Thanks for the help. I'm glad at least EclEmma/JaCoCo can be ruled out. Now it's back to the JUnit people...
Evgeny Mandrikov <[email protected]> ezt írta (időpont: 2020. márc. 2., H, 18:42): > Hm, however EclEmma and JaCoCo register hook outside of the test, > and the following example seems to work even with JUnit 4.13 > > import org.junit.Test; > > import org.junit.Rule; > import org.junit.rules.Timeout; > > import java.util.concurrent.TimeUnit; > > public class ExampleTest { > > static { > Runtime.getRuntime().addShutdownHook(new Thread(new Runnable() { > @Override > public void run() { > System.err.println("Hook executed"); > } > })); > } > > @Rule > public Timeout globalTimeout = new Timeout(5, TimeUnit.MINUTES); > > @Test > public final void test() { > } > } > > But not the following reduction of your > > FlowableGroupByTest.mapFactoryWithExpiringGuavaCacheDemonstrationCodeForUseInJavadoc > > package io.reactivex.rxjava3.internal.operators.flowable; > > import java.util.Map; > import java.util.concurrent.TimeUnit; > > import org.junit.Rule; > import org.junit.Test; > import org.junit.rules.Timeout; > import com.google.common.cache.CacheBuilder; > import io.reactivex.rxjava3.core.Flowable; > import io.reactivex.rxjava3.functions.Consumer; > import io.reactivex.rxjava3.functions.Function; > import io.reactivex.rxjava3.internal.functions.Functions; > > public class FlowableGroupByTest { > > static { > System.err.println("registering hook"); > Thread thread = new Thread(new Runnable() { > @Override > public void run() { > System.err.println("hook executed"); > } > }); > Runtime.getRuntime().addShutdownHook(thread); > System.err.println(thread.getThreadGroup()); > } > > @Rule > public Timeout globalTimeout = new Timeout(5, TimeUnit.MINUTES); > > private static final Function<Integer, Integer> mod5 = new > Function<Integer, Integer>() { > @Override > public Integer apply(Integer n) throws Exception { > return n % 5; > } > }; > > @Test > public void > mapFactoryWithExpiringGuavaCacheDemonstrationCodeForUseInJavadoc() { > Function<Consumer<Object>, Map<Integer, Object>> evictingMapFactory = new > Function<Consumer<Object>, Map<Integer, Object>>() { > @Override > public Map<Integer, Object> apply(final Consumer<Object> notify) throws > Exception { > return CacheBuilder.newBuilder() // > .<Integer, Object>build().asMap(); > } > }; > Flowable.range(1, 1) // > .groupBy(mod5, Functions.<Integer>identity(), true, 16, > evictingMapFactory) // > .test(); > } > > } > > > whose *execution without EclEmma* sometimes prints "hook executed" and > sometimes not. > > While the root cause is unclear, I hope that the above helps to understand > that the problem is not in EclEmma/JaCoCo and pinpoint the root cause. > > > Regards, > Evgeny > > On Monday, March 2, 2020 at 6:11:19 PM UTC+1, Dávid Karnok wrote: >> >> Hi and thanks for the detailed analysis. >> >> 2020. március 2., hétfő 17:22:29 UTC+1 időpontban Evgeny Mandrikov a >> következőt írta: >>> >>> Hi, >>> >>> JaCoCo and hence EclEmma rely on the execution of JVM shutdown hook. >>> The message "java.net.SocketException: Connection reset" might indicate >>> that JVM shutdown hook was not executed. >>> >>> Using >>> Eclipse Version: 2019-09 R (4.13.0), Build id: 20190917-1200 >>> >>> After clone of your project at >>> https://github.com/ReactiveX/RxJava/commits/5f6aafcbaa97f330ad2007f4ad02c5e33db8914b >>> >>> Execution of the following test *even without EclEmma* >>> >>> package io.reactivex.rxjava3.core; >>> >>> import org.junit.Test; >>> >>> public class ExampleTest extends RxJavaTest { >>> @Test >>> public void test() { >>> Runtime.getRuntime().addShutdownHook(new Thread(new Runnable() { >>> @Override >>> public void run() { >>> System.err.println("hook executed"); >>> } >>> })); >>> } >>> } >>> >>> >>> Won't print anything, whereas execution of the same test without >>> "extends RxJavaTest" will print "hook executed". >>> >>> The same test will also