Documentation

Monads that also support choice and failure.

The Monad class defines the basic operations over a monad, a concept from a branch of mathematics known as category theory. From the perspective of a Haskell programmer, however, it is best to think of a monad as an abstract datatype of actions. Haskell's do expressions provide a convenient syntax for writing monadic expressions.

Instances of Monad should satisfy the following:

Left identity

return a >>= k = k a

Right identity

m >>= return = m

Associativity

m >>= (\x -> k x >>= h) = (m >>= k) >>= h

Furthermore, the Monad and Applicative operations should relate as follows:

• pure = return

• m1 <*> m2 = m1 >>= (\x1 -> m2 >>= (\x2 -> return (x1 x2)))

The above laws imply:

• fmap f xs  =  xs >>= return . f

• (>>) = (*>)

and that pure and (<*>) satisfy the applicative functor laws.

The instances of Monad for lists, Maybe and IO defined in the Prelude satisfy these laws.

A type f is a Functor if it provides a function fmap which, given any types a and b lets you apply any function from (a -> b) to turn an f a into an f b, preserving the structure of f. Furthermore f needs to adhere to the following:

Identity

fmap id == id

Composition

fmap (f . g) == fmap f . fmap g

Note, that the second law follows from the free theorem of the type fmap and the first law, so you need only check that the former condition holds. See https://www.schoolofhaskell.com/user/edwardk/snippets/fmap or https://github.com/quchen/articles/blob/master/second_functor_law.md for an explanation.

When a value is bound in do-notation, the pattern on the left hand side of <- might not match. In this case, this class provides a function to recover.

A Monad without a MonadFail instance may only be used in conjunction with pattern that always match, such as newtypes, tuples, data types with only a single data constructor, and irrefutable patterns (~pat).

Instances of MonadFail should satisfy the following law: fail s should be a left zero for >>=,

fail s >>= f  =  fail s

If your Monad is also MonadPlus, a popular definition is

fail _ = mzero

fail s should be an action that runs in the monad itself, not an exception (except in instances of MonadIO). In particular, fail should not be implemented in terms of error.

Since: base-4.9.0.0

Map each element of a structure to a monadic action, evaluate these actions from left to right, and collect the results. For a version that ignores the results see mapM_.

Examples

Evaluate each monadic action in the structure from left to right, and collect the results. For a version that ignores the results see sequence_.

Examples

>>> sequence $ Right [1,2,3,4]
[Right 1,Right 2,Right 3,Right 4]

>>> sequence $ [Right 1,Right 2,Right 3,Right 4]
Right [1,2,3,4]

>>> sequence $ Left [1,2,3,4]
Left [1,2,3,4]

>>> sequence $ [Left 0, Right 1,Right 2,Right 3,Right 4]
Left 0

forM is mapM with its arguments flipped. For a version that ignores the results see forM_.

Repeat an action indefinitely.

Examples

echoServer :: Socket -> IO ()
echoServer socket = forever $ do
  client <- accept socket
  forkFinally (echo client) (\_ -> hClose client)
  where
    echo :: Handle -> IO ()
    echo client = forever $
      hGetLine client >>= hPutStrLn client

Promote a function to a monad.

Conditional failure of Alternative computations. Defined by

guard True  = pure ()
guard False = empty

Examples

>>> safeDiv 4 0
Nothing

>>> safeDiv 4 2
Just 2

safeDiv :: Int -> Int -> Maybe Int
safeDiv x y | y /= 0    = Just (x `div` y)
            | otherwise = Nothing

safeDiv :: Int -> Int -> Maybe Int
safeDiv x y = do
  guard (y /= 0)
  return (x `div` y)

The join function is the conventional monad join operator. It is used to remove one level of monadic structure, projecting its bound argument into the outer level.

'join bss' can be understood as the do expression

do bs <- bss
   bs

Examples

atomically :: STM a -> IO a

atomically :: STM (IO b) -> IO (IO b)
join       :: IO (IO b)  -> IO b

join . atomically :: STM (IO b) -> IO b

Same as >>=, but with the arguments interchanged.

