1GREP(1) General Commands Manual GREP(1)
2
6 grep, egrep, fgrep - print lines matching a pattern
7
9 grep [options] PATTERN [FILE...]
10 grep [options] [-e PATTERN | -f FILE] [FILE...]
11
13 Grep searches the named input FILEs (or standard input if no files are
14 named, or the file name - is given) for lines containing a match to the
15 given PATTERN. By default, grep prints the matching lines.
16 In addition, two variant programs egrep and fgrep are available. Egrep
18 is the same as grep -E. Fgrep is the same as grep -F.
19
21 -A NUM, --after-context=NUM
22 Print NUM lines of trailing context after matching lines.
23 Places a line containing -- between contiguous groups of
24 matches.
25 -a, --text
27 Process a binary file as if it were text; this is equivalent to
28 the --binary-files=text option.
29 -B NUM, --before-context=NUM
31 Print NUM lines of leading context before matching lines.
32 Places a line containing -- between contiguous groups of
33 matches.
34 -C NUM, --context=NUM
36 Print NUM lines of output context. Places a line containing --
37 between contiguous groups of matches.
38 -b, --byte-offset
40 Print the byte offset within the input file before each line of
41 output.
42 --binary-files=TYPE
44 If the first few bytes of a file indicate that the file contains
45 binary data, assume that the file is of type TYPE. By default,
46 TYPE is binary, and grep normally outputs either a one-line mes‐
47 sage saying that a binary file matches, or no message if there
48 is no match. If TYPE is without-match, grep assumes that a
49 binary file does not match; this is equivalent to the -I option.
50 If TYPE is text, grep processes a binary file as if it were
51 text; this is equivalent to the -a option. Warning: grep
52 --binary-files=text might output binary garbage, which can have
53 nasty side effects if the output is a terminal and if the termi‐
54 nal driver interprets some of it as commands.
55 --colour[=WHEN], --color[=WHEN]
57 Surround the matching string with the marker find in GREP_COLOR
58 environment variable. WHEN may be `never', `always', or `auto'
59 -c, --count
61 Suppress normal output; instead print a count of matching lines
62 for each input file. With the -v, --invert-match option (see
63 below), count non-matching lines.
64 -D ACTION, --devices=ACTION
66 If an input file is a device, FIFO or socket, use ACTION to
67 process it. By default, ACTION is read, which means that
68 devices are read just as if they were ordinary files. If ACTION
69 is skip, devices are silently skipped.
70 -d ACTION, --directories=ACTION
72 If an input file is a directory, use ACTION to process it. By
73 default, ACTION is read, which means that directories are read
74 just as if they were ordinary files. If ACTION is skip, direc‐
75 tories are silently skipped. If ACTION is recurse, grep reads
76 all files under each directory, recursively; this is equivalent
77 to the -r option.
78 -E, --extended-regexp
80 Interpret PATTERN as an extended regular expression (see below).
81 -e PATTERN, --regexp=PATTERN
83 Use PATTERN as the pattern; useful to protect patterns beginning
84 with -.
85 -F, --fixed-strings
87 Interpret PATTERN as a list of fixed strings, separated by new‐
88 lines, any of which is to be matched.
89 -P, --perl-regexp
91 Interpret PATTERN as a Perl regular expression.
92 -f FILE, --file=FILE
94 Obtain patterns from FILE, one per line. The empty file con‐
95 tains zero patterns, and therefore matches nothing.
96 -G, --basic-regexp
98 Interpret PATTERN as a basic regular expression (see below).
99 This is the default.
100 -H, --with-filename
102 Print the filename for each match.
103 -h, --no-filename
105 Suppress the prefixing of filenames on output when multiple
106 files are searched.
107 --help Output a brief help message.
109 -I Process a binary file as if it did not contain matching data;
111 this is equivalent to the --binary-files=without-match option.
112 -i, --ignore-case
114 Ignore case distinctions in both the PATTERN and the input
115 files.
116 -L, --files-without-match
118 Suppress normal output; instead print the name of each input
119 file from which no output would normally have been printed. The
120 scanning will stop on the first match.
