Debian Codegraph @

upstream/debian-unstable/man1/sed.1

raw
1 .\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.48.5.
2 .TH SED "1" "April 2026" "GNU sed 4.9" "User Commands"
3 .SH NAME
4 sed \- stream editor for filtering and transforming text
5 .SH SYNOPSIS
6 .nf
7 sed [-V] [--version] [--help] [-n] [--quiet] [--silent]
8 [-l N] [--line-length=N] [-u] [--unbuffered]
9 [-E] [-r] [--regexp-extended]
10 [-e script] [--expression=script]
11 [-f script-file] [--file=script-file]
12 [script-if-no-other-script]
13 [file...]
14 .fi
15 .SH DESCRIPTION
16 .ds sd \fIsed\fP
17 .ds Sd \fISed\fP
18 \*(Sd is a stream editor.
19 A stream editor is used to perform basic text
20 transformations on an input stream
21 (a file or input from a pipeline).
22 While in some ways similar to an editor which
23 permits scripted edits (such as \fIed\fP),
24 \*(sd works by making only one pass over the
25 input(s), and is consequently more efficient.
26 But it is \*(sd's ability to filter text in a pipeline
27 which particularly distinguishes it from other types of
28 editors.
29 .HP
30 \fB\-n\fR, \fB\-\-quiet\fR, \fB\-\-silent\fR
31 .IP
32 suppress automatic printing of pattern space
33 .HP
34 \fB\-\-debug\fR
35 .IP
36 annotate program execution
37 .HP
38 \fB\-e\fR script, \fB\-\-expression\fR=\fI\,script\/\fR
39 .IP
40 add the script to the commands to be executed
41 .HP
42 \fB\-f\fR script\-file, \fB\-\-file\fR=\fI\,script\-file\/\fR
43 .IP
44 add the contents of script\-file to the commands to be executed
45 .HP
46 \fB\-\-follow\-symlinks\fR
47 .IP
48 follow symlinks when processing in place
49 .HP
50 \fB\-i[SUFFIX]\fR, \fB\-\-in\-place\fR[=\fI\,SUFFIX\/\fR]
51 .IP
52 edit files in place (makes backup if SUFFIX supplied)
53 .HP
54 \fB\-l\fR N, \fB\-\-line\-length\fR=\fI\,N\/\fR
55 .IP
56 specify the desired line\-wrap length for the `l' command
57 .HP
58 \fB\-\-posix\fR
59 .IP
60 disable all GNU extensions.
61 .HP
62 \fB\-E\fR, \fB\-r\fR, \fB\-\-regexp\-extended\fR
63 .IP
64 use extended regular expressions in the script
65 (for portability use POSIX \fB\-E\fR).
66 .HP
67 \fB\-s\fR, \fB\-\-separate\fR
68 .IP
69 consider files as separate rather than as a single,
70 continuous long stream.
71 .HP
72 \fB\-\-sandbox\fR
73 .IP
74 operate in sandbox mode (disable e/r/w commands).
75 .HP
76 \fB\-u\fR, \fB\-\-unbuffered\fR
77 .IP
78 load minimal amounts of data from the input files and flush
79 the output buffers more often
80 .HP
81 \fB\-z\fR, \fB\-\-null\-data\fR
82 .IP
83 separate lines by NUL characters
84 .TP
85 \fB\-\-help\fR
86 display this help and exit
87 .TP
88 \fB\-\-version\fR
89 output version information and exit
90 .PP
91 If no \fB\-e\fR, \fB\-\-expression\fR, \fB\-f\fR, or \fB\-\-file\fR option is given, then the first
92 non\-option argument is taken as the sed script to interpret. All
93 remaining arguments are names of input files; if no input files are
94 specified, then the standard input is read.
95 .PP
96 GNU sed home page: <https://www.gnu.org/software/sed/>.
97 General help using GNU software: <https://www.gnu.org/gethelp/>.
98 E\-mail bug reports to: <bug\[email protected]>.
99 .PP
100 Packaged by Debian
101 Copyright \(co 2022 Free Software Foundation, Inc.
102 License GPLv3+: GNU GPL version 3 or later <https://gnu.org/licenses/gpl.html>.
103 This is free software: you are free to change and redistribute it.
104 There is NO WARRANTY, to the extent permitted by law.
105 .SH "COMMAND SYNOPSIS"
106 This is just a brief synopsis of \*(sd commands to serve as
107 a reminder to those who already know \*(sd;
108 other documentation (such as the texinfo document)
109 must be consulted for fuller descriptions.
