Re: Wrong test-case: rdfms-not-id-and-resource-attr/test003

>>>Jan Wielemaker said:
> I'm updating the SWI-Prolog RDF parser with the latest RDF revision
> (nodeID support), testing them on the test-cases I received from Jeremy
> Carroll (a snapshot of the jena CVS I believe).

A snapshot from Jeremy?  Could you please try them with the official
RDF Core test cases from the web site:
  http://www.w3.org/2000/10/rdf-tests/rdfcore/

However...

> 
> The test-case identified as
> 
> http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-not-id-and-resource-attr/Manifest.rdf#test003
> 
> seems to be wrong.  The test says:
> 
> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>          xmlns:eg="http://example.org/">
> 
>   <rdf:Description>
>     <eg:prop1  rdf:ID="reify" rdf:bagID="bag"/>
>   </rdf:Description>
> </rdf:RDF>
> 
> In the test-result this returns an blank node with `prop1' property
> and object "" (empty literal) and a reification thereof where the value
> is another blank node.  With a reified statement that differs from the
> real statement, this can't be right.
> 
> I think the specs says it should be two blank nodes related using the
> prop1 property and properly reified.

I actually noticed this problem when I was testing them myself last
Friday.  The property element eg:prop1 has an empty literal value,
not a blank node, and that is the statement that is reified with
ID "reify" (http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-not-id-and-resource-attr/test003.rdf#reify)
and associated with the bag ID "bag" (http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-not-id-and-resource-attr/test003.rdf#bag).

The fix I've approved for
  http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-not-id-and-resource-attr/test003.nt
is:

19c19
< <http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-not-id-and-resource-attr/test003.rdf#reify> <http://www.w3.org/1999/02/22-rdf-syntax-ns#object> _:j88097 .
---
> <http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-not-id-and-resource-attr/test003.rdf#reify> <http://www.w3.org/1999/02/22-rdf-syntax-ns#object> "" .


Can you confirm this fixes it for you?

> 
> 	Regards --- Jan
> 
> P.s.	Using SWI-Prolog from the CVS snapshot it processes the rest of
> 	the test-suite properly now.

Excellent!

Dave