Yeah sorry about this one - it was a private email that I tried to take back on-list but sent to the xsl-list instead of the Xalan list!

Anyway, the crux of the matter was that the user had a cdata section on the source xml:

<RDF>
  <description><![CDATA[<<<<MY TEST>>>>]]></description>
</RDF>

And wanted to get identical output from an identity transform.  To achieve this, you need to set the cdata-section-elements attribute on <xsl:output> which they did, but still ended up with:  

<RDF>
  <description>&lt;&lt;&lt;&lt;MY TEST&gt;&gt;&gt;&gt;</description>
</RDF>

...as if the cdata-section-elements had been ignored.  I think they were using XML Spy, so I suggested that they may be using a custom processor instead of Xalan, but I have no experience of XML Spy so Im just guessing.  

cheers
andrew

> 
> > <RDF>
> >   <description>&lt;&lt;&lt;&lt;MY TEST&gt;&gt;&gt;&gt;</description>
> > </RDF>
> > 
> 
> That is exactly equivalent to the version using CDATA marked section.
> So why is this output a problem? It should be the same kind 
> of thing as
> if attribute quotes change from a="b" to a='b' it is just a difference
> at the surface syntax, not in the underlying result tree.
> 
> Like all features of xsl:output,  cdata-section-elements="description"
> is only a hint to the serialiser, which may be ignored, and in
> particular if you pass the result tree directly as a tree to another
> process, then this XSLt process will not serialise the tree 
> at all, and
> so all these hints will not be used.
> 
> David
> 
 

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.467 / Virus Database: 266 - Release Date: 01/04/2003
 

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list