| Summary: | Inconsistent handling of error types | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | [Eclipse Project] JDT | Reporter: | Olivier Thomann <Olivier_Thomann> | ||||||
| Component: | APT | Assignee: | Olivier Thomann <Olivier_Thomann> | ||||||
| Status: | RESOLVED FIXED | QA Contact: | |||||||
| Severity: | normal | ||||||||
| Priority: | P3 | CC: | eclipse | ||||||
| Version: | 3.7 | ||||||||
| Target Milestone: | 3.7 M7 | ||||||||
| Hardware: | PC | ||||||||
| OS: | Windows 7 | ||||||||
| Whiteboard: | |||||||||
| Attachments: |
|
||||||||
|
Description
Olivier Thomann
I'll take care of this. I need to review all usage of TypeKind.ERROR and how the error type instances are being instantiated. We also need to make sure that any method declared on javax.lang.model.type.DeclaredType can be invoked on an error type without any errors. Created attachment 193576 [details]
Proposed fix
Created attachment 193835 [details]
Proposed fix + regression test
Updated regression tests. This is using ProcessingEnvironment.getSourceVersion() to switch between old implementation and new implementation for error types as interfaces.
Tests are passing on JDK6 and JDK7.
Walter, I'll release this for M7 so that we get a more consistent handling of error types inside the apt model. Released for 3.7M7. |