com.sun.xml.ws.api.server.AsyncProvider Maven / Gradle / Ivy
Show all versions of jaxws-rt Show documentation
/*
* Copyright (c) 1997, 2019 Oracle and/or its affiliates. All rights reserved.
*
* This program and the accompanying materials are made available under the
* terms of the Eclipse Distribution License v. 1.0, which is available at
* http://www.eclipse.org/org/documents/edl-v10.php.
*
* SPDX-License-Identifier: BSD-3-Clause
*/
package com.sun.xml.ws.api.server;
import com.sun.istack.NotNull;
import javax.xml.ws.Provider;
import javax.xml.ws.WebServiceContext;
import java.util.concurrent.Executor;
/**
* Asynchronous version of {@link Provider}.
*
*
* Applications that use the JAX-WS RI can implement this interface instead of
* {@link Provider} to implement asynchronous web services (AWS.) AWS enables
* applications to perform operations with long latency without blocking a thread,
* and thus particularly suitable for highly scalable service implementation,
* at the expesnce of implementation complexity.
*
*
Programming Model
*
* Whenever a new reuqest arrives, the JAX-WS RI invokes the {@link #invoke} method
* to notify the application. Normally, the application then schedules an execution
* of this request, and exit from this method immediately (the point of AWS is not
* to use this calling thread for request processing.)
*
*
* Unlike the synchronous version, which requires the response to be given as the return value,
* with AWS the JAX-WS RI will keep the connection with client open, until the application
* eventually notifies the JAX-WS RI via {@link AsyncProviderCallback}. When that
* happens that causes the JAX-WS RI to send back a response to the client.
*
*
* The following code shows a very simple AWS example:
*
*
* @WebService
* class MyAsyncEchoService implements AsyncProvider<Source> {
* private static final {@link Executor} exec = ...;
*
* public void invoke( final Source request, final AsyncProviderCallback<Source> callback, final WebServiceContext context) {
* exec.execute(new {@link Runnable}() {
* public void run() {
* Thread.sleep(1000); // kill time.
* callback.send(request); // just echo back
* }
* });
* }
* }
*
*
*
* Please also check the {@link Provider} and its programming model for general
* provider programming model.
*
*
*
WebServiceContext
*
* In synchronous web services, the injected {@link WebServiceContext} instance uses
* the calling {@link Thread} to determine which request it should return information about.
* This no longer works with AWS, as you may need to call {@link WebServiceContext}
* much later, possibly from entirely different thread.
*
*
* For this reason, {@link AsyncProvider} passes in {@link WebServiceContext} as
* a parameter. This object remains usable until you invoke {@link AsyncProviderCallback},
* and it can be invoked from any thread, even concurrently. AWS must not use the injected
* {@link WebServiceContext}, as its behavior is undefined.
*
* @see Provider
* @author Jitendra Kotamraju
* @author Kohsuke Kawaguchi
* @since 2.1
*/
public interface AsyncProvider {
/**
* Schedules an execution of a request.
*
* @param request
* Represents the request message or payload.
* @param callback
* Application must notify this callback interface when the processing
* of a request is complete.
* @param context
* The web service context instance that can be used to retrieve
* context information about the given request.
*/
public void invoke(
@NotNull T request,
@NotNull AsyncProviderCallback callback,
@NotNull WebServiceContext context);
}