[swift-evolution] Enhanced existential types proposal discussion

[Austin Zheng]

I think the problem here is that P == P is true, but P : P is not (a protocol does not conform to itself).

I think there was some discussion about it on the original "Completing Generics" thread from March. I'd probably ask on the swift-users list why P can't be made to conform to P, and then put together a proposal if there's no good reason.

Austin

> On May 29, 2016, at 3:13 PM, Charles Srstka <cocoadev at charlessoft.com> wrote:
> 
> Forgive me if this has already come up, but since we’re talking about fixing generics, I wonder if there is any solution in the pipeline for this problem:
> 
> --
> 
> protocol P { func foo() }
> struct S: P { func foo() { print("foo") } }
> 
> func doSomething<C: CollectionType where C.Generator.Element: P>(c: C) {
> 	for each in c {
> 		each.foo()
> 	}
> }
> 
> let arr: [P] = [S()]
> 
> doSomething(arr) // error: cannot invoke 'doSomething' with an argument list of type '([P])’
> 
> --
> 
> Why is this an error? The whole definition of [P] is basically an array of things that conform to P. Isn’t that exactly what “where Element: P” is asking for?
> 
> Changing Element: P to Element == P solves this particular issue, of course, but then passing something like [S] to the array will fail. The result is that you need to write two functions, and either have one eat the performance cost of constructing a new array that has the correct static type to pass to the other (since, unlike arrays, I can’t figure out a way to convert “Collection where Element: P” into “Collection where Element == P” with a simple cast), or just fill it with the dreaded copy-paste code. Neither seems ideal.
> 
> Charles
>