1 SED(1) User Commands SED(1)
2
3
4
5 NAME
6 sed - stream editor
7
8 SYNOPSIS
9 sed [-Ealnr] command [file ...]
10 sed [-Ealnr] [-e command] [-f command_file] [-I[extension] |
11 -i[extension]] [file ...]
12
13 DESCRIPTION
14 The sed utility reads the specified files, or the standard input if no
15 files are specified, modifying the input as specified by a list of
16 commands. The input is then written to the standard output.
17
18 A single command may be specified as the first argument to . Multiple
19 commands may be specified by using the -e or -f options. All commands
20 are applied to the input in the order they are specified regardless of
21 their origin.
22
23 The following options are available:
24
25 -E Interpret regular expressions as extended (modern) regular
26 expressions rather than basic regular expressions (BRE's). The
27 regex(5) manual page fully describes both formats.
28
29 -a The files listed as parameters for the ``w'' functions are
30 created (or truncated) before any processing begins, by default.
31 The -a option causes sed to delay opening each file until a
32 command containing the related ``w'' function is applied to a
33 line of input.
34
35 -e command
36 Append the editing commands specified by the command argument to
37 the list of commands.
38
39 -f command_file
40 Append the editing commands found in the file command_file to
41 the list of commands. The editing commands should each be
42 listed on a separate line.
43
44 -I[extension]
45 Edit files in-place, saving backups if extension was specified.
46 It is not recommended to omit saving backups when in-place
47 editing files, as you risk corruption or partial content in
48 situations where disk space is exhausted, etc.
49
50 Note that in-place editing with -I still takes place in a single
51 continuous line address space covering all files, although each
52 file preserves its individuality instead of forming one output
53 stream. The line counter is never reset between files, address
54 ranges can span file boundaries, and the ``$'' address matches
55 only the last line of the last file. (See Sed Addresses . )
56 That can lead to unexpected results in many cases of in-place
57 editing, where using -i is desired.
58
59 -i[extension]
60 Edit files in-place similarly to -I, but treat each file
61 independently from other files. In particular, line numbers in
62 each file start at 1, the ``$'' address matches the last line of
63 the current file, and address ranges are limited to the current
64 file. (See Sed Addresses . ) The net result is as though each
65 file were edited by a separate sed instance.
66
67 -l Make output line buffered.
68
69 -n By default, each line of input is echoed to the standard output
70 after all of the commands have been applied to it. The -n
71 option suppresses this behavior.
72
73 -r Same as -E for compatibility with GNU sed.
74
75 The form of a sed command is as follows:
76
77 [address[,address]]function[arguments]
78
79 Whitespace may be inserted before the first address and the
80 function portions of the command.
81
82 Normally, sed cyclically copies a line of input, not including
83 its terminating newline character, into a pattern space, (unless
84 there is something left after a ``D'' function), applies all of
85 the commands with addresses that select that pattern space,
86 copies the pattern space to the standard output, appending a
87 newline, and deletes the pattern space.
88
89 Some of the functions use a hold space to save all or part of
90 the pattern space for subsequent retrieval.
91
92 Sed Addresses
93 An address is not required, but if specified must have one of the
94 following formats:
95
96 o a number that counts input lines cumulatively across input files
97 (or in each file independently if a -i option is in effect);
98
99 o a dollar (``$'') character that addresses the last line of input
100 (or the last line of the current file if a -i option was
101 specified);
102
103 o a context address that consists of a regular expression preceded
104 and followed by a delimiter. The closing delimiter can also
105 optionally be followed by the ``I'' character, to indicate that
106 the regular expression is to be matched in a case-insensitive
107 way.
108
109 A command line with no addresses selects every pattern space.
110
111 A command line with one address selects all of the pattern
112 spaces that match the address.
113
114 A command line with two addresses selects an inclusive range.
115 This range starts with the first pattern space that matches the
116 first address. The end of the range is the next following
117 pattern space that matches the second address. If the second
118 address is a number less than or equal to the line number first
119 selected, only that line is selected. The number in the second
120 address may be prefixed with a (``+'') to specify the number of
121 lines to match after the first pattern. In the case when the
122 second address is a context address, sed does not re-match the
123 second address against the pattern space that matched the first
124 address. Starting at the first line following the selected
125 range, sed starts looking again for the first address.
126
127 Editing commands can be applied to non-selected pattern spaces
128 by use of the exclamation character (``!'') function.
