source: trunk/src/grep/doc/grep.texi@ 3603

\input texinfo  @c -*-texinfo-*-
@c %**start of header
@setfilename grep.info
@include version.texi
@settitle GNU Grep @value{VERSION}

@c Combine indices.
@syncodeindex ky cp
@syncodeindex pg cp
@syncodeindex tp cp
@defcodeindex op
@syncodeindex op cp
@syncodeindex vr cp
@c %**end of header

@documentencoding UTF-8
@c These two require Texinfo 5.0 or later, so use the older
@c equivalent @set variables supported in 4.11 and later.
@ignore
@codequotebacktick on
@codequoteundirected on
@end ignore
@set txicodequoteundirected
@set txicodequotebacktick
@iftex
@c TeX sometimes fails to hyphenate, so help it here.
@hyphenation{spec-i-fied}
@end iftex

@copying
This manual is for @command{grep}, a pattern matching engine.

Copyright @copyright{} 1999--2002, 2005, 2008--2021 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

@dircategory Text creation and manipulation
@direntry
* grep: (grep).                 Print lines that match patterns.
@end direntry

@titlepage
@title GNU Grep: Print lines that match patterns
@subtitle version @value{VERSION}, @value{UPDATED}
@author Alain Magloire et al.
@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage

@contents


@ifnottex
@node Top
@top grep

@command{grep} prints lines that contain a match for one or more patterns.

This manual is for version @value{VERSION} of GNU Grep.

@insertcopying
@end ifnottex

@menu
* Introduction::                Introduction.
* Invoking::                    Command-line options, environment, exit status.
* Regular Expressions::         Regular Expressions.
* Usage::                       Examples.
* Performance::                 Performance tuning.
* Reporting Bugs::              Reporting Bugs.
* Copying::                     License terms for this manual.
* Index::                       Combined index.
@end menu


@node Introduction
@chapter Introduction

@cindex searching for patterns

Given one or more patterns, @command{grep} searches input files
for matches to the patterns.
When it finds a match in a line,
it copies the line to standard output (by default),
or produces whatever other sort of output you have requested with options.

Though @command{grep} expects to do the matching on text,
it has no limits on input line length other than available memory,
and it can match arbitrary characters within a line.
If the final byte of an input file is not a newline,
@command{grep} silently supplies one.
Since newline is also a separator for the list of patterns,
there is no way to match newline characters in a text.


@node Invoking
@chapter Invoking @command{grep}

The general synopsis of the @command{grep} command line is

@example
grep [@var{option}...] [@var{patterns}] [@var{file}...]
@end example

@noindent
There can be zero or more @var{option} arguments, and zero or more
@var{file} arguments.  The @var{patterns} argument contains one or
more patterns separated by newlines, and is omitted when patterns are
given via the @samp{-e@ @var{patterns}} or @samp{-f@ @var{file}}
options.  Typically @var{patterns} should be quoted when
@command{grep} is used in a shell command.

@menu
* Command-line Options::        Short and long names, grouped by category.
* Environment Variables::       POSIX, GNU generic, and GNU grep specific.
* Exit Status::                 Exit status returned by @command{grep}.
* grep Programs::               @command{grep} programs.
@end menu

@node Command-line Options
@section Command-line Options

@command{grep} comes with a rich set of options:
some from POSIX and some being GNU extensions.
Long option names are always a GNU extension,
even for options that are from POSIX specifications.
Options that are specified by POSIX,
under their short names,
are explicitly marked as such
to facilitate POSIX-portable programming.
A few option names are provided
for compatibility with older or more exotic implementations.

@menu
* Generic Program Information::
* Matching Control::
* General Output Control::
* Output Line Prefix Control::
* Context Line Control::
* File and Directory Selection::
* Other Options::
@end menu

Several additional options control
which variant of the @command{grep} matching engine is used.
@xref{grep Programs}.

@node Generic Program Information
@subsection Generic Program Information

@table @option

@item --help
@opindex --help
@cindex usage summary, printing
Print a usage message briefly summarizing the command-line options
and the bug-reporting address, then exit.

@item -V
@itemx --version
@opindex -V
@opindex --version
@cindex version, printing
Print the version number of @command{grep} to the standard output stream.
This version number should be included in all bug reports.

@end table

@node Matching Control
@subsection Matching Control

@table @option

@item -e @var{patterns}
@itemx --regexp=@var{patterns}
@opindex -e
@opindex --regexp=@var{patterns}
@cindex patterns option
Use @var{patterns} as one or more patterns; newlines within
@var{patterns} separate each pattern from the next.
If this option is used multiple times or is combined with the
@option{-f} (@option{--file}) option, search for all patterns given.
Typically @var{patterns} should be quoted when @command{grep} is used
in a shell command.
(@option{-e} is specified by POSIX.)

@item -f @var{file}
@itemx --file=@var{file}
@opindex -f
@opindex --file
@cindex patterns from file
Obtain patterns from @var{file}, one per line.
If this option is used multiple times or is combined with the
@option{-e} (@option{--regexp}) option, search for all patterns given.
The empty file contains zero patterns, and therefore matches nothing.
(@option{-f} is specified by POSIX.)

@item -i
@itemx -y
@itemx --ignore-case
@opindex -i
@opindex -y
@opindex --ignore-case
@cindex case insensitive search
Ignore case distinctions in patterns and input data,
so that characters that differ only in case
match each other.  Although this is straightforward when letters
differ in case only via lowercase-uppercase pairs, the behavior is
unspecified in other situations.  For example, uppercase ``S'' has an
unusual lowercase counterpart ``Å¿'' (Unicode character U+017F, LATIN
SMALL LETTER LONG S) in many locales, and it is unspecified whether
this unusual character matches ``S'' or ``s'' even though uppercasing
it yields ``S''.  Another example: the lowercase German letter ``ß''
(U+00DF, LATIN SMALL LETTER SHARP S) is normally capitalized as the
two-character string ``SS'' but it does not match ``SS'', and it might
not match the uppercase letter ``ẞ'' (U+1E9E, LATIN CAPITAL LETTER
SHARP S) even though lowercasing the latter yields the former.

@option{-y} is an obsolete synonym that is provided for compatibility.
(@option{-i} is specified by POSIX.)

@item --no-ignore-case
@opindex --no-ignore-case
Do not ignore case distinctions in patterns and input data.  This is
the default.  This option is useful for passing to shell scripts that
already use @option{-i}, in order to cancel its effects because the
two options override each other.

@item -v
@itemx --invert-match
@opindex -v
@opindex --invert-match
@cindex invert matching
@cindex print non-matching lines
Invert the sense of matching, to select non-matching lines.
(@option{-v} is specified by POSIX.)

@item -w
@itemx --word-regexp
@opindex -w
@opindex --word-regexp
@cindex matching whole words
Select only those lines containing matches that form whole words.
The test is that the matching substring must either
be at the beginning of the line,
or preceded by a non-word constituent character.
Similarly,
it must be either at the end of the line
or followed by a non-word constituent character.
Word constituent characters are letters, digits, and the underscore.
This option has no effect if @option{-x} is also specified.

Because the @option{-w} option can match a substring that does not
begin and end with word constituents, it differs from surrounding a
regular expression with @samp{\<} and @samp{\>}.  For example, although
@samp{grep -w @@} matches a line containing only @samp{@@}, @samp{grep
'\<@@\>'} cannot match any line because @samp{@@} is not a
word constituent.  @xref{The Backslash Character and Special
Expressions}.

@item -x
@itemx --line-regexp
@opindex -x
@opindex --line-regexp
@cindex match the whole line
Select only those matches that exactly match the whole line.
For regular expression patterns, this is like parenthesizing each
pattern and then surrounding it with @samp{^} and @samp{$}.
(@option{-x} is specified by POSIX.)

@end table

@node General Output Control
@subsection General Output Control

@table @option

@item -c
@itemx --count
@opindex -c
@opindex --count
@cindex counting lines
Suppress normal output;
instead print a count of matching lines for each input file.
With the @option{-v} (@option{--invert-match}) option,
count non-matching lines.
(@option{-c} is specified by POSIX.)

