ls: support explicit --time=modification selection

[coreutils.git] / doc / coreutils.texi

\input texinfo
@c %**start of header
@setfilename coreutils.info
@include version.texi
@settitle GNU Coreutils @value{VERSION}
@documentencoding UTF-8
@set txicodequoteundirected
@set txicodequotebacktick
@allowcodebreaks false

@c %**end of header

@include constants.texi

@c Define new indices.
@defcodeindex op
@defcodeindex fl

@c Put everything in one index (arbitrarily chosen to be the concept index).
@syncodeindex fl cp
@syncodeindex fn cp
@syncodeindex ky cp
@syncodeindex op cp
@syncodeindex pg cp
@syncodeindex vr cp

@dircategory Basics
@direntry
* Coreutils: (coreutils).       Core GNU (file, text, shell) utilities.
* Common options: (coreutils)Common options.
* File permissions: (coreutils)File permissions.  Access modes.
* Date input formats: (coreutils)Date input formats.
@end direntry

@c FIXME: the following need documentation
@c * [: (coreutils)[ invocation.                   File/string tests.
@c * pinky: (coreutils)pinky invocation.           FIXME.

@dircategory Individual utilities
@direntry
* arch: (coreutils)arch invocation.             Print machine hardware name.
* b2sum: (coreutils)b2sum invocation.           Print or check BLAKE2 digests.
* base32: (coreutils)base32 invocation.         Base32 encode/decode data.
* base64: (coreutils)base64 invocation.         Base64 encode/decode data.
* basename: (coreutils)basename invocation.     Strip directory and suffix.
* basenc: (coreutils)basenc invocation.         Encoding/decoding of data.
* cat: (coreutils)cat invocation.               Concatenate and write files.
* chcon: (coreutils)chcon invocation.           Change SELinux CTX of files.
* chgrp: (coreutils)chgrp invocation.           Change file groups.
* chmod: (coreutils)chmod invocation.           Change access permissions.
* chown: (coreutils)chown invocation.           Change file owners and groups.
* chroot: (coreutils)chroot invocation.         Specify the root directory.
* cksum: (coreutils)cksum invocation.           Print POSIX CRC checksum.
* comm: (coreutils)comm invocation.             Compare sorted files by line.
* cp: (coreutils)cp invocation.                 Copy files.
* csplit: (coreutils)csplit invocation.         Split by context.
* cut: (coreutils)cut invocation.               Print selected parts of lines.
* date: (coreutils)date invocation.             Print/set system date and time.
* dd: (coreutils)dd invocation.                 Copy and convert a file.
* df: (coreutils)df invocation.                 Report file system usage.
* dir: (coreutils)dir invocation.               List directories briefly.
* dircolors: (coreutils)dircolors invocation.   Color setup for ls.
* dirname: (coreutils)dirname invocation.       Strip last file name component.
* du: (coreutils)du invocation.                 Report file usage.
* echo: (coreutils)echo invocation.             Print a line of text.
* env: (coreutils)env invocation.               Modify the environment.
* expand: (coreutils)expand invocation.         Convert tabs to spaces.
* expr: (coreutils)expr invocation.             Evaluate expressions.
* factor: (coreutils)factor invocation.         Print prime factors
* false: (coreutils)false invocation.           Do nothing, unsuccessfully.
* fmt: (coreutils)fmt invocation.               Reformat paragraph text.
* fold: (coreutils)fold invocation.             Wrap long input lines.
* groups: (coreutils)groups invocation.         Print group names a user is in.
* head: (coreutils)head invocation.             Output the first part of files.
* hostid: (coreutils)hostid invocation.         Print numeric host identifier.
* hostname: (coreutils)hostname invocation.     Print or set system name.
* id: (coreutils)id invocation.                 Print user identity.
* install: (coreutils)install invocation.       Copy files and set attributes.
* join: (coreutils)join invocation.             Join lines on a common field.
* kill: (coreutils)kill invocation.             Send a signal to processes.
* link: (coreutils)link invocation.             Make hard links between files.
* ln: (coreutils)ln invocation.                 Make links between files.
* logname: (coreutils)logname invocation.       Print current login name.
* ls: (coreutils)ls invocation.                 List directory contents.
* md5sum: (coreutils)md5sum invocation.         Print or check MD5 digests.
* mkdir: (coreutils)mkdir invocation.           Create directories.
* mkfifo: (coreutils)mkfifo invocation.         Create FIFOs (named pipes).
* mknod: (coreutils)mknod invocation.           Create special files.
* mktemp: (coreutils)mktemp invocation.         Create temporary files.
* mv: (coreutils)mv invocation.                 Rename files.
* nice: (coreutils)nice invocation.             Modify niceness.
* nl: (coreutils)nl invocation.                 Number lines and write files.
* nohup: (coreutils)nohup invocation.           Immunize to hangups.
* nproc: (coreutils)nproc invocation.           Print the number of processors.
* numfmt: (coreutils)numfmt invocation.         Reformat numbers.
* od: (coreutils)od invocation.                 Dump files in octal, etc.
* paste: (coreutils)paste invocation.           Merge lines of files.
* pathchk: (coreutils)pathchk invocation.       Check file name portability.
* pr: (coreutils)pr invocation.                 Paginate or columnate files.
* printenv: (coreutils)printenv invocation.     Print environment variables.
* printf: (coreutils)printf invocation.         Format and print data.
* ptx: (coreutils)ptx invocation.               Produce permuted indexes.
* pwd: (coreutils)pwd invocation.               Print working directory.
* readlink: (coreutils)readlink invocation.     Print referent of a symlink.
* realpath: (coreutils)realpath invocation.     Print resolved file names.
* rm: (coreutils)rm invocation.                 Remove files.
* rmdir: (coreutils)rmdir invocation.           Remove empty directories.
* runcon: (coreutils)runcon invocation.         Run in specified SELinux CTX.
* seq: (coreutils)seq invocation.               Print numeric sequences
* sha1sum: (coreutils)sha1sum invocation.       Print or check SHA-1 digests.
* sha2: (coreutils)sha2 utilities.              Print or check SHA-2 digests.
* shred: (coreutils)shred invocation.           Remove files more securely.
* shuf: (coreutils)shuf invocation.             Shuffling text files.
* sleep: (coreutils)sleep invocation.           Delay for a specified time.
* sort: (coreutils)sort invocation.             Sort text files.
* split: (coreutils)split invocation.           Split into pieces.
* stat: (coreutils)stat invocation.             Report file(system) status.
* stdbuf: (coreutils)stdbuf invocation.         Modify stdio buffering.
* stty: (coreutils)stty invocation.             Print/change terminal settings.
* sum: (coreutils)sum invocation.               Print traditional checksum.
* sync: (coreutils)sync invocation.             Sync files to stable storage.
* tac: (coreutils)tac invocation.               Reverse files.
* tail: (coreutils)tail invocation.             Output the last part of files.
* tee: (coreutils)tee invocation.               Redirect to multiple files.
* test: (coreutils)test invocation.             File/string tests.
* timeout: (coreutils)timeout invocation.       Run with time limit.
* touch: (coreutils)touch invocation.           Change file timestamps.
* tr: (coreutils)tr invocation.                 Translate characters.
* true: (coreutils)true invocation.             Do nothing, successfully.
* truncate: (coreutils)truncate invocation.     Shrink/extend size of a file.
* tsort: (coreutils)tsort invocation.           Topological sort.
* tty: (coreutils)tty invocation.               Print terminal name.
* uname: (coreutils)uname invocation.           Print system information.
* unexpand: (coreutils)unexpand invocation.     Convert spaces to tabs.
* uniq: (coreutils)uniq invocation.             Uniquify files.
* unlink: (coreutils)unlink invocation.         Removal via unlink(2).
* uptime: (coreutils)uptime invocation.         Print uptime and load.
* users: (coreutils)users invocation.           Print current user names.
* vdir: (coreutils)vdir invocation.             List directories verbosely.
* wc: (coreutils)wc invocation.                 Line, word, and byte counts.
* who: (coreutils)who invocation.               Print who is logged in.
* whoami: (coreutils)whoami invocation.         Print effective user ID.
* yes: (coreutils)yes invocation.               Print a string indefinitely.
@end direntry

@copying
This manual documents version @value{VERSION} of the GNU core
utilities, including the standard programs for text and file manipulation.

Copyright @copyright{} 1994--2022 Free Software Foundation, Inc.

@quotation
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''.
@end quotation
@end copying

@titlepage
@title GNU @code{Coreutils}
@subtitle Core GNU utilities
@subtitle for version @value{VERSION}, @value{UPDATED}
@author David MacKenzie et al.

@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage
@shortcontents
@contents

@ifnottex
@node Top
@top GNU Coreutils

@insertcopying
@end ifnottex

@cindex core utilities
@cindex text utilities
@cindex shell utilities
@cindex file utilities

@menu
* Introduction::                 Caveats, overview, and authors
* Common options::               Common options
* Output of entire files::       cat tac nl od base32 base64 basenc
* Formatting file contents::     fmt pr fold
* Output of parts of files::     head tail split csplit
* Summarizing files::            wc sum cksum b2sum md5sum sha1sum sha2
* Operating on sorted files::    sort shuf uniq comm ptx tsort
* Operating on fields::          cut paste join
* Operating on characters::      tr expand unexpand
* Directory listing::            ls dir vdir dircolors
* Basic operations::             cp dd install mv rm shred
* Special file types::         mkdir rmdir unlink mkfifo mknod ln link readlink
* Changing file attributes::     chgrp chmod chown touch
* File space usage::             df du stat sync truncate
* Printing text::                echo printf yes
* Conditions::                   false true test expr
* Redirection::                  tee
* File name manipulation::       dirname basename pathchk mktemp realpath
* Working context::              pwd stty printenv tty
* User information::             id logname whoami groups users who
* System context::               date arch nproc uname hostname hostid uptime
* SELinux context::              chcon runcon
* Modified command invocation::  chroot env nice nohup stdbuf timeout
* Process control::              kill
* Delaying::                     sleep
* Numeric operations::           factor numfmt seq
* File permissions::             Access modes
* File timestamps::              File timestamp issues
* Date input formats::           Specifying date strings
* Version sort ordering::        Details on version-sort algorithm
* Opening the software toolbox:: The software tools philosophy
* GNU Free Documentation License:: Copying and sharing this manual
* Concept index::                General index

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

Common Options

* Exit status::                  Indicating program success or failure
* Backup options::               Backup options
* Block size::                   Block size
* Floating point::               Floating point number representation
* Signal specifications::        Specifying signals
* Disambiguating names and IDs:: chgrp, chown, chroot, id: user and group syntax
* Random sources::               Sources of random data
* Target directory::             Target directory
* Trailing slashes::             Trailing slashes
* Traversing symlinks::          Traversing symlinks to directories
* Treating / specially::         Treating / specially
* Standards conformance::        Standards conformance
* Multi-call invocation::        Multi-call program invocation

Output of entire files

* cat invocation::               Concatenate and write files
* tac invocation::               Concatenate and write files in reverse
* nl invocation::                Number lines and write files
* od invocation::                Write files in octal or other formats
* base32 invocation::            Transform data into printable data
* base64 invocation::            Transform data into printable data
* basenc invocation::            Transform data into printable data

Formatting file contents

* fmt invocation::               Reformat paragraph text
* pr invocation::                Paginate or columnate files for printing
* fold invocation::              Wrap input lines to fit in specified width

Output of parts of files

* head invocation::              Output the first part of files
* tail invocation::              Output the last part of files
* split invocation::             Split a file into fixed-size pieces
* csplit invocation::            Split a file into context-determined pieces

Summarizing files

* wc invocation::                Print newline, word, and byte counts
* sum invocation::               Print checksum and block counts
* cksum invocation::             Print CRC checksum and byte counts
* b2sum invocation::             Print or check BLAKE2 digests
* md5sum invocation::            Print or check MD5 digests
* sha1sum invocation::           Print or check SHA-1 digests
* sha2 utilities::               Print or check SHA-2 digests

Operating on sorted files

* sort invocation::              Sort text files
* shuf invocation::              Shuffle text files
* uniq invocation::              Uniquify files
* comm invocation::              Compare two sorted files line by line
* ptx invocation::               Produce a permuted index of file contents
* tsort invocation::             Topological sort

@command{ptx}: Produce permuted indexes

* General options in ptx::       Options which affect general program behavior
* Charset selection in ptx::     Underlying character set considerations
* Input processing in ptx::      Input fields, contexts, and keyword selection
* Output formatting in ptx::     Types of output format, and sizing the fields
* Compatibility in ptx::         The GNU extensions to @command{ptx}

Operating on fields

* cut invocation::               Print selected parts of lines
* paste invocation::             Merge lines of files
* join invocation::              Join lines on a common field

Operating on characters

* tr invocation::                Translate, squeeze, and/or delete characters
* expand invocation::            Convert tabs to spaces
* unexpand invocation::          Convert spaces to tabs

@command{tr}: Translate, squeeze, and/or delete characters

* Character arrays::             Specifying arrays of characters
* Translating::                  Changing one set of characters to another
* Squeezing and deleting::       Removing characters

Directory listing

* ls invocation::                List directory contents
* dir invocation::               Briefly list directory contents
* vdir invocation::              Verbosely list directory contents
* dircolors invocation::         Color setup for @command{ls}

@command{ls}:  List directory contents

* Which files are listed::       Which files are listed
* What information is listed::   What information is listed
* Sorting the output::           Sorting the output
* General output formatting::    General output formatting
* Formatting the file names::    Formatting the file names

Basic operations

* cp invocation::                Copy files and directories
* dd invocation::                Convert and copy a file
* install invocation::           Copy files and set attributes
* mv invocation::                Move (rename) files
* rm invocation::                Remove files or directories
* shred invocation::             Remove files more securely

Special file types

* link invocation::              Make a hard link via the link syscall
* ln invocation::                Make links between files
* mkdir invocation::             Make directories
* mkfifo invocation::            Make FIFOs (named pipes)
* mknod invocation::             Make block or character special files
* readlink invocation::          Print value of a symlink or canonical file name
* rmdir invocation::             Remove empty directories
* unlink invocation::            Remove files via unlink syscall

Changing file attributes

* chown invocation::             Change file owner and group
* chgrp invocation::             Change group ownership
* chmod invocation::             Change access permissions
* touch invocation::             Change file timestamps

File space usage

* df invocation::                Report file system space usage
* du invocation::                Estimate file space usage
* stat invocation::              Report file or file system status
* sync invocation::              Synchronize cached writes to persistent storage
* truncate invocation::          Shrink or extend the size of a file

Printing text

* echo invocation::              Print a line of text
* printf invocation::            Format and print data
* yes invocation::               Print a string until interrupted

Conditions

* false invocation::             Do nothing, unsuccessfully
* true invocation::              Do nothing, successfully
* test invocation::              Check file types and compare values
* expr invocation::              Evaluate expressions

@command{test}: Check file types and compare values

* File type tests::              File type tests
* Access permission tests::      Access permission tests
* File characteristic tests::    File characteristic tests
* String tests::                 String tests
* Numeric tests::                Numeric tests

@command{expr}: Evaluate expression

* String expressions::           + : match substr index length
* Numeric expressions::          + - * / %
* Relations for expr::           | & < <= = == != >= >
* Examples of expr::             Examples of using @command{expr}

Redirection

* tee invocation::               Redirect output to multiple files or processes

File name manipulation

* basename invocation::          Strip directory and suffix from a file name
* dirname invocation::           Strip last file name component
* pathchk invocation::           Check file name validity and portability
* mktemp invocation::            Create temporary file or directory
* realpath invocation::          Print resolved file names

Working context

* pwd invocation::               Print working directory
* stty invocation::              Print or change terminal characteristics
* printenv invocation::          Print all or some environment variables
* tty invocation::               Print file name of terminal on standard input

@command{stty}: Print or change terminal characteristics

* Control::                      Control settings
* Input::                        Input settings
* Output::                       Output settings
* Local::                        Local settings
* Combination::                  Combination settings
* Characters::                   Special characters
* Special::                      Special settings

User information

* id invocation::                Print user identity
* logname invocation::           Print current login name
* whoami invocation::            Print effective user ID
* groups invocation::            Print group names a user is in
* users invocation::             Print login names of users currently logged in
* who invocation::               Print who is currently logged in

System context

* arch invocation::              Print machine hardware name
* date invocation::              Print or set system date and time
* nproc invocation::             Print the number of processors
* uname invocation::             Print system information
* hostname invocation::          Print or set system name
* hostid invocation::            Print numeric host identifier
* uptime invocation::            Print system uptime and load

@command{date}: Print or set system date and time

* Time conversion specifiers::   %[HIklMNpPrRsSTXzZ]
* Date conversion specifiers::   %[aAbBcCdDeFgGhjmuUVwWxyY]
* Literal conversion specifiers:: %[%nt]
* Padding and other flags::      Pad with zeros, spaces, etc.
* Setting the time::             Changing the system clock
* Options for date::             Instead of the current time
* Date input formats::           Specifying date strings
* Examples of date::             Examples

SELinux context

* chcon invocation::             Change SELinux context of file
* runcon invocation::            Run a command in specified SELinux context

Modified command invocation

* chroot invocation::            Run a command with a different root directory
* env invocation::               Run a command in a modified environment
* nice invocation::              Run a command with modified niceness
* nohup invocation::             Run a command immune to hangups
* stdbuf invocation::            Run a command with modified I/O buffering
* timeout invocation::           Run a command with a time limit

Process control

* kill invocation::              Sending a signal to processes.

Delaying

* sleep invocation::             Delay for a specified time

Numeric operations

* factor invocation::            Print prime factors
* numfmt invocation::            Reformat numbers
* seq invocation::               Print numeric sequences


File timestamps

* File timestamps::              File timestamp issues

File permissions

* Mode Structure::               Structure of file mode bits
* Symbolic Modes::               Mnemonic representation of file mode bits
* Numeric Modes::                File mode bits as octal numbers
* Directory Setuid and Setgid::  Set-user-ID and set-group-ID on directories

Date input formats

* General date syntax::          Common rules
* Calendar date items::          21 Jul 2020
* Time of day items::            9:20pm
* Time zone items::              UTC, -0700, +0900, @dots{}
* Combined date and time of day items:: 2020-07-21T20:02:00,000000-0400
* Day of week items::            Monday and others
* Relative items in date strings:: next tuesday, 2 years ago
* Pure numbers in date strings:: 20200721, 1440
* Seconds since the Epoch::      @@1595289600
* Specifying time zone rules::   TZ="America/New_York", TZ="UTC0"
* Authors of parse_datetime::    Bellovin, Eggert, Salz, Berets, et al.

Version sorting order

* Version sort overview::
* Version sort implementation::
* Differences from Debian version sort::
* Advanced version sort topics::

Opening the software toolbox

* Toolbox introduction::         Toolbox introduction
* I/O redirection::              I/O redirection
* The who command::              The @command{who} command
* The cut command::              The @command{cut} command
* The sort command::             The @command{sort} command
* The uniq command::             The @command{uniq} command
* Putting the tools together::   Putting the tools together

Copying This Manual

* GNU Free Documentation License::     Copying and sharing this manual

@end detailmenu
@end menu


@node Introduction
@chapter 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.

@cindex POSIX
The GNU utilities documented here are mostly compatible with the
POSIX standard.
@cindex bugs, reporting

Please report bugs to @email{bug-coreutils@@gnu.org}.
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.

If you have a problem with @command{sort} or @command{date}, try using the
@option{--debug} option, as it can often help find and fix problems without
having to wait for an answer to a bug report.  If the debug output
does not suffice to fix the problem on your own, please compress and
attach it to the rest of your bug report.

Although diffs are welcome,
please include a description of the problem as well, since this is
sometimes difficult to infer.  @xref{Bugs, , , gcc, Using and Porting GNU CC}.

@cindex Berry, K.
@cindex Paterson, R.
@cindex Stallman, R.
@cindex Pinard, F.
@cindex MacKenzie, D.
@cindex Meyering, J.
@cindex Youmans, B.
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 @command{fmt} man page was written by Ross Paterson.  Fran@,{c}ois
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.

@node Common options
@chapter Common options

@macro optBackup
@item -b
@itemx --backup[=@var{method}]
@opindex -b
@opindex --backup
@vindex VERSION_CONTROL
@cindex backups, making
@xref{Backup options}.
Make a backup of each file that would otherwise be overwritten or removed.
@end macro

@macro optBackupSuffix
@item -S @var{suffix}
@itemx --suffix=@var{suffix}
@opindex -S
@opindex --suffix
Append @var{suffix} to each backup file made with @option{-b}.
@xref{Backup options}.
@end macro

@macro optTargetDirectory
@item -t @var{directory}
@itemx --target-directory=@var{directory}
@opindex -t
@opindex --target-directory
@cindex target directory
@cindex destination directory
Specify the destination @var{directory}.
@xref{Target directory}.
@end macro

@macro optNoTargetDirectory
@item -T
@itemx --no-target-directory
@opindex -T
@opindex --no-target-directory
@cindex target directory
@cindex destination directory
Do not treat the last operand specially when it is a directory or a
symbolic link to a directory.  @xref{Target directory}.
@end macro

@macro outputNUL
@cindex output NUL-byte-terminated lines
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.
@end macro

@macro optNull
@item -0
@itemx --null
@opindex -0
@opindex --null
@outputNUL
@end macro

@macro optZero
@item -z
@itemx --zero
@opindex -z
@opindex --zero
@outputNUL
@end macro

@macro optZeroTerminated
@item -z
@itemx --zero-terminated
@opindex -z
@opindex --zero-terminated
@cindex process zero-terminated items
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 @samp{perl -0} or
@samp{find -print0} and @samp{xargs -0} which do the same in order to
reliably handle arbitrary file names (even those containing blanks
or other special characters).
@end macro

@macro optSi
@item --si
@opindex --si
@cindex SI output
Append an SI-style abbreviation to each size, such as @samp{M} for
megabytes.  Powers of 1000 are used, not 1024; @samp{M} stands for
1,000,000 bytes.  This option is equivalent to
@option{--block-size=si}.  Use the @option{-h} or
@option{--human-readable} option if
you prefer powers of 1024.
@end macro

@macro optHumanReadable
@item -h
@itemx --human-readable
@opindex -h
@opindex --human-readable
@cindex human-readable output
Append a size letter to each size, such as @samp{M} for mebibytes.
Powers of 1024 are used, not 1000; @samp{M} stands for 1,048,576 bytes.
This option is equivalent to @option{--block-size=human-readable}.
Use the @option{--si} option if you prefer powers of 1000.
@end macro

@macro optStripTrailingSlashes
@item --strip-trailing-slashes
@opindex --strip-trailing-slashes
@cindex stripping trailing slashes
Remove any trailing slashes from each @var{source} argument.
@xref{Trailing slashes}.
@end macro

@macro mayConflictWithShellBuiltIn{cmd}
@cindex conflicts with shell built-ins
@cindex built-in shell commands, conflicts with
Due to shell aliases and built-in @command{\cmd\} functions, using an
unadorned @command{\cmd\} interactively or in a script may get you
different functionality than that described here.  Invoke it via
@command{env} (i.e., @code{env \cmd\ @dots{}}) to avoid interference
from the shell.

@end macro

@macro multiplierSuffixes{varName}
@var{\varName\} may be, or may be an integer optionally followed by,
one of the following multiplicative suffixes:
@example
@samp{b}  =>            512 ("blocks")
@samp{KB} =>           1000 (KiloBytes)
@samp{K}  =>           1024 (KibiBytes)
@samp{MB} =>      1000*1000 (MegaBytes)
@samp{M}  =>      1024*1024 (MebiBytes)
@samp{GB} => 1000*1000*1000 (GigaBytes)
@samp{G}  => 1024*1024*1024 (GibiBytes)
@end example
and so on for @samp{T}, @samp{P}, @samp{E}, @samp{Z}, and @samp{Y}.
Binary prefixes can be used, too: @samp{KiB}=@samp{K}, @samp{MiB}=@samp{M},
and so on.
@end macro

@c FIXME: same as above, but no ``blocks'' line.
@macro multiplierSuffixesNoBlocks{varName}
@var{\varName\} may be, or may be an integer optionally followed by,
one of the following multiplicative suffixes:
@example
@samp{KB} =>           1000 (KiloBytes)
@samp{K}  =>           1024 (KibiBytes)
@samp{MB} =>      1000*1000 (MegaBytes)
@samp{M}  =>      1024*1024 (MebiBytes)
@samp{GB} => 1000*1000*1000 (GigaBytes)
@samp{G}  => 1024*1024*1024 (GibiBytes)
@end example
and so on for @samp{T}, @samp{P}, @samp{E}, @samp{Z}, and @samp{Y}.
Binary prefixes can be used, too: @samp{KiB}=@samp{K}, @samp{MiB}=@samp{M},
and so on.
@end macro

@cindex 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.)

@vindex POSIXLY_CORRECT
Normally options and operands can appear in any order, and programs act
as if all the options appear before any operands.  For example,
@samp{sort -r passwd -t :} acts like @samp{sort -r -t : passwd}, since
@samp{:} is an option-argument of @option{-t}.  However, if the
@env{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
@samp{-}.  With such a program, options must precede operands even if
@env{POSIXLY_CORRECT} is not set, and this fact is noted in the
program description.  For example, the @command{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, @samp{rmdir
--ignore-fail-on-non-empty} can be invoked as @samp{rmdir
--ignore-fail} or even @samp{rmdir --i}.  Ambiguous options, such as
@samp{ls --h}, are identified as such.

Some of these programs recognize the @option{--help} and @option{--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.

@table @samp

@item --help
@opindex --help
@cindex help, online
Print a usage message listing all available options, then exit successfully.

@item --version
@opindex --version
@cindex version number, finding
Print the version number, then exit successfully.

@item --
@opindex --
@cindex option delimiter
Delimit the option list.  Later arguments, if any, are treated as
operands even if they begin with @samp{-}.  For example, @samp{sort --
-r} reads from the file named @file{-r}.

@end table

@cindex standard input
@cindex standard output
A single @samp{-} 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, @samp{sort -}
reads from standard input, and is equivalent to plain @samp{sort}.  Unless
otherwise specified, a @samp{-} can appear as any operand that requires a file
name.

@menu
* Exit status::                 Indicating program success or failure.
* Backup options::              -b -S, in some programs.
* Block size::                  BLOCK_SIZE and --block-size, in some programs.
* Floating point::              Floating point number representation.
* Signal specifications::       Specifying signals using the --signal option.
* Disambiguating names and IDs:: chgrp, chown, chroot, id: user and group syntax
* Random sources::              --random-source, in some programs.
* Target directory::            Specifying a target directory, in some programs.
* Trailing slashes::            --strip-trailing-slashes, in some programs.
* Traversing symlinks::         -H, -L, or -P, in some programs.
* Treating / specially::        --preserve-root and --no-preserve-root.
* Special built-in utilities::  @command{break}, @command{:}, @dots{}
* Standards conformance::       Conformance to the POSIX standard.
* Multi-call invocation::       Multi-call program invocation.
@end menu


@node Exit status
@section Exit status

@macro exitstatus
An exit status of zero indicates success,
and a nonzero value indicates failure.
@end macro

Nearly every command invocation yields an integral @dfn{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
@samp{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 @samp{0} and @samp{1}.
Here are some of the exceptions:
@command{chroot}, @command{env}, @command{expr}, @command{nice},
@command{nohup}, @command{numfmt}, @command{printenv}, @command{sort},
@command{stdbuf}, @command{test}, @command{timeout}, @command{tty}.


@node Backup options
@section Backup options

@cindex backup options

Some GNU programs (at least @command{cp}, @command{install},
@command{ln}, and @command{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.

@table @samp

@item -b
@itemx --backup[=@var{method}]
@opindex -b
@opindex --backup
@vindex VERSION_CONTROL
@cindex backups, making
Make a backup of each file that would otherwise be overwritten or removed.
Without this option, the original versions are destroyed.
Use @var{method} to determine the type of backups to make.
When this option is used but @var{method} is not specified,
then the value of the @env{VERSION_CONTROL}
environment variable is used.  And if @env{VERSION_CONTROL} is not set,
the default backup type is @samp{existing}.

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

@vindex version-control @r{Emacs variable}
This option corresponds to the Emacs variable @samp{version-control};
the values for @var{method} are the same as those used in Emacs.
This option also accepts more descriptive names.
The valid @var{method}s are (unique abbreviations are accepted):

@table @samp
@item none
@itemx off
@opindex none @r{backup method}
Never make backups.

@item numbered
@itemx t
@opindex numbered @r{backup method}
Always make numbered backups.

@item existing
@itemx nil
@opindex existing @r{backup method}
Make numbered backups of files that already have them, simple backups
of the others.

@item simple
@itemx never
@opindex simple @r{backup method}
Always make simple backups.  Please note @samp{never} is not to be
confused with @samp{none}.

@end table

@item -S @var{suffix}
@itemx --suffix=@var{suffix}
@opindex -S
@opindex --suffix
@cindex backup suffix
@vindex SIMPLE_BACKUP_SUFFIX
Append @var{suffix} to each backup file made with @option{-b}.  If this
option is not specified, the value of the @env{SIMPLE_BACKUP_SUFFIX}
environment variable is used.  And if @env{SIMPLE_BACKUP_SUFFIX} is not
set, the default is @samp{~}, just as in Emacs.

@end table

@node Block size
@section Block size

@cindex block size

Some GNU programs (at least @command{df}, @command{du}, and
@command{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.

@opindex --block-size=@var{size}
@vindex BLOCKSIZE
@vindex BLOCK_SIZE
@vindex DF_BLOCK_SIZE
@vindex DU_BLOCK_SIZE
@vindex LS_BLOCK_SIZE
@vindex POSIXLY_CORRECT@r{, and block size}

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

@table @code

@item DF_BLOCK_SIZE
This specifies the default block size for the @command{df} command.
Similarly, @env{DU_BLOCK_SIZE} specifies the default for @command{du} and
@env{LS_BLOCK_SIZE} for @command{ls}.

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

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

@item POSIXLY_CORRECT
If neither @env{@var{command}_BLOCK_SIZE}, nor @env{BLOCK_SIZE}, nor
@env{BLOCKSIZE} is set, but this variable is set, the block size
defaults to 512.

@end table

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 @command{ls} file sizes, the block size
defaults to 1 byte.

@cindex human-readable output
@cindex SI output

A block size specification can be a positive integer specifying the number
of bytes per block, or it can be @code{human-readable} or @code{si} to
select a human-readable format.  Integers may be followed by suffixes
that are upward compatible with the
@uref{http://www.bipm.org/en/publications/si-brochure/chapter3.html,
SI prefixes}
for decimal multiples and with the
@uref{https://physics.nist.gov/cuu/Units/binary.html, 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 @samp{M} for megabytes.  @code{BLOCK_SIZE=human-readable} uses
powers of 1024; @samp{M} stands for 1,048,576 bytes.
@code{BLOCK_SIZE=si} is similar, but uses powers of 1000 and appends
@samp{B}; @samp{MB} stands for 1,000,000 bytes.

@vindex LC_NUMERIC
A block size specification preceded by @samp{'} causes output sizes to
be displayed with thousands separators.  The @env{LC_NUMERIC} locale
specifies the thousands separator and grouping.  For example, in an
American English locale, @samp{--block-size="'1kB"} would cause a size
of 1234000 bytes to be displayed as @samp{1,234}.  In the default C
locale, there is no thousands separator so a leading @samp{'} 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 @samp{iB}, specifies
a multiple using powers of 1024.  A size letter followed by @samp{B}
specifies powers of 1000 instead.  For example, @samp{1M} and
@samp{1MiB} are equivalent to @samp{1048576}, whereas @samp{1MB} is
equivalent to @samp{1000000}.

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

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

@table @samp
@item kB
@cindex kilobyte, definition of
kilobyte: @math{10^3 = 1000}.
@item k
@itemx K
@itemx KiB
@cindex kibibyte, definition of
kibibyte: @math{2^{10} = 1024}.  @samp{K} is special: the SI prefix is
@samp{k} and the ISO/IEC 80000-13 prefix is @samp{Ki}, but tradition and
POSIX use @samp{k} to mean @samp{KiB}.
@item MB
@cindex megabyte, definition of
megabyte: @math{10^6 = 1,000,000}.
@item M
@itemx MiB
@cindex mebibyte, definition of
mebibyte: @math{2^{20} = 1,048,576}.
@item GB
@cindex gigabyte, definition of
gigabyte: @math{10^9 = 1,000,000,000}.
@item G
@itemx GiB
@cindex gibibyte, definition of
gibibyte: @math{2^{30} = 1,073,741,824}.
@item TB
@cindex terabyte, definition of
terabyte:  @math{10^{12} = 1,000,000,000,000}.
@item T
@itemx TiB
@cindex tebibyte, definition of
tebibyte: @math{2^{40} = 1,099,511,627,776}.
@item PB
@cindex petabyte, definition of
petabyte: @math{10^{15} = 1,000,000,000,000,000}.
@item P
@itemx PiB
@cindex pebibyte, definition of
pebibyte: @math{2^{50} = 1,125,899,906,842,624}.
@item EB
@cindex exabyte, definition of
exabyte: @math{10^{18} = 1,000,000,000,000,000,000}.
@item E
@itemx EiB
@cindex exbibyte, definition of
exbibyte: @math{2^{60} = 1,152,921,504,606,846,976}.
@item ZB
@cindex zettabyte, definition of
zettabyte: @math{10^{21} = 1,000,000,000,000,000,000,000}
@item Z
@itemx ZiB
@math{2^{70} = 1,180,591,620,717,411,303,424}.
@item YB
@cindex yottabyte, definition of
yottabyte: @math{10^{24} = 1,000,000,000,000,000,000,000,000}.
@item Y
@itemx YiB
@math{2^{80} = 1,208,925,819,614,629,174,706,176}.
@end table

@opindex -k
@opindex -h
@opindex --block-size
@opindex --human-readable
@opindex --si

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

@node Floating point
@section Floating point numbers
@cindex floating point
@cindex IEEE floating point

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
@uref{https://@/docs.oracle.com/@/cd/@/E19957-01/@/806-3568/@/ncg_goldberg.html,
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 @code{strtod} and @code{strtold} to
convert from text to floating point numbers.  These floating point
numbers therefore can use scientific notation like @code{1.0e-34} and
@code{-10e100}.  Commands that parse floating point also understand
case-insensitive @code{inf}, @code{infinity}, and @code{NaN}, although
whether such values are useful depends on the command in question.
Modern C implementations also accept hexadecimal floating point
numbers such as @code{-0x.ep-3}, which stands for @minus{}14/16 times
@math{2^-3}, which equals @minus{}0.109375.  @xref{Parsing of
Floats,,, libc, The GNU C Library Reference Manual}.

@vindex LC_NUMERIC
Normally the @env{LC_NUMERIC} locale determines the decimal-point
character.  However, some commands' descriptions specify that they
accept numbers in either the current or the C locale; for example,
they treat @samp{3.14} like @samp{3,14} if the current locale uses
comma as a decimal point.

@node Signal specifications
@section Signal specifications
@cindex signals, specifying

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

@table @samp
@item HUP
1.  Hangup.
@item INT
2.  Terminal interrupt.
@item QUIT
3.  Terminal quit.
@item ABRT
6.  Process abort.
@item KILL
9.  Kill (cannot be caught or ignored).
@item ALRM
14.  Alarm Clock.
@item TERM
15.  Termination.
@end table

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

@table @samp
@item BUS
Access to an undefined portion of a memory object.
@item CHLD
Child process terminated, stopped, or continued.
@item CONT
Continue executing, if stopped.
@item FPE
Erroneous arithmetic operation.
@item ILL
Illegal Instruction.
@item PIPE
Write on a pipe with no one to read it.
@item SEGV
Invalid memory reference.
@item STOP
Stop executing (cannot be caught or ignored).
@item TSTP
Terminal stop.
@item TTIN
Background process attempting read.
@item TTOU
Background process attempting write.
@item URG
High bandwidth data is available at a socket.
@item USR1
User-defined signal 1.
@item USR2
User-defined signal 2.
@end table

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

@table @samp
@item POLL
Pollable event.
@item PROF
Profiling timer expired.
@item SYS
Bad system call.
@item TRAP
Trace/breakpoint trap.
@item VTALRM
Virtual timer expired.
@item XCPU
CPU time limit exceeded.
@item XFSZ
File size limit exceeded.
@end table

@noindent
POSIX 1003.1-2001 systems that support the XRT extension
also support at least eight real-time signals called @samp{RTMIN},
@samp{RTMIN+1}, @dots{}, @samp{RTMAX-1}, @samp{RTMAX}.

@node Disambiguating names and IDs
@section chown, chgrp, chroot, id: Disambiguating user names and IDs
@cindex user names, disambiguating
@cindex user IDs, disambiguating
@cindex group names, disambiguating
@cindex group IDs, disambiguating
@cindex disambiguating group names and IDs

Since the @var{user} and @var{group} arguments to these commands
may be specified as names or numeric IDs, there is an
apparent ambiguity.
What if a user or group @emph{name} is a string of digits?
Should the command interpret it as a user name or as an ID@?
(Using a number as a user name is common in some environments.)
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
@samp{42} is a user name that maps to some other user ID, say 1000.
Simply invoking @code{chown 42 F}, will set @file{F}s owner ID to
1000---not what you intended.

GNU @command{chown}, @command{chgrp}, @command{chroot}, and @command{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 @samp{+},
in order to force its interpretation as an integer:

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

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

@node Random sources
@section Sources of random data

@cindex random sources

The @command{shuf}, @command{shred}, and @command{sort} commands
sometimes need random data to do their work.  For example, @samp{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 @option{--random-source=@var{file}} option.
An error is reported if @var{file} does not contain enough bytes.

For example, the device file @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.

@file{/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 @file{/dev/random} or
@file{/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.
@cindex random seed
Rather than depending on a file, one can generate a reproducible
arbitrary amount of pseudo-random data given a seed value, using
for example:

@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)
@end example

@node Target directory
@section Target directory

@cindex target directory

The @command{cp}, @command{install}, @command{ln}, and @command{mv}
commands normally treat the last operand specially when it is a
directory or a symbolic link to a directory.  For example, @samp{cp
source dest} is equivalent to @samp{cp source dest/source} if
@file{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:

@table @samp

@item -T
@itemx --no-target-directory
@opindex --no-target-directory
@cindex target directory
@cindex destination 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
@samp{mv /tmp/source /tmp/dest} succeeds, there is no guarantee that
@file{/tmp/source} was renamed to @file{/tmp/dest}: it could have been
renamed to @file{/tmp/dest/source} instead, if some other process
created @file{/tmp/dest} as a directory.  However, if @file{mv
-T /tmp/source /tmp/dest} succeeds, there is no
question that @file{/tmp/source} was renamed to @file{/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 @option{--target-directory} (@option{-t}) option.

@item -t @var{directory}
@itemx --target-directory=@var{directory}
@opindex --target-directory
@cindex target directory
@cindex destination directory
Use @var{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 @command{xargs}
program is designed to work well with this convention.

The commands in the @command{mv}-family are unusual in that they take
a variable number of arguments with a special case at the @emph{end}
(namely, the target directory).  This makes it nontrivial to perform some
operations, e.g., ``move all files from here to ../d/'', because
@code{mv * ../d/} might exhaust the argument space, and @code{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 @option{--target-directory} (@option{-t}) option allows the @command{cp},
@command{install}, @command{ln}, and @command{mv} programs to be used
conveniently with @command{xargs}.  For example, you can move the files
from the current directory to a sibling directory, @code{d} like this:

@example
ls | xargs mv -t ../d --
@end example

However, this doesn't move files whose names begin with @samp{.}.
If you use the GNU @command{find} program, you can move those
files too, with this command:

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

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 @command{find} and GNU @command{xargs}:

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

@end table

@noindent
The @option{--target-directory} (@option{-t}) and
@option{--no-target-directory} (@option{-T})
options cannot be combined.

@node Trailing slashes
@section Trailing slashes

@cindex trailing slashes

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

This is useful when a @var{source} argument may have a trailing slash and
@c FIXME: mv's behavior in this case is system-dependent
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, @command{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 @emph{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.

@node Traversing symlinks
@section Traversing symlinks

@cindex symbolic link to directory, controlling traversal of

The following options modify how @command{chown} and @command{chgrp}
@c FIXME: note that 'du' has these options, too, but they have slightly
@c different meaning.
traverse a hierarchy when the @option{--recursive} (@option{-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 @option{--dereference} and
@option{--no-dereference} (@option{-h}), which control whether to modify
a symlink or its referent.

@table @samp

@macro choptH
@item -H
@opindex -H
@cindex symbolic link to directory, traverse if on the command line
If @option{--recursive} (@option{-R}) is specified and
a command line argument is a symbolic link to a directory, traverse it.
@end macro
@choptH

@macro choptL
@item -L
@opindex -L
@cindex symbolic link to directory, traverse each that is encountered
In a recursive traversal, traverse every symbolic link to a directory
that is encountered.
@end macro

@c Append the following warning to -L where appropriate (e.g. chown).
@macro warnOptDerefWithRec

Combining this dereferencing option with the @option{--recursive} option
may create a security risk:
During the traversal of the directory tree, an attacker may be able to
introduce a symlink to an arbitrary target; when the tool reaches that,
the operation will be performed on the target of that symlink,
possibly allowing the attacker to escalate privileges.

@end macro

@choptL

@macro choptP
@item -P
@opindex -P
@cindex symbolic link to directory, never traverse
Do not traverse any symbolic links.
This is the default if none of @option{-H}, @option{-L},
or @option{-P} is specified.
@end macro
@choptP

@end table


@node Treating / specially
@section Treating @file{/} specially

Certain commands can operate destructively on entire hierarchies.
For example, if a user with appropriate privileges mistakenly runs
@samp{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 @command{rm} normally declines to operate on any directory
that resolves to @file{/}.  If you really want to try to remove all
the files on your system, you can use the @option{--no-preserve-root}
option, but the default behavior, specified by the
@option{--preserve-root} option, is safer for most purposes.

The commands @command{chgrp}, @command{chmod} and @command{chown}
can also operate destructively on entire hierarchies, so they too
support these options.  Although, unlike @command{rm}, they don't
actually unlink files, these commands are arguably more dangerous
when operating recursively on @file{/}, 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 @file{/}, so they default to
@option{--no-preserve-root}, but using the @option{--preserve-root}
option makes them safer for most purposes.  For convenience you can
specify @option{--preserve-root} in an alias or in a shell function.

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

@node Special built-in utilities
@section Special built-in utilities

Some programs like @command{nice} can invoke other programs; for
example, the command @samp{nice cat file} invokes the program
@command{cat} by executing the command @samp{cat file}.  However,
@dfn{special built-in utilities} like @command{exit} cannot be invoked
this way.  For example, the command @samp{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.

@quotation
@t{.@: : break continue eval exec exit export readonly
return set shift times trap unset}
@end quotation

For example, because @samp{.}, @samp{:}, and @samp{exec} are special,
the commands @samp{nice . foo.sh}, @samp{nice :}, and @samp{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 @command{history}, and
@command{suspend}, and with Bash the command @samp{nice suspend}
generates an error message instead of suspending.

@node Standards conformance
@section Standards conformance

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

Newer versions of POSIX are occasionally incompatible with older
versions.  For example, older versions of POSIX required the
command @samp{sort +1} to sort based on the second and succeeding
fields in each input line, but in POSIX 1003.1-2001
the same command is required to sort the file named @file{+1}, and you
must instead use the command @samp{sort -k 2} to get the field-based
sort.  To complicate things further, POSIX 1003.1-2008 allows an
implementation to have either the old or the new behavior.

@vindex _POSIX2_VERSION
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 @env{_POSIX2_VERSION}
environment variable to a value of the form @var{yyyymm} specifying
the year and month the standard was adopted.  Three values are currently
supported for @env{_POSIX2_VERSION}: @samp{199209} stands for
POSIX 1003.2-1992, @samp{200112} stands for POSIX
1003.1-2001, and @samp{200809} stands for POSIX 1003.1-2008.
For example, if you have a POSIX 1003.1-2001 system but are running software
containing traditional usage like @samp{sort +1} or @samp{tail +10},
you can work around the compatibility problems by setting
@samp{_POSIX2_VERSION=200809} in your environment.

@c This node is named "Multi-call invocation", not the usual
@c "coreutils invocation", so that shell commands like
@c 'info coreutils "touch invocation"' work as expected.
@node Multi-call invocation
@section @command{coreutils}: Multi-call program

@pindex multicall
@cindex combined
@cindex calling combined multi-call program

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

@example
coreutils @option{--coreutils-prog=PROGRAM} @dots{}
@end example

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

@node Output of entire files
@chapter Output of entire files

@cindex output of entire files
@cindex entire files, output of

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

@menu
* cat invocation::              Concatenate and write files.
* tac invocation::              Concatenate and write files in reverse.
* nl invocation::               Number lines and write files.
* od invocation::               Write files in octal or other formats.
* base32 invocation::           Transform data into printable data.
* base64 invocation::           Transform data into printable data.
* basenc invocation::           Transform data into printable data.
@end menu

@node cat invocation
@section @command{cat}: Concatenate and write files

@pindex cat
@cindex concatenate and write files
@cindex copying files

@command{cat} copies each @var{file} (@samp{-} means standard input), or
standard input if none are given, to standard output.  Synopsis:

@example
cat [@var{option}] [@var{file}]@dots{}
@end example

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -A
@itemx --show-all
@opindex -A
@opindex --show-all
Equivalent to @option{-vET}.

@item -b
@itemx --number-nonblank
@opindex -b
@opindex --number-nonblank
Number all nonempty output lines, starting with 1.

@item -e
@opindex -e
Equivalent to @option{-vE}.

@item -E
@itemx --show-ends
@opindex -E
@opindex --show-ends
Display a @samp{$} after the end of each line.
The @code{\r\n} combination is shown as @samp{^M$}.

@item -n
@itemx --number
@opindex -n
@opindex --number
Number all output lines, starting with 1.  This option is ignored
if @option{-b} is in effect.

@item -s
@itemx --squeeze-blank
@opindex -s
@opindex --squeeze-blank
@cindex squeezing empty lines
@cindex squeezing blank lines
Suppress repeated adjacent blank lines; output just one empty line
instead of several.

@item -t
@opindex -t
Equivalent to @option{-vT}.

@item -T
@itemx --show-tabs
@opindex -T
@opindex --show-tabs
Display TAB characters as @samp{^I}.

@item -u
@opindex -u
Ignored; for POSIX compatibility.

@item -v
@itemx --show-nonprinting
@opindex -v
@opindex --show-nonprinting
Display control characters except for LFD and TAB using
@samp{^} notation and precede characters that have the high bit set with
@samp{M-}.

@end table

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

@exitstatus

Examples:

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

# Copy standard input to standard output.
cat
@end example


@node tac invocation
@section @command{tac}: Concatenate and write files in reverse

@pindex tac
@cindex reversing files

@command{tac} copies each @var{file} (@samp{-} means standard input), or
standard input if none are given, to standard output, reversing the
records (lines by default) in each separately.  Synopsis:

@example
tac [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@dfn{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 @ref{Common options}.

@table @samp

@item -b
@itemx --before
@opindex -b
@opindex --before
The separator is attached to the beginning of the record that it
precedes in the file.

@item -r
@itemx --regex
@opindex -r
@opindex --regex
Treat the separator string as a regular expression.

@item -s @var{separator}
@itemx --separator=@var{separator}
@opindex -s
@opindex --separator
Use @var{separator} as the record separator, instead of newline.
Note an empty @var{separator} is treated as a zero byte.
I.e., input and output items are delimited with ASCII NUL.

@end table

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

@exitstatus

Example:

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


@node nl invocation
@section @command{nl}: Number lines and write files

@pindex nl
@cindex numbering lines
@cindex line numbering

@command{nl} writes each @var{file} (@samp{-} 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:

@example
nl [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@cindex logical pages, numbering on
@command{nl} decomposes its input into (logical) page sections;
by default, the line number is reset to 1 at each logical page section.
@command{nl} treats all of the input files as a single document;
it does not reset line numbers or logical pages between files.

@cindex headers, numbering
@cindex body, numbering
@cindex footers, numbering
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:

@table @samp
@item \:\:\:
start of header;
@item \:\:
start of body;
@item \:
start of footer.
@end table

The characters from which these strings are made can be changed from
@samp{\} and @samp{:} via options (see below), but the pattern
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 @command{nl} treats a
file that contains no section delimiters as a single body section.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -b @var{style}
@itemx --body-numbering=@var{style}
@opindex -b
@opindex --body-numbering
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:

@table @samp
@item a
number all lines,
@item t
number only nonempty lines (default for body),
@item n
do not number lines (default for header and footer),
@item p@var{bre}
number only lines that contain a match for the basic regular
expression @var{bre}.
@xref{Regular Expressions, , Regular Expressions, grep, The GNU Grep Manual}.
@end table

@item -d @var{cd}
@itemx --section-delimiter=@var{cd}
@opindex -d
@opindex --section-delimiter
@cindex section delimiters of pages
Set the section delimiter characters to @var{cd}; default is
@samp{\:}.  If only @var{c} is given, the second remains @samp{:}.
As a GNU extension more than two characters can be specified,
and also if @var{cd} is empty (@option{-d ''}), then section
matching is disabled.
(Remember to protect @samp{\} or other metacharacters from shell
expansion with quotes or extra backslashes.)

@item -f @var{style}
@itemx --footer-numbering=@var{style}
@opindex -f
@opindex --footer-numbering
Analogous to @option{--body-numbering}.

@item -h @var{style}
@itemx --header-numbering=@var{style}
@opindex -h
@opindex --header-numbering
Analogous to @option{--body-numbering}.

@item -i @var{number}
@itemx --line-increment=@var{number}
@opindex -i
@opindex --line-increment
Increment line numbers by @var{number} (default 1).
@var{number} can be negative to decrement.

@item -l @var{number}
@itemx --join-blank-lines=@var{number}
@opindex -l
@opindex --join-blank-lines
@cindex empty lines, numbering
@cindex blank lines, numbering
Consider @var{number} (default 1) consecutive empty lines to be one
logical line for numbering, and only number the last one.  Where fewer
than @var{number} consecutive empty lines occur, do not number them.
An empty line is one that contains no characters, not even spaces
or tabs.

@item -n @var{format}
@itemx --number-format=@var{format}
@opindex -n
@opindex --number-format
Select the line numbering format (default is @code{rn}):

@table @samp
@item ln
@opindex ln @r{format for @command{nl}}
left justified, no leading zeros;
@item rn
@opindex rn @r{format for @command{nl}}
right justified, no leading zeros;
@item rz
@opindex rz @r{format for @command{nl}}
right justified, leading zeros.
@end table

@item -p
@itemx --no-renumber
@opindex -p
@opindex --no-renumber
Do not reset the line number at the start of a logical page.

@item -s @var{string}
@itemx --number-separator=@var{string}
@opindex -s
@opindex --number-separator
Separate the line number from the text line in the output with
@var{string} (default is the TAB character).

@item -v @var{number}
@itemx --starting-line-number=@var{number}
@opindex -v
@opindex --starting-line-number
Set the initial line number on each logical page to @var{number} (default 1).
The starting @var{number} can be negative.

@item -w @var{number}
@itemx --number-width=@var{number}
@opindex -w
@opindex --number-width
Use @var{number} characters for line numbers (default 6).

@end table

@exitstatus


@node od invocation
@section @command{od}: Write files in octal or other formats

@pindex od
@cindex octal dump of files
@cindex hex dump of files
@cindex ASCII dump of files
@cindex file contents, dumping unambiguously

@command{od} writes an unambiguous representation of each @var{file}
(@samp{-} means standard input), or standard input if none are given.
Synopses:

@example
od [@var{option}]@dots{} [@var{file}]@dots{}
od [-abcdfilosx]@dots{} [@var{file}] [[+]@var{offset}[.][b]]
od [@var{option}]@dots{} --traditional [@var{file}]@c
 [[+]@var{offset}[.][b] [[+]@var{label}[.][b]]]
@end example

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

If @var{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 @samp{0x} or @samp{0X} it is interpreted as a hexadecimal
number.  If there is a trailing @samp{b}, the number of bytes skipped
will be @var{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 @samp{+} or (if there are two
operands) a digit.  For example, in @samp{od foo 10} and @samp{od +10}
the @samp{10} is an offset, whereas in @samp{od 10} the @samp{10} is a
file name.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -A @var{radix}
@itemx --address-radix=@var{radix}
@opindex -A
@opindex --address-radix
@cindex radix for file offsets
@cindex file offset radix
Select the base in which file offsets are printed.  @var{radix} can
be one of the following:

@table @samp
@item d
decimal;
@item o
octal;
@item x
hexadecimal;
@item n
none (do not print offsets).
@end table

The default is octal.

@item --endian=@var{order}
@opindex --endian
@cindex byte-swapping
@cindex endianness
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 @option{--type} size and endian @var{order}, which can be
@samp{little} or @samp{big}.

@item -j @var{bytes}
@itemx --skip-bytes=@var{bytes}
@opindex -j
@opindex --skip-bytes
Skip @var{bytes} input bytes before formatting and writing.  If
@var{bytes} begins with @samp{0x} or @samp{0X}, it is interpreted in
hexadecimal; otherwise, if it begins with @samp{0}, in octal; otherwise,
in decimal.
@multiplierSuffixes{bytes}

@item -N @var{bytes}
@itemx --read-bytes=@var{bytes}
@opindex -N
@opindex --read-bytes
Output at most @var{bytes} bytes of the input.  Prefixes and suffixes on
@code{bytes} are interpreted as for the @option{-j} option.

@item -S @var{bytes}
@itemx --strings[=@var{bytes}]
@opindex -S
@opindex --strings
@cindex string constants, outputting
Instead of the normal output, output only @dfn{string constants}: at
least @var{bytes} consecutive ASCII graphic characters,
followed by a zero byte (ASCII NUL).
Prefixes and suffixes on @var{bytes} are interpreted as for the
@option{-j} option.

If @var{bytes} is omitted with @option{--strings}, the default is 3.

@item -t @var{type}
@itemx --format=@var{type}
@opindex -t
@opindex --format
Select the format in which to output the file data.  @var{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 @var{type}
string, or use this option more than once, @command{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.

@table @samp
@item a
named character, ignoring high-order bit
@item c
printable single byte character, C backslash escape
or a 3 digit octal sequence
@item d
signed decimal
@item f
floating point (@pxref{Floating point})
@item o
octal
@item u
unsigned decimal
@item x
hexadecimal
@end table

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

@cindex type size
Except for types @samp{a} and @samp{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 (@samp{d}, @samp{o},
@samp{u}, @samp{x}):

@table @samp
@item C
char
@item S
short
@item I
int
@item L
long
@end table

For floating point (@code{f}):

@table @asis
@item F
float
@item D
double
@item L
long double
@end table

@item -v
@itemx --output-duplicates
@opindex -v
@opindex --output-duplicates
Output consecutive lines that are identical.  By default, when two or
more consecutive output lines would be identical, @command{od} outputs only
the first line, and puts just an asterisk on the following line to
indicate the elision.

@item -w[@var{n}]
@itemx --width[=@var{n}]
@opindex -w
@opindex --width
Dump @code{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 @var{n} is
omitted, the default is 32.

@end table

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

@table @samp

@item -a
@opindex -a
Output as named characters.  Equivalent to @samp{-t a}.

@item -b
@opindex -b
Output as octal bytes.  Equivalent to @samp{-t o1}.

@item -c
@opindex -c
Output as printable single byte characters, C backslash escapes
or 3 digit octal sequences.  Equivalent to @samp{-t c}.

@item -d
@opindex -d
Output as unsigned decimal two-byte units.  Equivalent to @samp{-t u2}.

@item -f
@opindex -f
Output as floats.  Equivalent to @samp{-t fF}.

@item -i
@opindex -i
Output as decimal ints.  Equivalent to @samp{-t dI}.

@item -l
@opindex -l
Output as decimal long ints.  Equivalent to @samp{-t dL}.

@item -o
@opindex -o
Output as octal two-byte units.  Equivalent to @option{-t o2}.

@item -s
@opindex -s
Output as decimal two-byte units.  Equivalent to @option{-t d2}.

@item -x
@opindex -x
Output as hexadecimal two-byte units.  Equivalent to @samp{-t x2}.

@item --traditional
@opindex --traditional
Recognize the non-option label argument that traditional @command{od}
accepted.  The following syntax:

@example
od --traditional [@var{file}] [[+]@var{offset}[.][b] [[+]@var{label}[.][b]]]
@end example

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

@end table

@exitstatus


@node base32 invocation
@section @command{base32}: Transform data into printable data

@pindex base32
@cindex base32 encoding

@command{base32} transforms data read from a file, or standard input,
into (or from) base32 encoded form.  The base32 encoded form uses
printable ASCII characters to represent binary data.
The usage and options of this command are precisely the
same as for @command{base64}.  @xref{base64 invocation}.


@node base64 invocation
@section @command{base64}: Transform data into printable data

@pindex base64
@cindex base64 encoding

@command{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:

@example
base64 [@var{option}]@dots{} [@var{file}]
base64 --decode [@var{option}]@dots{} [@var{file}]
@end example

The base64 encoding expands data to roughly 133% of the original.
The base32 encoding expands data to roughly 160% of the original.
The format conforms to
@uref{https://tools.ietf.org/search/rfc4648, RFC 4648}.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -w @var{cols}
@itemx --wrap=@var{cols}
@opindex -w
@opindex --wrap
@cindex wrap data
@cindex column to wrap data after
During encoding, wrap lines after @var{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.

@item -d
@itemx --decode
@opindex -d
@opindex --decode
@cindex Decode base64 data
@cindex Base64 decoding
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.

@item -i
@itemx --ignore-garbage
@opindex -i
@opindex --ignore-garbage
@cindex Ignore garbage in base64 stream
When decoding, newlines are always accepted.
During decoding, ignore unrecognized bytes,
to permit distorted data to be decoded.

@end table

@exitstatus

@node basenc invocation
@section @command{basenc}: Transform data into printable data

@pindex basenc
@cindex base32 encoding

@command{basenc} transforms data read from a file, or standard input,
into (or from) various common encoding forms.  The encoded form uses
printable ASCII characters to represent binary data.

Synopses:

@example
basenc @var{encoding} [@var{option}]@dots{} [@var{file}]
basenc @var{encoding} --decode [@var{option}]@dots{} [@var{file}]
@end example

The @var{encoding} argument is required. If @var{file} is omitted,
@command{basenc} reads from standard input.
The @option{-w/--wrap},@option{-i/--ignore-garbage},
@option{-d/--decode} options of this command are precisely the same as
for @command{base64}. @xref{base64 invocation}.


Supported @var{encoding}s are:

@table @samp

@item --base64
@opindex --base64
Encode into (or decode from with @option{-d/--decode}) base64 form.
The format conforms to
@uref{https://tools.ietf.org/search/rfc4648#section-4, RFC 4648#4}.
Equivalent to the @command{base64} command.

@item --base64url
@opindex --base64url
Encode into (or decode from with @option{-d/--decode}) file-and-url-safe
base64 form (using @samp{_} and @samp{-} instead of @samp{+} and @samp{/}).
The format conforms to
@uref{https://tools.ietf.org/search/rfc4648#section-5, RFC 4648#5}.

@item --base32
@opindex --base32
Encode into (or decode from with @option{-d/--decode}) base32 form.
The encoded data uses the @samp{ABCDEFGHIJKLMNOPQRSTUVWXYZ234567=} characters.
The format conforms to
@uref{https://tools.ietf.org/search/rfc4648#section-6, RFC 4648#6}.
Equivalent to the @command{base32} command.

