FWIW, this is an implementation in Perl of an exotic sort algorithm, comb sort (or Dobosiewicz's sort - see https://en.wikipedia.org/wiki/Comb_sort for details), using prototypes to mimic Perl's internal sort function:

sub comb_sort (&\@) {
    my $code_ref = shift;
    my $v_ref = shift;
    my $max = scalar (@$v_ref);
    my $gap = $max;
    while (1) {
        my $swapped = 0;
        $gap = int ($gap / 1.3);
        $gap = 1 if $gap < 1;
        my $lmax = $max - $gap - 1;
        foreach my $i (0..$lmax) {
            local ($a, $b) = ($$v_ref[$i], $$v_ref[$i+$gap]);
            ($$v_ref[$i], $$v_ref[$i+$gap], $swapped) = ($$v_ref[$i+$g
+ap], $$v_ref[$i], 1)
                if $code_ref->($a, $b) > 0;
        }
        last if $gap == 1 and $swapped == 0;
    }
}
[download]

This sort subroutine can be called with a code block just as Perl's internal sort. For example:

#!/usr/bin/perl
use strict;
use warnings;

my @v;
my $max = 500;

$v[$_] = int rand(20000) foreach (0..$max);

comb_sort {$a<=>$b} @v;
print "@v";
[download]

I'm not sure that's what you're after, but you can see above a calling syntax with a simple code block (no need to build an actual subroutine).