1 {-# OPTIONS_GHC -cpp -fffi -fglasgow-exts #-}
4 -- Copyright : (c) The University of Glasgow 2001,
5 -- (c) David Roundy 2003-2005,
6 -- (c) Simon Marlow 2005
7 -- (c) Don Stewart 2005-2006
8 -- (c) Bjorn Bringert 2006
11 -- (c) 2001,2002 Manuel M T Chakravarty & Gabriele Keller
12 -- (c) 2006 Manuel M T Chakravarty & Roman Leshchinskiy
14 -- License : BSD-style
16 -- Maintainer : dons@cse.unsw.edu.au
17 -- Stability : experimental
18 -- Portability : portable, requires ffi and cpp
19 -- Tested with : GHC 6.4.1 and Hugs March 2005
23 -- | A time and space-efficient implementation of byte vectors using
24 -- packed Word8 arrays, suitable for high performance use, both in terms
25 -- of large data quantities, or high speed requirements. Byte vectors
26 -- are encoded as strict Word8 arrays of bytes, held in a ForeignPtr,
27 -- and can be passed between C and Haskell with little effort.
29 -- This module is intended to be imported @qualified@, to avoid name
30 -- clashes with Prelude functions. eg.
32 -- > import qualified Data.ByteString as B
34 -- Original GHC implementation by Bryan O\'Sullivan. Rewritten to use
35 -- UArray by Simon Marlow. Rewritten to support slices and use
36 -- ForeignPtr by David Roundy. Polished and extended by Don Stewart.
39 module Data.ByteString (
41 -- * The @ByteString@ type
42 ByteString(..), -- instances: Eq, Ord, Show, Read, Data, Typeable
44 -- * Introducing and eliminating 'ByteString's
45 empty, -- :: ByteString
46 packByte, -- :: Word8 -> ByteString
47 pack, -- :: [Word8] -> ByteString
48 unpack, -- :: ByteString -> [Word8]
49 packWith, -- :: (a -> Word8) -> [a] -> ByteString
50 unpackWith, -- :: (Word8 -> a) -> ByteString -> [a]
53 cons, -- :: Word8 -> ByteString -> ByteString
54 snoc, -- :: Word8 -> ByteString -> ByteString
55 null, -- :: ByteString -> Bool
56 length, -- :: ByteString -> Int
57 head, -- :: ByteString -> Word8
58 tail, -- :: ByteString -> ByteString
59 last, -- :: ByteString -> Word8
60 init, -- :: ByteString -> ByteString
61 append, -- :: ByteString -> ByteString -> ByteString
63 -- * Special ByteStrings
64 inits, -- :: ByteString -> [ByteString]
65 tails, -- :: ByteString -> [ByteString]
66 elems, -- :: ByteString -> [ByteString]
68 -- * Transformating ByteStrings
69 map, -- :: (Word8 -> Word8) -> ByteString -> ByteString
70 reverse, -- :: ByteString -> ByteString
71 intersperse, -- :: Word8 -> ByteString -> ByteString
72 transpose, -- :: [ByteString] -> [ByteString]
74 -- * Reducing 'ByteString's
75 foldl, -- :: (a -> Word8 -> a) -> a -> ByteString -> a
76 foldr, -- :: (Word8 -> a -> a) -> a -> ByteString -> a
77 foldl1, -- :: (Word8 -> Word8 -> Word8) -> ByteString -> Word8
78 foldr1, -- :: (Word8 -> Word8 -> Word8) -> ByteString -> Word8
81 concat, -- :: [ByteString] -> ByteString
82 concatMap, -- :: (Word8 -> ByteString) -> ByteString -> ByteString
83 any, -- :: (Word8 -> Bool) -> ByteString -> Bool
84 all, -- :: (Word8 -> Bool) -> ByteString -> Bool
85 maximum, -- :: ByteString -> Word8
86 minimum, -- :: ByteString -> Word8
87 mapIndexed, -- :: (Int -> Word8 -> Word8) -> ByteString -> ByteString
89 -- * Generating and unfolding ByteStrings
90 replicate, -- :: Int -> Word8 -> ByteString
91 unfoldrN, -- :: (Word8 -> Maybe (Word8, Word8)) -> Word8 -> ByteString
95 -- ** Breaking strings
96 take, -- :: Int -> ByteString -> ByteString
97 drop, -- :: Int -> ByteString -> ByteString
98 splitAt, -- :: Int -> ByteString -> (ByteString, ByteString)
99 takeWhile, -- :: (Word8 -> Bool) -> ByteString -> ByteString
100 dropWhile, -- :: (Word8 -> Bool) -> ByteString -> ByteString
101 break, -- :: (Word8 -> Bool) -> ByteString -> (ByteString, ByteString)
102 span, -- :: (Word8 -> Bool) -> ByteString -> (ByteString, ByteString)
103 spanEnd, -- :: (Word8 -> Bool) -> ByteString -> (ByteString, ByteString)
105 -- ** Breaking and dropping on specific bytes
106 breakByte, -- :: Word8 -> ByteString -> (ByteString, ByteString)
107 spanByte, -- :: Word8 -> ByteString -> (ByteString, ByteString)
108 breakFirst, -- :: Word8 -> ByteString -> Maybe (ByteString,ByteString)
109 breakLast, -- :: Word8 -> ByteString -> Maybe (ByteString,ByteString)
111 -- ** Breaking into many substrings
112 split, -- :: Word8 -> ByteString -> [ByteString]
113 splitWith, -- :: (Word8 -> Bool) -> ByteString -> [ByteString]
114 tokens, -- :: (Word8 -> Bool) -> ByteString -> [ByteString]
115 group, -- :: ByteString -> [ByteString]
116 groupBy, -- :: (Word8 -> Word8 -> Bool) -> ByteString -> [ByteString]
118 -- ** Joining strings
119 join, -- :: ByteString -> [ByteString] -> ByteString
120 joinWithByte, -- :: Word8 -> ByteString -> ByteString -> ByteString
