NAME
Regexp::Keep - filter to allow the \K
escape in regexes
SYNOPSIS
use Regexp::Keep;
# slow and inefficient
my $r = "abc.def.ghi.jkl";
$r =~ s/(.*)\..*/$1/;
# fast and efficient
my $s = "abc.def.ghi.jkl";
$s =~ s/.*\K\..*//;
DESCRIPTION
This allows you to use the \K
escape in your regexes, which fools the regex engine into thinking it has only just started matching your regex. This means you can turn the inefficient replace-with-itself construct
s/(save)delete/$1/;
into the more efficient
s/save\Kdelete//;
construct.
IMPLEMENTATION
What \K
filters into is .{0}(?{ Regexp::Keep::KEEP })
, which is an XS function call embedded into the regex. The function sets PL_regstartp[0]
to the current location in the string. This means that $&
now starts where \K
is seen. That means a replacement will begin being replaced there.
EXAMPLES
Here's are short examples to show you the abilities of \K
:
"alphabet" =~ /([^aeiou][a-z][aeiou])[a-z]/;
# $1 is "pha", $& is "phab"
"alphabet" =~ /\K([^aeiou][a-z][aeiou])[a-z]/;
# $1 is "pha", $& is "phab"
"alphabet" =~ /([^aeiou]\K[a-z][aeiou])[a-z]/;
# $1 is "pha", $& is "hab"
"alphabet" =~ /([^aeiou][a-z]\K[aeiou])[a-z]/;
# $1 is "pha", $& is "ab"
"alphabet" =~ /([^aeiou][a-z][aeiou])\K[a-z]/;
# $1 is "pha", $& is "b"
"alphabet" =~ /([^aeiou][a-z][aeiou])[a-z]\K/;
# $1 is "pha", $& is ""
BUGS
I fixed a bug where \K
following a simple thing (like a letter that isn't followed by a quantifier) didn't work properly.
If you're using this module, you don't have a version of Perl with the \K
escape built-in. Bummer. I should try to make it built-in.
HISTORY
AUTHOR
Jeff japhy
Pinyan, japhy@pobox.com
COPYRIGHT AND LICENSE
Copyright (C) 2004 by japhy
This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself, either Perl version 5.8.3 or, at your option, any later version of Perl 5 you may have available.