-- |
-- Module      : Streamly.Internal.Unicode.Stream
-- Copyright   : (c) 2018 Composewell Technologies
--               (c) Bjoern Hoehrmann 2008-2009
--
-- License     : BSD-3-Clause
-- Maintainer  : streamly@composewell.com
-- Stability   : experimental
-- Portability : GHC
--

module Streamly.Internal.Unicode.Stream
    (
    -- * Construction (Decoding)
      decodeLatin1

    -- ** UTF-8 Decoding
    , decodeUtf8
    , decodeUtf8'
    , decodeUtf8_

    -- ** Resumable UTF-8 Decoding
    , DecodeError(..)
    , DecodeState
    , CodePoint
    , decodeUtf8Either
    , resumeDecodeUtf8Either

    -- ** UTF-8 Array Stream Decoding
    , decodeUtf8Arrays
    , decodeUtf8Arrays'
    , decodeUtf8Arrays_

    -- * Elimination (Encoding)
    -- ** Latin1 Encoding
    , encodeLatin1
    , encodeLatin1'
    , encodeLatin1_

    -- ** UTF-8 Encoding
    , encodeUtf8
    , encodeUtf8'
    , encodeUtf8_
    , encodeStrings
    {-
    -- * Operations on character strings
    , strip -- (dropAround isSpace)
    , stripEnd
    -}

    -- * Transformation
    , stripHead
    , lines
    , words
    , unlines
    , unwords

    -- * StreamD UTF8 Encoding / Decoding transformations.
    , decodeUtf8D
    , decodeUtf8D'
    , decodeUtf8D_
    , encodeUtf8D
    , encodeUtf8D'
    , encodeUtf8D_
    , decodeUtf8EitherD
    , resumeDecodeUtf8EitherD
    , decodeUtf8ArraysD
    , decodeUtf8ArraysD'
    , decodeUtf8ArraysD_

    -- * Deprecations
    , decodeUtf8Lax
    , encodeLatin1Lax
    , encodeUtf8Lax
    )
where

#include "inline.hs"

import Control.Monad.IO.Class (MonadIO, liftIO)
import Data.Bits (shiftR, shiftL, (.|.), (.&.))
import Data.Char (chr, ord)
import Data.Word (Word8)
import Foreign.ForeignPtr (touchForeignPtr)
import Foreign.ForeignPtr.Unsafe (unsafeForeignPtrToPtr)
import Foreign.Storable (Storable(..))
import Fusion.Plugin.Types (Fuse(..))
import GHC.Base (assert, unsafeChr)
import GHC.ForeignPtr (ForeignPtr (..))
import GHC.IO.Encoding.Failure (isSurrogate)
import GHC.Ptr (Ptr (..), plusPtr)
import System.IO.Unsafe (unsafePerformIO)
import Streamly.Data.Fold (Fold)
import Streamly.Data.Array.Foreign (Array)
import Streamly.Internal.Data.Unfold (Unfold)
import Streamly.Internal.Data.SVar (adaptState)
import Streamly.Internal.Data.Stream.IsStream (IsStream)
import Streamly.Internal.Data.Stream.Serial (SerialT)
import Streamly.Internal.Data.Stream.StreamD (Stream(..), Step (..))
import Streamly.Internal.Data.Tuple.Strict (Tuple'(..))

import qualified Streamly.Internal.Data.Unfold as Unfold
import qualified Streamly.Internal.Data.Stream.Serial as Serial
import qualified Streamly.Internal.Data.Array.Foreign as Array
import qualified Streamly.Internal.Data.Array.Foreign.Type as A
import qualified Streamly.Internal.Data.Stream.IsStream as S
import qualified Streamly.Internal.Data.Stream.StreamD as D

import Prelude hiding (lines, words, unlines, unwords)

-- $setup
-- >>> :m
-- >>> import Prelude hiding (lines, words, unlines, unwords)
-- >>> import qualified Streamly.Prelude as Stream
-- >>> import qualified Streamly.Data.Fold as Fold
-- >>> import Streamly.Internal.Unicode.Stream

-------------------------------------------------------------------------------
-- Latin1 decoding
-------------------------------------------------------------------------------

-- | Decode a stream of bytes to Unicode characters by mapping each byte to a
-- corresponding Unicode 'Char' in 0-255 range.
--
-- /Since: 0.7.0 ("Streamly.Data.Unicode.Stream")/
--
-- @since 0.8.0
{-# INLINE decodeLatin1 #-}
decodeLatin1 :: (IsStream t, Monad m) => t m Word8 -> t m Char
decodeLatin1 :: t m Word8 -> t m Char
decodeLatin1 = (Word8 -> Char) -> t m Word8 -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
(a -> b) -> t m a -> t m b
S.map (Int -> Char
unsafeChr (Int -> Char) -> (Word8 -> Int) -> Word8 -> Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral)

-------------------------------------------------------------------------------
-- Latin1 encoding
-------------------------------------------------------------------------------

-- | Encode a stream of Unicode characters to bytes by mapping each character
-- to a byte in 0-255 range. Throws an error if the input stream contains
-- characters beyond 255.
--
-- @since 0.8.0
{-# INLINE encodeLatin1' #-}
encodeLatin1' :: (IsStream t, Monad m) => t m Char -> t m Word8
encodeLatin1' :: t m Char -> t m Word8
encodeLatin1' = (Char -> Word8) -> t m Char -> t m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
(a -> b) -> t m a -> t m b
S.map Char -> Word8
forall p. Num p => Char -> p
convert
    where
    convert :: Char -> p
convert Char
c =
        let codepoint :: Int
codepoint = Char -> Int
ord Char
c
        in if Int
codepoint Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
> Int
255
           then [Char] -> p
forall a. HasCallStack => [Char] -> a
error ([Char] -> p) -> [Char] -> p
forall a b. (a -> b) -> a -> b
$ [Char]
"Streamly.Unicode.encodeLatin1 invalid " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++
                      [Char]
"input char codepoint " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Int -> [Char]
forall a. Show a => a -> [Char]
show Int
codepoint
           else Int -> p
forall a b. (Integral a, Num b) => a -> b
fromIntegral Int
codepoint

-- XXX Should we instead replace the invalid chars by NUL or whitespace or some
-- other control char? That may affect the perf a bit but may be a better
-- behavior.
--
-- | Like 'encodeLatin1'' but silently maps input codepoints beyond 255 to
-- arbitrary Latin1 chars in 0-255 range. No error or exception is thrown when
-- such mapping occurs.
--
-- /Since: 0.7.0 ("Streamly.Data.Unicode.Stream")/
--
-- /Since: 0.8.0 (Lenient Behaviour)/
{-# INLINE encodeLatin1 #-}
encodeLatin1 :: (IsStream t, Monad m) => t m Char -> t m Word8
encodeLatin1 :: t m Char -> t m Word8
encodeLatin1 = (Char -> Word8) -> t m Char -> t m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
(a -> b) -> t m a -> t m b
S.map (Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> (Char -> Int) -> Char -> Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Char -> Int
ord)

-- | Like 'encodeLatin1' but drops the input characters beyond 255.
--
-- @since 0.8.0
{-# INLINE encodeLatin1_ #-}
encodeLatin1_ :: (IsStream t, Monad m) => t m Char -> t m Word8
encodeLatin1_ :: t m Char -> t m Word8
encodeLatin1_ = (Char -> Word8) -> t m Char -> t m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
(a -> b) -> t m a -> t m b
S.map (Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> (Char -> Int) -> Char -> Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Char -> Int
ord) (t m Char -> t m Word8)
-> (t m Char -> t m Char) -> t m Char -> t m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. (Char -> Bool) -> t m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
(a -> Bool) -> t m a -> t m a
S.filter (Char -> Char -> Bool
forall a. Ord a => a -> a -> Bool
<= Int -> Char
chr Int
255)

-- | Same as 'encodeLatin1'
--
{-# DEPRECATED encodeLatin1Lax "Please use 'encodeLatin1' instead" #-}
{-# INLINE encodeLatin1Lax #-}
encodeLatin1Lax :: (IsStream t, Monad m) => t m Char -> t m Word8
encodeLatin1Lax :: t m Char -> t m Word8
encodeLatin1Lax = t m Char -> t m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *).
(IsStream t, Monad m) =>
t m Char -> t m Word8
encodeLatin1

-------------------------------------------------------------------------------
-- UTF-8 decoding
-------------------------------------------------------------------------------

-- Int helps in cheaper conversion from Int to Char
type CodePoint = Int
type DecodeState = Word8

-- We can divide the errors in three general categories:
-- * A non-starter was encountered in a begin state
-- * A starter was encountered without completing a codepoint
-- * The last codepoint was not complete (input underflow)
--
-- Need to separate resumable and non-resumable error. In case of non-resumable
-- error we can also provide the failing byte. In case of resumable error the
-- state can be opaque.
--
data DecodeError = DecodeError !DecodeState !CodePoint deriving Int -> DecodeError -> [Char] -> [Char]
[DecodeError] -> [Char] -> [Char]
DecodeError -> [Char]
(Int -> DecodeError -> [Char] -> [Char])
-> (DecodeError -> [Char])
-> ([DecodeError] -> [Char] -> [Char])
-> Show DecodeError
forall a.
(Int -> a -> [Char] -> [Char])
-> (a -> [Char]) -> ([a] -> [Char] -> [Char]) -> Show a
showList :: [DecodeError] -> [Char] -> [Char]
$cshowList :: [DecodeError] -> [Char] -> [Char]
show :: DecodeError -> [Char]
$cshow :: DecodeError -> [Char]
showsPrec :: Int -> DecodeError -> [Char] -> [Char]
$cshowsPrec :: Int -> DecodeError -> [Char] -> [Char]
Show

-- See http://bjoern.hoehrmann.de/utf-8/decoder/dfa/ for details.

-- XXX Use names decodeSuccess = 0, decodeFailure = 12

decodeTable :: [Word8]
decodeTable :: [Word8]
decodeTable = [
   -- The first part of the table maps bytes to character classes that
   -- to reduce the size of the transition table and create bitmasks.
   Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,  Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,
   Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,  Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,
   Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,  Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,
   Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,  Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,Word8
0,
   Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,Word8
1,  Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,Word8
9,
   Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,  Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,Word8
7,
   Word8
8,Word8
8,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,  Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,Word8
2,
  Word8
10,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
3,Word8
4,Word8
3,Word8
3, Word8
11,Word8
6,Word8
6,Word8
6,Word8
5,Word8
8,Word8
8,Word8
8,Word8
8,Word8
8,Word8
8,Word8
8,Word8
8,Word8
8,Word8
8,Word8
8,

   -- The second part is a transition table that maps a combination
   -- of a state of the automaton and a character class to a state.
   Word8
0,Word8
12,Word8
24,Word8
36,Word8
60,Word8
96,Word8
84,Word8
12,Word8
12,Word8
12,Word8
48,Word8
72, Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,
  Word8
12, Word8
0,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12, Word8
0,Word8
12, Word8
0,Word8
12,Word8
12, Word8
12,Word8
24,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
24,Word8
12,Word8
24,Word8
12,Word8
12,
  Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
24,Word8
12,Word8
12,Word8
12,Word8
12, Word8
12,Word8
24,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
24,Word8
12,Word8
12,
  Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
36,Word8
12,Word8
36,Word8
12,Word8
12, Word8
12,Word8
36,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
36,Word8
12,Word8
36,Word8
12,Word8
12,
  Word8
12,Word8
36,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12,Word8
12
  ]

