Search in sources :

Example 1 with DelimitedDataTupleParserFactory

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

the class PigletMetadataProvider method getScannerRuntime.

@Override
public Pair<IOperatorDescriptor, AlgebricksPartitionConstraint> getScannerRuntime(IDataSource<String> dataSource, List<LogicalVariable> scanVariables, List<LogicalVariable> projectVariables, boolean projectPushed, List<LogicalVariable> minFilterVars, List<LogicalVariable> maxFilterVars, IOperatorSchema opSchema, IVariableTypeEnvironment typeEnv, JobGenContext context, JobSpecification jobSpec, Object implConfig) throws AlgebricksException {
    PigletFileDataSource ds = (PigletFileDataSource) dataSource;
    FileSplit[] fileSplits = ds.getFileSplits();
    String[] locations = new String[fileSplits.length];
    for (int i = 0; i < fileSplits.length; ++i) {
        locations[i] = fileSplits[i].getNodeName();
    }
    IFileSplitProvider fsp = new ConstantFileSplitProvider(fileSplits);
    Object[] colTypes = ds.getSchemaTypes();
    IValueParserFactory[] vpfs = new IValueParserFactory[colTypes.length];
    ISerializerDeserializer[] serDesers = new ISerializerDeserializer[colTypes.length];
    for (int i = 0; i < colTypes.length; ++i) {
        Type colType = (Type) colTypes[i];
        IValueParserFactory vpf;
        ISerializerDeserializer serDeser;
        switch(colType.getTag()) {
            case INTEGER:
                vpf = IntegerParserFactory.INSTANCE;
                serDeser = IntegerSerializerDeserializer.INSTANCE;
                break;
            case CHAR_ARRAY:
                vpf = UTF8StringParserFactory.INSTANCE;
                serDeser = new UTF8StringSerializerDeserializer();
                break;
            case FLOAT:
                vpf = FloatParserFactory.INSTANCE;
                serDeser = FloatSerializerDeserializer.INSTANCE;
                break;
            default:
                throw new UnsupportedOperationException();
        }
        vpfs[i] = vpf;
        serDesers[i] = serDeser;
    }
    ITupleParserFactory tpf = new DelimitedDataTupleParserFactory(vpfs, ',');
    RecordDescriptor rDesc = new RecordDescriptor(serDesers);
    IOperatorDescriptor scanner = new FileScanOperatorDescriptor(jobSpec, fsp, tpf, rDesc);
    AlgebricksAbsolutePartitionConstraint constraint = new AlgebricksAbsolutePartitionConstraint(locations);
    return new Pair<>(scanner, constraint);
}
Also used : IValueParserFactory(org.apache.hyracks.dataflow.common.data.parsers.IValueParserFactory) IFileSplitProvider(org.apache.hyracks.dataflow.std.file.IFileSplitProvider) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) ConstantFileSplitProvider(org.apache.hyracks.dataflow.std.file.ConstantFileSplitProvider) DelimitedDataTupleParserFactory(org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory) FileSplit(org.apache.hyracks.api.io.FileSplit) UTF8StringSerializerDeserializer(org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer) AlgebricksPartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksPartitionConstraint) AlgebricksAbsolutePartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksAbsolutePartitionConstraint) ISerializerDeserializer(org.apache.hyracks.api.dataflow.value.ISerializerDeserializer) Type(org.apache.hyracks.algebricks.examples.piglet.types.Type) ITupleParserFactory(org.apache.hyracks.dataflow.std.file.ITupleParserFactory) AlgebricksAbsolutePartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksAbsolutePartitionConstraint) IOperatorDescriptor(org.apache.hyracks.api.dataflow.IOperatorDescriptor) FileScanOperatorDescriptor(org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor) Pair(org.apache.hyracks.algebricks.common.utils.Pair)

Example 2 with DelimitedDataTupleParserFactory

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

the class AbstractBTreeOperatorTest method loadPrimaryIndex.

