Search in sources :

Example 1 with HashJoinType

use of org.apache.flink.table.runtime.operators.join.HashJoinType in project flink by apache.

the class BatchExecHashJoin method translateToPlanInternal.

@Override
@SuppressWarnings("unchecked")
protected Transformation<RowData> translateToPlanInternal(PlannerBase planner, ExecNodeConfig config) {
    ExecEdge leftInputEdge = getInputEdges().get(0);
    ExecEdge rightInputEdge = getInputEdges().get(1);
    Transformation<RowData> leftInputTransform = (Transformation<RowData>) leftInputEdge.translateToPlan(planner);
    Transformation<RowData> rightInputTransform = (Transformation<RowData>) rightInputEdge.translateToPlan(planner);
    // get input types
    RowType leftType = (RowType) leftInputEdge.getOutputType();
    RowType rightType = (RowType) rightInputEdge.getOutputType();
    JoinUtil.validateJoinSpec(joinSpec, leftType, rightType, false);
    int[] leftKeys = joinSpec.getLeftKeys();
    int[] rightKeys = joinSpec.getRightKeys();
    LogicalType[] keyFieldTypes = IntStream.of(leftKeys).mapToObj(leftType::getTypeAt).toArray(LogicalType[]::new);
    RowType keyType = RowType.of(keyFieldTypes);
    GeneratedJoinCondition condFunc = JoinUtil.generateConditionFunction(config.getTableConfig(), joinSpec.getNonEquiCondition().orElse(null), leftType, rightType);
    // projection for equals
    GeneratedProjection leftProj = ProjectionCodeGenerator.generateProjection(new CodeGeneratorContext(config.getTableConfig()), "HashJoinLeftProjection", leftType, keyType, leftKeys);
    GeneratedProjection rightProj = ProjectionCodeGenerator.generateProjection(new CodeGeneratorContext(config.getTableConfig()), "HashJoinRightProjection", rightType, keyType, rightKeys);
    Transformation<RowData> buildTransform;
    Transformation<RowData> probeTransform;
    GeneratedProjection buildProj;
    GeneratedProjection probeProj;
    int[] buildKeys;
    int[] probeKeys;
    RowType buildType;
    RowType probeType;
    int buildRowSize;
    long buildRowCount;
    long probeRowCount;
    boolean reverseJoin = !leftIsBuild;
    if (leftIsBuild) {
        buildTransform = leftInputTransform;
        buildProj = leftProj;
        buildType = leftType;
        buildRowSize = estimatedLeftAvgRowSize;
        buildRowCount = estimatedLeftRowCount;
        buildKeys = leftKeys;
        probeTransform = rightInputTransform;
        probeProj = rightProj;
        probeType = rightType;
        probeRowCount = estimatedLeftRowCount;
        probeKeys = rightKeys;
    } else {
        buildTransform = rightInputTransform;
        buildProj = rightProj;
        buildType = rightType;
        buildRowSize = estimatedRightAvgRowSize;
        buildRowCount = estimatedRightRowCount;
        buildKeys = rightKeys;
        probeTransform = leftInputTransform;
        probeProj = leftProj;
        probeType = leftType;
        probeRowCount = estimatedLeftRowCount;
        probeKeys = leftKeys;
    }
    // operator
    StreamOperatorFactory<RowData> operator;
    FlinkJoinType joinType = joinSpec.getJoinType();
    HashJoinType hashJoinType = HashJoinType.of(leftIsBuild, joinType.isLeftOuter(), joinType.isRightOuter(), joinType == FlinkJoinType.SEMI, joinType == FlinkJoinType.ANTI);
    if (LongHashJoinGenerator.support(hashJoinType, keyType, joinSpec.getFilterNulls())) {
        operator = LongHashJoinGenerator.gen(config.getTableConfig(), hashJoinType, keyType, buildType, probeType, buildKeys, probeKeys, buildRowSize, buildRowCount, reverseJoin, condFunc);
    } else {
        operator = SimpleOperatorFactory.of(HashJoinOperator.newHashJoinOperator(hashJoinType, condFunc, reverseJoin, joinSpec.getFilterNulls(), buildProj, probeProj, tryDistinctBuildRow, buildRowSize, buildRowCount, probeRowCount, keyType));
    }
    long managedMemory = config.get(ExecutionConfigOptions.TABLE_EXEC_RESOURCE_HASH_JOIN_MEMORY).getBytes();
    return ExecNodeUtil.createTwoInputTransformation(buildTransform, probeTransform, createTransformationName(config), createTransformationDescription(config), operator, InternalTypeInfo.of(getOutputType()), probeTransform.getParallelism(), managedMemory);
}
Also used : Transformation(org.apache.flink.api.dag.Transformation) ExecEdge(org.apache.flink.table.planner.plan.nodes.exec.ExecEdge) HashJoinType(org.apache.flink.table.runtime.operators.join.HashJoinType) CodeGeneratorContext(org.apache.flink.table.planner.codegen.CodeGeneratorContext) RowType(org.apache.flink.table.types.logical.RowType) LogicalType(org.apache.flink.table.types.logical.LogicalType) FlinkJoinType(org.apache.flink.table.runtime.operators.join.FlinkJoinType) RowData(org.apache.flink.table.data.RowData) GeneratedJoinCondition(org.apache.flink.table.runtime.generated.GeneratedJoinCondition) GeneratedProjection(org.apache.flink.table.runtime.generated.GeneratedProjection)

Aggregations

Transformation (org.apache.flink.api.dag.Transformation)1 RowData (org.apache.flink.table.data.RowData)1 CodeGeneratorContext (org.apache.flink.table.planner.codegen.CodeGeneratorContext)1 ExecEdge (org.apache.flink.table.planner.plan.nodes.exec.ExecEdge)1 GeneratedJoinCondition (org.apache.flink.table.runtime.generated.GeneratedJoinCondition)1 GeneratedProjection (org.apache.flink.table.runtime.generated.GeneratedProjection)1 FlinkJoinType (org.apache.flink.table.runtime.operators.join.FlinkJoinType)1 HashJoinType (org.apache.flink.table.runtime.operators.join.HashJoinType)1 LogicalType (org.apache.flink.table.types.logical.LogicalType)1 RowType (org.apache.flink.table.types.logical.RowType)1