Re: Copying a hash to another hash problem

First a question: Your data appears to be entirely numeric. Why are you using hashes instead of arrays?

Secondly, your %newhash should not be empty after the copy operation, but be advised that it will contain references to the ORIGINAL second level hashes from the first assignment. Example:

use strict;

use Data::Dumper;

my %first;
$first{'foo'}{'bar'} = 1;
$first{'foo'}{'baz'} = 2;
$first{'bah'}{'oof'} = 3;

print "First hash:\n";
print Dumper \%first;

my %second = %first;
print "\nSecond hash\n";
print Dumper \%second;

print "\nFirst nested hash address: $first{'foo'}\n";
print "Second nexted hash address: $second{'foo'}\n";
[download]

output:

First hash:
$VAR1 = {
          'bah' => {
                     'oof' => 3
                   },
          'foo' => {
                     'baz' => 2,
                     'bar' => 1
                   }
        };

Second hash
$VAR1 = {
          'bah' => {
                     'oof' => 3
                   },
          'foo' => {
                     'baz' => 2,
                     'bar' => 1
                   }
        };

First nested hash address: HASH(0x804ee40)
Second nexted hash address: HASH(0x804ee40)
[download]

Comment on Re: Copying a hash to another hash problem Select or Download Code

Replies are listed 'Best First'.
Re: Re: Copying a hash to another hash problem by Anonymous Monk on Oct 22, 2002 at 14:42 UTC
Thankyou all for your constructive replies. I can see my obvious mistake that the copy of one hash to another is fundamentally flawed in my code and has been pointed out, i'm using numeric values for both keys and values, so I will rewrite the code using arrays. Again, thanks for the advice. Cheers. Sean.	[reply]
Re: Re: Re: Copying a hash to another hash problem by demerphq (Chancellor) on Oct 22, 2002 at 16:11 UTC
i'm using numeric values for both keys and values Well, this isnt intrinsically a bad thing. But more often than not it is. The reason I say this is because if your indexes are sparse, ie 1, 500,623,7439,104029 or the like, then a hash is a much more suitable container type than an array is. (A hash would store the above usin 5 slots in a hash, an array would store at minimum 104029 elements.) However if the array is dense, 1..9,11..19,21..29 etc then an array is a more suitable approach. (Each slot in a hash is bigger than in an array, so even with the 4 wasted slots in the array would still be much smaller than the 26 allocated slots in the hash.) Also, the difference isnt so much due to what you store in the values, but what you store in the keys. Good luck and dont forget strict. :-) --- demerphq my friends call me, usually because I'm late....	[reply]
Re: Re: Re: Re: Copying a hash to another hash problem by Anonymous Monk on Oct 23, 2002 at 07:41 UTC
Maybe I should have given an explanation of my requirements. Basically, I have a NAS setup with 700+ disks contained within 45 drawers split over 7 frames that I want to watch the temp on. The code queries the drawers one by one and maps which drawers are in which frames and collects the current temp(also producing a pictorial map of the system). So the array would only ever contain a max of (7) for the frames and (1-8) for the drawers within the frames (as not all frames have the same no. of drawers). The copying (or snapshot) of the array was for comparing against the next time the prog went to check the temp to inform me of differences. Cheers Sean	[reply]