GitLab

Seems it's perfectly ok to have the data type of the same name in the same module, but different packages. Yet used in the same module:

{-# LANGUAGE PackageImports #-}
{-# LANGUAGE RankNTypes #-}
module Data.Foo where
 
import "void" Data.Void as A
import Data.Void as B
 
foo :: A.Void -> a
foo = absurd
 
bar :: B.Void -> a
bar x = x ()

{-# LANGUAGE RankNTypes #-}
module Data.Void where
 
type Void = forall a b. a -> b

https://gist.github.com/phadej/8b628d579ddf6958d937 motivated by: http://stackoverflow.com/questions/28159068/packagename-with-ghc-generics

I'd propose to change Datatype class to be:

class Datatype d where
  -- | The name of the datatype (unqualified)
  datatypeName :: t d (f :: * -> *) a -> [Char]
  -- | The fully-qualified name of the module where the type is declared
  moduleName   :: t d (f :: * -> *) a -> [Char]
  -- | The package name of the module where the type is declared
  packageName :: t d (f :: * -> *) a -> [Char]
  -- | Marks if the datatype is actually a newtype
  isNewtype    :: t d (f :: * -> *) a -> Bool
  isNewtype _ = False

Trac metadata

Trac field	Value
Version	7.8.4
Type	FeatureRequest
TypeOfFailure	OtherFailure
Priority	normal
Resolution	Unresolved
Component	libraries/base
Test case
Differential revisions
BlockedBy
Related
Blocking
CC	ekmett, hvr
Operating system
Architecture