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

Example 31 with FileScanOperatorDescriptor

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

the class TPCHCustomerOrderHashJoinTest method customerOrderCIDInMemoryHashLeftOuterJoin.

@Test
public void customerOrderCIDInMemoryHashLeftOuterJoin() throws Exception {
    JobSpecification spec = new JobSpecification();
    FileSplit[] custSplits = new FileSplit[] { new ManagedFileSplit(NC1_ID, "data" + File.separator + "tpch0.001" + File.separator + "customer.tbl") };
    IFileSplitProvider custSplitsProvider = new ConstantFileSplitProvider(custSplits);
    RecordDescriptor custDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    FileSplit[] ordersSplits = new FileSplit[] { new ManagedFileSplit(NC2_ID, "data" + File.separator + "tpch0.001" + File.separator + "orders.tbl") };
    IFileSplitProvider ordersSplitsProvider = new ConstantFileSplitProvider(ordersSplits);
    RecordDescriptor ordersDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    RecordDescriptor custOrderJoinDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitsProvider, 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, NC2_ID);
    FileScanOperatorDescriptor custScanner = new FileScanOperatorDescriptor(spec, custSplitsProvider, new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE }, '|'), custDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, custScanner, NC1_ID);
    IMissingWriterFactory[] nonMatchWriterFactories = new IMissingWriterFactory[ordersDesc.getFieldCount()];
    for (int j = 0; j < nonMatchWriterFactories.length; j++) {
        nonMatchWriterFactories[j] = NoopMissingWriterFactory.INSTANCE;
    }
    InMemoryHashJoinOperatorDescriptor join = new InMemoryHashJoinOperatorDescriptor(spec, new int[] { 0 }, new int[] { 1 }, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, null, custOrderJoinDesc, true, nonMatchWriterFactories, 128, 128);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, join, NC1_ID);
    ResultSetId rsId = new ResultSetId(1);
    spec.addResultSetId(rsId);
    IOperatorDescriptor printer = new ResultWriterOperatorDescriptor(spec, rsId, false, false, ResultSerializerFactoryProvider.INSTANCE.getResultSerializerFactoryProvider());
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);
    IConnectorDescriptor ordJoinConn = new MToNBroadcastConnectorDescriptor(spec);
    spec.connect(ordJoinConn, ordScanner, 0, join, 1);
    IConnectorDescriptor custJoinConn = new OneToOneConnectorDescriptor(spec);
    spec.connect(custJoinConn, custScanner, 0, join, 0);
    IConnectorDescriptor joinPrinterConn = new OneToOneConnectorDescriptor(spec);
    spec.connect(joinPrinterConn, join, 0, printer, 0);
    spec.addRoot(printer);
    runTest(spec);
}

Example 32 with FileScanOperatorDescriptor

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

the class TPCHCustomerOrderNestedLoopJoinTest method customerOrderCIDOuterJoinMulti.

@Test
public void customerOrderCIDOuterJoinMulti() throws Exception {
    JobSpecification spec = new JobSpecification();
    FileSplit[] custSplits = new FileSplit[] { new ManagedFileSplit(NC1_ID, "data" + File.separator + "tpch0.001" + File.separator + "customer-part1.tbl"), new ManagedFileSplit(NC2_ID, "data" + File.separator + "tpch0.001" + File.separator + "customer-part2.tbl") };
    IFileSplitProvider custSplitsProvider = new ConstantFileSplitProvider(custSplits);
    RecordDescriptor custDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    FileSplit[] ordersSplits = new FileSplit[] { new ManagedFileSplit(NC1_ID, "data" + File.separator + "tpch0.001" + File.separator + "orders-part1.tbl"), new ManagedFileSplit(NC2_ID, "data" + File.separator + "tpch0.001" + File.separator + "orders-part2.tbl") };
    IFileSplitProvider ordersSplitsProvider = new ConstantFileSplitProvider(ordersSplits);
    RecordDescriptor ordersDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    RecordDescriptor custOrderJoinDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() });
    FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitsProvider, 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);
    FileScanOperatorDescriptor custScanner = new FileScanOperatorDescriptor(spec, custSplitsProvider, new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE }, '|'), custDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, custScanner, NC1_ID, NC2_ID);
    IMissingWriterFactory[] nonMatchWriterFactories = new IMissingWriterFactory[ordersDesc.getFieldCount()];
    for (int j = 0; j < nonMatchWriterFactories.length; j++) {
        nonMatchWriterFactories[j] = NoopMissingWriterFactory.INSTANCE;
    }
    NestedLoopJoinOperatorDescriptor join = new NestedLoopJoinOperatorDescriptor(spec, new JoinComparatorFactory(PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY), 1, 0), custOrderJoinDesc, 5, true, nonMatchWriterFactories);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, join, NC1_ID, NC2_ID);
    ResultSetId rsId = new ResultSetId(1);
    spec.addResultSetId(rsId);
    IOperatorDescriptor printer = new ResultWriterOperatorDescriptor(spec, rsId, false, false, ResultSerializerFactoryProvider.INSTANCE.getResultSerializerFactoryProvider());
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);
    IConnectorDescriptor ordJoinConn = new OneToOneConnectorDescriptor(spec);
    spec.connect(ordJoinConn, ordScanner, 0, join, 0);
    IConnectorDescriptor custJoinConn = new MToNBroadcastConnectorDescriptor(spec);
    spec.connect(custJoinConn, custScanner, 0, join, 1);
    IConnectorDescriptor joinPrinterConn = new MToNBroadcastConnectorDescriptor(spec);
    spec.connect(joinPrinterConn, join, 0, printer, 0);
    spec.addRoot(printer);
    runTest(spec);
}

