Examples with FileSplit org.apache.hyracks.api.io.FileSplit used on opensource projects

Example 16 with FileSplit

use of org.apache.hyracks.api.io.FileSplit in project asterixdb by apache.

the class CountOfCountsTest method countOfCountsExternalSortMultiNC.

@Test
public void countOfCountsExternalSortMultiNC() throws Exception {
    JobSpecification spec = new JobSpecification();
    FileSplit[] splits = new FileSplit[] { new ManagedFileSplit(NC2_ID, "data" + File.separator + "words.txt") };
    IFileSplitProvider splitProvider = new ConstantFileSplitProvider(splits);
    RecordDescriptor desc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer() });
    FileScanOperatorDescriptor csvScanner = new FileScanOperatorDescriptor(spec, splitProvider, new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE }, ','), desc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, csvScanner, NC2_ID);
    ExternalSortOperatorDescriptor sorter = new ExternalSortOperatorDescriptor(spec, 3, new int[] { 0 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, desc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, sorter, NC1_ID, NC2_ID, NC1_ID, NC2_ID);
    RecordDescriptor desc2 = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE });
    PreclusteredGroupOperatorDescriptor group = new PreclusteredGroupOperatorDescriptor(spec, new int[] { 0 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new CountFieldAggregatorFactory(true) }), desc2);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, group, NC1_ID, NC2_ID, NC1_ID, NC2_ID);
    InMemorySortOperatorDescriptor sorter2 = new InMemorySortOperatorDescriptor(spec, new int[] { 1 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, desc2);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, sorter2, NC1_ID, NC2_ID);
    RecordDescriptor desc3 = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE, IntegerSerializerDeserializer.INSTANCE });
    PreclusteredGroupOperatorDescriptor group2 = new PreclusteredGroupOperatorDescriptor(spec, new int[] { 1 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new CountFieldAggregatorFactory(true) }), desc3);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, group2, NC1_ID, NC2_ID);
    ResultSetId rsId = new ResultSetId(1);
    IOperatorDescriptor printer = new ResultWriterOperatorDescriptor(spec, rsId, true, false, ResultSerializerFactoryProvider.INSTANCE.getResultSerializerFactoryProvider());
    spec.addResultSetId(rsId);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);
    IConnectorDescriptor conn1 = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(new int[] { 0 }, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }));
    spec.connect(conn1, csvScanner, 0, sorter, 0);
    IConnectorDescriptor conn2 = new OneToOneConnectorDescriptor(spec);
    spec.connect(conn2, sorter, 0, group, 0);
    IConnectorDescriptor conn3 = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(new int[] { 1 }, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }));
    spec.connect(conn3, group, 0, sorter2, 0);
    IConnectorDescriptor conn4 = new OneToOneConnectorDescriptor(spec);
    spec.connect(conn4, sorter2, 0, group2, 0);
    IConnectorDescriptor conn5 = new MToNBroadcastConnectorDescriptor(spec);
    spec.connect(conn5, group2, 0, printer, 0);
    spec.addRoot(printer);
    runTest(spec);
}

Example 17 with FileSplit

use of org.apache.hyracks.api.io.FileSplit in project asterixdb by apache.

the class AbstractRTreeOperatorTest method setup.

@Before
public void setup() throws Exception {
    testHelper = createTestHelper();
    primaryFileName = testHelper.getPrimaryIndexName();
    primarySplitProvider = new ConstantFileSplitProvider(new FileSplit[] { new ManagedFileSplit(NC1_ID, primaryFileName) });
    primaryHelperFactory = new IndexDataflowHelperFactory(storageManager, primarySplitProvider);
    secondaryFileName = testHelper.getSecondaryIndexName();
    secondarySplitProvider = new ConstantFileSplitProvider(new FileSplit[] { new ManagedFileSplit(NC1_ID, secondaryFileName) });
    secondaryHelperFactory = new IndexDataflowHelperFactory(storageManager, secondarySplitProvider);
    // field, type and key declarations for primary index
    primaryTypeTraits[0] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[1] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[2] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[3] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[4] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[5] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[6] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[7] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[8] = UTF8StringPointable.TYPE_TRAITS;
    primaryTypeTraits[9] = UTF8StringPointable.TYPE_TRAITS;
    primaryComparatorFactories[0] = PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY);
    // field, type and key declarations for secondary indexes
    secondaryTypeTraits[0] = DoublePointable.TYPE_TRAITS;
    secondaryTypeTraits[1] = DoublePointable.TYPE_TRAITS;
    secondaryTypeTraits[2] = DoublePointable.TYPE_TRAITS;
    secondaryTypeTraits[3] = DoublePointable.TYPE_TRAITS;
    secondaryTypeTraits[4] = UTF8StringPointable.TYPE_TRAITS;
    secondaryComparatorFactories[0] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
    secondaryComparatorFactories[1] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
    secondaryComparatorFactories[2] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
    secondaryComparatorFactories[3] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
    // This only used for LSMRTree
    int[] rtreeFields = null;
    int[] filterFields = null;
    ITypeTraits[] filterTypes = null;
    IBinaryComparatorFactory[] filterCmpFactories = null;
    if (rTreeType == RTreeType.LSMRTREE || rTreeType == RTreeType.LSMRTREE_WITH_ANTIMATTER) {
        rtreeFields = new int[] { 0, 1, 2, 3, 4 };
        filterFields = new int[] { 4 };
        filterTypes = new ITypeTraits[] { UTF8StringPointable.TYPE_TRAITS };
        filterCmpFactories = new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) };
    }
    int[] btreeFields = null;
    if (rTreeType == RTreeType.LSMRTREE) {
        btreeKeyFieldCount = 1;
        btreeComparatorFactories = new IBinaryComparatorFactory[btreeKeyFieldCount];
        btreeComparatorFactories[0] = PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY);
        btreeFields = new int[] { 4 };
    } else {
        btreeComparatorFactories[0] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
        btreeComparatorFactories[1] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
        btreeComparatorFactories[2] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
        btreeComparatorFactories[3] = PointableBinaryComparatorFactory.of(DoublePointable.FACTORY);
        btreeComparatorFactories[4] = PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY);
    }
    IPrimitiveValueProviderFactory[] secondaryValueProviderFactories = RTreeUtils.createPrimitiveValueProviderFactories(secondaryComparatorFactories.length, DoublePointable.FACTORY);
    rtreeFactory = createSecondaryResourceFactory(secondaryValueProviderFactories, RTreePolicyType.RSTARTREE, btreeComparatorFactories, LSMRTreeUtils.proposeBestLinearizer(secondaryTypeTraits, secondaryComparatorFactories.length), btreeFields);
}