121 -l, --files-with-matches
123 Suppress normal output; instead print the name of each input
124 file from which output would normally have been printed. The
125 scanning will stop on the first match.
126 -m NUM, --max-count=NUM
128 Stop reading a file after NUM matching lines. If the input is
129 standard input from a regular file, and NUM matching lines are
130 output, grep ensures that the standard input is positioned to
131 just after the last matching line before exiting, regardless of
132 the presence of trailing context lines. This enables a calling
133 process to resume a search. When grep stops after NUM matching
134 lines, it outputs any trailing context lines. When the -c or
135 --count option is also used, grep does not output a count
136 greater than NUM. When the -v or --invert-match option is also
137 used, grep stops after outputting NUM non-matching lines.
138 --mmap If possible, use the mmap(2) system call to read input, instead
140 of the default read(2) system call. In some situations, --mmap
141 yields better performance. However, --mmap can cause undefined
142 behavior (including core dumps) if an input file shrinks while
143 grep is operating, or if an I/O error occurs.
144 -n, --line-number
146 Prefix each line of output with the line number within its input
147 file.
148 -o, --only-matching
150 Show only the part of a matching line that matches PATTERN.
151 --label=LABEL
153 Displays input actually coming from standard input as input com‐
154 ing from file LABEL. This is especially useful for tools like
155 zgrep, e.g. gzip -cd foo.gz |grep --label=foo something
156 --line-buffered
158 Use line buffering, it can be a performance penality.
159 -q, --quiet, --silent
161 Quiet; do not write anything to standard output. Exit immedi‐
162 ately with zero status if any match is found, even if an error
163 was detected. Also see the -s or --no-messages option.
164 -R, -r, --recursive
166 Read all files under each directory, recursively; this is equiv‐
167 alent to the -d recurse option.
168 --include=PATTERN
170 Recurse in directories only searching file matching PATTERN.
171 --exclude=PATTERN
173 Recurse in directories skip file matching PATTERN.
174 -s, --no-messages
176 Suppress error messages about nonexistent or unreadable files.
177 Portability note: unlike GNU grep, traditional grep did not con‐
178 form to POSIX.2, because traditional grep lacked a -q option and
179 its -s option behaved like GNU grep's -q option. Shell scripts
180 intended to be portable to traditional grep should avoid both -q
181 and -s and should redirect output to /dev/null instead.
182 -U, --binary
184 Treat the file(s) as binary. By default, under MS-DOS and MS-
185 Windows, grep guesses the file type by looking at the contents
186 of the first 32KB read from the file. If grep decides the file
187 is a text file, it strips the CR characters from the original
188 file contents (to make regular expressions with ^ and $ work
189 correctly). Specifying -U overrules this guesswork, causing all
190 files to be read and passed to the matching mechanism verbatim;
191 if the file is a text file with CR/LF pairs at the end of each
192 line, this will cause some regular expressions to fail. This
193 option has no effect on platforms other than MS-DOS and MS-Win‐
194 dows.
195 -u, --unix-byte-offsets
197 Report Unix-style byte offsets. This switch causes grep to
198 report byte offsets as if the file were Unix-style text file,
199 i.e. with CR characters stripped off. This will produce results
200 identical to running grep on a Unix machine. This option has no
201 effect unless -b option is also used; it has no effect on plat‐
202 forms other than MS-DOS and MS-Windows.
203 -V, --version
205 Print the version number of grep to standard error. This ver‐
206 sion number should be included in all bug reports (see below).
207 -v, --invert-match
209 Invert the sense of matching, to select non-matching lines.
210 -w, --word-regexp
212 Select only those lines containing matches that form whole
213 words. The test is that the matching substring must either be
214 at the beginning of the line, or preceded by a non-word con‐
215 stituent character. Similarly, it must be either at the end of
216 the line or followed by a non-word constituent character. Word-
217 constituent characters are letters, digits, and the underscore.
218 -x, --line-regexp
220 Select only those matches that exactly match the whole line.
221 -y Obsolete synonym for -i.
