Search code examples
javalucenesparqljenaarq

Is there a way to create several indexes in LARQ and execute SPARQL on them

I want to index two different OntModel and execute sparql queries on them. For example:

String query = "PREFIX fise: <http://fise.iks-project.eu/ontology/>\n"
                   + "PREFIX pf: <http://jena.hpl.hp.com/ARQ/property#>\n"
                   + "PREFIX skos:   <http://www.w3.org/2004/02/skos/core#>\n"
                   + "SELECT distinct ?def\n" + "WHERE {\n"
                   + "  ?item a skos:Concept .\n"
                   + "  ?item skos:definition ?def.\n"
                   + "  (?label ?score)  pf:textMatch 'someKeyword'. \n"
                   + "};

ResultSet result = QueryExecutionFactory.create(query, ontModel1).execSelect();
ResultSet result2 = QueryExecutionFactory.create(query, ontModel2).execSelect();

How can I create different indexes for the two ont models or should I do that? Eventually, what I need is to query the two ont models using LARQ features.

As far as I understand to be able to use lucene features in while executing SPARQL, we need to implement a code like the following one.

IndexBuilderString larqBuilder = new IndexBuilderString();
larqBuilder.indexStatements(ontModel.listStatements());
IndexLARQ index = larqBuilder.getIndex();
LARQ.setDefaultIndex(index);

But this code allows setting a single index. Furthermore, I don't want to merge two ont models. There is a Concext concept but I didn't get how to use it.

Solution

  • It seems the answer is registering indexes per-query execution using the context of query.

     QueryExecution qe = QueryExecutionFactory.create(query, model) ;
 LARQ.setDefaultIndex(qe.getContext(), index) ;