Search in sources :

Example 11 with DataSource

use of com.linkedin.pinot.core.common.DataSource in project pinot by linkedin.

the class ChunkIndexCreationDriverImplTest method test3.

@Test
public void test3() throws Exception {
    final IndexSegmentImpl segment = (IndexSegmentImpl) Loaders.IndexSegment.load(INDEX_DIR.listFiles()[0], ReadMode.mmap);
    final DataSource ds = segment.getDataSource("column7");
    final Block bl = ds.nextBlock();
    final BlockValSet valSet = bl.getBlockValueSet();
    final int maxValue = ((SegmentMetadataImpl) segment.getSegmentMetadata()).getColumnMetadataFor("column7").getMaxNumberOfMultiValues();
    final BlockMultiValIterator it = (BlockMultiValIterator) valSet.iterator();
    while (it.hasNext()) {
        final int[] entry = new int[maxValue];
        it.nextIntVal(entry);
        LOGGER.trace(Arrays.toString(entry));
    }
}
Also used : BlockMultiValIterator(com.linkedin.pinot.core.common.BlockMultiValIterator) IndexSegmentImpl(com.linkedin.pinot.core.segment.index.IndexSegmentImpl) Block(com.linkedin.pinot.core.common.Block) BlockValSet(com.linkedin.pinot.core.common.BlockValSet) DataSource(com.linkedin.pinot.core.common.DataSource) Test(org.testng.annotations.Test)

Example 12 with DataSource

use of com.linkedin.pinot.core.common.DataSource in project pinot by linkedin.

the class ChunkIndexCreationDriverImplTest method test5.

@Test(enabled = false)
public void test5() throws Exception {
    final IndexSegmentImpl segment = (IndexSegmentImpl) Loaders.IndexSegment.load(INDEX_DIR.listFiles()[0], ReadMode.mmap);
    final List<String> rhs = new ArrayList<String>();
    rhs.add("-100");
    final Predicate p = new EqPredicate("column1", rhs);
    final DataSource ds = segment.getDataSource("column1", p);
    final Block bl = ds.nextBlock();
    final BlockDocIdSet idSet = bl.getBlockDocIdSet();
    final BlockDocIdIterator it = idSet.iterator();
    int docId = it.next();
    final StringBuilder b = new StringBuilder();
    while (docId != Constants.EOF) {
        b.append(docId + ",");
        docId = it.next();
    }
//    System.out.println(b.toString());
}
Also used : IndexSegmentImpl(com.linkedin.pinot.core.segment.index.IndexSegmentImpl) BlockDocIdSet(com.linkedin.pinot.core.common.BlockDocIdSet) ArrayList(java.util.ArrayList) Block(com.linkedin.pinot.core.common.Block) EqPredicate(com.linkedin.pinot.core.common.predicate.EqPredicate) EqPredicate(com.linkedin.pinot.core.common.predicate.EqPredicate) Predicate(com.linkedin.pinot.core.common.Predicate) DataSource(com.linkedin.pinot.core.common.DataSource) BlockDocIdIterator(com.linkedin.pinot.core.common.BlockDocIdIterator) Test(org.testng.annotations.Test)

Example 13 with DataSource

use of com.linkedin.pinot.core.common.DataSource in project pinot by linkedin.

the class ChunkIndexCreationDriverImplTest method test2.

@Test
public void test2() throws Exception {
    final IndexSegmentImpl segment = (IndexSegmentImpl) Loaders.IndexSegment.load(INDEX_DIR.listFiles()[0], ReadMode.mmap);
    //    System.out.println("INdex dir:" + INDEX_DIR);
    final DataSource ds = segment.getDataSource("column1");
    final Block bl = ds.nextBlock();
    final BlockValSet valSet = bl.getBlockValueSet();
    final BlockSingleValIterator it = (BlockSingleValIterator) valSet.iterator();
    // TODO: FIXME - load segment with known data and verify that it exists
    while (it.hasNext()) {
        LOGGER.trace(Integer.toString(it.nextIntVal()));
    }
}
Also used : IndexSegmentImpl(com.linkedin.pinot.core.segment.index.IndexSegmentImpl) BlockSingleValIterator(com.linkedin.pinot.core.common.BlockSingleValIterator) Block(com.linkedin.pinot.core.common.Block) BlockValSet(com.linkedin.pinot.core.common.BlockValSet) DataSource(com.linkedin.pinot.core.common.DataSource) Test(org.testng.annotations.Test)

Example 14 with DataSource

use of com.linkedin.pinot.core.common.DataSource in project pinot by linkedin.

the class ChunkIndexCreationDriverImplTest method test6.

