RE: Request for help with semantics of XML Schema Datatypes

Jeremy:
Some answers inline.

All the best, Ashok


-----Original Message-----
From: w3c-xml-query-wg-request@w3.org
[mailto:w3c-xml-query-wg-request@w3.org]On Behalf Of Jeremy Carroll
Sent: Monday, October 04, 2004 8:34 AM
To: w3c-xml-query-wg@w3.org; w3c-xsl-wg@w3.org; SWBPD list
Subject: Request for help with semantics of XML Schema Datatypes




Hello,

the Semantic Web Best Practices and Deployement WG has a task force 
clarifying the correct use of XML Schema Datatypes within the Semantic Web.

The task force description can be found here:

http://lists.w3.org/Archives/Public/public-swbp-wg/2004Sep/att-0100/

The XML Schema WG suggested that you may be able to help concerning the 
denotational semantics of the XML Schema datatypes.

 From the TF description, it is this issue:

[[
What is the relationship between the value spaces of the various XML 
Schema built-in simple types when used within RDF and OWL ?
]]

Key questions are for example:

"0"^^xsd:float the same as or different from "0"^^xsd:int
[AM] These values are distinct but compare equal.
In fact "-0"^^xsd:float also compares equal to "0"^^xsd:int

"3.2"^^xsd:float the same as or different from "3.2"^^xsd:int
[AM] There is no legal value such as "3.2"^^xsd:int.  Are you thinking
of some kind of truncation semantics?

"http://www.example.org/"^^xsd:string the same as or different from 
"http://www.example.org/"^^xsd:anyURI
[AM] This is currently under discussion.  I can let you know when
we have a firm position.

where the ^^ denotes the value arising from the lexical form on the left 
according to the datatype on the right.

I have looked at F&O and the XPath 2.0 WDs and believe that an answer 
offered by your WGs is the eq operator in XPath. (Modulo difficulties to 
do with NaN, and numeric approximation). This would suggest that all the 
numeric types have a shared value space of numbers, understood as in 
mathematics; whereas all the other datatypes are disjoint, and 
understood more as in a sorted type system (is that the right term?). I 
haven't yet understood the rules for dates under eq.
[AM] In XML Schema you cannot compare values from different datatypes
but the F&O defines ordering and equality semantics for some of them.

This message is to ask whether:
- any participants of either XSLT or XQuery WGs would like to 
participate in this Task Force
[AM] I'm not sure how much time I can commit to, but I can certainly answer questions.
- you would like me to attend one of your telecons to explain in more 
detail what are the concerns

BTW, the Semantic Web Best Practices and Deployment WG, will be having a 
face to face in Bristol on 1st and 2nd Nov. For this TF, the f2f goal is 
to agree an outline first note, outlining the issues and possible solutions.

Also note that the SWBPD's archive is public. So any member confidential 
reply should not be cc-ed to that WG.

Jeremy Carroll

Received on Monday, 4 October 2004 16:28:01 UTC