my ($first_key, $first_val) = each(%{$main_hash{$k}[$x]};
[download]

One other thing that stands out to me... The following line lets the loop go one iteration too many. You should use $x instead of $x + 1:

    last if ($x + 1 >= scalar(@{$main_hash{$k}});
[download]

Hello, please refer above for the corrected version. Regarding the other thing that you mentioned, nope, I think I need  $x+1, that is because when I reach the last item in the array, I want to exit the inner 'for loop', if not  my ($next_key, $next_val) = each(%{$main_hash{$k}[$x+1]}); would give me 'Uninitialized values' errors.

Hello jokerzy,

(I used Data::Dumper to print the hash out, and I think it's a hash of arrays of (ref to) hash, correct?)

Actually, it’s a hash of refs to anonymous arrays, the elements of which are refs to anonymous hashes — but you’ve got the idea! :-)

As to the code, frozenwithjoy had the right intuition: the logic of your inner loop is incorrect. However, the fix requires a different approach:

#! perl
use strict;
use warnings;
use Data::Dumper;

my %main_hash = (
                    keyA => [
                                { keya => 'vala' },
                                { keyb => 'valb' },
                                { keyc => 'valc' },
                                { keyd => 'vald' },
                            ],

                    keyB => [
                                { keye => 'vale' },
                                { keyf => 'valf' },
                                { keyg => 'valg' },
                                { keyh => 'valh' },
                            ],

                    keyC => [
                                { keyi => 'vali' },
                                { keyj => 'valj' },
                            ],
                );

print Dumper(%main_hash), "\n";

foreach my $k (sort keys %main_hash)
{
    my @array = @{ $main_hash{$k} };

    foreach my $i (1 .. $#array)
    {
        my %inner_hash_first = %{ $array[$i - 1] };
        my %inner_hash_next  = %{ $array[$i    ] };

        my ($first_key, $first_val) = each( %inner_hash_first );
        my ($next_key,  $next_val)  = each( %inner_hash_next  );

        print "Now compare ($first_key : $first_val) with ($next_key :
+ $next_val)\n";
    }
}
[download]

Note in the above script that dereferencing has been decomposed into separate stages by using intermediate variables (@array, %inner_hash_first, and %inner_hash_next). This is a good idea for two reasons:

1. It makes the code clearer, and so easier to follow and debug.
2. In some cases it also promotes efficiency: for example, @{ $main_hash{$k} } is now re-calculated only once for each new value of $k.

The output from the above script is:

Now compare (keya : vala) with (keyb : valb)
Now compare (keyb : valb) with (keyc : valc)
Now compare (keyc : valc) with (keyd : vald)
Now compare (keye : vale) with (keyf : valf)
Now compare (keyf : valf) with (keyg : valg)
Now compare (keyg : valg) with (keyh : valh)
Now compare (keyi : vali) with (keyj : valj)
[download]

which shows that each value of each inner hash is available for comparison with its successor (if it has one), as required.

(You should probably also add a check to verify that each inner hash contains exactly one key/value pair.)

Update

Is that a pseudo-hash I'm using here?

No, according to perldoc, pseudohashes “have been removed from Perl.” See the old thread My views on pseudohashes.

HTH,

Hi Athanasius, <== My first link to another user :)

Thanks for the help. I like to way your code is written, especially how they're broken into different sections; it was very easy to grasp. Thanks also for the update about my pseudo-hash question.

Signing off,
jokerzy