io.cdap.cdap.api.data.batch

Class Split

java.lang.Object

io.cdap.cdap.api.data.batch.Split

Direct Known Subclasses:

SimpleSplit, TableSplit, TimeseriesTable.InputSplit

public abstract class Split
extends Object

Defines split of the dataset. Typically a dataset is spit into multiple chunks that are fed into a batch job. Sub-class can implements the writeExternal(DataOutput) and the readExternal(DataInput) and have a public default constructor (constructor without arguments) to provide a more efficient serialization mechanism.

Constructor Summary

Constructors

Constructor and Description
Split()

Method Summary

Modifier and Type	Method and Description
long	getLength() By default assume that the size of each split is roughly the same.
void	readExternal(DataInput in) Deserialize this split.
void	writeExternal(DataOutput out) Serializing this split.

Methods inherited from class java.lang.Object

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

Constructor Detail

Split

public Split()

Method Detail

getLength

public long getLength()

By default assume that the size of each split is roughly the same.

Returns:

Optional split length. Used only for tracking split consuming completion percentile.

writeExternal

public void writeExternal(DataOutput out)
                   throws IOException

Serializing this split.

Parameters:

out - the DataOutput for writing out this split information

Throws:

IOException - if failed to serialize

readExternal

public void readExternal(DataInput in)
                  throws IOException

Deserialize this split.

Parameters:

in - the DataInput to read back the split information.

Throws:

IOException - if failed to deserialize