Data.Conduit

If this is your first time with conduit, you should probably start with the tutorial: https://haskell.fpcomplete.com/user/snoyberg/library-documentation/conduit-overview.

Core interface

Types

type Source m o = ConduitM () o m ()

Provides a stream of output values, without consuming any input or producing a final result.

Since 0.5.0

type Conduit i m o = ConduitM i o m ()

Consumes a stream of input values and produces a stream of output values, without producing a final result.

Since 0.5.0

type Sink i = ConduitM i Void

Consumes a stream of input values and produces a final result, without producing any output.

type Sink i m r = ConduitM i Void m r

Since 0.5.0

data ConduitM i o m r

Core datatype of the conduit package. This type represents a general component which can consume a stream of input values i, produce a stream of output values o, perform actions in the m monad, and produce a final result r. The type synonyms provided here are simply wrappers around this type.

Since 1.0.0

Instances

MonadRWS r w s m => MonadRWS r w s (ConduitM i o m)

MonadBase base m => MonadBase base (ConduitM i o m)

MonadError e m => MonadError e (ConduitM i o m)

MonadReader r m => MonadReader r (ConduitM i o m)

MonadWriter w m => MonadWriter w (ConduitM i o m)

MonadState s m => MonadState s (ConduitM i o m)

MFunctor (ConduitM i o)

MonadTrans (ConduitM i o)

Monad (ConduitM i o m)

Functor (ConduitM i o m)

Applicative (ConduitM i o m)

MonadThrow m => MonadThrow (ConduitM i o m)

MonadCatch m => MonadCatch (ConduitM i o m)

MonadResource m => MonadResource (ConduitM i o m)

MonadIO m => MonadIO (ConduitM i o m)

Monad m => Monoid (ConduitM i o m ())

Connect/fuse operators

($$) :: Monad m => Source m a -> Sink a m b -> m b infixr 0

The connect operator, which pulls data from a source and pushes to a sink. If you would like to keep the Source open to be used for other operations, use the connect-and-resume operator $$+.

Since 0.4.0

($=) :: Monad m => Conduit a m b -> ConduitM b c m r -> ConduitM a c m r infixl 1

Left fuse, combining a source and a conduit together into a new source.

Both the Source and Conduit will be closed when the newly-created Source is closed.

Leftover data from the Conduit will be discarded.

Note: Since version 1.0.18, this operator has been generalized to be identical to =$=.

Since 0.4.0

(=$) :: Monad m => Conduit a m b -> ConduitM b c m r -> ConduitM a c m r infixr 2

Right fuse, combining a conduit and a sink together into a new sink.

Both the Conduit and Sink will be closed when the newly-created Sink is closed.

Leftover data returned from the Sink will be discarded.

Note: Since version 1.0.18, this operator has been generalized to be identical to =$=.

Since 0.4.0

(=$=) :: Monad m => Conduit a m b -> ConduitM b c m r -> ConduitM a c m r infixr 2

Fusion operator, combining two Conduits together into a new Conduit.

Both Conduits will be closed when the newly-created Conduit is closed.

Leftover data returned from the right Conduit will be discarded.

Since 0.4.0

connect :: Monad m => Source m a -> Sink a m b -> m b

Named function synonym for $$.

Since 1.2.3

fuse :: Monad m => Conduit a m b -> ConduitM b c m r -> ConduitM a c m r

Named function synonym for =$=.

Since 1.2.3

Fuse with upstream results

fuseBoth :: Monad m => ConduitM a b m r1 -> ConduitM b c m r2 -> ConduitM a c m (r1, r2)

Fuse two ConduitMs together, and provide the return value of both. Note that this will force the entire upstream ConduitM to be run to produce the result value, even if the downstream terminates early.

Since 1.1.5

fuseUpstream :: Monad m => ConduitM a b m r -> Conduit b m c -> ConduitM a c m r

Same as fuseBoth, but ignore the return value from the downstream Conduit. Same caveats of forced consumption apply.

Since 1.1.5

Primitives

await :: Monad m => Consumer i m (Maybe i)

Wait for a single input value from upstream. If no data is available, returns Nothing.

Since 0.5.0

yield

Arguments

:: Monad m

=> o

output value

-> ConduitM i o m ()

Send a value downstream to the next component to consume. If the downstream component terminates, this call will never return control. If you would like to register a cleanup function, please use yieldOr instead.

