* {@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.) * * *
* 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. * * * * *
* 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. * * * *
* The lifecycle of pipelines is designed to allow a {@link Tube} to store various * state in easily accessible fashion. * * *
* 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. * *
* 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. * * *
* 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. * * * *
* {@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. * *
* 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); }