129
130 Sed Regular Expressions
131 The regular expressions used in , by default, are basic regular
132 expressions (BREs, see regex(5) for more information), but extended
133 (modern) regular expressions can be used instead if the -E flag is
134 given. In addition, sed has the following two additions to regular
135 expressions:
136
137
138 1. In a context address, any character other than a backslash
139 (``\'') or newline character may be used to delimit the regular
140 expression. The opening delimiter needs to be preceded by a
141 backslash unless it is a slash. For example, the context
142 address \xabcx is equivalent to /abc/ . Also, putting a
143 backslash character before the delimiting character within the
144 regular expression causes the character to be treated literally.
145 For example, in the context address \xabc\xdefx , the RE
146 delimiter is an ``x'' and the second ``x'' stands for itself, so
147 that the regular expression is ``abcxdef''.
148
149
150 2. The escape sequence \n matches a newline character embedded in
151 the pattern space. You cannot, however, use a literal newline
152 character in an address or in the substitute command.
153
154 One special feature of sed regular expressions is that they can
155 default to the last regular expression used. If a regular
156 expression is empty, i.e., just the delimiter characters are
157 specified, the last regular expression encountered is used
158 instead. The last regular expression is defined as the last
159 regular expression used as part of an address or substitute
160 command, and at run-time, not compile-time. For example, the
161 command ``/abc/s//XXX/'' will substitute ``XXX'' for the pattern
162 ``abc''.
163
164 Sed Functions
165 In the following list of commands, the maximum number of permissible
166 addresses for each command is indicated by [0addr], [1addr], or
167 [2addr], representing zero, one, or two addresses.
168
169 The argument text consists of one or more lines. To embed a newline in
170 the text, precede it with a backslash. Other backslashes in text are
171 deleted and the following character taken literally.
172
173 The ``r'' and ``w'' functions take an optional file parameter, which
174 should be separated from the function letter by white space. Each file
175 given as an argument to sed is created (or its contents truncated)
176 before any input processing begins.
177
178 The ``b'', ``r'', ``s'', ``t'', ``w'', ``y'', ``!'', and ``:''
179 functions all accept additional arguments. The following synopses
180 indicate which arguments have to be separated from the function letters
181 by white space characters.
182
183 Two of the functions take a function-list. This is a list of sed
184 functions separated by newlines, as follows:
185
186 { function
187 function
188 ...
189 function
190 }
191
192 The ``{'' can be preceded by white space and can be followed by white
193 space. The function can be preceded by white space. The terminating
194 ``}'' must be preceded by a newline or optional white space.
195
196
197 [2addr] function-list
198 Execute function-list only when the pattern space is selected.
199
200
201 [1addr]a\
202
203 text Write text to standard output immediately before each attempt to
204 read a line of input, whether by executing the ``N'' function or
205 by beginning a new cycle.
206
207
208 [2addr]b[label]
209 Branch to the ``:'' function with the specified label. If the
210 label is not specified, branch to the end of the script.
211
212
213 [2addr]c\
214
215 text Delete the pattern space. With 0 or 1 address or at the end of
216 a 2-address range, text is written to the standard output.
217
218
219 [2addr]d
220 Delete the pattern space and start the next cycle.
221
222
223 [2addr]D
224 Delete the initial segment of the pattern space through the
225 first newline character and start the next cycle.
226
227
228 [2addr]g
229 Replace the contents of the pattern space with the contents of
230 the hold space.
231
232
233 [2addr]G
234 Append a newline character followed by the contents of the hold
235 space to the pattern space.
236
237
238 [2addr]h
239 Replace the contents of the hold space with the contents of the
240 pattern space.
241
242
243 [2addr]H
244 Append a newline character followed by the contents of the
245 pattern space to the hold space.
246
247
248 [1addr]i\
249
250 text Write text to the standard output.
251
252
253 [2addr]l
254 (The letter ell.) Write the pattern space to the standard
255 output in a visually unambiguous form. This form is as follows:
256
257
258 backslash
259 \\
260
261 alert \a
262
263 form-feed
264 \f
265
266 carriage-return
267 \r
268
269 tab \t
270
271 vertical tab
272 \v
273
274 Nonprintable characters are written as three-digit octal numbers
275 (with a preceding backslash) for each byte in the character
276 (most significant byte first). Long lines are folded, with the
277 point of folding indicated by displaying a backslash followed by
278 a newline. The end of each line is marked with a ``$''.
279
280
281 [2addr]n
282 Write the pattern space to the standard output if the default
283 output has not been suppressed, and replace the pattern space
284 with the next line of input.
285
286
287 [2addr]N
288 Append the next line of input to the pattern space, using an
289 embedded newline character to separate the appended material
290 from the original contents. Note that the current line number
291 changes.
292
293
294 [2addr]p
295 Write the pattern space to standard output.
296
297
298 [2addr]P
299 Write the pattern space, up to the first newline character to
300 the standard output.