@item --base32hex
@opindex --base32hex
Encode into (or decode from with @option{-d/--decode}) Extended Hex Alphabet
base32 form. The encoded data uses the
@samp{0123456789ABCDEFGHIJKLMNOPQRSTUV=} characters. The format conforms to
@uref{https://tools.ietf.org/search/rfc4648#section-7, RFC 4648#7}.

@item --base16
@opindex --base16
Encode into (or decode from with @option{-d/--decode}) base16 (hexadecimal)
form. The encoded data uses the @samp{0123456789ABCDEF} characters. The format
conforms to
@uref{https://tools.ietf.org/search/rfc4648#section-8, RFC 4648#8}.

@item --base2lsbf
@opindex --base2lsbf
Encode into (or decode from with @option{-d/--decode}) binary string form
(@samp{0} and @samp{1}) with the @emph{least} significant bit of every byte
first.

@item --base2msbf
@opindex --base2msbf
Encode into (or decode from with @option{-d/--decode}) binary string form
(@samp{0} and @samp{1}) with the @emph{most} significant bit of every byte
first.

@item --z85
@opindex --z85
Encode into (or decode from with @option{-d/--decode}) Z85 form
(a modified Ascii85 form). The encoded data uses the
@samp{0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTU@
VWXYZ.-:+=^!/*?&<>()[]@{@}@@%$#}.
characters. The format conforms to
@uref{https://rfc.zeromq.org/spec:32/Z85/, ZeroMQ spec:32/Z85}.

When encoding with @option{--z85}, input length must be a multiple of 4;
when decoding with @option{--z85}, input length must be a multiple of 5.

@end table



Encoding/decoding examples:

@example
$ printf '\376\117\202' | basenc --base64
/k+C

$ printf '\376\117\202' | basenc --base64url
_k-C

$ printf '\376\117\202' | basenc --base32
7ZHYE===

$ printf '\376\117\202' | basenc --base32hex
VP7O4===

$ printf '\376\117\202' | basenc --base16
FE4F82

$ printf '\376\117\202' | basenc --base2lsbf
011111111111001001000001

$ printf '\376\117\202' | basenc --base2msbf
111111100100111110000010

$ printf '\376\117\202\000' | basenc --z85
@@.FaC

$ printf 01010100 | basenc --base2msbf --decode
T

$ printf 01010100 | basenc --base2lsbf --decode
*
@end example



@node Formatting file contents
@chapter Formatting file contents

@cindex formatting file contents

These commands reformat the contents of files.

@menu
* fmt invocation::              Reformat paragraph text.
* pr invocation::               Paginate or columnate files for printing.
* fold invocation::             Wrap input lines to fit in specified width.
@end menu


@node fmt invocation
@section @command{fmt}: Reformat paragraph text

@pindex fmt
@cindex reformatting paragraph text
@cindex paragraphs, reformatting
@cindex text, reformatting

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

@example
fmt [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@command{fmt} reads from the specified @var{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.

@cindex line-breaking
@cindex sentences and line-breaking
@cindex Knuth, Donald E.
@cindex Plass, Michael F.
@command{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 @dfn{sentence break} is defined as either the end
of a paragraph or a word ending in any of @samp{.?!}, followed by two
spaces or end of line, ignoring any intervening parentheses or quotes.
Like @TeX{}, @command{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'',
@cite{Software---Practice & Experience} @b{11}, 11 (November 1981),
1119--1184.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -c
@itemx --crown-margin
@opindex -c
@opindex --crown-margin
@cindex crown margin
@dfn{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.

@item -t
@itemx --tagged-paragraph
@opindex -t
@opindex --tagged-paragraph
@cindex tagged paragraphs
@dfn{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.

@item -s
@itemx --split-only
@opindex -s
@opindex --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.

@item -u
@itemx --uniform-spacing
@opindex -u
@opindex --uniform-spacing
Uniform spacing.  Reduce spacing between words to one space, and spacing
between sentences to two spaces.

@item -@var{width}
@itemx -w @var{width}
@itemx --width=@var{width}
@opindex -@var{width}
@opindex -w
@opindex --width
Fill output lines up to @var{width} characters (default 75 or @var{goal}
plus 10, if @var{goal} is provided).

@item -g @var{goal}
@itemx --goal=@var{goal}
@opindex -g
@opindex --goal
@command{fmt} initially tries to make lines @var{goal} characters wide.
By default, this is 7% shorter than @var{width}.

@item -p @var{prefix}
@itemx --prefix=@var{prefix}
Only lines beginning with @var{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.

@end table

@exitstatus

@node pr invocation
@section @command{pr}: Paginate or columnate files for printing

@pindex pr
@cindex printing, preparing files for
@cindex multicolumn output, generating
@cindex merging files in parallel

@command{pr} writes each @var{file} (@samp{-} means standard input), or
standard input if none are given, to standard output, paginating and
optionally outputting in multicolumn format; optionally merges all
@var{file}s, printing all in parallel, one per column.  Synopsis:

@example
pr [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@vindex LC_MESSAGES
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 @var{page_length} is 66
lines.  The default number of text lines is therefore 56.
The text line of the header takes the form
@samp{@var{date} @var{string} @var{page}}, with spaces inserted around
@var{string} so that the line takes up the full @var{page_width}.  Here,
@var{date} is the date (see the @option{-D} or @option{--date-format}
option for details), @var{string} is the centered header string, and
@var{page} identifies the page number.  The @env{LC_MESSAGES} locale
category affects the spelling of @var{page}; in the default C locale, it
is @samp{Page @var{number}} where @var{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 @samp{space}).  For multicolumn output, lines will always be truncated to
@var{page_width} (default 72), unless you use the @option{-J} option.
For single
column output no line truncation occurs by default.  Use @option{-W} option to
truncate lines in that case.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item +@var{first_page}[:@var{last_page}]
@itemx --pages=@var{first_page}[:@var{last_page}]
@c The two following @opindex lines evoke warnings because they contain ':'
@c The 'info' spec does not permit that.  If we use those lines, we end
@c up with truncated index entries that don't work.
@c @opindex +@var{first_page}[:@var{last_page}]
@c @opindex --pages=@var{first_page}[:@var{last_page}]
@opindex +@var{page_range}
@opindex --pages=@var{page_range}
Begin printing with page @var{first_page} and stop with @var{last_page}.
Missing @samp{:@var{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 @samp{+@var{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 @option{-N}
option.

@item -@var{column}
@itemx --columns=@var{column}
@opindex -@var{column}
@opindex --columns
@cindex down columns
With each single @var{file}, produce @var{column} columns of output
(default is 1) and print columns down, unless @option{-a} is used.  The
column width is automatically decreased as @var{column} increases; unless
you use the @option{-W/-w} option to increase @var{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 @option{-e}
and @option{-i} are on for multiple text-column output.  Together with
@option{-J} option column alignment and line truncation is turned off.
Lines of full length are joined in a free field format and @option{-S}
option may set field separators.  @option{-@var{column}} may not be used
with @option{-m} option.

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

@item -c
@itemx --show-control-chars
@opindex -c
@opindex --show-control-chars
Print control characters using hat notation (e.g., @samp{^G}); print
other nonprinting characters in octal backslash notation.  By default,
nonprinting characters are not changed.

@item -d
@itemx --double-space
@opindex -d
@opindex --double-space
@cindex double spacing
Double space the output.

@item -D @var{format}
@itemx --date-format=@var{format}
@cindex time formats
@cindex formatting times
Format header dates using @var{format}, using the same conventions as
for the command @samp{date +@var{format}}.  @xref{date invocation}.
Except for directives, which start with
@samp{%}, characters in @var{format} are printed unchanged.  You can use
this option to specify an arbitrary string in place of the header date,
e.g., @option{--date-format="Monday morning"}.

@vindex POSIXLY_CORRECT
@vindex LC_TIME
The default date format is @samp{%Y-%m-%d %H:%M} (for example,
@samp{2020-07-09 23:59});
but if the @env{POSIXLY_CORRECT} environment variable is set
and the @env{LC_TIME} locale category specifies the POSIX
locale, the default is @samp{%b %e %H:%M %Y} (for example,
@samp{Jul@ @ 9 23:59 2020}.

@vindex TZ
Timestamps are listed according to the time zone rules specified by
the @env{TZ} environment variable, or by the system default rules if
@env{TZ} is not set.  @xref{TZ Variable,, Specifying the Time Zone
with @env{TZ}, libc, The GNU C Library Reference Manual}.

@item -e[@var{in-tabchar}[@var{in-tabwidth}]]
@itemx --expand-tabs[=@var{in-tabchar}[@var{in-tabwidth}]]
@opindex -e
@opindex --expand-tabs
@cindex input tabs
Expand @var{tab}s to spaces on input.  Optional argument @var{in-tabchar} is
the input tab character (default is the TAB character).  Second optional
argument @var{in-tabwidth} is the input tab character's width (default
is 8).

@item -f
@itemx -F
@itemx --form-feed
@opindex -F
@opindex -f
@opindex --form-feed
Use a form feed instead of newlines to separate output pages.  This does
not alter the default page length of 66 lines.

@item -h @var{header}
@itemx --header=@var{header}
@opindex -h
@opindex --header
Replace the file name in the header with the centered string @var{header}.
When using the shell, @var{header} should be quoted and should be
separated from @option{-h} by a space.

@item -i[@var{out-tabchar}[@var{out-tabwidth}]]
@itemx --output-tabs[=@var{out-tabchar}[@var{out-tabwidth}]]
@opindex -i
@opindex --output-tabs
@cindex output tabs
Replace spaces with @var{tab}s on output.  Optional argument @var{out-tabchar}
is the output tab character (default is the TAB character).  Second optional
argument @var{out-tabwidth} is the output tab character's width (default
is 8).

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


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

@item -m
@itemx --merge
@opindex -m
@opindex --merge
Merge and print all @var{file}s in parallel, one in each column.  If a
line is too long to fit in a column, it is truncated, unless the @option{-J}
option is used.  @option{--sep-string[=@var{string}]} may be used.
Empty pages in
some @var{file}s (form feeds set) produce empty columns, still marked
by @var{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
@samp{@var{date} @var{page}} with spaces inserted in the middle; this
may be used with the @option{-h} or @option{--header} option to fill up
the middle blank part.

@item -n[@var{number-separator}[@var{digits}]]
@itemx --number-lines[=@var{number-separator}[@var{digits}]]
@opindex -n
@opindex --number-lines
Provide @var{digits} digit line numbering (default for @var{digits} is
5).  With multicolumn output the number occupies the first @var{digits}
column positions of each text column or only each line of @option{-m}
output.  With single column output the number precedes each line just as
@option{-m} does.  Default counting of the line numbers starts with the
first line of the input file (not the first line printed, compare the
@option{--page} option and @option{-N} option).
Optional argument @var{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 @var{margin} specified
by @option{-o} option.  With multicolumn output priority is given to
@samp{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 @var{margin}.  That means a
fixed number of spaces is always printed in the place of the
@var{number-separator} TAB@.  The tabification depends upon the output
position.

@item -N @var{line_number}
@itemx --first-line-number=@var{line_number}
@opindex -N
@opindex --first-line-number
Start line counting with the number @var{line_number} at first line of
first page printed (in most cases not the first line of the input file).

@item -o @var{margin}
@itemx --indent=@var{margin}
@opindex -o
@opindex --indent
@cindex indenting lines
@cindex left margin
Indent each line with a margin @var{margin} spaces wide (default is zero).
The total page width is the size of the margin plus the @var{page_width}
set with the @option{-W/-w} option.  A limited overflow may occur with
numbered single column output (compare @option{-n} option).

@item -r
@itemx --no-file-warnings
@opindex -r
@opindex --no-file-warnings
Do not print a warning message when an argument @var{file} cannot be
opened.  (The exit status will still be nonzero, however.)

@item -s[@var{char}]
@itemx --separator[=@var{char}]
@opindex -s
@opindex --separator
Separate columns by a single character @var{char}.  The default for
@var{char} is the TAB character without @option{-w} and @samp{no
character} with @option{-w}.  Without @option{-s} the default separator
@samp{space} is set.  @option{-s[char]} turns off line truncation of all
three column options (@option{-COLUMN}|@option{-a -COLUMN}|@option{-m}) unless
@option{-w} is set.  This is a POSIX-compliant formulation.


@item -S[@var{string}]
@itemx --sep-string[=@var{string}]
@opindex -S
@opindex --sep-string
Use @var{string} to separate output columns.  The @option{-S} option doesn't
affect the @option{-W/-w} option, unlike the @option{-s} option which does.  It
does not affect line truncation or column alignment.
Without @option{-S}, and with @option{-J}, @command{pr} uses the default output
separator, TAB@.
Without @option{-S} or @option{-J}, @command{pr} uses a @samp{space}
(same as @option{-S"@w{ }"}).
If no @samp{@var{string}} argument is specified, @samp{""} is assumed.

@item -t
@itemx --omit-header
@opindex -t
@opindex --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.  @option{-t} or @option{-T} may be
useful together with other options; e.g.: @option{-t -e4}, expand TAB characters
in the input file to 4 spaces but don't make any other changes.  Use of
@option{-t} overrides @option{-h}.

@item -T
@itemx --omit-pagination
@opindex -T
@opindex --omit-pagination
Do not print header [and footer].  In addition eliminate all form feeds
set in the input files.

@item -v
@itemx --show-nonprinting
@opindex -v
@opindex --show-nonprinting
Print nonprinting characters in octal backslash notation.

@item -w @var{page_width}
@itemx --width=@var{page_width}
@opindex -w
@opindex --width
Set page width to @var{page_width} characters for multiple text-column
output only (default for @var{page_width} is 72).  The specified
@var{page_width} is rounded down so that columns have equal width.
@option{-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 @var{page_width} setting is possible with single column output.
A POSIX-compliant formulation.

@item -W @var{page_width}
@itemx --page_width=@var{page_width}
@opindex -W
@opindex --page_width
Set the page width to @var{page_width} characters, honored with and
without a column option.  With a column option, the specified @var{page_width}
is rounded down so that columns have equal width.  Text lines are truncated,
unless @option{-J} is used. Together with one of the three column options
(@option{-@var{column}}, @option{-a -@var{column}} or @option{-m}) column
alignment is always used.  The separator options @option{-S} or @option{-s}
don't disable the @option{-W} option.  Default is 72 characters.  Without
@option{-W @var{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 @option{-W 72 -J}@.  The header
line is never truncated.

@end table

@exitstatus


@node fold invocation
@section @command{fold}: Wrap input lines to fit in specified width

@pindex fold
@cindex wrapping long input lines
@cindex folding long input lines

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

@example
fold [@var{option}]@dots{} [@var{file}]@dots{}
@end example

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

@cindex screen columns
@command{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 @ref{Common options}.

@table @samp

@item -b
@itemx --bytes
@opindex -b
@opindex --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.

@item -s
@itemx --spaces
@opindex -s
@opindex --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.

@item -w @var{width}
@itemx --width=@var{width}
@opindex -w
@opindex --width
Use a maximum line length of @var{width} columns instead of 80.

For compatibility @command{fold} supports an obsolete option syntax
@option{-@var{width}}.  New scripts should use @option{-w @var{width}}
instead.

@end table

@exitstatus


@node Output of parts of files
@chapter Output of parts of files

@cindex output of parts of files
@cindex parts of files, output of

These commands output pieces of the input.

@menu
* head invocation::             Output the first part of files.
* tail invocation::             Output the last part of files.
* split invocation::            Split a file into pieces.
* csplit invocation::           Split a file into context-determined pieces.
@end menu

@node head invocation
@section @command{head}: Output the first part of files

@pindex head
@cindex initial part of files, outputting
@cindex first part of files, outputting

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

@example
head [@var{option}]@dots{} [@var{file}]@dots{}
@end example

If more than one @var{file} is specified, @command{head} prints a
one-line header consisting of:

@example
==> @var{file name} <==
@end example

@noindent
before the output for each @var{file}.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -c [-]@var{num}
@itemx --bytes=[-]@var{num}
@opindex -c
@opindex --bytes
Print the first @var{num} bytes, instead of initial lines.
However, if @var{num} is prefixed with a @samp{-},
print all but the last @var{num} bytes of each file.
@multiplierSuffixes{num}

@item -n [-]@var{num}
@itemx --lines=[-]@var{num}
@opindex -n
@opindex --lines
Output the first @var{num} lines.
However, if @var{num} is prefixed with a @samp{-},
print all but the last @var{num} lines of each file.
Size multiplier suffixes are the same as with the @option{-c} option.

@item -q
@itemx --quiet
@itemx --silent
@opindex -q
@opindex --quiet
@opindex --silent
Never print file name headers.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Always print file name headers.

@optZeroTerminated

@end table

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

@exitstatus


@node tail invocation
@section @command{tail}: Output the last part of files

@pindex tail
@cindex last part of files, outputting

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

@example
tail [@var{option}]@dots{} [@var{file}]@dots{}
@end example

If more than one @var{file} is specified, @command{tail} prints a
one-line header before the output for each @var{file}, consisting of:

@example
==> @var{file name} <==
@end example

For further processing of tail output, it can be useful to convert the
file headers to line prefixes, which can be done like:

@example
tail @dots{} |
awk '
  /^==> .* <==$/ @{prefix=substr($0,5,length-8)":"; next@}
  @{print prefix$0@}
' | @dots{}
@end example

@cindex BSD @command{tail}
GNU @command{tail} can output any amount of data (some other versions of
@command{tail} cannot).  It also has no @option{-r} option (print in
reverse), since reversing a file is really a different job from printing
the end of a file; BSD @command{tail} (which is the one with @option{-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 @command{tac} command.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -c [+]@var{num}
@itemx --bytes=[+]@var{num}
@opindex -c
@opindex --bytes
Output the last @var{num} bytes, instead of final lines.
However, if @var{num} is prefixed with a @samp{+}, start printing with
byte @var{num} from the start of each file, instead of from the end.
@multiplierSuffixes{num}

@item -f
@itemx --follow[=@var{how}]
@opindex -f
@opindex --follow
@cindex growing files
@vindex name @r{follow option}
@vindex descriptor @r{follow option}
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, @command{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 @option{--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
@option{--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, @command{tail} prints a message saying the file has been truncated
and resumes tracking from the start of the file, assuming it has been
truncated to 0, which is the usual truncation operation for log files.

When a file is removed, @command{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 @option{--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 @samp{descriptor} and @samp{name} may be specified only
with the long form of the option, not with @option{-f}.

The @option{-f} option is ignored if
no @var{file} operand is specified and standard input is a FIFO or a pipe.
Likewise, the @option{-f} option has no effect for any
operand specified as @samp{-}, 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, @command{tail} sleeps for one second between checks---
use @option{--sleep-interval=@var{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:

@example
alias tail='tail -s.1'
@end example

@item -F
@opindex -F
This option is the same as @option{--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.

@item --max-unchanged-stats=@var{n}
@opindex --max-unchanged-stats
When tailing a file by name, if there have been @var{n} (default
n=@value{DEFAULT_MAX_N_UNCHANGED_STATS_BETWEEN_OPENS}) consecutive
iterations for which the file has not changed, then
@code{open}/@code{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.

@item -n [+]@var{num}
@itemx --lines=[+]@var{}
@opindex -n
@opindex --lines
Output the last @var{num} lines.
However, if @var{num} is prefixed with a @samp{+}, start printing with
line @var{num} from the start of each file, instead of from the end.
Size multiplier suffixes are the same as with the @option{-c} option.

@item --pid=@var{pid}
@opindex --pid
When following by name or by descriptor, you may specify the process ID,
@var{pid}, of the sole writer of all @var{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 @command{make} and @command{tail}
like this then the tail process will stop when your build completes.
Without this option, you would have had to kill the @code{tail -f}
process yourself.

@example
$ make >& makerr & tail --pid=$! -f makerr
@end example

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

@item -q
@itemx --quiet
@itemx --silent
@opindex -q
@opindex --quiet
@opindex --silent
Never print file name headers.

@item --retry
@opindex --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 @option{--follow=descriptor}),
this option only affects the initial open of the file, as after a successful
open, @command{tail} will start following the file descriptor.

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

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

@item -s @var{number}
@itemx --sleep-interval=@var{number}
@opindex -s
@opindex --sleep-interval
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.
When @command{tail} uses inotify, this polling-related option
is usually ignored.  However, if you also specify @option{--pid=@var{p}},
@command{tail} checks whether process @var{p} is alive at least
every @var{number} seconds.
The @var{number} must be non-negative and can be a floating-point number
in either the current or the C locale.  @xref{Floating point}.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Always print file name headers.

@optZeroTerminated

@end table

For compatibility @command{tail} also supports an obsolete usage
@samp{tail -[@var{num}][bcl][f] [@var{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, @var{num} is an optional decimal number optionally
followed by a size letter (@samp{b}, @samp{c}, @samp{l}) to mean count
by 512-byte blocks, bytes, or lines, optionally followed by @samp{f}
which has the same meaning as @option{-f}.

@vindex _POSIX2_VERSION
On systems not conforming to POSIX 1003.1-2001, the leading @samp{-}
can be replaced by @samp{+} in the traditional option syntax with the
same meaning as in counts, and on obsolete systems predating POSIX
1003.1-2001 traditional usage overrides normal usage when the two
conflict.  This behavior can be controlled with the
@env{_POSIX2_VERSION} environment variable (@pxref{Standards
conformance}).

Scripts intended for use on standard hosts should avoid traditional
syntax and should use @option{-c @var{num}[b]}, @option{-n
@var{num}}, and/or @option{-f} instead.  If your script must also
run on hosts that support only the traditional syntax, you can often
rewrite it to avoid problematic usages, e.g., by using @samp{sed -n
'$p'} rather than @samp{tail -1}.  If that's not possible, the script
can use a test like @samp{if tail -c +1 </dev/null >/dev/null 2>&1;
then @dots{}} 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 @samp{tail - main.c}, since it might be
interpreted as either @samp{tail main.c} or as @samp{tail -- -
main.c}; avoid @samp{tail -c 4}, since it might mean either @samp{tail
-c4} or @samp{tail -c 10 4}; and avoid @samp{tail +4}, since it might
mean either @samp{tail ./+4} or @samp{tail -n +4}.

@exitstatus


@node split invocation
@section @command{split}: Split a file into pieces.

@pindex split
@cindex splitting a file into pieces
@cindex pieces, splitting a file into

@command{split} creates output files containing consecutive or interleaved
sections of @var{input}  (standard input if none is given or @var{input}
is @samp{-}).  Synopsis:

@example
split [@var{option}] [@var{input} [@var{prefix}]]
@end example

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

@cindex output file name prefix
The output files' names consist of @var{prefix} (@samp{x} by default)
followed by a group of characters (@samp{aa}, @samp{ab}, @dots{} by
default), such that concatenating the output files in traditional
sorted order by file name produces the original input file (except
@option{-nr/@var{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.
(@samp{yz}, @samp{zaaa}, @samp{zaab}, @dots{}).  In this way an arbitrary
number of output files are supported, which sort as described above,
even in the presence of an @option{--additional-suffix} option.
If the @option{-a} option is specified and the output file names are
exhausted, @command{split} reports an error without deleting the
output files that it did create.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -l @var{lines}
@itemx --lines=@var{lines}
@opindex -l
@opindex --lines
Put @var{lines} lines of @var{input} into each output file.
If @option{--separator} is specified, then @var{lines} determines
the number of records.

For compatibility @command{split} also supports an obsolete
option syntax @option{-@var{lines}}.  New scripts should use
@option{-l @var{lines}} instead.

@item -b @var{size}
@itemx --bytes=@var{size}
@opindex -b
@opindex --bytes
Put @var{size} bytes of @var{input} into each output file.
@multiplierSuffixes{size}

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

@item --filter=@var{command}
@opindex --filter
With this option, rather than simply writing to each output file,
write through a pipe to the specified shell @var{command} for each output file.
@var{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 secondary storage,
yet you must split it into individually-compressed pieces
of a more manageable size.
To do that, you might run this command:

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

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

@item -n @var{chunks}
@itemx --number=@var{chunks}
@opindex -n
@opindex --number

Split @var{input} to @var{chunks} output files where @var{chunks} may be:

@example
@var{n}      generate @var{n} files based on current size of @var{input}
@var{k}/@var{n}    output only @var{k}th of @var{n} to standard output
l/@var{n}    generate @var{n} files without splitting lines or records
l/@var{k}/@var{n}  likewise but output only @var{k}th of @var{n} to stdout
r/@var{n}    like @samp{l} but use round robin distribution
r/@var{k}/@var{n}  likewise but output only @var{k}th of @var{n} to stdout
@end example

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

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

For @samp{l} mode, chunks are approximately @var{input} size / @var{n}.
The @var{input} is partitioned into @var{n} equal sized portions, with
the last assigned any excess.  If a line @emph{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 @samp{r} mode, the size of @var{input} is irrelevant,
and so can be a pipe for example.

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

@item -d
@itemx --numeric-suffixes[=@var{from}]
@opindex -d
@opindex --numeric-suffixes
Use digits in suffixes rather than lower-case letters.  The numerical
suffix counts from @var{from} if specified, 0 otherwise.

@var{from} is supported with the long form option, and 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 @option{-a} to allow
suffixes beyond @samp{99}.  Note if option @option{--number} is specified and
the number of files is less than @var{from}, a single run is assumed and the
minimum suffix length required is automatically determined.

@item -x
@itemx --hex-suffixes[=@var{from}]
@opindex -x
@opindex --hex-suffixes
Like @option{--numeric-suffixes}, but use hexadecimal numbers (in lower case).

@item --additional-suffix=@var{suffix}
@opindex --additional-suffix
Append an additional @var{suffix} to output file names.  @var{suffix}
must not contain slash.

@item -e
@itemx --elide-empty-files
@opindex -e
@opindex --elide-empty-files
Suppress the generation of zero-length output files.  This can happen
with the @option{--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.

@item -t @var{separator}
@itemx --separator=@var{separator}
@opindex -t
@opindex --separator
@cindex line separator character
@cindex record separator character
Use character @var{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 @samp{\0},
e.g., @samp{split -t '\0'}.

@item -u
@itemx --unbuffered
@opindex -u
@opindex --unbuffered
Immediately copy input to output in @option{--number r/@dots{}} mode,
which is a much slower mode of operation.

@item --verbose
@opindex --verbose
Write a diagnostic just before each output file is opened.

@end table

@exitstatus

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

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

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

08
0
==> xac <==
9
10
@end example

Use the "l/" modifier to suppress that:

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

==> xab <==
08
09

==> xac <==
10
@end example

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

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

==> xab <==
07
10

==> xac <==
08
@end example

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

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


@node csplit invocation
@section @command{csplit}: Split a file into context-determined pieces

@pindex csplit
@cindex context splitting
@cindex splitting a file into pieces by context

@command{csplit} creates zero or more output files containing sections of
@var{input} (standard input if @var{input} is @samp{-}).  Synopsis:

@example
csplit [@var{option}]@dots{} @var{input} @var{pattern}@dots{}
@end example

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

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

The types of pattern arguments are:

@table @samp

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

@item /@var{regexp}/[@var{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
@var{regexp}.  The optional @var{offset} is an integer, that can
be preceded by @samp{+} or @samp{-}.
If it is given, the input up to (but not including) the
matching line plus or minus @var{offset} is put into the output file,
and the line after that begins the next section of input.
Note lines within a negative offset of a regexp pattern
are not matched in subsequent regexp patterns.

@item %@var{regexp}%[@var{offset}]
Like the previous type, except that it does not create an output
file, so that section of the input file is effectively ignored.

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

@end table

The output files' names consist of a prefix (@samp{xx} by default)
followed by a suffix.  By default, the suffix is an ascending sequence
of two-digit decimal numbers from @samp{00} to @samp{99}.  In any case,
concatenating the output files in sorted order by file name produces the
original input file, excluding portions skipped with a %@var{regexp}%
pattern or the @option{--suppress-matched} option.

By default, if @command{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 @ref{Common options}.

@table @samp

@item -f @var{prefix}
@itemx --prefix=@var{prefix}
@opindex -f
@opindex --prefix
@cindex output file name prefix
Use @var{prefix} as the output file name prefix.

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

@item -n @var{digits}
@itemx --digits=@var{digits}
@opindex -n
@opindex --digits
Use output file names containing numbers that are @var{digits} digits
long instead of the default 2.

@item -k
@itemx --keep-files
@opindex -k
@opindex --keep-files
Do not remove output files when errors are encountered.

@item --suppress-matched
@opindex --suppress-matched
Do not output lines matching the specified @var{pattern}.
I.e., suppress the boundary line from the start of the second
and subsequent splits.

@item -z
@itemx --elide-empty-files
@opindex -z
@opindex --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.

@item -s
@itemx -q
@itemx --silent
@itemx --quiet
@opindex -s
@opindex -q
@opindex --silent
@opindex --quiet
Do not print counts of output file sizes.

@end table

@exitstatus

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

@example
$ mkdir d && cd d
@end example

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

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

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

@example
$ ls
xx00  xx01  xx02
@end example

Use @command{head} to show their contents:

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

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

==> xx02 <==
10
11
12
13
14
@end example

Example of splitting input by empty lines:

@example
$ csplit --suppress-matched @var{input.txt} '/^$/' '@{*@}'
@end example

@c
@c TODO: "uniq" already supports "--group".
@c        when it gets the "--key" option, uncomment this example.
@c
@c Example of splitting input file, based on the value of column 2:
@c
@c @example
@c $ cat @var{input.txt} |
@c       sort -k2,2 |
@c       uniq --group -k2,2 |
@c       csplit -m '/^$/' '@{*@}'
@c @end example

@node Summarizing files
@chapter Summarizing files

@cindex summarizing files

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

@menu
* wc invocation::               Print newline, word, and byte counts.
* sum invocation::              Print checksum and block counts.
* cksum invocation::            Print CRC checksum and byte counts.
* b2sum invocation::            Print or check BLAKE2 digests.
* md5sum invocation::           Print or check MD5 digests.
* sha1sum invocation::          Print or check SHA-1 digests.
* sha2 utilities::              Print or check SHA-2 digests.
@end menu


@node wc invocation
@section @command{wc}: Print newline, word, and byte counts

@pindex wc
@cindex byte count
@cindex character count
@cindex word count
@cindex line count

@command{wc} counts the number of bytes, characters, words, and newlines
in each given @var{file}, or standard input if none are given
or for a @var{file} of @samp{-}.  A word is a nonzero length
sequence of printable characters delimited by white space.  Synopsis:

@example
wc [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@cindex total counts
@command{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 @var{file} is given, @command{wc} prints a final line
containing the cumulative counts, with the file name @file{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, @command{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

@example
wc --bytes --words
@end example

@noindent
prints both the byte counts and the word counts.

With the @option{--max-line-length} option, @command{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 @ref{Common options}.

@table @samp

@item -c
@itemx --bytes
@opindex -c
@opindex --bytes
Print only the byte counts.

@item -m
@itemx --chars
@opindex -m
@opindex --chars
Print only the character counts, as per the current locale.
Invalid characters are not counted.

@item -w
@itemx --words
@opindex -w
@opindex --words
Print only the word counts.  A word is a nonzero length
sequence of printable characters separated by white space.

@item -l
@itemx --lines
@opindex -l
@opindex --lines
Print only the newline character counts.
Note a file without a trailing newline character,
will not have that last portion included in the line count.

@item -L
@itemx --max-line-length
@opindex -L
@opindex --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.

@macro filesZeroFromOption{cmd,withTotalOption,subListOutput}
@item --files0-from=@var{file}
@opindex --files0-from=@var{file}
@c This is commented out to avoid a texi2dvi failure.
@c texi2dvi (GNU Texinfo 4.11) 1.104
@c @cindex including files from @command{\cmd\}
Disallow processing files named on the command line, and instead process
those named in file @var{file}; each name being terminated by a zero byte
(ASCII NUL).
This is useful \withTotalOption\
when the list of file names is so long that it may exceed a command line
length limitation.
In such cases, running @command{\cmd\} via @command{xargs} is undesirable
because it splits the list into pieces and makes @command{\cmd\} print
\subListOutput\ for each sublist rather than for the entire list.
One way to produce a list of ASCII NUL terminated file
names is with GNU
@command{find}, using its @option{-print0} predicate.
If @var{file} is @samp{-} then the ASCII NUL terminated
file names are read from standard input.
@end macro
@filesZeroFromOption{wc,,a total}

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

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

@end table

@exitstatus


@node sum invocation
@section @command{sum}: Print checksum and block counts

@pindex sum
@cindex 16-bit checksum
@cindex checksum, 16-bit

@command{sum} computes a 16-bit checksum for each given @var{file}, or
standard input if none are given or for a @var{file} of @samp{-}.  Synopsis:

@example
sum [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@command{sum} prints the checksum for each @var{file} followed by the
number of blocks in the file (rounded up).  If at least one @var{file}
is given, file names are also printed.

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

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -r
@opindex -r
@cindex BSD @command{sum}
Use the default (BSD compatible) algorithm.  This option is included for
compatibility with the System V @command{sum}.  Unless @option{-s} was also
given, it has no effect.

@item -s
@itemx --sysv
@opindex -s
@opindex --sysv
@cindex System V @command{sum}
Compute checksums using an algorithm compatible with System V
@command{sum}'s default, and print file sizes in units of 512-byte blocks.

@end table

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

@exitstatus


@node cksum invocation
@section @command{cksum}: Print and verify file checksums

@pindex cksum
@cindex cyclic redundancy check
@cindex CRC checksum
@cindex digest

@command{cksum} by default computes a cyclic redundancy check (CRC) checksum
for each given @var{file}, or standard input if none are given or for a
@var{file} of @samp{-}.

cksum also supports the @option{-a,--algorithm} option to select the
digest algorithm to use. @command{cksum} is the preferred interface
to these digests, subsuming the other standalone checksumming utilities,
which can be emulated using @code{cksum -a md5 --untagged "$@@"} etc.
Synopsis:

@example
cksum [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@command{cksum} is typically used to ensure that files have not been corrupted,
by comparing the @command{cksum} output for the received files with the
@command{cksum} output for the original files (typically given in the
distribution).

@command{cksum} by default prints the POSIX standard CRC checksum
for each file along with the number of bytes in the file,
and the file name unless no arguments were given.

The same usage and options as the @command{b2sum}
command are supported.  @xref{b2sum invocation}.
In addition @command{cksum} supports the following options.

@table @samp

@item -a
@itemx --algorithm
@opindex -a
@opindex --algorithm
@cindex digest algorithm
Compute checksums using the specified digest algorithm.

Supported legacy checksums (which are not supported by @option{--check}):
@example
@samp{sysv}      equivalent to @command{sum -s}
@samp{bsd}       equivalent to @command{sum -r}
@samp{crc}       equivalent to @command{cksum} (the default)
@end example

Supported more modern digest algorithms are:
@example
@samp{md5}       equivalent to @command{md5sum}
@samp{sha1}      equivalent to @command{sha1sum}
@samp{sha224}    equivalent to @command{sha224sum}
@samp{sha256}    equivalent to @command{sha256sum}
@samp{sha384}    equivalent to @command{sha384sum}
@samp{sha512}    equivalent to @command{sha512sum}
@samp{blake2b}   equivalent to @command{b2sum}
@samp{sm3}       only available through @command{cksum}
@end example

@item --debug
@opindex --debug
Output extra information to stderr, like the checksum implementation being used.

@item --untagged
@opindex --untagged
Output using the original Coreutils format used by the other
standalone checksum utilities like @command{md5sum} for example.
This format has the checksum at the start of the line, and may be
more amenable to further processing by other utilities,
especially in combination with the @option{--zero} option.
Note this does not identify the digest algorithm used for the checksum.
@xref{md5sum invocation} for details of this format.
@end table


@node b2sum invocation
@section @command{b2sum}: Print or check BLAKE2 digests

@pindex b2sum
@cindex BLAKE2
@cindex 512-bit checksum
@cindex checksum, 512-bit
@cindex fingerprint, 512-bit
@cindex message-digest, 512-bit

@command{b2sum} computes a 512-bit checksum for each specified
@var{file}.  The same usage and options as the @command{md5sum}
command are supported.  @xref{md5sum invocation}.
In addition @command{b2sum} supports the following options.

@table @samp
@item -l
@itemx --length
@opindex -l
@opindex --length
@cindex BLAKE2 hash length
Change (shorten) the default digest length.
This is specified in bits and thus must be a multiple of 8.
This option is ignored when @option{--check} is specified,
as the length is automatically determined when checking.
@end table

@node md5sum invocation
@section @command{md5sum}: Print or check MD5 digests

@pindex md5sum
@cindex MD5
@cindex 128-bit checksum
@cindex checksum, 128-bit
@cindex fingerprint, 128-bit
@cindex message-digest, 128-bit

@command{md5sum} computes a 128-bit checksum (or @dfn{fingerprint} or
@dfn{message-digest}) for each specified @var{file}.

@macro weakHash{hash}
Note: The \hash\ digest is more reliable than a simple CRC (provided by
the @command{cksum} command) for detecting accidental file corruption,
as the chances of accidentally having two files with identical \hash\
are vanishingly small.  However, it should not be considered secure
against malicious tampering: although finding a file with a given \hash\
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 \hash\ digest.  For more secure hashes,
consider using SHA-2, or the newer @command{b2sum} command.
@xref{sha2 utilities}. @xref{b2sum invocation}.
@end macro
@weakHash{MD5}

If a @var{file} is specified as @samp{-} or if no files are given
@command{md5sum} computes the checksum for the standard input.
@command{md5sum} can also determine whether a file and checksum are
consistent.  Synopsis:

@example
md5sum [@var{option}]@dots{} [@var{file}]@dots{}
@end example

For each @var{file}, @samp{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 @samp{*}, text mode with @samp{ } (space).
Binary mode is the default on systems where it's significant,
otherwise text mode is the default.  The @command{cksum} command always
uses binary mode and a @samp{ } (space) flag.

Without @option{--zero}, if @var{file} contains a backslash, newline,
or carriage return, 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 @var{file} is omitted or specified as @samp{-}, standard input is read.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -b
@itemx --binary
@opindex -b
@opindex --binary
@cindex binary input files
Note this option is not supported by the @command{cksum} command,
as it operates in binary mode exclusively.
Treat each input file as binary, by reading it in binary mode and
outputting a @samp{*} flag.  This is the inverse of @option{--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.

@item -c
@itemx --check
Read file names and checksum information (not data) from each
@var{file} (or from standard input if no @var{file} was specified) and report
whether the checksums match the contents of the named files.
The input to this mode of @command{md5sum} is usually the output of
a prior, checksum-generating run of @samp{md5sum}.

Three input formats are supported.  Either the default output
format described above, the @option{--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 the @command{cksum} command, the @option{--check} option
supports auto-detecting the digest algorithm to use,
when presented with checksum information in the @option{--tag} output format.

Output with @option{--zero} enabled is not supported by @option{--check}.
@sp 1
For each such line, @command{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 @option{--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, @command{md5sum} exits with nonzero status.  Otherwise,
it exits successfully.
Note the @command{cksum} command doesn't support @option{--check}
with the older @samp{sysv}, @samp{bsd}, or @samp{crc} algorithms.

@item --ignore-missing
@opindex --ignore-missing
@cindex verifying MD5 checksums
This option is useful only when verifying checksums.
When verifying checksums, don't fail or report any status
for missing files.  This is useful when verifying a subset
of downloaded files given a larger list of checksums.

@item --quiet
@opindex --quiet
@cindex verifying MD5 checksums
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.

@item --status
@opindex --status
@cindex verifying MD5 checksums
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.

@item --tag
@opindex --tag
@cindex BSD output
Output BSD style checksums, which indicate the checksum algorithm used.
As a GNU extension, if @option{--zero} is not used, file names with problematic
characters are escaped as described above, with the same escaping indicator of
@samp{\} at the start of the line, being used.
The @option{--tag} option implies binary mode, and is disallowed with
@option{--text} mode as supporting that would unnecessarily complicate
the output format, while providing little benefit.
The @command{cksum} command, uses @option{--tag} as its default output format.

@item -t
@itemx --text
@opindex -t
@opindex --text
@cindex text input files
Note this option is not supported by the @command{cksum} command.
Treat each input file as text, by reading it in text mode and
outputting a @samp{ } flag.  This is the inverse of @option{--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 @option{--tag} is used.

@item -w
@itemx --warn
@opindex -w
@opindex --warn
@cindex verifying MD5 checksums
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.

@item --strict
@opindex --strict
@cindex verifying MD5 checksums
When verifying checksums,
if one or more input line is invalid,
exit nonzero after all warnings have been issued.

@optZero
Also file name escaping is not used.
@end table

@exitstatus


@node sha1sum invocation
@section @command{sha1sum}: Print or check SHA-1 digests

@pindex sha1sum
@cindex SHA-1
@cindex 160-bit checksum
@cindex checksum, 160-bit
@cindex fingerprint, 160-bit
@cindex message-digest, 160-bit

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

@weakHash{SHA-1}


@node sha2 utilities
@section sha2 utilities: Print or check SHA-2 digests

@pindex sha224sum
@pindex sha256sum
@pindex sha384sum
@pindex sha512sum
@cindex SHA-2
@cindex 224-bit checksum
@cindex 256-bit checksum
@cindex 384-bit checksum
@cindex 512-bit checksum
@cindex checksum, 224-bit
@cindex checksum, 256-bit
@cindex checksum, 384-bit
@cindex checksum, 512-bit
@cindex fingerprint, 224-bit
@cindex fingerprint, 256-bit
@cindex fingerprint, 384-bit
@cindex fingerprint, 512-bit
@cindex message-digest, 224-bit
@cindex message-digest, 256-bit
@cindex message-digest, 384-bit
@cindex message-digest, 512-bit

The commands @command{sha224sum}, @command{sha256sum},
@command{sha384sum} and @command{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 @command{md5sum}
and @command{sha1sum}.
@xref{md5sum invocation}.


@node Operating on sorted files
@chapter Operating on sorted files

@cindex operating on sorted files
@cindex sorted files, operations on

These commands work with (or produce) sorted files.

@menu
* sort invocation::             Sort text files.
* shuf invocation::             Shuffle text files.
* uniq invocation::             Uniquify files.
* comm invocation::             Compare two sorted files line by line.
* ptx invocation::              Produce a permuted index of file contents.
* tsort invocation::            Topological sort.
@end menu


@node sort invocation
@section @command{sort}: Sort text files

@pindex sort
@cindex sorting files

@command{sort} sorts, merges, or compares all the lines from the given
files, or standard input if none are given or for a @var{file} of
@samp{-}.  By default, @command{sort} writes the results to standard
output.  Synopsis:

@example
sort [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@cindex sort stability
@cindex sort's last-resort comparison
Many options affect how @command{sort} compares lines; if the results
are unexpected, try the @option{--debug} option to see what happened.
A pair of lines is compared as follows:
@command{sort} compares each pair of fields (see @option{--key}), 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, @command{sort} uses a default key of
the entire line.  Finally, as a last resort when all keys compare
equal, @command{sort} compares entire lines as if no ordering options
other than @option{--reverse} (@option{-r}) were specified.  The
@option{--stable} (@option{-s}) option disables this @dfn{last-resort
comparison} so that lines in which all fields compare equal are left
in their original relative order.  The @option{--unique}
(@option{-u}) option also disables the last-resort comparison.
@vindex LC_ALL
@vindex LC_COLLATE

Unless otherwise specified, all comparisons use the character collating
sequence specified by the @env{LC_COLLATE} locale.@footnote{If you
use a non-POSIX locale (e.g., by setting @env{LC_ALL}
to @samp{en_US}), then @command{sort} may produce output that is sorted
differently than you're accustomed to.  In that case, set the @env{LC_ALL}
environment variable to @samp{C}@.  Note that setting only @env{LC_COLLATE}
has two problems.  First, it is ineffective if @env{LC_ALL} is also set.
Second, it has undefined behavior if @env{LC_CTYPE} (or @env{LANG}, if
@env{LC_CTYPE} is unset) is set to an incompatible value.  For example,
you get undefined behavior if @env{LC_CTYPE} is @code{ja_JP.PCK} but
@env{LC_COLLATE} is @code{en_US.UTF-8}.}
A line's trailing newline is not part of the line for comparison
purposes.  If the final byte of an input file is not a newline, GNU
@command{sort} silently supplies one.  GNU @command{sort} (as
specified for all GNU utilities) has no limit on input line length or
restrictions on bytes allowed within lines.

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

@table @samp

@item -c
@itemx --check
@itemx --check=diagnose-first
@opindex -c
@opindex --check
@cindex checking for sortedness
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.

@item -C
@itemx --check=quiet
@itemx --check=silent
@opindex -c
@opindex --check
@cindex checking for sortedness
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 @option{-c}, except it does not print a diagnostic.

@item -m
@itemx --merge
@opindex -m
@opindex --merge
@cindex merging sorted files
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.

@end table

@cindex exit status of @command{sort}
Exit status:

@display
0 if no error occurred
1 if invoked with @option{-c} or @option{-C} and the input is not sorted
2 if an error occurred
@end display

@vindex TMPDIR
If the environment variable @env{TMPDIR} is set, @command{sort} uses its
value as the directory for temporary files instead of @file{/tmp}.  The
@option{--temporary-directory} (@option{-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 @command{sort}, global options affect only later key fields,
so portable shell scripts should specify global options first.

@table @samp

@item -b
@itemx --ignore-leading-blanks
@opindex -b
@opindex --ignore-leading-blanks
@cindex blanks, ignoring leading
@vindex LC_CTYPE
Ignore leading blanks when finding sort keys in each line.
By default a blank is a space or a tab, but the @env{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 @option{-k} option.

@item -d
@itemx --dictionary-order
@opindex -d
@opindex --dictionary-order
@cindex dictionary order
@cindex phone directory order
@cindex telephone directory order
@vindex LC_CTYPE
Sort in @dfn{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 @env{LC_CTYPE} locale can change this.

@item -f
@itemx --ignore-case
@opindex -f
@opindex --ignore-case
@cindex ignoring case
@cindex case folding
@vindex LC_CTYPE
Fold lowercase characters into the equivalent uppercase characters when
comparing so that, for example, @samp{b} and @samp{B} sort as equal.
The @env{LC_CTYPE} locale determines character types.
When used with @option{--unique} those lower case equivalent lines are
thrown away.  (There is currently no way to throw away the upper case
equivalent instead.  (Any @option{--reverse} given would only affect
the final result, after the throwing away.))

@item -g
@itemx --general-numeric-sort
@itemx --sort=general-numeric
@opindex -g
@opindex --general-numeric-sort
@opindex --sort
@cindex general numeric sort
@vindex LC_NUMERIC
Sort numerically, converting a prefix of each line to a long
double-precision floating point number.  @xref{Floating point}.
Do not report overflow, underflow, or conversion errors.
Use the following collating sequence:

@itemize @bullet
@item
Lines that do not start with numbers (all considered to be equal).
@item
NaNs (``Not a Number'' values, in IEEE floating point arithmetic)
in a consistent but machine-dependent order.
@item
Minus infinity.
@item
Finite numbers in ascending numeric order (with @math{-0} and @math{+0} equal).
@item
Plus infinity.
@end itemize

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

You can use this option to sort hexadecimal numbers prefixed with
@samp{0x} or @samp{0X}, where those numbers are not fixed width,
or of varying case.  However for hex numbers of consistent case,
and left padded with @samp{0} to a consistent width, a standard
lexicographic sort will be faster.

@item -h
@itemx --human-numeric-sort
@itemx --sort=human-numeric
@opindex -h
@opindex --human-numeric-sort
@opindex --sort
@cindex human numeric sort
@vindex LC_NUMERIC
Sort numerically, first by numeric sign (negative, zero, or positive);
then by SI suffix (either empty, or @samp{k} or @samp{K}, or
one of @samp{MGTPEZY}, in that order; @pxref{Block size}); and finally
by numeric value.  For example, @samp{1023M} sorts before @samp{1G}
because @samp{M} (mega) precedes @samp{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 @command{df}, @command{du}, or @command{ls} commands that are
invoked with their @option{--human-readable} or @option{--si} options.
The syntax for numbers is the same as for the @option{--numeric-sort}
option; the SI suffix must immediately follow the number.
Note also the @command{numfmt} command, which can be used to reformat
numbers to human format @emph{after} the sort, thus often allowing
sort to operate on more accurate numbers.

@item -i
@itemx --ignore-nonprinting
@opindex -i
@opindex --ignore-nonprinting
@cindex nonprinting characters, ignoring
@cindex unprintable characters, ignoring
@vindex LC_CTYPE
Ignore nonprinting characters.
The @env{LC_CTYPE} locale determines character types.
This option has no effect if the stronger @option{--dictionary-order}
(@option{-d}) option is also given.

@item -M
@itemx --month-sort
@itemx --sort=month
@opindex -M
@opindex --month-sort
@opindex --sort
@cindex months, sorting by
@vindex LC_TIME
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 @samp{JAN} < @samp{FEB} < @dots{} < @samp{DEC}@.
Invalid names compare low to valid names.  The @env{LC_TIME} locale
category determines the month spellings.
By default a blank is a space or a tab, but the @env{LC_CTYPE} locale
can change this.

@item -n
@itemx --numeric-sort
@itemx --sort=numeric
@opindex -n
@opindex --numeric-sort
@opindex --sort
@cindex numeric sort
@vindex LC_NUMERIC
Sort numerically.  The number begins each line and consists
of optional blanks, an optional @samp{-} 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 @samp{0}.  The @env{LC_NUMERIC}
locale specifies the decimal-point character and thousands separator.
By default a blank is a space or a tab, but the @env{LC_CTYPE} locale
can change this.

Comparison is exact; there is no rounding error.

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

@item -V
@itemx --version-sort
@opindex -V
@opindex --version-sort
@cindex version number 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.  (@xref{Version sort ordering}.)

@item -r
@itemx --reverse
@opindex -r
@opindex --reverse
@cindex reverse sorting
Reverse the result of comparison, so that lines with greater key values
appear earlier in the output instead of later.

@item -R
@itemx --random-sort
@itemx --sort=random
@opindex -R
@opindex --random-sort
@opindex --sort
@cindex random sort
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 (@pxref{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 @command{sort} more
than once.

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

@end table

Other options are:

@table @samp

@item --compress-program=@var{prog}
Compress any temporary files with the program @var{prog}.

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

Terminate with an error if @var{prog} exits with nonzero status.

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

@filesZeroFromOption{sort,,sorted output}

@item -k @var{pos1}[,@var{pos2}]
@itemx --key=@var{pos1}[,@var{pos2}]
@opindex -k
@opindex --key
@cindex sort field
Specify a sort field that consists of the part of the line between
@var{pos1} and @var{pos2} (or the end of the line, if @var{pos2} is
omitted), @emph{inclusive}.

In its simplest form @var{pos} specifies a field number (starting with 1),
with fields being separated by runs of blank characters, and by default
those blanks being included in the comparison at the start of each field.
To adjust the handling of blank characters see the @option{-b} and
@option{-t} options.

More generally,
each @var{pos} has the form @samp{@var{f}[.@var{c}][@var{opts}]},
where @var{f} is the number of the field to use, and @var{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
@var{pos2} indicates the field's last character.  If @samp{.@var{c}} is
omitted from @var{pos1}, it defaults to 1 (the beginning of the field);
if omitted from @var{pos2}, it defaults to 0 (the end of the field).
@var{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 @option{--key=2,2}
(@option{-k 2,2}).  See below for more notes on keys and more examples.
See also the @option{--debug} option to help determine the part
of the line being used in the sort.

@item --debug
Highlight the portion of each line used for sorting.
Also issue warnings about questionable usage to standard error.

@item --batch-size=@var{nmerge}
@opindex --batch-size
@cindex number of inputs to merge, nmerge
Merge at most @var{nmerge} inputs at once.

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

A large value of @var{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 @var{nmerge} may reduce memory
requirements and I/O at the expense of temporary storage consumption and
merge performance.

The value of @var{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 @var{nmerge} may be bounded by a resource limit for open
file descriptors.  The commands @samp{ulimit -n} or @samp{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 @var{nmerge} exceeds the resource limit, @command{sort}
silently uses a smaller value.

@item -o @var{output-file}
@itemx --output=@var{output-file}
@opindex -o
@opindex --output
@cindex overwriting of input, allowed
Write output to @var{output-file} instead of standard output.
Normally, @command{sort} reads all input before opening
@var{output-file}, so you can sort a file in place by using
commands like @code{sort -o F F} and @code{cat F | sort -o F}@.
However, it is often safer to output to an otherwise-unused file, as
data may be lost if the system crashes or @command{sort} encounters
an I/O or other serious error while a file is being sorted in place.
Also, @command{sort} with @option{--merge} (@option{-m}) can open
the output file before reading all input, so a command like @code{cat
F | sort -m -o F - G} is not safe as @command{sort} might start
writing @file{F} before @command{cat} is done reading it.

@vindex POSIXLY_CORRECT
On newer systems, @option{-o} cannot appear after an input file if
@env{POSIXLY_CORRECT} is set, e.g., @samp{sort F -o F}@.  Portable
scripts should specify @option{-o @var{output-file}} before any input
files.

@item --random-source=@var{file}
@opindex --random-source
@cindex random source for sorting
Use @var{file} as a source of random data used to determine which
random hash function to use with the @option{-R} option.  @xref{Random
sources}.

@item -s
@itemx --stable
@opindex -s
@opindex --stable
@cindex sort stability
@cindex sort's last-resort comparison

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

@item -S @var{size}
@itemx --buffer-size=@var{size}
@opindex -S
@opindex --buffer-size
@cindex size for main memory sorting
Use a main-memory sort buffer of the given @var{size}.  By default,
@var{size} is in units of 1024 bytes.  Appending @samp{%} causes
@var{size} to be interpreted as a percentage of physical memory.
Appending @samp{K} multiplies @var{size} by 1024 (the default),
@samp{M} by 1,048,576, @samp{G} by 1,073,741,824, and so on for
@samp{T}, @samp{P}, @samp{E}, @samp{Z}, and @samp{Y}@.  Appending
@samp{b} causes @var{size} to be interpreted as a byte count, with no
multiplication.

This option can improve the performance of @command{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 @var{size} if @command{sort} encounters input lines larger
than @var{size}.

@item -t @var{separator}
@itemx --field-separator=@var{separator}
@opindex -t
@opindex --field-separator
@cindex field separator character
Use character @var{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 @env{LC_CTYPE} locale
can change this.

That is, given the input line @w{@samp{ foo bar}}, @command{sort} breaks it
into fields @w{@samp{ foo}} and @w{@samp{ bar}}.  The field separator is
not considered to be part of either the field preceding or the field
following, so with @samp{sort @w{-t " "}} the same input line has
three fields: an empty field, @samp{foo}, and @samp{bar}.
However, fields that extend to the end of the line,
as @option{-k 2}, or fields consisting of a range, as @option{-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 @samp{\0}, e.g., @samp{sort -t '\0'}.

@item -T @var{tempdir}
@itemx --temporary-directory=@var{tempdir}
@opindex -T
@opindex --temporary-directory
@cindex temporary directory
@vindex TMPDIR
Use directory @var{tempdir} to store temporary files, overriding the
@env{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
file systems.

@item --parallel=@var{n}
@opindex --parallel
@cindex multithreaded sort
Set the number of sorts run in parallel to @var{n}. By default,
@var{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 @var{n} threads increases the memory usage by
a factor of log @var{n}.  Also see @ref{nproc invocation}.

@item -u
@itemx --unique
@opindex -u
@opindex --unique
@cindex uniquifying output

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

This option also disables the default last-resort comparison.

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

@optZeroTerminated
@macro newlineFieldSeparator
Note with @option{-z} the newline character is treated as a field separator.
@end macro

@end table

Historical (BSD and System V) implementations of @command{sort} have
differed in their interpretation of some options, particularly
@option{-b}, @option{-f}, and @option{-n}.
GNU sort follows the POSIX
behavior, which is usually (but not always!) like the System V behavior.
According to POSIX, @option{-n} no longer implies @option{-b}.  For
consistency, @option{-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 @option{-b}.

A position in a sort field specified with @option{-k} may have any
of the option letters @samp{MbdfghinRrV} appended to it, in which case no
global ordering options are inherited by that particular field.  The
@option{-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 @option{-t}
is not used, then @option{-k} is typically combined with @option{-b} or
an option that implicitly ignores leading blanks (@samp{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 @option{-b}
option was specified, the @samp{.@var{c}} part of a field specification
is counted from the first nonblank character of the field.

@vindex _POSIX2_VERSION
@vindex POSIXLY_CORRECT
On systems not conforming to POSIX 1003.1-2001,
@command{sort} supports a traditional origin-zero
syntax @samp{+@var{pos1} [-@var{pos2}]} for specifying sort keys.
The traditional command @samp{sort +@var{a}.@var{x} -@var{b}.@var{y}}
is equivalent to @samp{sort -k @var{a+1}.@var{x+1},@var{b}} if @var{y}
is @samp{0} or absent, otherwise it is equivalent to @samp{sort -k
@var{a+1}.@var{x+1},@var{b+1}.@var{y}}.

This traditional behavior can be controlled with the
@env{_POSIX2_VERSION} environment variable (@pxref{Standards
conformance}); it can also be enabled when @env{POSIXLY_CORRECT} is
not set by using the traditional syntax with @samp{-@var{pos2}} present.

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

Here are some examples to illustrate various combinations of options.

@itemize @bullet

@item
Sort in descending (reverse) numeric order.

@example
sort -n -r
@end example

@item
Run no more than 4 sorts concurrently, using a buffer size of 10M.

@example
sort --parallel=4 -S 10M
@end example

@item
Sort alphabetically, omitting the first and second fields
and the blanks at the start of the third field.
This uses a single key composed of the characters beginning
at the start of the first nonblank character in field three
and extending to the end of each line.

@example
sort -k 3b
@end example

@item
Sort numerically on the second field and resolve ties by sorting
alphabetically on the third and fourth characters of field five.
Use @samp{:} as the field delimiter.

@example
sort -t : -k 2,2n -k 5.3,5.4
@end example

Note that if you had written @option{-k 2n} instead of @option{-k 2,2n}
@command{sort} would have used all characters beginning in the second field
and extending to the end of the line as the primary @emph{numeric}
key.  For the large majority of applications, treating keys spanning
more than one field as numeric will not do what you expect.

Also note that the @samp{n} modifier was applied to the field-end
specifier for the first key.  It would have been equivalent to
specify @option{-k 2n,2} or @option{-k 2n,2n}.  All modifiers except
@samp{b} apply to the associated @emph{field}, regardless of whether
the modifier character is attached to the field-start and/or the
field-end part of the key specifier.

@item
Sort the password file on the fifth field and ignore any
leading blanks.  Sort lines with equal values in field five
on the numeric user ID in field three.  Fields are separated
by @samp{:}.

@example
sort -t : -k 5b,5 -k 3,3n /etc/passwd
sort -t : -n -k 5b,5 -k 3,3 /etc/passwd
sort -t : -b -k 5,5 -k 3,3n /etc/passwd
@end example

These three commands have equivalent effect.  The first specifies that
the first key's start position ignores leading blanks and the second
key is sorted numerically.  The other two commands rely on global
options being inherited by sort keys that lack modifiers.  The inheritance
works in this case because @option{-k 5b,5b} and @option{-k 5b,5} are
equivalent, as the location of a field-end lacking a @samp{.@var{c}}
character position is not affected by whether initial blanks are
skipped.

