Summary: | Unexpected error: can't determine implemented interfaces of missing type | ||
---|---|---|---|
Product: | [Tools] AspectJ | Reporter: | Ron Bodkin <rbodkin+LISTS> |
Component: | Compiler | Assignee: | aspectj inbox <aspectj-inbox> |
Status: | RESOLVED FIXED | QA Contact: | |
Severity: | normal | ||
Priority: | P2 | ||
Version: | DEVELOPMENT | ||
Target Milestone: | 1.5.0RC1 | ||
Hardware: | PC | ||
OS: | Windows XP | ||
Whiteboard: |
Description
Ron Bodkin
2005-11-14 14:16:01 EST
If I use call(public * javax..xml.rpc.Call.invoke*(..)) then the errors all disappear. That's an ugly workaround though... Interestingly, I seem to only get the linkage error from bug # 116255 when this bug happens, although I believe that has to do with changing the execution path of the system (i.e., it's correlated without causation). Actually the only correlation between the two issues was that I had forgotten I was using the Nov. 9 development build which didn't have the sax parser factory code in it. However, this bug is present in HEAD and the ugly workaround also works in HEAD. just one quick comment to myself, from this line in the stack trace: KindedPointcut.warnOnConfusingSig I can tell its only going down this code path in order to work out whether to give you a lint warning. Hmmm... it would be nice to isolate this kind of exception from breaking the weave (sounds like a great use for aspects). I think I put code in before to stop this going wrong if the hierarchy couldn't be determined - but that was before 'Missing' was introduced as a real type as opposed to just a broken signature. I'll take a look tomorrow. As I thought... the missing logic I put in ages and ages ago is broken by the new MissingResolvedTypeWithKnownSignature I'm about to put in a fix for this case fix checked in - waiting on build Fix available. Can you try it Ron as I can't recreate easily here? I'm closing as the fix is available - (just to get the bug numbers down!!) - please reopen if you still have problems with this Ron. |