Examples with IFileSplitProvider org.apache.hyracks.dataflow.std.file.IFileSplitProvider used on opensource projects

Example 21 with IFileSplitProvider

use of org.apache.hyracks.dataflow.std.file.IFileSplitProvider in project asterixdb by apache.

the class PushRuntimeTest method scanSplitWrite.

@Test
public void scanSplitWrite() throws Exception {
    final int outputArity = 2;
    JobSpecification spec = new JobSpecification(FRAME_SIZE);
    String[] inputFileName = { "data" + File.separator + "simple" + File.separator + "int-string-part1.tbl", "data" + File.separator + "simple" + File.separator + "int-string-part1-split-0.tbl", "data" + File.separator + "simple" + File.separator + "int-string-part1-split-1.tbl" };
    File[] inputFiles = new File[inputFileName.length];
    for (int i = 0; i < inputFileName.length; i++) {
        inputFiles[i] = new File(inputFileName[i]);
    }
    File[] outputFile = new File[outputArity];
    FileSplit[] outputFileSplit = new FileSplit[outputArity];
    for (int i = 0; i < outputArity; i++) {
        outputFileSplit[i] = createFile(AlgebricksHyracksIntegrationUtil.nc1);
        outputFile[i] = outputFileSplit[i].getFile(AlgebricksHyracksIntegrationUtil.nc1.getIoManager());
    }
    FileSplit[] inputSplits = new FileSplit[] { new ManagedFileSplit(AlgebricksHyracksIntegrationUtil.NC1_ID, inputFileName[0]) };
    IFileSplitProvider intSplitProvider = new ConstantFileSplitProvider(inputSplits);
    RecordDescriptor scannerDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer() });
    IValueParserFactory[] valueParsers = new IValueParserFactory[] { IntegerParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE };
    FileScanOperatorDescriptor intScanner = new FileScanOperatorDescriptor(spec, intSplitProvider, new DelimitedDataTupleParserFactory(valueParsers, '|'), scannerDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, intScanner, DEFAULT_NODES);
    SplitOperatorDescriptor splitOp = new SplitOperatorDescriptor(spec, scannerDesc, outputArity, new TupleFieldEvaluatorFactory(0), BinaryIntegerInspectorImpl.FACTORY);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, splitOp, new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    IOperatorDescriptor[] outputOp = new IOperatorDescriptor[outputFile.length];
    for (int i = 0; i < outputArity; i++) {
        outputOp[i] = new LineFileWriteOperatorDescriptor(spec, new FileSplit[] { outputFileSplit[i] });
        PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, outputOp[i], new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    }
    spec.connect(new OneToOneConnectorDescriptor(spec), intScanner, 0, splitOp, 0);
    for (int i = 0; i < outputArity; i++) {
        spec.connect(new OneToOneConnectorDescriptor(spec), splitOp, i, outputOp[i], 0);
    }
    for (int i = 0; i < outputArity; i++) {
        spec.addRoot(outputOp[i]);
    }
    AlgebricksHyracksIntegrationUtil.runJob(spec);
    for (int i = 0; i < outputArity; i++) {
        compareFiles("data" + File.separator + "device0" + File.separator + inputFileName[i + 1], outputFile[i].getAbsolutePath());
    }
}

Example 22 with IFileSplitProvider

use of org.apache.hyracks.dataflow.std.file.IFileSplitProvider in project asterixdb by apache.

the class PushRuntimeTest method scanLimitWrite.

