and is treated like any other string.

-n

Do not output the trailing newline.

-e

Enable interpretation of the following backslash-escaped characters in each string:

\a

alert (bell)

\b

backspace

\c

produce no further output

\e

escape

\f

form feed

\n

newline

\r

carriage return

\t

horizontal tab

\v

vertical tab

\\

backslash

\0nnn

the eight-bit value that is the octal number nnn (zero to three octal digits), if nnn is a nine-bit value, the ninth bit is ignored

\nnn

the eight-bit value that is the octal number nnn (one to three octal digits), if nnn is a nine-bit value, the ninth bit is ignored

\xhh

the eight-bit value that is the hexadecimal number hh (one or two hexadecimal digits)

-E

Disable interpretation of backslash escapes in each string. This is the default. If -e and -E are both specified, the last one given takes effect.

If the POSIXLY_CORRECT environment variable is set, then when echo’s first argument is not -n it outputs option-like arguments instead of treating them as options. For example, echo -ne hello outputs ‘-ne hello’ instead of plain ‘hello’.

POSIX does not require support for any options, and says that the behavior of echo is implementation-defined if any string contains a backslash or if the first argument is -n. Portable programs can use the printf command if they need to omit trailing newlines or output control characters or backslashes. See printf invocation.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Printing text   [Contents][Index]

15.2 printf: Format and print data

printf does formatted printing of text. Synopsis:

printf format [argument]…

printf prints the format string, interpreting ‘%’ directives and ‘\’ escapes to format numeric and string arguments in a way that is mostly similar to the C ‘printf’ function. See printf format directives in The GNU C Library Reference Manual, for details. The differences are listed below.

Due to shell aliases and built-in printf functions, using an unadorned printf interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env printf …) to avoid interference from the shell.

A floating-point argument must use a period before any fractional digits, but is printed according to the LC_NUMERIC category of the current locale. For example, in a locale whose radix character is a comma, the command ‘printf %g 3.14’ outputs ‘3,14’ whereas the command ‘printf %g 3,14’ is an error. See Floating point.

printf interprets ‘\ooo’ in format as an octal number (if ooo is 1 to 3 octal digits) specifying a byte to print, and ‘\xhh’ as a hexadecimal number (if hh is 1 to 2 hex digits) specifying a character to print. Note however that when ‘\ooo’ specifies a number larger than 255, printf ignores the ninth bit. For example, ‘printf '\400'’ is equivalent to ‘printf '\0'’.

printf interprets two character syntaxes introduced in ISO C 99: ‘\u’ for 16-bit Unicode (ISO/IEC 10646) characters, specified as four hexadecimal digits hhhh, and ‘\U’ for 32-bit Unicode characters, specified as eight hexadecimal digits hhhhhhhh. printf outputs the Unicode characters according to the LC_CTYPE locale. Unicode characters in the ranges U+0000…U+009F, U+D800…U+DFFF cannot be specified by this syntax, except for U+0024 ($), U+0040 (@), and U+0060 ()`.

The processing of ‘\u’ and ‘\U’ requires a full-featured iconv facility. It is activated on systems with glibc 2.2 (or newer), or when libiconv is installed prior to this package. Otherwise ‘\u’ and ‘\U’ will print as-is.

The only options are a lone --help or --version. See Common options. Options must precede operands.

The Unicode character syntaxes are useful for writing strings in a locale independent way. For example, a string containing the Euro currency symbol

$ env printf '\u20AC 14.95'

will be output correctly in all locales supporting the Euro symbol (ISO-8859-15, UTF-8, and others). Similarly, a Chinese string

$ env printf '\u4e2d\u6587'

will be output correctly in all Chinese locales (GB2312, BIG5, UTF-8, etc).

Note that in these examples, the printf command has been invoked via env to ensure that we run the program found via your shell’s search path, and not a shell alias or a built-in function.

For larger strings, you don’t need to look up the hexadecimal code values of each character one by one. ASCII characters mixed with \u escape sequences is also known as the JAVA source file encoding. You can use GNU recode 3.5c (or newer) to convert strings to this encoding. Here is how to convert a piece of text into a shell script which will output this text in a locale-independent way:

$ LC_CTYPE=zh_CN.big5 /usr/local/bin/printf \
    '\u4e2d\u6587\n' > sample.txt
$ recode BIG5..JAVA < sample.txt \
    | sed -e "s|^|/usr/local/bin/printf '|" -e "s|$|\\\\n'|" \
    > sample.sh

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: Printing text   [Contents][Index]

15.3 yes: Print a string until interrupted

yes prints the command line arguments, separated by spaces and followed by a newline, forever until it is killed. If no arguments are given, it prints ‘y’ followed by a newline forever until killed.

Upon a write error, yes exits with status ‘1’.

The only options are a lone --help or --version. To output an argument that begins with ‘-’, precede it with --, e.g., ‘yes -- --help’. See Common options.


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

16 Conditions

This section describes commands that are primarily useful for their exit status, rather than their output. Thus, they are often used as the condition of shell if statements, or as the last command in a pipeline.


Next: , Up: Conditions   [Contents][Index]

16.1 false: Do nothing, unsuccessfully

false does nothing except return an exit status of 1, meaning failure. It can be used as a place holder in shell scripts where an unsuccessful command is needed. In most modern shells, false is a built-in command, so when you use ‘false’ in a script, you’re probably using the built-in command, not the one documented here.

false honors the --help and --version options.

This version of false is implemented as a C program, and is thus more secure and faster than a shell script implementation, and may safely be used as a dummy shell for the purpose of disabling accounts.

Note that false (unlike all other programs documented herein) exits unsuccessfully, even when invoked with --help or --version.

Portable programs should not assume that the exit status of false is 1, as it is greater than 1 on some non-GNU hosts.


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

16.2 true: Do nothing, successfully

true does nothing except return an exit status of 0, meaning success. It can be used as a place holder in shell scripts where a successful command is needed, although the shell built-in command : (colon) may do the same thing faster. In most modern shells, true is a built-in command, so when you use ‘true’ in a script, you’re probably using the built-in command, not the one documented here.

true honors the --help and --version options.

Note, however, that it is possible to cause true to exit with nonzero status: with the --help or --version option, and with standard output already closed or redirected to a file that evokes an I/O error. For example, using a Bourne-compatible shell:

$ ./true --version >&-
./true: write error: Bad file number
$ ./true --version > /dev/full
./true: write error: No space left on device

This version of true is implemented as a C program, and is thus more secure and faster than a shell script implementation, and may safely be used as a dummy shell for the purpose of disabling accounts.


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

16.3 test: Check file types and compare values

test returns a status of 0 (true) or 1 (false) depending on the evaluation of the conditional expression expr. Each part of the expression must be a separate argument.

test has file status checks, string operators, and numeric comparison operators.

test has an alternate form that uses opening and closing square brackets instead a leading ‘test’. For example, instead of ‘test -d /’, you can write ‘[ -d / ]’. The square brackets must be separate arguments; for example, ‘[-d /]’ does not have the desired effect. Since ‘test expr’ and ‘[ expr ]’ have the same meaning, only the former form is discussed below.

Synopses:

test expression
test
[ expression ]
[ ]
[ option

Due to shell aliases and built-in test functions, using an unadorned test interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env test …) to avoid interference from the shell.

If expression is omitted, test returns false. If expression is a single argument, test returns false if the argument is null and true otherwise. The argument can be any string, including strings like ‘-d’, ‘-1’, ‘--’, ‘--help’, and ‘--version’ that most other programs would treat as options. To get help and version information, invoke the commands ‘[ --help’ and ‘[ --version’, without the usual closing brackets. See Common options.

Exit status:

0 if the expression is true,
1 if the expression is false,
2 if an error occurred.

Next: , Up: test invocation   [Contents][Index]

16.3.1 File type tests

These options test for particular types of files. (Everything’s a file, but not all files are the same!)

-b file

True if file exists and is a block special device.

-c file

True if file exists and is a character special device.

-d file

True if file exists and is a directory.

-f file

True if file exists and is a regular file.

-h file
-L file

True if file exists and is a symbolic link. Unlike all other file-related tests, this test does not dereference file if it is a symbolic link.

-p file

True if file exists and is a named pipe.

-S file

True if file exists and is a socket.

-t fd

True if fd is a file descriptor that is associated with a terminal.


Next: , Previous: , Up: test invocation   [Contents][Index]

16.3.2 Access permission tests

These options test for particular access permissions.

-g file

True if file exists and has its set-group-ID bit set.

-k file

True if file exists and has its sticky bit set.

-r file

True if file exists and read permission is granted.

-u file

True if file exists and has its set-user-ID bit set.

-w file

True if file exists and write permission is granted.

-x file

True if file exists and execute permission is granted (or search permission, if it is a directory).

-O file

True if file exists and is owned by the current effective user ID.

-G file

True if file exists and is owned by the current effective group ID.


Next: , Previous: , Up: test invocation   [Contents][Index]

16.3.3 File characteristic tests

These options test other file characteristics.

-e file

True if file exists.

-s file

True if file exists and has a size greater than zero.

file1 -nt file2

True if file1 is newer (according to modification date) than file2, or if file1 exists and file2 does not.

file1 -ot file2

True if file1 is older (according to modification date) than file2, or if file2 exists and file1 does not.

file1 -ef file2

True if file1 and file2 have the same device and inode numbers, i.e., if they are hard links to each other.


Next: , Previous: , Up: test invocation   [Contents][Index]

16.3.4 String tests

These options test string characteristics. You may need to quote string arguments for the shell. For example:

test -n "$V"

The quotes here prevent the wrong arguments from being passed to test if ‘$V’ is empty or contains special characters.

-z string

True if the length of string is zero.

-n string
string

True if the length of string is nonzero.

string1 = string2

True if the strings are equal.

string1 == string2

True if the strings are equal (synonym for =).

string1 != string2

True if the strings are not equal.


Next: , Previous: , Up: test invocation   [Contents][Index]

16.3.5 Numeric tests

Numeric relational operators. The arguments must be entirely numeric (possibly negative), or the special expression -l string, which evaluates to the length of string.

arg1 -eq arg2
arg1 -ne arg2
arg1 -lt arg2
arg1 -le arg2
arg1 -gt arg2
arg1 -ge arg2

These arithmetic binary operators return true if arg1 is equal, not-equal, less-than, less-than-or-equal, greater-than, or greater-than-or-equal than arg2, respectively.

For example:

test -1 -gt -2 && echo yes
⇒ yes
test -l abc -gt 1 && echo yes
⇒ yes
test 0x100 -eq 1
error→ test: integer expression expected before -eq

Previous: , Up: test invocation   [Contents][Index]

16.3.6 Connectives for test

The usual logical connectives.

! expr

True if expr is false.

expr1 -a expr2

True if both expr1 and expr2 are true.

expr1 -o expr2

True if either expr1 or expr2 is true.


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

16.4 expr: Evaluate expressions

expr evaluates an expression and writes the result on standard output. Each token of the expression must be a separate argument.

Operands are either integers or strings. Integers consist of one or more decimal digits, with an optional leading ‘-’. expr converts anything appearing in an operand position to an integer or a string depending on the operation being applied to it.

Strings are not quoted for expr itself, though you may need to quote them to protect characters with special meaning to the shell, e.g., spaces. However, regardless of whether it is quoted, a string operand should not be a parenthesis or any of expr’s operators like +, so you cannot safely pass an arbitrary string $str to expr merely by quoting it to the shell. One way to work around this is to use the GNU extension +, (e.g., + "$str" = foo); a more portable way is to use " $str" and to adjust the rest of the expression to take the leading space into account (e.g., " $str" = " foo").

You should not pass a negative integer or a string with leading ‘-’ as expr’s first argument, as it might be misinterpreted as an option; this can be avoided by parenthesization. Also, portable scripts should not use a string operand that happens to take the form of an integer; this can be worked around by inserting leading spaces as mentioned above.

Operators may be given as infix symbols or prefix keywords. Parentheses may be used for grouping in the usual manner. You must quote parentheses and many operators to avoid the shell evaluating them, however.

When built with support for the GNU MP library, expr uses arbitrary-precision arithmetic; otherwise, it uses native arithmetic types and may fail due to arithmetic overflow.

The only options are --help and --version. See Common options. Options must precede operands.

Exit status:

0 if the expression is neither null nor 0,
1 if the expression is null or 0,
2 if the expression is invalid,
3 if an internal error occurred (e.g., arithmetic overflow).

Next: , Up: expr invocation   [Contents][Index]

16.4.1 String expressions

expr supports pattern matching and other string operators. These have higher precedence than both the numeric and relational operators (in the next sections).

string : regex

Perform pattern matching. The arguments are converted to strings and the second is considered to be a (basic, a la GNU grep) regular expression, with a ^ implicitly prepended. The first argument is then matched against this regular expression.

If the match succeeds and regex uses ‘\(’ and ‘\)’, the : expression returns the part of string that matched the subexpression; otherwise, it returns the number of characters matched.

If the match fails, the : operator returns the null string if ‘\(’ and ‘\)’ are used in regex, otherwise 0.

Only the first ‘\( … \)’ pair is relevant to the return value; additional pairs are meaningful only for grouping the regular expression operators.

In the regular expression, \+, \?, and \| are operators which respectively match one or more, zero or one, or separate alternatives. SunOS and other expr’s treat these as regular characters. (POSIX allows either behavior.) See Regular Expression Library in Regex, for details of regular expression syntax. Some examples are in Examples of expr.

match string regex

An alternative way to do pattern matching. This is the same as ‘string : regex.

substr string position length

Returns the substring of string beginning at position with length at most length. If either position or length is negative, zero, or non-numeric, returns the null string.

index string charset

Returns the first position in string where the first character in charset was found. If no character in charset is found in string, return 0.

length string

Returns the length of string.

+ token

Interpret token as a string, even if it is a keyword like match or an operator like /. This makes it possible to test expr length + "$x" or expr + "$x" : '.*/\(.\)' and have it do the right thing even if the value of $x happens to be (for example) / or index. This operator is a GNU extension. Portable shell scripts should use " $token" : ' \(.*\)' instead of + "$token".