@item --color[=@var{WHEN}]
@itemx --colour[=@var{WHEN}]
@opindex --color
@opindex --colour
@cindex highlight, color, colour
Surround the matched (non-empty) strings, matching lines, context lines,
file names, line numbers, byte offsets, and separators (for fields and
groups of context lines) with escape sequences to display them in color
on the terminal.
The colors are defined by the environment variable @env{GREP_COLORS}
and default to @samp{ms=01;31:mc=01;31:sl=:cx=:fn=35:ln=32:bn=32:se=36}
for bold red matched text, magenta file names, green line numbers,
green byte offsets, cyan separators, and default terminal colors otherwise.
The deprecated environment variable @env{GREP_COLOR} is still supported,
but its setting does not have priority;
it defaults to @samp{01;31} (bold red)
which only covers the color for matched text.
@var{WHEN} is @samp{never}, @samp{always}, or @samp{auto}.

@item -L
@itemx --files-without-match
@opindex -L
@opindex --files-without-match
@cindex files which don't match
Suppress normal output;
instead print the name of each input file from which
no output would normally have been printed.

@item -l
@itemx --files-with-matches
@opindex -l
@opindex --files-with-matches
@cindex names of matching files
Suppress normal output;
instead print the name of each input file from which
output would normally have been printed.
Scanning each input file stops upon first match.
(@option{-l} is specified by POSIX.)

@item -m @var{num}
@itemx --max-count=@var{num}
@opindex -m
@opindex --max-count
@cindex max-count
Stop after the first @var{num} selected lines.
If the input is standard input from a regular file,
and @var{num} selected lines are output,
@command{grep} ensures that the standard input is positioned
just after the last selected line before exiting,
regardless of the presence of trailing context lines.
This enables a calling process to resume a search.
For example, the following shell script makes use of it:

@example
while grep -m 1 'PATTERN'
do
  echo xxxx
done < FILE
@end example

But the following probably will not work because a pipe is not a regular
file:

@example
# This probably will not work.
cat FILE |
while grep -m 1 'PATTERN'
do
  echo xxxx
done
@end example

@cindex context lines
When @command{grep} stops after @var{num} selected lines,
it outputs any trailing context lines.
When the @option{-c} or @option{--count} option is also used,
@command{grep} does not output a count greater than @var{num}.
When the @option{-v} or @option{--invert-match} option is also used,
@command{grep} stops after outputting @var{num} non-matching lines.

@item -o
@itemx --only-matching
@opindex -o
@opindex --only-matching
@cindex only matching
Print only the matched (non-empty) parts of matching lines,
with each such part on a separate output line.
Output lines use the same delimiters as input, and delimiters are null
bytes if @option{-z} (@option{--null-data}) is also used (@pxref{Other
Options}).

@item -q
@itemx --quiet
@itemx --silent
@opindex -q
@opindex --quiet
@opindex --silent
@cindex quiet, silent
Quiet; do not write anything to standard output.
Exit immediately with zero status if any match is found,
even if an error was detected.
Also see the @option{-s} or @option{--no-messages} option.
(@option{-q} is specified by POSIX.)

@item -s
@itemx --no-messages
@opindex -s
@opindex --no-messages
@cindex suppress error messages
Suppress error messages about nonexistent or unreadable files.
Portability note:
unlike GNU @command{grep},
7th Edition Unix @command{grep} did not conform to POSIX,
because it lacked @option{-q}
and its @option{-s} option behaved like
GNU @command{grep}'s @option{-q} option.@footnote{Of course, 7th Edition
Unix predated POSIX by several years!}
USG-style @command{grep} also lacked @option{-q}
but its @option{-s} option behaved like GNU @command{grep}'s.
Portable shell scripts should avoid both
@option{-q} and @option{-s} and should redirect
standard and error output to @file{/dev/null} instead.
(@option{-s} is specified by POSIX.)

@end table

@node Output Line Prefix Control
@subsection Output Line Prefix Control

When several prefix fields are to be output,
the order is always file name, line number, and byte offset,
regardless of the order in which these options were specified.

@table @option

@item -b
@itemx --byte-offset
@opindex -b
@opindex --byte-offset
@cindex byte offset
Print the 0-based byte offset within the input file
before each line of output.
If @option{-o} (@option{--only-matching}) is specified,
print the offset of the matching part itself.

@item -H
@itemx --with-filename
@opindex -H
@opindex --with-filename
@cindex with filename prefix
Print the file name for each match.
This is the default when there is more than one file to search.

@item -h
@itemx --no-filename
@opindex -h
@opindex --no-filename
@cindex no filename prefix
Suppress the prefixing of file names on output.
This is the default when there is only one file
(or only standard input) to search.

@item --label=@var{LABEL}
@opindex --label
@cindex changing name of standard input
Display input actually coming from standard input
as input coming from file @var{LABEL}.
This can be useful for commands that transform a file's contents
before searching; e.g.:

@example
gzip -cd foo.gz | grep --label=foo -H 'some pattern'
@end example

@item -n
@itemx --line-number
@opindex -n
@opindex --line-number
@cindex line numbering
Prefix each line of output with the 1-based line number within its input file.
(@option{-n} is specified by POSIX.)

@item -T
@itemx --initial-tab
@opindex -T
@opindex --initial-tab
@cindex tab-aligned content lines
Make sure that the first character of actual line content lies on a tab stop,
so that the alignment of tabs looks normal.
This is useful with options that prefix their output to the actual content:
@option{-H}, @option{-n}, and @option{-b}.
This may also prepend spaces to output line numbers and byte offsets
so that lines from a single file all start at the same column.

@item -Z
@itemx --null
@opindex -Z
@opindex --null
@cindex zero-terminated file names
Output a zero byte (the ASCII NUL character)
instead of the character that normally follows a file name.
For example,
@samp{grep -lZ} outputs a zero byte after each file name
instead of the usual newline.
This option makes the output unambiguous,
even in the presence of file names containing unusual characters like newlines.
This option can be used with commands like
@samp{find -print0}, @samp{perl -0}, @samp{sort -z}, and @samp{xargs -0}
to process arbitrary file names,
even those that contain newline characters.

@end table

@node Context Line Control
@subsection Context Line Control

@cindex context lines
@dfn{Context lines} are non-matching lines that are near a matching line.
They are output only if one of the following options are used.
Regardless of how these options are set,
@command{grep} never outputs any given line more than once.
If the @option{-o} (@option{--only-matching}) option is specified,
these options have no effect and a warning is given upon their use.

@table @option

@item -A @var{num}
@itemx --after-context=@var{num}
@opindex -A
@opindex --after-context
@cindex after context
@cindex context lines, after match
Print @var{num} lines of trailing context after matching lines.

@item -B @var{num}
@itemx --before-context=@var{num}
@opindex -B
@opindex --before-context
@cindex before context
@cindex context lines, before match
Print @var{num} lines of leading context before matching lines.

@item -C @var{num}
@itemx -@var{num}
@itemx --context=@var{num}
@opindex -C
@opindex --context
@opindex -@var{num}
@cindex context lines
Print @var{num} lines of leading and trailing output context.

@item --group-separator=@var{string}
@opindex --group-separator
@cindex group separator
When @option{-A}, @option{-B} or @option{-C} are in use,
print @var{string} instead of @option{--} between groups of lines.

@item --no-group-separator
@opindex --group-separator
@cindex group separator
When @option{-A}, @option{-B} or @option{-C} are in use,
do not print a separator between groups of lines.

@end table

Here are some points about how @command{grep} chooses
the separator to print between prefix fields and line content:

@itemize @bullet
@item
Matching lines normally use @samp{:} as a separator
between prefix fields and actual line content.

@item
Context (i.e., non-matching) lines use @samp{-} instead.

@item
When context is not specified,
matching lines are simply output one right after another.

@item
When context is specified,
lines that are adjacent in the input form a group
and are output one right after another, while
by default a separator appears between non-adjacent groups.

@item
The default separator
is a @samp{--} line; its presence and appearance
can be changed with the options above.

@item
Each group may contain
several matching lines when they are close enough to each other
that two adjacent groups connect and can merge into a single
contiguous one.
@end itemize

@node File and Directory Selection
@subsection File and Directory Selection

@table @option

@item -a
@itemx --text
@opindex -a
@opindex --text
@cindex suppress binary data
@cindex binary files
Process a binary file as if it were text;
this is equivalent to the @samp{--binary-files=text} option.