{-# NOINLINE utf8d #-}
utf8d :: A.Array Word8
utf8d :: Array Word8
utf8d =
      IO (Array Word8) -> Array Word8
forall a. IO a -> a
unsafePerformIO
    -- Aligning to cacheline makes a barely noticeable difference
    -- XXX currently alignment is not implemented for unmanaged allocation
    (IO (Array Word8) -> Array Word8)
-> IO (Array Word8) -> Array Word8
forall a b. (a -> b) -> a -> b
$ Fold IO Word8 (Array Word8) -> Stream IO Word8 -> IO (Array Word8)
forall (m :: * -> *) a b.
Monad m =>
Fold m a b -> Stream m a -> m b
D.fold (Int -> Int -> Fold IO Word8 (Array Word8)
forall (m :: * -> *) a.
(MonadIO m, Storable a) =>
Int -> Int -> Fold m a (Array a)
A.writeNAlignedUnmanaged Int
64 ([Word8] -> Int
forall (t :: * -> *) a. Foldable t => t a -> Int
length [Word8]
decodeTable))
              ([Word8] -> Stream IO Word8
forall (m :: * -> *) a. Applicative m => [a] -> Stream m a
D.fromList [Word8]
decodeTable)

-- | Return element at the specified index without checking the bounds.
-- and without touching the foreign ptr.
{-# INLINE_NORMAL unsafePeekElemOff #-}
unsafePeekElemOff :: forall a. Storable a => Ptr a -> Int -> a
unsafePeekElemOff :: Ptr a -> Int -> a
unsafePeekElemOff Ptr a
p Int
i = let !x :: a
x = IO a -> a
forall a. IO a -> a
A.unsafeInlineIO (IO a -> a) -> IO a -> a
forall a b. (a -> b) -> a -> b
$ Ptr a -> Int -> IO a
forall a. Storable a => Ptr a -> Int -> IO a
peekElemOff Ptr a
p Int
i in a
x

-- decode is split into two separate cases to avoid branching instructions.
-- From the higher level flow we already know which case we are in so we can
-- call the appropriate decode function.
--
-- When the state is 0
{-# INLINE decode0 #-}
decode0 :: Ptr Word8 -> Word8 -> Tuple' DecodeState CodePoint
decode0 :: Ptr Word8 -> Word8 -> Tuple' Word8 Int
decode0 Ptr Word8
table Word8
byte =
    let !t :: Word8
t = Ptr Word8
table Ptr Word8 -> Int -> Word8
forall a. Storable a => Ptr a -> Int -> a
`unsafePeekElemOff` Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
byte
        !codep' :: Int
codep' = (Int
0xff Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftR` Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
t) Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
byte
        !state' :: Word8
state' = Ptr Word8
table Ptr Word8 -> Int -> Word8
forall a. Storable a => Ptr a -> Int -> a
`unsafePeekElemOff` (Int
256 Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
t)
     in Bool -> Tuple' Word8 Int -> Tuple' Word8 Int
forall a. HasCallStack => Bool -> a -> a
assert ((Word8
byte Word8 -> Word8 -> Bool
forall a. Ord a => a -> a -> Bool
> Word8
0x7f Bool -> Bool -> Bool
|| [Char] -> Bool
forall a. HasCallStack => [Char] -> a
error [Char]
showByte)
                Bool -> Bool -> Bool
&& (Word8
state' Word8 -> Word8 -> Bool
forall a. Eq a => a -> a -> Bool
/= Word8
0 Bool -> Bool -> Bool
|| [Char] -> Bool
forall a. HasCallStack => [Char] -> a
error ([Char]
showByte [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
showTable)))
               (Word8 -> Int -> Tuple' Word8 Int
forall a b. a -> b -> Tuple' a b
Tuple' Word8
state' Int
codep')

    where

    utf8table :: Array Word8
utf8table =
        let !(Ptr Addr#
addr) = Ptr Word8
table
            end :: Ptr b
end = Ptr Word8
table Ptr Word8 -> Int -> Ptr b
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
364
        in ForeignPtr Word8 -> Ptr Word8 -> Array Word8
forall a. ForeignPtr a -> Ptr a -> Array a
A.Array (Addr# -> ForeignPtrContents -> ForeignPtr Word8
forall a. Addr# -> ForeignPtrContents -> ForeignPtr a
ForeignPtr Addr#
addr ForeignPtrContents
forall a. HasCallStack => a
undefined) Ptr Word8
forall b. Ptr b
end :: A.Array Word8
    showByte :: [Char]
showByte = [Char]
"Streamly: decode0: byte: " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Word8 -> [Char]
forall a. Show a => a -> [Char]
show Word8
byte
    showTable :: [Char]
showTable = [Char]
" table: " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Array Word8 -> [Char]
forall a. Show a => a -> [Char]
show Array Word8
utf8table

-- When the state is not 0
{-# INLINE decode1 #-}
decode1
    :: Ptr Word8
    -> DecodeState
    -> CodePoint
    -> Word8
    -> Tuple' DecodeState CodePoint
decode1 :: Ptr Word8 -> Word8 -> Int -> Word8 -> Tuple' Word8 Int
decode1 Ptr Word8
table Word8
state Int
codep Word8
byte =
    -- Remember codep is Int type!
    -- Can it be unsafe to convert the resulting Int to Char?
    let !t :: Word8
t = Ptr Word8
table Ptr Word8 -> Int -> Word8
forall a. Storable a => Ptr a -> Int -> a
`unsafePeekElemOff` Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
byte
        !codep' :: Int
codep' = (Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
byte Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Int
0x3f) Int -> Int -> Int
forall a. Bits a => a -> a -> a
.|. (Int
codep Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftL` Int
6)
        !state' :: Word8
state' = Ptr Word8
table Ptr Word8 -> Int -> Word8
forall a. Storable a => Ptr a -> Int -> a
`unsafePeekElemOff`
                    (Int
256 Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
state Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
t)
     in Bool -> Tuple' Word8 Int -> Tuple' Word8 Int
forall a. HasCallStack => Bool -> a -> a
assert (Int
codep' Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
<= Int
0x10FFFF
                    Bool -> Bool -> Bool
|| [Char] -> Bool
forall a. HasCallStack => [Char] -> a
error ([Char]
showByte [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Word8 -> Int -> [Char]
forall a a. (Show a, Show a) => a -> a -> [Char]
showState Word8
state Int
codep))
               (Word8 -> Int -> Tuple' Word8 Int
forall a b. a -> b -> Tuple' a b
Tuple' Word8
state' Int
codep')
    where

    utf8table :: Array Word8
utf8table =
        let !(Ptr Addr#
addr) = Ptr Word8
table
            end :: Ptr b
end = Ptr Word8
table Ptr Word8 -> Int -> Ptr b
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
364
        in ForeignPtr Word8 -> Ptr Word8 -> Array Word8
forall a. ForeignPtr a -> Ptr a -> Array a
A.Array (Addr# -> ForeignPtrContents -> ForeignPtr Word8
forall a. Addr# -> ForeignPtrContents -> ForeignPtr a
ForeignPtr Addr#
addr ForeignPtrContents
forall a. HasCallStack => a
undefined) Ptr Word8
forall b. Ptr b
end :: A.Array Word8
    showByte :: [Char]
showByte = [Char]
"Streamly: decode1: byte: " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Word8 -> [Char]
forall a. Show a => a -> [Char]
show Word8
byte
    showState :: a -> a -> [Char]
showState a
st a
cp =
        [Char]
" state: " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ a -> [Char]
forall a. Show a => a -> [Char]
show a
st [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++
        [Char]
" codepoint: " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ a -> [Char]
forall a. Show a => a -> [Char]
show a
cp [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++
        [Char]
" table: " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Array Word8 -> [Char]
forall a. Show a => a -> [Char]
show Array Word8
utf8table

-------------------------------------------------------------------------------
-- Resumable UTF-8 decoding
-------------------------------------------------------------------------------

{-# ANN type UTF8DecodeState Fuse #-}
data UTF8DecodeState s a
    = UTF8DecodeInit s
    | UTF8DecodeInit1 s Word8
    | UTF8DecodeFirst s Word8
    | UTF8Decoding s !DecodeState !CodePoint
    | YieldAndContinue a (UTF8DecodeState s a)
    | Done

{-# INLINE_NORMAL resumeDecodeUtf8EitherD #-}
resumeDecodeUtf8EitherD
    :: Monad m
    => DecodeState
    -> CodePoint
    -> Stream m Word8
    -> Stream m (Either DecodeError Char)
resumeDecodeUtf8EitherD :: Word8
-> Int -> Stream m Word8 -> Stream m (Either DecodeError Char)
resumeDecodeUtf8EitherD Word8
dst Int
codep (Stream State Stream m Word8 -> s -> m (Step s Word8)
step s
state) =
    let A.Array ForeignPtr Word8
p Ptr Word8
_ = Array Word8
utf8d
        !ptr :: Ptr Word8
ptr = ForeignPtr Word8 -> Ptr Word8
forall a. ForeignPtr a -> Ptr a
unsafeForeignPtrToPtr ForeignPtr Word8
p
        stt :: UTF8DecodeState s a
stt =
            if Word8
dst Word8 -> Word8 -> Bool
forall a. Eq a => a -> a -> Bool
== Word8
0
            then s -> UTF8DecodeState s a
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
state
            else s -> Word8 -> Int -> UTF8DecodeState s a
forall s a. s -> Word8 -> Int -> UTF8DecodeState s a
UTF8Decoding s
state Word8
dst Int
codep
    in (State Stream m (Either DecodeError Char)
 -> UTF8DecodeState s (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> UTF8DecodeState s (Either DecodeError Char)
-> Stream m (Either DecodeError Char)
forall (m :: * -> *) a s.
(State Stream m a -> s -> m (Step s a)) -> s -> Stream m a
Stream (Ptr Word8
-> State Stream m (Either DecodeError Char)
-> UTF8DecodeState s (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a.
Ptr Word8
-> State Stream m a
-> UTF8DecodeState s (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
step' Ptr Word8
ptr) UTF8DecodeState s (Either DecodeError Char)
forall a. UTF8DecodeState s a
stt
  where
    {-# INLINE_LATE step' #-}
    step' :: Ptr Word8
-> State Stream m a
-> UTF8DecodeState s (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
step' Ptr Word8
_ State Stream m a
gst (UTF8DecodeInit s
st) = do
        Step s Word8
r <- State Stream m Word8 -> s -> m (Step s Word8)
step (State Stream m a -> State Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a (n :: * -> *) b.
State t m a -> State t n b
adaptState State Stream m a
gst) s
st
        Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$ case Step s Word8
r of
            Yield Word8
x s
s -> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (s -> Word8 -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> Word8 -> UTF8DecodeState s a
UTF8DecodeInit1 s
s Word8
x)
            Skip s
s -> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (s -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
s)
            Step s Word8
Stop   -> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip UTF8DecodeState s (Either DecodeError Char)
forall s a. UTF8DecodeState s a
Done

    step' Ptr Word8
_ State Stream m a
_ (UTF8DecodeInit1 s
st Word8
x) = do
        -- Note: It is important to use a ">" instead of a "<=" test
        -- here for GHC to generate code layout for default branch
        -- prediction for the common case. This is fragile and might
        -- change with the compiler versions, we need a more reliable
        -- "likely" primitive to control branch predication.
        case Word8
x Word8 -> Word8 -> Bool
forall a. Ord a => a -> a -> Bool
> Word8
0x7f of
            Bool
False ->
                Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$ UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (UTF8DecodeState s (Either DecodeError Char)
 -> Step
      (UTF8DecodeState s (Either DecodeError Char))
      (Either DecodeError Char))
-> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall a b. (a -> b) -> a -> b
$ Either DecodeError Char
-> UTF8DecodeState s (Either DecodeError Char)
-> UTF8DecodeState s (Either DecodeError Char)
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue
                    (Char -> Either DecodeError Char
forall a b. b -> Either a b
Right (Char -> Either DecodeError Char)
-> Char -> Either DecodeError Char
forall a b. (a -> b) -> a -> b
$ Int -> Char
unsafeChr (Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
x))
                    (s -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
st)
            -- Using a separate state here generates a jump to a
            -- separate code block in the core which seems to perform
            -- slightly better for the non-ascii case.
            Bool
True -> Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$ UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (UTF8DecodeState s (Either DecodeError Char)
 -> Step
      (UTF8DecodeState s (Either DecodeError Char))
      (Either DecodeError Char))
-> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall a b. (a -> b) -> a -> b
$ s -> Word8 -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> Word8 -> UTF8DecodeState s a
UTF8DecodeFirst s
st Word8
x