print "hook executed" if you remove >>> "globalTimeout" defined in "RJavaTest". >>> >>> From the above IMO clear that the issue is not in EclEmma. >>> >>> Furthermore, problem with execution of JVM shutdown hooks >>> in presence of org.junit.rules.Timeout from JUnit 4.13 can be >>> demonstrated >>> without Eclipse >>> using following "build.gradle" >>> >>> apply plugin: 'java' >>> >>> repositories { >>> mavenCentral() >>> } >>> >>> test { >>> testLogging.showStandardStreams = true >>> } >>> >>> dependencies { >>> testCompile 'junit:junit:4.13' >>> } >>> >>> and following "src/test/java/Example.java" >>> >>> import org.junit.Test; >>> >>> import org.junit.Rule; >>> import org.junit.Test; >>> import org.junit.rules.Timeout; >>> >>> import java.util.concurrent.TimeUnit; >>> >>> public class ExampleTest { >>> @Rule >>> public Timeout globalTimeout = new Timeout(5, TimeUnit.MINUTES); >>> >>> @Test >>> public final void test() { >>> Runtime.getRuntime().addShutdownHook(new Thread(new Runnable() { >>> @Override >>> public void run() { >>> System.err.println("hook executed"); >>> } >>> })); >>> } >>> } >>> >>> >>> And since the problem goes away in all the above cases after downgrade >>> to JUnit 4.12, >>> then IMO it is caused by changes in JUnit 4.13. >>> >>> >>> Regards, >>> Evgeny >>> >>> >>> On Sunday, February 16, 2020 at 6:56:33 PM UTC+1, Dávid Karnok wrote: >>>> >>>> Hi. I've run into an odd coverage failure with JUnit 4.13, Eclipse, >>>> EclEmma and test methods referencing Google Guava's Cache. >>>> >>>> I was advised by the JUnit project to ask about the situation with this >>>> project (and/or with Eclipse / Google Guava). >>>> >>>> >>>> #### Environment: >>>> >>>> Eclipse: Version: 2019-09 R (4.13.0), Build id: 20190917-1200 >>>> EclEmma 3.1.2.201903112331 org.eclipse.eclemma.feature.feature.group >>>> Eclipse >>>> EclEmma >>>> JUnit 4.13 >>>> Windows 10 x64 >>>> Java 1.8u241 >>>> >>>> #### Issue: >>>> >>>> I've run into an odd error while running coverage on an unit test >>>> having some methods use Google Guava's CacheBuilder in Eclipse. The >>>> coverage fails with the error code 5013 and error report indicating >>>> java.net.SocketException: Connection reset. If I remove these methods, the >>>> coverage succeeds. >>>> >>>> If I revert back to JUnit 4.12, both the test file and the individual >>>> test succeeds in producing the coverage data. >>>> >>>> Example method: >>>> >>>> https://github.com/ReactiveX/RxJava/blob/5f6aafcbaa97f330ad2007f4ad02c5e33db8914b/src/test/java/io/reactivex/rxjava3/internal/operators/flowable/FlowableGroupByTest.java#L1917 >>>> >>>> This is where the cache is created: >>>> >>>> https://github.com/ReactiveX/RxJava/blob/5f6aafcbaa97f330ad2007f4ad02c5e33db8914b/src/test/java/io/reactivex/rxjava3/internal/operators/flowable/FlowableGroupByTest.java#L2194 >>>> >>>> #### Additional details: >>>> >>>> - If I run "gradlew test jacocoTestReport" on the entire project, it >>>> succeeds in creating the report HTML apparently. >>>> - If I run "Coverage As > JUnit test" on the entire project from >>>> Eclipse, it fails with the same 5013 error. >>>> - If I run any other test class/package/method in the project, the >>>> coverage is created successfully. >>>> >>>> Please advise. >>>> >>> -- > You received this message because you are subscribed to a topic in the > Google Groups "JaCoCo and EclEmma Users" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/jacoco/SHw3R6cvZSs/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/jacoco/5b1f5db4-62f0-42da-8273-b2719f688cd5%40googlegroups.com > <https://groups.google.com/d/msgid/jacoco/5b1f5db4-62f0-42da-8273-b2719f688cd5%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- Best regards, David Karnok -- You received this message because you are subscribed to the Google Groups "JaCoCo and EclEmma Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/jacoco/CAAWwtm9UKOzO8Bv3qUEsqo9o2BcFN1-Q1bFPTF-1TBPkUTi55A%40mail.gmail.com.
