org.apache.hadoop.hive.ql.optimizer
Class GenMRUnion1

java.lang.Object
  org.apache.hadoop.hive.ql.optimizer.GenMRUnion1

All Implemented Interfaces:

NodeProcessor

public class GenMRUnion1
extends Object
implements NodeProcessor

Processor for the rule - TableScan followed by Union.

Constructor Summary

GenMRUnion1()

Method Summary

Object

process(Node nd, Stack<Node> stack, NodeProcessorCtx opProcCtx, Object... nodeOutputs) Union Operator encountered .

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

GenMRUnion1

public GenMRUnion1()

Method Detail

process

public Object process(Node nd,
                      Stack<Node> stack,
                      NodeProcessorCtx opProcCtx,
                      Object... nodeOutputs)
               throws SemanticException

Union Operator encountered . Currently, the algorithm is pretty simple: If all the sub-queries are map-only, don't do anything. However, if there is a mapjoin followed by the union, merge at the union Otherwise, insert a FileSink on top of all the sub-queries. This can be optimized later on.

Specified by:

process in interface NodeProcessor

Parameters:

nd - the file sink operator encountered

opProcCtx - context

nodeOutputs - A variable argument list of outputs from other nodes in the walk

Returns:

Object to be returned by the process call

Throws:

SemanticException