ReqT
- type of message sent one or more times to the server.RespT
- type of message received one or more times from the server.public abstract class ClientCall<ReqT,RespT> extends Object
Instances are created
by a Channel
and used by stubs to invoke their remote behavior.
More advanced usages may consume this interface directly as opposed to using a stub. Common reasons for doing so would be the need to interact with flow-control or when acting as a generic proxy for arbitrary operations.
start(io.grpc.ClientCall.Listener<RespT>, io.grpc.Metadata)
must be called prior to calling any other methods, with the exception of
cancel(java.lang.String, java.lang.Throwable)
. Whereas cancel(java.lang.String, java.lang.Throwable)
must not be followed by any other methods,
but can be called more than once, while only the first one has effect.
No generic method for determining message receipt or providing acknowledgement is provided. Applications are expected to utilize normal payload messages for such signals, as a response naturally acknowledges its request.
Methods are guaranteed to be non-blocking. Not thread-safe except for request(int)
, which
may be called from any thread.
There is no interaction between the states on the Listener
and ClientCall
, i.e., if Listener.onClose()
is called, it has no bearing on
the permitted operations on ClientCall
(but it may impact whether they do anything).
There is a race between cancel(java.lang.String, java.lang.Throwable)
and the completion/failure of the RPC in other ways.
If cancel(java.lang.String, java.lang.Throwable)
won the race, Listener.onClose()
is called with
CANCELLED
. Otherwise, Listener.onClose()
is
called with whatever status the RPC was finished. We ensure that at most one is called.
call = channel.newCall(unaryMethod, callOptions); call.start(listener, headers); call.sendMessage(message); call.halfClose(); call.request(1); // wait for listener.onMessage()
The following snippet demonstrates a bi-directional streaming case, where the client sends
requests produced by a fictional makeNextRequest()
in a flow-control-compliant
manner, and notifies gRPC library to receive additional response after one is consumed by
a fictional processResponse()
.
call = channel.newCall(bidiStreamingMethod, callOptions); listener = new ClientCall.Listener<FooResponse>() { @Override public void onMessage(FooResponse response) { processResponse(response); // Notify gRPC to receive one additional response. call.request(1); } @Override public void onReady() { while (call.isReady()) { FooRequest nextRequest = makeNextRequest(); if (nextRequest == null) { // No more requests to send call.halfClose(); return; } call.sendMessage(nextRequest); } } } call.start(listener, headers); // Notify gRPC to receive one response. Without this line, onMessage() would never be called. call.request(1);
DO NOT MOCK: Use InProcessServerBuilder and make a test server instead.
Modifier and Type | Class and Description |
---|---|
static class |
ClientCall.Listener<T>
Callbacks for receiving metadata, response messages and completion status from the server.
|
Constructor and Description |
---|
ClientCall() |
Modifier and Type | Method and Description |
---|---|
abstract void |
cancel(String message,
Throwable cause)
Prevent any further processing for this
ClientCall . |
Attributes |
getAttributes()
Returns additional properties of the call.
|
abstract void |
halfClose()
Close the call for request message sending.
|
boolean |
isReady()
If
true , indicates that the call is capable of sending additional messages
without requiring excessive buffering internally. |
abstract void |
request(int numMessages)
Requests up to the given number of messages from the call to be delivered to
ClientCall.Listener.onMessage(Object) . |
abstract void |
sendMessage(ReqT message)
Send a request message to the server.
|
void |
setMessageCompression(boolean enabled)
Enables per-message compression, if an encoding type has been negotiated.
|
abstract void |
start(ClientCall.Listener<RespT> responseListener,
Metadata headers)
Start a call, using
responseListener for processing response messages. |
public abstract void start(ClientCall.Listener<RespT> responseListener, Metadata headers)
responseListener
for processing response messages.
It must be called prior to any other method on this class, except for cancel(java.lang.String, java.lang.Throwable)
which
may be called at any time.
Since Metadata
is not thread-safe, the caller must not access (read or write) headers
after this point.
responseListener
- receives response messagesheaders
- which can contain extra call metadata, e.g. authentication credentials.IllegalStateException
- if a method (including start()
) on this class has been
called.public abstract void request(int numMessages)
ClientCall.Listener.onMessage(Object)
. No additional messages will be delivered.
Message delivery is guaranteed to be sequential in the order received. In addition, the listener methods will not be accessed concurrently. While it is not guaranteed that the same thread will always be used, it is guaranteed that only a single thread will access the listener at a time.
If it is desired to bypass inbound flow control, a very large number of messages can be
specified (e.g. Integer.MAX_VALUE
).
If called multiple times, the number of messages able to delivered will be the sum of the calls.
This method is safe to call from multiple threads without external synchronization.
numMessages
- the requested number of messages to be delivered to the listener. Must be
non-negative.IllegalStateException
- if call is not start()
edIllegalArgumentException
- if numMessages is negativepublic abstract void cancel(@Nullable String message, @Nullable Throwable cause)
ClientCall
. No further messages may be sent or
will be received. The server is informed of cancellations, but may not stop processing the
call. Cancellation is permitted if previously halfClose()
d. Cancelling an already cancel()
ed ClientCall
has no effect.
No other methods on this class can be called after this method has been called.
It is recommended that at least one of the arguments to be non-null
, to provide
useful debug information. Both argument being null may log warnings and result in suboptimal
performance. Also note that the provided information will not be sent to the server.
message
- if not null
, will appear as the description of the CANCELLED statuscause
- if not null
, will appear as the cause of the CANCELLED statuspublic abstract void halfClose()
IllegalStateException
- if call is already halfClose()
d or cancel(java.lang.String, java.lang.Throwable)
edpublic abstract void sendMessage(ReqT message)
message
- message to be sent to the server.IllegalStateException
- if call is halfClose()
d or explicitly cancel(java.lang.String, java.lang.Throwable)
edpublic boolean isReady()
true
, indicates that the call is capable of sending additional messages
without requiring excessive buffering internally. This event is
just a suggestion and the application is free to ignore it, however doing so may
result in excessive buffering within the call.
If false
, ClientCall.Listener.onReady()
will be called after isReady()
transitions to true
.
If the type of the call is either MethodDescriptor.MethodType.UNARY
or
MethodDescriptor.MethodType.SERVER_STREAMING
, this method may persistently return
false. Calls that send exactly one message should not check this method.
This abstract class's implementation always returns true
. Implementations generally
override the method.
@ExperimentalApi(value="https://github.com/grpc/grpc-java/issues/1703") public void setMessageCompression(boolean enabled)
@ExperimentalApi(value="https://github.com/grpc/grpc-java/issues/2607") @Grpc.TransportAttr public Attributes getAttributes()
ClientCall.Listener.onHeaders(io.grpc.Metadata)
or ClientCall.Listener.onClose(io.grpc.Status, io.grpc.Metadata)
. If called prematurely, the implementation may throw IllegalStateException
or return arbitrary Attributes
.null
attributesIllegalStateException
- (optional) if called before permitted