hyracks/hyracks-examples/hyracks-integration-tests/src/test/java/edu/uci/ics/hyracks/tests/integration/SortMergeTest.java

/*
 * Copyright 2009-2010 by The Regents of the University of California
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * you may obtain a copy of the License from
 * 
 *     http://www.apache.org/licenses/LICENSE-2.0
 * 
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
package edu.uci.ics.hyracks.tests.integration;

import java.io.File;

import org.junit.Test;

import edu.uci.ics.hyracks.api.constraints.PartitionConstraintHelper;
import edu.uci.ics.hyracks.api.dataflow.IOperatorDescriptor;
import edu.uci.ics.hyracks.api.dataflow.value.IBinaryComparatorFactory;
import edu.uci.ics.hyracks.api.dataflow.value.IBinaryHashFunctionFactory;
import edu.uci.ics.hyracks.api.dataflow.value.ISerializerDeserializer;
import edu.uci.ics.hyracks.api.dataflow.value.RecordDescriptor;
import edu.uci.ics.hyracks.api.dataset.ResultSetId;
import edu.uci.ics.hyracks.api.io.FileReference;
import edu.uci.ics.hyracks.api.job.JobSpecification;
import edu.uci.ics.hyracks.data.std.accessors.PointableBinaryComparatorFactory;
import edu.uci.ics.hyracks.data.std.accessors.PointableBinaryHashFunctionFactory;
import edu.uci.ics.hyracks.data.std.primitive.UTF8StringPointable;
import edu.uci.ics.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer;
import edu.uci.ics.hyracks.dataflow.common.data.parsers.IValueParserFactory;
import edu.uci.ics.hyracks.dataflow.common.data.parsers.UTF8StringParserFactory;
import edu.uci.ics.hyracks.dataflow.common.data.partition.FieldHashPartitionComputerFactory;
import edu.uci.ics.hyracks.dataflow.std.connectors.MToNPartitioningMergingConnectorDescriptor;
import edu.uci.ics.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor;
import edu.uci.ics.hyracks.dataflow.std.file.ConstantFileSplitProvider;
import edu.uci.ics.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory;
import edu.uci.ics.hyracks.dataflow.std.file.FileScanOperatorDescriptor;
import edu.uci.ics.hyracks.dataflow.std.file.FileSplit;
import edu.uci.ics.hyracks.dataflow.std.file.IFileSplitProvider;
import edu.uci.ics.hyracks.dataflow.std.result.ResultWriterOperatorDescriptor;
import edu.uci.ics.hyracks.dataflow.std.sort.ExternalSortOperatorDescriptor;
import edu.uci.ics.hyracks.dataflow.std.sort.InMemorySortOperatorDescriptor;
import edu.uci.ics.hyracks.tests.util.ResultSerializerFactoryProvider;

public class SortMergeTest extends AbstractIntegrationTest {
    @Test
    public void sortMergeTest01() throws Exception {
        JobSpecification spec = new JobSpecification();

        FileSplit[] ordersSplits = new FileSplit[] {
                new FileSplit(NC1_ID, new FileReference(new File("data/tpch0.001/orders-part1.tbl"))),
                new FileSplit(NC2_ID, new FileReference(new File("data/tpch0.001/orders-part2.tbl"))) };
        IFileSplitProvider ordersSplitProvider = new ConstantFileSplitProvider(ordersSplits);
        RecordDescriptor ordersDesc = new RecordDescriptor(new ISerializerDeserializer[] {
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE });

        FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitProvider,
                new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE }, '|'), ordersDesc);
        PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, ordScanner, NC1_ID, NC2_ID);

        InMemorySortOperatorDescriptor sorter = new InMemorySortOperatorDescriptor(spec, new int[] { 1 },
                new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) },
                ordersDesc);
        PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, sorter, NC1_ID, NC2_ID);

        ResultSetId rsId = new ResultSetId(1);
        spec.addResultSetId(rsId);

        IOperatorDescriptor printer = new ResultWriterOperatorDescriptor(spec, rsId, true,
                ResultSerializerFactoryProvider.INSTANCE.getResultSerializerFactoryProvider());
        PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);

        spec.connect(new OneToOneConnectorDescriptor(spec), ordScanner, 0, sorter, 0);

        spec.connect(
                new MToNPartitioningMergingConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(
                        new int[] { 1 }, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory
                                .of(UTF8StringPointable.FACTORY) }), new int[] { 1 },
                        new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory
                                .of(UTF8StringPointable.FACTORY) }), sorter, 0, printer, 0);

        runTest(spec);
    }

    @Test
    public void sortMergeTest02() throws Exception {
        JobSpecification spec = new JobSpecification();

        FileSplit[] ordersSplits = new FileSplit[] {
                new FileSplit(NC1_ID, new FileReference(new File("data/tpch0.001/orders-part1.tbl"))),
                new FileSplit(NC2_ID, new FileReference(new File("data/tpch0.001/orders-part2.tbl"))) };
        IFileSplitProvider ordersSplitProvider = new ConstantFileSplitProvider(ordersSplits);
        RecordDescriptor ordersDesc = new RecordDescriptor(new ISerializerDeserializer[] {
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE, UTF8StringSerializerDeserializer.INSTANCE,
                UTF8StringSerializerDeserializer.INSTANCE });

        FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitProvider,
                new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE,
                        UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE }, '|'), ordersDesc);
        PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, ordScanner, NC1_ID, NC2_ID);

        ExternalSortOperatorDescriptor sorter = new ExternalSortOperatorDescriptor(spec, 4, new int[] { 1, 0 },
                new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY),
                        PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, ordersDesc);
        PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, sorter, NC1_ID, NC2_ID);

        ResultSetId rsId = new ResultSetId(1);
        spec.addResultSetId(rsId);

        IOperatorDescriptor printer = new ResultWriterOperatorDescriptor(spec, rsId, false,
                ResultSerializerFactoryProvider.INSTANCE.getResultSerializerFactoryProvider());
        PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);

        spec.connect(new OneToOneConnectorDescriptor(spec), ordScanner, 0, sorter, 0);

        spec.connect(
                new MToNPartitioningMergingConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(new int[] {
                        1, 0 }, new IBinaryHashFunctionFactory[] {
                        PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY),
                        PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }), new int[] { 1, 0 },
                        new IBinaryComparatorFactory[] {
                                PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY),
                                PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }), sorter, 0,
                printer, 0);

        runTest(spec);
    }
}