co.cask.cdap.data.stream

Class StreamDataFileReader

java.lang.Object

co.cask.cdap.data.stream.StreamDataFileReader

All Implemented Interfaces:

FileReader<PositionStreamEvent,Long>, PositionReporter<Long>, Closeable, AutoCloseable

@NotThreadSafe
public final class StreamDataFileReader
extends Object
implements FileReader<PositionStreamEvent,Long>

Class for reading data file written by StreamDataFileWriter.

See Also:

StreamDataFileWriter

Method Summary

Methods

Modifier and Type	Method and Description
void	close()
static StreamDataFileReader	create(com.google.common.io.InputSupplier<? extends SeekableInputStream> eventInputSupplier) Opens a new StreamDataFileReader with the given inputs.
static StreamDataFileReader	createByStartTime(com.google.common.io.InputSupplier<? extends SeekableInputStream> eventInputSupplier, com.google.common.io.InputSupplier<? extends InputStream> indexInputSupplier, long startTime) Opens a new StreamDataFileReader with the given inputs that starts reading events that are written at or after the given timestamp.
static StreamDataFileReader	createWithOffset(com.google.common.io.InputSupplier<? extends SeekableInputStream> eventInputSupplier, com.google.common.io.InputSupplier<? extends InputStream> indexInputSupplier, long offset) Opens a new StreamDataFileReader with the given inputs, which starts reading events at a the smallest event position that is larger than or equal to the given offset.
Long	getPosition() Returns the position information.
void	initialize() Opens this reader to prepare for consumption.
int	read(Collection<? super PositionStreamEvent> events, int maxEvents, long timeout, TimeUnit unit) Reads as much data as possible until maxEvents.
int	read(Collection<? super PositionStreamEvent> events, int maxEvents, long timeout, TimeUnit unit, ReadFilter readFilter) Reads as much data as possible that passes the given filter until maxEvents.

Methods inherited from class java.lang.Object

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

Method Detail

create

public static StreamDataFileReader create(com.google.common.io.InputSupplier<? extends SeekableInputStream> eventInputSupplier)

Opens a new StreamDataFileReader with the given inputs.

Parameters:

eventInputSupplier - An InputSupplier for providing the stream to read events.

Returns:

A new instance of StreamDataFileReader.

createByStartTime

public static StreamDataFileReader createByStartTime(com.google.common.io.InputSupplier<? extends SeekableInputStream> eventInputSupplier,
                                     @Nullable
                                     com.google.common.io.InputSupplier<? extends InputStream> indexInputSupplier,
                                     long startTime)

Opens a new StreamDataFileReader with the given inputs that starts reading events that are written at or after the given timestamp.

Parameters:

eventInputSupplier - An InputSupplier for providing the stream to read events.

indexInputSupplier - An InputSupplier for providing the stream to read event index.

startTime - Timestamp in milliseconds for the event time to start reading with.

Returns:

A new instance of StreamDataFileReader.

createWithOffset

public static StreamDataFileReader createWithOffset(com.google.common.io.InputSupplier<? extends SeekableInputStream> eventInputSupplier,
                                    @Nullable
                                    com.google.common.io.InputSupplier<? extends InputStream> indexInputSupplier,
                                    long offset)

Opens a new StreamDataFileReader with the given inputs, which starts reading events at a the smallest event position that is larger than or equal to the given offset.

Parameters:

eventInputSupplier - An InputSupplier for providing the stream to read events.

indexInputSupplier - An InputSupplier for providing the stream to read event index.

offset - An arbitrary event file offset.

Returns:

A new instance of StreamDataFileReader.

getPosition

public Long getPosition()

Description copied from interface: PositionReporter

Returns the position information.

Specified by:

getPosition in interface PositionReporter<Long>

Returns:

An instance of type <P>

initialize

public void initialize()
                throws IOException

Opens this reader to prepare for consumption. Calling this method is optional as the read(java.util.Collection, int, long, java.util.concurrent.TimeUnit, co.cask.cdap.data.file.ReadFilter) method would do the initialization if this method hasn't been called.

Specified by:

initialize in interface FileReader<PositionStreamEvent,Long>

Throws:

IOException - If there is error initializing.

close

public void close()
           throws IOException

Specified by:

close in interface Closeable

Specified by:

close in interface AutoCloseable

Throws:

IOException

read

public int read(Collection<? super PositionStreamEvent> events,
       int maxEvents,
       long timeout,
       TimeUnit unit)
         throws IOException,
                InterruptedException

Description copied from interface: FileReader

Reads as much data as possible until maxEvents.

Specified by:

read in interface FileReader<PositionStreamEvent,Long>

Parameters:

events - Collection for storing data read.

maxEvents - Maximum number of events to read.

timeout - Maximum of time to spend on trying to read events

unit - Unit for the timeout.

Returns:

Number of events read, could be 0. If -1 is returned, meaning it reached EOF and all sub-sequence call to this method will also return -1.

Throws:

IOException - If failed to read.

InterruptedException - If the reading is interrupted.

read

public int read(Collection<? super PositionStreamEvent> events,
       int maxEvents,
       long timeout,
       TimeUnit unit,
       ReadFilter readFilter)
         throws IOException,
                InterruptedException

Description copied from interface: FileReader

Reads as much data as possible that passes the given filter until maxEvents.

Specified by:

read in interface FileReader<PositionStreamEvent,Long>

Parameters:

events - Collection for storing data read.

maxEvents - Maximum number of events to read.

timeout - Maximum of time to spend on trying to read events

unit - Unit for the timeout.

readFilter - Filter to apply during reading

Returns:

Number of events read, could be 0. If -1 is returned, meaning it reached EOF and all sub-sequence call to this method will also return -1.

Throws:

IOException - If failed to read.

InterruptedException - If the reading is interrupted.