Permalink
1089 lines (899 sloc) 35.3 KB
{-# LANGUAGE CPP, ForeignFunctionInterface, DeriveDataTypeable #-}
#if __GLASGOW_HASKELL__ >= 702
{-# LANGUAGE DeriveGeneric #-}
#endif
#if __GLASGOW_HASKELL__ >= 706
{-# LANGUAGE CApiFFI #-}
#endif
-----------------------------------------------------------------------------
-- |
-- Copyright : (c) 2006-2015 Duncan Coutts
-- License : BSD-style
--
-- Maintainer : duncan@community.haskell.org
--
-- Zlib wrapper layer
--
-----------------------------------------------------------------------------
module Codec.Compression.Zlib.Stream (
-- * The Zlib state monad
Stream,
State,
mkState,
runStream,
unsafeLiftIO,
finalise,
-- * Initialisation
deflateInit,
inflateInit,
-- ** Initialisation parameters
Format(..),
gzipFormat,
zlibFormat,
rawFormat,
gzipOrZlibFormat,
formatSupportsDictionary,
CompressionLevel(..),
defaultCompression,
noCompression,
bestSpeed,
bestCompression,
compressionLevel,
Method(..),
deflateMethod,
WindowBits(..),
defaultWindowBits,
windowBits,
MemoryLevel(..),
defaultMemoryLevel,
minMemoryLevel,
maxMemoryLevel,
memoryLevel,
CompressionStrategy(..),
defaultStrategy,
filteredStrategy,
huffmanOnlyStrategy,
-- * The buisness
deflate,
inflate,
Status(..),
Flush(..),
ErrorCode(..),
-- ** Special operations
inflateReset,
-- * Buffer management
-- ** Input buffer
pushInputBuffer,
inputBufferEmpty,
popRemainingInputBuffer,
-- ** Output buffer
pushOutputBuffer,
popOutputBuffer,
outputBufferBytesAvailable,
outputBufferSpaceRemaining,
outputBufferFull,
-- ** Dictionary
deflateSetDictionary,
inflateSetDictionary,
-- ** Dictionary hashes
DictionaryHash,
dictionaryHash,
zeroDictionaryHash,
#ifdef DEBUG
-- * Debugging
consistencyCheck,
dump,
trace,
#endif
) where
-- Note we don't use the MIN_VERSION_* macros here for compatability with
-- old Cabal versions that come with old GHC, that didn't provide these
-- macros for .hsc files. So we use __GLASGOW_HASKELL__ as a proxy.
import Foreign
( Word8, Ptr, nullPtr, plusPtr, peekByteOff, pokeByteOff
, ForeignPtr, FinalizerPtr, mallocForeignPtrBytes, addForeignPtrFinalizer
, withForeignPtr, touchForeignPtr, minusPtr )
#if __GLASGOW_HASKELL__ >= 702
import Foreign.ForeignPtr.Unsafe ( unsafeForeignPtrToPtr )
import System.IO.Unsafe ( unsafePerformIO )
#else
import Foreign ( unsafeForeignPtrToPtr, unsafePerformIO )
#endif
#ifdef __GLASGOW_HASKELL__
import Foreign
( finalizeForeignPtr )
#endif
import Foreign.C
import Data.ByteString.Internal (nullForeignPtr)
import qualified Data.ByteString.Unsafe as B
import Data.ByteString (ByteString)
#if !(__GLASGOW_HASKELL__ >= 710)
import Control.Applicative (Applicative(..))
#endif
import Control.Monad (ap,liftM)
#if __GLASGOW_HASKELL__ >= 702
#if __GLASGOW_HASKELL__ >= 708
import Control.Monad.ST.Strict
#else
import Control.Monad.ST.Strict hiding (unsafeIOToST)
#endif
import Control.Monad.ST.Unsafe
#else
import Control.Monad.ST.Strict
#endif
import Control.Exception (assert)
import Data.Typeable (Typeable)
#if __GLASGOW_HASKELL__ >= 702
import GHC.Generics (Generic)
#endif
#ifdef DEBUG
import System.IO (hPutStrLn, stderr)
#endif
import Prelude hiding (length)
#include "zlib.h"
pushInputBuffer :: ForeignPtr Word8 -> Int -> Int -> Stream ()
pushInputBuffer inBuf' offset length = do
-- must not push a new input buffer if the last one is not used up
inAvail <- getInAvail
assert (inAvail == 0) $ return ()
-- Now that we're setting a new input buffer, we can be sure that zlib no
-- longer has a reference to the old one. Therefore this is the last point
-- at which the old buffer had to be retained. It's safe to release now.
inBuf <- getInBuf
unsafeLiftIO $ touchForeignPtr inBuf
-- now set the available input buffer ptr and length
setInBuf inBuf'
setInAvail length
setInNext (unsafeForeignPtrToPtr inBuf' `plusPtr` offset)
-- Note the 'unsafe'. We are passing the raw ptr inside inBuf' to zlib.
