1.\" Copyright (c) 1992, 1993 2.\" The Regents of the University of California. All rights reserved. 3.\" 4.\" This code is derived from software contributed to Berkeley by 5.\" the Institute of Electrical and Electronics Engineers, Inc. 6.\" 7.\" Redistribution and use in source and binary forms, with or without 8.\" modification, are permitted provided that the following conditions 9.\" are met: 10.\" 1. Redistributions of source code must retain the above copyright 11.\" notice, this list of conditions and the following disclaimer. 12.\" 2. Redistributions in binary form must reproduce the above copyright 13.\" notice, this list of conditions and the following disclaimer in the 14.\" documentation and/or other materials provided with the distribution. 15.\" 4. Neither the name of the University nor the names of its contributors 16.\" may be used to endorse or promote products derived from this software 17.\" without specific prior written permission. 18.\" 19.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 20.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 21.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 22.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 23.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 24.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 25.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 26.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 27.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 28.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 29.\" SUCH DAMAGE. 30.\" 31.\" @(#)sed.1 8.2 (Berkeley) 12/30/93 32.\" $FreeBSD$ 33.\" 34.Dd May 24, 2009 35.Dt SED 1 36.Os 37.Sh NAME 38.Nm sed 39.Nd stream editor 40.Sh SYNOPSIS 41.Nm 42.Op Fl Ealnr 43.Ar command 44.Op Ar 45.Nm 46.Op Fl Ealnr 47.Op Fl e Ar command 48.Op Fl f Ar command_file 49.Op Fl I Ar extension 50.Op Fl i Ar extension 51.Op Ar 52.Sh DESCRIPTION 53The 54.Nm 55utility reads the specified files, or the standard input if no files 56are specified, modifying the input as specified by a list of commands. 57The input is then written to the standard output. 58.Pp 59A single command may be specified as the first argument to 60.Nm . 61Multiple commands may be specified by using the 62.Fl e 63or 64.Fl f 65options. 66All commands are applied to the input in the order they are specified 67regardless of their origin. 68.Pp 69The following options are available: 70.Bl -tag -width indent 71.It Fl E 72Interpret regular expressions as extended (modern) regular expressions 73rather than basic regular expressions (BRE's). 74The 75.Xr re_format 7 76manual page fully describes both formats. 77.It Fl a 78The files listed as parameters for the 79.Dq w 80functions are created (or truncated) before any processing begins, 81by default. 82The 83.Fl a 84option causes 85.Nm 86to delay opening each file until a command containing the related 87.Dq w 88function is applied to a line of input. 89.It Fl e Ar command 90Append the editing commands specified by the 91.Ar command 92argument 93to the list of commands. 94.It Fl f Ar command_file 95Append the editing commands found in the file 96.Ar command_file 97to the list of commands. 98The editing commands should each be listed on a separate line. 99.It Fl I Ar extension 100Edit files in-place, saving backups with the specified 101.Ar extension . 102If a zero-length 103.Ar extension 104is given, no backup will be saved. 105It is not recommended to give a zero-length 106.Ar extension 107when in-place editing files, as you risk corruption or partial content 108in situations where disk space is exhausted, etc. 109.Pp 110Note that in-place editing with 111.Fl I 112still takes place in a single continuous line address space covering 113all files, although each file preserves its individuality instead of 114forming one output stream. 115The line counter is never reset between files, address ranges can span 116file boundaries, and the 117.Dq $ 118address matches only the last line of the last file. 119(See 120.Sx "Sed Addresses" . ) 121That can lead to unexpected results in many cases of in-place editing, 122where using 123.Fl i 124is desired. 125.It Fl i Ar extension 126Edit files in-place similarly to 127.Fl I , 128but treat each file independently from other files. 129In particular, line numbers in each file start at 1, 130the 131.Dq $ 132address matches the last line of the current file, 133and address ranges are limited to the current file. 134(See 135.Sx "Sed Addresses" . ) 136The net result is as though each file were edited by a separate 137.Nm 138instance. 139.It Fl l 140Make output line buffered. 