    -- XXX should we merge it with UTF8DecodeInit1?
    step' Ptr Word8
table State Stream m a
_ (UTF8DecodeFirst s
st Word8
x) = do
        let (Tuple' Word8
sv Int
cp) = Ptr Word8 -> Word8 -> Tuple' Word8 Int
decode0 Ptr Word8
table Word8
x
        Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$
            case Word8
sv of
                Word8
12 ->
                    UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (UTF8DecodeState s (Either DecodeError Char)
 -> Step
      (UTF8DecodeState s (Either DecodeError Char))
      (Either DecodeError Char))
-> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall a b. (a -> b) -> a -> b
$ Either DecodeError Char
-> UTF8DecodeState s (Either DecodeError Char)
-> UTF8DecodeState s (Either DecodeError Char)
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue (DecodeError -> Either DecodeError Char
forall a b. a -> Either a b
Left (DecodeError -> Either DecodeError Char)
-> DecodeError -> Either DecodeError Char
forall a b. (a -> b) -> a -> b
$ Word8 -> Int -> DecodeError
DecodeError Word8
0 (Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
x))
                                            (s -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
st)
                Word8
0 -> [Char]
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall a. HasCallStack => [Char] -> a
error [Char]
"unreachable state"
                Word8
_ -> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (s -> Word8 -> Int -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> Word8 -> Int -> UTF8DecodeState s a
UTF8Decoding s
st Word8
sv Int
cp)

    -- We recover by trying the new byte x a starter of a new codepoint.
    -- XXX on error need to report the next byte "x" as well.
    -- XXX need to use the same recovery in array decoding routine as well
    step' Ptr Word8
table State Stream m a
gst (UTF8Decoding s
st Word8
statePtr Int
codepointPtr) = do
        Step s Word8
r <- State Stream m Word8 -> s -> m (Step s Word8)
step (State Stream m a -> State Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a (n :: * -> *) b.
State t m a -> State t n b
adaptState State Stream m a
gst) s
st
        case Step s Word8
r of
            Yield Word8
x s
s -> do
                let (Tuple' Word8
sv Int
cp) = Ptr Word8 -> Word8 -> Int -> Word8 -> Tuple' Word8 Int
decode1 Ptr Word8
table Word8
statePtr Int
codepointPtr Word8
x
                Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$
                    case Word8
sv of
                        Word8
0 -> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (UTF8DecodeState s (Either DecodeError Char)
 -> Step
      (UTF8DecodeState s (Either DecodeError Char))
      (Either DecodeError Char))
-> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall a b. (a -> b) -> a -> b
$ Either DecodeError Char
-> UTF8DecodeState s (Either DecodeError Char)
-> UTF8DecodeState s (Either DecodeError Char)
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue (Char -> Either DecodeError Char
forall a b. b -> Either a b
Right (Char -> Either DecodeError Char)
-> Char -> Either DecodeError Char
forall a b. (a -> b) -> a -> b
$ Int -> Char
unsafeChr Int
cp)
                                        (s -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
s)
                        Word8
12 ->
                            UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (UTF8DecodeState s (Either DecodeError Char)
 -> Step
      (UTF8DecodeState s (Either DecodeError Char))
      (Either DecodeError Char))
-> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall a b. (a -> b) -> a -> b
$ Either DecodeError Char
-> UTF8DecodeState s (Either DecodeError Char)
-> UTF8DecodeState s (Either DecodeError Char)
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue (DecodeError -> Either DecodeError Char
forall a b. a -> Either a b
Left (DecodeError -> Either DecodeError Char)
-> DecodeError -> Either DecodeError Char
forall a b. (a -> b) -> a -> b
$ Word8 -> Int -> DecodeError
DecodeError Word8
statePtr Int
codepointPtr)
                                        (s -> Word8 -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> Word8 -> UTF8DecodeState s a
UTF8DecodeInit1 s
s Word8
x)
                        Word8
_ -> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (s -> Word8 -> Int -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> Word8 -> Int -> UTF8DecodeState s a
UTF8Decoding s
s Word8
sv Int
cp)
            Skip s
s -> Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$ UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (s -> Word8 -> Int -> UTF8DecodeState s (Either DecodeError Char)
forall s a. s -> Word8 -> Int -> UTF8DecodeState s a
UTF8Decoding s
s Word8
statePtr Int
codepointPtr)
            Step s Word8
Stop -> Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$ UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. s -> Step s a
Skip (UTF8DecodeState s (Either DecodeError Char)
 -> Step
      (UTF8DecodeState s (Either DecodeError Char))
      (Either DecodeError Char))
-> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall a b. (a -> b) -> a -> b
$ Either DecodeError Char
-> UTF8DecodeState s (Either DecodeError Char)
-> UTF8DecodeState s (Either DecodeError Char)
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue (DecodeError -> Either DecodeError Char
forall a b. a -> Either a b
Left (DecodeError -> Either DecodeError Char)
-> DecodeError -> Either DecodeError Char
forall a b. (a -> b) -> a -> b
$ Word8 -> Int -> DecodeError
DecodeError Word8
statePtr Int
codepointPtr) UTF8DecodeState s (Either DecodeError Char)
forall s a. UTF8DecodeState s a
Done

    step' Ptr Word8
_ State Stream m a
_ (YieldAndContinue Either DecodeError Char
c UTF8DecodeState s (Either DecodeError Char)
s) = Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return (Step
   (UTF8DecodeState s (Either DecodeError Char))
   (Either DecodeError Char)
 -> m (Step
         (UTF8DecodeState s (Either DecodeError Char))
         (Either DecodeError Char)))
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall a b. (a -> b) -> a -> b
$ Either DecodeError Char
-> UTF8DecodeState s (Either DecodeError Char)
-> Step
     (UTF8DecodeState s (Either DecodeError Char))
     (Either DecodeError Char)
forall s a. a -> s -> Step s a
Yield Either DecodeError Char
c UTF8DecodeState s (Either DecodeError Char)
s
    step' Ptr Word8
_ State Stream m a
_ UTF8DecodeState s (Either DecodeError Char)
Done = Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
-> m (Step
        (UTF8DecodeState s (Either DecodeError Char))
        (Either DecodeError Char))
forall (m :: * -> *) a. Monad m => a -> m a
return Step
  (UTF8DecodeState s (Either DecodeError Char))
  (Either DecodeError Char)
forall s a. Step s a
Stop

-- XXX We can use just one API, and define InitState = 0 and InitCodePoint = 0
-- to use as starting state.
--
{-# INLINE_NORMAL decodeUtf8EitherD #-}
decodeUtf8EitherD :: Monad m
    => Stream m Word8 -> Stream m (Either DecodeError Char)
decodeUtf8EitherD :: Stream m Word8 -> Stream m (Either DecodeError Char)
decodeUtf8EitherD = Word8
-> Int -> Stream m Word8 -> Stream m (Either DecodeError Char)
forall (m :: * -> *).
Monad m =>
Word8
-> Int -> Stream m Word8 -> Stream m (Either DecodeError Char)
resumeDecodeUtf8EitherD Word8
0 Int
0

-- |
--
-- /Pre-release/
{-# INLINE decodeUtf8Either #-}
decodeUtf8Either :: (Monad m, IsStream t)
    => t m Word8 -> t m (Either DecodeError Char)
decodeUtf8Either :: t m Word8 -> t m (Either DecodeError Char)
decodeUtf8Either = Stream m (Either DecodeError Char) -> t m (Either DecodeError Char)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m (Either DecodeError Char)
 -> t m (Either DecodeError Char))
-> (t m Word8 -> Stream m (Either DecodeError Char))
-> t m Word8
-> t m (Either DecodeError Char)
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m Word8 -> Stream m (Either DecodeError Char)
forall (m :: * -> *).
Monad m =>
Stream m Word8 -> Stream m (Either DecodeError Char)
decodeUtf8EitherD (Stream m Word8 -> Stream m (Either DecodeError Char))
-> (t m Word8 -> Stream m Word8)
-> t m Word8
-> Stream m (Either DecodeError Char)
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Word8 -> Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

-- |
--
-- /Pre-release/
{-# INLINE resumeDecodeUtf8Either #-}
resumeDecodeUtf8Either
    :: (Monad m, IsStream t)
    => DecodeState
    -> CodePoint
    -> t m Word8
    -> t m (Either DecodeError Char)
resumeDecodeUtf8Either :: Word8 -> Int -> t m Word8 -> t m (Either DecodeError Char)
resumeDecodeUtf8Either Word8
st Int
cp =
    Stream m (Either DecodeError Char) -> t m (Either DecodeError Char)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m (Either DecodeError Char)
 -> t m (Either DecodeError Char))
-> (t m Word8 -> Stream m (Either DecodeError Char))
-> t m Word8
-> t m (Either DecodeError Char)
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Word8
-> Int -> Stream m Word8 -> Stream m (Either DecodeError Char)
forall (m :: * -> *).
Monad m =>
Word8
-> Int -> Stream m Word8 -> Stream m (Either DecodeError Char)
resumeDecodeUtf8EitherD Word8
st Int
cp (Stream m Word8 -> Stream m (Either DecodeError Char))
-> (t m Word8 -> Stream m Word8)
-> t m Word8
-> Stream m (Either DecodeError Char)
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Word8 -> Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

-------------------------------------------------------------------------------
-- One shot decoding
-------------------------------------------------------------------------------

data CodingFailureMode
    = TransliterateCodingFailure
    | ErrorOnCodingFailure
    | DropOnCodingFailure
    deriving (Int -> CodingFailureMode -> [Char] -> [Char]
[CodingFailureMode] -> [Char] -> [Char]
CodingFailureMode -> [Char]
(Int -> CodingFailureMode -> [Char] -> [Char])
-> (CodingFailureMode -> [Char])
-> ([CodingFailureMode] -> [Char] -> [Char])
-> Show CodingFailureMode
forall a.
(Int -> a -> [Char] -> [Char])
-> (a -> [Char]) -> ([a] -> [Char] -> [Char]) -> Show a
showList :: [CodingFailureMode] -> [Char] -> [Char]
$cshowList :: [CodingFailureMode] -> [Char] -> [Char]
show :: CodingFailureMode -> [Char]
$cshow :: CodingFailureMode -> [Char]
showsPrec :: Int -> CodingFailureMode -> [Char] -> [Char]
$cshowsPrec :: Int -> CodingFailureMode -> [Char] -> [Char]
Show)

{-# INLINE replacementChar #-}
replacementChar :: Char
replacementChar :: Char
replacementChar = Char
'\xFFFD'