-- To make this safe we need to hold on to the ForeignPtr for at least as
-- long as zlib is using the underlying raw ptr.
inputBufferEmpty :: Stream Bool
inputBufferEmpty = getInAvail >>= return . (==0)
popRemainingInputBuffer :: Stream (ForeignPtr Word8, Int, Int)
popRemainingInputBuffer = do
inBuf <- getInBuf
inNext <- getInNext
inAvail <- getInAvail
-- there really should be something to pop, otherwise it's silly
assert (inAvail > 0) $ return ()
setInAvail 0
return (inBuf, inNext `minusPtr` unsafeForeignPtrToPtr inBuf, inAvail)
pushOutputBuffer :: ForeignPtr Word8 -> Int -> Int -> Stream ()
pushOutputBuffer outBuf' offset length = do
--must not push a new buffer if there is still data in the old one
outAvail <- getOutAvail
assert (outAvail == 0) $ return ()
-- Note that there may still be free space in the output buffer, that's ok,
-- you might not want to bother completely filling the output buffer say if
-- there's only a few free bytes left.
outBuf <- getOutBuf
unsafeLiftIO $ touchForeignPtr outBuf
-- now set the available input buffer ptr and length
setOutBuf outBuf'
setOutFree length
setOutNext (unsafeForeignPtrToPtr outBuf' `plusPtr` offset)
setOutOffset offset
setOutAvail 0
-- get that part of the output buffer that is currently full
-- (might be 0, use outputBufferBytesAvailable to check)
-- this may leave some space remaining in the buffer, use
-- outputBufferSpaceRemaining to check.
popOutputBuffer :: Stream (ForeignPtr Word8, Int, Int)
popOutputBuffer = do
outBuf <- getOutBuf
outOffset <- getOutOffset
outAvail <- getOutAvail
-- there really should be something to pop, otherwise it's silly
assert (outAvail > 0) $ return ()
setOutOffset (outOffset + outAvail)
setOutAvail 0
return (outBuf, outOffset, outAvail)
-- this is the number of bytes available in the output buffer
outputBufferBytesAvailable :: Stream Int
outputBufferBytesAvailable = getOutAvail
-- you needen't get all the output immediately, you can continue until
-- there is no more output space available, this tells you that amount
outputBufferSpaceRemaining :: Stream Int
outputBufferSpaceRemaining = getOutFree
-- you only need to supply a new buffer when there is no more output buffer
-- space remaining
outputBufferFull :: Stream Bool
outputBufferFull = liftM (==0) outputBufferSpaceRemaining
-- you can only run this when the output buffer is not empty
-- you can run it when the input buffer is empty but it doesn't do anything
-- after running deflate either the output buffer will be full
-- or the input buffer will be empty (or both)
deflate :: Flush -> Stream Status
deflate flush = do
outFree <- getOutFree
-- deflate needs free space in the output buffer
assert (outFree > 0) $ return ()
result <- deflate_ flush
outFree' <- getOutFree
-- number of bytes of extra output there is available as a result of
-- the call to deflate:
let outExtra = outFree - outFree'
outAvail <- getOutAvail
setOutAvail (outAvail + outExtra)
return result
inflate :: Flush -> Stream Status
inflate flush = do
outFree <- getOutFree
-- inflate needs free space in the output buffer
assert (outFree > 0) $ return ()
result <- inflate_ flush
outFree' <- getOutFree
-- number of bytes of extra output there is available as a result of
-- the call to inflate:
let outExtra = outFree - outFree'
outAvail <- getOutAvail
setOutAvail (outAvail + outExtra)
return result
inflateReset :: Stream ()
inflateReset = do
outAvail <- getOutAvail
inAvail <- getInAvail
-- At the point where this is used, all the output should have been consumed
-- and any trailing input should be extracted and resupplied explicitly, not
-- just left.
assert (outAvail == 0 && inAvail == 0) $ return ()
err <- withStreamState $ \zstream ->
c_inflateReset zstream
failIfError err
deflateSetDictionary :: ByteString -> Stream Status
deflateSetDictionary dict = do
err <- withStreamState $ \zstream ->
B.unsafeUseAsCStringLen dict $ \(ptr, len) ->
c_deflateSetDictionary zstream ptr (fromIntegral len)
toStatus err
inflateSetDictionary :: ByteString -> Stream Status
inflateSetDictionary dict = do
err <- withStreamState $ \zstream -> do
B.unsafeUseAsCStringLen dict $ \(ptr, len) ->
c_inflateSetDictionary zstream ptr (fromIntegral len)
toStatus err
-- | A hash of a custom compression dictionary. These hashes are used by
-- zlib as dictionary identifiers.
-- (The particular hash function used is Adler32.)
--
newtype DictionaryHash = DictHash CULong
deriving (Eq, Ord, Read, Show)
-- | Update a running 'DictionaryHash'. You can generate a 'DictionaryHash'
-- from one or more 'ByteString's by starting from 'zeroDictionaryHash', e.g.
