Search in sources :

Example 1 with AccumuloIndexSetProvider

use of org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider in project incubator-rya by apache.

the class PCJOptionalTestIT method testSimpleOptionalTest1.

@Test
public void testSimpleOptionalTest1() throws Exception {
    final String query = // 
    "" + // 
    "SELECT ?u ?s ?t " + // 
    "{" + // 
    "  ?s a ?t ." + // 
    "  OPTIONAL{?t <http://www.w3.org/2000/01/rdf-schema#label> ?u } ." + // 
    "  ?u <uri:talksTo> ?s . " + // 
    "}";
    final SPARQLParser parser = new SPARQLParser();
    final ParsedQuery pq1 = parser.parseQuery(query, null);
    final SimpleExternalTupleSet extTup1 = new SimpleExternalTupleSet((Projection) pq1.getTupleExpr().clone());
    final List<ExternalTupleSet> list = new ArrayList<ExternalTupleSet>();
    list.add(extTup1);
    final List<QueryModelNode> optTupNodes = Lists.newArrayList();
    optTupNodes.add(extTup1);
    final PCJOptimizer pcj = new PCJOptimizer(list, true, new AccumuloIndexSetProvider(new Configuration(), list));
    final TupleExpr te = pq1.getTupleExpr();
    pcj.optimize(te, null, null);
    final NodeCollector nc = new NodeCollector();
    te.visit(nc);
    final List<QueryModelNode> qNodes = nc.getNodes();
    Assert.assertEquals(qNodes.size(), optTupNodes.size());
    for (final QueryModelNode node : qNodes) {
        Assert.assertTrue(optTupNodes.contains(node));
    }
}
Also used : SPARQLParser(org.openrdf.query.parser.sparql.SPARQLParser) Configuration(org.apache.hadoop.conf.Configuration) ParsedQuery(org.openrdf.query.parser.ParsedQuery) AccumuloIndexSetProvider(org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider) ArrayList(java.util.ArrayList) QueryModelNode(org.openrdf.query.algebra.QueryModelNode) TupleExpr(org.openrdf.query.algebra.TupleExpr) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) ExternalTupleSet(org.apache.rya.indexing.external.tupleSet.ExternalTupleSet) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) PCJOptimizer(org.apache.rya.indexing.pcj.matching.PCJOptimizer) NodeCollector(org.apache.rya.indexing.external.PrecompJoinOptimizerTest.NodeCollector) Test(org.junit.Test)

Example 2 with AccumuloIndexSetProvider

use of org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider in project incubator-rya by apache.

the class AccumuloPcjIT method testEvaluateThreeIndexValidate.

