I have a large text file that contains several fields in a fixed format. Among the fields are a unique numerical ID, and a text ID that ideally should be unique. The text id also should not start with a capital letter, but sometimes does.

My script reads the file, detects IDs that start with a capital letter, and also detects if any of the textual IDs are duplicated. The numerical IDs are always unique.

I find duplicates by storing a count in a hash where the text id is the key. After I have filled the hash, I then traverse it to find values higher than one. If such a value is found, I run through the entire file again to find the numerical IDs of the duplicates.

This means that I first go through the file once to detect duplicates, and then go through the file again once for each duplicate found. I can't help but think that there is a more elegant and efficient way of doing things. My code is shown below:

#!perl -w
use strict;
my(@lines,%dup, $id, $key);
chomp(@lines=<STDIN>);
# read file, put count of key in hash
print "Keys with initial caps:\n";
foreach(@lines) {
   if (m/PROBABLECAUSE\w*\((\d+),\s*\w*,\s+(\w*)/) {
      ($id, $key)= ($1, $2);
      $dup{$key} += 1;
      # check if any key has init caps (not allowed)
      if ($key =~ m/^[A-Z]\w*/) {
         print "Id: $id - $key\n";
      }
   }
}
# check hash for duplicates, if found, display positions in file
print "Duplicated keys:\n";
foreach $key (keys %dup) {
  if ($dup{$key}>1) {
     print "$key ($dup{$key})\n";
     foreach(@lines) {
        if (m/PROBABLECAUSE\w*\((\d+),\s*\w*,\s+($key),/) {
           print "Id: $1\n";
        }
     }
  }
}
[download]

PROBABLECAUSE(0, probablecauseUndefined, undefined, Unknown, indetermi
+nate, prim, false, "", UNIDENTIFIED, Y)
[download]

update

Using the tips I received, the solution is now cleaner and more elegant, and as an added bonus, I have learned about perl references. Thank you, monks!