--
-- > dictionaryHash zeroDictionaryHash :: ByteString -> DictionaryHash
--
-- or
--
-- > foldl' dictionaryHash zeroDictionaryHash :: [ByteString] -> DictionaryHash
--
dictionaryHash :: DictionaryHash -> ByteString -> DictionaryHash
dictionaryHash (DictHash adler) dict =
unsafePerformIO $
B.unsafeUseAsCStringLen dict $ \(ptr, len) ->
liftM DictHash $ c_adler32 adler ptr (fromIntegral len)
-- | A zero 'DictionaryHash' to use as the initial value with 'dictionaryHash'.
--
zeroDictionaryHash :: DictionaryHash
zeroDictionaryHash = DictHash 0
----------------------------
-- Stream monad
--
newtype Stream a = Z {
unZ :: ForeignPtr StreamState
-> ForeignPtr Word8
-> ForeignPtr Word8
-> Int -> Int
-> IO (ForeignPtr Word8
,ForeignPtr Word8
,Int, Int, a)
}
instance Functor Stream where
fmap = liftM
instance Applicative Stream where
pure = returnZ
(<*>) = ap
(*>) = thenZ_
instance Monad Stream where
(>>=) = thenZ
-- m >>= f = (m `thenZ` \a -> consistencyCheck `thenZ_` returnZ a) `thenZ` f
(>>) = (*>)
return = pure
fail = (finalise >>) . failZ
returnZ :: a -> Stream a
returnZ a = Z $ \_ inBuf outBuf outOffset outLength ->
return (inBuf, outBuf, outOffset, outLength, a)
{-# INLINE returnZ #-}
thenZ :: Stream a -> (a -> Stream b) -> Stream b
thenZ (Z m) f =
Z $ \stream inBuf outBuf outOffset outLength ->
m stream inBuf outBuf outOffset outLength >>=
\(inBuf', outBuf', outOffset', outLength', a) ->
unZ (f a) stream inBuf' outBuf' outOffset' outLength'
{-# INLINE thenZ #-}
thenZ_ :: Stream a -> Stream b -> Stream b
thenZ_ (Z m) f =
Z $ \stream inBuf outBuf outOffset outLength ->
m stream inBuf outBuf outOffset outLength >>=
\(inBuf', outBuf', outOffset', outLength', _) ->
unZ f stream inBuf' outBuf' outOffset' outLength'
{-# INLINE thenZ_ #-}
failZ :: String -> Stream a
failZ msg = Z (\_ _ _ _ _ -> fail ("Codec.Compression.Zlib: " ++ msg))
data State s = State !(ForeignPtr StreamState)
!(ForeignPtr Word8)
!(ForeignPtr Word8)
{-# UNPACK #-} !Int
{-# UNPACK #-} !Int
mkState :: ST s (State s)
mkState = unsafeIOToST $ do
stream <- mallocForeignPtrBytes (#{const sizeof(z_stream)})
withForeignPtr stream $ \ptr -> do
#{poke z_stream, msg} ptr nullPtr
#{poke z_stream, zalloc} ptr nullPtr
#{poke z_stream, zfree} ptr nullPtr
#{poke z_stream, opaque} ptr nullPtr
#{poke z_stream, next_in} ptr nullPtr
#{poke z_stream, next_out} ptr nullPtr
#{poke z_stream, avail_in} ptr (0 :: CUInt)
#{poke z_stream, avail_out} ptr (0 :: CUInt)
return (State stream nullForeignPtr nullForeignPtr 0 0)
runStream :: Stream a -> State s -> ST s (a, State s)
runStream (Z m) (State stream inBuf outBuf outOffset outLength) =
unsafeIOToST $
m stream inBuf outBuf outOffset outLength >>=
\(inBuf', outBuf', outOffset', outLength', a) ->
return (a, State stream inBuf' outBuf' outOffset' outLength')
-- This is marked as unsafe because runStream uses unsafeIOToST so anything
-- lifted here can end up being unsafePerformIO'd.