122 -- * Indexing ByteStrings
123 index, -- :: ByteString -> Int -> Word8
124 elemIndex, -- :: Word8 -> ByteString -> Maybe Int
125 elemIndices, -- :: Word8 -> ByteString -> [Int]
126 elemIndexLast, -- :: Word8 -> ByteString -> Maybe Int
127 findIndex, -- :: (Word8 -> Bool) -> ByteString -> Maybe Int
128 findIndices, -- :: (Word8 -> Bool) -> ByteString -> [Int]
129 count, -- :: Word8 -> ByteString -> Int
131 -- * Ordered ByteStrings
132 sort, -- :: ByteString -> ByteString
134 -- * Searching ByteStrings
136 -- ** Searching by equality
137 -- | These functions use memchr(3) to efficiently search the ByteString
139 elem, -- :: Word8 -> ByteString -> Bool
140 notElem, -- :: Word8 -> ByteString -> Bool
141 filterByte, -- :: Word8 -> ByteString -> ByteString
142 filterNotByte, -- :: Word8 -> ByteString -> ByteString
144 -- ** Searching with a predicate
145 filter, -- :: (Word8 -> Bool) -> ByteString -> ByteString
146 find, -- :: (Word8 -> Bool) -> ByteString -> Maybe Word8
148 -- ** Prefixes and suffixes
149 -- | These functions use memcmp(3) to efficiently compare substrings
150 isPrefixOf, -- :: ByteString -> ByteString -> Bool
151 isSuffixOf, -- :: ByteString -> ByteString -> Bool
153 -- ** Search for arbitrary substrings
154 isSubstringOf, -- :: ByteString -> ByteString -> Bool
155 findSubstring, -- :: ByteString -> ByteString -> Maybe Int
156 findSubstrings, -- :: ByteString -> ByteString -> [Int]
158 -- * Zipping and unzipping ByteStrings
159 zip, -- :: ByteString -> ByteString -> [(Word8,Word8)]
160 zipWith, -- :: (Word8 -> Word8 -> c) -> ByteString -> ByteString -> [c]
161 unzip, -- :: [(Word8,Word8)] -> (ByteString,ByteString)
163 -- * Unchecked access
164 unsafeHead, -- :: ByteString -> Word8
165 unsafeTail, -- :: ByteString -> ByteString
166 unsafeIndex, -- :: ByteString -> Int -> Word8
168 -- * Low level introduction and elimination
169 generate, -- :: Int -> (Ptr Word8 -> IO Int) -> IO ByteString
170 create, -- :: Int -> (Ptr Word8 -> IO ()) -> ByteString
171 fromForeignPtr, -- :: ForeignPtr Word8 -> Int -> ByteString
172 toForeignPtr, -- :: ByteString -> (ForeignPtr Word8, Int, Int)
173 skipIndex, -- :: ByteString -> Int
175 -- ** Packing CStrings and pointers
176 packCString, -- :: CString -> ByteString
177 packCStringLen, -- :: CString -> ByteString
178 packMallocCString, -- :: CString -> ByteString
180 #if defined(__GLASGOW_HASKELL__)
181 packCStringFinalizer, -- :: Ptr Word8 -> Int -> IO () -> IO ByteString
182 packAddress, -- :: Addr# -> ByteString
183 unsafePackAddress, -- :: Int -> Addr# -> ByteString
184 unsafeFinalize, -- :: ByteString -> IO ()
187 -- ** Using ByteStrings as CStrings
188 useAsCString, -- :: ByteString -> (CString -> IO a) -> IO a
189 unsafeUseAsCString, -- :: ByteString -> (CString -> IO a) -> IO a
190 unsafeUseAsCStringLen, -- :: ByteString -> (CStringLen -> IO a) -> IO a
192 -- ** Copying ByteStrings
193 -- | These functions perform memcpy(3) operations
194 copy, -- :: ByteString -> ByteString
195 copyCString, -- :: CString -> ByteString
196 copyCStringLen, -- :: CStringLen -> ByteString
198 -- * I\/O with @ByteString@s
200 -- ** Standard input and output
202 #if defined(__GLASGOW_HASKELL__)
203 getLine, -- :: IO ByteString
205 getContents, -- :: IO ByteString
206 putStr, -- :: ByteString -> IO ()
207 putStrLn, -- :: ByteString -> IO ()
210 readFile, -- :: FilePath -> IO ByteString
211 writeFile, -- :: FilePath -> ByteString -> IO ()
212 -- mmapFile, -- :: FilePath -> IO ByteString
214 -- ** I\/O with Handles
215 #if defined(__GLASGOW_HASKELL__)
216 getArgs, -- :: IO [ByteString]
217 hGetLine, -- :: Handle -> IO ByteString
218 hGetNonBlocking, -- :: Handle -> Int -> IO ByteString
220 hGetContents, -- :: Handle -> IO ByteString
221 hGet, -- :: Handle -> Int -> IO ByteString
222 hPut, -- :: Handle -> ByteString -> IO ()
224 -- * Fusion utilities
225 #if defined(__GLASGOW_HASKELL__)
226 unpackList, -- eek, otherwise it gets thrown away by the simplifier
229 noAL, NoAL, loopArr, loopAcc, loopSndAcc,
230 loopU, mapEFL, filterEFL, foldEFL,
235 import qualified Prelude as P
236 import Prelude hiding (reverse,head,tail,last,init,null
237 ,length,map,lines,foldl,foldr,unlines
238 ,concat,any,take,drop,splitAt,takeWhile
239 ,dropWhile,span,break,elem,filter,maximum
240 ,minimum,all,concatMap,foldl1,foldr1
241 ,readFile,writeFile,replicate
242 ,getContents,getLine,putStr,putStrLn
243 ,zip,zipWith,unzip,notElem)
245 import qualified Data.List as List
248 import Data.Word (Word8)
249 import Data.Maybe (listToMaybe)
250 import Data.Array (listArray)
251 import qualified Data.Array as Array ((!))