-- XXX write it as a parser and use parseMany to decode a stream, need to check
-- if that preserves the same performance. Or we can use a resumable parser
-- that parses a chunk at a time.
--
-- XXX Implement this in terms of decodeUtf8Either. Need to make sure that
-- decodeUtf8Either preserves the performance characterstics.
--
{-# INLINE_NORMAL decodeUtf8WithD #-}
decodeUtf8WithD :: Monad m
    => CodingFailureMode -> Stream m Word8 -> Stream m Char
decodeUtf8WithD :: CodingFailureMode -> Stream m Word8 -> Stream m Char
decodeUtf8WithD CodingFailureMode
cfm (Stream State Stream m Word8 -> s -> m (Step s Word8)
step s
state) =
    let A.Array ForeignPtr Word8
p Ptr Word8
_ = Array Word8
utf8d
        !ptr :: Ptr Word8
ptr = ForeignPtr Word8 -> Ptr Word8
forall a. ForeignPtr a -> Ptr a
unsafeForeignPtrToPtr ForeignPtr Word8
p
    in (State Stream m Char
 -> UTF8DecodeState s Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> UTF8DecodeState s Char -> Stream m Char
forall (m :: * -> *) a s.
(State Stream m a -> s -> m (Step s a)) -> s -> Stream m a
Stream (Ptr Word8
-> State Stream m Char
-> UTF8DecodeState s Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a.
Ptr Word8
-> State Stream m a
-> UTF8DecodeState s Char
-> m (Step (UTF8DecodeState s Char) Char)
step' Ptr Word8
ptr) (s -> UTF8DecodeState s Char
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
state)

    where

    prefix :: [Char]
prefix = [Char]
"Streamly.Internal.Data.Stream.StreamD.decodeUtf8With: "

    {-# INLINE handleError #-}
    handleError :: [Char] -> UTF8DecodeState s Char -> UTF8DecodeState s Char
handleError [Char]
e UTF8DecodeState s Char
s =
        case CodingFailureMode
cfm of
            CodingFailureMode
ErrorOnCodingFailure -> [Char] -> UTF8DecodeState s Char
forall a. HasCallStack => [Char] -> a
error [Char]
e
            CodingFailureMode
TransliterateCodingFailure -> Char -> UTF8DecodeState s Char -> UTF8DecodeState s Char
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue Char
replacementChar UTF8DecodeState s Char
s
            CodingFailureMode
DropOnCodingFailure -> UTF8DecodeState s Char
s

    {-# INLINE handleUnderflow #-}
    handleUnderflow :: UTF8DecodeState s Char
handleUnderflow =
        case CodingFailureMode
cfm of
            CodingFailureMode
ErrorOnCodingFailure -> [Char] -> UTF8DecodeState s Char
forall a. HasCallStack => [Char] -> a
error ([Char] -> UTF8DecodeState s Char)
-> [Char] -> UTF8DecodeState s Char
forall a b. (a -> b) -> a -> b
$ [Char]
prefix [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
"Not enough input"
            CodingFailureMode
TransliterateCodingFailure -> Char -> UTF8DecodeState s Char -> UTF8DecodeState s Char
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue Char
replacementChar UTF8DecodeState s Char
forall s a. UTF8DecodeState s a
Done
            CodingFailureMode
DropOnCodingFailure -> UTF8DecodeState s Char
forall s a. UTF8DecodeState s a
Done

    {-# INLINE_LATE step' #-}
    step' :: Ptr Word8
-> State Stream m a
-> UTF8DecodeState s Char
-> m (Step (UTF8DecodeState s Char) Char)
step' Ptr Word8
_ State Stream m a
gst (UTF8DecodeInit s
st) = do
        Step s Word8
r <- State Stream m Word8 -> s -> m (Step s Word8)
step (State Stream m a -> State Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a (n :: * -> *) b.
State t m a -> State t n b
adaptState State Stream m a
gst) s
st
        Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$ case Step s Word8
r of
            Yield Word8
x s
s -> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (s -> Word8 -> UTF8DecodeState s Char
forall s a. s -> Word8 -> UTF8DecodeState s a
UTF8DecodeInit1 s
s Word8
x)
            Skip s
s -> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (s -> UTF8DecodeState s Char
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
s)
            Step s Word8
Stop   -> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip UTF8DecodeState s Char
forall s a. UTF8DecodeState s a
Done

    step' Ptr Word8
_ State Stream m a
_ (UTF8DecodeInit1 s
st Word8
x) = do
        -- Note: It is important to use a ">" instead of a "<=" test
        -- here for GHC to generate code layout for default branch
        -- prediction for the common case. This is fragile and might
        -- change with the compiler versions, we need a more reliable
        -- "likely" primitive to control branch predication.
        case Word8
x Word8 -> Word8 -> Bool
forall a. Ord a => a -> a -> Bool
> Word8
0x7f of
            Bool
False ->
                Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$ UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char)
-> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall a b. (a -> b) -> a -> b
$ Char -> UTF8DecodeState s Char -> UTF8DecodeState s Char
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue
                    (Int -> Char
unsafeChr (Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
x))
                    (s -> UTF8DecodeState s Char
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
st)
            -- Using a separate state here generates a jump to a
            -- separate code block in the core which seems to perform
            -- slightly better for the non-ascii case.
            Bool
True -> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$ UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char)
-> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall a b. (a -> b) -> a -> b
$ s -> Word8 -> UTF8DecodeState s Char
forall s a. s -> Word8 -> UTF8DecodeState s a
UTF8DecodeFirst s
st Word8
x

    -- XXX should we merge it with UTF8DecodeInit1?
    step' Ptr Word8
table State Stream m a
_ (UTF8DecodeFirst s
st Word8
x) = do
        let (Tuple' Word8
sv Int
cp) = Ptr Word8 -> Word8 -> Tuple' Word8 Int
decode0 Ptr Word8
table Word8
x
        Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$
            case Word8
sv of
                Word8
12 ->
                    let msg :: [Char]
msg = [Char]
prefix [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
"Invalid first UTF8 byte " [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Word8 -> [Char]
forall a. Show a => a -> [Char]
show Word8
x
                     in UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char)
-> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall a b. (a -> b) -> a -> b
$ [Char] -> UTF8DecodeState s Char -> UTF8DecodeState s Char
forall s.
[Char] -> UTF8DecodeState s Char -> UTF8DecodeState s Char
handleError [Char]
msg (s -> UTF8DecodeState s Char
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
st)
                Word8
0 -> [Char] -> Step (UTF8DecodeState s Char) Char
forall a. HasCallStack => [Char] -> a
error [Char]
"unreachable state"
                Word8
_ -> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (s -> Word8 -> Int -> UTF8DecodeState s Char
forall s a. s -> Word8 -> Int -> UTF8DecodeState s a
UTF8Decoding s
st Word8
sv Int
cp)

    -- We recover by trying the new byte x as a starter of a new codepoint.
    -- XXX need to use the same recovery in array decoding routine as well
    step' Ptr Word8
table State Stream m a
gst (UTF8Decoding s
st Word8
statePtr Int
codepointPtr) = do
        Step s Word8
r <- State Stream m Word8 -> s -> m (Step s Word8)
step (State Stream m a -> State Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a (n :: * -> *) b.
State t m a -> State t n b
adaptState State Stream m a
gst) s
st
        case Step s Word8
r of
            Yield Word8
x s
s -> do
                let (Tuple' Word8
sv Int
cp) = Ptr Word8 -> Word8 -> Int -> Word8 -> Tuple' Word8 Int
decode1 Ptr Word8
table Word8
statePtr Int
codepointPtr Word8
x
                Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$ case Word8
sv of
                    Word8
0 -> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char)
-> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall a b. (a -> b) -> a -> b
$ Char -> UTF8DecodeState s Char -> UTF8DecodeState s Char
forall s a. a -> UTF8DecodeState s a -> UTF8DecodeState s a
YieldAndContinue
                            (Int -> Char
unsafeChr Int
cp) (s -> UTF8DecodeState s Char
forall s a. s -> UTF8DecodeState s a
UTF8DecodeInit s
s)
                    Word8
12 ->
                        let msg :: [Char]
msg = [Char]
prefix
                                [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
"Invalid subsequent UTF8 byte "
                                [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Word8 -> [Char]
forall a. Show a => a -> [Char]
show Word8
x
                                [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
" in state "
                                [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Word8 -> [Char]
forall a. Show a => a -> [Char]
show Word8
statePtr
                                [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
" accumulated value "
                                [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ Int -> [Char]
forall a. Show a => a -> [Char]
show Int
codepointPtr
                         in UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char)
-> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall a b. (a -> b) -> a -> b
$ [Char] -> UTF8DecodeState s Char -> UTF8DecodeState s Char
forall s.
[Char] -> UTF8DecodeState s Char -> UTF8DecodeState s Char
handleError [Char]
msg (s -> Word8 -> UTF8DecodeState s Char
forall s a. s -> Word8 -> UTF8DecodeState s a
UTF8DecodeInit1 s
s Word8
x)
                    Word8
_ -> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (s -> Word8 -> Int -> UTF8DecodeState s Char
forall s a. s -> Word8 -> Int -> UTF8DecodeState s a
UTF8Decoding s
s Word8
sv Int
cp)
            Skip s
s -> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$
                UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip (s -> Word8 -> Int -> UTF8DecodeState s Char
forall s a. s -> Word8 -> Int -> UTF8DecodeState s a
UTF8Decoding s
s Word8
statePtr Int
codepointPtr)
            Step s Word8
Stop -> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$ UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. s -> Step s a
Skip UTF8DecodeState s Char
forall s. UTF8DecodeState s Char
handleUnderflow

    step' Ptr Word8
_ State Stream m a
_ (YieldAndContinue Char
c UTF8DecodeState s Char
s) = Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (UTF8DecodeState s Char) Char
 -> m (Step (UTF8DecodeState s Char) Char))
-> Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall a b. (a -> b) -> a -> b
$ Char
-> UTF8DecodeState s Char -> Step (UTF8DecodeState s Char) Char
forall s a. a -> s -> Step s a
Yield Char
c UTF8DecodeState s Char
s
    step' Ptr Word8
_ State Stream m a
_ UTF8DecodeState s Char
Done = Step (UTF8DecodeState s Char) Char
-> m (Step (UTF8DecodeState s Char) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return Step (UTF8DecodeState s Char) Char
forall s a. Step s a
Stop

{-# INLINE decodeUtf8D #-}
decodeUtf8D :: Monad m => Stream m Word8 -> Stream m Char
decodeUtf8D :: Stream m Word8 -> Stream m Char
decodeUtf8D = CodingFailureMode -> Stream m Word8 -> Stream m Char
forall (m :: * -> *).
Monad m =>
CodingFailureMode -> Stream m Word8 -> Stream m Char
decodeUtf8WithD CodingFailureMode
TransliterateCodingFailure

-- | Decode a UTF-8 encoded bytestream to a stream of Unicode characters.
-- Any invalid codepoint encountered is replaced with the unicode replacement
-- character.
--
-- /Since: 0.7.0 ("Streamly.Data.Unicode.Stream")/
--
-- /Since: 0.8.0 (Lenient Behaviour)/
{-# INLINE decodeUtf8 #-}
decodeUtf8 :: (Monad m, IsStream t) => t m Word8 -> t m Char
decodeUtf8 :: t m Word8 -> t m Char
decodeUtf8 = Stream m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Char -> t m Char)
-> (t m Word8 -> Stream m Char) -> t m Word8 -> t m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m Word8 -> Stream m Char
forall (m :: * -> *). Monad m => Stream m Word8 -> Stream m Char
decodeUtf8D (Stream m Word8 -> Stream m Char)
-> (t m Word8 -> Stream m Word8) -> t m Word8 -> Stream m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Word8 -> Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

{-# INLINE decodeUtf8D' #-}
decodeUtf8D' :: Monad m => Stream m Word8 -> Stream m Char
decodeUtf8D' :: Stream m Word8 -> Stream m Char
decodeUtf8D' = CodingFailureMode -> Stream m Word8 -> Stream m Char
forall (m :: * -> *).
Monad m =>
CodingFailureMode -> Stream m Word8 -> Stream m Char
decodeUtf8WithD CodingFailureMode
ErrorOnCodingFailure

-- | Decode a UTF-8 encoded bytestream to a stream of Unicode characters.
-- The function throws an error if an invalid codepoint is encountered.
--
-- @since 0.8.0
{-# INLINE decodeUtf8' #-}
decodeUtf8' :: (Monad m, IsStream t) => t m Word8 -> t m Char
decodeUtf8' :: t m Word8 -> t m Char
decodeUtf8' = Stream m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Char -> t m Char)
-> (t m Word8 -> Stream m Char) -> t m Word8 -> t m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m Word8 -> Stream m Char
forall (m :: * -> *). Monad m => Stream m Word8 -> Stream m Char
decodeUtf8D' (Stream m Word8 -> Stream m Char)
-> (t m Word8 -> Stream m Word8) -> t m Word8 -> Stream m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Word8 -> Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