unsafeLiftIO :: IO a -> Stream a
unsafeLiftIO m = Z $ \_stream inBuf outBuf outOffset outLength -> do
a <- m
return (inBuf, outBuf, outOffset, outLength, a)
getStreamState :: Stream (ForeignPtr StreamState)
getStreamState = Z $ \stream inBuf outBuf outOffset outLength -> do
return (inBuf, outBuf, outOffset, outLength, stream)
getInBuf :: Stream (ForeignPtr Word8)
getInBuf = Z $ \_stream inBuf outBuf outOffset outLength -> do
return (inBuf, outBuf, outOffset, outLength, inBuf)
getOutBuf :: Stream (ForeignPtr Word8)
getOutBuf = Z $ \_stream inBuf outBuf outOffset outLength -> do
return (inBuf, outBuf, outOffset, outLength, outBuf)
getOutOffset :: Stream Int
getOutOffset = Z $ \_stream inBuf outBuf outOffset outLength -> do
return (inBuf, outBuf, outOffset, outLength, outOffset)
getOutAvail :: Stream Int
getOutAvail = Z $ \_stream inBuf outBuf outOffset outLength -> do
return (inBuf, outBuf, outOffset, outLength, outLength)
setInBuf :: ForeignPtr Word8 -> Stream ()
setInBuf inBuf = Z $ \_stream _ outBuf outOffset outLength -> do
return (inBuf, outBuf, outOffset, outLength, ())
setOutBuf :: ForeignPtr Word8 -> Stream ()
setOutBuf outBuf = Z $ \_stream inBuf _ outOffset outLength -> do
return (inBuf, outBuf, outOffset, outLength, ())
setOutOffset :: Int -> Stream ()
setOutOffset outOffset = Z $ \_stream inBuf outBuf _ outLength -> do
return (inBuf, outBuf, outOffset, outLength, ())
setOutAvail :: Int -> Stream ()
setOutAvail outLength = Z $ \_stream inBuf outBuf outOffset _ -> do
return (inBuf, outBuf, outOffset, outLength, ())
----------------------------
-- Debug stuff
--
#ifdef DEBUG
trace :: String -> Stream ()
trace = unsafeLiftIO . hPutStrLn stderr
dump :: Stream ()
dump = do
inNext <- getInNext
inAvail <- getInAvail
outNext <- getOutNext
outFree <- getOutFree
outAvail <- getOutAvail
outOffset <- getOutOffset
unsafeLiftIO $ hPutStrLn stderr $
"Stream {\n" ++
" inNext = " ++ show inNext ++ ",\n" ++
" inAvail = " ++ show inAvail ++ ",\n" ++
"\n" ++
" outNext = " ++ show outNext ++ ",\n" ++
" outFree = " ++ show outFree ++ ",\n" ++
" outAvail = " ++ show outAvail ++ ",\n" ++
" outOffset = " ++ show outOffset ++ "\n" ++
"}"
consistencyCheck
consistencyCheck :: Stream ()
consistencyCheck = do
outBuf <- getOutBuf
outOffset <- getOutOffset
outAvail <- getOutAvail
outNext <- getOutNext
let outBufPtr = unsafeForeignPtrToPtr outBuf
assert (outBufPtr `plusPtr` (outOffset + outAvail) == outNext) $ return ()
#endif
----------------------------
-- zlib wrapper layer
--
data Status =
Ok
| StreamEnd
| Error ErrorCode String
data ErrorCode =
NeedDict DictionaryHash
| FileError
| StreamError
| DataError
| MemoryError
| BufferError -- ^ No progress was possible or there was not enough room in
-- the output buffer when 'Finish' is used. Note that
-- 'BuferError' is not fatal, and 'inflate' can be called
-- again with more input and more output space to continue.
| VersionError
| Unexpected
toStatus :: CInt -> Stream Status
toStatus errno = case errno of
(#{const Z_OK}) -> return Ok
(#{const Z_STREAM_END}) -> return StreamEnd
(#{const Z_NEED_DICT}) -> do
adler <- withStreamPtr (#{peek z_stream, adler})
err (NeedDict (DictHash adler)) "custom dictionary needed"
(#{const Z_BUF_ERROR}) -> err BufferError "buffer error"
(#{const Z_ERRNO}) -> err FileError "file error"
(#{const Z_STREAM_ERROR}) -> err StreamError "stream error"
(#{const Z_DATA_ERROR}) -> err DataError "data error"
(#{const Z_MEM_ERROR}) -> err MemoryError "insufficient memory"
(#{const Z_VERSION_ERROR}) -> err VersionError "incompatible zlib version"
other -> return $ Error Unexpected
("unexpected zlib status: " ++ show other)
where
err errCode altMsg = liftM (Error errCode) $ do
msgPtr <- withStreamPtr (#{peek z_stream, msg})
if msgPtr /= nullPtr
then unsafeLiftIO (peekCAString msgPtr)
else return altMsg
failIfError :: CInt -> Stream ()
failIfError errno = toStatus errno >>= \status -> case status of
(Error _ msg) -> fail msg
_ -> return ()
data Flush =
NoFlush
| SyncFlush
| FullFlush
| Finish
-- | Block -- only available in zlib 1.2 and later, uncomment if you need it.
fromFlush :: Flush -> CInt
fromFlush NoFlush = #{const Z_NO_FLUSH}
fromFlush SyncFlush = #{const Z_SYNC_FLUSH}
fromFlush FullFlush = #{const Z_FULL_FLUSH}
fromFlush Finish = #{const Z_FINISH}
-- fromFlush Block = #{const Z_BLOCK}
-- | The format used for compression or decompression. There are three
-- variations.