Example 33 with FileScanOperatorDescriptor

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

the class UnionTest method createUnionJobSpec.

public static JobSpecification createUnionJobSpec() throws Exception {
    JobSpecification spec = new JobSpecification();
    IFileSplitProvider splitProvider = new ConstantFileSplitProvider(new FileSplit[] { new ManagedFileSplit(NC2_ID, "data" + File.separator + "words.txt"), new ManagedFileSplit(NC1_ID, "data" + File.separator + "nc1" + File.separator + "words.txt") });
    RecordDescriptor desc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer() });
    FileScanOperatorDescriptor csvScanner01 = new FileScanOperatorDescriptor(spec, splitProvider, new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE }, ','), desc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, csvScanner01, NC2_ID, NC1_ID);
    FileScanOperatorDescriptor csvScanner02 = new FileScanOperatorDescriptor(spec, splitProvider, new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE }, ','), desc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, csvScanner02, NC2_ID, NC1_ID);
    UnionAllOperatorDescriptor unionAll = new UnionAllOperatorDescriptor(spec, 2, desc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, unionAll, NC2_ID, NC1_ID);
    ResultSetId rsId = new ResultSetId(1);
    spec.addResultSetId(rsId);
    IOperatorDescriptor printer = new ResultWriterOperatorDescriptor(spec, rsId, false, false, ResultSerializerFactoryProvider.INSTANCE.getResultSerializerFactoryProvider());
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC2_ID, NC1_ID);
    spec.connect(new OneToOneConnectorDescriptor(spec), csvScanner01, 0, unionAll, 0);
    spec.connect(new OneToOneConnectorDescriptor(spec), csvScanner02, 0, unionAll, 1);
    spec.connect(new OneToOneConnectorDescriptor(spec), unionAll, 0, printer, 0);
    spec.addRoot(printer);
    return spec;
}

Example 34 with FileScanOperatorDescriptor

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

the class WordCountMain method createJob.

private static JobSpecification createJob(FileSplit[] inSplits, FileSplit[] outSplits, String algo, int htSize, int frameLimit, String format, int frameSize) {
    JobSpecification spec = new JobSpecification(frameSize);
    IFileSplitProvider splitsProvider = new ConstantFileSplitProvider(inSplits);
    RecordDescriptor wordDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer() });
    FileScanOperatorDescriptor wordScanner = new FileScanOperatorDescriptor(spec, splitsProvider, new WordTupleParserFactory(), wordDesc);
    createPartitionConstraint(spec, wordScanner, inSplits);
    RecordDescriptor groupResultDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE });
    IOperatorDescriptor gBy;
    int[] keys = new int[] { 0 };
    if ("hash".equalsIgnoreCase(algo)) {
        gBy = new ExternalGroupOperatorDescriptor(spec, htSize, fileSize, keys, frameLimit, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, new UTF8StringNormalizedKeyComputerFactory(), new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(1, false), new IntSumFieldAggregatorFactory(3, false), new FloatSumFieldAggregatorFactory(5, false) }), new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(1, false), new IntSumFieldAggregatorFactory(2, false), new FloatSumFieldAggregatorFactory(3, false) }), groupResultDesc, groupResultDesc, new HashSpillableTableFactory(new IBinaryHashFunctionFamily[] { UTF8StringBinaryHashFunctionFamily.INSTANCE }));
        createPartitionConstraint(spec, gBy, outSplits);
        IConnectorDescriptor scanGroupConn = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(keys, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }));
        spec.connect(scanGroupConn, wordScanner, 0, gBy, 0);
    } else {
        IBinaryComparatorFactory[] cfs = new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) };
        IOperatorDescriptor sorter = "memsort".equalsIgnoreCase(algo) ? new InMemorySortOperatorDescriptor(spec, keys, new UTF8StringNormalizedKeyComputerFactory(), cfs, wordDesc) : new ExternalSortOperatorDescriptor(spec, frameLimit, keys, new UTF8StringNormalizedKeyComputerFactory(), cfs, wordDesc);
        createPartitionConstraint(spec, sorter, outSplits);
        IConnectorDescriptor scanSortConn = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(keys, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }));
        spec.connect(scanSortConn, wordScanner, 0, sorter, 0);
        gBy = new PreclusteredGroupOperatorDescriptor(spec, keys, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new CountFieldAggregatorFactory(true) }), groupResultDesc);
        createPartitionConstraint(spec, gBy, outSplits);
        OneToOneConnectorDescriptor sortGroupConn = new OneToOneConnectorDescriptor(spec);
        spec.connect(sortGroupConn, sorter, 0, gBy, 0);
    }
    IFileSplitProvider outSplitProvider = new ConstantFileSplitProvider(outSplits);
    IOperatorDescriptor writer = "text".equalsIgnoreCase(format) ? new PlainFileWriterOperatorDescriptor(spec, outSplitProvider, ",") : new FrameFileWriterOperatorDescriptor(spec, outSplitProvider);
    createPartitionConstraint(spec, writer, outSplits);
    IConnectorDescriptor gbyPrinterConn = new OneToOneConnectorDescriptor(spec);
    spec.connect(gbyPrinterConn, gBy, 0, writer, 0);
    spec.addRoot(writer);
    return spec;
}

Example 35 with FileScanOperatorDescriptor

use of org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor 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;
}