110 .SS
111 Zero-address ``commands''
112 .TP
113 .RI :\ label
114 Label for
115 .B b
116 and
117 .B t
118 commands.
119 .TP
120 .RI # comment
121 The comment extends until the next newline (or the end of a
122 .B \-e
123 script fragment).
124 .TP
125 }
126 The closing bracket of a { } block.
127 .SS
128 Zero- or One- address commands
129 .TP
130 =
131 Print the current line number.
132 .TP
133 a \e
134 .TP
135 .I text
136 Append
137 .IR text ,
138 which has each embedded newline preceded by a backslash.
139 .TP
140 i \e
141 .TP
142 .I text
143 Insert
144 .IR text ,
145 which has each embedded newline preceded by a backslash.
146 .TP
147 q [\fIexit-code\fR]
148 Immediately quit the \*(sd script without processing
149 any more input, except that if auto-print is not disabled
150 the current pattern space will be printed. The exit code
151 argument is a GNU extension.
152 .TP
153 Q [\fIexit-code\fR]
154 Immediately quit the \*(sd script without processing
155 any more input. This is a GNU extension.
156 .TP
157 .RI r\ filename
158 Append text read from
159 .IR filename .
160 .TP
161 .RI R\ filename
162 Append a line read from
163 .IR filename .
164 Each invocation of the command reads a line from the file.
165 This is a GNU extension.
166 .SS
167 Commands which accept address ranges
168 .TP
169 {
170 Begin a block of commands (end with a }).
171 .TP
172 .RI b\ label
173 Branch to
174 .IR label ;
175 if
176 .I label
177 is omitted, branch to end of script.
178 .TP
179 c \e
180 .TP
181 .I text
182 Replace the selected lines with
183 .IR text ,
184 which has each embedded newline preceded by a backslash.
185 .TP
186 d
187 Delete pattern space.
188 Start next cycle.
189 .TP
190 D
191 If pattern space contains no newline, start a normal new cycle as if
192 the d command was issued. Otherwise, delete text in the pattern
193 space up to the first newline, and restart cycle with the resultant
194 pattern space, without reading a new line of input.
195 .TP
196 h H
197 Copy/append pattern space to hold space.
198 .TP
199 g G
200 Copy/append hold space to pattern space.
201 .TP
202 l
203 List out the current line in a ``visually unambiguous'' form.
204 .TP
205 .RI l\ width
206 List out the current line in a ``visually unambiguous'' form,
207 breaking it at
208 .I width
209 characters. This is a GNU extension.
210 .TP
211 n N
212 Read/append the next line of input into the pattern space.
213 .TP
214 p
215 Print the current pattern space.
216 .TP
217 P
218 Print up to the first embedded newline of the current pattern space.
219 .TP
220 .RI s/ regexp / replacement /
221 Attempt to match
222 .I regexp
223 against the pattern space.
224 If successful, replace that portion matched
225 with
226 .IR replacement .
227 The
228 .I replacement
229 may contain the special character
230 .B &
231 to refer to that portion of the pattern space which matched,
232 and the special escapes \e1 through \e9 to refer to the
233 corresponding matching sub-expressions in the
234 .IR regexp .
235 .TP
236 .RI t\ label
237 If a s/// has done a successful substitution since the
238 last input line was read and since the last t or T
239 command, then branch to
240 .IR label ;
241 if
242 .I label
243 is omitted, branch to end of script.
244 .TP
245 .RI T\ label
246 If no s/// has done a successful substitution since the
247 last input line was read and since the last t or T
248 command, then branch to
249 .IR label ;
250 if
251 .I label
252 is omitted, branch to end of script. This is a GNU
253 extension.
254 .TP
255 .RI w\ filename
256 Write the current pattern space to
257 .IR filename .
258 .TP
259 .RI W\ filename
260 Write the first line of the current pattern space to
261 .IR filename .
262 This is a GNU extension.
263 .TP
264 x
265 Exchange the contents of the hold and pattern spaces.
266 .TP
267 .RI y/ source / dest /
268 Transliterate the characters in the pattern space which appear in
269 .I source
270 to the corresponding character in
271 .IR dest .
272 .SH
273 Addresses
274 \*(Sd commands can be given with no addresses, in which
275 case the command will be executed for all input lines;
276 with one address, in which case the command will only be executed
277 for input lines which match that address; or with two
278 addresses, in which case the command will be executed
279 for all input lines which match the inclusive range of
280 lines starting from the first address and continuing to
281 the second address.
