Implements java.io.Closeable.close.

Doc from java.io.Closeable.close.

Closes this stream and releases any system resources associated with it. If the stream is already closed then invoking this method has no effect.

As noted in AutoCloseable#close(), cases where the close may fail require careful attention. It is strongly advised to relinquish the underlying resources and to internally mark the Closeable as closed, prior to throwing the IOException.

Annotations

@Override

Implements org.eclipse.jetty.io.Content.Source.demand.

Doc from org.eclipse.jetty.io.Content.Source.demand.

Demands to invoke the given demand callback parameter when a chunk of content is available.

See how to use this method idiomatically.

Implementations guarantee that calls to this method are safely reentrant so that stack overflows are avoided in the case of mutual recursion between the execution of the Runnable callback and a call to this method. Invocations of the passed Runnable are serialized and a callback for demand call is not invoked until any previous demand callback has returned. Thus the Runnable should not block waiting for a callback of a future demand call.

The demand callback may be invoked spuriously: a subsequent call to read() may return null.

Calling this method establishes a pending demand, which is fulfilled when the demand callback is invoked.

Calling this method when there is already a pending demand results in an IllegalStateException to be thrown.

If the invocation of the demand callback throws an exception, then fail(Throwable) is called.

Parameters

demandCallback:Runnable

the demand callback to invoke where there is a content chunk available

Annotations

@Override

Implements org.eclipse.jetty.io.Content.Source.fail.

Doc from org.eclipse.jetty.io.Content.Source.fail.

Fails this content source with a last failure chunk, failing and discarding accumulated content chunks that were not yet read.

The failure may be notified to the content reader at a later time, when the content reader reads a content chunk, via a Chunk instance with a non null Chunk#getFailure().

If read() has returned a last chunk, this is a no operation.

Typical failure: the content being aborted by user code, or idle timeouts.

If this method has already been called, then it is a no operation.

Parameters

failure:Throwable

the cause of the failure

Annotations

@Override

Overrides default org.eclipse.jetty.io.Content.Source.fail.

Doc from org.eclipse.jetty.io.Content.Source.fail.

Fails this content source with a failure chunk that may or not may be last. If last is true, then the failure is persistent and a call to this method acts as fail(Throwable). Otherwise the failure is transient and a failure chunk will be read in order with content chunks, and subsequent calls to read may produce other content.

A Content.Source or its reader may treat a transient failure as persistent.

Parameters

failure:Throwable

A failure.

last:boolean

true if the failure is persistent, false if the failure is transient.

Annotations

@Override

Overrides default org.eclipse.jetty.io.Content.Source.getLength.

Returns:long

Doc from org.eclipse.jetty.io.Content.Source.getLength.

the content length, if known, or -1 if the content length is unknown

Annotations

@Override

Implements org.eclipse.jetty.io.Content.Source.read.

Doc from org.eclipse.jetty.io.Content.Source.read.

Reads a chunk of content.

See how to use this method idiomatically.

The returned chunk could be:

• null, to signal that there isn't a chunk of content available
• an Chunk instance with non null Chunk#getFailure(), to signal that there was a failure trying to produce a chunk of content, or that the content production has been failed externally
• a Chunk instance, containing the chunk of content.

Once a read returns an Chunk instance with non-null Chunk#getFailure() then if the failure is last further reads will continue to return the same failure chunk instance, otherwise further read() operations may return different non-failure chunks.

Once a read returns a last chunk, further reads will continue to return a last chunk (although the instance may be different).

The content reader code must ultimately arrange for a call to Chunk#release() on the returned Chunk.

Additionally, prior to the ultimate call to Chunk#release(), the reader code may make additional calls to Chunk#retain(), that must ultimately be matched by a correspondent number of calls to Chunk#release().

Concurrent reads from different threads are not recommended, as they are inherently in a race condition.

Reads performed outside the invocation context of a demand callback are allowed. However, reads performed with a pending demand are inherently in a race condition (the thread that reads with the thread that invokes the demand callback).

Returns:Content.Chunk

a chunk of content, possibly a failure instance, or null

Annotations

@Override

Implements org.eclipse.jetty.io.Content.Sink.write.

Doc from org.eclipse.jetty.io.Content.Sink.write.

Writes the given ByteBuffer, notifying the Callback when the write is complete.

Implementations guarantee that calls to this method are safely reentrant so that stack overflows are avoided in the case of mutual recursion between the execution of the Callback and a call to this method.

The write completes:

• immediately with a failure when this instance is closed or already has a failure
• successfully when a non empty Content.Chunk returned by read() is released
• successfully just before the Content.Chunk is returned by read(), for any empty chunk Content.Chunk.

Parameters

last:boolean

whether the ByteBuffer is the last to write

byteBuffer:ByteBuffer

the ByteBuffer to write

callback:Callback

the callback to notify when the write operation is complete

Annotations

@Override