Benchmark of String Comparison Methods

The fastest way of identifying (non)empty strings in Perl is the direct method ($foo eq ''), the slowest is the regular expression ($foo =~ /^$/):

	#!/usr/bin/env perl
	use strict;
	use warnings;
	use Benchmark ( 'cmpthese' );
	my $foo = '';
	cmpthese( 10000000, {
	'direct' => sub { return $foo eq '' },
	'length' => sub { return length $foo },
	'regexp' => sub { return $foo =~ /^$/ },
	});
	$foo = 'something';
	cmpthese( 10000000, {
	'direct' => sub { return $foo eq '' },
	'length' => sub { return length $foo },
	'regexp' => sub { return $foo =~ /^$/ },
	});

view raw benchmark-empty-strings.pl hosted with ❤ by GitHub

My own results:

             Rate regexp direct length
regexp  4219409/s     --   -76%   -78%
direct 17543860/s   316%     --    -7%
length 18867925/s   347%     8%     --
             Rate regexp length direct
regexp  5128205/s     --   -73%   -74%
length 18867925/s   268%     --    -6%
direct 20000000/s   290%     6%     --