Bio.Iteratee.List

Monadic Iteratees: incremental input parsers, processors and transformers

This module provides many basic iteratees from which more complicated iteratees can be built. In general these iteratees parallel those in Data.List, with some additions.

Iteratees

Iteratee Utilities

isFinished :: Nullable s => Iteratee s m Bool

Check if a stream has received EOF.

stream2list :: Monad m => Iteratee [el] m [el]

Read a stream to the end and return all of its elements as a list. This iteratee returns all data from the stream *strictly*.

stream2stream :: (Monad m, Nullable s, Monoid s) => Iteratee s m s

Read a stream to the end and return all of its elements as a stream. This iteratee returns all data from the stream *strictly*.

Basic Iteratees

dropWhileStream :: (el -> Bool) -> Iteratee [el] m ()

Skip all elements while the predicate is true.

The analogue of List.dropWhile

dropStream :: Int -> Iteratee [el] m ()

Drop n elements of the stream, if there are that many.

The analogue of List.drop

headStream :: Iteratee [el] m el

Attempt to read the next element of the stream and return it Raise a (recoverable) error if the stream is terminated.

The analogue of List.head

Because head can raise an error, it shouldn't be used when constructing iteratees for convStream. Use tryHead instead.

tryHead :: Iteratee [el] m (Maybe el)

Similar to headStream, except it returns Nothing if the stream is terminated.

lastStream :: Iteratee [el] m el

Attempt to read the last element of the stream and return it Raise a (recoverable) error if the stream is terminated

The analogue of List.last

heads :: (Monad m, Eq el) => [el] -> Iteratee [el] m Int

Given a sequence of characters, attempt to match them against the characters on the stream. Return the count of how many characters matched. The matched characters are removed from the stream. For example, if the stream contains abd, then (heads abc) will remove the characters ab and return 2.

peekStream :: Iteratee [el] m (Maybe el)

Look ahead at the next element of the stream, without removing it from the stream. Return Just c if successful, return Nothing if the stream is terminated by EOF.

roll

Arguments

:: Monad m

=> Int

length of chunk (t)

-> Int

amount to consume (d)

-> Iteratee [el] m [[el]]

Return a chunk of t elements length while consuming d elements from the stream. Useful for creating a 'rolling average' with convStream.

lengthStream :: Num a => Iteratee [el] m a

Return the total length of the remaining part of the stream.

This forces evaluation of the entire stream.

The analogue of List.length

chunkLength :: Iteratee [el] m (Maybe Int)

Get the length of the current chunk, or Nothing if EOF.

This function consumes no input.

takeFromChunk :: Int -> Iteratee [el] m [el]

Take n elements from the current chunk, or the whole chunk if n is greater.

Nested iteratee combinators

breakStream :: (el -> Bool) -> Iteratee [el] m [el]

Takes an element predicate and returns the (possibly empty) prefix of the stream. None of the characters in the string satisfy the character predicate. If the stream is not terminated, the first character of the remaining stream satisfies the predicate.

N.B. breakE should be used in preference to breakStream. breakStream will retain all data until the predicate is met, which may result in a space leak.

The analogue of List.break

breakE :: (el -> Bool) -> Enumeratee [el] [el] m a

Takes an element predicate and an iteratee, running the iteratee on all elements of the stream until the predicate is met.

the following rule relates break to breakE break pred === joinI (breakE pred stream2stream)

breakE should be used in preference to break whenever possible.

takeStream

Arguments

:: Monad m

=> Int

number of elements to consume

-> Enumeratee [el] [el] m a

Read n elements from a stream and apply the given iteratee to the stream of the read elements. Unless the stream is terminated early, we read exactly n elements, even if the iteratee has accepted fewer.

The analogue of List.take

takeUpTo :: Monad m => Int -> Enumeratee [el] [el] m a

Read n elements from a stream and apply the given iteratee to the stream of the read elements. If the given iteratee accepted fewer elements, we stop. This is the variation of takeStream with the early termination of processing of the outer stream once the processing of the inner stream finished early.

Iteratees composed with takeUpTo will consume only enough elements to reach a done state. Any remaining data will be available in the outer stream.

> let iter = do
h <- joinI $ takeUpTo 5 I.head
t <- stream2list
return (h,t)

> enumPureNChunk [1..10::Int] 3 iter >>= run >>= print
(1,[2,3,4,5,6,7,8,9,10])

> enumPureNChunk [1..10::Int] 7 iter >>= run >>= print
(1,[2,3,4,5,6,7,8,9,10])

in each case, I.head consumes only one element, returning the remaining 4 elements to the outer stream

takeWhileE :: (el -> Bool) -> Enumeratee [el] [el] m a

Takes an element predicate and an iteratee, running the iteratee on all elements of the stream while the predicate is met.

This is preferred to takeWhile.

mapStream :: (el -> el') -> Enumeratee [el] [el'] m a

Map a function over an Iteratee. This one is reimplemented and differs from the the one in Data.Iteratee.ListLike in so far that it doesn't pass on an EOF received in the input, which is the expected behavior.

concatMapStream :: Monoid t => (a -> t) -> Enumeratee [a] t m r

Apply a function to the elements of a stream, concatenate the results into a stream. No giant intermediate list is produced.

concatMapStreamM :: Monad m => (a -> m t) -> Enumeratee [a] t m r

Apply a monadic function to the elements of a stream, concatenate the results into a stream. No giant intermediate list is produced.

mapMaybeStream :: (a -> Maybe b) -> Enumeratee [a] [b] m r

filterStream :: (el -> Bool) -> Enumeratee [el] [el] m a

Creates an enumeratee with only elements from the stream that satisfy the predicate function. The outer stream is completely consumed.