protected void loadPrimaryIndex() throws Exception {
    JobSpecification spec = new JobSpecification();
    FileSplit[] ordersSplits = new FileSplit[] { new ManagedFileSplit(NC1_ID, "data" + File.separator + "tpch0.001" + File.separator + "orders-part1.tbl") };
    IFileSplitProvider ordersSplitProvider = new ConstantFileSplitProvider(ordersSplits);
    RecordDescriptor ordersDesc = inputRecordDesc;
    FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitProvider, new DelimitedDataTupleParserFactory(inputParserFactories, '|'), ordersDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, ordScanner, NC1_ID);
    ExternalSortOperatorDescriptor sorter = new ExternalSortOperatorDescriptor(spec, 1000, new int[] { 0 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, ordersDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, sorter, NC1_ID);
    int[] fieldPermutation = { 0, 1, 2, 4, 5, 7 };
    TreeIndexBulkLoadOperatorDescriptor primaryBtreeBulkLoad = new TreeIndexBulkLoadOperatorDescriptor(spec, primaryRecDesc, fieldPermutation, 0.7f, true, 1000L, true, primaryHelperFactory);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, primaryBtreeBulkLoad, NC1_ID);
    NullSinkOperatorDescriptor nsOpDesc = new NullSinkOperatorDescriptor(spec);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, nsOpDesc, NC1_ID);
    spec.connect(new OneToOneConnectorDescriptor(spec), ordScanner, 0, sorter, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), sorter, 0, primaryBtreeBulkLoad, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), primaryBtreeBulkLoad, 0, nsOpDesc, 0);
    spec.addRoot(nsOpDesc);
    runTest(spec);
}
Also used : NullSinkOperatorDescriptor(org.apache.hyracks.dataflow.std.misc.NullSinkOperatorDescriptor) IFileSplitProvider(org.apache.hyracks.dataflow.std.file.IFileSplitProvider) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) ConstantFileSplitProvider(org.apache.hyracks.dataflow.std.file.ConstantFileSplitProvider) DelimitedDataTupleParserFactory(org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory) OneToOneConnectorDescriptor(org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor) FileSplit(org.apache.hyracks.api.io.FileSplit) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) FileScanOperatorDescriptor(org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor) ExternalSortOperatorDescriptor(org.apache.hyracks.dataflow.std.sort.ExternalSortOperatorDescriptor) JobSpecification(org.apache.hyracks.api.job.JobSpecification) TreeIndexBulkLoadOperatorDescriptor(org.apache.hyracks.storage.am.common.dataflow.TreeIndexBulkLoadOperatorDescriptor)

Example 3 with DelimitedDataTupleParserFactory

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

the class AbstractBTreeOperatorTest method insertPipeline.

