3-- Copyright : (c) 2017 Composewell Technologies
6-- Maintainer : streamly@composewell.com
7-- Stability : experimental
10-- Streamly is a general purpose programming framework using cocnurrent data
11-- flow programming paradigm. It can be considered as a generalization of
12-- Haskell lists to monadic streaming with concurrent composition capability.
13-- The serial stream type in streamly @SerialT m a@ is like the list type @[a]@
14-- parameterized by the monad @m@. For example, @SerialT IO a@ is a moral
15-- equivalent of @[a]@ in the IO monad. Streams are constructed very much like
16-- lists, except that they use 'nil' and 'cons' instead of '[]' and ':'.
20-- > import "Streamly.Prelude" (cons, consM)
21-- > import qualified "Streamly.Prelude" as S
23-- > S.toList $ 1 \`cons` 2 \`cons` 3 \`cons` nil
27-- Unlike lists, streams can be constructed from monadic effects:
30-- > S.'toList' $ 'getLine' \`consM` 'getLine' \`consM` S.'nil'
36-- Streams are processed just like lists, with list like combinators, except
37-- that they are monadic and work in a streaming fashion. Here is a simple
38-- console echo program example:
41-- > S.drain $ S.repeatM getLine & S.mapM putStrLn
44-- @SerialT Identity a@ is a moral equivalent of pure lists. Streamly utilizes
45-- fusion for high performance, therefore, we can represent and process strings
46-- as streams of 'Char', encode and decode the streams to/from UTF8 and
47-- serialize them to @Array Word8@ obviating the need for special purpose
48-- libraries like @bytestring@ and @text@.
51-- please see the "Streamly.Tutorial" module and the examples directory in this
54{-# LANGUAGE NoMonomorphismRestriction #-}
55{-# OPTIONS_GHC -Wno-orphans #-}
59module Streamly {-# DEPRECATED "Please use \"Streamly.Prelude\" instead." #-}
61 -- -- * Concepts Overview
77 -- * Stream transformers
78 -- | A stream represents a sequence of pure or effectful actions. The
79 -- `cons` and `consM` operations and the corresponding operators '.:' and
80 -- '|:' can be used to join pure values or effectful actions in a sequence.
81 -- The effects in the stream can be executed in many different ways
82 -- depending on the type of stream. In other words, the behavior of 'consM'
83 -- depends on the type of the stream.
85 -- There are three high level categories of streams, /spatially ordered/
86 -- streams, /speculative/ streams and /time ordered/ streams. Spatially
87 -- ordered streams, 'SerialT' and 'WSerialT', execute the effects in serial
88 -- order i.e. one at a time and present the outputs of those effects to the
89 -- consumer in the same order. Speculative streams, 'AheadT', may execute
90 -- many effects concurrently but present the outputs to the consumer in the
91 -- specified spatial order. Time ordered streams, 'AsyncT', 'WAsyncT' and
92 -- 'ParallelT', may execute many effects concurrently and present the
93 -- outputs of those effects to the consumer in time order i.e. as soon as
94 -- the output is generated.
96 -- We described above how the effects in a sequence are executed for
97 -- different types of streams. The behvavior of the 'Semigroup' and 'Monad'
98 -- instances follow the behavior of 'consM'. Stream generation operations
99 -- like 'repeatM' also execute the effects differently for different
100 -- streams, providing a concurrent generation capability when used with
101 -- stream types that execute effects concurrently. Similarly, effectful
102 -- transformation operations like 'mapM' also execute the transforming
103 -- effects differently for different types of streams.
110 -- ** Speculative Streams
114 -- ** Asynchronous Streams
120 -- ** Zipping Streams
125 -- * Parallel Function Application
142 -- * Concurrency Control
155 -- * Stream Type Adapters
179 -- ** Folding Containers of Streams
180 -- | These are variants of standard 'Foldable' fold functions that use a
181 -- polymorphic stream sum operation (e.g. 'async' or 'wSerial') to fold a
182 -- finite container of streams. Note that these are just special cases of
183 -- the more general 'concatMapWith' operation.
212import Data.Semigroup (Semigroup(..))
