Fwd: Single Abstract Method for SAMs cannot be polymorphic (original) (raw)

Grégoire Neuville gregoire.neuville at gmail.com
Wed Apr 3 03:58:04 PDT 2013

A generic SAM is not a valid target type for a lambda, but it may be one for a method reference.

Ok, understood.

You can have generic (polymorphic) SAMs. You just cannot initialize them with lambdas. Use method refs instead.

Yup, got it, read you loud and clear :-) Thanks !

(BTW, lambda support in IntelliJ is very impressive but still new, so there are bound to be differences between what IntelliJ accepts and what javac does.)

Yeah, I noticed that, indeed. It's especially evident as far as type inference is concerned : most of the times javac is smarter in the matter, but sometimes - quite rarely actually - IntelliJ prevails.

On 4/2/2013 4:16 PM, Grégoire Neuville wrote:

I meant 'the annotation in itself is marked as erroneous' : so does the code '(a, g) -> new Gen<>()'.

Note though that the sole annotation (i.e if 'CoArbitrary<?> coArb = (a, g) -> new Gen<>();' is commented) doesn't prevent the code from being compiled by javac : should it ?

On 2 April 2013 22:04, Grégoire Neuville <gregoire.neuville at gmail.com>** wrote: Hi all, The below code : public class TestGenericSAM { class Gen {} @FunctionalInterface interface CoArbitrary { abstract Gen coarbitrary(A a, Gen g); } void test() { CoArbitrary<?> coArb = (a, g) -> new Gen<>(); } } doesn't compile. I guess this is by design (the annotation alone is marked as erroneous by IntelliJ), but I'm just wondering why. Thanks a lot for any explanation ! -- Grégoire Neuville

-- Grégoire Neuville

-- Grégoire Neuville

More information about the lambda-dev mailing list