Examples with RepeatedValueVector org.apache.drill.exec.vector.complex.RepeatedValueVector used on opensource projects

Example 1 with RepeatedValueVector

use of org.apache.drill.exec.vector.complex.RepeatedValueVector in project drill by apache.

the class ParquetColumnMetadata method makeRepeatedFixedWidthReader.

@SuppressWarnings("resource")
FixedWidthRepeatedReader makeRepeatedFixedWidthReader(ParquetRecordReader reader, int recordsPerBatch) throws Exception {
    final RepeatedValueVector repeatedVector = RepeatedValueVector.class.cast(vector);
    ColumnReader<?> dataReader = ColumnReaderFactory.createFixedColumnReader(reader, true, column, columnChunkMetaData, recordsPerBatch, repeatedVector.getDataVector(), se);
    return new FixedWidthRepeatedReader(reader, dataReader, getTypeLengthInBits(column.getType()), UNDEFINED_LENGTH, column, columnChunkMetaData, false, repeatedVector, se);
}

Example 2 with RepeatedValueVector

use of org.apache.drill.exec.vector.complex.RepeatedValueVector in project drill by axbaretto.

the class FlattenRecordBatch method getFlattenFieldTransferPair.

/**
 * The data layout is the same for the actual data within a repeated field, as it is in a scalar vector for
 * the same sql type. For example, a repeated int vector has a vector of offsets into a regular int vector to
 * represent the lists. As the data layout for the actual values in the same in the repeated vector as in the
 * scalar vector of the same type, we can avoid making individual copies for the column being flattened, and just
 * use vector copies between the inner vector of the repeated field to the resulting scalar vector from the flatten
 * operation. This is completed after we determine how many records will fit (as we will hit either a batch end, or
 * the end of one of the other vectors while we are copying the data of the other vectors alongside each new flattened
 * value coming out of the repeated field.)
 */
@SuppressWarnings("resource")
private TransferPair getFlattenFieldTransferPair(FieldReference reference) {
    final TypedFieldId fieldId = incoming.getValueVectorId(popConfig.getColumn());
    final Class<?> vectorClass = incoming.getSchema().getColumn(fieldId.getFieldIds()[0]).getValueClass();
    final ValueVector flattenField = incoming.getValueAccessorById(vectorClass, fieldId.getFieldIds()).getValueVector();
    TransferPair tp = null;
    if (flattenField instanceof RepeatedMapVector) {
        tp = ((RepeatedMapVector) flattenField).getTransferPairToSingleMap(reference.getAsNamePart().getName(), oContext.getAllocator());
    } else if (!(flattenField instanceof RepeatedValueVector)) {
        if (incoming.getRecordCount() != 0) {
            throw UserException.unsupportedError().message("Flatten does not support inputs of non-list values.").build(logger);
        }
        logger.error("Cannot cast {} to RepeatedValueVector", flattenField);
        // when incoming recordCount is 0, don't throw exception since the type being seen here is not solid
        final ValueVector vv = new RepeatedMapVector(flattenField.getField(), oContext.getAllocator(), null);
        tp = RepeatedValueVector.class.cast(vv).getTransferPair(reference.getAsNamePart().getName(), oContext.getAllocator());
    } else {
        final ValueVector vvIn = RepeatedValueVector.class.cast(flattenField).getDataVector();
        // vvIn may be null because of fast schema return for repeated list vectors
        if (vvIn != null) {
            tp = vvIn.getTransferPair(reference.getAsNamePart().getName(), oContext.getAllocator());
        }
    }
    return tp;
}

Example 3 with RepeatedValueVector

use of org.apache.drill.exec.vector.complex.RepeatedValueVector in project drill by axbaretto.

the class FlattenRecordBatch method setFlattenVector.

@SuppressWarnings("resource")
private void setFlattenVector() {
    final TypedFieldId typedFieldId = incoming.getValueVectorId(popConfig.getColumn());
    final MaterializedField field = incoming.getSchema().getColumn(typedFieldId.getFieldIds()[0]);
    final RepeatedValueVector vector;
    final ValueVector inVV = incoming.getValueAccessorById(field.getValueClass(), typedFieldId.getFieldIds()).getValueVector();
    if (!(inVV instanceof RepeatedValueVector)) {
        if (incoming.getRecordCount() != 0) {
            throw UserException.unsupportedError().message("Flatten does not support inputs of non-list values.").build(logger);
        }
        // when incoming recordCount is 0, don't throw exception since the type being seen here is not solid
        logger.error("setFlattenVector cast failed and recordcount is 0, create empty vector anyway.");
        vector = new RepeatedMapVector(field, oContext.getAllocator(), null);
    } else {
        vector = RepeatedValueVector.class.cast(inVV);
    }
    flattener.setFlattenField(vector);
}

Example 4 with RepeatedValueVector

use of org.apache.drill.exec.vector.complex.RepeatedValueVector in project drill by axbaretto.

the class ParquetColumnMetadata method makeRepeatedFixedWidthReader.