@item --binary-files=@var{type}
@opindex --binary-files
@cindex binary files
If a file's data or metadata
indicate that the file contains binary data,
assume that the file is of type @var{type}.
Non-text bytes indicate binary data; these are either output bytes that are
improperly encoded for the current locale (@pxref{Environment
Variables}), or null input bytes when the
@option{-z} (@option{--null-data}) option is not given (@pxref{Other
Options}).

By default, @var{type} is @samp{binary}, and @command{grep}
suppresses output after null input binary data is discovered,
and suppresses output lines that contain improperly encoded data.
When some output is suppressed, @command{grep} follows any output
with a one-line message saying that a binary file matches.

If @var{type} is @samp{without-match},
when @command{grep} discovers null input binary data
it assumes that the rest of the file does not match;
this is equivalent to the @option{-I} option.

If @var{type} is @samp{text},
@command{grep} processes binary data as if it were text;
this is equivalent to the @option{-a} option.

When @var{type} is @samp{binary}, @command{grep} may treat non-text
bytes as line terminators even without the @option{-z}
(@option{--null-data}) option.  This means choosing @samp{binary}
versus @samp{text} can affect whether a pattern matches a file.  For
example, when @var{type} is @samp{binary} the pattern @samp{q$} might
match @samp{q} immediately followed by a null byte, even though this
is not matched when @var{type} is @samp{text}.  Conversely, when
@var{type} is @samp{binary} the pattern @samp{.} (period) might not
match a null byte.

@emph{Warning:} The @option{-a} (@option{--binary-files=text}) option
might output binary garbage, which can have nasty side effects if the
output is a terminal and if the terminal driver interprets some of it
as commands.  On the other hand, when reading files whose text
encodings are unknown, it can be helpful to use @option{-a} or to set
@samp{LC_ALL='C'} in the environment, in order to find more matches
even if the matches are unsafe for direct display.

@item -D @var{action}
@itemx --devices=@var{action}
@opindex -D
@opindex --devices
@cindex device search
If an input file is a device, FIFO, or socket, use @var{action} to process it.
If @var{action} is @samp{read},
all devices are read just as if they were ordinary files.
If @var{action} is @samp{skip},
devices, FIFOs, and sockets are silently skipped.
By default, devices are read if they are on the command line or if the
@option{-R} (@option{--dereference-recursive}) option is used, and are
skipped if they are encountered recursively and the @option{-r}
(@option{--recursive}) option is used.
This option has no effect on a file that is read via standard input.

@item -d @var{action}
@itemx --directories=@var{action}
@opindex -d
@opindex --directories
@cindex directory search
@cindex symbolic links
If an input file is a directory, use @var{action} to process it.
By default, @var{action} is @samp{read},
which means that directories are read just as if they were ordinary files
(some operating systems and file systems disallow this,
and will cause @command{grep}
to print error messages for every directory or silently skip them).
If @var{action} is @samp{skip}, directories are silently skipped.
If @var{action} is @samp{recurse},
@command{grep} reads all files under each directory, recursively,
following command-line symbolic links and skipping other symlinks;
this is equivalent to the @option{-r} option.

@item --exclude=@var{glob}
@opindex --exclude
@cindex exclude files
@cindex searching directory trees
Skip any command-line file with a name suffix that matches the pattern
@var{glob}, using wildcard matching; a name suffix is either the whole
name, or a trailing part that starts with a non-slash character
immediately after a slash (@samp{/}) in the name.
When searching recursively, skip any subfile whose base
name matches @var{glob}; the base name is the part after the last
slash.  A pattern can use
@samp{*}, @samp{?}, and @samp{[}...@samp{]} as wildcards,
and @code{\} to quote a wildcard or backslash character literally.

@item --exclude-from=@var{file}
@opindex --exclude-from
@cindex exclude files
@cindex searching directory trees
Skip files whose name matches any of the patterns
read from @var{file} (using wildcard matching as described
under @option{--exclude}).

@item --exclude-dir=@var{glob}
@opindex --exclude-dir
@cindex exclude directories
Skip any command-line directory with a name suffix that matches the
pattern @var{glob}.  When searching recursively, skip any subdirectory
whose base name matches @var{glob}.  Ignore any redundant trailing
slashes in @var{glob}.

@item -I
Process a binary file as if it did not contain matching data;
this is equivalent to the @samp{--binary-files=without-match} option.

@item --include=@var{glob}
@opindex --include
@cindex include files
@cindex searching directory trees
Search only files whose name matches @var{glob},
using wildcard matching as described under @option{--exclude}.
If contradictory @option{--include} and @option{--exclude} options are
given, the last matching one wins.  If no @option{--include} or
@option{--exclude} options match, a file is included unless the first
such option is @option{--include}.

@item -r
@itemx --recursive
@opindex -r
@opindex --recursive
@cindex recursive search
@cindex searching directory trees
@cindex symbolic links
For each directory operand,
read and process all files in that directory, recursively.
Follow symbolic links on the command line, but skip symlinks
that are encountered recursively.
Note that if no file operand is given, grep searches the working directory.
This is the same as the @samp{--directories=recurse} option.

@item -R
@itemx --dereference-recursive
@opindex -R
@opindex --dereference-recursive
@cindex recursive search
@cindex searching directory trees
@cindex symbolic links
For each directory operand, read and process all files in that
directory, recursively, following all symbolic links.

@end table

@node Other Options
@subsection Other Options

@table @option

@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{grep PAT --
-file1 file2} searches for the pattern PAT in the files named @file{-file1}
and @file{file2}.

@item --line-buffered
@opindex --line-buffered
@cindex line buffering
Use line buffering for standard output, regardless of output device.
By default, standard output is line buffered for interactive devices,
and is fully buffered otherwise.  With full buffering, the output
buffer is flushed when full; with line buffering, the buffer is also
flushed after every output line.  The buffer size is system dependent.

@item -U
@itemx --binary
@opindex -U
@opindex --binary
@cindex MS-Windows binary I/O
@cindex binary I/O
On platforms that distinguish between text and binary I/O,
use the latter when reading and writing files other
than the user's terminal, so that all input bytes are read and written
as-is.  This overrides the default behavior where @command{grep}
follows the operating system's advice whether to use text or binary
I/O@.  On MS-Windows when @command{grep} uses text I/O it reads a
carriage return--newline pair as a newline and a Control-Z as
end-of-file, and it writes a newline as a carriage return--newline
pair.

When using text I/O @option{--byte-offset} (@option{-b}) counts and
@option{--binary-files} heuristics apply to input data after text-I/O
processing.  Also, the @option{--binary-files} heuristics need not agree
with the @option{--binary} option; that is, they may treat the data as
text even if @option{--binary} is given, or vice versa.
@xref{File and Directory Selection}.

This option has no effect on GNU and other POSIX-compatible platforms,
which do not distinguish text from binary I/O.

@item -z
@itemx --null-data
@opindex -z
@opindex --null-data
@cindex zero-terminated lines
Treat input and output data as sequences of lines, each terminated by
a zero byte (the ASCII NUL character) instead of a newline.
Like the @option{-Z} or @option{--null} option,
this option can be used with commands like
@samp{sort -z} to process arbitrary file names.

@end table

@node Environment Variables
@section Environment Variables

The behavior of @command{grep} is affected
by the following environment variables.

@vindex LANGUAGE @r{environment variable}
@vindex LC_ALL @r{environment variable}
@vindex LC_MESSAGES @r{environment variable}
@vindex LANG @r{environment variable}
The locale for category @w{@code{LC_@var{foo}}}
is specified by examining the three environment variables
@env{LC_ALL}, @w{@env{LC_@var{foo}}}, and @env{LANG},
in that order.
The first of these variables that is set specifies the locale.
For example, if @env{LC_ALL} is not set,
but @env{LC_COLLATE} is set to @samp{pt_BR},
then the Brazilian Portuguese locale is used
for the @env{LC_COLLATE} category.
As a special case for @env{LC_MESSAGES} only, the environment variable
@env{LANGUAGE} can contain a colon-separated list of languages that
overrides the three environment variables that ordinarily specify
the @env{LC_MESSAGES} category.
The @samp{C} locale is used if none of these environment variables are set,
if the locale catalog is not installed,
or if @command{grep} was not compiled
with national language support (NLS).
The shell command @code{locale -a} lists locales that are currently available.

