Hi all

I am trying to parse a multi-level XML document, and then adding certain nodes to a hash. The problem is that with the many different levels I am struggling with the referencing of the nodes that I want to add

This is the code I have

#!/usr/bin/perl

use XML::Simple;
use Data::Dumper;
use utf8;
open (INPUT, "<:utf8", "w.xml") or die "Can't open";

my $xml = new XML::Simple (KeyAttr=>[]);
my $afWN = $xml->XMLin("wnafr.xml");

open (OUTPUT, ">:utf8", "Output.txt") or die "Can't open";

print OUTPUT Dumper($W);

foreach my $e (@{$W->{XML}})
{
        print "I.D.:", $e->{ID}, "\n";
    print "Part of Speech: ", $e->{POS}, "\n";
        print "Literal: ", "$e->{SYNONYM}{LITERAL}{content}", "\n";
    print "\n";
}

close (OUTPUT);
[download]

The XML looks like this

<XML><ID>ENG20-00001740-a</ID><POS>a</POS><SYNONYM><LITERAL sense="1">in staat</LITERAL><WORD>in</WORD><WORD>staat</WORD></SYNONYM><ILR type="be_in_state">ENG20-05295659-n</ILR><ILR type="be_in_state">ENG20-04904666-n</ILR><ILR type="near_antonym">ENG20-00002062-a</ILR><DEF/><USAGE/><BCS>3</BCS><DOMAIN>quality</DOMAIN><SUMO type="=">Breathing</SUMO></SYNSET></XML>

<XML><ID>ENG20-00001740-v</ID><POS>v</POS><SYNONYM><LITERAL sense="1">asem</LITERAL><WORD>asem</WORD><LITERAL sense="1">respireer</LITERAL><WORD>respireer</WORD><LITERAL sense="1">asem skep</LITERAL><WORD>asem</WORD><WORD>skep</WORD><LITERAL sense="1">asemhaal</LITERAL><WORD>asemhaal</WORD></SYNONYM><ILR type="verb_group">ENG20-00002536-v</ILR><ILR type="verb_group">ENG20-00002307-v</ILR><ILR type="also_see">ENG20-00004923-v</ILR><ILR type="also_see">ENG20-00004127-v</ILR><ILR type="subevent">ENG20-00004923-v</ILR><ILR type="subevent">ENG20-00004127-v</ILR><DEF/><USAGE/><BCS>3</BCS><DOMAIN>medicine</DOMAIN><SUMO type="=">Breathing</SUMO></SYNSET></XML>

I want to put all the Literal->content in a separate hash, but the problem is that there can occur more that 1. If there is more than 1 Literal->content the XML parser stores them in a array of hashes. If there is only 1 Literal->content it is simply stores as a hash.

Any ideas?