253 -- Control.Exception.bracket not available in yhc or nhc
254 import Control.Exception (bracket)
255 import Control.Monad (when)
257 import Foreign.C.String (CString, CStringLen)
258 import Foreign.C.Types (CSize, CInt)
259 import Foreign.ForeignPtr
260 import Foreign.Marshal.Array
262 import Foreign.Storable (Storable(..))
264 -- hGetBuf and hPutBuf not available in yhc or nhc
265 import System.IO (stdin,stdout,hClose,hFileSize
266 ,hGetBuf,hPutBuf,openBinaryFile
269 #if !defined(__GLASGOW_HASKELL__)
270 import System.IO.Unsafe
273 #if defined(__GLASGOW_HASKELL__)
275 import Data.Generics (Data(..), Typeable(..))
277 import System.IO (hGetBufNonBlocking)
278 import System.IO.Error (isEOFError)
280 import Foreign.Marshal (alloca)
281 import qualified Foreign.Concurrent as FC (newForeignPtr)
284 import GHC.Prim (realWorld#, Addr#, Word#, (+#), writeWord8OffAddr#)
285 import GHC.Base (build, unsafeChr)
286 import GHC.Word hiding (Word8)
287 import GHC.Ptr (Ptr(..))
288 import GHC.ST (ST(..))
293 -- CFILES stuff is Hugs only
294 {-# CFILES cbits/fpstring.c #-}
296 -- -----------------------------------------------------------------------------
298 -- Useful macros, until we have bang patterns
301 #define STRICT1(f) f a | a `seq` False = undefined
302 #define STRICT2(f) f a b | a `seq` b `seq` False = undefined
303 #define STRICT3(f) f a b c | a `seq` b `seq` c `seq` False = undefined
304 #define STRICT4(f) f a b c d | a `seq` b `seq` c `seq` d `seq` False = undefined
305 #define STRICT5(f) f a b c d e | a `seq` b `seq` c `seq` d `seq` e `seq` False = undefined
307 -- -----------------------------------------------------------------------------
309 -- | A space-efficient representation of a Word8 vector, supporting many
310 -- efficient operations. A 'ByteString' contains 8-bit characters only.
312 -- Instances of Eq, Ord, Read, Show, Data, Typeable
314 data ByteString = PS {-# UNPACK #-} !(ForeignPtr Word8)
318 #if defined(__GLASGOW_HASKELL__)
319 deriving (Data, Typeable)
322 instance Eq ByteString
325 instance Ord ByteString
326 where compare = compareBytes
328 instance Show ByteString where
329 showsPrec p ps r = showsPrec p (unpackWith w2c ps) r
331 instance Read ByteString where
332 readsPrec p str = [ (packWith c2w x, y) | (x, y) <- readsPrec p str ]
335 instance Arbitrary PackedString where
336 arbitrary = P.pack `fmap` arbitrary
337 coarbitrary s = coarbitrary (P.unpack s)
340 -- | /O(n)/ Equality on the 'ByteString' type.
341 eq :: ByteString -> ByteString -> Bool
342 eq a@(PS p s l) b@(PS p' s' l')
343 | l /= l' = False -- short cut on length
344 | p == p' && s == s' = True -- short cut for the same string
345 | otherwise = compareBytes a b == EQ
348 -- | /O(n)/ 'compareBytes' provides an 'Ordering' for 'ByteStrings' supporting slices.
