Examples with TestDataRecordReader com.linkedin.pinot.util.TestDataRecordReader used on opensource projects

Example 1 with TestDataRecordReader

use of com.linkedin.pinot.util.TestDataRecordReader 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>());
}

Example 2 with TestDataRecordReader

use of com.linkedin.pinot.util.TestDataRecordReader in project pinot by linkedin.

the class DataFetcherTest method setup.

@BeforeClass
private void setup() throws Exception {
    GenericRow[] segmentData = new GenericRow[NUM_ROWS];
    // Generate random dimension and metric values.
    for (int i = 0; i < NUM_ROWS; i++) {
        double randomDouble = _random.nextDouble();
        String randomDoubleString = String.valueOf(randomDouble);
        _dimensionValues[i] = randomDoubleString;
        _intMetricValues[i] = (int) randomDouble;
        _longMetricValues[i] = (long) randomDouble;
        _floatMetricValues[i] = (float) randomDouble;
        _doubleMetricValues[i] = randomDouble;
        HashMap<String, Object> map = new HashMap<>();
        map.put(DIMENSION_NAME, _dimensionValues[i]);
        map.put(INT_METRIC_NAME, _intMetricValues[i]);
        map.put(LONG_METRIC_NAME, _longMetricValues[i]);
        map.put(FLOAT_METRIC_NAME, _floatMetricValues[i]);
        map.put(DOUBLE_METRIC_NAME, _doubleMetricValues[i]);
        map.put(NO_DICT_INT_METRIC_NAME, _intMetricValues[i]);
        map.put(NO_DICT_LONG_METRIC_NAME, _longMetricValues[i]);
        map.put(NO_DICT_FLOAT_METRIC_NAME, _floatMetricValues[i]);
        map.put(NO_DICT_DOUBLE_METRIC_NAME, _doubleMetricValues[i]);
        GenericRow genericRow = new GenericRow();
        genericRow.init(map);
        segmentData[i] = genericRow;
    }
    // Create an index segment with the random dimension and metric values.
    final Schema schema = new Schema();
    schema.addField(new DimensionFieldSpec(DIMENSION_NAME, FieldSpec.DataType.STRING, true));
    schema.addField(new MetricFieldSpec(INT_METRIC_NAME, FieldSpec.DataType.INT));
    schema.addField(new MetricFieldSpec(LONG_METRIC_NAME, FieldSpec.DataType.LONG));
    schema.addField(new MetricFieldSpec(FLOAT_METRIC_NAME, FieldSpec.DataType.FLOAT));
    schema.addField(new MetricFieldSpec(DOUBLE_METRIC_NAME, FieldSpec.DataType.DOUBLE));
    schema.addField(new MetricFieldSpec(NO_DICT_INT_METRIC_NAME, FieldSpec.DataType.INT));
    schema.addField(new MetricFieldSpec(NO_DICT_LONG_METRIC_NAME, FieldSpec.DataType.LONG));
    schema.addField(new MetricFieldSpec(NO_DICT_FLOAT_METRIC_NAME, FieldSpec.DataType.FLOAT));
    schema.addField(new MetricFieldSpec(NO_DICT_DOUBLE_METRIC_NAME, FieldSpec.DataType.DOUBLE));
    SegmentGeneratorConfig config = new SegmentGeneratorConfig(schema);
    FileUtils.deleteQuietly(new File(INDEX_DIR_PATH));
    config.setOutDir(INDEX_DIR_PATH);
    config.setSegmentName(SEGMENT_NAME);
    config.setRawIndexCreationColumns(Arrays.asList(NO_DICT_INT_METRIC_NAME, NO_DICT_LONG_METRIC_NAME, NO_DICT_FLOAT_METRIC_NAME, NO_DICT_DOUBLE_METRIC_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);
    Map<String, BaseOperator> dataSourceMap = new HashMap<>();
    for (String column : indexSegment.getColumnNames()) {
        dataSourceMap.put(column, indexSegment.getDataSource(column));
    }
    // Get a data fetcher for the index segment.
    _dataFetcher = new DataFetcher(dataSourceMap);
}