@item
Sort a set of log files, primarily by IPv4 address and secondarily by
timestamp.  If two lines' primary and secondary keys are identical,
output the lines in the same order that they were input.  The log
files contain lines that look like this:

@example
4.150.156.3 - - [01/Apr/2020:06:31:51 +0000] message 1
211.24.3.231 - - [24/Apr/2020:20:17:39 +0000] message 2
@end example

Fields are separated by exactly one space.  Sort IPv4 addresses
lexicographically, e.g., 212.61.52.2 sorts before 212.129.233.201
because 61 is less than 129.

@example
sort -s -t ' ' -k 4.9n -k 4.5M -k 4.2n -k 4.14,4.21 file*.log |
sort -s -t '.' -k 1,1n -k 2,2n -k 3,3n -k 4,4n
@end example

This example cannot be done with a single @command{sort} invocation,
since IPv4 address components are separated by @samp{.} while dates
come just after a space.  So it is broken down into two invocations of
@command{sort}: the first sorts by timestamp and the second by IPv4
address.  The timestamp is sorted by year, then month, then day, and
finally by hour-minute-second field, using @option{-k} to isolate each
field.  Except for hour-minute-second there's no need to specify the
end of each key field, since the @samp{n} and @samp{M} modifiers sort
based on leading prefixes that cannot cross field boundaries.  The
IPv4 addresses are sorted lexicographically.  The second sort uses
@samp{-s} so that ties in the primary key are broken by the secondary
key; the first sort uses @samp{-s} so that the combination of the two
sorts is stable.

@item
Generate a tags file in case-insensitive sorted order.

@example
find src -type f -print0 | sort -z -f | xargs -0 etags --append
@end example

The use of @option{-print0}, @option{-z}, and @option{-0} in this case means
that file names that contain blanks or other special characters are
not broken up
by the sort operation.

@c This example is a bit contrived and needs more explanation.
@c @item
@c Sort records separated by an arbitrary string by using a pipe to convert
@c each record delimiter string to @samp{\0}, then using sort's -z option,
@c and converting each @samp{\0} back to the original record delimiter.
@c
@c @example
@c printf 'c\n\nb\n\na\n' |
@c perl -0pe 's/\n\n/\n\0/g' |
@c sort -z |
@c perl -0pe 's/\0/\n/g'
@c @end example

@item
Use the common DSU, Decorate Sort Undecorate idiom to
sort lines according to their length.

@example
awk '@{print length, $0@}' /etc/passwd | sort -n | cut -f2- -d' '
@end example

In general this technique can be used to sort data that the @command{sort}
command does not support, or is inefficient at, sorting directly.

@item
Shuffle a list of directories, but preserve the order of files within
each directory.  For instance, one could use this to generate a music
playlist in which albums are shuffled but the songs of each album are
played in order.

@example
ls */* | sort -t / -k 1,1R -k 2,2
@end example

@end itemize


@node shuf invocation
@section @command{shuf}: Shuffling text

@pindex shuf
@cindex shuffling files

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

@example
shuf [@var{option}]@dots{} [@var{file}]
shuf -e [@var{option}]@dots{} [@var{arg}]@dots{}
shuf -i @var{lo}-@var{hi} [@var{option}]@dots{}
@end example

@command{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:

@table @samp

@item -e
@itemx --echo
@opindex -c
@opindex --echo
@cindex command-line operands to shuffle
Treat each command-line operand as an input line.

@item -i @var{lo}-@var{hi}
@itemx --input-range=@var{lo}-@var{hi}
@opindex -i
@opindex --input-range
@cindex input range to shuffle
Act as if input came from a file containing the range of unsigned
decimal integers @var{lo}@dots{}@var{hi}, one per line.

@end table

@command{shuf}'s other options can affect its behavior in all
operation modes:

@table @samp

@item -n @var{count}
@itemx --head-count=@var{count}
@opindex -n
@opindex --head-count
@cindex head of output
Output at most @var{count} lines.  By default, all input lines are
output.

@item -o @var{output-file}
@itemx --output=@var{output-file}
@opindex -o
@opindex --output
@cindex overwriting of input, allowed
Write output to @var{output-file} instead of standard output.
@command{shuf} reads all input before opening
@var{output-file}, so you can safely shuffle a file in place by using
commands like @code{shuf -o F <F} and @code{cat F | shuf -o F}.

@item --random-source=@var{file}
@opindex --random-source
@cindex random source for shuffling
Use @var{file} as a source of random data used to determine which
permutation to generate.  @xref{Random sources}.

@item -r
@itemx --repeat
@opindex -r
@opindex --repeat
@cindex repeat output values
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 @option{--head-count}; if
@option{--head-count} is not given, @command{shuf} repeats
indefinitely.

@optZeroTerminated

@end table

For example:

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

@noindent
might produce the output

@example
Panama!
A man,
a canal:
a plan,
@end example

@noindent
Similarly, the command:

@example
shuf -e clubs hearts diamonds spades
@end example

@noindent
might output:

@example
clubs
diamonds
spades
hearts
@end example

@noindent
and the command @samp{shuf -i 1-4} might output:

@example
4
2
1
3
@end example

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

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

@example
shuf -r -n 50 -i 0-9
@end example

@noindent
To simulate 100 coin flips, use:

@example
shuf -r -n 100 -e Head Tail
@end example

@exitstatus


@node uniq invocation
@section @command{uniq}: Uniquify files

@pindex uniq
@cindex uniquify files

@command{uniq} writes the unique lines in the given @file{input}, or
standard input if nothing is given or for an @var{input} name of
@samp{-}.  Synopsis:

@example
uniq [@var{option}]@dots{} [@var{input} [@var{output}]]
@end example

By default, @command{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 @code{sort -u}.
@xref{sort invocation}.

@vindex LC_COLLATE
Comparisons honor the rules specified by the @env{LC_COLLATE}
locale category.

If no @var{output} file is specified, @command{uniq} writes to standard
output.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -f @var{n}
@itemx --skip-fields=@var{n}
@opindex -f
@opindex --skip-fields
Skip @var{n} fields on each line before checking for uniqueness.  Use
a null string for comparison if a line has fewer than @var{n} fields.
Fields are a sequence of blank characters followed by non-blank characters.
Field numbers are one based, i.e., @option{-f 1} will skip the first
field (which may optionally have leading blanks).

For compatibility @command{uniq} supports a traditional option syntax
@option{-@var{n}}.  New scripts should use @option{-f @var{n}} instead.

@item -s @var{n}
@itemx --skip-chars=@var{n}
@opindex -s
@opindex --skip-chars
Skip @var{n} characters before checking for uniqueness.  Use a null string
for comparison if a line has fewer than @var{n} characters.  If you use both
the field and character skipping options, fields are skipped over first.

@vindex _POSIX2_VERSION
On systems not conforming to POSIX 1003.1-2001,
@command{uniq} supports a traditional option syntax
@option{+@var{n}}.
Although this traditional behavior can be controlled with the
@env{_POSIX2_VERSION} environment variable (@pxref{Standards
conformance}), portable scripts should avoid commands whose
behavior depends on this variable.
For example, use @samp{uniq ./+10} or @samp{uniq -s 10} rather than
the ambiguous @samp{uniq +10}.

@item -c
@itemx --count
@opindex -c
@opindex --count
Print the number of times each line occurred along with the line.

@item -i
@itemx --ignore-case
@opindex -i
@opindex --ignore-case
Ignore differences in case when comparing lines.

@item -d
@itemx --repeated
@opindex -d
@opindex --repeated
@cindex repeated lines, outputting
Discard lines that are not repeated.  When used by itself, this option
causes @command{uniq} to print the first copy of each repeated line,
and nothing else.

@item -D
@itemx --all-repeated[=@var{delimit-method}]
@opindex -D
@opindex --all-repeated
@cindex all repeated lines, outputting
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 @var{delimit-method}, supported with the long form option,
specifies how to delimit groups of repeated lines, and must be one of the
following:

@table @samp

@item none
Do not delimit groups of repeated lines.
This is equivalent to @option{--all-repeated} (@option{-D}).

@item prepend
Output a newline before each group of repeated lines.
@macro nulOutputNote
With @option{--zero-terminated} (@option{-z}), use a zero
byte (ASCII NUL) instead of a newline as the delimiter.
@end macro
@nulOutputNote

@item separate
Separate groups of repeated lines with a single newline.
This is the same as using @samp{prepend}, except that
no delimiter is inserted before the first group, and hence
may be better suited for output direct to users.
@nulOutputNote
@end table

@macro ambiguousGroupNote
Note that when groups are delimited and the input stream contains
blank lines, then the output is ambiguous.
To avoid that, filter the input through @samp{tr -s '\\n'} to
remove blank lines.
@end macro
@ambiguousGroupNote

This is a GNU extension.
@c FIXME: give an example showing *how* it's useful

@item --group[=@var{delimit-method}]
@opindex --group
@cindex all lines, grouping
Output all lines, and delimit each unique group.
@nulOutputNote
The optional @var{delimit-method} specifies how to delimit
groups, and must be one of the following:

@table @samp

@item 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.

@item prepend
Output a delimiter before each group of unique items.

@item append
Output a delimiter after each group of unique items.

@item both
Output a delimiter around each group of unique items.
@end table

@ambiguousGroupNote

This is a GNU extension.

@item -u
@itemx --unique
@opindex -u
@opindex --unique
@cindex unique lines, outputting
Discard the last line that would be output for a repeated input group.
When used by itself, this option causes @command{uniq} to print unique
lines, and nothing else.

@item -w @var{n}
@itemx --check-chars=@var{n}
@opindex -w
@opindex --check-chars
Compare at most @var{n} characters on each line (after skipping any specified
fields and characters).  By default the entire rest of the lines are
compared.

@optZeroTerminated
@newlineFieldSeparator

@end table

@exitstatus


@node comm invocation
@section @command{comm}: Compare two sorted files line by line

@pindex comm
@cindex line-by-line comparison
@cindex comparing sorted files

@command{comm} writes to standard output lines that are common, and lines
that are unique, to two input files; a file name of @samp{-} means
standard input.  Synopsis:

@example
comm [@var{option}]@dots{} @var{file1} @var{file2}
@end example

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

@cindex differing lines
@cindex common lines
With no options, @command{comm} produces three-column output.  Column one
contains lines unique to @var{file1}, column two contains lines unique
to @var{file2}, and column three contains lines common to both files.
Columns are separated by a single TAB character.
@c FIXME: when there's an option to supply an alternative separator
@c string, append "by default" to the above sentence.

@opindex -1
@opindex -2
@opindex -3
The options @option{-1}, @option{-2}, and @option{-3} suppress printing of
the corresponding columns (and separators).  Also see @ref{Common options}.

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

@macro checkOrderOption{cmd}
If the @option{--check-order} option is given, unsorted inputs will
cause a fatal error message.  If the option @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
@ifset JOIN_COMMAND
lines, and when both input files are non empty.
@end ifset
@ifclear JOIN_COMMAND
lines.
@end ifclear
If an input file is diagnosed as being unsorted, the @command{\cmd\}
command will exit with a nonzero status (and the output should not be used).

Forcing @command{\cmd\} to process wrongly sorted input files
containing unpairable lines by specifying @option{--nocheck-order} is
not guaranteed to produce any particular output.  The output will
probably not correspond with whatever you hoped it would be.
@end macro
@checkOrderOption{comm}

@table @samp

@item --check-order
Fail with an error message if either input file is wrongly ordered.

@item --nocheck-order
Do not check that both input files are in sorted order.

Other options are:

@item --output-delimiter=@var{str}
Print @var{str} between adjacent output columns,
rather than the default of a single TAB character.

The delimiter @var{str} may not be empty.

@item --total
Output a summary at the end.

Similar to the regular output,
column one contains the total number of lines unique to @var{file1},
column two contains the total number of lines unique to @var{file2}, and
column three contains the total number of lines common to both files,
followed by the word @samp{total} in the additional column four.

In the following example, @command{comm} omits the regular output
(@option{-123}), thus just printing the summary:

@example
$ printf '%s\n' a b c d e     > file1
$ printf '%s\n'   b c d e f g > file2
$ comm --total -123 file1 file2
1       2       4       total
@end example

This option is a GNU extension.  Portable scripts should use @command{wc} to
get the totals, e.g. for the above example files:

@example
$ comm -23 file1 file2 | wc -l    # number of lines only in file1
1
$ comm -13 file1 file2 | wc -l    # number of lines only in file2
2
$ comm -12 file1 file2 | wc -l    # number of lines common to both files
4
@end example

@optZeroTerminated

@end table

@node ptx invocation
@section @command{ptx}: Produce permuted indexes

@pindex ptx

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

@example
ptx [@var{option} @dots{}] [@var{file} @dots{}]
ptx -G [@var{option} @dots{}] [@var{input} [@var{output}]]
@end example

The @option{-G} (or its equivalent: @option{--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 @option{-G} is not specified, GNU extensions are always enabled.
GNU extensions to @command{ptx} are documented wherever appropriate in this
document.  @xref{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
@var{file}s after the options.  If there is no @var{file}, the program
reads the standard input.  If there is one or several @var{file}s, 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 @emph{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 @var{input} to be read
instead of the standard input.  If two parameters are given, they give
respectively the name of the @var{input} file to read and the name of
the @var{output} file to produce.  @emph{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 @command{ptx}
compatibility; GNU Standards normally discourage output parameters not
introduced by an option.

Note that for @emph{any} file named as the value of an option or as an
input text file, a single dash @samp{-} 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.

@menu
* General options in ptx::      Options which affect general program behavior.
* Charset selection in ptx::    Underlying character set considerations.
* Input processing in ptx::     Input fields, contexts, and keyword selection.
* Output formatting in ptx::    Types of output format, and sizing the fields.
* Compatibility in ptx::
@end menu


@node General options in ptx
@subsection General options

@table @samp

@item -G
@itemx --traditional
As already explained, this option disables all GNU extensions to
@command{ptx} and switches to traditional mode.

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

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

@end table

@exitstatus


@node Charset selection in ptx
@subsection Charset selection

As it is set up now, @command{ptx} assumes that the input file is coded
using 8-bit characters, and it may not work well in multibyte locales.
In a single-byte locale, 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.

The output of @command{ptx} assumes the locale's character encoding.
For example, with @command{ptx}'s @option{-T} option, if the locale
uses the Latin-1 encoding you may need a LaTeX directive like
@samp{\usepackage[latin1]@{inputenc@}} to render non-ASCII characters
correctly.

@table @samp

@item -f
@itemx --ignore-case
@opindex -f
@opindex --ignore-case
Fold lower case letters to upper case for sorting.

@end table


@node Input processing in ptx
@subsection Word selection and input processing

@table @samp

@item -b @var{file}
@itemx --break-file=@var{file}
@opindex -b
@opindex --break-file

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

@item -i @var{file}
@itemx --ignore-file=@var{file}
@opindex -i
@opindex --ignore-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
@dfn{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
@option{-S} option.

@item -o @var{file}
@itemx --only-file=@var{file}
@opindex -o
@opindex --only-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 @dfn{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 @option{-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.

@item -r
@itemx --references
@opindex -r
@opindex --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.
@xref{Output formatting in ptx},
for more information about reference production.
Using this option changes the default value for option @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
@emph{when} the context ends exactly at the newline.  If option
@option{-r} is used with @option{-S} default value, or when GNU extensions
are disabled, this condition is always met and references are completely
excluded from the output contexts.

@item -S @var{regexp}
@itemx --sentence-regexp=@var{regexp}
@opindex -S
@opindex --sentence-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 @option{-r} option is not
used, end of sentences are used.  In this case, this @var{regex} is
imported from GNU Emacs:

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

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

@example
\n
@end example

Using an empty @var{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 @option{-F
""}.  @xref{Regexps, , Syntax of Regular Expressions, emacs, 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 @command{ptx} itself.

@item -W @var{regexp}
@itemx --word-regexp=@var{regexp}
@opindex -W
@opindex --word-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 @var{regexp} used is @samp{\w+}.  When GNU extensions are
disabled, a word is by default anything which ends with a space, a tab
or a newline; the @var{regexp} used is @samp{[^ \t\n]+}.

An empty @var{regexp} is equivalent to not using this option.
@xref{Regexps, , Syntax of Regular Expressions, emacs, 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 @command{ptx} itself.

@end table


@node Output formatting in ptx
@subsection Output formatting

Output format is mainly controlled by the @option{-O} and @option{-T} options
described in the table below.  When neither @option{-O} nor @option{-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 @option{-A} and are output before the
left context, that is, if option @option{-R} is @emph{not} selected, then
a colon is added after the reference; this nicely interfaces with GNU
Emacs @code{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.

@table @samp

@item -g @var{number}
@itemx --gap-size=@var{number}
@opindex -g
@opindex --gap-size

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

@item -w @var{number}
@itemx --width=@var{number}
@opindex -w
@opindex --width

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 @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.

@item -A
@itemx --auto-reference
@opindex -A
@opindex --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 @option{-A} and @option{-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.

@item -R
@itemx --right-side-refs
@opindex -R
@opindex --right-side-refs

In the default output format, when option @option{-R} is not used, any
references produced by the effect of options @option{-r} or @option{-A} are
placed to the far right of output lines, after the right context.  With
default output format, when the @option{-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 @option{-R} is
ignored, with one exception:  with @option{-R} the width of references
is @emph{not} taken into account in total output width given by @option{-w}.

This option is automatically selected whenever GNU extensions are
disabled.

@item -F @var{string}
@itemx --flag-truncation=@var{string}
@opindex -F
@opindex --flag-truncation

This option will request that any truncation in the output be reported
using the string @var{string}.  Most output fields theoretically extend
towards the beginning or the end of the current line, or current
sentence, as selected with option @option{-S}@.  But there is a maximum
allowed output line width, changeable through option @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 @option{-F /}.

@var{string} may have more than one character, as in @option{-F @dots{}}.
Also, in the particular case when @var{string} is empty (@option{-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 @command{ptx} itself.

@item -M @var{string}
@itemx --macro-name=@var{string}
@opindex -M
@opindex --macro-name

Select another @var{string} to be used instead of @samp{xx}, while
generating output suitable for @command{nroff}, @command{troff} or @TeX{}.

@item -O
@itemx --format=roff
@opindex -O
@opindex --format=roff

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

@example
.xx "@var{tail}" "@var{before}" "@var{keyword_and_after}"@c
 "@var{head}" "@var{ref}"
@end example

so it will be possible to write a @samp{.xx} roff macro to take care of
the output typesetting.  This is the default output format when GNU
extensions are disabled.  Option @option{-M} can be used to change
@samp{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 @samp{"} is doubled
so it will be correctly processed by @command{nroff} or @command{troff}.

@item -T
@itemx --format=tex
@opindex -T
@opindex --format=tex

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

@example
\xx @{@var{tail}@}@{@var{before}@}@{@var{keyword}@}@c
@{@var{after}@}@{@var{head}@}@{@var{ref}@}
@end example

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

In this output format, some special characters, like @samp{$}, @samp{%},
@samp{&}, @samp{#} and @samp{_} are automatically protected with a
backslash.  Curly brackets @samp{@{}, @samp{@}} are protected with a
backslash and a pair of dollar signs (to force mathematical mode).  The
backslash itself produces the sequence @code{\backslash@{@}}.
Circumflex and tilde diacritical marks produce the sequence @code{^\@{ @}} and
@code{~\@{ @}} 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{}.

@end table


@node Compatibility in ptx
@subsection The GNU extensions to @command{ptx}

This version of @command{ptx} contains a few features which do not exist in
System V @command{ptx}.  These extra features are suppressed by using the
@option{-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 @option{-G} if you care about GNU extensions.
Here are the differences between this program and System V @command{ptx}.

@itemize @bullet

@item
This program can read many input files at once, it always writes the
resulting concordance on standard output.  On the other hand, System V
@command{ptx} reads only one file and sends the result to standard output
or, if a second @var{file} parameter is given on the command, to that
@var{file}.

Having output parameters not introduced by options is a dangerous
practice which GNU avoids as far as possible.  So, for using @command{ptx}
portably between GNU and System V, you should always use it with a
single input file, and always expect the result on standard output.  You
might also want to automatically configure in a @option{-G} option to
@command{ptx} calls in products using @command{ptx}, if the configurator finds
that the installed @command{ptx} accepts @option{-G}.

@item
The only options available in System V @command{ptx} are options @option{-b},
@option{-f}, @option{-g}, @option{-i}, @option{-o}, @option{-r}, @option{-t} and
@option{-w}.  All other options are GNU extensions and are not repeated in
this enumeration.  Moreover, some options have a slightly different
meaning when GNU extensions are enabled, as explained below.

@item
By default, concordance output is not formatted for @command{troff} or
@command{nroff}.  It is rather formatted for a dumb terminal.  @command{troff}
or @command{nroff} output may still be selected through option @option{-O}.

@item
Unless @option{-R} option is used, the maximum reference width is
subtracted from the total output line width.  With GNU extensions
disabled, width of references is not taken into account in the output
line width computations.

@item
All 256 bytes, even ASCII NUL bytes, are always read and
processed from input file with no adverse effect, even if GNU extensions
are disabled.  However, System V @command{ptx} does not accept 8-bit
characters, a few control characters are rejected, and the tilde
@samp{~} is also rejected.

@item
Input line length is only limited by available memory, even if GNU
extensions are disabled.  However, System V @command{ptx} processes only
the first 200 characters in each line.

@item
The break (non-word) characters default to be every character except all
letters of the underlying character set, diacriticized or not.  When GNU
extensions are disabled, the break characters default to space, tab and
newline only.

@item
The program makes better use of output line width.  If GNU extensions
are disabled, the program rather tries to imitate System V @command{ptx},
but still, there are some slight disposition glitches this program does
not completely reproduce.

@item
The user can specify both an Ignore file and an Only file.  This is not
allowed with System V @command{ptx}.

@end itemize


@node tsort invocation
@section @command{tsort}: Topological sort

@pindex tsort
@cindex topological sort

@command{tsort} performs a topological sort on the given @var{file}, or
standard input if no input file is given or for a @var{file} of
@samp{-}.  For more details and some history, see @ref{tsort background}.
Synopsis:

@example
tsort [@var{option}] [@var{file}]
@end example

@command{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

@example
tsort <<EOF
a b c
d
e f
b c d e
EOF
@end example

@noindent
will produce the output

@example
a
b
c
d
e
f
@end example

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 @code{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.

@example
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
@end example

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

@example
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
@end example

@command{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
@code{parse_options} may be placed anywhere in the list as long as it
precedes @code{main}.

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.

@exitstatus

@menu
* tsort background::            Where tsort came from.
@end menu

@node tsort background
@subsection @command{tsort}: Background

@command{tsort} exists because very early versions of the Unix linker processed
an archive file exactly once, and in order.  As @command{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, @code{scanf} probably calls @code{read}.  That means
that in a single pass through an archive, it was important for @code{scanf.o}
to appear before read.o, because otherwise a program which calls
@code{scanf} but not @code{read} might end up with an unexpected unresolved
reference to @code{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 @command{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 @command{tsort} over the @command{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
@command{ranlib}, now generally built by @command{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.


@node Operating on fields
@chapter Operating on fields

@menu
* cut invocation::              Print selected parts of lines.
* paste invocation::            Merge lines of files.
* join invocation::             Join lines on a common field.
@end menu


@node cut invocation
@section @command{cut}: Print selected parts of lines

@pindex cut
@command{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
@samp{-}.  Synopsis:

@example
cut @var{option}@dots{} [@var{file}]@dots{}
@end example

In the table which follows, the @var{byte-list}, @var{character-list},
and @var{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: @option{-@var{m}} means @samp{1-@var{m}}; @samp{@var{n}-} means
@samp{@var{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 @ref{Common
options}.

@table @samp

@item -b @var{byte-list}
@itemx --bytes=@var{byte-list}
@opindex -b
@opindex --bytes
Select for printing only the bytes in positions listed in
@var{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 @option{--output-delimiter}), then output that
string between ranges of selected bytes.

@item -c @var{character-list}
@itemx --characters=@var{character-list}
@opindex -c
@opindex --characters
Select for printing only the characters in positions listed in
@var{character-list}.  The same as @option{-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
@option{--output-delimiter}), then output that string between ranges
of selected bytes.

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

Note @command{awk} supports more sophisticated field processing,
like reordering fields, and handling fields aligned with blank characters.
By default @command{awk} uses (and discards) runs of blank characters
to separate fields, and ignores leading and trailing blanks.
@example
@verbatim
awk '{print $2}'      # print the second field
awk '{print $(NF-1)}' # print the penultimate field
awk '{print $2,$1}'   # reorder the first two fields
@end verbatim
@end example
Note while @command{cut} accepts field specifications in
arbitrary order, output is always in the order encountered in the file.

In the unlikely event that @command{awk} is unavailable,
one can use the @command{join} command, to process blank
characters as @command{awk} does above.
@example
@verbatim
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
@end verbatim
@end example

@item -d @var{input_delim_byte}
@itemx --delimiter=@var{input_delim_byte}
@opindex -d
@opindex --delimiter
With @option{-f}, use the first byte of @var{input_delim_byte} as
the input fields separator (default is TAB).

@item -n
@opindex -n
Do not split multi-byte characters (no-op for now).

@item -s
@itemx --only-delimited
@opindex -s
@opindex --only-delimited
For @option{-f}, do not print lines that do not contain the field separator
character.  Normally, any line without a field separator is printed verbatim.

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

@item --complement
@opindex --complement
This option is a GNU extension.
Select for printing the complement of the bytes, characters or fields
selected with the @option{-b}, @option{-c} or @option{-f} options.
In other words, do @emph{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.

@optZeroTerminated

@end table

@exitstatus


@node paste invocation
@section @command{paste}: Merge lines of files

@pindex paste
@cindex merging files

@command{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 @samp{-} or if no input files
are given.

Synopsis:

@example
paste [@var{option}]@dots{} [@var{file}]@dots{}
@end example

For example, with:
@example
$ cat num2
1
2
$ cat let3
a
b
c
@end example

Take lines sequentially from each file:
@example
$ paste num2 let3
1       a
2       b
       @ c
@end example

Duplicate lines from a file:
@example
$ paste num2 let3 num2
1       a      1
2       b      2
       @ c
@end example

Intermix lines from standard input:
@example
$ paste - let3 - < num2
1       a      2
       @ b
       @ c
@end example

Join consecutive lines with a space:
@example
$ seq 4 | paste -d ' ' - -
1 2
3 4
@end example

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -s
@itemx --serial
@opindex -s
@opindex --serial
Paste the lines of one file at a time rather than one line from each
file.  Using the above example data:

@example
$ paste -s num2 let3
1       2
a       b       c
@end example

@item -d @var{delim-list}
@itemx --delimiters=@var{delim-list}
@opindex -d
@opindex --delimiters
Consecutively use the characters in @var{delim-list} instead of
TAB to separate merged lines.  When @var{delim-list} is
exhausted, start again at its beginning.  Using the above example data:

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

@optZeroTerminated

@end table

@exitstatus


@node join invocation
@section @command{join}: Join lines on a common field

@pindex join
@cindex common field, joining on

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

@example
join [@var{option}]@dots{} @var{file1} @var{file2}
@end example

Either @var{file1} or @var{file2} (but not both) can be @samp{-},
meaning standard input.  @var{file1} and @var{file2} should be
sorted on the join fields.

@example
@group
$ cat file1
a 1
b 2
e 5

$ cat file2
a X
e Y
f Z

$ join file1 file2
a 1 X
e 5 Y
@end group
@end example


@noindent
@command{join}'s default behavior (when no options are given):
@itemize
@item the join field is the first field in each line;
@item fields in the input are separated by one or more blanks, with leading
blanks on the line ignored;
@item fields in the output are separated by a space;
@item each output line consists of the join field, the remaining
fields from @var{file1}, then the remaining fields from @var{file2}.
@end itemize


@menu
* General options in join::      Options which affect general program behavior.
* Sorting files for join::       Using @command{sort} before @command{join}.
* Working with fields::          Joining on different fields.
* Paired and unpaired lines::    Controlling @command{join}'s field matching.
* Header lines::                 Working with header lines in files.
* Set operations::               Union, Intersection and Difference of files.
@end menu

@node General options in join
@subsection General options
The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -a @var{file-number}
@opindex -a
Print a line for each unpairable line in file @var{file-number} (either
@samp{1} or @samp{2}), in addition to the normal output.

@item --check-order
Fail with an error message if either input file is wrongly ordered.

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

@item -e @var{string}
@opindex -e
Replace those output fields that are missing in the input with @var{string}.
I.e., missing fields specified with the @option{-12jo} options.

@item --header
@opindex --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 @option{-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
@option{--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.

@item -i
@itemx --ignore-case
@opindex -i
@opindex --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 @samp{sort -f} to produce this ordering.

@item -1 @var{field}
@opindex -1
Join on field @var{field} (a positive integer) of file 1.

@item -2 @var{field}
@opindex -2
Join on field @var{field} (a positive integer) of file 2.

@item -j @var{field}
Equivalent to @option{-1 @var{field} -2 @var{field}}.

@item -o @var{field-list}
@itemx -o auto
If the keyword @samp{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 @option{-e} option and extra fields
are discarded.

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

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

The elements in @var{field-list}
are separated by commas or blanks.
Blank separators typically need to be quoted for the shell.  For
example, the commands @samp{join -o 1.2,2.2} and @samp{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 @var{field-list}.

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

@item -v @var{file-number}
Print a line for each unpairable line in file @var{file-number}
(either @samp{1} or @samp{2}), instead of the normal output.

@optZeroTerminated
@newlineFieldSeparator

@end table

@exitstatus

@set JOIN_COMMAND
@checkOrderOption{join}
@clear JOIN_COMMAND



@node Sorting files for join
@subsection Pre-sorting

@command{join} requires sorted input files. Each input file should be
sorted according to the key (=field/column number) used in
@command{join}. The recommended sorting option is @samp{sort -k 1b,1}
(assuming the desired key is in the first column).

@noindent Typical usage:
@example
@group
$ sort -k 1b,1 file1 > file1.sorted
$ sort -k 1b,1 file2 > file2.sorted
$ join file1.sorted file2.sorted > file3
@end group
@end example

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

@example
@group
$ sort -k 1bf,1 file1 > file1.sorted
$ sort -k 1bf,1 file2 > file2.sorted
$ join --ignore-case file1.sorted file2.sorted > file3
@end group
@end example

The @command{sort} and @command{join} commands should use consistent
locales and options if the output of @command{sort} is fed to
@command{join}.  You can use a command like @samp{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 @command{join} and @command{sort}.

@noindent To avoid any locale-related issues, it is recommended to use the
@samp{C} locale for both commands:

@example
@group
$ LC_ALL=C sort -k 1b,1 file1 > file1.sorted
$ LC_ALL=C sort -k 1b,1 file2 > file2.sorted
$ LC_ALL=C join file1.sorted file2.sorted > file3
@end group
@end example


@node Working with fields
@subsection Working with fields

Use @option{-1},@option{-2} to set the key fields for each of the input files.
Ensure the preceding @command{sort} commands operated on the same fields.

@noindent
The following example joins two files, using the values from seventh field
of the first file and the third field of the second file:

@example
@group
$ sort -k 7b,7 file1 > file1.sorted
$ sort -k 3b,3 file2 > file2.sorted
$ join -1 7 -2 3 file1.sorted file2.sorted > file3
@end group
@end example

@noindent
If the field number is the same for both files, use @option{-j}:

@example
@group
$ sort -k4b,4 file1 > file1.sorted
$ sort -k4b,4 file2 > file2.sorted
$ join -j4    file1.sorted file2.sorted > file3
@end group
@end example

@noindent
Both @command{sort} and @command{join} operate of whitespace-delimited
fields. To specify a different delimiter, use @option{-t} in @emph{both}:

@example
@group
$ sort -t, -k3b,3 file1 > file1.sorted
$ sort -t, -k3b,3 file2 > file2.sorted
$ join -t, -j3    file1.sorted file2.sorted > file3
@end group
@end example

@noindent
To specify a tab (@sc{ascii} 0x09) character instead of whitespace,
use:@footnote{the @code{$'\t'} is supported in most modern shells.
For older shells, use a literal tab.}

@example
@group
$ sort -t$'\t' -k3b,3 file1 > file1.sorted
$ sort -t$'\t' -k3b,3 file2 > file2.sorted
$ join -t$'\t' -j3    file1.sorted file2.sorted > file3
@end group
@end example


@noindent
If @samp{join -t ''} is specified then the whole line is considered which
matches the default operation of sort:

@example
@group
$ sort file1 > file1.sorted
$ sort file2 > file2.sorted
$ join -t '' file1.sorted file2.sorted > file3
@end group
@end example


@node Paired and unpaired lines
@subsection Controlling @command{join}'s field matching

In this section the @command{sort} commands are omitted for brevity.
Sorting the files before joining is still required.

@command{join}'s default behavior is to print only lines common to
both input files. Use @option{-a} and @option{-v} to print unpairable lines
from one or both files.

@noindent
All examples below use the following two (pre-sorted) input files:

@multitable @columnfractions .5 .5
@item
@example
$ cat file1
a 1
b 2
@end example

@tab
@example
$ cat file2
a A
c C
@end example
@end multitable


@c TODO: Find better column widths that work for both HTML and PDF
@c       and disable indentation of @example.
@multitable @columnfractions 0.5 0.5

@headitem Command @tab Outcome


@item
@example
$ join file1 file2
a 1 A
@end example
@tab
common lines
(@emph{intersection})



@item
@example
$ join -a 1 file1 file2
a 1 A
b 2
@end example
@tab
common lines @emph{and} unpaired
lines from the first file


@item
@example
$ join -a 2 file1 file2
a 1 A
c C
@end example
@tab
common lines @emph{and} unpaired lines from the second file


@item
@example
$ join -a 1 -a 2 file1 file2
a 1 A
b 2
c C
@end example
@tab
all lines (paired and unpaired) from both files
(@emph{union}).
@*
see note below regarding @code{-o auto}.


@item
@example
$ join -v 1 file1 file2
b 2
@end example
@tab
unpaired lines from the first file
(@emph{difference})


@item
@example
$ join -v 2 file1 file2
c C
@end example
@tab
unpaired lines from the second file
(@emph{difference})


@item
@example
$ join -v 1 -v 2 file1 file2
b 2
c C
@end example
@tab
unpaired lines from both files, omitting common lines
(@emph{symmetric difference}).


@end multitable

@noindent
The @option{-o auto -e X} options are useful when dealing with unpaired lines.
The following example prints all lines (common and unpaired) from both files.
Without @option{-o auto} it is not easy to discern which fields originate from
which file:

@example
$ join -a 1 -a 2 file1 file2
a 1 A
b 2
c C

$ join -o auto -e X -a 1 -a 2 file1 file2
a 1 A
b 2 X
c X C
@end example


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:

@example
@group
$ 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
@end group
@end example


@node Header lines
@subsection Header lines

The @option{--header} option can be used when the files to join
have a header line which is not sorted:

@example
@group
$ cat file1
Name     Age
Alice    25
Charlie  34

$ cat file2
Name   Country
Alice  France
Bob    Spain

$ join --header -o auto -e NA -a1 -a2 file1 file2
Name     Age   Country
Alice    25    France
Bob      NA    Spain
Charlie  34    NA
@end group
@end example


To sort a file with a header line, use GNU @command{sed -u}.
The following example sort the files but keeps the first line of each
file in place:

@example
@group
$ ( sed -u 1q ; sort -k2b,2 ) < file1 > file1.sorted
$ ( sed -u 1q ; sort -k2b,2 ) < file2 > file2.sorted
$ join --header -o auto -e NA -a1 -a2 file1.sorted file2.sorted > file3
@end group
@end example

@node Set operations
@subsection Union, Intersection and Difference of files

Combine @command{sort}, @command{uniq} and @command{join} to
perform the equivalent of set operations on files:

@c From https://www.pixelbeat.org/cmdline.html#sets
@multitable @columnfractions 0.5 0.5
@headitem Command @tab outcome
@item @code{sort -u file1 file2}
@tab Union of unsorted files

@item @code{sort file1 file2 | uniq -d}
@tab Intersection of unsorted files

@item @code{sort file1 file1 file2 | uniq -u}
@tab Difference of unsorted files

@item @code{sort file1 file2 | uniq -u}
@tab Symmetric Difference of unsorted files

@item @code{join -t '' -a1 -a2 file1 file2}
@tab Union of sorted files

@item @code{join -t '' file1 file2}
@tab Intersection of sorted files

@item @code{join -t '' -v2 file1 file2}
@tab Difference of sorted files

@item @code{join -t '' -v1 -v2 file1 file2}
@tab Symmetric Difference of sorted files

@end multitable

All examples above operate on entire lines and not on specific fields:
@command{sort} without @option{-k} and @command{join -t ''} both consider
entire lines as the key.


@node Operating on characters
@chapter Operating on characters

@cindex operating on characters

These commands operate on individual characters.

@menu
* tr invocation::               Translate, squeeze, and/or delete characters.
* expand invocation::           Convert tabs to spaces.
* unexpand invocation::         Convert spaces to tabs.
@end menu


@node tr invocation
@section @command{tr}: Translate, squeeze, and/or delete characters

@pindex tr

Synopsis:

@example
tr [@var{option}]@dots{} @var{string1} [@var{string2}]
@end example

@command{tr} copies standard input to standard output, performing
one of the following operations:

@itemize @bullet
@item
translate, and optionally squeeze repeated characters in the result,
@item
squeeze repeated characters,
@item
delete characters,
@item
delete characters, then squeeze repeated characters from the result.
@end itemize

The @var{string1} and @var{string2} operands define arrays of
characters @var{array1} and @var{array2}.  By default @var{array1}
lists input characters that @command{tr} operates on, and @var{array2}
lists corresponding translations.  In some cases the second operand is
omitted.

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands.

@table @samp

@item -c
@itemx -C
@itemx --complement
@opindex -c
@opindex -C
@opindex --complement
Instead of @var{array1}, use its complement (all characters not
specified by @var{string1}), in ascending order.  Use this option with
caution in multibyte locales where its meaning is not always clear
or portable; see @ref{Character arrays}.

@item -d
@itemx --delete
@opindex -d
@opindex --delete
Delete characters in @var{array1}; do not translate.

@item -s
@itemx --squeeze-repeats
@opindex -s
@opindex --squeeze-repeats
Replace each sequence of a repeated character that is listed in
the last specified @var{array}, with a single occurrence of that character.

@item -t
@itemx --truncate-set1
@opindex -t
@opindex --truncate-set1
Truncate @var{array1} to the length of @var{array2}.

@end table


@exitstatus

@menu
* Character arrays::            Specifying arrays of characters.
* Translating::                 Changing characters to other characters.
* Squeezing and deleting::      Removing characters.
@end menu


@node Character arrays
@subsection Specifying arrays of characters

@cindex arrays of characters in @command{tr}

The @var{string1} and @var{string2} operands are not regular
expressions, even though they may look similar.  Instead, they
merely represent arrays of characters.  As a GNU extension to POSIX,
an empty string operand represents an empty array of characters.

The interpretation of @var{string1} and @var{string2} depends on locale.
GNU @command{tr} fully supports only safe single-byte locales,
where each possible input byte represents a single character.
Unfortunately, this means GNU @command{tr} will not handle commands
like @samp{tr @"o @L{}} the way you might expect,
since (assuming a UTF-8 encoding) this is equivalent to
@samp{tr '\303\266' '\305\201'} and GNU @command{tr} will
simply transliterate all @samp{\303} bytes to @samp{\305} bytes, etc.
POSIX does not clearly specify the behavior of @command{tr} in locales
where characters are represented by byte sequences instead of by
individual bytes, or where data might contain invalid bytes that are
encoding errors.  To avoid problems in this area, you can run
@command{tr} in a safe single-byte locale by using a shell command
like @samp{LC_ALL=C tr} instead of plain @command{tr}.

Although most characters simply represent themselves in @var{string1}
and @var{string2}, the strings can contain shorthands listed below,
for convenience.  Some shorthands can be used only in @var{string1} or
@var{string2}, as noted below.

@table @asis

@item Backslash escapes
@cindex backslash escapes

The following backslash escape sequences are recognized:

@table @samp
@item \a
Bell (BEL, Control-G).
@item \b
Backspace (BS, Control-H).
@item \f
Form feed (FF, Control-L).
@item \n
Newline (LF, Control-J).
@item \r
Carriage return (CR, Control-M).
@item \t
Tab (HT, Control-I).
@item \v
Vertical tab (VT, Control-K).
@item \@var{ooo}
The eight-bit byte with the value given by @var{ooo}, which is the longest
sequence of one to three octal digits following the backslash.
For portability, @var{ooo} should represent a value that fits in eight bits.
As a GNU extension to POSIX, if the value would not fit, then only the
first two digits of @var{ooo} are used, e.g., @samp{\400}
is equivalent to @samp{\0400} and represents a two-byte sequence.
@item \\
A backslash.
@end table

It is an error if no character follows an unescaped backslash.
As a GNU extension, a backslash followed by a character not listed
above is interpreted as that character, removing any special
significance; this can be used to escape the characters
@samp{[} and @samp{-} when they would otherwise be special.

@item Ranges
@cindex ranges

The notation @samp{@var{m}-@var{n}} expands to the characters
from @var{m} through @var{n}, in ascending order.  @var{m} should
not collate after @var{n}; if it does, an error results.  As an example,
@samp{0-9} is the same as @samp{0123456789}.

GNU @command{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, @samp{tr -d '[0-9]'} deletes brackets
as well as digits.

Many historically common and even accepted uses of ranges are not fully
portable.  For example, on EBCDIC hosts using the @samp{A-Z}
range will not do what most would expect because @samp{A} through @samp{Z}
are not contiguous as they are in ASCII@.
One 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.

@item Repeated characters
@cindex repeated characters

The notation @samp{[@var{c}*@var{n}]} in @var{string2} expands to @var{n}
copies of character @var{c}.  Thus, @samp{[y*6]} is the same as
@samp{yyyyyy}.  The notation @samp{[@var{c}*]} in @var{string2} expands
to as many copies of @var{c} as are needed to make @var{array2} as long as
@var{array1}.  If @var{n} begins with @samp{0}, it is interpreted in
octal, otherwise in decimal.  A zero-valued @var{n} is treated as if
it were absent.

@item Character classes
@cindex character classes

The notation @samp{[:@var{class}:]} expands to all characters in
the (predefined) class @var{class}.  When the @option{--delete} (@option{-d})
and @option{--squeeze-repeats} (@option{-s}) options are both given, any
character class can be used in @var{string2}.  Otherwise, only the
character classes @code{lower} and @code{upper} are accepted in
@var{string2}, and then only if the corresponding character class
(@code{upper} and @code{lower}, respectively) is specified in the same
relative position in @var{string1}.  Doing this specifies case conversion.
Except for case conversion, a class's characters appear in no particular order.
The class names are given below; an error results when an invalid class
name is given.

@table @code
@item alnum
@opindex alnum
Letters and digits.
@item alpha
@opindex alpha
Letters.
@item blank
@opindex blank
Horizontal whitespace.
@item cntrl
@opindex cntrl
Control characters.
@item digit
@opindex digit
Digits.
@item graph
@opindex graph
Printable characters, not including space.
@item lower
@opindex lower
Lowercase letters.
@item print
@opindex print
Printable characters, including space.
@item punct
@opindex punct
Punctuation characters.
@item space
@opindex space
Horizontal or vertical whitespace.
@item upper
@opindex upper
Uppercase letters.
@item xdigit
@opindex xdigit
Hexadecimal digits.
@end table

@item Equivalence classes
@cindex equivalence classes

The syntax @samp{[=@var{c}=]} expands to all characters equivalent to
@var{c}, in no particular order.  These equivalence classes are
allowed in @var{string2} only when @option{--delete} (@option{-d}) and
@option{--squeeze-repeats} @option{-s} are both given.

Although equivalence classes are intended to support non-English alphabets,
there seems to be no standard way to define them or determine their
contents.  Therefore, they are not fully implemented in GNU @command{tr};
each character's equivalence class consists only of that character,
which is of no particular use.

@end table


@node Translating
@subsection Translating

@cindex translating characters

@command{tr} performs translation when @var{string1} and @var{string2} are
both given and the @option{--delete} (@option{-d}) option is not given.
@command{tr} translates each character of its input that is in @var{array1}
to the corresponding character in @var{array2}.  Characters not in
@var{array1} are passed through unchanged.

As a GNU extension to POSIX, when a character appears more than once
in @var{array1}, only the final instance is used.  For example, these
two commands are equivalent:

@example
tr aaa xyz
tr a z
@end example

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

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

@noindent
However, ranges like @code{a-z} are not portable outside the C locale.

When @command{tr} is performing translation, @var{array1} and @var{array2}
typically have the same length.  If @var{array1} is shorter than
@var{array2}, the extra characters at the end of @var{array2} are ignored.

On the other hand, making @var{array1} longer than @var{array2} is not
portable; POSIX says that the result is undefined.  In this situation,
BSD @command{tr} pads @var{array2} to the length of @var{array1} by repeating
the last character of @var{array2} as many times as necessary.  System V
@command{tr} truncates @var{array1} to the length of @var{array2}.

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

Acting like System V @command{tr} in this case breaks the relatively common
BSD idiom:

@example
tr -cs A-Za-z0-9 '\012'
@end example

@noindent
because it converts only zero bytes (the first element in the
complement of @var{array1}), rather than all non-alphanumerics, to
newlines.

@noindent
By the way, the above idiom is not portable because it uses ranges, and
it assumes that the octal code for newline is 012.  Here is a better
way to write it:

@example
tr -cs '[:alnum:]' '[\n*]'
@end example


@node Squeezing and deleting
@subsection Squeezing repeats and deleting

@cindex squeezing repeat characters
@cindex deleting characters
@cindex removing characters

When given just the @option{--delete} (@option{-d}) option, @command{tr}
removes any input characters that are in @var{array1}.

When given just the @option{--squeeze-repeats} (@option{-s}) option
and not translating, @command{tr} replaces each input sequence of a
repeated character that is in @var{array1} with a single occurrence of
that character.

When given both @option{--delete} and @option{--squeeze-repeats}, @command{tr}
first performs any deletions using @var{array1}, then squeezes repeats
from any remaining characters using @var{array2}.

The @option{--squeeze-repeats} option may also be used when translating,
in which case @command{tr} first performs translation, then squeezes
repeats from any remaining characters using @var{array2}.

Here are some examples to illustrate various combinations of options:

@itemize @bullet

@item
Remove all zero bytes:

@example
tr -d '\0'
@end example

@item
Put all words on lines by themselves.  This converts all
non-alphanumeric characters to newlines, then squeezes each string
of repeated newlines into a single newline:

@example
tr -cs '[:alnum:]' '[\n*]'
@end example

@item
Convert each sequence of repeated newlines to a single newline.
I.e., delete empty lines:

@example
tr -s '\n'
@end example

@item
Find doubled occurrences of words in a document.
@c Separate the following two "the"s, so typo checkers don't complain.
For example, people often write ``the @w{}the'' with the repeated words
separated by a newline.  The Bourne shell script below works first
by converting each sequence of punctuation and blank characters to a
single newline.  That puts each ``word'' on a line by itself.
Next it maps all uppercase characters to lower case, and finally it
runs @command{uniq} with the @option{-d} option to print out only the words
that were repeated.

@example
#!/bin/sh
cat -- "$@@" \
  | tr -s '[:punct:][:blank:]' '[\n*]' \
  | tr '[:upper:]' '[:lower:]' \
  | uniq -d
@end example

@item
Deleting a small set of characters is usually straightforward.  For example,
to remove all @samp{a}s, @samp{x}s, and @samp{M}s you would do this:

@example
tr -d axM
@end example

However, when @samp{-} is one of those characters, it can be tricky because
@samp{-} has special meanings.  Performing the same task as above but also
removing all @samp{-} characters, we might try @code{tr -d -axM}, but
that would fail because @command{tr} would try to interpret @option{-a} as
a command-line option.  Alternatively, we could try putting the hyphen
inside the string, @code{tr -d a-xM}, but that wouldn't work either because
it would make @command{tr} interpret @code{a-x} as the range of characters
@samp{a}@dots{}@samp{x} rather than the three.
One way to solve the problem is to put the hyphen at the end of the list
of characters:

@example
tr -d axM-
@end example

Or you can use @samp{--} to terminate option processing:

@example
tr -d -- -axM
@end example

@end itemize


@node expand invocation
@section @command{expand}: Convert tabs to spaces

@pindex expand
@cindex tabs to spaces, converting
@cindex converting tabs to spaces

@command{expand} writes the contents of each given @var{file}, or standard
input if none are given or for a @var{file} of @samp{-}, to standard
output, with tab characters converted to the appropriate number of
spaces.  Synopsis:

@example
expand [@var{option}]@dots{} [@var{file}]@dots{}
@end example

By default, @command{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 @option{-t 8} (set
tabs every 8 columns).

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -t @var{tab1}[,@var{tab2}]@dots{}
@itemx --tabs=@var{tab1}[,@var{tab2}]@dots{}
@opindex -t
@opindex --tabs
@cindex tab stops, setting
If only one tab stop is given, set the tabs @var{tab1} spaces apart
(default is 8).  Otherwise, set the tabs at columns @var{tab1},
@var{tab2}, @dots{} (numbered from 0), and replace any tabs beyond the
last tab stop given with single spaces.
@macro gnuExpandTabs
Tab stops can be separated by blanks as well as by commas.

As a GNU extension the last @var{tab} specified can be prefixed
with a @samp{/} to indicate a tab size to use for remaining positions.
For example, @option{--tabs=2,4,/8} will set tab stops at position 2 and 4,
and every multiple of 8 after that.

Also the last @var{tab} specified can be prefixed with a @samp{+} to indicate
a tab size to use for remaining positions, offset from the final explicitly
specified tab stop.
For example, to ignore the 1 character gutter present in diff output,
one can specify a 1 character offset using @option{--tabs=1,+8},
which will set tab stops at positions 1,9,17,@dots{}
@end macro
@gnuExpandTabs


For compatibility, GNU @command{expand} also accepts the obsolete
option syntax, @option{-@var{t1}[,@var{t2}]@dots{}}.  New scripts
should use @option{-t @var{t1}[,@var{t2}]@dots{}} instead.

@item -i
@itemx --initial
@opindex -i
@opindex --initial
@cindex initial tabs, converting
Only convert initial tabs (those that precede all non-space or non-tab
characters) on each line to spaces.

@end table

@exitstatus


@node unexpand invocation
@section @command{unexpand}: Convert spaces to tabs

@pindex unexpand

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

@example
unexpand [@var{option}]@dots{} [@var{file}]@dots{}
@end example

By default, @command{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 @ref{Common options}.

@table @samp

@item -t @var{tab1}[,@var{tab2}]@dots{}
@itemx --tabs=@var{tab1}[,@var{tab2}]@dots{}
@opindex -t
@opindex --tabs
If only one tab stop is given, set the tabs @var{tab1} columns apart
instead of the default 8.  Otherwise, set the tabs at columns
@var{tab1}, @var{tab2}, @dots{} (numbered from 0), and leave blanks
beyond the tab stops given unchanged.
@gnuExpandTabs

This option implies the @option{-a} option.

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

@item -a
@itemx --all
@opindex -a
@opindex --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.

@end table

@exitstatus


@node Directory listing
@chapter Directory listing

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

@menu
* ls invocation::               List directory contents.
* dir invocation::              Briefly ls.
* vdir invocation::             Verbosely ls.
* dircolors invocation::        Color setup for ls, etc.
@end menu


@node ls invocation
@section @command{ls}: List directory contents

@pindex ls
@cindex directory listing

The @command{ls} program lists information about files (of any type,
including directories).  Options and file arguments can be intermixed
arbitrarily, as usual.  Later options override earlier options that
are incompatible.

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

@vindex LC_ALL
By default, the output is sorted alphabetically, according to the locale
settings in effect.@footnote{If you use a non-POSIX
locale (e.g., by setting @env{LC_ALL} to @samp{en_US}), then @command{ls} may
produce output that is sorted differently than you're accustomed to.
In that case, set the @env{LC_ALL} environment variable to @samp{C}.}
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 @command{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 @command{ls}'s operation.

@cindex exit status of @command{ls}
Exit status:

@display
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)
@end display

Also see @ref{Common options}.

@menu
* Which files are listed::
* What information is listed::
* Sorting the output::
* General output formatting::
* Formatting file timestamps::
* Formatting the file names::
@end menu


@node Which files are listed
@subsection Which files are listed

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

@table @samp

@item -a
@itemx --all
@opindex -a
@opindex --all
In directories, do not ignore file names that start with @samp{.}.

@item -A
@itemx --almost-all
@opindex -A
@opindex --almost-all
In directories, do not ignore all file names that start with @samp{.};
ignore only @file{.} and @file{..}.  The @option{--all} (@option{-a})
option overrides this option.

@item -B
@itemx --ignore-backups
@opindex -B
@opindex --ignore-backups
@cindex backup files, ignoring
In directories, ignore files that end with @samp{~}.  This option is
equivalent to @samp{--ignore='*~' --ignore='.*~'}.

@item -d
@itemx --directory
@opindex -d
@opindex --directory
List just the names of directories, as with other types of files, rather
than listing their contents.
@c The following sentence is the same as the one for -F.
Do not follow symbolic links listed on the
command line unless the @option{--dereference-command-line} (@option{-H}),
@option{--dereference} (@option{-L}), or
@option{--dereference-command-line-symlink-to-dir} options are specified.

@item -H
@itemx --dereference-command-line
@opindex -H
@opindex --dereference-command-line
@cindex symbolic links, dereferencing
If a command line argument specifies a symbolic link, show information
for the file the link references rather than for the link itself.

@item --dereference-command-line-symlink-to-dir
@opindex --dereference-command-line-symlink-to-dir
@cindex symbolic links, dereferencing
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 unless long format is being used
or any of the following options is in effect:
@option{--classify} (@option{-F}),
@option{--directory} (@option{-d}),
@option{--dereference} (@option{-L}), or
@option{--dereference-command-line} (@option{-H})).

@item --group-directories-first
@opindex --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 @option{--sort=none}
(@option{-U}) disables this option altogether.

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

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

@item -I @var{pattern}
@itemx --ignore=@var{pattern}
@opindex -I
@opindex --ignore=@var{pattern}
In directories, ignore files whose names match the shell pattern
(not regular expression) @var{pattern}.  As
in the shell, an initial @samp{.} in a file name does not match a
wildcard at the start of @var{pattern}.  Sometimes it is useful
to give this option several times.  For example,

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

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

@item -L
@itemx --dereference
@opindex -L
@opindex --dereference
@cindex symbolic links, dereferencing
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, @command{ls} still prints the name
of the link itself, not the name of the file that the link points to.

@item -R
@itemx --recursive
@opindex -R
@opindex --recursive
@cindex recursive directory listing
@cindex directory listing, recursive
List the contents of all directories recursively.

@end table


@node What information is listed
@subsection What information is listed

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

@table @samp

@item --author
@opindex --author
@cindex hurd, author, printing
In long format, list each file's author.
In GNU/Hurd, file authors can differ from their owners, but in other
operating systems the two are the same.

@item -D
@itemx --dired
@opindex -D
@opindex --dired
@cindex dired Emacs mode support
Print an additional line after the main output:

@example
//DIRED// @var{beg1} @var{end1} @var{beg2} @var{end2} @dots{}
@end example

@noindent
The @var{begn} and @var{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 via
@option{--recursive} (@option{-R}), output a similar
line with offsets for each subdirectory name:

@example
//SUBDIRED// @var{beg1} @var{end1} @dots{}
@end example

Finally, output a line of the form:

@example
//DIRED-OPTIONS// --quoting-style=@var{word}
@end example

@noindent
where @var{word} is the quoting style (@pxref{Formatting the file names}).

Here is an actual example:

@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
@end example

The pairs of offsets on the @samp{//DIRED//} line above delimit
these names: @file{f1}, @file{f2}, @file{sub}, @file{sub2}, @file{deeper},
@file{file}.
The offsets on the @samp{//SUBDIRED//} line delimit the following
directory names: @file{a}, @file{a/sub}, @file{a/sub/deeper}, @file{a/sub2}.

Here is an example of how to extract the fifth entry name, @samp{deeper},
corresponding to the pair of offsets, 222 and 228:

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

Although the listing above includes a trailing slash
for the @samp{deeper} entry, the offsets select the name without
the trailing slash.  However, if you invoke @command{ls} with @option{--dired}
(@option{-D}) along with an option like
@option{--escape} (@option{-b}) and operate
on a file whose name contains special characters, the backslash
@emph{is} included:

@example
$ 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
@end example

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

The @option{--dired} (@option{-D}) option has well-defined behavior
only when long format is in effect and hyperlinks are disabled (e.g.,
@option{--hyperlink=none}).

@item --full-time
@opindex --full-time
Produce long format, and list times in full.  It is
equivalent to using @option{--format=long} (@option{-l}) with
@option{--time-style=full-iso} (@pxref{Formatting file timestamps}).

@item -g
@opindex -g
Produce long format, but omit owner information.

@item -G
@itemx --no-group
@opindex -G
@opindex --no-group
Inhibit display of group information in long format.
(This is the default in some non-GNU versions of @command{ls}, so we
provide this option for compatibility.)

@optHumanReadable

@item -i
@itemx --inode
@opindex -i
@opindex --inode
@cindex inode number, printing
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.)

@item -l
@itemx --format=long
@itemx --format=verbose
@opindex -l
@opindex --format
@opindex long ls @r{format}
@opindex verbose ls @r{format}
Produce long format.
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 (@pxref{Formatting file timestamps}), normally
the modification timestamp (the mtime, @pxref{File timestamps}).
If the owner or group name cannot be determined, print
the owner or group ID instead, right-justified as a cue
that it is a number rather than a textual name.
Print question marks for other information that
cannot be determined.

Normally the size is printed as a byte count without punctuation, but
this can be overridden (@pxref{Block size}).
For example, @option{--human-readable} (@option{-h})
prints an abbreviated, human-readable count, and
@samp{--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
@samp{total @var{blocks}}, where @var{blocks} is the file system allocation
for all files in that directory.  The block size currently defaults to 1024
bytes, but this can be overridden (@pxref{Block size}).
The @var{blocks} computed counts each hard link separately;
this is arguably a deficiency.

The file type is one of the following characters:

@c The commented-out entries are ones we're not sure about.

@table @samp
@item -
regular file
@item b
block special file
@item c
character special file
@item C
high performance (``contiguous data'') file
@item d
directory
@item D
door (Solaris)
@c @item F
@c semaphore, if this is a distinct file type
@item l
symbolic link
@c @item m
@c multiplexed file (7th edition Unix; obsolete)
@item M
off-line (``migrated'') file (Cray DMF)
@item n
network special file (HP-UX)
@item p
FIFO (named pipe)
@item P
port (Solaris)
@c @item Q
@c message queue, if this is a distinct file type
@item s
socket
@c @item S
@c shared memory object, if this is a distinct file type
@c @item T
@c typed memory object, if this is a distinct file type
@c @item w
@c whiteout (4.4BSD; not implemented)
@item ?
some other file type
@end table

@cindex permissions, output by @command{ls}
The file mode bits listed are similar to symbolic mode specifications
(@pxref{Symbolic Modes}).  But @command{ls} combines multiple bits into the
third character of each set of permissions as follows:

@table @samp
@item s
If the set-user-ID or set-group-ID bit and the corresponding executable bit
are both set.

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

@item 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.  @xref{Mode Structure}.

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

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

@item -
Otherwise.
@end table

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 @command{ls} uses a @samp{.} 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 @samp{+} character.

