Re: ISSUE-126 (Revisit Datatypes): The list of normative datatypes should be revisited

On Jun 18, 2008, at 3:23 PM, Boris Motik wrote:

> 2. We define xsd:anyURI to be a subset of xsd:string.

I thought this isn't a problem, but now I think it might be. The  
issue is that the lexical to value space mapping for xsd:anyURI is  
different from that for xsd:string.

For xsd:anyURI, my understanding is that

1. http://neurocommons.org/page/Main%5FPage
2. http://neurocommons.org/page/Main_Page

Have the same value

However if interpreted as xsd:string these are different values.

The issues are:

If we have a pattern facet, is the pattern applied in the value space  
or the lexical space? If in the value space then this is really  
painful, as one has to do schema dependent unescaping.

If in the lexical space, then this is at variance with the approach  
that OWL operates on the values.

This also explains why xsd:anyURI isn't a subtype of xsd:string (even  
if the value space of xsd:anyURI is a subset of the value space of  
xsd:string).

-Alan

http://www.w3.org/TR/xmlschema-2/#anyURI
http://www.cs.tut.fi/~jkorpela/rfc/2396/full.html#2.4.2