org.apache.hadoop.mapreduce.lib.join
Class CompositeRecordReader.JoinCollector

java.lang.Object
  org.apache.hadoop.mapreduce.lib.join.CompositeRecordReader.JoinCollector

Enclosing class:

CompositeRecordReader<K extends org.apache.hadoop.io.WritableComparable<?>,V extends org.apache.hadoop.io.Writable,X extends org.apache.hadoop.io.Writable>

public class CompositeRecordReader.JoinCollector
extends Object

Collector for join values. This accumulates values for a given key from the child RecordReaders. If one or more child RR contain duplicate keys, this will emit the cross product of the associated values until exhausted.

Constructor Summary

CompositeRecordReader.JoinCollector(int card)
Construct a collector capable of handling the specified number of children.

Method Summary

void	add(int id, ResetableIterator<X> i) Register a given iterator at position id.
void	clear() Clear all state information.
void	close() Close all child iterators.
boolean	flush(TupleWritable value) Write the next value into key, value as accepted by the operation associated with this set of RecordReaders.
boolean	hasNext() Returns false if exhausted or if reset(K) has not been called.
K	key() Return the key associated with this collection.
protected boolean	next(TupleWritable val) Populate Tuple from iterators.
boolean	replay(TupleWritable val) Replay the last Tuple emitted.
void	reset(K key) Codify the contents of the collector to be iterated over.

Methods inherited from class java.lang.Object

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

Constructor Detail

CompositeRecordReader.JoinCollector

public CompositeRecordReader.JoinCollector(int card)

Construct a collector capable of handling the specified number of children.

Method Detail

add

public void add(int id,
                ResetableIterator<X> i)
         throws IOException

Register a given iterator at position id.

Throws:

IOException

key

public K key()

Return the key associated with this collection.

reset

public void reset(K key)

Codify the contents of the collector to be iterated over. When this is called, all RecordReaders registered for this key should have added ResetableIterators.

clear

public void clear()

Clear all state information.

hasNext

public boolean hasNext()

Returns false if exhausted or if reset(K) has not been called.

next

protected boolean next(TupleWritable val)
                throws IOException

Populate Tuple from iterators. It should be the case that, given iterators i_1...i_n over values from sources s_1...s_n sharing key k, repeated calls to next should yield I x I.

Throws:

IOException

replay

public boolean replay(TupleWritable val)
               throws IOException

Replay the last Tuple emitted.

Throws:

IOException

close

public void close()
           throws IOException

Close all child iterators.

Throws:

IOException

flush

public boolean flush(TupleWritable value)
              throws IOException

Write the next value into key, value as accepted by the operation associated with this set of RecordReaders.

Throws:

IOException