--
data Format = GZip | Zlib | Raw | GZipOrZlib
deriving (Eq, Ord, Enum, Bounded, Show, Typeable
#if __GLASGOW_HASKELL__ >= 702
, Generic
#endif
)
{-# DEPRECATED GZip "Use gzipFormat. Format constructors will be hidden in version 0.7" #-}
{-# DEPRECATED Zlib "Use zlibFormat. Format constructors will be hidden in version 0.7" #-}
{-# DEPRECATED Raw "Use rawFormat. Format constructors will be hidden in version 0.7" #-}
{-# DEPRECATED GZipOrZlib "Use gzipOrZlibFormat. Format constructors will be hidden in version 0.7" #-}
-- | The gzip format uses a header with a checksum and some optional meta-data
-- about the compressed file. It is intended primarily for compressing
-- individual files but is also sometimes used for network protocols such as
-- HTTP. The format is described in detail in RFC #1952
-- <http://www.ietf.org/rfc/rfc1952.txt>
--
gzipFormat :: Format
gzipFormat = GZip
-- | The zlib format uses a minimal header with a checksum but no other
-- meta-data. It is especially designed for use in network protocols. The
-- format is described in detail in RFC #1950
-- <http://www.ietf.org/rfc/rfc1950.txt>
--
zlibFormat :: Format
zlibFormat = Zlib
-- | The \'raw\' format is just the compressed data stream without any
-- additional header, meta-data or data-integrity checksum. The format is
-- described in detail in RFC #1951 <http://www.ietf.org/rfc/rfc1951.txt>
--
rawFormat :: Format
rawFormat = Raw
-- | This is not a format as such. It enabled zlib or gzip decoding with
-- automatic header detection. This only makes sense for decompression.
--
gzipOrZlibFormat :: Format
gzipOrZlibFormat = GZipOrZlib
formatSupportsDictionary :: Format -> Bool
formatSupportsDictionary Zlib = True
formatSupportsDictionary Raw = True
formatSupportsDictionary _ = False
-- | The compression method
--
data Method = Deflated
deriving (Eq, Ord, Enum, Bounded, Show, Typeable
#if __GLASGOW_HASKELL__ >= 702
, Generic
#endif
)
{-# DEPRECATED Deflated "Use deflateMethod. Method constructors will be hidden in version 0.7" #-}
-- | \'Deflate\' is the only method supported in this version of zlib.
-- Indeed it is likely to be the only method that ever will be supported.
--
deflateMethod :: Method
deflateMethod = Deflated
fromMethod :: Method -> CInt
fromMethod Deflated = #{const Z_DEFLATED}
-- | The compression level parameter controls the amount of compression. This
-- is a trade-off between the amount of compression and the time required to do
-- the compression.
--
data CompressionLevel =
DefaultCompression
| NoCompression
| BestSpeed
| BestCompression
| CompressionLevel Int
deriving (Eq, Show, Typeable
#if __GLASGOW_HASKELL__ >= 702
, Generic
#endif
)
{-# DEPRECATED DefaultCompression "Use defaultCompression. CompressionLevel constructors will be hidden in version 0.7" #-}
{-# DEPRECATED NoCompression "Use noCompression. CompressionLevel constructors will be hidden in version 0.7" #-}
{-# DEPRECATED BestSpeed "Use bestSpeed. CompressionLevel constructors will be hidden in version 0.7" #-}
{-# DEPRECATED BestCompression "Use bestCompression. CompressionLevel constructors will be hidden in version 0.7" #-}
--FIXME: cannot deprecate constructor named the same as the type
{- DEPRECATED CompressionLevel "Use compressionLevel. CompressionLevel constructors will be hidden in version 0.7" -}
-- | The default compression level is 6 (that is, biased towards higher
-- compression at expense of speed).
defaultCompression :: CompressionLevel
defaultCompression = DefaultCompression
-- | No compression, just a block copy.
noCompression :: CompressionLevel
noCompression = CompressionLevel 0
-- | The fastest compression method (less compression)
bestSpeed :: CompressionLevel
bestSpeed = CompressionLevel 1
-- | The slowest compression method (best compression).
bestCompression :: CompressionLevel
bestCompression = CompressionLevel 9
-- | A specific compression level between 0 and 9.
compressionLevel :: Int -> CompressionLevel
compressionLevel n
| n >= 0 && n <= 9 = CompressionLevel n
| otherwise = error "CompressionLevel must be in the range 0..9"
fromCompressionLevel :: CompressionLevel -> CInt
fromCompressionLevel DefaultCompression = -1
fromCompressionLevel NoCompression = 0
fromCompressionLevel BestSpeed = 1
fromCompressionLevel BestCompression = 9
fromCompressionLevel (CompressionLevel n)
| n >= 0 && n <= 9 = fromIntegral n
| otherwise = error "CompressLevel must be in the range 1..9"
-- | This specifies the size of the compression window. Larger values of this
-- parameter result in better compression at the expense of higher memory
-- usage.
--
-- The compression window size is the value of the the window bits raised to
-- the power 2. The window bits must be in the range @9..15@ which corresponds
-- to compression window sizes of 512b to 32Kb. The default is 15 which is also
-- the maximum size.
--
-- The total amount of memory used depends on the window bits and the
-- 'MemoryLevel'. See the 'MemoryLevel' for the details.