Many of the environment variables in the following list let you
control highlighting using
Select Graphic Rendition (SGR)
commands interpreted by the terminal or terminal emulator.
(See the
section
in the documentation of your text terminal
for permitted values and their meanings as character attributes.)
These substring values are integers in decimal representation
and can be concatenated with semicolons.
@command{grep} takes care of assembling the result
into a complete SGR sequence (@samp{\33[}...@samp{m}).
Common values to concatenate include
@samp{1} for bold,
@samp{4} for underline,
@samp{5} for blink,
@samp{7} for inverse,
@samp{39} for default foreground color,
@samp{30} to @samp{37} for foreground colors,
@samp{90} to @samp{97} for 16-color mode foreground colors,
@samp{38;5;0} to @samp{38;5;255}
for 88-color and 256-color modes foreground colors,
@samp{49} for default background color,
@samp{40} to @samp{47} for background colors,
@samp{100} to @samp{107} for 16-color mode background colors,
and @samp{48;5;0} to @samp{48;5;255}
for 88-color and 256-color modes background colors.

The two-letter names used in the @env{GREP_COLORS} environment variable
(and some of the others) refer to terminal ``capabilities,'' the ability
of a terminal to highlight text, or change its color, and so on.
These capabilities are stored in an online database and accessed by
the @code{terminfo} library.

@cindex environment variables

@table @env

@item GREP_COLOR
@vindex GREP_COLOR @r{environment variable}
@cindex highlight markers
This variable specifies the color used to highlight matched (non-empty) text.
It is deprecated in favor of @env{GREP_COLORS}, but still supported.
The @samp{mt}, @samp{ms}, and @samp{mc} capabilities of @env{GREP_COLORS}
have priority over it.
It can only specify the color used to highlight
the matching non-empty text in any matching line
(a selected line when the @option{-v} command-line option is omitted,
or a context line when @option{-v} is specified).
The default is @samp{01;31},
which means a bold red foreground text on the terminal's default background.

@item GREP_COLORS
@vindex GREP_COLORS @r{environment variable}
@cindex highlight markers
This variable specifies the colors and other attributes
used to highlight various parts of the output.
Its value is a colon-separated list of @code{terminfo} capabilities
that defaults to @samp{ms=01;31:mc=01;31:sl=:cx=:fn=35:ln=32:bn=32:se=36}
with the @samp{rv} and @samp{ne} boolean capabilities omitted (i.e., false).
Supported capabilities are as follows.

@table @code
@item sl=
@vindex sl GREP_COLORS @r{capability}
SGR substring for whole selected lines
(i.e.,
matching lines when the @option{-v} command-line option is omitted,
or non-matching lines when @option{-v} is specified).
If however the boolean @samp{rv} capability
and the @option{-v} command-line option are both specified,
it applies to context matching lines instead.
The default is empty (i.e., the terminal's default color pair).

@item cx=
@vindex cx GREP_COLORS @r{capability}
SGR substring for whole context lines
(i.e.,
non-matching lines when the @option{-v} command-line option is omitted,
or matching lines when @option{-v} is specified).
If however the boolean @samp{rv} capability
and the @option{-v} command-line option are both specified,
it applies to selected non-matching lines instead.
The default is empty (i.e., the terminal's default color pair).

@item rv
@vindex rv GREP_COLORS @r{capability}
Boolean value that reverses (swaps) the meanings of
the @samp{sl=} and @samp{cx=} capabilities
when the @option{-v} command-line option is specified.
The default is false (i.e., the capability is omitted).

@item mt=01;31
@vindex mt GREP_COLORS @r{capability}
SGR substring for matching non-empty text in any matching line
(i.e.,
a selected line when the @option{-v} command-line option is omitted,
or a context line when @option{-v} is specified).
Setting this is equivalent to setting both @samp{ms=} and @samp{mc=}
at once to the same value.
The default is a bold red text foreground over the current line background.

@item ms=01;31
@vindex ms GREP_COLORS @r{capability}
SGR substring for matching non-empty text in a selected line.
(This is used only when the @option{-v} command-line option is omitted.)
The effect of the @samp{sl=} (or @samp{cx=} if @samp{rv}) capability
remains active when this takes effect.
The default is a bold red text foreground over the current line background.

@item mc=01;31
@vindex mc GREP_COLORS @r{capability}
SGR substring for matching non-empty text in a context line.
(This is used only when the @option{-v} command-line option is specified.)
The effect of the @samp{cx=} (or @samp{sl=} if @samp{rv}) capability
remains active when this takes effect.
The default is a bold red text foreground over the current line background.

@item fn=35
@vindex fn GREP_COLORS @r{capability}
SGR substring for file names prefixing any content line.
The default is a magenta text foreground over the terminal's default background.

@item ln=32
@vindex ln GREP_COLORS @r{capability}
SGR substring for line numbers prefixing any content line.
The default is a green text foreground over the terminal's default background.

@item bn=32
@vindex bn GREP_COLORS @r{capability}
SGR substring for byte offsets prefixing any content line.
The default is a green text foreground over the terminal's default background.

@item se=36
@vindex fn GREP_COLORS @r{capability}
SGR substring for separators that are inserted
between selected line fields (@samp{:}),
between context line fields (@samp{-}),
and between groups of adjacent lines
when nonzero context is specified (@samp{--}).
The default is a cyan text foreground over the terminal's default background.

@item ne
@vindex ne GREP_COLORS @r{capability}
Boolean value that prevents clearing to the end of line
using Erase in Line (EL) to Right (@samp{\33[K})
each time a colorized item ends.
This is needed on terminals on which EL is not supported.
It is otherwise useful on terminals
for which the @code{back_color_erase}
(@code{bce}) boolean @code{terminfo} capability does not apply,
when the chosen highlight colors do not affect the background,
or when EL is too slow or causes too much flicker.
The default is false (i.e., the capability is omitted).
@end table

Note that boolean capabilities have no @samp{=}... part.
They are omitted (i.e., false) by default and become true when specified.


@item LC_ALL
@itemx LC_COLLATE
@itemx LANG
@vindex LC_ALL @r{environment variable}
@vindex LC_COLLATE @r{environment variable}
@vindex LANG @r{environment variable}
@cindex character type
@cindex national language support
@cindex NLS
These variables specify the locale for the @env{LC_COLLATE} category,
which might affect how range expressions like @samp{[a-z]} are
interpreted.

@item LC_ALL
@itemx LC_CTYPE
@itemx LANG
@vindex LC_ALL @r{environment variable}
@vindex LC_CTYPE @r{environment variable}
@vindex LANG @r{environment variable}
@cindex encoding error
@cindex null character
These variables specify the locale for the @env{LC_CTYPE} category,
which determines the type of characters,
e.g., which characters are whitespace.
This category also determines the character encoding.
@xref{Character Encoding}.

@item LANGUAGE
@itemx LC_ALL
@itemx LC_MESSAGES
@itemx LANG
@vindex LANGUAGE @r{environment variable}
@vindex LC_ALL @r{environment variable}
@vindex LC_MESSAGES @r{environment variable}
@vindex LANG @r{environment variable}
@cindex language of messages
@cindex message language
@cindex national language support
@cindex translation of message language
These variables specify the locale for the @env{LC_MESSAGES} category,
which determines the language that @command{grep} uses for messages.
The default @samp{C} locale uses American English messages.

@item POSIXLY_CORRECT
@vindex POSIXLY_CORRECT @r{environment variable}
If set, @command{grep} behaves as POSIX requires; otherwise,
@command{grep} behaves more like other GNU programs.
POSIX
requires that options that
follow file names must be treated as file names;
by default,
such options are permuted to the front of the operand list
and are treated as options.
Also, @env{POSIXLY_CORRECT} disables special handling of an
invalid bracket expression.  @xref{invalid-bracket-expr}.