282 Three things to note about address ranges:
283 the syntax is
284 .IR addr1 , addr2
285 (i.e., the addresses are separated by a comma);
286 the line which
287 .I addr1
288 matched will always be accepted,
289 even if
290 .I addr2
291 selects an earlier line;
292 and if
293 .I addr2
294 is a
295 .IR regexp ,
296 it will not be tested against the line that
297 .I addr1
298 matched.
299 .PP
300 After the address (or address-range),
301 and before the command, a
302 .B !
303 may be inserted,
304 which specifies that the command shall only be
305 executed if the address (or address-range) does
306 .B not
307 match.
308 .PP
309 The following address types are supported:
310 .TP
311 .I number
312 Match only the specified line
313 .IR number
314 (which increments cumulatively across files, unless the
315 .B \-s
316 option is specified on the command line).
317 .TP
318 .IR first ~ step
319 Match every
320 .IR step 'th
321 line starting with line
322 .IR first .
323 For example, ``sed \-n 1~2p'' will print all the odd-numbered lines in
324 the input stream, and the address 2~5 will match every fifth line,
325 starting with the second.
326 .I first
327 can be zero; in this case, \*(sd operates as if it were equal to
328 .IR step .
329 (This is an extension.)
330 .TP
331 $
332 Match the last line.
333 .TP
334 .RI / regexp /
335 Match lines matching the regular expression
336 .IR regexp .
337 Matching is performed on the current pattern space, which
338 can be modified with commands such as ``s///''.
339 .TP
340 .BI \fR\e\fPc regexp c
341 Match lines matching the regular expression
342 .IR regexp .
343 The
344 .B c
345 may be any character.
346 .PP
347 GNU \*(sd also supports some special 2-address forms:
348 .TP
349 .RI 0, addr2
350 Start out in "matched first address" state, until
351 .I addr2
352 is found.
353 This is similar to
354 .RI 1, addr2 ,
355 except that if
356 .I addr2
357 matches the very first line of input the
358 .RI 0, addr2
359 form will be at the end of its range, whereas the
360 .RI 1, addr2
361 form will still be at the beginning of its range.
362 This works only when
363 .I addr2
364 is a regular expression.
365 .TP
366 .IR addr1 ,+ N
367 Will match
368 .I addr1
369 and the
370 .I N
371 lines following
372 .IR addr1 .
373 .TP
374 .IR addr1 ,~ N
375 Will match
376 .I addr1
377 and the lines following
378 .I addr1
379 until the next line whose input line number is a multiple of
380 .IR N .
381 .SH "REGULAR EXPRESSIONS"
382 POSIX.2 BREs
383 .I should
384 be supported, but they aren't completely because of performance
385 problems.
386 The
387 .B \en
388 sequence in a regular expression matches the newline character,
389 and similarly for
390 .BR \ea ,
391 .BR \et ,
392 and other sequences.
393 The \fI-E\fP option switches to using extended regular expressions instead;
394 it has been supported for years by GNU sed, and is now
395 included in POSIX.
396 .SH BUGS
397 .PP
398 E-mail bug reports to
399 .BR [email protected] .
400 Also, please include the output of ``sed \-\-version'' in the body
401 of your report if at all possible.
402 .SH AUTHOR
403 Written by Jay Fenlason, Tom Lord, Ken Pizzini,
404 Paolo Bonzini, Jim Meyering, and Assaf Gordon.
405 .PP
406 This sed program was built with SELinux support.
407 SELinux is disabled on this system.
408 .PP
409 GNU sed home page: <https://www.gnu.org/software/sed/>.
410 General help using GNU software: <https://www.gnu.org/gethelp/>.
411 E\-mail bug reports to: <bug\[email protected]>.
412 .SH "SEE ALSO"
413 .BR awk (1),
414 .BR ed (1),
415 .BR grep (1),
416 .BR tr (1),
417 .BR perlre (1),
418 sed.info,
419 any of various books on \*(sd,
420 .na
421 the \*(sd FAQ (http://sed.sf.net/grabbag/tutorials/sedfaq.txt),
422 http://sed.sf.net/grabbag/.
423 .PP
424 The full documentation for
425 .B sed
426 is maintained as a Texinfo manual. If the
427 .B info
428 and
429 .B sed
430 programs are properly installed at your site, the command
431 .IP
432 .B info sed
433 .PP
434 should give you access to the complete manual.