Re: Test cases, quick update.

> I'd like feedback from Jos, Pat, JJC in particular on the false document
> issue herein.
>
>
> I'm bringing the manifest up-to-date, but I'd like some quick feedback
> on a couple of issues (one of which involves the "always false document"
> notion).
>
>
> Firstly: despite wandering off somewhat and highlighting that we need to
> feed back to XML schema,
>
http://www.w3.org/2000/10/rdf-tests/rdfcore/datatypes/Manifest.rdf#semantic-equivalence-between-datatypes
> was not approved at the 2002-11-22 telecon. However, everyone seems
> convinced of its truth (including DanC, who we take to be our voice of
> the XMLSchema WG).
>
> That is, that
> http://www.w3.org/2000/10/rdf-tests/rdfcore/datatypes/test005a.nt
> [[
> <http://example.org/foo> <http://example.org/bar> "10"
^^<http://www.w3.org/2001/XMLSchema#integer> .
> ]]
> entails
> http://www.w3.org/2000/10/rdf-tests/rdfcore/datatypes/test005b.nt
> [[
> <http://example.org/foo> <http://example.org/bar> "10.0"
^^<http://www.w3.org/2001/XMLSchema#decimal> .
> ]]
>
>
> If everyone's happy with this as being noncontroversial, can I approve
> this?


I am happy with that one
we also have running code for that testcase


> Secondly: I was actioned to create a test case where a well-formed DT
> clashes with a range constraint. (Such a document comes out false
> according to the MT). This brings me to address the definition of
> entailment test cases; I'd like them (and so would the mathematicians
> present, I get the impression) to actually reflect _entailments_, not
> some approximation thereof.
>
> To represent the test case I've been tasked to write, I therefore need
> to add the notion of a "false document" to the entailment test case
> definition. Currently the premise and conclusion document declarations
> in an entailment test case can take one of the following two forms:
>
> [[
> <test:premiseDocument>
>   <test:NT-Document rdf:about
="http://www.w3.org/2000/10/rdf-tests/rdfcore/datatypes/test005a.nt" />
> </test:premiseDocument>
> ]]
>
> or
>
> [[
> <test:conclusionDocument>
>  <test:RDF-XML-Document rdf:about="http://example.org/blah.rdf" />
> </test:conclusionDocument>
> ]]
>
> I would like to add the following construct as a premise or conclusion
> "document":
>
> [[
> <test:conclusionDocument>
>   <test:False-Document />
> </test:conclusionDocument>
> ]]
>
> That is, a document that always "comes out false" under any
> interpretation. It's not constructible using "plain" RDF so I'd
> rather just add support in this format. Jos, can you add support for
> this?


yes, see below


> That being OK, I'm going to add the following test case:
>
> [[
> <ex:a> <ex:b> "25"^^xsd:decimal .
> <ex:b> rdfs:range xsd:string .
> ]]
>
> ENTAILS
>
> { a test:False-Document }
>


Jan, for the moment we've done that in
http://www.agfa.com/w3c/euler/xsd-rules
(but it's very incomplete...)

[[[
### inconsistency/incompleteness detections @@

{ :xsdI1 . ?p rdfs:range xsd:string . ?s ?p ?x^^xsd:decimal } log:implies
{ _:x a test:False-Document } .
]]]

and then indeed we can derive from above that

_:y a test:False-Document .


hope this helps...


>
> I'll take another pass over the DT test cases tomorrow; in particular,
> the ill-formed-literal test case can be simplified as it currently
> stands.


O great
BTw, could you temporarily remove/escape the double quotes
in "badly-formed" in /2000/10/rdf-tests/rdfcore/Manifest.rdf
(this is of course not strictly needed, but I am getting
as input corrupted triples from the Jena-1.6.0 N3 output)


> Cheers,
> jan
>
> --
> jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/
> Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/
> Random act of violence against bread: whole pint.
>   -- extract from the "Hawk the Slayer" drinking game

-- ,
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

Received on Wednesday, 11 December 2002 14:11:22 UTC