RE: Extending "choice" model group in a xsd schema

Michael Kay wrote:
> 
> 
> Derivation by extension allows you to create a type that allows the
> original
> content, with new content added at the end. It doesn't allow any other
> modification that produces content models which the original type won't
> accept.
> 

That's exactly the situation i am into, now. I am trying to extend an
already published (3rd party) xsd. And one of the difficulties is trying to
extend it in such a way that the new order makes sense. Not easy to achieve
for the reason you mentioned.

To overcome this, i was thinking of trying to do a "redefine" of the
element. But then again, apart from other rules, redefinition needs to be an
"extension" (or restriction) of the original. 

Later down this mail discussion, there's a suggestion of using named model
group for creating a composite element instead of going for inheritance. But
that would require the original xsd to be composed of (reusable) named model
groups.
 
-- 
View this message in context: http://old.nabble.com/Extending-%22choice%22-model-group-in-a-xsd-schema-tp26967149p27141856.html
Sent from the w3.org - xmlschema-dev mailing list archive at Nabble.com.

Received on Wednesday, 13 January 2010 09:25:26 UTC