Examples with UTF8StringSerializerDeserializer org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer used on opensource projects

Example 6 with UTF8StringSerializerDeserializer

use of org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer 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 7 with UTF8StringSerializerDeserializer

use of org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer 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 8 with UTF8StringSerializerDeserializer

use of org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer 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 9 with UTF8StringSerializerDeserializer

use of org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer in project asterixdb by apache.

the class LSMBTreeFilterMergeTestDriver method runTest.

@Override
protected void runTest(ISerializerDeserializer[] fieldSerdes, int numKeys, BTreeLeafFrameType leafType, ITupleReference lowKey, ITupleReference highKey, ITupleReference prefixLowKey, ITupleReference prefixHighKey) throws Exception {
    OrderedIndexTestContext ctx = createTestContext(fieldSerdes, numKeys, leafType, true);
    ctx.getIndex().create();
    ctx.getIndex().activate();
    // to determine which field types to generate.
    if (fieldSerdes[0] instanceof IntegerSerializerDeserializer) {
        orderedIndexTestUtils.bulkLoadIntTuples(ctx, numTuplesToInsert, true, getRandom());
    } else if (fieldSerdes[0] instanceof UTF8StringSerializerDeserializer) {
        orderedIndexTestUtils.bulkLoadStringTuples(ctx, numTuplesToInsert, true, getRandom());
    }
    int maxTreesToMerge = AccessMethodTestsConfig.LSM_BTREE_MAX_TREES_TO_MERGE;
    ILSMIndexAccessor accessor = (ILSMIndexAccessor) ctx.getIndexAccessor();
    IBinaryComparator comp = ctx.getComparatorFactories()[0].createBinaryComparator();
    for (int i = 0; i < maxTreesToMerge; i++) {
        int flushed = 0;
        for (; flushed < i; flushed++) {
            Pair<ITupleReference, ITupleReference> minMax = null;
            if (fieldSerdes[0] instanceof IntegerSerializerDeserializer) {
                minMax = orderedIndexTestUtils.insertIntTuples(ctx, numTuplesToInsert, true, getRandom());
            } else {
                minMax = orderedIndexTestUtils.insertStringTuples(ctx, numTuplesToInsert, true, getRandom());
            }
            if (minMax != null) {
                ILSMComponentFilter f = ((LSMBTree) ctx.getIndex()).getCurrentMemoryComponent().getLSMComponentFilter();
                Pair<ITupleReference, ITupleReference> obsMinMax = filterToMinMax(f);
                Assert.assertEquals(0, TreeIndexTestUtils.compareFilterTuples(obsMinMax.getLeft(), minMax.getLeft(), comp));
                Assert.assertEquals(0, TreeIndexTestUtils.compareFilterTuples(obsMinMax.getRight(), minMax.getRight(), comp));
            }
            StubIOOperationCallback stub = new StubIOOperationCallback();
            BlockingIOOperationCallbackWrapper waiter = new BlockingIOOperationCallbackWrapper(stub);
            accessor.scheduleFlush(waiter);
            waiter.waitForIO();
            if (minMax != null) {
                Pair<ITupleReference, ITupleReference> obsMinMax = filterToMinMax(stub.getLastNewComponent().getLSMComponentFilter());
                Assert.assertEquals(0, TreeIndexTestUtils.compareFilterTuples(obsMinMax.getLeft(), minMax.getLeft(), comp));
                Assert.assertEquals(0, TreeIndexTestUtils.compareFilterTuples(obsMinMax.getRight(), minMax.getRight(), comp));
            }
        }
        List<ILSMDiskComponent> flushedComponents = ((LSMBTree) ctx.getIndex()).getImmutableComponents();
        MutablePair<ITupleReference, ITupleReference> expectedMergeMinMax = null;
        for (ILSMDiskComponent f : flushedComponents) {
            Pair<ITupleReference, ITupleReference> componentMinMax = filterToMinMax(f.getLSMComponentFilter());
            if (expectedMergeMinMax == null) {
                expectedMergeMinMax = MutablePair.of(componentMinMax.getLeft(), componentMinMax.getRight());
            }
            if (TreeIndexTestUtils.compareFilterTuples(expectedMergeMinMax.getLeft(), componentMinMax.getLeft(), comp) > 0) {
                expectedMergeMinMax.setLeft(componentMinMax.getLeft());
            }
            if (TreeIndexTestUtils.compareFilterTuples(expectedMergeMinMax.getRight(), componentMinMax.getRight(), comp) < 0) {
                expectedMergeMinMax.setRight(componentMinMax.getRight());
            }
        }
        accessor.scheduleMerge(NoOpIOOperationCallbackFactory.INSTANCE.createIoOpCallback(), ((LSMBTree) ctx.getIndex()).getImmutableComponents());
        flushedComponents = ((LSMBTree) ctx.getIndex()).getImmutableComponents();
        Pair<ITupleReference, ITupleReference> mergedMinMax = filterToMinMax(flushedComponents.get(0).getLSMComponentFilter());
        Assert.assertEquals(0, TreeIndexTestUtils.compareFilterTuples(expectedMergeMinMax.getLeft(), mergedMinMax.getLeft(), comp));
        Assert.assertEquals(0, TreeIndexTestUtils.compareFilterTuples(expectedMergeMinMax.getRight(), mergedMinMax.getRight(), comp));
        orderedIndexTestUtils.checkPointSearches(ctx);
        orderedIndexTestUtils.checkScan(ctx);
        orderedIndexTestUtils.checkDiskOrderScan(ctx);
        orderedIndexTestUtils.checkRangeSearch(ctx, lowKey, highKey, true, true);
        if (prefixLowKey != null && prefixHighKey != null) {
            orderedIndexTestUtils.checkRangeSearch(ctx, prefixLowKey, prefixHighKey, true, true);
        }
    }
    ctx.getIndex().deactivate();
    ctx.getIndex().destroy();
}

