Modifier and Type	Class and Description
private class	DefaultOutputSink.FlushCompletionHandler Flush `CompletionHandler`, which will be passed on each `Http2Stream` write to make sure the data reached the wires.
private static class	DefaultOutputSink.OutputQueueRecord

Modifier and Type	Field and Description
private final AtomicInteger	availStreamWindowSize
private BundleQueue<CompletionHandler<Http2Stream>>	flushHandlersQueue
private final Object	flushHandlersSync
private final Http2Session	http2Session
private volatile boolean	isLastFrameQueued
private static final Logger	LOGGER
private static final int	MAX_OUTPUT_QUEUE_SIZE
pack-priv final TaskQueue<DefaultOutputSink.OutputQueueRecord>	outputQueue
private final Http2Stream	stream
private static final DefaultOutputSink.OutputQueueRecord	TERMINATING_QUEUE_RECORD
private Termination	terminationFlag
private final AtomicInteger	unflushedWritesCounter
private static final int	ZERO_QUEUE_RECORD_SIZE

Access	Constructor and Description
pack-priv	DefaultOutputSink(final Http2Stream stream)

Method Summary

Modifier and Type	Method and Description
private void	addOutputQueueRecord(DefaultOutputSink.OutputQueueRecord outputQueueRecord)
private void	assertReady()
public boolean	canWrite() Implements org.glassfish.grizzly.http2.StreamOutputSink.canWrite.
private int	Returns: the amount of data that may be written. checkOutputWindow(final long check the provided size against the window size limit. size) The method is responsible for checking the current output window size.
public synchronized void	close() Implements org.glassfish.grizzly.http2.StreamOutputSink.close. Closes the output sink by adding last DataFrame with the FIN flag to a queue.
private DefaultOutputSink.OutputQueueRecord	createOutputQueueRecord(final Buffer data, final HttpContent httpContent, final DefaultOutputSink.FlushCompletionHandler flushCompletionHandler, boolean isLast, final boolean isZeroSizeData)
public void	flush(final CompletionHandler<Http2Stream> `CompletionHandler` to be notified completionHandler) Implements org.glassfish.grizzly.http2.StreamOutputSink.flush. Flush `Http2Stream` output and notify `CompletionHandler` once all output data has been flushed.
private void	flushToConnectionOutputSink(final List<Http2Frame> headerFrames, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner, final boolean sendFIN)
private void	flushToConnectionOutputSink(final Buffer data, final DefaultOutputSink.FlushCompletionHandler flushCompletionHandler, final boolean sendFIN)
private void	flushToConnectionOutputSink(final List<Http2Frame> headerFrames, final Buffer data, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner, final boolean sendFIN)
public int	getUnflushedWritesCount() Implements org.glassfish.grizzly.http2.StreamOutputSink.getUnflushedWritesCount. Returns: the number of writes (not bytes), that haven't reached network layer
public boolean	isClosed() Implements org.glassfish.grizzly.http2.StreamOutputSink.isClosed.
private boolean	isTerminated()
private boolean	isWantToWrite()
public void	notifyWritePossible(final WriteHandler writeHandler) Implements org.glassfish.grizzly.http2.StreamOutputSink.notifyWritePossible.
public void	onPeerWindowUpdate(final int the delta. delta) Implements org.glassfish.grizzly.http2.StreamOutputSink.onPeerWindowUpdate. The method is called by HTTP2 Filter once WINDOW_UPDATE message comes for this `Http2Stream`.
private Buffer	prepareDataToSend(final boolean isRecordNull, final boolean isLast, final Buffer data, final boolean isZeroSizeData)
private void	releaseWriteQueueSpace(final int justSentBytes, final boolean isAtomic, final boolean isEndOfChunk)
private int	reserveWriteQueueSpace(final int spaceToReserve)
private void	sendTrailers(final CompletionHandler<WriteResult<K, L>> completionHandler, final HttpTrailer httpContent)
private Buffer	splitOutputBufferIfNeeded(final Buffer buffer, final int length)
public synchronized void	terminate(final Termination terminationFlag) Implements org.glassfish.grizzly.http2.StreamOutputSink.terminate. Unlike `close()` this method forces the output sink termination by setting termination flag and canceling all the pending writes.
public synchronized void	writeDownStream(final HttpPacket `HttpPacket` to send httpPacket, final FilterChainContext ctx, final CompletionHandler<WriteResult<K, L>> the `CompletionHandler`, which will be notified about write progress. completionHandler, final MessageCloner<Buffer> the `MessageCloner`, which will be able to clone the message in case it can't be completely written in the current thread. messageCloner) Implements org.glassfish.grizzly.http2.StreamOutputSink.writeDownStream. Send an `HttpPacket` to the `Http2Stream`.
private DefaultOutputSink.OutputQueueRecord	writeDownStream0(final HttpPacket httpPacket, final FilterChainContext ctx, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner)
private void	writeEmptyFin()

