Re^3: file name parsing to get the original file name (regex performance)


Perl: the Markov chain saw
	PerlMonks

Re^3: file name parsing to get the original file name (regex performance)

by Aristotle (Chancellor)

on Aug 24, 2003 at 05:43 UTC ( [id://286147]=note: print w/replies, xml )

Need Help??

in reply to Re: Re: file name parsing to get the original file name
in thread file name parsing to get the original file name

Enter sexeger.

Add this to Abigail's benchmark.

    aristotle => 'foreach my $f (@files) {
                      my ($fn) = reverse($f) =~ m!^(.*?)/?!s;
                      $fn = reverse $fn;
                  }',
[download]


              Rate     markm   abigail    markm2 aristotle
markm      39625/s        --      -56%      -58%      -61%
abigail    89688/s      126%        --       -4%      -11%
markm2     93877/s      137%        5%        --       -7%
aristotle 100885/s      155%       12%        7%        --

Reversing the string (twice!) may be costly, but the simplicity of the regex offsets this. Note that [^/]+ would have been much slower. .*? has been treated to special optimizations.

Makeshifts last the longest.