{-# INLINE decodeUtf8D_ #-}
decodeUtf8D_ :: Monad m => Stream m Word8 -> Stream m Char
decodeUtf8D_ :: Stream m Word8 -> Stream m Char
decodeUtf8D_ = CodingFailureMode -> Stream m Word8 -> Stream m Char
forall (m :: * -> *).
Monad m =>
CodingFailureMode -> Stream m Word8 -> Stream m Char
decodeUtf8WithD CodingFailureMode
DropOnCodingFailure

-- | Decode a UTF-8 encoded bytestream to a stream of Unicode characters.
-- Any invalid codepoint encountered is dropped.
--
-- @since 0.8.0
{-# INLINE decodeUtf8_ #-}
decodeUtf8_ :: (Monad m, IsStream t) => t m Word8 -> t m Char
decodeUtf8_ :: t m Word8 -> t m Char
decodeUtf8_ = Stream m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Char -> t m Char)
-> (t m Word8 -> Stream m Char) -> t m Word8 -> t m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m Word8 -> Stream m Char
forall (m :: * -> *). Monad m => Stream m Word8 -> Stream m Char
decodeUtf8D_ (Stream m Word8 -> Stream m Char)
-> (t m Word8 -> Stream m Word8) -> t m Word8 -> Stream m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Word8 -> Stream m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

-- | Same as 'decodeUtf8'
--
{-# DEPRECATED decodeUtf8Lax "Please use 'decodeUtf8' instead" #-}
{-# INLINE decodeUtf8Lax #-}
decodeUtf8Lax :: (IsStream t, Monad m) => t m Word8 -> t m Char
decodeUtf8Lax :: t m Word8 -> t m Char
decodeUtf8Lax = t m Word8 -> t m Char
forall (m :: * -> *) (t :: (* -> *) -> * -> *).
(Monad m, IsStream t) =>
t m Word8 -> t m Char
decodeUtf8

-------------------------------------------------------------------------------
-- Decoding Array Streams
-------------------------------------------------------------------------------

{-# ANN type FlattenState Fuse #-}
data FlattenState s a
    = OuterLoop s !(Maybe (DecodeState, CodePoint))
    | InnerLoopDecodeInit s (ForeignPtr a) !(Ptr a) !(Ptr a)
    | InnerLoopDecodeFirst s (ForeignPtr a) !(Ptr a) !(Ptr a) Word8
    | InnerLoopDecoding s (ForeignPtr a) !(Ptr a) !(Ptr a)
        !DecodeState !CodePoint
    | YAndC !Char (FlattenState s a) -- These constructors can be
                                     -- encoded in the UTF8DecodeState
                                     -- type, I prefer to keep these
                                     -- flat even though that means
                                     -- coming up with new names
    | D

-- The normal decodeUtf8 above should fuse with flattenArrays
-- to create this exact code but it doesn't for some reason, as of now this
-- remains the fastest way I could figure out to decodeUtf8.
--
-- XXX Add Proper error messages
{-# INLINE_NORMAL decodeUtf8ArraysWithD #-}
decodeUtf8ArraysWithD ::
       MonadIO m
    => CodingFailureMode
    -> Stream m (A.Array Word8)
    -> Stream m Char
decodeUtf8ArraysWithD :: CodingFailureMode -> Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysWithD CodingFailureMode
cfm (Stream State Stream m (Array Word8) -> s -> m (Step s (Array Word8))
step s
state) =
    let A.Array ForeignPtr Word8
p Ptr Word8
_ = Array Word8
utf8d
        !ptr :: Ptr Word8
ptr = ForeignPtr Word8 -> Ptr Word8
forall a. ForeignPtr a -> Ptr a
unsafeForeignPtrToPtr ForeignPtr Word8
p
    in (State Stream m Char
 -> FlattenState s Word8 -> m (Step (FlattenState s Word8) Char))
-> FlattenState s Word8 -> Stream m Char
forall (m :: * -> *) a s.
(State Stream m a -> s -> m (Step s a)) -> s -> Stream m a
Stream (Ptr Word8
-> State Stream m Char
-> FlattenState s Word8
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a.
Ptr Word8
-> State Stream m a
-> FlattenState s Word8
-> m (Step (FlattenState s Word8) Char)
step' Ptr Word8
ptr) (s -> Maybe (Word8, Int) -> FlattenState s Word8
forall s a. s -> Maybe (Word8, Int) -> FlattenState s a
OuterLoop s
state Maybe (Word8, Int)
forall a. Maybe a
Nothing)
  where
    {-# INLINE transliterateOrError #-}
    transliterateOrError :: [Char] -> FlattenState s a -> FlattenState s a
transliterateOrError [Char]
e FlattenState s a
s =
        case CodingFailureMode
cfm of
            CodingFailureMode
ErrorOnCodingFailure -> [Char] -> FlattenState s a
forall a. HasCallStack => [Char] -> a
error [Char]
e
            CodingFailureMode
TransliterateCodingFailure -> Char -> FlattenState s a -> FlattenState s a
forall s a. Char -> FlattenState s a -> FlattenState s a
YAndC Char
replacementChar FlattenState s a
s
            CodingFailureMode
DropOnCodingFailure -> FlattenState s a
s
    {-# INLINE inputUnderflow #-}
    inputUnderflow :: FlattenState s a
inputUnderflow =
        case CodingFailureMode
cfm of
            CodingFailureMode
ErrorOnCodingFailure ->
                [Char] -> FlattenState s a
forall a. HasCallStack => [Char] -> a
error ([Char] -> FlattenState s a) -> [Char] -> FlattenState s a
forall a b. (a -> b) -> a -> b
$
                [Char] -> [Char]
forall a. Show a => a -> [Char]
show [Char]
"Streamly.Internal.Data.Stream.StreamD."
                [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
"decodeUtf8ArraysWith: Input Underflow"
            CodingFailureMode
TransliterateCodingFailure -> Char -> FlattenState s a -> FlattenState s a
forall s a. Char -> FlattenState s a -> FlattenState s a
YAndC Char
replacementChar FlattenState s a
forall s a. FlattenState s a
D
            CodingFailureMode
DropOnCodingFailure -> FlattenState s a
forall s a. FlattenState s a
D
    {-# INLINE_LATE step' #-}
    step' :: Ptr Word8
-> State Stream m a
-> FlattenState s Word8
-> m (Step (FlattenState s Word8) Char)
step' Ptr Word8
_ State Stream m a
gst (OuterLoop s
st Maybe (Word8, Int)
Nothing) = do
        Step s (Array Word8)
r <- State Stream m (Array Word8) -> s -> m (Step s (Array Word8))
step (State Stream m a -> State Stream m (Array Word8)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a (n :: * -> *) b.
State t m a -> State t n b
adaptState State Stream m a
gst) s
st
        Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$
            case Step s (Array Word8)
r of
                Yield A.Array {Ptr Word8
ForeignPtr Word8
aEnd :: forall a. Array a -> Ptr a
aStart :: forall a. Array a -> ForeignPtr a
aEnd :: Ptr Word8
aStart :: ForeignPtr Word8
..} s
s ->
                    let p :: Ptr Word8
p = ForeignPtr Word8 -> Ptr Word8
forall a. ForeignPtr a -> Ptr a
unsafeForeignPtrToPtr ForeignPtr Word8
aStart
                     in FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> FlattenState s Word8
forall s a. s -> ForeignPtr a -> Ptr a -> Ptr a -> FlattenState s a
InnerLoopDecodeInit s
s ForeignPtr Word8
aStart Ptr Word8
p Ptr Word8
aEnd)
                Skip s
s -> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (s -> Maybe (Word8, Int) -> FlattenState s Word8
forall s a. s -> Maybe (Word8, Int) -> FlattenState s a
OuterLoop s
s Maybe (Word8, Int)
forall a. Maybe a
Nothing)
                Step s (Array Word8)
Stop -> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip FlattenState s Word8
forall s a. FlattenState s a
D
    step' Ptr Word8
_ State Stream m a
gst (OuterLoop s
st dst :: Maybe (Word8, Int)
dst@(Just (Word8
ds, Int
cp))) = do
        Step s (Array Word8)
r <- State Stream m (Array Word8) -> s -> m (Step s (Array Word8))
step (State Stream m a -> State Stream m (Array Word8)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a (n :: * -> *) b.
State t m a -> State t n b
adaptState State Stream m a
gst) s
st
        Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$
            case Step s (Array Word8)
r of
                Yield A.Array {Ptr Word8
ForeignPtr Word8
aEnd :: Ptr Word8
aStart :: ForeignPtr Word8
aEnd :: forall a. Array a -> Ptr a
aStart :: forall a. Array a -> ForeignPtr a
..} s
s ->
                    let p :: Ptr Word8
p = ForeignPtr Word8 -> Ptr Word8
forall a. ForeignPtr a -> Ptr a
unsafeForeignPtrToPtr ForeignPtr Word8
aStart
                     in FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> Word8
-> Int
-> FlattenState s Word8
forall s a.
s
-> ForeignPtr a
-> Ptr a
-> Ptr a
-> Word8
-> Int
-> FlattenState s a
InnerLoopDecoding s
s ForeignPtr Word8
aStart Ptr Word8
p Ptr Word8
aEnd Word8
ds Int
cp)
                Skip s
s -> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (s -> Maybe (Word8, Int) -> FlattenState s Word8
forall s a. s -> Maybe (Word8, Int) -> FlattenState s a
OuterLoop s
s Maybe (Word8, Int)
dst)
                Step s (Array Word8)
Stop -> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip FlattenState s Word8
forall s a. FlattenState s a
inputUnderflow
    step' Ptr Word8
_ State Stream m a
_ (InnerLoopDecodeInit s
st ForeignPtr Word8
startf Ptr Word8
p Ptr Word8
end)
        | Ptr Word8
p Ptr Word8 -> Ptr Word8 -> Bool
forall a. Eq a => a -> a -> Bool
== Ptr Word8
end = do
            IO () -> m ()
forall (m :: * -> *) a. MonadIO m => IO a -> m a
liftIO (IO () -> m ()) -> IO () -> m ()
forall a b. (a -> b) -> a -> b
$ ForeignPtr Word8 -> IO ()
forall a. ForeignPtr a -> IO ()
touchForeignPtr ForeignPtr Word8
startf
            Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$ FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (FlattenState s Word8 -> Step (FlattenState s Word8) Char)
-> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall a b. (a -> b) -> a -> b
$ s -> Maybe (Word8, Int) -> FlattenState s Word8
forall s a. s -> Maybe (Word8, Int) -> FlattenState s a
OuterLoop s
st Maybe (Word8, Int)
forall a. Maybe a
Nothing
    step' Ptr Word8
_ State Stream m a
_ (InnerLoopDecodeInit s
st ForeignPtr Word8
startf Ptr Word8
p Ptr Word8
end) = do
        Word8
x <- IO Word8 -> m Word8
forall (m :: * -> *) a. MonadIO m => IO a -> m a
liftIO (IO Word8 -> m Word8) -> IO Word8 -> m Word8
forall a b. (a -> b) -> a -> b
$ Ptr Word8 -> IO Word8
forall a. Storable a => Ptr a -> IO a
peek Ptr Word8
p
        -- Note: It is important to use a ">" instead of a "<=" test here for
        -- GHC to generate code layout for default branch prediction for the
        -- common case. This is fragile and might change with the compiler
        -- versions, we need a more reliable "likely" primitive to control
        -- branch predication.
        case Word8
x Word8 -> Word8 -> Bool
forall a. Ord a => a -> a -> Bool
> Word8
0x7f of
            Bool
False ->
                Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$ FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (FlattenState s Word8 -> Step (FlattenState s Word8) Char)
-> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall a b. (a -> b) -> a -> b
$ Char -> FlattenState s Word8 -> FlattenState s Word8
forall s a. Char -> FlattenState s a -> FlattenState s a
YAndC
                    (Int -> Char
unsafeChr (Word8 -> Int
forall a b. (Integral a, Num b) => a -> b
fromIntegral Word8
x))
                    (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> FlattenState s Word8
forall s a. s -> ForeignPtr a -> Ptr a -> Ptr a -> FlattenState s a
InnerLoopDecodeInit s
st ForeignPtr Word8
startf (Ptr Word8
p Ptr Word8 -> Int -> Ptr Word8
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
1) Ptr Word8
end)
            -- Using a separate state here generates a jump to a separate code
            -- block in the core which seems to perform slightly better for the
            -- non-ascii case.
            Bool
True -> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$ FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (FlattenState s Word8 -> Step (FlattenState s Word8) Char)
-> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall a b. (a -> b) -> a -> b
$ s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> Word8
-> FlattenState s Word8
forall s a.
s -> ForeignPtr a -> Ptr a -> Ptr a -> Word8 -> FlattenState s a
InnerLoopDecodeFirst s
st ForeignPtr Word8
startf Ptr Word8
p Ptr Word8
end Word8
x

