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

Example 16 with PlainFileWriterOperatorDescriptor

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

the class Groupby method createJob.

private static JobSpecification createJob(FileSplit[] inSplits, FileSplit[] outSplits, int htSize, long fileSize, int frameLimit, int frameSize, String alg, boolean outPlain) {
    JobSpecification spec = new JobSpecification(frameSize);
    IFileSplitProvider splitsProvider = new ConstantFileSplitProvider(inSplits);
    FileScanOperatorDescriptor fileScanner = new FileScanOperatorDescriptor(spec, splitsProvider, new DelimitedDataTupleParserFactory(lineitemParserFactories, '|'), lineitemDesc);
    createPartitionConstraint(spec, fileScanner, inSplits);
    // Output: each unique string with an integer count
    RecordDescriptor outDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE, // IntegerSerializerDeserializer.INSTANCE,
    IntegerSerializerDeserializer.INSTANCE });
    // Specify the grouping key, which will be the string extracted during
    // the scan.
    int[] keys = new int[] { 0 };
    AbstractOperatorDescriptor grouper;
    if (alg.equalsIgnoreCase("hash")) {
        // external hash graph
        grouper = new ExternalGroupOperatorDescriptor(spec, htSize, fileSize, keys, frameLimit, new IBinaryComparatorFactory[] { // PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY),
        PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, new IntegerNormalizedKeyComputerFactory(), new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new CountFieldAggregatorFactory(false) }), new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(keys.length, false) }), outDesc, outDesc, new HashSpillableTableFactory(new IBinaryHashFunctionFamily[] { MurmurHash3BinaryHashFunctionFamily.INSTANCE }));
        createPartitionConstraint(spec, grouper, outSplits);
    } else if (alg.equalsIgnoreCase("sort")) {
        grouper = new SortGroupByOperatorDescriptor(spec, frameLimit, keys, keys, new IntegerNormalizedKeyComputerFactory(), new IBinaryComparatorFactory[] { // PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY),
        PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new CountFieldAggregatorFactory(true) }), new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(keys.length, true) }), outDesc, outDesc, false);
        createPartitionConstraint(spec, grouper, outSplits);
    } else {
        System.err.println("unknow groupby alg:" + alg);
        return null;
    }
    // Connect scanner with the grouper
    IConnectorDescriptor scanGroupConnDef2 = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(keys, new IBinaryHashFunctionFactory[] { // PointableBinaryHashFunctionFactory.of(IntegerPointable.FACTORY),
    PointableBinaryHashFunctionFactory.of(IntegerPointable.FACTORY) }));
    spec.connect(scanGroupConnDef2, fileScanner, 0, grouper, 0);
    IFileSplitProvider outSplitProvider = new ConstantFileSplitProvider(outSplits);
    AbstractSingleActivityOperatorDescriptor writer = outPlain ? new PlainFileWriterOperatorDescriptor(spec, outSplitProvider, "|") : new FrameFileWriterOperatorDescriptor(spec, outSplitProvider);
    createPartitionConstraint(spec, writer, outSplits);
    IConnectorDescriptor groupOutConn = new OneToOneConnectorDescriptor(spec);
    spec.connect(groupOutConn, grouper, 0, writer, 0);
    spec.addRoot(writer);
    return spec;
}

Example 17 with PlainFileWriterOperatorDescriptor

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

the class Sort method createJob.