Since 0.5.0

leftover :: i -> ConduitM i o m ()

Provide a single piece of leftover input to be consumed by the next component in the current monadic binding.

Note: it is highly encouraged to only return leftover values from input already consumed from upstream.

Since 0.5.0

runConduit :: Monad m => ConduitM () Void m r -> m r

Run a pipeline until processing completes.

Since 1.2.1

Finalization

bracketP :: MonadResource m => IO a -> (a -> IO ()) -> (a -> ConduitM i o m r) -> ConduitM i o m r

Perform some allocation and run an inner component. Two guarantees are given about resource finalization:

1. It will be prompt. The finalization will be run as early as possible.
2. It is exception safe. Due to usage of resourcet, the finalization will be run in the event of any exceptions.

Since 0.5.0

addCleanup :: Monad m => (Bool -> m ()) -> ConduitM i o m r -> ConduitM i o m r

Add some code to be run when the given component cleans up.

The supplied cleanup function will be given a True if the component ran to completion, or False if it terminated early due to a downstream component terminating.

Note that this function is not exception safe. For that, please use bracketP.

Since 0.4.1

yieldOr

Arguments

:: Monad m

=> o

-> m ()

finalizer

-> ConduitM i o m ()

Similar to yield, but additionally takes a finalizer to be run if the downstream component terminates.

Since 0.5.0

Exception handling

catchC :: (MonadBaseControl IO m, Exception e) => ConduitM i o m r -> (e -> ConduitM i o m r) -> ConduitM i o m r

Catch all exceptions thrown by the current component of the pipeline.

Note: this will not catch exceptions thrown by other components! For example, if an exception is thrown in a Source feeding to a Sink, and the Sink uses catchC, the exception will not be caught.

Due to this behavior (as well as lack of async exception handling), you should not try to implement combinators such as onException in terms of this primitive function.

Note also that the exception handling will not be applied to any finalizers generated by this conduit.

Since 1.0.11

handleC :: (MonadBaseControl IO m, Exception e) => (e -> ConduitM i o m r) -> ConduitM i o m r -> ConduitM i o m r

The same as flip catchC.

Since 1.0.11

tryC :: (MonadBaseControl IO m, Exception e) => ConduitM i o m r -> ConduitM i o m (Either e r)

A version of try for use within a pipeline. See the comments in catchC for more details.

Since 1.0.11

Generalized conduit types

type Producer m o = forall i. ConduitM i o m ()

A component which produces a stream of output values, regardless of the input stream. A Producer is a generalization of a Source, and can be used as either a Source or a Conduit.

Since 1.0.0

type Consumer i m r = forall o. ConduitM i o m r

A component which consumes a stream of input values and produces a final result, regardless of the output stream. A Consumer is a generalization of a Sink, and can be used as either a Sink or a Conduit.

Since 1.0.0

toProducer :: Monad m => Source m a -> Producer m a

Generalize a Source to a Producer.

Since 1.0.0

toConsumer :: Monad m => Sink a m b -> Consumer a m b

Generalize a Sink to a Consumer.

Since 1.0.0

Utility functions

awaitForever :: Monad m => (i -> ConduitM i o m r) -> ConduitM i o m ()

Wait for input forever, calling the given inner component for each piece of new input. Returns the upstream result type.

This function is provided as a convenience for the common pattern of awaiting input, checking if it's Just and then looping.

Since 0.5.0

transPipe :: Monad m => (forall a. m a -> n a) -> ConduitM i o m r -> ConduitM i o n r

Transform the monad that a ConduitM lives in.

Note that the monad transforming function will be run multiple times, resulting in unintuitive behavior in some cases. For a fuller treatment, please see:

https://github.com/snoyberg/conduit/wiki/Dealing-with-monad-transformers

This function is just a synonym for hoist.

Since 0.4.0

mapOutput :: Monad m => (o1 -> o2) -> ConduitM i o1 m r -> ConduitM i o2 m r

Apply a function to all the output values of a ConduitM.

This mimics the behavior of fmap for a Source and Conduit in pre-0.4 days. It can also be simulated by fusing with the map conduit from Data.Conduit.List.

Since 0.4.1

mapOutputMaybe :: Monad m => (o1 -> Maybe o2) -> ConduitM i o1 m r -> ConduitM i o2 m r

Same as mapOutput, but use a function that returns Maybe values.

