co.cask.format.text.input

Class CombineTextInputFormat

java.lang.Object

org.apache.hadoop.mapreduce.InputFormat<K,V>

org.apache.hadoop.mapreduce.lib.input.FileInputFormat<K,V>

org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat<org.apache.hadoop.io.NullWritable,co.cask.cdap.api.data.format.StructuredRecord>

co.cask.format.text.input.CombineTextInputFormat

public class CombineTextInputFormat
extends org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat<org.apache.hadoop.io.NullWritable,co.cask.cdap.api.data.format.StructuredRecord>

Combined input format that tracks which file each text record was read from and optionally emits a file header as the first record for each split.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	CombineTextInputFormat.WrapperReader A wrapper class that's responsible for delegating to a corresponding RecordReader in PathTrackingInputFormat.

Nested classes/interfaces inherited from class org.apache.hadoop.mapreduce.lib.input.FileInputFormat

org.apache.hadoop.mapreduce.lib.input.FileInputFormat.Counter

Field Summary

Fields inherited from class org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat

SPLIT_MINSIZE_PERNODE, SPLIT_MINSIZE_PERRACK

Fields inherited from class org.apache.hadoop.mapreduce.lib.input.FileInputFormat

INPUT_DIR, INPUT_DIR_RECURSIVE, NUM_INPUT_FILES, PATHFILTER_CLASS, SPLIT_MAXSIZE, SPLIT_MINSIZE

Constructor Summary

Constructors

Constructor and Description
CombineTextInputFormat()

Method Summary

Modifier and Type	Method and Description
org.apache.hadoop.mapreduce.RecordReader<org.apache.hadoop.io.NullWritable,co.cask.cdap.api.data.format.StructuredRecord>	createRecordReader(org.apache.hadoop.mapreduce.InputSplit split, org.apache.hadoop.mapreduce.TaskAttemptContext context) Creates a RecordReader that delegates to some other RecordReader for each path in the input split.
List<org.apache.hadoop.mapreduce.InputSplit>	getSplits(org.apache.hadoop.mapreduce.JobContext job) Converts the CombineFileSplits derived by CombineFileInputFormat into CombineHeaderFileSplits that optionally keep track of the header for each file.

Methods inherited from class org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat

createPool, createPool, getFileBlockLocations, isSplitable, setMaxSplitSize, setMinSplitSizeNode, setMinSplitSizeRack

Methods inherited from class org.apache.hadoop.mapreduce.lib.input.FileInputFormat

addInputPath, addInputPathRecursively, addInputPaths, computeSplitSize, getBlockIndex, getFormatMinSplitSize, getInputDirRecursive, getInputPathFilter, getInputPaths, getMaxSplitSize, getMinSplitSize, listStatus, makeSplit, setInputDirRecursive, setInputPathFilter, setInputPaths, setInputPaths, setMaxInputSplitSize, setMinInputSplitSize

Methods inherited from class java.lang.Object

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

Constructor Detail

CombineTextInputFormat

public CombineTextInputFormat()

Method Detail

getSplits

public List<org.apache.hadoop.mapreduce.InputSplit> getSplits(org.apache.hadoop.mapreduce.JobContext job)
                                                       throws IOException

Converts the CombineFileSplits derived by CombineFileInputFormat into CombineHeaderFileSplits that optionally keep track of the header for each file. It is assumed that every file has the same header. It would be possible to read the header for each individual file, but the use case for that is unclear and it could potentially add a decent amount of overhead if there are a bunch of small files.

Overrides:

getSplits in class org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat<org.apache.hadoop.io.NullWritable,co.cask.cdap.api.data.format.StructuredRecord>

Throws:

IOException

createRecordReader

public org.apache.hadoop.mapreduce.RecordReader<org.apache.hadoop.io.NullWritable,co.cask.cdap.api.data.format.StructuredRecord> createRecordReader(org.apache.hadoop.mapreduce.InputSplit split,
                                                                                                                                                    org.apache.hadoop.mapreduce.TaskAttemptContext context)
                                                                                                                                             throws IOException

Creates a RecordReader that delegates to some other RecordReader for each path in the input split. The header for each file is set in the context Configuration to make it available to the delegate RecordReaders.

Specified by:

createRecordReader in class org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat<org.apache.hadoop.io.NullWritable,co.cask.cdap.api.data.format.StructuredRecord>

Throws:

IOException