The analogue of List.filter

filterStreamM :: Monad m => (a -> m Bool) -> Enumeratee [a] [a] m r

Apply a monadic filter predicate to an Iteratee.

groupStreamBy :: Monad m => (t -> t -> Bool) -> m (Iteratee [t] m t2) -> Enumeratee [t] [t2] m a

Grouping on Iteratees. groupStreamBy cmp inner outer executes inner to obtain an Iteratee i, then passes elements e to i as long as cmp e0 e, where e0 is some preceeding element, is true. Else, the result of run i is passed to outer and groupStreamBy restarts. At end of input, the resulting outer is returned.

groupStreamOn :: (Monad m, Eq t1) => (e -> t1) -> (t1 -> m (Iteratee [e] m t2)) -> Enumeratee [e] [(t1, t2)] m a

Grouping on Iteratees. groupStreamOn proj inner outer executes inner (proj e), where e is the first input element, to obtain an Iteratee i, then passes elements e to i as long as proj e produces the same result. If proj e changes or the input ends, the pair of proj e and the result of run i is passed to outer. At end of input, the resulting outer is returned.

mergeStreams :: Monad m => (el1 -> el2 -> b) -> Enumeratee [el2] b (Iteratee [el1] m) a

mergeStreams offers another way to nest iteratees: as a monad stack. This allows for the possibility of interleaving data from multiple streams.

-- print each element from a stream of lines.
logger :: (MonadIO m) => Iteratee [ByteString] m ()
logger = mapStreamM_ (liftIO . putStrLn . B.unpack)

-- combine alternating lines from two sources
-- To see how this was derived, follow the types from
-- 'ileaveLines logger' and work outwards.
run =<< enumFile 10 "file1" (joinI $ enumLinesBS $
          ( enumFile 10 "file2" . joinI . enumLinesBS $ joinI
                (ileaveLines logger)) >>= run)

ileaveLines :: (Functor m, Monad m)
  => Enumeratee [ByteString] [ByteString] (Iteratee [ByteString] m)
       [ByteString]
ileaveLines = mergeStreams (\l1 l2 ->
   [B.pack "f1:\n\t" ,l1 ,B.pack "f2:\n\t" ,l2 ]

mergeByChunks

Arguments

:: Monad m

=> ([el1] -> [el2] -> c3)

merge function

-> ([el1] -> c3)

-> ([el2] -> c3)

-> Enumeratee [el2] c3 (Iteratee [el1] m) a

A version of mergeStreams which operates on chunks instead of elements.

mergeByChunks offers more control than mergeStreams. mergeStreams terminates when the first stream terminates, however mergeByChunks will continue until both streams are exhausted.

mergeByChunks guarantees that both chunks passed to the merge function will have the same number of elements, although that number may vary between calls.

Folds

foldStream :: (a -> el -> a) -> a -> Iteratee [el] m a

Left-associative fold that is strict in the accumulator. This function should be used in preference to foldl whenever possible.

The analogue of List.foldl'.

Enumerators

Basic enumerators

enumPureNChunk :: Monad m => [el] -> Int -> Enumerator [el] m a

The pure n-chunk enumerator It passes a given stream of elements to the iteratee in n-sized chunks.

Enumerator Combinators

enumWith :: Monad m => Iteratee [el] m a -> Iteratee [el] m b -> Iteratee [el] m (a, b)

Enumerate over two iteratees in parallel as long as the first iteratee is still consuming input. The second iteratee will be terminated with EOF when the first iteratee has completed. An example use is to determine how many elements an iteratee has consumed:

snd <$> enumWith (dropWhile (<5)) length

Compare to zipStreams

zipStreams :: Monad m => Iteratee [el] m a -> Iteratee [el] m b -> Iteratee [el] m (a, b)

Enumerate two iteratees over a single stream simultaneously.

Compare to List.zip.

zipStreams3 :: Monad m => Iteratee [el] m a -> Iteratee [el] m b -> Iteratee [el] m c -> Iteratee [el] m (a, b, c)

zipStreams4 :: Monad m => Iteratee [el] m a -> Iteratee [el] m b -> Iteratee [el] m c -> Iteratee [el] m d -> Iteratee [el] m (a, b, c, d)

zipStreams5 :: Monad m => Iteratee [el] m a -> Iteratee [el] m b -> Iteratee [el] m c -> Iteratee [el] m d -> Iteratee [el] m e -> Iteratee [el] m (a, b, c, d, e)

sequenceStreams_ :: Monad m => [Iteratee [el] m a] -> Iteratee [el] m ()

Enumerate a list of iteratees over a single stream simultaneously and discard the results. This is a different behavior than Prelude's sequence_ which runs iteratees in the list one after the other.

Compare to Prelude.sequence_.

countConsumed :: (Monad m, Integral n) => Iteratee [el] m a -> Iteratee [el] m (a, n)

Transform an iteratee into one that keeps track of how much data it consumes.

Monadic functions

mapStreamM :: Monad m => (el -> m el') -> Enumeratee [el] [el'] m a

Maps a monadic function over an Iteratee.

mapStreamM_ :: Monad m => (el -> m b) -> Iteratee [el] m ()

Maps a monadic function over the elements of the stream and ignores the result.

foldStreamM :: Monad m => (b -> a -> m b) -> b -> Iteratee [a] m b

Folds a monadic function over an Iteratee.

Re-exported modules

module Bio.Iteratee.Iteratee