@Test(enabled = false)
public void test6() throws Exception {
    final IndexSegmentImpl segment = (IndexSegmentImpl) Loaders.IndexSegment.load(INDEX_DIR.listFiles()[0], ReadMode.mmap);
    final ImmutableDictionaryReader d = segment.getDictionaryFor("column7");
    final List<String> rhs = new ArrayList<String>();
    rhs.add(d.get(new Random().nextInt(d.length())).toString());
    final Predicate p = new EqPredicate("column7", rhs);
    final DataSource ds = segment.getDataSource("column7", p);
    final Block bl = ds.nextBlock();
    final BlockDocIdSet idSet = bl.getBlockDocIdSet();
    final BlockDocIdIterator it = idSet.iterator();
    int docId = it.next();
    final StringBuilder b = new StringBuilder();
    while (docId != Constants.EOF) {
        b.append(docId + ",");
        docId = it.next();
    }
//    System.out.println(b.toString());
}
Also used : ImmutableDictionaryReader(com.linkedin.pinot.core.segment.index.readers.ImmutableDictionaryReader) ArrayList(java.util.ArrayList) EqPredicate(com.linkedin.pinot.core.common.predicate.EqPredicate) EqPredicate(com.linkedin.pinot.core.common.predicate.EqPredicate) Predicate(com.linkedin.pinot.core.common.Predicate) DataSource(com.linkedin.pinot.core.common.DataSource) BlockDocIdIterator(com.linkedin.pinot.core.common.BlockDocIdIterator) IndexSegmentImpl(com.linkedin.pinot.core.segment.index.IndexSegmentImpl) Random(java.util.Random) BlockDocIdSet(com.linkedin.pinot.core.common.BlockDocIdSet) Block(com.linkedin.pinot.core.common.Block) Test(org.testng.annotations.Test)

Example 15 with DataSource

use of com.linkedin.pinot.core.common.DataSource in project pinot by linkedin.

the class DefaultGroupKeyGeneratorTest method setup.

