RE: Getting a list of node Text()

> Yes, I agree.  The source I have no control over, other than 
> to submit error reports and hope for the best.  It's coming 
> out of an app where I don't control the output.
> 
> But having investigated this I came to the same conclusion, 
> and I'll submit some reports about it and see what happens.

Meanwhile you could write a SAX filter to preprocess the document, since SAX
applications (unlike XSLT) can see CDATA boundaries.

Michael Kay
http://www.saxonica.com/