213import Streamly.Internal.Control.Concurrent (MonadAsync)
214import Streamly.Internal.Data.SVar (Rate(..))
215import Streamly.Internal.Data.Stream.Ahead
216import Streamly.Internal.Data.Stream.Async
217import Streamly.Internal.Data.Stream.IsStream.Combinators
218import Streamly.Internal.Data.Stream.IsStream.Expand
219import Streamly.Internal.Data.Stream.IsStream.Type
220import Streamly.Internal.Data.Stream.Serial (StreamT, InterleavedT)
221import Streamly.Internal.Data.Stream.Zip
223import qualified Streamly.Prelude as P
224import qualified Streamly.Internal.Data.Stream.IsStream as IP
225import qualified Streamly.Internal.Data.Stream.IsStream.Transform as Transform
227-- XXX provide good succinct examples of pipelining, merging, splitting etc.
232-- A program is expressed as a network of streams and folds. A stream is a
233-- source or generator of data elements and a fold is a consumer of data elements that
234-- reduces multiple input elements to a single value.
236-- In the following example, a 'Word8' stream is generated by using
237-- 'Streamly.FileSystem.Handle.read' on a file handle, then the
238-- 'Streamly.Prelude.splitBySuffix' transformation splits the stream on
239-- newlines (ascii value 10); it uses the 'Streamly.Data.Fold.drain' fold to reduce
240-- the resulting lines to unit values (@()@), 'Streamly.Prelude.length' fold
241-- then counts the unit elements in the resulting stream which gives us the
242-- number of lines in the file:
244-- > S.length $ S.splitOnSuffix FL.drain 10 $ FH.read fh
246-- The following example folds the lines to arrays of 'Word8' using the
247-- 'Streamly.Data.Array.Foreign.writeF' fold and then wraps the lines in square
248-- brackets before writing them to standard output using
249-- 'Streamly.FileSystem.Handle.write':
251-- > wrapLine ln = S.fromList "[" <> A.read ln <> S.fromList "]\n"
252-- > readLines = S.splitOnSuffix A.writeF 10
253-- > FH.write stdout $ S.concatMap wrapLine $ readLines fh1
255-- One stream can be appended after another:
257-- > FH.write stdout $ S.concatMap wrapLine $ readLines fh1 <> readLines fh2
259-- The following example reads two files concurrently, merges the lines from
260-- the two streams and writes the resulting stream to another file:
262-- > FH.write stdout $ S.concatMap wrapLine $ readLines fh1 `parallel` readLines fh2
264-- There are many ways to generate, merge, zip, transform and fold data
265-- streams. Many transformations can be chained in a stream pipeline. See
266-- "Streamly.Prelude" module for combinators to manipulate streams.
270-- The way stream types in this module like 'SerialT' represent data sources,
271-- the same way the 'Fold' type from "Streamly.Data.Fold" represents data sinks or
272-- reducers of streams. Reducers can be combined to consume a stream source in
273-- many ways. The simplest is to reduce a stream source using a fold e.g.:
275-- > S.fold FL.length $ S.enumerateTo 100
277-- Folds are consumers of streams and can be used to split a stream into
278-- multiple independent flows. Grouping transforms a stream by applying a fold
279-- on segments of a stream, distributing applies multiple folds in parallel on
280-- the same stream and combines them, partitioning sends different elements of
281-- a stream to different folds, unzipping divides the elements of a stream into
282-- parts and sends them through different folds. Parsers are nothing but a
283-- particular type of folds. Transformations can be applied contravariantly on
284-- the input of a fold.
288-- |---transform----Fold m a b--------|
289-- ---stream m a-->transform-->| |---f b c ...
290-- |---transform----Fold m a c--------|
298-- Streamly arrays (See "Streamly.Data.Array.Foreign") complement streams to provide an
299-- efficient computing paradigm. Streams are suitable for immutable
300-- transformations of /potentially infinite/ data using /sequential access/ and
301-- pipelined transformations whereas arrays are suitable for in-place
302-- transformations of /necessarily finite/ data using /random access/. Streams
303-- are synonymous with /sequential pipelined processing/ whereas arrays are
304-- synonymous with /efficient buffering and random access/.
