Search in sources :

Example 1 with InMemorySortRuntimeFactory

use of org.apache.hyracks.algebricks.runtime.operators.sort.InMemorySortRuntimeFactory in project asterixdb by apache.

the class InMemoryStableSortPOperator method contributeRuntimeOperator.

@Override
public void contributeRuntimeOperator(IHyracksJobBuilder builder, JobGenContext context, ILogicalOperator op, IOperatorSchema opSchema, IOperatorSchema[] inputSchemas, IOperatorSchema outerPlanSchema) throws AlgebricksException {
    RecordDescriptor recDescriptor = JobGenHelper.mkRecordDescriptor(context.getTypeEnvironment(op), opSchema, context);
    int n = sortColumns.length;
    int[] sortFields = new int[n];
    IBinaryComparatorFactory[] comps = new IBinaryComparatorFactory[n];
    int i = 0;
    INormalizedKeyComputerFactoryProvider nkcfProvider = context.getNormalizedKeyComputerFactoryProvider();
    INormalizedKeyComputerFactory nkcf = null;
    IVariableTypeEnvironment env = context.getTypeEnvironment(op);
    for (OrderColumn oc : sortColumns) {
        LogicalVariable var = oc.getColumn();
        sortFields[i] = opSchema.findVariable(var);
        Object type = env.getVarType(var);
        OrderKind order = oc.getOrder();
        if (i == 0 && nkcfProvider != null && type != null) {
            nkcf = nkcfProvider.getNormalizedKeyComputerFactory(type, order == OrderKind.ASC);
        }
        IBinaryComparatorFactoryProvider bcfp = context.getBinaryComparatorFactoryProvider();
        comps[i] = bcfp.getBinaryComparatorFactory(type, oc.getOrder() == OrderKind.ASC);
        i++;
    }
    IPushRuntimeFactory runtime = new InMemorySortRuntimeFactory(sortFields, nkcf, comps, null);
    builder.contributeMicroOperator(op, runtime, recDescriptor);
    ILogicalOperator src = op.getInputs().get(0).getValue();
    builder.contributeGraphEdge(src, 0, op, 0);
}
Also used : LogicalVariable(org.apache.hyracks.algebricks.core.algebra.base.LogicalVariable) InMemorySortRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.sort.InMemorySortRuntimeFactory) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) OrderColumn(org.apache.hyracks.algebricks.core.algebra.properties.OrderColumn) ILogicalOperator(org.apache.hyracks.algebricks.core.algebra.base.ILogicalOperator) IBinaryComparatorFactory(org.apache.hyracks.api.dataflow.value.IBinaryComparatorFactory) IBinaryComparatorFactoryProvider(org.apache.hyracks.algebricks.data.IBinaryComparatorFactoryProvider) IPushRuntimeFactory(org.apache.hyracks.algebricks.runtime.base.IPushRuntimeFactory) INormalizedKeyComputerFactory(org.apache.hyracks.api.dataflow.value.INormalizedKeyComputerFactory) OrderKind(org.apache.hyracks.algebricks.core.algebra.operators.logical.OrderOperator.IOrder.OrderKind) INormalizedKeyComputerFactoryProvider(org.apache.hyracks.algebricks.data.INormalizedKeyComputerFactoryProvider) IVariableTypeEnvironment(org.apache.hyracks.algebricks.core.algebra.expressions.IVariableTypeEnvironment)

Example 2 with InMemorySortRuntimeFactory

use of org.apache.hyracks.algebricks.runtime.operators.sort.InMemorySortRuntimeFactory in project asterixdb by apache.

the class PushRuntimeTest method scanMicroSortWrite.

@Test
public void scanMicroSortWrite() 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 + "nation.tbl");
    IFileSplitProvider splitProvider = new ConstantFileSplitProvider(fileSplits);
    RecordDescriptor scannerDesc = new RecordDescriptor(new ISerializerDeserializer[] { IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer() });
    IValueParserFactory[] valueParsers = new IValueParserFactory[] { IntegerParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, IntegerParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE };
    FileScanOperatorDescriptor scanner = new FileScanOperatorDescriptor(spec, splitProvider, new DelimitedDataTupleParserFactory(valueParsers, '|'), scannerDesc);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, scanner, new String[] { AlgebricksHyracksIntegrationUtil.NC1_ID });
    // the algebricks op.
    InMemorySortRuntimeFactory sort = new InMemorySortRuntimeFactory(new int[] { 1 }, null, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, null);
    RecordDescriptor sortDesc = scannerDesc;
    String fileName = "scanMicroSortWrite.out";
    String filePath = PATH_ACTUAL + SEPARATOR + fileName;
    String resultFilePath = PATH_EXPECTED + SEPARATOR + fileName;
    File outFile = new File(filePath);
    SinkWriterRuntimeFactory writer = new SinkWriterRuntimeFactory(new int[] { 0, 1, 2, 3 }, new IPrinterFactory[] { IntegerPrinterFactory.INSTANCE, UTF8StringPrinterFactory.INSTANCE, IntegerPrinterFactory.INSTANCE, UTF8StringPrinterFactory.INSTANCE }, outFile, PrinterBasedWriterFactory.INSTANCE, sortDesc);
    AlgebricksMetaOperatorDescriptor algebricksOp = new AlgebricksMetaOperatorDescriptor(spec, 1, 0, new IPushRuntimeFactory[] { sort, writer }, new RecordDescriptor[] { sortDesc, null });
    PartitionConstraintHelper.addPartitionCountConstraint(spec, algebricksOp, 1);
    spec.connect(new OneToOneConnectorDescriptor(spec), scanner, 0, algebricksOp, 0);
    spec.addRoot(algebricksOp);
    AlgebricksHyracksIntegrationUtil.runJob(spec);
    compareFiles(filePath, resultFilePath);
    outFile.delete();
}
Also used : InMemorySortRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.sort.InMemorySortRuntimeFactory) 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) UTF8StringSerializerDeserializer(org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer) 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 3 with InMemorySortRuntimeFactory

