[jsdl-wg] Version 1.0 of the XSD Schema for JSDL

Donal K. Fellows donal.k.fellows at manchester.ac.uk
Wed Nov 24 03:52:13 CST 2004


A Stephen McGough wrote:
> What is the difference between xsd:QName and xsd:NMTOKEN? We've changed 
> things over to xsd:NMTOKEN. It's easy to change to xsd:QName but I think 
> we need to know why.

QName is syntactically a subset of NMTOKEN, and semantically subject to
the same XML namespace rules as the elements and attributes of the document.

> We have them in now - so no more work (for me anyway). If we've missed 
> any could people send the ones missed to the list. To make my life easy 
> could you send the items needed to be added and not references to 
> documents!

I'm not sure that it's a very useful distinction though. In terms of
what is likely to be useful, it's probably better to stick to the
general instruction set classes that we've got. The problem is really
that there is a partial order on the potential values for PAE but we
don't really have the time to fix that (and nor is it a closed set). Yuck.

>>   O.S.T.E. is the defined set of string values that correspond to the
>>   defined CIM_OperatingSystem.OSType values (*not* the numeric forms).
> 
> Ditto.

That's what I sent to Darren at the meeting. :^) OTOH, this is probably
another case where my comments just above apply; e.g. for many apps,
Win32 is just Win32 and the distinction between variants is irrelevant,
but the JSDL consumer is a very specific type of OS (well, unless it is
some kind of higher-level brokering/scheduling thing of course.)

>> Also, rangeValue could derive from xsd:token and not just xsd:string so
>> we could pick up the normalization rules.
> 
> Hmm, but then people couldn't use spaces - is this going to cause people 
> problems? I tend to write lists "2, 5, 6". Could easily ditch the spaces 
> thoug. Also can you have commas in xsd:tokens? Thoughts please folks.

According to <URL:http://www.w3.org/TR/xmlschema-2/#token> commas are
fine in token and non-standard whitespace is collapsed, so that allows
the RE for ranges to become:
RE:    (\d+( ?- ?\d*)?|- ?\d+)( ?, ?(\d+( ?- ?\d*)?|- ?\d+))*
         ^^^^^^^^^^^^^^ ^^^^^^  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Expl:   D or D- or D-D  or -D   or a comma-sep list of them

(Yes, the \d is allowed in XSD; I looked it up specifically.
<URL:http://www.w3.org/TR/xmlschema-2/#dt-regex> The XSD rules for REs
in Appendix F are among the clearest RE syntax descriptions I've seen.)

Donal.





More information about the jsdl-wg mailing list