This is the mail archive of the kawa@sourceware.cygnus.com mailing list for the Kawa project. See the Kawa home page for more information.
> 1. I fixed the problem with AWT/Swing methods: in general, if class <A> > defines method m, and class <B> inherits from A, > primitive-virtual-method must be called with <A> and "m", not <B> and > "m" (this is maybe obvious, but I thought it didn't matter as long as > m was visible in class <B>). There are two sets of methods Kawa can use (I may have this backwards): java.lang.Class.getMethod(s) - returns public methods in the given class *or* its ancestors. java.lang.Class.getDeclaredMethod(s) - returns all methods in the given class *only*. One of the patches I sent switched this. Using both methods, or explicitly searching back htrough the base classes seems undesirable, though it is an option. > 2. I tried with a functional version of `toplevel' instead of the > macro one; I also tried with `define-syntax' instead of `defmacro' but > without success. define-syntax is preferable. Since it is a data structure, it can be used in the same compilation using with less trouble. (Though all of this needs to be re-done ...) > 3. Just to repeat, if I type in Kawa's top repl something close to the > expansion of a call to `toplevel' (the only difference is that I used > global definitions instead of a `let*'), it works. This is not really clear. Send me an explicit testcase including instructions on how to reproduce the problem, and I can try it. (You sent me something earlier, but it seems like it might have changed.) --Per Bothner Cygnus Solutions bothner@cygnus.com http://www.cygnus.com/~bothner