The elements have the strange definition shown below:
  <![CDATA[   .....snip.....   ]]>


That isn't an element, it's a CDATA section and is invisible to an XSLT
processor. You just extract the data from the surrounding element in the
usual way.

CDATA is just an authoring convenience, within the marked region < acts
like &lt; and & acts like &amp; so if you have a large chunk of XML that
you want to quote as data rather than as part of the XML tree, you can
do

<x><![[CDATA[<p>zzz <span>dddd &nbsp; ...</span></p>]]></x>

but XSLT will see the same input as if you had gone

<x>&lt;p>zzz &lt;span>dddd &amp;nbsp; ...&lt;/span>&lt;/p></x>

David


________________________________________________________________________
This e-mail has been scanned for all viruses by Star. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________