306-- In general, a data processing pipeline reads data from some IO device, does
307-- some processing on it and finally writes the output to another IO device.
308-- Streams provide the overall framework of sequential processing pipeline in
309-- which arrays are used as buffering elements in the middle. In addition to
310-- buffering in the middle, arrays can also be used at the boundaries of the
311-- pipeline to efficiently interface with external storage systems like memory,
312-- files and network. If streams are the pipes in a water pipeline network
313-- then arrays are like the storage tanks in the middle. On the input side,
314-- think of arrays as buckets to fetch water to feed the pipeline and on the
315-- output side buckets to remove the processed water.
317-- 'ByteString' data type from the 'bytestring' package and the 'Text' data
318-- type from the 'text' package are special cases of arrays. 'ByteString' is
319-- like @Array Word8@ and 'Text' is like @utf16@ encoded @Array Word8@.
320-- Streamly arrays can be transformed as efficiently as @bytestring@ or @text@
321-- by using stream operations on them.
323-- Streams and arrays are equally important in computing. They are computing
324-- duals of each other.
327-- The basic stream type is 'Serial', it represents a sequence of IO actions,
328-- and is a 'Monad'. The type 'SerialT' is a monad transformer that can
329-- represent a sequence of actions in an arbitrary monad. The type 'Serial' is
330-- in fact a synonym for @SerialT IO@. There are a few more types similar to
331-- 'SerialT', all of them represent a stream and differ only in the
332-- 'Semigroup', 'Applicative' and 'Monad' compositions of the stream. 'Serial'
333-- and 'WSerial' types compose serially whereas 'Async' and 'WAsync'
334-- types compose concurrently. All these types can be freely inter-converted
335-- using type combinators without any cost. You can freely switch to any type
336-- of composition at any point in the program. When no type annotation or
337-- explicit stream type combinators are used, the default stream type is
338-- inferred as 'Serial'.
340-- This module exports stream types, instances and combinators for:
342-- * converting between different stream types
343-- * appending and concurrently merging streams
344-- * Concurrency control
345-- * Concurrent function application
346-- * Stream rate control
348-- This module is designed to be imported unqualified:
354-- See the "Streamly.Prelude" module for APIs for construction,
355-- generation, elimination and transformation of streams.
357------------------------------------------------------------------------------
358-- Eliminating a stream
359------------------------------------------------------------------------------
361-- | Same as 'runStream'
364{-# DEPRECATED runStreaming "Please use runStream instead." #-}
365runStreaming :: (Monad m, IsStream t) => t m a -> m ()
366runStreaming = P.drain . adapt
368-- | Same as @runStream@.
371{-# DEPRECATED runStreamT "Please use runStream instead." #-}
372runStreamT :: Monad m => SerialT m a -> m ()
375-- | Same as "Streamly.Prelude.runStream".
377{-# DEPRECATED runStream "Please use Streamly.Prelude.drain instead." #-}
378runStream :: Monad m => SerialT m a -> m ()
381-- | Same as @drain . fromWSerial@.
384{-# DEPRECATED runInterleavedT "Please use 'drain . interleaving' instead." #-}
385runInterleavedT :: Monad m => WSerialT m a -> m ()
386runInterleavedT = P.drain . adapt
388-- | Same as @drain . fromParallel@.
391{-# DEPRECATED runParallelT "Please use 'drain . fromParallel' instead." #-}
392runParallelT :: Monad m => ParallelT m a -> m ()
393runParallelT = P.drain . adapt
395-- | Same as @drain . fromAsync@.
398{-# DEPRECATED runAsyncT "Please use 'drain . fromAsync' instead." #-}
399runAsyncT :: Monad m => AsyncT m a -> m ()
400runAsyncT = P.drain . adapt
402-- | Same as @drain . zipping@.
405{-# DEPRECATED runZipStream "Please use 'drain . fromZipSerial instead." #-}
406runZipStream :: Monad m => ZipSerialM m a -> m ()
407runZipStream = P.drain . adapt
409-- | Same as @drain . zippingAsync@.
