This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
Summary: | OpenEditorEnablesEditMenuFactoryTest.testIfProjectFactoryInstalled failed seven times in a row | ||
---|---|---|---|
Product: | platform | Reporter: | Jaroslav Tulach <jtulach> |
Component: | Text | Assignee: | Miloslav Metelka <mmetelka> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | mmetelka |
Priority: | P1 | Keywords: | 7.4_HR_FIX, TEST |
Version: | 7.4 | ||
Hardware: | PC | ||
OS: | Linux | ||
Issue Type: | DEFECT | Exception Reporter: | |
Bug Depends on: | 228991 | ||
Bug Blocks: | |||
Attachments: | List of changes |
I can see two fixes for bug 228991 has been integrated. Adding dependency. Mílo, are you running openide.text unit tests as part of build validation? You should I see random failures of various openide.text tests that started this week. Like the one below. I am afraid all these instabilities are caused by something in openide.text - passing to you. org.openide.text.NotifyModifiedOnNbEditorLikeKitTest.testReloadWithoutModifiedIssue57104 Failing for the past 1 build (Since Unstable#4446 ) Took 0.32 sec. add description Error Message Length it new expected:<10> but was:<0> Stacktrace junit.framework.AssertionFailedError: Length it new expected:<10> but was:<0> at org.openide.text.NotifyModifiedTest.runTest(NotifyModifiedTest.java:117) at org.netbeans.junit.NbTestCase.access$200(NbTestCase.java:97) at org.netbeans.junit.NbTestCase$2.doSomething(NbTestCase.java:431) at org.netbeans.junit.NbTestCase$1Guard.run(NbTestCase.java:357) at java.lang.Thread.run(Thread.java:722) Caused by: junit.framework.AssertionFailedError: Length it new expected:<10> but was:<0> at org.openide.text.NotifyModifiedTest.testReloadWithoutModifiedIssue57104(NotifyModifiedTest.java:581) at org.openide.text.NotifyModifiedTest.runTest(NotifyModifiedTest.java:115) The ide.ergonomics test seems to fail with: java.lang.IllegalStateException: BaseDocument.insertString not in EDT: offset=0, text= org.netbeans.editor.BaseDocument.insertString(BaseDocument.java:743) org.netbeans.editor.BaseDocument.read(BaseDocument.java:1451) org.netbeans.editor.BaseKit.read(BaseKit.java:663) org.openide.text.DataEditorSupport.loadFromStreamToKit(:490) org.openide.text.DocumentOpenClose$DocumentLoad.atomicLockedRun(:760) org.openide.text.DocumentOpenClose$DocumentLoad.run(DocumentOpenClose.java:611) which has been added at 180341:781d41acc048 - mmetelka 21.10.2010 #183219 - The characters move theirs position during selection. am I not sure why it fails the test now. (In reply to Jaroslav Tulach from comment #2) > Mílo, are you running openide.text unit tests as part of build validation? Yes, sure I run unit tests of openide.text editor (some tests inherited from openide.text and tested with BaseDocument impl) openide.loaders None of them was failing with the final integration of the issue 228991. But it's certainly caused by it and there must be some race condition which causes the tests to fail. I'll check that. Regarding java.lang.IllegalStateException: BaseDocument.insertString not in EDT: offset=0, text= org.netbeans.editor.BaseDocument.insertString(BaseDocument.java:743) ... it looks like you have all the loggers turned on for some reason since this only happens for a logger Logger.getLogger(BaseDocument.class.getName() + "-EDT") (or parent loggers). I'll probably remove the LOG_EDT logger and accompanying code completely since it's a bit strange for a logger to be able to trigger an ISE. I have disabled the EDT logger check in BaseDocument and the test now passes fine. http://hg.netbeans.org/jet-main/rev/5386d4a46f3b The mentioned failing openide.text tests now pass fine for me after the fix of issue 234781. I would like to ask QE for verification. 5386d4a46f3b transplanted to c7cd87dab565 |
Created attachment 139232 [details] List of changes Together with this test also other tests in openide.text are randomly failing. I am attaching list of integrations that happened at the time of first failure. There is no commit of mine, so I expect that something has been changed in the general text infrastructure.