To make expr interpret keywords as strings, you must use the quote operator.


Next: , Previous: , Up: expr invocation   [Contents][Index]

16.4.2 Numeric expressions

expr supports the usual numeric operators, in order of increasing precedence. These numeric operators have lower precedence than the string operators described in the previous section, and higher precedence than the connectives (next section).

+ -

Addition and subtraction. Both arguments are converted to integers; an error occurs if this cannot be done.

* / %

Multiplication, division, remainder. Both arguments are converted to integers; an error occurs if this cannot be done.


Next: , Previous: , Up: expr invocation   [Contents][Index]

16.4.3 Relations for expr

expr supports the usual logical connectives and relations. These have lower precedence than the string and numeric operators (previous sections). Here is the list, lowest-precedence operator first.

|

Returns its first argument if that is neither null nor zero, otherwise its second argument if it is neither null nor zero, otherwise 0. It does not evaluate its second argument if its first argument is neither null nor zero.

&

Return its first argument if neither argument is null or zero, otherwise 0. It does not evaluate its second argument if its first argument is null or zero.

< <= = == != >= >

Compare the arguments and return 1 if the relation is true, 0 otherwise. == is a synonym for =. expr first tries to convert both arguments to integers and do a numeric comparison; if either conversion fails, it does a lexicographic comparison using the character collating sequence specified by the LC_COLLATE locale.


Previous: , Up: expr invocation   [Contents][Index]

16.4.4 Examples of using expr

Here are a few examples, including quoting for shell metacharacters.

To add 1 to the shell variable foo, in Bourne-compatible shells:

foo=$(expr $foo + 1)

To print the non-directory part of the file name stored in $fname, which need not contain a /:

expr $fname : '.*/\(.*\)' '|' $fname

An example showing that \+ is an operator:

expr aaa : 'a\+'
⇒ 3
expr abc : 'a\(.\)c'
⇒ b
expr index abcdef cz
⇒ 3
expr index index a
error→ expr: syntax error
expr index + index a
⇒ 0

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

17 Redirection

Unix shells commonly provide several forms of redirection—ways to change the input source or output destination of a command. But one useful redirection is performed by a separate command, not by the shell; it’s described here.


Up: Redirection   [Contents][Index]

17.1 tee: Redirect output to multiple files or processes

The tee command copies standard input to standard output and also to any files given as arguments. This is useful when you want not only to send some data down a pipe, but also to save a copy. Synopsis:

tee [option]… [file]…

If a file being written to does not already exist, it is created. If a file being written to already exists, the data it previously contained is overwritten unless the -a option is used.

In previous versions of GNU coreutils (v5.3.0 - v8.23), a file of ‘-’ caused tee to send another copy of input to standard output. However, as the interleaved output was not very useful, tee now conforms to POSIX which explicitly mandates it to treat ‘-’ as a file with such name.

The program accepts the following options. Also see Common options.

-a
--append

Append standard input to the given files rather than overwriting them.

-i
--ignore-interrupts

Ignore interrupt signals.

-p
--output-error[=mode]

Select the behavior with errors on the outputs, where mode is one of the following:

warn

Warn on error opening or writing any output, including pipes. Writing is continued to still open files/pipes. Exit status indicates failure if any output has an error.

warn-nopipe

Warn on error opening or writing any output, except pipes. Writing is continued to still open files/pipes. Exit status indicates failure if any non pipe output had an error. This is the default mode when not specified.

exit

Exit on error opening or writing any output, including pipes.

exit-nopipe

Exit on error opening or writing any output, except pipes.

The tee command is useful when you happen to be transferring a large amount of data and also want to summarize that data without reading it a second time. For example, when you are downloading a DVD image, you often want to verify its signature or checksum right away. The inefficient way to do it is simply:

wget http://example.com/some.iso && sha1sum some.iso

One problem with the above is that it makes you wait for the download to complete before starting the time-consuming SHA1 computation. Perhaps even more importantly, the above requires reading the DVD image a second time (the first was from the network).

The efficient way to do it is to interleave the download and SHA1 computation. Then, you’ll get the checksum for free, because the entire process parallelizes so well:

# slightly contrived, to demonstrate process substitution
wget -O - http://example.com/dvd.iso \
  | tee >(sha1sum > dvd.sha1) > dvd.iso

That makes tee write not just to the expected output file, but also to a pipe running sha1sum and saving the final checksum in a file named dvd.sha1.

Note, however, that this example relies on a feature of modern shells called process substitution (the ‘>(command)’ syntax, above; See Process Substitution in The Bash Reference Manual.), so it works with zsh, bash, and ksh, but not with /bin/sh. So if you write code like this in a shell script, be sure to start the script with ‘#!/bin/bash’.

Since the above example writes to one file and one process, a more conventional and portable use of tee is even better:

wget -O - http://example.com/dvd.iso \
  | tee dvd.iso | sha1sum > dvd.sha1

You can extend this example to make tee write to two processes, computing MD5 and SHA1 checksums in parallel. In this case, process substitution is required:

wget -O - http://example.com/dvd.iso \
  | tee >(sha1sum > dvd.sha1) \
        >(md5sum > dvd.md5) \
  > dvd.iso

This technique is also useful when you want to make a compressed copy of the contents of a pipe. Consider a tool to graphically summarize disk usage data from ‘du -ak’. For a large hierarchy, ‘du -ak’ can run for a long time, and can easily produce terabytes of data, so you won’t want to rerun the command unnecessarily. Nor will you want to save the uncompressed output.

Doing it the inefficient way, you can’t even start the GUI until after you’ve compressed all of the du output:

du -ak | gzip -9 > /tmp/du.gz
gzip -d /tmp/du.gz | xdiskusage -a

With tee and process substitution, you start the GUI right away and eliminate the decompression completely:

du -ak | tee >(gzip -9 > /tmp/du.gz) | xdiskusage -a

Finally, if you regularly create more than one type of compressed tarball at once, for example when make dist creates both gzip-compressed and bzip2-compressed tarballs, there may be a better way. Typical automake-generated Makefile rules create the two compressed tar archives with commands in sequence, like this (slightly simplified):

tardir=your-pkg-M.N
tar chof - "$tardir" | gzip  -9 -c > your-pkg-M.N.tar.gz
tar chof - "$tardir" | bzip2 -9 -c > your-pkg-M.N.tar.bz2

However, if the hierarchy you are archiving and compressing is larger than a couple megabytes, and especially if you are using a multi-processor system with plenty of memory, then you can do much better by reading the directory contents only once and running the compression programs in parallel:

tardir=your-pkg-M.N
tar chof - "$tardir" \
  | tee >(gzip -9 -c > your-pkg-M.N.tar.gz) \
  | bzip2 -9 -c > your-pkg-M.N.tar.bz2

An exit status of zero indicates success, and a nonzero value indicates failure.


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

18 File name manipulation

This section describes commands that manipulate file names.


Next: , Up: File name manipulation   [Contents][Index]

18.1 basename: Strip directory and suffix from a file name

basename removes any leading directory components from name. Synopsis:

basename name [suffix]
basename optionname

If suffix is specified and is identical to the end of name, it is removed from name as well. Note that since trailing slashes are removed prior to suffix matching, suffix will do nothing if it contains slashes. basename prints the result on standard output.

Together, basename and dirname are designed such that if ‘ls "$name"’ succeeds, then the command sequence ‘cd "$(dirname "$name")"; ls "$(basename "$name")"’ will, too. This works for everything except file names containing a trailing newline.

POSIX allows the implementation to define the results if name is empty or ‘//’. In the former case, GNU basename returns the empty string. In the latter case, the result is ‘//’ on platforms where // is distinct from /, and ‘/’ on platforms where there is no difference.

The program accepts the following options. Also see Common options. Options must precede operands.

-a
--multiple

Support more than one argument. Treat every argument as a name. With this, an optional suffix must be specified using the -s option.

-s suffix
--suffix=suffix

Remove a trailing suffix. This option implies the -a option.

-z
--zero

Output a zero byte (ASCII NUL) at the end of each line, rather than a newline. This option enables other programs to parse the output even when that output would contain data with embedded newlines.

An exit status of zero indicates success, and a nonzero value indicates failure.

Examples:

# Output "sort".
basename /usr/bin/sort

# Output "stdio".
basename include/stdio.h .h

# Output "stdio".
basename -s .h include/stdio.h

# Output "stdio" followed by "stdlib"
basename -a -s .h include/stdio.h include/stdlib.h

Next: , Previous: , Up: File name manipulation   [Contents][Index]

18.2 dirname: Strip last file name component

dirname prints all but the final slash-delimited component of each name. Slashes on either side of the final component are also removed. If the string contains no slash, dirname prints ‘.’ (meaning the current directory). Synopsis:

dirname [option] name

name need not be a file name, but if it is, this operation effectively lists the directory that contains the final component, including the case when the final component is itself a directory.

Together, basename and dirname are designed such that if ‘ls "$name"’ succeeds, then the command sequence ‘cd "$(dirname "$name")"; ls "$(basename "$name")"’ will, too. This works for everything except file names containing a trailing newline.

POSIX allows the implementation to define the results if name is ‘//’. With GNU dirname, the result is ‘//’ on platforms where // is distinct from /, and ‘/’ on platforms where there is no difference.

The program accepts the following option. Also see Common options.

-z
--zero

Output a zero byte (ASCII NUL) at the end of each line, rather than a newline. This option enables other programs to parse the output even when that output would contain data with embedded newlines.

An exit status of zero indicates success, and a nonzero value indicates failure.

Examples:

# Output "/usr/bin".
dirname /usr/bin/sort
dirname /usr/bin//.//

# Output "dir1" followed by "dir2"
dirname dir1/str dir2/str

# Output ".".
dirname stdio.h

Next: , Previous: , Up: File name manipulation   [Contents][Index]

18.3 pathchk: Check file name validity and portability

pathchk checks validity and portability of file names. Synopsis:

pathchk [option]… name

For each name, pathchk prints an error message if any of these conditions is true:

  1. One of the existing directories in name does not have search (execute) permission,
  2. The length of name is larger than the maximum supported by the operating system.
  3. The length of one component of name is longer than its file system’s maximum.

A nonexistent name is not an error, so long a file with that name could be created under the above conditions.

The program accepts the following options. Also see Common options. Options must precede operands.

-p

Instead of performing checks based on the underlying file system, print an error message if any of these conditions is true:

  1. A file name is empty.
  2. A file name contains a character outside the POSIX portable file name character set, namely, the ASCII letters and digits, ‘.’, ‘_’, ‘-’, and ‘/’.
  3. The length of a file name or one of its components exceeds the POSIX minimum limits for portability.
-P

Print an error message if a file name is empty, or if it contains a component that begins with ‘-’.

--portability

Print an error message if a file name is not portable to all POSIX hosts. This option is equivalent to ‘-p -P’.

Exit status:

0 if all specified file names passed all checks,
1 otherwise.

Next: , Previous: , Up: File name manipulation   [Contents][Index]

18.4 mktemp: Create temporary file or directory

mktemp manages the creation of temporary files and directories. Synopsis:

mktemp [option]… [template]

Safely create a temporary file or directory based on template, and print its name. If given, template must include at least three consecutive ‘X’s in the last component. If omitted, the template ‘tmp.XXXXXXXXXX’ is used, and option --tmpdir is implied. The final run of ‘X’s in the template will be replaced by alpha-numeric characters; thus, on a case-sensitive file system, and with a template including a run of n instances of ‘X’, there are ‘62**n’ potential file names.

Older scripts used to create temporary files by simply joining the name of the program with the process id (‘$$’) as a suffix. However, that naming scheme is easily predictable, and suffers from a race condition where the attacker can create an appropriately named symbolic link, such that when the script then opens a handle to what it thought was an unused file, it is instead modifying an existing file. Using the same scheme to create a directory is slightly safer, since the mkdir will fail if the target already exists, but it is still inferior because it allows for denial of service attacks. Therefore, modern scripts should use the mktemp command to guarantee that the generated name will be unpredictable, and that knowledge of the temporary file name implies that the file was created by the current script and cannot be modified by other users.

When creating a file, the resulting file has read and write permissions for the current user, but no permissions for the group or others; these permissions are reduced if the current umask is more restrictive.

Here are some examples (although note that if you repeat them, you will most likely get different file names):

The program accepts the following options. Also see Common options.

-d
--directory

Create a directory rather than a file. The directory will have read, write, and search permissions for the current user, but no permissions for the group or others; these permissions are reduced if the current umask is more restrictive.

-q
--quiet

Suppress diagnostics about failure to create a file or directory. The exit status will still reflect whether a file was created.

-u
--dry-run

Generate a temporary name that does not name an existing file, without changing the file system contents. Using the output of this command to create a new file is inherently unsafe, as there is a window of time between generating the name and using it where another process can create an object by the same name.

-p dir
--tmpdir[=dir]

Treat template relative to the directory dir. If dir is not specified (only possible with the long option --tmpdir) or is the empty string, use the value of TMPDIR if available, otherwise use ‘/tmp’. If this is specified, template must not be absolute. However, template can still contain slashes, although intermediate directories must already exist.

--suffix=suffix

Append suffix to the template. suffix must not contain slash. If --suffix is specified, template must end in ‘X’; if it is not specified, then an appropriate --suffix is inferred by finding the last ‘X’ in template. This option exists for use with the default template and for the creation of a suffix that starts with ‘X’.

-t

Treat template as a single file relative to the value of TMPDIR if available, or to the directory specified by -p, otherwise to ‘/tmp’. template must not contain slashes. This option is deprecated; the use of -p without -t offers better defaults (by favoring the command line over TMPDIR) and more flexibility (by allowing intermediate directories).

Exit status:

0 if the file was created,
1 otherwise.

