Re: [Bigdata-developers] Jena Adapter?

[Bryan T. <br...@sy...>]

Laurent,

This exception would be from using the wrong Sesame JARs.  I think we probably failed to update the POM when we updated the JARs.

java.lang.NoSuchFieldError: NQUADS
at com.bigdata.rdf.rio.nquads.NQuadsParserFactory.getRDFFormat(NQuadsParserFactory.java:47)

I have just committed a fix for this.  It is building in CI now.  A new snapshot JAR should be posted shortly.

Truth maintenance is only supported for triples, not quads.  There are several discussions on the forum concerning what would be required to support truth maintenance and inference over quads.  Briefly, the problem is specifying which named graph(s) are combined and where the entailments are written.  There are also several suggestions of alternative mechanisms involving triples that may be better suited to some applications, especially the triples plus statement level metadata database mode.

Thanks,
Bryan

From: Laurent Pellegrino <lau...@gm...<mailto:lau...@gm...>>
Date: Friday, March 1, 2013 6:41 AM
To: Bryan Thompson <br...@sy...<mailto:br...@sy...>>
Subject: Re: [Bigdata-developers] Jena Adapter?

Hello,

Thank you for the answers. I tried to create a Bigdata repository with support for quads with the following piece of code:

  Properties properties =
                new Properties();
  File journal = File.createTempFile("bigdata", ".jnl");

  properties.setProperty(
                BigdataSail.Options.FILE, journal.getAbsolutePath());
  properties.setProperty(
                BigdataSail.Options.QUADS_MODE, "true");
  properties.setProperty(
                BigdataSail.Options.TRUTH_MAINTENANCE, "false");

  BigdataSail sail = new BigdataSail(properties);
  Repository repo = new BigdataSailRepository(sail);
  repo.initialize();
  repo.shutDown();


but the repository initialization seems not to work. I get the following error :

[main] ERROR rio.RDFParserRegistry  - Failed to instantiate service
java.lang.NoSuchFieldError: NQUADS
at com.bigdata.rdf.rio.nquads.NQuadsParserFactory.getRDFFormat(NQuadsParserFactory.java:47)
at org.openrdf.rio.RDFParserRegistry.getKey(RDFParserRegistry.java:38)
at org.openrdf.rio.RDFParserRegistry.getKey(RDFParserRegistry.java:15)
at info.aduna.lang.service.ServiceRegistry.add(ServiceRegistry.java:74)
at info.aduna.lang.service.ServiceRegistry.<init>(ServiceRegistry.java:44)
at info.aduna.lang.service.FileFormatServiceRegistry.<init>(FileFormatServiceRegistry.java:20)
at org.openrdf.rio.RDFParserRegistry.<init>(RDFParserRegistry.java:33)
at org.openrdf.rio.RDFParserRegistry.getInstance(RDFParserRegistry.java:26)
at com.bigdata.rdf.ServiceProviderHook.forceLoad(ServiceProviderHook.java:109)
at com.bigdata.rdf.ServiceProviderHook.<clinit>(ServiceProviderHook.java:84)
at com.bigdata.rdf.store.AbstractTripleStore.<init>(AbstractTripleStore.java:1266)
at com.bigdata.rdf.store.AbstractLocalTripleStore.<init>(AbstractLocalTripleStore.java:57)
at com.bigdata.rdf.store.LocalTripleStore.<init>(LocalTripleStore.java:161)
at com.bigdata.rdf.sail.BigdataSail.createLTS(BigdataSail.java:726)
at com.bigdata.rdf.sail.BigdataSail.createLTS(BigdataSail.java:653)
at com.bigdata.rdf.sail.BigdataSail.<init>(BigdataSail.java:630)
at fr.inria.eventcloud.stash.rdfstores.benchmark.BigdataRepository.createBigDataRepository(BigdataRepository.java:53)
at fr.inria.eventcloud.stash.rdfstores.benchmark.RDFStoreBenchmark.testBigdataSequential(RDFStoreBenchmark.java:45)
at fr.inria.eventcloud.stash.rdfstores.benchmark.RDFStoreBenchmark.main(RDFStoreBenchmark.java:77)


I have set truth maintenance to false given that I get the following message if I let it to the default value when quads mode is set to true:

java.lang.UnsupportedOperationException: com.bigdata.rdf.sail.truthMaintenance is not supported with quads (com.bigdata.rdf.store.AbstractTripleStore.quads)

My test was performed by using bigdata version 1.2.2-SNAPSHOT from your maven repository.

I cannot figure out what is wrong in the previous piece of code. Could you help me?

Kind Regards,

Laurent


On Thu, Feb 28, 2013 at 1:12 PM, Bryan Thompson <br...@sy...<mailto:br...@sy...>> wrote:
Laurent,

Unfortunately not.   This has been discussed several times, but never implemented.  However, the advent of SPARQL UPDATE makes it much easier to write portable applications.  If you like, feel free to post questions about how to achieve certain things under bigdata that you currently accomplish with Jena.

Thanks,
Bryan

From: Laurent Pellegrino <lau...@gm...<mailto:lau...@gm...>>
Date: Thursday, February 28, 2013 4:27 AM
To: "big...@li...<mailto:big...@li...>" <big...@li...<mailto:big...@li...>>
Subject: [Bigdata-developers] Jena Adapter?

Hello all,

I am investigating the use of Bigdata in one application that currently uses Jena TDB. I would like to test Bigdata without rewriting all, that's why I wonder if there exists a jena adapter on top of Bigdata?

Kind Regards,

Laurent