Since 0.5.0

mapInput

Arguments

:: Monad m

=> (i1 -> i2)

map initial input to new input

-> (i2 -> Maybe i1)

map new leftovers to initial leftovers

-> ConduitM i2 o m r

-> ConduitM i1 o m r

Apply a function to all the input values of a ConduitM.

Since 0.5.0

passthroughSink

Arguments

:: Monad m

=> Sink i m r

-> (r -> m ())

finalizer

-> Conduit i m i

Turn a Sink into a Conduit in the following way:

• All input passed to the Sink is yielded downstream.
• When the Sink finishes processing, the result is passed to the provided to the finalizer function.

Note that the Sink will stop receiving input as soon as the downstream it is connected to shuts down.

An example usage would be to write the result of a Sink to some mutable variable while allowing other processing to continue.

Since 1.1.0

Connect-and-resume

data ResumableSource m o

A Source which has been started, but has not yet completed.

This type contains both the current state of the Source, and the finalizer to be run to close it.

Since 0.5.0

Instances

MFunctor ResumableSource

Since 1.0.13

newResumableSource :: Monad m => Source m o -> ResumableSource m o

Turn a Source into a ResumableSource with no attached finalizer.

Since 1.1.4

($$+) :: Monad m => Source m a -> Sink a m b -> m (ResumableSource m a, b) infixr 0

The connect-and-resume operator. This does not close the Source, but instead returns it to be used again. This allows a Source to be used incrementally in a large program, without forcing the entire program to live in the Sink monad.

Mnemonic: connect + do more.

Since 0.5.0

($$++) :: Monad m => ResumableSource m a -> Sink a m b -> m (ResumableSource m a, b) infixr 0

Continue processing after usage of $$+.

Since 0.5.0

($$+-) :: Monad m => ResumableSource m a -> Sink a m b -> m b infixr 0

Complete processing of a ResumableSource. This will run the finalizer associated with the ResumableSource. In order to guarantee process resource finalization, you must use this operator after using $$+ and $$++.

Since 0.5.0

($=+) :: Monad m => ResumableSource m a -> Conduit a m b -> ResumableSource m b infixl 1

Left fusion for a resumable source.

Since 1.0.16

unwrapResumable :: MonadIO m => ResumableSource m o -> m (Source m o, m ())

Unwraps a ResumableSource into a Source and a finalizer.

A ResumableSource represents a Source which has already been run, and therefore has a finalizer registered. As a result, if we want to turn it into a regular Source, we need to ensure that the finalizer will be run appropriately. By appropriately, I mean:

• If a new finalizer is registered, the old one should not be called.
• If the old one is called, it should not be called again.

This function returns both a Source and a finalizer which ensures that the above two conditions hold. Once you call that finalizer, the Source is invalidated and cannot be used.

Since 0.5.2

closeResumableSource :: Monad m => ResumableSource m a -> m ()

Execute the finalizer associated with a ResumableSource, rendering the ResumableSource invalid for further use.

This is just a more explicit version of $$+- return ().

Since 1.1.3

For Conduits

data ResumableConduit i m o

A generalization of ResumableSource. Allows to resume an arbitrary conduit, keeping its state and using it later (or finalizing it).

Since 1.0.17

newResumableConduit :: Monad m => Conduit i m o -> ResumableConduit i m o

Turn a Conduit into a ResumableConduit with no attached finalizer.

Since 1.1.4

(=$$+) :: Monad m => Conduit a m b -> Sink b m r -> Sink a m (ResumableConduit a m b, r) infixr 0

The connect-and-resume operator. This does not close the Conduit, but instead returns it to be used again. This allows a Conduit to be used incrementally in a large program, without forcing the entire program to live in the Sink monad.

Leftover data returned from the Sink will be discarded.

Mnemonic: connect + do more.

Since 1.0.17

(=$$++) :: Monad m => ResumableConduit i m o -> Sink o m r -> Sink i m (ResumableConduit i m o, r) infixr 0

Continue processing after usage of =$$+. Connect a ResumableConduit to a sink and return the output of the sink together with a new ResumableConduit.