Conditional execution of Applicative expressions. For example,

when debug (putStrLn "Debugging")

will output the string Debugging if the Boolean value debug is True, and otherwise do nothing.

Promote a function to a monad, scanning the monadic arguments from left to right. For example,

liftM2 (+) [0,1] [0,2] = [0,2,1,3]
liftM2 (+) (Just 1) Nothing = Nothing

Promote a function to a monad, scanning the monadic arguments from left to right (cf. liftM2).

Promote a function to a monad, scanning the monadic arguments from left to right (cf. liftM2).

Promote a function to a monad, scanning the monadic arguments from left to right (cf. liftM2).

In many situations, the liftM operations can be replaced by uses of ap, which promotes function application.

return f `ap` x1 `ap` ... `ap` xn

is equivalent to

liftMn f x1 x2 ... xn

void value discards or ignores the result of evaluation, such as the return value of an IO action.

Examples

>>> void Nothing
Nothing
>>> void (Just 3)
Just ()

>>> void (Left 8675309)
Left 8675309
>>> void (Right 8675309)
Right ()

>>> void [1,2,3]
[(),(),()]

>>> void (1,2)
(1,())

>>> mapM print [1,2]
1
2
[(),()]
>>> void $ mapM print [1,2]
1
2

Map each element of a structure to a monadic action, evaluate these actions from left to right, and ignore the results. For a version that doesn't ignore the results see mapM.

mapM_ is just like traverse_, but specialised to monadic actions.

forM_ is mapM_ with its arguments flipped. For a version that doesn't ignore the results see forM.

forM_ is just like for_, but specialised to monadic actions.

Evaluate each monadic action in the structure from left to right, and ignore the results. For a version that doesn't ignore the results see sequence.

sequence_ is just like sequenceA_, but specialised to monadic actions.

The sum of a collection of actions using (<|>), generalizing concat.

msum is just like asum, but specialised to MonadPlus.

Examples

>>> msum [Just "Hello", Nothing, Just "World"]
Just "Hello"

This generalizes the list-based filter function.

Left-to-right composition of Kleisli arrows.

'(bs >=> cs) a' can be understood as the do expression

do b <- bs a
   cs b

Right-to-left composition of Kleisli arrows. (>=>), with the arguments flipped.

Note how this operator resembles function composition (.):

(.)   ::            (b ->   c) -> (a ->   b) -> a ->   c
(<=<) :: Monad m => (b -> m c) -> (a -> m b) -> a -> m c

The mapAndUnzipM function maps its first argument over a list, returning the result as a pair of lists. This function is mainly used with complicated data structures or a state monad.

The zipWithM function generalizes zipWith to arbitrary applicative functors.

zipWithM_ is the extension of zipWithM which ignores the final result.

The foldM function is analogous to foldl, except that its result is encapsulated in a monad. Note that foldM works from left-to-right over the list arguments. This could be an issue where (>>) and the `folded function' are not commutative.

foldM f a1 [x1, x2, ..., xm]

==

do
  a2 <- f a1 x1
  a3 <- f a2 x2
  ...
  f am xm

If right-to-left evaluation is required, the input list should be reversed.

Note: foldM is the same as foldlM

Like foldM, but discards the result.

replicateM n act performs the action act n times, and then returns the list of results:

Examples

>>> import Control.Monad.State
>>> runState (replicateM 3 $ state $ \s -> (s, s + 1)) 1
([1,2,3],4)

Like replicateM, but discards the result.

Examples

>>> replicateM_ 3 (putStrLn "a")
a
a
a

The reverse of when.

Strict version of <$>.

Since: base-4.8.0.0

Direct MonadPlus equivalent of filter.

Examples

filter = ( mfilter :: (a -> Bool) -> [a] -> [a] )

>>> mfilter odd (Just 1)
Just 1
>>> mfilter odd (Just 2)
Nothing