-- Hoogle documentation, generated by Haddock
-- See Hoogle, http://www.haskell.org/hoogle/


-- | streaming FASTA parser
--   
--   Stream-based handling of FASTA files. The user selects a window size,
--   the library then handles the window. For each window, the previous
--   (past) window is available, in case some data sits on the boundary
--   between windows.
--   
--   Greg Schwartz' <a>http://hackage.haskell.org/package/fasta</a> package
--   is a lot more complete. This one is mostly tailored to my usage
--   requirements (and may at some point use his library).
@package BiobaseFasta
@version 0.3.0.0


-- | Streaming Fasta handling via the <tt>streaming</tt> library.
--   
--   The functions in here should be streaming in constant memory.
--   
--   TODO Check if this is actually true with some unit tests.
module Biobase.Fasta.Streaming

-- | Control structure for <a>streamingFasta</a>.
data FindHeader
FindHeader :: [ByteString] -> !Int -> FindHeader

-- | the collected header parts (in reverse order)
[headerParts] :: FindHeader -> [ByteString]

-- | accumulated header length
[headerLength] :: FindHeader -> !Int
HasHeader :: !ByteString -> !ByteString -> [ByteString] -> !Int -> !Int -> FindHeader

-- | the (size-truncated) header for this fasta file
[fhHeader] :: FindHeader -> !ByteString

-- | overlap (if any) from earlier parts of the fasta file
[dataOverlap] :: FindHeader -> !ByteString

-- | collection of dataParts, in reverse order!
[dataParts] :: FindHeader -> [ByteString]

-- | total length of data parts, simplifies checking if enough data was
--   collected
[dataLength] :: FindHeader -> !Int

-- | count how many entries we have seen
[entries] :: FindHeader -> !Int
newtype CurrentSize
CurrentSize :: Int -> CurrentSize
newtype OverlapSize
OverlapSize :: Int -> OverlapSize
newtype HeaderSize
HeaderSize :: Int -> HeaderSize

-- | lens into the unique id / first word of the header.
fastaUid :: Lens' (SequenceIdentifier w) ByteString

-- | Fully stream a fasta file, making sure to never exceed a constant
--   amount of memory. The <tt>go</tt> function yields values of type
--   <tt>a</tt> down the line for continued streaming.
--   
--   <pre>
--   r4 = toList . streamingFasta (HeaderSize 2) (OverlapSize 1) (CurrentSize 2) go . S8.fromStrict $ BS.pack t0
--    where go (Header h) (Overlap o) (Current c) = yield (h,o,c)
--   </pre>
streamingFasta :: forall m w ty k r a. Monad m => HeaderSize -> OverlapSize -> CurrentSize -> ByteString m r -> Stream (Of (BioSequenceWindow w ty k)) m r
instance GHC.Show.Show Biobase.Fasta.Streaming.CurrentSize
instance GHC.Classes.Ord Biobase.Fasta.Streaming.CurrentSize
instance GHC.Classes.Eq Biobase.Fasta.Streaming.CurrentSize
instance GHC.Show.Show Biobase.Fasta.Streaming.OverlapSize
instance GHC.Classes.Ord Biobase.Fasta.Streaming.OverlapSize
instance GHC.Classes.Eq Biobase.Fasta.Streaming.OverlapSize
instance GHC.Show.Show Biobase.Fasta.Streaming.HeaderSize
instance GHC.Classes.Ord Biobase.Fasta.Streaming.HeaderSize
instance GHC.Classes.Eq Biobase.Fasta.Streaming.HeaderSize


-- | A convenience module for *small* <tt>Fasta</tt> entries, that are
--   completely in memory and *not* to be streamed.
--   
--   The <tt>Data.ByteString.Strict.Lens</tt> module is very helpful for
--   further handling of <a>Fasta</a> entries.
--   
--   For convenience, the <a>convertString</a> function from
--   <tt>string-conversions</tt> is supplied.
module Biobase.Fasta.Strict

-- | A *strict* <tt>Fasta</tt> entry.
data Fasta which ty
Fasta :: !SequenceIdentifier which -> !BioSequence ty -> Fasta which ty
[_header] :: Fasta which ty -> !SequenceIdentifier which
[_fasta] :: Fasta which ty -> !BioSequence ty

-- | If you don't want to deal with the phantom types.
type FastaUntyped = Fasta Void Void
fasta :: forall which_anQa ty_anQb ty_apOX. Lens (Fasta which_anQa ty_anQb) (Fasta which_anQa ty_apOX) (BioSequence ty_anQb) (BioSequence ty_apOX)
header :: forall which_anQa ty_anQb which_apOY. Lens (Fasta which_anQa ty_anQb) (Fasta which_apOY ty_anQb) (SequenceIdentifier which_anQa) (SequenceIdentifier which_apOY)

-- | Render a <a>Fasta</a> entry to a <a>ByteString</a>. Will end with a
--   final <tt>n</tt> in any case.
fastaToByteString :: Int -> Fasta which ty -> ByteString

-- | Render a <a>Fasta</a> entry to a <tt>Builder</tt>. Will end with a
--   final <tt>n</tt> in any case.
fastaToBuilder :: Int -> Fasta which ty -> Builder

-- | Try to parse a <a>ByteString</a> as a <a>Fasta</a>, failing with
--   <a>Left</a>, succees with <a>Right</a>.
byteStringToFasta :: ByteString -> Either String (Fasta which ty)

-- | Try to parse a <a>ByteString</a> as multiple <a>Fasta</a> entries.
--   Even though this is using the underlying streaming interface, this is
--   not streaming.
byteStringToMultiFasta :: ByteString -> [Fasta which ty]

-- | A lens that goes from a <a>BioSequenceWindow</a> to a <a>Fasta</a>.
windowedFasta :: Lens' (BioSequenceWindow w ty k) (Fasta w ty)

-- | A prism from a <a>ByteString</a> to a <a>Fasta</a>. Note that this
--   will only be an identity if the underlying fasta file is rendered with
--   <tt>k</tt> characters per line.
rawFasta :: Int -> Prism' ByteString (Fasta which ty)
convertString :: ConvertibleStrings a b => a -> b
instance forall k1 (which :: k1) k2 (ty :: k2). GHC.Generics.Generic (Biobase.Fasta.Strict.Fasta which ty)
instance forall k1 (which :: k1) k2 (ty :: k2). GHC.Show.Show (Biobase.Fasta.Strict.Fasta which ty)
instance forall k1 (which :: k1) k2 (ty :: k2). GHC.Read.Read (Biobase.Fasta.Strict.Fasta which ty)
instance forall k1 (which :: k1) k2 (ty :: k2). GHC.Classes.Ord (Biobase.Fasta.Strict.Fasta which ty)
instance forall k1 (which :: k1) k2 (ty :: k2). GHC.Classes.Eq (Biobase.Fasta.Strict.Fasta which ty)