Community
Participate
Working Groups
Umbrella bug for known opened Java search / Call Hierarchy issues.
"call hierarchy": https://bugs.eclipse.org/bugs/buglist.cgi?quicksearch=call%20hierarchy&component=Core&product=JDT 47 bugs found. "search" https://bugs.eclipse.org/bugs/buglist.cgi?quicksearch=search&component=Core&product=JDT 173 bugs found. Its hard to find a point where to start
(In reply to Jörg Kubitz from comment #1) > "call hierarchy": > https://bugs.eclipse.org/bugs/buglist. > cgi?quicksearch=call%20hierarchy&component=Core&product=JDT > > 47 bugs found. > > "search" > https://bugs.eclipse.org/bugs/buglist. > cgi?quicksearch=search&component=Core&product=JDT > > 173 bugs found. > > Its hard to find a point where to start There are for sure many duplicates, as no one really tracked this :-( I would propose to start with bug 565512, because the move from "classpath" to "module" hell in Java 9+ broke a lot, and there is already a starting point given (see bug 565512 comment 4). A fix for the root cause of search issues in "module" based environments will probably cover a lot of other follow up bugs.
(In reply to Andrey Loskutov from comment #2) > I would propose to start with bug 565512, Thats what i thought too. till i made a junit and found that i´the search does not even come that far... calls in .class files seem to be not implemented at all. It looks just random that if the caller has the same method name that a caller is found inside a jar. JDT has a big pool of search tests but they all(?) ran only on open projects while there are at least 3 different search implementations for binaries, modular binaries, and non binaries.
(In reply to Jörg Kubitz from comment #3) > (In reply to Andrey Loskutov from comment #2) > > > I would propose to start with bug 565512, > > Thats what i thought too. till i made a junit and found that i´the search > does not even come that far... calls in .class files seem to be not > implemented at all. It looks just random that if the caller has the same > method name that a caller is found inside a jar. I haven't checked the patch, but I assume you are searching for methods/fields. I would start to search for types - once they can be found, one can check if the underlined symbols can be also found.
(In reply to Andrey Loskutov from comment #4) > I haven't checked the patch, but I assume you are searching for > methods/fields. > I would start to search for types - once they can be found, one can check if > the underlined symbols can be also found. we think alike: I made sure it finds the classes and methods DECLARATIONS. Both the callers and the callees, the implementations and the superclasses. But it does not find IMPLEMENTORS (of types) nor REFERENCES (of methods).
From this list, I've moved (to https://github.com/eclipse-jdt/eclipse.jdt.core) what I think we might fix/work on: https://bugs.eclipse.org/bugs/buglist.cgi?quicksearch=call%20hierarchy&list_id=21337922 I didn't move bugs with unclear reproduction and performances issues (for performance issues that Advantest runs into, we'll be opening issues). If anyone is interested in the remaining issues, please evaluate and move.
