OBO::Parser::OBOParser?

Who "woulda thunk it!" The problem-domain breadth of the treasures to be found on CPAN continues to amaze me! Thanks, karlgoethebier, for bringing that module to light.

My, I didn't know about this one, but I am so grateful for your reply, thanks!!

Hi,
If I may add to what others have said here.
Please and please read up perldsc - Perl Data Structures Cookbook
You will be glad you did!

Thank you all for your replies, will check out that book and the CPAN module!

I don't quite get your problem but as far as i understood you want soimething like this:

use strict;
use Data::Dumper;

open(IN, "<", "in.obo") || die "$!";

my %hash;
my $term;

while(<IN>){
  chomp;
  if(/^id\:\s+(.*)/){
    $term = $1;
  }elsif(/^is\_a\:\s+.*?\s+\!\s+(.*)/){
    push(@{$hash{$term}},$1);
  }
}
close IN;
print Dumper(\%hash);
[download]

PS: code not tested

Thank you, I will test this code as well as the CPAN module, which seems like the easiest way to do it. Best!

I am not aware of the OBO data set , but i was just checking the data set and thought i can share some technical inputs on it. Your requirement was clear enough to understand in that you wish to build a data structure like a hash of arrays, or any such data structure from which you can readily extract your data based on the id key.
However as i mentioned your data set has some interesting features. I hope the OBO Parser solves your problems, coz if you had to work with such data sets and write code from scratch it is hardly extensible.
But in any case, i was able to find a way to structure your dataset using the space delimiter and the '!' character delimiter. Actually if you consider it , it might not be a great approach, but again that is your data set :D
If you were to practically wish to achieve this you would need a thorough understanding of the data structures in Perl. But not to worry, you can read the docs and figure it out . perldata, perlreftut to begin with.
The key feature i would say i found that was needed for me to make an extraction , was the use of anonymous arrays and references. Now even though you might say the code works, i can only surmise it is hardly extensible in case your requirement changes, and if you were asked to analyse a dataset of a million or so records, i think it is best you have someone use the standard module (like OBO::Parser )
Note - I created a file of your data in the OP and passed it as an argument to this script below

#!/usr/bin/perl
use strict;

my (%hash,$hash_id);
my $isa_array_ref;
open(my $fh,"<",$ARGV[0])
  ||  die "$0: can't open $ARGV[0] for reading: $!";
LINE: while(<$fh>){
  chomp($_);
  next LINE if ($_ eq "Term");

#split on first blank space
  my @TermRow = split(/ /,$_,2);
  if($TermRow[0] eq 'id:'){
    $hash_id = $TermRow[1];
    $isa_array_ref = undef;
  }
  elsif($TermRow[0] eq 'is_a:'){
      my @TermISAText = split(/!/,$TermRow[1]);

#checking if anonymous array reference already exists
      if($isa_array_ref){
         my @temp_array = @{$isa_array_ref};
         push(@temp_array,$TermISAText[1]);
         $isa_array_ref = \@temp_array;
          $hash{$hash_id} = $isa_array_ref;
        }   
       else{

#creating an anonymous array reference
          $isa_array_ref= [$TermISAText[1]];
          $hash{$hash_id} = $isa_array_ref;
        }
  }

}
close($fh);

print "Result of Extraction:\n ";

my @id_keys = keys %hash;
  foreach(@id_keys){
    print "key : $_";
    print "list of values \n";
    foreach(@{$hash{$_}}){
      print $_,"\n";
    }
    print "\n";
  }
[download]

XXXXXX:progs$ perl term_reader.pl ./term.txt
Result of Extraction:
 key : HP:0000008list of values 
 Abnormal internal genitalia
 Abnormality of the female genitalia

key : HP:0000007list of values 
 Mode of inheritance
[download]

In situations like this, let the use of the data structure dictate its design. How will you want to access this data?