@Test
public void scanLimitWrite() throws Exception {
    JobSpecification spec = new JobSpecification(FRAME_SIZE);
    // the scanner
    FileSplit[] fileSplits = new FileSplit[1];
    fileSplits[0] = new ManagedFileSplit(AlgebricksHyracksIntegrationUtil.NC1_ID, "data" + File.separator + "tpch0.001" + File.separator + "customer.tbl");
    IFileSplitProvider splitProvider = new ConstantFileSplitProvider(fileSplits);
    RecordDescriptor scannerDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), FloatSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    IValueParserFactory[] valueParsers = new IValueParserFactory[] { IntegerParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, IntegerParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, FloatParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE };
    FileScanOperatorDescriptor scanner = new FileScanOperatorDescriptor(spec, splitProvider, new DelimitedDataTupleParserFactory(valueParsers, '|'), scannerDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, scanner, DEFAULT_NODES);
    // the algebricks op.
    StreamLimitRuntimeFactory limit = new StreamLimitRuntimeFactory(new IntegerConstantEvalFactory(2), null, new int[] { 0 }, BinaryIntegerInspectorImpl.FACTORY);
    RecordDescriptor limitDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE });
    String filePath = PATH_ACTUAL + SEPARATOR + "scanLimitWrite.out";
    File outFile = new File(filePath);
    SinkWriterRuntimeFactory writer = new SinkWriterRuntimeFactory(new int[] { 0 }, new IPrinterFactory[] { IntegerPrinterFactory.INSTANCE }, outFile, PrinterBasedWriterFactory.INSTANCE, limitDesc);
    AlgebricksMetaOperatorDescriptor algebricksOp = new AlgebricksMetaOperatorDescriptor(spec, 1, 0, new IPushRuntimeFactory[] { limit, writer }, new RecordDescriptor[] { limitDesc, null });
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, algebricksOp, new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    spec.connect(new OneToOneConnectorDescriptor(spec), scanner, 0, algebricksOp, 0);
    spec.addRoot(algebricksOp);
    AlgebricksHyracksIntegrationUtil.runJob(spec);
    StringBuilder buf = new StringBuilder();
    readFileToString(outFile, buf);
    Assert.assertEquals("12", buf.toString());
    outFile.delete();
}

Example 23 with IFileSplitProvider

use of org.apache.hyracks.dataflow.std.file.IFileSplitProvider in project asterixdb by apache.

the class PushRuntimeTest method scanSortGbySelectWrite.

@Test
public void scanSortGbySelectWrite() throws Exception {
    JobSpecification spec = new JobSpecification(FRAME_SIZE);
    // the scanner
    FileSplit[] fileSplits = new FileSplit[1];
    fileSplits[0] = new ManagedFileSplit(AlgebricksHyracksIntegrationUtil.NC1_ID, "data" + File.separator + "tpch0.001" + File.separator + "customer.tbl");
    IFileSplitProvider splitProvider = new ConstantFileSplitProvider(fileSplits);
    RecordDescriptor scannerDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), FloatSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    IValueParserFactory[] valueParsers = new IValueParserFactory[] { IntegerParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, IntegerParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, FloatParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE };
    FileScanOperatorDescriptor scanner = new FileScanOperatorDescriptor(spec, splitProvider, new DelimitedDataTupleParserFactory(valueParsers, '|'), scannerDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, scanner, new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    // the sort (by nation id)
    RecordDescriptor sortDesc = scannerDesc;
    InMemorySortOperatorDescriptor sort = new InMemorySortOperatorDescriptor(spec, new int[] { 3 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, sortDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, sort, new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    // the group-by
    NestedTupleSourceRuntimeFactory nts = new NestedTupleSourceRuntimeFactory();
    RecordDescriptor ntsDesc = sortDesc;
    AggregateRuntimeFactory agg = new AggregateRuntimeFactory(new IAggregateEvaluatorFactory[] { new TupleCountAggregateFunctionFactory() });
    RecordDescriptor aggDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE });
    AlgebricksPipeline pipeline = new AlgebricksPipeline(new IPushRuntimeFactory[] { nts, agg }, new RecordDescriptor[] { ntsDesc, aggDesc });
    NestedPlansAccumulatingAggregatorFactory npaaf = new NestedPlansAccumulatingAggregatorFactory(new AlgebricksPipeline[] { pipeline }, new int[] { 3 }, new int[] {});
    RecordDescriptor gbyDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE, IntegerSerializerDeserializer.INSTANCE });
    PreclusteredGroupOperatorDescriptor gby = new PreclusteredGroupOperatorDescriptor(spec, new int[] { 3 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, npaaf, gbyDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, gby, new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    // the algebricks op.
    IScalarEvaluatorFactory cond = new IntegerEqualsEvalFactory(new IntegerConstantEvalFactory(3), // Canadian customers
    new TupleFieldEvaluatorFactory(0));
    StreamSelectRuntimeFactory select = new StreamSelectRuntimeFactory(cond, new int[] { 1 }, BinaryBooleanInspectorImpl.FACTORY, false, -1, null);
    RecordDescriptor selectDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE });
    String filePath = PATH_ACTUAL + SEPARATOR + "scanSortGbySelectWrite.out";
    File outFile = new File(filePath);
    SinkWriterRuntimeFactory writer = new SinkWriterRuntimeFactory(new int[] { 0 }, new IPrinterFactory[] { IntegerPrinterFactory.INSTANCE }, outFile, PrinterBasedWriterFactory.INSTANCE, selectDesc);
    AlgebricksMetaOperatorDescriptor algebricksOp = new AlgebricksMetaOperatorDescriptor(spec, 1, 0, new IPushRuntimeFactory[] { select, writer }, new RecordDescriptor[] { selectDesc, null });
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, algebricksOp, new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    spec.connect(new OneToOneConnectorDescriptor(spec), scanner, 0, sort, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), sort, 0, gby, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), gby, 0, algebricksOp, 0);
    spec.addRoot(algebricksOp);
    AlgebricksHyracksIntegrationUtil.runJob(spec);
    StringBuilder buf = new StringBuilder();
    readFileToString(outFile, buf);
    Assert.assertEquals("9", buf.toString());
    outFile.delete();
}