Inherited from java.lang.Object:: clone equals finalize getClass hashCode notify notifyAll toString wait wait wait

Field Detail

availStreamWindowSize	back to summary
private final AtomicInteger availStreamWindowSize

flushHandlersQueue	back to summary
private BundleQueue<CompletionHandler<Http2Stream>> flushHandlersQueue

flushHandlersSync	back to summary
private final Object flushHandlersSync

http2Session	back to summary
private final Http2Session http2Session

isLastFrameQueued	back to summary
private volatile boolean isLastFrameQueued

LOGGER	back to summary
private static final Logger LOGGER

MAX_OUTPUT_QUEUE_SIZE	back to summary
private static final int MAX_OUTPUT_QUEUE_SIZE

outputQueue	back to summary
pack-priv final TaskQueue<DefaultOutputSink.OutputQueueRecord> outputQueue

stream	back to summary
private final Http2Stream stream

TERMINATING_QUEUE_RECORD	back to summary
private static final DefaultOutputSink.OutputQueueRecord TERMINATING_QUEUE_RECORD

terminationFlag	back to summary
private Termination terminationFlag

unflushedWritesCounter	back to summary
private final AtomicInteger unflushedWritesCounter

ZERO_QUEUE_RECORD_SIZE	back to summary
private static final int ZERO_QUEUE_RECORD_SIZE

Constructor Detail

DefaultOutputSink	back to summary
pack-priv DefaultOutputSink(final Http2Stream stream)

Method Detail

addOutputQueueRecord	back to summary
private void addOutputQueueRecord(DefaultOutputSink.OutputQueueRecord outputQueueRecord) throws Http2StreamException

assertReady	back to summary
private void assertReady() throws IOException

canWrite	back to summary
public boolean canWrite() Implements org.glassfish.grizzly.http2.StreamOutputSink.canWrite. Annotations @Override

canWrite

back to summary

public boolean canWrite()

Implements org.glassfish.grizzly.http2.StreamOutputSink.canWrite.

Annotations

@Override

checkOutputWindow	back to summary
private int checkOutputWindow(final long size) The method is responsible for checking the current output window size. The returned integer value is the size of the data, which could be sent now. Parameters size:long check the provided size against the window size limit. Returns:int the amount of data that may be written.

checkOutputWindow

back to summary

private int checkOutputWindow(final long size)

The method is responsible for checking the current output window size. The returned integer value is the size of the data, which could be sent now.

Parameters

size:long: check the provided size against the window size limit.

Returns:int

the amount of data that may be written.

close	back to summary
public synchronized void close() Implements org.glassfish.grizzly.http2.StreamOutputSink.close. Closes the output sink by adding last DataFrame with the FIN flag to a queue. If the output sink is already closed - method does nothing. Annotations @Override

back to summary

public synchronized void close()

Implements org.glassfish.grizzly.http2.StreamOutputSink.close.

Closes the output sink by adding last DataFrame with the FIN flag to a queue. If the output sink is already closed - method does nothing.

Annotations

@Override

createOutputQueueRecord	back to summary
private DefaultOutputSink.OutputQueueRecord createOutputQueueRecord(final Buffer data, final HttpContent httpContent, final DefaultOutputSink.FlushCompletionHandler flushCompletionHandler, boolean isLast, final boolean isZeroSizeData)

flush back to summary

flush	back to summary
public void flush(final CompletionHandler<Http2Stream> completionHandler) Implements org.glassfish.grizzly.http2.StreamOutputSink.flush. Flush `Http2Stream` output and notify `CompletionHandler` once all output data has been flushed. Parameters completionHandler:CompletionHandler<Http2Stream> `CompletionHandler` to be notified Annotations @Override

public void flush(final CompletionHandler<Http2Stream> completionHandler)

Implements org.glassfish.grizzly.http2.StreamOutputSink.flush.

Flush Http2Stream output and notify CompletionHandler once all output data has been flushed.

Parameters

completionHandler:CompletionHandler<Http2Stream>: CompletionHandler to be notified

Annotations

@Override

flushToConnectionOutputSink	back to summary
private void flushToConnectionOutputSink(final List<Http2Frame> headerFrames, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner, final boolean sendFIN)

flushToConnectionOutputSink	back to summary
private void flushToConnectionOutputSink(final Buffer data, final DefaultOutputSink.FlushCompletionHandler flushCompletionHandler, final boolean sendFIN)

flushToConnectionOutputSink	back to summary
private void flushToConnectionOutputSink(final List<Http2Frame> headerFrames, final Buffer data, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner, final boolean sendFIN)

getUnflushedWritesCount	back to summary
public int getUnflushedWritesCount() Implements org.glassfish.grizzly.http2.StreamOutputSink.getUnflushedWritesCount. Returns:int the number of writes (not bytes), that haven't reached network layer Annotations @Override

getUnflushedWritesCount

back to summary

public int getUnflushedWritesCount()

Implements org.glassfish.grizzly.http2.StreamOutputSink.getUnflushedWritesCount.