Previous: , Up: File name manipulation   [Contents][Index]

18.5 realpath: Print the resolved file name.

realpath expands all symbolic links and resolves references to ‘/./’, ‘/../’ and extra ‘/’ characters. By default, all but the last component of the specified files must exist. Synopsis:

realpath [option]… file

The program accepts the following options. Also see Common options.

-e
--canonicalize-existing

Ensure that all components of the specified file names exist. If any component is missing or unavailable, realpath will output a diagnostic unless the -q option is specified, and exit with a nonzero exit code. A trailing slash requires that the name resolve to a directory.

-m
--canonicalize-missing

If any component of a specified file name is missing or unavailable, treat it as a directory.

-L
--logical

Symbolic links are resolved in the specified file names, but they are resolved after any subsequent ‘..’ components are processed.

-P
--physical

Symbolic links are resolved in the specified file names, and they are resolved before any subsequent ‘..’ components are processed. This is the default mode of operation.

-q
--quiet

Suppress diagnostic messages for specified file names.

--relative-to=file

Print the resolved file names relative to the specified file. Note this option honors the -m and -e options pertaining to file existence.

--relative-base=base

This option is valid when used with --relative-to, and will restrict the output of --relative-to so that relative names are output, only when files are descendants of base. Otherwise output the absolute file name. If --relative-to was not specified, then the descendants of base are printed relative to base. If --relative-to is specified, then that directory must be a descendant of base for this option to have an effect. Note: this option honors the -m and -e options pertaining to file existence. For example:

realpath --relative-to=/usr /tmp /usr/bin
⇒ ../tmp
⇒ bin
realpath --relative-base=/usr /tmp /usr/bin
⇒ /tmp
⇒ bin
-s
--strip
--no-symlinks

Do not resolve symbolic links. Only resolve references to ‘/./’, ‘/../’ and remove extra ‘/’ characters. When combined with the -m option, realpath operates only on the file name, and does not touch any actual file.

-z
--zero

Output a zero byte (ASCII NUL) at the end of each line, rather than a newline. This option enables other programs to parse the output even when that output would contain data with embedded newlines.

Exit status:

0 if all file names were printed without issue.
1 otherwise.

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

19 Working context

This section describes commands that display or alter the context in which you are working: the current directory, the terminal settings, and so forth. See also the user-related commands in the next section.


Next: , Up: Working context   [Contents][Index]

19.1 pwd: Print working directory

pwd prints the name of the current directory. Synopsis:

pwd [option]…

The program accepts the following options. Also see Common options.

-L
--logical

If the contents of the environment variable PWD provide an absolute name of the current directory with no ‘.’ or ‘..’ components, but possibly with symbolic links, then output those contents. Otherwise, fall back to default -P handling.

-P
--physical

Print a fully resolved name for the current directory. That is, all components of the printed name will be actual directory names—none will be symbolic links.

If -L and -P are both given, the last one takes precedence. If neither option is given, then this implementation uses -P as the default unless the POSIXLY_CORRECT environment variable is set.

Due to shell aliases and built-in pwd functions, using an unadorned pwd interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env pwd …) to avoid interference from the shell.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Working context   [Contents][Index]

19.2 stty: Print or change terminal characteristics

stty prints or changes terminal characteristics, such as baud rate. Synopses:

stty [option] [setting]…
stty [option]

If given no line settings, stty prints the baud rate, line discipline number (on systems that support it), and line settings that have been changed from the values set by ‘stty sane’. By default, mode reading and setting are performed on the tty line connected to standard input, although this can be modified by the --file option.

stty accepts many non-option arguments that change aspects of the terminal line operation, as described below.

The program accepts the following options. Also see Common options.

-a
--all

Print all current settings in human-readable form. This option may not be used in combination with any line settings.

-F device
--file=device

Set the line opened by the file name specified in device instead of the tty line connected to standard input. This option is necessary because opening a POSIX tty requires use of the O_NONDELAY flag to prevent a POSIX tty from blocking until the carrier detect line is high if the clocal flag is not set. Hence, it is not always possible to allow the shell to open the device in the traditional manner.

-g
--save

Print all current settings in a form that can be used as an argument to another stty command to restore the current settings. This option may not be used in combination with any line settings.

Many settings can be turned off by preceding them with a ‘-’. Such arguments are marked below with “May be negated” in their description. The descriptions themselves refer to the positive case, that is, when not negated (unless stated otherwise, of course).

Some settings are not available on all POSIX systems, since they use extensions. Such arguments are marked below with “Non-POSIX” in their description. On non-POSIX systems, those or other settings also may not be available, but it’s not feasible to document all the variations: just try it and see.

stty is installed only on platforms with the POSIX terminal interface, so portable scripts should not rely on its existence on non-POSIX platforms.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Up: stty invocation   [Contents][Index]

19.2.1 Control settings

Control settings:

parenb

Generate parity bit in output and expect parity bit in input. May be negated.

parodd

Set odd parity (even if negated). May be negated.

cmspar

Use "stick" (mark/space) parity. If parodd is set, the parity bit is always 1; if parodd is not set, the parity bit is always zero. Non-POSIX. May be negated.

cs5
cs6
cs7
cs8

Set character size to 5, 6, 7, or 8 bits.

hup
hupcl

Send a hangup signal when the last process closes the tty. May be negated.

cstopb

Use two stop bits per character (one if negated). May be negated.

cread

Allow input to be received. May be negated.

clocal

Disable modem control signals. May be negated.

crtscts

Enable RTS/CTS flow control. Non-POSIX. May be negated.

cdtrdsr

Enable DTR/DSR flow control. Non-POSIX. May be negated.


Next: , Previous: , Up: stty invocation   [Contents][Index]

19.2.2 Input settings

These settings control operations on data received from the terminal.

ignbrk

Ignore break characters. May be negated.

brkint

Make breaks cause an interrupt signal. May be negated.

ignpar

Ignore characters with parity errors. May be negated.

parmrk

Mark parity errors (with a 255-0-character sequence). May be negated.

inpck

Enable input parity checking. May be negated.

istrip

Clear high (8th) bit of input characters. May be negated.

inlcr

Translate newline to carriage return. May be negated.

igncr

Ignore carriage return. May be negated.

icrnl

Translate carriage return to newline. May be negated.

iutf8

Assume input characters are UTF-8 encoded. May be negated.

ixon

Enable XON/XOFF flow control (that is, Ctrl-S/Ctrl-Q). May be negated.

ixoff
tandem

Enable sending of stop character when the system input buffer is almost full, and start character when it becomes almost empty again. May be negated.

iuclc

Translate uppercase characters to lowercase. Non-POSIX. May be negated. Note ilcuc is not implemented, as one would not be able to issue almost any (lowercase) Unix command, after invoking it.

ixany

Allow any character to restart output (only the start character if negated). Non-POSIX. May be negated.

imaxbel

Enable beeping and not flushing input buffer if a character arrives when the input buffer is full. Non-POSIX. May be negated.


Next: , Previous: , Up: stty invocation   [Contents][Index]

19.2.3 Output settings

These settings control operations on data sent to the terminal.

opost

Postprocess output. May be negated.

olcuc

Translate lowercase characters to uppercase. Non-POSIX. May be negated. (Note ouclc is not currently implemented.)

ocrnl

Translate carriage return to newline. Non-POSIX. May be negated.

onlcr

Translate newline to carriage return-newline. Non-POSIX. May be negated.

onocr

Do not print carriage returns in the first column. Non-POSIX. May be negated.

onlret

Newline performs a carriage return. Non-POSIX. May be negated.

ofill

Use fill (padding) characters instead of timing for delays. Non-POSIX. May be negated.

ofdel

Use ASCII DEL characters for fill instead of ASCII NUL characters. Non-POSIX. May be negated.

nl1
nl0

Newline delay style. Non-POSIX.

cr3
cr2
cr1
cr0

Carriage return delay style. Non-POSIX.

tab3
tab2
tab1
tab0

Horizontal tab delay style. Non-POSIX.

bs1
bs0

Backspace delay style. Non-POSIX.

vt1
vt0

Vertical tab delay style. Non-POSIX.

ff1
ff0

Form feed delay style. Non-POSIX.


Next: , Previous: , Up: stty invocation   [Contents][Index]

19.2.4 Local settings

isig

Enable interrupt, quit, and suspend special characters. May be negated.

icanon

Enable erase, kill, werase, and rprnt special characters. May be negated.

iexten

Enable non-POSIX special characters. May be negated.

echo

Echo input characters. May be negated.

echoe
crterase

Echo erase characters as backspace-space-backspace. May be negated.

echok

Echo a newline after a kill character. May be negated.

echonl

Echo newline even if not echoing other characters. May be negated.

noflsh

Disable flushing after interrupt and quit special characters. May be negated.

xcase

Enable input and output of uppercase characters by preceding their lowercase equivalents with ‘\’, when icanon is set. Non-POSIX. May be negated.

tostop

Stop background jobs that try to write to the terminal. Non-POSIX. May be negated.

echoprt
prterase

Echo erased characters backward, between ‘\’ and ‘/’. Non-POSIX. May be negated.

echoctl
ctlecho

Echo control characters in hat notation (‘^c’) instead of literally. Non-POSIX. May be negated.

echoke
crtkill

Echo the kill special character by erasing each character on the line as indicated by the echoprt and echoe settings, instead of by the echoctl and echok settings. Non-POSIX. May be negated.

extproc

Enable ‘LINEMODE’, which is used to avoid echoing each character over high latency links. See also ftp://ftp.rfc-editor.org/in-notes/rfc1116.txt">Internet RFC 1116. Non-POSIX. May be negated.


Next: , Previous: , Up: stty invocation   [Contents][Index]

19.2.5 Combination settings

Combination settings:

evenp
parity

Same as parenb -parodd cs7. May be negated. If negated, same as -parenb cs8.

oddp

Same as parenb parodd cs7. May be negated. If negated, same as -parenb cs8.

nl

Same as -icrnl -onlcr. May be negated. If negated, same as icrnl -inlcr -igncr onlcr -ocrnl -onlret.

ek

Reset the erase and kill special characters to their default values.

sane

Same as:

cread -ignbrk brkint -inlcr -igncr icrnl
icanon iexten echo echoe echok -echonl -noflsh
-ixoff -iutf8 -iuclc -ixany imaxbel -xcase -olcuc -ocrnl
opost -ofill onlcr -onocr -onlret nl0 cr0 tab0 bs0 vt0 ff0
isig -tostop -ofdel -echoprt echoctl echoke -extproc

and also sets all special characters to their default values.

cooked

Same as brkint ignpar istrip icrnl ixon opost isig icanon, plus sets the eof and eol characters to their default values if they are the same as the min and time characters. May be negated. If negated, same as raw.

raw

Same as:

-ignbrk -brkint -ignpar -parmrk -inpck -istrip
-inlcr -igncr -icrnl -ixon -ixoff -icanon -opost
-isig -iuclc -ixany -imaxbel -xcase min 1 time 0

May be negated. If negated, same as cooked.

cbreak

Same as -icanon. May be negated. If negated, same as icanon.

pass8

Same as -parenb -istrip cs8. May be negated. If negated, same as parenb istrip cs7.

litout

Same as -parenb -istrip -opost cs8. May be negated. If negated, same as parenb istrip opost cs7.

decctlq

Same as -ixany. Non-POSIX. May be negated.

tabs

Same as tab0. Non-POSIX. May be negated. If negated, same as tab3.

lcase
LCASE

Same as xcase iuclc olcuc. Non-POSIX. May be negated. (Used for terminals with uppercase characters only.)

crt

Same as echoe echoctl echoke.

dec

Same as echoe echoctl echoke -ixany intr ^C erase ^? kill C-u.


Next: , Previous: , Up: stty invocation   [Contents][Index]

19.2.6 Special characters

The special characters’ default values vary from system to system. They are set with the syntax ‘name value’, where the names are listed below and the value can be given either literally, in hat notation (‘^c’), or as an integer which may start with ‘0x’ to indicate hexadecimal, ‘0’ to indicate octal, or any other digit to indicate decimal.

For GNU stty, giving a value of ^- or undef disables that special character. (This is incompatible with Ultrix stty, which uses a value of ‘u’ to disable a special character. GNU stty treats a value ‘u’ like any other, namely to set that special character to U.)

intr

Send an interrupt signal.

quit

Send a quit signal.

erase

Erase the last character typed.

kill

Erase the current line.

eof

Send an end of file (terminate the input).

eol

End the line.

eol2

Alternate character to end the line. Non-POSIX.

discard

Alternate character to toggle discarding of output. Non-POSIX.

swtch

Switch to a different shell layer. Non-POSIX.

status

Send an info signal. Not currently supported on Linux. Non-POSIX.

start

Restart the output after stopping it.

stop

Stop the output.

susp

Send a terminal stop signal.

dsusp

Send a terminal stop signal after flushing the input. Non-POSIX.

rprnt

Redraw the current line. Non-POSIX.

werase

Erase the last word typed. Non-POSIX.

lnext

Enter the next character typed literally, even if it is a special character. Non-POSIX.


Previous: , Up: stty invocation   [Contents][Index]

19.2.7 Special settings

min n

Set the minimum number of characters that will satisfy a read until the time value has expired, when -icanon is set.

time n

Set the number of tenths of a second before reads time out if the minimum number of characters have not been read, when -icanon is set.

ispeed n

Set the input speed to n.

ospeed n

Set the output speed to n.

rows n

Tell the tty kernel driver that the terminal has n rows. Non-POSIX.

cols n
columns n

Tell the kernel that the terminal has n columns. Non-POSIX.

size

Print the number of rows and columns that the kernel thinks the terminal has. (Systems that don’t support rows and columns in the kernel typically use the environment variables LINES and COLUMNS instead; however, GNU stty does not know anything about them.) Non-POSIX.

line n

Use line discipline n. Non-POSIX.

speed

Print the terminal speed.

n

Set the input and output speeds to n. n can be one of: 0 50 75 110 134 134.5 150 200 300 600 1200 1800 2400 4800 9600 19200 38400 exta extb. exta is the same as 19200; extb is the same as 38400. Many systems, including GNU/Linux, support higher speeds. The stty command includes support for speeds of 57600, 115200, 230400, 460800, 500000, 576000, 921600, 1000000, 1152000, 1500000, 2000000, 2500000, 3000000, 3500000, or 4000000 where the system supports these. 0 hangs up the line if -clocal is set.