    step' Ptr Word8
table State Stream m a
_ (InnerLoopDecodeFirst s
st ForeignPtr Word8
startf Ptr Word8
p Ptr Word8
end Word8
x) = do
        let (Tuple' Word8
sv Int
cp) = Ptr Word8 -> Word8 -> Tuple' Word8 Int
decode0 Ptr Word8
table Word8
x
        Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$
            case Word8
sv of
                Word8
12 ->
                    FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (FlattenState s Word8 -> Step (FlattenState s Word8) Char)
-> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall a b. (a -> b) -> a -> b
$
                    [Char] -> FlattenState s Word8 -> FlattenState s Word8
forall s a. [Char] -> FlattenState s a -> FlattenState s a
transliterateOrError
                        (
                           [Char]
"Streamly.Internal.Data.Stream.StreamD."
                        [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
"decodeUtf8ArraysWith: Invalid UTF8"
                        [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
" codepoint encountered"
                        )
                        (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> FlattenState s Word8
forall s a. s -> ForeignPtr a -> Ptr a -> Ptr a -> FlattenState s a
InnerLoopDecodeInit s
st ForeignPtr Word8
startf (Ptr Word8
p Ptr Word8 -> Int -> Ptr Word8
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
1) Ptr Word8
end)
                Word8
0 -> [Char] -> Step (FlattenState s Word8) Char
forall a. HasCallStack => [Char] -> a
error [Char]
"unreachable state"
                Word8
_ -> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> Word8
-> Int
-> FlattenState s Word8
forall s a.
s
-> ForeignPtr a
-> Ptr a
-> Ptr a
-> Word8
-> Int
-> FlattenState s a
InnerLoopDecoding s
st ForeignPtr Word8
startf (Ptr Word8
p Ptr Word8 -> Int -> Ptr Word8
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
1) Ptr Word8
end Word8
sv Int
cp)
    step' Ptr Word8
_ State Stream m a
_ (InnerLoopDecoding s
st ForeignPtr Word8
startf Ptr Word8
p Ptr Word8
end Word8
sv Int
cp)
        | Ptr Word8
p Ptr Word8 -> Ptr Word8 -> Bool
forall a. Eq a => a -> a -> Bool
== Ptr Word8
end = do
            IO () -> m ()
forall (m :: * -> *) a. MonadIO m => IO a -> m a
liftIO (IO () -> m ()) -> IO () -> m ()
forall a b. (a -> b) -> a -> b
$ ForeignPtr Word8 -> IO ()
forall a. ForeignPtr a -> IO ()
touchForeignPtr ForeignPtr Word8
startf
            Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$ FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (FlattenState s Word8 -> Step (FlattenState s Word8) Char)
-> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall a b. (a -> b) -> a -> b
$ s -> Maybe (Word8, Int) -> FlattenState s Word8
forall s a. s -> Maybe (Word8, Int) -> FlattenState s a
OuterLoop s
st ((Word8, Int) -> Maybe (Word8, Int)
forall a. a -> Maybe a
Just (Word8
sv, Int
cp))
    step' Ptr Word8
table State Stream m a
_ (InnerLoopDecoding s
st ForeignPtr Word8
startf Ptr Word8
p Ptr Word8
end Word8
statePtr Int
codepointPtr) = do
        Word8
x <- IO Word8 -> m Word8
forall (m :: * -> *) a. MonadIO m => IO a -> m a
liftIO (IO Word8 -> m Word8) -> IO Word8 -> m Word8
forall a b. (a -> b) -> a -> b
$ Ptr Word8 -> IO Word8
forall a. Storable a => Ptr a -> IO a
peek Ptr Word8
p
        let (Tuple' Word8
sv Int
cp) = Ptr Word8 -> Word8 -> Int -> Word8 -> Tuple' Word8 Int
decode1 Ptr Word8
table Word8
statePtr Int
codepointPtr Word8
x
        Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$
            case Word8
sv of
                Word8
0 ->
                    FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (FlattenState s Word8 -> Step (FlattenState s Word8) Char)
-> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall a b. (a -> b) -> a -> b
$
                    Char -> FlattenState s Word8 -> FlattenState s Word8
forall s a. Char -> FlattenState s a -> FlattenState s a
YAndC
                        (Int -> Char
unsafeChr Int
cp)
                        (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> FlattenState s Word8
forall s a. s -> ForeignPtr a -> Ptr a -> Ptr a -> FlattenState s a
InnerLoopDecodeInit s
st ForeignPtr Word8
startf (Ptr Word8
p Ptr Word8 -> Int -> Ptr Word8
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
1) Ptr Word8
end)
                Word8
12 ->
                    FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip (FlattenState s Word8 -> Step (FlattenState s Word8) Char)
-> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall a b. (a -> b) -> a -> b
$
                    [Char] -> FlattenState s Word8 -> FlattenState s Word8
forall s a. [Char] -> FlattenState s a -> FlattenState s a
transliterateOrError
                        (
                           [Char]
"Streamly.Internal.Data.Stream.StreamD."
                        [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
"decodeUtf8ArraysWith: Invalid UTF8"
                        [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
" codepoint encountered"
                        )
                        (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> FlattenState s Word8
forall s a. s -> ForeignPtr a -> Ptr a -> Ptr a -> FlattenState s a
InnerLoopDecodeInit s
st ForeignPtr Word8
startf (Ptr Word8
p Ptr Word8 -> Int -> Ptr Word8
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
1) Ptr Word8
end)
                Word8
_ ->
                    FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. s -> Step s a
Skip
                    (s
-> ForeignPtr Word8
-> Ptr Word8
-> Ptr Word8
-> Word8
-> Int
-> FlattenState s Word8
forall s a.
s
-> ForeignPtr a
-> Ptr a
-> Ptr a
-> Word8
-> Int
-> FlattenState s a
InnerLoopDecoding s
st ForeignPtr Word8
startf (Ptr Word8
p Ptr Word8 -> Int -> Ptr Word8
forall a b. Ptr a -> Int -> Ptr b
`plusPtr` Int
1) Ptr Word8
end Word8
sv Int
cp)
    step' Ptr Word8
_ State Stream m a
_ (YAndC Char
c FlattenState s Word8
s) = Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (FlattenState s Word8) Char
 -> m (Step (FlattenState s Word8) Char))
-> Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall a b. (a -> b) -> a -> b
$ Char -> FlattenState s Word8 -> Step (FlattenState s Word8) Char
forall s a. a -> s -> Step s a
Yield Char
c FlattenState s Word8
s
    step' Ptr Word8
_ State Stream m a
_ FlattenState s Word8
D = Step (FlattenState s Word8) Char
-> m (Step (FlattenState s Word8) Char)
forall (m :: * -> *) a. Monad m => a -> m a
return Step (FlattenState s Word8) Char
forall s a. Step s a
Stop

{-# INLINE decodeUtf8ArraysD #-}
decodeUtf8ArraysD ::
       MonadIO m
    => Stream m (A.Array Word8)
    -> Stream m Char
decodeUtf8ArraysD :: Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysD = CodingFailureMode -> Stream m (Array Word8) -> Stream m Char
forall (m :: * -> *).
MonadIO m =>
CodingFailureMode -> Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysWithD CodingFailureMode
TransliterateCodingFailure

-- |
--
-- /Pre-release/
{-# INLINE decodeUtf8Arrays #-}
decodeUtf8Arrays ::
       (MonadIO m, IsStream t) => t m (Array Word8) -> t m Char
decodeUtf8Arrays :: t m (Array Word8) -> t m Char
decodeUtf8Arrays =
    Stream m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Char -> t m Char)
-> (t m (Array Word8) -> Stream m Char)
-> t m (Array Word8)
-> t m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m (Array Word8) -> Stream m Char
forall (m :: * -> *).
MonadIO m =>
Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysD (Stream m (Array Word8) -> Stream m Char)
-> (t m (Array Word8) -> Stream m (Array Word8))
-> t m (Array Word8)
-> Stream m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m (Array Word8) -> Stream m (Array Word8)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

{-# INLINE decodeUtf8ArraysD' #-}
decodeUtf8ArraysD' ::
       MonadIO m
    => Stream m (A.Array Word8)
    -> Stream m Char
decodeUtf8ArraysD' :: Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysD' = CodingFailureMode -> Stream m (Array Word8) -> Stream m Char
forall (m :: * -> *).
MonadIO m =>
CodingFailureMode -> Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysWithD CodingFailureMode
ErrorOnCodingFailure

-- |
--
-- /Pre-release/
{-# INLINE decodeUtf8Arrays' #-}
decodeUtf8Arrays' :: (MonadIO m, IsStream t) => t m (Array Word8) -> t m Char
decodeUtf8Arrays' :: t m (Array Word8) -> t m Char
decodeUtf8Arrays' = Stream m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Char -> t m Char)
-> (t m (Array Word8) -> Stream m Char)
-> t m (Array Word8)
-> t m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m (Array Word8) -> Stream m Char
forall (m :: * -> *).
MonadIO m =>
Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysD' (Stream m (Array Word8) -> Stream m Char)
-> (t m (Array Word8) -> Stream m (Array Word8))
-> t m (Array Word8)
-> Stream m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m (Array Word8) -> Stream m (Array Word8)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

{-# INLINE decodeUtf8ArraysD_ #-}
decodeUtf8ArraysD_ ::
       MonadIO m
    => Stream m (A.Array Word8)
    -> Stream m Char
decodeUtf8ArraysD_ :: Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysD_ = CodingFailureMode -> Stream m (Array Word8) -> Stream m Char
forall (m :: * -> *).
MonadIO m =>
CodingFailureMode -> Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysWithD CodingFailureMode
DropOnCodingFailure

-- |
--
-- /Pre-release/
{-# INLINE decodeUtf8Arrays_ #-}
decodeUtf8Arrays_ ::
       (MonadIO m, IsStream t) => t m (Array Word8) -> t m Char
decodeUtf8Arrays_ :: t m (Array Word8) -> t m Char
decodeUtf8Arrays_ =
    Stream m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Char -> t m Char)
-> (t m (Array Word8) -> Stream m Char)
-> t m (Array Word8)
-> t m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m (Array Word8) -> Stream m Char
forall (m :: * -> *).
MonadIO m =>
Stream m (Array Word8) -> Stream m Char
decodeUtf8ArraysD_ (Stream m (Array Word8) -> Stream m Char)
-> (t m (Array Word8) -> Stream m (Array Word8))
-> t m (Array Word8)
-> Stream m Char
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m (Array Word8) -> Stream m (Array Word8)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

-------------------------------------------------------------------------------
-- Encoding Unicode (UTF-8) Characters
-------------------------------------------------------------------------------

data WList = WCons !Word8 !WList | WNil

-- UTF-8 primitives, Lifted from GHC.IO.Encoding.UTF8.