Returns:int

the number of writes (not bytes), that haven't reached network layer

Annotations

@Override

isClosed	back to summary
public boolean isClosed() Implements org.glassfish.grizzly.http2.StreamOutputSink.isClosed. Annotations @Override

isClosed

back to summary

public boolean isClosed()

Implements org.glassfish.grizzly.http2.StreamOutputSink.isClosed.

Annotations

@Override

isTerminated	back to summary
private boolean isTerminated()

isWantToWrite	back to summary
private boolean isWantToWrite()

notifyWritePossible	back to summary
public void notifyWritePossible(final WriteHandler writeHandler) Implements org.glassfish.grizzly.http2.StreamOutputSink.notifyWritePossible. Annotations @Override

notifyWritePossible

back to summary

public void notifyWritePossible(final WriteHandler writeHandler)

Implements org.glassfish.grizzly.http2.StreamOutputSink.notifyWritePossible.

Annotations

@Override

onPeerWindowUpdate back to summary

onPeerWindowUpdate	back to summary
public void onPeerWindowUpdate(final int delta) throws Http2StreamException Implements org.glassfish.grizzly.http2.StreamOutputSink.onPeerWindowUpdate. The method is called by HTTP2 Filter once WINDOW_UPDATE message comes for this `Http2Stream`. Parameters delta:int the delta. Annotations @Override Exceptions Http2StreamException: if an error occurs processing the window update.

public void onPeerWindowUpdate(final int delta) throws Http2StreamException

Implements org.glassfish.grizzly.http2.StreamOutputSink.onPeerWindowUpdate.

The method is called by HTTP2 Filter once WINDOW_UPDATE message comes for this Http2Stream.

Parameters

delta:int: the delta.

Annotations

@Override

Exceptions

Http2StreamException:: if an error occurs processing the window update.

prepareDataToSend	back to summary
private Buffer prepareDataToSend(final boolean isRecordNull, final boolean isLast, final Buffer data, final boolean isZeroSizeData)

releaseWriteQueueSpace	back to summary
private void releaseWriteQueueSpace(final int justSentBytes, final boolean isAtomic, final boolean isEndOfChunk)

reserveWriteQueueSpace	back to summary
private int reserveWriteQueueSpace(final int spaceToReserve)

sendTrailers	back to summary
private void sendTrailers(final CompletionHandler<WriteResult<K, L>> completionHandler, final HttpTrailer httpContent)

splitOutputBufferIfNeeded	back to summary
private Buffer splitOutputBufferIfNeeded(final Buffer buffer, final int length)

terminate back to summary

terminate	back to summary
public synchronized void terminate(final Termination terminationFlag) Implements org.glassfish.grizzly.http2.StreamOutputSink.terminate. Unlike `close()` this method forces the output sink termination by setting termination flag and canceling all the pending writes. Annotations @Override

public synchronized void terminate(final Termination terminationFlag)

Implements org.glassfish.grizzly.http2.StreamOutputSink.terminate.

Unlike close() this method forces the output sink termination by setting termination flag and canceling all the pending writes.

Annotations

@Override

writeDownStream back to summary

writeDownStream	back to summary
public synchronized void writeDownStream(final HttpPacket httpPacket, final FilterChainContext ctx, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner) throws IOException Implements org.glassfish.grizzly.http2.StreamOutputSink.writeDownStream. Send an `HttpPacket` to the `Http2Stream`. The writeDownStream(...) methods have to be synchronized with shutdown(). Parameters httpPacket:HttpPacket `HttpPacket` to send completionHandler:CompletionHandler<WriteResult<K, L>> the `CompletionHandler`, which will be notified about write progress. messageCloner:MessageCloner<Buffer> the `MessageCloner`, which will be able to clone the message in case it can't be completely written in the current thread. Annotations @Override Exceptions IOException: if an error occurs with the write operation.

public synchronized void writeDownStream(final HttpPacket httpPacket, final FilterChainContext ctx, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner) throws IOException

Implements org.glassfish.grizzly.http2.StreamOutputSink.writeDownStream.

Send an HttpPacket to the Http2Stream. The writeDownStream(...) methods have to be synchronized with shutdown().

Parameters

httpPacket:HttpPacket: HttpPacket to send
completionHandler:CompletionHandler<WriteResult<K, L>>: the CompletionHandler, which will be notified about write progress.
messageCloner:MessageCloner<Buffer>: the MessageCloner, which will be able to clone the message in case it can't be completely written in the current thread.

