Bugzilla – Full Text Bug Listing
| Summary: | @AJ and formal binding impl | ||
|---|---|---|---|
| Product: | [Tools] AspectJ | Reporter: | Alexandre Vasseur <avasseur> |
| Component: | Compiler | Assignee: | Alexandre Vasseur <avasseur> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | enhancement | ||
| Priority: | P3 | ||
| Version: | DEVELOPMENT | ||
| Target Milestone: | 1.5.0 M3 | ||
| Hardware: | All | ||
| OS: | All | ||
| Whiteboard: | |||
reviewed and fixed + test, all bindings and arg binding are supported and error is supposed to be seen when formal binding is missing close when M3 ships was M3 remind was M3 remind |
Some notes on formal binding impl for @AJ In code style, the advice signature is always looking the same (..bindings.., org.aspectj.lang.Part thisJoinPointStaticPart, JoinPoint thisJoinPoint, org.aspectj.lang.Part thisEnclosingJoinPointStaticPart) In @ style, it is user defined. The current impl is handling the formal binding in a way that only args/this/target can be bound. I had to add some conditionals to handles cases like that: void myAdvice(JoinPoint jp, Object target) {..} where target() binding is at index 1, while the index 0 is not an unbound but actually "virtually" bound to the joinpoint. Pointcut are thus having a virtuallyBoundedNames String[] that is the list of arguments for which we don't want any complain if it not bound. This one is populated during @ extraction, and Pointcut.concretize() make sure the info is not lost when composition occurs (with PerClause f.e.) The only issue is that currently, binding JP/SJP/ESJP themselves is not possible f.e. "... && args(jp)" myAdvice(JoinPoint jp, JoinPoint currentJp) Is that a big issue ? May be impact adviceexecution() and could impact some use cases around mock testing.