Next: , Previous: , Up: Working context   [Contents][Index]

19.3 printenv: Print all or some environment variables

printenv prints environment variable values. Synopsis:

printenv [option] [variable]…

If no variables are specified, printenv prints the value of every environment variable. Otherwise, it prints the value of each variable that is set, and nothing for those that are not set.

The program accepts the following option. Also see Common options.

-0
--null

Output a zero byte (ASCII NUL) at the end of each line, rather than a newline. This option enables other programs to parse the output even when that output would contain data with embedded newlines.

Exit status:

0 if all variables specified were found
1 if at least one specified variable was not found
2 if a write error occurred

Previous: , Up: Working context   [Contents][Index]

19.4 tty: Print file name of terminal on standard input

tty prints the file name of the terminal connected to its standard input. It prints ‘not a tty’ if standard input is not a terminal. Synopsis:

tty [option]…

The program accepts the following option. Also see Common options.

-s
--silent
--quiet

Print nothing; only return an exit status.

Exit status:

0 if standard input is a terminal
1 if standard input is not a terminal
2 if given incorrect arguments
3 if a write error occurs

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

20 User information

This section describes commands that print user-related information: logins, groups, and so forth.


Next: , Up: User information   [Contents][Index]

20.1 id: Print user identity

id prints information about the given user, or the process running it if no user is specified. Synopsis:

id [option]… [user]

user can be either a user ID or a name, with name look-up taking precedence unless the ID is specified with a leading ‘+’. See Disambiguating names and IDs.

By default, it prints the real user ID, real group ID, effective user ID if different from the real user ID, effective group ID if different from the real group ID, and supplemental group IDs. In addition, if SELinux is enabled and the POSIXLY_CORRECT environment variable is not set, then print ‘context=c’, where c is the security context.

Each of these numeric values is preceded by an identifying string and followed by the corresponding user or group name in parentheses.

The options cause id to print only part of the above information. Also see Common options.

-g
--group

Print only the group ID.

-G
--groups

Print only the group ID and the supplementary groups.

-n
--name

Print the user or group name instead of the ID number. Requires -u, -g, or -G.

-r
--real

Print the real, instead of effective, user or group ID. Requires -u, -g, or -G.

-u
--user

Print only the user ID.

-Z
--context

Print only the security context of the process, which is generally the user’s security context inherited from the parent process. If neither SELinux or SMACK is enabled then print a warning and set the exit status to 1.

-z
--zero

Delimit output items with NUL characters. This option is not permitted when using the default format.

Example:

$ id -Gn --zero
users <NUL> devs <NUL>

Primary and supplementary groups for a process are normally inherited from its parent and are usually unchanged since login. This means that if you change the group database after logging in, id will not reflect your changes within your existing login session. Running id with a user argument causes the user and group database to be consulted afresh, and so will give a different result.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: User information   [Contents][Index]

20.2 logname: Print current login name

logname prints the calling user’s name, as found in a system-maintained file (often /var/run/utmp or /etc/utmp), and exits with a status of 0. If there is no entry for the calling process, logname prints an error message and exits with a status of 1.

The only options are --help and --version. See Common options.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: User information   [Contents][Index]

20.3 whoami: Print effective user ID

whoami prints the user name associated with the current effective user ID. It is equivalent to the command ‘id -un’.

The only options are --help and --version. See Common options.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: User information   [Contents][Index]

20.4 groups: Print group names a user is in

groups prints the names of the primary and any supplementary groups for each given username, or the current process if no names are given. If more than one name is given, the name of each user is printed before the list of that user’s groups and the user name is separated from the group list by a colon. Synopsis:

groups [username]…

The group lists are equivalent to the output of the command ‘id -Gn’.

The only options are --help and --version. See Common options.

Primary and supplementary groups for a process are normally inherited from its parent and are usually unchanged since login. This means that if you change the group database after logging in, groups will not reflect your changes within your existing login session. Running groups with a list of users causes the user and group database to be consulted afresh, and so will give a different result.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: User information   [Contents][Index]

20.5 users: Print login names of users currently logged in

users prints on a single line a blank-separated list of user names of users currently logged in to the current host. Each user name corresponds to a login session, so if a user has more than one login session, that user’s name will appear the same number of times in the output. Synopsis:

users [file]

With no file argument, users extracts its information from a system-maintained file (often /var/run/utmp or /etc/utmp). If a file argument is given, users uses that file instead. A common choice is /var/log/wtmp.

The only options are --help and --version. See Common options.

The users command is installed only on platforms with the POSIX <utmpx.h> include file or equivalent, so portable scripts should not rely on its existence on non-POSIX platforms.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: User information   [Contents][Index]

20.6 who: Print who is currently logged in

who prints information about users who are currently logged on. Synopsis:

who [option] [file] [am i]

If given no non-option arguments, who prints the following information for each user currently logged on: login name, terminal line, login time, and remote hostname or X display.

If given one non-option argument, who uses that instead of a default system-maintained file (often /var/run/utmp or /etc/utmp) as the name of the file containing the record of users logged on. /var/log/wtmp is commonly given as an argument to who to look at who has previously logged on.

If given two non-option arguments, who prints only the entry for the user running it (determined from its standard input), preceded by the hostname. Traditionally, the two arguments given are ‘am i’, as in ‘who am i’.

Time stamps are listed according to the time zone rules specified by the TZ environment variable, or by the system default rules if TZ is not set. See Specifying the Time Zone with TZ in The GNU C Library Reference Manual.

The program accepts the following options. Also see Common options.

-a
--all

Same as ‘-b -d --login -p -r -t -T -u’.

-b
--boot

Print the date and time of last system boot.

-d
--dead

Print information corresponding to dead processes.

-H
--heading

Print a line of column headings.

-l
--login

List only the entries that correspond to processes via which the system is waiting for a user to login. The user name is always ‘LOGIN’.

--lookup

Attempt to canonicalize hostnames found in utmp through a DNS lookup. This is not the default because it can cause significant delays on systems with automatic dial-up internet access.

-m

Same as ‘who am i’.

-p
--process

List active processes spawned by init.

-q
--count

Print only the login names and the number of users logged on. Overrides all other options.

-r
--runlevel

Print the current (and maybe previous) run-level of the init process.

-s

Ignored; for compatibility with other versions of who.

-t
--time

Print last system clock change.

-u

After the login time, print the number of hours and minutes that the user has been idle. ‘.’ means the user was active in the last minute. ‘old’ means the user has been idle for more than 24 hours.

-w
-T
--mesg
--message
--writable

After each login name print a character indicating the user’s message status:

+’ allowing write messages
‘-’ disallowing write messages
‘?’ cannot find terminal device

The who command is installed only on platforms with the POSIX <utmpx.h> include file or equivalent, so portable scripts should not rely on its existence on non-POSIX platforms.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

21 System context

This section describes commands that print or change system-wide information.


Next: , Up: System context   [Contents][Index]

21.1 date: Print or set system date and time

Synopses:

date [option]… [+format]
date [-u|--utc|--universal] [ MMDDhhmm[[CC]YY][.ss] ]

Invoking date with no format argument is equivalent to invoking it with a default format that depends on the LC_TIME locale category. In the default C locale, this format is ‘'+%a %b %e %H:%M:%S %Z %Y'’, so the output looks like ‘Thu Mar  3 13:47:51 PST 2005’.

Normally, date uses the time zone rules indicated by the TZ environment variable, or the system default rules if TZ is not set. See Specifying the Time Zone with TZ in The GNU C Library Reference Manual.

If given an argument that starts with a ‘+’, date prints the current date and time (or the date and time specified by the --date option, see below) in the format defined by that argument, which is similar to that of the strftime function. Except for conversion specifiers, which start with ‘%’, characters in the format string are printed unchanged. The conversion specifiers are described below.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Up: date invocation   [Contents][Index]

21.1.1 Time conversion specifiers

date conversion specifiers related to times.

%H

hour (‘00’…‘23’)

%I

hour (‘01’…‘12’)

%k

hour, space padded (‘ 0’…‘23’); equivalent to ‘%_H’. This is a GNU extension.

%l

hour, space padded (‘ 1’…‘12’); equivalent to ‘%_I’. This is a GNU extension.

%M

minute (‘00’…‘59’)

%N

nanoseconds (‘000000000’…‘999999999’). This is a GNU extension.

%p

locale’s equivalent of either ‘AM’ or ‘PM’; blank in many locales. Noon is treated as ‘PM’ and midnight as ‘AM’.

%P

like ‘%p’, except lower case. This is a GNU extension.

%r

locale’s 12-hour clock time (e.g., ‘11:11:04 PM’)

%R

24-hour hour and minute. Same as ‘%H:%M’.

%s

seconds since the epoch, i.e., since 1970-01-01 00:00:00 UTC. Leap seconds are not counted unless leap second support is available. See %s-examples, for examples. This is a GNU extension.

%S

second (‘00’…‘60’). This may be ‘60’ if leap seconds are supported.

%T

24-hour hour, minute, and second. Same as ‘%H:%M:%S’.

%X

locale’s time representation (e.g., ‘23:13:48’)

%z

RFC 2822/ISO 8601 style numeric time zone (e.g., ‘-0600’ or ‘+0530’), or nothing if no time zone is determinable. This value reflects the numeric time zone appropriate for the current time, using the time zone rules specified by the TZ environment variable. The time (and optionally, the time zone rules) can be overridden by the --date option.

%:z

RFC 3339/ISO 8601 style numeric time zone with ‘:’ (e.g., ‘-06:00’ or ‘+05:30’), or nothing if no time zone is determinable. This is a GNU extension.

%::z

Numeric time zone to the nearest second with ‘:’ (e.g., ‘-06:00:00’ or ‘+05:30:00’), or nothing if no time zone is determinable. This is a GNU extension.

%:::z

Numeric time zone with ‘:’ using the minimum necessary precision (e.g., ‘-06’, ‘+05:30’, or ‘-04:56:02’), or nothing if no time zone is determinable. This is a GNU extension.

%Z

alphabetic time zone abbreviation (e.g., ‘EDT’), or nothing if no time zone is determinable. See ‘%z’ for how it is determined.


Next: , Previous: , Up: date invocation   [Contents][Index]

21.1.2 Date conversion specifiers

date conversion specifiers related to dates.

%a

locale’s abbreviated weekday name (e.g., ‘Sun’)

%A

locale’s full weekday name, variable length (e.g., ‘Sunday’)

%b

locale’s abbreviated month name (e.g., ‘Jan’)

%B

locale’s full month name, variable length (e.g., ‘January’)

%c

locale’s date and time (e.g., ‘Thu Mar  3 23:05:25 2005’)

%C

century. This is like ‘%Y’, except the last two digits are omitted. For example, it is ‘20’ if ‘%Y’ is ‘2000’, and is ‘-0’ if ‘%Y’ is ‘-001’. It is normally at least two characters, but it may be more.

%d

day of month (e.g., ‘01’)

%D

date; same as ‘%m/%d/%y

%e

day of month, space padded; same as ‘%_d

%F

full date in ISO 8601 format; same as ‘%Y-%m-%d’. This is a good choice for a date format, as it is standard and is easy to sort in the usual case where years are in the range 0000…9999.

%g

year corresponding to the ISO week number, but without the century (range ‘00’ through ‘99’). This has the same format and value as ‘%y’, except that if the ISO week number (see ‘%V’) belongs to the previous or next year, that year is used instead.

%G

year corresponding to the ISO week number. This has the same format and value as ‘%Y’, except that if the ISO week number (see ‘%V’) belongs to the previous or next year, that year is used instead. It is normally useful only if ‘%V’ is also used; for example, the format ‘%G-%m-%d’ is probably a mistake, since it combines the ISO week number year with the conventional month and day.

%h

same as ‘%b

%j

day of year (‘001’…‘366’)

%m

month (‘01’…‘12’)

%u

day of week (‘1’…‘7’) with ‘1’ corresponding to Monday

%U

week number of year, with Sunday as the first day of the week (‘00’…‘53’). Days in a new year preceding the first Sunday are in week zero.

%V

ISO week number, that is, the week number of year, with Monday as the first day of the week (‘01’…‘53’). If the week containing January 1 has four or more days in the new year, then it is considered week 1; otherwise, it is week 53 of the previous year, and the next week is week 1. (See the ISO 8601 standard.)

%w

day of week (‘0’…‘6’) with 0 corresponding to Sunday

%W

week number of year, with Monday as first day of week (‘00’…‘53’). Days in a new year preceding the first Monday are in week zero.

%x

locale’s date representation (e.g., ‘12/31/99’)

%y

last two digits of year (‘00’…‘99’)

%Y

year. This is normally at least four characters, but it may be more. Year ‘0000’ precedes year ‘0001’, and year ‘-001’ precedes year ‘0000’.


Next: , Previous: , Up: date invocation   [Contents][Index]

21.1.3 Literal conversion specifiers

date conversion specifiers that produce literal strings.

%%

a literal %

%n

a newline

%t

a horizontal tab


Next: , Previous: , Up: date invocation   [Contents][Index]

21.1.4 Padding and other flags

Unless otherwise specified, date normally pads numeric fields with zeros, so that, for example, numeric months are always output as two digits. Seconds since the epoch are not padded, though, since there is no natural width for them.

As a GNU extension, date recognizes any of the following optional flags after the ‘%’:

-

(hyphen) Do not pad the field; useful if the output is intended for human consumption.

_

(underscore) Pad with spaces; useful if you need a fixed number of characters in the output, but zeros are too distracting.

0

(zero) Pad with zeros even if the conversion specifier would normally pad with spaces.

^

Use upper case characters if possible.

#

Use opposite case characters if possible. A field that is normally upper case becomes lower case, and vice versa.