{-# INLINE ord2 #-}
ord2 :: Char -> WList
ord2 :: Char -> WList
ord2 Char
c = Bool -> WList -> WList
forall a. HasCallStack => Bool -> a -> a
assert (Int
n Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
>= Int
0x80 Bool -> Bool -> Bool
&& Int
n Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
<= Int
0x07ff) (Word8 -> WList -> WList
WCons Word8
x1 (Word8 -> WList -> WList
WCons Word8
x2 WList
WNil))
  where
    n :: Int
n = Char -> Int
ord Char
c
    x1 :: Word8
x1 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ (Int
n Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftR` Int
6) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0xC0
    x2 :: Word8
x2 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ (Int
n Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Int
0x3F) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0x80

{-# INLINE ord3 #-}
ord3 :: Char -> WList
ord3 :: Char -> WList
ord3 Char
c = Bool -> WList -> WList
forall a. HasCallStack => Bool -> a -> a
assert (Int
n Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
>= Int
0x0800 Bool -> Bool -> Bool
&& Int
n Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
<= Int
0xffff) (Word8 -> WList -> WList
WCons Word8
x1 (Word8 -> WList -> WList
WCons Word8
x2 (Word8 -> WList -> WList
WCons Word8
x3 WList
WNil)))
  where
    n :: Int
n = Char -> Int
ord Char
c
    x1 :: Word8
x1 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ (Int
n Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftR` Int
12) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0xE0
    x2 :: Word8
x2 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ ((Int
n Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftR` Int
6) Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Int
0x3F) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0x80
    x3 :: Word8
x3 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ (Int
n Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Int
0x3F) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0x80

{-# INLINE ord4 #-}
ord4 :: Char -> WList
ord4 :: Char -> WList
ord4 Char
c = Bool -> WList -> WList
forall a. HasCallStack => Bool -> a -> a
assert (Int
n Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
>= Int
0x10000)  (Word8 -> WList -> WList
WCons Word8
x1 (Word8 -> WList -> WList
WCons Word8
x2 (Word8 -> WList -> WList
WCons Word8
x3 (Word8 -> WList -> WList
WCons Word8
x4 WList
WNil))))
  where
    n :: Int
n = Char -> Int
ord Char
c
    x1 :: Word8
x1 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ (Int
n Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftR` Int
18) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0xF0
    x2 :: Word8
x2 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ ((Int
n Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftR` Int
12) Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Int
0x3F) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0x80
    x3 :: Word8
x3 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ ((Int
n Int -> Int -> Int
forall a. Bits a => a -> Int -> a
`shiftR` Int
6) Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Int
0x3F) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0x80
    x4 :: Word8
x4 = Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Int -> Word8) -> Int -> Word8
forall a b. (a -> b) -> a -> b
$ (Int
n Int -> Int -> Int
forall a. Bits a => a -> a -> a
.&. Int
0x3F) Int -> Int -> Int
forall a. Num a => a -> a -> a
+ Int
0x80

{-# ANN type EncodeState Fuse #-}
data EncodeState s = EncodeState s !WList

{-# ANN type InvalidAction Fuse #-}
data InvalidAction =
    DropInvalid | ErrorInvalid | IgnoreInvalid | ReplaceInvalid

replaceInvalid :: s -> Step (EncodeState s) a
replaceInvalid :: s -> Step (EncodeState s) a
replaceInvalid s
s =
    EncodeState s -> Step (EncodeState s) a
forall s a. s -> Step s a
Skip (EncodeState s -> Step (EncodeState s) a)
-> EncodeState s -> Step (EncodeState s) a
forall a b. (a -> b) -> a -> b
$ s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s (Word8 -> WList -> WList
WCons Word8
239 (Word8 -> WList -> WList
WCons Word8
191 (Word8 -> WList -> WList
WCons Word8
189 WList
WNil)))

dropInvalid :: s -> Step (EncodeState s) a
dropInvalid :: s -> Step (EncodeState s) a
dropInvalid s
s = EncodeState s -> Step (EncodeState s) a
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s WList
WNil)

errorOnInvalid :: s -> Step (EncodeState s) a
errorOnInvalid :: s -> Step (EncodeState s) a
errorOnInvalid s
_ =
    [Char] -> Step (EncodeState s) a
forall a. HasCallStack => [Char] -> a
error ([Char] -> Step (EncodeState s) a)
-> [Char] -> Step (EncodeState s) a
forall a b. (a -> b) -> a -> b
$
    [Char] -> [Char]
forall a. Show a => a -> [Char]
show [Char]
"Streamly.Internal.Data.Stream.StreamD.encodeUtf8:"
    [Char] -> [Char] -> [Char]
forall a. [a] -> [a] -> [a]
++ [Char]
"Encountered a surrogate"

{-# INLINE_NORMAL encodeUtf8DGeneric #-}
encodeUtf8DGeneric ::
       Monad m
    => InvalidAction
    -> Stream m Char
    -> Stream m Word8
encodeUtf8DGeneric :: InvalidAction -> Stream m Char -> Stream m Word8
encodeUtf8DGeneric InvalidAction
act (Stream State Stream m Char -> s -> m (Step s Char)
step s
state) =
    (State Stream m Word8
 -> EncodeState s -> m (Step (EncodeState s) Word8))
-> EncodeState s -> Stream m Word8
forall (m :: * -> *) a s.
(State Stream m a -> s -> m (Step s a)) -> s -> Stream m a
Stream State Stream m Word8
-> EncodeState s -> m (Step (EncodeState s) Word8)
forall (m :: * -> *) a.
State Stream m a -> EncodeState s -> m (Step (EncodeState s) Word8)
step' (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
state WList
WNil)

    where

    {-# INLINE_LATE step' #-}
    step' :: State Stream m a -> EncodeState s -> m (Step (EncodeState s) Word8)
step' State Stream m a
gst (EncodeState s
st WList
WNil) = do
        Step s Char
r <- State Stream m Char -> s -> m (Step s Char)
step (State Stream m a -> State Stream m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a (n :: * -> *) b.
State t m a -> State t n b
adaptState State Stream m a
gst) s
st
        Step (EncodeState s) Word8 -> m (Step (EncodeState s) Word8)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (EncodeState s) Word8 -> m (Step (EncodeState s) Word8))
-> Step (EncodeState s) Word8 -> m (Step (EncodeState s) Word8)
forall a b. (a -> b) -> a -> b
$
            case Step s Char
r of
                Yield Char
c s
s ->
                    case Char -> Int
ord Char
c of
                        Int
x | Int
x Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
<= Int
0x7F ->
                                Word8 -> EncodeState s -> Step (EncodeState s) Word8
forall s a. a -> s -> Step s a
Yield (Int -> Word8
forall a b. (Integral a, Num b) => a -> b
fromIntegral Int
x) (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s WList
WNil)
                            | Int
x Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
<= Int
0x7FF -> EncodeState s -> Step (EncodeState s) Word8
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s (Char -> WList
ord2 Char
c))
                            | Int
x Int -> Int -> Bool
forall a. Ord a => a -> a -> Bool
<= Int
0xFFFF ->
                                case InvalidAction
act of
                                    InvalidAction
DropInvalid ->
                                        if Char -> Bool
isSurrogate Char
c
                                        then s -> Step (EncodeState s) Word8
forall s a. s -> Step (EncodeState s) a
dropInvalid s
s
                                        else EncodeState s -> Step (EncodeState s) Word8
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s (Char -> WList
ord3 Char
c))

                                    InvalidAction
ErrorInvalid ->
                                        if Char -> Bool
isSurrogate Char
c
                                        then s -> Step (EncodeState s) Word8
forall s a. s -> Step (EncodeState s) a
errorOnInvalid s
s
                                        else EncodeState s -> Step (EncodeState s) Word8
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s (Char -> WList
ord3 Char
c))

                                    InvalidAction
IgnoreInvalid ->
                                        EncodeState s -> Step (EncodeState s) Word8
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s (Char -> WList
ord3 Char
c))

                                    InvalidAction
ReplaceInvalid ->
                                        if Char -> Bool
isSurrogate Char
c
                                        then s -> Step (EncodeState s) Word8
forall s a. s -> Step (EncodeState s) a
replaceInvalid s
s
                                        else EncodeState s -> Step (EncodeState s) Word8
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s (Char -> WList
ord3 Char
c))

                            | Bool
otherwise -> EncodeState s -> Step (EncodeState s) Word8
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s (Char -> WList
ord4 Char
c))
                Skip s
s -> EncodeState s -> Step (EncodeState s) Word8
forall s a. s -> Step s a
Skip (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s WList
WNil)
                Step s Char
Stop -> Step (EncodeState s) Word8
forall s a. Step s a
Stop
    step' State Stream m a
_ (EncodeState s
s (WCons Word8
x WList
xs)) = Step (EncodeState s) Word8 -> m (Step (EncodeState s) Word8)
forall (m :: * -> *) a. Monad m => a -> m a
return (Step (EncodeState s) Word8 -> m (Step (EncodeState s) Word8))
-> Step (EncodeState s) Word8 -> m (Step (EncodeState s) Word8)
forall a b. (a -> b) -> a -> b
$ Word8 -> EncodeState s -> Step (EncodeState s) Word8
forall s a. a -> s -> Step s a
Yield Word8
x (s -> WList -> EncodeState s
forall s. s -> WList -> EncodeState s
EncodeState s
s WList
xs)

-- More yield points improve performance, but I am not sure if they can cause
-- too much code bloat or some trouble with fusion. So keeping only two yield
-- points for now, one for the ascii chars (fast path) and one for all other
-- paths (slow path).
{-# INLINE_NORMAL encodeUtf8D' #-}
encodeUtf8D' :: Monad m => Stream m Char -> Stream m Word8
encodeUtf8D' :: Stream m Char -> Stream m Word8
encodeUtf8D' = InvalidAction -> Stream m Char -> Stream m Word8
forall (m :: * -> *).
Monad m =>
InvalidAction -> Stream m Char -> Stream m Word8
encodeUtf8DGeneric InvalidAction
ErrorInvalid

-- | Encode a stream of Unicode characters to a UTF-8 encoded bytestream. When
-- any invalid character (U+D800-U+D8FF) is encountered in the input stream the
-- function errors out.
--
-- @since 0.8.0
{-# INLINE encodeUtf8' #-}
encodeUtf8' :: (Monad m, IsStream t) => t m Char -> t m Word8
encodeUtf8' :: t m Char -> t m Word8
encodeUtf8' = Stream m Word8 -> t m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Word8 -> t m Word8)
-> (t m Char -> Stream m Word8) -> t m Char -> t m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m Char -> Stream m Word8
forall (m :: * -> *). Monad m => Stream m Char -> Stream m Word8
encodeUtf8D' (Stream m Char -> Stream m Word8)
-> (t m Char -> Stream m Char) -> t m Char -> Stream m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Char -> Stream m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

-- | See section "3.9 Unicode Encoding Forms" in
-- https://www.unicode.org/versions/Unicode13.0.0/UnicodeStandard-13.0.pdf
--
{-# INLINE_NORMAL encodeUtf8D #-}
encodeUtf8D :: Monad m => Stream m Char -> Stream m Word8
encodeUtf8D :: Stream m Char -> Stream m Word8
encodeUtf8D = InvalidAction -> Stream m Char -> Stream m Word8
forall (m :: * -> *).
Monad m =>
InvalidAction -> Stream m Char -> Stream m Word8
encodeUtf8DGeneric InvalidAction
ReplaceInvalid