349 compareBytes :: ByteString -> ByteString -> Ordering
350 compareBytes (PS x1 s1 l1) (PS x2 s2 l2)
351 | l1 == 0 && l2 == 0 = EQ -- short cut for empty strings
352 | x1 == x2 && s1 == s2 && l1 == l2 = EQ -- short cut for the same string
353 | otherwise = inlinePerformIO $
354 withForeignPtr x1 $ \p1 ->
355 withForeignPtr x2 $ \p2 -> do
356 i <- memcmp (p1 `plusPtr` s1) (p2 `plusPtr` s2) (min l1 l2)
357 return $ case i `compare` 0 of
358 EQ -> l1 `compare` l2
360 {-# INLINE compareBytes #-}
364 -- About 4x slower over 32M
366 compareBytes :: ByteString -> ByteString -> Ordering
367 compareBytes (PS fp1 off1 len1) (PS fp2 off2 len2) = inlinePerformIO $
368 withForeignPtr fp1 $ \p1 ->
369 withForeignPtr fp2 $ \p2 ->
370 cmp (p1 `plusPtr` off1)
371 (p2 `plusPtr` off2) 0 len1 len2
373 cmp :: Ptr Word8 -> Ptr Word8 -> Int -> Int -> Int-> IO Ordering
375 cmp p1 p2 n len1 len2
376 | n == len1 = if n == len2 then return EQ else return LT
377 | n == len2 = return GT
379 (a :: Word8) <- peekByteOff p1 n
380 (b :: Word8) <- peekByteOff p2 n
381 case a `compare` b of
382 EQ -> cmp p1 p2 (n+1) len1 len2
385 {-# INLINE compareBytes #-}
388 -- -----------------------------------------------------------------------------
389 -- Introducing and eliminating 'ByteString's
391 -- | /O(1)/ The empty 'ByteString'
393 empty = inlinePerformIO $ mallocByteString 1 >>= \fp -> return $ PS fp 0 0
394 {-# NOINLINE empty #-}
396 -- | /O(1)/ Convert a 'Word8' into a 'ByteString'
397 packByte :: Word8 -> ByteString
398 packByte c = unsafePerformIO $ mallocByteString 2 >>= \fp -> do
399 withForeignPtr fp $ \p -> poke p c
401 {-# INLINE packByte #-}
404 -- XXX The unsafePerformIO is critical!
408 -- packByte 255 `compare` packByte 127
412 -- case mallocByteString 2 of
413 -- ForeignPtr f internals ->
414 -- case writeWord8OffAddr# f 0 255 of _ ->
415 -- case writeWord8OffAddr# f 0 127 of _ ->
416 -- case eqAddr# f f of
417 -- False -> case compare (GHC.Prim.plusAddr# f 0)
418 -- (GHC.Prim.plusAddr# f 0)
422 -- | /O(n)/ Convert a '[Word8]' into a 'ByteString'.
424 -- For applications with large numbers of string literals, pack can be a
425 -- bottleneck. In such cases, consider using packAddress (GHC only).
426 pack :: [Word8] -> ByteString
428 #if !defined(__GLASGOW_HASKELL__)
430 pack str = create (P.length str) $ \p -> go p str
433 go p (x:xs) = poke p x >> go (p `plusPtr` 1) xs -- less space than pokeElemOff
435 #else /* hack away */
437 pack str = create (P.length str) $ \(Ptr p) -> stToIO (go p 0# str)
439 go _ _ [] = return ()
440 go p i (W8# c:cs) = writeByte p i c >> go p (i +# 1#) cs
442 writeByte p i c = ST $ \s# ->
443 case writeWord8OffAddr# p i c s# of s2# -> (# s2#, () #)
447 -- | /O(n)/ Converts a 'ByteString' to a '[Word8]'.
448 unpack :: ByteString -> [Word8]
450 #if !defined(__GLASGOW_HASKELL__)
452 unpack (PS _ _ 0) = []
453 unpack (PS ps s l) = inlinePerformIO $ withForeignPtr ps $ \p ->
454 go (p `plusPtr` s) (l - 1) []
457 go p 0 acc = peek p >>= \e -> return (e : acc)
458 go p n acc = peekByteOff p n >>= \e -> go p (n-1) (e : acc)
459 {-# INLINE unpack #-}
463 unpack ps = build (unpackFoldr ps)
464 {-# INLINE unpack #-}
466 unpackList :: ByteString -> [Word8]
467 unpackList (PS fp off len) = withPtr fp $ \p -> do
469 loop _ (-1) acc = return acc
472 loop q (n-1) (a : acc)
473 loop (p `plusPtr` off) (len-1) []
476 "unpack-list" [1] forall p . unpackFoldr p (:) [] = unpackList p
479 unpackFoldr :: ByteString -> (Word8 -> a -> a) -> a -> a
480 unpackFoldr (PS fp off len) f ch = withPtr fp $ \p -> do
482 loop _ (-1) acc = return acc
485 loop q (n-1) (a `f` acc)
486 loop (p `plusPtr` off) (len-1) ch
487 {-# INLINE [0] unpackFoldr #-}
491 ------------------------------------------------------------------------
493 -- | /O(n)/ Convert a '[a]' into a 'ByteString' using some
494 -- conversion function
495 packWith :: (a -> Word8) -> [a] -> ByteString
496 packWith k str = create (P.length str) $ \p -> go p str
500 go p (x:xs) = poke p (k x) >> go (p `plusPtr` 1) xs -- less space than pokeElemOff
501 {-# INLINE packWith #-}
502 {-# SPECIALIZE packWith :: (Char -> Word8) -> [Char] -> ByteString #-}
504 -- | /O(n)/ Converts a 'ByteString' to a '[a]', using a conversion function.
