1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 | {-# LANGUAGE KindSignatures, FlexibleContexts #-} import Control.Applicative import Control.Monad import Control.Monad.Trans.Either {-| An EitherLike is a Kleisli category over the flipped monad, with: > returnE >|> f = f > f >|> returnE = f > (f >|> g) >|> h = f >|> (g >|> h) Two functors define the interaction between the ordinary and flipped monad. (fmap k .) defines a functor from one Kleisli category of the flipped monad to another Kleisli category of the flipped monad: > forall k . (fmap k .) returnE = returnE > forall k . (fmap k .) (f >|> g) = (fmap k .) f >|> (fmap k .) g (fmapE k .) defines a functor from one Kleisli category of the ordinary monad to another Kleisli category of the ordinary monad. > forall k . (fmapE k .) return = return > forall k . (fmapE k .) (f >=> g) = (fmapE k .) f >=> (fmapE k .) g -} class EitherLike e where returnE :: (Monad m) => a -> e a m r (>|>) :: (Monad m) => (a -> e b m r) -> (b -> e c m r) -> (a -> e c m r) instance EitherLike EitherT where returnE = left f >|> g = \x -> EitherT $ do y <- runEitherT (f x) runEitherT $ case y of Left z -> g z Right r -> right r throwE :: (EitherLike e, Monad m) => a -> e a m r throwE = returnE catchE :: (EitherLike e, Monad m) => e a m r -> (a -> e b m r) -> e b m r m `catchE` h = ((\() -> m) >|> h) () {-| Two functors define how an EitherLike may be lifted. (liftE .) defines a functor from the Kleisli category of the ordinary lower monad to the Kleisli category of the ordinary transformed monad: > (liftE .) return = return > (liftE .) (f >=> g) = (liftE .) f >=> (liftE .) g (liftE .) defines a functor from the Kleisli category of the flipped lower monad to the Kleisli category of the flipped transformed monad: > (liftE .) returnE = returnE > (liftE .) (f >|> g) = (liftE .) f >|> (liftE .) g -} class EitherTrans t where liftE :: (Monad (e l m), EitherLike e) => e l m r -> t e l m r newtype StateE s e l (m :: * -> *) r = StateE { runStateE :: s -> e l m (r, s) } instance (Monad (e l m)) => Functor (StateE s e l m) where fmap = liftM instance (Monad (e l m)) => Applicative (StateE s e l m) where pure = return (<*>) = ap instance (Monad (e l m)) => Monad (StateE s e l m) where return r = StateE $ \s -> return (r, s) m >>= f = StateE $ \s -> do (a, s') <- runStateE m s runStateE (f a) s' instance (EitherLike e) => EitherLike (StateE s e) where returnE l = StateE $ \s -> returnE l f >|> g = \a -> StateE $ \s -> (((`runStateE` s) . f) >|> ((`runStateE` s) . g)) a instance EitherTrans (StateE s) where liftE m = StateE $ \s -> liftM (\r -> (r, s)) m get :: (Monad (e l m)) => StateE s e l m s get = StateE $ \s -> return (s, s) put :: (Monad (e l m)) => s -> StateE s e l m () put s = StateE $ \_ -> return ((), s) {- Proof that StateE satisfies the EitherLike laws: -- Note that this is not a required law, but I'm using it as a helper proof Identity law for fmap = ((`runStateE` s) .), id = returnE: (`runStateE` s) . returnE = returnE Proof: (`runStateE` s) . returnE = \a -> runStateE (returnE a) s = \a -> runStateE (StateE $ \s -> returnE