protected void insertPipeline(boolean useUpsert) throws Exception {
    IndexOperation pipelineOperation = useUpsert ? IndexOperation.UPSERT : IndexOperation.INSERT;
    JobSpecification spec = new JobSpecification();
    FileSplit[] ordersSplits = new FileSplit[] { new ManagedFileSplit(NC1_ID, "data" + File.separator + "tpch0.002" + File.separator + "orders-part2.tbl") };
    IFileSplitProvider ordersSplitProvider = new ConstantFileSplitProvider(ordersSplits);
    RecordDescriptor ordersDesc = inputRecordDesc;
    FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitProvider, new DelimitedDataTupleParserFactory(inputParserFactories, '|'), ordersDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, ordScanner, NC1_ID);
    // insert into primary index
    TreeIndexInsertUpdateDeleteOperatorDescriptor primaryBtreeInsertOp = new TreeIndexInsertUpdateDeleteOperatorDescriptor(spec, ordersDesc, primaryFieldPermutation, pipelineOperation, primaryHelperFactory, null, NoOpOperationCallbackFactory.INSTANCE);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, primaryBtreeInsertOp, NC1_ID);
    // first secondary index
    int[] fieldPermutationB = secondaryFieldPermutationB;
    TreeIndexInsertUpdateDeleteOperatorDescriptor secondaryInsertOp = new TreeIndexInsertUpdateDeleteOperatorDescriptor(spec, ordersDesc, fieldPermutationB, pipelineOperation, secondaryHelperFactory, null, NoOpOperationCallbackFactory.INSTANCE);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, secondaryInsertOp, NC1_ID);
    NullSinkOperatorDescriptor nullSink = new NullSinkOperatorDescriptor(spec);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, nullSink, NC1_ID);
    spec.connect(new OneToOneConnectorDescriptor(spec), ordScanner, 0, primaryBtreeInsertOp, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), primaryBtreeInsertOp, 0, secondaryInsertOp, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), secondaryInsertOp, 0, nullSink, 0);
    spec.addRoot(nullSink);
    runTest(spec);
}
Also used : NullSinkOperatorDescriptor(org.apache.hyracks.dataflow.std.misc.NullSinkOperatorDescriptor) IFileSplitProvider(org.apache.hyracks.dataflow.std.file.IFileSplitProvider) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) ConstantFileSplitProvider(org.apache.hyracks.dataflow.std.file.ConstantFileSplitProvider) DelimitedDataTupleParserFactory(org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory) OneToOneConnectorDescriptor(org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor) FileSplit(org.apache.hyracks.api.io.FileSplit) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) IndexOperation(org.apache.hyracks.storage.am.common.ophelpers.IndexOperation) TreeIndexInsertUpdateDeleteOperatorDescriptor(org.apache.hyracks.storage.am.common.dataflow.TreeIndexInsertUpdateDeleteOperatorDescriptor) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) FileScanOperatorDescriptor(org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor) JobSpecification(org.apache.hyracks.api.job.JobSpecification)

Example 4 with DelimitedDataTupleParserFactory

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

the class PushRuntimeTest method scanSelectWrite.

@Test
public void scanSelectWrite() throws Exception {
    JobSpecification spec = new JobSpecification(FRAME_SIZE);
    // the scanner
    FileSplit[] intFileSplits = new FileSplit[1];
    intFileSplits[0] = new ManagedFileSplit(AlgebricksHyracksIntegrationUtil.NC1_ID, "data" + File.separator + "simple" + File.separator + "int-part1.tbl");
    IFileSplitProvider intSplitProvider = new ConstantFileSplitProvider(intFileSplits);
    RecordDescriptor intScannerDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE });
    IValueParserFactory[] valueParsers = new IValueParserFactory[] { IntegerParserFactory.INSTANCE };
    FileScanOperatorDescriptor intScanner = new FileScanOperatorDescriptor(spec, intSplitProvider, new DelimitedDataTupleParserFactory(valueParsers, '|'), intScannerDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, intScanner, DEFAULT_NODES);
    // the algebricks op.
    IScalarEvaluatorFactory cond = new IntegerGreaterThanEvalFactory(new IntegerConstantEvalFactory(2), new TupleFieldEvaluatorFactory(0));
    StreamSelectRuntimeFactory select = new StreamSelectRuntimeFactory(cond, new int[] { 0 }, BinaryBooleanInspectorImpl.FACTORY, false, -1, null);
    RecordDescriptor selectDesc = intScannerDesc;
    String filePath = PATH_ACTUAL + SEPARATOR + "scanSelectWrite.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.addPartitionCountConstraint(spec, algebricksOp, 1);
    spec.connect(new OneToOneConnectorDescriptor(spec), intScanner, 0, algebricksOp, 0);
    spec.addRoot(algebricksOp);
    AlgebricksHyracksIntegrationUtil.runJob(spec);
    StringBuilder buf = new StringBuilder();
    readFileToString(outFile, buf);
    Assert.assertEquals("0", buf.toString());
    outFile.delete();
}
Also used : TupleFieldEvaluatorFactory(org.apache.hyracks.algebricks.runtime.evaluators.TupleFieldEvaluatorFactory) IValueParserFactory(org.apache.hyracks.dataflow.common.data.parsers.IValueParserFactory) IFileSplitProvider(org.apache.hyracks.dataflow.std.file.IFileSplitProvider) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) ConstantFileSplitProvider(org.apache.hyracks.dataflow.std.file.ConstantFileSplitProvider) AlgebricksMetaOperatorDescriptor(org.apache.hyracks.algebricks.runtime.operators.meta.AlgebricksMetaOperatorDescriptor) DelimitedDataTupleParserFactory(org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory) OneToOneConnectorDescriptor(org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor) FileSplit(org.apache.hyracks.api.io.FileSplit) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) IScalarEvaluatorFactory(org.apache.hyracks.algebricks.runtime.base.IScalarEvaluatorFactory) StreamSelectRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.std.StreamSelectRuntimeFactory) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) SinkWriterRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.std.SinkWriterRuntimeFactory) FileScanOperatorDescriptor(org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor) JobSpecification(org.apache.hyracks.api.job.JobSpecification) File(java.io.File) Test(org.junit.Test)

