Re: [resteasy-dev] Testsuite question

Maybe solution would be to use undertow arquillian adapter http://arquillian.org/modules/undertow-embedded-container-adapter/ to test again undertow? On Wed, Apr 19, 2017 at 4:27 PM, Ron Sigal <rsigal(a)redhat.com <mailto:rsigal@redhat.com>> wrote: On 04/19/2017 06:32 AM, Katerina Novotna wrote: > ----- Original Message ----- >> This question arises from RESTEASY-1617 "Add support for injection in >> constraint validators on Java SE". We already use a CDI aware >> ValidatorFactory in the context of Wildfly / EAP, but not in Java SE. >> >> My question is: where do I test this new facility? I've got a test that >> makes an invocation to an embedded instance of Undertow. We have > Could you please point me to your branch with the test? The branch is at https://github.com/ronsigal/Resteasy/tree/1617 <https://github.com/ronsigal/Resteasy/tree/1617>;. The test, temporarily, is at https://github.com/ronsigal/Resteasy/blob/1617/testsuite/integration-test... <https://github.com/ronsigal/Resteasy/blob/1617/testsuite/integration-test...;. I have a lot of tests (in my local branches) in integration-tests/src/test/java/org/jboss/resteasy/experiment/undertow, by the way. I usually start with an embedded undertow server while I'm still investigating or testing. >> resteasy-integration-tests for running tests on Wildfly, which isn't >> appropriate in this case. We also have resteasy-unit-tests, but my >> understanding is that it's not for tests that make network invocations. >> >> We could create a new test module for running tests with embedded >> servers. Or I could just put it in resteasy-integration-tests and forget >> about it. >> >> It's not a big deal, but I'm open to suggestions. > My opinion is, that it could be placed into resteasy-integration-tests into new package to separate the tests from other integration tests. So, maybe, .../testsuite/integration-tests/src/test/java/org/jboss/resteasy/undertow ? > If in the future there should be more tests for different server adapters, then new module would be better. Well, that's an interesting issue. I see that resteasy-undertow, resteasy-netty, and resteasy-netty4, for example, have a src/test/java directory. I haven't looked too closely, but I suppose those tests are specifically for issues related to the particular server adaptors. In the case of my test for RESTEASY-1617, on the other hand, I just want to run with *some* server outside of Wildfly; that is, I want to run the test with Java SE. So, I think the question is, how often does Resteasy behavior potentially differ when running with Java SE compared to running with JEE? I don't have a good answer, off the top of my head. I guess at least one relevant situation would be when there are different dependencies for Java SE and JEE. Note that in this case, I added > <dependency> > <groupId>org.hibernate</groupId> > <artifactId>hibernate-validator-cdi</artifactId> > <version>5.4.1.Final</version> > </dependency> > > <dependency> > <groupId>org.jboss.weld.se > <http://org.jboss.weld.se></groupId> > <artifactId>weld-se</artifactId> > <version>2.4.3.Final</version> > </dependency> to pom.xml in resteasy-integration-tests. Note that org.jboss.weld.se <http://org.jboss.weld.se> doesn't exist in Wildfly. So, how common is that? > -- Katka -- My company's smarter than your company (unless you work for Red Hat) _______________________________________________ resteasy-dev mailing list resteasy-dev(a)lists.jboss.org <mailto:resteasy-dev@lists.jboss.org> https://lists.jboss.org/mailman/listinfo/resteasy-dev <https://lists.jboss.org/mailman/listinfo/resteasy-dev>