@Test
public void testEvaluateThreeIndexValidate() throws Exception {
    final URI superclass = new URIImpl("uri:superclass");
    final URI superclass2 = new URIImpl("uri:superclass2");
    final URI sub = new URIImpl("uri:entity");
    subclass = new URIImpl("uri:class");
    obj = new URIImpl("uri:obj");
    talksTo = new URIImpl("uri:talksTo");
    final URI howlsAt = new URIImpl("uri:howlsAt");
    final URI subType = new URIImpl("uri:subType");
    final URI superSuperclass = new URIImpl("uri:super_superclass");
    conn.add(subclass, RDF.TYPE, superclass);
    conn.add(subclass2, RDF.TYPE, superclass2);
    conn.add(obj, RDFS.LABEL, new LiteralImpl("label"));
    conn.add(obj2, RDFS.LABEL, new LiteralImpl("label2"));
    conn.add(sub, howlsAt, superclass);
    conn.add(superclass, subType, superSuperclass);
    conn.add(obj, RDFS.LABEL, new LiteralImpl("label"));
    conn.add(obj2, RDFS.LABEL, new LiteralImpl("label2"));
    final String indexSparqlString = // 
    "" + // 
    "SELECT ?dog ?pig ?duck  " + // 
    "{" + // 
    "  ?pig a ?dog . " + // 
    "  ?pig <http://www.w3.org/2000/01/rdf-schema#label> ?duck " + // 
    "}";
    final String indexSparqlString2 = // 
    "" + // 
    "SELECT ?o ?f ?e ?c ?l  " + // 
    "{" + // 
    "  ?e <uri:talksTo> ?o . " + // 
    "  ?o <http://www.w3.org/2000/01/rdf-schema#label> ?l. " + // 
    "  ?c a ?f . " + // 
    "}";
    final String indexSparqlString3 = // 
    "" + // 
    "SELECT ?wolf ?sheep ?chicken  " + // 
    "{" + // 
    "  ?wolf <uri:howlsAt> ?sheep . " + // 
    "  ?sheep <uri:subType> ?chicken. " + // 
    "}";
    final String queryString = // 
    "" + // 
    "SELECT ?e ?c ?l ?f ?o " + // 
    "{" + // 
    "  ?e a ?c . " + // 
    "  ?e <http://www.w3.org/2000/01/rdf-schema#label> ?l. " + // 
    "  ?e <uri:talksTo> ?o . " + // 
    "  ?o <http://www.w3.org/2000/01/rdf-schema#label> ?l. " + // 
    "  ?c a ?f . " + // 
    "  ?e <uri:howlsAt> ?f. " + // 
    "  ?f <uri:subType> ?o. " + // 
    "}";
    PcjIntegrationTestingUtil.createAndPopulatePcj(conn, accCon, tablename + 1, indexSparqlString, new String[] { "dog", "pig", "duck" }, Optional.<PcjVarOrderFactory>absent());
    final AccumuloIndexSet ais1 = new AccumuloIndexSet(conf, tablename + 1);
    PcjIntegrationTestingUtil.createAndPopulatePcj(conn, accCon, tablename + 2, indexSparqlString2, new String[] { "o", "f", "e", "c", "l" }, Optional.<PcjVarOrderFactory>absent());
    final AccumuloIndexSet ais2 = new AccumuloIndexSet(conf, tablename + 2);
    PcjIntegrationTestingUtil.createAndPopulatePcj(conn, accCon, tablename + 3, indexSparqlString3, new String[] { "wolf", "sheep", "chicken" }, Optional.<PcjVarOrderFactory>absent());
    final AccumuloIndexSet ais3 = new AccumuloIndexSet(conf, tablename + 3);
    final List<ExternalTupleSet> index = new ArrayList<>();
    index.add(ais1);
    index.add(ais3);
    index.add(ais2);
    ParsedQuery pq = null;
    final SPARQLParser sp = new SPARQLParser();
    pq = sp.parseQuery(queryString, null);
    final List<TupleExpr> teList = Lists.newArrayList();
    final TupleExpr te = pq.getTupleExpr();
    final PCJOptimizer pcj = new PCJOptimizer(index, false, new AccumuloIndexSetProvider(conf));
    pcj.optimize(te, null, null);
    teList.add(te);
    final IndexPlanValidator ipv = new IndexPlanValidator(false);
    Assert.assertTrue(ipv.isValid(te));
}
Also used : SPARQLParser(org.openrdf.query.parser.sparql.SPARQLParser) AccumuloIndexSet(org.apache.rya.indexing.external.tupleSet.AccumuloIndexSet) ParsedQuery(org.openrdf.query.parser.ParsedQuery) AccumuloIndexSetProvider(org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider) ArrayList(java.util.ArrayList) IndexPlanValidator(org.apache.rya.indexing.IndexPlanValidator.IndexPlanValidator) URIImpl(org.openrdf.model.impl.URIImpl) URI(org.openrdf.model.URI) TupleExpr(org.openrdf.query.algebra.TupleExpr) ExternalTupleSet(org.apache.rya.indexing.external.tupleSet.ExternalTupleSet) LiteralImpl(org.openrdf.model.impl.LiteralImpl) PCJOptimizer(org.apache.rya.indexing.pcj.matching.PCJOptimizer) Test(org.junit.Test)

Example 3 with AccumuloIndexSetProvider

use of org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider in project incubator-rya by apache.

the class PrecompJoinOptimizerVarToConstTest method testFreeTestIndexFunction.