use of org.apache.hyracks.algebricks.runtime.operators.sort.InMemorySortRuntimeFactory in project asterixdb by apache.

the class PushRuntimeTest method scanMicroSortGbySelectWrite.

@Test
public void scanMicroSortGbySelectWrite() 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;
    InMemorySortRuntimeFactory sort = new InMemorySortRuntimeFactory(new int[] { 3 }, null, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, null);
    // 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 });
    MicroPreClusteredGroupRuntimeFactory gby = new MicroPreClusteredGroupRuntimeFactory(new int[] { 3 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(IntegerPointable.FACTORY) }, npaaf, sortDesc, gbyDesc, null);
    // 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[] { sort, gby, select, writer }, new RecordDescriptor[] { sortDesc, gbyDesc, selectDesc, 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("9", buf.toString());
    outFile.delete();
}
Also used : InMemorySortRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.sort.InMemorySortRuntimeFactory) AggregateRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.aggreg.AggregateRuntimeFactory) RunningAggregateRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.std.RunningAggregateRuntimeFactory) MicroPreClusteredGroupRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.group.MicroPreClusteredGroupRuntimeFactory) IFileSplitProvider(org.apache.hyracks.dataflow.std.file.IFileSplitProvider) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) 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) TupleCountAggregateFunctionFactory(org.apache.hyracks.algebricks.runtime.aggregators.TupleCountAggregateFunctionFactory) NestedTupleSourceRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.std.NestedTupleSourceRuntimeFactory) ManagedFileSplit(org.apache.hyracks.api.io.ManagedFileSplit) FileScanOperatorDescriptor(org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor) JobSpecification(org.apache.hyracks.api.job.JobSpecification) AlgebricksPipeline(org.apache.hyracks.algebricks.runtime.base.AlgebricksPipeline) TupleFieldEvaluatorFactory(org.apache.hyracks.algebricks.runtime.evaluators.TupleFieldEvaluatorFactory) IValueParserFactory(org.apache.hyracks.dataflow.common.data.parsers.IValueParserFactory) 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) NestedPlansAccumulatingAggregatorFactory(org.apache.hyracks.algebricks.runtime.operators.aggreg.NestedPlansAccumulatingAggregatorFactory) IScalarEvaluatorFactory(org.apache.hyracks.algebricks.runtime.base.IScalarEvaluatorFactory) StreamSelectRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.std.StreamSelectRuntimeFactory) SinkWriterRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.std.SinkWriterRuntimeFactory) File(java.io.File) Test(org.junit.Test)

Aggregations

InMemorySortRuntimeFactory (org.apache.hyracks.algebricks.runtime.operators.sort.InMemorySortRuntimeFactory)3 RecordDescriptor (org.apache.hyracks.api.dataflow.value.RecordDescriptor)3 File (java.io.File)2 AlgebricksMetaOperatorDescriptor (org.apache.hyracks.algebricks.runtime.operators.meta.AlgebricksMetaOperatorDescriptor)2 SinkWriterRuntimeFactory (org.apache.hyracks.algebricks.runtime.operators.std.SinkWriterRuntimeFactory)2 FileSplit (org.apache.hyracks.api.io.FileSplit)2 ManagedFileSplit (org.apache.hyracks.api.io.ManagedFileSplit)2 JobSpecification (org.apache.hyracks.api.job.JobSpecification)2 UTF8StringSerializerDeserializer (org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer)2 IValueParserFactory (org.apache.hyracks.dataflow.common.data.parsers.IValueParserFactory)2 OneToOneConnectorDescriptor (org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor)2 ConstantFileSplitProvider (org.apache.hyracks.dataflow.std.file.ConstantFileSplitProvider)2 DelimitedDataTupleParserFactory (org.apache.hyracks.dataflow.std.file.DelimitedDataTupleParserFactory)2 FileScanOperatorDescriptor (org.apache.hyracks.dataflow.std.file.FileScanOperatorDescriptor)2 IFileSplitProvider (org.apache.hyracks.dataflow.std.file.IFileSplitProvider)2 Test (org.junit.Test)2 ILogicalOperator (org.apache.hyracks.algebricks.core.algebra.base.ILogicalOperator)1 LogicalVariable (org.apache.hyracks.algebricks.core.algebra.base.LogicalVariable)1 IVariableTypeEnvironment (org.apache.hyracks.algebricks.core.algebra.expressions.IVariableTypeEnvironment)1 OrderKind (org.apache.hyracks.algebricks.core.algebra.operators.logical.OrderOperator.IOrder.OrderKind)1