private static JobSpecification createJob(FileSplit[] ordersSplits, FileSplit[] outputSplit, String memBufferAlg, int frameLimit, int frameSize, int limit, boolean usingHeapSorter) {
    JobSpecification spec = new JobSpecification();
    spec.setFrameSize(frameSize);
    IFileSplitProvider ordersSplitProvider = new ConstantFileSplitProvider(ordersSplits);
    FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitProvider, new DelimitedDataTupleParserFactory(orderParserFactories, '|'), ordersDesc);
    createPartitionConstraint(spec, ordScanner, ordersSplits);
    AbstractSorterOperatorDescriptor sorter;
    if (usingHeapSorter && limit < Integer.MAX_VALUE) {
        sorter = new TopKSorterOperatorDescriptor(spec, frameLimit, limit, SortFields, null, SortFieldsComparatorFactories, ordersDesc);
    } else {
        if (memBufferAlg.equalsIgnoreCase("bestfit")) {
            sorter = new ExternalSortOperatorDescriptor(spec, frameLimit, SortFields, null, SortFieldsComparatorFactories, ordersDesc, Algorithm.MERGE_SORT, EnumFreeSlotPolicy.SMALLEST_FIT, limit);
        } else if (memBufferAlg.equalsIgnoreCase("biggestfit")) {
            sorter = new ExternalSortOperatorDescriptor(spec, frameLimit, SortFields, null, SortFieldsComparatorFactories, ordersDesc, Algorithm.MERGE_SORT, EnumFreeSlotPolicy.BIGGEST_FIT, limit);
        } else {
            sorter = new ExternalSortOperatorDescriptor(spec, frameLimit, SortFields, null, SortFieldsComparatorFactories, ordersDesc, Algorithm.MERGE_SORT, EnumFreeSlotPolicy.LAST_FIT, limit);
        }
    }
    createPartitionConstraint(spec, sorter, ordersSplits);
    IFileSplitProvider outputSplitProvider = new ConstantFileSplitProvider(outputSplit);
    IOperatorDescriptor printer = new PlainFileWriterOperatorDescriptor(spec, outputSplitProvider, "|");
    createPartitionConstraint(spec, printer, outputSplit);
    spec.connect(new OneToOneConnectorDescriptor(spec), ordScanner, 0, sorter, 0);
    spec.connect(new MToNPartitioningMergingConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(SortFields, orderBinaryHashFunctionFactories), SortFields, SortFieldsComparatorFactories, new UTF8StringNormalizedKeyComputerFactory()), sorter, 0, printer, 0);
    spec.addRoot(printer);
    return spec;
}

Example 18 with PlainFileWriterOperatorDescriptor

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

the class RTreeSecondaryIndexInsertOperatorTest method searchUpdatedSecondaryIndexTest.

@Test
public void searchUpdatedSecondaryIndexTest() 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, secondaryRecDesc, keyFields, true, true, secondaryHelperFactory, false, false, null, NoOpOperationCallbackFactory.INSTANCE, null, null, false);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, secondarySearchOp, NC1_ID);
    // fifth field from the tuples coming from secondary index
    int[] primaryLowKeyFields = { 4 };
    // fifth field from the tuples coming from secondary index
    int[] primaryHighKeyFields = { 4 };
    // search primary index
    BTreeSearchOperatorDescriptor primarySearchOp = new BTreeSearchOperatorDescriptor(spec, primaryRecDesc, primaryLowKeyFields, primaryHighKeyFields, true, true, primaryHelperFactory, false, false, null, NoOpOperationCallbackFactory.INSTANCE, null, null, false);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, primarySearchOp, 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, primarySearchOp, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), primarySearchOp, 0, printer, 0);
    spec.addRoot(printer);
    runTest(spec);
}

Example 19 with PlainFileWriterOperatorDescriptor

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

the class RTreeSecondaryIndexScanOperatorTest method scanPrimaryIndexTest.

@Test
public void scanPrimaryIndexTest() throws Exception {
    JobSpecification spec = new JobSpecification();
    // build dummy tuple
    ArrayTupleBuilder tb = new ArrayTupleBuilder(secondaryKeyFieldCount);
    DataOutput dos = tb.getDataOutput();
    tb.reset();
    DoubleSerializerDeserializer.INSTANCE.serialize(0.0, dos);
    tb.addFieldEndOffset();
    DoubleSerializerDeserializer.INSTANCE.serialize(0.0, dos);
    tb.addFieldEndOffset();
    DoubleSerializerDeserializer.INSTANCE.serialize(0.0, dos);
    tb.addFieldEndOffset();
    DoubleSerializerDeserializer.INSTANCE.serialize(0.0, 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 = null;
    RTreeSearchOperatorDescriptor secondarySearchOp = new RTreeSearchOperatorDescriptor(spec, secondaryRecDesc, 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);
}

Example 20 with PlainFileWriterOperatorDescriptor

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

the class RTreeSecondaryIndexStatsOperatorTest method showPrimaryIndexStats.

@Test
public void showPrimaryIndexStats() throws Exception {
    JobSpecification spec = new JobSpecification();
    TreeIndexStatsOperatorDescriptor secondaryStatsOp = new TreeIndexStatsOperatorDescriptor(spec, storageManager, secondaryHelperFactory);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, secondaryStatsOp, 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), secondaryStatsOp, 0, printer, 0);
    spec.addRoot(printer);
    runTest(spec);
}