Since 1.0.17

(=$$+-) :: Monad m => ResumableConduit i m o -> Sink o m r -> Sink i m r infixr 0

Complete processing of a ResumableConduit. This will run the finalizer associated with the ResumableConduit. In order to guarantee process resource finalization, you must use this operator after using =$$+ and =$$++.

Since 1.0.17

unwrapResumableConduit :: MonadIO m => ResumableConduit i m o -> m (Conduit i m o, m ())

Unwraps a ResumableConduit into a Conduit and a finalizer.

Since unwrapResumable for more information.

Since 1.0.17

Fusion with leftovers

fuseLeftovers :: Monad m => ([b] -> [a]) -> ConduitM a b m () -> ConduitM b c m r -> ConduitM a c m r

Similar to fuseReturnLeftovers, but use the provided function to convert downstream leftovers to upstream leftovers.

Since 1.0.17

fuseReturnLeftovers :: Monad m => ConduitM a b m () -> ConduitM b c m r -> ConduitM a c m (r, [b])

Same as normal fusion (e.g. =$=), except instead of discarding leftovers from the downstream component, return them.

Since 1.0.17

Flushing

data Flush a

Provide for a stream of data that can be flushed.

A number of Conduits (e.g., zlib compression) need the ability to flush the stream at some point. This provides a single wrapper datatype to be used in all such circumstances.

Since 0.3.0

Constructors

Chunk a

Flush

Instances

Functor Flush

Eq a => Eq (Flush a)

Ord a => Ord (Flush a)

Show a => Show (Flush a)

Newtype wrappers

ZipSource

newtype ZipSource m o

A wrapper for defining an Applicative instance for Sources which allows to combine sources together, generalizing zipSources. A combined source will take input yielded from each of its Sources until any of them stop producing output.

Since 1.0.13

Constructors

ZipSource

Fields

getZipSource :: Source m o

Instances

Monad m => Functor (ZipSource m)

Monad m => Applicative (ZipSource m)

sequenceSources :: (Traversable f, Monad m) => f (Source m o) -> Source m (f o)

Coalesce all values yielded by all of the Sources.

Implemented on top of ZipSource, see that data type for more details.

Since 1.0.13

ZipSink

newtype ZipSink i m r

A wrapper for defining an Applicative instance for Sinks which allows to combine sinks together, generalizing zipSinks. A combined sink distributes the input to all its participants and when all finish, produces the result. This allows to define functions like

sequenceSinks :: (Monad m)
          => [Sink i m r] -> Sink i m [r]
sequenceSinks = getZipSink . sequenceA . fmap ZipSink

Note that the standard Applicative instance for conduits works differently. It feeds one sink with input until it finishes, then switches to another, etc., and at the end combines their results.

Since 1.0.13

Constructors

ZipSink

Fields

getZipSink :: Sink i m r

Instances

Monad m => Functor (ZipSink i m)

Monad m => Applicative (ZipSink i m)

sequenceSinks :: (Traversable f, Monad m) => f (Sink i m r) -> Sink i m (f r)

Send incoming values to all of the Sink providing, and ultimately coalesce together all return values.

Implemented on top of ZipSink, see that data type for more details.

Since 1.0.13

ZipConduit

newtype ZipConduit i o m r

Provides an alternative Applicative instance for ConduitM. In this instance, every incoming value is provided to all ConduitMs, and output is coalesced together. Leftovers from individual ConduitMs will be used within that component, and then discarded at the end of their computation. Output and finalizers will both be handled in a left-biased manner.

As an example, take the following program:

main :: IO ()
main = do
    let src = mapM_ yield [1..3 :: Int]
        conduit1 = CL.map (+1)
        conduit2 = CL.concatMap (replicate 2)
        conduit = getZipConduit $ ZipConduit conduit1 <* ZipConduit conduit2
        sink = CL.mapM_ print
    src $$ conduit =$ sink

It will produce the output: 2, 1, 1, 3, 2, 2, 4, 3, 3

Since 1.0.17

Constructors

ZipConduit

Fields

getZipConduit :: ConduitM i o m r

Instances

Functor (ZipConduit i o m)

Monad m => Applicative (ZipConduit i o m)

sequenceConduits :: (Traversable f, Monad m) => f (ConduitM i o m r) -> ConduitM i o m (f r)

Provide identical input to all of the Conduits and combine their outputs into a single stream.

Implemented on top of ZipConduit, see that data type for more details.

Since 1.0.17