Here are some examples of padding:

date +%d/%m -d "Feb 1"
⇒ 01/02
date +%-d/%-m -d "Feb 1"
⇒ 1/2
date +%_d/%_m -d "Feb 1"
⇒  1/ 2

As a GNU extension, you can specify the field width (after any flag, if present) as a decimal number. If the natural size of the output of the field has less than the specified number of characters, the result is written right adjusted and padded to the given size. For example, ‘%9B’ prints the right adjusted month name in a field of width 9.

An optional modifier can follow the optional flag and width specification. The modifiers are:

E

Use the locale’s alternate representation for date and time. This modifier applies to the ‘%c’, ‘%C’, ‘%x’, ‘%X’, ‘%y’ and ‘%Y’ conversion specifiers. In a Japanese locale, for example, ‘%Ex’ might yield a date format based on the Japanese Emperors’ reigns.

O

Use the locale’s alternate numeric symbols for numbers. This modifier applies only to numeric conversion specifiers.

If the format supports the modifier but no alternate representation is available, it is ignored.


Next: , Previous: , Up: date invocation   [Contents][Index]

21.1.5 Setting the time

If given an argument that does not start with ‘+’, date sets the system clock to the date and time specified by that argument (as described below). You must have appropriate privileges to set the system clock. Note for changes to persist across a reboot, the hardware clock may need to be updated from the system clock, which might not happen automatically on your system.

The argument must consist entirely of digits, which have the following meaning:

MM

month

DD

day within month

hh

hour

mm

minute

CC

first two digits of year (optional)

YY

last two digits of year (optional)

ss

second (optional)

Note, the --date and --set options may not be used with an argument in the above format. The --universal option may be used with such an argument to indicate that the specified date and time are relative to Coordinated Universal Time rather than to the local time zone.


Next: , Previous: , Up: date invocation   [Contents][Index]

21.1.6 Options for date

The program accepts the following options. Also see Common options.

-d datestr
--date=datestr

Display the date and time specified in datestr instead of the current date and time. datestr can be in almost any common format. It can contain month names, time zones, ‘am’ and ‘pm’, ‘yesterday’, etc. For example, --date="2004-02-27 14:19:13.489392193 +0530" specifies the instant of time that is 489,392,193 nanoseconds after February 27, 2004 at 2:19:13 PM in a time zone that is 5 hours and 30 minutes east of UTC.
Note: input currently must be in locale independent format. E.g., the LC_TIME=C below is needed to print back the correct date in many locales:

date -d "$(LC_TIME=C date)"

See Date input formats.

-f datefile
--file=datefile

Parse each line in datefile as with -d and display the resulting date and time. If datefile is ‘-’, use standard input. This is useful when you have many dates to process, because the system overhead of starting up the date executable many times can be considerable.

-I[timespec]
--iso-8601[=timespec]

Display the date using the ISO 8601 format, ‘%Y-%m-%d’.

The argument timespec specifies the number of additional terms of the time to include. It can be one of the following:

auto

Print just the date. This is the default if timespec is omitted.

hours

Append the hour of the day to the date.

minutes

Append the hours and minutes.

seconds

Append the hours, minutes and seconds.

ns

Append the hours, minutes, seconds and nanoseconds.

If showing any time terms, then include the time zone using the format ‘%z’.

-r file
--reference=file

Display the date and time of the last modification of file, instead of the current date and time.

-R
--rfc-822
--rfc-2822

Display the date and time using the format ‘%a, %d %b %Y %H:%M:%S %z’, evaluated in the C locale so abbreviations are always in English. For example:

Fri, 09 Sep 2005 13:51:39 -0700

This format conforms to ftp://ftp.rfc-editor.org/in-notes/rfc2822.txt">Internet RFCs 2822 and ftp://ftp.rfc-editor.org/in-notes/rfc822.txt">822, the current and previous standards for Internet email.

--rfc-3339=timespec

Display the date using a format specified by ftp://ftp.rfc-editor.org/in-notes/rfc3339.txt">Internet RFC 3339. This is a subset of the ISO 8601 format, except that it also permits applications to use a space rather than a ‘T’ to separate dates from times. Unlike the other standard formats, RFC 3339 format is always suitable as input for the --date (-d) and --file (-f) options, regardless of the current locale.

The argument timespec specifies how much of the time to include. It can be one of the following:

date

Print just the full-date, e.g., ‘2005-09-14’. This is equivalent to the format ‘%Y-%m-%d’.

seconds

Print the full-date and full-time separated by a space, e.g., ‘2005-09-14 00:56:06+05:30’. The output ends with a numeric time-offset; here the ‘+05:30’ means that local time is five hours and thirty minutes east of UTC. This is equivalent to the format ‘%Y-%m-%d %H:%M:%S%:z’.

ns

Like ‘seconds’, but also print nanoseconds, e.g., ‘2005-09-14 00:56:06.998458565+05:30’. This is equivalent to the format ‘%Y-%m-%d %H:%M:%S.%N%:z’.

-s datestr
--set=datestr

Set the date and time to datestr. See -d above. See also Setting the time.

-u
--utc
--universal

Use Coordinated Universal Time (UTC) by operating as if the TZ environment variable were set to the string ‘UTC0’. Coordinated Universal Time is often called “Greenwich Mean Time” (GMT) for historical reasons. Typically, systems ignore leap seconds and thus implement an approximation to UTC rather than true UTC.


Previous: , Up: date invocation   [Contents][Index]

21.1.7 Examples of date

Here are a few examples. Also see the documentation for the -d option in the previous section.


Next: , Previous: , Up: System context   [Contents][Index]

21.2 arch: Print machine hardware name

arch prints the machine hardware name, and is equivalent to ‘uname -m’. Synopsis:

arch [option]

The program accepts the Common options only.

arch is not installed by default, so portable scripts should not rely on its existence.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: System context   [Contents][Index]

21.3 nproc: Print the number of available processors

Print the number of processing units available to the current process, which may be less than the number of online processors. If this information is not accessible, then print the number of processors installed. If the OMP_NUM_THREADS environment variable is set, then it will determine the returned value. The result is guaranteed to be greater than zero. Synopsis:

nproc [option]

The program accepts the following options. Also see Common options.

--all

Print the number of installed processors on the system, which may be greater than the number online or available to the current process. The OMP_NUM_THREADS environment variable is not honored in this case.

--ignore=number

If possible, exclude this number of processing units.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: System context   [Contents][Index]

21.4 uname: Print system information

uname prints information about the machine and operating system it is run on. If no options are given, uname acts as if the -s option were given. Synopsis:

uname [option]…

If multiple options or -a are given, the selected information is printed in this order:

kernel-name nodename kernel-release kernel-version
machine processor hardware-platform operating-system

The information may contain internal spaces, so such output cannot be parsed reliably. In the following example, release is ‘2.2.18ss.e820-bda652a #4 SMP Tue Jun 5 11:24:08 PDT 2001’:

uname -a
⇒ Linux dumdum 2.2.18 #4 SMP Tue Jun 5 11:24:08 PDT 2001 i686 unknown unknown GNU/Linux

The program accepts the following options. Also see Common options.

-a
--all

Print all of the below information, except omit the processor type and the hardware platform name if they are unknown.

-i
--hardware-platform

Print the hardware platform name (sometimes called the hardware implementation). Print ‘unknown’ if the kernel does not make this information easily available, as is the case with Linux kernels.

-m
--machine

Print the machine hardware name (sometimes called the hardware class or hardware type).

-n
--nodename

Print the network node hostname.

-p
--processor

Print the processor type (sometimes called the instruction set architecture or ISA). Print ‘unknown’ if the kernel does not make this information easily available, as is the case with Linux kernels.

-o
--operating-system

Print the name of the operating system.

-r
--kernel-release

Print the kernel release.

-s
--kernel-name

Print the kernel name. POSIX 1003.1-2001 (see Standards conformance) calls this “the implementation of the operating system”, because the POSIX specification itself has no notion of “kernel”. The kernel name might be the same as the operating system name printed by the -o or --operating-system option, but it might differ. Some operating systems (e.g., FreeBSD, HP-UX) have the same name as their underlying kernels; others (e.g., GNU/Linux, Solaris) do not.

-v
--kernel-version

Print the kernel version.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: System context   [Contents][Index]

21.5 hostname: Print or set system name

With no arguments, hostname prints the name of the current host system. With one argument, it sets the current host name to the specified string. You must have appropriate privileges to set the host name. Synopsis:

hostname [name]

The only options are --help and --version. See Common options.

hostname is not installed by default, and other packages also supply a hostname command, so portable scripts should not rely on its existence or on the exact behavior documented above.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: System context   [Contents][Index]

21.6 hostid: Print numeric host identifier

hostid prints the numeric identifier of the current host in hexadecimal. This command accepts no arguments. The only options are --help and --version. See Common options.

For example, here’s what it prints on one system I use:

$ hostid
1bac013d

On that system, the 32-bit quantity happens to be closely related to the system’s Internet address, but that isn’t always the case.

hostid is installed only on systems that have the gethostid function, so portable scripts should not rely on its existence.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: System context   [Contents][Index]

21.7 uptime: Print system uptime and load

uptime prints the current time, the system’s uptime, the number of logged-in users and the current load average.

If an argument is specified, it is used as the file to be read to discover how many users are logged in. If no argument is specified, a system default is used (uptime --help indicates the default setting).

The only options are --help and --version. See Common options.

For example, here’s what it prints right now on one system I use:

$ uptime
 14:07  up   3:35,  3 users,  load average: 1.39, 1.15, 1.04

The precise method of calculation of load average varies somewhat between systems. Some systems calculate it as the average number of runnable processes over the last 1, 5 and 15 minutes, but some systems also include processes in the uninterruptible sleep state (that is, those processes which are waiting for disk I/O). The Linux kernel includes uninterruptible processes.

uptime is installed only on platforms with infrastructure for obtaining the boot time, and other packages also supply an uptime command, so portable scripts should not rely on its existence or on the exact behavior documented above.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

22 SELinux context

This section describes commands for operations with SELinux contexts.


Next: , Up: SELinux context   [Contents][Index]

22.1 chcon: Change SELinux context of file

chcon changes the SELinux security context of the selected files. Synopses:

chcon [option]… context file…
chcon [option]… [-u user] [-r role] [-l range] [-t type] file…
chcon [option]… --reference=rfile file

Change the SELinux security context of each file to context. With --reference, change the security context of each file to that of rfile.

The program accepts the following options. Also see Common options.

--dereference

Do not affect symbolic links but what they refer to; this is the default.

-h
--no-dereference

Affect the symbolic links themselves instead of any referenced file.

--reference=rfile

Use rfile’s security context rather than specifying a context value.

-R
--recursive

Operate on files and directories recursively.

--preserve-root

Refuse to operate recursively on the root directory, /, when used together with the --recursive option. See Treating / specially.

--no-preserve-root

Do not treat the root directory, /, specially when operating recursively; this is the default. See Treating / specially.

-H

If --recursive (-R) is specified and a command line argument is a symbolic link to a directory, traverse it. See Traversing symlinks.

-L

In a recursive traversal, traverse every symbolic link to a directory that is encountered. See Traversing symlinks.

-P

Do not traverse any symbolic links. This is the default if none of -H, -L, or -P is specified. See Traversing symlinks.

-v
--verbose

Output a diagnostic for every file processed.

-u user
--user=user

Set user user in the target security context.

-r role
--role=role

Set role role in the target security context.

-t type
--type=type

Set type type in the target security context.

-l range
--range=range

Set range range in the target security context.

An exit status of zero indicates success, and a nonzero value indicates failure.


Previous: , Up: SELinux context   [Contents][Index]

22.2 runcon: Run a command in specified SELinux context

runcon runs file in specified SELinux security context.

Synopses:

runcon context command [args]
runcon [ -c ] [-u user] [-r role] [-t type] [-l range] command [args]

Run command with completely-specified context, or with current or transitioned security context modified by one or more of level, role, type and user.

If none of -c, -t, -u, -r, or -l is specified, the first argument is used as the complete context. Any additional arguments after command are interpreted as arguments to the command.

With neither context nor command, print the current security context.

The program accepts the following options. Also see Common options.

-c
--compute

Compute process transition context before modifying.

-u user
--user=user

Set user user in the target security context.

-r role
--role=role

Set role role in the target security context.

-t type
--type=type

Set type type in the target security context.

-l range
--range=range

Set range range in the target security context.

Exit status:

126 if command is found but cannot be invoked
127 if runcon itself fails or if command cannot be found
the exit status of command otherwise

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

23 Modified command invocation

This section describes commands that run other commands in some context different than the current one: a modified environment, as a different user, etc.


Next: , Up: Modified command invocation   [Contents][Index]

23.1 chroot: Run a command with a different root directory

chroot runs a command with a specified root directory. On many systems, only the super-user can do this.4. Synopses:

chroot option newroot [command [args]…]
chroot option

Ordinarily, file names are looked up starting at the root of the directory structure, i.e., /. chroot changes the root to the directory newroot (which must exist), then changes the working directory to /, and finally runs command with optional args. If command is not specified, the default is the value of the SHELL environment variable or /bin/sh if not set, invoked with the -i option. command must not be a special built-in utility (see Special built-in utilities).

The program accepts the following options. Also see Common options. Options must precede operands.

--groups=groups

Use this option to override the supplementary groups to be used by the new process. The items in the list (names or numeric IDs) must be separated by commas. Use ‘--groups=''’ to disable the supplementary group look-up implicit in the --userspec option.

--userspec=user[:group]

By default, command is run with the same credentials as the invoking process. Use this option to run it as a different user and/or with a different primary group. If a user is specified then the supplementary groups are set according to the system defined list for that user, unless overridden with the --groups option.

--skip-chdir

Use this option to not change the working directory to / after changing the root directory to newroot, i.e., inside the chroot. This option is only permitted when newroot is the old / directory, and therefore is mostly useful together with the --groups and --userspec options to retain the previous working directory.

