Attributes, namespace partitions, and schemas

[Ray Waldin]

OK, I think I've answered my own questions here by re-reading the spec a few
times and digging through a ton of posts on this subject from earlier this year.
A little confirmation would help tho...

> Given the following XML:
> 
> <a xmlns="uri1" xmlns:b="uri1" c="d" b:c="e"/>
...
> Is it possible for a schema to indicate that the above is allowed or
> disallowed?

Yes, the unconstrained wildcard attribute specification <anyAttribute/> allows
this, but there is no way to explicitly allow a particular global attribute yet. 

> Should SAX2 report two attributes with the same localName and namespaceURI?

No. The per-element attribute (c) should have a null namespaceUri and the global
attribute (b:c) should have a namespaceUri of "uri1".

> Which value for c should an application honor if it encounters such a case?

Depends on whether the application is looking for a per-element attribute or a
global attribute.  Maybe both.

So the bottom line is, there are two "classes" of attribute: global and
per-element.  Global attributes and per-element attributes may belong to the
same namespace indirectly, but never share namespaceUris directly as per-element
attributes don't really have namespaceUris of their own.  

This also explains why "default namespaces do not apply directly to attributes".
Unqualified attributes *always* belong *indirectly* to whatever namespace their
parent element belongs to, so they can never *directly* belong to the default
(unqualified) namespace as global attributes.  

Is this even close to correct?  TIA...

-Ray

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)