--
data WindowBits = WindowBits Int
| DefaultWindowBits -- This constructor must be last to make
-- the Ord instance work. The Ord instance
-- is used by the tests.
-- It makse sense because the default value
-- is is also the max value at 15.
deriving (Eq, Ord, Show, Typeable
#if __GLASGOW_HASKELL__ >= 702
, Generic
#endif
)
{-# DEPRECATED DefaultWindowBits "Use defaultWindowBits. WindowBits constructors will be hidden in version 0.7" #-}
--FIXME: cannot deprecate constructor named the same as the type
{- DEPRECATED WindowBits "Use windowBits. WindowBits constructors will be hidden in version 0.7" -}
-- | The default 'WindowBits' is 15 which is also the maximum size.
--
defaultWindowBits :: WindowBits
defaultWindowBits = WindowBits 15
-- | A specific compression window size, specified in bits in the range @9..15@
--
windowBits :: Int -> WindowBits
windowBits n
| n >= 9 && n <= 15 = WindowBits n
| otherwise = error "WindowBits must be in the range 9..15"
fromWindowBits :: Format -> WindowBits-> CInt
fromWindowBits format bits = (formatModifier format) (checkWindowBits bits)
where checkWindowBits DefaultWindowBits = 15
checkWindowBits (WindowBits n)
| n >= 9 && n <= 15 = fromIntegral n
| otherwise = error "WindowBits must be in the range 9..15"
formatModifier Zlib = id
formatModifier GZip = (+16)
formatModifier GZipOrZlib = (+32)
formatModifier Raw = negate
-- | The 'MemoryLevel' parameter specifies how much memory should be allocated
-- for the internal compression state. It is a tradoff between memory usage,
-- compression ratio and compression speed. Using more memory allows faster
-- compression and a better compression ratio.
--
-- The total amount of memory used for compression depends on the 'WindowBits'
-- and the 'MemoryLevel'. For decompression it depends only on the
-- 'WindowBits'. The totals are given by the functions:
--
-- > compressTotal windowBits memLevel = 4 * 2^windowBits + 512 * 2^memLevel
-- > decompressTotal windowBits = 2^windowBits
--
-- For example, for compression with the default @windowBits = 15@ and
-- @memLevel = 8@ uses @256Kb@. So for example a network server with 100
-- concurrent compressed streams would use @25Mb@. The memory per stream can be
-- halved (at the cost of somewhat degraded and slower compressionby) by
-- reducing the @windowBits@ and @memLevel@ by one.
--
-- Decompression takes less memory, the default @windowBits = 15@ corresponds
-- to just @32Kb@.
--
data MemoryLevel =
DefaultMemoryLevel
| MinMemoryLevel
| MaxMemoryLevel
| MemoryLevel Int
deriving (Eq, Show, Typeable
#if __GLASGOW_HASKELL__ >= 702
, Generic
#endif
)
{-# DEPRECATED DefaultMemoryLevel "Use defaultMemoryLevel. MemoryLevel constructors will be hidden in version 0.7" #-}
{-# DEPRECATED MinMemoryLevel "Use minMemoryLevel. MemoryLevel constructors will be hidden in version 0.7" #-}
{-# DEPRECATED MaxMemoryLevel "Use maxMemoryLevel. MemoryLevel constructors will be hidden in version 0.7" #-}
--FIXME: cannot deprecate constructor named the same as the type
{- DEPRECATED MemoryLevel "Use memoryLevel. MemoryLevel constructors will be hidden in version 0.7" -}
-- | The default memory level. (Equivalent to @'memoryLevel' 8@)
--
defaultMemoryLevel :: MemoryLevel
defaultMemoryLevel = MemoryLevel 8
-- | Use minimum memory. This is slow and reduces the compression ratio.
-- (Equivalent to @'memoryLevel' 1@)
--
minMemoryLevel :: MemoryLevel
minMemoryLevel = MemoryLevel 1
-- | Use maximum memory for optimal compression speed.
-- (Equivalent to @'memoryLevel' 9@)
--
maxMemoryLevel :: MemoryLevel
maxMemoryLevel = MemoryLevel 9
-- | A specific level in the range @1..9@
--
memoryLevel :: Int -> MemoryLevel
memoryLevel n
| n >= 1 && n <= 9 = MemoryLevel n
| otherwise = error "MemoryLevel must be in the range 1..9"
fromMemoryLevel :: MemoryLevel -> CInt
fromMemoryLevel DefaultMemoryLevel = 8
fromMemoryLevel MinMemoryLevel = 1
fromMemoryLevel MaxMemoryLevel = 9
fromMemoryLevel (MemoryLevel n)
| n >= 1 && n <= 9 = fromIntegral n
| otherwise = error "MemoryLevel must be in the range 1..9"
-- | The strategy parameter is used to tune the compression algorithm.
--
-- The strategy parameter only affects the compression ratio but not the
-- correctness of the compressed output even if it is not set appropriately.
