Re: [closed] issue #pfps-07 typed literals with lang tags

From: pat hayes <phayes@ai.uwf.edu>
Subject: Re: [closed] issue #pfps-07 typed literals with lang tags
Date: Thu, 1 May 2003 10:07:55 -0500

> >From: pat hayes <phayes@ai.uwf.edu>
> >Subject: [closed] issue #pfps-07 typed literals with lang tags
> >Date: Tue, 29 Apr 2003 16:06:56 -0500
> >
> >>  Peter,
> >>
> >>  In
> >>
> >>    http://lists.w3.org/Archives/Public/www-rdf-comments/2003JanMar/0088.html
> >>
> >>  you raised a last call comment on the RDFCore WD's which was recorded as:
> >>
> >>  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-07
> >>
> >>  The RDFCore WG has accepted your comment:
> >>
> >>    http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Apr/0361.html
> >>
> >>  and the wording of the current editor's draft of the semantics document
> >>
> >  > http://www.w3.org/2001/sw/RDFCore/TR/WD-rdf-mt-20030117/
> >>
> >>  incorporates conditions and rules for the case you describe.
> >>
> >>  Please reply to this message, copying www-rdf-comments@w3.org,
> >>  indicating whether this decision is acceptable.
> >>
> >>  Pat Hayes
> >
> >This message is extraordinarily cryptic. 
> >
> >I would appreciate some indication of just what has been done to address
> >whatever concern I had here.
> 
> OK.
> 
> The omission of this case (non-XML typed literals containing lang 
> tags) from the last call document was a editorial mistake. Putting 
> this case back in has necessitated making changes to the semantic 
> conditions and to the closure rules to cover this case.
> 
> The semantic conditions have been altered so that lang tags  in 
> non-XML typed literals are required to not influence the 
> interpretation of the literal, ie they are effectively ignored by the 
> semantics.  (Section 3.1, 
> http://www.w3.org/2001/sw/RDFCore/TR/WD-rdf-mt-20030117/#rdfinterp)
> 
> The closure rules (see 
> http://www.w3.org/2001/sw/RDFCore/TR/WD-rdf-mt-20030117/#rdf_entail ) 
> now include two rules which sanction the removal, insertion or 
> replacement of lang tags in any non-XML typed literal. There is a 
> comment in the subsequent text to the effect that it would be 
> computationally more efficient to simply strip out lang tags from all 
> non-XML literals as a syntactic normalization.
> 
> In both cases, these are stated as RDF conditions, so hold in all 
> semantic extensions which apply to the lang tag literal syntax.  The 
> conditions on datatypes are stated so that the lexical form to which 
> the datatype mapping is applied consists of the literal string, and 
> does not include the lang tag (in non-XML literals). This is in line 
> with the above rules, of course.
> 
> The case of lang tags in XML typed literals is covered by other rules 
> and conditions. This is somewhat more complicated and may still be 
> modified. I presume that your comment was not intended to apply to 
> this case as the last call document covered this case (inadequately, 
> but that was a different question which is still being addressed.)
> 
> Pat

These changes do provide a meaning for literals such as "13"@fr^^xsd:int.

However, the changes also serve to point out the differences between
rdf:XMLLiteral and all other datatypes.  I view this difference is as a
serious mistake.


peter

Received on Thursday, 1 May 2003 15:43:09 UTC