Search in sources :

Example 26 with KeywordMatcherSourceOperator

use of edu.uci.ics.texera.dataflow.keywordmatcher.KeywordMatcherSourceOperator in project textdb by TextDB.

the class LogicalPlanTest method testLogicalPlan3.

/*
     * Test a valid operator graph.
     * 
     *                  --> RegexMatcher -->
     *                  |                    >-- Join1
     * KeywordSource --< -> NlpEntityOperator -->          >-- Join2 --> TupleSink
     *                  |                           /
     *                  --> FuzzyTokenMatcher ----->
     * 
     */
@Test
public void testLogicalPlan3() throws Exception {
    LogicalPlan logicalPlan = getLogicalPlan3();
    Plan queryPlan = logicalPlan.buildQueryPlan();
    ISink tupleSink = queryPlan.getRoot();
    Assert.assertTrue(tupleSink instanceof TupleSink);
    IOperator join2 = ((TupleSink) tupleSink).getInputOperator();
    Assert.assertTrue(join2 instanceof Join);
    IOperator join2Input1 = ((Join) join2).getOuterInputOperator();
    Assert.assertTrue(join2Input1 instanceof Join);
    IOperator join2Input2 = ((Join) join2).getInnerInputOperator();
    Assert.assertTrue(join2Input2 instanceof FuzzyTokenMatcher);
    IOperator join1Input1 = ((Join) join2Input1).getInnerInputOperator();
    Assert.assertTrue(join1Input1 instanceof RegexMatcher);
    IOperator join1Input2 = ((Join) join2Input1).getOuterInputOperator();
    Assert.assertTrue(join1Input2 instanceof NlpEntityOperator);
    IOperator connectorOut1 = ((RegexMatcher) join1Input1).getInputOperator();
    Assert.assertTrue(connectorOut1 instanceof ConnectorOutputOperator);
    IOperator connectorOut2 = ((NlpEntityOperator) join1Input2).getInputOperator();
    Assert.assertTrue(connectorOut2 instanceof ConnectorOutputOperator);
    IOperator connectorOut3 = ((FuzzyTokenMatcher) join2Input2).getInputOperator();
    Assert.assertTrue(connectorOut3 instanceof ConnectorOutputOperator);
    HashSet<Integer> connectorIndices = new HashSet<>();
    connectorIndices.add(((ConnectorOutputOperator) connectorOut1).getOutputIndex());
    connectorIndices.add(((ConnectorOutputOperator) connectorOut2).getOutputIndex());
    connectorIndices.add(((ConnectorOutputOperator) connectorOut3).getOutputIndex());
    Assert.assertEquals(connectorIndices.size(), 3);
    OneToNBroadcastConnector connector1 = ((ConnectorOutputOperator) connectorOut1).getOwnerConnector();
    OneToNBroadcastConnector connector2 = ((ConnectorOutputOperator) connectorOut2).getOwnerConnector();
    OneToNBroadcastConnector connector3 = ((ConnectorOutputOperator) connectorOut3).getOwnerConnector();
    Assert.assertSame(connector1, connector2);
    Assert.assertSame(connector1, connector3);
    IOperator keywordSource = connector1.getInputOperator();
    Assert.assertTrue(keywordSource instanceof KeywordMatcherSourceOperator);
}
Also used : TupleSink(edu.uci.ics.texera.dataflow.sink.tuple.TupleSink) IOperator(edu.uci.ics.texera.api.dataflow.IOperator) Join(edu.uci.ics.texera.dataflow.join.Join) Plan(edu.uci.ics.texera.api.engine.Plan) FuzzyTokenMatcher(edu.uci.ics.texera.dataflow.fuzzytokenmatcher.FuzzyTokenMatcher) KeywordMatcherSourceOperator(edu.uci.ics.texera.dataflow.keywordmatcher.KeywordMatcherSourceOperator) ISink(edu.uci.ics.texera.api.dataflow.ISink) ConnectorOutputOperator(edu.uci.ics.texera.dataflow.connector.OneToNBroadcastConnector.ConnectorOutputOperator) NlpEntityOperator(edu.uci.ics.texera.dataflow.nlp.entity.NlpEntityOperator) RegexMatcher(edu.uci.ics.texera.dataflow.regexmatcher.RegexMatcher) OneToNBroadcastConnector(edu.uci.ics.texera.dataflow.connector.OneToNBroadcastConnector) HashSet(java.util.HashSet) Test(org.junit.Test)

Aggregations

KeywordMatcherSourceOperator (edu.uci.ics.texera.dataflow.keywordmatcher.KeywordMatcherSourceOperator)26 Test (org.junit.Test)22 Tuple (edu.uci.ics.texera.api.tuple.Tuple)20 JoinDistancePredicate (edu.uci.ics.texera.dataflow.join.JoinDistancePredicate)19 Span (edu.uci.ics.texera.api.span.Span)11 ArrayList (java.util.ArrayList)11 IField (edu.uci.ics.texera.api.field.IField)9 IntegerField (edu.uci.ics.texera.api.field.IntegerField)9 StringField (edu.uci.ics.texera.api.field.StringField)9 TextField (edu.uci.ics.texera.api.field.TextField)9 Schema (edu.uci.ics.texera.api.schema.Schema)9 KeywordSourcePredicate (edu.uci.ics.texera.dataflow.keywordmatcher.KeywordSourcePredicate)4 IOperator (edu.uci.ics.texera.api.dataflow.IOperator)3 ISink (edu.uci.ics.texera.api.dataflow.ISink)3 Plan (edu.uci.ics.texera.api.engine.Plan)3 Join (edu.uci.ics.texera.dataflow.join.Join)3 RegexMatcher (edu.uci.ics.texera.dataflow.regexmatcher.RegexMatcher)3 TupleSink (edu.uci.ics.texera.dataflow.sink.tuple.TupleSink)3 OneToNBroadcastConnector (edu.uci.ics.texera.dataflow.connector.OneToNBroadcastConnector)2 ConnectorOutputOperator (edu.uci.ics.texera.dataflow.connector.OneToNBroadcastConnector.ConnectorOutputOperator)2