Subject:XSLT/XPath Help Author:(Deleted User) Date:08 Dec 2008 02:52 PM
Hi Bruno,
whatever is inside a CDATA node, is a single text fragment; what CDATA does is to avoid the manual escaping of reserved characters, but in reality, <tag><![CDATA[<>]]></tag> is identical to <tag><></tag>.
So the only way you can extract a piece of that CDATA is if you use a custom XSLT function that parses the text fragment, on which you can then apply the XPath expression to locate it.