$ echo foobar | pcre2grep -o1 -o2 -o3 --om-separator=, '(.)(.)(.)'
f,o,o
b,a,r

$ echo foobar | perl -C -lne 'print "$1,$2,$3" while /(.)(.)(.)/g'
f,o,o
b,a,r

     -o Always print filename headers with output lines.

-o Print each match, but only the match, not the entire line.

$ echo 'a -b- c' | grep '\<-b-\>'
$ echo 'a -b- c' | grep -we -b-
a -b- c

$ echo 'a--b--c' | grep -we -b-
a--b--c

$ echo aaa | sed 's/\<a/<a/g
<a<a<a
$ echo aaa | grep -o '\<a'
a
a
a

1. GNU grep version 3.11 (Linux)


2. macOS grep version 13.7.8 (BSD-based)


3. OpenBSD grep version 7.7 (the only one that doesn’t support <code>-C NUM</code> and requires <code>-CNUM</code>)


4. FreeBSD grep version 14.3. See: https://man.freebsd.org/cgi/man.cgi?query=grep&apropos=0&sektion=0&manpath=FreeBSD+14.3-RELEASE+and+Ports&arch=default&format=html


5. NetBSD grep version 10.1


6. Oracle Solaris 11 grep Documentation: https://docs.oracle.com/cd/E88353_01/html/E37839/grep-1.html


7. BusyBox grep version 1.1.36


8. ToyBox grep version 0.8.9


9. Illumos grep - source https://github.com/illumos/illumos-gate/blob/master/usr/src/cmd/grep/grep.c.

• For busybox and toybox I installed their packages on Debian and
  then ran “busybox grep” and “toybox grep” to check their support.


• I haven’t been able to fully install OpenBSD on my system, but from
  the installer I could drop to a shell and use its “grep”.

grep [-C[num]]

grep [-C num]

As requested, I've tried to rewrite this to a single proposed change. I tried to make it sound "POSIX-like". Improvements welcome.

Desired Action

Update the grep utility specification to add the following widely-implemented options that improve grep's functionality across systems, specifically -A -B -C -G -H -L -h -m -o with majority semantics, extending the existing POSIX grep specification here: https://pubs.opengroup.org/onlinepubs/9799919799/utilities/grep.html

This is the proposed resolution for
https://austingroupbugs.net/view.php?id=1941

Synopsis Changes:

Modify the synopsis to include the new options:

grep [-E|-F|-G] [-c|-l|-L|-q] [-H|-h] [-inosvx] [-A num] [-B num] [-C num] [-m num] -e pattern_list [-e pattern_list]... [-f pattern_file]... [file...]

grep [-E|-F|-G] [-c|-l|-L|-q] [-H|-h] [-inosvx] [-A num] [-B num] [-C num] [-m num] [-e pattern_list]... -f pattern_file [-f pattern_file]... [file...]

grep [-E|-F|-G] [-c|-l|-L|-q] [-H|-h] [-inosvx] [-A num] [-B num] [-C num] [-m num] pattern [file...]

Option Descriptions:

Add the following option descriptions to the OPTIONS section:

-A num

Write num lines of context after each selected line.

-B num

Write num lines of context before each selected line.

-C num

Write num lines of context before and after each selected line. Equivalent to specifying both -A num and -B num.

-G

Interpret patterns as basic regular expressions. This is the default behavior.

-H

Prefix each output line with the filename followed by a colon (or dash if it is a context line and is not selected). Default when multiple files are specified.

-L

Write only names of files containing no selected lines, one per line. Empty files are included if processed.
If standard input was processed its name is considered "(standard input)".
If the same filename was processed multiple times, and it is to be output,
it's implementation-defined whether its name will be output only once or once for each occurance.

-h

Suppress filename prefixes. Default when reading standard input or a single file.

-m num

Stop reading each file for matching after num selected lines. Complete context for the last match should be written after stopping, but any lines written afterwards are not considered matches. Implementations may instead stop writing lines for that file after writing the corresponding information about the last match, without the trailing context. This is applied independently to each file. num shall be a non-negative decimal integer; if zero, behavior is implementation-defined.

-o

Write only matching non-empty portions of selected lines, possibly prefixed with filename and line number information (see STDOUT). Multiple non-overlapping matches for one pattern are written as separate lines, with the same filename and/or line number if they are written. Combining -o with context options (-A, -B, -C) or -v is implementation-defined.

For context options (-A, -B, -C), if there are overlapping or adjacent regions of context, the lines are written in sequence without duplication, and the group separator ('--') is written to separate non-adjacent context regions. When writing a filename and/or line number, non-matching context lines use dash separators while selected lines use colon separators (see STDOUT section). If a context line also matches the pattern, it shall be displayed as a selected line and counted toward the -m limit. A context does not extend before the beginning or after the end of a file. For all numeric arguments, num shall be a non-negative decimal integer.

STDOUT Changes:

If -q is specified, there is no output to standard out regardless of other options.

Otherwise, if the -l or -L option is in effect, the following shall be written for each file containing at least one selected input line:

Otherwise, if the -H option is selected or more than one file argument appears (and -q, -l, and -L are not specified), the grep utility shall prefix each output line by this, where <context-marker> is ':' if this line is selected and '-' if this line was not selected but is instead only being displayed as a context line:

"%s%s", <file>, <context-marker>

The remainder of each output line shall depend on the other options specified:

If the -c option is in effect, the remainder of each output line shall contain:

"%d\n", <count>

Otherwise, if -c is not in effect and the -n option is in effect, the following shall be written to standard output, where <context-marker> is the same as described above:

"%d%s", <line number>, <context-marker>

Finally, if the '-o' option was selected, the following shall be written to standard output, with multiple lines for each selected-part that matches:

"%s\n", <selected-part>

Otherwise, if the -o option was not selected, the following shall be written to standard output, which may either be a selected line, or one of the context lines before or after a selected line:

"%s\n", <line contents>

Context group separators shall consist of a single line containing only "--" when context from different matches would otherwise be adjacent and there is at least one non-context line between the context regions in the original file.

Option Interactions:

• Context options -A and -B may be combined; the result is the union of the specified ranges.


• The impact of -C combined with -A and/or -B is implementation-defined; implementations are encouraged to use left-to-right application where last one(s) given take precedence.


• Pattern interpretation options (-E, -F, -G), filename options (-H, -h), and output format options (-l, -L, -o) are mutually exclusive within each group; the last specified takes effect.


• Context and -o options have no effect with -c, -l, -L, or -q.


• Context options with -v display context around non-matching selected lines.


• Binary file treatment is implementation-defined.

Error Conditions:

Invalid numeric arguments shall cause grep to write a diagnostic message to standard error and exit with status greater than 1.