com.sun.xml.ws.api.pipe.Tube Maven / Gradle / Ivy
/*
* Copyright (c) 1997, 2021 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.pipe;
import com.sun.istack.NotNull;
import com.sun.xml.ws.api.message.Message;
import com.sun.xml.ws.api.message.Packet;
import com.sun.xml.ws.api.pipe.helper.AbstractFilterTubeImpl;
import com.sun.xml.ws.api.pipe.helper.AbstractTubeImpl;
import com.sun.xml.ws.api.server.Adapter;
import javax.annotation.PreDestroy;
import javax.xml.ws.Dispatch;
import javax.xml.ws.Provider;
import javax.xml.ws.WebServiceException;
import javax.xml.ws.handler.LogicalHandler;
import javax.xml.ws.handler.soap.SOAPHandler;
import java.text.SimpleDateFormat;
/**
* Abstraction of the intermediate layers in the processing chain
* and transport.
*
* What is a {@link Tube}?
*
* {@link Tube} is a basic processing unit that represents SOAP-level
* protocol handling code. Mutliple tubes are often put together in
* a line (it needs not one dimensional — more later), and act on
* {@link Packet}s in a sequential fashion.
*
*
* {@link Tube}s run asynchronously. That is, there is no guarantee that
* {@link #processRequest(Packet)} and {@link #processResponse(Packet)} runs
* in the same thread, nor is there any guarantee that this tube and next
* tube runs in the same thread. Furthermore, one thread may be used to
* run multiple pipeline in turn (just like a real CPU runs multiple
* threads in turn.)
*
*
*
Tube examples
*
* Transport is a kind of tube. It sends the {@link Packet}
* through, say, HTTP connection, and receives the data back into another {@link Packet}.
*
*
* More often, a tube works like a filter. It acts on a packet,
* and then it tells the JAX-WS that the packet should be passed into another
* tube. It can do the same on the way back.
*
*
* For example, XWSS will be a {@link Tube}. It will act on a request
* {@link Packet}, then perhaps wrap it into
* another {@link Packet} to encrypt the body and add a header, then
* the processing will go on to the next tube.
*
*
* Yet another kind of filter tube is those that wraps {@link LogicalHandler}
* and {@link SOAPHandler}. These tubes are heavy-weight; they often consume
* a message in a packet and create a new one, and then pass it to the next tube.
*
*
* There would be a {@link Tube} implementation that invokes {@link Provider}.
* There would be a {@link Tube} implementation that invokes a service method
* on the user's code.
* There would be a {@link Dispatch} implementation that invokes a {@link Tube}.
*
*
* WS-MEX can be implemented as a {@link Tube} that looks for
* {@link Message#getPayloadNamespaceURI()} and serves the request.
*
*
*
*
*
Tube Lifecycle
* Pipeline is expensive to set up, so once it's created it will be reused.
* A pipeline is not reentrant; one pipeline is used to process one request/response
* at at time. The same pipeline instance may serve multiple request/response,
* if one comes after another and they don't overlap.
*
* Where a need arises to process multiple requests concurrently, a pipeline
* gets cloned through {@link TubeCloner}. Note that this need may happen on
* both server (because it quite often serves multiple requests concurrently)
* and client (because it needs to support asynchronous method invocations.)
*
* Created pipelines (including cloned ones and the original) may be discarded and GC-ed
* at any time at the discretion of whoever owns pipelines. Tubes can, however, expect
* at least one copy (or original) of pipeline to live at any given time while a pipeline
* owner is interested in the given pipeline configuration (in more concerete terms,
* for example, as long as a dispatch object lives, it's going to keep at least one
* copy of a pipeline alive.)
*
* Before a pipeline owner dies, it may invoke {@link #preDestroy()} on the last
* remaining pipeline. It is "may" for pipeline owners that live in the client-side
* of JAX-WS (such as dispatches and proxies), but it is a "must" for pipeline owners
* that live in the server-side of JAX-WS.
*
* This last invocation gives a chance for some pipes to clean up any state/resource
* acquired (such as WS-RM's sequence, WS-Trust's SecurityToken), although as stated above,
* this is not required for clients.
