Good day folks,
Here is a test case that exposes a certain weakness in PolyKinds
error messaging. In the following snippet (T W) :: * is malkinded:
> {-# LANGUAGE PolyKinds #-}
> {-# LANGUAGE UnicodeSyntax #-}
>
> class Eq a ⇒ EqSyn a where
> data T w
> data W f
>
> x = (undefined ∷ EqSyn w ⇒ T w → ()) (undefined ∷ T W)
8.0.1rc2 yields:
> ../ghc-playground.hs:8:42: error:
> • Expected kind ‘T W’, but ‘undefined :: T W’ has kind ‘T W’
> • In the first argument of ‘undefined ::
> EqSyn w => T w -> ()’, namely
> ‘(undefined :: T W)’
> In the expression:
> (undefined :: EqSyn w => T w -> ()) (undefined :: T W)
> In an equation for ‘main’:
> main = (undefined :: EqSyn w => T w -> ()) (undefined :: T W)
..which is.. not informative.
With PolyKinds removed:
> ../ghc-playground.hs:7:56: error:
> • Expecting one more argument to ‘W’
> Expected a type, but ‘W’ has kind ‘* -> *’
> • In the first argument of ‘T’, namely ‘W’
> In an expression type signature: T W
> In the first argument of ‘undefined ::
> EqSyn w => T w -> ()’, namely
> ‘(undefined :: T W)’
I'm not sure if this was reported, so I'm sorry for the noise if it was.
--
с уважениeм / respectfully,
Косырев Сергей
_______________________________________________
ghc-devs mailing list
ghc-devs@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs
