RE: (bbq) RE: Re: Another flatfile thingy

[perlmonkey]

Hmm, did merlyn happen to mention a better construct, the s/// for ... it is something we put into production. I guess an alternative would be  map { s/^\s+|\s+$//g } @headers But I doubt it would be significantly better. I guess I will have to benchmark.

Thanks!

[BBQ]

Nope, he just made a note of it. (Better than running with scissors though) :)

I'd like to see the results of whatever you come up with in place of s/// for. (either that or I have to RTFM and learn once and for all how to use benchmark;)

#!/home/bbq/bin/perl
# Trust no1!

[btrott]

Doing it in two steps is more efficient. From perlfaq4:

    It is much better faster to do this in two steps: 

        $string =~ s/^\s+//;
        $string =~ s/\s+$//;

    Or more nicely written as: 

        for ($string) {
            s/^\s+//;
            s/\s+$//;
        }

Here's a benchmark:

    use Benchmark;

    timethese(500_000, {
            one => sub {
        my $foo = "    perl monks    ";
            $foo =~ s/^\s+|\s+$//g;
        },

        two => sub {
            my $foo = "    perl monks    ";
            for ($foo) {
                s/^\s+//;
                s/\s+$//;
            }
        },
    });

And here are the results:

    Benchmark: timing 500000 iterations of one, two...
       one: 11 secs (11.75 usr  0.00 sys = 11.75 cpu)
       two:  8 secs ( 7.57 usr  0.00 sys =  7.57 cpu)