223 -Z, --null
225 Output a zero byte (the ASCII NUL character) instead of the
226 character that normally follows a file name. For example, grep
227 -lZ outputs a zero byte after each file name instead of the
228 usual newline. This option makes the output unambiguous, even
229 in the presence of file names containing unusual characters like
230 newlines. This option can be used with commands like find
231 -print0, perl -0, sort -z, and xargs -0 to process arbitrary
232 file names, even those that contain newline characters.
233
235 A regular expression is a pattern that describes a set of strings.
236 Regular expressions are constructed analogously to arithmetic expres‐
237 sions, by using various operators to combine smaller expressions.
238 Grep understands two different versions of regular expression syntax:
240 “basic” and “extended.” In GNU grep, there is no difference in avail‐
241 able functionality using either syntax. In other implementations,
242 basic regular expressions are less powerful. The following description
243 applies to extended regular expressions; differences for basic regular
244 expressions are summarized afterwards.
245 The fundamental building blocks are the regular expressions that match
247 a single character. Most characters, including all letters and digits,
248 are regular expressions that match themselves. Any metacharacter with
249 special meaning may be quoted by preceding it with a backslash.
250 A bracket expression is a list of characters enclosed by [ and ]. It
252 matches any single character in that list; if the first character of
253 the list is the caret ^ then it matches any character not in the list.
254 For example, the regular expression [0123456789] matches any single
255 digit.
256 Within a bracket expression, a range expression consists of two charac‐
258 ters separated by a hyphen. It matches any single character that sorts
259 between the two characters, inclusive, using the locale's collating
260 sequence and character set. For example, in the default C locale,
261 [a-d] is equivalent to [abcd]. Many locales sort characters in dictio‐
262 nary order, and in these locales [a-d] is typically not equivalent to
263 [abcd]; it might be equivalent to [aBbCcDd], for example. To obtain
264 the traditional interpretation of bracket expressions, you can use the
265 C locale by setting the LC_ALL environment variable to the value C.
266 Finally, certain named classes of characters are predefined within
268 bracket expressions, as follows. Their names are self explanatory, and
269 they are [:alnum:], [:alpha:], [:cntrl:], [:digit:], [:graph:],
270 [:lower:], [:print:], [:punct:], [:space:], [:upper:], and [:xdigit:].
271 For example, [[:alnum:]] means [0-9A-Za-z], except the latter form
272 depends upon the C locale and the ASCII character encoding, whereas the
273 former is independent of locale and character set. (Note that the
274 brackets in these class names are part of the symbolic names, and must
275 be included in addition to the brackets delimiting the bracket list.)
276 Most metacharacters lose their special meaning inside lists. To
277 include a literal ] place it first in the list. Similarly, to include
278 a literal ^ place it anywhere but first. Finally, to include a literal
279 - place it last.
280 The period . matches any single character. The symbol \w is a synonym
282 for [[:alnum:]] and \W is a synonym for [^[:alnum]].
283 The caret ^ and the dollar sign $ are metacharacters that respectively
285 match the empty string at the beginning and end of a line. The symbols
286 \< and \> respectively match the empty string at the beginning and end
287 of a word. The symbol \b matches the empty string at the edge of a
288 word, and \B matches the empty string provided it's not at the edge of
289 a word.
290 A regular expression may be followed by one of several repetition oper‐
292 ators:
293 ? The preceding item is optional and matched at most once.
294 * The preceding item will be matched zero or more times.
295 + The preceding item will be matched one or more times.
296 {n} The preceding item is matched exactly n times.
297 {n,} The preceding item is matched n or more times.
298 {n,m} The preceding item is matched at least n times, but not more
299 than m times.
300 Two regular expressions may be concatenated; the resulting regular
302 expression matches any string formed by concatenating two substrings
303 that respectively match the concatenated subexpressions.
304 Two regular expressions may be joined by the infix operator |; the
306 resulting regular expression matches any string matching either subex‐
307 pression.
308 Repetition takes precedence over concatenation, which in turn takes
310 precedence over alternation. A whole subexpression may be enclosed in
311 parentheses to override these precedence rules.
312 The backreference \n, where n is a single digit, matches the substring
314 previously matched by the nth parenthesized subexpression of the regu‐
315 lar expression.
316 In basic regular expressions the metacharacters ?, +, {, |, (, and )
318 lose their special meaning; instead use the backslashed versions \?,
319 \+, \{, \|, \(, and \).
320 Traditional egrep did not support the { metacharacter, and some egrep
322 implementations support \{ instead, so portable scripts should avoid {
323 in egrep patterns and should use [{] to match a literal {.
324 GNU egrep attempts to support traditional usage by assuming that { is
326 not special if it would be the start of an invalid interval specifica‐
327 tion. For example, the shell command egrep '{1' searches for the two-
328 character string {1 instead of reporting a syntax error in the regular
329 expression. POSIX.2 allows this behavior as an extension, but portable
330 scripts should avoid it.
331
333 Grep's behavior is affected by the following environment variables.
334 A locale LC_foo is specified by examining the three environment vari‐
336 ables LC_ALL, LC_foo, LANG, in that order. The first of these vari‐
337 ables that is set specifies the locale. For example, if LC_ALL is not
338 set, but LC_MESSAGES is set to pt_BR, then Brazilian Portuguese is used
339 for the LC_MESSAGES locale. The C locale is used if none of these
340 environment variables are set, or if the locale catalog is not
341 installed, or if grep was not compiled with national language support
342 (NLS).
343 GREP_OPTIONS
345 This variable specifies default options to be placed in front of
346 any explicit options. For example, if GREP_OPTIONS is
347 '--binary-files=without-match --directories=skip', grep behaves
348 as if the two options --binary-files=without-match and --direc‐
349 tories=skip had been specified before any explicit options.
350 Option specifications are separated by whitespace. A backslash
351 escapes the next character, so it can be used to specify an
352 option containing whitespace or a backslash.
353 GREP_COLOR
355 Specifies the marker for highlighting.
356 LC_ALL, LC_COLLATE, LANG
358 These variables specify the LC_COLLATE locale, which determines
359 the collating sequence used to interpret range expressions like
360 [a-z].
361 LC_ALL, LC_CTYPE, LANG
363 These variables specify the LC_CTYPE locale, which determines
364 the type of characters, e.g., which characters are whitespace.
365 LC_ALL, LC_MESSAGES, LANG
367 These variables specify the LC_MESSAGES locale, which determines
368 the language that grep uses for messages. The default C locale
369 uses American English messages.
370 POSIXLY_CORRECT
372 If set, grep behaves as POSIX.2 requires; otherwise, grep
373 behaves more like other GNU programs. POSIX.2 requires that
374 options that follow file names must be treated as file names; by
375 default, such options are permuted to the front of the operand
376 list and are treated as options. Also, POSIX.2 requires that
377 unrecognized options be diagnosed as “illegal”, but since they
378 are not really against the law the default is to diagnose them
379 as “invalid”. POSIXLY_CORRECT also disables _N_GNU_nonop‐
380 tion_argv_flags_, described below.
381 _N_GNU_nonoption_argv_flags_
383 (Here N is grep's numeric process ID.) If the ith character of
384 this environment variable's value is 1, do not consider the ith
385 operand of grep to be an option, even if it appears to be one.
386 A shell can put this variable in the environment for each com‐
387 mand it runs, specifying which operands are the results of file
388 name wildcard expansion and therefore should not be treated as
389 options. This behavior is available only with the GNU C
390 library, and only when POSIXLY_CORRECT is not set.
391
393 Normally, exit status is 0 if selected lines are found and 1 otherwise.
394 But the exit status is 2 if an error occurred, unless the -q or --quiet
395 or --silent option is used and a selected line is found.
396
398 Email bug reports to bug-grep@gnu.org.
399 Large repetition counts in the {n,m} construct may cause grep to use
401 lots of memory. In addition, certain other obscure regular expressions
402 require exponential time and space, and may cause grep to run out of
403 memory.
404 Backreferences are very slow, and may require exponential time.
406GNU Project 2002/01/22 GREP(1)