Return to grep.1 CVS log

Up to [Research Unix] / researchv10dc / man / manx

researchv10 Dan Cross

.TH GREP 1
.CT 1 files
.SH NAME
grep, egrep, fgrep \(mi search a file for a pattern
.SH SYNOPSIS
.B grep
[
.I option ...
]
.I expression
[
.I file ...
]
.PP
.B egrep
[
.I option ...
]
.I expression
[
.I file ...
]
.PP
.B fgrep
[
.I option ...
]
.I strings
[
.I file ...
]
.SH DESCRIPTION
Commands of the
.I grep\^
family search the input
.I files\^
(standard input default)
for lines matching
a pattern.
Normally, each line found
is copied to the standard output.
.I Grep\^
patterns are limited regular
.I expressions
in the style of
.IR ed (1);
it uses a compact non-deterministic algorithm.
.I Egrep\^
patterns are full regular
.IR expressions ;
it uses a fast deterministic algorithm.
.I Fgrep\^
patterns are fixed
.IR strings ;
it is fast and compact.
The following
.I options\^
are recognized:
.TP
.BI -f " file\^"
Read the pattern from
.IR file ;
there is no pattern argument
.IR ( egrep
and
.IR fgrep ).
.PD 0
.TP
.B -v
Reverse: print lines that do not match.
.TP
.B -i
Ignore alphabetic case distinctions.
.TP
.B -n
Mark each printed line with its line number counted in its file.
.TP
.B -x
Exact: print only lines matched in their entirety
.RI ( fgrep\^
only).
.TP
.B -c
Print only a count of matching lines.
.TP
.B -l
Print only the names of files with matching lines (once).
.TP
.B -b
Mark each printed line with its block number counted in its file.
This is sometimes useful in locating
disk block numbers by context.
A block is 1024 bytes.
.TP
.B -h
Do not print filename headers with output lines.
.TP
.B -s
Produce no output, but return status.
.TP
.BI -e " expression\^"
Same as a simple
.I expression\^
argument; useful when the
.I expression\^
begins with a
.BR - .
.PD
.PP
Output lines are tagged by filename when there is more than one
input file.
Care should be taken when
using the shell metacharacters
.B $*[^|()\e
in
.IR expression ;
it is safest to enclose the
entire
.I expression\^
in single quotes
.BR \&\|\(fm \|.\|.\|.\| \(fm .
.PP
.I Fgrep\^
searches for lines that contain any of the
.I strings,
which appear as a single argument with
embedded newlines.
.PP
.I Egrep\^
accepts regular expressions as in
.IR ed (1),
with the following changes:
.TP
1.
There is no backreferencing
.RB ( \e(\e)\e1\e2 ...).
.PD 0
.TP
2.
The characters
.LR ^ ,
.LR $ ,
and
.L *
are always metacharacters unless quoted or contained in brackets
.BR [\^] .
.TP
3.
A regular expression followed by
.B +
matches one or more occurrences of the
expression.
.TP
4.
A regular expression followed by
.B ?
matches 0 or 1 occurrence.
.TP
5.
Two regular expressions separated by \(bv or
newline match occurrences of either.
.TP
6.
Parentheses
.B (\|)
specify grouping.
.PD
.PP
The order of precedence of operators is
.BR [\|] ,
then
.BR \(**\|?\|+ ,
then concatenation, then \(bv and new-line.
.SH SEE ALSO
.IR ed (1),
.IR sed (1),
.IR sh (1).
.SH DIAGNOSTICS
Exit status is 0 if any matches are found,
1 if none, 2 for syntax errors or inaccessible files
(even if matches were found).
.SH BUGS
Ideally there should be only one
.IR grep ,
but we do not know a single algorithm that spans a wide enough
range of space-time tradeoffs.
.br
Lines
are truncated at
.B BUFSIZ
characters; see
.IR setbuf (3).
Null characters behave as end-of-line in matches.
.\"	@(#)grep.1	6.2 of 9/2/83

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.