Hi again Peter and all,
I did eventually get the JenaSesame code working (I had forgotten to add VM
arguments to the JUnit test to increase the heap size), and was able to
list the statements in the remote Sesame repository that was connected to a
Jena model using JenaSesame.createModel(repositoryConnection).
However, when I added the Jena model connected to the OWLIM Sesame
repository as a submodel to the model passed into the SPINInferences.run()
method, and ran SPIN inferencing, there were no inferencing results.
Here is the relevant code:
****************************************
SPINModuleRegistry.get().init();
Model inputModel = ModelFactory.createDefaultModel();
OntModel spinModel = ModelFactory.createOntologyModel(
OntModelSpec.OWL_MEM );
//add system triples
spinModel.add(SystemTriples.getVocabularyModel());
//add SPIN rules
spinModel.read("http://localhost/tmp/rules.spin.rdf";);
//register locally defined functions
SPINModuleRegistry.get().registerAll(spinModel, null);
//connect to OWLIM repository
repositoryManager = new RemoteRepositoryManager("
http://localhost:8080/openrdf-sesame";);
try {
// Initialise the repository manager
repositoryManager.initialize();
// Get repository
repository = repositoryManager.getRepository("test");
// Open a connection to this repository
repositoryConnection = repository.getConnection();
//create the DatasetGraph instance
//dataset = new SesameDataset(repositoryConnection);
owlimModel = JenaSesame.createModel(repositoryConnection);
} catch (RepositoryException e) {
e.printStackTrace();
} catch (RepositoryConfigException e) {
e.printStackTrace();
}
//add owlim model
spinModel.addSubModel(owlimModel);
//add input model
spinModel.addSubModel(inputModel);
//create output model to contain inferred statements
Model outputModel = ModelFactory.createDefaultModel();
//add output model to SPIN model
spinModel.addSubModel(outputModel);
//run SPIN inferences
SPINInferences.run(spinModel, outputModel, null, null, false, null);
//get number of inferred statements
//System.out.println("Inferred triples: " + outputModel.size());
//get inferred statements
StmtIterator itr = outputModel.listStatements();
System.out.println("Iterating over statements");
while(itr.hasNext()){
Statement s = itr.next();
System.out.println(s.toString());
}
****************************************
So, I need another solution ... are there any working examples of using the
SPIN API to execute SPIN rules over a Jena model that is connected to a
remote Sesame repository? Presumably TopBraid Composer does something
similar in order to manage inferencing over large sets of triples, but is
there a way to reproduce this behaviour using the SPIN API?
Any suggestions are much appreciated!
Alison
On Thu, Feb 2, 2012 at 4:04 PM, Peter Ansell <[email protected]> wrote:
> HI Alison,
>
> The GraphRepository that is created by JenaSesame in the background is
> definitely a streaming wrapper, using iterators to access the results,
> so it is less likely that it is the wrapper causing the issue,
> although it could be.
>
> I don't think any of the current JenaSesame tests use listNameSpaces,
> so we will need to do some testing to see why that is happening. The
> nature of that method is such that Jena itself would need to store all
> of the results in memory to make sure the results set is distinct
> before emitting any results to free up memory. Have you tested the
> performance of some of the other Model methods? Does listSubjects
> (also distinct-based so may have the same issue) or listStatements
> (should stream results cleanly) give you results?
>
> Note that JenaSesame is setup to use jena-2.7.0-incubating, so it will
> not work with SPIN-API just yet per the focus on getting a jena-2.6.4
> based release out right now. I have done some experiments with getting
> SPIN-API to work with jena-2.7.0-incubating [1] to get around this but
> they haven't gone far just yet
>
> Cheers,
>
> Peter
>
> [1] https://github.com/ansell/spin/tree/feature/jena27upgrade
>
> On 3 February 2012 04:05, Alison Callahan <[email protected]>
> wrote:
> > Hi Peter,
> >
> > Thanks for your help. I have forked your JenaSesame repository and am
> trying
> > it out with my code.
> >
> > I have instantiated a Jena model from a remote OWLIM repository using
> > JenaSesame.createModel(repositoryConnection), and tried to do a simple
> test
> > using the model.listNameSpaces() method, with the result that after 3200
> > seconds I got a Java OutOfMemory error. I had given Eclipse 4GB to work
> > with.