@item -n
@itemx --numeric-uid-gid
@opindex -n
@opindex --numeric-uid-gid
@cindex numeric uid and gid
@cindex numeric user and group IDs
Produce long format, but
display right-justified numeric user and group IDs
instead of left-justified owner and group names.

@item -o
@opindex -o
Produce long format, but omit group information.
It is equivalent to using @option{--format=long} (@option{-l})
with @option{--no-group} (@option{-G}).

@item -s
@itemx --size
@opindex -s
@opindex --size
@cindex file system allocation
@cindex size of files, reporting
Print the file system allocation of each file to the left of the file name.
This is the amount of file system 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 allocation is printed in units of
1024 bytes, but this can be overridden (@pxref{Block size}).

@cindex NFS mounts from BSD to HP-UX
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 @command{ls} program.

@optSi

@item -Z
@itemx --context
@opindex -Z
@opindex --context
@cindex SELinux
@cindex security context
Display the SELinux security context or @samp{?} if none is found.
In long format, print the security context to the left of the size column.

@end table


@node Sorting the output
@subsection Sorting the output

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

@table @samp

@item -c
@itemx --time=ctime
@itemx --time=status
@opindex -c
@opindex --time
@opindex ctime@r{, printing or sorting by}
@opindex status time@r{, printing or sorting by}
@opindex use time@r{, printing or sorting files by}
In long format,
print the status change timestamp (the ctime) instead of the mtime.
When sorting by time or when not using long format,
sort according to the ctime.  @xref{File timestamps}.

@item -f
@opindex -f
@cindex unsorted directory listing
@cindex directory order, listing by
Produce an unsorted directory listing.
This is equivalent to the combination of @option{--all} (@option{-a}),
@option{--sort=none} (@option{-U}), @option{-1},
@option{--color=none}, and @option{--hyperlink=none},
while also disabling any previous use of @option{--size} (@option{-s}).

@item -r
@itemx --reverse
@opindex -r
@opindex --reverse
@cindex reverse sorting
Reverse whatever the sorting method is---e.g., list files in reverse
alphabetical order, youngest first, smallest first, or whatever.
This option has no effect when @option{--sort=none} (@option{-U})
is in effect.

@item -S
@itemx --sort=size
@opindex -S
@opindex --sort
@opindex size of files@r{, sorting files by}
Sort by file size, largest first.

@item -t
@itemx --sort=time
@opindex -t
@opindex --sort
@opindex modification timestamp@r{, sorting files by}
Sort by modification timestamp (mtime) by default, newest first.
The timestamp to order by can be changed with the @option{--time} option.
@xref{File timestamps}.

@item -u
@itemx --time=atime
@itemx --time=access
@itemx --time=use
@opindex -u
@opindex --time
@opindex use time@r{, printing or sorting files by}
@opindex atime@r{, printing or sorting files by}
@opindex access timestamp@r{, printing or sorting files by}
In long format, print the last access timestamp (the atime).
When sorting by time or when not using long format,
sort according to the atime.
@xref{File timestamps}.

@item --time=mtime
@itemx --time=modification
@opindex --time
@opindex data modification time@r{, printing or sorting files by}
@opindex mtime@r{, printing or sorting files by}
This is the default timestamp display and sorting mode.
In long format, print the last data modification timestamp (the mtime).
When sorting by time or when not using long format,
sort according to the mtime.
@xref{File timestamps}.

@item --time=birth
@itemx --time=creation
@opindex --time
@opindex birth time@r{, printing or sorting files by}
@opindex creation timestamp@r{, printing or sorting files by}
In long format, print the file creation timestamp if available,
falling back to the file modification timestamp (mtime) if not.
When sorting by time or when not using long format,
sort according to the birth time.
@xref{File timestamps}.

@item -U
@itemx --sort=none
@opindex -U
@opindex --sort
@opindex none@r{, sorting option for @command{ls}}
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 @option{-f} does.)  This can be useful when listing large
directories, where sorting can take some time.

@item -v
@itemx --sort=version
@opindex -v
@opindex --sort
@opindex version@r{, sorting option for @command{ls}}
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.  @xref{Version sort ordering}.

@item --sort=width
@opindex --sort
@opindex width@r{, sorting option for @command{ls}}
Sort by printed width of file names.
This can be useful with the @option{--format=vertical} (@option{-C})
output format, to most densely display the listed files.

@item -X
@itemx --sort=extension
@opindex -X
@opindex --sort
@opindex extension@r{, sorting files by}
Sort directory contents alphabetically by file extension (characters
after the last @samp{.}); files with no extension are sorted first.

@end table


@node General output formatting
@subsection General output formatting

These options affect the appearance of the overall output.

@table @samp

@item --format=single-column
@opindex --format
@opindex single-column @r{output of files}
List one file name per line, with no other information.
This is the default for @command{ls} when standard
output is not a terminal.  See also the @option{--escape} (@option{-b}),
@option{--hide-control-chars} (@option{-q}), and @option{--zero} options
to disambiguate output of file names containing newline characters.

@item -1
@opindex -1
List one file per line.  This is like @option{--format=single-column}
except that it has no effect if long format is also in effect.

@item -C
@itemx --format=vertical
@opindex -C
@opindex --format
@opindex vertical @r{sorted files in columns}
List files in columns, sorted vertically, with no other information.
This is the default for @command{ls} if standard output is a terminal.
It is always the default for the @command{dir} program.
GNU @command{ls} uses variable width columns to display as many files as
possible in the fewest lines.

@item --color [=@var{when}]
@opindex --color
@cindex color, distinguishing file types with
Specify whether to use color for distinguishing file types; @var{when}
may be omitted, or one of:
@itemize @bullet
@item none
@vindex none @r{color option}
- Do not use color at all.  This is the default.
@item auto
@vindex auto @r{color option}
@cindex terminal, using color iff
- Only use color if standard output is a terminal.
@item always
@vindex always @r{color option}
- Always use color.
@end itemize
Specifying @option{--color} and no @var{when} is equivalent to
@option{--color=always}.
If piping a colored listing through a pager like @command{less},
use the pager's @option{-R} option to pass the color codes to the terminal.

@vindex LS_COLORS
@vindex SHELL @r{environment variable, and color}
Using the @option{--color} option may incur a noticeable
performance penalty when run in a large directory,
because the default settings require that @command{ls} @code{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
@command{dircolors} to set the @env{LS_COLORS} environment variable like this,
@example
eval $(dircolors -p | perl -pe \
  's/^((CAP|S[ET]|O[TR]|M|E)\w+).*/$1 00/' | dircolors -)
@end example
and on a @code{dirent.d_type}-capable file system, @command{ls}
will perform only one @code{stat} call per command line argument.

@item -F
@itemx --classify [=@var{when}]
@itemx --indicator-style=classify
@opindex -F
@opindex --classify
@opindex --indicator-style
@cindex file type and executables, marking
@cindex executables and file type, marking
Append a character to each file name indicating the file type.  Also,
for regular files that are executable, append @samp{*}.  The file type
indicators are @samp{/} for directories, @samp{@@} for symbolic links,
@samp{|} for FIFOs, @samp{=} for sockets, @samp{>} for doors,
and nothing for regular files.
@var{when} may be omitted, or one of:
@itemize @bullet
@item none
@vindex none @r{classify option}
- Do not classify.  This is the default.
@item auto
@vindex auto @r{classify option}
@cindex terminal, using classify iff
- Only classify if standard output is a terminal.
@item always
@vindex always @r{classify option}
- Always classify.
@end itemize
Specifying @option{--classify} and no @var{when} is equivalent to
@option{--classify=always}.
@c The following sentence is the same as the one for -d.
Do not follow symbolic links listed on the
command line unless the @option{--dereference-command-line} (@option{-H}),
@option{--dereference} (@option{-L}), or
@option{--dereference-command-line-symlink-to-dir} options are specified.

@item --file-type
@itemx --indicator-style=file-type
@opindex --file-type
@opindex --indicator-style
@cindex file type, marking
Append a character to each file name indicating the file type.  This is
like @option{--classify} (@option{-F}, except that executables are not marked.

@item --hyperlink [=@var{when}]
@opindex --hyperlink
@cindex hyperlink, linking to files
Output codes recognized by some terminals to link
to files using the @samp{file://} URI format.
@var{when} may be omitted, or one of:
@itemize @bullet
@item none
@vindex none @r{hyperlink option}
- Do not use hyperlinks at all.  This is the default.
@item auto
@vindex auto @r{hyperlink option}
@cindex terminal, using hyperlink iff
- Only use hyperlinks if standard output is a terminal.
@item always
@vindex always @r{hyperlink option}
- Always use hyperlinks.
@end itemize
Specifying @option{--hyperlink} and no @var{when} is equivalent to
@option{--hyperlink=always}.

@item --indicator-style=@var{word}
@opindex --indicator-style
Append a character indicator with style @var{word} to entry names,
as follows:

@table @samp
@item none
Do not append any character indicator; this is the default.
@item slash
Append @samp{/} for directories.  This is the same as the @option{-p}
option.
@item file-type
Append @samp{/} for directories, @samp{@@} for symbolic links, @samp{|}
for FIFOs, @samp{=} for sockets, and nothing for regular files.  This is
the same as the @option{--file-type} option.
@item classify
Append @samp{*} for executable regular files, otherwise behave as for
@samp{file-type}.  This is the same as the @option{--classify}
(@option{-F}) option.
@end table

@item -k
@itemx --kibibytes
@opindex -k
@opindex --kibibytes
Set the default block size to its normal value of 1024 bytes,
overriding any contrary specification in environment variables
(@pxref{Block size}).  If @option{--block-size},
@option{--human-readable} (@option{-h}), or @option{--si} options are used,
they take precedence even if @option{--kibibytes} (@option{-k}) is placed after

The @option{--kibibytes} (@option{-k}) option affects the
per-directory block count written in long format,
and the file system allocation written by the @option{--size} (@option{-s})
option.  It does not affect the file size in bytes that is written in
long format.

@item -m
@itemx --format=commas
@opindex -m
@opindex --format
@opindex commas@r{, outputting between files}
List files horizontally, with as many as will fit on each line,
separated by @samp{, } (a comma and a space),
and with no other information.

@item -p
@itemx --indicator-style=slash
@opindex -p
@opindex --indicator-style
@cindex file type, marking
Append a @samp{/} to directory names.

@item -x
@itemx --format=across
@itemx --format=horizontal
@opindex -x
@opindex --format
@opindex across@r{, listing files}
@opindex horizontal@r{, listing files}
List the files in columns, sorted horizontally.

@item -T @var{cols}
@itemx --tabsize=@var{cols}
@opindex -T
@opindex --tabsize
Assume that each tab stop is @var{cols} columns wide.  The default is 8.
@command{ls} uses tabs where possible in the output, for efficiency.  If
@var{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
@option{-T0} option or put @code{TABSIZE=0} in your environment, to tell
@command{ls} to align using spaces, not tabs.

@item -w @var{cols}
@itemx --width=@var{cols}
@opindex -w
@opindex --width
@vindex COLUMNS
Assume the screen is @var{cols} columns wide.  The default is taken
from the terminal settings if possible; otherwise the environment
variable @env{COLUMNS} is used if it is set; otherwise the default
is 80.  With a @var{cols} value of @samp{0}, there is no limit on
the length of the output line, and that single output line will
be delimited with spaces, not tabs.

@item --zero
@opindex --zero
@outputNUL
This option is incompatible with the @option{--dired} (@option{-D}) option.
This option also implies the options @option{--show-control-chars},
@option{-1}, @option{--color=none}, and
@option{--quoting-style=literal} (@option{-N}).

@end table


@node Formatting file timestamps
@subsection Formatting file timestamps

By default, file timestamps are listed in abbreviated form, using
a date like @samp{Mar 30@ @ 2020} for non-recent timestamps, and a
date-without-year and time like @samp{Mar 30 23:45} for recent timestamps.
This format can change depending on the current locale as detailed below.

@cindex clock skew
A timestamp is considered to be @dfn{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 @command{make} that rely on file timestamps.
@xref{File timestamps}.

@vindex TZ
Timestamps are listed according to the time zone rules specified by
the @env{TZ} environment variable, or by the system default rules if
@env{TZ} is not set.  @xref{TZ Variable,, Specifying the Time Zone
with @env{TZ}, libc, The GNU C Library Reference Manual}.

The following option changes how file timestamps are printed.

@table @samp
@item --time-style=@var{style}
@opindex --time-style
@cindex time style
List timestamps in style @var{style}.  The @var{style} should
be one of the following:

@table @samp
@item +@var{format}
@vindex LC_TIME
List timestamps using @var{format}, where @var{format} is interpreted
like the format argument of @command{date} (@pxref{date invocation}).
For example, @option{--time-style="+%Y-%m-%d %H:%M:%S"} causes
@command{ls} to list timestamps like @samp{2020-03-30 23:45:56}.  As
with @command{date}, @var{format}'s interpretation is affected by the
@env{LC_TIME} locale category.

If @var{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.

@item full-iso
List timestamps in full using ISO 8601-like date, time, and time zone
components with nanosecond precision, e.g., @samp{2020-07-21
23:45:56.477817180 -0400}.  This style is equivalent to
@samp{+%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 @command{make}'s behavior, since GNU @command{make}
uses the full timestamp to determine whether a file is out of date.

@item long-iso
List ISO 8601 date and time components with minute precision, e.g.,
@samp{2020-03-30 23:45}.  These timestamps are shorter than
@samp{full-iso} timestamps, and are usually good enough for everyday
work.  This style is equivalent to @samp{+%Y-%m-%d %H:%M}.

@item iso
List ISO 8601 dates for non-recent timestamps (e.g.,
@samp{2020-03-30@ }), and ISO 8601-like month, day, hour, and
minute for recent timestamps (e.g., @samp{03-30 23:45}).  These
timestamps are uglier than @samp{long-iso} timestamps, but they carry
nearly the same information in a smaller space and their brevity helps
@command{ls} output fit within traditional 80-column output lines.
The following two @command{ls} invocations are equivalent:

@example
newline='
'
ls -l --time-style="+%Y-%m-%d $newline%m-%d %H:%M"
ls -l --time-style="iso"
@end example

@item locale
@vindex LC_TIME
List timestamps in a locale-dependent form.  For example, a French
locale might list non-recent timestamps like @samp{30 mars@ @ @ 2020}
and recent timestamps like @samp{30 mars@ @ 23:45}.  Locale-dependent
timestamps typically consume more space than @samp{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 @env{LC_TIME} locale category specifies the timestamp format.  The
default POSIX locale uses timestamps like @samp{Mar 30@
@ 2020} and @samp{Mar 30 23:45}; in this locale, the following two
@command{ls} invocations are equivalent:

@example
newline='
'
ls -l --time-style="+%b %e  %Y$newline%b %e %H:%M"
ls -l --time-style="locale"
@end example

Other locales behave differently.  For example, in a German locale,
@option{--time-style="locale"} might be equivalent to
@option{--time-style="+%e. %b %Y $newline%e. %b %H:%M"}
and might generate timestamps like @samp{30. M@"ar 2020@ } and
@samp{30. M@"ar 23:45}.

@item posix-@var{style}
@vindex LC_TIME
List POSIX-locale timestamps if the @env{LC_TIME} locale
category is POSIX, @var{style} timestamps otherwise.  For
example, the @samp{posix-long-iso} style lists
timestamps like @samp{Mar 30@ @ 2020} and @samp{Mar 30 23:45} when in
the POSIX locale, and like @samp{2020-03-30 23:45} otherwise.
@end table
@end table

@vindex TIME_STYLE
You can specify the default value of the @option{--time-style} option
with the environment variable @env{TIME_STYLE}; if @env{TIME_STYLE} is not set
the default style is @samp{locale}.  GNU Emacs 21.3 and
later use the @option{--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
@samp{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.


@node Formatting the file names
@subsection Formatting the file names

These options change how file names themselves are printed.

@table @samp

@item -b
@itemx --escape
@itemx --quoting-style=escape
@opindex -b
@opindex --escape
@opindex --quoting-style
@cindex backslash sequences for file names
Quote nongraphic characters in file names using alphabetic and octal
backslash sequences like those used in C.

@item -N
@itemx --literal
@itemx --quoting-style=literal
@opindex -N
@opindex --literal
@opindex --quoting-style
Do not quote file names.  However, with @command{ls} nongraphic
characters are still printed as question marks if the output is a
terminal and you do not specify the @option{--show-control-chars}
option.

@item -q
@itemx --hide-control-chars
@opindex -q
@opindex --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
@command{ls}.

@item -Q
@itemx --quote-name
@itemx --quoting-style=c
@opindex -Q
@opindex --quote-name
@opindex --quoting-style
Enclose file names in double quotes and quote nongraphic characters as
in C.

@item --quoting-style=@var{word}
@opindex --quoting-style
@cindex quoting style
Use style @var{word} to quote file names and other strings that may
contain arbitrary characters.  The @var{word} should
be one of the following:

@macro quotingStyles
@table @samp
@item literal
Output strings as-is; this is the same as the @option{--literal} (@option{-N})
option.
@item 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
@command{bash}, but it does not always work for incompatible shells
like @command{csh}.
@item shell-always
Quote strings for the shell, even if they would normally not require quoting.
@item shell-escape
Like @samp{shell}, but also quoting non-printable characters using the POSIX
proposed @samp{$''} syntax suitable for most shells.
@item shell-escape-always
Like @samp{shell-escape}, but quote strings even if they would
normally not require quoting.
@item c
Quote strings as for C character string literals, including the
surrounding double-quote characters; this is the same as the
@option{--quote-name} (@option{-Q}) option.
@item escape
Quote strings as for C character string literals, except omit the
surrounding double-quote
characters; this is the same as the @option{--escape} (@option{-b}) option.
@item clocale
Quote strings as for C character string literals, except use
surrounding quotation marks appropriate for the
locale.
@item locale
@c Use @t instead of @samp to avoid duplicate quoting in some output styles.
Quote strings as for C character string literals, except use
surrounding quotation marks appropriate for the locale, and quote
@t{'like this'} instead of @t{"like
this"} in the default C locale.  This looks nicer on many displays.
@end table
@end macro
@quotingStyles

You can specify the default value of the @option{--quoting-style} option
with the environment variable @env{QUOTING_STYLE}@.  If that environment
variable is not set, the default value is @samp{shell-escape} when the
output is a terminal, and @samp{literal} otherwise.

@item --show-control-chars
@opindex --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
@command{ls}.

@end table


@node dir invocation
@section @command{dir}: Briefly list directory contents

@pindex dir
@cindex directory listing, brief

@command{dir} is equivalent to @code{ls -C
-b}; that is, by default files are listed in columns, sorted vertically,
and special characters are represented by backslash escape sequences.

@xref{ls invocation, @command{ls}}.


@node vdir invocation
@section @command{vdir}: Verbosely list directory contents

@pindex vdir
@cindex directory listing, verbose

@command{vdir} is equivalent to @code{ls -l
-b}; that is, by default files are listed in long format and special
characters are represented by backslash escape sequences.

@xref{ls invocation, @command{ls}}.

@node dircolors invocation
@section @command{dircolors}: Color setup for @command{ls}

@pindex dircolors
@cindex color setup
@cindex setup for color

@command{dircolors} outputs a sequence of shell commands to set up the
terminal for color output from @command{ls} (and @command{dir}, etc.).
Typical usage:

@example
eval "$(dircolors [@var{option}]@dots{} [@var{file}])"
@end example

If @var{file} is specified, @command{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 @samp{dircolors --print-database}.

To make @command{dircolors} read a @file{~/.dircolors} file if it
exists, you can put the following lines in your @file{~/.bashrc} (or
adapt them to your favorite shell):

@example
d=.dircolors
test -r $d && eval "$(dircolors $d)"
@end example

@vindex LS_COLORS
@vindex SHELL @r{environment variable, and color}
The output is a shell command to set the @env{LS_COLORS} environment
variable.  You can specify the shell syntax to use on the command line,
or @command{dircolors} will guess it from the value of the @env{SHELL}
environment variable.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp
@item -b
@itemx --sh
@itemx --bourne-shell
@opindex -b
@opindex --sh
@opindex --bourne-shell
@cindex Bourne shell syntax for color setup
@cindex @command{sh} syntax for color setup
Output Bourne shell commands.  This is the default if the @env{SHELL}
environment variable is set and does not end with @samp{csh} or
@samp{tcsh}.

@item -c
@itemx --csh
@itemx --c-shell
@opindex -c
@opindex --csh
@opindex --c-shell
@cindex C shell syntax for color setup
@cindex @command{csh} syntax for color setup
Output C shell commands.  This is the default if @code{SHELL} ends with
@command{csh} or @command{tcsh}.

@item -p
@itemx --print-database
@opindex -p
@opindex --print-database
@cindex color database, printing
@cindex database for color setup, printing
@cindex printing color database
Print the (compiled-in) default color configuration database.  This
output is itself a valid configuration file, and is fairly descriptive
of the possibilities.

@item --print-ls-colors
@opindex --print-ls-colors
@cindex printing ls colors
Print the LS_COLORS entries on separate lines,
each colored as per the color they represent.

@end table

@exitstatus


@node Basic operations
@chapter Basic operations

@cindex manipulating files

This chapter describes the commands for basic file manipulation:
copying, moving (renaming), and deleting (removing).

@menu
* cp invocation::               Copy files.
* dd invocation::               Convert and copy a file.
* install invocation::          Copy files and set attributes.
* mv invocation::               Move (rename) files.
* rm invocation::               Remove files or directories.
* shred invocation::            Remove files more securely.
@end menu


@node cp invocation
@section @command{cp}: Copy files and directories

@pindex cp
@cindex copying files and directories
@cindex files, copying
@cindex directories, copying

@command{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:

@example
cp [@var{option}]@dots{} [-T] @var{source} @var{dest}
cp [@var{option}]@dots{} @var{source}@dots{} @var{directory}
cp [@var{option}]@dots{} -t @var{directory} @var{source}@dots{}
@end example

@itemize @bullet
@item
If two file names are given, @command{cp} copies the first file to the
second.

@item
If the @option{--target-directory} (@option{-t}) option is given, or
failing that if the last file is a directory and the
@option{--no-target-directory} (@option{-T}) option is not given,
@command{cp} copies each @var{source} file to the specified directory,
using the @var{source}s' names.
@end itemize

Generally, files are written just as they are read.  For exceptions,
see the @option{--sparse} option below.

By default, @command{cp} does not copy directories.  However, the
@option{-R}, @option{-a}, and @option{-r} options cause @command{cp} to
copy recursively by descending into source directories and copying files
to corresponding destination directories.

When copying from a symbolic link, @command{cp} normally follows the
link only when not copying recursively or when @option{--link}
(@option{-l}) is used.  This default can be overridden with the
@option{--archive} (@option{-a}), @option{-d}, @option{--dereference}
(@option{-L}), @option{--no-dereference} (@option{-P}), and
@option{-H} options.  If more than one of these options is specified,
the last one silently overrides the others.

When copying to a symbolic link, @command{cp} follows the
link only when it refers to an existing regular file.
However, when copying to a dangling symbolic link, @command{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 @env{POSIXLY_CORRECT} to make @command{cp} attempt to create
the target of a dangling destination symlink, in spite of the possible risk.
Also, when an option like
@option{--backup} or @option{--link} acts to rename or remove the
destination before copying, @command{cp} renames or removes the
symbolic link rather than the file it points to.

By default, @command{cp} copies the contents of special files only
when not copying recursively.  This default can be overridden with the
@option{--copy-contents} option.

@cindex self-backups
@cindex backups, making only
@command{cp} generally refuses to copy a file onto itself, with the
following exception: if @option{--force --backup} is specified with
@var{source} and @var{dest} identical, and referring to a regular file,
@command{cp} will make a backup file, either regular or numbered, as
specified in the usual ways (@pxref{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 @ref{Common options}.

@table @samp
@item -a
@itemx --archive
@opindex -a
@opindex --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., @samp{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 @option{-dR --preserve=all} with the reduced diagnostics.

@item --attributes-only
@opindex --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 @option{--preserve} option for controlling which attributes to copy.

@item -b
@itemx --backup[=@var{method}]
@opindex -b
@opindex --backup
@vindex VERSION_CONTROL
@cindex backups, making
@xref{Backup options}.
Make a backup of each file that would otherwise be overwritten or removed.
As a special case, @command{cp} makes a backup of @var{source} when the force
and backup options are given and @var{source} and @var{dest} are the same
name for an existing, regular file.  One useful application of this
combination of options is this tiny Bourne shell script:

@example
#!/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
@end example

@item --copy-contents
@cindex directories, copying recursively
@cindex copying directories recursively
@cindex recursively copying directories
@cindex non-directories, copying as special files
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 @file{/dev} directory.  In most cases,
@code{cp -R --copy-contents} will hang indefinitely trying to read
from FIFOs and special files like @file{/dev/console}, and it will
fill up your destination file system if you use it to copy @file{/dev/zero}.
This option has no effect unless copying recursively, and it does not
affect the copying of symbolic links.

@item -d
@opindex -d
@cindex symbolic links, copying
@cindex hard links, preserving
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 @option{--no-dereference --preserve=links}.

@item -f
@itemx --force
@opindex -f
@opindex --force
When copying without this option and an existing destination file cannot
be opened for writing, the copy fails.  However, with @option{--force},
when a destination file cannot be opened, @command{cp} then
tries to recreate the file by first removing it.  Note @option{--force}
alone will not remove dangling symlinks.
When this option is combined with
@option{--link} (@option{-l}) or @option{--symbolic-link}
(@option{-s}), the destination link is replaced, and unless
@option{--backup} (@option{-b}) is also given there is no brief
moment when the destination does not exist.  Also see the
description of @option{--remove-destination}.

This option is independent of the @option{--interactive} or
@option{-i} option: neither cancels the effect of the other.

This option is ignored when the @option{--no-clobber} or @option{-n} option
is also used.

@item -H
@opindex -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.

@item -i
@itemx --interactive
@opindex -i
@opindex --interactive
When copying a file other than a directory, prompt whether to
overwrite an existing destination file.  The @option{-i} option overrides
a previous @option{-n} option.

@item -l
@itemx --link
@opindex -l
@opindex --link
Make hard links instead of copies of non-directories.

@item -L
@itemx --dereference
@opindex -L
@opindex --dereference
Follow symbolic links when copying from them.
With this option, @command{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.

@item -n
@itemx --no-clobber
@opindex -n
@opindex --no-clobber
Do not overwrite an existing file; silently do nothing instead.
This option overrides a previous
@option{-i} option.  This option is mutually exclusive with @option{-b} or
@option{--backup} option.

@item -P
@itemx --no-dereference
@opindex -P
@opindex --no-dereference
@cindex symbolic links, copying
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.

@item -p
@itemx --preserve[=@var{attribute_list}]
@opindex -p
@opindex --preserve
@cindex file information, preserving, extended attributes, xattr
Preserve the specified attributes of the original files.
If specified, the @var{attribute_list} must be a comma-separated list
of one or more of the following strings:

@table @samp
@item mode
Preserve the file mode bits and access control lists.
@item 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.
@item 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 @code{utimensat} function,
which makes it possible even for symbolic links.
@item links
Preserve in the destination files
any links between corresponding source files.
Note that with @option{-L} or @option{-H}, this option can convert
symbolic links to hard links.  For example,
@example
$ mkdir c; : > a; ln -s a b; cp -aH a b c; ls -i1 c
74161745 a
74161745 b
@end example
@noindent
Note the inputs: @file{b} is a symlink to regular file @file{a},
yet the files in destination directory, @file{c/}, are hard-linked.
Since @option{-a} implies @option{--no-dereference} it would copy the symlink,
but the later @option{-H} tells @command{cp} to dereference the command line
arguments where it then sees two files with the same inode number.
Then the @option{--preserve=links} option also implied by @option{-a}
will preserve the perceived hard link.

Here is a similar example that exercises @command{cp}'s @option{-L} option:
@example
$ mkdir b c; (cd b; : > a; ln -s a b); cp -aL b c; ls -i1 c/b
74163295 a
74163295 b
@end example

@item context
Preserve SELinux security context of the file, or fail with full diagnostics.
@item xattr
Preserve extended attributes of the file, or fail with full diagnostics.
If @command{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 @option{--preserve=mode} or @option{--preserve=context}.
@item 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 @command{cp}'s exit status.  In contrast to @option{-a},
all but @samp{Operation not supported} warnings are output.
@end table

Using @option{--preserve} with no @var{attribute_list} is equivalent
to @option{--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.
@xref{File permissions}.

@item --no-preserve=@var{attribute_list}
@cindex file information, preserving
Do not preserve the specified attributes.  The @var{attribute_list}
has the same form as for @option{--preserve}.

@item --parents
@opindex --parents
@cindex parent directories and @command{cp}
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 @command{cp} must be the name of an existing directory.
For example, the command:

@example
cp --parents a/b/c existing_dir
@end example

@noindent
copies the file @file{a/b/c} to @file{existing_dir/a/b/c}, creating
any missing intermediate directories.

@item -R
@itemx -r
@itemx --recursive
@opindex -R
@opindex -r
@opindex --recursive
@cindex directories, copying recursively
@cindex copying directories recursively
@cindex recursively copying directories
@cindex non-directories, copying as special files
Copy directories recursively.  By default, do not follow symbolic
links in the source unless used together with the @option{--link}
(@option{-l}) option; see the @option{--archive} (@option{-a}), @option{-d},
@option{--dereference} (@option{-L}), @option{--no-dereference}
(@option{-P}), and @option{-H} options.  Special files are copied by
creating a destination file of the same type as the source; see the
@option{--copy-contents} option.  It is not portable to use
@option{-r} to copy symbolic links or special files.  On some
non-GNU systems, @option{-r} implies the equivalent of
@option{-L} and @option{--copy-contents} for historical reasons.
Also, it is not portable to use @option{-R} to copy symbolic links
unless you also specify @option{-P}, as POSIX allows
implementations that dereference symbolic links by default.

@item --reflink[=@var{when}]
@opindex --reflink[=@var{when}]
@cindex COW
@cindex clone
@cindex copy on write
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 data blocks as long as they remain unmodified.
Thus, if an I/O error affects data blocks of one of the files,
the other suffers the same fate.

The @var{when} value can be one of the following:

@table @samp
@item always
If the copy-on-write operation is not supported
then report the failure for each file and exit with a failure status.
Plain @option{--reflink} is equivalent to @option{--reflink=always}.

@item auto
If the copy-on-write operation is not supported then fall back
to the standard copy behavior.
This is the default if no @option{--reflink} option is given.

@item never
Disable copy-on-write operation and use the standard copy behavior.
@end table

This option is overridden by the @option{--link}, @option{--symbolic-link}
and @option{--attributes-only} options, thus allowing it to be used
to configure the default data copying behavior for @command{cp}.

@item --remove-destination
@opindex --remove-destination
Remove each existing destination file before attempting to open it
(contrast with @option{-f} above).

@item --sparse=@var{when}
@opindex --sparse=@var{when}
@cindex sparse files, copying
@cindex holes, copying files with
@findex read @r{system call, and holes}
A @dfn{sparse file} contains @dfn{holes}---a sequence of zero bytes that
does not occupy any file system blocks; the @samp{read} system call
reads these as zeros.  This can both save considerable space and
increase speed, since many binary files contain lots of consecutive zero
bytes.  By default, @command{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 @var{when} value can be one of the following:

@table @samp
@item 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.

@item 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,  @samp{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, @command{cp} does not even try to make it sparse.

@item never
Never make the output file sparse.
This is useful in creating a file for use with the @command{mkswap} command,
since such a file must not have any holes.
@end table

For example, with the following alias, @command{cp} will use the
minimum amount of space supported by the file system.
(Older versions of @command{cp} can also benefit from
@option{--reflink=auto} here.)

@example
alias cp='cp --sparse=always'
@end example

@optStripTrailingSlashes

@item -s
@itemx --symbolic-link
@opindex -s
@opindex --symbolic-link
@cindex symbolic links, copying with
Make symbolic links instead of copies of non-directories.  All source
file names must be absolute (starting with @samp{/}) 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.

@optBackupSuffix

@optTargetDirectory

@optNoTargetDirectory

@item -u
@itemx --update
@opindex -u
@opindex --update
@cindex newer files, copying only
Do not copy a non-directory that has an existing destination with the
same or newer modification timestamp.  If timestamps are being preserved,
the comparison is to the source timestamp truncated to the
resolutions of the destination file system and of the system calls
used to update timestamps; this avoids duplicate work if several
@samp{cp -pu} commands are executed with the same source and destination.
This option is ignored if the @option{-n} or @option{--no-clobber}
option is also specified.
Also, if @option{--preserve=links} is also specified (like with @samp{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.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Print the name of each file before copying it.

@item -x
@itemx --one-file-system
@opindex -x
@opindex --one-file-system
@cindex file systems, omitting copying to different
Skip subdirectories that are on different file systems from the one that
the copy started on.
However, mount point directories @emph{are} copied.

@macro optContext
@item -Z
@itemx --context[=@var{context}]
@opindex -Z
@opindex --context
@cindex SELinux, setting/restoring security context
@cindex security context
Without a specified @var{context}, adjust the SELinux security context according
to the system default type for destination files, similarly to the
@command{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.
@end macro
@optContext
This option is mutually exclusive with the @option{--preserve=context}
option, and overrides the @option{--preserve=all} and @option{-a} options.

@end table

@exitstatus


@node dd invocation
@section @command{dd}: Convert and copy a file

@pindex dd
@cindex converting while copying a file

@command{dd} copies input to output with a changeable I/O block size,
while optionally performing conversions on the data.  Synopses:

@example
dd [@var{operand}]@dots{}
dd @var{option}
@end example

The only options are @option{--help} and @option{--version}.
@xref{Common options}.

By default, @command{dd} copies standard input to standard output.
To copy, @command{dd} repeatedly does the following steps in order:

@enumerate
@item
Read an input block.

@item
If converting via @samp{sync}, pad as needed to meet the input block size.
Pad with spaces if converting via @samp{block} or @samp{unblock}, NUL
bytes otherwise.

@item
If @samp{bs=} is given and no conversion mentioned in steps (4) or (5)
is given, output the data as a single block and skip all remaining steps.

@item
If the @samp{swab} conversion is given, swap each pair of input bytes.
If the input data length is odd, preserve the last input byte
(since there is nothing to swap it with).

@item
If any of the conversions @samp{swab}, @samp{block}, @samp{unblock},
@samp{lcase}, @samp{ucase}, @samp{ascii}, @samp{ebcdic} and @samp{ibm}
are given, do these conversions.  These conversions operate
independently of input blocking, and might deal with records that span
block boundaries.

@item
Aggregate the resulting data into output blocks of the specified size,
and output each output block in turn.  Do not pad the last output block;
it can be shorter than usual.
@end enumerate

@command{dd} accepts the following operands,
whose syntax was inspired by the DD (data definition) statement of
OS/360 JCL.

@table @samp

@item if=@var{file}
@opindex if
Read from @var{file} instead of standard input.

@item of=@var{file}
@opindex of
Write to @var{file} instead of standard output.  Unless
@samp{conv=notrunc} is given, truncate @var{file} before writing it.

@item ibs=@var{bytes}
@opindex ibs
@cindex block size of input
@cindex input block size
Set the input block size to @var{bytes}.
This makes @command{dd} read @var{bytes} per block.
The default is 512 bytes.

@item obs=@var{bytes}
@opindex obs
@cindex block size of output
@cindex output block size
Set the output block size to @var{bytes}.
This makes @command{dd} write @var{bytes} per block.
The default is 512 bytes.

@item bs=@var{bytes}
@opindex bs
@cindex block size
Set both input and output block sizes to @var{bytes}.
This makes @command{dd} read and write @var{bytes} per block,
overriding any @samp{ibs} and @samp{obs} settings.
In addition, if no data-transforming @option{conv} operand is specified,
input is copied to the output as soon as it's read,
even if it is smaller than the block size.

@item cbs=@var{bytes}
@opindex cbs
@cindex block size of conversion
@cindex conversion block size
@cindex fixed-length records, converting to variable-length
@cindex variable-length records, converting to fixed-length
Set the conversion block size to @var{bytes}.
When converting variable-length records to fixed-length ones
(@option{conv=block}) or the reverse (@option{conv=unblock}),
use @var{bytes} as the fixed record length.

@item skip=@var{n}
@itemx iseek=@var{n}
@opindex skip
@opindex iseek
Skip @var{n} @samp{ibs}-byte blocks in the input file before copying.
If @var{n} ends in the letter @samp{B}, interpret @var{n}
as a byte count rather than a block count.
(@samp{B} and the @samp{iseek=} spelling are GNU extensions to POSIX.)

@item seek=@var{n}
@itemx oseek=@var{n}
@opindex seek
@opindex oseek
Skip @var{n} @samp{obs}-byte blocks in the output file before
truncating or copying.
If @var{n} ends in the letter @samp{B}, interpret @var{n}
as a byte count rather than a block count.
(@samp{B} and the @samp{oseek=} spelling are GNU extensions to POSIX.)

@item count=@var{n}
@opindex count
Copy @var{n} @samp{ibs}-byte blocks from the input file, instead
of everything until the end of the file.
If @var{n} ends in the letter @samp{B},
interpret @var{n} as a byte count rather than a block count;
this is a GNU extension to POSIX.
If short reads occur, as could be the case
when reading from a pipe for example, @samp{iflag=fullblock}
ensures that @samp{count=} counts complete input blocks
rather than input read operations.
As an extension to POSIX, @samp{count=0} copies zero blocks
instead of copying all blocks.

@item status=@var{level}
@opindex status
Specify the amount of information printed.
If this operand is given multiple times, the last one takes precedence.
The @var{level} value can be one of the following:

@table @samp

@item none
@opindex none @r{dd status=}
Do not print any informational or warning messages to standard error.
Error messages are output as normal.

@item noxfer
@opindex noxfer @r{dd status=}
Do not print the final transfer rate and volume statistics
that normally make up the last status line.

@item progress
@opindex progress @r{dd status=}
Print the transfer rate and volume statistics on standard error,
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.

@end table

Transfer information is normally output to standard error upon
receipt of the @samp{INFO} signal or when @command{dd} exits,
and defaults to the following form in the C locale:

@example
7287+1 records in
116608+0 records out
59703296 bytes (60 MB, 57 MiB) copied, 0.0427974 s, 1.4 GB/s
@end example

The notation @samp{@var{w}+@var{p}} stands for @var{w} whole blocks
and @var{p} partial blocks.  A partial block occurs when a read or
write operation succeeds but transfers less data than the block size.
An additional line like @samp{1 truncated record} or @samp{10
truncated records} is output after the @samp{records out} line if
@samp{conv=block} processing truncated one or more input records.

The @samp{status=} operand is a GNU extension to POSIX.

@item conv=@var{conversion}[,@var{conversion}]@dots{}
@opindex conv
Convert the file as specified by the @var{conversion} argument(s).
(No spaces around any comma(s).)

Conversions:

@table @samp

@item ascii
@opindex ascii@r{, converting to}
Convert EBCDIC to ASCII,
using the conversion table specified by POSIX@.
This provides a 1:1 translation for all 256 bytes.
This implies @samp{conv=unblock}; input is converted to
ASCII before trailing spaces are deleted.

@item ebcdic
@opindex ebcdic@r{, converting to}
Convert ASCII to EBCDIC@.
This is the inverse of the @samp{ascii} conversion.
This implies @samp{conv=block}; trailing spaces are added
before being converted to EBCDIC@.

@item ibm
@opindex alternate ebcdic@r{, converting to}
This acts like @samp{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 @samp{~}, @samp{[}, and @samp{]}.

The @samp{ascii}, @samp{ebcdic}, and @samp{ibm} conversions are
mutually exclusive.  If you use any of these conversions, you should also
use the @samp{cbs=} operand.

@item block
@opindex block @r{(space-padding)}
For each line in the input, output @samp{cbs} bytes, replacing the
input newline with a space and truncating or padding input lines with
spaces as necessary.

@item unblock
@opindex unblock
Remove any trailing spaces in each @samp{cbs}-sized input block,
and append a newline.

The @samp{block} and @samp{unblock} conversions are mutually exclusive.
If you use either of these conversions, you should also use the
@samp{cbs=} operand.

@item lcase
@opindex lcase@r{, converting to}
Change uppercase letters to lowercase.

@item ucase
@opindex ucase@r{, converting to}
Change lowercase letters to uppercase.

The @samp{lcase} and @samp{ucase} conversions are mutually exclusive.

@item sparse
@opindex 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 conversion in conjunction with
@samp{conv=notrunc} or @samp{oflag=append}.
With @samp{conv=notrunc}, existing data in the output file
corresponding to NUL blocks from the input, will be untouched.
With @samp{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 conversion
is most useful with virtual or pre zeroed devices.

The @samp{sparse} conversion is a GNU extension to POSIX.

@item swab
@opindex swab @r{(byte-swapping)}
@cindex byte-swapping
Swap every pair of input bytes.

@item sync
@opindex sync @r{(padding with ASCII NULs)}
Pad every input block to size of @samp{ibs} with trailing zero bytes.
When used with @samp{block} or @samp{unblock}, pad with spaces instead of
zero bytes.

@end table

The following ``conversions'' are really file flags
and don't affect internal processing:

@table @samp
@item excl
@opindex excl
@cindex creating output file, requiring
Fail if the output file already exists; @command{dd} must create the
output file itself.

@item nocreat
@opindex nocreat
@cindex creating output file, avoiding
Do not create the output file; the output file must already exist.

The @samp{excl} and @samp{nocreat} conversions are mutually exclusive,
and are GNU extensions to POSIX.

@item notrunc
@opindex notrunc
@cindex truncating output file, avoiding
Do not truncate the output file.

@item noerror
@opindex noerror
@cindex read errors, ignoring
Continue after read errors.

@item fdatasync
@opindex fdatasync
@cindex synchronized data writes, before finishing
Synchronize output data just before finishing,
even if there were write errors.
This forces a physical write of output data,
so that even if power is lost the output data will be preserved.
If neither this nor @samp{fsync} are specified, output is treated as
usual with file systems, i.e., output data and metadata may be cached
in primary memory for some time before the operating system physically
writes it, and thus output data and metadata may be lost if power is lost.
@xref{sync invocation}.
This conversion is a GNU extension to POSIX.

@item fsync
@opindex fsync
@cindex synchronized data and metadata writes, before finishing
Synchronize output data and metadata just before finishing,
even if there were write errors.
This acts like @samp{fdatasync} except it also preserves output metadata,
such as the last-modified time of the output file; for this reason it
may be a bit slower than @samp{fdatasync} although the performance
difference is typically insignificant for @command{dd}.
This conversion is a GNU extension to POSIX.

@end table

@item iflag=@var{flag}[,@var{flag}]@dots{}
@opindex iflag
Access the input file using the flags specified by the @var{flag}
argument(s).  (No spaces around any comma(s).)

@item oflag=@var{flag}[,@var{flag}]@dots{}
@opindex oflag
Access the output file using the flags specified by the @var{flag}
argument(s).  (No spaces around any comma(s).)

Here are the flags.

@table @samp

@item append
@opindex append
@cindex appending to the output file
Write in append mode, so that even if some other process is writing to
this file, every @command{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 @samp{of=@var{file}} operand,
you should also specify @samp{conv=notrunc} unless you want the
output file to be truncated before being appended to.

@item cio
@opindex cio
@cindex concurrent I/O
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.

@item direct
@opindex direct
@cindex direct I/O
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 @samp{oflag=direct} will cause writes to fail with @code{EINVAL} if the
output buffer size is not a multiple of 512.

@item directory
@opindex directory
@cindex directory I/O

Fail unless the file is a directory.  Most operating systems do not
allow I/O to a directory, so this flag has limited utility.

@item dsync
@opindex dsync
@cindex synchronized data reads
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.

@item sync
@opindex sync
@cindex synchronized data and metadata I/O
Use synchronized I/O for both data and metadata.

@item nocache
@opindex nocache
@cindex discarding file cache
Request to discard the system data cache for a file.
When count=0 all cached data for the file is specified,
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.

Note data that is not already persisted to storage will not
be discarded from cache, so note the use of the @samp{sync} conversions
in the examples below, which are used to maximize the
effectiveness of the @samp{nocache} flag.

Here are some usage examples:

@example
# 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

# Advise to drop cache for part of file
# Note the kernel will only consider complete and
# already persisted pages.
dd if=ifile iflag=nocache skip=10 count=10 of=/dev/null

# Stream data using just the read-ahead cache.
# See also the @samp{direct} flag.
dd if=ifile of=ofile iflag=nocache oflag=nocache,sync
@end example

@item nonblock
@opindex nonblock
@cindex nonblocking I/O
Use non-blocking I/O.

@item noatime
@opindex noatime
@cindex access timestamp
Do not update the file's access timestamp.
@xref{File timestamps}.
Some older file systems silently ignore this flag, so it is a good
idea to test it on your files before relying on it.

@item noctty
@opindex noctty
@cindex controlling terminal
Do not assign the file to be a controlling terminal for @command{dd}.
This has no effect when the file is not a terminal.
On many hosts (e.g., GNU/Linux hosts), this flag has no effect
at all.

@item nofollow
@opindex nofollow
@cindex symbolic links, following
Do not follow symbolic links.

@item nolinks
@opindex nolinks
@cindex hard links
Fail if the file has multiple hard links.

@item binary
@opindex binary
@cindex binary I/O
Use binary I/O@.  This flag has an effect only on nonstandard
platforms that distinguish binary from text I/O.

@item text
@opindex text
@cindex text I/O
Use text I/O@.  Like @samp{binary}, this flag has no effect on
standard platforms.

@item fullblock
@opindex fullblock
Accumulate full blocks from input.  The @code{read} system call
may return early if a full block is not available.
When that happens, continue calling @code{read} to fill the remainder
of the block.
This flag can be used only with @code{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 @samp{count=} argument is
interpreted as a block count rather than a count of read operations.

@end table

These flags are all GNU extensions to POSIX.
They are not supported on all systems, and @samp{dd} rejects
attempts to use them when they are not supported.  When reading from
standard input or writing to standard output, the @samp{nofollow} and
@samp{noctty} flags should not be specified, and the other flags
(e.g., @samp{nonblock}) can affect how other processes behave with the
affected file descriptors, even after @command{dd} exits.

@end table

The behavior of @command{dd} is unspecified if operands other than
@samp{conv=}, @samp{iflag=}, @samp{oflag=}, and @samp{status=} are
specified more than once.

@cindex multipliers after numbers
The numeric-valued strings above (@var{n} and @var{bytes})
are unsigned decimal integers that
can be followed by a multiplier: @samp{b}=512, @samp{c}=1,
@samp{w}=2, @samp{x@var{m}}=@var{m}, or any of the
standard block size suffixes like @samp{k}=1024 (@pxref{Block size}).
These multipliers are GNU extensions to POSIX, except that
POSIX allows @var{bytes} to be followed by @samp{k}, @samp{b}, and
@samp{x@var{m}}.
Block sizes (i.e., specified by @var{bytes} strings) must be nonzero.

Any block size you specify via @samp{bs=}, @samp{ibs=}, @samp{obs=}, @samp{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 with offset or size that is not a multiple of the I/O
block size, you can use a numeric string @var{n} that ends in the
letter @samp{B}.
For example, the following shell commands copy data
in 1 MiB blocks between a flash drive and a tape, but do not save
or restore a 512-byte area at the start of the flash drive:

@example
flash=/dev/sda
tape=/dev/st0

# Copy all but the initial 512 bytes from flash to tape.
dd if=$flash iseek=512B bs=1MiB of=$tape

# Copy from tape back to flash, leaving initial 512 bytes alone.
dd if=$tape bs=1MiB of=$flash oseek=512B
@end example

@cindex ddrescue
@cindex storage devices, failing
For failing storage devices, other tools come with a great variety of extra
functionality to ease the saving of as much data as possible before the
device finally dies, e.g.
@uref{https://www.gnu.org/software/ddrescue/, GNU @command{ddrescue}}.
However, in some cases such a tool is not available or the administrator
feels more comfortable with the handling of @command{dd}.
As a simple rescue method, call @command{dd} as shown in the following
example: the operand @samp{conv=noerror,sync} is used to continue
after read errors and to pad out bad reads with NULs, while
@samp{iflag=fullblock} caters for short reads (which traditionally never
occur on flash or similar devices):

@example
# Rescue data from an (unmounted!) partition of a failing device.
dd conv=noerror,sync iflag=fullblock </dev/sda1 > /mnt/rescue.img
@end example

Sending an @samp{INFO} signal (or @samp{USR1} signal where that is unavailable)
to a running @command{dd} process makes it print I/O statistics to
standard error and then resume copying.   In the example below,
@command{dd} is run in the background to copy 5GB of data.
The @command{kill} command makes it output intermediate I/O statistics,
and when @command{dd} completes normally or is killed by the
@code{SIGINT} signal, it outputs the final statistics.

@example
# 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 second.
while kill -s USR1 $pid 2>/dev/null; do sleep 1; done
@end example

The above script will output in the following format:

@example
3441325+0 records in
3441325+0 records out
3441325000 bytes (3.4 GB, 3.2 GiB) copied, 1.00036 s, 3.4 GB/s
5000000+0 records in
5000000+0 records out
5000000000 bytes (5.0 GB, 4.7 GiB) copied, 1.44433 s, 3.5 GB/s
@end example

The @samp{status=progress} operand periodically updates the last line
of the transfer statistics above.

@vindex POSIXLY_CORRECT
On systems lacking the @samp{INFO} signal @command{dd} responds to the
@samp{USR1} signal instead, unless the @env{POSIXLY_CORRECT}
environment variable is set.

@exitstatus


@node install invocation
@section @command{install}: Copy files and set attributes

@pindex install
@cindex copying files and setting attributes

@command{install} copies files while setting their file mode bits and, if
possible, their owner and group.  Synopses:

@example
install [@var{option}]@dots{} [-T] @var{source} @var{dest}
install [@var{option}]@dots{} @var{source}@dots{} @var{directory}
install [@var{option}]@dots{} -t @var{directory} @var{source}@dots{}
install [@var{option}]@dots{} -d @var{directory}@dots{}
@end example

@itemize @bullet
@item
If two file names are given, @command{install} copies the first file to the
second.

@item
If the @option{--target-directory} (@option{-t}) option is given, or
failing that if the last file is a directory and the
@option{--no-target-directory} (@option{-T}) option is not given,
@command{install} copies each @var{source} file to the specified
directory, using the @var{source}s' names.

@item
If the @option{--directory} (@option{-d}) option is given,
@command{install} creates each @var{directory} and any missing parent
directories.  Parent directories are created with mode
@samp{u=rwx,go=rx} (755), regardless of the @option{-m} option or the
current umask.  @xref{Directory Setuid and Setgid}, for how the
set-user-ID and set-group-ID bits of parent directories are inherited.
@end itemize

@cindex Makefiles, installing programs in
@command{install} is similar to @command{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.

@cindex extended attributes, xattr
@command{install} never preserves extended attributes (xattr).

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@optBackup

@item -C
@itemx --compare
@opindex -C
@opindex --compare
Compare content of source and destination files, and if there would be no
change to the destination content, owner, group, permissions, and possibly
SELinux context, then do not modify the destination at all.
Note this option is best used in conjunction with @option{--user},
@option{--group} and @option{--mode} options, lest @command{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.

@item -c
@opindex -c
Ignored; for compatibility with old Unix versions of @command{install}.

@item -D
@opindex -D
Create any missing parent directories of @var{dest},
then copy @var{source} to @var{dest}.
Explicitly specifying the @option{--target-directory=@var{dir}} will similarly
ensure the presence of that hierarchy before copying @var{source} arguments.

@item -d
@itemx --directory
@opindex -d
@opindex --directory
@cindex directories, creating with given attributes
@cindex parent directories, creating missing
@cindex leading directories, creating missing
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.

@item -g @var{group}
@itemx --group=@var{group}
@opindex -g
@opindex --group
@cindex group ownership of installed files, setting
Set the group ownership of installed files or directories to
@var{group}.  The default is the process's current group.  @var{group}
may be either a group name or a numeric group ID.

@item -m @var{mode}
@itemx --mode=@var{mode}
@opindex -m
@opindex --mode
@cindex permissions of installed files, setting
Set the file mode bits for the installed file or directory to @var{mode},
which can be either an octal number, or a symbolic mode as in
@command{chmod}, with @samp{a=} (no access allowed to anyone) as the
point of departure (@pxref{File permissions}).
The default mode is @samp{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 @samp{755}, since it disables
instead of preserving set-user-ID and set-group-ID on directories.
@xref{Directory Setuid and Setgid}.

@item -o @var{owner}
@itemx --owner=@var{owner}
@opindex -o
@opindex --owner
@cindex ownership of installed files, setting
@cindex appropriate privileges
@vindex root @r{as default owner}
If @command{install} has appropriate privileges (is run as root), set the
ownership of installed files or directories to @var{owner}.  The default
is @code{root}.  @var{owner} may be either a user name or a numeric user
ID.

@item --preserve-context
@opindex --preserve-context
@cindex SELinux
@cindex security 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.

@item -p
@itemx --preserve-timestamps
@opindex -p
@opindex --preserve-timestamps
@cindex timestamps of installed files, preserving
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 timestamps are both set to the time of installation.
This option is useful if you want to use the last modification timestamps
of installed files to keep track of when they were last built as opposed
to when they were last installed.

@item -s
@itemx --strip
@opindex -s
@opindex --strip
@cindex symbol table information, stripping
@cindex stripping symbol table information
Strip the symbol tables from installed binary executables.

@item --strip-program=@var{program}
@opindex --strip-program
@cindex symbol table information, stripping, program
Program used to strip binaries.

@optBackupSuffix

@optTargetDirectory
Also specifying the @option{-D} option will ensure the directory is present.

@optNoTargetDirectory

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Print the name of each file before copying it.

@optContext
This option is mutually exclusive with the @option{--preserve-context} option.


@end table

@exitstatus


@node mv invocation
@section @command{mv}: Move (rename) files

@pindex mv

@command{mv} moves or renames files (or directories).  Synopses:

@example
mv [@var{option}]@dots{} [-T] @var{source} @var{dest}
mv [@var{option}]@dots{} @var{source}@dots{} @var{directory}
mv [@var{option}]@dots{} -t @var{directory} @var{source}@dots{}
@end example

@itemize @bullet
@item
If two file names are given, @command{mv} moves the first file to the
second.

@item
If the @option{--target-directory} (@option{-t}) option is given, or
failing that if the last file is a directory and the
@option{--no-target-directory} (@option{-T}) option is not given,
@command{mv} moves each @var{source} file to the specified
directory, using the @var{source}s' names.
@end itemize

@command{mv} can move any type of file from one file system to another.
Prior to version @code{4.0} of the fileutils,
@command{mv} could move only regular files between file systems.
For example, now @command{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 @code{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.

@cindex extended attributes, xattr
@command{mv} always tries to copy extended attributes (xattr), which may
include SELinux context, ACLs or Capabilities.
Upon failure all but @samp{Operation not supported} warnings are output.

@cindex prompting, and @command{mv}
If a destination file exists but is normally unwritable, standard input
is a terminal, and the @option{-f} or @option{--force} option is not given,
@command{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.

@emph{Warning}: Avoid specifying a source name with a trailing slash,
when it might be a symlink to a directory.
Otherwise, @command{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
@code{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.
@xref{Trailing slashes}.

@emph{Note}: @command{mv} will only replace empty directories in the
destination.  Conflicting populated directories are skipped with a diagnostic.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@optBackup

@item -f
@itemx --force
@opindex -f
@opindex --force
@cindex prompts, omitting
Do not prompt the user before removing a destination file.
@macro mvOptsIfn
If you specify more than one of the @option{-i}, @option{-f}, @option{-n}
options, only the final one takes effect.
@end macro
@mvOptsIfn

@item -i
@itemx --interactive
@opindex -i
@opindex --interactive
@cindex prompts, forcing
Prompt whether to overwrite each existing destination file, regardless
of its permissions.
If the response is not affirmative, the file is skipped.
@mvOptsIfn

@item -n
@itemx --no-clobber
@opindex -n
@opindex --no-clobber
@cindex prompts, omitting
Do not overwrite an existing file; silently do nothing instead.
@mvOptsIfn
This option is mutually exclusive with @option{-b} or @option{--backup} option.

@item -u
@itemx --update
@opindex -u
@opindex --update
@cindex newer files, moving only
Do not move a non-directory that has an existing destination with the
same or newer modification timestamp.
If the move is across file system boundaries, the comparison is to the
source timestamp truncated to the resolutions of the destination file
system and of the system calls used to update timestamps; this avoids
duplicate work if several @samp{mv -u} commands are executed with the
same source and destination.
This option is ignored if the @option{-n} or @option{--no-clobber}
option is also specified.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Print the name of each file before moving it.

@optStripTrailingSlashes

@optBackupSuffix

@optTargetDirectory

@optNoTargetDirectory

@item -Z
@itemx --context
@opindex -Z
@opindex --context
@cindex SELinux, restoring security context
@cindex security context
This option functions similarly to the @command{restorecon} command,
by adjusting the SELinux security context according
to the system default type for destination files and each created directory.

@end table

@exitstatus


@node rm invocation
@section @command{rm}: Remove files or directories

@pindex rm
@cindex removing files or directories

@command{rm} removes each given @var{file}.  By default, it does not remove
directories.  Synopsis:

@example
rm [@var{option}]@dots{} [@var{file}]@dots{}
@end example

@cindex prompting, and @command{rm}
If the @option{-I} or @option{--interactive=once} option is given,
and there are more than three files or the @option{-r}, @option{-R},
or @option{--recursive} are given, then @command{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 @option{-f} or @option{--force} option is not given, or the
@option{-i} or @option{--interactive=always} option @emph{is} given,
@command{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
@file{.} or @file{..} is rejected without any prompting, as mandated
by POSIX.

@emph{Warning}: If you use @command{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 unrecoverable, consider using @command{shred}.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -d
@itemx --dir
@opindex -d
@opindex --dir
@cindex directories, removing
Remove the listed directories if they are empty.

@item -f
@itemx --force
@opindex -f
@opindex --force
Ignore nonexistent files and missing operands, and never prompt the user.
Ignore any previous @option{--interactive} (@option{-i}) option.

@item -i
@opindex -i
Prompt whether to remove each file.
If the response is not affirmative, the file is skipped.
Ignore any previous @option{--force} (@option{-f}) option.
Equivalent to @option{--interactive=always}.

@item -I
@opindex -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 @option{--force} (@option{-f}) option.  Equivalent to
@option{--interactive=once}.

@item --interactive [=@var{when}]
@opindex --interactive
Specify when to issue an interactive prompt.  @var{when} may be
omitted, or one of:
@itemize @bullet
@item never
@vindex never @r{interactive option}
- Do not prompt at all.
@item once
@vindex once @r{interactive option}
- Prompt once if more than three files are named or if a recursive
removal is requested.  Equivalent to @option{-I}.
@item always
@vindex always @r{interactive option}
- Prompt for every file being removed.  Equivalent to @option{-i}.
@end itemize
@option{--interactive} with no @var{when} is equivalent to
@option{--interactive=always}.

@item --one-file-system
@opindex --one-file-system
@cindex one file system, restricting @command{rm} to
When removing a hierarchy recursively, skip any directory that is on a
file system different from that of the corresponding command line argument.
@cindex bind mount
This option is useful when removing a build ``chroot'' hierarchy,
which normally contains no valuable data.  However, it is not uncommon
to bind-mount @file{/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 @file{/home}.  Then, when you use @command{rm -rf} to remove
your normally throw-away chroot, that command will remove everything
under @file{/home}, too.
Use the @option{--one-file-system} option, and it will
warn about and skip directories on other file systems.
Of course, this will not save your @file{/home} if it and your
chroot happen to be on the same file system.
See also @option{--preserve-root=all} to protect command line arguments
themselves.