505 unpackWith :: (Word8 -> a) -> ByteString -> [a]
506 unpackWith _ (PS _ _ 0) = []
507 unpackWith k (PS ps s l) = inlinePerformIO $ withForeignPtr ps $ \p ->
508 go (p `plusPtr` s) (l - 1) []
511 go p 0 acc = peek p >>= \e -> return (k e : acc)
512 go p n acc = peekByteOff p n >>= \e -> go p (n-1) (k e : acc)
513 {-# INLINE unpackWith #-}
514 {-# SPECIALIZE unpackWith :: (Word8 -> Char) -> ByteString -> [Char] #-}
516 -- ---------------------------------------------------------------------
519 -- | /O(1)/ Test whether a ByteString is empty.
520 null :: ByteString -> Bool
521 null (PS _ _ l) = l == 0
524 -- | /O(1)/ 'length' returns the length of a ByteString as an 'Int'.
525 length :: ByteString -> Int
526 length (PS _ _ l) = l
527 {-# INLINE length #-}
529 -- | /O(n)/ 'cons' is analogous to (:) for lists, but of different
530 -- complexity, as it requires a memcpy.
531 cons :: Word8 -> ByteString -> ByteString
532 cons c (PS x s l) = create (l+1) $ \p -> withForeignPtr x $ \f -> do
533 memcpy (p `plusPtr` 1) (f `plusPtr` s) l
539 -- | /O(n)/ Append a byte to the end of a 'ByteString'
540 snoc :: ByteString -> Word8 -> ByteString
541 snoc (PS x s l) c = create (l+1) $ \p -> withForeignPtr x $ \f -> do
542 memcpy p (f `plusPtr` s) l
543 poke (p `plusPtr` l) c
548 -- | /O(1)/ Extract the first element of a ByteString, which must be non-empty.
549 head :: ByteString -> Word8
551 | null ps = errorEmptyList "head"
552 | otherwise = inlinePerformIO $ withForeignPtr x $ \p -> peekByteOff p s
555 -- | /O(1)/ Extract the elements after the head of a ByteString, which must be non-empty.
556 tail :: ByteString -> ByteString
558 | l <= 0 = errorEmptyList "tail"
559 | otherwise = PS p (s+1) (l-1)
562 -- | /O(1)/ Extract the last element of a ByteString, which must be finite and non-empty.
563 last :: ByteString -> Word8
565 | null ps = errorEmptyList "last"
566 | otherwise = inlinePerformIO $ withForeignPtr x $ \p -> peekByteOff p (s+l-1)
569 -- | /O(1)/ Return all the elements of a 'ByteString' except the last one.
570 init :: ByteString -> ByteString
572 | l <= 0 = errorEmptyList "init"
573 | otherwise = PS p s (l-1)
576 -- | /O(n)/ Append two ByteStrings
577 append :: ByteString -> ByteString -> ByteString
578 append xs ys | null xs = ys
580 | otherwise = concat [xs,ys]
581 {-# INLINE append #-}
583 -- ---------------------------------------------------------------------
586 -- | /O(n)/ 'map' @f xs@ is the ByteString obtained by applying @f@ to each
587 -- element of @xs@. This function is subject to array fusion.
588 map :: (Word8 -> Word8) -> ByteString -> ByteString
589 map f = loopArr . loopU (mapEFL f) noAL
592 -- | /O(n)/ Like 'map', but not fuseable. The benefit is that it is
593 -- slightly faster for one-shot cases.
594 mapF :: (Word8 -> Word8) -> ByteString -> ByteString
596 mapF f (PS fp s len) = inlinePerformIO $ withForeignPtr fp $ \a -> do
597 np <- mallocByteString (len+1)
598 withForeignPtr np $ \p -> do
599 map_ 0 (a `plusPtr` s) p
602 map_ :: Int -> Ptr Word8 -> Ptr Word8 -> IO ()
605 | n >= len = return ()
607 x <- peekByteOff p1 n
608 pokeByteOff p2 n (f x)
612 -- | /O(n)/ 'reverse' @xs@ efficiently returns the elements of @xs@ in reverse order.
613 reverse :: ByteString -> ByteString
614 reverse (PS x s l) = create l $ \p -> withForeignPtr x $ \f ->
615 c_reverse p (f `plusPtr` s) l
618 reverse = pack . P.reverse . unpack
621 -- | /O(n)/ The 'intersperse' function takes a 'Word8' and a
622 -- 'ByteString' and \`intersperses\' that byte between the elements of
623 -- the 'ByteString'. It is analogous to the intersperse function on
625 intersperse :: Word8 -> ByteString -> ByteString
626 intersperse c ps@(PS x s l)
628 | otherwise = create (2*l-1) $ \p -> withForeignPtr x $ \f ->
629 c_intersperse p (f `plusPtr` s) l c
632 intersperse c = pack . List.intersperse c . unpack
635 -- | The 'transpose' function transposes the rows and columns of its
636 -- 'ByteString' argument.
637 transpose :: [ByteString] -> [ByteString]
638 transpose ps = P.map pack (List.transpose (P.map unpack ps))
640 -- ---------------------------------------------------------------------
641 -- Reducing 'ByteString's
643 -- | 'foldl', applied to a binary operator, a starting value (typically
644 -- the left-identity of the operator), and a ByteString, reduces the
645 -- ByteString using the binary operator, from left to right.
646 -- This function is subject to array fusion.
647 foldl :: (a -> Word8 -> a) -> a -> ByteString -> a
648 foldl f z = loopAcc . loopU (foldEFL f) z
653 -- About twice as fast with 6.4.1, but not fuseable
654 -- A simple fold . map is enough to make it worth while.