141.It Fl n 142By default, each line of input is echoed to the standard output after 143all of the commands have been applied to it. 144The 145.Fl n 146option suppresses this behavior. 147.It Fl r 148Same as 149.Fl E 150for compatibility with GNU sed. 151.El 152.Pp 153The form of a 154.Nm 155command is as follows: 156.Pp 157.Dl [address[,address]]function[arguments] 158.Pp 159Whitespace may be inserted before the first address and the function 160portions of the command. 161.Pp 162Normally, 163.Nm 164cyclically copies a line of input, not including its terminating newline 165character, into a 166.Em "pattern space" , 167(unless there is something left after a 168.Dq D 169function), 170applies all of the commands with addresses that select that pattern space, 171copies the pattern space to the standard output, appending a newline, and 172deletes the pattern space. 173.Pp 174Some of the functions use a 175.Em "hold space" 176to save all or part of the pattern space for subsequent retrieval. 177.Sh "Sed Addresses" 178An address is not required, but if specified must have one of the 179following formats: 180.Bl -bullet -offset indent 181.It 182a number that counts 183input lines 184cumulatively across input files (or in each file independently 185if a 186.Fl i 187option is in effect); 188.It 189a dollar 190.Pq Dq $ 191character that addresses the last line of input (or the last line 192of the current file if a 193.Fl i 194option was specified); 195.It 196a context address 197that consists of a regular expression preceded and followed by a 198delimiter. The closing delimiter can also optionally be followed by the 199.Dq I 200character, to indicate that the regular expression is to be matched 201in a case-insensitive way. 202.El 203.Pp 204A command line with no addresses selects every pattern space. 205.Pp 206A command line with one address selects all of the pattern spaces 207that match the address. 208.Pp 209A command line with two addresses selects an inclusive range. 210This 211range starts with the first pattern space that matches the first 212address. 213The end of the range is the next following pattern space 214that matches the second address. 215If the second address is a number 216less than or equal to the line number first selected, only that 217line is selected. 218The number in the second address may be prefixed with a 219.Pq Dq \&+ 220to specify the number of lines to match after the first pattern. 221In the case when the second address is a context 222address, 223.Nm 224does not re-match the second address against the 225pattern space that matched the first address. 226Starting at the 227first line following the selected range, 228.Nm 229starts looking again for the first address. 230.Pp 231Editing commands can be applied to non-selected pattern spaces by use 232of the exclamation character 233.Pq Dq \&! 234function. 235.Sh "Sed Regular Expressions" 236The regular expressions used in 237.Nm , 238by default, are basic regular expressions (BREs, see 239.Xr re_format 7 240for more information), but extended (modern) regular expressions can be used 241instead if the 242.Fl E 243flag is given. 244In addition, 245.Nm 246has the following two additions to regular expressions: 247.Pp 248.Bl -enum -compact 249.It 250In a context address, any character other than a backslash 251.Pq Dq \e 252or newline character may be used to delimit the regular expression. 253The opening delimiter needs to be preceded by a backslash 254unless it is a slash. 255For example, the context address 256.Li \exabcx 257is equivalent to 258.Li /abc/ . 259Also, putting a backslash character before the delimiting character 260within the regular expression causes the character to be treated literally. 261For example, in the context address 262.Li \exabc\exdefx , 263the RE delimiter is an 264.Dq x 265and the second 266.Dq x 267stands for itself, so that the regular expression is 268.Dq abcxdef . 269.Pp 270.It 271The escape sequence \en matches a newline character embedded in the 272pattern space. 273You cannot, however, use a literal newline character in an address or 274in the substitute command. 275.El 276.Pp 277One special feature of 278.Nm 279regular expressions is that they can default to the last regular 280expression used. 281If a regular expression is empty, i.e., just the delimiter characters 282are specified, the last regular expression encountered is used instead. 283The last regular expression is defined as the last regular expression 284used as part of an address or substitute command, and at run-time, not 285compile-time. 286For example, the command 287.Dq /abc/s//XXX/ 288will substitute 289.Dq XXX 290for the pattern 291.Dq abc . 