412{-# DEPRECATED runZipAsync "Please use 'drain . fromZipAsync instead." #-}
413runZipAsync :: Monad m => ZipAsyncM m a -> m ()
414runZipAsync = P.drain . adapt
417-- | Same as "Streamly.Prelude.foldWith".
419{-# DEPRECATED foldWith "Please use Streamly.Prelude.foldWith instead." #-}
420{-# INLINABLE foldWith #-}
421foldWith :: (IsStream t, Foldable f)
422 => (t m a -> t m a -> t m a) -> f (t m a) -> t m a
425-- | Same as "Streamly.Prelude.foldMapWith".
427{-# DEPRECATED foldMapWith "Please use Streamly.Prelude.foldMapWith instead." #-}
428{-# INLINABLE foldMapWith #-}
429foldMapWith :: (IsStream t, Foldable f)
430 => (t m b -> t m b -> t m b) -> (a -> t m b) -> f a -> t m b
431foldMapWith = P.foldMapWith
433-- | Same as "Streamly.Prelude.forEachWith".
435{-# DEPRECATED forEachWith "Please use Streamly.Prelude.forEachWith instead." #-}
436{-# INLINABLE forEachWith #-}
437forEachWith :: (IsStream t, Foldable f)
438 => (t m b -> t m b -> t m b) -> f a -> (a -> t m b) -> t m b
439forEachWith = P.forEachWith
442-- XXX Deprecate it in 0.8.0
444-- | Make a stream asynchronous, triggers the computation and returns a stream
445-- in the underlying monad representing the output generated by the original
446-- computation. The returned action is exhaustible and must be drained once. If
447-- not drained fully we may have a thread blocked forever and once exhausted it
448-- will always return 'empty'.
451{-# INLINABLE mkAsync #-}
452mkAsync :: (IsStream t, MonadAsync m) => t m a -> m (t m a)
453mkAsync = return . Transform.mkAsync
455------------------------------------------------------------------------------
457------------------------------------------------------------------------------
461-- When a stream consumer demands an element from a serial stream constructed
462-- as @a \`consM` b \`consM` ... nil@, the action @a@ at the head of the stream
463-- sequence is executed and the result is supplied to the consumer. When the
464-- next element is demanded, the action @b@ is executed and its result is
465-- supplied. Thus, the effects are performed and results are consumed strictly
466-- in a serial order. Serial streams can be considered as /spatially ordered/
467-- streams as the order of execution and consumption is the same as the spatial
468-- order in which the actions are composed by the programmer.
470-- Serial streams enforce the side effects as well as the results of the
471-- actions to be in the same order in which the actions are added to the
472-- stream. Therefore, the semigroup operation for serial streams is not
476-- a <> b is not the same as b <> a
479-- There are two serial stream types 'SerialT' and 'WSerialT'. The stream
480-- evaluation of both the variants works in the same way as described above,
481-- they differ only in the 'Semigroup' and 'Monad' implementaitons.
485-- When a stream consumer demands an element from a speculative stream
486-- constructed as @a \`consM` b \`consM` ... nil@, the action @a@ at the head
487-- of the stream is executed and the output of the action is supplied to the
488-- consumer. However, in addition to the action at the head multiple actions
489-- following it may also be executed concurrently and the results buffered.
490-- When the next element is demanded it may be served from the buffer and we
491-- may execute the next action in the sequence to keep the buffer adequately
492-- filled. Thus, the actions are executed concurrently but results consumed in
493-- serial order just like serial streams. `consM` can be used to fold an
494-- infinite lazy container of effects, as the number of concurrent executions
497-- Similar to 'consM', the monadic stream generation (e.g. replicateM) and
498-- transformation operations (e.g. mapM) on speculative streams can execute
499-- multiple effects concurrently in a speculative manner.
501-- How many effects can be executed concurrently and how many results can be
502-- buffered are controlled by 'maxThreads' and 'maxBuffer' combinators
503-- respectively. The actual number of concurrent threads is adjusted according
504-- to the rate at which the consumer is consuming the stream. It may even
505-- execute actions serially in a single thread if that is enough to match the
508-- Speculative streams enforce ordering of the results of actions in the stream
509-- but the side effects are only partially ordered. Therefore, the semigroup
510-- operation for speculative streams is not commutative from the pure outputs
511-- perspective but commutative from side effects perspective.