656 foldl f v (PS x s l) = inlinePerformIO $ withForeignPtr x $ \ptr ->
657 lgo v (ptr `plusPtr` s) (ptr `plusPtr` (s+l))
660 lgo z p q | p == q = return z
661 | otherwise = do c <- peek p
662 lgo (f z c) (p `plusPtr` 1) q
665 -- | 'foldr', applied to a binary operator, a starting value
666 -- (typically the right-identity of the operator), and a ByteString,
667 -- reduces the ByteString using the binary operator, from right to left.
668 foldr :: (Word8 -> a -> a) -> a -> ByteString -> a
669 foldr k z (PS x s l) = inlinePerformIO $ withForeignPtr x $ \ptr ->
670 go (ptr `plusPtr` s) (ptr `plusPtr` (s+l))
673 go p q | p == q = return z
674 | otherwise = do c <- peek p
675 ws <- go (p `plusPtr` 1) q
678 -- | 'foldl1' is a variant of 'foldl' that has no starting value
679 -- argument, and thus must be applied to non-empty 'ByteStrings'.
680 -- This function is subject to array fusion.
681 foldl1 :: (Word8 -> Word8 -> Word8) -> ByteString -> Word8
683 | null ps = errorEmptyList "foldl1"
684 | otherwise = foldl f (unsafeHead ps) (unsafeTail ps)
686 -- | 'foldr1' is a variant of 'foldr' that has no starting value argument,
687 -- and thus must be applied to non-empty 'ByteString's
688 foldr1 :: (Word8 -> Word8 -> Word8) -> ByteString -> Word8
690 | null ps = errorEmptyList "foldr1"
691 | otherwise = foldr f (last ps) (init ps)
693 -- ---------------------------------------------------------------------
696 -- | /O(n)/ Concatenate a list of ByteStrings.
697 concat :: [ByteString] -> ByteString
700 concat xs = create len $ \ptr -> go xs ptr
701 where len = P.sum . P.map length $ xs
704 go (PS p s l:ps) ptr = do
705 withForeignPtr p $ \fp -> memcpy ptr (fp `plusPtr` s) l
706 go ps (ptr `plusPtr` l)
708 -- | Map a function over a 'ByteString' and concatenate the results
709 concatMap :: (Word8 -> ByteString) -> ByteString -> ByteString
710 concatMap f = foldr (append . f) empty
712 -- | /O(n)/ Applied to a predicate and a ByteString, 'any' determines if
713 -- any element of the 'ByteString' satisfies the predicate.
714 any :: (Word8 -> Bool) -> ByteString -> Bool
715 any _ (PS _ _ 0) = False
716 any f (PS x s l) = inlinePerformIO $ withForeignPtr x $ \ptr ->
717 go (ptr `plusPtr` s) (ptr `plusPtr` (s+l))
720 go p q | p == q = return False
721 | otherwise = do c <- peek p
722 if f c then return True
723 else go (p `plusPtr` 1) q
727 -- | /O(n)/ Applied to a predicate and a 'ByteString', 'all' determines
728 -- if all elements of the 'ByteString' satisfy the predicate.
729 all :: (Word8 -> Bool) -> ByteString -> Bool
730 all _ (PS _ _ 0) = True
731 all f (PS x s l) = inlinePerformIO $ withForeignPtr x $ \ptr ->
732 go (ptr `plusPtr` s) (ptr `plusPtr` (s+l))
735 go p q | p == q = return True -- end of list
736 | otherwise = do c <- peek p
738 then go (p `plusPtr` 1) q
742 -- | /O(n)/ 'maximum' returns the maximum value from a 'ByteString'
743 maximum :: ByteString -> Word8
744 maximum xs@(PS x s l)
745 | null xs = errorEmptyList "maximum"
746 | otherwise = inlinePerformIO $ withForeignPtr x $ \p ->
747 return $ c_maximum (p `plusPtr` s) l
748 {-# INLINE maximum #-}
750 -- | /O(n)/ 'minimum' returns the minimum value from a 'ByteString'
751 minimum :: ByteString -> Word8
752 minimum xs@(PS x s l)
753 | null xs = errorEmptyList "minimum"
754 | otherwise = inlinePerformIO $ withForeignPtr x $ \p ->
755 return $ c_minimum (p `plusPtr` s) l
756 {-# INLINE minimum #-}
758 -- fusion is too slow here (10x)
761 maximum xs@(PS x s l)
762 | null xs = errorEmptyList "maximum"
763 | otherwise = inlinePerformIO $ withForeignPtr x $ \p -> do
765 maximum_ (p `plusPtr` s) 0 l w
767 maximum_ :: Ptr Word8 -> Int -> Int -> Word8 -> IO Word8
771 | otherwise = do w <- peekByteOff ptr n
772 maximum_ ptr (n+1) m (if w > c then w else c)
774 minimum xs@(PS x s l)
775 | null xs = errorEmptyList "minimum"
776 | otherwise = inlinePerformIO $ withForeignPtr x $ \p -> do
778 minimum_ (p `plusPtr` s) 0 l w
780 minimum_ :: Ptr Word8 -> Int -> Int -> Word8 -> IO Word8
784 | otherwise = do w <- peekByteOff ptr n
785 minimum_ ptr (n+1) m (if w < c then w else c)
788 -- | /O(n)/ map Word8 functions, provided with the index at each position
789 mapIndexed :: (Int -> Word8 -> Word8) -> ByteString -> ByteString
790 mapIndexed k (PS ps s l) = create l $ \p -> withForeignPtr ps $ \f ->
791 go 0 (f `plusPtr` s) p (f `plusPtr` s `plusPtr` l)
793 go :: Int -> Ptr Word8 -> Ptr Word8 -> Ptr Word8 -> IO ()
795 go n f t p | f == p = return ()
796 | otherwise = do w <- peek f
798 go (n+1) (f `plusPtr` 1) (t `plusPtr` 1) p
800 -- ---------------------------------------------------------------------
801 -- Unfolds and replicates
803 -- | /O(n)/ 'replicate' @n x@ is a ByteString of length @n@ with @x@
804 -- the value of every element. The following holds:
806 -- > replicate w c = unfoldr w (\u -> Just (u,u)) c
808 -- This implemenation uses @memset(3)@
809 replicate :: Int -> Word8 -> ByteString
810 replicate w c = create w $ \ptr -> memset ptr c (fromIntegral w) >> return ()
814 replicate w c = inlinePerformIO $ generate w $ \ptr -> go ptr w
818 go ptr n = poke ptr c >> go (ptr `plusPtr` 1) (n-1)
821 -- | /O(n)/ The 'unfoldrN' function is analogous to the List \'unfoldr\'.