The user and group name look-up performed by the --userspec and --groups options, is done both outside and inside the chroot, with successful look-ups inside the chroot taking precedence. If the specified user or group items are intended to represent a numeric ID, then a name to ID resolving step is avoided by specifying a leading ‘+’. See Disambiguating names and IDs.

Here are a few tips to help avoid common problems in using chroot. To start with a simple example, make command refer to a statically linked binary. If you were to use a dynamically linked executable, then you’d have to arrange to have the shared libraries in the right place under your new root directory.

For example, if you create a statically linked ls executable, and put it in /tmp/empty, you can run this command as root:

$ chroot /tmp/empty /ls -Rl /

Then you’ll see output like this:

/:
total 1023
-rwxr-xr-x 1 0 0 1041745 Aug 16 11:17 ls

If you want to use a dynamically linked executable, say bash, then first run ‘ldd bash’ to see what shared objects it needs. Then, in addition to copying the actual binary, also copy the listed files to the required positions under your intended new root directory. Finally, if the executable requires any other files (e.g., data, state, device files), copy them into place, too.

chroot is installed only on systems that have the chroot function, so portable scripts should not rely on its existence.

Exit status:

125 if chroot itself fails
126 if command is found but cannot be invoked
127 if command cannot be found
the exit status of command otherwise

Next: , Previous: , Up: Modified command invocation   [Contents][Index]

23.2 env: Run a command in a modified environment

env runs a command with a modified environment. Synopses:

env [option]… [name=value]… [command [args]…]
env

Operands of the form ‘variable=value’ set the environment variable variable to value value. value may be empty (‘variable=’). Setting a variable to an empty value is different from unsetting it. These operands are evaluated left-to-right, so if two operands mention the same variable the earlier is ignored.

Environment variable names can be empty, and can contain any characters other than ‘=’ and ASCII NUL. However, it is wise to limit yourself to names that consist solely of underscores, digits, and ASCII letters, and that begin with a non-digit, as applications like the shell do not work well with other names.

The first operand that does not contain the character ‘=’ specifies the program to invoke; it is searched for according to the PATH environment variable. Any remaining arguments are passed as arguments to that program. The program should not be a special built-in utility (see Special built-in utilities).

Modifications to PATH take effect prior to searching for command. Use caution when reducing PATH; behavior is not portable when PATH is undefined or omits key directories such as /bin.

In the rare case that a utility contains a ‘=’ in the name, the only way to disambiguate it from a variable assignment is to use an intermediate command for command, and pass the problematic program name via args. For example, if ./prog= is an executable in the current PATH:

env prog= true # runs 'true', with prog= in environment
env ./prog= true # runs 'true', with ./prog= in environment
env -- prog= true # runs 'true', with prog= in environment
env sh -c '\prog= true' # runs 'prog=' with argument 'true'
env sh -c 'exec "$@"' sh prog= true # also runs 'prog='

If no command name is specified following the environment specifications, the resulting environment is printed. This is like specifying the printenv program.

For some examples, suppose the environment passed to env contains ‘LOGNAME=rms’, ‘EDITOR=emacs’, and ‘PATH=.:/gnubin:/hacks’:

The program accepts the following options. Also see Common options. Options must precede operands.

-0
--null

Output a zero byte (ASCII NUL) at the end of each line, rather than a newline. This option enables other programs to parse the output even when that output would contain data with embedded newlines.

-u name
--unset=name

Remove variable name from the environment, if it was in the environment.

-
-i
--ignore-environment

Start with an empty environment, ignoring the inherited environment.

Exit status:

0   if no command is specified and the environment is output
125 if env itself fails
126 if command is found but cannot be invoked
127 if command cannot be found
the exit status of command otherwise

Next: , Previous: , Up: Modified command invocation   [Contents][Index]

23.3 nice: Run a command with modified niceness

nice prints a process’s niceness, or runs a command with modified niceness. niceness affects how favorably the process is scheduled in the system. Synopsis:

nice [option]… [command [arg]…]

If no arguments are given, nice prints the current niceness. Otherwise, nice runs the given command with its niceness adjusted. By default, its niceness is incremented by 10.

Niceness values range at least from -20 (process has high priority and gets more resources, thus slowing down other processes) through 19 (process has lower priority and runs slowly itself, but has less impact on the speed of other running processes). Some systems may have a wider range of niceness values; conversely, other systems may enforce more restrictive limits. An attempt to set the niceness outside the supported range is treated as an attempt to use the minimum or maximum supported value.

A niceness should not be confused with a scheduling priority, which lets applications determine the order in which threads are scheduled to run. Unlike a priority, a niceness is merely advice to the scheduler, which the scheduler is free to ignore. Also, as a point of terminology, POSIX defines the behavior of nice in terms of a nice value, which is the non-negative difference between a niceness and the minimum niceness. Though nice conforms to POSIX, its documentation and diagnostics use the term “niceness” for compatibility with historical practice.

command must not be a special built-in utility (see Special built-in utilities).

Due to shell aliases and built-in nice functions, using an unadorned nice interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env nice …) to avoid interference from the shell.

Note to change the niceness of an existing process, one needs to use the renice command.

The program accepts the following option. Also see Common options. Options must precede operands.

-n adjustment
--adjustment=adjustment

Add adjustment instead of 10 to the command’s niceness. If adjustment is negative and you lack appropriate privileges, nice issues a warning but otherwise acts as if you specified a zero adjustment.

For compatibility nice also supports an obsolete option syntax -adjustment. New scripts should use -n adjustment instead.

nice is installed only on systems that have the POSIX setpriority function, so portable scripts should not rely on its existence on non-POSIX platforms.

Exit status:

0   if no command is specified and the niceness is output
125 if nice itself fails
126 if command is found but cannot be invoked
127 if command cannot be found
the exit status of command otherwise

It is sometimes useful to run a non-interactive program with reduced niceness.

$ nice factor 4611686018427387903

Since nice prints the current niceness, you can invoke it through itself to demonstrate how it works.

The default behavior is to increase the niceness by ‘10’:

$ nice
0
$ nice nice
10
$ nice -n 10 nice
10

The adjustment is relative to the current niceness. In the next example, the first nice invocation runs the second one with niceness 10, and it in turn runs the final one with a niceness that is 3 more:

$ nice nice -n 3 nice
13

Specifying a niceness larger than the supported range is the same as specifying the maximum supported value:

$ nice -n 10000000000 nice
19

Only a privileged user may run a process with lower niceness:

$ nice -n -1 nice
nice: cannot set niceness: Permission denied
0
$ sudo nice -n -1 nice
-1

Next: , Previous: , Up: Modified command invocation   [Contents][Index]

23.4 nohup: Run a command immune to hangups

nohup runs the given command with hangup signals ignored, so that the command can continue running in the background after you log out. Synopsis:

nohup command [arg]…

If standard input is a terminal, redirect it so that terminal sessions do not mistakenly consider the terminal to be used by the command. Make the substitute file descriptor unreadable, so that commands that mistakenly attempt to read from standard input can report an error. This redirection is a GNU extension; programs intended to be portable to non-GNU hosts can use ‘nohup command [arg]… 0>/dev/null’ instead.

If standard output is a terminal, the command’s standard output is appended to the file nohup.out; if that cannot be written to, it is appended to the file $HOME/nohup.out; and if that cannot be written to, the command is not run. Any nohup.out or $HOME/nohup.out file created by nohup is made readable and writable only to the user, regardless of the current umask settings.

If standard error is a terminal, it is normally redirected to the same file descriptor as the (possibly-redirected) standard output. However, if standard output is closed, standard error terminal output is instead appended to the file nohup.out or $HOME/nohup.out as above.

To capture the command’s output to a file other than nohup.out you can redirect it. For example, to capture the output of make:

nohup make > make.log

nohup does not automatically put the command it runs in the background; you must do that explicitly, by ending the command line with an ‘&’. Also, nohup does not alter the niceness of command; use nice for that, e.g., ‘nohup nice command’.

command must not be a special built-in utility (see Special built-in utilities).

The only options are --help and --version. See Common options. Options must precede operands.

Exit status:

125 if nohup itself fails, and POSIXLY_CORRECT is not set
126 if command is found but cannot be invoked
127 if command cannot be found
the exit status of command otherwise

If POSIXLY_CORRECT is set, internal failures give status 127 instead of 125.


Next: , Previous: , Up: Modified command invocation   [Contents][Index]

23.5 stdbuf: Run a command with modified I/O stream buffering

stdbuf allows one to modify the buffering operations of the three standard I/O streams associated with a program. Synopsis:

stdbuf optioncommand

command must start with the name of a program that

  1. uses the ISO C FILE streams for input/output (note the programs dd and cat don’t do that),
  2. does not adjust the buffering of its standard streams (note the program tee is not in this category).

Any additional args are passed as additional arguments to the command.

The program accepts the following options. Also see Common options.

-i mode
--input=mode

Adjust the standard input stream buffering.

-o mode
--output=mode

Adjust the standard output stream buffering.

-e mode
--error=mode

Adjust the standard error stream buffering.

The mode can be specified as follows:

L

Set the stream to line buffered mode. In this mode data is coalesced until a newline is output or input is read from any stream attached to a terminal device. This option is invalid with standard input.

0

Disable buffering of the selected stream. In this mode, data is output immediately and only the amount of data requested is read from input. Note the difference in function for input and output. Disabling buffering for input will not influence the responsiveness or blocking behavior of the stream input functions. For example fread will still block until EOF or error, even if the underlying read returns less data than requested.

size

Specify the size of the buffer to use in fully buffered mode. size may be, or may be an integer optionally followed by, one of the following multiplicative suffixes:

KB’ =>           1000 (KiloBytes)
‘K’  =>           1024 (KibiBytes)
‘MB’ =>      1000*1000 (MegaBytes)
‘M’  =>      1024*1024 (MebiBytes)
‘GB’ => 1000*1000*1000 (GigaBytes)
‘G’  => 1024*1024*1024 (GibiBytes)

and so on for ‘T’, ‘P’, ‘E’, ‘Z’, and ‘Y’.

stdbuf is installed only on platforms that use the Executable and Linkable Format (ELF) and support the constructor attribute, so portable scripts should not rely on its existence.

Exit status:

125 if stdbuf itself fails
126 if command is found but cannot be invoked
127 if command cannot be found
the exit status of command otherwise

Previous: , Up: Modified command invocation   [Contents][Index]

23.6 timeout: Run a command with a time limit

timeout runs the given command and kills it if it is still running after the specified time interval. Synopsis:

timeout [option] duration command [arg]…

command must not be a special built-in utility (see Special built-in utilities).

The program accepts the following options. Also see Common options. Options must precede operands.

--preserve-status

Return the exit status of the managed command on timeout, rather than a specific exit status indicating a timeout. This is useful if the managed command supports running for an indeterminate amount of time.

--foreground

Don’t create a separate background program group, so that the managed command can use the foreground TTY normally. This is needed to support timing out commands not started directly from an interactive shell, in two situations.

  1. command is interactive and needs to read from the terminal for example
  2. the user wants to support sending signals directly to command from the terminal (like Ctrl-C for example)

Note in this mode of operation, any children of command will not be timed out. Also SIGCONT will not be sent to command, as it’s generally not needed with foreground processes, and can cause intermittent signal delivery issues with programs that are monitors themselves (like GDB for example).

-k duration
--kill-after=duration

Ensure the monitored command is killed by also sending a ‘KILL’ signal, after the specified duration. Without this option, if the selected signal proves not to be fatal, timeout does not kill the command.

-s signal
--signal=signal

Send this signal to command on timeout, rather than the default ‘TERM’ signal. signal may be a name like ‘HUP’ or a number. See Signal specifications.

duration is a floating point number followed by an optional unit:

s’ for seconds (the default)
‘m’ for minutes
‘h’ for hours
‘d’ for days

A duration of 0 disables the associated timeout. Note that the actual timeout duration is dependent on system conditions, which should be especially considered when specifying sub-second timeouts.

Exit status:

124 if command times out
125 if timeout itself fails
126 if command is found but cannot be invoked
127 if command cannot be found
137 if command is sent the KILL(9) signal (128+9)
the exit status of command otherwise

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

24 Process control


Up: Process control   [Contents][Index]

24.1 kill: Send a signal to processes

The kill command sends a signal to processes, causing them to terminate or otherwise act upon receiving the signal in some way. Alternatively, it lists information about signals. Synopses:

kill [-s signal | --signal signal | -signal] pid…
kill [-l | --list | -t | --table] [signal]…

Due to shell aliases and built-in kill functions, using an unadorned kill interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env kill …) to avoid interference from the shell.

The first form of the kill command sends a signal to all pid arguments. The default signal to send if none is specified is ‘TERM’. The special signal number ‘0’ does not denote a valid signal, but can be used to test whether the pid arguments specify processes to which a signal could be sent.

If pid is positive, the signal is sent to the process with the process ID pid. If pid is zero, the signal is sent to all processes in the process group of the current process. If pid is -1, the signal is sent to all processes for which the user has permission to send a signal. If pid is less than -1, the signal is sent to all processes in the process group that equals the absolute value of pid.

If pid is not positive, a system-dependent set of system processes is excluded from the list of processes to which the signal is sent.

If a negative pid argument is desired as the first one, it should be preceded by --. However, as a common extension to POSIX, -- is not required with ‘kill -signal -pid’. The following commands are equivalent:

kill -15 -1
kill -TERM -1
kill -s TERM -- -1
kill -- -1

The first form of the kill command succeeds if every pid argument specifies at least one process that the signal was sent to.

The second form of the kill command lists signal information. Either the -l or --list option, or the -t or --table option must be specified. Without any signal argument, all supported signals are listed. The output of -l or --list is a list of the signal names, one per line; if signal is already a name, the signal number is printed instead. The output of -t or --table is a table of signal numbers, names, and descriptions. This form of the kill command succeeds if all signal arguments are valid and if there is no output error.

The kill command also supports the --help and --version options. See Common options.

