# MonadCont done right

### From HaskellWiki

(Difference between revisions)

(Replaced obsolete hawiki links) |
|||

(One intermediate revision by one user not shown) | |||

Line 1: | Line 1: | ||

− | The Cont class [[MonadCont]] [http://haskell.org/ | + | The Cont class [[MonadCont]] [http://www.haskell.org/ghc/docs/latest/html/libraries/mtl/Control-Monad-Cont-Class.html] defined in the [[monad template library]] [http://en.wikibooks.org/wiki/Haskell/Monad_transformers#Transformers_are_cousins] could be improved if you are willing to use [[rank two polymorphism]] [http://en.wikibooks.org/wiki/Haskell/Polymorphism#Higher_rank_types]. |

Notice the change in the signature of <hask>callCC</hask>. This allows one to use the passed continuation in different situations inside a <hask>callCC</hask> block. However, you will have to provide an explicit signature for the function you are calling <hask>callCC</hask> with. | Notice the change in the signature of <hask>callCC</hask>. This allows one to use the passed continuation in different situations inside a <hask>callCC</hask> block. However, you will have to provide an explicit signature for the function you are calling <hask>callCC</hask> with. | ||

− | == Possible | + | == Possible implementation == |

<haskell> | <haskell> | ||

Line 18: | Line 18: | ||

</haskell> | </haskell> | ||

− | == Alternative | + | == Alternative implementation == |

This implementation has the advantage that it provides a polymorphic version of <hask>callCC</hask> for all instances of MonadCont from <hask>Control.Monad.Cont</hask>. I also added <hask>shift</hask> and <hask>reset</hask> functions for using ComposableContinuations. | This implementation has the advantage that it provides a polymorphic version of <hask>callCC</hask> for all instances of MonadCont from <hask>Control.Monad.Cont</hask>. I also added <hask>shift</hask> and <hask>reset</hask> functions for using ComposableContinuations. | ||

## Latest revision as of 21:27, 6 December 2008

The Cont class MonadCont [1] defined in the monad template library [2] could be improved if you are willing to use rank two polymorphism [3].

Notice the change in the signature ofcallCC

callCC

callCC

## [edit] 1 Possible implementation

newtype Cont r a = Cont { runCont :: ((a -> r) -> r) } -- r is the final result type of the whole computation class (Monad m) => MonadCont m where callCC :: ((a -> (forall b. m b)) -> m a) -> m a instance Monad (Cont r) where return a = Cont (\k -> k a) -- i.e. return a = \k -> k a (Cont c) >>= f = Cont (\k -> c (\a -> runCont (f a) k)) -- i.e. c >>= f = \k -> c (\a -> f a k) instance MonadCont (Cont r) where callCC f = Cont (\k -> runCont (f (\a -> Cont (\_ -> k a))) k)

## [edit] 2 Alternative implementation

This implementation has the advantage that it provides a polymorphic version ofcallCC

Control.Monad.Cont

shift

reset

{-# OPTIONS -fglasgow-exts -fno-warn-unused-binds -cpp #-} module ContExts ( callCC', shift, reset, shiftT, resetT, ) where import Control.Monad.Cont -- Cont' m a is the type of a continuation expecting an a within the -- continuation monad Cont m type Cont' m a = forall r. a -> m r callCC' :: forall a m. MonadCont m => (Cont' m a -> m a) -> m a #if __GLASGOW_HASKELL__ > 602 callCC' f = callCC f' where #else callCC' (f :: ((a -> (forall b. m b)) -> m a) ) = callCC f' where #endif f' :: (a -> m (EmptyMonad m)) -> m a f' g = f g' where g' :: a -> m b g' = (=<<) runEmptyMonad . g -- ghc doesn't allow something like m (forall c. m c) newtype EmptyMonad m = EmptyMonad { runEmptyMonad :: forall c. m c } -- shift/reset for the Cont monad shift :: ((a -> Cont s r) -> Cont r r) -> Cont r a shift e = Cont $ \k -> e (return . k) `runCont` id reset :: Cont a a -> Cont r a reset e = return $ e `runCont` id -- shiftT/resetT for the ContT monad transformer shiftT :: Monad m => ((a -> ContT r m s) -> ContT s m s) -> ContT s m a shiftT e = ContT $ \k -> e (lift . k) `runContT` return resetT :: Monad m => ContT a m a -> ContT r m a resetT e = lift $ e `runContT` return

All of this is presumably meant to be under the MIT license, since it wants to be in the library. --SamB 22:44, 30 October 2006 (UTC)