@item _@var{N}_GNU_nonoption_argv_flags_
@vindex _@var{N}_GNU_nonoption_argv_flags_ @r{environment variable}
(Here @code{@var{N}} is @command{grep}'s numeric process ID.)
If the @var{i}th character of this environment variable's value is @samp{1},
do not consider the @var{i}th operand of @command{grep} to be an option,
even if it appears to be one.
A shell can put this variable in the environment for each command it runs,
specifying which operands are the results of file name wildcard expansion
and therefore should not be treated as options.
This behavior is available only with the GNU C library,
and only when @env{POSIXLY_CORRECT} is not set.

@end table

The @env{GREP_OPTIONS} environment variable of @command{grep} 2.20 and
earlier is no longer supported, as it caused problems when writing
portable scripts.  To make arbitrary changes to how @command{grep}
works, you can use an alias or script instead.  For example, if
@command{grep} is in the directory @samp{/usr/bin} you can prepend
@file{$HOME/bin} to your @env{PATH} and create an executable script
@file{$HOME/bin/grep} containing the following:

@example
#! /bin/sh
export PATH=/usr/bin
exec grep --color=auto --devices=skip "$@@"
@end example


@node Exit Status
@section Exit Status
@cindex exit status
@cindex return status

Normally the exit status is 0 if a line is selected, 1 if no lines
were selected, and 2 if an error occurred.  However, if the
@option{-q} or @option{--quiet} or @option{--silent} option is used
and a line is selected, the exit status is 0 even if an error
occurred.  Other @command{grep} implementations may exit with status
greater than 2 on error.

@node grep Programs
@section @command{grep} Programs
@cindex @command{grep} programs
@cindex variants of @command{grep}

@command{grep} searches the named input files
for lines containing a match to the given patterns.
By default, @command{grep} prints the matching lines.
A file named @file{-} stands for standard input.
If no input is specified, @command{grep} searches the working
directory @file{.} if given a command-line option specifying
recursion; otherwise, @command{grep} searches standard input.
There are four major variants of @command{grep},
controlled by the following options.

@table @option

@item -G
@itemx --basic-regexp
@opindex -G
@opindex --basic-regexp
@cindex matching basic regular expressions
Interpret patterns as basic regular expressions (BREs).
This is the default.

@item -E
@itemx --extended-regexp
@opindex -E
@opindex --extended-regexp
@cindex matching extended regular expressions
Interpret patterns as extended regular expressions (EREs).
(@option{-E} is specified by POSIX.)

@item -F
@itemx --fixed-strings
@opindex -F
@opindex --fixed-strings
@cindex matching fixed strings
Interpret patterns as fixed strings, not regular expressions.
(@option{-F} is specified by POSIX.)

@item -P
@itemx --perl-regexp
@opindex -P
@opindex --perl-regexp
@cindex matching Perl-compatible regular expressions
Interpret patterns as Perl-compatible regular expressions (PCREs).
PCRE support is here to stay, but consider this option experimental when
combined with the @option{-z} (@option{--null-data}) option, and note that
@samp{grep@ -P} may warn of unimplemented features.
@xref{Other Options}.

@end table

In addition,
two variant programs @command{egrep} and @command{fgrep} are available.
@command{egrep} is the same as @samp{grep@ -E}.
@command{fgrep} is the same as @samp{grep@ -F}.
Direct invocation as either
@command{egrep} or @command{fgrep} is deprecated,
but is provided to allow historical applications
that rely on them to run unmodified.


@node Regular Expressions
@chapter Regular Expressions
@cindex regular expressions

A @dfn{regular expression} is a pattern that describes a set of strings.
Regular expressions are constructed analogously to arithmetic expressions,
by using various operators to combine smaller expressions.
@command{grep} understands
three different versions of regular expression syntax:
basic (BRE), extended (ERE), and Perl-compatible (PCRE).
In GNU @command{grep},
there is no difference in available functionality between the basic and
extended syntaxes.
In other implementations, basic regular expressions are less powerful.
The following description applies to extended regular expressions;
differences for basic regular expressions are summarized afterwards.
Perl-compatible regular expressions give additional functionality, and
are documented in the @i{pcresyntax}(3) and @i{pcrepattern}(3) manual
pages, but work only if PCRE is available in the system.

@menu
* Fundamental Structure::
* Character Classes and Bracket Expressions::
* The Backslash Character and Special Expressions::
* Anchoring::
* Back-references and Subexpressions::
* Basic vs Extended::
* Character Encoding::
* Matching Non-ASCII::
@end menu

@node Fundamental Structure
@section Fundamental Structure

@cindex ordinary characters
@cindex special characters
In regular expressions, the characters @samp{.?*+@{|()[\^$} are
@dfn{special characters} and have uses described below.  All other
characters are @dfn{ordinary characters}, and each ordinary character
is a regular expression that matches itself.

@opindex .
@cindex dot
@cindex period
The period @samp{.} matches any single character.
It is unspecified whether @samp{.} matches an encoding error.

@cindex interval expressions
A regular expression may be followed by one of several
repetition operators; the operators beginning with @samp{@{}
are called @dfn{interval expressions}.

@table @samp

@item ?
@opindex ?
@cindex question mark
@cindex match expression at most once
The preceding item is optional and is matched at most once.

@item *
@opindex *
@cindex asterisk
@cindex match expression zero or more times
The preceding item is matched zero or more times.

@item +
@opindex +
@cindex plus sign
@cindex match expression one or more times
The preceding item is matched one or more times.

@item @{@var{n}@}
@opindex @{@var{n}@}
@cindex braces, one argument
@cindex match expression @var{n} times
The preceding item is matched exactly @var{n} times.

@item @{@var{n},@}
@opindex @{@var{n},@}
@cindex braces, second argument omitted
@cindex match expression @var{n} or more times
The preceding item is matched @var{n} or more times.

@item @{,@var{m}@}
@opindex @{,@var{m}@}
@cindex braces, first argument omitted
@cindex match expression at most @var{m} times
The preceding item is matched at most @var{m} times.
This is a GNU extension.

@item @{@var{n},@var{m}@}
@opindex @{@var{n},@var{m}@}
@cindex braces, two arguments
@cindex match expression from @var{n} to @var{m} times
The preceding item is matched at least @var{n} times, but not more than
@var{m} times.

@end table

The empty regular expression matches the empty string.
Two regular expressions may be concatenated;
the resulting regular expression
matches any string formed by concatenating two substrings
that respectively match the concatenated expressions.

Two regular expressions may be joined by the infix operator @samp{|};
the resulting regular expression
matches any string matching either alternate expression.

Repetition takes precedence over concatenation,
which in turn takes precedence over alternation.
A whole expression may be enclosed in parentheses
to override these precedence rules and form a subexpression.
An unmatched @samp{)} matches just itself.

@node Character Classes and Bracket Expressions
@section Character Classes and Bracket Expressions

@cindex bracket expression
@cindex character class
A @dfn{bracket expression} is a list of characters enclosed by @samp{[} and
@samp{]}.
It matches any single character in that list.
If the first character of the list is the caret @samp{^},
then it matches any character @strong{not} in the list,
and it is unspecified whether it matches an encoding error.
For example, the regular expression
@samp{[0123456789]} matches any single digit,
whereas @samp{[^()]} matches any single character that is not
an opening or closing parenthesis, and might or might not match an
encoding error.

@cindex range expression
Within a bracket expression, a @dfn{range expression} consists of two
characters separated by a hyphen.
It matches any single character that
sorts between the two characters, inclusive.
In the default C locale, the sorting sequence is the native character
order; for example, @samp{[a-d]} is equivalent to @samp{[abcd]}.
In other locales, the sorting sequence is not specified, and
@samp{[a-d]} might be equivalent to @samp{[abcd]} or to
@samp{[aBbCcDd]}, or it might fail to match any character, or the set of
characters that it matches might even be erratic.
To obtain the traditional interpretation
of bracket expressions, you can use the @samp{C} locale by setting the
@env{LC_ALL} environment variable to the value @samp{C}.

Finally, certain named classes of characters are predefined within
bracket expressions, as follows.
Their interpretation depends on the @env{LC_CTYPE} locale;
for example, @samp{[[:alnum:]]} means the character class of numbers and letters
in the current locale.

@cindex classes of characters
@cindex character classes
@table @samp

@item [:alnum:]
@opindex alnum @r{character class}
@cindex alphanumeric characters
Alphanumeric characters:
@samp{[:alpha:]} and @samp{[:digit:]}; in the @samp{C} locale and ASCII
character encoding, this is the same as @samp{[0-9A-Za-z]}.

