Hi Mitko,
On 10/2/07, Mitko Iliev <imi...@openlinksw.co.uk> wrote:
> I guess you didn't specified the graph(s) from where to fetch the data,
> For testing i suppose you put test data in some graph(s), then should
> use FROM e.g. "select * from <graph-iri> where { ?x ?y ?z }" .
Yes, in fact that is what I am doing.
The problem is that the Soprano Model unit tests expect the
model/triple store to be fully empty for each test case. So, the
virtrdf graph keeps interfering with it :)
There are also tests for select-ing triples with a context wildcard
(e.g., <given subject> ?p ?o ?g, or ?s <given predicate> ?o ?g>),
where the above problem is manifested.
I guess I will have to modify the unit tests to match the problem in
Virtuoso. Unless you can suggest some better ways ... ?
Right now, all the test "pass"; some are passed by the test case
(around 10-12) and rest (8 or so) 'fail' but inspecting the failure
reason we can see that in fact if it werent for the virtrdf context,
it will actually work.
Btw, thanks very much for your work on Virtuoso Triple Store! It is
really awesome and standards compliant.
Regards
Rajeev J Sebastian
