peg - Perl _expression_ (GNU) grep script

[Clarendon4]

Hope others find "peg" as useful as it has been to me. Give it a try.

Announce - "peg", yet another (GNU) grep like Perl program.

The latest version is available at:

http://cpan.mirrors.uk2.net/authors/id/A/AD/ADAVIES/

Features:

• Basic synopsis:  peg [OPTION]... PERLEXPR [FILE]...
• Uses Perl expressions, NOT regular expression patterns.

    eg% peg /needle/i haystack

    eg% peg "!(/^=\w/ .. /^=cut/) and /whatever/" *pm

• ...but automatically converts simple "string" expressions to "/string/".

These are identical:

    eg% peg  needle  haystack
    eg% peg /needle/ haystack

• Supports, or has alternatives to, all the commonly used GNU grep options.

    eg% peg -inHC1Tm1 FOO words.txt
    words.txt-420-    afloat
    words.txt:421:    afoot
    words.txt-422-    aforementioned

• If no files are specified, does The Right Thing

    eg% cat haystack | peg -i needle

...searches the input stream, while

    eg% peg -i needle

...recursively searches each file beneath the cwd.

• Is configurable through variables set in a configuration file "peg_ini.pl".
• Options to restrict files searched based on name and last modification time.

    eg% peg -p "/\.p[ml]$/i" -M 48h whatever

...looks in .pm & .pl files modified in the last 2 days.

Commonly used -p tests of file extensions can be defined:

    $Peg_p{p} = 'pl:pm:t'; # peg_ini.pl

...and then used as:

    eg% peg -p p -M 2d whatever

• The "-z" option prints 'context'.

    eg% peg -z "/^sub \w+/" foo *pm

...prints the last matched "sub" line for any lines matching "foo" ie. (probably) the subroutine that contains "foo".

Again, common 'contexts' can be defined:

    $Peg_z{p} = '/^(?:\s*sub\s+\w|=head|__(?:END|DATA)__)/'; # peg_ini
+.pl

...and then used as "-z p".

    eg% peg chmod Temp.pm -nz p
    **** (331) sub _gettemp {
    523:    chmod(0600, $path);
    545:    chmod(0700, $path);
    **** (645) sub _force_writable {
    647:  chmod 0600, $file;

• Customizable command line options.

    # peg_ini.pl
    $Peg_longopt{'ignore-dir'} = sub {
        my $argv_ref = shift;
        my $dir_name = shift @$argv_ref or die;
        unshift @$argv_ref, "-p", qq{ \$File !~ m:(^|/)$dir_name/: };
    };

...would then enable

    eg% peg foo --ignore-dir CVS

...to search for foo, ignoring all files in or beneath a directory called CVS.

• Windows friendly - does filename globbing.
• Coloured output. (On Win32, this requires Win32::Console::ANSI.)
• Fast find...

As an optimization of the common case of doing a recursive search, an external program can be configured to feed peg the list of files beneath the current directory. The C program qfind.c which is also available at:

http://cpan.mirrors.uk2.net/authors/id/A/AD/ADAVIES/

...is written specifically for this, and provides a significant speed up on Win32 compared to the default of traversing the file system with Perl's File::Find::find. Additionally peg performs parallel processing to maximize performance.

• Can read and write encoded data.
• Lots more options. See the "-help".