292.Sh "Sed Functions" 293In the following list of commands, the maximum number of permissible 294addresses for each command is indicated by [0addr], [1addr], or [2addr], 295representing zero, one, or two addresses. 296.Pp 297The argument 298.Em text 299consists of one or more lines. 300To embed a newline in the text, precede it with a backslash. 301Other backslashes in text are deleted and the following character 302taken literally. 303.Pp 304The 305.Dq r 306and 307.Dq w 308functions take an optional file parameter, which should be separated 309from the function letter by white space. 310Each file given as an argument to 311.Nm 312is created (or its contents truncated) before any input processing begins. 313.Pp 314The 315.Dq b , 316.Dq r , 317.Dq s , 318.Dq t , 319.Dq w , 320.Dq y , 321.Dq \&! , 322and 323.Dq \&: 324functions all accept additional arguments. 325The following synopses indicate which arguments have to be separated from 326the function letters by white space characters. 327.Pp 328Two of the functions take a function-list. 329This is a list of 330.Nm 331functions separated by newlines, as follows: 332.Bd -literal -offset indent 333{ function 334 function 335 ... 336 function 337} 338.Ed 339.Pp 340The 341.Dq { 342can be preceded by white space and can be followed by white space. 343The function can be preceded by white space. 344The terminating 345.Dq } 346must be preceded by a newline, and may also be preceded by white space. 347.Pp 348.Bl -tag -width "XXXXXX" -compact 349.It [2addr] function-list 350Execute function-list only when the pattern space is selected. 351.Pp 352.It [1addr]a\e 353.It text 354Write 355.Em text 356to standard output immediately before each attempt to read a line of input, 357whether by executing the 358.Dq N 359function or by beginning a new cycle. 360.Pp 361.It [2addr]b[label] 362Branch to the 363.Dq \&: 364function with the specified label. 365If the label is not specified, branch to the end of the script. 366.Pp 367.It [2addr]c\e 368.It text 369Delete the pattern space. 370With 0 or 1 address or at the end of a 2-address range, 371.Em text 372is written to the standard output. 373.Pp 374.It [2addr]d 375Delete the pattern space and start the next cycle. 376.Pp 377.It [2addr]D 378Delete the initial segment of the pattern space through the first 379newline character and start the next cycle. 380.Pp 381.It [2addr]g 382Replace the contents of the pattern space with the contents of the 383hold space. 384.Pp 385.It [2addr]G 386Append a newline character followed by the contents of the hold space 387to the pattern space. 388.Pp 389.It [2addr]h 390Replace the contents of the hold space with the contents of the 391pattern space. 392.Pp 393.It [2addr]H 394Append a newline character followed by the contents of the pattern space 395to the hold space. 396.Pp 397.It [1addr]i\e 398.It text 399Write 400.Em text 401to the standard output. 402.Pp 403.It [2addr]l 404(The letter ell.) 405Write the pattern space to the standard output in a visually unambiguous 406form. 407This form is as follows: 408.Pp 409.Bl -tag -width "carriage-returnXX" -offset indent -compact 410.It backslash 411\e\e 412.It alert 413\ea 414.It form-feed 415\ef 416.It carriage-return 417\er 418.It tab 419\et 420.It vertical tab 421\ev 422.El 423.Pp 424Nonprintable characters are written as three-digit octal numbers (with a 425preceding backslash) for each byte in the character (most significant byte 426first). 427Long lines are folded, with the point of folding indicated by displaying 428a backslash followed by a newline. 429The end of each line is marked with a 430.Dq $ . 431.Pp 432.It [2addr]n 433Write the pattern space to the standard output if the default output has 434not been suppressed, and replace the pattern space with the next line of 435input. 436.Pp 437.It [2addr]N 438Append the next line of input to the pattern space, using an embedded 439newline character to separate the appended material from the original 440contents. 441Note that the current line number changes. 442.Pp 443.It [2addr]p 444Write the pattern space to standard output. 445.Pp 446.It [2addr]P 447Write the pattern space, up to the first newline character to the 448standard output. 449.Pp 450.It [1addr]q 451Branch to the end of the script and quit without starting a new cycle. 452.Pp 453.It [1addr]r file 454Copy the contents of 455.Em file 456to the standard output immediately before the next attempt to read a 457line of input. 458If 459.Em file 460cannot be read for any reason, it is silently ignored and no error 461condition is set. 462.Pp 463.It [2addr]s/regular expression/replacement/flags 464Substitute the replacement string for the first instance of the regular 465expression in the pattern space. 