Louis,
You will have to forgive me if this is wrong - my brain has been missing for the last couple of weeks (Philippines) and I just found it (back to work).

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

my $file = $ARGV[0] || 'input.txt';
open (INPUT, '<', $file) or die "Unable to open $file for reading : $!
+";
my %data;

while ( <INPUT> ) {
    chomp;
    my @info = split /\|\|/;
    push @{$data{$info[3]}}, $info[1] if ! $data{$info[3]} || ! grep /
+^$info[1]$/ , @{$data{$info[3]}};
}

for my $id ( keys %data ) {
    print "$id : " , scalar @{$data{$id}}, "\n";
}
[download]

It assumes you only want unique ids counted - it is much easier if that is not the case : $data{$name}++

Cheers - L~R