[weld-dev] CDI 2.0 TCK ObserverMethodWithoutNotifyMethodTest failure - ShouldThrow...Definition or Deployment Exception? - spec inconsistency??
Martin Kouba
mkouba at redhat.com
Fri Apr 27 02:41:45 EDT 2018
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-base/src/main/java/org/jboss/arquillian/container/impl/client/container/DeploymentExceptionHandler.java#L63-L66
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/org/jboss/weld/tck/wildfly8/WildFly8DeploymentExceptionTransformer.java
>
> 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 at gmail.com>
>> To: weld-dev at 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_bean_discovery
>> (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 at 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$Worker.run(ThreadPoolExecutor.java:624)
>> at java.lang.Thread.run(Thread.java:748)
>> Caused by: org.jboss.weld.exceptions.DeploymentException: WELD-000179:
>> ObserverMethodConfigurator created by
>> org.jboss.cdi.tck.tests.extensions.lifecycle.broken.observerMethod.AfterBeanDiscoveryObserver at d31423c
>> cannot be processed
>> at
>> org.jboss.weld.bootstrap.events.AfterBeanDiscoveryImpl$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 at 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 at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/weld-dev
> _______________________________________________
> weld-dev mailing list
> weld-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/weld-dev
>
--
Martin Kouba
Senior Software Engineer
Red Hat, Czech Republic
More information about the weld-dev
mailing list