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: | Broken test compilation | ||
---|---|---|---|
Product: | java | Reporter: | Jesse Glick <jglick> |
Component: | Hints | Assignee: | Jan Lahoda <jlahoda> |
Status: | RESOLVED FIXED | ||
Severity: | blocker | CC: | mmetelka |
Priority: | P2 | Keywords: | TEST |
Version: | 6.x | ||
Hardware: | All | ||
OS: | All | ||
URL: | http://deadlock.netbeans.org/hudson/job/test-compilation/169/testReport/org.netbeans.nbbuild/SubAntJUnitReport/_hudson_workdir_jobs_test_compilation_workspace_java_hints/ | ||
Issue Type: | DEFECT | Exception Reporter: |
Description
Jesse Glick
2007-11-06 07:39:56 UTC
Should be OK now: Checking in HintsTestBase.java; /cvs/java/hints/test/unit/src/org/netbeans/modules/java/hints/infrastructure/HintsTestBase.java,v <-- HintsTestBase.java new revision: 1.7; previous revision: 1.6 done Should consider merging this to branch in case someone wants to run tests on release60. BTW I am not sure what the purpose of LanguageProvider.java 1.12 was; Language, LanguageEmbedding, and Token all have to take a type parameter extending TokenId anyway. I am a little surprised javac even lets you refer to Token<?>, as if e.g. Token<String> were even possible. I was trying to simplify <? extends TokenId> to just <?> where possible since if class Language<T extends TokenId> then IIUC Language<?> should be equivalent to Language<? extends TokenId>. Javac seems to have no problem to compile Language<? extends TokenId> lang = Language.find("text/plain"); where public static Language<?> find(String mimeType) { ... } within lexer module's code (although now I see that an incremental building problem gets hit in this case too). However when compiling the same code in other module then javac always fails with incompatible types problem. So it's likely a difference between compiling against jar file versus sources. Anyway it seems fixed in JDK6 - the code compiles anywhere. I have noticed generally that a lot of generics-related bugs in JDK 5's javac are fixed in JDK 6. |