public class GrpcWriteStream<T> extends Object implements WriteStream<T>
| Constructor and Description |
|---|
GrpcWriteStream(io.grpc.stub.StreamObserver<T> observer) |
| Modifier and Type | Method and Description |
|---|---|
WriteStream<T> |
drainHandler(Handler<Void> hndlr)
Set a drain handler on the stream.
|
void |
end(Handler<AsyncResult<Void>> hndlr)
Same as
WriteStream.end() but with an handler called when the operation completes |
WriteStream<T> |
exceptionHandler(Handler<Throwable> hndlr)
Set an exception handler on the write stream.
|
WriteStream<T> |
setWriteQueueMaxSize(int i)
Set the maximum size of the write queue to
maxSize. |
io.grpc.stub.StreamObserver<T> |
streamObserver()
Low level control of the observer for advanced use cases.
|
Future<Void> |
write(T data)
Write some data to the stream.
|
void |
write(T data,
Handler<AsyncResult<Void>> hndlr)
Same as
#write(T) but with an handler called when the operation completes |
boolean |
writeQueueFull()
This will return
true if there are more bytes in the write queue than the value set using WriteStream.setWriteQueueMaxSize(int) |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, waitend, end, endpublic GrpcWriteStream(io.grpc.stub.StreamObserver<T> observer)
public WriteStream<T> exceptionHandler(Handler<Throwable> hndlr)
WriteStreamexceptionHandler in interface StreamBaseexceptionHandler in interface WriteStream<T>hndlr - the exception handlerpublic Future<Void> write(T data)
WriteStream The data is usually put on an internal write queue, and the write actually happens
asynchronously. To avoid running out of memory by putting too much on the write queue,
check the WriteStream.writeQueueFull() method before writing. This is done automatically if
using a Pipe.
When the data is moved from the queue to the actual medium, the returned
Future will be completed with the write result, e.g the future is succeeded
when a server HTTP response buffer is written to the socket and failed if the remote
client has closed the socket while the data was still pending for write.
write in interface WriteStream<T>data - the data to writepublic void write(T data, Handler<AsyncResult<Void>> hndlr)
WriteStream#write(T) but with an handler called when the operation completeswrite in interface WriteStream<T>public void end(Handler<AsyncResult<Void>> hndlr)
WriteStreamWriteStream.end() but with an handler called when the operation completesend in interface WriteStream<T>public WriteStream<T> setWriteQueueMaxSize(int i)
WriteStreammaxSize. You will still be able to write to the stream even
if there is more than maxSize items in the write queue. This is used as an indicator by classes such as
Pipe to provide flow control.
The value is defined by the implementation of the stream, e.g in bytes for a
NetSocket, etc...setWriteQueueMaxSize in interface WriteStream<T>i - the max size of the write streampublic boolean writeQueueFull()
WriteStreamtrue if there are more bytes in the write queue than the value set using WriteStream.setWriteQueueMaxSize(int)writeQueueFull in interface WriteStream<T>true if write queue is fullpublic WriteStream<T> drainHandler(Handler<Void> hndlr)
WriteStreamPipe for an example of this being used.
The stream implementation defines when the drain handler, for example it could be when the queue size has been
reduced to maxSize / 2.
drainHandler in interface WriteStream<T>hndlr - the handlerpublic io.grpc.stub.StreamObserver<T> streamObserver()
Copyright © 2021 Eclipse. All rights reserved.