@item [:alpha:]
@opindex alpha @r{character class}
@cindex alphabetic characters
Alphabetic characters:
@samp{[:lower:]} and @samp{[:upper:]}; in the @samp{C} locale and ASCII
character encoding, this is the same as @samp{[A-Za-z]}.

@item [:blank:]
@opindex blank @r{character class}
@cindex blank characters
Blank characters:
space and tab.

@item [:cntrl:]
@opindex cntrl @r{character class}
@cindex control characters
Control characters.
In ASCII, these characters have octal codes 000
through 037, and 177 (DEL).
In other character sets, these are
the equivalent characters, if any.

@item [:digit:]
@opindex digit @r{character class}
@cindex digit characters
@cindex numeric characters
Digits: @code{0 1 2 3 4 5 6 7 8 9}.

@item [:graph:]
@opindex graph @r{character class}
@cindex graphic characters
Graphical characters:
@samp{[:alnum:]} and @samp{[:punct:]}.

@item [:lower:]
@opindex lower @r{character class}
@cindex lower-case letters
Lower-case letters; in the @samp{C} locale and ASCII character
encoding, this is
@code{a b c d e f g h i j k l m n o p q r s t u v w x y z}.

@item [:print:]
@opindex print @r{character class}
@cindex printable characters
Printable characters:
@samp{[:alnum:]}, @samp{[:punct:]}, and space.

@item [:punct:]
@opindex punct @r{character class}
@cindex punctuation characters
Punctuation characters; in the @samp{C} locale and ASCII character
encoding, this is
@code{!@: " # $ % & ' ( ) * + , - .@: / : ; < = > ?@: @@ [ \ ] ^ _ ` @{ | @} ~}.

@item [:space:]
@opindex space @r{character class}
@cindex space characters
@cindex whitespace characters
Space characters: in the @samp{C} locale, this is
tab, newline, vertical tab, form feed, carriage return, and space.
@xref{Usage}, for more discussion of matching newlines.

@item [:upper:]
@opindex upper @r{character class}
@cindex upper-case letters
Upper-case letters: in the @samp{C} locale and ASCII character
encoding, this is
@code{A B C D E F G H I J K L M N O P Q R S T U V W X Y Z}.

@item [:xdigit:]
@opindex xdigit @r{character class}
@cindex xdigit class
@cindex hexadecimal digits
Hexadecimal digits:
@code{0 1 2 3 4 5 6 7 8 9 A B C D E F a b c d e f}.

@end table
Note that the brackets in these class names are
part of the symbolic names, and must be included in addition to
the brackets delimiting the bracket expression.

@anchor{invalid-bracket-expr}
If you mistakenly omit the outer brackets, and search for say, @samp{[:upper:]},
GNU @command{grep} prints a diagnostic and exits with status 2, on
the assumption that you did not intend to search for the nominally
equivalent regular expression: @samp{[:epru]}.
Set the @env{POSIXLY_CORRECT} environment variable to disable this feature.

Special characters lose their special meaning inside bracket expressions.

@table @samp
@item ]
ends the bracket expression if it's not the first list item.
So, if you want to make the @samp{]} character a list item,
you must put it first.

@item [.
represents the open collating symbol.

@item .]
represents the close collating symbol.

@item [=
represents the open equivalence class.

@item =]
represents the close equivalence class.

@item [:
represents the open character class symbol, and should be followed by a
valid character class name.

@item :]
represents the close character class symbol.

@item -
represents the range if it's not first or last in a list or the ending point
of a range.

@item ^
represents the characters not in the list.
If you want to make the @samp{^}
character a list item, place it anywhere but first.

@end table

@node The Backslash Character and Special Expressions
@section The Backslash Character and Special Expressions
@cindex backslash

The @samp{\} character followed by a special character is a regular
expression that matches the special character.
The @samp{\} character,
when followed by certain ordinary characters,
takes a special meaning:

@table @samp

@item \b
Match the empty string at the edge of a word.

@item \B
Match the empty string provided it's not at the edge of a word.

@item \<
Match the empty string at the beginning of a word.

@item \>
Match the empty string at the end of a word.

@item \w
Match word constituent, it is a synonym for @samp{[_[:alnum:]]}.

@item \W
Match non-word constituent, it is a synonym for @samp{[^_[:alnum:]]}.

@item \s
Match whitespace, it is a synonym for @samp{[[:space:]]}.

@item \S
Match non-whitespace, it is a synonym for @samp{[^[:space:]]}.

@end table

For example, @samp{\brat\b} matches the separate word @samp{rat},
@samp{\Brat\B} matches @samp{crate} but not @samp{furry rat}.

@node Anchoring
@section Anchoring
@cindex anchoring

The caret @samp{^} and the dollar sign @samp{$} are special characters that
respectively match the empty string at the beginning and end of a line.
They are termed @dfn{anchors}, since they force the match to be ``anchored''
to beginning or end of a line, respectively.

@node Back-references and Subexpressions
@section Back-references and Subexpressions
@cindex subexpression
@cindex back-reference

The back-reference @samp{\@var{n}},
where @var{n} is a single nonzero digit, matches
the substring previously matched by the @var{n}th parenthesized subexpression
of the regular expression.
For example, @samp{(a)\1} matches @samp{aa}.
If the parenthesized subexpression does not participate in the match,
the back-reference makes the whole match fail;
for example, @samp{(a)*\1} fails to match @samp{a}.
If the parenthesized subexpression matches more than one substring,
the back-reference refers to the last matched substring;
for example, @samp{^(ab*)*\1$} matches @samp{ababbabb} but not @samp{ababbab}.
When multiple regular expressions are given with
@option{-e} or from a file (@samp{-f @var{file}}),
back-references are local to each expression.

@xref{Known Bugs}, for some known problems with back-references.

@node Basic vs Extended
@section Basic vs Extended Regular Expressions
@cindex basic regular expressions

In basic regular expressions the characters @samp{?}, @samp{+},
@samp{@{}, @samp{|}, @samp{(}, and @samp{)} lose their special meaning;
instead use the backslashed versions @samp{\?}, @samp{\+}, @samp{\@{},
@samp{\|}, @samp{\(}, and @samp{\)}.  Also, a backslash is needed
before an interval expression's closing @samp{@}}, and an unmatched
@code{\)} is invalid.

Portable scripts should avoid the following constructs, as
POSIX says they produce undefined results:

@itemize @bullet
@item
Extended regular expressions that use back-references.
@item
Basic regular expressions that use @samp{\?}, @samp{\+}, or @samp{\|}.
@item
Empty parenthesized regular expressions like @samp{()}.
@item
Empty alternatives (as in, e.g, @samp{a|}).
@item
Repetition operators that immediately follow empty expressions,
unescaped @samp{$}, or other repetition operators.
@item
A backslash escaping an ordinary character (e.g., @samp{\S}),
unless it is a back-reference.
@item
An unescaped @samp{[} that is not part of a bracket expression.
@item
In extended regular expressions, an unescaped @samp{@{} that is not
part of an interval expression.
@end itemize

@cindex interval expressions
Traditional @command{egrep} did not support interval expressions and
some @command{egrep} implementations use @samp{\@{} and @samp{\@}} instead, so
portable scripts should avoid interval expressions in @samp{grep@ -E} patterns
and should use @samp{[@{]} to match a literal @samp{@{}.

GNU @command{grep@ -E} attempts to support traditional usage by
assuming that @samp{@{} is not special if it would be the start of an
invalid interval expression.
For example, the command
@samp{grep@ -E@ '@{1'} searches for the two-character string @samp{@{1}
instead of reporting a syntax error in the regular expression.
POSIX allows this behavior as an extension, but portable scripts
should avoid it.

@node Character Encoding
@section Character Encoding
@cindex character encoding

The @env{LC_CTYPE} locale specifies the encoding of characters in
patterns and data, that is, whether text is encoded in UTF-8, ASCII,
or some other encoding.  @xref{Environment Variables}.