@Test
public void testFreeTestIndexFunction() throws Exception {
    final SPARQLParser parser1 = new SPARQLParser();
    final SPARQLParser parser2 = new SPARQLParser();
    final ParsedQuery pq1 = parser1.parseQuery(q22, null);
    final ParsedQuery pq2 = parser2.parseQuery(q24, null);
    System.out.println("Query is " + pq1.getTupleExpr());
    System.out.println("Index is " + pq2.getTupleExpr());
    final SimpleExternalTupleSet extTup = new SimpleExternalTupleSet(new Projection(pq2.getTupleExpr()));
    final List<ExternalTupleSet> list = new ArrayList<ExternalTupleSet>();
    list.add(extTup);
    final TupleExpr tup = pq1.getTupleExpr().clone();
    final PCJOptimizer pcj = new PCJOptimizer(list, false, new AccumuloIndexSetProvider(new Configuration(), list));
    pcj.optimize(tup, null, null);
    System.out.println("Processed query is " + tup);
    final Set<StatementPattern> qSet = Sets.newHashSet(StatementPatternCollector.process(pq1.getTupleExpr()));
    final Set<QueryModelNode> eTupSet = PcjIntegrationTestingUtil.getTupleSets(tup);
    final Set<StatementPattern> set = Sets.newHashSet();
    Assert.assertEquals(2, eTupSet.size());
    for (final QueryModelNode s : eTupSet) {
        final Set<StatementPattern> tempSet = Sets.newHashSet(StatementPatternCollector.process(((ExternalTupleSet) s).getTupleExpr()));
        set.addAll(tempSet);
    }
    Assert.assertTrue(qSet.containsAll(set));
}
Also used : SPARQLParser(org.openrdf.query.parser.sparql.SPARQLParser) Configuration(org.apache.hadoop.conf.Configuration) ParsedQuery(org.openrdf.query.parser.ParsedQuery) AccumuloIndexSetProvider(org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider) ArrayList(java.util.ArrayList) Projection(org.openrdf.query.algebra.Projection) QueryModelNode(org.openrdf.query.algebra.QueryModelNode) TupleExpr(org.openrdf.query.algebra.TupleExpr) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) ExternalTupleSet(org.apache.rya.indexing.external.tupleSet.ExternalTupleSet) StatementPattern(org.openrdf.query.algebra.StatementPattern) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) PCJOptimizer(org.apache.rya.indexing.pcj.matching.PCJOptimizer) Test(org.junit.Test)

Example 4 with AccumuloIndexSetProvider

use of org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider in project incubator-rya by apache.

the class PrecompJoinOptimizerVarToConstTest method testContextFilterFourIndex.

@Test
public void testContextFilterFourIndex() throws Exception {
    final SPARQLParser parser1 = new SPARQLParser();
    final SPARQLParser parser3 = new SPARQLParser();
    final SPARQLParser parser4 = new SPARQLParser();
    final ParsedQuery pq1 = parser1.parseQuery(q19, null);
    final ParsedQuery pq3 = parser3.parseQuery(q17, null);
    final ParsedQuery pq4 = parser4.parseQuery(q18, null);
    System.out.println("Query is " + pq1.getTupleExpr());
    System.out.println("Indexes are " + pq3.getTupleExpr() + " , " + pq4.getTupleExpr());
    final SimpleExternalTupleSet extTup2 = new SimpleExternalTupleSet(new Projection(pq3.getTupleExpr()));
    final SimpleExternalTupleSet extTup3 = new SimpleExternalTupleSet(new Projection(pq4.getTupleExpr()));
    final List<ExternalTupleSet> list = new ArrayList<ExternalTupleSet>();
    list.add(extTup3);
    list.add(extTup2);
    final TupleExpr tup = pq1.getTupleExpr().clone();
    final PCJOptimizer pcj = new PCJOptimizer(list, false, new AccumuloIndexSetProvider(new Configuration(), list));
    pcj.optimize(tup, null, null);
    System.out.println("Processed query is " + tup);
    final Set<StatementPattern> qSet = Sets.newHashSet(StatementPatternCollector.process(pq1.getTupleExpr()));
    final Set<QueryModelNode> eTupSet = PcjIntegrationTestingUtil.getTupleSets(tup);
    Assert.assertEquals(2, eTupSet.size());
    final Set<StatementPattern> set = Sets.newHashSet();
    for (final QueryModelNode s : eTupSet) {
        final Set<StatementPattern> tempSet = Sets.newHashSet(StatementPatternCollector.process(((ExternalTupleSet) s).getTupleExpr()));
        set.addAll(tempSet);
    }
    Assert.assertTrue(qSet.containsAll(set));
}
Also used : SPARQLParser(org.openrdf.query.parser.sparql.SPARQLParser) Configuration(org.apache.hadoop.conf.Configuration) ParsedQuery(org.openrdf.query.parser.ParsedQuery) AccumuloIndexSetProvider(org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider) ArrayList(java.util.ArrayList) Projection(org.openrdf.query.algebra.Projection) QueryModelNode(org.openrdf.query.algebra.QueryModelNode) TupleExpr(org.openrdf.query.algebra.TupleExpr) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) ExternalTupleSet(org.apache.rya.indexing.external.tupleSet.ExternalTupleSet) StatementPattern(org.openrdf.query.algebra.StatementPattern) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) PCJOptimizer(org.apache.rya.indexing.pcj.matching.PCJOptimizer) Test(org.junit.Test)

