org.apache.hadoop.hive.ql.exec
Class ExecMapperContext

java.lang.Object
  org.apache.hadoop.hive.ql.exec.ExecMapperContext

public class ExecMapperContext
extends Object

Field Summary

static org.apache.commons.logging.Log

l4j

Constructor Summary

ExecMapperContext()

Method Summary

void	clear()
String	getCurrentBigBucketFile()
String	getCurrentInputFile()
Map<String,FetchOperator>	getFetchOperators()
String	getFileId()
IOContext	getIoCxt()
org.apache.hadoop.mapred.JobConf	getJc()
String	getLastInputFile()
MapredLocalWork	getLocalWork()
boolean	inputFileChanged() For CompbineFileInputFormat, the mapper's input file will be changed on the fly, and the input file name is passed to jobConf by shims/initNextRecordReader.
void	resetRow() Reset the execution context for each new row.
void	setCurrentBigBucketFile(String currentBigBucketFile)
void	setCurrentInputFile(String currentInputFile)
void	setFetchOperators(Map<String,FetchOperator> fetchOperators)
void	setFileId(String fileId)
void	setIoCxt(IOContext ioCxt)
void	setJc(org.apache.hadoop.mapred.JobConf jc)
void	setLastInputFile(String lastInputFile)
void	setLocalWork(MapredLocalWork localWork)

Methods inherited from class java.lang.Object

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

Field Detail

l4j

public static final org.apache.commons.logging.Log l4j

Constructor Detail

ExecMapperContext

public ExecMapperContext()

Method Detail

getCurrentBigBucketFile

public String getCurrentBigBucketFile()

setCurrentBigBucketFile

public void setCurrentBigBucketFile(String currentBigBucketFile)

clear

public void clear()

inputFileChanged

public boolean inputFileChanged()

For CompbineFileInputFormat, the mapper's input file will be changed on the fly, and the input file name is passed to jobConf by shims/initNextRecordReader. If the map local work has any mapping depending on the current mapper's input file, the work need to clear context and re-initialization after the input file changed. This is first introduced to process bucket map join.

Returns:

is the input file changed?

resetRow

public void resetRow()

Reset the execution context for each new row. This function should be called only once at the root of the operator tree -- ExecMapper.map(). Note: this function should be kept minimum since it is called for each input row.

getLastInputFile

public String getLastInputFile()

setLastInputFile

public void setLastInputFile(String lastInputFile)

getCurrentInputFile

public String getCurrentInputFile()

setCurrentInputFile

public void setCurrentInputFile(String currentInputFile)

getJc

public org.apache.hadoop.mapred.JobConf getJc()

setJc

public void setJc(org.apache.hadoop.mapred.JobConf jc)

getLocalWork

public MapredLocalWork getLocalWork()

setLocalWork

public void setLocalWork(MapredLocalWork localWork)

getFileId

public String getFileId()

setFileId

public void setFileId(String fileId)

getFetchOperators

public Map<String,FetchOperator> getFetchOperators()

setFetchOperators

public void setFetchOperators(Map<String,FetchOperator> fetchOperators)

getIoCxt

public IOContext getIoCxt()

setIoCxt

public void setIoCxt(IOContext ioCxt)