Fetching cdata content of tags with Mojo::DOM?

[Charlie Brady]

I have some translation files which are XML documents, which contain some
cdata elements:

...
<entry>
<base>ACTIVATE</base>
<trans>Activate</trans>
</entry>
<entry>
<base>UNREG_DESC</base>
<trans>
<![CDATA[
To obtain more information about blah, please visit our website
<A HREF="http://www.domain.com/" TARGET=_top>
http://www.domain.com/</A>
]]>
</trans>
</entry>
...

I'm trying to turn this into a hash from which I can look up the
translations:

my %lexicon = map
{ $entry->base->content => $entry->trans->content}
$dom->find('lexicon entry')->each;

but then I find that $lexicon{UNREG_DESC} gives me '<![CDATA[ ...' when I
want 'To obtain more information ...'. I think I need to initialise a new
Mojo::DOM object with ''<![CDATA[ ...' but I can't figure out what to do
next. Hints please.

[Charlie Brady]

On Mon, 21 Jul 2014, Charlie Brady wrote:

> I'm trying to turn this into a hash from which I can look up the
> translations:
>
> my %lexicon = map
> { $entry->base->content => $entry->trans->content}
> $dom->find('lexicon entry')->each;
>
> but then I find that $lexicon{UNREG_DESC} gives me '<![CDATA[ ...' when I
> want 'To obtain more information ...'. I think I need to initialise a new
> Mojo::DOM object with ''<![CDATA[ ...' but I can't figure out what to do
> next. Hints please.

I can give my own hint:

http://mojolicio.us/perldoc/Mojo/DOM#contents

So I can do:

my %lexicon = map
{ $entry->base->content => $entry->trans->contents->first->content}

[Charlie Brady]

On Mon, 21 Jul 2014, Charlie Brady wrote:

> So I can do:

Without the stupid c&p error:

my %lexicon = map
{ my $entry = $_ ;

[Charlie Brady]

No, that doesn't work for me, because of whitespace.

cf.

$dom->parse('<!-- test --><b>123</b>')

and:

$dom->parse(' <!-- test --><b>123</b>')

Can anybody suggest anything, should of using s/// and parsing a new
Mojo::DOM object? I can see Mojo::DOM::text, but dont see how I could use
it.

[Charlie Brady]

Sorry, everyone. I need more sleep. That should say:

.. short of using s/// and parsing a new Mojo::DOM object.

I have my code working that way now, but it seems excessive to parse those
fragments all over again.

[sri]

It's nice that you want to share your solutions, but please don't spam the list.

--

sebastian

[Charlie Brady]

On Mon, 21 Jul 2014, sri wrote:

> It's nice that you want to share your solutions, but please don't spam the
> list.

Sorry if you thought I did that. I thought I was asking for help. I'd
still like some help if someone is able to provide it.

[sri]

> It's nice that you want to share your solutions, but please don't spam the
> list.

Sorry if you thought I did that.

I tried to be nice about the warning, but your passive aggressive response makes me believe that you're just not taking it serious. Therefore i'm afraid i'm forced to temporarily ban you from this list. In the future, please be nice when participating here.

--

sebastian