RE: Is "Expanded QName" underspecified?

> I'm trying to prototype the xf:node-name function. Problem 
> is, I really 
> can't tell what the behavior of the returned value should be. It's 
> possible I'm not looking at the most recent documents, but:
> 
> 
> Operators says this returns an expanded QName and points to 
> the Data Model 
> spec.
> 
> Data Model says "An expanded QName is in the value space of 
> xs:QName, and 
> consists of a namespace URI and a local name". But that 
> doesn't tell me 
> how this value behaves when I actually try to operate on it 
> -- to take one 
> trivial example, it doesn't tell me what the lexical 
> representation of an 
> _expanded_ QName should be.

An expanded QName is a (URI, localname) pair. 

The current situation is that there is no way of converting an expanded
QName to a string (see
http://www.w3.org/TR/xquery-operators/#casting-from-primitive-to-primitive)
and therefore there is no way of writing a QName to the result tree.

The operations you can perform on an expanded QName are to extract the
namespace URI and the local name. You can also compare two expanded QNames
for equality.

This isn't very satisfactory, but it's the best we could do for the time
being.

Michael Kay
Software AG




> 
> You've got an open issue (Issue-0063: Is prefix preserved?) 
> which suggests 
> that the non-expanded lexical value -- the Namespaces-style 
> qualified name 
> -- _might_ be available. But it isn't clear that this is the intended 
> lexical value under all cases, or which cases it applies 
> to... and the 
> issue itself indicates that you hadn't decided where the 
> prefix would be 
> taken from.
> 
> Checked the XSLT2 spec; didn't find any help there.
> 
> 
> Is there really a hole in the spec where someone made an 
> assumption, or 
> did I miss something? 
> 
> ______________________________________
> Joe Kesselman  / IBM Research
> 
> 

Received on Monday, 26 August 2002 13:20:18 UTC