# @groups = groups1($s)
sub groups1 {
    return map { substr $_[0], $-[$_], $+[$_] - $-[$_] } 1..$#- 
}

##</code><code>##

# educated_foo
sub groups2 {
    no strict 'refs'; return map { $$_ } 1..$#- 
}

##</code><code>##

# demerphq
sub groups3 {
    return eval '($'.join(',$',1..$#-).')'
}

##</code><code>##

             s/iter     demerphq    @- and @- educated_foo
demerphq       6.65           --         -44%         -50%
@- and @-      3.72          79%           --         -10%
educated_foo   3.34          99%          11%           --

##</code><code>##

#!/usr/bin/perl

# @groups = groups1($s)
sub groups1 {
    return map { substr $_[0], $-[$_], $+[$_] - $-[$_] } 1..$#- 
}

# educated_foo
sub groups2 {
    no strict 'refs'; return map { $$_ } 1..$#- 
}

# demerphq
sub groups3 {
    return eval '($'.join(',$',1..$#-).')'
#    return '($'.join(',$',1..$#-).')'
}

use constant MAX_GROUPS => 5; # use < 26 which corresponds to 'z'
use constant MAX_REPEAT => 20;

use constant ITERATIONS => 20000;
use constant RANDOM_SEED => 666;

# $re = random_re($n)
#   a regex with $n groups qr/(a*)(b*)...(z*)/
#   use $n >= 1 to avoid the subtleties of the empty regex qr//
sub random_re {
    my $n = shift;
    my $re = join '', map { "($_*)" } 'a' .. chr(ord('a')+$n-1);
    return qr/$re/
}

# $text = random_text($n) 
#   a text which matches the random regex produced by random_re
sub random_text {
    my $n = shift;
    return join '', map { $_ x int(rand(MAX_REPEAT)) } 'a' .. chr(ord('a')+$n-1)
}

sub run_code {
    my $sub = shift;

    srand(RANDOM_SEED); # always the same seed - so that it is reproducible

    for (1..ITERATIONS) {
        $n = int(rand(MAX_GROUPS))+1;
        $re = random_re($n);
        $text = random_text($n);
#        print "n: $n, re: $re, text: $text\n";

        die "not ok: $text =~ /$re/" unless $text =~ /$re/; # assertion

        @groups = $sub->($text);
#        print "groups 1: @groups\n";
    }
}

#run_code(\&groups2);
#exit;

use Benchmark qw(cmpthese :hireswallclock );

use constant MIN_RUNTIME => 30; # benchmark minimum time: ? CPU seconds
cmpthese(-MIN_RUNTIME(), {
    '@- and @-' => sub { run_code(\&groups1) },
    'educated_foo' => sub { run_code(\&groups2) },
    'demerphq' => sub { run_code(\&groups3) },
});