RE: [F&O] Casting to xs:NOTATION, xs:QName not clarified

According to Ashok's mail from 5 Nov 2004
is it "possible to construct xs:QName and xs:NOTATION from string 

If the casting table 
applies for both casts and constructors, then there is a contradiction
in the F&O spec.


Michael Rys wrote:
> The casting table applies to both forms:
> So if the casting tables says that you cannot cast from one type to
> another, it applies to both cast as Type and Type().
> Best regards
> Michael
>>-----Original Message-----
>>From: [mailto:public-qt-comments-
>>] On Behalf Of Frank Legler
>>Sent: Monday, January 17, 2005 10:24 AM
>>To: Laurens Holst
>>Subject: Re: [F&O] Casting to xs:NOTATION, xs:QName not clarified
>>At Saturday, 15 January 2005, Laurens Holst <lholst@students.cs.uu.
>>nl> wrote:
>>>Frank Legler wrote:
>>>>Hello Ashok,
>>>>Does that mean that it is possible to construct xs:QName and 
>>>>from xs:string,
>>>>but it is not possible to cast xs:QName and xs:NOTATION from a xs:
>>>From section 17.6 of XQuery 1.0 and XPath 2.0 Functions and Operators:

>>>"Casting from xs:string and xdt:untypedAtomic to xs:QName, xs:
>>>or types derived from them is not possible because it requires 
>>>of the namespace context. Constructor functions are available to
>>>construct values of these types provided that the lexical representation 

>>>is known statically."
>>I still do not understand why we differentiate between casting and
>>construction here.
>>Given the following examples
>>  (1) "a" cast as xs:QName   (: invalid according to the spec :)
>>    and
>>  (2) xs:QName("a")          (: valid according to the spec :)
>>Don't both examples require knowledge of the namespace context?
>>And why is a static evaluation possible for (2) but not for (1)?

Mail was send from
EASY and FREE access to your email anywhere:

Received on Tuesday, 18 January 2005 02:00:02 UTC