LGTM
http://gwt-code-reviews.appspot.com/1450806/diff/1/dev/core/src/com/google/gwt/dev/javac/rebind/RebindRuleResolver.java File dev/core/src/com/google/gwt/dev/javac/rebind/RebindRuleResolver.java (right): http://gwt-code-reviews.appspot.com/1450806/diff/1/dev/core/src/com/google/gwt/dev/javac/rebind/RebindRuleResolver.java#newcode22 dev/core/src/com/google/gwt/dev/javac/rebind/RebindRuleResolver.java:22: boolean checkRebindRuleResolvable(String typeName); On 2011/05/27 23:11:20, jbrosenberg wrote:
Well, I guess I don't want to expose the full RebindOracle interface
to the
StandardGeneratorContext (e.g. generators shouldn't be able to call
rebind()
directly). Also, other things implement the RebindOracle interface,
such as
ModuleSpaceHost (and by extension ShellModuleSpaceHost). It doesn't
seem
sensible for those to implement checkRebindRuleResolvable(), etc. No?
What I'd meant was not exposing this to users, but at a first glance it seemed like more of the unnecessary abstraction that's plagued this part of the system (for example, there's effectively only one real implementor of RebindOracle and RebindPermutationOracle). But now that I look closer, there's really no other easy way to do what you want. Oh well, just another reason to come through here sometime holistically and make it all more sane. http://gwt-code-reviews.appspot.com/1450806/ -- http://groups.google.com/group/Google-Web-Toolkit-Contributors