515-- /Scheduling and execution:/ In an asynchronous stream @a \`consM` b \`consM`
516-- c ...@, the actions @a@, @b@, and @c@ are executed concurrently with the
517-- consumer of the stream. The actions are /scheduled/ for execution in the
518-- same order as they are specified in the stream. Multiple scheduled actions
519-- may be /executed/ concurrently in parallel threads of execution. The
520-- actions may be executed out of order and they may complete at arbitrary
521-- times. Therefore, the /effects/ of the actions may be observed out of
524-- /Buffering:/ The /results/ from multiple threads of execution are queued in
525-- a buffer as soon as they become available. The consumer of the stream is
526-- served from this buffer. Therefore, the consumer may observe the results to
527-- be out of order. In other words, an asynchronous stream is an unordered
528-- stream i.e. order does not matter.
530-- /Concurrency control:/ Threads are suspended if the `maxBuffer` limit is
531-- reached, and resumed when the consumer makes space in the buffer. The
532-- maximum number of concurrent threads depends on `maxThreads`. Number of
533-- threads is increased or decreased based on the speed of the consumer.
535-- /Generation operations:/ Concurrent stream generation operations e.g.
536-- 'Streamly.Prelude.replicateM' when used in async style schedule and execute
537-- the stream generating actions in the manner described above. The generation
538-- actions run concurrently, effects and results of the actions as observed by
539-- the consumer of the stream may be out of order.
541-- /Transformation operations:/ Concurrent stream transformation operations
542-- e.g. 'Streamly.Prelude.mapM', when used in async style, schedule and
543-- execute transformation actions in the manner described above. Transformation
544-- actions run concurrently, effects and results of the actions may be
545-- observed by the consumer out of order.
547-- /Variants:/ There are two asynchronous stream types 'AsyncT' and 'WAsyncT'.
548-- They are identical with respect to single stream evaluation behavior. Their
549-- behaviors differ in how they combine multiple streams using 'Semigroup' or
550-- 'Monad' composition. Since the order of elements does not matter in
551-- asynchronous streams the 'Semigroup' operation is effectively commutative.
555-- 'ZipSerialM' and 'ZipAsyncM', provide 'Applicative' instances for zipping the
556-- corresponding elements of two streams together. Note that these types are
561-- Stream processing functions can be composed in a chain using function
562-- application with or without the '$' operator, or with reverse function
563-- application operator '&'. Streamly provides concurrent versions of these
564-- operators applying stream processing functions such that each stage of the
565-- stream can run in parallel. The operators start with a @|@; we can read '|$'
566-- as "@parallel dollar@" to remember that @|@ comes before '$'.
568-- Imports for the code snippets below:
572-- import qualified Streamly.Prelude as S
573-- import Control.Concurrent
577-- The 'Semigroup' operation '<>' of each stream type combines two streams in a
578-- type specific manner. This section provides polymorphic versions of '<>'
579-- which can be used to combine two streams in a predetermined way irrespective
582-- XXX An alternative design choice would be to let a control parameter affect
583-- the nearest SVar only and then it gets cleared. The benefit of the current
584-- choice is that it is simply just like global configuration, just like state
585-- behaves, so should be easy to comprehend. But it has the downside of leaking
586-- to undesired actions, that is we can forget to reset it.
590-- These combinators can be used at any point in a stream composition to set
591-- parameters to control the concurrency of the /argument stream/. A control
592-- parameter set at any point remains effective for any concurrent combinators
593-- used in the argument stream until it is reset by using the combinator again.
594-- These control parameters have no effect on non-concurrent combinators in the
595-- stream, or on non-concurrent streams.
