First you might also try perusing perldsc for good info on Hashes of Hashes, Arrays of Arrays, Arrays of Hashes,etc.

If the files are not all that big (to justify using a database, as I would do if the files were big, and then run an SQL Query), or I wanted a little practice honing up my hashes of hashes skills I might do the following (I have removed some code to let you work out some details, this is probably not the optimal way but it is one way):

My logic goes as follows:

1. open <FILE2> and place all the cities in a hash keyed by zipcode something like:

  
  chomp($line)             #line from FILE2
  my ($zip_code,$city) = split /\|/,$line;
  $zip_city{$zip_code} = $city; #insert city into hash
[download]

2. open <FILE1> again split it and compare the keys in my original hash setting the state and city in a hash again keyed against zip (the delete $zip_city{$current_zipcode}; part is so that I don't continue to go over zips I already have a state for, you could also put a last; so it would not continue reading from <FILE1> once the hash %zip_city contained no keys).

 chomp($line);
 my ($state,$min,$max) = split /,/,$line;
 foreach my $current_zipcode ( keys %zip_city )
 {
   if ( $current_zipcode >= $min and $current_zipcode <= $max )
   {
    $zip_city_state{$current_zipcode}{city}  = $zip_city{$current_zipc
+ode};
    $zip_city_state{$current_zipcode}{state} = $state;
    delete $zip_city{$current_zipcode};
   }
 }
[download]

3. Lastly for each line in <FILE3> I would see if the zipcode (assuming you put it in a variable called $zipfrom3)existed in the second hash (%zip_city_state hash) and if it does print the info you want:

print "$zipfrom3,$zipzip_city_state{$zipfrom3}{city}, $zip_city_state{
+$zipfrom3}{state}\n";
[download]

or else complain that it did not exist in the hash (if you use warnings; it will do a good job of complaining if you don't check for the existance of a zip in the hash first, but you should check for the existance anyhow).

-enlil