466Any character other than backslash or newline can be used instead of 467a slash to delimit the RE and the replacement. 468Within the RE and the replacement, the RE delimiter itself can be used as 469a literal character if it is preceded by a backslash. 470.Pp 471An ampersand 472.Pq Dq & 473appearing in the replacement is replaced by the string matching the RE. 474The special meaning of 475.Dq & 476in this context can be suppressed by preceding it by a backslash. 477The string 478.Dq \e# , 479where 480.Dq # 481is a digit, is replaced by the text matched 482by the corresponding backreference expression (see 483.Xr re_format 7 ) . 484.Pp 485A line can be split by substituting a newline character into it. 486To specify a newline character in the replacement string, precede it with 487a backslash. 488.Pp 489The value of 490.Em flags 491in the substitute function is zero or more of the following: 492.Bl -tag -width "XXXXXX" -offset indent 493.It Ar N 494Make the substitution only for the 495.Ar N Ns 'th 496occurrence of the regular expression in the pattern space. 497.It g 498Make the substitution for all non-overlapping matches of the 499regular expression, not just the first one. 500.It p 501Write the pattern space to standard output if a replacement was made. 502If the replacement string is identical to that which it replaces, it 503is still considered to have been a replacement. 504.It w Em file 505Append the pattern space to 506.Em file 507if a replacement was made. 508If the replacement string is identical to that which it replaces, it 509is still considered to have been a replacement. 510.It I 511Match the regular expression in a case-insensitive way. 512.El 513.Pp 514.It [2addr]t [label] 515Branch to the 516.Dq \&: 517function bearing the label if any substitutions have been made since the 518most recent reading of an input line or execution of a 519.Dq t 520function. 521If no label is specified, branch to the end of the script. 522.Pp 523.It [2addr]w Em file 524Append the pattern space to the 525.Em file . 526.Pp 527.It [2addr]x 528Swap the contents of the pattern and hold spaces. 529.Pp 530.It [2addr]y/string1/string2/ 531Replace all occurrences of characters in 532.Em string1 533in the pattern space with the corresponding characters from 534.Em string2 . 535Any character other than a backslash or newline can be used instead of 536a slash to delimit the strings. 537Within 538.Em string1 539and 540.Em string2 , 541a backslash followed by any character other than a newline is that literal 542character, and a backslash followed by an ``n'' is replaced by a newline 543character. 544.Pp 545.It [2addr]!function 546.It [2addr]!function-list 547Apply the function or function-list only to the lines that are 548.Em not 549selected by the address(es). 550.Pp 551.It [0addr]:label 552This function does nothing; it bears a label to which the 553.Dq b 554and 555.Dq t 556commands may branch. 557.Pp 558.It [1addr]= 559Write the line number to the standard output followed by a newline 560character. 561.Pp 562.It [0addr] 563Empty lines are ignored. 564.Pp 565.It [0addr]# 566The 567.Dq # 568and the remainder of the line are ignored (treated as a comment), with 569the single exception that if the first two characters in the file are 570.Dq #n , 571the default output is suppressed. 572This is the same as specifying the 573.Fl n 574option on the command line. 575.El 576.Sh ENVIRONMENT 577The 578.Ev COLUMNS , LANG , LC_ALL , LC_CTYPE 579and 580.Ev LC_COLLATE 581environment variables affect the execution of 582.Nm 583as described in 584.Xr environ 7 . 585.Sh EXIT STATUS 586.Ex -std 587.Sh SEE ALSO 588.Xr awk 1 , 589.Xr ed 1 , 590.Xr grep 1 , 591.Xr regex 3 , 592.Xr re_format 7 593.Sh STANDARDS 594The 595.Nm 596utility is expected to be a superset of the 597.St -p1003.2 598specification. 599.Pp 600The 601.Fl E , I , a 602and 603.Fl i 604options, the prefixing 605.Dq \&+ 606in the second member of an address range, 607as well as the 608.Dq I 609flag to the address regular expression and substitution command are 610non-standard 611.Fx 612extensions and may not be available on other operating systems. 613.Sh HISTORY 614A 615.Nm 616command, written by 617.An L. E. McMahon , 618appeared in 619.At v7 . 620.Sh AUTHORS 621.An "Diomidis D. Spinellis" Aq dds@FreeBSD.org 622.Sh BUGS 623Multibyte characters containing a byte with value 0x5C 624.Tn ( ASCII 625.Ql \e ) 626may be incorrectly treated as line continuation characters in arguments to the 627.Dq a , 628.Dq c 629and 630.Dq i 631commands. 632Multibyte characters cannot be used as delimiters with the 633.Dq s 634and 635.Dq y 636commands. 637