> >
> > All that to say, can you tell me more about how the
> > JenaSesame.createModel(RepositoryConnection connection) method works?
> Does
> > it populate a model with all of the statements in a OWLIM Sesame
> repository,
> > or does it access them as needed?
> >
> > Thanks again,
> >
> > Alison
> >
> >
> > On Tue, Jan 31, 2012 at 6:19 PM, Peter Ansell <[email protected]>
> > wrote:
> >>
> >> Hi Allison,
> >>
> >> I am not sure if there is the could try the JenaSesame library from
> >> either my fork at https://github.com/ansell/JenaSesame or Andy
> >> Seaborne https://github.com/afs/JenaSesame . I haven't quite completed
> >> and tested the sesame-jena implementation but when I do I will send a
> >> pull request back to Andys fork, but you shouldn't need to use
> >> sesame-jena from what I can tell.
> >>
> >> The jena-sesame module which should be very similar in implementation
> >> between my fork and Andys, (as opposed to the sesame-jena module)
> >> enables the creation and use of either a Jena DatasetGraph or a Jena
> >> Model as a wrapper around an existing Sesame RepositoryConnection. It
> >> is very alpha software though, but all of the tests that exist pass.
> >> Feel free to fork it and send changes back if you find bugs.
> >>
> >> Hope that helps,
> >>
> >> Peter
> >>
> >> On 1 February 2012 08:17, Alison Callahan <[email protected]>
> >> wrote:
> >> > Hello all,
> >> >
> >> > I am using the SPIN API 1.2.0 to execute SPIN rules using the
> >> > SPINInferences.run(com.hp.hpl.jena.rdf.model.Model queryModel,
> >> >
> >> >
> com.hp.hpl.jena.rdf.model.Model newTriples,SPINExplanations explanations,
> List<SPINStatistics> statistics,
> >> > boolean singlePass, ProgressMonitor monito) method. However, rather
> than
> >> > having to load all of my SPIN rules/linked data/ontologies into a Jena
> >> > model
> >> > (quite memory intensive) I would like to instead connect to an OWLIM
> >> > Sesame
> >> > repository, as is done by TopBraid Composer.
> >> >
> >> > I am able to connect to an OWLIM Sesame repository using Jena using
> the
> >> > Jena
> >> > DatasetGraph interface
> >> >
> >> > (see
> http://owlim.ontotext.com/display/OWLIMv43/OWLIM-SE+Installation#OWLIM-SEInstallation-InstantiateOWLIMSEadapterusingtheprovidedAssembler
> ),
> >> > but it would defeat the purpose of connecting to an OWLIM Sesame
> >> > repository
> >> > to then load the DatasetGraph into a Jena model to pass into
> >> > SPINInferences.run().
> >> >
> >> > Can anyone provide suggestions as to how to pass a Jena DatasetGraph,
> or
> >> > otherwise connect to a Sesame repository such that I can use it with
> >> > SPINInferences.run()?
> >> >
> >> > Thanks,
> >> >
> >> > Alison
> >> >
> >> > --
> >> > You received this message because you are subscribed to the Google
> >> > Group "TopBraid Suite Users", the topics of which include Enterprise
> >> > Vocabulary Network (EVN), TopBraid Composer,
> >> > TopBraid Live, TopBraid Ensemble, SPARQLMotion and SPIN.
> >> > To post to this group, send email to
> >> > [email protected]
> >> > To unsubscribe from this group, send email to
> >> > [email protected]
> >> > For more options, visit this group at
> >> > http://groups.google.com/group/topbraid-users?hl=en
> >>
> >> --
> >> You received this message because you are subscribed to the Google
> >> Group "TopBraid Suite Users", the topics of which include Enterprise
> >> Vocabulary Network (EVN), TopBraid Composer,
> >> TopBraid Live, TopBraid Ensemble, SPARQLMotion and SPIN.
> >> To post to this group, send email to
> >> [email protected]
> >> To unsubscribe from this group, send email to
> >> [email protected]
> >> For more options, visit this group at
> >> http://groups.google.com/group/topbraid-users?hl=en
> >
> >
>
--
You received this message because you are subscribed to the Google
Group "TopBraid Suite Users", the topics of which include Enterprise Vocabulary
Network (EVN), TopBraid Composer,
TopBraid Live, TopBraid Ensemble, SPARQLMotion and SPIN.
To post to this group, send email to
[email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/topbraid-users?hl=en