--
data CompressionStrategy =
DefaultStrategy
| Filtered
| HuffmanOnly
deriving (Eq, Ord, Enum, Bounded, Show, Typeable
#if __GLASGOW_HASKELL__ >= 702
, Generic
#endif
)
{-
-- -- only available in zlib 1.2 and later, uncomment if you need it.
| RLE -- ^ Use 'RLE' to limit match distances to one (run-length
-- encoding). 'RLE' is designed to be almost as fast as
-- 'HuffmanOnly', but give better compression for PNG
-- image data.
| Fixed -- ^ 'Fixed' prevents the use of dynamic Huffman codes,
-- allowing for a simpler decoder for special applications.
-}
{-# DEPRECATED DefaultStrategy "Use defaultStrategy. CompressionStrategy constructors will be hidden in version 0.7" #-}
{-# DEPRECATED Filtered "Use filteredStrategy. CompressionStrategy constructors will be hidden in version 0.7" #-}
{-# DEPRECATED HuffmanOnly "Use huffmanOnlyStrategy. CompressionStrategy constructors will be hidden in version 0.7" #-}
-- | Use this default compression strategy for normal data.
--
defaultStrategy :: CompressionStrategy
defaultStrategy = DefaultStrategy
-- | Use the filtered compression strategy for data produced by a filter (or
-- predictor). Filtered data consists mostly of small values with a somewhat
-- random distribution. In this case, the compression algorithm is tuned to
-- compress them better. The effect of this strategy is to force more Huffman
-- coding and less string matching; it is somewhat intermediate between
-- 'defaultCompressionStrategy' and 'huffmanOnlyCompressionStrategy'.
--
filteredStrategy :: CompressionStrategy
filteredStrategy = Filtered
-- | Use the Huffman-only compression strategy to force Huffman encoding only
-- (no string match).
--
huffmanOnlyStrategy :: CompressionStrategy
huffmanOnlyStrategy = HuffmanOnly
fromCompressionStrategy :: CompressionStrategy -> CInt
fromCompressionStrategy DefaultStrategy = #{const Z_DEFAULT_STRATEGY}
fromCompressionStrategy Filtered = #{const Z_FILTERED}
fromCompressionStrategy HuffmanOnly = #{const Z_HUFFMAN_ONLY}
--fromCompressionStrategy RLE = #{const Z_RLE}
--fromCompressionStrategy Fixed = #{const Z_FIXED}
withStreamPtr :: (Ptr StreamState -> IO a) -> Stream a
withStreamPtr f = do
stream <- getStreamState
unsafeLiftIO (withForeignPtr stream f)
withStreamState :: (StreamState -> IO a) -> Stream a
withStreamState f = do
stream <- getStreamState
unsafeLiftIO (withForeignPtr stream (f . StreamState))
setInAvail :: Int -> Stream ()
setInAvail val = withStreamPtr $ \ptr ->
#{poke z_stream, avail_in} ptr (fromIntegral val :: CUInt)
getInAvail :: Stream Int
getInAvail = liftM (fromIntegral :: CUInt -> Int) $
withStreamPtr (#{peek z_stream, avail_in})
setInNext :: Ptr Word8 -> Stream ()
setInNext val = withStreamPtr (\ptr -> #{poke z_stream, next_in} ptr val)
getInNext :: Stream (Ptr Word8)
getInNext = withStreamPtr (#{peek z_stream, next_in})
setOutFree :: Int -> Stream ()
setOutFree val = withStreamPtr $ \ptr ->
#{poke z_stream, avail_out} ptr (fromIntegral val :: CUInt)
getOutFree :: Stream Int
getOutFree = liftM (fromIntegral :: CUInt -> Int) $
withStreamPtr (#{peek z_stream, avail_out})
setOutNext :: Ptr Word8 -> Stream ()
setOutNext val = withStreamPtr (\ptr -> #{poke z_stream, next_out} ptr val)
#ifdef DEBUG
getOutNext :: Stream (Ptr Word8)
getOutNext = withStreamPtr (#{peek z_stream, next_out})
#endif
inflateInit :: Format -> WindowBits -> Stream ()
inflateInit format bits = do
checkFormatSupported format
err <- withStreamState $ \zstream ->
c_inflateInit2 zstream (fromIntegral (fromWindowBits format bits))
failIfError err
getStreamState >>= unsafeLiftIO . addForeignPtrFinalizer c_inflateEnd
deflateInit :: Format
-> CompressionLevel
-> Method
-> WindowBits
-> MemoryLevel
-> CompressionStrategy
-> Stream ()
deflateInit format compLevel method bits memLevel strategy = do
checkFormatSupported format
err <- withStreamState $ \zstream ->
c_deflateInit2 zstream
(fromCompressionLevel compLevel)
(fromMethod method)
(fromWindowBits format bits)
(fromMemoryLevel memLevel)
(fromCompressionStrategy strategy)
failIfError err
getStreamState >>= unsafeLiftIO . addForeignPtrFinalizer c_deflateEnd
inflate_ :: Flush -> Stream Status
inflate_ flush = do
err <- withStreamState $ \zstream ->
c_inflate zstream (fromFlush flush)
toStatus err
deflate_ :: Flush -> Stream Status
deflate_ flush = do
err <- withStreamState $ \zstream ->
c_deflate zstream (fromFlush flush)
toStatus err
-- | This never needs to be used as the stream's resources will be released
-- automatically when no longer needed, however this can be used to release
-- them early. Only use this when you can guarantee that the stream will no
-- longer be needed, for example if an error occurs or if the stream ends.