-- | Encode a stream of Unicode characters to a UTF-8 encoded bytestream. Any
-- Invalid characters (U+D800-U+D8FF) in the input stream are replaced by the
-- Unicode replacement character U+FFFD.
--
-- /Since: 0.7.0 ("Streamly.Data.Unicode.Stream")/
--
-- /Since: 0.8.0 (Lenient Behaviour)/
{-# INLINE encodeUtf8 #-}
encodeUtf8 :: (Monad m, IsStream t) => t m Char -> t m Word8
encodeUtf8 :: t m Char -> t m Word8
encodeUtf8 = Stream m Word8 -> t m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Word8 -> t m Word8)
-> (t m Char -> Stream m Word8) -> t m Char -> t m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m Char -> Stream m Word8
forall (m :: * -> *). Monad m => Stream m Char -> Stream m Word8
encodeUtf8D (Stream m Char -> Stream m Word8)
-> (t m Char -> Stream m Char) -> t m Char -> Stream m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Char -> Stream m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

{-# INLINE_NORMAL encodeUtf8D_ #-}
encodeUtf8D_ :: Monad m => Stream m Char -> Stream m Word8
encodeUtf8D_ :: Stream m Char -> Stream m Word8
encodeUtf8D_  = InvalidAction -> Stream m Char -> Stream m Word8
forall (m :: * -> *).
Monad m =>
InvalidAction -> Stream m Char -> Stream m Word8
encodeUtf8DGeneric InvalidAction
DropInvalid

-- | Encode a stream of Unicode characters to a UTF-8 encoded bytestream. Any
-- Invalid characters (U+D800-U+D8FF) in the input stream are dropped.
--
-- @since 0.8.0
{-# INLINE encodeUtf8_ #-}
encodeUtf8_ :: (Monad m, IsStream t) => t m Char -> t m Word8
encodeUtf8_ :: t m Char -> t m Word8
encodeUtf8_ = Stream m Word8 -> t m Word8
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
Stream m a -> t m a
D.fromStreamD (Stream m Word8 -> t m Word8)
-> (t m Char -> Stream m Word8) -> t m Char -> t m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Stream m Char -> Stream m Word8
forall (m :: * -> *). Monad m => Stream m Char -> Stream m Word8
encodeUtf8D_ (Stream m Char -> Stream m Word8)
-> (t m Char -> Stream m Char) -> t m Char -> Stream m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. t m Char -> Stream m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
t m a -> Stream m a
D.toStreamD

-- | Same as 'encodeUtf8'
--
{-# DEPRECATED encodeUtf8Lax "Please use 'encodeUtf8' instead" #-}
{-# INLINE encodeUtf8Lax #-}
encodeUtf8Lax :: (IsStream t, Monad m) => t m Char -> t m Word8
encodeUtf8Lax :: t m Char -> t m Word8
encodeUtf8Lax = t m Char -> t m Word8
forall (m :: * -> *) (t :: (* -> *) -> * -> *).
(Monad m, IsStream t) =>
t m Char -> t m Word8
encodeUtf8

-------------------------------------------------------------------------------
-- Encode streams of containers
-------------------------------------------------------------------------------

-- | Encode a container to @Array Word8@ provided an unfold to covert it to a
-- Char stream and an encoding function.
--
-- /Internal/
{-# INLINE encodeObject #-}
encodeObject :: MonadIO m =>
       (SerialT m Char -> SerialT m Word8)
    -> Unfold m a Char
    -> a
    -> m (Array Word8)
encodeObject :: (SerialT m Char -> SerialT m Word8)
-> Unfold m a Char -> a -> m (Array Word8)
encodeObject SerialT m Char -> SerialT m Word8
encode Unfold m a Char
u = Fold m Word8 (Array Word8) -> SerialT m Word8 -> m (Array Word8)
forall (m :: * -> *) a b.
Monad m =>
Fold m a b -> SerialT m a -> m b
S.fold Fold m Word8 (Array Word8)
forall (m :: * -> *) a.
(MonadIO m, Storable a) =>
Fold m a (Array a)
Array.write (SerialT m Word8 -> m (Array Word8))
-> (a -> SerialT m Word8) -> a -> m (Array Word8)
forall b c a. (b -> c) -> (a -> b) -> a -> c
. SerialT m Char -> SerialT m Word8
encode (SerialT m Char -> SerialT m Word8)
-> (a -> SerialT m Char) -> a -> SerialT m Word8
forall b c a. (b -> c) -> (a -> b) -> a -> c
. Unfold m a Char -> a -> SerialT m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
Unfold m a b -> a -> t m b
S.unfold Unfold m a Char
u

-- | Encode a stream of container objects using the supplied encoding scheme.
-- Each object is encoded as an @Array Word8@.
--
-- /Internal/
{-# INLINE encodeObjects #-}
encodeObjects :: (MonadIO m, IsStream t) =>
       (SerialT m Char -> SerialT m Word8)
    -> Unfold m a Char
    -> t m a
    -> t m (Array Word8)
encodeObjects :: (SerialT m Char -> SerialT m Word8)
-> Unfold m a Char -> t m a -> t m (Array Word8)
encodeObjects SerialT m Char -> SerialT m Word8
encode Unfold m a Char
u = (a -> m (Array Word8)) -> t m a -> t m (Array Word8)
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
(a -> m b) -> t m a -> t m b
Serial.mapM ((SerialT m Char -> SerialT m Word8)
-> Unfold m a Char -> a -> m (Array Word8)
forall (m :: * -> *) a.
MonadIO m =>
(SerialT m Char -> SerialT m Word8)
-> Unfold m a Char -> a -> m (Array Word8)
encodeObject SerialT m Char -> SerialT m Word8
encode Unfold m a Char
u)

-- | Encode a stream of 'String' using the supplied encoding scheme. Each
-- string is encoded as an @Array Word8@.
--
-- @since 0.8.0
{-# INLINE encodeStrings #-}
encodeStrings :: (MonadIO m, IsStream t) =>
    (SerialT m Char -> SerialT m Word8) -> t m String -> t m (Array Word8)
encodeStrings :: (SerialT m Char -> SerialT m Word8)
-> t m [Char] -> t m (Array Word8)
encodeStrings SerialT m Char -> SerialT m Word8
encode = (SerialT m Char -> SerialT m Word8)
-> Unfold m [Char] Char -> t m [Char] -> t m (Array Word8)
forall (m :: * -> *) (t :: (* -> *) -> * -> *) a.
(MonadIO m, IsStream t) =>
(SerialT m Char -> SerialT m Word8)
-> Unfold m a Char -> t m a -> t m (Array Word8)
encodeObjects SerialT m Char -> SerialT m Word8
encode Unfold m [Char] Char
forall (m :: * -> *) a. Monad m => Unfold m [a] a
Unfold.fromList

{-
-------------------------------------------------------------------------------
-- Utility operations on strings
-------------------------------------------------------------------------------

strip :: IsStream t => t m Char -> t m Char
strip = undefined

stripTail :: IsStream t => t m Char -> t m Char
stripTail = undefined
-}

-- | Remove leading whitespace from a string.
--
-- > stripHead = S.dropWhile isSpace
--
-- /Pre-release/
{-# INLINE stripHead #-}
stripHead :: (Monad m, IsStream t) => t m Char -> t m Char
stripHead :: t m Char -> t m Char
stripHead = (Char -> Bool) -> t m Char -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a.
(IsStream t, Monad m) =>
(a -> Bool) -> t m a -> t m a
S.dropWhile Char -> Bool
isSpace

-- | Fold each line of the stream using the supplied 'Fold'
-- and stream the result.
--
-- >>> Stream.toList $ lines Fold.toList (Stream.fromList "lines\nthis\nstring\n\n\n")
-- ["lines","this","string","",""]
--
-- > lines = S.splitOnSuffix (== '\n')
--
-- /Pre-release/
{-# INLINE lines #-}
lines :: (Monad m, IsStream t) => Fold m Char b -> t m Char -> t m b
lines :: Fold m Char b -> t m Char -> t m b
lines = (Char -> Bool) -> Fold m Char b -> t m Char -> t m b
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
(a -> Bool) -> Fold m a b -> t m a -> t m b
S.splitOnSuffix (Char -> Char -> Bool
forall a. Eq a => a -> a -> Bool
== Char
'\n')

foreign import ccall unsafe "u_iswspace"
  iswspace :: Int -> Int

-- | Code copied from base/Data.Char to INLINE it
{-# INLINE isSpace #-}
isSpace :: Char -> Bool
isSpace :: Char -> Bool
isSpace Char
c
  | Word
uc Word -> Word -> Bool
forall a. Ord a => a -> a -> Bool
<= Word
0x377 = Word
uc Word -> Word -> Bool
forall a. Eq a => a -> a -> Bool
== Word
32 Bool -> Bool -> Bool
|| Word
uc Word -> Word -> Word
forall a. Num a => a -> a -> a
- Word
0x9 Word -> Word -> Bool
forall a. Ord a => a -> a -> Bool
<= Word
4 Bool -> Bool -> Bool
|| Word
uc Word -> Word -> Bool
forall a. Eq a => a -> a -> Bool
== Word
0xa0
  | Bool
otherwise = Int -> Int
iswspace (Char -> Int
ord Char
c) Int -> Int -> Bool
forall a. Eq a => a -> a -> Bool
/= Int
0
  where
    uc :: Word
uc = Int -> Word
forall a b. (Integral a, Num b) => a -> b
fromIntegral (Char -> Int
ord Char
c) :: Word

-- | Fold each word of the stream using the supplied 'Fold'
-- and stream the result.
--
-- >>>  Stream.toList $ words Fold.toList (Stream.fromList "fold these     words")
-- ["fold","these","words"]
--
-- > words = S.wordsBy isSpace
--
-- /Pre-release/
{-# INLINE words #-}
words :: (Monad m, IsStream t) => Fold m Char b -> t m Char -> t m b
words :: Fold m Char b -> t m Char -> t m b
words = (Char -> Bool) -> Fold m Char b -> t m Char -> t m b
forall (t :: (* -> *) -> * -> *) (m :: * -> *) a b.
(IsStream t, Monad m) =>
(a -> Bool) -> Fold m a b -> t m a -> t m b
S.wordsBy Char -> Bool
isSpace

-- | Unfold a stream to character streams using the supplied 'Unfold'
-- and concat the results suffixing a newline character @\\n@ to each stream.
--
-- @
-- unlines = Stream.interposeSuffix '\n'
-- unlines = Stream.intercalateSuffix Unfold.fromList "\n"
-- @
--
-- /Pre-release/
{-# INLINE unlines #-}
unlines :: (MonadIO m, IsStream t) => Unfold m a Char -> t m a -> t m Char
unlines :: Unfold m a Char -> t m a -> t m Char
unlines = Char -> Unfold m a Char -> t m a -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) c b.
(IsStream t, Monad m) =>
c -> Unfold m b c -> t m b -> t m c
S.interposeSuffix Char
'\n'

-- | Unfold the elements of a stream to character streams using the supplied
-- 'Unfold' and concat the results with a whitespace character infixed between
-- the streams.
--
-- @
-- unwords = Stream.interpose ' '
-- unwords = Stream.intercalate Unfold.fromList " "
-- @
--
-- /Pre-release/
{-# INLINE unwords #-}
unwords :: (MonadIO m, IsStream t) => Unfold m a Char -> t m a -> t m Char
unwords :: Unfold m a Char -> t m a -> t m Char
unwords = Char -> Unfold m a Char -> t m a -> t m Char
forall (t :: (* -> *) -> * -> *) (m :: * -> *) c b.
(IsStream t, Monad m) =>
c -> Unfold m b c -> t m b -> t m c
S.interpose Char
' '