597-- /Pitfall:/ Remember that 'maxBuffer' in the following example applies to
598-- 'mapM' and any other combinators that may follow it, and it does not apply
599-- to the combinators before it:
608-- If we use '&' instead of '$' the situation will reverse, in the following
609-- example, 'maxBuffer' does not apply to 'mapM', it applies to combinators
610-- that come before it, because those are the arguments to 'maxBuffer':
621-- You may want to use different stream composition styles at different points
622-- in your program. Stream types can be freely converted or adapted from one
623-- type to another. The 'IsStream' type class facilitates type conversion of
624-- one stream type to another. It is not used directly, instead the type
625-- combinators provided below are used for conversions.
627-- To adapt from one monomorphic type (e.g. 'AsyncT') to another monomorphic
628-- type (e.g. 'SerialT') use the 'adapt' combinator. To give a polymorphic code
629-- a specific interpretation or to adapt a specific type to a polymorphic type
630-- use the type specific combinators e.g. 'fromAsync' or 'fromWSerial'. You
631-- cannot adapt polymorphic code to polymorphic code, as the compiler would not know
632-- which specific type you are converting from or to. If you see a an
633-- @ambiguous type variable@ error then most likely you are using 'adapt'
634-- unnecessarily on polymorphic code.
636-- | Same as 'Streamly.Prelude.concatFoldableWith'
639{-# DEPRECATED foldWith "Please use 'Streamly.Prelude.concatFoldableWith' instead." #-}
640{-# INLINEABLE foldWith #-}
641foldWith :: (IsStream t, Foldable f) => (t m a -> t m a -> t m a) -> f (t m a) -> t m a
642foldWith = P.concatFoldableWith
644-- | Same as 'Streamly.Prelude.concatMapFoldableWith'
647{-# DEPRECATED foldMapWith "Please use 'Streamly.Prelude.concatMapFoldableWith' instead." #-}
648{-# INLINEABLE foldMapWith #-}
649foldMapWith :: (IsStream t, Foldable f) => (t m b -> t m b -> t m b) -> (a -> t m b) -> f a -> t m b
650foldMapWith = P.concatMapFoldableWith
652-- | Same as 'Streamly.Prelude.concatForFoldableWith'
655{-# DEPRECATED forEachWith "Please use 'Streamly.Prelude.concatForFoldableWith' instead." #-}
656{-# INLINEABLE forEachWith #-}
657forEachWith :: (IsStream t, Foldable f) => (t m b -> t m b -> t m b) -> f a -> (a -> t m b) -> t m b
658forEachWith = P.concatForFoldableWith
660{-# DEPRECATED serially "Please use 'Streamly.Prelude.fromSerial' instead." #-}
661serially :: IsStream t => SerialT m a -> t m a
664{-# DEPRECATED wSerially "Please use 'Streamly.Prelude.fromWSerial' instead." #-}
665wSerially :: IsStream t => WSerialT m a -> t m a
666wSerially = fromWSerial
668{-# DEPRECATED asyncly "Please use 'Streamly.Prelude.fromAsync' instead." #-}
669asyncly :: IsStream t => AsyncT m a -> t m a
672{-# DEPRECATED aheadly "Please use 'Streamly.Prelude.fromAhead' instead." #-}
673aheadly :: IsStream t => AheadT m a -> t m a
676{-# DEPRECATED wAsyncly "Please use 'Streamly.Prelude.fromWAsync' instead." #-}
677wAsyncly :: IsStream t => WAsyncT m a -> t m a
680{-# DEPRECATED parallely "Please use 'Streamly.Prelude.fromParallel' instead." #-}
681parallely :: IsStream t => ParallelT m a -> t m a
682parallely = fromParallel
684{-# DEPRECATED zipSerially "Please use 'Streamly.Prelude.fromZipSerial' instead." #-}
685zipSerially :: IsStream t => ZipSerialM m a -> t m a
686zipSerially = fromZipSerial
688{-# DEPRECATED zipAsyncly "Please use 'Streamly.Prelude.fromZipAsync' instead." #-}
689zipAsyncly :: IsStream t => ZipAsyncM m a -> t m a
690zipAsyncly = fromZipAsync
keysesc to close
- j k
- move down, up
- enter
- open
- esc
- back, or up a level
- 1 2 3 4
- readme, files, commits, refs
- /
- filter
- y
- copy the clone command
- c
- compare the chosen commit with another (or shift-click both)
- ?
- show or hide this list