*
*
*
*
Tube and state
*
* The lifecycle of pipelines is designed to allow a {@link Tube} to store various
* state in easily accessible fashion.
*
*
*
Per-packet state
*
* Any information that changes from a packet to packet should be
* stored in {@link Packet} (if such informaton is specific to your problem domain,
* then most likely {@link Packet#invocationProperties}.)
* This includes information like transport-specific headers.
*
*
Per-thread state
*
* Any expensive-to-create objects that are non-reentrant can be stored
* either in instance variables of a {@link Tube}, or a static {@link ThreadLocal}.
*
*
* The first approach works, because {@link Tube} is
* non reentrant. When a tube is copied, new instances should be allocated
* so that two {@link Tube} instances don't share thread-unsafe resources.
*
* Similarly the second approach works, since {@link ThreadLocal} guarantees
* that each thread gets its own private copy.
*
*
* The former is faster to access, and you need not worry about clean up.
* On the other hand, because there can be many more concurrent requests
* than # of threads, you may end up holding onto more resources than necessary.
*
*
* This includes state like canonicalizers, JAXB unmarshallers,
* {@link SimpleDateFormat}, etc.
*
*
*
Per-proxy/per-endpoint state
*
* Information that is tied to a particular proxy/dispatch can be stored
* in a separate object that is referenced from a tube. When
* a new tube is copied, you can simply hand out a reference to the newly
* created one, so that all copied tubes refer to the same instance.
* See the following code as an example:
*
*
* class TubeImpl {
* // this object stores per-proxy state
* class DataStore {
* int counter;
* }
*
* private DataStore ds;
*
* // create a fresh new pipe
* public TubeImpl(...) {
* ....
* ds = new DataStore();
* }
*
* // copy constructor
* private TubeImpl(TubeImpl that, PipeCloner cloner) {
* cloner.add(that,this);
* ...
* this.ds = that.ds;
* }
*
* public TubeImpl copy(PipeCloner pc) {
* return new TubeImpl(this,pc);
* }
* }
*
*
*
* Note that access to such resource may need to be synchronized,
* since multiple copies of pipelines may execute concurrently.
*
*
*
*
VM-wide state
*
* {@code static} is always there for you to use.
*
*
*
* @see AbstractTubeImpl
* @see AbstractFilterTubeImpl
*
* @author Kohsuke Kawaguchi
* @author Jitendra Kotamraju
*/
public interface Tube {
/**
* Acts on a request and perform some protocol specific operation.
*
* TODO: exception handling semantics need more discussion
*
* @throws WebServiceException
* On the server side, this signals an error condition where
* a fault reply is in order (or the exception gets eaten by
* the top-most transport {@link Adapter} if it's one-way.)
* This frees each {@link Tube} from try/catching a
* {@link WebServiceException} in every layer.
*
* Note that this method is also allowed to return
* {@link NextAction#returnWith(Packet)} with
* a {@link Packet} that has a fault as the payload.
*
*
* On the client side, the {@link WebServiceException} thrown
* will be propagated all the way back to the calling client
* applications. (The consequence of that is that if you are
* a filtering {@link Tube}, you must not eat the exception
* that was given to {@link #processException(Throwable)} .
*
* @throws RuntimeException
* Other runtime exception thrown by this method must
* be treated as a bug in the tube implementation,
* and therefore should not be converted into a fault.
* (Otherwise it becomes very difficult to debug implementation
* problems.)
*
*
* On the server side, this exception should be most likely
* just logged. On the client-side it gets propagated to the
* client application.
*
*
* The consequence of this is that if a pipe calls
* into an user application (such as {@link SOAPHandler}
* or {@link LogicalHandler}), where a {@link RuntimeException}
* is *not* a bug in the JAX-WS implementation, it must be catched
* and wrapped into a {@link WebServiceException}.
*
* @param request
* The packet that represents a request message.
* If the packet has a non-null message, it must be a valid
* unconsumed {@link Message}. This message represents the
* SOAP message to be sent as a request.