@item --preserve-root [=all]
@opindex --preserve-root
@cindex root directory, disallow recursive destruction
Fail upon any attempt to remove the root directory, @file{/},
when used with the @option{--recursive} option.
This is the default behavior.
@xref{Treating / specially}.
When @samp{all} is specified, reject any command line argument
that is not on the same file system as its parent.

@item --no-preserve-root
@opindex --no-preserve-root
@cindex root directory, allow recursive destruction
Do not treat @file{/} specially when removing recursively.
This option is not recommended unless you really want to
remove all the files on your computer.
@xref{Treating / specially}.

@item -r
@itemx -R
@itemx --recursive
@opindex -r
@opindex -R
@opindex --recursive
@cindex directories, removing (recursively)
Remove the listed directories and their contents recursively.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Print the name of each file before removing it.

@end table

@cindex files beginning with @samp{-}, removing
@cindex @samp{-}, removing files beginning with
One common question is how to remove files whose names begin with a
@samp{-}.  GNU @command{rm}, like every program that uses the @code{getopt}
function to parse its arguments, lets you use the @samp{--} option to
indicate that all following arguments are non-options.  To remove a file
called @file{-f} in the current directory, you could type either:

@example
rm -- -f
@end example

@noindent
or:

@example
rm ./-f
@end example

@opindex - @r{and Unix @command{rm}}
The Unix @command{rm} program's use of a single @samp{-} for this purpose
predates the development of the @code{getopt} standard syntax.

@exitstatus


@node shred invocation
@section @command{shred}: Remove files more securely

@pindex shred
@cindex data, erasing
@cindex erasing data

@command{shred} overwrites devices or files, to help prevent even
extensive forensics from recovering the data.

Ordinarily when you remove a file (@pxref{rm invocation}), its data
and metadata are not actually destroyed.  Only the file's directory
entry is removed, and the file's storage is reclaimed only when no
process has the file open and no other directory entry links to the
file.  And even if file's data and metadata's storage space is freed
for further reuse, there are undelete utilities that will attempt to
reconstruct the file from the data in freed storage, and that can
bring the file back if the storage was not rewritten.

On a busy system with a nearly-full device, space can get reused in a few
seconds.  But there is no way to know for sure.  And although the
undelete utilities and already-existing processes require insider or
superuser access, you may be wary of the superuser,
of processes running on your behalf, or of attackers
that can physically access the storage device.  So if you have sensitive
data, you may want to be sure that recovery is not possible
by plausible attacks like these.

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
this is often the preferred method.  However, some storage devices
are expensive or are harder to destroy, so the @command{shred} utility tries
to achieve a similar effect non-destructively, by overwriting the file
with non-sensitive data.

@strong{Please note} that @command{shred} relies on a crucial
assumption: that the file system and hardware overwrite data in place.
Although this is common and is the traditional
way to do things, but many modern file system designs do not satisfy this
assumption.  Exceptions include:

@itemize @bullet

@item
Log-structured or journaled file systems, such as ext3/ext4 (in
@code{data=journal} mode), Btrfs, NTFS, ReiserFS, XFS, ZFS, file
systems supplied with AIX and Solaris, etc., when they are configured to
journal data.

@item
File systems that write redundant data and carry on even if some writes
fail, such as RAID-based file systems.

@item
File systems that make snapshots, such as Network Appliance's NFS server.

@item
File systems that cache in temporary locations, such as NFS version 3
clients.

@item
Compressed file systems.
@end itemize

For ext3 and ext4 file systems, @command{shred} is less effective
when the file system is in @code{data=journal}
mode, which journals file data in addition to just metadata.  In both
the @code{data=ordered} (default) and @code{data=writeback} modes,
@command{shred} works as usual.  The ext3/ext4 journaling modes can be changed
by adding the @code{data=something} option to the mount options for a
particular file system in the @file{/etc/fstab} file, as documented in
the @command{mount} man page (@samp{man mount}).  Alternatively, if
you know how large the journal is, you can shred the journal by
shredding enough file data so that the journal cycles around and fills
up with shredded data.

If you are not sure how your file system operates, then you should assume
that it does not overwrite data in place, which means @command{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 file system design issues mentioned above.
However, devices are also problematic for shredding, for reasons
such as the following:

@itemize @bullet

@item
Solid-state storage devices (SSDs) typically do wear leveling to
prolong service life, and this means writes are distributed to other
blocks by the hardware, so ``overwritten'' data blocks are still
present in the underlying device.

@item
Most storage devices map out bad blocks invisibly to
the application; if the bad blocks contain sensitive data,
@command{shred} won't be able to destroy it.

@item
With some obsolete storage technologies,
it may be possible to take (say) a floppy 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.  With these older technologies, if the file has been
overwritten only once, it's reputedly not even that hard.  Luckily,
this kind of data recovery has become difficult, and there is no
public evidence that today's higher-density storage devices can be
analyzed in this way.

The @command{shred} command can use many overwrite passes,
with data patterns chosen to
maximize the damage they do to the old data.
By default the patterns are designed for best effect on hard drives using
now-obsolete technology; for newer devices, a single pass should suffice.
For more details, see the source code and Peter Gutmann's paper
@uref{https://www.cs.auckland.ac.nz/~pgut001/pubs/secure_del.html,
@cite{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).
@end itemize

@command{shred} makes no attempt to detect or report these problems, just as
it makes no attempt to do anything about backups.  However, since it is
more reliable to shred devices than files, @command{shred} by default does
not deallocate or remove the output file.  This default is more suitable
for devices, which typically cannot be deallocated 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 @command{shred}, be sure that it is not backed up or mirrored.

@example
shred [@var{option}]@dots{} @var{file}[@dots{}]
@end example

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -f
@itemx --force
@opindex -f
@opindex --force
@cindex force deletion
Override file permissions if necessary to allow overwriting.

@item -n @var{number}
@itemx --iterations=@var{number}
@opindex -n @var{number}
@opindex --iterations=@var{number}
@cindex iterations, selecting the number of
By default, @command{shred} uses @value{SHRED_DEFAULT_PASSES} 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.

@item --random-source=@var{file}
@opindex --random-source
@cindex random source for shredding
Use @var{file} as a source of random data used to overwrite and to
choose pass ordering.  @xref{Random sources}.

@item -s @var{bytes}
@itemx --size=@var{bytes}
@opindex -s @var{bytes}
@opindex --size=@var{bytes}
@cindex size of file to shred
Shred the first @var{bytes} bytes of the file.  The default is to shred
the whole file.  @var{bytes} can be followed by a size specification like
@samp{K}, @samp{M}, or @samp{G} to specify a multiple.  @xref{Block size}.

@item -u
@itemx --remove[=@var{how}]
@opindex -u
@opindex --remove
@opindex --remove=unlink
@opindex --remove=wipe
@opindex --remove=wipesync
@cindex removing files after shredding
After shredding a file, deallocate 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 @var{how} parameter, supported with the long form option,
gives control of how to more efficiently remove each directory entry.
The @samp{unlink} parameter will just use a standard unlink call,
@samp{wipe} will also first obfuscate bytes in the name, and
@samp{wipesync} will also sync each obfuscated byte in the name to
the file system.
Note @samp{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.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Display to standard error all status updates as sterilization proceeds.

@item -x
@itemx --exact
@opindex -x
@opindex --exact
By default, @command{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 @option{--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.

@item -z
@itemx --zero
@opindex -z
@opindex --zero
Normally, the last pass that @command{shred} writes is made up of
random data.  If this would be conspicuous on your storage device (for
example, because it looks like encrypted data), or you just think
it's tidier, the @option{--zero} option adds an additional overwrite pass with
all zero bits.  This is in addition to the number of passes specified
by the @option{--iterations} option.

@end table

You might use the following command to erase the file system you
created on a USB flash drive.  This command typically takes several
minutes, depending on the drive's size and write speed.  On modern
storage devices a single pass should be adequate, and will take one
third the time of the default three-pass approach.

@example
shred -v -n 1 /dev/sdd1
@end example

Similarly, to erase all data on a selected partition of
your device, you could give a command like the following.

@example
# 1 pass, write pseudo-random data; 3x faster than the default
shred -v -n1 /dev/sda5
@end example

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 @samp{-n0 --zero},
in case some device 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 @var{file} of @samp{-} denotes standard output.
The intended use of this is to shred a removed temporary file.
For example:

@example
i=$(mktemp)
exec 3<>"$i"
rm -- "$i"
echo "Hello, world" >&3
shred - >&3
exec 3>-
@end example

However, the command @samp{shred - >file} does not shred the contents
of @var{file}, since the shell truncates @var{file} before invoking
@command{shred}.  Use the command @samp{shred file} or (if using a
Bourne-compatible shell) the command @samp{shred - 1<>file} instead.

@exitstatus


@node Special file types
@chapter Special file types

@cindex special file types
@cindex file types, special

This chapter describes commands which create special types of files (and
@command{rmdir}, which removes directories, one special file type).

@cindex special file types
@cindex file types
Although Unix-like operating systems have markedly fewer special file
types than others, not @emph{everything} can be treated only as the
undifferentiated byte stream of @dfn{normal files}.  For example, when a
file is created or removed, the system must record this information,
which it does in a @dfn{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 @dfn{special files}.

@menu
* link invocation::             Make a hard link via the link syscall
* ln invocation::               Make links between files.
* mkdir invocation::            Make directories.
* mkfifo invocation::           Make FIFOs (named pipes).
* mknod invocation::            Make block or character special files.
* readlink invocation::         Print value of a symlink or canonical file name.
* rmdir invocation::            Remove empty directories.
* unlink invocation::           Remove files via the unlink syscall
@end menu


@node link invocation
@section @command{link}: Make a hard link via the link syscall

@pindex link
@cindex links, creating
@cindex hard links, creating
@cindex creating links (hard only)

@command{link} creates a single hard link at a time.
It is a minimalist interface to the system-provided
@code{link} function.  @xref{Hard Links, , , libc,
The GNU C Library Reference Manual}.
It avoids the bells and whistles of the more commonly-used
@command{ln} command (@pxref{ln invocation}).
Synopsis:

@example
link @var{filename} @var{linkname}
@end example

@var{filename} must specify an existing file, and @var{linkname}
must specify a nonexistent entry in an existing directory.
@command{link} simply calls @code{link (@var{filename}, @var{linkname})}
to create the link.

On a GNU system, this command acts like @samp{ln --directory
--no-target-directory @var{filename} @var{linkname}}.  However, the
@option{--directory} and @option{--no-target-directory} options are
not specified by POSIX, and the @command{link} command is
more portable in practice.

If @var{filename} is a symbolic link, it is unspecified whether
@var{linkname} will be a hard link to the symbolic link or to the
target of the symbolic link.  Use @command{ln -P} or @command{ln -L}
to specify which behavior is desired.

@exitstatus


@node ln invocation
@section @command{ln}: Make links between files

@pindex ln
@cindex links, creating
@cindex hard links, creating
@cindex symbolic (soft) links, creating
@cindex creating links (hard or soft)

@cindex file systems and hard links
@command{ln} makes links between files.  By default, it makes hard links;
with the @option{-s} option, it makes symbolic (or @dfn{soft}) links.
Synopses:

@example
ln [@var{option}]@dots{} [-T] @var{target} @var{linkname}
ln [@var{option}]@dots{} @var{target}
ln [@var{option}]@dots{} @var{target}@dots{} @var{directory}
ln [@var{option}]@dots{} -t @var{directory} @var{target}@dots{}
@end example

@itemize @bullet

@item
If two file names are given, @command{ln} creates a link to the first
file from the second.

@item
If one @var{target} is given, @command{ln} creates a link to that file
in the current directory.

@item
If the @option{--target-directory} (@option{-t}) option is given, or
failing that if the last file is a directory and the
@option{--no-target-directory} (@option{-T}) option is not given,
@command{ln} creates a link to each @var{target} file in the specified
directory, using the @var{target}s' names.

@end itemize

Normally @command{ln} does not replace existing files.  Use the
@option{--force} (@option{-f}) option to replace them unconditionally,
the @option{--interactive} (@option{-i}) option to replace them
conditionally, and the @option{--backup} (@option{-b}) option to
rename them.  Unless the @option{--backup} (@option{-b}) option is
used there is no brief moment when the destination does not exist;
this is an extension to POSIX.

@cindex hard link, defined
@cindex inode, and hard links
A @dfn{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 @emph{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.)

@cindex dereferencing symbolic links
@cindex symbolic link, defined
@dfn{Symbolic links} (@dfn{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 @dfn{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.  @xref{Symbolic Links,,,
libc, The GNU C Library Reference Manual}.

Symbolic links can contain arbitrary strings; a @dfn{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 @ref{Common options}.

@table @samp

@optBackup

@item -d
@itemx -F
@itemx --directory
@opindex -d
@opindex -F
@opindex --directory
@cindex hard links to directories
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.

@item -f
@itemx --force
@opindex -f
@opindex --force
Remove existing destination files.

@item -i
@itemx --interactive
@opindex -i
@opindex --interactive
@cindex prompting, and @command{ln}
Prompt whether to remove existing destination files.

@item -L
@itemx --logical
@opindex -L
@opindex --logical
If @option{-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.

@item -n
@itemx --no-dereference
@opindex -n
@opindex --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.  @command{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, @command{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 @option{--no-target-directory}
(@option{-T}) option, so it has no effect if both options are given.

@item -P
@itemx --physical
@opindex -P
@opindex --physical
If @option{-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.

@item -r
@itemx --relative
@opindex -r
@opindex --relative
Make symbolic links relative to the link location.
This option is only valid with the @option{--symbolic} option.

Example:

@example
ln -srv /a/file /tmp
'/tmp/file' -> '../a/file'
@end example

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.
@xref{realpath invocation}, which gives greater control
over relative file name generation, as demonstrated in the following example:

@example
@verbatim
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"
}
@end verbatim
@end example

@item -s
@itemx --symbolic
@opindex -s
@opindex --symbolic
Make symbolic links instead of hard links.  This option merely produces
an error message on systems that do not support symbolic links.

@optBackupSuffix

@optTargetDirectory

@optNoTargetDirectory

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Print the name of each file after linking it successfully.

@end table

@cindex hard links to symbolic links
@cindex symbolic links and @command{ln}
If @option{-L} and @option{-P} are both given, the last one takes
precedence.  If @option{-s} is also given, @option{-L} and @option{-P}
are silently ignored.  If neither option is given, then this
implementation defaults to @option{-P} if the system @code{link} supports
hard links to symbolic links (such as the GNU system), and @option{-L}
if @code{link} follows symbolic links (such as on BSD).

@exitstatus

Examples:

@example
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
@end example


@node mkdir invocation
@section @command{mkdir}: Make directories

@pindex mkdir
@cindex directories, creating
@cindex creating directories

@command{mkdir} creates directories with the specified names.  Synopsis:

@example
mkdir [@var{option}]@dots{} @var{name}@dots{}
@end example

@command{mkdir} creates each directory @var{name} in the order given.
It reports an error if @var{name} already exists, unless the
@option{-p} option is given and @var{name} is a directory.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -m @var{mode}
@itemx --mode=@var{mode}
@opindex -m
@opindex --mode
@cindex modes of created directories, setting
Set the file permission bits of created directories to @var{mode},
which uses the same syntax as
in @command{chmod} and uses @samp{a=rwx} (read, write and execute allowed for
everyone) for the point of the departure.  @xref{File permissions}.
This option affects only directories given on the command line;
it does not affect any parents that may be created via the @option{-p} option.

Normally the directory has the desired file mode bits at the moment it
is created.  As a GNU extension, @var{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.  @xref{Directory Setuid and Setgid}, for how the
set-user-ID and set-group-ID bits of directories are inherited unless
overridden in this way.

@item -p
@itemx --parents
@opindex -p
@opindex --parents
@cindex parent directories, creating
Make any missing parent directories for each argument, setting their
file permission bits to @samp{=rwx,u+wx},
that is, with the umask modified by @samp{u+wx}.  Ignore
existing parent directories, and do not change their file permission
bits.

If the @option{-m} option is also given, it does not affect
file permission bits of any newly-created parent directories.
To control these bits, set the
umask before invoking @command{mkdir}.  For example, if the shell
command @samp{(umask u=rwx,go=rx; mkdir -p P/Q)} creates the parent
@file{P} it sets the parent's file permission bits to @samp{u=rwx,go=rx}.
(The umask must include @samp{u=wx} for this method to work.)
To set a parent's special mode bits as well, you can invoke
@command{chmod} after @command{mkdir}.  @xref{Directory Setuid and
Setgid}, for how the set-user-ID and set-group-ID bits of
newly-created parent directories are inherited.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Print a message for each created directory.  This is most useful with
@option{--parents}.

@optContext

@end table

@exitstatus


@node mkfifo invocation
@section @command{mkfifo}: Make FIFOs (named pipes)

@pindex mkfifo
@cindex FIFOs, creating
@cindex named pipes, creating
@cindex creating FIFOs (named pipes)

@command{mkfifo} creates FIFOs (also called @dfn{named pipes}) with the
specified names.  Synopsis:

@example
mkfifo [@var{option}] @var{name}@dots{}
@end example

A @dfn{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 @ref{Common options}.

@table @samp

@item -m @var{mode}
@itemx --mode=@var{mode}
@opindex -m
@opindex --mode
@cindex modes of created FIFOs, setting
Set the mode of created FIFOs to @var{mode}, which is symbolic as in
@command{chmod} and uses @samp{a=rw} (read and write allowed for everyone)
for the point of departure.  @var{mode} should specify only file
permission bits.  @xref{File permissions}.

@optContext

@end table

@exitstatus


@node mknod invocation
@section @command{mknod}: Make block or character special files

@pindex mknod
@cindex block special files, creating
@cindex character special files, creating

@command{mknod} creates a FIFO, character special file, or block special
file with the specified name.  Synopsis:

@example
mknod [@var{option}]@dots{} @var{name} @var{type} [@var{major} @var{minor}]
@end example

@cindex special files
@cindex block special files
@cindex character special files
Unlike the phrase ``special file type'' above, the term @dfn{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 flash drive.  (These files are typically created at
system-configuration time.)  The @command{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
@dfn{block special} files and @dfn{character special} files.

@c mknod is a shell built-in at least with OpenBSD's /bin/sh
@mayConflictWithShellBuiltIn{mknod}

The arguments after @var{name} specify the type of file to make:

@table @samp

@item p
@opindex p @r{for FIFO file}
for a FIFO

@item b
@opindex b @r{for block special file}
for a block special file

@item c
@c Don't document the 'u' option -- it's just a synonym for 'c'.
@c Do *any* versions of mknod still use it?
@c @itemx u
@opindex c @r{for character special file}
@c @opindex u @r{for character special file}
for a character special file

@end table

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 @samp{0x} or @samp{0X},
it is interpreted as hexadecimal; otherwise, if it begins with @samp{0},
as octal; otherwise, as decimal.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -m @var{mode}
@itemx --mode=@var{mode}
@opindex -m
@opindex --mode
Set the mode of created files to @var{mode}, which is symbolic as in
@command{chmod} and uses @samp{a=rw} as the point of departure.
@var{mode} should specify only file permission bits.
@xref{File permissions}.

@optContext

@end table

@exitstatus


@node readlink invocation
@section @command{readlink}: Print value of a symlink or canonical file name

@pindex readlink
@cindex displaying value of a symbolic link
@cindex canonical file name
@cindex canonicalize a file name
@cindex realpath

@command{readlink} may work in one of two supported modes:

@table @samp

@item Readlink mode

@command{readlink} outputs the value of the given symbolic links.
If @command{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.

@item Canonicalize mode

@command{readlink} outputs the absolute name of the given files which contain
no @file{.}, @file{..} components nor any repeated separators
(@file{/}) or symbolic links.  Note the @command{realpath} command is the
preferred command to use for canonicalization.  @xref{realpath invocation}.

@end table

@example
readlink [@var{option}]@dots{} @var{file}@dots{}
@end example

By default, @command{readlink} operates in readlink mode.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -f
@itemx --canonicalize
@opindex -f
@opindex --canonicalize
Activate canonicalize mode.
If any component of the file name except the last one is missing or unavailable,
@command{readlink} produces no output and exits with a nonzero exit
code.  A trailing slash is ignored.

@item -e
@itemx --canonicalize-existing
@opindex -e
@opindex --canonicalize-existing
Activate canonicalize mode.
If any component is missing or unavailable, @command{readlink} produces
no output and exits with a nonzero exit code.  A trailing slash
requires that the name resolve to a directory.

@item -m
@itemx --canonicalize-missing
@opindex -m
@opindex --canonicalize-missing
Activate canonicalize mode.
If any component is missing or unavailable, @command{readlink} treats it
as a directory.

@item -n
@itemx --no-newline
@opindex -n
@opindex --no-newline
Do not print the output delimiter, when a single @var{file} is specified.
Print a warning if specified along with multiple @var{file}s.

@item -s
@itemx -q
@itemx --silent
@itemx --quiet
@opindex -s
@opindex -q
@opindex --silent
@opindex --quiet
Suppress most error messages.  On by default.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Report error messages.

@optZero

@end table

The @command{readlink} utility first appeared in OpenBSD 2.1.

The @command{realpath} command without options, operates like
@command{readlink} in canonicalize mode.

@exitstatus


@node rmdir invocation
@section @command{rmdir}: Remove empty directories

@pindex rmdir
@cindex removing empty directories
@cindex directories, removing empty

@command{rmdir} removes empty directories.  Synopsis:

@example
rmdir [@var{option}]@dots{} @var{directory}@dots{}
@end example

If any @var{directory} argument does not refer to an existing empty
directory, it is an error.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item --ignore-fail-on-non-empty
@opindex --ignore-fail-on-non-empty
@cindex directory deletion, ignoring failures
Ignore each failure to remove a directory that is non-empty.

@item -p
@itemx --parents
@opindex -p
@opindex --parents
@cindex parent directories, removing
Remove @var{directory}, then try to remove each component of @var{directory}.
So, for example, @samp{rmdir -p a/b/c} is similar to @samp{rmdir a/b/c a/b a}.
As such, it fails if any of those directories turns out not to be empty.
Use the @option{--ignore-fail-on-non-empty} option to make it so such
a failure does not evoke a diagnostic and does not cause @command{rmdir} to
exit unsuccessfully.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
@cindex directory deletion, reporting
Give a diagnostic for each successful removal.
@var{directory} is removed.

@end table

@xref{rm invocation}, for how to remove non-empty directories recursively.

To remove all empty directories under @var{dirname}, including
directories that become empty because other directories are removed,
you can use either of the following commands:

@example
# This uses GNU extensions.
find @var{dirname} -type d -empty -delete

# This runs on any POSIX platform.
find @var{dirname} -depth -type d -exec rmdir @{@} +
@end example

@exitstatus


@node unlink invocation
@section @command{unlink}: Remove files via the unlink syscall

@pindex unlink
@cindex removing files or directories (via the unlink syscall)

@command{unlink} deletes a single specified file name.
It is a minimalist interface to the system-provided
@code{unlink} function.  @xref{Deleting Files, , , libc,
The GNU C Library Reference Manual}.  Synopsis:
It avoids the bells and whistles of the more commonly-used
@command{rm} command (@pxref{rm invocation}).

@example
unlink @var{filename}
@end example

On some systems @code{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 @code{unlink} can never delete the name of a directory.

The @command{unlink} command honors the @option{--help} and
@option{--version} options.  To remove a file whose name begins with
@samp{-}, prefix the name with @samp{./}, e.g., @samp{unlink ./--help}.

@exitstatus


@node Changing file attributes
@chapter Changing file attributes

@cindex changing file attributes
@cindex file attributes, changing
@cindex attributes, file

A file is not merely its contents, a name, and a file type
(@pxref{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
@dfn{attributes}.

These commands change file attributes.

@menu
* chown invocation::            Change file owners and groups.
* chgrp invocation::            Change file groups.
* chmod invocation::            Change access permissions.
* touch invocation::            Change file timestamps.
@end menu


@node chown invocation
@section @command{chown}: Change file owner and group

@pindex chown
@cindex file ownership, changing
@cindex group ownership, changing
@cindex changing file ownership
@cindex changing group ownership

@command{chown} changes the user and/or group ownership of each given @var{file}
to @var{new-owner} or to the user and group of an existing reference file.
Synopsis:

@example
chown [@var{option}]@dots{} @{@var{new-owner} | --reference=@var{ref_file}@}@c
 @var{file}@dots{}
@end example

If used, @var{new-owner} specifies the new owner and/or group as follows
(with no embedded white space):

@example
[@var{owner}] [ : [@var{group}] ]
@end example

Specifically:

@table @var
@item owner
If only an @var{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.

@item owner@samp{:}group
If the @var{owner} is followed by a colon and a @var{group} (a
group name or numeric group ID), with no spaces between them, the group
ownership of the files is changed as well (to @var{group}).

@item owner@samp{:}
If a colon but no group name follows @var{owner}, that user is
made the owner of the files and the group of the files is changed to
@var{owner}'s login group.

@item @samp{:}group
If the colon and following @var{group} are given, but the owner
is omitted, only the group of the files is changed; in this case,
@command{chown} performs the same function as @command{chgrp}.

@item @samp{:}
If only a colon is given, or if @var{new-owner} is empty, neither the
owner nor the group is changed.

@end table

If @var{owner} or @var{group} is intended to represent a numeric user
or group ID, then you may specify it with a leading @samp{+}.
@xref{Disambiguating names and IDs}.

Some older scripts may still use @samp{.} in place of the @samp{:} separator.
POSIX 1003.1-2001 (@pxref{Standards conformance}) does not
require support for that, but for backward compatibility GNU
@command{chown} supports @samp{.} so long as no ambiguity results,
although it issues a warning and support may be removed in future versions.
New scripts should avoid the use of @samp{.} because it is not
portable, and because it has undesirable results if the entire
@var{owner@samp{.}group} happens to identify a user whose name
contains @samp{.}.

@macro chownGroupRestrictions
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.
@end macro
@chownGroupRestrictions

The @command{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 @code{chown} system call, which may
make system-dependent file mode modifications outside the control of
the @command{chown} command.  For example, the @command{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 @ref{Common options}.

@table @samp

@item -c
@itemx --changes
@opindex -c
@opindex --changes
@cindex changed owners, verbosely describing
Verbosely describe the action for each @var{file} whose ownership
actually changes.

@item -f
@itemx --silent
@itemx --quiet
@opindex -f
@opindex --silent
@opindex --quiet
@cindex error messages, omitting
Do not print error messages about files whose ownership cannot be
changed.

@item --from=@var{old-owner}
@opindex --from
@cindex symbolic links, changing owner
Change a @var{file}'s ownership only if it has current attributes specified
by @var{old-owner}.  @var{old-owner} has the same form as @var{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, @code{root} might run

@example
find / -owner OLDUSER -print0 | xargs -0 chown -h NEWUSER
@end example

But that is dangerous because the interval between when the @command{find}
tests the existing file's owner and when the @command{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:

@example
find / -owner OLDUSER -exec chown -h NEWUSER @{@} \;
@end example

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:

@example
chown -h -R --from=OLDUSER NEWUSER /
@end example

@item --dereference
@opindex --dereference
@cindex symbolic links, changing owner
@findex lchown
Do not act on symbolic links themselves but rather on what they point to.
This is the default when not operating recursively.
@warnOptDerefWithRec

@item -h
@itemx --no-dereference
@opindex -h
@opindex --no-dereference
@cindex symbolic links, changing owner
@findex lchown
Act on symbolic links themselves instead of what they point to.
This mode relies on the @code{lchown} system call.
On systems that do not provide the @code{lchown} system call,
@command{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 @option{--verbose}.

@item --preserve-root
@opindex --preserve-root
@cindex root directory, disallow recursive modification
Fail upon any attempt to recursively change the root directory, @file{/}.
Without @option{--recursive}, this option has no effect.
@xref{Treating / specially}.

@item --no-preserve-root
@opindex --no-preserve-root
@cindex root directory, allow recursive modification
Cancel the effect of any preceding @option{--preserve-root} option.
@xref{Treating / specially}.

@item --reference=@var{ref_file}
@opindex --reference
Change the user and group of each @var{file} to be the same as those of
@var{ref_file}.  If @var{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.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Output a diagnostic for every file processed.
If a symbolic link is encountered during a recursive traversal
on a system without the @code{lchown} system call, and @option{--no-dereference}
is in effect, then issue a diagnostic saying neither the symbolic link nor
its referent is being changed.

@item -R
@itemx --recursive
@opindex -R
@opindex --recursive
@cindex recursively changing file ownership
Recursively change ownership of directories and their contents.

@choptH
@xref{Traversing symlinks}.

@choptL
@warnOptDerefWithRec
@xref{Traversing symlinks}.

@choptP
@xref{Traversing symlinks}.

@end table

@exitstatus

Examples:

@example
# 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
@end example


@node chgrp invocation
@section @command{chgrp}: Change group ownership

@pindex chgrp
@cindex group ownership, changing
@cindex changing group ownership

@command{chgrp} changes the group ownership of each given @var{file}
to @var{group} (which can be either a group name or a numeric group ID)
or to the group of an existing reference file.  @xref{chown invocation}.
Synopsis:

@example
chgrp [@var{option}]@dots{} @{@var{group} | --reference=@var{ref_file}@}@c
 @var{file}@dots{}
@end example

If @var{group} is intended to represent a
numeric group ID, then you may specify it with a leading @samp{+}.
@xref{Disambiguating names and IDs}.

@chownGroupRestrictions

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -c
@itemx --changes
@opindex -c
@opindex --changes
@cindex changed files, verbosely describing
Verbosely describe the action for each @var{file} whose group actually
changes.

@item -f
@itemx --silent
@itemx --quiet
@opindex -f
@opindex --silent
@opindex --quiet
@cindex error messages, omitting
Do not print error messages about files whose group cannot be
changed.

@item --dereference
@opindex --dereference
@cindex symbolic links, changing owner
@findex lchown
Do not act on symbolic links themselves but rather on what they point to.
This is the default when not operating recursively.
@warnOptDerefWithRec

@item -h
@itemx --no-dereference
@opindex -h
@opindex --no-dereference
@cindex symbolic links, changing group
@findex lchown
Act on symbolic links themselves instead of what they point to.
This mode relies on the @code{lchown} system call.
On systems that do not provide the @code{lchown} system call,
@command{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 @option{--verbose}.

@item --preserve-root
@opindex --preserve-root
@cindex root directory, disallow recursive modification
Fail upon any attempt to recursively change the root directory, @file{/}.
Without @option{--recursive}, this option has no effect.
@xref{Treating / specially}.

@item --no-preserve-root
@opindex --no-preserve-root
@cindex root directory, allow recursive modification
Cancel the effect of any preceding @option{--preserve-root} option.
@xref{Treating / specially}.

@item --reference=@var{ref_file}
@opindex --reference
Change the group of each @var{file} to be the same as that of
@var{ref_file}.  If @var{ref_file} is a symbolic link, do not use the
group of the symbolic link, but rather that of the file it refers to.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Output a diagnostic for every file processed.
If a symbolic link is encountered during a recursive traversal
on a system without the @code{lchown} system call, and @option{--no-dereference}
is in effect, then issue a diagnostic saying neither the symbolic link nor
its referent is being changed.

@item -R
@itemx --recursive
@opindex -R
@opindex --recursive
@cindex recursively changing group ownership
Recursively change the group ownership of directories and their contents.

@choptH
@xref{Traversing symlinks}.

@choptL
@warnOptDerefWithRec
@xref{Traversing symlinks}.

@choptP
@xref{Traversing symlinks}.

@end table

@exitstatus

Examples:

@example
# Change the group of /u to "staff".
chgrp staff /u

# Change the group of /u and subfiles to "staff".
chgrp -hR staff /u
@end example


@node chmod invocation
@section @command{chmod}: Change access permissions

@pindex chmod
@cindex changing access permissions
@cindex access permissions, changing
@cindex permissions, changing access

@command{chmod} changes the access permissions of the named files.  Synopsis:

@example
chmod [@var{option}]@dots{} @{@var{mode} | --reference=@var{ref_file}@}@c
 @var{file}@dots{}
@end example

@cindex symbolic links, permissions of
@command{chmod} never changes the permissions of symbolic links, since
the @command{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, @command{chmod} changes the permissions of the pointed-to file.
In contrast, @command{chmod} ignores symbolic links encountered during
recursive directory traversals.

Only a process whose effective user ID matches the user ID of the file,
or a process with appropriate privileges, is permitted to change the
file mode bits of a file.

A successful use of @command{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 @var{mode} or
@var{ref_file} to be ignored.  This behavior depends on the policy and
functionality of the underlying @code{chmod} system call.  When in
doubt, check the underlying system behavior.

If used, @var{mode} specifies the new file mode bits.
For details, see the section on @ref{File permissions}.
If you really want @var{mode} to have a leading @samp{-}, you should
use @option{--} first, e.g., @samp{chmod -- -w file}.  Typically,
though, @samp{chmod a-w file} is preferable, and @command{chmod -w
file} (without the @option{--}) complains if it behaves differently
from what @samp{chmod a-w file} would do.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -c
@itemx --changes
@opindex -c
@opindex --changes
Verbosely describe the action for each @var{file} whose permissions
actually change.

@item -f
@itemx --silent
@itemx --quiet
@opindex -f
@opindex --silent
@opindex --quiet
@cindex error messages, omitting
Do not print error messages about files whose permissions cannot be
changed.

@item --preserve-root
@opindex --preserve-root
@cindex root directory, disallow recursive modification
Fail upon any attempt to recursively change the root directory, @file{/}.
Without @option{--recursive}, this option has no effect.
@xref{Treating / specially}.

@item --no-preserve-root
@opindex --no-preserve-root
@cindex root directory, allow recursive modification
Cancel the effect of any preceding @option{--preserve-root} option.
@xref{Treating / specially}.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Verbosely describe the action or non-action taken for every @var{file}.

@item --reference=@var{ref_file}
@opindex --reference
Change the mode of each @var{file} to be the same as that of @var{ref_file}.
@xref{File permissions}.
If @var{ref_file} is a symbolic link, do not use the mode
of the symbolic link, but rather that of the file it refers to.

@item -R
@itemx --recursive
@opindex -R
@opindex --recursive
@cindex recursively changing access permissions
Recursively change permissions of directories and their contents.

@end table

@exitstatus

Examples:

@smallexample
# Change file permissions of FOO to be world readable
# and user writable, with no other permissions.
chmod 644 foo
chmod a=r,u+w foo

# Add user and group execute permissions to FOO.
chmod +110 file
chmod ug+x file

# Set file permissions of DIR and subsidiary files to
# be the umask default, assuming execute permissions for
# directories and for files already executable.
chmod -R a=,+rwX dir
@end smallexample


@node touch invocation
@section @command{touch}: Change file timestamps

@pindex touch
@cindex changing file timestamps
@cindex file timestamps, changing
@cindex timestamps, changing file

@command{touch} changes the access and/or modification timestamps of the
specified files.  Synopsis:

@example
touch [@var{option}]@dots{} @var{file}@dots{}
@end example

@cindex empty files, creating
Any @var{file} argument that does not exist is created empty, unless
option @option{--no-create} (@option{-c}) or @option{--no-dereference}
(@option{-h}) was in effect.

A @var{file} argument string of @samp{-} is handled specially and
causes @command{touch} to change the times of the file associated with
standard output.

By default, @command{touch} sets file timestamps to the current time.
Because @command{touch} acts on its operands left to right, the
resulting timestamps of earlier and later operands may disagree.

@cindex permissions, for changing file timestamps
When setting file timestamps to the current time, @command{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 timestamps are being set to the
current time.

The @command{touch} command cannot set a file's status change timestamp to
a user-specified value, and cannot change the file's birth time (if
supported) at all.  Also, @command{touch} has issues similar to those
affecting all programs that update file timestamps.  For example,
@command{touch} may set a file's timestamp to a value that differs
slightly from the requested time.  @xref{File timestamps}.

@vindex TZ
Timestamps assume the time zone rules specified by the @env{TZ}
environment variable, or by the system default rules if @env{TZ} is
not set.  @xref{TZ Variable,, Specifying the Time Zone with @env{TZ},
libc, The GNU C Library Reference Manual}.
You can avoid ambiguities during
daylight saving transitions by using UTC timestamps.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -a
@itemx --time=atime
@itemx --time=access
@itemx --time=use
@opindex -a
@opindex --time
@opindex atime@r{, changing}
@opindex access @r{time, changing}
@opindex use @r{time, changing}
Change the access timestamp only.  @xref{File timestamps}.

@item -c
@itemx --no-create
@opindex -c
@opindex --no-create
Do not warn about or create files that do not exist.

@item -d @var{time}
@itemx --date=@var{time}
@opindex -d
@opindex --date
@opindex time
Use @var{time} instead of the current time.  It can contain month names,
time zones, @samp{am} and @samp{pm}, @samp{yesterday}, etc.  For
example, @option{--date="2020-07-21 14:19:13.489392193 +0530"}
specifies the instant of time that is 489,392,193 nanoseconds after
July 21, 2020 at 2:19:13 PM in a time zone that is 5 hours and 30
minutes east of UTC@.  @xref{Date input formats}.
File systems that do not support high-resolution timestamps
silently ignore any excess precision here.

@item -f
@opindex -f
@cindex BSD @command{touch} compatibility
Ignored; for compatibility with BSD versions of @command{touch}.

@item -h
@itemx --no-dereference
@opindex -h
@opindex --no-dereference
@cindex symbolic links, changing time
@findex lutimes
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 @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
timestamp, such that only changes to the modification timestamp will persist
long enough to be observable.  When coupled with option @option{-r}, a
reference timestamp is taken from a symbolic link rather than the file
it refers to.

@item -m
@itemx --time=mtime
@itemx --time=modify
@opindex -m
@opindex --time
@opindex mtime@r{, changing}
@opindex modify @r{time, changing}
Change the modification timestamp only.

@item -r @var{file}
@itemx --reference=@var{file}
@opindex -r
@opindex --reference
Use the times of the reference @var{file} instead of the current time.
If this option is combined with the @option{--date=@var{time}}
(@option{-d @var{time}}) option, the reference @var{file}'s time is
the origin for any relative @var{time}s given, but is otherwise ignored.
For example, @samp{-r foo -d '-5 seconds'} specifies a timestamp
equal to five seconds before the corresponding timestamp for @file{foo}.
If @var{file} is a symbolic link, the reference timestamp is taken
from the target of the symlink, unless @option{-h} was also in effect.

@item -t [[@var{cc}]@var{yy}]@var{mmddhhmm}[.@var{ss}]
@cindex leap seconds
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 @var{cc}
is 20 for years in the range 0 @dots{} 68, and 19 for years in
69 @dots{} 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, @var{ss} may be
@samp{60}.

@end table

@vindex _POSIX2_VERSION
On systems predating POSIX 1003.1-2001,
@command{touch} supports an obsolete syntax, as follows.
If no timestamp is given with any of the @option{-d}, @option{-r}, or
@option{-t} options, and if there are two or more @var{file}s and the
first @var{file} is of the form @samp{@var{mmddhhmm}[@var{yy}]} and this
would be a valid argument to the @option{-t} option (if the @var{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.
Although this obsolete behavior can be controlled with the
@env{_POSIX2_VERSION} environment variable (@pxref{Standards
conformance}), portable scripts should avoid commands whose
behavior depends on this variable.
For example, use @samp{touch ./12312359 main.c} or @samp{touch -t
12312359 main.c} rather than the ambiguous @samp{touch 12312359 main.c}.

@exitstatus


@node File space usage
@chapter File space usage

@cindex File space usage
@cindex disk usage

No file system can hold an infinite amount of data.  These commands report
how much storage is in use or available, report other file and
file status information, and write buffers to file systems.

@menu
* df invocation::               Report file system space usage.
* du invocation::               Estimate file space usage.
* stat invocation::             Report file or file system status.
* sync invocation::             Synchronize cached writes to persistent storage.
* truncate invocation::         Shrink or extend the size of a file.
@end menu


@node df invocation
@section @command{df}: Report file system space usage

@pindex df
@cindex file system usage
@cindex disk usage by file system

@command{df} reports the amount of space used and available on
file systems.  Synopsis:

@example
df [@var{option}]@dots{} [@var{file}]@dots{}
@end example

With no arguments, @command{df} reports the space used and available on all
currently mounted file systems (of all types).  Otherwise, @command{df}
reports on the file system containing each argument @var{file}.

Normally the space is printed in units of
1024 bytes, but this can be overridden (@pxref{Block size}).
Non-integer quantities are rounded up to the next higher unit.

For bind mounts and without arguments, @command{df} only outputs the statistics
for that device with the shortest mount point name in the list of file systems
(@var{mtab}), i.e., it hides duplicate entries, unless the @option{-a} option is
specified.

With the same logic, @command{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 @samp{rootfs} is
not shown per default when already the real root device has been mounted.

@cindex disk device file
@cindex device file
If an argument @var{file} resolves to a special file containing
a mounted file system, @command{df} shows the space available on that
file system rather than on the file system containing the device node.
GNU @command{df} does not attempt to determine the 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 @ref{Common options}.

@table @samp

@item -a
@itemx --all
@opindex -a
@opindex --all
@cindex ignore file systems
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 @samp{/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.

@item -B @var{size}
@itemx --block-size=@var{size}
@opindex -B
@opindex --block-size
@cindex file system sizes
Scale sizes by @var{size} before printing them (@pxref{Block size}).
For example, @option{-BG} prints sizes in units of 1,073,741,824 bytes.

@optHumanReadable

@item -H
@opindex -H
Equivalent to @option{--si}.

@item -i
@itemx --inodes
@opindex -i
@opindex --inodes
@cindex inode usage
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 file system.

@item -k
@opindex -k
@cindex kibibytes for file system sizes
Print sizes in 1024-byte blocks, overriding the default block size
(@pxref{Block size}).
This option is equivalent to @option{--block-size=1K}.

@item -l
@itemx --local
@opindex -l
@opindex --local
@cindex file system types, limiting output to certain
Limit the listing to local file systems.  By default, remote file systems
are also listed.

@item --no-sync
@opindex --no-sync
@cindex file system space, retrieving old data more quickly
Do not invoke the @code{sync} system call before getting any usage data.
This may make @command{df} run significantly faster on systems with many
file systems, but on some systems (notably Solaris) the results may be slightly
out of date.  This is the default.

@item --output
@itemx --output[=@var{field_list}]
@opindex --output
Use the output format defined by @var{field_list}, or print all fields if
@var{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 @option{--output} together with each of the options @option{-i},
@option{-P}, and @option{-T} is mutually exclusive.

FIELD_LIST is a comma-separated list of columns to be included in @command{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 @var{field_list} are:
@table @samp
@item source
The source of the mount point, usually a device.
@item fstype
File system type.

@item itotal
Total number of inodes.
@item iused
Number of used inodes.
@item iavail
Number of available inodes.
@item ipcent
Percentage of @var{iused} divided by @var{itotal}.

@item size
Total number of blocks.
@item used
Number of used blocks.
@item avail
Number of available blocks.
@item pcent
Percentage of @var{used} divided by @var{size}.

@item file
The file name if specified on the command line.
@item target
The mount point.
@end table

The fields for block and inodes statistics are affected by the scaling
options like @option{-h} as usual.

The definition of the @var{field_list} can even be split among several
@option{--output} uses.

@example
#!/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
@end example


@item -P
@itemx --portability
@opindex -P
@opindex --portability
@cindex one-line output format
@cindex POSIX output format
@cindex portable output format
@cindex output format, portable
Use the POSIX output format.  This is like the default format except
for the following:

@enumerate
@item
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.

@item
The labels in the header output line are changed to conform to POSIX.

@item
The default block size and output format are unaffected by the
@env{DF_BLOCK_SIZE}, @env{BLOCK_SIZE} and @env{BLOCKSIZE} environment
variables.  However, the default block size is still affected by
@env{POSIXLY_CORRECT}: it is 512 if @env{POSIXLY_CORRECT} is set, 1024
otherwise.  @xref{Block size}.
@end enumerate

@optSi

@item --sync
@opindex --sync
@cindex file system space, retrieving current data more slowly
Invoke the @code{sync} system call before getting any usage data.  On
some systems (notably Solaris), doing this yields more up to date results,
but in general this option makes @command{df} much slower, especially when
there are many or very busy file systems.

@item --total
@opindex --total
@cindex grand total of file system size, usage and available space
Print a grand total of all arguments after all arguments have
been processed.  This can be used to find out the total 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, @command{df} prints @samp{"total"} into the
@var{source} column, and @samp{"-"} into the @var{target} column.
If there is no @var{source} column (see @option{--output}), then
@command{df} prints @samp{"total"} into the @var{target} column,
if present.

@item -t @var{fstype}
@itemx --type=@var{fstype}
@opindex -t
@opindex --type
@cindex file system types, limiting output to certain
Limit the listing to file systems of type @var{fstype}.  Multiple
file system types can be specified by giving multiple @option{-t} options.
By default, nothing is omitted.

@item -T
@itemx --print-type
@opindex -T
@opindex --print-type
@cindex file system types, printing
Print each file system's type.  The types printed here are the same ones
you can include or exclude with @option{-t} and @option{-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):

@table @samp

@item nfs
@cindex NFS file system type
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.

@item ext2@r{, }ext3@r{, }ext4@r{, }xfs@r{, }btrfs@dots{}
@cindex Linux file system types
@cindex local file system types
@opindex ext2 @r{file system type}
@opindex ext3 @r{file system type}
@opindex ext4 @r{file system type}
@opindex xfs @r{file system type}
@opindex btrfs @r{file system type}
A file system on a locally-mounted device.  (The system might even
support more than one type here; GNU/Linux does.)

@item iso9660@r{, }cdfs
@cindex CD-ROM file system type
@cindex DVD file system type
@cindex ISO9660 file system type
@opindex iso9660 @r{file system type}
@opindex cdfs @r{file system type}
A file system on a CD or DVD drive.  HP-UX uses @samp{cdfs}, most other
systems use @samp{iso9660}.

@item ntfs@r{,}fat
@cindex NTFS file system
@cindex DOS file system
@cindex MS-DOS file system
@cindex MS-Windows file system
@opindex ntfs @r{file system file}
@opindex fat @r{file system file}
File systems used by MS-Windows / MS-DOS.

@end table

@item -x @var{fstype}
@itemx --exclude-type=@var{fstype}
@opindex -x
@opindex --exclude-type
Limit the listing to file systems not of type @var{fstype}.
Multiple file system types can be eliminated by giving multiple
@option{-x} options.  By default, no file system types are omitted.

@item -v
Ignored; for compatibility with System V versions of @command{df}.

@end table

@command{df} is installed only on systems that have usable mount tables,
so portable scripts should not rely on its existence.

@exitstatus
Failure includes the case where no output is generated, so you can
inspect the exit status of a command like @samp{df -t ext3 -t reiserfs
@var{dir}} to test whether @var{dir} is on a file system of type
@samp{ext3} or @samp{reiserfs}.

Since the list of file systems (@var{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 @option{-a}, @option{-l}, @option{-t}
or @option{-x} is used together with a file name argument.


@node du invocation
@section @command{du}: Estimate file space usage

@pindex du
@cindex file space usage
@cindex disk usage for files

@command{du} reports the amount of file system space used by the set
of specified files and for each subdirectory (of directory arguments).
Synopsis:

@example
du [@var{option}]@dots{} [@var{file}]@dots{}
@end example

With no arguments, @command{du} reports the file system space for the current
directory.  Normally the space is printed in units of
1024 bytes, but this can be overridden (@pxref{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 @var{file} argument order affects which links
are counted, and changing the argument order may change the numbers
and entries that @command{du} outputs.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@optNull

@item -a
@itemx --all
@opindex -a
@opindex --all
Show counts for all files, not just directories.

@item --apparent-size
@opindex --apparent-size
Print apparent sizes, rather than file system usage.  The apparent size of a
file is the number of bytes reported by @code{wc -c} on regular files,
or more generally, @code{ls -l --block-size=1} or @code{stat --format=%s}.
For example, a file containing the word @samp{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 file system space, depending on
the type and configuration of the file system on which the file resides.
However, a sparse file created with this command:

@example
dd bs=1 seek=2GiB if=/dev/null of=big
@end example

@noindent
has an apparent size of 2 GiB, yet on most modern
file systems, it actually uses almost no space.

@item -B @var{size}
@itemx --block-size=@var{size}
@opindex -B
@opindex --block-size
@cindex file sizes
Scale sizes by @var{size} before printing them (@pxref{Block size}).
For example, @option{-BG} prints sizes in units of 1,073,741,824 bytes.

@item -b
@itemx --bytes
@opindex -b
@opindex --bytes
Equivalent to @code{--apparent-size --block-size=1}.

@item -c
@itemx --total
@opindex -c
@opindex --total
@cindex grand total of file system space
Print a grand total of all arguments after all arguments have
been processed.  This can be used to find out the total file system usage of
a given set of files or directories.

@item -D
@itemx --dereference-args
@opindex -D
@opindex --dereference-args
Dereference symbolic links that are command line arguments.
Does not affect other symbolic links.  This is helpful for finding
out the file system usage of directories, such as @file{/usr/tmp}, which
are often symbolic links.

@item -d @var{depth}
@itemx --max-depth=@var{depth}
@opindex -d @var{depth}
@opindex --max-depth=@var{depth}
@cindex limiting output of @command{du}
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 @code{du --max-depth=0} is equivalent to @code{du -s}.

@c --files0-from=FILE
@filesZeroFromOption{du,, with the @option{--total} (@option{-c}) option}

@item -H
@opindex -H
Equivalent to @option{--dereference-args} (@option{-D}).

@optHumanReadable

@item --inodes
@opindex --inodes
@cindex inode usage, dereferencing in @command{du}
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 @command{df},
option @option{--inodes}).
It can well be combined with the options @option{-a}, @option{-c},
@option{-h}, @option{-l}, @option{-s}, @option{-S}, @option{-t} and
@option{-x}; however, passing other options regarding the block size, for
example @option{-b}, @option{-m} and @option{--apparent-size}, is ignored.

@item -k
@opindex -k
@cindex kibibytes for file sizes
Print sizes in 1024-byte blocks, overriding the default block size
(@pxref{Block size}).
This option is equivalent to @option{--block-size=1K}.

@item -L
@itemx --dereference
@opindex -L
@opindex --dereference
@cindex symbolic links, dereferencing in @command{du}
Dereference symbolic links (show the file system space used by the file
or directory that the link points to instead of the space used by
the link).

@item -l
@itemx --count-links
@opindex -l
@opindex --count-links
@cindex hard links, counting in @command{du}
Count the size of all files, even if they have appeared already (as a
hard link).

@item -m
@opindex -m
@cindex mebibytes for file sizes
Print sizes in 1,048,576-byte blocks, overriding the default block size
(@pxref{Block size}).
This option is equivalent to @option{--block-size=1M}.

@item -P
@itemx --no-dereference
@opindex -P
@opindex --no-dereference
@cindex symbolic links, dereferencing in @command{du}
For each symbolic link encountered by @command{du},
consider the file system space used by the symbolic link itself.

@item -S
@itemx --separate-dirs
@opindex -S
@opindex --separate-dirs
Normally, in the output of @command{du} (when not using @option{--summarize}),
the size listed next to a directory name, @var{d}, represents the sum
of sizes of all entries beneath @var{d} as well as the size of @var{d} itself.
With @option{--separate-dirs}, the size reported for a directory name,
@var{d}, will exclude the size of any subdirectories.

@optSi

@item -s
@itemx --summarize
@opindex -s
@opindex --summarize
Display only a total for each argument.

@item -t @var{size}
@itemx --threshold=@var{size}
@opindex -t
@opindex --threshold
Exclude entries based on a given @var{size}.  The @var{size} refers to used
blocks in normal mode (@pxref{Block size}), or inodes count in conjunction
with the @option{--inodes} option.

If @var{size} is positive, then @command{du} will only print entries with a size
greater than or equal to that.

If @var{size} is negative, then @command{du} will only print entries with a size
smaller than or equal to that.

Although GNU @command{find} can be used to find files of a certain size,
@command{du}'s @option{--threshold} option can be used to also filter
directories based on a given size.

Please note that the @option{--threshold} option can be combined with the
@option{--apparent-size} option, and in this case would elide entries based on
its apparent size.

Please note that the @option{--threshold} option can be combined with the
@option{--inodes} option, and in this case would elide entries based on
its inodes count.

Here's how you would use @option{--threshold} to find directories with a size
greater than or equal to 200 megabytes:

@example
du --threshold=200MB
@end example

Here's how you would use @option{--threshold} to find directories and files -
note the @option{-a} - with an apparent size smaller than or equal to 500 bytes:

@example
du -a -t -500 --apparent-size
@end example

Here's how you would use @option{--threshold} to find directories on the root
file system with more than 20000 inodes used in the directory tree below:

@example
du --inodes -x --threshold=20000 /
@end example


@item --time
@opindex --time
@cindex last modified dates, displaying in @command{du}
Show the most recent modification timestamp (mtime) of any file in the
directory, or any of its subdirectories.  @xref{File timestamps}.

@item --time=ctime
@itemx --time=status
@itemx --time=use
@opindex --time
@opindex ctime@r{, show the most recent}
@opindex status time@r{, show the most recent}
@opindex use time@r{, show the most recent}
Show the most recent status change timestamp (ctime) of any file in
the directory, or any of its subdirectories.  @xref{File timestamps}.

@item --time=atime
@itemx --time=access
@opindex --time
@opindex atime@r{, show the most recent}
@opindex access timestamp@r{, show the most recent}
Show the most recent access timestamp (atime) of any file in the
directory, or any of its subdirectories.  @xref{File timestamps}.

@item --time-style=@var{style}
@opindex --time-style
@cindex time style
List timestamps in style @var{style}.  This option has an effect only if
the @option{--time} option is also specified.  The @var{style} should
be one of the following:

@table @samp
@item +@var{format}
@vindex LC_TIME
List timestamps using @var{format}, where @var{format} is interpreted
like the format argument of @command{date} (@pxref{date invocation}).
For example, @option{--time-style="+%Y-%m-%d %H:%M:%S"} causes
@command{du} to list timestamps like @samp{2020-07-21 23:45:56}.  As
with @command{date}, @var{format}'s interpretation is affected by the
@env{LC_TIME} locale category.

@item full-iso
List timestamps in full using ISO 8601-like date, time, and time zone
components with nanosecond precision, e.g., @samp{2020-07-21
23:45:56.477817180 -0400}.  This style is equivalent to
@samp{+%Y-%m-%d %H:%M:%S.%N %z}.

@item long-iso
List ISO 8601 date and time components with minute precision, e.g.,
@samp{2020-07-21 23:45}.  These timestamps are shorter than
@samp{full-iso} timestamps, and are usually good enough for everyday
work.  This style is equivalent to @samp{+%Y-%m-%d %H:%M}.

@item iso
List ISO 8601 dates for timestamps, e.g., @samp{2020-07-21}.
This style is equivalent to @samp{+%Y-%m-%d}.
@end table

@vindex TIME_STYLE
You can specify the default value of the @option{--time-style} option
with the environment variable @env{TIME_STYLE}; if @env{TIME_STYLE} is not set
the default style is @samp{long-iso}.  For compatibility with @command{ls},
if @env{TIME_STYLE} begins with @samp{+} and contains a newline,
the newline and any later characters are ignored; if @env{TIME_STYLE}
begins with @samp{posix-} the @samp{posix-} is ignored; and if
@env{TIME_STYLE} is @samp{locale} it is ignored.

@item -X @var{file}
@itemx --exclude-from=@var{file}
@opindex -X @var{file}
@opindex --exclude-from=@var{file}
@cindex excluding files from @command{du}
Like @option{--exclude}, except take the patterns to exclude from @var{file},
one per line.  If @var{file} is @samp{-}, take the patterns from standard
input.

@item --exclude=@var{pattern}
@opindex --exclude=@var{pattern}
@cindex excluding files from @command{du}
When recursing, skip subdirectories or files matching @var{pattern}.
For example, @code{du --exclude='*.o'} excludes files whose names
end in @samp{.o}.

@item -x
@itemx --one-file-system
@opindex -x
@opindex --one-file-system
@cindex one file system, restricting @command{du} to
Skip directories that are on different file systems from the one that
the argument being processed is on.

@end table

@cindex NFS mounts from BSD to HP-UX
On BSD systems, @command{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 @command{du} program.

@exitstatus


@node stat invocation
@section @command{stat}: Report file or file system status

@pindex stat
@cindex file status
@cindex file system status

@command{stat} displays information about the specified file(s).  Synopsis:

@example
stat [@var{option}]@dots{} [@var{file}]@dots{}
@end example

With no option, @command{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, @command{stat} can
also give information about the files the links point to.

@mayConflictWithShellBuiltIn{stat}

@table @samp

@item -L
@itemx --dereference
@opindex -L
@opindex --dereference
@cindex symbolic links, dereferencing in @command{stat}
Change how @command{stat} treats symbolic links.
With this option, @command{stat} acts on the file referenced
by each symbolic link argument.
Without it, @command{stat} acts on any symbolic link argument directly.

@item -f
@itemx --file-system
@opindex -f
@opindex --file-system
@cindex file systems
Report information about the file systems where the given files are located
instead of information about the files themselves.
This option implies the @option{-L} option.

@item --cached=@var{mode}
@opindex --cached=@var{mode}
@cindex attribute caching
Control how attributes are read from the file system;
if supported by the system.  This allows one to
control the trade-off between freshness and efficiency
of attribute access, especially useful with remote file systems.
@var{mode} can be:

@table @samp
@item always
Always read the already cached attributes if available.

@item never
Always sychronize with the latest file system attributes.
This also mounts automounted files.

@item default
Leave the caching behavior to the underlying file system.

@end table

@item -c
@itemx --format=@var{format}
@opindex -c
@opindex --format=@var{format}
@cindex output format
Use @var{format} rather than the default format.
@var{format} is automatically newline-terminated, so
running a command like the following with two or more @var{file}
operands produces a line of output for each operand:
@example
$ stat --format=%d:%i / /usr
2050:2
2057:2
@end example

@item --printf=@var{format}
@opindex --printf=@var{format}
@cindex output format
Use @var{format} rather than the default format.
Like @option{--format}, but interpret backslash escapes,
and do not output a mandatory trailing newline.
If you want a newline, include @samp{\n} in the @var{format}.
Here's how you would use @option{--printf} to print the device
and inode numbers of @file{/} and @file{/usr}:
@example
$ stat --printf='%d:%i\n' / /usr
2050:2
2057:2
@end example

@item -t
@itemx --terse
@opindex -t
@opindex --terse
@cindex terse output
Print the information in terse form, suitable for parsing by other programs.

