Bugzilla – Full Text Bug Listing
| Summary: | [1.5][compiler] Interfaces and generics | ||
|---|---|---|---|
| Product: | [Eclipse Project] JDT | Reporter: | Sven Köhler <sven.koehler> |
| Component: | Core | Assignee: | Philipe Mulet <philippe_mulet> |
| Status: | RESOLVED INVALID | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | ||
| Version: | 3.1 | ||
| Target Milestone: | 3.2 M3 | ||
| Hardware: | PC | ||
| OS: | Windows XP | ||
| Whiteboard: | |||
This is the essence of raw types and erasure mechanism. What specific portion of the JLS 3rd edition do you think our compiler contradict ? If this is a language semantic debate, then this is the wrong location for complaining; since the spec is not owned by Eclipse. Can you please clarify? Closing as invalid. Added GenericTypeTest#test828. |
Hi, assume these two interfaces: public interface Transformable<T extends Transformable> { public T transform(); } public interface Volume<V extends Volume> extends Transformable<V> { // public V transform(); } Note that the line "public V transform();" is only a comment. The interface Volume inherits the method transform from Transformable<V> - so transform() should return something that extends Volume since V is known to extend Vector. Instead the code: Volume v1 = null; Volume v2 = v1.transform(); does not compile. v1.transform() returns Transformable instead of Volume. I know that the code Volume<Volume> v1 = null; Volume v2 = v1.transform(); compiles fine, but if you comment out the "public V transform();", you will see that the first code-example compiles clean again. I think that this is a bug in both compilers: javac and JDT. The "public V transform()" should be generated automatically if this is the only possibility for generating proper byte-code. Currently JDT confirms to the behaviour of javac.