Example 5 with AccumuloIndexSetProvider

use of org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider in project incubator-rya by apache.

the class PCJOptimizerBenchmark method makeChainedPCJOptimizer.

private static PCJOptimizer makeChainedPCJOptimizer(final BenchmarkParams params) throws Exception {
    final Queue<String> varQueue = Lists.newLinkedList(variables);
    final SPARQLParser parser = new SPARQLParser();
    final List<ExternalTupleSet> indices = new ArrayList<>();
    // Create the first PCJ.
    final List<String> pcjVars = new ArrayList<>();
    pcjVars.add(varQueue.remove());
    pcjVars.add(varQueue.remove());
    for (int spI = 1; spI < params.getPCJSPCount(); spI++) {
        pcjVars.add(varQueue.remove());
    }
    String pcjSparql = buildChainedSPARQL(pcjVars);
    Projection projection = (Projection) parser.parseQuery(pcjSparql, null).getTupleExpr();
    indices.add(new SimpleExternalTupleSet(projection));
    // Add the rest of the PCJs.
    for (int pcjI = 1; pcjI < params.getNumPCJS(); pcjI++) {
        // Remove the previous PCJs first variable.
        pcjVars.remove(0);
        // And add a new one to the end of it.
        pcjVars.add(varQueue.remove());
        // Build the index.
        pcjSparql = buildChainedSPARQL(pcjVars);
        projection = (Projection) parser.parseQuery(pcjSparql, null).getTupleExpr();
        indices.add(new SimpleExternalTupleSet(projection));
    }
    // Create the optimizer.
    return new PCJOptimizer(indices, false, new AccumuloIndexSetProvider(new Configuration()));
}
Also used : SPARQLParser(org.openrdf.query.parser.sparql.SPARQLParser) Configuration(org.apache.hadoop.conf.Configuration) AccumuloIndexSetProvider(org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider) ArrayList(java.util.ArrayList) Projection(org.openrdf.query.algebra.Projection) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) ExternalTupleSet(org.apache.rya.indexing.external.tupleSet.ExternalTupleSet) SimpleExternalTupleSet(org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet) PCJOptimizer(org.apache.rya.indexing.pcj.matching.PCJOptimizer)

Aggregations

AccumuloIndexSetProvider (org.apache.rya.indexing.pcj.matching.provider.AccumuloIndexSetProvider)13 ExternalTupleSet (org.apache.rya.indexing.external.tupleSet.ExternalTupleSet)12 PCJOptimizer (org.apache.rya.indexing.pcj.matching.PCJOptimizer)12 SPARQLParser (org.openrdf.query.parser.sparql.SPARQLParser)12 ArrayList (java.util.ArrayList)11 Configuration (org.apache.hadoop.conf.Configuration)10 SimpleExternalTupleSet (org.apache.rya.indexing.external.tupleSet.SimpleExternalTupleSet)10 Test (org.junit.Test)10 TupleExpr (org.openrdf.query.algebra.TupleExpr)10 ParsedQuery (org.openrdf.query.parser.ParsedQuery)10 Projection (org.openrdf.query.algebra.Projection)7 QueryModelNode (org.openrdf.query.algebra.QueryModelNode)7 StatementPattern (org.openrdf.query.algebra.StatementPattern)5 IndexPlanValidator (org.apache.rya.indexing.IndexPlanValidator.IndexPlanValidator)2 NodeCollector (org.apache.rya.indexing.external.PrecompJoinOptimizerTest.NodeCollector)2 AccumuloIndexSet (org.apache.rya.indexing.external.tupleSet.AccumuloIndexSet)2 URI (org.openrdf.model.URI)2 LiteralImpl (org.openrdf.model.impl.LiteralImpl)2 URIImpl (org.openrdf.model.impl.URIImpl)2 MongoPcjIndexSetProvider (org.apache.rya.indexing.mongodb.pcj.MongoPcjIndexSetProvider)1