Example 5 with DelimitedDataTupleParserFactory

use of org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory 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());
    }
}
Also used : TupleFieldEvaluatorFactory(org.apache.hyracks.algebricks.runtime.evaluators.TupleFieldEvaluatorFactory) IValueParserFactory(org.apache.hyracks.dataflow.common.data.parsers.IValueParserFactory) SplitOperatorDescriptor(org.apache.hyracks.algebricks.runtime.operators.std.SplitOperatorDescriptor) IFileSplitProvider(org.apache.hyracks.dataflow.std.file.IFileSplitProvider) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) ConstantFileSplitProvider(org.apache.hyracks.dataflow.std.file.ConstantFileSplitProvider) DelimitedDataTupleParserFactory(org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory) OneToOneConnectorDescriptor(org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor) FileSplit(org.apache.hyracks.api.io.FileSplit) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) UTF8StringSerializerDeserializer(org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer) LineFileWriteOperatorDescriptor(org.apache.hyracks.dataflow.std.file.LineFileWriteOperatorDescriptor) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) IOperatorDescriptor(org.apache.hyracks.api.dataflow.IOperatorDescriptor) FileScanOperatorDescriptor(org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor) JobSpecification(org.apache.hyracks.api.job.JobSpecification) File(java.io.File) Test(org.junit.Test)

Aggregations

DelimitedDataTupleParserFactory (org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory)46 JobSpecification (org.apache.hyracks.api.job.JobSpecification)45 ConstantFileSplitProvider (org.apache.hyracks.dataflow.std.file.ConstantFileSplitProvider)45 FileScanOperatorDescriptor (org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor)45 IFileSplitProvider (org.apache.hyracks.dataflow.std.file.IFileSplitProvider)43 RecordDescriptor (org.apache.hyracks.api.dataflow.value.RecordDescriptor)41 ManagedFileSplit (org.apache.hyracks.api.io.ManagedFileSplit)41 FileSplit (org.apache.hyracks.api.io.FileSplit)39 OneToOneConnectorDescriptor (org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor)39 UTF8StringSerializerDeserializer (org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer)36 IValueParserFactory (org.apache.hyracks.dataflow.common.data.parsers.IValueParserFactory)36 Test (org.junit.Test)34 IOperatorDescriptor (org.apache.hyracks.api.dataflow.IOperatorDescriptor)33 IConnectorDescriptor (org.apache.hyracks.api.dataflow.IConnectorDescriptor)24 ResultSetId (org.apache.hyracks.api.dataset.ResultSetId)24 ResultWriterOperatorDescriptor (org.apache.hyracks.dataflow.std.result.ResultWriterOperatorDescriptor)23 FieldHashPartitionComputerFactory (org.apache.hyracks.dataflow.common.data.partition.FieldHashPartitionComputerFactory)18 MToNBroadcastConnectorDescriptor (org.apache.hyracks.dataflow.std.connectors.MToNBroadcastConnectorDescriptor)18 IBinaryHashFunctionFactory (org.apache.hyracks.api.dataflow.value.IBinaryHashFunctionFactory)12 MToNPartitioningConnectorDescriptor (org.apache.hyracks.dataflow.std.connectors.MToNPartitioningConnectorDescriptor)12