A signal may be a signal name like ‘HUP’, or a signal number like ‘1’, or an exit status of a process terminated by the signal. A signal name can be given in canonical form or prefixed by ‘SIG’. The case of the letters is ignored, except for the -signal option which must use upper case to avoid ambiguity with lower case option letters. See Signal specifications, for a list of supported signal names and numbers.


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

25 Delaying


Up: Delaying   [Contents][Index]

25.1 sleep: Delay for a specified time

sleep pauses for an amount of time specified by the sum of the values of the command line arguments. Synopsis:

sleep number[smhd]…

Each argument is a number followed by an optional unit; the default is seconds. The units are:

s

seconds

m

minutes

h

hours

d

days

Historical implementations of sleep have required that number be an integer, and only accepted a single argument without a suffix. However, GNU sleep accepts arbitrary floating point numbers. See Floating point.

The only options are --help and --version. See Common options.

Due to shell aliases and built-in sleep functions, using an unadorned sleep interactively or in a script may get you different functionality than that described here. Invoke it via env (i.e., env sleep …) to avoid interference from the shell.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

26 Numeric operations

These programs do numerically-related operations.


Next: , Up: Numeric operations   [Contents][Index]

26.1 factor: Print prime factors

factor prints prime factors. Synopses:

factor [number]…
factor option

If no number is specified on the command line, factor reads numbers from standard input, delimited by newlines, tabs, or spaces.

The factor command supports only a small number of options:

--help

Print a short help on standard output, then exit without further processing.

--version

Print the program version on standard output, then exit without further processing.

Factoring the product of the eighth and ninth Mersenne primes takes about 30 milliseconds of CPU time on a 2.2 GHz Athlon.

M8=$(echo 2^31-1|bc)
M9=$(echo 2^61-1|bc)
n=$(echo "$M8 * $M9" | bc)
/usr/bin/time -f %U factor $n
4951760154835678088235319297: 2147483647 2305843009213693951
0.03

Similarly, factoring the eighth Fermat number 2^{256}+1 takes about 20 seconds on the same machine.

Factoring large numbers is, in general, hard. The Pollard Rho algorithm used by factor is particularly effective for numbers with relatively small factors. If you wish to factor large numbers which do not have small factors (for example, numbers which are the product of two large primes), other methods are far better.

If factor is built without using GNU MP, only single-precision arithmetic is available, and so large numbers (typically 2^{64} and above) will not be supported. The single-precision code uses an algorithm which is designed for factoring smaller numbers.

An exit status of zero indicates success, and a nonzero value indicates failure.


Next: , Previous: , Up: Numeric operations   [Contents][Index]

26.2 numfmt: Reformat numbers

numfmt reads numbers in various representations and reformats them as requested. The most common usage is converting numbers to/from human representation (e.g. ‘4G’ → ‘4,000,000,000’).

numfmt [option]… [number]

numfmt converts each number on the command-line according to the specified options (see below). If no numbers are given, it reads numbers from standard input. numfmt can optionally extract numbers from specific columns, maintaining proper line padding and alignment.

An exit status of zero indicates success, and a nonzero value indicates failure.

See --invalid for additional information regarding exit status.

26.2.1 General options

The program accepts the following options. Also see Common options.

--debug

Print (to standard error) warning messages about possible erroneous usage.

-d d
--delimiter=d

Use the character d as input field separator (default: whitespace). Note: Using non-default delimiter turns off automatic padding.

--field=fields

Convert the number in input field fields (default: 1). fields supports cut style field ranges:

N    N'th field, counted from 1
N-   from N'th field, to end of line
N-M  from N'th to M'th field (inclusive)
-M   from first to M'th field (inclusive)
-    all fields
--format=format

Use printf-style floating FORMAT string. The format string must contain one ‘%f’ directive, optionally with ‘'’, ‘-’, ‘0’, width or precision modifiers. The ‘'’ modifier will enable --grouping, the ‘-’ modifier will enable left-aligned --padding and the width modifier will enable right-aligned --padding. The ‘0’ width modifier (without the ‘-’ modifier) will generate leading zeros on the number, up to the specified width. A precision specification like ‘%.1f’ will override the precision determined from the input data or set due to --to option auto scaling.

--from=unit

Auto-scales input numbers according to unit. See UNITS below. The default is no scaling, meaning suffixes (e.g. ‘M’, ‘G’) will trigger an error.

--from-unit=n

Specify the input unit size (instead of the default 1). Use this option when the input numbers represent other units (e.g. if the input number ‘10’ represents 10 units of 512 bytes, use ‘--from-unit=512’). Suffixes are handled as with ‘--from=auto’.

--grouping

Group digits in output numbers according to the current locale’s grouping rules (e.g Thousands Separator character, commonly ‘.’ (dot) or ‘,’ comma). This option has no effect in ‘POSIX/C’ locale.

--header[=n]

Print the first n (default: 1) lines without any conversion.

--invalid=mode

The default action on input errors is to exit immediately with status code 2. --invalid=abort explicitly specifies this default mode. With a mode of ‘fail’, print a warning for each conversion error, and exit with status 2. With a mode of ‘warn’, exit with status 0, even in the presence of conversion errors, and with a mode of ‘ignore’ do not even print diagnostics.

--padding=n

Pad the output numbers to n characters, by adding spaces. If n is a positive number, numbers will be right-aligned. If n is a negative number, numbers will be left-aligned. By default, numbers are automatically aligned based on the input line’s width (only with the default delimiter).

--round=method

When converting number representations, round the number according to method, which can be ‘up’, ‘down’, ‘from-zero’ (the default), ‘towards-zero’, ‘nearest’.

--suffix=suffix

Add ‘SUFFIX’ to the output numbers, and accept optional ‘SUFFIX’ in input numbers.

--to=unit

Auto-scales output numbers according to unit. See Units below. The default is no scaling, meaning all the digits of the number are printed.

--to-unit=n

Specify the output unit size (instead of the default 1). Use this option when the output numbers represent other units (e.g. to represent ‘4,000,000’ bytes in blocks of 1KB, use ‘--to=si --to-unit=1000’). Suffixes are handled as with ‘--from=auto’.

26.2.2 Possible units:

The following are the possible unit options with --from=UNITS and --to=UNITS:

none

No scaling is performed. For input numbers, no suffixes are accepted, and any trailing characters following the number will trigger an error. For output numbers, all digits of the numbers will be printed.

si

Auto-scale numbers according to the International System of Units (SI) standard. For input numbers, accept one of the following suffixes. For output numbers, values larger than 1000 will be rounded, and printed with one of the following suffixes:

K’  =>  1000^1 = 10^3 (Kilo)
‘M’  =>  1000^2 = 10^6 (Mega)
‘G’  =>  1000^3 = 10^9 (Giga)
‘T’  =>  1000^4 = 10^{12} (Tera)
‘P’  =>  1000^5 = 10^{15} (Peta)
‘E’  =>  1000^6 = 10^{18} (Exa)
‘Z’  =>  1000^7 = 10^{21} (Zetta)
‘Y’  =>  1000^8 = 10^{24} (Yotta)
iec

Auto-scale numbers according to the International Electrotechnical Commission (IEC) standard. For input numbers, accept one of the following suffixes. For output numbers, values larger than 1024 will be rounded, and printed with one of the following suffixes:

K’  =>  1024^1 = 2^{10} (Kibi)
‘M’  =>  1024^2 = 2^{20} (Mebi)
‘G’  =>  1024^3 = 2^{30} (Gibi)
‘T’  =>  1024^4 = 2^{40} (Tebi)
‘P’  =>  1024^5 = 2^{50} (Pebi)
‘E’  =>  1024^6 = 2^{60} (Exbi)
‘Z’  =>  1024^7 = 2^{70} (Zebi)
‘Y’  =>  1024^8 = 2^{80} (Yobi)

The iec option uses a single letter suffix (e.g. ‘G’), which is not fully standard, as the iec standard recommends a two-letter symbol (e.g ‘Gi’) - but in practice, this method common. Compare with the iec-i option.

iec-i

Auto-scale numbers according to the International Electrotechnical Commission (IEC) standard. For input numbers, accept one of the following suffixes. For output numbers, values larger than 1024 will be rounded, and printed with one of the following suffixes:

Ki’  =>  1024^1 = 2^{10} (Kibi)
‘Mi’  =>  1024^2 = 2^{20} (Mebi)
‘Gi’  =>  1024^3 = 2^{30} (Gibi)
‘Ti’  =>  1024^4 = 2^{40} (Tebi)
‘Pi’  =>  1024^5 = 2^{50} (Pebi)
‘Ei’  =>  1024^6 = 2^{60} (Exbi)
‘Zi’  =>  1024^7 = 2^{70} (Zebi)
‘Yi’  =>  1024^8 = 2^{80} (Yobi)

The iec-i option uses a two-letter suffix symbol (e.g. ‘Gi’), as the iec standard recommends, but this is not always common in practice. Compare with the iec option.

auto

auto’ can only be used with --from. With this method, numbers with ‘K’,‘M’,‘G’,‘T’,‘P’,‘E’,‘Z’,‘Y’ suffixes are interpreted as SI values, and numbers with ‘Ki’, ‘Mi’,‘Gi’,‘Ti’,‘Pi’,‘Ei’,‘Zi’,‘Yi’ suffixes are interpreted as IEC values.

26.2.3 Examples of using numfmt

Converting a single number from/to human representation:

$ numfmt --to=si 500000
500K

$ numfmt --to=iec 500000
489K

$ numfmt --to=iec-i 500000
489Ki

$ numfmt --from=si 1M
1000000

$ numfmt --from=iec 1M
1048576

# with '--from=auto', M=Mega, Mi=Mebi
$ numfmt --from=auto 1M
1000000
$ numfmt --from=auto 1Mi
1048576

Converting from ‘SI’ to ‘IEC’ scales (e.g. when a harddisk capacity is advertised as ‘1TB’, while checking the drive’s capacity gives lower values):

$ numfmt --from=si --to=iec 1T
932G

Converting a single field from an input file / piped input (these contrived examples are for demonstration purposes only, as both ls and df support the --human-readable option to output sizes in human-readable format):

# Third field (file size) will be shown in SI representation
$ ls -log | numfmt --field 3 --header --to=si | head -n4
-rw-r--r--  1     94K Aug 23  2011 ABOUT-NLS
-rw-r--r--  1    3.7K Jan  7 16:15 AUTHORS
-rw-r--r--  1     36K Jun  1  2011 COPYING
-rw-r--r--  1       0 Jan  7 15:15 ChangeLog

# Second field (size) will be shown in IEC representation
$ df --block-size=1 | numfmt --field 2 --header --to=iec | head -n4
File system   1B-blocks        Used  Available Use% Mounted on
rootfs             132G   104741408   26554036  80% /
tmpfs              794M        7580     804960   1% /run/shm
/dev/sdb1          694G   651424756   46074696  94% /home

Output can be tweaked using --padding or --format:

# Pad to 10 characters, right-aligned
$ du -s * | numfmt --to=si --padding=10
      2.5K config.log
       108 config.status
      1.7K configure
        20 configure.ac

# Pad to 10 characters, left-aligned
$ du -s * | numfmt --to=si --padding=-10
2.5K       config.log
108        config.status
1.7K       configure
20         configure.ac

# Pad to 10 characters, left-aligned, using 'format'
$ du -s * | numfmt --to=si --format="%10f"
      2.5K config.log
       108 config.status
      1.7K configure
        20 configure.ac

# Pad to 10 characters, left-aligned, using 'format'
$ du -s * | numfmt --to=si --padding="%-10f"
2.5K       config.log
108        config.status
1.7K       configure
20         configure.ac

With locales that support grouping digits, using --grouping or --format enables grouping. In ‘POSIX’ locale, grouping is silently ignored:

$ LC_ALL=C numfmt --from=iec --grouping 2G
2147483648

$ LC_ALL=en_US.utf8 numfmt --from=iec --grouping 2G
2,147,483,648

$ LC_ALL=ta_IN numfmt --from=iec --grouping 2G
2,14,74,83,648

$ LC_ALL=C ./src/numfmt --from=iec --format="==%'15f==" 2G
==     2147483648==

$ LC_ALL=en_US.utf8 ./src/numfmt --from=iec --format="==%'15f==" 2G
==  2,147,483,648==

$ LC_ALL=en_US.utf8 ./src/numfmt --from=iec --format="==%'-15f==" 2G
==2,147,483,648  ==

$ LC_ALL=ta_IN ./src/numfmt --from=iec --format="==%'15f==" 2G
== 2,14,74,83,648==

Previous: , Up: Numeric operations   [Contents][Index]

26.3 seq: Print numeric sequences

seq prints a sequence of numbers to standard output. Synopses:

seq [option]… last
seq [option]… first last
seq [option]… first increment last

seq prints the numbers from first to last by increment. By default, each number is printed on a separate line. When increment is not specified, it defaults to ‘1’, even when first is larger than last. first also defaults to ‘1’. So seq 1 prints ‘1’, but seq 0 and seq 10 5 produce no output. The sequence of numbers ends when the sum of the current number and increment would become greater than last, so seq 1 10 10 only produces ‘1’. Floating-point numbers may be specified. See Floating point.

The program accepts the following options. Also see Common options. Options must precede operands.

-f format
--format=format

Print all numbers using format. format must contain exactly one of the ‘printf’-style floating point conversion specifications ‘%a’, ‘%e’, ‘%f’, ‘%g’, ‘%A’, ‘%E’, ‘%F’, ‘%G’. The ‘%’ may be followed by zero or more flags taken from the set ‘-+#0 '’, then an optional width containing one or more digits, then an optional precision consisting of a ‘.’ followed by zero or more digits. format may also contain any number of ‘%%’ conversion specifications. All conversion specifications have the same meaning as with ‘printf’.

The default format is derived from first, step, and last. If these all use a fixed point decimal representation, the default format is ‘%.pf’, where p is the minimum precision that can represent the output numbers exactly. Otherwise, the default format is ‘%g’.

