Re: Test cases - ARP report.

>>>Jeremy Carroll said:
> Running ARP against the zip
> 
> all_20021213.zip

(I'm using the CVS/web which might be different)

> rdfms-nested-bagIDs 008 old nt file is correct, new one is wrong.

The current
  http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-nested-bagIDs/test008.nt
CVS 1.3, Jan having just changed it is not correct.

It uses rdf:ID to identify the object of the statement of the
property element some:prop.  This is incorrect, rdf:ID on a property
element always identifies the reified node.  A blank node is
generated in such cases, when there are property attributes.

Firstly, the some:prop property element is about the blank node from
the rdf:Description and has as an object, the generated blank node,
let's call it _:j2

So change:
  _:j0 <http://example.org/prop> <http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-nested-bagIDs/test008.rdf#aReification> .
to
  _:j0 <http://example.org/prop> _:j2 .

The reification of the above triple is missing (+4 more triples)
since it is pointed at from the bag.  Let's call that node _:j4.

The pointer from the bag is wrong:
  <http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-nested-bagIDs/test008.rdf#bag> <http://www.w3.org/1999/02/22-rdf-syntax-ns#_1> _:j3 .
should point at the reification of the above
  <http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-nested-bagIDs/test008.rdf#bag> <http://www.w3.org/1999/02/22-rdf-syntax-ns#_1> _:j4 .


The following other changes are needed:

The statement from the property attribute hangs off the blank node (_:j2)
not the 'aReification' node.

from
  <http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-nested-bagIDs/test008.rdf#aReification> <http://example.org/propAttr> "val" .
to
  _:j2 <http://example.org/propAttr> "val" .

the corresponding reification of that triple in the bag 
  _:j3 <http://www.w3.org/1999/02/22-rdf-syntax-ns#subject> <http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-nested-bagIDs/test008.rdf#aReification> .
to
  _:j3 <http://www.w3.org/1999/02/22-rdf-syntax-ns#subject> _:j2 .

<snip/>

i.e in summary, we should revert to CVS 1.2 which has all these
changes and I previously reported, works fine.

Dave

Received on Monday, 16 December 2002 13:11:38 UTC