Re: [weld-dev] CDI 2.0 TCK ObserverMethodWithoutNotifyMethodTest failure - ShouldThrow...Definition or Deployment Exception? - spec inconsistency??

From: "Martin Kouba" <mkouba(a)redhat.com&gt; To: "Matej Novotny" <manovotn(a)redhat.com&gt;, "Gordon Hutchison" <gordon.hutchison(a)gmail.com&gt; Cc: weld-dev(a)lists.jboss.org Sent: Friday, April 27, 2018 8:41:45 AM Subject: Re: [weld-dev] CDI 2.0 TCK ObserverMethodWithoutNotifyMethodTest failure - ShouldThrow...Definition or Deployment Exception? - spec inconsistency?? Dne 27.4.2018 v 08:26 Matej Novotny napsal(a): > Hi Gordon, > > running the test on WFLY (or actually without EE container) gives me the > same stacktrace, however the test passes for me. > What is it you fail on exactly? > > Debugging a bit shows that @ShouldThrowException annotation (which is > guarding the exception in the test) goes through a whole chain of > exceptions - having it wrapped should still pass the test. > E.g. it looks at the cause, see if that's assignable, and if it isn't then > the method is recursively invoked for Exception.getCause(). > Here is a link to Arq. code doing that - > https://github.com/arquillian/arquillian-core/blob/master/container/impl-... In fact, we're using a customized org.jboss.arquillian.container.spi.client.container.DeploymentExceptionTransformer for WildFly: https://github.com/weld/core/blob/master/jboss-tck-runner/src/test/java/o...

> > As for spec point of view... > Wrapping those exceptions meets both criteria mentioned in spec as long as > you can get to both of them. > Also, having DefinitionException on top makes sense from user point of view > (at least in my opinion) because is it ultimately developer's mistake that > he/she did not add the notify() method. > And that subsequently leads to DeploymentException. > Does it make sense? > > I'd like to help more but I am not sure whether you are battling some > actual test failure or the somewhat clashing sentences in spec itself? > > Matej > > ----- Original Message ----- >> From: "Gordon Hutchison" <gordon.hutchison(a)gmail.com&gt; >> To: weld-dev(a)lists.jboss.org >> Sent: Thursday, April 26, 2018 5:17:55 PM >> Subject: [weld-dev] CDI 2.0 TCK ObserverMethodWithoutNotifyMethodTest >> failure - ShouldThrow...Definition or >> Deployment Exception? - spec inconsistency?? >> >> Hi Folks, >> >> I work in the IBM CDI implementation team. >> >> We are running the CDI 2.0 TCK/CDS and get a test failure on: >> >> org.jboss.cdi.tck.tests.extensions.lifecycle.broken.observerMethod.ObserverMethodWithoutNotifyMethodTest >> >> that I need some advice on. >> >> The test has: >> >> import javax.enterprise.inject.spi.De ploymentException; >> >> @ShouldThrowException(DeploymentException.class) >> >> >> // unable to process ObserverMethodConfigurator due to missing notify >> method >> impl >> @Test >> @SpecAssertions({ @SpecAssertion(section = AFTER_BEAN_DISCOVERY, id = >> "ee") >> }) >> public void unableToProcessObserverMethodConfigurator() { >> } >> >> This behaviour is specified in >> https://docs.jboss.org/cdi/spec/2.0/cdi-spec-with-assertions.html#after_b... >> (search for "ee)" ) >> >> ee) If the container is unable to process the ObserverMethodConfigurator >> it >> automatically detects the problem and treats it as a deployment problem. >> ObserverMethodWithoutNotifyMethodTest.unableToProcessObserverMethodConfigurator() >> >> >> However what we see coming back from Weld is not a javax >> DeploymentException >> but a javax DefinitionException wrapping a weld DeploymentException cause >> >> Stack Dump = javax.enterprise.inject.spi.De finitionException: >> org.jboss.weld.exceptions.DeploymentException: WELD-000179: >> ObserverMethodConfigurator created by >> org.jboss.cdi.tck.tests.extensions.lifecycle.broken.observerMethod.AfterBeanDiscoveryObserver@d31423c >> cannot be processed >> at >> org.jboss.weld.bootstrap.events.AfterBeanDiscoveryImpl.finish(AfterBeanDiscoveryImpl.java:180) >> <<<<<<< Problem???? >> at >> org.jboss.weld.bootstrap.events.AfterBeanDiscoveryImpl.fire(AfterBeanDiscoveryImpl.java:76) >> at org.jboss.weld.bootstrap.WeldStartup.deployBeans(WeldStartup.java:446) >> at >> org.jboss.weld.bootstrap.WeldBootstrap.deployBeans(WeldBootstrap.java:83) >> at >> com.ibm.ws.cdi.impl.CDIContainerImpl.startInitialization(CDIContainerImpl.java:149) >> at >> com.ibm.ws.cdi.liberty.CDIRuntimeImpl.applicationStarting(CDIRuntimeImpl.java:400) >> at >> com.ibm.ws.container.service.state.internal.ApplicationStateManager.fireStarting(ApplicationStateManager.java:28) >> at >> com.ibm.ws.container.service.state.internal.StateChangeServiceImpl.fireApplicationStarting(StateChangeServiceImpl.java:50) >> at >> com.ibm.ws.app.manager.module.internal.DeployedAppInfoBase.preDeployApp(DeployedAppInfoBase.java:383) >> at >> com.ibm.ws.app.manager.module.internal.DeployedAppInfoBase.deployApp(DeployedAppInfoBase.java:412) >> at com.ibm.ws.app.manager.war.int >> ernal.WARApplicationHandlerImpl.install(WARApplicationHandlerImpl.java:65) >> at >> com.ibm.ws.app.manager.internal.statemachine.StartAction.execute(StartAction.java:140) >> at >> com.ibm.ws.app.manager.internal.statemachine.ApplicationStateMachineImpl.enterState(ApplicationStateMachineImpl.java:1258) >> at >> com.ibm.ws.app.manager.internal.statemachine.ApplicationStateMachineImpl.run(ApplicationStateMachineImpl.java:873) >> at >> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) >> at >> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) >> at java.lang.Thread.run(Thread.java:748) >> Caused by: org.jboss.weld.exceptions.DeploymentException: WELD-000179: >> ObserverMethodConfigurator created by >> org.jboss.cdi.tck.tests.extensions.lifecycle.broken.observerMethod.AfterBeanDiscoveryObserver@d31423c >> cannot be processed >> at >> org.jboss.weld.bootstrap.events.AfterBeanDiscoveryImpl$ObserverRegistration.initObserverMethod(AfterBeanDiscoveryImpl.java:368) >> at >> org.jboss.weld.bootstrap.events.AfterBeanDiscoveryImpl.processObserverRegistration(AfterBeanDiscoveryImpl.java:268) >> at >> org.jboss.weld.bootstrap.events.AfterBeanDiscoveryImpl.finish(AfterBeanDiscoveryImpl.java:177) >> ... 16 more >> Caused by: org.jboss.weld.exceptions.DefinitionException: WELD-000415: >> Custom >> implementation of observer method does not override either notify(T) or >> notify(EventContext<T>): >> org.jboss.weld.bootstrap.events.configurator.ObserverMethodConfiguratorImpl@7e164c01 >> at >> org.jboss.weld.bootstrap.events.configurator.ObserverMethodConfiguratorImpl$ImmutableObserverMethod.<init>(ObserverMethodConfiguratorImpl.java:272) >> at >> org.jboss.weld.bootstrap.events.configurator.ObserverMethodConfiguratorImpl.complete(ObserverMethodConfiguratorImpl.java:234) >> at >> org.jboss.weld.bootstrap.events.AfterBeanDiscoveryImpl$ObserverRegistration.initObserverMethod(AfterBeanDiscoveryImpl.java:366) >> ... 18 more >> >> >> We can see the AfterBeanDiscoveryImpl.java:180 being >> >> /** >> * Bean and observer registration is delayed until after all {@link >> AfterBeanDiscovery} observers are notified. >> */ >> private void finish() { >> try { >> GlobalEnablementBuilder globalEnablementBuilder = >> getBeanManager().getServices().get(GlobalEnablementBuilder.class); >> for (BeanRegistration registration : additionalBeans) { >> processBeanRegistration(registration, globalEnablementBuilder); >> } >> for (ObserverRegistration registration : additionalObservers) { >> processObserverRegistration(registration); >> } >> } catch (Exception e) { >> throw new DefinitionException(e); <<<<<< line 180 >> } >> } >> >> >> I am also aware of, in http://docs.jboss.org/cdi/spec/2.0/cdi-spec.pdf >> 11.5.3., >> "If any observer method of the AfterBeanDiscovery event throws an >> exception, >> the exception is treated as a definition error by the container" >> >> So why is the test expecting DeploymentException >> >> import javax.enterprise.inject.spi.De ploymentException; >> ... @ShouldThrowException(DeploymentException.class) >> >> >> It seems like the test is seeing >> @SpecAssertions({ @SpecAssertion(section = AFTER_BEAN_DISCOVERY, id = >> "ee") >> }) >> If the container is unable to process the ObserverMethodConfigurator it >> automatically detects the problem and treats it as a deployment problem. >> as dominant but the weld implementation is seeing >> http://docs.jboss.org/cdi/spec/2.0/cdi-spec.pdf : 11.5.3 >> If any observer method of the AfterBeanDiscovery event throws an >> exception, >> the exception is treated as a definition error by the container" >> as dominant. >> >> ....and the two do not agree. >> >> So the bottom line is the CDI CTS/TCK test >> org.jboss.cdi.tck.tests.extensions.lifecycle.broken.observerMethod.ObserverMethodWithoutNotifyMethodTest >> fails due to the inconsistency. >> >> Have I missunderstood the spec contradiction or missed some part of Weld >> integration needed? >> >> Gordon Hutchison >> >> _______________________________________________ >> weld-dev mailing list >> weld-dev(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/weld-dev > _______________________________________________ > weld-dev mailing list > weld-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/weld-dev > -- Martin Kouba Senior Software Engineer Red Hat, Czech Republic