You want to compare each line to every other line and print only those with more than one term in common, right?
Assuming that's what you're looking for, you don't need to worry about combinations. Compare the first element with every other element. Then compare the second with every other element except the first. Repeat until complete.
Here's a possibly cleaner way to the same result. If your data is more complex (it probably is ;) you may need to tweak the regex.
#!/usr/bin/perl
use strict;
use warnings;
chomp (my @lines = <DATA>);
for my $i (0 .. $#lines) {
(my $re = $lines[$i]) =~ s/_/|/g; # create $re outside the inner l
+oop
for my $j ($i + 1 .. $#lines) {
my $count = () = $lines[$j] =~ /$re/g;
if ($count >= 2) {
print "$lines[$i] and $lines[$j]\n";
}
}
}
__DATA__
one_two
one_three_two
three_one
one_four
four_three_one