Manual / sed
sed(1)
| SED(1) | General Commands Manual | SED(1) |
NAME
sed - stream editor
SYNOPSIS
sed [ -n ] script [ file ] ...
sed [ -n ] [ -e script ] ... [ -f sfile ] ... [
file ] ...
DESCRIPTION
Sed copies the named files (standard input default) to the standard output, edited according to a script of commands. Each -e adds its argument to the script and each -f adds the contents of sfile; the pieces are joined with newlines, in order. With neither, the first operand is the script. The -n option suppresses the default output, as does a first script line that is exactly `#n'.
A script consists of editing commands, one per line or separated by `;', of the form
- [address [, address] ] [!] function [arguments]
Blanks may precede the function, and a `#' begins a comment that runs to the end of the line.
In normal operation sed cyclically copies a line of input into a pattern space (unless there is something left after a `D' command), applies in sequence all commands whose addresses select that pattern space, and at the end of the script copies the pattern space to the standard output (except under -n) and deletes it. A second buffer, the hold space, is empty at the start and is touched only by the `g', `G', `h', `H' and `x' commands. Each space holds up to 2047 characters; a space grown past that by `s', `N', `G' or `H' is a fatal error, not a truncation. An input line longer than 255 characters is truncated there.
An address is either a decimal number that counts input lines cumulatively across files, a `$' that addresses the last line of input, or a context address `/regular expression/' - or `\cREc' with any other single character c as the delimiter - in the style of ed(1), where the escape `\n' matches the newline embedded in the pattern space by `N' or `G' and an empty regular expression `//' means the last one used.
A command with no addresses selects every pattern space. A command with one address selects each pattern space that matches the address. A command with two addresses selects the inclusive range from the first pattern space that matches the first address through the next pattern space that matches the second. (If the second address is a number less than or equal to the line number first selected, only one line is selected.) Thereafter the process is repeated, looking again for the first address. A `!' after the addresses applies the function to the pattern spaces not selected.
In the following list the maximum number of addresses each function takes is in parentheses. An argument denoted text consists of one or more lines, all but the last of which end with `\' to hide the newline; any other `\' in text is removed and the next character taken literally. The V7 form `a\' followed by a newline and the text is the portable one; `a text' on one line is also accepted (leading blanks dropped; `a\text' keeps them). An argument denoted rfile or wfile must terminate the command line. Each wfile is created before processing begins; `/dev/stdout' names the standard output. There can be at most 4 distinct wfile arguments, and at most 48 commands and 16 labels in a script.
- (1)a\
-
- text
- Append. Place text on the output before reading the next input line.
- (2)b label
- Branch to the `:' command bearing the label. If label is empty, branch to the end of the script. A label ends at the end of the line or at a `;'.
- (2)c\
-
- text
- Change. Delete the pattern space. With 0 or 1 address or at the end of a 2-address range, place text on the output. Start the next cycle.
- (2)d
- Delete the pattern space. Start the next cycle.
- (2)D
- Delete the initial segment of the pattern space through the first newline. Start the next cycle without reading new input.
- (2)g
- Replace the contents of the pattern space by the contents of the hold space.
- (2)G
- Append the contents of the hold space to the pattern space.
- (2)h
- Replace the contents of the hold space by the contents of the pattern space.
- (2)H
- Append the contents of the pattern space to the hold space.
- (1)i\
-
- text
- Insert. Place text on the standard output.
- (2)l
- List the pattern space on the standard output in an unambiguous form: `\\', `\a', `\b', `\f', `\n', `\r', `\t' and `\v' as those escapes, other non-printing characters as three-digit octal `\ddd', a `$' at the end, and long lines folded at 69 characters with a `\'.
- (2)n
- Copy the pattern space to the standard output (unless under -n). Replace the pattern space with the next line of input. If there is no next line, quit without running the rest of the script.
- (2)N
- Append the next line of input to the pattern space with an embedded newline. (The current line number changes.) If there is no next line, quit without copying the pattern space to the output; `$!N' is the portable spelling.
- (2)p
- Print. Copy the pattern space to the standard output.
- (2)P
- Copy the initial segment of the pattern space through the first newline to the standard output.
- (1)q [exit-code]
- Quit. Branch to the end of the script. Do not start a new cycle. The exit status is exit-code, or 0.
- (2)r rfile
- Read the contents of rfile. Place them on the output before reading the next input line. A file that cannot be read is silently nothing.
- (2)s/regular expression/replacement/flags
- Substitute the replacement string for instances of the regular expression in the pattern space. Any character may be used instead of `/'. For a fuller description see ed(1) and REGULAR EXPRESSIONS below. Flags is zero or more of
- g
- Global. Substitute for all nonoverlapping instances of the regular expression rather than just the first one.
- p
- Print the pattern space if a replacement was made.
- n
- n = 1 - 512. Substitute for just the nth occurrence of the regular expression; with g as well, the nth and every later one.
- w wfile
- Write. Append the pattern space to wfile if a replacement was made.
- In the replacement an `&' stands for the matched text, `\1' to `\9' for the text matched by the corresponding `\(' `\)' subexpression, `\n' for a newline, and a `\' before any other character (a newline included) for that character itself.
- (2)t label
- Test. Branch to the `:' command bearing the label if any substitutions have been made since the most recent reading of an input line or execution of a `t'. If label is empty, branch to the end of the script.
- (2)w wfile
- Write. Append the pattern space to wfile.
- (2)x
- Exchange the contents of the pattern and hold spaces.
- (2)y/string1/string2/
- Transform. Replace all occurrences of characters in string1 with the corresponding character in string2. The lengths of string1 and string2 must be equal. `\n' is a newline, `\\' a backslash.
- (2)! function
- Don't. Apply the function (or group, if function is `{') only to lines not selected by the address(es).
- (0): label
- This command does nothing; it bears a label for `b' and `t' commands to branch to.
- (1)=
- Place the current line number on the standard output as a line.
- (2){
- Execute the following commands through a matching `}' only when the pattern space is selected.
- (0)
- An empty command is ignored.
REGULAR EXPRESSIONS
The regular expressions are POSIX basic regular expressions (the one engine shared with ed(1), ex(1), vi(1) and grep(1)): an ordinary character matches itself; `.' matches any character; `[set]' matches any character of the set, `[^set]' any character not in it, with ranges `a-z' and the classes [:alpha:] [:digit:] [:alnum:] [:upper:] [:lower:] [:space:] [:blank:] [:punct:] [:print:] [:cntrl:] [:graph:] [:xdigit:]; `\(' `\)' group, and `\1' to `\9' match what the corresponding group matched; `*' matches zero or more of the preceding item, `\{n,m\}' between n and m of it (`\{n\}' exactly n, `\{n,\}' at least n), and `\+' and `\?' one or more and zero or one; `^' at the start and `$' at the end anchor the match, `\n' matches the embedded newline, and `\' before any other character takes it literally. A bare `{', `+', `?', `(' or `)' is an ordinary character. The match is the leftmost one, and the longest at that position. There is no alternation and no `-E'.
FILES
/usr/man/SED this page
SEE ALSO
DIAGNOSTICS
A script error is reported as `sed: message near char N', N being the offset into the assembled script text; the exit status is 2. Exit status is otherwise 0, the argument of `q', or 2 when a file could not be opened (the remaining input is still processed) or a space overflowed.
BUGS
An input line is cut at 255 characters. `N' at the end of input follows POSIX and BSD (no output), not GNU. A `\' before the delimiter inside a bracket expression is not special; the delimiter itself is literal there.
| September 11, 2026 | maux |