>>118
そのコードで何を計測しているつもりなんだ？

#!/usr/bin/perl -w
use strict;
use Benchmark qw(timethese);

my $i = 0;
my $str = join '', 'a'..'m';
my @line;
push(@line, $str), substr($str, 0, 0, chop $str) for 1..1000;

timethese(-5, {
　　re1 => sub { /^(?=(?:abcdef|ghijklm))/ and ++$i for @line },
　　re2 => sub { /^(?=abcdef)/ || /^(?=ghijklm)/ and ++$i for @line },
　　re3 => sub { /^abcdef/ || /^ghijklm/ and ++$i for @line },
　　idx => sub { index($_, 'abcdef', 0) == 0 || index($_, 'ghijklm', 0) == 0 and ++$i for @line },
});

@line = ();
push @line, join '', map +('a','b')[ rand 2 ], 1..10 for 1..1000;

timethese(-5, {
　　re1 => sub { /^(?=ab..)(?=..ab)/ and ++$i for @line },
　　re2 => sub { /^(?=ab..)/ && /^(?=..ab)/ and ++$i for @line },
　　re3 => sub { /^ab../ && /^..ab/ and ++$i for @line },
　　idx => sub { index($_, 'ab', 0) == 0 && index($_, 'ab', 2) == 2 and ++$i for @line },
});