true. So if order matters, $base and $compare should not be switched. Anyway the first field must be unique in the $base file then, otherwise it wouldn't make much sense, so the code is still not that wrong. Thanks.

I think the first field could still duplicate, because the second disambiguates.

That gives me another idea on doing it, too: store the concatenation of the first two fields, and use that as the lookup. Then we don't need to do a second comparison.

So this would be a very open method:

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

my $base = shift or die;
my $cmpdepth = 2;
my %base = ();

{
  open my $fh ,'<',$base or die "opening '$base' failed: $!";
  while(defined(local $_=<$fh>)){chomp;
     my $c = \%base;
     for(split){
       $c = ($c->{$_} = {});
     }
  }
  close $fh;
}

while(defined(local $_ =<>)){chomp;
  my @s = split;
  my $f = 1;
  my $c = \%base;
  for( 0 .. ($cmpdepth - 1) ){
     $f-- and last unless
      exists $c->{ $s[$_] } and $c = $c->{$s[$_]};
  }
  print "$_\n" if $f;
}
[download]

that gives the possibility to check up to any number of fields.

--
http://fruiture.de