a) s = \a -> (\s -> returnE a) s = \a -> returnE a = returnE Left identity law for (.) = (>|>), id = returnE: returnE >|> g = g Proof: returnE >|> g = \a -> StateE $ \s -> (((`runStateE` s) . returnE) >|> ((`runStateE` s) . g)) a = \a -> StateE $ \s -> (returnE >|> ((`runStateE` s) . g)) a = \a -> StateE $ \s -> ((`runStateE` s) . g) a = \a -> StateE $ \s -> runStateE (g a) s = \a -> StateE $ runStateE (g a) = \a -> g a = g Right identity law for (.) = (>|>), id = returnE: f >|> returnE = f Proof: f >|> returnE = \a -> StateE $ \s -> (((`runStateE` s) . f) >|> ((`runStateE` s) . returnE)) a = \a -> StateE $ \s -> (((`runStateE` s) . f) >|> returnE) a = \a -> StateE $ \s -> ((`runStateE` s) . f) a = \a -> StateE $ \s -> runStateE (f a) s = \a -> StateE $ runStateE (f a) = \a -> f a = f -- Note that this is not a required law, but I'm using it as a helper proof Composition law for fmap = ((`runStateE` s) .), (.) = (>|>): (`runStateE` s) . (f >|> g) = ((`runStateE` s) . f) >|> ((`runStateE` s) . g) Proof: (`runStateE` s) . (f >|> g) = (`runStateE` s) . (\b -> StateE $ \s -> ((`runStateE` s) . f) >|> ((`runStateE` s) . g) b ) = \b -> (`runStateE` s) $ StateE $ \s -> ((`runStateE` s) . f) >|> ((`runStateE` s) . g) b ) = \b -> ((`runStateE` s) . f) >|> ((`runStateE` s) . g) b = ((`runStateE` s) . f) >|> ((`runStateE` s) . g) Associativity law for (.) = (>|>): f >|> (g >|> h) = (f >|> g) >|> h Proof: f >|> (g >|> h) = \a -> StateE $ \s -> ( ((`runStateE` s) . f) >|> ((`runStateE` s) . (g >|> h)) ) a = \a -> StateE $ \s -> ( ((`runStateE` s) . f) >|> ((`runStateE` s) . (g >|> h)) ) a = \a -> StateE $ \s -> ( ((`runStateE` s) . f) >|> ( ((`runStateE` s) . g) >|> ((`runStateE` s) . h) ) ) a = \a -> StateE $ \s -> ( ( ((`runStateE` s) . f) >|> ((`runStateE` s) . g) ) >|> ((`runStateE` s) . h) ) a = \a -> StateE $ \s -> ( ((`runStateE` s) . (f >|> g)) >|> ((`runStateE` s) . h) = (f >|> g) >|> h -- STILL MISSING: The functor law proofs -} {- Proof that this satisfies the EitherTrans laws: The monad transformer laws are just the same proofs as for StateT, since it is implemented identically. Identity law for fmap = (liftE .) and id = returnE (liftE .) returnE = returnE Proof: liftE . returnE = \l -> StateE $ \s -> liftM (\r -> (r, s)) (returnE l) = \l -> StateE $ \s -> returnE l = returnE Composition law for fmap = (liftE .) and (.) = (>|>) (liftE .) f >|> (liftE .) g = (liftE .) (f >|> g) Proof: liftE . f >|> liftE . g = (\a -> StateE $ \s -> liftM (\r -> (r, s)) (f a)) >|> (\b -> StateE $ \s -> liftM (\r -> (r, s)) (g b)) = \a -> StateE $ \s -> ( ((`runStateE` s) . (\a -> StateE $ \s -> liftM (\r -> (r, s)) (f a))) >|> ((`runStateE` s) . (\b -> StateE $ \s -> liftM (\r -> (r, s)) (g b))) ) a = \a -> StateE $ \s -> ( (\a -> liftM (\r -> (r, s)) (f a)) >|> (\b -> liftM (\r -> (r, s)) (g b)) ) a = \a -> StateE $ \s -> ( (liftM (\r -> (r, s)) . f) >|> (liftM (\r -> (r, s)) . g) ) a = \a -> StateE $ \s -> liftM (\r -> (r, s)) . (f >|> g) = liftE . (f >|> g) -} |
31:47: Warning: Redundant bracket
Found:
(b -> e c m r) -> (a -> e c m r)
Why not:
(b -> e c m r) -> a -> e c m r