In the @samp{C} or @samp{POSIX} locale, every character is encoded as
a single byte and every byte is a valid character.  In more-complex
encodings such as UTF-8, a sequence of multiple bytes may be needed to
represent a character, and some bytes may be encoding errors that do
not contribute to the representation of any character.  POSIX does not
specify the behavior of @command{grep} when patterns or input data
contain encoding errors or null characters, so portable scripts should
avoid such usage.  As an extension to POSIX, GNU @command{grep} treats
null characters like any other character.  However, unless the
@option{-a} (@option{--binary-files=text}) option is used, the
presence of null characters in input or of encoding errors in output
causes GNU @command{grep} to treat the file as binary and suppress
details about matches.  @xref{File and Directory Selection}.

Regardless of locale, the 103 characters in the POSIX Portable
Character Set (a subset of ASCII) are always encoded as a single byte,
and the 128 ASCII characters have their usual single-byte encodings on
all but oddball platforms.

@node Matching Non-ASCII
@section Matching Non-ASCII and Non-printable Characters
@cindex non-ASCII matching
@cindex non-printable matching

In a regular expression, non-ASCII and non-printable characters other
than newline are not special, and represent themselves.  For example,
in a locale using UTF-8 the command @samp{grep 'Λ@tie{}ω'} (where the
white space between @samp{Λ} and the @samp{ω} is a tab character)
searches for @samp{Λ} (Unicode character U+039B GREEK CAPITAL LETTER
LAMBDA), followed by a tab (U+0009 TAB), followed by @samp{ω} (U+03C9
GREEK SMALL LETTER OMEGA).

Suppose you want to limit your pattern to only printable characters
(or even only printable ASCII characters) to keep your script readable
or portable, but you also want to match specific non-ASCII or non-null
non-printable characters.  If you are using the @option{-P}
(@option{--perl-regexp}) option, PCREs give you several ways to do
this.  Otherwise, if you are using Bash, the GNU project's shell, you
can represent these characters via ANSI-C quoting.  For example, the
Bash commands @samp{grep $'Λ\tω'} and @samp{grep $'\u039B\t\u03C9'}
both search for the same three-character string @samp{Λ@tie{}ω}
mentioned earlier.  However, because Bash translates ANSI-C quoting
before @command{grep} sees the pattern, this technique should not be
used to match printable ASCII characters; for example, @samp{grep
$'\u005E'} is equivalent to @samp{grep '^'} and matches any line, not
just lines containing the character @samp{^} (U+005E CIRCUMFLEX
ACCENT).

Since PCREs and ANSI-C quoting are GNU extensions to POSIX, portable
shell scripts written in ASCII should use other methods to match
specific non-ASCII characters.  For example, in a UTF-8 locale the
command @samp{grep "$(printf '\316\233\t\317\211\n')"} is a portable
albeit hard-to-read alternative to Bash's @samp{grep $'Λ\tω'}.
However, none of these techniques will let you put a null character
directly into a command-line pattern; null characters can appear only
in a pattern specified via the @option{-f} (@option{--file}) option.

@node Usage
@chapter Usage

@cindex usage, examples
Here is an example command that invokes GNU @command{grep}:

@example
grep -i 'hello.*world' menu.h main.c
@end example

@noindent
This lists all lines in the files @file{menu.h} and @file{main.c} that
contain the string @samp{hello} followed by the string @samp{world};
this is because @samp{.*} matches zero or more characters within a line.
@xref{Regular Expressions}.
The @option{-i} option causes @command{grep}
to ignore case, causing it to match the line @samp{Hello, world!}, which
it would not otherwise match.

Here is a more complex example,
showing the location and contents of any line
containing @samp{f} and ending in @samp{.c},
within all files in the current directory whose names
start with non-@samp{.}, contain @samp{g}, and end in @samp{.h}.
The @option{-n} option outputs line numbers, the @option{--} argument
treats any later arguments as file names not options even if
@code{*g*.h} expands to a file name that starts with @samp{-},
and the empty file @file{/dev/null} causes file names to be output
even if only one file name happens to be of the form @samp{*g*.h}.

@example
grep -n -- 'f.*\.c$' *g*.h /dev/null
@end example

@noindent
Note that the regular expression syntax used in the pattern differs
from the globbing syntax that the shell uses to match file names.

@xref{Invoking}, for more details about
how to invoke @command{grep}.

@cindex using @command{grep}, Q&A
@cindex FAQ about @command{grep} usage
Here are some common questions and answers about @command{grep} usage.

@enumerate

@item
How can I list just the names of matching files?

@example
grep -l 'main' test-*.c
@end example

@noindent
lists names of @samp{test-*.c} files in the current directory whose contents
mention @samp{main}.

@item
How do I search directories recursively?

@example
grep -r 'hello' /home/gigi
@end example

@noindent
searches for @samp{hello} in all files
under the @file{/home/gigi} directory.
For more control over which files are searched,
use @command{find} and @command{grep}.
For example, the following command searches only C files:

@example
find /home/gigi -name '*.c' ! -type d \
  -exec grep -H 'hello' '@{@}' +
@end example

This differs from the command:

@example
grep -H 'hello' /home/gigi/*.c
@end example

which merely looks for @samp{hello} in non-hidden C files in
@file{/home/gigi} whose names end in @samp{.c}.
The @command{find} command line above is more similar to the command:

@example
grep -r --include='*.c' 'hello' /home/gigi
@end example

@item
What if a pattern or file has a leading @samp{-}?

@example
grep -- '--cut here--' *
@end example

@noindent
searches for all lines matching @samp{--cut here--}.
Without @option{--},
@command{grep} would attempt to parse @samp{--cut here--} as a list of
options, and there would be similar problems with any file names
beginning with @samp{-}.

Alternatively, you can prevent misinterpretation of leading @samp{-}
by using @option{-e} for patterns and leading @samp{./} for files:

@example
grep -e '--cut here--' ./*
@end example

@item
Suppose I want to search for a whole word, not a part of a word?

@example
grep -w 'hello' test*.log
@end example

@noindent
searches only for instances of @samp{hello} that are entire words;
it does not match @samp{Othello}.
For more control, use @samp{\<} and
@samp{\>} to match the start and end of words.
For example:

@example
grep 'hello\>' test*.log
@end example

@noindent
searches only for words ending in @samp{hello}, so it matches the word
@samp{Othello}.

@item
How do I output context around the matching lines?

@example
grep -C 2 'hello' test*.log
@end example

@noindent
prints two lines of context around each matching line.

@item
How do I force @command{grep} to print the name of the file?

Append @file{/dev/null}:

@example
grep 'eli' /etc/passwd /dev/null
@end example

gets you:

@example
/etc/passwd:eli:x:2098:1000:Eli Smith:/home/eli:/bin/bash
@end example

Alternatively, use @option{-H}, which is a GNU extension:

@example
grep -H 'eli' /etc/passwd
@end example

@item
Why do people use strange regular expressions on @command{ps} output?

@example
ps -ef | grep '[c]ron'
@end example

If the pattern had been written without the square brackets, it would
have matched not only the @command{ps} output line for @command{cron},
but also the @command{ps} output line for @command{grep}.
Note that on some platforms,
@command{ps} limits the output to the width of the screen;
@command{grep} does not have any limit on the length of a line
except the available memory.

@item
Why does @command{grep} report ``Binary file matches''?

If @command{grep} listed all matching ``lines'' from a binary file, it
would probably generate output that is not useful, and it might even
muck up your display.
So GNU @command{grep} suppresses output from
files that appear to be binary files.
To force GNU @command{grep}
to output lines even from files that appear to be binary, use the
@option{-a} or @samp{--binary-files=text} option.
To eliminate the
``Binary file matches'' messages, use the @option{-I} or
@samp{--binary-files=without-match} option,
or the @option{-s} or @option{--no-messages} option.

@item
Why doesn't @samp{grep -lv} print non-matching file names?

@samp{grep -lv} lists the names of all files containing one or more
lines that do not match.
To list the names of all files that contain no
matching lines, use the @option{-L} or @option{--files-without-match}
option.

@item
I can do ``OR'' with @samp{|}, but what about ``AND''?

@example
grep 'paul' /etc/motd | grep 'franc,ois'
@end example

@noindent
finds all lines that contain both @samp{paul} and @samp{franc,ois}.

@item
Why does the empty pattern match every input line?

