note
roboticus
[id://546129]:<p>
Here's my stab at it. I took out the hash because my method doesn't generate duplicates. Instead I just return the list of results:<p>
<code>
#!/usr/bin/perl -w
use strict;
use warnings;
# initial source in sequence order
my $source = 'horse:cow:dog:cat';
function($source);
# and now the results
foreach (function($source)) {
print "$_\n";
}
# generate array of combinations
sub function {
my @t = split /:/, shift;
my @res=(shift @t);
for my $i (@t) {
@res=($i, @res, map{$_.':'.$i} @res);
}
return @res;
}
</code>
<b>UPDATE:</b> I didn't benchmark it because I've never used that module before. (I'll have to go install it and read up on it.) But I submitted it because I suspect that generating the strings from components might be faster than removing chunks. Any benchmarking ninjas out there wanna help me out?<p>
--roboticus
554374
554374