Install the Perl module from CPAN, Devel::NYTProf, write a test script that invokes this subroutine, and then profile it. I would not suggest using NYTProf on code that also uses Benchmark -- just write a script that invokes the subroutine a few times with varying but realistic data and run it through the profiler.
By doing this you'll gain a better appreciation for where the hotspots are. Without doing this, we really have to just use intuition, which is not as reliable as real data.
Dave
In reply to Re: Speeding up a Perl code. Can it be faster?
by davido
in thread Speeding up a Perl code. Can it be faster?
by $h4X4_|=73}{
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |