Hi guys,

I'm still exploring some design space for DSLs, following our interesting discussion.

I'm trying to write the evaluator for the DSL (see below).

For the general case, the evaluator looks like:

eval :: Nomex r a -> State Game a

This eval function takes an expression (called Nomex), that can possibly have effects.

It returns a state monad, to allow you to modify the game state.

But for effectless instructions, it would be better to run the evaluator in the reader monad:

evalNoEffect :: Nomex NoEffect a -> Reader Game a

So you can have additional guaranties that evaluating your expression will not have effects.

I tried (see below), but it doesn't work for the moment:

> {-# LANGUAGE GADTs #-}

> {-# LANGUAGE KindSignatures, DataKinds, ScopedTypeVariables,

>   MultiParamTypeClasses, FunctionalDependencies, FlexibleInstances, UndecidableInstances #-}

> module DSLEffects where

> import Control.Monad.Error
> import Control.Monad.State
> import Control.Monad.Reader
> import Data.Typeable

This is the DSL:
   
> data Effects = Effect | NoEffect

> data Nomex :: Effects -> * -> * where
>   ReadAccount  :: Nomex r Int                --ReadAccount has no effect: it can be run in whatever monad
>   WriteAccount :: Int -> Nomex Effect ()  --WriteAccount has effect
>   SetVictory   :: Nomex NoEffect Bool -> Nomex Effect () --SetVictory don't accept effectful computations
>   Bind         :: Nomex m a -> (a -> Nomex m b) -> Nomex m b
>   Return       :: a -> Nomex r a  --wrapping a constant has no effect

> instance Monad (Nomex a) where
>   return = Return
>   (>>=) = Bind


> noEff :: Nomex NoEffect ()
> noEff = return ()

> hasEffect :: Nomex Effect ()
> hasEffect = do
>    a <- ReadAccount
>    WriteAccount a

> data Game = Game { victory :: Nomex NoEffect Bool,
>                    account :: Int}


> eval :: Nomex r a -> State Game a
> eval a@ReadAccount    = liftEval $ evalNoEffect a
> eval (WriteAccount a) = modify (\g -> g{account = a})

> eval (SetVictory v)   = modify (\g -> g{victory = v})

> eval a@(Return _)     = liftEval $ evalNoEffect a
> eval (Bind exp f)     = eval exp >>= eval . f

> evalNoEffect :: Nomex NoEffect a -> Reader Game a
> evalNoEffect ReadAccount  = asks account
> evalNoEffect (Return a)   = return a
> evalNoEffect (Bind exp f) = evalNoEffect exp >>= evalNoEffect . f

> liftEval :: Reader Game a -> State Game a
> liftEval r = get >>= return . runReader r

This is not compiling:

exceptEffect.lhs:60:15:
    Couldn't match type 'NoEffect with 'Effect
    Inaccessible code in
      a pattern with constructor
        WriteAccount :: Int -> Nomex 'Effect (),
      in an equation for `evalEffect'
    In the pattern: WriteAccount a
    In an equation for `evalEffect':
        evalEffect (WriteAccount a) = modify (\ g -> g {account = a})

It seems that the type of effectless computations (NoEffect) leaks in the type of effectful ones (due to the pattern matching)...

Thanks,

Corentin

On Mon, Feb 3, 2014 at 12:44 PM, Corentin Dupont <corentin.dupont@gmail.com> wrote:

I saw that to write liftQD you decontruct (unwrap) the type and reconstruct it.

I don't know if I can do that for my Exp (which is a full DSL)...

Anyway, there should be a way to encode the Effect/NoEffect semantic at type level...

Using Oleg's parametrized monad idea (http://hackage.haskell.org/package/monad-param-0.0.2/docs/Control-Monad-Parameterized.html), I tried:

> {-# LANGUAGE KindSignatures, DataKinds, ScopedTypeVariables, GADTs
>   MultiParamTypeClasses, FunctionalDependencies, FlexibleInstances, UndecidableInstances #-}

> module DSLEffects where
> import Prelude hiding (return, (>>), (>>=))
> import Control.Monad.Parameterized

This data type will be promoted to kind level (thanks to DataKinds):

> data Eff = Effect | NoEffect

This class allows to specify the semantic on Effects (Effect + NoEffect = Effect):

> class Effects (m :: Eff) (n::Eff) (r::Eff) | m n -> r
> instance Effects Effect n Effect
> instance Effects NoEffect n n

This is the DSL:

> data Exp :: Eff -> * -> * where

>   ReadAccount  :: Exp NoEffect Int      --ReadAccount has no effect
>   WriteAccount :: Int -> Exp Effect ()  --WriteAccount has effect
>   Const        :: a -> Exp r a
>   Bind         :: Effects m n r => Exp m a -> (a -> Exp n b) -> Exp r b --Bind comes with a semantic on effects
>   Fmap         :: (a -> b) -> Exp m a -> Exp m b

> instance Functor (Exp r) where
>   fmap = Fmap

> instance Return (Exp r) where
>    returnM = Const

> instance (Effects m n r) => Bind (Exp m) (Exp n) (Exp r) where
>    (>>=) = Bind

> noEff :: Exp NoEffect ()
> noEff = returnM ()

> hasEffect :: Exp Effect ()
> hasEffect = ReadAccount >> (returnM () :: Exp Effect ())

This is working more or less, however I am obliged to put the type signature on the returnM (last line): why?

Furthermore, I cannot write directly:

> hasEffect :: Exp Effect ()
> hasEffect = ReadAccount

Do you have a better idea?

On Sun, Feb 2, 2014 at 8:55 PM, Lindsey Kuper <lindsey@composition.al> wrote:

On Sun, Feb 2, 2014 at 2:42 PM, Corentin Dupont
<corentin.dupont@gmail.com> wrote:
> you should be able to run an effectless monad in an effectful one.
> How to encode this semantic?

In LVish we just have a `liftQD` operation that will let you lift a
deterministic computation to a quasi-deterministic one (recall that
deterministic computations can perform fewer effects):

  liftQD :: Par Det s a -> Par QuasiDet s a

So, analogously, you could have a `liftEff` and then write `liftEff
noEff`.  This is also a little bit ugly, but you may find you don't
have to do it very often (we rarely use `liftQD`).

Lindsey