@BeforeClass
private void setup() throws Exception {
    GenericRow[] segmentData = new GenericRow[NUM_ROWS];
    int value = _random.nextInt(MAX_STEP_LENGTH);
    // Generate random values for the segment.
    for (int i = 0; i < UNIQUE_ROWS; i++) {
        Map<String, Object> map = new HashMap<>();
        for (String singleValueColumn : SINGLE_VALUE_COLUMNS) {
            map.put(singleValueColumn, value);
            value += 1 + _random.nextInt(MAX_STEP_LENGTH);
        }
        for (String multiValueColumn : MULTI_VALUE_COLUMNS) {
            int numMultiValues = 1 + _random.nextInt(MAX_NUM_MULTI_VALUES);
            Integer[] values = new Integer[numMultiValues];
            for (int k = 0; k < numMultiValues; k++) {
                values[k] = value;
                value += 1 + _random.nextInt(MAX_STEP_LENGTH);
            }
            map.put(multiValueColumn, values);
        }
        GenericRow genericRow = new GenericRow();
        genericRow.init(map);
        segmentData[i] = genericRow;
    }
    for (int i = UNIQUE_ROWS; i < NUM_ROWS; i += UNIQUE_ROWS) {
        System.arraycopy(segmentData, 0, segmentData, i, UNIQUE_ROWS);
    }
    // Create an index segment with the random values.
    Schema schema = new Schema();
    for (String singleValueColumn : SINGLE_VALUE_COLUMNS) {
        DimensionFieldSpec dimensionFieldSpec = new DimensionFieldSpec(singleValueColumn, FieldSpec.DataType.INT, true);
        schema.addField(dimensionFieldSpec);
    }
    for (String multiValueColumn : MULTI_VALUE_COLUMNS) {
        DimensionFieldSpec dimensionFieldSpec = new DimensionFieldSpec(multiValueColumn, FieldSpec.DataType.INT, false);
        schema.addField(dimensionFieldSpec);
    }
    SegmentGeneratorConfig config = new SegmentGeneratorConfig(schema);
    FileUtils.deleteQuietly(new File(INDEX_DIR_PATH));
    config.setOutDir(INDEX_DIR_PATH);
    config.setSegmentName(SEGMENT_NAME);
    SegmentIndexCreationDriverImpl driver = new SegmentIndexCreationDriverImpl();
    driver.init(config, new TestDataRecordReader(schema, segmentData));
    driver.build();
    IndexSegment indexSegment = Loaders.IndexSegment.load(new File(INDEX_DIR_PATH, SEGMENT_NAME), ReadMode.heap);
    // Get a data fetcher for the index segment.
    Map<String, BaseOperator> dataSourceMap = new HashMap<>();
    Map<String, Block> blockMap = new HashMap<>();
    for (String column : indexSegment.getColumnNames()) {
        DataSource dataSource = indexSegment.getDataSource(column);
        dataSourceMap.put(column, dataSource);
        blockMap.put(column, dataSource.getNextBlock());
    }
    // Generate a random test doc id set.
    int num1 = _random.nextInt(50);
    int num2 = num1 + 1 + _random.nextInt(50);
    for (int i = 0; i < 20; i += 2) {
        _testDocIdSet[i] = num1 + 50 * i;
        _testDocIdSet[i + 1] = num2 + 50 * i;
    }
    DataFetcher dataFetcher = new DataFetcher(dataSourceMap);
    DocIdSetBlock docIdSetBlock = new DocIdSetBlock(_testDocIdSet, _testDocIdSet.length);
    ProjectionBlock projectionBlock = new ProjectionBlock(blockMap, new DataBlockCache(dataFetcher), docIdSetBlock);
    _transformBlock = new TransformBlock(projectionBlock, new HashMap<String, BlockValSet>());
}
Also used : BaseOperator(com.linkedin.pinot.core.operator.BaseOperator) HashMap(java.util.HashMap) Schema(com.linkedin.pinot.common.data.Schema) GenericRow(com.linkedin.pinot.core.data.GenericRow) ProjectionBlock(com.linkedin.pinot.core.operator.blocks.ProjectionBlock) TransformBlock(com.linkedin.pinot.core.operator.blocks.TransformBlock) SegmentGeneratorConfig(com.linkedin.pinot.core.indexsegment.generator.SegmentGeneratorConfig) DimensionFieldSpec(com.linkedin.pinot.common.data.DimensionFieldSpec) IndexSegment(com.linkedin.pinot.core.indexsegment.IndexSegment) DocIdSetBlock(com.linkedin.pinot.core.operator.blocks.DocIdSetBlock) SegmentIndexCreationDriverImpl(com.linkedin.pinot.core.segment.creator.impl.SegmentIndexCreationDriverImpl) DataSource(com.linkedin.pinot.core.common.DataSource) Block(com.linkedin.pinot.core.common.Block) DocIdSetBlock(com.linkedin.pinot.core.operator.blocks.DocIdSetBlock) TransformBlock(com.linkedin.pinot.core.operator.blocks.TransformBlock) ProjectionBlock(com.linkedin.pinot.core.operator.blocks.ProjectionBlock) DataFetcher(com.linkedin.pinot.core.common.DataFetcher) DataBlockCache(com.linkedin.pinot.core.common.DataBlockCache) File(java.io.File) TestDataRecordReader(com.linkedin.pinot.util.TestDataRecordReader) BeforeClass(org.testng.annotations.BeforeClass)

Aggregations

DataSource (com.linkedin.pinot.core.common.DataSource)28 Block (com.linkedin.pinot.core.common.Block)21 Test (org.testng.annotations.Test)15 BlockSingleValIterator (com.linkedin.pinot.core.common.BlockSingleValIterator)14 Predicate (com.linkedin.pinot.core.common.Predicate)13 ArrayList (java.util.ArrayList)12 BlockDocIdIterator (com.linkedin.pinot.core.common.BlockDocIdIterator)11 EqPredicate (com.linkedin.pinot.core.common.predicate.EqPredicate)11 RealtimeSegmentImplTest (com.linkedin.pinot.core.realtime.impl.kafka.RealtimeSegmentImplTest)10 NEqPredicate (com.linkedin.pinot.core.common.predicate.NEqPredicate)8 RangePredicate (com.linkedin.pinot.core.common.predicate.RangePredicate)8 FieldSpec (com.linkedin.pinot.common.data.FieldSpec)6 BlockMetadata (com.linkedin.pinot.core.common.BlockMetadata)6 BlockValSet (com.linkedin.pinot.core.common.BlockValSet)5 BitmapBasedFilterOperator (com.linkedin.pinot.core.operator.filter.BitmapBasedFilterOperator)5 ScanBasedFilterOperator (com.linkedin.pinot.core.operator.filter.ScanBasedFilterOperator)5 IndexSegmentImpl (com.linkedin.pinot.core.segment.index.IndexSegmentImpl)5 Dictionary (com.linkedin.pinot.core.segment.index.readers.Dictionary)4 HashMap (java.util.HashMap)4 BlockDocIdSet (com.linkedin.pinot.core.common.BlockDocIdSet)3