Example 24 with IFileSplitProvider

use of org.apache.hyracks.dataflow.std.file.IFileSplitProvider in project asterixdb by apache.

the class SecondaryTreeIndexOperationsHelper method buildCompactJobSpec.

@Override
public JobSpecification buildCompactJobSpec() throws AlgebricksException {
    JobSpecification spec = RuntimeUtils.createJobSpecification(metadataProvider.getApplicationContext());
    Pair<IFileSplitProvider, AlgebricksPartitionConstraint> splitsAndConstraint = metadataProvider.getSplitProviderAndConstraints(dataset, index.getIndexName());
    IIndexDataflowHelperFactory dataflowHelperFactory = new IndexDataflowHelperFactory(metadataProvider.getStorageComponentProvider().getStorageManager(), splitsAndConstraint.first);
    LSMTreeIndexCompactOperatorDescriptor compactOp = new LSMTreeIndexCompactOperatorDescriptor(spec, dataflowHelperFactory);
    AlgebricksPartitionConstraintHelper.setPartitionConstraintInJobSpec(spec, compactOp, secondaryPartitionConstraint);
    spec.addRoot(compactOp);
    spec.setConnectorPolicyAssignmentPolicy(new ConnectorPolicyAssignmentPolicy());
    return spec;
}

Example 25 with IFileSplitProvider

use of org.apache.hyracks.dataflow.std.file.IFileSplitProvider in project asterixdb by apache.

the class LSMRTreeWithAntiMatterTuplesSecondaryIndexSearchOperatorTest method shouldWriteFilterValueIfAppendFilterIsTrue.

@Test
public void shouldWriteFilterValueIfAppendFilterIsTrue() throws Exception {
    JobSpecification spec = new JobSpecification();
    // build tuple
    ArrayTupleBuilder tb = new ArrayTupleBuilder(secondaryKeyFieldCount);
    DataOutput dos = tb.getDataOutput();
    tb.reset();
    DoubleSerializerDeserializer.INSTANCE.serialize(61.2894, dos);
    tb.addFieldEndOffset();
    DoubleSerializerDeserializer.INSTANCE.serialize(-149.624, dos);
    tb.addFieldEndOffset();
    DoubleSerializerDeserializer.INSTANCE.serialize(61.8894, dos);
    tb.addFieldEndOffset();
    DoubleSerializerDeserializer.INSTANCE.serialize(-149.024, dos);
    tb.addFieldEndOffset();
    ISerializerDeserializer[] keyRecDescSers = { DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE };
    RecordDescriptor keyRecDesc = new RecordDescriptor(keyRecDescSers);
    ConstantTupleSourceOperatorDescriptor keyProviderOp = new ConstantTupleSourceOperatorDescriptor(spec, keyRecDesc, tb.getFieldEndOffsets(), tb.getByteArray(), tb.getSize());
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, keyProviderOp, NC1_ID);
    int[] keyFields = { 0, 1, 2, 3 };
    RTreeSearchOperatorDescriptor secondarySearchOp = new RTreeSearchOperatorDescriptor(spec, secondaryWithFilterRecDesc, keyFields, true, true, secondaryHelperFactory, false, false, null, NoOpOperationCallbackFactory.INSTANCE, null, null, false);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, secondarySearchOp, NC1_ID);
    IFileSplitProvider outSplits = new ConstantFileSplitProvider(new FileSplit[] { createFile(nc1) });
    IOperatorDescriptor printer = new PlainFileWriterOperatorDescriptor(spec, outSplits, ",");
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);
    spec.connect(new OneToOneConnectorDescriptor(spec), keyProviderOp, 0, secondarySearchOp, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), secondarySearchOp, 0, printer, 0);
    spec.addRoot(printer);
    runTest(spec);
}