-s string
--separator=string

Separate numbers with string; default is a newline. The output always terminates with a newline.

-w
--equal-width

Print all numbers with the same width, by padding with leading zeros. first, step, and last should all use a fixed point decimal representation. (To have other kinds of padding, use --format).

You can get finer-grained control over output with -f:

$ seq -f '(%9.2E)' -9e5 1.1e6 1.3e6
(-9.00E+05)
( 2.00E+05)
( 1.30E+06)

If you want hexadecimal integer output, you can use printf to perform the conversion:

$ printf '%x\n' $(seq 1048575 1024 1050623)
fffff
1003ff
1007ff

For very long lists of numbers, use xargs to avoid system limitations on the length of an argument list:

$ seq 1000000 | xargs printf '%x\n' | tail -n 3
f423e
f423f
f4240

To generate octal output, use the printf %o format instead of %x.

On most systems, seq can produce whole-number output for values up to at least 2^{53}. Larger integers are approximated. The details differ depending on your floating-point implementation. See Floating point. A common case is that seq works with integers through 2^{64}, and larger integers may not be numerically correct:

$ seq 50000000000000000000 2 50000000000000000004
50000000000000000000
50000000000000000000
50000000000000000004

However, note that when limited to non-negative whole numbers, an increment of 1 and no format-specifying option, seq can print arbitrarily large numbers.

Be careful when using seq with outlandish values: otherwise you may see surprising results, as seq uses floating point internally. For example, on the x86 platform, where the internal representation uses a 64-bit fraction, the command:

seq 1 0.0000000000000000001 1.0000000000000000009

outputs 1.0000000000000000007 twice and skips 1.0000000000000000008.

An exit status of zero indicates success, and a nonzero value indicates failure.


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

27 File permissions

Each file has a set of file mode bits that control the kinds of access that users have to that file. They can be represented either in symbolic form or as an octal number.


Next: , Up: File permissions   [Contents][Index]

27.1 Structure of File Mode Bits

The file mode bits have two parts: the file permission bits, which control ordinary access to the file, and special mode bits, which affect only some files.

There are three kinds of permissions that a user can have for a file:

  1. permission to read the file. For directories, this means permission to list the contents of the directory.
  2. permission to write to (change) the file. For directories, this means permission to create and remove files in the directory.
  3. permission to execute the file (run it as a program). For directories, this means permission to access files in the directory.

There are three categories of users who may have different permissions to perform any of the above operations on a file:

  1. the file’s owner;
  2. other users who are in the file’s group;
  3. everyone else.

Files are given an owner and group when they are created. Usually the owner is the current user and the group is the group of the directory the file is in, but this varies with the operating system, the file system the file is created on, and the way the file is created. You can change the owner and group of a file by using the chown and chgrp commands.

In addition to the three sets of three permissions listed above, the file mode bits have three special components, which affect only executable files (programs) and, on most systems, directories:

  1. Set the process’s effective user ID to that of the file upon execution (called the set-user-ID bit, or sometimes the setuid bit). For directories on a few systems, give files created in the directory the same owner as the directory, no matter who creates them, and set the set-user-ID bit of newly-created subdirectories.
  2. Set the process’s effective group ID to that of the file upon execution (called the set-group-ID bit, or sometimes the setgid bit). For directories on most systems, give files created in the directory the same group as the directory, no matter what group the user who creates them is in, and set the set-group-ID bit of newly-created subdirectories.
  3. Prevent unprivileged users from removing or renaming a file in a directory unless they own the file or the directory; this is called the restricted deletion flag for the directory, and is commonly found on world-writable directories like /tmp.

    For regular files on some older systems, save the program’s text image on the swap device so it will load more quickly when run; this is called the sticky bit.

In addition to the file mode bits listed above, there may be file attributes specific to the file system, e.g., access control lists (ACLs), whether a file is compressed, whether a file can be modified (immutability), and whether a file can be dumped. These are usually set using programs specific to the file system. For example:

ext2

On GNU and GNU/Linux the file attributes specific to the ext2 file system are set using chattr.

FFS

On FreeBSD the file flags specific to the FFS file system are set using chflags.

Even if a file’s mode bits allow an operation on that file, that operation may still fail, because:

For example, if the immutable attribute is set on a file, it cannot be modified, regardless of the fact that you may have just run chmod a+w FILE.


Next: , Previous: , Up: File permissions   [Contents][Index]

27.2 Symbolic Modes

Symbolic modes represent changes to files’ mode bits as operations on single-character symbols. They allow you to modify either all or selected parts of files’ mode bits, optionally based on their previous values, and perhaps on the current umask as well (see Umask and Protection).

The format of symbolic modes is:

[ugoa…][-+=]perms[,…]

where perms is either zero or more letters from the set ‘rwxXst’, or a single letter from the set ‘ugo’.

The following sections describe the operators and other details of symbolic modes.


Next: , Up: Symbolic Modes   [Contents][Index]

27.2.1 Setting Permissions

The basic symbolic operations on a file’s permissions are adding, removing, and setting the permission that certain users have to read, write, and execute or search the file. These operations have the following format:

users operation permissions

The spaces between the three parts above are shown for readability only; symbolic modes cannot contain spaces.

The users part tells which users’ access to the file is changed. It consists of one or more of the following letters (or it can be empty; see Umask and Protection, for a description of what happens then). When more than one of these letters is given, the order that they are in does not matter.

u

the user who owns the file;

g

other users who are in the file’s group;

o

all other users;

a

all users; the same as ‘ugo’.

The operation part tells how to change the affected users’ access to the file, and is one of the following symbols:

+

to add the permissions to whatever permissions the users already have for the file;

-

to remove the permissions from whatever permissions the users already have for the file;

=

to make the permissions the only permissions that the users have for the file.

The permissions part tells what kind of access to the file should be changed; it is normally zero or more of the following letters. As with the users part, the order does not matter when more than one letter is given. Omitting the permissions part is useful only with the ‘=’ operation, where it gives the specified users no access at all to the file.

r

the permission the users have to read the file;

w

the permission the users have to write to the file;

x

the permission the users have to execute the file, or search it if it is a directory.

For example, to give everyone permission to read and write a regular file, but not to execute it, use:

a=rw

To remove write permission for all users other than the file’s owner, use:

go-w

The above command does not affect the access that the owner of the file has to it, nor does it affect whether other users can read or execute the file.

To give everyone except a file’s owner no permission to do anything with that file, use the mode below. Other users could still remove the file, if they have write permission on the directory it is in.

go=

Another way to specify the same thing is:

og-rwx

Next: , Previous: , Up: Symbolic Modes   [Contents][Index]

27.2.2 Copying Existing Permissions

You can base a file’s permissions on its existing permissions. To do this, instead of using a series of ‘r’, ‘w’, or ‘x’ letters after the operator, you use the letter ‘u’, ‘g’, or ‘o’. For example, the mode

o+g

adds the permissions for users who are in a file’s group to the permissions that other users have for the file. Thus, if the file started out as mode 664 (‘rw-rw-r--’), the above mode would change it to mode 666 (‘rw-rw-rw-’). If the file had started out as mode 741 (‘rwxr----x’), the above mode would change it to mode 745 (‘rwxr--r-x’). The ‘-’ and ‘=’ operations work analogously.


Next: , Previous: , Up: Symbolic Modes   [Contents][Index]

27.2.3 Changing Special Mode Bits

In addition to changing a file’s read, write, and execute/search permissions, you can change its special mode bits. See Mode Structure, for a summary of these special mode bits.

To change the file mode bits to set the user ID on execution, use ‘u’ in the users part of the symbolic mode and ‘s’ in the permissions part.

To change the file mode bits to set the group ID on execution, use ‘g’ in the users part of the symbolic mode and ‘s’ in the permissions part.

To set both user and group ID on execution, omit the users part of the symbolic mode (or use ‘a’) and use ‘s’ in the permissions part.

To change the file mode bits to set the restricted deletion flag or sticky bit, omit the users part of the symbolic mode (or use ‘a’) and use ‘t’ in the permissions part.

For example, to set the set-user-ID mode bit of a program, you can use the mode:

u+s

To remove both set-user-ID and set-group-ID mode bits from it, you can use the mode:

a-s

To set the restricted deletion flag or sticky bit, you can use the mode:

+t

The combination ‘o+s’ has no effect. On GNU systems the combinations ‘u+t’ and ‘g+t’ have no effect, and ‘o+t’ acts like plain ‘+t’.

The ‘=’ operator is not very useful with special mode bits. For example, the mode:

o=t

does set the restricted deletion flag or sticky bit, but it also removes all read, write, and execute/search permissions that users not in the file’s group might have had for it.

See Directory Setuid and Setgid, for additional rules concerning set-user-ID and set-group-ID bits and directories.


Next: , Previous: , Up: Symbolic Modes   [Contents][Index]

27.2.4 Conditional Executability

There is one more special type of symbolic permission: if you use ‘X’ instead of ‘x’, execute/search permission is affected only if the file is a directory or already had execute permission.

For example, this mode:

a+X

gives all users permission to search directories, or to execute files if anyone could execute them before.


Next: , Previous: , Up: Symbolic Modes   [Contents][Index]

27.2.5 Making Multiple Changes

The format of symbolic modes is actually more complex than described above (see Setting Permissions). It provides two ways to make multiple changes to files’ mode bits.

The first way is to specify multiple operation and permissions parts after a users part in the symbolic mode.

For example, the mode:

og+rX-w

gives users other than the owner of the file read permission and, if it is a directory or if someone already had execute permission to it, gives them execute/search permission; and it also denies them write permission to the file. It does not affect the permission that the owner of the file has for it. The above mode is equivalent to the two modes:

og+rX
og-w

The second way to make multiple changes is to specify more than one simple symbolic mode, separated by commas. For example, the mode:

a+r,go-w

gives everyone permission to read the file and removes write permission on it for all users except its owner. Another example:

u=rwx,g=rx,o=

sets all of the permission bits for the file explicitly. (It gives users who are not in the file’s group no permission at all for it.)

The two methods can be combined. The mode:

a+r,g+x-w

gives all users permission to read the file, and gives users who are in the file’s group permission to execute/search it as well, but not permission to write to it. The above mode could be written in several different ways; another is:

u+r,g+rx,o+r,g-w

Previous: , Up: Symbolic Modes   [Contents][Index]

27.2.6 The Umask and Protection

If the users part of a symbolic mode is omitted, it defaults to ‘a’ (affect all users), except that any permissions that are set in the system variable umask are not affected. The value of umask can be set using the umask command. Its default value varies from system to system.

Omitting the users part of a symbolic mode is generally not useful with operations other than ‘+’. It is useful with ‘+’ because it allows you to use umask as an easily customizable protection against giving away more permission to files than you intended to.

As an example, if umask has the value 2, which removes write permission for users who are not in the file’s group, then the mode:

+w

adds permission to write to the file to its owner and to other users who are in the file’s group, but not to other users. In contrast, the mode:

a+w

ignores umask, and does give write permission for the file to all users.


Next: , Previous: , Up: File permissions   [Contents][Index]

27.3 Numeric Modes

As an alternative to giving a symbolic mode, you can give an octal (base 8) number that represents the mode. This number is always interpreted in octal; you do not have to add a leading ‘0’, as you do in C. Mode ‘0055’ is the same as mode ‘55’. (However, modes of five digits or more, such as ‘00055’, are sometimes special. See Directory Setuid and Setgid.)

A numeric mode is usually shorter than the corresponding symbolic mode, but it is limited in that normally it cannot take into account the previous file mode bits; it can only set them absolutely. The set-user-ID and set-group-ID bits of directories are an exception to this general limitation. See Directory Setuid and Setgid. Also, operator numeric modes can take previous file mode bits into account. See Operator Numeric Modes.

The permissions granted to the user, to other users in the file’s group, and to other users not in the file’s group each require three bits, which are represented as one octal digit. The three special mode bits also require one bit each, and they are as a group represented as another octal digit. Here is how the bits are arranged, starting with the lowest valued bit:

Value in  Corresponding
Mode      Mode Bit

          Other users not in the file's group:
   1      Execute/search
   2      Write
   4      Read

          Other users in the file's group:
  10      Execute/search
  20      Write
  40      Read

          The file's owner:
 100      Execute/search
 200      Write
 400      Read

          Special mode bits:
1000      Restricted deletion flag or sticky bit
2000      Set group ID on execution
4000      Set user ID on execution

For example, numeric mode ‘4755’ corresponds to symbolic mode ‘u=rwxs,go=rx’, and numeric mode ‘664’ corresponds to symbolic mode ‘ug=rw,o=r’. Numeric mode ‘0’ corresponds to symbolic mode ‘a=’.


Next: , Previous: , Up: File permissions   [Contents][Index]

27.4 Operator Numeric Modes

An operator numeric mode is a numeric mode that is prefixed by a ‘-’, ‘+’, or ‘=’ operator, which has the same interpretation as in symbolic modes. For example, ‘+440’ enables read permission for the file’s owner and group, ‘-1’ disables execute permission for other users, and ‘=600’ clears all permissions except for enabling read-write permissions for the file’s owner. Operator numeric modes can be combined with symbolic modes by separating them with a comma; for example, ‘=0,u+r’ clears all permissions except for enabling read permission for the file’s owner.

The commands ‘chmod =755 dir’ and ‘chmod 755 dir’ differ in that the former clears the directory dir’s setuid and setgid bits, whereas the latter preserves them. See Directory Setuid and Setgid.

Operator numeric modes are a GNU extension.


Previous: , Up: File permissions   [Contents][Index]

27.5 Directories and the Set-User-ID and Set-Group-ID Bits

On most systems, if a directory’s set-group-ID bit is set, newly created subfiles inherit the same group as the directory, and newly created subdirectories inherit the set-group-ID bit of the parent directory. On a few systems, a directory’s set-user-ID bit has a similar effect on the ownership of new subfiles and the set-user-ID bits of new subdirectories. These mechanisms let users share files more easily, by lessening the need to use chmod or chown to share new files. <