Unicode Collation Algorithm in Perl

unicode_cmp.pl

use strict;                                                                                                                                                  
use warnings;
use utf8::all;

use Test::More;
use Unicode::Collate;

my $a1 = "\N{U+212B}";              # Å
my $a2 = "\N{U+00C5}";              # Å
my $a3 = "\N{U+0041}\N{U+030A}";    # Å

subtest "Perl's eq() operator" => sub {
    ok !( $a1 cmp $a1 ), "U+212B and U+212B should evaluate the same";
    ok !( $a2 cmp $a2 ), "U+00C5 and U+00C5 should evaluate the same";
    ok !( $a3 cmp $a3 ), "U+0041 U+030A and U+0041 U+030A should evaluate the same";
  TODO: {
        local $TODO = "Perl's cmp operator is broken";
        ok !( $a1 cmp $a2 ), "U+212B and U+00C5 should evaluate the same";
        ok !( $a1 cmp $a3 ), "U+212B and U+0041 U+030A should evaluate the same";
        ok !( $a2 cmp $a3 ), "U+00C5 and U+0041 U+030A should evaluate the same";
    }
};

subtest "Unicode Collation Algorithm" => sub {
    my $c = Unicode::Collate->new;
    ok !( $c->cmp( $a1, $a1 ) ), "U+212B and U+212B should evaluate the same";
    ok !( $c->cmp( $a2, $a2 ) ), "U+00C5 and U+00C5 should evaluate the same";
    ok !( $c->cmp( $a3, $a3 ) ), "U+0041 U+030A and U+0041 U+030A should evaluate the same";

    ok !( $c->cmp( $a1, $a2 ) ), "U+212B and U+00C5 should evaluate the same";
    ok !( $c->cmp( $a1, $a3 ) ), "U+212B and U+0041 U+030A should evaluate the same";
    ok !( $c->cmp( $a2, $a3 ) ), "U+00C5 and U+0041 U+030A should evaluate the same";
};

done_testing;