The output of the following commands are identical and the @option{--format}
also identifies the items printed (in fuller form) in the default format.
Note the format string would include another @samp{%C} at the end with an
active SELinux security context.
@example
$ stat --format="%n %s %b %f %u %g %D %i %h %t %T %X %Y %Z %W %o" ...
$ stat --terse ...
@end example

The same illustrating terse output in @option{--file-system} mode:
@example
$ stat -f --format="%n %i %l %t %s %S %b %f %a %c %d" ...
$ stat -f --terse ...
@end example
@end table

The valid @var{format} directives for files with @option{--format} and
@option{--printf} are:

@itemize @bullet
@item %a - Permission bits in octal (note @samp{#} and @samp{0} printf flags)
@item %A - Permission bits in symbolic form (similar to @command{ls -ld})
@item %b - Number of blocks allocated (see @samp{%B})
@item %B - The size in bytes of each block reported by @samp{%b}
@item %C - The SELinux security context of a file, if available
@item %d - Device number in decimal (st_dev)
@item %D - Device number in hex (st_dev)
@item %Hd - Major device number in decimal
@item %Ld - Minor device number in decimal
@item %f - Raw mode in hex
@item %F - File type
@item %g - Group ID of owner
@item %G - Group name of owner
@item %h - Number of hard links
@item %i - Inode number
@item %m - Mount point (See note below)
@item %n - File name
@item %N - Quoted file name with dereference if symbolic link (see below)
@item %o - Optimal I/O transfer size hint
@item %s - Total size, in bytes
@item %r - Device type in decimal (st_rdev)
@item %R - Device type in hex (st_rdev)
@item %Hr - Major device type in decimal (see below)
@item %Lr - Minor device type in decimal (see below)
@item %t - Major device type in hex (see below)
@item %T - Minor device type in hex (see below)
@item %u - User ID of owner
@item %U - User name of owner
@item %w - Time of file birth, or @samp{-} if unknown
@item %W - Time of file birth as seconds since Epoch, or @samp{0}
@item %x - Time of last access
@item %X - Time of last access as seconds since Epoch
@item %y - Time of last data modification
@item %Y - Time of last data modification as seconds since Epoch
@item %z - Time of last status change
@item %Z - Time of last status change as seconds since Epoch
@end itemize

The @samp{%a} format prints the octal mode, and so it is useful
to control the zero padding of the output with the @samp{#} and @samp{0}
printf flags. For example to pad to at least 3 wide while making larger
numbers unambiguously octal, you can use @samp{%#03a}.

The @samp{%N} format can be set with the environment variable
@env{QUOTING_STYLE}@.  If that environment variable is not set,
the default value is @samp{shell-escape-always}.  Valid quoting styles are:
@quotingStyles

The @samp{r}, @samp{R}, @samp{%t}, and @samp{%T} formats operate on the st_rdev
member of the stat(2) structure, i.e., the represented device rather than
the containing device, and so 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 @samp{%W}, @samp{%X}, @samp{%Y}, and @samp{%Z} formats accept a
precision preceded by a period to specify the number of digits to
print after the decimal point.  For example, @samp{%.3X} outputs the
access timestamp to millisecond precision.  If a period is given but no
precision, @command{stat} uses 9 digits, so @samp{%.X} is equivalent to
@samp{%.9X}@.  When discarding excess precision, timestamps are truncated
toward minus infinity.

@example
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]
@end example

The mount point printed by @samp{%m} is similar to that output
by @command{df}, except that:
@itemize @bullet
@item
stat does not dereference symlinks by default
(unless @option{-L} is specified)
@item
stat does not search for specified device nodes in the
file system list, instead operating on them directly
@item
@cindex bind mount
stat outputs the alias for a bind mounted file, rather than
the initial mount point of its backing device.
One can recursively call stat until there is no change in output,
to get the current base mount point
@end itemize

When listing file system information (@option{--file-system} (@option{-f})),
you must use a different set of @var{format} directives:

@itemize @bullet
@item %a - Free blocks available to non-super-user
@item %b - Total data blocks in file system
@item %c - Total file nodes in file system
@item %d - Free file nodes in file system
@item %f - Free blocks in file system
@item %i - File System ID in hex
@item %l - Maximum length of file names
@item %n - File name
@item %s - Block size (for faster transfers)
@item %S - Fundamental block size (for block counts)
@item %t - Type in hex
@item %T - Type in human readable form
@end itemize

@vindex TZ
Timestamps are listed according to the time zone rules specified by
the @env{TZ} environment variable, or by the system default rules if
@env{TZ} is not set.  @xref{TZ Variable,, Specifying the Time Zone
with @env{TZ}, libc, The GNU C Library Reference Manual}.

@exitstatus


@node sync invocation
@section @command{sync}: Synchronize cached writes to persistent storage

@pindex sync
@cindex synchronize file system and memory
@cindex Synchronize cached writes to persistent storage

@command{sync} synchronizes in memory files or file systems to persistent
storage.  Synopsis:

@example
sync [@var{option}] [@var{file}]@dots{}
@end example

@cindex superblock, writing
@cindex inodes, written buffered
@command{sync} writes any data buffered in memory out to the storage device.
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 @command{sync} program does nothing but exercise the @code{sync},
@code{syncfs}, @code{fsync}, and @code{fdatasync} system calls.

@cindex crashes and corruption
The kernel keeps data in memory to avoid doing (relatively slow) device
reads and writes.  This improves performance, but if the computer
crashes, data may be lost or the file system corrupted as a
result.  The @command{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
@ref{Common options}.

@table @samp
@item -d
@itemx --data
@opindex --data
Use fdatasync(2) to sync only the data for the file,
and any metadata required to maintain file system consistency.

@item -f
@itemx --file-system
@opindex --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 @emph{not} specify
this option if passing a device node like @samp{/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.
@end table

@exitstatus


@node truncate invocation
@section @command{truncate}: Shrink or extend the size of a file

@pindex truncate
@cindex truncating, file sizes

@command{truncate} shrinks or extends the size of each @var{file} to the
specified size.  Synopsis:

@example
truncate @var{option}@dots{} @var{file}@dots{}
@end example

@cindex files, creating
Any @var{file} that does not exist is created.

@cindex sparse files, creating
@cindex holes, creating files with
If a @var{file} is larger than the specified size, the extra data is lost.
If a @var{file} is shorter, it is extended and the sparse extended part
(or hole) reads as zero bytes.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -c
@itemx --no-create
@opindex -c
@opindex --no-create
Do not create files that do not exist.

@item -o
@itemx --io-blocks
@opindex -o
@opindex --io-blocks
Treat @var{size} as number of I/O blocks of the @var{file} rather than bytes.

@item -r @var{rfile}
@itemx --reference=@var{rfile}
@opindex -r
@opindex --reference
Base the size of each @var{file} on the size of @var{rfile}.

@item -s @var{size}
@itemx --size=@var{size}
@opindex -s
@opindex --size
Set or adjust the size of each @var{file} according to @var{size}.
@var{size} is in bytes unless @option{--io-blocks} is specified.
@multiplierSuffixesNoBlocks{size}

@var{size} may also be prefixed by one of the following to adjust
the size of each @var{file} based on its current size:
@example
@samp{+}  => extend by
@samp{-}  => reduce by
@samp{<}  => at most
@samp{>}  => at least
@samp{/}  => round down to multiple of
@samp{%}  => round up to multiple of
@end example

@end table

@exitstatus


@node Printing text
@chapter Printing text

@cindex printing text, commands for
@cindex commands for printing text

This section describes commands that display text strings.

@menu
* echo invocation::             Print a line of text.
* printf invocation::           Format and print data.
* yes invocation::              Print a string until interrupted.
@end menu


@node echo invocation
@section @command{echo}: Print a line of text

@pindex echo
@cindex displaying text
@cindex printing text
@cindex text, displaying
@cindex arbitrary text, displaying

@command{echo} writes each given @var{string} to standard output, with a
space between each and a newline after the last one.  Synopsis:

@example
echo [@var{option}]@dots{} [@var{string}]@dots{}
@end example

@mayConflictWithShellBuiltIn{echo}

Due to historical and backwards compatibility reasons, certain bare option-like
strings cannot be passed to @command{echo} as non-option arguments.
It is therefore not advisable to use @command{echo} for printing unknown or
variable arguments.  The @command{printf} command is recommended as a more
portable and flexible replacement for tasks historically performed by
@command{echo}.  @xref{printf invocation}.

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands, and the normally-special argument
@samp{--} has no special meaning and is treated like any other
@var{string}.

@table @samp
@item -n
@opindex -n
Do not output the trailing newline.

@item -e
@opindex -e
@cindex backslash escapes
Enable interpretation of the following backslash-escaped characters in
each @var{string}:

@table @samp
@item \a
alert (bell)
@item \b
backspace
@item \c
produce no further output
@item \e
escape
@item \f
form feed
@item \n
newline
@item \r
carriage return
@item \t
horizontal tab
@item \v
vertical tab
@item \\
backslash
@item \0@var{nnn}
the eight-bit value that is the octal number @var{nnn}
(zero to three octal digits), if @var{nnn} is
a nine-bit value, the ninth bit is ignored
@item \@var{nnn}
the eight-bit value that is the octal number @var{nnn}
(one to three octal digits), if @var{nnn} is
a nine-bit value, the ninth bit is ignored
@item \x@var{hh}
the eight-bit value that is the hexadecimal number @var{hh}
(one or two hexadecimal digits)
@end table

@item -E
@opindex -E
@cindex backslash escapes
Disable interpretation of backslash escapes in each @var{string}.
This is the default.  If @option{-e} and @option{-E} are both
specified, the last one given takes effect.

@end table

@vindex POSIXLY_CORRECT
If the @env{POSIXLY_CORRECT} environment variable is set, then when
@command{echo}'s first argument is not @option{-n} it outputs
option-like arguments instead of treating them as options.  For
example, @code{echo -ne hello} outputs @samp{-ne hello} instead of
plain @samp{hello}.  Also backslash escapes are always enabled.
Note to echo the string @samp{-n}, one of the characters
can be escaped in either octal or hexadecimal representation.
For example, @code{echo -e '\x2dn'}.

POSIX does not require support for any options, and says
that the behavior of @command{echo} is implementation-defined if any
@var{string} contains a backslash or if the first argument is @option{-n}.
Portable programs should use the @command{printf} command instead.
@xref{printf invocation}.

@exitstatus


@node printf invocation
@section @command{printf}: Format and print data

@pindex printf
@command{printf} does formatted printing of text.  Synopsis:

@example
printf @var{format} [@var{argument}]@dots{}
@end example

@command{printf} prints the @var{format} string, interpreting @samp{%}
directives and @samp{\} escapes to format numeric and string arguments
in a way that is mostly similar to the C @samp{printf} function.
@xref{Output Conversion Syntax,, @command{printf} format directives,
libc, The GNU C Library Reference Manual}, for details.
The differences are listed below.

@mayConflictWithShellBuiltIn{printf}

@itemize @bullet

@item
The @var{format} argument is reused as necessary to convert all the
given @var{argument}s.  For example, the command @samp{printf %s a b}
outputs @samp{ab}.

@item
Missing @var{argument}s are treated as null strings or as zeros,
depending on whether the context expects a string or a number.  For
example, the command @samp{printf %sx%d} prints @samp{x0}.

@item
@kindex \c
An additional escape, @samp{\c}, causes @command{printf} to produce no
further output.  For example, the command @samp{printf 'A%sC\cD%sF' B
E} prints @samp{ABC}.

@item
The hexadecimal escape sequence @samp{\x@var{hh}} has at most two
digits, as opposed to C where it can have an unlimited number of
digits.  For example, the command @samp{printf '\x07e'} prints two
bytes, whereas the C statement @samp{printf ("\x07e")} prints just
one.

@item
@kindex %b
An additional directive @samp{%b}, prints its
argument string with @samp{\} escapes interpreted in the same way as in
the @var{format} string, except that octal escapes are of the form
@samp{\0@var{ooo}} where @var{ooo} is 0 to 3 octal digits.  If
@samp{\@var{ooo}} is nine-bit value, ignore the ninth bit.
If a precision is also given, it limits the number of bytes printed
from the converted string.

@item
@kindex %q
An additional directive @samp{%q}, prints its argument string
in a format that can be reused as input by most shells.
Non-printable characters are escaped with the POSIX proposed @samp{$''} syntax,
and shell metacharacters are quoted appropriately.
This is an equivalent format to @command{ls --quoting=shell-escape} output.

@item
Numeric arguments must be single C constants, possibly with leading
@samp{+} or @samp{-}.  For example, @samp{printf %.4d -3} outputs
@samp{-0003}.

@item
@vindex POSIXLY_CORRECT
If the leading character of a numeric argument is @samp{"} or @samp{'}
then its value is the numeric value of the immediately following
character.  Any remaining characters are silently ignored if the
@env{POSIXLY_CORRECT} environment variable is set; otherwise, a
warning is printed.  For example, @samp{printf "%d" "'a"} outputs
@samp{97} on hosts that use the ASCII character set, since
@samp{a} has the numeric value 97 in ASCII.

@end itemize

@vindex LC_NUMERIC
A floating point argument is interpreted according to
the @env{LC_NUMERIC} category of either the current or the C locale,
and is printed according to the current locale.
For example, in a locale whose decimal point character is a comma,
the command @samp{printf '%g %g' 2,5 2.5} outputs @samp{2,5 2,5}.
@xref{Floating point}.

@kindex \@var{ooo}
@kindex \x@var{hh}
@command{printf} interprets @samp{\@var{ooo}} in @var{format} as an octal number
(if @var{ooo} is 1 to 3 octal digits) specifying a byte to print,
and @samp{\x@var{hh}} as a hexadecimal number (if @var{hh} is 1 to 2 hex
digits) specifying a character to print.
Note however that when @samp{\@var{ooo}} specifies a number larger than 255,
@command{printf} ignores the ninth bit.
For example, @samp{printf '\400'} is equivalent to @samp{printf '\0'}.

@kindex \uhhhh
@kindex \Uhhhhhhhh
@cindex Unicode
@cindex ISO/IEC 10646
@vindex LC_CTYPE
@command{printf} interprets two character syntaxes introduced in
ISO C 99:
@samp{\u} for 16-bit Unicode (ISO/IEC 10646)
characters, specified as
four hexadecimal digits @var{hhhh}, and @samp{\U} for 32-bit Unicode
characters, specified as eight hexadecimal digits @var{hhhhhhhh}.
@command{printf} outputs the Unicode characters
according to the @env{LC_CTYPE} locale.  Unicode characters in the ranges
U+0000@dots{}U+009F, U+D800@dots{}U+DFFF cannot be specified by this syntax,
except for U+0024 ($), U+0040 (@@), and U+0060 (@`).

The processing of @samp{\u} and @samp{\U} requires a full-featured
@code{iconv} facility.  It is activated on systems with glibc 2.2 (or newer),
or when @code{libiconv} is installed prior to this package.  Otherwise
@samp{\u} and @samp{\U} will print as-is.

The only options are a lone @option{--help} or
@option{--version}.  @xref{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

@example
$ env printf '\u20AC 14.95'
@end example

@noindent
will be output correctly in all locales supporting the Euro symbol
(ISO-8859-15, UTF-8, and others).  Similarly, a Chinese string

@example
$ env printf '\u4e2d\u6587'
@end example

@noindent
will be output correctly in all Chinese locales (GB2312, BIG5, UTF-8, etc).

Note that in these examples, the @command{printf} command has been
invoked via @command{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:

@example
$ 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
@end example

@exitstatus


@node yes invocation
@section @command{yes}: Print a string until interrupted

@pindex yes
@cindex repeated output of a string

@command{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 @samp{y} followed by a newline forever until killed.

Upon a write error, @command{yes} exits with status @samp{1}.

The only options are a lone @option{--help} or @option{--version}.
To output an argument that begins with
@samp{-}, precede it with @option{--}, e.g., @samp{yes -- --help}.
@xref{Common options}.


@node Conditions
@chapter Conditions

@cindex conditions
@cindex commands for exit status
@cindex exit status commands

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 @code{if} statements, or as the last command in a
pipeline.

@menu
* false invocation::            Do nothing, unsuccessfully.
* true invocation::             Do nothing, successfully.
* test invocation::             Check file types and compare values.
* expr invocation::             Evaluate expressions.
@end menu


@node false invocation
@section @command{false}: Do nothing, unsuccessfully

@pindex false
@cindex do nothing, unsuccessfully
@cindex failure exit status
@cindex exit status of @command{false}

@command{false} does nothing except return an exit status of 1, meaning
@dfn{failure}.  It can be used as a place holder in shell scripts
where an unsuccessful command is needed.
In most modern shells, @command{false} is a built-in command, so when
you use @samp{false} in a script, you're probably using the built-in
command, not the one documented here.

@command{false} honors the @option{--help} and @option{--version} options.

This version of @command{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 @command{false} (unlike all other programs documented herein)
exits unsuccessfully, even when invoked with
@option{--help} or @option{--version}.

Portable programs should not assume that the exit status of
@command{false} is 1, as it is greater than 1 on some
non-GNU hosts.


@node true invocation
@section @command{true}: Do nothing, successfully

@pindex true
@cindex do nothing, successfully
@cindex no-op
@cindex successful exit
@cindex exit status of @command{true}

@command{true} does nothing except return an exit status of 0, meaning
@dfn{success}.  It can be used as a place holder in shell scripts
where a successful command is needed, although the shell built-in
command @code{:} (colon) may do the same thing faster.
In most modern shells, @command{true} is a built-in command, so when
you use @samp{true} in a script, you're probably using the built-in
command, not the one documented here.

@command{true} honors the @option{--help} and @option{--version} options.

Note, however, that it is possible to cause @command{true}
to exit with nonzero status: with the @option{--help} or @option{--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:

@example
$ ./true --version >&-
./true: write error: Bad file number
$ ./true --version > /dev/full
./true: write error: No space left on device
@end example

This version of @command{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.

@node test invocation
@section @command{test}: Check file types and compare values

@pindex test
@cindex check file types
@cindex compare values
@cindex expression evaluation

@command{test} returns a status of 0 (true) or 1 (false) depending on the
evaluation of the conditional expression @var{expr}.  Each part of the
expression must be a separate argument.

@command{test} has file status checks, string operators, and numeric
comparison operators.

@command{test} has an alternate form that uses opening and closing
square brackets instead a leading @samp{test}.  For example, instead
of @samp{test -d /}, you can write @samp{[ -d / ]}.  The square
brackets must be separate arguments; for example, @samp{[-d /]} does
not have the desired effect.  Since @samp{test @var{expr}} and @samp{[
@var{expr} ]} have the same meaning, only the former form is discussed
below.

Synopses:

@example
test @var{expression}
test
[ @var{expression} ]
[ ]
[ @var{option}
@end example

@mayConflictWithShellBuiltIn{test}

If @var{expression} is omitted, @command{test} returns false.
If @var{expression} is a single argument,
@command{test} returns false if the argument is null and true
otherwise.  The argument
can be any string, including strings like @samp{-d}, @samp{-1},
@samp{--}, @samp{--help}, and @samp{--version} that most other
programs would treat as options.  To get help and version information,
invoke the commands @samp{[ --help} and @samp{[ --version}, without
the usual closing brackets.  @xref{Common options}.

@cindex exit status of @command{test}
Exit status:

@display
0 if the expression is true,
1 if the expression is false,
2 if an error occurred.
@end display

@menu
* File type tests::             -[bcdfhLpSt]
* Access permission tests::     -[gkruwxOG]
* File characteristic tests::   -e -s -nt -ot -ef
* String tests::                -z -n = == !=
* Numeric tests::               -eq -ne -lt -le -gt -ge
* Connectives for test::        ! -a -o
@end menu


@node File type tests
@subsection File type tests

@cindex file type tests

These options test for particular types of files.  (Everything's a file,
but not all files are the same!)

@table @samp

@item -b @var{file}
@opindex -b
@cindex block special check
True if @var{file} exists and is a block special device.

@item -c @var{file}
@opindex -c
@cindex character special check
True if @var{file} exists and is a character special device.

@item -d @var{file}
@opindex -d
@cindex directory check
True if @var{file} exists and is a directory.

@item -f @var{file}
@opindex -f
@cindex regular file check
True if @var{file} exists and is a regular file.

@item -h @var{file}
@itemx -L @var{file}
@opindex -L
@opindex -h
@cindex symbolic link check
True if @var{file} exists and is a symbolic link.
Unlike all other file-related tests, this test does not dereference
@var{file} if it is a symbolic link.

@item -p @var{file}
@opindex -p
@cindex named pipe check
True if @var{file} exists and is a named pipe.

@item -S @var{file}
@opindex -S
@cindex socket check
True if @var{file} exists and is a socket.

@item -t @var{fd}
@opindex -t
@cindex terminal check
True if @var{fd} is a file descriptor that is associated with a
terminal.

@end table


@node Access permission tests
@subsection Access permission tests

@cindex access permission tests
@cindex permission tests

These options test for particular access permissions.

@table @samp

@item -g @var{file}
@opindex -g
@cindex set-group-ID check
True if @var{file} exists and has its set-group-ID bit set.

@item -k @var{file}
@opindex -k
@cindex sticky bit check
True if @var{file} exists and has its @dfn{sticky} bit set.

@item -r @var{file}
@opindex -r
@cindex readable file check
True if @var{file} exists and the user has read access.

@item -u @var{file}
@opindex -u
@cindex set-user-ID check
True if @var{file} exists and has its set-user-ID bit set.

@item -w @var{file}
@opindex -w
@cindex writable file check
True if @var{file} exists and the user has write access.

@item -x @var{file}
@opindex -x
@cindex executable file check
True if @var{file} exists and the user has execute access
(or search permission, if it is a directory).

@item -O @var{file}
@opindex -O
@cindex owned by effective user ID check
True if @var{file} exists and is owned by the current effective user ID.

@item -G @var{file}
@opindex -G
@cindex owned by effective group ID check
True if @var{file} exists and is owned by the current effective group ID.

@end table

@node File characteristic tests
@subsection File characteristic tests

@cindex file characteristic tests

These options test other file characteristics.

@table @samp

@item -e @var{file}
@opindex -e
@cindex existence-of-file check
True if @var{file} exists.

@item -s @var{file}
@opindex -s
@cindex nonempty file check
True if @var{file} exists and has a size greater than zero.

@item @var{file1} -nt @var{file2}
@opindex -nt
@cindex newer-than file check
True if @var{file1} is newer (according to modification date) than
@var{file2}, or if @var{file1} exists and @var{file2} does not.

@item @var{file1} -ot @var{file2}
@opindex -ot
@cindex older-than file check
True if @var{file1} is older (according to modification date) than
@var{file2}, or if @var{file2} exists and @var{file1} does not.

@item @var{file1} -ef @var{file2}
@opindex -ef
@cindex same file check
@cindex hard link check
True if @var{file1} and @var{file2} have the same device and inode
numbers, i.e., if they are hard links to each other.

@item -N @var{file}
@opindex -N
@cindex mtime-greater-atime file check
True if @var{file} exists and has been modified (mtime) since it was
last read (atime).

@end table


@node String tests
@subsection String tests

@cindex string tests

These options test string characteristics.  You may need to quote
@var{string} arguments for the shell.  For example:

@example
test -n "$V"
@end example

The quotes here prevent the wrong arguments from being passed to
@command{test} if @samp{$V} is empty or contains special characters.

@table @samp

@item -z @var{string}
@opindex -z
@cindex zero-length string check
True if the length of @var{string} is zero.

@item -n @var{string}
@itemx @var{string}
@opindex -n
@cindex nonzero-length string check
True if the length of @var{string} is nonzero.

@item @var{string1} = @var{string2}
@opindex =
@cindex equal string check
True if the strings are equal.

@item @var{string1} == @var{string2}
@opindex ==
@cindex equal string check
True if the strings are equal (synonym for =).
Note this form is not as portable to other
shells and systems.

@item @var{string1} != @var{string2}
@opindex !=
@cindex not-equal string check
True if the strings are not equal.

@end table


@node Numeric tests
@subsection Numeric tests

@cindex numeric tests
@cindex arithmetic tests

Numeric relational operators.  The arguments must be entirely numeric
(possibly negative), or the special expression @w{@code{-l @var{string}}},
which evaluates to the length of @var{string}.

@table @samp

@item @var{arg1} -eq @var{arg2}
@itemx @var{arg1} -ne @var{arg2}
@itemx @var{arg1} -lt @var{arg2}
@itemx @var{arg1} -le @var{arg2}
@itemx @var{arg1} -gt @var{arg2}
@itemx @var{arg1} -ge @var{arg2}
@opindex -eq
@opindex -ne
@opindex -lt
@opindex -le
@opindex -gt
@opindex -ge
These arithmetic binary operators return true if @var{arg1} is equal,
not-equal, less-than, less-than-or-equal, greater-than, or
greater-than-or-equal than @var{arg2}, respectively.

@end table

For example:

@example
test -1 -gt -2 && echo yes
@result{} yes
test -l abc -gt 1 && echo yes
@result{} yes
test 0x100 -eq 1
@error{} test: integer expression expected before -eq
@end example


@node Connectives for test
@subsection Connectives for @command{test}

@cindex logical connectives
@cindex connectives, logical

Note it's preferred to use shell logical primitives
rather than these logical connectives internal to @command{test},
because an expression may become ambiguous
depending on the expansion of its parameters.

For example, this becomes ambiguous when @samp{$1}
is set to @samp{'!'} and @samp{$2} to the empty string @samp{''}:

@example
test "$1" -a "$2"
@end example

and should be written as:

@example
test "$1" && test "$2"
@end example

Note the shell logical primitives also benefit from
short circuit operation, which can be significant
for file attribute tests.

@table @samp

@item ! @var{expr}
@opindex !
True if @var{expr} is false.
@samp{!} has lower precedence than all parts of @var{expr}.
Note @samp{!} needs to be specified to the left
of a binary expression, I.e., @samp{'!' 1 -gt 2}
rather than @samp{1 '!' -gt 2}.
Also @samp{!} is often a shell special character
and is best used quoted.


@item @var{expr1} -a @var{expr2}
@opindex -a
@cindex logical and operator
@cindex and operator
True if both @var{expr1} and @var{expr2} are true.
@samp{-a} is left associative,
and has a higher precedence than @samp{-o}.

@item @var{expr1} -o @var{expr2}
@opindex -o
@cindex logical or operator
@cindex or operator
True if either @var{expr1} or @var{expr2} is true.
@samp{-o} is left associative.

@end table


@node expr invocation
@section @command{expr}: Evaluate expressions

@pindex expr
@cindex expression evaluation
@cindex evaluation of expressions

@command{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 @samp{-}.
@command{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 @command{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 @command{expr}'s
operators like @code{+}, so you cannot safely pass an arbitrary string
@code{$str} to expr merely by quoting it to the shell.  One way to
work around this is to use the GNU extension @code{+},
(e.g., @code{+ "$str" = foo}); a more portable way is to use
@code{@w{" $str"}} and to adjust the rest of the expression to take
the leading space into account (e.g., @code{@w{" $str" = " foo"}}).

You should not pass a negative integer or a string with leading
@samp{-} as @command{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.

@cindex parentheses for grouping
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.

Because @command{expr} uses multiple-precision arithmetic, it works
with integers wider than those of machine registers.

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.  Options must precede operands.

@cindex exit status of @command{expr}
Exit status:

@display
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).
@end display

@menu
* String expressions::          + : match substr index length
* Numeric expressions::         + - * / %
* Relations for expr::          | & < <= = == != >= >
* Examples of expr::            Examples.
@end menu


@node String expressions
@subsection String expressions

@cindex string expressions
@cindex expressions, string

@command{expr} supports pattern matching and other string operators.  These
have higher precedence than both the numeric and relational operators (in
the next sections).

@table @samp

@item @var{string} : @var{regex}
@cindex pattern matching
@cindex regular expression matching
@cindex matching patterns
Perform pattern matching.  The arguments are converted to strings and the
second is considered to be a (basic, a la GNU @code{grep}) regular
expression, with a @code{^} implicitly prepended.  The first argument is
then matched against this regular expression.

If @var{regex} does not use @samp{\(} and @samp{\)}, the @code{:}
expression returns the number of characters matched, or 0 if the match
fails.

If @var{regex} uses @samp{\(} and @samp{\)}, the @code{:} expression
returns the part of @var{string} that matched the subexpression, or
the null string if the match failed or the subexpression did not
contribute to the match.

@kindex \( @r{regexp operator}
Only the first @samp{\( @dots{} \)} pair is relevant to the return
value; additional pairs are meaningful only for grouping the regular
expression operators.

@kindex \+ @r{regexp operator}
@kindex \? @r{regexp operator}
@kindex \| @r{regexp operator}
In the regular expression, @code{\+}, @code{\?}, and @code{\|} are
operators which respectively match one or more, zero or one, or separate
alternatives.  These operators are GNU extensions.  @xref{Regular Expressions,,
Regular Expressions, grep, The GNU Grep Manual}, for details of
regular expression syntax.  Some examples are in @ref{Examples of expr}.

@item match @var{string} @var{regex}
@findex match
An alternative way to do pattern matching.  This is the same as
@w{@samp{@var{string} : @var{regex}}}.

@item substr @var{string} @var{position} @var{length}
@findex substr
Returns the substring of @var{string} beginning at @var{position}
with length at most @var{length}.  If either @var{position} or
@var{length} is negative, zero, or non-numeric, returns the null string.

@item index @var{string} @var{charset}
@findex index
Returns the first position in @var{string} where the first character in
@var{charset} was found.  If no character in @var{charset} is found in
@var{string}, return 0.

@item length @var{string}
@findex length
Returns the length of @var{string}.

@item + @var{token}
@kindex +
Interpret @var{token} as a string, even if it is a keyword like @var{match}
or an operator like @code{/}.
This makes it possible to test @code{expr length + "$x"} or
@code{expr + "$x" : '.*/\(.\)'} and have it do the right thing even if
the value of @var{$x} happens to be (for example) @code{/} or @code{index}.
This operator is a GNU extension.  Portable shell scripts should use
@code{@w{" $token"} : @w{' \(.*\)'}} instead of @code{+ "$token"}.

@end table

To make @command{expr} interpret keywords as strings, you must use the
@code{quote} operator.


@node Numeric expressions
@subsection Numeric expressions

@cindex numeric expressions
@cindex expressions, numeric

@command{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).

@table @samp

@item + -
@kindex +
@kindex -
@cindex addition
@cindex subtraction
Addition and subtraction.  Both arguments are converted to integers;
an error occurs if this cannot be done.

@item * / %
@kindex *
@kindex /
@kindex %
@cindex multiplication
@cindex division
@cindex remainder
Multiplication, division, remainder.  Both arguments are converted to
integers; an error occurs if this cannot be done.

@end table


@node Relations for expr
@subsection Relations for @command{expr}

@cindex connectives, logical
@cindex logical connectives
@cindex relations, numeric or string

@command{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.

@table @samp

@item |
@kindex |
@cindex logical or operator
@cindex or operator
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.

@item &
@kindex &
@cindex logical and operator
@cindex and operator
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.

@item < <= = == != >= >
@kindex <
@kindex <=
@kindex =
@kindex ==
@kindex >
@kindex >=
@cindex comparison operators
@vindex LC_COLLATE
Compare the arguments and return 1 if the relation is true, 0 otherwise.
@code{==} is a synonym for @code{=}.  @command{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 @env{LC_COLLATE} locale.

@end table


@node Examples of expr
@subsection Examples of using @command{expr}

@cindex examples of @command{expr}
Here are a few examples, including quoting for shell metacharacters.

To add 1 to the shell variable @code{foo}, in Bourne-compatible shells:

@example
foo=$(expr $foo + 1)
@end example

To print the non-directory part of the file name stored in
@code{$fname}, which need not contain a @code{/}:

@example
expr $fname : '.*/\(.*\)' '|' $fname
@end example

An example showing that @code{\+} is an operator:

@example
expr aaa : 'a\+'
@result{} 3
@end example

@example
expr abc : 'a\(.\)c'
@result{} b
expr index abcdef cz
@result{} 3
expr index index a
@error{} expr: syntax error
expr index + index a
@result{} 0
@end example


@node Redirection
@chapter Redirection

@cindex redirection
@cindex commands for redirection

Unix shells commonly provide several forms of @dfn{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.

@menu
* tee invocation::              Redirect output to multiple files or processes.
@end menu


@node tee invocation
@section @command{tee}: Redirect output to multiple files or processes

@pindex tee
@cindex pipe fitting
@cindex destinations, multiple output
@cindex read from standard input and write to standard output and files

The @command{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:

@example
tee [@var{option}]@dots{} [@var{file}]@dots{}
@end example

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 @option{-a} option is used.

In previous versions of GNU Coreutils (v5.3.0 -- v8.23),
a @var{file} of @samp{-}
caused @command{tee} to send another copy of input to standard output.
However, as the interleaved output was not very useful, @command{tee} now
conforms to POSIX and treats @samp{-} as a file name.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp
@item -a
@itemx --append
@opindex -a
@opindex --append
Append standard input to the given files rather than overwriting
them.

@item -i
@itemx --ignore-interrupts
@opindex -i
@opindex --ignore-interrupts
Ignore interrupt signals.

@item -p
@itemx --output-error[=@var{mode}]
@opindex -p
@opindex --output-error
Adjust the behavior with errors on the outputs,
with the long form option supporting selection
between the following @var{mode}s:

@table @samp
@item 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.

@item warn-nopipe
This is the default @var{mode} when not specified,
or when the short form @option{-p} is used.
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.

@item exit
Exit on error opening or writing any output, including pipes.

@item exit-nopipe
Exit on error opening or writing any output, except pipes.
@end table

@end table

The @command{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:

@example
wget https://example.com/some.iso && sha1sum some.iso
@end example

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:

@example
# slightly contrived, to demonstrate process substitution
wget -O - https://example.com/dvd.iso \
  | tee >(sha1sum > dvd.sha1) > dvd.iso
@end example

That makes @command{tee} write not just to the expected output file,
but also to a pipe running @command{sha1sum} and saving the final
checksum in a file named @file{dvd.sha1}.

Note, however, that this example relies on a feature of modern shells
called @dfn{process substitution}
(the @samp{>(command)} syntax, above;
@xref{Process Substitution,,Process Substitution, bash,
The Bash Reference Manual}.),
so it works with @command{zsh}, @command{bash}, and @command{ksh},
but not with @command{/bin/sh}.  So if you write code like this
in a shell script, be sure to start the script with @samp{#!/bin/bash}.

Note also that if any of the process substitutions (or piped standard output)
might exit early without consuming all the data, the @option{-p} option
is needed to allow @command{tee} to continue to process the input
to any remaining outputs.

Since the above example writes to one file and one process,
a more conventional and portable use of @command{tee} is even better:

@example
wget -O - https://example.com/dvd.iso \
  | tee dvd.iso | sha1sum > dvd.sha1
@end example

You can extend this example to make @command{tee} write to two processes,
computing MD5 and SHA1 checksums in parallel.  In this case,
process substitution is required:

@example
wget -O - https://example.com/dvd.iso \
  | tee >(sha1sum > dvd.sha1) \
        >(md5sum > dvd.md5) \
  > dvd.iso
@end example

This technique is also useful when you want to make a @emph{compressed}
copy of the contents of a pipe.
Consider a tool to graphically summarize file system usage data from
@samp{du -ak}.
For a large hierarchy, @samp{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 @command{du} output:

@example
du -ak | gzip -9 > /tmp/du.gz
gzip -d /tmp/du.gz | checkspace -a
@end example

With @command{tee} and process substitution, you start the GUI
right away and eliminate the decompression completely:

@example
du -ak | tee >(gzip -9 > /tmp/du.gz) | checkspace -a
@end example

Finally, if you regularly create more than one type of
compressed tarball at once, for example when @code{make dist} creates
both @command{gzip}-compressed and @command{bzip2}-compressed tarballs,
there may be a better way.
Typical @command{automake}-generated @file{Makefile} rules create
the two compressed tar archives with commands in sequence, like this
(slightly simplified):

@example
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
@end example

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:

@example
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
@end example

If you want to further process the output from process substitutions,
and those processes write atomically (i.e., write less than the system's
PIPE_BUF size at a time), that's possible with a construct like:

@example
tardir=your-pkg-M.N
tar chof - "$tardir" \
  | tee >(md5sum --tag) > >(sha256sum --tag) \
  | sort | gpg --clearsign > your-pkg-M.N.tar.sig
@end example

@exitstatus


@node File name manipulation
@chapter File name manipulation

@cindex file name manipulation
@cindex manipulation of file names
@cindex commands for file name manipulation

This section describes commands that manipulate file names.

@menu
* basename invocation::         Strip directory and suffix from a file name.
* dirname invocation::          Strip last file name component.
* pathchk invocation::          Check file name validity and portability.
* mktemp invocation::           Create temporary file or directory.
* realpath invocation::         Print resolved file names.
@end menu


@node basename invocation
@section @command{basename}: Strip directory and suffix from a file name

@pindex basename
@cindex strip directory and suffix from file names
@cindex directory, stripping from file names
@cindex suffix, stripping from file names
@cindex file names, stripping directory and suffix
@cindex leading directory components, stripping

@command{basename} removes any leading directory components from
@var{name}.  Synopsis:

@example
basename @var{name} [@var{suffix}]
basename @var{option}@dots{} @var{name}@dots{}
@end example

If @var{suffix} is specified and is identical to the end of @var{name},
it is removed from @var{name} as well.  Note that since trailing slashes
are removed prior to suffix matching, @var{suffix} will do nothing if it
contains slashes.  @command{basename} prints the result on standard
output.

@c This test is used both here and in the section on dirname.
@macro basenameAndDirname
Together, @command{basename} and @command{dirname} are designed such
that if @samp{ls "$name"} succeeds, then the command sequence @samp{cd
"$(dirname "$name")"; ls "$(basename "$name")"} will, too.  This works
for everything except file names containing a trailing newline.
@end macro
@basenameAndDirname

POSIX allows the implementation to define the results if
@var{name} is empty or @samp{//}.  In the former case, GNU
@command{basename} returns the empty string.  In the latter case, the
result is @samp{//} on platforms where @var{//} is distinct from
@var{/}, and @samp{/} on platforms where there is no difference.

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands.

@table @samp

@item -a
@itemx --multiple
@opindex -a
@opindex --multiple
Support more than one argument.  Treat every argument as a @var{name}.
With this, an optional @var{suffix} must be specified using the
@option{-s} option.

@item -s @var{suffix}
@itemx --suffix=@var{suffix}
@opindex -s
@opindex --suffix
Remove a trailing @var{suffix}.
This option implies the @option{-a} option.

@optZero

@end table

@exitstatus

Examples:

@example
# 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
@end example


@node dirname invocation
@section @command{dirname}: Strip last file name component

@pindex dirname
@cindex directory components, printing
@cindex stripping non-directory suffix
@cindex non-directory suffix, stripping

@command{dirname} prints all but the final slash-delimited component
of each @var{name}.  Slashes on either side of the final component are
also removed.  If the string contains no slash, @command{dirname}
prints @samp{.} (meaning the current directory).  Synopsis:

@example
dirname [@var{option}] @var{name}@dots{}
@end example

@var{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.

@basenameAndDirname

POSIX allows the implementation to define the results if
@var{name} is @samp{//}.  With GNU @command{dirname}, the
result is @samp{//} on platforms where @var{//} is distinct from
@var{/}, and @samp{/} on platforms where there is no difference.

The program accepts the following option.  Also see @ref{Common options}.

@table @samp

@optZero

@end table

@exitstatus

Examples:

@example
# Output "/usr/bin".
dirname /usr/bin/sort
dirname /usr/bin//.//

# Output "dir1" followed by "dir2"
dirname dir1/str dir2/str

# Output ".".
dirname stdio.h
@end example


@node pathchk invocation
@section @command{pathchk}: Check file name validity and portability

@pindex pathchk
@cindex file names, checking validity and portability
@cindex valid file names, checking for
@cindex portable file names, checking for

@command{pathchk} checks validity and portability of file names.  Synopsis:

@example
pathchk [@var{option}]@dots{} @var{name}@dots{}
@end example

For each @var{name}, @command{pathchk} prints an error message if any of
these conditions is true:

@enumerate
@item
One of the existing directories in @var{name} does not have search
(execute) permission,
@item
The length of @var{name} is larger than the maximum supported by the
operating system.
@item
The length of one component of @var{name} is longer than
its file system's maximum.
@end enumerate

A nonexistent @var{name} is not an error, so long as a file with that
name could be created under the above conditions.

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands.

@table @samp

@item -p
@opindex -p
Instead of performing checks based on the underlying file system,
print an error message if any of these conditions is true:

@enumerate
@item
A file name is empty.

@item
A file name contains a character outside the POSIX portable file
name character set, namely, the ASCII letters and digits, @samp{.},
@samp{_}, @samp{-}, and @samp{/}.

@item
The length of a file name or one of its components exceeds the
POSIX minimum limits for portability.
@end enumerate

@item -P
@opindex -P
Print an error message if a file name is empty, or if it contains a component
that begins with @samp{-}.

@item --portability
@opindex --portability
Print an error message if a file name is not portable to all POSIX
hosts.  This option is equivalent to @samp{-p -P}.

@end table

@cindex exit status of @command{pathchk}
Exit status:

@display
0 if all specified file names passed all checks,
1 otherwise.
@end display

@node mktemp invocation
@section @command{mktemp}: Create temporary file or directory

@pindex mktemp
@cindex file names, creating temporary
@cindex directory, creating temporary
@cindex temporary files and directories

@command{mktemp} manages the creation of temporary files and
directories.  Synopsis:

@example
mktemp [@var{option}]@dots{} [@var{template}]
@end example

Safely create a temporary file or directory based on @var{template},
and print its name.  If given, @var{template} must include at least
three consecutive @samp{X}s in the last component.  If omitted, the template
@samp{tmp.XXXXXXXXXX} is used, and option @option{--tmpdir} is
implied.  The final run of @samp{X}s in the @var{template} will be replaced
by alpha-numeric characters; thus, on a case-sensitive file system,
and with a @var{template} including a run of @var{n} instances of @samp{X},
there are @samp{62**@var{n}} potential file names.

Older scripts used to create temporary files by simply joining the
name of the program with the process id (@samp{$$}) 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 @command{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 @command{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):

@itemize @bullet

@item
Create a temporary file in the current directory.
@example
$ mktemp file.XXXX
file.H47c
@end example

@item
Create a temporary file with a known suffix.
@example
$ mktemp --suffix=.txt file-XXXX
file-H08W.txt
$ mktemp file-XXXX-XXXX.txt
file-XXXX-eI9L.txt
@end example

@item
Create a secure fifo relative to the user's choice of @env{TMPDIR},
but falling back to the current directory rather than @file{/tmp}.
Note that @command{mktemp} does not create fifos, but can create a
secure directory in which the fifo can live.  Exit the shell if the
directory or fifo could not be created.
@example
$ dir=$(mktemp -p "$@{TMPDIR:-.@}" -d dir-XXXX) || exit 1
$ fifo=$dir/fifo
$ mkfifo "$fifo" || @{ rmdir "$dir"; exit 1; @}
@end example

@item
Create and use a temporary file if possible, but ignore failure.  The
file will reside in the directory named by @env{TMPDIR}, if specified,
or else in @file{/tmp}.
@example
$ file=$(mktemp -q) && @{
>   # Safe to use $file only within this block.  Use quotes,
>   # since $TMPDIR, and thus $file, may contain whitespace.
>   echo ... > "$file"
>   rm "$file"
> @}
@end example

@item
Act as a semi-random character generator (it is not fully random,
since it is impacted by the contents of the current directory).  To
avoid security holes, do not use the resulting names to create a file.
@example
$ mktemp -u XXX
Gb9
$ mktemp -u XXX
nzC
@end example

@end itemize

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -d
@itemx --directory
@opindex -d
@opindex --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.

@item -q
@itemx --quiet
@opindex -q
@opindex --quiet
Suppress diagnostics about failure to create a file or directory.  The
exit status will still reflect whether a file was created.

@item -u
@itemx --dry-run
@opindex -u
@opindex --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.

@item -p @var{dir}
@itemx --tmpdir[=@var{dir}]
@opindex -p
@opindex --tmpdir
Treat @var{template} relative to the directory @var{dir}.  If
@var{dir} is not specified (only possible with the long option
@option{--tmpdir}) or is the empty string, use the value of
@env{TMPDIR} if available, otherwise use @samp{/tmp}.  If this is
specified, @var{template} must not be absolute.  However,
@var{template} can still contain slashes, although intermediate
directories must already exist.

@item --suffix=@var{suffix}
@opindex --suffix
Append @var{suffix} to the @var{template}.  @var{suffix} must not
contain slash.  If @option{--suffix} is specified, @var{template} must
end in @samp{X}; if it is not specified, then an appropriate
@option{--suffix} is inferred by finding the last @samp{X} in
@var{template}.  This option exists for use with the default
@var{template} and for the creation of a @var{suffix} that starts with
@samp{X}.

@item -t
@opindex -t
Treat @var{template} as a single file relative to the value of
@env{TMPDIR} if available, or to the directory specified by
@option{-p}, otherwise to @samp{/tmp}.  @var{template} must not
contain slashes.  This option is deprecated; the use of @option{-p}
without @option{-t} offers better defaults (by favoring the command
line over @env{TMPDIR}) and more flexibility (by allowing intermediate
directories).

@end table

@cindex exit status of @command{mktemp}
Exit status:

@display
0 if the file was created,
1 otherwise.
@end display


@node realpath invocation
@section @command{realpath}: Print the resolved file name.

@pindex realpath
@cindex file names, canonicalization
@cindex symlinks, resolution
@cindex canonical file name
@cindex canonicalize a file name
@pindex realpath
@findex realpath

@command{realpath} expands all symbolic links and resolves references to
@samp{/./}, @samp{/../} and extra @samp{/} characters.  By default,
all but the last component of the specified files must exist.  Synopsis:

@example
realpath [@var{option}]@dots{} @var{file}@dots{}
@end example

The file name canonicalization functionality overlaps with that of the
@command{readlink} command.  This is the preferred command for
canonicalization as it's a more suitable and standard name.  In addition
this command supports relative file name processing functionality.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -e
@itemx --canonicalize-existing
@opindex -e
@opindex --canonicalize-existing
Ensure that all components of the specified file names exist.
If any component is missing or unavailable, @command{realpath} will output
a diagnostic unless the @option{-q} option is specified, and exit with a
nonzero exit code.  A trailing slash requires that the name resolve to a
directory.

@item -m
@itemx --canonicalize-missing
@opindex -m
@opindex --canonicalize-missing
If any component of a specified file name is missing or unavailable,
treat it as a directory.

@item -L
@itemx --logical
@opindex -L
@opindex --logical
Symbolic links are resolved in the specified file names,
but they are resolved after any subsequent @samp{..} components are processed.

@item -P
@itemx --physical
@opindex -P
@opindex --physical
Symbolic links are resolved in the specified file names,
and they are resolved before any subsequent @samp{..} components are processed.
This is the default mode of operation.

@item -q
@itemx --quiet
@opindex -q
@opindex --quiet
Suppress diagnostic messages for specified file names.

@item --relative-to=@var{dir}
@opindex --relative-to
@cindex relpath
Print the resolved file names relative to the specified directory.
Note this option honors the @option{-m} and @option{-e} options
pertaining to file existence.

@item --relative-base=@var{dir}
@opindex --relative-base
Print the resolved file names as relative @emph{if} the files
are descendants of @var{dir}.
Otherwise, print the resolved file names as absolute.
Note this option honors the @option{-m} and @option{-e} options
pertaining to file existence.
For details about combining @option{--relative-to} and @option{--relative-base},
@pxref{Realpath usage examples}.

@item -s
@itemx --strip
@itemx --no-symlinks
@opindex -s
@opindex --strip
@opindex --no-symlinks
Do not resolve symbolic links.  Only resolve references to
@samp{/./}, @samp{/../} and remove extra @samp{/} characters.
When combined with the @option{-m} option, realpath operates
only on the file name, and does not touch any actual file.

@optZero

@end table

@cindex exit status of @command{realpath}
Exit status:

@display
0 if all file names were printed without issue.
1 otherwise.
@end display

@menu
* Realpath usage examples::              Realpath usage examples.
@end menu


@node Realpath usage examples
@subsection Realpath usage examples

@opindex --relative-to
@opindex --relative-base

By default, @command{realpath} prints the absolute file name of given files
(symlinks are resolved, @file{words} is resolved to @file{american-english}):

@example
@group
cd /home/user
realpath /usr/bin/sort /tmp/foo /usr/share/dict/words 1.txt
@result{} /usr/bin/sort
@result{} /tmp/foo
@result{} /usr/share/dict/american-english
@result{} /home/user/1.txt
@end group
@end example

With @option{--relative-to}, file names are printed relative to
the given directory:

@example
@group
realpath --relative-to=/usr/bin \
         /usr/bin/sort /tmp/foo /usr/share/dict/words 1.txt
@result{} sort
@result{} ../../tmp/foo
@result{} ../share/dict/american-english
@result{} ../../home/user/1.txt
@end group
@end example

With @option{--relative-base}, relative file names are printed @emph{if}
the resolved file name is below the given base directory. For files outside the
base directory absolute file names are printed:

@example
@group
realpath --relative-base=/usr \
         /usr/bin/sort /tmp/foo /usr/share/dict/words 1.txt
@result{} bin/sort
@result{} /tmp/foo
@result{} share/dict/american-english
@result{} /home/user/1.txt
@end group
@end example

When both @option{--relative-to=DIR1} and @option{--relative-base=DIR2}
are used, file names are printed relative to @var{dir1} @emph{if} they are
located below @var{dir2}. If the files are not below @var{dir2}, they are
printed as absolute file names:

@example
@group
realpath --relative-to=/usr/bin --relative-base=/usr \
         /usr/bin/sort /tmp/foo /usr/share/dict/words 1.txt
@result{} sort
@result{} /tmp/foo
@result{} ../share/dict/american-english
@result{} /home/user/1.txt
@end group
@end example

When both @option{--relative-to=DIR1} and @option{--relative-base=DIR2}
are used, @var{dir1} @emph{must} be a subdirectory of @var{dir2}. Otherwise,
@command{realpath} prints absolutes file names.


@node Working context
@chapter Working context

@cindex working context
@cindex commands for printing the 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.

@menu
* pwd invocation::              Print working directory.
* stty invocation::             Print or change terminal characteristics.
* printenv invocation::         Print environment variables.
* tty invocation::              Print file name of terminal on standard input.
@end menu


@node pwd invocation
@section @command{pwd}: Print working directory

@pindex pwd
@cindex print name of current directory
@cindex current working directory, printing
@cindex working directory, printing


@command{pwd} prints the name of the current directory.  Synopsis:

@example
pwd [@var{option}]@dots{}
@end example

The program accepts the following options.  Also see @ref{Common options}.

@table @samp
@item -L
@itemx --logical
@opindex -L
@opindex --logical
If the contents of the environment variable @env{PWD} provide an
absolute name of the current directory with no @samp{.} or @samp{..}
components, but possibly with symbolic links, then output those
contents.  Otherwise, fall back to default @option{-P} handling.

@item -P
@itemx --physical
@opindex -P
@opindex --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.
@end table

@cindex symbolic links and @command{pwd}
If @option{-L} and @option{-P} are both given, the last one takes
precedence.  If neither option is given, then this implementation uses
@option{-P} as the default unless the @env{POSIXLY_CORRECT}
environment variable is set.

@mayConflictWithShellBuiltIn{pwd}

@exitstatus


@node stty invocation
@section @command{stty}: Print or change terminal characteristics

@pindex stty
@cindex change or print terminal settings
@cindex terminal settings
@cindex line settings of terminal

@command{stty} prints or changes terminal characteristics, such as baud rate.
Synopses:

@example
stty [@var{option}] [@var{setting}]@dots{}
stty [@var{option}]
@end example

If given no line settings, @command{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 @samp{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
@option{--file} option.

@command{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 @ref{Common options}.

@table @samp
@item -a
@itemx --all
@opindex -a
@opindex --all
Print all current settings in human-readable form.  This option may not
be used in combination with any line settings.

@item -F @var{device}
@itemx --file=@var{device}
@opindex -F
@opindex --file
Set the line opened by the file name specified in @var{device} instead of
the tty line connected to standard input.  This option is necessary
because opening a POSIX tty requires use of the
@code{O_NONDELAY} flag to prevent a POSIX tty from blocking
until the carrier detect line is high if
the @code{clocal} flag is not set.  Hence, it is not always possible
to allow the shell to open the device in the traditional manner.

@item -g
@itemx --save
@opindex -g
@opindex --save
@cindex machine-readable @command{stty} output
Print all current settings in a form that can be used as an argument to
another @command{stty} command to restore the current settings.  This option
may not be used in combination with any line settings.

@end table

Many settings can be turned off by preceding them with a @samp{-}.
Such arguments are marked below with ``May be negated'' in their
description.  The descriptions themselves refer to the positive
case, that is, when @emph{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.

@command{stty} is installed only on platforms with the POSIX terminal
interface, so portable scripts should not rely on its existence on
non-POSIX platforms.

@exitstatus

@menu
* Control::                     Control settings
* Input::                       Input settings
* Output::                      Output settings
* Local::                       Local settings
* Combination::                 Combination settings
* Characters::                  Special characters
* Special::                     Special settings
@end menu


@node Control
@subsection Control settings

@cindex control settings
Control settings:

@table @samp
@item parenb
@opindex parenb
@cindex two-way parity
Generate parity bit in output and expect parity bit in input.
May be negated.

@item parodd
@opindex parodd
@cindex odd parity
@cindex even parity
Set odd parity (even if negated).  May be negated.

@item cmspar
@opindex cmspar
@cindex constant parity
@cindex stick parity
@cindex mark parity
@cindex space parity
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.

@item cs5
@itemx cs6
@itemx cs7
@itemx cs8
@opindex cs@var{n}
@cindex character size
@cindex eight-bit characters
Set character size to 5, 6, 7, or 8 bits.

@item hup
@itemx hupcl
@opindex hup[cl]
Send a hangup signal when the last process closes the tty.  May be
negated.

@item cstopb
@opindex cstopb
@cindex stop bits
Use two stop bits per character (one if negated).  May be negated.

@item cread
@opindex cread
Allow input to be received.  May be negated.

@item clocal
@opindex clocal
@cindex modem control
Disable modem control signals.  May be negated.

@item crtscts
@opindex crtscts
@cindex hardware flow control
@cindex flow control, hardware
@cindex RTS/CTS flow control
Enable RTS/CTS flow control.  Non-POSIX@.  May be negated.

@item cdtrdsr
@opindex cdtrdsr
@cindex hardware flow control
@cindex flow control, hardware
@cindex DTR/DSR flow control
Enable DTR/DSR flow control. Non-POSIX@.  May be negated.
@end table


@node Input
@subsection Input settings

@cindex input settings
These settings control operations on data received from the terminal.

@table @samp
@item ignbrk
@opindex ignbrk
@cindex breaks, ignoring
Ignore break characters.  May be negated.

@item brkint
@opindex brkint
@cindex breaks, cause interrupts
Make breaks cause an interrupt signal.  May be negated.

@item ignpar
@opindex ignpar
@cindex parity, ignoring
Ignore characters with parity errors.  May be negated.

@item parmrk
@opindex parmrk
@cindex parity errors, marking
Mark parity errors (with a 255-0-character sequence).  May be negated.

@item inpck
@opindex inpck
Enable input parity checking.  May be negated.

@item istrip
@opindex istrip
@cindex eight-bit input
Clear high (8th) bit of input characters.  May be negated.

@item inlcr
@opindex inlcr
@cindex newline, translating to return
Translate newline to carriage return.  May be negated.

@item igncr
@opindex igncr
@cindex return, ignoring
Ignore carriage return.  May be negated.

@item icrnl
@opindex icrnl
@cindex return, translating to newline
Translate carriage return to newline.  May be negated.

@item iutf8
@opindex iutf8
@cindex input encoding, UTF-8
Assume input characters are UTF-8 encoded.  May be negated.

@item ixon
@opindex ixon
@kindex C-s/C-q flow control
@cindex XON/XOFF flow control
Enable XON/XOFF flow control (that is, @kbd{Ctrl-S}/@kbd{Ctrl-Q}).  May
be negated.

@item ixoff
@itemx tandem
@opindex ixoff
@opindex tandem
@cindex software flow control
@cindex flow control, software
Enable sending of @code{stop} character when the system input buffer
is almost full, and @code{start} character when it becomes almost
empty again.  May be negated.

@item iuclc
@opindex iuclc
@cindex uppercase, translating to lowercase
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.

@item ixany
@opindex ixany
Allow any character to restart output (only the start character
if negated).  Non-POSIX@.  May be negated.

@item imaxbel
@opindex imaxbel
@cindex beeping at input buffer full
Enable beeping and not flushing input buffer if a character arrives
when the input buffer is full.  Non-POSIX@.  May be negated.
@end table


@node Output
@subsection Output settings

@cindex output settings
These settings control operations on data sent to the terminal.

@table @samp
@item opost
@opindex opost
Postprocess output.  May be negated.

@item olcuc
@opindex olcuc
@cindex lowercase, translating to output
Translate lowercase characters to uppercase.  Non-POSIX@.  May be
negated.  (Note ouclc is not currently implemented.)

@item ocrnl
@opindex ocrnl
@cindex return, translating to newline
Translate carriage return to newline.  Non-POSIX@.  May be negated.

@item onlcr
@opindex onlcr
@cindex newline, translating to crlf
Translate newline to carriage return-newline.  Non-POSIX@.  May be
negated.

@item onocr
@opindex onocr
Do not print carriage returns in the first column.  Non-POSIX@.
May be negated.

@item onlret
@opindex onlret
Newline performs a carriage return.  Non-POSIX@.  May be negated.

@item ofill
@opindex ofill
@cindex pad instead of timing for delaying
Use fill (padding) characters instead of timing for delays.
Non-POSIX@.
May be negated.

@item ofdel
@opindex ofdel
@cindex pad character
Use ASCII DEL characters for fill instead of
ASCII NUL characters.  Non-POSIX@.
May be negated.

@item nl1
@itemx nl0
@opindex nl@var{n}
Newline delay style.  Non-POSIX.

@item cr3
@itemx cr2
@itemx cr1
@itemx cr0
@opindex cr@var{n}
Carriage return delay style.  Non-POSIX.

@item tab3
@itemx tab2
@itemx tab1
@itemx tab0
@opindex tab@var{n}
Horizontal tab delay style.  Non-POSIX.

@item bs1
@itemx bs0
@opindex bs@var{n}
Backspace delay style.  Non-POSIX.

@item vt1
@itemx vt0
@opindex vt@var{n}
Vertical tab delay style.  Non-POSIX.

@item ff1
@itemx ff0
@opindex ff@var{n}
Form feed delay style.  Non-POSIX.
@end table


@node Local
@subsection Local settings

@cindex local settings

@table @samp
@item isig
@opindex isig
Enable @code{interrupt}, @code{quit}, and @code{suspend} special
characters.  May be negated.

@item icanon
@opindex icanon
Enable @code{erase}, @code{kill}, @code{werase}, and @code{rprnt}
special characters.  May be negated.

@item iexten
@opindex iexten
Enable non-POSIX special characters.  May be negated.

@item echo
@opindex echo
Echo input characters.  May be negated.

@item echoe
@itemx crterase
@opindex echoe
@opindex crterase
Echo @code{erase} characters as backspace-space-backspace.  May be
negated.

@item echok
@opindex echok
@cindex newline echoing after @code{kill}
Echo a newline after a @code{kill} character.  May be negated.

@item echonl
@opindex echonl
@cindex newline, echoing
Echo newline even if not echoing other characters.  May be negated.

@item noflsh
@opindex noflsh
@cindex flushing, disabling
Disable flushing after @code{interrupt} and @code{quit} special
characters.  May be negated.

@item xcase
@opindex xcase
@cindex case translation
Enable input and output of uppercase characters by preceding their
lowercase equivalents with @samp{\}, when @code{icanon} is set.
Non-POSIX@.  May be negated.

@item tostop
@opindex tostop
@cindex background jobs, stopping at terminal write
Stop background jobs that try to write to the terminal.  Non-POSIX@.
May be negated.

