First, XMLSpy's ability to handle complex XSDs has *many* bugs in it (don't ask me to list them, and IAC it will matter which version of XMLSpy you happen to have, but they all have some bugs).

The fact that XMLSpy doesn't think this document is valid against Adobe's schema is, basically, meaningless. You may want to another validator when XMLSpy complains.

Second, this document is well-formed, so whether it is valid against whatever version of Adobe's schema they put in its header is immaterial, from Fox's POV.

Don, how would you like the result to look in the cursor? Basically, design the cursor you want as the target and show us a CREATE TABLE statement or something. From there, this is doable.

José, personally I don't much care whether we use XMLTOCURSOR or XMLAdapter here. XMLTOCURSOR uses a simpler format, so I'd go for that, unless Don indicates that the documents are very large, in which case XMLAdapter is definitely a better choice.

Don, you may want to post to me e-mail, I don't read here very much, but you will be welcome to re-post my solution on UT.

>L<