--
finalise :: Stream ()
#ifdef __GLASGOW_HASKELL__
--TODO: finalizeForeignPtr is ghc-only
finalise = getStreamState >>= unsafeLiftIO . finalizeForeignPtr
#else
finalise = return ()
#endif
checkFormatSupported :: Format -> Stream ()
checkFormatSupported format = do
version <- unsafeLiftIO (peekCAString =<< c_zlibVersion)
case version of
('1':'.':'1':'.':_)
| format == GZip
|| format == GZipOrZlib
-> fail $ "version 1.1.x of the zlib C library does not support the"
++ " 'gzip' format via the in-memory api, only the 'raw' and "
++ " 'zlib' formats."
_ -> return ()
----------------------
-- The foreign imports
newtype StreamState = StreamState (Ptr StreamState)
-- inflateInit2 and deflateInit2 are actually defined as macros in zlib.h
-- They are defined in terms of inflateInit2_ and deflateInit2_ passing two
-- additional arguments used to detect compatability problems. They pass the
-- version of zlib as a char * and the size of the z_stream struct.
-- If we compile via C then we can avoid this hassle however thats not really
-- kosher since the Haskell FFI is defined at the C ABI level, not the C
-- language level. There is no requirement to compile via C and pick up C
-- headers. So it's much better if we can make it work properly and that'd
-- also allow compiling via ghc's ncg which is a good thing since the C
-- backend is not going to be around forever.
--
-- So we define c_inflateInit2 and c_deflateInit2 here as wrappers around
-- their _ counterparts and pass the extra args.
--
-- As of GHC 7.6, we can import macros directly using the CApiFFI extension.
-- This avoids the need for the hsc2hs #{const_str} construct, which means
-- hsc2hs can run in cross-compilation mode.
##ifdef NON_BLOCKING_FFI
##define SAFTY safe
##else
##define SAFTY unsafe
##endif
#if __GLASGOW_HASKELL__ >= 706
foreign import capi unsafe "zlib.h inflateInit2"
c_inflateInit2 :: StreamState -> CInt -> IO CInt
foreign import capi unsafe "zlib.h deflateInit2"
c_deflateInit2 :: StreamState
-> CInt -> CInt -> CInt -> CInt -> CInt -> IO CInt
#else
foreign import ccall unsafe "zlib.h inflateInit2_"
c_inflateInit2_ :: StreamState -> CInt -> Ptr CChar -> CInt -> IO CInt
c_inflateInit2 :: StreamState -> CInt -> IO CInt
c_inflateInit2 z n =
withCAString #{const_str ZLIB_VERSION} $ \versionStr ->
c_inflateInit2_ z n versionStr (#{const sizeof(z_stream)} :: CInt)
foreign import ccall unsafe "zlib.h deflateInit2_"
c_deflateInit2_ :: StreamState
-> CInt -> CInt -> CInt -> CInt -> CInt
-> Ptr CChar -> CInt
-> IO CInt
c_deflateInit2 :: StreamState
-> CInt -> CInt -> CInt -> CInt -> CInt -> IO CInt
c_deflateInit2 z a b c d e =
withCAString #{const_str ZLIB_VERSION} $ \versionStr ->
c_deflateInit2_ z a b c d e versionStr (#{const sizeof(z_stream)} :: CInt)
#endif
foreign import ccall SAFTY "zlib.h inflate"
c_inflate :: StreamState -> CInt -> IO CInt
foreign import ccall unsafe "zlib.h &inflateEnd"
c_inflateEnd :: FinalizerPtr StreamState
foreign import ccall unsafe "zlib.h inflateReset"
c_inflateReset :: StreamState -> IO CInt
foreign import ccall unsafe "zlib.h deflateSetDictionary"
c_deflateSetDictionary :: StreamState
-> Ptr CChar
-> CUInt
-> IO CInt
foreign import ccall unsafe "zlib.h inflateSetDictionary"
c_inflateSetDictionary :: StreamState
-> Ptr CChar
-> CUInt
-> IO CInt
foreign import ccall SAFTY "zlib.h deflate"
c_deflate :: StreamState -> CInt -> IO CInt
foreign import ccall unsafe "zlib.h &deflateEnd"
c_deflateEnd :: FinalizerPtr StreamState
foreign import ccall unsafe "zlib.h zlibVersion"
c_zlibVersion :: IO CString
foreign import ccall unsafe "zlib.h adler32"
c_adler32 :: CULong
-> Ptr CChar
-> CUInt
-> IO CULong