hash of a hash

[cckramer]

I have this info' in a file.

blockA/input1 to block A/output1 delayXX
blockA/input1 to blockA/output2 delayYY
blockA/input1 to blockA/output3 delayZZ

blockB/input1 to blockB/output1 delayPP

blockC/input1 to blockC/output1 delayQQ
blockC/input2 to blockC/output1 delayRR

how to store the inputs and outputs for each block as shown above ? in 2 hashes or arrays ? I am not interested in the delay shown.I just need to capture the possible combination of inputs and outputs for each block.

[nithinpes]

You can probably create a hash with block names as keys. Let the value for each key be an anonymous hash in which inputs are keys and the possible outputs for each input(stored in anonymous array) as values.
You end up in the following structure which is 'hash of hash of an array'.

%myhash = ( 'blockA' => {'input1'=>['output1','output2']},
                     'blockB' => {'input1'=>['output1']},
                      'blockC' => {'input1'=>['output1'], 'input2'=>['output1']}   ) ;
 

[numberwhun]

Being the late hour that it is I just don't have it in me (mentally) to produce full code at the moment, but what I can say is that you can first use a regular expression to parse out of each line what you are needing as far as text. Then, store it in the hash, as such:

my $text = "blockA/input1 to blockA/output1 delayXX";
my $hash;
 
if($text ~= /^(\w)+\/(\w)+\s+\w+\s+(\w)+\/(\w)+\s+(\w)+)
$hash{$1}{$2} = $5;
 

There are a few different ways to handle hashes of hashes, but at work we tend to use something similar to this and it works great. Granted, at this late hour I may have something incorrect in the assignment and also, it is untested, but you should get the idea.

Also, search google for hash of hashes and you may find some good links.

Regards,

Jeff

[cckramer]

Thanks. I shall give it a shot..also I tried this rough code.Is this okay ?

# check if line starts with "delay"
 if ($line =~ /^delay/) {
 
# regular expression to get the inputs & outputs
   $line =~ m/\-from\s+\[\s+(.*)\]\s+\-to\s+\[\s+(.*)\]/;
 
# inputs & block information
   $in = $1;
   @in1 = split(/\//, $in);
   $blockin = $in1[0];
   $inport = $in1[1];
 
# outputs & block information
 
   $out = $2;
   @out1 = split(/\//, $out);
   $blockout = $out1[0];
   $outport = $out1[1];
 
# construct Hash of Hash of Hash function
   $HoHoH{$blockin}{$inport}{$outport} = 1;
}
}
 
 
# check if Hash of Hash of Hash has been correctly stored
 
# $blocks gives first level
for my $blocks (keys %HoHoH) {
 
# inputs gives the second level
  for my $inputs (keys %{$HoHoH{$blocks}}) {
 
# outputs gives the third level
   for my $outputs (keys %{$HoHoH{$blocks}{$inputs}}) {
   print "^^ $blocks/$inputs -> $blocks/$outputs ^^\n";
}
}
}
 

[numberwhun]

You need to please use code tags when posting code. If you notice, the rest of us use them and when you don't, we have to clean up behind you.

This is your nice warning.

Regards,

Jeff

[cckramer]

Thank you.I shall keep that in mind for the future

[KevinADC]

@cckramer

Hard to say because the data your code is parsing is not the same as the sample data you posted. On the face of it, the code looks OK to me. But I think a hash of arrays might work better. Whats the actual data look like you are parsing?

[cckramer]

Thank for all those who helped.My code that I posted earlier to create a hash of hash of hash worked.

(keys{$ blocks{$inputs}{$outputs}})