@item echoprt
@itemx prterase
@opindex echoprt
@opindex prterase
Echo erased characters backward, between @samp{\} and @samp{/}.
Non-POSIX@.  May be negated.

@item echoctl
@itemx ctlecho
@opindex echoctl
@opindex ctlecho
@cindex control characters, using @samp{^@var{c}}
@cindex hat notation for control characters
Echo control characters in hat notation (@samp{^@var{c}}) instead
of literally.  Non-POSIX@.  May be negated.

@item echoke
@itemx crtkill
@opindex echoke
@opindex crtkill
Echo the @code{kill} special character by erasing each character on
the line as indicated by the @code{echoprt} and @code{echoe} settings,
instead of by the @code{echoctl} and @code{echok} settings.
Non-POSIX@.
May be negated.

@item extproc
@opindex extproc
Enable @samp{LINEMODE}, which is used to avoid echoing
each character over high latency links.  See also
@uref{https://tools.ietf.org/search/rfc1116, Internet RFC 1116}.
Non-POSIX@.
May be negated.

@item flusho
@opindex flusho
Discard output.
Note this setting is currently ignored on GNU/Linux systems.
Non-POSIX@.
May be negated.
@end table


@node Combination
@subsection Combination settings

@cindex combination settings
Combination settings:

@table @samp
@item evenp
@opindex evenp
@itemx parity
@opindex parity
Same as @code{parenb -parodd cs7}.  May be negated.  If negated, same
as @code{-parenb cs8}.

@item oddp
@opindex oddp
Same as @code{parenb parodd cs7}.  May be negated.  If negated, same
as @code{-parenb cs8}.

@item nl
@opindex nl
Same as @code{-icrnl -onlcr}.  May be negated.  If negated, same as
@code{icrnl -inlcr -igncr onlcr -ocrnl -onlret}.

@item ek
@opindex ek
Reset the @code{erase} and @code{kill} special characters to their default
values.

@item sane
@opindex sane
Same as:

@c This is too long to write inline.
@example
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
@end example

@noindent
and also sets all special characters to their default values.

@item cooked
@opindex cooked
Same as @code{brkint ignpar istrip icrnl ixon opost isig icanon}, plus
sets the @code{eof} and @code{eol} characters to their default values
if they are the same as the @code{min} and @code{time} characters.
May be negated.  If negated, same as @code{raw}.

@item raw
@opindex raw
Same as:

@example
-ignbrk -brkint -ignpar -parmrk -inpck -istrip
-inlcr -igncr -icrnl -ixon -ixoff -icanon -opost
-isig -iuclc -ixany -imaxbel -xcase min 1 time 0
@end example

@noindent
May be negated.  If negated, same as @code{cooked}.

@item cbreak
@opindex cbreak
Same as @option{-icanon}.  May be negated.  If negated, same as
@code{icanon}.

@item pass8
@opindex pass8
@cindex eight-bit characters
Same as @code{-parenb -istrip cs8}.  May be negated.  If negated,
same as @code{parenb istrip cs7}.

@item litout
@opindex litout
Same as @option{-parenb -istrip -opost cs8}.  May be negated.
If negated, same as @code{parenb istrip opost cs7}.

@item decctlq
@opindex decctlq
Same as @option{-ixany}.  Non-POSIX@.  May be negated.

@item tabs
@opindex tabs
Same as @code{tab0}.  Non-POSIX@.  May be negated.  If negated, same
as @code{tab3}.

@item lcase
@itemx LCASE
@opindex lcase
@opindex LCASE
Same as @code{xcase iuclc olcuc}.  Non-POSIX@.  May be negated.
(Used for terminals with uppercase characters only.)

@item crt
@opindex crt
Same as @code{echoe echoctl echoke}.

@item dec
@opindex dec
Same as @code{echoe echoctl echoke -ixany intr ^C erase ^? kill C-u}.
@end table


@node Characters
@subsection Special characters

@cindex special characters
@cindex characters, special

The special characters' default values vary from system to system.
They are set with the syntax @samp{name value}, where the names are
listed below and the value can be given either literally, in hat
notation (@samp{^@var{c}}), or as an integer which may start with
@samp{0x} to indicate hexadecimal, @samp{0} to indicate octal, or
any other digit to indicate decimal.

@cindex disabling special characters
@kindex u@r{, and disabling special characters}
For GNU stty, giving a value of @code{^-} or @code{undef} disables that
special character.  (This is incompatible with Ultrix @command{stty},
which uses  a value of @samp{u} to disable a special character.  GNU
@command{stty} treats a value @samp{u} like any other, namely to set that
special character to @key{U}.)

@table @samp

@item intr
@opindex intr
Send an interrupt signal.

@item quit
@opindex quit
Send a quit signal.

@item erase
@opindex erase
Erase the last character typed.

@item kill
@opindex kill
Erase the current line.

@item eof
@opindex eof
Send an end of file (terminate the input).

@item eol
@opindex eol
End the line.

@item eol2
@opindex eol2
Alternate character to end the line.  Non-POSIX.

@item discard
@opindex discard
@opindex flush
Alternate character to toggle discarding of output.  Non-POSIX.

@item swtch
@opindex swtch
Switch to a different shell layer.  Non-POSIX.

@item status
@opindex status
Send an info signal.  Not currently supported on GNU/Linux.  Non-POSIX.

@item start
@opindex start
Restart the output after stopping it.

@item stop
@opindex stop
Stop the output.

@item susp
@opindex susp
Send a terminal stop signal.

@item dsusp
@opindex dsusp
Send a terminal stop signal after flushing the input.  Non-POSIX.

@item rprnt
@opindex rprnt
Redraw the current line.  Non-POSIX.

@item werase
@opindex werase
Erase the last word typed.  Non-POSIX.

@item lnext
@opindex lnext
Enter the next character typed literally, even if it is a special
character.  Non-POSIX.
@end table


@node Special
@subsection Special settings

@cindex special settings

@table @samp
@item min @var{n}
@opindex min
Set the minimum number of characters that will satisfy a read until
the time value has expired, when @option{-icanon} is set.

@item time @var{n}
@opindex time
Set the number of tenths of a second before reads time out if the minimum
number of characters have not been read, when @option{-icanon} is set.

@item ispeed @var{n}
@opindex ispeed
Set the input speed to @var{n}.

@item ospeed @var{n}
@opindex ospeed
Set the output speed to @var{n}.

@item rows @var{n}
@opindex rows
Tell the tty kernel driver that the terminal has @var{n} rows.
Non-POSIX.

@item cols @var{n}
@itemx columns @var{n}
@opindex cols
@opindex columns
Tell the kernel that the terminal has @var{n} columns.  Non-POSIX.

@item drain
@opindex drain
@cindex nonblocking @command{stty} setting
Apply settings after first waiting for pending output to be transmitted.
This is enabled by default for GNU @command{stty}.
It is useful to disable this option
in cases where the system may be in a state where serial transmission
is not possible.
For example, if the system has received the @samp{DC3} character
with @code{ixon} (software flow control) enabled, then @command{stty} would
block without @code{-drain} being specified.
May be negated. Non-POSIX.

@item size
@opindex size
@vindex LINES
@vindex COLUMNS
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 @env{LINES} and @env{COLUMNS}
instead; however, GNU @command{stty} does not know anything about them.)
Non-POSIX.

@item line @var{n}
@opindex line
Use line discipline @var{n}.  Non-POSIX.

@item speed
@opindex speed
Print the terminal speed.

@item @var{n}
@cindex baud rate, setting
Set the input and output speeds to @var{n}.  @var{n} can be one of: 0
50 75 110 134 134.5 150 200 300 600 1200 1800 2400 4800 9600 19200
38400 @code{exta} @code{extb}.  @code{exta} is the same as 19200;
@code{extb} is the same as 38400.  Many systems, including GNU/Linux,
support higher speeds.  The @command{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 @option{-clocal} is set.
@end table


@node printenv invocation
@section @command{printenv}: Print all or some environment variables

@pindex printenv
@cindex printing all or some environment variables
@cindex environment variables, printing

@command{printenv} prints environment variable values.  Synopsis:

@example
printenv [@var{option}] [@var{variable}]@dots{}
@end example

If no @var{variable}s are specified, @command{printenv} prints the value of
every environment variable.  Otherwise, it prints the value of each
@var{variable} that is set, and nothing for those that are not set.

The program accepts the following option.  Also see @ref{Common options}.

@table @samp

@optNull

@end table

@cindex exit status of @command{printenv}
Exit status:

@display
0 if all variables specified were found
1 if at least one specified variable was not found
2 if a write error occurred
@end display


@node tty invocation
@section @command{tty}: Print file name of terminal on standard input

@pindex tty
@cindex print terminal file name
@cindex terminal file name, printing

@command{tty} prints the file name of the terminal connected to its standard
input.  It prints @samp{not a tty} if standard input is not a terminal.
Synopsis:

@example
tty [@var{option}]@dots{}
@end example

The program accepts the following option.  Also see @ref{Common options}.

@table @samp

@item -s
@itemx --silent
@itemx --quiet
@opindex -s
@opindex --silent
@opindex --quiet
Print nothing; only return an exit status.

@end table

@cindex exit status of @command{tty}
Exit status:

@display
0 if standard input is a terminal
1 if standard input is a non-terminal file
2 if given incorrect arguments
3 if a write error occurs
@end display


@node User information
@chapter User information

@cindex user information, commands for
@cindex commands for printing user information

This section describes commands that print user-related information:
logins, groups, and so forth.

@menu
* id invocation::               Print user identity.
* logname invocation::          Print current login name.
* whoami invocation::           Print effective user ID.
* groups invocation::           Print group names a user is in.
* users invocation::            Print login names of users currently logged in.
* who invocation::              Print who is currently logged in.
@end menu


@node id invocation
@section @command{id}: Print user identity

@pindex id
@cindex real user and group IDs, printing
@cindex effective user and group IDs, printing
@cindex printing real and effective user and group IDs

@command{id} prints information about the given user, or the process
running it if no user is specified.  Synopsis:

@example
id [@var{option}]@dots{} [@var{user}]@dots{}
@end example

@var{user} can be either a user ID or a name, with name look-up
taking precedence unless the ID is specified with a leading @samp{+}.
@xref{Disambiguating names and IDs}.

@vindex POSIXLY_CORRECT
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 @env{POSIXLY_CORRECT} environment variable is not set,
then print @samp{context=@var{c}}, where @var{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 @command{id} to print only part of the above information.
Also see @ref{Common options}.

@table @samp
@item -g
@itemx --group
@opindex -g
@opindex --group
Print only the group ID.

@item -G
@itemx --groups
@opindex -G
@opindex --groups
Print only the group ID and the supplementary groups.

@item -n
@itemx --name
@opindex -n
@opindex --name
Print the user or group name instead of the ID number.  Requires
@option{-u}, @option{-g}, or @option{-G}.

@item -r
@itemx --real
@opindex -r
@opindex --real
Print the real, instead of effective, user or group ID@.  Requires
@option{-u}, @option{-g}, or @option{-G}.

@item -u
@itemx --user
@opindex -u
@opindex --user
Print only the user ID.

@item -Z
@itemx --context
@opindex -Z
@opindex --context
@cindex SELinux
@cindex security 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.

@item -z
@itemx --zero
@opindex -z
@opindex --zero
Delimit output items with ASCII NUL characters.
This option is not permitted when using the default format.
When multiple users are specified, and the @option{--groups} option
is also in effect, groups are delimited with a single NUL character,
while users are delimited with two NUL characters.

Example:
@example
$ id -Gn --zero
users <NUL> devs <NUL>
@end example

@end table

@macro primaryAndSupplementaryGroups{cmd,arg}
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, @command{\cmd\}
will not reflect your changes within your existing login session.
Running @command{\cmd\} with a \arg\ causes the user and group
database to be consulted afresh, and so will give a different result.
@end macro
@primaryAndSupplementaryGroups{id,user argument}

@exitstatus

@node logname invocation
@section @command{logname}: Print current login name

@pindex logname
@cindex printing user's login name
@cindex login name, printing
@cindex user name, printing

@flindex utmp
@command{logname} prints the calling user's name, as found in a
system-maintained file (often @file{/var/run/utmp} or
@file{/etc/utmp}), and exits with a status of 0.  If there is no entry
for the calling process, @command{logname} prints
an error message and exits with a status of 1.

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.

@exitstatus


@node whoami invocation
@section @command{whoami}: Print effective user name

@pindex whoami
@cindex effective user name, printing
@cindex printing the effective user ID

@command{whoami} prints the user name associated with the current
effective user ID@.  It is equivalent to the command @samp{id -un}.

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.

@exitstatus


@node groups invocation
@section @command{groups}: Print group names a user is in

@pindex groups
@cindex printing groups a user is in
@cindex supplementary groups, printing

@command{groups} prints the names of the primary and any supplementary
groups for each given @var{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:

@example
groups [@var{username}]@dots{}
@end example

The group lists are equivalent to the output of the command @samp{id -Gn}.

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.

@primaryAndSupplementaryGroups{groups,list of users}

@exitstatus

@node users invocation
@section @command{users}: Print login names of users currently logged in

@pindex users
@cindex printing current usernames
@cindex usernames, printing current

@cindex login sessions, printing users with
@command{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:

@example
users [@var{file}]
@end example

@flindex utmp
@flindex wtmp
With no @var{file} argument, @command{users} extracts its information from
a system-maintained file (often @file{/var/run/utmp} or
@file{/etc/utmp}).  If a file argument is given, @command{users} uses
that file instead.  A common choice is @file{/var/log/wtmp}.

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.

The @command{users} command is installed only on platforms with the
POSIX @code{<utmpx.h>} include file or equivalent, so portable scripts
should not rely on its existence on non-POSIX platforms.

@exitstatus


@node who invocation
@section @command{who}: Print who is currently logged in

@pindex who
@cindex printing current user information
@cindex information, about current users

@command{who} prints information about users who are currently logged on.
Synopsis:

@example
@command{who} [@var{option}] [@var{file}] [am i]
@end example

@cindex terminal lines, currently used
@cindex login time
@cindex remote hostname
If given no non-option arguments, @command{who} prints the following
information for each user currently logged on: login name, terminal
line, login time, and remote hostname or X display.

@flindex utmp
@flindex wtmp
If given one non-option argument, @command{who} uses that instead of
a default system-maintained file (often @file{/var/run/utmp} or
@file{/etc/utmp}) as the name of the file containing the record of
users logged on.  @file{/var/log/wtmp} is commonly given as an argument
to @command{who} to look at who has previously logged on.

@opindex am i
@opindex who am i
If given two non-option arguments, @command{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 @samp{am
i}, as in @samp{who am i}.

@vindex TZ
Timestamps are listed according to the time zone rules specified by
the @env{TZ} environment variable, or by the system default rules if
@env{TZ} is not set.  @xref{TZ Variable,, Specifying the Time Zone
with @env{TZ}, libc, The GNU C Library Reference Manual}.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -a
@itemx --all
@opindex -a
@opindex --all
Same as @samp{-b -d --login -p -r -t -T -u}.

@item -b
@itemx --boot
@opindex -b
@opindex --boot
Print the date and time of last system boot.

@item -d
@itemx --dead
@opindex -d
@opindex --dead
Print information corresponding to dead processes.

@item -H
@itemx --heading
@opindex -H
@opindex --heading
Print a line of column headings.

@item -l
@itemx --login
@opindex -l
@opindex --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 @samp{LOGIN}.

@item --lookup
@opindex --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.

@item -m
@opindex -m
Same as @samp{who am i}.

@item -p
@itemx --process
@opindex -p
@opindex --process
List active processes spawned by init.

@item -q
@itemx --count
@opindex -q
@opindex --count
Print only the login names and the number of users logged on.
Overrides all other options.

@item -r
@itemx --runlevel
@opindex -r
@opindex --runlevel
Print the current (and maybe previous) run-level of the init process.

@item -s
@opindex -s
Ignored; for compatibility with other versions of @command{who}.

@item -t
@itemx --time
@opindex -t
@opindex --time
Print last system clock change.

@item -u
@opindex -u
@cindex idle time
After the login time, print the number of hours and minutes that the
user has been idle.  @samp{.} means the user was active in the last minute.
@samp{old} means the user has been idle for more than 24 hours.

@item -w
@itemx -T
@itemx --mesg
@itemx --message
@itemx --writable
@opindex -w
@opindex -T
@opindex --mesg
@opindex --message
@opindex --writable
@cindex message status
@pindex write@r{, allowed}
After each login name print a character indicating the user's message status:

@display
@samp{+} allowing @code{write} messages
@samp{-} disallowing @code{write} messages
@samp{?} cannot find terminal device
@end display

@end table

The @command{who} command is installed only on platforms with the
POSIX @code{<utmpx.h>} include file or equivalent, so portable scripts
should not rely on its existence on non-POSIX platforms.

@exitstatus


@node System context
@chapter System context

@cindex system context
@cindex context, system
@cindex commands for system context

This section describes commands that print or change system-wide
information.

@menu
* date invocation::             Print or set system date and time.
* arch invocation::             Print machine hardware name.
* nproc invocation::            Print the number of processors.
* uname invocation::            Print system information.
* hostname invocation::         Print or set system name.
* hostid invocation::           Print numeric host identifier.
* uptime invocation::           Print system uptime and load.
@end menu

@node date invocation
@section @command{date}: Print or set system date and time

@pindex date
@cindex time, printing or setting
@cindex printing the current time

Synopses:

@example
date [@var{option}]@dots{} [+@var{format}]
date [-u|--utc|--universal] @c this avoids a newline in the output
[ MMDDhhmm[[CC]YY][.ss] ]
@end example

The @command{date} command displays the date and time.
With the @option{--set} (@option{-s}) option, or with
@samp{MMDDhhmm[[CC]YY][.ss]},
it sets the date and time.

@vindex LC_TIME
Invoking @command{date} with no @var{format} argument is equivalent to invoking
it with a default format that depends on the @env{LC_TIME} locale category.
In the default C locale, this format is @samp{'+%a %b %e %H:%M:%S %Z %Y'},
so the output looks like @samp{Thu Jul @ 9 17:00:00 EDT 2020}.

@vindex TZ
Normally, @command{date} uses the time zone rules indicated by the
@env{TZ} environment variable, or the system default rules if @env{TZ}
is not set.  @xref{TZ Variable,, Specifying the Time Zone with
@env{TZ}, libc, The GNU C Library Reference Manual}.

@findex strftime @r{and @command{date}}
@cindex time formats
@cindex formatting times
If given an argument that starts with a @samp{+}, @command{date} prints the
current date and time (or the date and time specified by the
@option{--date} option, see below) in the format defined by that argument,
which is similar to that of the @code{strftime} function.  Except for
conversion specifiers, which start with @samp{%}, characters in the
format string are printed unchanged.  The conversion specifiers are
described below.

@exitstatus

@menu
* Time conversion specifiers::     %[HIklMNpPrRsSTXzZ]
* Date conversion specifiers::     %[aAbBcCdDeFgGhjmuUVwWxyY]
* Literal conversion specifiers::  %[%nt]
* Padding and other flags::        Pad with zeros, spaces, etc.
* Setting the time::               Changing the system clock.
* Options for date::               Instead of the current time.
@detailmenu
* Date input formats::             Specifying date strings.
@end detailmenu
* Examples of date::               Examples.
@end menu

@node Time conversion specifiers
@subsection Time conversion specifiers

@cindex time conversion specifiers
@cindex conversion specifiers, time

@command{date} conversion specifiers related to times.

@table @samp
@item %H
hour (@samp{00}@dots{}@samp{23})
@item %I
hour (@samp{01}@dots{}@samp{12})
@item %k
hour, space padded (@samp{ 0}@dots{}@samp{23}); equivalent to @samp{%_H}@.
This is a GNU extension.
@item %l
hour, space padded (@samp{ 1}@dots{}@samp{12}); equivalent to @samp{%_I}@.
This is a GNU extension.
@item %M
minute (@samp{00}@dots{}@samp{59})
@item %N
nanoseconds (@samp{000000000}@dots{}@samp{999999999}).
This is a GNU extension.
@item %p
locale's equivalent of either @samp{AM} or @samp{PM};
blank in many locales.
Noon is treated as @samp{PM} and midnight as @samp{AM}.
@item %P
like @samp{%p}, except lower case.
This is a GNU extension.
@item %r
locale's 12-hour clock time (e.g., @samp{11:11:04 PM})
@item %R
24-hour hour and minute.  Same as @samp{%H:%M}.
@item %s
@cindex Epoch, seconds since
@cindex seconds since the Epoch
@cindex beginning of time
@cindex leap seconds
seconds since the Epoch, i.e., since 1970-01-01 00:00 UTC@.
Leap seconds are not counted unless leap second support is available.
@xref{%s-examples}, for examples.
This is a GNU extension.
@item %S
@cindex leap seconds
second (@samp{00}@dots{}@samp{60}).
This may be @samp{60} if leap seconds are supported.
@item %T
24-hour hour, minute, and second.  Same as @samp{%H:%M:%S}.
@item %X
locale's time representation (e.g., @samp{23:13:48})
@item %z
Four-digit numeric time zone, e.g., @samp{-0600} or @samp{+0530}, or
@samp{-0000} 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 @env{TZ} environment variable.  A time zone is not determinable if
its numeric offset is zero and its abbreviation begins with @samp{-}.
The time (and optionally, the time zone rules) can be overridden
by the @option{--date} option.
@item %:z
Numeric time zone with @samp{:}, e.g., @samp{-06:00} or
@samp{+05:30}), or @samp{-00:00} if no time zone is determinable.
This is a GNU extension.
@item %::z
Numeric time zone to the nearest second with @samp{:} (e.g.,
@samp{-06:00:00} or @samp{+05:30:00}), or @samp{-00:00:00} if no time zone is
determinable.
This is a GNU extension.
@item %:::z
Numeric time zone with @samp{:} using the minimum necessary precision
(e.g., @samp{-06}, @samp{+05:30}, or @samp{-04:56:02}), or @samp{-00} if
no time zone is determinable.
This is a GNU extension.
@item %Z
alphabetic time zone abbreviation (e.g., @samp{EDT}), or nothing if no
time zone is determinable.  See @samp{%z} for how it is determined.
@end table


@node Date conversion specifiers
@subsection Date conversion specifiers

@cindex date conversion specifiers
@cindex conversion specifiers, date

@command{date} conversion specifiers related to dates.

@table @samp
@item %a
locale's abbreviated weekday name (e.g., @samp{Sun})
@item %A
locale's full weekday name, variable length (e.g., @samp{Sunday})
@item %b
locale's abbreviated month name (e.g., @samp{Jan})
@item %B
locale's full month name, variable length (e.g., @samp{January})
@item %c
locale's date and time (e.g., @samp{Thu Mar @ 3 23:05:25 2020})
@item %C
century.  This is like @samp{%Y}, except the last two digits are omitted.
For example, it is @samp{20} if @samp{%Y} is @samp{2019},
and is @samp{-0} if @samp{%Y} is @samp{-001}.
It is normally at least two characters, but it may be more.
@item %d
day of month (e.g., @samp{01})
@item %D
date; same as @samp{%m/%d/%y}
@item %e
day of month, space padded; same as @samp{%_d}
@item %F
full date in ISO 8601 format; like @samp{%+4Y-%m-%d}
except that any flags or field width override the @samp{+}
and (after subtracting 6) the @samp{4}.
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@dots{}9999.
@item %g
year corresponding to the ISO week number, but without the century
(range @samp{00} through @samp{99}).  This has the same format and value
as @samp{%y}, except that if the ISO week number (see
@samp{%V}) belongs
to the previous or next year, that year is used instead.
@item %G
year corresponding to the ISO week number.  This has the
same format and value as @samp{%Y}, except that if the ISO
week number (see
@samp{%V}) belongs to the previous or next year, that year is used
instead.
It is normally useful only if @samp{%V} is also used;
for example, the format @samp{%G-%m-%d} is probably a mistake,
since it combines the ISO week number year with the conventional month and day.
@item %h
same as @samp{%b}
@item %j
day of year (@samp{001}@dots{}@samp{366})
@item %m
month (@samp{01}@dots{}@samp{12})
@item %q
quarter of year (@samp{1}@dots{}@samp{4})
@item %u
day of week (@samp{1}@dots{}@samp{7}) with @samp{1} corresponding to Monday
@item %U
week number of year, with Sunday as the first day of the week
(@samp{00}@dots{}@samp{53}).
Days in a new year preceding the first Sunday are in week zero.
@item %V
ISO week number, that is, the
week number of year, with Monday as the first day of the week
(@samp{01}@dots{}@samp{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.)
@item %w
day of week (@samp{0}@dots{}@samp{6}) with 0 corresponding to Sunday
@item %W
week number of year, with Monday as first day of week
(@samp{00}@dots{}@samp{53}).
Days in a new year preceding the first Monday are in week zero.
@item %x
locale's date representation (e.g., @samp{12/31/99})
@item %y
last two digits of year (@samp{00}@dots{}@samp{99})
@item %Y
year.  This is normally at least four characters, but it may be more.
Year @samp{0000} precedes year @samp{0001}, and year @samp{-001}
precedes year @samp{0000}.
@end table


@node Literal conversion specifiers
@subsection Literal conversion specifiers

@cindex literal conversion specifiers
@cindex conversion specifiers, literal

@command{date} conversion specifiers that produce literal strings.

@table @samp
@item %%
a literal %
@item %n
a newline
@item %t
a horizontal tab
@end table


@node Padding and other flags
@subsection Padding and other flags

@cindex numeric field padding
@cindex padding of numeric fields
@cindex fields, padding numeric

Unless otherwise specified, @command{date} normally pads numeric fields
with zeros, so that, for
example, numeric months are always output as two digits.
Most numeric fields are padded on the left.
However, nanoseconds are padded on the right since they are commonly
used after decimal points in formats like @samp{%s.%-N}.
Also, seconds since the Epoch are not padded
since there is no natural width for them.

The following optional flags can appear after the @samp{%}:

@table @samp
@item -
(hyphen) Do not pad the field; useful if the output is intended for
human consumption.
This is a GNU extension.
As a special case, @samp{%-N} outputs only enough trailing digits to
not lose information, assuming that the timestamp's resolution is the
same as the current hardware clock.  For example, if the hardware
clock resolution is 1 microsecond, @samp{%s.%-N} outputs something
like @samp{1640890100.395710}.

@item _
(underscore) Pad with spaces; useful if you need a fixed
number of characters in the output, but zeros are too distracting.
This is a GNU extension.
@item 0
(zero) Pad with zeros even if the conversion specifier
would normally pad with spaces.
@item +
Pad with zeros, like @samp{0}.  In addition, precede any year number
with @samp{+} if it exceeds 9999 or if its field width exceeds 4;
similarly, precede any century number with @samp{+} if it exceeds 99
or if its field width exceeds 2.  This supports ISO 8601 formats
for dates far in the future; for example, the command @code{date
--date=12019-02-25 +%+13F} outputs the string @samp{+012019-02-25}.
@item ^
Use upper case characters if possible.
This is a GNU extension.
@item #
Use opposite case characters if possible.
A field that is normally upper case becomes lower case, and vice versa.
This is a GNU extension.
@end table

@noindent
Here are some examples of padding:

@example
date +%d/%m -d "Feb 1"
@result{} 01/02
date +%-d/%-m -d "Feb 1"
@result{} 1/2
date +%_d/%_m -d "Feb 1"
@result{}  1/ 2
@end example

You can optionally 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 normally written right adjusted and padded to the given
size.  For example, @samp{%9B} prints the right adjusted month name in
a field of width 9.  Nanoseconds are left adjusted, and are truncated
or padded to the field width.

An optional modifier can follow the optional flag and width
specification.  The modifiers are:

@table @samp
@item E
Use the locale's alternate representation for date and time.  This
modifier applies to the @samp{%c}, @samp{%C}, @samp{%x}, @samp{%X},
@samp{%y} and @samp{%Y} conversion specifiers.  In a Japanese locale, for
example, @samp{%Ex} might yield a date format based on the Japanese
Emperors' reigns.

@item O
Use the locale's alternate numeric symbols for numbers.  This modifier
applies only to numeric conversion specifiers.
@end table

If the format supports the modifier but no alternate representation
is available, it is ignored.

POSIX specifies the behavior of flags and field widths only for
@samp{%C}, @samp{%F}, @samp{%G}, and @samp{%Y} (all without
modifiers), and requires a flag to be present if and only if a field
width is also present.  Other combinations of flags, field widths and
modifiers are GNU extensions.


@node Setting the time
@subsection Setting the time

@cindex setting the time
@cindex time setting
@cindex appropriate privileges

You must have appropriate privileges to set the
system clock.  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.

To set the clock, you can use the @option{--set} (@option{-s}) option
(@pxref{Options for date}).  To set the clock without using GNU
extensions, you can give @command{date} an argument of the form
@samp{MMDDhhmm[[CC]YY][.ss]} where each two-letter
component stands for two digits with the following meanings:

@table @var
@item MM
month
@item DD
day within month
@item hh
hour
@item mm
minute
@item CC
first two digits of year (optional)
@item YY
last two digits of year (optional)
@item ss
second (optional)
@end table

Note, the @option{--date} and @option{--set} options may not be used with an
argument in the above format.  The @option{--universal} option may be used
with such an argument to indicate that the specified date and time are
relative to Universal Time rather than to the local time zone.


@node Options for date
@subsection Options for @command{date}

@cindex @command{date} options
@cindex options for @command{date}

The program accepts the following options.  Also see @ref{Common options}.
Except for @option{-u}, these options are all GNU extensions to POSIX.

All options that specify the date to display are mutually exclusive.
I.e.: @option{--date}, @option{--file}, @option{--reference},
@option{--resolution}.

@table @samp

@item -d @var{datestr}
@itemx --date=@var{datestr}
@opindex -d
@opindex --date
@cindex parsing date strings
@cindex date strings, parsing
@cindex arbitrary date strings, parsing
@opindex yesterday
@opindex tomorrow
@opindex next @var{day}
@opindex last @var{day}
Display the date and time specified in @var{datestr} instead of the
current date and time.  @var{datestr} can be in almost any common
format.  It can contain month names, time zones, @samp{am} and @samp{pm},
@samp{yesterday}, etc.  For example, @option{--date="2020-07-21
14:19:13.489392193 +0530"} specifies the instant of time that is
489,392,193 nanoseconds after July 21, 2020 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:
@example
date -d "$(LC_TIME=C date)"
@end example
@xref{Date input formats}.

@item --debug
@opindex --debug
@cindex debugging date strings
@cindex date strings, debugging
@cindex arbitrary date strings, debugging
Annotate the parsed date, display the effective time zone, and warn about
potential misuse.

@item -f @var{datefile}
@itemx --file=@var{datefile}
@opindex -f
@opindex --file
Parse each line in @var{datefile} as with @option{-d} and display the
resulting date and time.  If @var{datefile} is @samp{-}, use standard
input.  This is useful when you have many dates to process, because the
system overhead of starting up the @command{date} executable many times can
be considerable.

@item -I[@var{timespec}]
@itemx --iso-8601[=@var{timespec}]
@opindex -I[@var{timespec}]
@opindex --iso-8601[=@var{timespec}]
Display the date using an ISO 8601 format, @samp{%Y-%m-%d}.

The argument @var{timespec} specifies the number of additional
terms of the time to include.  It can be one of the following:
@table @samp
@item auto
Print just the date.  This is the default if @var{timespec} is omitted.

@item hours
Append the hour of the day to the date.

@item minutes
Append the hours and minutes.

@item seconds
Append the hours, minutes and seconds.

@item ns
Append the hours, minutes, seconds and nanoseconds.
@end table

If showing any time terms, then include the time zone using the format
@samp{%:z}.
@macro dateParseNote
This format is always suitable as input
for the @option{--date} (@option{-d}) and @option{--file}
(@option{-f}) options, regardless of the current locale.
@end macro
@dateParseNote

@item -r @var{file}
@itemx --reference=@var{file}
@opindex -r
@opindex --reference
Display the date and time of the last modification of @var{file},
instead of the current date and time.

@item --resolution
@opindex --resolution
Display the timestamp resolution instead of the time.
Current clock timestamps that are output by @command{date}
are integer multiples of the timestamp resolution.
With this option, the format defaults to @samp{%s.%N}.
For example, if the clock resolution is 1 millsecond,
the output is:

@example
0.001000000
@end example

@item -R
@itemx --rfc-email
@opindex -R
@opindex --rfc-email
Display the date and time using the format @samp{%a, %d %b %Y %H:%M:%S
%z}, evaluated in the C locale so abbreviations are always in English.
For example:

@example
Mon, 09 Jul 2020 17:00:00 -0400
@end example

@opindex --rfc-822
@opindex --rfc-2822
This format conforms to Internet RFCs
@uref{https://tools.ietf.org/search/rfc5322, 5322},
@uref{https://tools.ietf.org/search/rfc2822, 2822} and
@uref{https://tools.ietf.org/search/rfc822, 822}, the
current and previous standards for Internet email.
For compatibility with older versions of @command{date},
@option{--rfc-2822} and @option{--rfc-822} are aliases for
@option{--rfc-email}.

@item --rfc-3339=@var{timespec}
@opindex --rfc-3339=@var{timespec}
Display the date using a format specified by
@uref{https://tools.ietf.org/search/rfc3339, Internet
RFC 3339}.  This is like @option{--iso-8601}, except that a space rather
than a @samp{T} separates dates from times, and a period rather than
a comma separates seconds from subseconds.
@dateParseNote

The argument @var{timespec} specifies how much of the time to include.
It can be one of the following:

@table @samp
@item date
Print just the full-date, e.g., @samp{2020-07-21}.
This is equivalent to the format @samp{%Y-%m-%d}.

@item seconds
Print the full-date and full-time separated by a space, e.g.,
@samp{2020-07-21 04:30:37+05:30}.  The output ends with a numeric
time-offset; here the @samp{+05:30} means that local time is five
hours and thirty minutes east of UTC@.  This is equivalent to
the format @samp{%Y-%m-%d %H:%M:%S%:z}.

@item ns
Like @samp{seconds}, but also print nanoseconds, e.g.,
@samp{2020-07-21 04:30:37.998458565+05:30}.
This is equivalent to the format @samp{%Y-%m-%d %H:%M:%S.%N%:z}.

@end table

@item -s @var{datestr}
@itemx --set=@var{datestr}
@opindex -s
@opindex --set
Set the date and time to @var{datestr}.  See @option{-d} above.
See also @ref{Setting the time}.

@item -u
@itemx --utc
@itemx --universal
@opindex -u
@opindex --utc
@opindex --universal
@cindex Coordinated Universal Time
@cindex UTC
@cindex Greenwich Mean Time
@cindex GMT
@cindex leap seconds
@vindex TZ
@cindex Universal Time
Use Universal Time by operating as if the
@env{TZ} environment variable were set to the string @samp{UTC0}.
UTC stands for Coordinated Universal Time, established in 1960.
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.
@end table


@node Examples of date
@subsection Examples of @command{date}

@cindex examples of @command{date}

Here are a few examples.  Also see the documentation for the @option{-d}
option in the previous section.

@itemize @bullet

@item
To print the date of the day before yesterday:

@example
date --date='2 days ago'
@end example

@item
To print the date of the day three months and one day hence:

@example
date --date='3 months 1 day'
@end example

@item
To print the day of year of Christmas in the current year:

@example
date --date='25 Dec' +%j
@end example

@item
To print the current full month name and the day of the month:

@example
date '+%B %d'
@end example

But this may not be what you want because for the first nine days of
the month, the @samp{%d} expands to a zero-padded two-digit field,
for example @samp{date -d 1may '+%B %d'} will print @samp{May 01}.

@item
To print a date without the leading zero for one-digit days
of the month, you can use the (GNU extension)
@samp{-} flag to suppress
the padding altogether:

@example
date -d 1may '+%B %-d'
@end example

@item
To print the current date and time in the format required by many
non-GNU versions of @command{date} when setting the system clock:

@example
date +%m%d%H%M%Y.%S
@end example

@item
To set the system clock forward by two minutes:

@example
date --set='+2 minutes'
@end example

@item
To print the date in Internet RFC 5322 format,
use @samp{date --rfc-email}.  Here is some example output:

@example
Tue, 09 Jul 2020 19:00:37 -0400
@end example

@anchor{%s-examples}
@item
To convert a date string to the number of seconds since the Epoch
(which is 1970-01-01 00:00 UTC), use the @option{--date} option with
the @samp{%s} format.  That can be useful in sorting and/or graphing
and/or comparing data by date.  The following command outputs the
number of the seconds since the Epoch for the time two minutes after the
Epoch:

@example
date --date='1970-01-01 00:02:00 +0000' +%s
120
@end example

If you do not specify time zone information in the date string,
@command{date} uses your computer's idea of the time zone when
interpreting the string.  For example, if your computer's time zone is
that of Cambridge, Massachusetts, which was then 5 hours (i.e., 18,000
seconds) behind UTC:

@example
# local time zone used
date --date='1970-01-01 00:02:00' +%s
18120
@end example

@item
If you're sorting or graphing dated data, your raw date values may be
represented as seconds since the Epoch.  But few people can look at
the date @samp{1577836800} and casually note ``Oh, that's the first
second of the year 2020 in Greenwich, England.''

@example
date --date='2020-01-01 UTC' +%s
1577836800
@end example

An alternative is to use the @option{--utc} (@option{-u}) option.
Then you may omit @samp{UTC} from the date string.  Although this
produces the same result for @samp{%s} and many other format sequences,
with a time zone offset different from zero, it would give a different
result for zone-dependent formats like @samp{%z}.

@example
date -u --date=2020-07-21 +%s
1595289600
@end example

To convert such an unwieldy number of seconds back to
a more readable form, use a command like this:

@example
date -d @@1595289600 +"%F %T %z"
2020-07-20 20:00:00 -0400
@end example

Often it is better to output UTC-relative date and time:

@example
date -u -d @@1595289600 +"%F %T %z"
2020-07-21 00:00:00 +0000
@end example

@item
@cindex leap seconds
Typically the seconds count omits leap seconds, but some systems are
exceptions.  Because leap seconds are not predictable, the mapping
between the seconds count and a future timestamp is not reliable on
the atypical systems that include leap seconds in their counts.

Here is how the two kinds of systems handle the leap second at
the end of the year 2016:

@example
# Typical systems ignore leap seconds:
date --date='2016-12-31 23:59:59 +0000' +%s
1483228799
date --date='2016-12-31 23:59:60 +0000' +%s
date: invalid date '2016-12-31 23:59:60 +0000'
date --date='2017-01-01 00:00:00 +0000' +%s
1483228800
@end example

@example
# Atypical systems count leap seconds:
date --date='2016-12-31 23:59:59 +0000' +%s
1483228825
date --date='2016-12-31 23:59:60 +0000' +%s
1483228826
date --date='2017-01-01 00:00:00 +0000' +%s
1483228827
@end example

@end itemize


@node arch invocation
@section @command{arch}: Print machine hardware name

@pindex arch
@cindex print machine hardware name
@cindex system information, printing

@command{arch} prints the machine hardware name,
and is equivalent to @samp{uname -m}.
Synopsis:

@example
arch [@var{option}]
@end example

The program accepts the @ref{Common options} only.

@command{arch} is not installed by default, so portable scripts should
not rely on its existence.

@exitstatus


@node nproc invocation
@section @command{nproc}: Print the number of available processors

@pindex nproc
@cindex Print the number of processors
@cindex system information, printing

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 @env{OMP_NUM_THREADS} or @env{OMP_THREAD_LIMIT}
environment variables are set, then they will determine the minimum
and maximum returned value respectively.  The result is guaranteed to be
greater than zero.  Synopsis:

@example
nproc [@var{option}]
@end example

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item --all
@opindex --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 @env{OMP_NUM_THREADS} or @env{OMP_THREAD_LIMIT} environment variables
are not honored in this case.

@item --ignore=@var{number}
@opindex --ignore
If possible, exclude this @var{number} of processing units.

@end table

@exitstatus


@node uname invocation
@section @command{uname}: Print system information

@pindex uname
@cindex print system information
@cindex system information, printing

@command{uname} prints information about the machine and operating system
it is run on.  If no options are given, @command{uname} acts as if the
@option{-s} option were given.  Synopsis:

@example
uname [@var{option}]@dots{}
@end example

If multiple options or @option{-a} are given, the selected information is
printed in this order:

@example
@var{kernel-name} @var{nodename} @var{kernel-release} @var{kernel-version}
@var{machine} @var{processor} @var{hardware-platform} @var{operating-system}
@end example

The information may contain internal spaces, so such output cannot be
parsed reliably.  In the following example, @var{kernel-version} is
@samp{#1 SMP Fri Jul 17 17:18:38 UTC 2020}:

@example
uname -a
@result{} Linux dumdum.example.org 5.9.16-200.fc33.x86_64@c
 #1 SMP Mon Dec 21 14:08:22 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
@end example


The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -a
@itemx --all
@opindex -a
@opindex --all
Print all of the below information, except omit the processor type
and the hardware platform name if they are unknown.

@item -i
@itemx --hardware-platform
@opindex -i
@opindex --hardware-platform
@cindex implementation, hardware
@cindex hardware platform
@cindex platform, hardware
Print the hardware platform name
(sometimes called the hardware implementation).
Print @samp{unknown} if this information is not available.
Note this is non-portable (even across GNU/Linux distributions).

@item -m
@itemx --machine
@opindex -m
@opindex --machine
@cindex machine type
@cindex hardware class
@cindex hardware type
Print the machine hardware name (sometimes called the hardware class
or hardware type).

@item -n
@itemx --nodename
@opindex -n
@opindex --nodename
@cindex hostname
@cindex node name
@cindex network node name
Print the network node hostname.

@item -p
@itemx --processor
@opindex -p
@opindex --processor
@cindex host processor type
Print the processor type (sometimes called the instruction set
architecture or ISA).
Print @samp{unknown} if this information is not available.
Note this is non-portable (even across GNU/Linux distributions).

@item -o
@itemx --operating-system
@opindex -o
@opindex --operating-system
@cindex operating system name
Print the name of the operating system.

@item -r
@itemx --kernel-release
@opindex -r
@opindex --kernel-release
@cindex kernel release
@cindex release of kernel
Print the kernel release.

@item -s
@itemx --kernel-name
@opindex -s
@opindex --kernel-name
@cindex kernel name
@cindex name of kernel
Print the kernel name.
POSIX 1003.1-2001 (@pxref{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 @option{-o} or @option{--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.

@item -v
@itemx --kernel-version
@opindex -v
@opindex --kernel-version
@cindex kernel version
@cindex version of kernel
Print the kernel version.

@end table

@exitstatus


@node hostname invocation
@section @command{hostname}: Print or set system name

@pindex hostname
@cindex setting the hostname
@cindex printing the hostname
@cindex system name, printing
@cindex appropriate privileges

With no arguments, @command{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:

@example
hostname [@var{name}]
@end example

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.

@command{hostname} is not installed by default, and other packages
also supply a @command{hostname} command, so portable scripts should
not rely on its existence or on the exact behavior documented above.

@exitstatus


@node hostid invocation
@section @command{hostid}: Print numeric host identifier

@pindex hostid
@cindex printing the host identifier

@command{hostid} prints the numeric identifier of the current host
in hexadecimal.  This command accepts no arguments.
The only options are @option{--help} and @option{--version}.
@xref{Common options}.

For example, here's what it prints on one system I use:

@example
$ hostid
1bac013d
@end example

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.

@command{hostid} is installed only on systems that have the
@code{gethostid} function, so portable scripts should not rely on its
existence.

@exitstatus

@node uptime invocation
@section @command{uptime}: Print system uptime and load

@pindex uptime
@cindex printing the system uptime and load

@command{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 (@command{uptime --help} indicates
the default setting).

The only options are @option{--help} and @option{--version}.
@xref{Common options}.

For example, here's what it prints right now on one system I use:

@example
$ uptime
 14:07  up   3:35,  3 users,  load average: 1.39, 1.15, 1.04
@end example

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 device I/O).  The Linux kernel
includes uninterruptible processes.

@command{uptime} is installed only on platforms with infrastructure
for obtaining the boot time, and other packages also supply an
@command{uptime} command, so portable scripts should not rely on its
existence or on the exact behavior documented above.

@exitstatus

@node SELinux context
@chapter SELinux context

@cindex SELinux context
@cindex SELinux, context
@cindex commands for SELinux context

This section describes commands for operations with SELinux
contexts.

@menu
* chcon invocation::            Change SELinux context of file
* runcon invocation::           Run a command in specified SELinux context
@end menu

@node chcon invocation
@section @command{chcon}: Change SELinux context of file

@pindex chcon
@cindex changing security context
@cindex change SELinux context

@command{chcon} changes the SELinux security context of the selected files.
Synopses:

@example
chcon [@var{option}]@dots{} @var{context} @var{file}@dots{}
chcon [@var{option}]@dots{} [-u @var{user}] [-r @var{role}] [-l @var{range}]@c
 [-t @var{type}] @var{file}@dots{}
chcon [@var{option}]@dots{} --reference=@var{rfile} @var{file}@dots{}
@end example

Change the SELinux security context of each @var{file} to @var{context}.
With @option{--reference}, change the security context of each @var{file}
to that of @var{rfile}.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item --dereference
@opindex --dereference
Do not affect symbolic links but what they refer to; this is the default.

@item -h
@itemx --no-dereference
@opindex -h
@opindex --no-dereference
@cindex no dereference
Affect the symbolic links themselves instead of any referenced file.

@item --reference=@var{rfile}
@opindex --reference
@cindex reference file
Use @var{rfile}'s security context rather than specifying a @var{context} value.

@item -R
@itemx --recursive
@opindex -R
@opindex --recursive
Operate on files and directories recursively.

@item --preserve-root
@opindex --preserve-root
Refuse to operate recursively on the root directory, @file{/},
when used together with the @option{--recursive} option.
@xref{Treating / specially}.

@item --no-preserve-root
@opindex --no-preserve-root
Do not treat the root directory, @file{/}, specially when operating
recursively; this is the default.
@xref{Treating / specially}.

@choptH
@xref{Traversing symlinks}.

@choptL
@xref{Traversing symlinks}.

@choptP
@xref{Traversing symlinks}.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
@cindex diagnostic
Output a diagnostic for every file processed.

@item -u @var{user}
@itemx --user=@var{user}
@opindex -u
@opindex --user
Set user @var{user} in the target security context.

@item -r @var{role}
@itemx --role=@var{role}
@opindex -r
@opindex --role
Set role @var{role} in the target security context.

@item -t @var{type}
@itemx --type=@var{type}
@opindex -t
@opindex --type
Set type @var{type} in the target security context.

@item -l @var{range}
@itemx --range=@var{range}
@opindex -l
@opindex --range
Set range @var{range} in the target security context.

@end table

@exitstatus

@node runcon invocation
@section @command{runcon}: Run a command in specified SELinux context

@pindex runcon
@cindex run with security context


@command{runcon} runs file in specified SELinux security context.

Synopses:
@example
runcon @var{context} @var{command} [@var{args}]
runcon [ -c ] [-u @var{user}] [-r @var{role}] [-t @var{type}]@c
 [-l @var{range}] @var{command} [@var{args}]
@end example

Run @var{command} with completely-specified @var{context}, or with
current or transitioned security context modified by one or more of @var{level},
@var{role}, @var{type} and @var{user}.

If none of @option{-c}, @option{-t}, @option{-u}, @option{-r}, or @option{-l}
is specified, the first argument is used as the complete context.
Any additional arguments after @var{command}
are interpreted as arguments to the command.

With neither @var{context} nor @var{command}, print the current
security context.

@cindex restricted security context
@cindex NO_NEW_PRIVS
Note also the @command{setpriv} command which can be used to set the
NO_NEW_PRIVS bit using @command{setpriv --no-new-privs runcon ...},
thus disallowing usage of a security context with more privileges
than the process would normally have.

@command{runcon} accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -c
@itemx --compute
@opindex -c
@opindex --compute
Compute process transition context before modifying.

@item -u @var{user}
@itemx --user=@var{user}
@opindex -u
@opindex --user
Set user @var{user} in the target security context.

@item -r @var{role}
@itemx --role=@var{role}
@opindex -r
@opindex --role
Set role @var{role} in the target security context.

@item -t @var{type}
@itemx --type=@var{type}
@opindex -t
@opindex --type
Set type @var{type} in the target security context.

@item -l @var{range}
@itemx --range=@var{range}
@opindex -l
@opindex --range
Set range @var{range} in the target security context.

@end table

@cindex exit status of @command{runcon}
Exit status:

@display
126 if @var{command} is found but cannot be invoked
127 if @command{runcon} itself fails or if @var{command} cannot be found
the exit status of @var{command} otherwise
@end display

@node Modified command invocation
@chapter Modified command invocation

@cindex modified command invocation
@cindex invocation of commands, modified
@cindex commands for invoking other commands

This section describes commands that run other commands in some context
different than the current one: a modified environment, as a different
user, etc.

@menu
* chroot invocation::           Modify the root directory.
* env invocation::              Modify environment variables.
* nice invocation::             Modify niceness.
* nohup invocation::            Immunize to hangups.
* stdbuf invocation::           Modify buffering of standard streams.
* timeout invocation::          Run with time limit.
@end menu


@node chroot invocation
@section @command{chroot}: Run a command with a different root directory

@pindex chroot
@cindex running a program in a specified root directory
@cindex root directory, running a program in a specified

@command{chroot} runs a command with a specified root directory.
On many systems, only the super-user can do this.@footnote{However,
some systems (e.g., FreeBSD) can be configured to allow certain regular
users to use the @code{chroot} system call, and hence to run this program.
Also, on Cygwin, anyone can run the @command{chroot} command, because the
underlying function is non-privileged due to lack of support in MS-Windows.
Furthermore, the @command{chroot} command avoids the @code{chroot} system call
when @var{newroot} is identical to the old @file{/} directory for consistency
with systems where this is allowed for non-privileged users.}.
Synopses:

@example
chroot @var{option} @var{newroot} [@var{command} [@var{args}]@dots{}]
chroot @var{option}
@end example

Ordinarily, file names are looked up starting at the root of the
directory structure, i.e., @file{/}.  @command{chroot} changes the root to
the directory @var{newroot} (which must exist), then changes the working
directory to @file{/}, and finally runs @var{command} with optional @var{args}.
If @var{command} is not specified, the default is the value of the @env{SHELL}
environment variable or @command{/bin/sh} if not set, invoked with the
@option{-i} option.
@var{command} must not be a special built-in utility
(@pxref{Special built-in utilities}).

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands.

@table @samp

@item --groups=@var{groups}
@opindex --groups
Use this option to override the supplementary @var{groups} to be
used by the new process.
The items in the list (names or numeric IDs) must be separated by commas.
Use @samp{--groups=''} to disable the supplementary group look-up
implicit in the @option{--userspec} option.

@item --userspec=@var{user}[:@var{group}]
@opindex --userspec
By default, @var{command} is run with the same credentials
as the invoking process.
Use this option to run it as a different @var{user} and/or with a
different primary @var{group}.
If a @var{user} is specified then the supplementary groups
are set according to the system defined list for that user,
unless overridden with the @option{--groups} option.

@item --skip-chdir
@opindex --skip-chdir
Use this option to not change the working directory to @file{/} after changing
the root directory to @var{newroot}, i.e., inside the chroot.
This option is only permitted when @var{newroot} is the old @file{/} directory,
and therefore is mostly useful together with the @option{--groups} and
@option{--userspec} options to retain the previous working directory.

@end table

The user and group name look-up performed by the @option{--userspec}
and @option{--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 @samp{+}.
@xref{Disambiguating names and IDs}.

Here are a few tips to help avoid common problems in using chroot.
To start with a simple example, make @var{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 @command{ls} executable,
and put it in @file{/tmp/empty}, you can run this command as root:

@example
$ chroot /tmp/empty /ls -Rl /
@end example

Then you'll see output like this:

@example
/:
total 1023
-rwxr-xr-x 1 0 0 1041745 Aug 16 11:17 ls
@end example

If you want to use a dynamically linked executable, say @command{bash},
then first run @samp{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.

@command{chroot} is installed only on systems that have the
@code{chroot} function, so portable scripts should not rely on its
existence.

@cindex exit status of @command{chroot}
Exit status:

@display
125 if @command{chroot} itself fails
126 if @var{command} is found but cannot be invoked
127 if @var{command} cannot be found
the exit status of @var{command} otherwise
@end display


@node env invocation
@section @command{env}: Run a command in a modified environment

@pindex env
@cindex environment, running a program in a modified
@cindex modified environment, running a program in a
@cindex running a program in a modified environment

@command{env} runs a command with a modified environment.  Synopses:

@example
env [@var{option}]@dots{} [@var{name}=@var{value}]@dots{} @c
[@var{command} [@var{args}]@dots{}]
env -[v]S'[@var{option}]@dots{} [@var{name}=@var{value}]@dots{} @c
[@var{command} [@var{args}]@dots{}]'
env
@end example

@command{env} is commonly used on first line of scripts (shebang line):
@example
#!/usr/bin/env @var{command}
#!/usr/bin/env -[v]S[@var{option}]@dots{} [@var{name}=@var{value}]@dots{} @c
@var{command} [@var{args}]@dots{}
@end example

Operands of the form @samp{@var{variable}=@var{value}} set
the environment variable @var{variable} to value @var{value}.
@var{value} may be empty (@samp{@var{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 @samp{=} 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.

@vindex PATH
The first operand that does not contain the character @samp{=}
specifies the program to invoke; it is
searched for according to the @env{PATH} environment variable.  Any
remaining arguments are passed as arguments to that program.
The program should not be a special built-in utility
(@pxref{Special built-in utilities}).

Modifications to @env{PATH} take effect prior to searching for
@var{command}.  Use caution when reducing @env{PATH}; behavior is
not portable when @env{PATH} is undefined or omits key directories
such as @file{/bin}.

In the rare case that a utility contains a @samp{=} in the name, the
only way to disambiguate it from a variable assignment is to use an
intermediate command for @var{command}, and pass the problematic
program name via @var{args}.  For example, if @file{./prog=} is an
executable in the current @env{PATH}:

@example
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='
@end example

@cindex environment, printing

If no command name is specified following the environment
specifications, the resulting environment is printed.  This is like
specifying the @command{printenv} program.

For some examples, suppose the environment passed to @command{env}
contains @samp{LOGNAME=rms}, @samp{EDITOR=emacs}, and
@samp{PATH=.:/gnubin:/hacks}:

@itemize @bullet

@item
Output the current environment.
@example
$ env | LC_ALL=C sort
EDITOR=emacs
LOGNAME=rms
PATH=.:/gnubin:/hacks
@end example

@item
Run @command{foo} with a reduced environment, preserving only the
original @env{PATH} to avoid problems in locating @command{foo}.
@example
env - PATH="$PATH" foo
@end example

@item
Run @command{foo} with the environment containing @samp{LOGNAME=rms},
@samp{EDITOR=emacs}, and @samp{PATH=.:/gnubin:/hacks}, and guarantees
that @command{foo} was found in the file system rather than as a shell
built-in.
@example
env foo
@end example

@item
Run @command{nemacs} with the environment containing @samp{LOGNAME=foo},
@samp{EDITOR=emacs}, @samp{PATH=.:/gnubin:/hacks}, and
@samp{DISPLAY=gnu:0}.
@example
env DISPLAY=gnu:0 LOGNAME=foo nemacs
@end example

@item
Attempt to run the program @command{/energy/--} (as that is the only
possible path search result); if the command exists, the environment
will contain @samp{LOGNAME=rms} and @samp{PATH=/energy}, and the
arguments will be @samp{e=mc2}, @samp{bar}, and @samp{baz}.
@example
env -u EDITOR PATH=/energy -- e=mc2 bar baz
@end example

@end itemize


@subsection General options

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands.

@table @samp

@optNull

@item -u @var{name}
@itemx --unset=@var{name}
@opindex -u
@opindex --unset
Remove variable @var{name} from the environment, if it was in the
environment.

@item -
@itemx -i
@itemx --ignore-environment
@opindex -
@opindex -i
@opindex --ignore-environment
Start with an empty environment, ignoring the inherited environment.

@item -C @var{dir}
@itemx --chdir=@var{dir}
@opindex -C
@opindex --chdir
Change the working directory to @var{dir} before invoking @var{command}.
This differs from the shell built-in @command{cd} in that it starts
@var{command} as a subprocess rather than altering the shell's own working
directory; this allows it to be chained with other commands that run commands
in a different context.  For example:

@example
# Run 'true' with /chroot as its root directory and /srv as its working
# directory.
chroot /chroot env --chdir=/srv true
# Run 'true' with /build as its working directory, FOO=bar in its
# environment, and a time limit of five seconds.
env --chdir=/build FOO=bar timeout 5 true
@end example

@item --default-signal[=@var{sig}]
Unblock and reset signal @var{sig} to its default signal handler.
Without @var{sig} all known signals are unblocked and reset to their defaults.
Multiple signals can be comma-separated. An empty @var{sig} argument is a no-op.
The following command runs @command{seq} with SIGINT and SIGPIPE set to their
default (which is to terminate the program):

@example
env --default-signal=PIPE,INT seq 1000 | head -n1
@end example

In the following example, we see how this is not
possible to do with traditional shells.
Here the first trap command sets SIGPIPE to ignore.
The second trap command ostensibly sets it back to its default,
but POSIX mandates that the shell must not change inherited
state of the signal - so it is a no-op.

@example
trap '' PIPE && sh -c 'trap - PIPE ; seq inf | head -n1'
@end example

Using @option{--default-signal=PIPE} we can
ensure the signal handling is set to its default behavior:

@example
trap '' PIPE && sh -c 'env --default-signal=PIPE seq inf | head -n1'
@end example


@item --ignore-signal[=@var{sig}]
Ignore signal @var{sig} when running a program.  Without @var{sig} all
known signals are set to ignore.  Multiple signals can be comma-separated.
An empty @var{sig} argument is a no-op. The following command runs @command{seq}
with SIGINT set to be ignored - pressing @kbd{Ctrl-C} will not terminate it:

@example
env --ignore-signal=INT seq inf > /dev/null
@end example

@samp{SIGCHLD} is special, in that @option{--ignore-signal=CHLD} might have
no effect (POSIX says it's unspecified).

Most operating systems do not allow ignoring @samp{SIGKILL}, @samp{SIGSTOP}
(and possibly other signals).  Attempting to ignore these signals will fail.

Multiple (and contradictory) @option{--default-signal=SIG} and
@option{--ignore-signal=SIG} options are processed left-to-right,
with the latter taking precedence.  In the following example, @samp{SIGPIPE} is
set to default while @samp{SIGINT} is ignored:

@example
env --default-signal=INT,PIPE --ignore-signal=INT
@end example

@item --block-signal[=@var{sig}]
Block signal(s) @var{sig} from being delivered.  Without @var{sig} all
known signals are set to blocked.  Multiple signals can be comma-separated.
An empty @var{sig} argument is a no-op.

@item --list-signal-handling
List blocked or ignored signals to standard error, before executing a command.

@item -v
@itemx --debug
@opindex -v
@opindex --debug
Show verbose information for each processing step.

@example
$ env -v -uTERM A=B uname -s
unset:    TERM
setenv:   A=B
executing: uname
   arg[0]= 'uname'
   arg[1]= '-s'
Linux
@end example

When combined with @option{-S} it is recommended to list @option{-v}
first, e.g. @command{env -vS'string'}.