822 -- 'unfoldrN' builds a ByteString from a seed value. The function takes
823 -- the element and returns 'Nothing' if it is done producing the
824 -- ByteString or returns 'Just' @(a,b)@, in which case, @a@ is a
825 -- prepending to the ByteString and @b@ is used as the next element in a
828 -- To preven unfoldrN having /O(n^2)/ complexity (as prepending a
829 -- character to a ByteString is /O(n)/, this unfoldr requires a maximum
830 -- final size of the ByteString as an argument. 'cons' can then be
831 -- implemented in /O(1)/ (i.e. a 'poke'), and the unfoldr itself has
832 -- linear complexity. The depth of the recursion is limited to this
833 -- size, but may be less. For lazy, infinite unfoldr, use
834 -- 'Data.List.unfoldr' (from 'Data.List').
838 -- > unfoldrN 10 (\x -> Just (x, chr (ord x + 1))) '0' == "0123456789"
840 -- The following equation connects the depth-limited unfoldr to the List unfoldr:
842 -- > unfoldrN n == take n $ List.unfoldr
843 unfoldrN :: Int -> (Word8 -> Maybe (Word8, Word8)) -> Word8 -> ByteString
844 unfoldrN i f w = inlinePerformIO $ generate i $ \p -> go p w 0
847 go q c n | n == i = return n -- stop if we reach `i'
848 | otherwise = case f c of
852 go (q `plusPtr` 1) new_c (n+1)
854 -- ---------------------------------------------------------------------
857 -- | /O(1)/ 'take' @n@, applied to a ByteString @xs@, returns the prefix
858 -- of @xs@ of length @n@, or @xs@ itself if @n > 'length' xs@.
859 take :: Int -> ByteString -> ByteString
863 | otherwise = PS x s n
866 -- | /O(1)/ 'drop' @n xs@ returns the suffix of @xs@ after the first @n@
867 -- elements, or @[]@ if @n > 'length' xs@.
868 drop :: Int -> ByteString -> ByteString
872 | otherwise = PS x (s+n) (l-n)
875 -- | /O(1)/ 'splitAt' @n xs@ is equivalent to @('take' n xs, 'drop' n xs)@.
876 splitAt :: Int -> ByteString -> (ByteString, ByteString)
877 splitAt n ps = (take n ps, drop n ps)
878 {-# INLINE splitAt #-}
880 -- | 'takeWhile', applied to a predicate @p@ and a ByteString @xs@,
881 -- returns the longest prefix (possibly empty) of @xs@ of elements that
883 takeWhile :: (Word8 -> Bool) -> ByteString -> ByteString
884 takeWhile f ps = take (findIndexOrEnd (not . f) ps) ps
885 {-# INLINE takeWhile #-}
887 -- | 'dropWhile' @p xs@ returns the suffix remaining after 'takeWhile' @p xs@.
888 dropWhile :: (Word8 -> Bool) -> ByteString -> ByteString
889 dropWhile f ps = drop (findIndexOrEnd (not . f) ps) ps
890 {-# INLINE dropWhile #-}
892 -- | 'break' @p@ is equivalent to @'span' ('not' . p)@.
893 break :: (Word8 -> Bool) -> ByteString -> (ByteString, ByteString)
894 break p ps = case findIndexOrEnd p ps of n -> (take n ps, drop n ps)
897 -- | 'breakByte' breaks its ByteString argument at the first occurence
898 -- of the specified byte. It is more efficient than 'break' as it is
899 -- implemented with @memchr(3)@. I.e.
