RE: namespaces support decision table

I don’t think you understood my question(s) but thanks anyway.

Shlomo.

Your questions seem to suggest a lot of confusion, and I don't think that can be solved by answering them directly.

An element (instance) has a name (the "expanded name") that's in two parts: namesapce URI and local name. The local name is as written in the source, the namespace URI is obtained by looking for the namespace binding of its prefix and/or the default namespace.

A schema defines rules for elements in terms of their expanded name. When you use an element name in name="x" its local name is x and it's namespace URI is either null or the target namespace depending on the values of elementForm and elementFormDefault. If you use an element name in ref="[p:]x" then its local name is x and its namespace URI is obtained by looking for the namespace binding of p or the default namespace.

If the expanded name in the schema is different from the expanded name in the instance (whether it's the local names that differ or the URIs) then they're simply talking about different elements.

Michael Kay

http://www.saxonica.com/

 

---

From: Shlomo Yona [mailto:S.Yona@F...]
Sent: 30 May 2007 06:33
To: xml-dev@l...
Subject: namespaces support decision table

Hello,

 

I am trying to understand the impact of schema hints and evidence from XML documents on the proper handling of namespaces. Here is what I came up with

 

a. Name is qualified in XML Schema: yes/no

b. Name is prefixed in XML document: yes/no (default namespace is also considered as a prefix, the empty string prefix, for this discussion’s sake)

c. Prefix for name is defined in XML document: yes/no

 

There are 8 possibilities for the above criterions.

 

An error should be issued (not necessarily same error in all cases) in all cases except for the following three:

 

a==b==c==0

a==1 b==0 c==1

a==1 b==1 c==1

 

My question is, are these indeed all possibilities? Are there other factors that need to be considered? Is my understanding correct?

 

Thanks.

 

Shlomo.