Example 18 with FileSplit

use of org.apache.hyracks.api.io.FileSplit in project asterixdb by apache.

the class FileSplitDomain method toString.

@Override
public String toString() {
    StringBuilder sb = new StringBuilder();
    sb.append("FileSplitDomain[");
    boolean fst = true;
    for (FileSplit fs : splits) {
        if (fst) {
            fst = false;
        } else {
            sb.append(", ");
        }
        sb.append(fs.getNodeName() + ":" + fs.getPath());
    }
    sb.append(']');
    return sb.toString();
}

Example 19 with FileSplit

use of org.apache.hyracks.api.io.FileSplit in project asterixdb by apache.

the class WordCountMain method parseFileSplits.

private static FileSplit[] parseFileSplits(String fileSplits) {
    String[] splits = fileSplits.split(",");
    FileSplit[] fSplits = new FileSplit[splits.length];
    for (int i = 0; i < splits.length; ++i) {
        String s = splits[i].trim();
        int idx = s.indexOf(':');
        if (idx < 0) {
            throw new IllegalArgumentException("File split " + s + " not well formed");
        }
        File file = new File(s.substring(idx + 1));
        fSplits[i] = new UnmanagedFileSplit(s.substring(0, idx), file.getAbsolutePath());
        fileSize += file.length();
    }
    return fSplits;
}

Example 20 with FileSplit

use of org.apache.hyracks.api.io.FileSplit in project asterixdb by apache.

the class TPCHCustomerOptimizedHybridHashJoinTest method customerOrderCIDHybridHashJoin_Case1.

@Test
public void customerOrderCIDHybridHashJoin_Case1() throws Exception {
    JobSpecification spec = new JobSpecification();
    FileSplit[] custSplits = new FileSplit[] { new ManagedFileSplit(NC1_ID, "data" + File.separator + "tpch0.001" + File.separator + "customer4.tbl") };
    IFileSplitProvider custSplitsProvider = new ConstantFileSplitProvider(custSplits);
    FileSplit[] ordersSplits = new FileSplit[] { new ManagedFileSplit(NC2_ID, "data" + File.separator + "tpch0.001" + File.separator + "orders4.tbl") };
    IFileSplitProvider ordersSplitsProvider = new ConstantFileSplitProvider(ordersSplits);
    FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitsProvider, new DelimitedDataTupleParserFactory(orderValueParserFactories, '|'), ordersDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, ordScanner, NC2_ID);
    FileScanOperatorDescriptor custScanner = new FileScanOperatorDescriptor(spec, custSplitsProvider, new DelimitedDataTupleParserFactory(custValueParserFactories, '|'), custDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, custScanner, NC1_ID);
    OptimizedHybridHashJoinOperatorDescriptor join = new OptimizedHybridHashJoinOperatorDescriptor(spec, 15, 243, 1.2, new int[] { 0 }, new int[] { 1 }, new IBinaryHashFunctionFamily[] { UTF8StringBinaryHashFunctionFamily.INSTANCE }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, custOrderJoinDesc, new JoinComparatorFactory(PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY), 0, 1), new JoinComparatorFactory(PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY), 1, 0), null);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, join, NC1_ID);
    String path = getClass().getName() + File.separator + "case1";
    IOperatorDescriptor printer = getPrinter(spec, path);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);
    IConnectorDescriptor custJoinConn = new OneToOneConnectorDescriptor(spec);
    spec.connect(custJoinConn, custScanner, 0, join, 0);
    IConnectorDescriptor ordJoinConn = new MToNBroadcastConnectorDescriptor(spec);
    spec.connect(ordJoinConn, ordScanner, 0, join, 1);
    IConnectorDescriptor joinPrinterConn = new OneToOneConnectorDescriptor(spec);
    spec.connect(joinPrinterConn, join, 0, printer, 0);
    spec.addRoot(printer);
    runTest(spec);
    System.out.println("output to " + path);
}