Example 10 with UTF8StringSerializerDeserializer

use of org.apache.hyracks.dataflow.common.data.marshalling.UTF8StringSerializerDeserializer in project asterixdb by apache.

the class BloomFilterTest method multiFieldTest.

@Test
public void multiFieldTest() throws Exception {
    if (LOGGER.isLoggable(Level.INFO)) {
        LOGGER.info("TESTING BLOOM FILTER");
    }
    IBufferCache bufferCache = harness.getBufferCache();
    int numElements = 10000;
    int[] keyFields = { 2, 4, 1 };
    BloomFilter bf = new BloomFilter(bufferCache, harness.getFileMapProvider(), harness.getFileReference(), keyFields);
    double acceptanleFalsePositiveRate = 0.1;
    int maxBucketsPerElement = BloomCalculations.maxBucketsPerElement(numElements);
    BloomFilterSpecification bloomFilterSpec = BloomCalculations.computeBloomSpec(maxBucketsPerElement, acceptanleFalsePositiveRate);
    bf.create();
    bf.activate();
    IIndexBulkLoader builder = bf.createBuilder(numElements, bloomFilterSpec.getNumHashes(), bloomFilterSpec.getNumBucketsPerElements());
    int fieldCount = 5;
    ISerializerDeserializer[] fieldSerdes = { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer() };
    ArrayTupleBuilder tupleBuilder = new ArrayTupleBuilder(fieldCount);
    ArrayTupleReference tuple = new ArrayTupleReference();
    int maxLength = 20;
    ArrayList<String> s1 = new ArrayList<>();
    ArrayList<String> s2 = new ArrayList<>();
    ArrayList<String> s3 = new ArrayList<>();
    ArrayList<String> s4 = new ArrayList<>();
    for (int i = 0; i < numElements; ++i) {
        s1.add(randomString(rnd.nextInt() % maxLength, rnd));
        s2.add(randomString(rnd.nextInt() % maxLength, rnd));
        s3.add(randomString(rnd.nextInt() % maxLength, rnd));
        s4.add(randomString(rnd.nextInt() % maxLength, rnd));
    }
    for (int i = 0; i < numElements; ++i) {
        TupleUtils.createTuple(tupleBuilder, tuple, fieldSerdes, s1.get(i), s2.get(i), i, s3.get(i), s4.get(i));
        builder.add(tuple);
    }
    builder.end();
    long[] hashes = new long[2];
    for (int i = 0; i < numElements; ++i) {
        TupleUtils.createTuple(tupleBuilder, tuple, fieldSerdes, s1.get(i), s2.get(i), i, s3.get(i), s4.get(i));
        Assert.assertTrue(bf.contains(tuple, hashes));
    }
    bf.deactivate();
    bf.destroy();
}