The @command{grep} command searches for lines that contain strings
that match a pattern.  Every line contains the empty string, so an
empty pattern causes @command{grep} to find a match on each line.  It
is not the only such pattern: @samp{^}, @samp{$}, and many
other patterns cause @command{grep} to match every line.

To match empty lines, use the pattern @samp{^$}.  To match blank
lines, use the pattern @samp{^[[:blank:]]*$}.  To match no lines at
all, use the command @samp{grep -f /dev/null}.

@item
How can I search in both standard input and in files?

Use the special file name @samp{-}:

@example
cat /etc/passwd | grep 'alain' - /etc/motd
@end example

@item
Why is this back-reference failing?

@example
echo 'ba' | grep -E '(a)\1|b\1'
@end example

This outputs an error message, because the second @samp{\1}
has nothing to refer back to, meaning it will never match anything.

@item
How can I match across lines?

Standard grep cannot do this, as it is fundamentally line-based.
Therefore, merely using the @code{[:space:]} character class does not
match newlines in the way you might expect.

With the GNU @command{grep} option @option{-z} (@option{--null-data}), each
input and output ``line'' is null-terminated; @pxref{Other Options}.  Thus,
you can match newlines in the input, but typically if there is a match
the entire input is output, so this usage is often combined with
output-suppressing options like @option{-q}, e.g.:

@example
printf 'foo\nbar\n' | grep -z -q 'foo[[:space:]]\+bar'
@end example

If this does not suffice, you can transform the input
before giving it to @command{grep}, or turn to @command{awk},
@command{sed}, @command{perl}, or many other utilities that are
designed to operate across lines.

@item
What do @command{grep}, @command{fgrep}, and @command{egrep} stand for?

The name @command{grep} comes from the way line editing was done on Unix.
For example,
@command{ed} uses the following syntax
to print a list of matching lines on the screen:

@example
global/regular expression/print
g/re/p
@end example

@command{fgrep} stands for Fixed @command{grep};
@command{egrep} stands for Extended @command{grep}.

@end enumerate


@node Performance
@chapter Performance

@cindex performance
Typically @command{grep} is an efficient way to search text.  However,
it can be quite slow in some cases, and it can search large files
where even minor performance tweaking can help significantly.
Although the algorithm used by @command{grep} is an implementation
detail that can change from release to release, understanding its
basic strengths and weaknesses can help you improve its performance.

The @command{grep} command operates partly via a set of automata that
are designed for efficiency, and partly via a slower matcher that
takes over when the fast matchers run into unusual features like
back-references.  When feasible, the Boyer--Moore fast string
searching algorithm is used to match a single fixed pattern, and the
Aho--Corasick algorithm is used to match multiple fixed patterns.

@cindex locales
Generally speaking @command{grep} operates more efficiently in
single-byte locales, since it can avoid the special processing needed
for multi-byte characters.  If your patterns will work just as well
that way, setting @env{LC_ALL} to a single-byte locale can help
performance considerably.  Setting @samp{LC_ALL='C'} can be
particularly efficient, as @command{grep} is tuned for that locale.

@cindex case insensitive search
Outside the @samp{C} locale, case-insensitive search, and search for
bracket expressions like @samp{[a-z]} and @samp{[[=a=]b]}, can be
surprisingly inefficient due to difficulties in fast portable access to
concepts like multi-character collating elements.

@cindex back-references
A back-reference such as @samp{\1} can hurt performance significantly
in some cases, since back-references cannot in general be implemented
via a finite state automaton, and instead trigger a backtracking
algorithm that can be quite inefficient.  For example, although the
pattern @samp{^(.*)\1@{14@}(.*)\2@{13@}$} matches only lines whose
lengths can be written as a sum @math{15x + 14y} for nonnegative
integers @math{x} and @math{y}, the pattern matcher does not perform
linear Diophantine analysis and instead backtracks through all
possible matching strings, using an algorithm that is exponential in
the worst case.

@cindex holes in files
On some operating systems that support files with holes---large
regions of zeros that are not physically present on secondary
storage---@command{grep} can skip over the holes efficiently without
needing to read the zeros.  This optimization is not available if the
@option{-a} (@option{--binary-files=text}) option is used (@pxref{File and
Directory Selection}), unless the @option{-z} (@option{--null-data})
option is also used (@pxref{Other Options}).

For more about the algorithms used by @command{grep} and about
related string matching algorithms, see:

@frenchspacing on
@itemize @bullet
@item
Aho AV. Algorithms for finding patterns in strings.
In: van Leeuwen J. @emph{Handbook of Theoretical Computer Science}, vol. A.
New York: Elsevier; 1990. p. 255--300.
This surveys classic string matching algorithms, some of which are
used by @command{grep}.

@item
Aho AV, Corasick MJ. Efficient string matching: an aid to bibliographic search.
@emph{CACM}. 1975;18(6):333--40.
@url{https://dx.doi.org/10.1145/360825.360855}.
This introduces the Aho--Corasick algorithm.

@item
Boyer RS, Moore JS. A fast string searching algorithm.
@emph{CACM}. 1977;20(10):762--72.
@url{https://dx.doi.org/10.1145/359842.359859}.
This introduces the Boyer--Moore algorithm.

@item
Faro S, Lecroq T. The exact online string matching problem: a review
of the most recent results.
@emph{ACM Comput Surv}. 2013;45(2):13.
@url{https://dx.doi.org/10.1145/2431211.2431212}.
This surveys string matching algorithms that might help improve the
performance of @command{grep} in the future.
@end itemize
@frenchspacing off

@node Reporting Bugs
@chapter Reporting bugs

@cindex bugs, reporting
Bug reports can be found at the
@url{https://debbugs.gnu.org/cgi/pkgreport.cgi?package=grep,
GNU bug report logs for @command{grep}}.
If you find a bug not listed there, please email it to
@email{bug-grep@@gnu.org} to create a new bug report.

@menu
* Known Bugs::
@end menu

@node Known Bugs
@section Known Bugs
@cindex Bugs, known

Large repetition counts in the @samp{@{n,m@}} construct may cause
@command{grep} to use lots of memory.
In addition, certain other
obscure regular expressions require exponential time and
space, and may cause @command{grep} to run out of memory.

Back-references can greatly slow down matching, as they can generate
exponentially many matching possibilities that can consume both time
and memory to explore.  Also, the POSIX specification for
back-references is at times unclear.  Furthermore, many regular
expression implementations have back-reference bugs that can cause
programs to return incorrect answers or even crash, and fixing these
bugs has often been low-priority: for example, as of 2021 the
@url{https://sourceware.org/bugzilla/,GNU C library bug database}
contained back-reference bugs
@url{https://sourceware.org/bugzilla/show_bug.cgi?id=52,,52},
@url{https://sourceware.org/bugzilla/show_bug.cgi?id=10844,,10844},
@url{https://sourceware.org/bugzilla/show_bug.cgi?id=11053,,11053},
@url{https://sourceware.org/bugzilla/show_bug.cgi?id=24269,,24269}
and @url{https://sourceware.org/bugzilla/show_bug.cgi?id=25322,,25322},
with little sign of forthcoming fixes.  Luckily,
back-references are rarely useful and it should be little trouble to
avoid them in practical applications.


@node Copying
@chapter Copying
@cindex copying

GNU @command{grep} is licensed under the GNU GPL, which makes it @dfn{free
software}.

The ``free'' in ``free software'' refers to liberty, not price.  As
some GNU project advocates like to point out, think of ``free speech''
rather than ``free beer''.  In short, you have the right (freedom) to
run and change @command{grep} and distribute it to other people, and---if you
want---charge money for doing either.  The important restriction is
that you have to grant your recipients the same rights and impose the
same restrictions.

This general method of licensing software is sometimes called
@dfn{open source}.  The GNU project prefers the term ``free software''
for reasons outlined at
@url{https://www.gnu.org/philosophy/open-source-misses-the-point.html}.

This manual is free documentation in the same sense.  The
documentation license is included below.  The license for the program
is available with the source code, or at
@url{https://www.gnu.org/licenses/gpl.html}.

@menu
* GNU Free Documentation License::
@end menu

@node GNU Free Documentation License
@section GNU Free Documentation License

@include fdl.texi


@node Index
@unnumbered Index

@printindex cp

@bye