XML::Complicated

[hacker]

In the process of working on another problem, I found the 'keyattr' bit of XML::Simple, and the FAQ that talks about 'array folding'. I cooked up this small snippet, which causes XML::Simple to throw an Out of memory! error. If I set 'forcearray' to 0, disabling it (against the recommended behavior), the error goes away.

What I want to know is.. why?

Obligatory Code

use strict;
use warnings;
use XML::Simple;
use Data::Dumper;

my $xml = XMLin(\*DATA,         keyattr    => ['url'],
                                forcearray => '1');   

print Dumper($xml);
print "URL: $xml->{bar}->{blort}->{quux}->{url}\n";

__DATA__
<?xml version="1.0" encoding="iso-8859-2"?>
<foo>
  <bar>
    <blort>
      <quux>
        <url>http://f.co/index.html</url>
      </quux>
    </blort> 
  </bar>
</foo>

[Chmrr]

This is a bug in Perl relating to pseudohashes, and is fixed in 5.9.0-to-be. There, it gives the more edifying message "Not a HASH reference at ..."

perl -pe '"I lo*`+$^X$\"$]!$/"=~m%(.*)%s;$_=$1;y^`+*^e v^#$&V"+@( NO CARRIER'

[grantm]

I cooked up this small snippet, which causes XML::Simple to throw an Out of memory!

No, Perl stops with an out of memory error when you treat an arrayref as a hashref. The print Dumper($xml) line output clearly shows that if you turn on forcearray, you get arrays.

[jeffa]

The reason i used the keyattr in my previous reply was due to a strange quirk within XML::Simple. Consider this example:

use XML::Simple;
use Data::Dumper;

my $xml = XMLin(\*DATA);
print Dumper $xml;

__DATA__
<documents>
   <document>
      <stayonhost>1</stayonhost>
      <staybelow>1</staybelow>
      <maxdepth>2</maxdepth>
      <name>Swedish Foo</name>
      <home_url>http://www.foo.se/bar/</home_url>
      <category>International</category>
   </document>
   <document>
      <stayonhost>0</stayonhost>
      <staybelow>0</staybelow>
      <maxdepth>3</maxdepth>
      <name>Swedish Bar</name>
      <home_url>http://www.bar.se/baz/</home_url>
      <category>National</category>
   </document>
</documents>

and it's output:

$VAR1 = {
  'document' => {
    'Swedish Foo' => {
      'category' => 'International',
      'maxdepth' => '2',
      'stayonhost' => '1',
      'staybelow' => '1',
      'home_url' => 'http://www.foo.se/bar/'
    },
    'Swedish Bar' => {
      'category' => 'National',
      'maxdepth' => '3',
      'stayonhost' => '0',
      'staybelow' => '0',
      'home_url' => 'http://www.bar.se/baz/'
    }
  }
};

What happened to the name element? I don't know for sure, but i am guessing that XML::Simple treats elements named name differently.* So, i read some docs, figured that keyattr might fix the problem and it did. But first i tried forcearray, which didn't fix the problem. I didn't think it would, but i didn't have a spare chicken to sacrifice at the time. ;) In case you are wondering ... forcearray is useful for keeping XML elements in order, remember that hashes lose order. forcearray is also useful for keeping your code consistent - you can always (99.99%?) expect an array, no guessing with ref needed.

I guess my point is that XML::Simple can handle tasks that are really beyond its scope, and you should be wary of that. It's great for getting stuff done quick, but you probably should pick a more industrial strength module for serious production code - like XML::Twig.

* yep ... here's what the docs say:

   The default value for 'keyattr' is ['name', 'key',
   'id'].  Setting this option to an empty list will
   disable the array folding feature.

jeffa

L-LL-L--L-LL-L--L-LL-L--
-R--R-RR-R--R-RR-R--R-RR
B--B--B--B--B--B--B--B--
H---H---H---H---H---H---
(the triplet paradiddle with high-hat)

[Aristotle]

Just posting to second the notion to use XML::Twig. I don't even bother with XML::Simple these days.

Makeshifts last the longest.