*
* The packet is also allowed to carry no message, which indicates
* that this is an output-only request.
* (that's called "solicit", right? - KK)
*
* @return
* A {@link NextAction} object that represents the next action
* to be taken by the JAX-WS runtime.
*/
@NotNull NextAction processRequest(@NotNull Packet request);
/**
* Acts on a response and performs some protocol specific operation.
*
*
* Once a {@link #processRequest(Packet)} is invoked, this method
* will be always invoked with the response, before this {@link Tube}
* processes another request.
*
* @param response
* If the packet has a non-null message, it must be
* a valid unconsumed {@link Message}. This message represents
* a response to the request message passed to
* {@link #processRequest(Packet)} earlier.
*
* The packet is also allowed to carry no message, which indicates
* that there was no response. This is used for things like
* one-way message and/or one-way transports.
*
* TODO: exception handling semantics need more discussion
*
* @return
* A {@link NextAction} object that represents the next action
* to be taken by the JAX-WS runtime.
*/
@NotNull NextAction processResponse(@NotNull Packet response);
/**
* Acts on a exception and performs some clean up operations.
*
*
* If a {@link #processRequest(Packet)}, {@link #processResponse(Packet)},
* {@link #processException(Throwable)} throws an exception, this method
* will be always invoked on all the {@link Tube}s in the remaining
* {@link NextAction}s.
*
*
* On the server side, the {@link Throwable} thrown will be propagated to the
* top-most transport. The transport converts the exception to fault reply or
* simply logs in case of one-way MEP. If you are a filtering {@link Tube} like
* {@link AbstractTubeImpl}, you don't have to override the implementation). On
* the other hand, any intermediate {@link Tube} may want to convert the exception
* to a fault message.
*
*
* On the client side, the {@link Throwable} thrown
* will be propagated all the way back to the calling client
* applications. (The consequence of that is that if you are
* a filtering {@link Tube} like {@link AbstractTubeImpl}, you don't have to
* override the implementation)
*
* @param t
*
* @return
* A {@link NextAction} object that represents the next action
* to be taken by the JAX-WS runtime.
*/
@NotNull NextAction processException(@NotNull Throwable t);
/**
* Invoked before the last copy of the pipeline is about to be discarded,
* to give {@link Tube}s a chance to clean up any resources.
*
*
* This can be used to invoke {@link PreDestroy} lifecycle methods
* on user handler. The invocation of it is optional on the client side,
* but mandatory on the server side.
*
*
* When multiple copies of pipelines are created, this method is called
* only on one of them.
*
* @throws WebServiceException
* If the clean up fails, {@link WebServiceException} can be thrown.
* This exception will be propagated to users (if this is client),
* or recorded (if this is server.)
*/
void preDestroy();
/**
* Creates an identical clone of this {@link Tube}.
*
*
* This method creates an identical pipeline that can be used
* concurrently with this pipeline. When the caller of a pipeline
* is multi-threaded and need concurrent use of the same pipeline,
* it can do so by creating copies through this method.
*
*
* Implementation Note
*
* It is the implementation's responsibility to call
* {@link TubeCloner#add(Tube,Tube)} to register the copied pipe
* with the original. This is required before you start copying
* the other {@link Tube} references you have, or else there's a
* risk of infinite recursion.
*
* For most {@link Tube} implementations that delegate to another
* {@link Tube}, this method requires that you also copy the {@link Tube}
* that you delegate to.
*
* For limited number of {@link Tube}s that do not maintain any
* thread unsafe resource, it is allowed to simply return {@code this}
* from this method (notice that even if you are stateless, if you
* got a delegating {@link Tube} and that one isn't stateless, you
* still have to copy yourself.)
*
*
* Note that this method might be invoked by one thread while another
* thread is executing the other process method. See
* the {@link Codec#copy()} for more discussion about this.
*
* @param cloner
* Use this object (in particular its {@link TubeCloner#copy(Tube)} method
* to clone other pipe references you have
* in your pipe. See {@link TubeCloner} for more discussion
* about why.
*
* @return
* always non-null {@link Tube}.
*/
Tube copy(TubeCloner cloner);
}