@item -S @var{string}
@itemx --split-string=@var{string}
@opindex -S
@opindex --split-string
@cindex shebang arguments
@cindex scripts arguments
@cindex env in scripts
process and split @var{string} into separate arguments used to pass
multiple arguments on shebang lines.  @command{env} supports FreeBSD's
syntax of several escape sequences and environment variable
expansions. See below for details and examples.

@end table

@cindex exit status of @command{env}
Exit status:

@display
0   if no @var{command} is specified and the environment is output
125 if @command{env} itself fails
126 if @var{command} is found but cannot be invoked
127 if @var{command} cannot be found
the exit status of @var{command} otherwise
@end display

@subsection @option{-S}/@option{--split-string} usage in scripts

The @option{-S}/@option{--split-string} option enables use of multiple
arguments on the first line of scripts (the shebang line, @samp{#!}).

When a script's interpreter is in a known location, scripts typically
contain the absolute file name in their first line:

@multitable {Python Script:} {#!/usr/bin/python3}
@item Shell script:
@tab
@example
#!/bin/sh
echo hello
@end example

@item Perl script:
@tab
@example
#!/usr/bin/perl
print "hello\n";
@end example

@item Python script:
@tab
@example
#!/usr/bin/python3
print("hello")
@end example

@end multitable

When a script's interpreter is in a non-standard location
in the @env{PATH} environment variable, it is recommended
to use @command{env} on the first line of the script to
find the executable and run it:

@multitable {Python Script:} {#!/usr/bin/env python3}
@item Shell script:
@tab
@example
#!/usr/bin/env bash
echo hello
@end example

@item Perl script:
@tab
@example
#!/usr/bin/env perl
print "hello\n";
@end example

@item Python script:
@tab
@example
#!/usr/bin/env python3
print("hello")
@end example

@end multitable

Most operating systems (e.g. GNU/Linux, BSDs) treat all text after the
first space as a single argument. When using @command{env} in a script
it is thus not possible to specify multiple arguments.

In the following example:
@example
#!/usr/bin/env perl -T -w
print "hello\n";
@end example

The operating system treats @samp{perl -T -w} as one argument (the
program's name), and executing the script fails with:

@example
/usr/bin/env: 'perl -T -w': No such file or directory
@end example

The @option{-S} option instructs @command{env} to split the single string
into multiple arguments. The following example works as expected:

@example
$ cat hello.pl
#!/usr/bin/env -S perl -T -w
print "hello\n";

$ chmod a+x hello.pl
$ ./hello.pl
hello
@end example

And is equivalent to running @command{perl -T -w hello.pl} on the command line
prompt.

@unnumberedsubsubsec Testing and troubleshooting

@cindex single quotes, and @command{env -S}
@cindex @command{env -S}, and single quotes
@cindex @option{-S}, env and single quotes
To test @command{env -S} on the command line, use single quotes for the
@option{-S} string to emulate a single paramter. Single quotes are not
needed when using @command{env -S} in a shebang line on the first line of a
script (the operating system already treats it as one argument).

The following command is equivalent to the @file{hello.pl} script above:

@example
$ env -S'perl -T -w' hello.pl
@end example

@cindex @command{env -S}, debugging
@cindex debugging, @command{env -S}

To troubleshoot @option{-S} usage add the @option{-v} as the first
argument (before @option{-S}).

Using @option{-vS} on a shebang line in a script:

@example
$ cat hello-debug.pl
#!/usr/bin/env -vS perl -T -w
print "hello\n";

$ chmod a+x hello-debug.pl
$ ./hello-debug.pl
split -S:  'perl -T -w'
 into:    'perl'
     &    '-T'
     &    '-w'
executing: perl
   arg[0]= 'perl'
   arg[1]= '-T'
   arg[2]= '-w'
   arg[3]= './hello-debug.pl'
hello
@end example

Using @option{-vS} on the command line prompt (adding single quotes):

@example
$ env -vS'perl -T -w' hello-debug.pl
split -S:  'perl -T -w'
 into:    'perl'
     &    '-T'
     &    '-w'
executing: perl
   arg[0]= 'perl'
   arg[1]= '-T'
   arg[2]= '-w'
   arg[3]= 'hello-debug.pl'
hello
@end example

@subsection @option{-S}/@option{--split-string} syntax

@unnumberedsubsubsec Splitting arguments by whitespace

Running @command{env -Sstring} splits the @var{string} into
arguments based on unquoted spaces or tab characters.
(Newlines, carriage returns, vertical tabs and form feeds are treated
like spaces and tabs.)

In the following contrived example the @command{awk} variable
@samp{OFS} will be @code{<space>xyz<space>} as these spaces are inside
double quotes. The other space characters are used as argument separators:

@example
$ cat one.awk
#!/usr/bin/env -S awk -v OFS=" xyz " -f
BEGIN @{print 1,2,3@}

$ chmod a+x one.awk
$ ./one.awk
1 xyz 2 xyz 3
@end example

When using @option{-S} on the command line prompt, remember to add
single quotes around the entire string:

@example
$ env -S'awk -v OFS=" xyz " -f' one.awk
1 xyz 2 xyz 3
@end example

@unnumberedsubsubsec Escape sequences

@command{env} supports several escape sequences. These sequences
are processed when unquoted or inside double quotes (unless otherwise noted).
Single quotes disable escape sequences except @samp{\'} and @samp{\\}.

@multitable @columnfractions .10 .90

@item @code{\c}
@tab Ignore the remaining characters in the string.
Cannot be used inside double quotes.

@item @code{\f}
@tab form-feed character (ASCII 0x0C)

@item @code{\n}
@tab new-line character (ASCII 0x0A)

@item @code{\r}
@tab carriage-return character (ASCII 0x0D)

@item @code{\t}
@tab tab character (ASCII 0x09)

@item @code{\v}
@tab vertical tab character (ASCII 0x0B)

@item @code{\#}
@tab A hash @samp{#} character. Used when a @samp{#} character
is needed as the first character of an argument (see 'comments' section
below).

@item @code{\$}
@tab A dollar-sign character @samp{$}. Unescaped @samp{$} characters
are used to expand environment variables (see 'variables' section below).

@item @code{\_}
@tab Inside double-quotes, replaced with a single space character.
Outside quotes, treated as an argument separator. @samp{\_} can be used
to avoid space characters in a shebang line (see examples below).

@item @code{\"}
@tab A double-quote character.

@item @code{\'}
@tab A single-quote character.
This escape sequence works inside single-quoted strings.

@item @code{\\}
@tab A backslash character.
This escape sequence works inside single-quoted strings.

@end multitable

The following @command{awk} script will use tab character as input and output
field separator (instead of spaces and tabs):

@example
$ cat tabs.awk
#!/usr/bin/env -S awk -v FS="\t" -v OFS="\t" -f
...
@end example

@unnumberedsubsubsec Comments

The escape sequence @samp{\c} (used outside single/double quotes)
causes @command{env} to ignore the rest of the string.

The @samp{#} character causes @command{env} to ignore the rest of
the string when it appears as the first character of an argument.
Use @samp{\#} to reverse this behavior.

@example
$ env -S'printf %s\n A B C'
A
B
C

$ env -S'printf %s\n A# B C'
A#
B
C

$ env -S'printf %s\n A #B C'
A

$ env -S'printf %s\n A \#B C'
A
#B
C

$ env -S'printf %s\n A\cB C'
A
@end example

NOTE: The above examples use single quotes as they are executed
on the command-line.



@unnumberedsubsubsec Environment variable expansion

The pattern @samp{$@{VARNAME@}} is used to substitute a value from
the environment variable. The pattern must include the curly braces
(@samp{@{},@samp{@}}). Without them @command{env} will reject the string.
Special shell variables (such as @samp{$@@}, @samp{$*}, @samp{$$}) are
not supported.

If the environment variable is empty or not set, the pattern will be replaced
by an empty string. The value of @samp{$@{VARNAME@}} will be that of
the executed @command{env}, before any modifications using
@option{-i}/@option{--ignore-environment}/@option{-u}/@option{--unset} or
setting new values using @samp{VAR=VALUE}.

The following python script prepends @file{/opt/custom/modules} to the python
module search path environment variable (@samp{PYTHONPATH}):

@example
$ cat custom.py
#!/usr/bin/env -S PYTHONPATH=/opt/custom/modules/:$@{PYTHONPATH@} python
print "hello"
...
@end example

The expansion of @samp{$@{PYTHONPATH@}} is performed by @command{env},
not by a shell. If the curly braces are omitted, @command{env} will fail:

@example
$ cat custom.py
#!/usr/bin/env -S PYTHONPATH=/opt/custom/modules/:$PYTHONPATH python
print "hello"
...

$ chmod a+x custom.py
$ custom.py
/usr/bin/env: only $@{VARNAME@} expansion is supported, error at: $PYTHONPATH @c
python
@end example

Environment variable expansion happens before clearing the environment
(with @option{-i}) or unsetting specific variables (with @option{-u}):

@example
$ env -S'-i OLDUSER=$@{USER@} env'
OLDUSER=gordon
@end example

Use @option{-v} to diagnose the operations step-by-step:

@example
$ env -vS'-i OLDUSER=$@{USER@} env'
expanding $@{USER@} into 'gordon'
split -S:  '-i OLDUSER=$@{USER@} env'
 into:    '-i'
     &    'OLDUSER=gordon'
     &    'env'
cleaning environ
setenv:   OLDUSER=gordon
executing: env
   arg[0]= 'env'
OLDUSER=gordon
@end example



@node nice invocation
@section @command{nice}: Run a command with modified niceness

@pindex nice
@cindex niceness
@cindex scheduling, affecting
@cindex appropriate privileges

@command{nice} prints a process's @dfn{niceness}, or runs
a command with modified niceness.  @dfn{niceness} affects how
favorably the process is scheduled in the system.
Synopsis:

@example
nice [@var{option}]@dots{} [@var{command} [@var{arg}]@dots{}]
@end example

If no arguments are given, @command{nice} prints the current niceness.
Otherwise, @command{nice} runs the given @var{command} with its
niceness adjusted.  By default, its niceness is incremented by 10.

Niceness values range at least from @minus{}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 @command{nice} in
terms of a @dfn{nice value}, which is the non-negative difference
between a niceness and the minimum niceness.  Though @command{nice}
conforms to POSIX, its documentation and diagnostics use the
term ``niceness'' for compatibility with historical practice.

@var{command} must not be a special built-in utility (@pxref{Special
built-in utilities}).

@mayConflictWithShellBuiltIn{nice}

Note to change the @dfn{niceness} of an existing process,
one needs to use the @command{renice} command.

The program accepts the following option.  Also see @ref{Common options}.
Options must precede operands.

@table @samp
@item -n @var{adjustment}
@itemx --adjustment=@var{adjustment}
@opindex -n
@opindex --adjustment
Add @var{adjustment} instead of 10 to the command's niceness.  If
@var{adjustment} is negative and you lack appropriate privileges,
@command{nice} issues a warning but otherwise acts as if you specified
a zero adjustment.

For compatibility @command{nice} also supports an obsolete
option syntax @option{-@var{adjustment}}.  New scripts should use
@option{-n @var{adjustment}} instead.

@end table

@command{nice} is installed only on systems that have the POSIX
@code{setpriority} function, so portable scripts should not rely on
its existence on non-POSIX platforms.

@cindex exit status of @command{nice}
Exit status:

@display
0   if no @var{command} is specified and the niceness is output
125 if @command{nice} itself fails
126 if @var{command} is found but cannot be invoked
127 if @var{command} cannot be found
the exit status of @var{command} otherwise
@end display

It is sometimes useful to run a non-interactive program with reduced niceness.

@example
$ nice factor 4611686018427387903
@end example

Since @command{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 @samp{10}:

@example
$ nice
0
$ nice nice
10
$ nice -n 10 nice
10
@end example

The @var{adjustment} is relative to the current niceness.  In the
next example, the first @command{nice} invocation runs the second one
with niceness 10, and it in turn runs the final one with a niceness
that is 3 more:

@example
$ nice nice -n 3 nice
13
@end example

Specifying a niceness larger than the supported range
is the same as specifying the maximum supported value:

@example
$ nice -n 10000000000 nice
19
@end example

Only a privileged user may run a process with lower niceness:

@example
$ nice -n -1 nice
nice: cannot set niceness: Permission denied
0
$ sudo nice -n -1 nice
-1
@end example


@node nohup invocation
@section @command{nohup}: Run a command immune to hangups

@pindex nohup
@cindex hangups, immunity to
@cindex immunity to hangups
@cindex logging out and continuing to run

@flindex nohup.out
@command{nohup} runs the given @var{command} with hangup signals ignored,
so that the command can continue running in the background after you log
out.  Synopsis:

@example
nohup @var{command} [@var{arg}]@dots{}
@end example

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 @samp{nohup @var{command} [@var{arg}]@dots{}
0>/dev/null} instead.

@flindex nohup.out
If standard output is a terminal, the command's standard output is appended
to the file @file{nohup.out}; if that cannot be written to, it is appended
to the file @file{$HOME/nohup.out}; and if that cannot be written to, the
command is not run.
Any @file{nohup.out} or @file{$HOME/nohup.out} file created by
@command{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 @file{nohup.out} or
@file{$HOME/nohup.out} as above.

To capture the command's output to a file other than @file{nohup.out}
you can redirect it.  For example, to capture the output of
@command{make}:

@example
nohup make > make.log
@end example

@command{nohup} does not automatically put the command it runs in the
background; you must do that explicitly, by ending the command line
with an @samp{&}.  Also, @command{nohup} does not alter the
niceness of @var{command}; use @command{nice} for that,
e.g., @samp{nohup nice @var{command}}.

@var{command} must not be a special built-in utility (@pxref{Special
built-in utilities}).

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.  Options must precede operands.

@cindex exit status of @command{nohup}
Exit status:

@display
125 if @command{nohup} itself fails, and @env{POSIXLY_CORRECT} is not set
126 if @var{command} is found but cannot be invoked
127 if @var{command} cannot be found
the exit status of @var{command} otherwise
@end display

If @env{POSIXLY_CORRECT} is set, internal failures give status 127
instead of 125.


@node stdbuf invocation
@section @command{stdbuf}: Run a command with modified I/O stream buffering

@pindex stdbuf
@cindex standard streams, buffering
@cindex line buffered

@command{stdbuf} allows one to modify the buffering operations of the
three standard I/O streams associated with a program.  Synopsis:

@example
stdbuf @var{option}@dots{} @var{command}
@end example

@var{command} must start with the name of a program that
@enumerate
@item
uses the ISO C @code{FILE} streams for input/output (note the
programs @command{dd} and @command{cat} don't do that),

@item
does not adjust the buffering of its standard streams (note the
program @command{tee} is not in this category).
@end enumerate

Any additional @var{arg}s are passed as additional arguments to the
@var{command}.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item -i @var{mode}
@itemx --input=@var{mode}
@opindex -i
@opindex --input
Adjust the standard input stream buffering.

@item -o @var{mode}
@itemx --output=@var{mode}
@opindex -o
@opindex --output
Adjust the standard output stream buffering.

@item -e @var{mode}
@itemx --error=@var{mode}
@opindex -e
@opindex --error
Adjust the standard error stream buffering.

@end table

The @var{mode} can be specified as follows:

@table @samp

@item 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.

@item 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 @code{fread} will still block until @code{EOF} or error,
even if the underlying @code{read} returns less data than requested.

@item @var{size}
Specify the size of the buffer to use in fully buffered mode.
@multiplierSuffixesNoBlocks{size}

@end table

@command{stdbuf} is installed only on platforms that use the
Executable and Linkable Format (ELF) and support the
@code{constructor} attribute, so portable scripts should not rely on
its existence.

@cindex exit status of @command{stdbuf}
Exit status:

@display
125 if @command{stdbuf} itself fails
126 if @var{command} is found but cannot be invoked
127 if @var{command} cannot be found
the exit status of @var{command} otherwise
@end display


@node timeout invocation
@section @command{timeout}: Run a command with a time limit

@pindex timeout
@cindex time limit
@cindex run commands with bounded time

@command{timeout} runs the given @var{command} and kills it if it is
still running after the specified time interval.  Synopsis:

@example
timeout [@var{option}] @var{duration} @var{command} [@var{arg}]@dots{}
@end example

@var{command} must not be a special built-in utility (@pxref{Special
built-in utilities}).

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands.

@table @samp
@item --preserve-status
@opindex --preserve-status
Return the exit status of the managed @var{command} on timeout, rather than
a specific exit status indicating a timeout.  This is useful if the
managed @var{command} supports running for an indeterminate amount of time.

@item --foreground
@opindex --foreground
Don't create a separate background program group, so that
the managed @var{command} can use the foreground TTY normally.
This is needed to support two situations when timing out commands,
when not invoking @command{timeout} from an interactive shell.
@enumerate
@item
@var{command} is interactive and needs to read from the terminal for example
@item
the user wants to support sending signals directly to @var{command}
from the terminal (like Ctrl-C for example)
@end enumerate

Note in this mode of operation, any children of @var{command}
will not be timed out.  Also SIGCONT will not be sent to @var{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).

@item -k @var{duration}
@itemx --kill-after=@var{duration}
@opindex -k
@opindex --kill-after
Ensure the monitored @var{command} is killed by also sending a @samp{KILL}
signal.

The specified @var{duration} starts from the point in time when
@command{timeout} sends the initial signal to @var{command}, i.e.,
not from the beginning when the @var{command} is started.

This option has no effect if either the main @var{duration}
of the @command{timeout} command, or the @var{duration} specified
to this option, is 0.

This option may be useful if the selected signal did not kill the @var{command},
either because the signal was blocked or ignored, or if the @var{command} takes
too long (e.g. for cleanup work) to terminate itself within a certain amount
of time.

@item -s @var{signal}
@itemx --signal=@var{signal}
@opindex -s
@opindex --signal
Send this @var{signal} to @var{command} on timeout, rather than the
default @samp{TERM} signal.  @var{signal} may be a name like @samp{HUP}
or a number.  @xref{Signal specifications}.

@item -v
@itemx --verbose
@opindex -v
@opindex --verbose
Diagnose to standard error, any signal sent upon timeout.
@end table

@cindex time units
@var{duration} is a floating point number in either the current or the
C locale (@pxref{Floating point}) followed by an optional unit:
@display
@samp{s} for seconds (the default)
@samp{m} for minutes
@samp{h} for hours
@samp{d} for days
@end display
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.

@cindex exit status of @command{timeout}
Exit status:

@display
124 if @var{command} times out, and @option{--preserve-status} is not specified
125 if @command{timeout} itself fails
126 if @var{command} is found but cannot be invoked
127 if @var{command} cannot be found
137 if @var{command} or @command{timeout} is sent the KILL(9) signal (128+9)
the exit status of @var{command} otherwise
@end display

In the case of the @samp{KILL(9)} signal, @command{timeout} returns with
exit status 137, regardless of whether that signal is sent to @var{command}
or to @command{timeout} itself, i.e., these cases cannot be distinguished.
In the latter case, the @var{command} process may still be alive after
@command{timeout} has forcefully been terminated.

Examples:

@example
# Send the default TERM signal after 20s to a short-living 'sleep 1'.
# As that terminates long before the given duration, 'timeout' returns
# with the same exit status as the command, 0 in this case.
timeout 20 sleep 1

# Send the INT signal after 5s to the 'sleep' command.  Returns after
# 5 seconds with exit status 124 to indicate the sending of the signal.
timeout -s INT 5 sleep 20

# Likewise, but the command ignoring the INT signal due to being started
# via 'env --ignore-signal'.  Thus, 'sleep' terminates regularly after
# the full 20 seconds, still 'timeout' returns with exit status 124.
timeout -s INT 5s env --ignore-signal=INT sleep 20

# Likewise, but sending the KILL signal 3 seconds after the initial
# INT signal.  Hence, 'sleep' is forcefully terminated after about
# 8 seconds (5+3), and 'timeout' returns with an exit status of 137.
timeout -s INT -k 3s 5s env --ignore-signal=INT sleep 20
@end example

@node Process control
@chapter Process control

@cindex processes, commands for controlling
@cindex commands for controlling processes

@menu
* kill invocation::             Sending a signal to processes.
@end menu


@node kill invocation
@section @command{kill}: Send a signal to processes

@pindex kill
@cindex send a signal to processes

The @command{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:

@example
kill [-s @var{signal} | --signal @var{signal} | -@var{signal}] @var{pid}@dots{}
kill [-l | --list | -t | --table] [@var{signal}]@dots{}
@end example

@mayConflictWithShellBuiltIn{kill}

The first form of the @command{kill} command sends a signal to all
@var{pid} arguments.  The default signal to send if none is specified
is @samp{TERM}@.  The special signal number @samp{0} does not denote a
valid signal, but can be used to test whether the @var{pid} arguments
specify processes to which a signal could be sent.

If @var{pid} is positive, the signal is sent to the process with the
process ID @var{pid}.  If @var{pid} is zero, the signal is sent to all
processes in the process group of the current process.  If @var{pid}
is @minus{}1, the signal is sent to all processes for which the user has
permission to send a signal.  If @var{pid} is less than @minus{}1, the signal
is sent to all processes in the process group that equals the absolute
value of @var{pid}.

If @var{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 @var{pid} argument is desired as the first one, it
should be preceded by @option{--}.  However, as a common extension to
POSIX, @option{--} is not required with @samp{kill
-@var{signal} -@var{pid}}.  The following commands are equivalent:

@example
kill -15 -1
kill -TERM -1
kill -s TERM -- -1
kill -- -1
@end example

The first form of the @command{kill} command succeeds if every @var{pid}
argument specifies at least one process that the signal was sent to.

The second form of the @command{kill} command lists signal information.
Either the @option{-l} or @option{--list} option, or the @option{-t}
or @option{--table} option must be specified.  Without any
@var{signal} argument, all supported signals are listed.  The output
of @option{-l} or @option{--list} is a list of the signal names, one
per line; if @var{signal} is already a name, the signal number is
printed instead.  The output of @option{-t} or @option{--table} is a
table of signal numbers, names, and descriptions.  This form of the
@command{kill} command succeeds if all @var{signal} arguments are valid
and if there is no output error.

The @command{kill} command also supports the @option{--help} and
@option{--version} options.  @xref{Common options}.

A @var{signal} may be a signal name like @samp{HUP}, or a signal
number like @samp{1}, or an exit status of a process terminated by the
signal.  A signal name can be given in canonical form or prefixed by
@samp{SIG}@.  The case of the letters is ignored, except for the
@option{-@var{signal}} option which must use upper case to avoid
ambiguity with lower case option letters.
@xref{Signal specifications}, for a list of supported
signal names and numbers.

@node Delaying
@chapter Delaying

@cindex delaying commands
@cindex commands for delaying

@c Perhaps @command{wait} or other commands should be described here also?

@menu
* sleep invocation::            Delay for a specified time.
@end menu


@node sleep invocation
@section @command{sleep}: Delay for a specified time

@pindex sleep
@cindex delay for a specified time

@command{sleep} pauses for an amount of time specified by the sum of
the values of the command line arguments.
Synopsis:

@example
sleep @var{number}[smhd]@dots{}
@end example

@cindex time units
Each argument is a non-negative number followed by an optional unit; the default
is seconds.  The units are:

@table @samp
@item s
seconds
@item m
minutes
@item h
hours
@item d
days
@end table

Although portable POSIX scripts must give @command{sleep} a single
non-negative integer argument without a suffix, GNU @command{sleep}
also accepts two or more arguments, unit suffixes, and floating-point
numbers in either the current or the C locale.  @xref{Floating point}.

For instance, the following could be used to @command{sleep} for
1 second, 234 milli-, 567 micro- and 890 nanoseconds:

@example
sleep 1234e-3 567.89e-6
@end example

Also one could sleep indefinitely like:

@example
sleep inf
@end example

The only options are @option{--help} and @option{--version}.  @xref{Common
options}.

@c sleep is a shell built-in at least with Solaris 11's /bin/sh
@mayConflictWithShellBuiltIn{sleep}

@exitstatus


@node Numeric operations
@chapter Numeric operations

@cindex numeric operations
These programs do numerically-related operations.

@menu
* factor invocation::              Show factors of numbers.
* numfmt invocation::              Reformat numbers.
* seq invocation::                 Print sequences of numbers.
@end menu


@node factor invocation
@section @command{factor}: Print prime factors

@pindex factor
@cindex prime factors

@command{factor} prints prime factors.  Synopsis:

@example
factor [@var{option}]@dots{} [@var{number}]@dots{}
@end example

If no @var{number} is specified on the command line, @command{factor} reads
numbers from standard input, delimited by newlines, tabs, or spaces.

The program accepts the following options.  Also see @ref{Common options}.

@table @samp
@item -h
@itemx --exponents
@opindex -h
@opindex --exponents
print factors in the form @math{p^e}, rather than repeating
the prime @samp{p}, @samp{e} times. If the exponent @samp{e} is 1,
then it is omitted.

@example
$ factor --exponents 3000
3000: 2^3 3 5^3
@end example
@end table

If the number to be factored is small (less than @math{2^{127}} on
typical machines), @command{factor} uses a faster algorithm.
For example, on a circa-2017 Intel Xeon Silver 4116, factoring the
product of the eighth and ninth Mersenne primes (approximately
@math{2^{92}}) takes about 4 ms of CPU time:

@example
$ M8=$(echo 2^31-1 | bc)
$ M9=$(echo 2^61-1 | bc)
$ n=$(echo "$M8 * $M9" | bc)
$ bash -c "time factor $n"
4951760154835678088235319297: 2147483647 2305843009213693951

real    0m0.004s
user    0m0.004s
sys     0m0.000s
@end example

For larger numbers, @command{factor} uses a slower algorithm.  On the
same platform, factoring the eighth Fermat number @math{2^{256} + 1}
takes about 14 seconds, and the slower algorithm would have taken
about 750 ms to factor @math{2^{127} - 3} instead of the 50 ms needed by
the faster algorithm.

Factoring large numbers is, in general, hard.  The Pollard-Brent rho
algorithm used by @command{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.

@exitstatus


@node numfmt invocation
@section @command{numfmt}: Reformat numbers

@pindex numfmt

@command{numfmt} reads numbers in various representations and reformats them
as requested.  The most common usage is converting numbers to/from @emph{human}
representation (e.g. @samp{4G} @expansion{} @samp{4,000,000,000}).

@example
numfmt [@var{option}]@dots{} [@var{number}]
@end example

@command{numfmt} converts each @var{number} on the command-line according to the
specified options (see below).  If no @var{number}s are given, it reads numbers
from standard input.  @command{numfmt} can optionally extract numbers from
specific columns, maintaining proper line padding and alignment.

@exitstatus

See @option{--invalid} for additional information regarding exit status.

@subsection General options

The program accepts the following options.  Also see @ref{Common options}.

@table @samp

@item --debug
@opindex --debug
Print (to standard error) warning messages about possible erroneous usage.

@item -d @var{d}
@itemx --delimiter=@var{d}
@opindex -d
@opindex --delimiter
Use the character @var{d} as input field separator (default: whitespace).
@emph{Note}: Using non-default delimiter turns off automatic padding.

@item --field=@var{fields}
@opindex --field
Convert the number in input field @var{fields} (default: 1).
@var{fields} supports @command{cut} style field ranges:

@example
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
@end example


@item --format=@var{format}
@opindex --format
Use printf-style floating FORMAT string.  The @var{format} string must contain
one @samp{%f} directive, optionally with @samp{'}, @samp{-}, @samp{0}, width
or precision modifiers.  The @samp{'} modifier will enable @option{--grouping},
the @samp{-} modifier will enable left-aligned @option{--padding} and the width
modifier will enable right-aligned @option{--padding}.  The @samp{0} width
modifier (without the @samp{-} modifier) will generate leading zeros on the
number, up to the specified width.  A precision specification like @samp{%.1f}
will override the precision determined from the input data or set due to
@option{--to} option auto scaling.

@item --from=@var{unit}
@opindex --from
Auto-scales input numbers according to @var{unit}.  See UNITS below.
The default is no scaling, meaning suffixes (e.g. @samp{M}, @samp{G}) will
trigger an error.

@item --from-unit=@var{n}
@opindex --from-unit
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 @samp{10}
represents 10 units of 512 bytes, use @samp{--from-unit=512}).
Suffixes are handled as with @samp{--from=auto}.

@item --grouping
@opindex --grouping
Group digits in output numbers according to the current locale's grouping rules
(e.g @emph{Thousands Separator} character, commonly @samp{.} (dot) or @samp{,}
comma).  This option has no effect in @samp{POSIX/C} locale.

@item --header[=@var{n}]
@opindex --header
@opindex --header=N
Print the first @var{n} (default: 1) lines without any conversion.

@item --invalid=@var{mode}
@opindex --invalid
The default action on input errors is to exit immediately with status code 2.
@option{--invalid=@samp{abort}} explicitly specifies this default mode.
With a @var{mode} of @samp{fail}, print a warning for @emph{each} conversion
error, and exit with status 2.  With a @var{mode} of @samp{warn}, exit with
status 0, even in the presence of conversion errors, and with a @var{mode} of
@samp{ignore} do not even print diagnostics.

@item --padding=@var{n}
@opindex --padding
Pad the output numbers to @var{n} characters, by adding spaces.  If @var{n} is
a positive number, numbers will be right-aligned.  If @var{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).

@item --round=@var{method}
@opindex --round
@opindex --round=up
@opindex --round=down
@opindex --round=from-zero
@opindex --round=towards-zero
@opindex --round=nearest
When converting number representations, round the number according to
@var{method}, which can be @samp{up}, @samp{down},
@samp{from-zero} (the default), @samp{towards-zero}, @samp{nearest}.

@item --suffix=@var{suffix}
@opindex --suffix
Add @samp{SUFFIX} to the output numbers, and accept optional @samp{SUFFIX} in
input numbers.

@item --to=@var{unit}
@opindex --to
Auto-scales output numbers according to @var{unit}.  See @emph{Units} below.
The default is no scaling, meaning all the digits of the number are printed.

@item --to-unit=@var{n}
@opindex --to-unit
Specify the output unit size (instead of the default 1).  Use this option when
the output numbers represent other units (e.g. to represent @samp{4,000,000}
bytes in blocks of 1KB, use @samp{--to=si --to-unit=1000}).
Suffixes are handled as with @samp{--from=auto}.

@optZeroTerminated
@newlineFieldSeparator

@end table

@subsection Possible @var{unit}s:

The following are the possible @var{unit} options with @option{--from=UNITS} and
@option{--to=UNITS}:

@table @var

@item 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.

@item si
Auto-scale numbers according to the @emph{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:

@example
@samp{K}  =>  @math{1000^1 = 10^3} (Kilo)
@samp{M}  =>  @math{1000^2 = 10^6} (Mega)
@samp{G}  =>  @math{1000^3 = 10^9} (Giga)
@samp{T}  =>  @math{1000^4 = 10^{12}} (Tera)
@samp{P}  =>  @math{1000^5 = 10^{15}} (Peta)
@samp{E}  =>  @math{1000^6 = 10^{18}} (Exa)
@samp{Z}  =>  @math{1000^7 = 10^{21}} (Zetta)
@samp{Y}  =>  @math{1000^8 = 10^{24}} (Yotta)
@end example

@item iec
Auto-scale numbers according to the @emph{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:

@example
@samp{K}  =>  @math{1024^1 = 2^{10}} (Kibi)
@samp{M}  =>  @math{1024^2 = 2^{20}} (Mebi)
@samp{G}  =>  @math{1024^3 = 2^{30}} (Gibi)
@samp{T}  =>  @math{1024^4 = 2^{40}} (Tebi)
@samp{P}  =>  @math{1024^5 = 2^{50}} (Pebi)
@samp{E}  =>  @math{1024^6 = 2^{60}} (Exbi)
@samp{Z}  =>  @math{1024^7 = 2^{70}} (Zebi)
@samp{Y}  =>  @math{1024^8 = 2^{80}} (Yobi)
@end example

The @option{iec} option uses a single letter suffix (e.g. @samp{G}), which is
not fully standard, as the @emph{iec} standard recommends a two-letter symbol
(e.g @samp{Gi}) - but in practice, this method common.  Compare with
the @option{iec-i} option.

@item iec-i
Auto-scale numbers according to the @emph{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:

@example
@samp{Ki}  =>  @math{1024^1 = 2^{10}} (Kibi)
@samp{Mi}  =>  @math{1024^2 = 2^{20}} (Mebi)
@samp{Gi}  =>  @math{1024^3 = 2^{30}} (Gibi)
@samp{Ti}  =>  @math{1024^4 = 2^{40}} (Tebi)
@samp{Pi}  =>  @math{1024^5 = 2^{50}} (Pebi)
@samp{Ei}  =>  @math{1024^6 = 2^{60}} (Exbi)
@samp{Zi}  =>  @math{1024^7 = 2^{70}} (Zebi)
@samp{Yi}  =>  @math{1024^8 = 2^{80}} (Yobi)
@end example

The @option{iec-i} option uses a two-letter suffix symbol (e.g. @samp{Gi}),
as the @emph{iec} standard recommends, but this is not always common in
practice.  Compare with the @option{iec} option.

@item auto
@samp{auto} can only be used with @option{--from}.  With this method, numbers
with @samp{K},@samp{M},@samp{G},@samp{T},@samp{P},@samp{E},@samp{Z},@samp{Y}
suffixes are interpreted as @emph{SI} values, and numbers with @samp{Ki},
@samp{Mi},@samp{Gi},@samp{Ti},@samp{Pi},@samp{Ei},@samp{Zi},@samp{Yi} suffixes
are interpreted as @emph{IEC} values.

@end table

@subsection Examples of using @command{numfmt}

Converting a single number from/to @emph{human} representation:
@example
$ 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
@end example

Converting from @samp{SI} to @samp{IEC} scales (e.g. when a drive's capacity is
advertised as @samp{1TB}, while checking the drive's capacity gives lower
values):

@example
$ numfmt --from=si --to=iec 1T
932G
@end example


Converting a single field from an input file / piped input (these contrived
examples are for demonstration purposes only, as both @command{ls} and
@command{df} support the @option{--human-readable} option to
output sizes in human-readable format):

@example
# 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
@end example


Output can be tweaked using @option{--padding} or @option{--format}:

@example
# 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
@end example

With locales that support grouping digits, using @option{--grouping} or
@option{--format} enables grouping.  In @samp{POSIX} locale, grouping is
silently ignored:

@example
$ 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 numfmt --from=iec --format="==%'15f==" 2G
==     2147483648==

$ LC_ALL=en_US.utf8 numfmt --from=iec --format="==%'15f==" 2G
==  2,147,483,648==

$ LC_ALL=en_US.utf8 numfmt --from=iec --format="==%'-15f==" 2G
==2,147,483,648  ==

$ LC_ALL=ta_IN numfmt --from=iec --format="==%'15f==" 2G
== 2,14,74,83,648==
@end example


@node seq invocation
@section @command{seq}: Print numeric sequences

@pindex seq
@cindex numeric sequences
@cindex sequence of numbers

@command{seq} prints a sequence of numbers to standard output.  Synopses:

@example
seq [@var{option}]@dots{} @var{last}
seq [@var{option}]@dots{} @var{first} @var{last}
seq [@var{option}]@dots{} @var{first} @var{increment} @var{last}
@end example

@command{seq} prints the numbers from @var{first} to @var{last} by
@var{increment}.  By default, each number is printed on a separate line.
When @var{increment} is not specified, it defaults to @samp{1},
even when @var{first} is larger than @var{last}.
@var{first} also defaults to @samp{1}.  So @code{seq 1} prints
@samp{1}, but @code{seq 0} and @code{seq 10 5} produce no output.
The sequence of numbers ends when the sum of the current number and
@var{increment} would become greater than @var{last},
so @code{seq 1 10 10} only produces @samp{1}.
@var{increment} must not be @samp{0}; use the tool @command{yes} to get
repeated output of a constant number.
@var{first}, @var{increment} and @var{last} must not be @code{NaN},
but @code{inf} is supported.
Floating-point numbers may be specified in either the current or
the C locale.  @xref{Floating point}.

The program accepts the following options.  Also see @ref{Common options}.
Options must precede operands.

@table @samp
@item -f @var{format}
@itemx --format=@var{format}
@opindex -f
@opindex --format
@cindex formatting of numbers in @command{seq}
Print all numbers using @var{format}.
@var{format} must contain exactly one of the @samp{printf}-style
floating point conversion specifications @samp{%a}, @samp{%e},
@samp{%f}, @samp{%g}, @samp{%A}, @samp{%E}, @samp{%F}, @samp{%G}@.
The @samp{%} may be followed by zero or more flags taken from the set
@samp{-+#0 '}, then an optional width containing one or more digits,
then an optional precision consisting of a @samp{.} followed by zero
or more digits.  @var{format} may also contain any number of @samp{%%}
conversion specifications.  All conversion specifications have the
same meaning as with @samp{printf}.

The default format is derived from @var{first}, @var{step}, and
@var{last}.  If these all use a fixed point decimal representation,
the default format is @samp{%.@var{p}f}, where @var{p} is the minimum
precision that can represent the output numbers exactly.  Otherwise,
the default format is @samp{%g}.

@item -s @var{string}
@itemx --separator=@var{string}
@opindex -s
@opindex --separator
@cindex separator for numbers in @command{seq}
Separate numbers with @var{string}; default is a newline.
The output always terminates with a newline.

@item -w
@itemx --equal-width
@opindex -w
@opindex --equal-width
Print all numbers with the same width, by padding with leading zeros.
@var{first}, @var{step}, and @var{last} should all use a fixed point
decimal representation.
(To have other kinds of padding, use @option{--format}).

@end table

You can get finer-grained control over output with @option{-f}:

@example
$ seq -f '(%9.2E)' -9e5 1.1e6 1.3e6
(-9.00E+05)
( 2.00E+05)
( 1.30E+06)
@end example

If you want hexadecimal integer output, you can use @command{printf}
to perform the conversion:

@example
$ printf '%x\n' $(seq 1048575 1024 1050623)
fffff
1003ff
1007ff
@end example

For very long lists of numbers, use xargs to avoid
system limitations on the length of an argument list:

@example
$ seq 1000000 | xargs printf '%x\n' | tail -n 3
f423e
f423f
f4240
@end example

To generate octal output, use the printf @code{%o} format instead
of @code{%x}.

On most systems, seq can produce whole-number output for values up to
at least @math{2^{53}}.  Larger integers are approximated.  The details
differ depending on your floating-point implementation.
@xref{Floating point}.  A common
case is that @command{seq} works with integers through @math{2^{64}},
and larger integers may not be numerically correct:

@example
$ seq 50000000000000000000 2 50000000000000000004
50000000000000000000
50000000000000000000
50000000000000000004
@end example

However, note that when limited to non-negative whole numbers,
an increment of less than 200, and no format-specifying option,
seq can print arbitrarily large numbers.
Therefore @command{seq inf} can be used to
generate an infinite sequence of numbers.

Be careful when using @command{seq} with outlandish values: otherwise
you may see surprising results, as @command{seq} uses floating point
internally.  For example, on the x86 platform, where the internal
representation uses a 64-bit fraction, the command:

@example
seq 1 0.0000000000000000001 1.0000000000000000009
@end example

outputs 1.0000000000000000007 twice and skips 1.0000000000000000008.

@exitstatus


@node File permissions
@chapter File permissions
@include perm.texi


@node File timestamps
@chapter File timestamps

@cindex atime
@cindex birthtime
@cindex ctime
@cindex mtime
Standard POSIX files have three timestamps: the access timestamp
(atime) of the last read, the modification timestamp (mtime) of the
last write, and the status change timestamp (ctime) of the last change
to the file's meta-information.  Some file systems support a
fourth time: the birth timestamp (birthtime) of when the file was
created; by definition, birthtime never changes.

One common example of a ctime change is when the permissions of a file
change.  Changing the permissions doesn't access the file, so 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.

Naively, a file's atime, mtime, and ctime are set to the current time
whenever you read, write, or change the attributes of the file
respectively, and searching a directory counts as reading it.  A
file's atime and mtime can also be set directly, via the
@command{touch} command (@pxref{touch invocation}).  In practice,
though, timestamps are not updated quite that way.

For efficiency reasons, many systems are lazy about updating atimes:
when a program accesses a file, they may delay updating the file's
atime, or may not update the file's atime if the file has been
accessed recently, or may not update the atime at all.  Similar
laziness, though typically not quite so extreme, applies to mtimes and
ctimes.

Some systems emulate timestamps instead of supporting them directly,
and these emulations may disagree with the naive interpretation.  For
example, a system may fake an atime or ctime by using the mtime.

@cindex clock skew
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
updates 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''.

@cindex file timestamp resolution
When the system updates a file timestamp to a desired time @var{t}
(which is either the current time, or a time specified via the
@command{touch} command), there are several reasons the file's
timestamp may be set to a value that differs from @var{t}.  First,
@var{t} 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
timestamp and 100-nanosecond resolution for modification timestamp, and the
operating system might use nanosecond resolution for the current time
and microsecond resolution for the primitive that @command{touch} uses
to set a file's timestamp to an arbitrary value.


@include parse-datetime.texi

@include sort-version.texi

@c              What's GNU?
@c              Arnold Robbins
@node Opening the software toolbox
@chapter Opening the Software Toolbox

An earlier version of this chapter appeared in
@uref{https://www.linuxjournal.com/article.php?sid=2762, the
@cite{What's GNU@?} column of the June 1994 @cite{Linux Journal}}.
It was written by Arnold Robbins.

@menu
* Toolbox introduction::        Toolbox introduction
* I/O redirection::             I/O redirection
* The who command::             The @command{who} command
* The cut command::             The @command{cut} command
* The sort command::            The @command{sort} command
* The uniq command::            The @command{uniq} command
* Putting the tools together::  Putting the tools together
@end menu


@node Toolbox introduction
@unnumberedsec Toolbox Introduction

This month's column is only peripherally related to the GNU Project, in
that it describes a number of the GNU tools on your GNU/Linux system
and how they
might be used.  What it's really about is the ``Software Tools'' philosophy
of program development and usage.

The software tools philosophy was an important and integral concept
in the initial design and development of Unix (of which GNU/Linux and GNU are
essentially clones).  Unfortunately, in the modern day press of
Internetworking and flashy GUIs, it seems to have fallen by the
wayside.  This is a shame, since it provides a powerful mental model
for solving many kinds of problems.

Many people carry a Swiss Army knife around in their pants pockets (or
purse).  A Swiss Army knife is a handy tool to have: it has several knife
blades, a screwdriver, tweezers, toothpick, nail file, corkscrew, and perhaps
a number of other things on it.  For the everyday, small miscellaneous jobs
where you need a simple, general purpose tool, it's just the thing.

On the other hand, an experienced carpenter doesn't build a house using
a Swiss Army knife.  Instead, he has a toolbox chock full of specialized
tools---a saw, a hammer, a screwdriver, a plane, and so on.  And he knows
exactly when and where to use each tool; you won't catch him hammering nails
with the handle of his screwdriver.

The Unix developers at Bell Labs were all professional programmers and trained
computer scientists.  They had found that while a one-size-fits-all program
might appeal to a user because there's only one program to use, in practice
such programs are

@enumerate a
@item
difficult to write,

@item
difficult to maintain and
debug, and

@item
difficult to extend to meet new situations.
@end enumerate

Instead, they felt that programs should be specialized tools.  In short, each
program ``should do one thing well.''  No more and no less.  Such programs are
simpler to design, write, and get right---they only do one thing.

Furthermore, they found that with the right machinery for hooking programs
together, that the whole was greater than the sum of the parts.  By combining
several special purpose programs, you could accomplish a specific task
that none of the programs was designed for, and accomplish it much more
quickly and easily than if you had to write a special purpose program.
We will see some (classic) examples of this further on in the column.
(An important additional point was that, if necessary, take a detour
and build any software tools you may need first, if you don't already
have something appropriate in the toolbox.)

@node I/O redirection
@unnumberedsec I/O Redirection

Hopefully, you are familiar with the basics of I/O redirection in the
shell, in particular the concepts of ``standard input,'' ``standard output,''
and ``standard error''.  Briefly, ``standard input'' is a data source, where
data comes from.  A program should not need to either know or care if the
data source is a regular file, a keyboard, a magnetic tape, or even a punched
card reader.  Similarly, ``standard output'' is a data sink, where data goes
to.  The program should neither know nor care where this might be.
Programs that only read their standard input, do something to the data,
and then send it on, are called @dfn{filters}, by analogy to filters in a
water pipeline.

With the Unix shell, it's very easy to set up data pipelines:

@example
program_to_create_data | filter1 | ... | filterN > final.pretty.data
@end example

We start out by creating the raw data; each filter applies some successive
transformation to the data, until by the time it comes out of the pipeline,
it is in the desired form.

This is fine and good for standard input and standard output.  Where does the
standard error come in to play?  Well, think about @command{filter1} in
the pipeline above.  What happens if it encounters an error in the data it
sees?  If it writes an error message to standard output, it will just
disappear down the pipeline into @command{filter2}'s input, and the
user will probably never see it.  So programs need a place where they can send
error messages so that the user will notice them.  This is standard error,
and it is usually connected to your console or window, even if you have
redirected standard output of your program away from your screen.

For filter programs to work together, the format of the data has to be
agreed upon.  The most straightforward and easiest format to use is simply
lines of text.  Unix data files are generally just streams of bytes, with
lines delimited by the ASCII LF (Line Feed) character,
conventionally called a ``newline'' in the Unix literature.  (This is
@code{'\n'} if you're a C programmer.)  This is the format used by all
the traditional filtering programs.  (Many earlier operating systems
had elaborate facilities and special purpose programs for managing
binary data.  Unix has always shied away from such things, under the
philosophy that it's easiest to simply be able to view and edit your
data with a text editor.)

OK, enough introduction.  Let's take a look at some of the tools, and then
we'll see how to hook them together in interesting ways.   In the following
discussion, we will only present those command line options that interest
us.  As you should always do, double check your system documentation
for the full story.

@node The who command
@unnumberedsec The @command{who} Command

The first program is the @command{who} command.  By itself, it generates a
list of the users who are currently logged in.  Although I'm writing
this on a single-user system, we'll pretend that several people are
logged in:

@example
$ who
@print{} arnold   console Jan 22 19:57
@print{} miriam   ttyp0   Jan 23 14:19(:0.0)
@print{} bill     ttyp1   Jan 21 09:32(:0.0)
@print{} arnold   ttyp2   Jan 23 20:48(:0.0)
@end example

Here, the @samp{$} is the usual shell prompt, at which I typed @samp{who}.
There are three people logged in, and I am logged in twice.  On traditional
Unix systems, user names are never more than eight characters long.  This
little bit of trivia will be useful later.  The output of @command{who} is nice,
but the data is not all that exciting.

@node The cut command
@unnumberedsec The @command{cut} Command

The next program we'll look at is the @command{cut} command.  This program
cuts out columns or fields of input data.  For example, we can tell it
to print just the login name and full name from the @file{/etc/passwd}
file.  The @file{/etc/passwd} file has seven fields, separated by
colons:

@example
arnold:xyzzy:2076:10:Arnold D. Robbins:/home/arnold:/bin/bash
@end example

To get the first and fifth fields, we would use @command{cut} like this:

@example
$ cut -d: -f1,5 /etc/passwd
@print{} root:Operator
@dots{}
@print{} arnold:Arnold D. Robbins
@print{} miriam:Miriam A. Robbins
@dots{}
@end example

With the @option{-c} option, @command{cut} will cut out specific characters
(i.e., columns) in the input lines.  This is useful for input data
that has fixed width fields, and does not have a field separator.  For
example, list the Monday dates for the current month:

@c Is using cal ok?  Looked at gcal, but I don't like it.
@example
$ cal | cut -c 3-5
@print{}Mo
@print{}
@print{}  6
@print{} 13
@print{} 20
@print{} 27
@end example

@node The sort command
@unnumberedsec The @command{sort} Command

Next we'll look at the @command{sort} command.  This is one of the most
powerful commands on a Unix-style system; one that you will often find
yourself using when setting up fancy data plumbing.

The @command{sort}
command reads and sorts each file named on the command line.  It then
merges the sorted data and writes it to standard output.  It will read
standard input if no files are given on the command line (thus
making it into a filter).  The sort is based on the character collating
sequence or based on user-supplied ordering criteria.


@node The uniq command
@unnumberedsec The @command{uniq} Command

Finally (at least for now), we'll look at the @command{uniq} program.  When
sorting data, you will often end up with duplicate lines, lines that
are identical.  Usually, all you need is one instance of each line.
This is where @command{uniq} comes in.  The @command{uniq} program reads its
standard input.  It prints only one
copy of each repeated line.  It does have several options.  Later on,
we'll use the @option{-c} option, which prints each unique line, preceded
by a count of the number of times that line occurred in the input.


@node Putting the tools together
@unnumberedsec Putting the Tools Together

Now, let's suppose this is a large ISP server system with dozens of users
logged in.  The management wants the system administrator to write a
program that will
generate a sorted list of logged in users.  Furthermore, even if a user
is logged in multiple times, his or her name should only show up in the
output once.

The administrator could sit down with the system documentation and write a C
program that did this.  It would take perhaps a couple of hundred lines
of code and about two hours to write it, test it, and debug it.
However, knowing the software toolbox, the administrator can instead start out
by generating just a list of logged on users:

@example
$ who | cut -c1-8
@print{} arnold
@print{} miriam
@print{} bill
@print{} arnold
@end example

Next, sort the list:

@example
$ who | cut -c1-8 | sort
@print{} arnold
@print{} arnold
@print{} bill
@print{} miriam
@end example

Finally, run the sorted list through @command{uniq}, to weed out duplicates:

@example
$ who | cut -c1-8 | sort | uniq
@print{} arnold
@print{} bill
@print{} miriam
@end example

The @command{sort} command actually has a @option{-u} option that does what
@command{uniq} does.  However, @command{uniq} has other uses for which one
cannot substitute @samp{sort -u}.

The administrator puts this pipeline into a shell script, and makes it
available for
all the users on the system (@samp{#} is the system administrator,
or @code{root}, prompt):

@example
# cat > /usr/local/bin/listusers
who | cut -c1-8 | sort | uniq
^D
# chmod +x /usr/local/bin/listusers
@end example

There are four major points to note here.  First, with just four
programs, on one command line, the administrator was able to save about two
hours worth of work.  Furthermore, the shell pipeline is just about as
efficient as the C program would be, and it is much more efficient in
terms of programmer time.  People time is much more expensive than
computer time, and in our modern ``there's never enough time to do
everything'' society, saving two hours of programmer time is no mean
feat.

Second, it is also important to emphasize that with the
@emph{combination} of the tools, it is possible to do a special
purpose job never imagined by the authors of the individual programs.

Third, it is also valuable to build up your pipeline in stages, as we did here.
This allows you to view the data at each stage in the pipeline, which helps
you acquire the confidence that you are indeed using these tools correctly.

Finally, by bundling the pipeline in a shell script, other users can use
your command, without having to remember the fancy plumbing you set up for
them.  In terms of how you run them, shell scripts and compiled programs are
indistinguishable.

After the previous warm-up exercise, we'll look at two additional, more
complicated pipelines.  For them, we need to introduce two more tools.

The first is the @command{tr} command, which stands for ``transliterate.''
The @command{tr} command works on a character-by-character basis, changing
characters.  Normally it is used for things like mapping upper case to
lower case:

@example
$ echo ThIs ExAmPlE HaS MIXED case! | tr '[:upper:]' '[:lower:]'
@print{} this example has mixed case!
@end example

There are several options of interest:

@table @code
@item -c
work on the complement of the listed characters, i.e.,
operations apply to characters not in the given set

@item -d
delete characters in the first set from the output

@item -s
squeeze repeated characters in the output into just one character.
@end table

We will be using all three options in a moment.

The other command we'll look at is @command{comm}.  The @command{comm}
command takes two sorted input files as input data, and prints out the
files' lines in three columns.  The output columns are the data lines
unique to the first file, the data lines unique to the second file, and
the data lines that are common to both.  The @option{-1}, @option{-2}, and
@option{-3} command line options @emph{omit} the respective columns.  (This is
non-intuitive and takes a little getting used to.)  For example:

@example
$ cat f1
@print{} 11111
@print{} 22222
@print{} 33333
@print{} 44444
$ cat f2
@print{} 00000
@print{} 22222
@print{} 33333
@print{} 55555
$ comm f1 f2
@print{}         00000
@print{} 11111
@print{}                 22222
@print{}                 33333
@print{} 44444
@print{}         55555
@end example

The file name @file{-} tells @command{comm} to read standard input
instead of a regular file.

Now we're ready to build a fancy pipeline.  The first application is a word
frequency counter.  This helps an author determine if he or she is over-using
certain words.

The first step is to change the case of all the letters in our input file
to one case.  ``The'' and ``the'' are the same word when doing counting.

@example
$ tr '[:upper:]' '[:lower:]' < whats.gnu | ...
@end example

The next step is to get rid of punctuation.  Quoted words and unquoted words
should be treated identically; it's easiest to just get the punctuation out of
the way.

@example
$ tr '[:upper:]' '[:lower:]' < whats.gnu | tr -cd '[:alnum:]_ \n' | ...
@end example

The second @command{tr} command operates on the complement of the listed
characters, which are all the letters, the digits, the underscore, and
the blank.  The @samp{\n} represents the newline character; it has to
be left alone.  (The ASCII tab character should also be included for
good measure in a production script.)

At this point, we have data consisting of words separated by blank space.
The words only contain alphanumeric characters (and the underscore).  The
next step is break the data apart so that we have one word per line.  This
makes the counting operation much easier, as we will see shortly.

@example
$ tr '[:upper:]' '[:lower:]' < whats.gnu | tr -cd '[:alnum:]_ \n' |
> tr -s ' ' '\n' | ...
@end example

This command turns blanks into newlines.  The @option{-s} option squeezes
multiple newline characters in the output into just one, removing
blank lines.  (The @samp{>} is the shell's ``secondary prompt.''
This is what the shell prints when it notices you haven't finished
typing in all of a command.)

We now have data consisting of one word per line, no punctuation, all one
case.  We're ready to count each word:

@example
$ tr '[:upper:]' '[:lower:]' < whats.gnu | tr -cd '[:alnum:]_ \n' |
> tr -s ' ' '\n' | sort | uniq -c | ...
@end example

At this point, the data might look something like this:

@example
     60 a
      2 able
      6 about
      1 above
      2 accomplish
      1 acquire
      1 actually
      2 additional
@end example

The output is sorted by word, not by count!  What we want is the most
frequently used words first.  Fortunately, this is easy to accomplish,
with the help of two more @command{sort} options:

@table @code
@item -n
do a numeric sort, not a textual one

@item -r
reverse the order of the sort
@end table

The final pipeline looks like this:

@example
$ tr '[:upper:]' '[:lower:]' < whats.gnu | tr -cd '[:alnum:]_ \n' |
> tr -s ' ' '\n' | sort | uniq -c | sort -n -r
@print{}    156 the
@print{}     60 a
@print{}     58 to
@print{}     51 of
@print{}     51 and
@dots{}
@end example

Whew!  That's a lot to digest.  Yet, the same principles apply.  With six
commands, on two lines (really one long one split for convenience), we've
created a program that does something interesting and useful, in much
less time than we could have written a C program to do the same thing.

A minor modification to the above pipeline can give us a simple spelling
checker!  To determine if you've spelled a word correctly, all you have to
do is look it up in a dictionary.  If it is not there, then chances are
that your spelling is incorrect.  So, we need a dictionary.
The conventional location for a dictionary is @file{/usr/share/dict/words}.

Now, how to compare our file with the dictionary?  As before, we generate
a sorted list of words, one per line:

@example
$ tr '[:upper:]' '[:lower:]' < whats.gnu | tr -cd '[:alnum:]_ \n' |
> tr -s ' ' '\n' | sort -u | ...
@end example

Now, all we need is a list of words that are @emph{not} in the
dictionary.  Here is where the @command{comm} command comes in.
Unfortunately @command{comm} operates on sorted input and
@file{/usr/share/dict/words} is not sorted the way that @command{sort}
and @command{comm} normally use, so we first create a properly-sorted
copy of the dictionary and then run a pipeline that uses the copy.

@example
$ sort /usr/share/dict/words > sorted-words
$ tr '[:upper:]' '[:lower:]' < whats.gnu | tr -cd '[:alnum:]_ \n' |
> tr -s ' ' '\n' | sort -u |
> comm -23 - sorted-words
@end example

The @option{-2} and @option{-3} options eliminate lines that are only in the
dictionary (the second file), and lines that are in both files.  Lines
only in the first file (standard input, our stream of words), are
words that are not in the dictionary.  These are likely candidates for
spelling errors.  This pipeline was the first cut at a production
spelling checker on Unix.

There are some other tools that deserve brief mention.

@table @command
@item grep
search files for text that matches a regular expression

@item wc
count lines, words, characters

@item tee
a T-fitting for data pipes, copies data to files and to standard output

@item sed
the stream editor, an advanced tool

@item awk
a data manipulation language, another advanced tool
@end table

The software tools philosophy also espoused the following bit of
advice: ``Let someone else do the hard part.''  This means, take
something that gives you most of what you need, and then massage it the
rest of the way until it's in the form that you want.

To summarize:

@enumerate 1
@item
Each program should do one thing well.  No more, no less.

@item
Combining programs with appropriate plumbing leads to results where
the whole is greater than the sum of the parts.  It also leads to novel
uses of programs that the authors might never have imagined.

@item
Programs should never print extraneous header or trailer data, since these
could get sent on down a pipeline.  (A point we didn't mention earlier.)

@item
Let someone else do the hard part.

@item
Know your toolbox!  Use each program appropriately.  If you don't have an
appropriate tool, build one.
@end enumerate

All the programs discussed are available as described in
@uref{https://www.gnu.org/software/coreutils/coreutils.html,
GNU core utilities}.

None of what I have presented in this column is new.  The Software Tools
philosophy was first introduced in the book @cite{Software Tools}, by
Brian Kernighan and P.J. Plauger (Addison-Wesley, ISBN 0-201-03669-X).
This book showed how to write and use software tools.  It was written in
1976, using a preprocessor for FORTRAN named @command{ratfor} (RATional
FORtran).  At the time, C was not as ubiquitous as it is now; FORTRAN
was.  The last chapter presented a @command{ratfor} to FORTRAN
processor, written in @command{ratfor}.  @command{ratfor} looks an awful
lot like C; if you know C, you won't have any problem following the
code.

In 1981, the book was updated and made available as @cite{Software Tools
in Pascal} (Addison-Wesley, ISBN 0-201-10342-7).  Both books are
still in print and are well worth
reading if you're a programmer.  They certainly made a major change in
how I view programming.

The programs in both books are available from
@uref{https://www.cs.princeton.edu/~bwk/, Brian Kernighan's home page}.
For a number of years, there was an active
Software Tools Users Group, whose members had ported the original
@command{ratfor} programs to essentially every computer system with a
FORTRAN compiler.  The popularity of the group waned in the middle 1980s
as Unix began to spread beyond universities.

With the current proliferation of GNU code and other clones of Unix programs,
these programs now receive little attention; modern C versions are
much more efficient and do more than these programs do.  Nevertheless, as
exposition of good programming style, and evangelism for a still-valuable
philosophy, these books are unparalleled, and I recommend them highly.

Acknowledgment: I would like to express my gratitude to Brian Kernighan
of Bell Labs, the original Software Toolsmith, for reviewing this column.

@node GNU Free Documentation License
@appendix GNU Free Documentation License

@include fdl.texi

@node Concept index
@unnumbered Index

@printindex cp

@bye

@c Local variables:
@c texinfo-column-for-description: 32
@c End: