PROPOSAL: Method and Field Literals (original) (raw)

Rémi Forax forax at univ-mlv.fr
Thu Mar 12 14:24:05 PDT 2009


Mark Mahieu a écrit :

On 12 Mar 2009, at 20:55, Rémi Forax wrote:

Neal Gafter a écrit :

I may regret saying this later, but I am not concerned about the potential conflict. I believe we can use the exact same syntax and distinguish whether it should be a java.lang.reflect.Method or a closure from context. Since we're doing Method first, it would take priority (just like a method invocation that requires no boxing is preferred during overload resolution to one that requires some argument to be boxed).

Neal, I don't know if you will regret saying this but I will regret to not saying that it will create lot of puzzlers like this one: class A { static int f(int) { ... } } ... {int => int} c=A#f(int); A#f(int).equals(c) // false because A#f(int) is a java.lang.reflect.Method Rémi I expect that would rather depend on how closure equality is defined (or not).

Here equals() is called on Method.

Mark

Rémi



More information about the coin-dev mailing list