If I understand correctly, you want to look at a substring of three characters as position '0' in your haystack, and see if it matches some permutation of the characters in (A,A,C). If so, do something. Then advance your position to 1, and check again. This is possible with a regex, but I think that it reads clearer (and is probably no less efficient) if you use substr to check each position within your haystack against a hash holding all permutations of your "needle" characters:
use Algorithm::Permute; my $needle_chars = 'AAC'; my $haystack = 'AAACACAA'; my $p = Algorithm::Permute->new( [ split //, $needle_chars ], 3 ); my %perms; while( my @perm = $p->next ) { $perms{ join '', @perm }++; } my $pos = 0; while( $pos + 3 < length $haystack ) { my $pos_chars = substr $haystack, $pos, 3; print $pos_chars, " found at $pos\n" if exists $perms{ $pos_chars }; $pos++; }
I don't know how big your set of needle characters really is. Perhaps instead of three it's really 100 characters, in which case holding onto all the permutations is impractical. But if the number of characters you are permuting is not too big this is an efficient solution.
Dave
In reply to Re: Regex: finding all possible substrings
by davido
in thread Regex: finding all possible substrings
by ssc11008
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |