What have you tried?

Actually for something like that I might use XML::Parser instead. Try a super search and see what you can find, there's been a bunch of relevant stuff recently.

Some of the answers to trying to understand xml::twig and also trying to learn how to extract attribute ought to be of help. In fact with minor work the code in Re: trying to understand xml::twig and also trying to learn how to extract attribute pretty much oes what you want already.

I agree with GrandFather here. In that specific case you don't use at all the fact that XML::Twig gives you trees. You are interested in the tags as such. XML::Parser should be enough for that.

but suppose I am already using xml::twig .. is it common for script to also use xml::parser in one script?

Reread the links given in my previous reply and ask questions about that code if you need to. Ya ain't never gonna lern ifn ya don try!

---

Perl reduces RSI - it saves typing

use XML::Rules;

XML::Rules->new(
    stripspaces => 7,
    rules => {
        lease => sub {return 'lease' => $_[1]->{true}},
        extra_device => sub {return 'extra_device' => $_[1]->{value}},
        computer => sub {
            if ($_[1]->{extra_device}) {
                print "computer id: $_[1]->{id} type: $_[1]->{type} os
+: $_[1]->{os} lease:$_[1]->{lease} extra_device=$_[1]->{extra_device}
+\n";
            } else {
                print "computer id: $_[1]->{id} type: $_[1]->{type} os
+: $_[1]->{os} lease:$_[1]->{lease}\n";
            }
            return;
        }
    }
)->parse(\*DATA);

__DATA__
<config>
  <computer id="one" type="mac" os="XP" >
        <lease true="yes" />
        <extra_device value="scanner"/>
  </computer>
  <computer id="two" type="pc" os="NT" >
        <lease true="no" />
  </computer>
</config>
[download]

Sometimes it's better to reach for a different hammer. Even if you already have a hammer in your hand.