Chris,

Thanks, and I agree, I also use Xselerator. It's the clients who are messing
about with the XML messages that are into spy... probably cus it's easier to
view the message (??)

Dan.


"Chris Barber" <chris@b...> wrote in message
news:%23j9z%23THAEHA.3456@T......
XMLSpy has a fairly long history of changing XSLT, XML, and sometimes
getting it wrong. Although it has a comprehensive suite of functionality I
have generally found it faster and easier to work with Marrowsoft Xselerator
since it uses the 'actual' XSLT engine that you are targeting and it *never*
makes unrequested changes to your source XML or XSL. It also seems to result
in a better understanding of XSLT principles and methods that XMLSpy may be
hiding from you by being too 'user' oriented as opposed to 'developer'
oriented.
It also provides step-through debugging which I'm not 100% sure that XMLSpy
does (please correct me if I'm wrong).

Don't get me wrong, XMLSpy">XMLSpy is a good product and certainly has its staunch
supporters. However, some of the best XSLT experts here seem to prefer
Xselerator - I think that indicates something.

Chris.

"Daniel Bass" <DanielBass TAKE at OUT CAPS WORDS Postmaster.co.uk> wrote in
message news:exMZghFAEHA.1032@T......

> AFAIK, XMLSpy">XMLSpy has bad habit of "adjusting" XML instance, when
> associating it with XSD Schema. If the schema has target namespace, but
> XML instance doesn't, XMLSpy">XMLSpy "fixes" XML instance.
> To avoid it, remove target namespace from your schemas. Your schema
> doesn't fit your XML. If your XML doesn't have namespace, why its schema
> has?

Hmm, good question. My clients provided me with a message (without xmlns),
and an XSD, and it's only when you "add" the schema to the XML via XMLSpy">XMLSpy
that you see that it is in the correct syntax. Only thing is it's not in the
correct syntax because spy fixes it for you!

i have a feeling they're creating it all in spy, and spy's filling in some
blanks i don't want to see filled. Fantastic.

>> I tried the local-name() method but it didn't work, the mapping appeared
to
>> return a match, but then none of the data that is scanned within the
>> template returns valid branches...
>You have also to fix *all* XPath expressions in your stylesheet! Instead
>of "a/b" - *[local-name()='a']/*[local-name()='b']. Crazy indeed.

You're kidding, that's not crazy, that's plain insane!

>> Oh, it's so confuzzling. **8-$
>Agree. Namespaces are probably the most confusing stuff in the whole XML :(

Now he tells me, DOH! ;o)
thanks for your help.
Dan.