Annotations

@Override

Exceptions

IOException:: if an error occurs with the write operation.

writeDownStream0	back to summary
private DefaultOutputSink.OutputQueueRecord writeDownStream0(final HttpPacket httpPacket, final FilterChainContext ctx, final CompletionHandler<WriteResult<K, L>> completionHandler, final MessageCloner<Buffer> messageCloner) throws IOException

writeEmptyFin	back to summary
private void writeEmptyFin()

org.glassfish.grizzly.http2 back to summary

private final Class DefaultOutputSink.FlushCompletionHandler

Class Inheritance

extends ChunkedCompletionHandler

java.lang.Object
org.glassfish.grizzly.http2.utils.ChunkedCompletionHandler
org.glassfish.grizzly.http2.DefaultOutputSink.FlushCompletionHandler

Flush CompletionHandler, which will be passed on each Http2Stream write to make sure the data reached the wires. Usually FlushCompletionHandler is also used as a wrapper for custom CompletionHandler provided by users. The parent class has an internal state, so be careful with reuses of the same instance!

Constructor Summary

Access	Constructor and Description
public	FlushCompletionHandler(final CompletionHandler<WriteResult<K, L>> parentCompletionHandler)

Method Summary

Modifier and Type	Method and Description
protected void	done0() Overrides org.glassfish.grizzly.http2.utils.ChunkedCompletionHandler.done0. Notifies all flush handlers about the completition.

Inherited from org.glassfish.grizzly.http2.utils.ChunkedCompletionHandler:: cancelled completed failed incChunks updated

Constructor Detail

FlushCompletionHandler	back to summary
public FlushCompletionHandler(final CompletionHandler<WriteResult<K, L>> parentCompletionHandler)

Method Detail

done0	back to summary
protected void done0() Overrides org.glassfish.grizzly.http2.utils.ChunkedCompletionHandler.done0. Notifies all flush handlers about the completition. Annotations @Override

done0

back to summary

protected void done0()

Overrides org.glassfish.grizzly.http2.utils.ChunkedCompletionHandler.done0.

Notifies all flush handlers about the completition.

Annotations

@Override

org.glassfish.grizzly.http2 back to summary

private Class DefaultOutputSink.OutputQueueRecord

Class Inheritance

extends AsyncQueueRecord<WriteResult<K, L>>

java.lang.Object
org.glassfish.grizzly.asyncqueue.AsyncQueueRecord
org.glassfish.grizzly.http2.DefaultOutputSink.OutputQueueRecord

Field Summary

Modifier and Type	Field and Description
private DefaultOutputSink.FlushCompletionHandler	chunkedCompletionHandler
private boolean	isLast
private final boolean	isZeroSizeData
private Source	resource
private final HttpTrailer	trailer

Inherited from org.glassfish.grizzly.asyncqueue.AsyncQueueRecord:: completionHandler connection isRecycled message recycleTrack

Constructor Summary

Access	Constructor and Description
public	OutputQueueRecord(final Source resource, final DefaultOutputSink.FlushCompletionHandler completionHandler, final boolean isLast, final boolean isZeroSizeData)
public	OutputQueueRecord(final Source resource, final DefaultOutputSink.FlushCompletionHandler completionHandler, final HttpTrailer trailer, final boolean isZeroDataSize)

Method Summary

Modifier and Type	Method and Description
public WriteResult<K, L>	getCurrentResult() Implements abstract org.glassfish.grizzly.asyncqueue.AsyncQueueRecord.getCurrentResult. Returns the current record result object.
private void	incChunksCounter()
public void	notifyFailure(final Throwable e) Overrides org.glassfish.grizzly.asyncqueue.AsyncQueueRecord.notifyFailure.
public void	recycle() Implements org.glassfish.grizzly.Cacheable.recycle.
public void	release()
private void	reset(final Source resource, final DefaultOutputSink.FlushCompletionHandler completionHandler, final boolean last)

Inherited from org.glassfish.grizzly.asyncqueue.AsyncQueueRecord:: checkRecycled getConnection getMessage notifyIncomplete set setMessage

Field Detail

chunkedCompletionHandler	back to summary
private DefaultOutputSink.FlushCompletionHandler chunkedCompletionHandler

isLast	back to summary
private boolean isLast

isZeroSizeData	back to summary
private final boolean isZeroSizeData

resource	back to summary
private Source resource

trailer	back to summary
private final HttpTrailer trailer

Constructor Detail

OutputQueueRecord	back to summary
public OutputQueueRecord(final Source resource, final DefaultOutputSink.FlushCompletionHandler completionHandler, final boolean isLast, final boolean isZeroSizeData)

OutputQueueRecord	back to summary
public OutputQueueRecord(final Source resource, final DefaultOutputSink.FlushCompletionHandler completionHandler, final HttpTrailer trailer, final boolean isZeroDataSize)