The code is easier to read this way. So easy that I can see a BUG! (I left it in with a comment).

The program could be made even clearer and further optimized, but this is a start. For example you can replace all the print DAT $c; with $all .= $c; and, wait till the end to open the file for append, print DAT $all;, and close.

By the way, if you can, you should test this in a standalone program on your computer, not just on the web.

# 1. Only open/close the file once to append, instead of $maxid times.
# 2. Use temporary values.
# 3. Delete stuff you don't need immediately, not next time around loo
+p.
# 4. In this case, ($c) is same as (defined $c) (cosmetic).

open(DAT,">>data/parsed-all.txt");
for(my $i = 1; $i <= $maxid; $i++) {
  my $currid = changeID($i);
  my ($c1,$c2,$c3) = ($compare1{$currid}, $compare2{$currid}, $compare
+3{$currid});
  delete $compare1{$currid}, $compare2{$currid}, $compare3{$currid};
  next if( $c1 && $c2 && $c3 && $c1 ne $c2 && $c1 ne $c3 && $c2 ne $c3
+ );
  if( $c1 && !$c2 && !$c2 ) {
    print DAT $c1;
    next;
  }
  if( $c2 && !$c1 && !$c3 ) {
    print DAT $c2;
    next;
  }
  if( $c3 && !$c1 && !$c2 ) {
    print DAT $c2; # <-- BUG HERE!
    next;
  }
  if( $c1 && $c2 ) {
    if( $c1 eq $c2 ) {
      print DAT $c1;
      next;
    }
  }
  if( $c1 && $c3 ) {
    if( $c1 eq $c3 ) {
      print DAT $c1;
      next;
    }
  }
  if( $c2 && $c3 ) {
    if( $c2 eq $c3 ) {
      print DAT $c2;
      next;
    }
  }
}
close(DAT);
[download]