- From: Gregory Williams <greg@evilfunhouse.com>
- Date: Thu, 9 Jun 2011 17:44:15 -0400
- To: Carlos Buil Aranda <cbuil@fi.upm.es>
- Cc: SPARQL Working Group <public-rdf-dawg@w3.org>
On May 23, 2011, at 8:30 AM, Carlos Buil Aranda wrote: > Dear all, > > I uploaded a set of tests for SERVICE. This is for ACTION-274. I'm assuming no one has run these tests yet because there are syntax errors in the manifest and in one of the result files, and as far as I can tell there's no way to tell from the manifest which qt:data file is meant to be available through which endpoint. The manifest's mf:entries lists 4 tests which don't have corresponding manifest information. Also, the tests are listed as dawgt:Approved, though I don't think we've actually approved them yet. Here's an example from the manifest: :b1 rdf:type mf:QueryEvaluationTest ; mf:name "" ; dawgt:approval dawgt:Approved ; mf:action [ qt:query <service01.rq> ; qt:data <data01.ttl> ; qt:data <data01endpoint.ttl> ] ; mf:result <service01.srx> . From reading the query and expected results, I can tell that the RDF in <data01.ttl> is meant to be loaded into the local dataset, and <data01endpoint.ttl> is meant to be "available" via the endpoint <http://example.org/sparql>. I think we will need new manifest vocab terms to indicate this (or we might be able to repurpose some of the service description terms...). As a first attempt, I'll suggest something like this: :service1 rdf:type mf:QueryEvaluationTest ; mf:name "SERVICE test 1" ; dawgt:approval dawgt:NotClassified ; mf:requires mf:BasicFederation ; mf:action [ qt:query <service01.rq> ; qt:data <data01.ttl> ; qt:serviceData [ qt:endpoint <http://example.org/sparql> ; qt:data <data01endpoint.ttl> ] ] ; mf:result <service01.srx> . This allows associating qt:data (and qt:graphData) to a specific endpoint for a test. I've also added an mf:requires triple indicating that basic federation support is required, added an mf:name, and renamed the test IRI (if we're going to use IRIs and not blank nodes, my preference is for IRIs that are slightly more useful than "b1"). Thoughts? thanks, .greg
Received on Thursday, 9 June 2011 21:45:15 UTC