301
302
303 [1addr]q
304 Branch to the end of the script and quit without starting a new
305 cycle.
306
307
308 [1addr]r file
309 Copy the contents of file to the standard output immediately
310 before the next attempt to read a line of input. If file cannot
311 be read for any reason, it is silently ignored and no error
312 condition is set.
313
314
315 [2addr]s/regular expression/replacement/flags
316 Substitute the replacement string for the first instance of the
317 regular expression in the pattern space. Any character other
318 than backslash or newline can be used instead of a slash to
319 delimit the RE and the replacement. Within the RE and the
320 replacement, the RE delimiter itself can be used as a literal
321 character if it is preceded by a backslash.
322
323 An ampersand (``&'') appearing in the replacement is replaced by
324 the string matching the RE. The special meaning of ``&'' in
325 this context can be suppressed by preceding it by a backslash.
326 The string ``\#'', where ``#'' is a digit, is replaced by the
327 text matched by the corresponding backreference expression (see
328 regex(5)) .
329
330 A line can be split by substituting a newline character into it.
331 To specify a newline character in the replacement string,
332 precede it with a backslash.
333
334 The value of flags in the substitute function is zero or more of
335 the following:
336
337 N Make the substitution only for the N'th occurrence of the
338 regular expression in the pattern space.
339
340 g Make the substitution for all non-overlapping matches of the
341 regular expression, not just the first one.
342
343 p Write the pattern space to standard output if a replacement was
344 made. If the replacement string is identical to that which it
345 replaces, it is still considered to have been a replacement.
346
347 w file Append the pattern space to file if a replacement was made. If
348 the replacement string is identical to that which it replaces,
349 it is still considered to have been a replacement.
350
351 I Match the regular expression in a case-insensitive way.
352
353
354 [2addr]t [label]
355 Branch to the ``:'' function bearing the label if any
356 substitutions have been made since the most recent reading of an
357 input line or execution of a ``t'' function. If no label is
358 specified, branch to the end of the script.
359
360
361 [2addr]w file
362 Append the pattern space to the file.
363
364
365 [2addr]x
366 Swap the contents of the pattern and hold spaces.
367
368
369 [2addr]y/string1/string2/
370 Replace all occurrences of characters in string1 in the pattern
371 space with the corresponding characters from string2. Any
372 character other than a backslash or newline can be used instead
373 of a slash to delimit the strings. Within string1 and string2,
374 a backslash followed by any character other than a newline is
375 that literal character, and a backslash followed by an ``n'' is
376 replaced by a newline character.
377
378
379 [2addr]!function
380
381 [2addr]!function-list
382 Apply the function or function-list only to the lines that are
383 not selected by the address(es).
384
385
386 [0addr]:label
387 This function does nothing; it bears a label to which the ``b''
388 and ``t'' commands may branch.
389
390
391 [1addr]=
392 Write the line number to the standard output followed by a
393 newline character.
394
395
396 [0addr]
397 Empty lines are ignored.
398
399
400 [0addr]#
401 The ``#'' and the remainder of the line are ignored (treated as
402 a comment), with the single exception that if the first two
403 characters in the file are ``#n'', the default output is
404 suppressed. This is the same as specifying the -n option on the
405 command line.
406
407 ENVIRONMENT
408 The COLUMNS,LANG,LC_ALL,LC_CTYPE and LC_COLLATE environment variables
409 affect the execution of sed as described in environ(5).
410
411 EXIT STATUS
412 The sed utility exits 0 on success, and >0 if an error occurs.
413
414 SEE ALSO
415 awk(1), ed(1), grep(1), regex(5)
416
417 STANDARDS
418 The sed utility is expected to be a superset of the IEEE Std 1003.2
419 (``POSIX.2'') specification.
420
421 The -E, I , a and -i options, the prefixing ``+'' in the second member
422 of an address range, as well as the ``I'' flag to the address regular
423 expression and substitution command are non-standard extensions and may
424 not be available on other operating systems.
425
426 HISTORY
427 A sed command, written by L. E. McMahon, appeared in Version 7 AT&T
428 UNIX.
429
430 AUTHORS
431 "Diomidis D. Spinellis" <dds@FreeBSD.org>
432
433 BUGS
434 Multibyte characters containing a byte with value 0x5C (ASCII `\') may
435 be incorrectly treated as line continuation characters in arguments to
436 the ``a'', ``c'' and ``i'' commands. Multibyte characters cannot be
437 used as delimiters with the ``s'' and ``y'' commands.
438
439
440
441 February 14, 2015 SED(1)