"Count the number of unique" problems are common; in general they can be addressed with something like this:

my %classification;

# Scan and "bucket" the items
for each $thing (@source_of_things) {
  # If the thing needs to be transformed, truncated, or pattern-matche
+d,
  # do that here to get the unique part you're interested in.
  $classification{$unique_part_of_thing}++;
}

# Analyze and report.
foreach $item (keys %classification) {
  print "$item found $classification{$item} time",
         ($classification{$item}==1 ? '' : "s"), "\n";
  # Or whatever you need to do with each unique item and the number of
  # times it was found.
}
[download]

Note that your source of things can be anything: the keys of a hash, the lines read from a filehandle, items in an array. This is a good general pattern of scan-and-count-unique that wil help in a lot of contexts. Sorting into frequency of ocurrence, etc. left as an exercise to the reader.