2009/3/6 Rickard Öberg <[email protected]> > Niclas Hedhman wrote: > > On Fri, Mar 6, 2009 at 9:25 AM, Rickard Öberg <[email protected]> > wrote: > >> Doesn't that lead to the > >> same lookup as we are talking about? > > > > Question to Stuart; I understand that it is a implementation detail > > whether there is one Method instance or many. But in the Sun > > implementation, can we perform a "==" instead of hashing for method > > lookup? I.e. loop through an array, since as Michael points out, the > > method count is relatively small, and "==" is fairly cheap. > > The problem, I think, is that the Method we have during assembly is > different from the one we get during Proxy invocation, so there's no way > to do == checks. :-( Otherwise we could use an IdentityMap and be done > with it... >
How accurate does the method matching have to be? Is there any possibility of using the method signature (say method.toString()) to do the hashmap lookup instead of the actual method object itself? You'd get clashes for hiding/overriding methods which have the same signature, but this might not occur in the limited world of the proxy... also as it's been a while since I looked at the Qi4j runtime internals, could someone point me to the particular piece of code that needs optimizing :) /Rickard > _______________________________________________ > qi4j-dev mailing list > [email protected] > http://lists.ops4j.org/mailman/listinfo/qi4j-dev > -- Cheers, Stuart
_______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
