$k3 = %{$hashr->{ $k1 }->{ $k2 }};
[download]

is analogues to

$k=%hash
[download]

$k3 is not a hashref but the scalar-value of a hash (... which produces something mostly useless like 1/8 indicating the memory use)

what you really want is either

 
$k3 = $hashr->{ $k1 }->{ $k2 };
[download]

to get the hashref

or

 
%h3 = %{$hashr->{ $k1 }->{ $k2 }};
[download]

to copy the nested sub-hash!

You have used the $comment as a key in the hash and not the value to be assigned to the hash ... I will show an example at the bottom, probably taking a retrospective look can prove beneficial ...

using the module Data::Dumper can allow you to view the data structure representation and understand or decide how to approach it since it makes that data structure stringified...

The Data Structure Cookbook is fantastic, I hope you have consulted it and that you can revert to it over and over, it is so useful...

I found Referencing in advanced data structures a very good source of clearing my confusion regarding many things...They seem difficult to understand but once you practiced them enough and you would find yourself starting to make sense of it all.. Also, Check the Tutorials section, for many related things have been graciously contributed by the beloved monks....

in the code example, notice the two lines:

$hash{$name}{$id}= $comment;
push @{$hash{$name}{$id}}, $comment;  #this one has been commented out
[download]

my %hash;
my ($name, $id, $comment);

while(<DATA>){
              chomp;
              my ($name, $id, $comment)=split;
              $hash{$name}{$id}= $comment;          #adding comment to
+ related $name/$id..


              #push @{$hash{$name}{$id}}, $comment; #create an anonymo
+us array
                                                    #in case more than
+ one $comment or other variables related to same $name/$id
              }
[download]

Name1 12 firstname1 some more info
[download]

The name, the ID and then followed by more than one item to be associated with $name,$id, you would configure your split to break properly and then treat $hash{$name}{$id} as an anonymous array reference that would accept qw(some more info), using push or other array manipulation techniques...

Iterating is done with respect to the type of reference and it's location within the structure, the function ref or the Data::Dumper output can be handy at this time

for my $key1(sort keys %hash){
    print "$key1\t->$hash{$key1}\t";        #$hash{$key} hash referenc
+e,to access it dereference it.....
    foreach $key2(sort keys %{$hash{$key1}}){
            print "$key2\t->$hash{$key1}{$key2}\t\n";
            }
    }
[download]

I hope this has been as concise and worthy for you to get to grabs, I haven't mentioned about iterating through @{$hash{$name}{$id}} for my confidence that you would figure it out. Let us know if you get any confusion...this is the entire code example

use strict;
use warnings;
my %hash;
my ($name, $id, $comment);
while(<DATA>){
              chomp;
              my ($name, $id, $comment)=split;
              $hash{$name}{$id}= $comment;


              #push @{$hash{$name}{$id}}, $comment;
              }


print "Key\t\t Val\t\t Key\t Val\n";
print "---------------------------------------------------\n";


for my $key1(sort keys %hash){
    print "$key1\t->$hash{$key1}\t";
    foreach my $key2(sort keys %{$hash{$key1}}){
            print "$key2\t->$hash{$key1}{$key2}\t\n";
            }
    }

use Data::Dumper;
print Dumper(\%hash);
__DATA__
Name1 12 firstname1
Name2 55 firstname2
Name3 20 firstname3
[download]

my $k3 = $hashr->{$k1}->{$k2};
[download]

I found References quick reference helped me a lot with this kind of problem.

The following nodes may be helpful:

• intro to references
• References quick reference
• Arrays and hashes only with references
• How can I visualize my complex data structure?