Here is OWTDI - but I bet by the time I finish this post, some other monk will come up with a faster, more Perlish way.

#!/usr/bin/perl -w
use strict;

print UniqueCount("1234567812") . "\n";

sub UniqueCount {
    my $string = shift;
    if ($string) {
        my %unique;
        return grep { !$unique{$_}++ } split //, $string;
    }
    else {
        return 0;
    }
}
[download]

You will want to modify if you do not want to count spaces or if you want to be case insensitive

Hope this helps - happy hacking - L~R

Update:Dropped un-needed @array in grep