@SuppressWarnings("resource")
FixedWidthRepeatedReader makeRepeatedFixedWidthReader(ParquetRecordReader reader, int recordsPerBatch) throws Exception {
    final RepeatedValueVector repeatedVector = RepeatedValueVector.class.cast(vector);
    ColumnReader<?> dataReader = ColumnReaderFactory.createFixedColumnReader(reader, true, column, columnChunkMetaData, recordsPerBatch, repeatedVector.getDataVector(), se);
    return new FixedWidthRepeatedReader(reader, dataReader, getTypeLengthInBits(column.getType()), UNDEFINED_LENGTH, column, columnChunkMetaData, false, repeatedVector, se);
}

Example 5 with RepeatedValueVector

use of org.apache.drill.exec.vector.complex.RepeatedValueVector in project drill by axbaretto.

the class TestRecordBatchSizer method testSizerRepeatedFixedWidth.

@Test
public void testSizerRepeatedFixedWidth() {
    BatchSchema schema = new SchemaBuilder().addArray("a", MinorType.BIGINT).addArray("b", MinorType.FLOAT8).build();
    RowSetBuilder builder = fixture.rowSetBuilder(schema);
    for (long i = 0; i < 10; i++) {
        builder.addRow(new long[] { 1, 2, 3, 4, 5 }, new double[] { i * 0.1, i * 0.1, i * 0.1, i * 0.2, i * 0.3 });
    }
    RowSet rows = builder.build();
    // Run the record batch sizer on the resulting batch.
    RecordBatchSizer sizer = new RecordBatchSizer(rows.container());
    assertEquals(2, sizer.columns().size());
    /**
     * stdDataSize:8*10, stdNetSize:8*10+4, dataSizePerEntry:5*8, netSizePerEntry:5*8+4,
     * totalDataSize:5*8*10, totalNetSize:5*8*10+5*8, valueCount:10,
     * elementCount:50, estElementCountPerArray:5, isVariableWidth:false
     */
    verifyColumnValues(sizer.columns().get("a"), 80, 84, 40, 44, 400, 440, 10, 50, 5, false);
    verifyColumnValues(sizer.columns().get("b"), 80, 84, 40, 44, 400, 440, 10, 50, 5, false);
    SingleRowSet empty = fixture.rowSet(schema);
    VectorAccessible accessible = empty.vectorAccessible();
    UInt4Vector offsetVector;
    ValueVector dataVector;
    for (VectorWrapper<?> vw : accessible) {
        ValueVector v = vw.getValueVector();
        RecordBatchSizer.ColumnSize colSize = sizer.getColumn(v.getField().getName());
        // Allocates to nearest power of two
        colSize.allocateVector(v, testRowCount);
        offsetVector = ((RepeatedValueVector) v).getOffsetVector();
        assertEquals((Integer.highestOneBit(testRowCount) << 1), offsetVector.getValueCapacity());
        dataVector = ((RepeatedValueVector) v).getDataVector();
        assertEquals(Integer.highestOneBit((testRowCount * 5) << 1), dataVector.getValueCapacity());
        v.clear();
        // Allocates the same as value passed since it is already power of two.
        // -1 is done for adjustment needed for offset vector.
        colSize.allocateVector(v, testRowCountPowerTwo - 1);
        offsetVector = ((RepeatedValueVector) v).getOffsetVector();
        assertEquals(testRowCountPowerTwo, offsetVector.getValueCapacity());
        dataVector = ((RepeatedValueVector) v).getDataVector();
        assertEquals(Integer.highestOneBit((testRowCountPowerTwo - 1) * 5) << 1, dataVector.getValueCapacity());
        v.clear();
        // Allocate for max rows.
        colSize.allocateVector(v, ValueVector.MAX_ROW_COUNT - 1);
        offsetVector = ((RepeatedValueVector) v).getOffsetVector();
        assertEquals(ValueVector.MAX_ROW_COUNT, offsetVector.getValueCapacity());
        dataVector = ((RepeatedValueVector) v).getDataVector();
        assertEquals(Integer.highestOneBit(((ValueVector.MAX_ROW_COUNT - 1) * 5) << 1), dataVector.getValueCapacity());
        v.clear();
        // Allocate for 0 rows. should atleast do allocation for 1 row.
        colSize.allocateVector(v, 0);
        offsetVector = ((RepeatedValueVector) v).getOffsetVector();
        assertEquals(ValueVector.MIN_ROW_COUNT + 1, offsetVector.getValueCapacity());
        dataVector = ((RepeatedValueVector) v).getDataVector();
        assertEquals(ValueVector.MIN_ROW_COUNT, dataVector.getValueCapacity());
        v.clear();
    }
    empty.clear();
    rows.clear();
}