901 -- > break (=='c') "abcd" == breakByte 'c' "abcd"
903 breakByte :: Word8 -> ByteString -> (ByteString, ByteString)
904 breakByte c p = case elemIndex c p of
906 Just n -> (take n p, drop n p)
907 {-# INLINE breakByte #-}
909 -- | 'spanByte' breaks its ByteString argument at the first
910 -- occurence of a byte other than its argument. It is more efficient
913 -- > span (=='c') "abcd" == spanByte 'c' "abcd"
915 spanByte :: Word8 -> ByteString -> (ByteString, ByteString)
916 spanByte c ps@(PS x s l) = inlinePerformIO $ withForeignPtr x $ \p ->
920 go p i | i >= l = return (ps, empty)
921 | otherwise = do c' <- peekByteOff p i
923 then return (take i ps, drop i ps)
925 {-# INLINE spanByte #-}
927 -- | /O(n)/ 'breakFirst' breaks the given ByteString on the first
928 -- occurence of @w@. It behaves like 'break', except the delimiter is
929 -- not returned, and @Nothing@ is returned if the delimiter is not in
930 -- the ByteString. I.e.
932 -- > breakFirst 'b' "aabbcc" == Just ("aa","bcc")
934 -- > breakFirst c xs ==
935 -- > let (x,y) = break (== c) xs
936 -- > in if null y then Nothing else Just (x, drop 1 y))
938 breakFirst :: Word8 -> ByteString -> Maybe (ByteString,ByteString)
939 breakFirst c p = case elemIndex c p of
941 Just n -> Just (take n p, drop (n+1) p)
942 {-# INLINE breakFirst #-}
944 -- | /O(n)/ 'breakLast' behaves like breakFirst, but from the end of the
947 -- > breakLast ('b') (pack "aabbcc") == Just ("aab","cc")
949 -- and the following are equivalent:
951 -- > breakLast 'c' "abcdef"
952 -- > let (x,y) = break (=='c') (reverse "abcdef")
953 -- > in if null x then Nothing else Just (reverse (drop 1 y), reverse x)
955 breakLast :: Word8 -> ByteString -> Maybe (ByteString,ByteString)
956 breakLast c p = case elemIndexLast c p of
958 Just n -> Just (take n p, drop (n+1) p)
959 {-# INLINE breakLast #-}
961 -- | 'span' @p xs@ breaks the ByteString into two segments. It is
962 -- equivalent to @('takeWhile' p xs, 'dropWhile' p xs)@
963 span :: (Word8 -> Bool) -> ByteString -> (ByteString, ByteString)
964 span p ps = break (not . p) ps
967 -- | 'spanEnd' behaves like 'span' but from the end of the 'ByteString'.
970 -- > spanEnd (not.isSpace) "x y z" == ("x y ","z")
974 -- > spanEnd (not . isSpace) ps
976 -- > let (x,y) = span (not.isSpace) (reverse ps) in (reverse y, reverse x)
978 spanEnd :: (Word8 -> Bool) -> ByteString -> (ByteString, ByteString)
979 spanEnd p ps = splitAt (findFromEndUntil (not.p) ps) ps
981 -- | /O(n)/ Splits a 'ByteString' into components delimited by
982 -- separators, where the predicate returns True for a separator element.
983 -- The resulting components do not contain the separators. Two adjacent
984 -- separators result in an empty component in the output. eg.
986 -- > splitWith (=='a') "aabbaca" == ["","","bb","c",""]
987 -- > splitWith (=='a') [] == []
989 splitWith :: (Word8 -> Bool) -> ByteString -> [ByteString]
991 #if defined(__GLASGOW_HASKELL__)
992 splitWith _pred (PS _ _ 0) = []
993 splitWith pred_ (PS fp off len) = splitWith' pred# off len fp
994 where pred# c# = pred_ (W8# c#)
996 splitWith' pred' off' len' fp' = withPtr fp $ \p ->
997 splitLoop pred' p 0 off' len' fp'
999 splitLoop :: (Word# -> Bool)
1001 -> Int -> Int -> Int
1005 splitLoop pred' p idx' off' len' fp'
1006 | pred' `seq` p `seq` idx' `seq` off' `seq` len' `seq` fp' `seq` False = undefined
1007 | idx' >= len' = return [PS fp' off' idx']
1009 w <- peekElemOff p (off'+idx')
1010 if pred' (case w of W8# w# -> w#)
1011 then return (PS fp' off' idx' :
1012 splitWith' pred' (off'+idx'+1) (len'-idx'-1) fp')
1013 else splitLoop pred' p (idx'+1) off' len' fp'
1014 {-# INLINE splitWith #-}
1017 splitWith _ (PS _ _ 0) = []
1018 splitWith p ps = splitWith' p ps
1021 splitWith' q qs = if null rest then [chunk]
1022 else chunk : splitWith' q (unsafeTail rest)
1023 where (chunk,rest) = break q qs
1026 -- | /O(n)/ Break a 'ByteString' into pieces separated by the byte
1027 -- argument, consuming the delimiter. I.e.
1029 -- > split '\n' "a\nb\nd\ne" == ["a","b","d","e"]
1030 -- > split 'a' "aXaXaXa" == ["","X","X","X"]
1031 -- > split 'x' "x" == ["",""]
1035 -- > join [c] . split c == id
1036 -- > split == splitWith . (==)
1038 -- As for all splitting functions in this library, this function does
1039 -- not copy the substrings, it just constructs new 'ByteStrings' that
1040 -- are slices of the original.
1042 split :: Word8 -> ByteString -> [ByteString]
1043 split _ (PS _ _ 0) = []
1044 split w (PS x s l) = inlinePerformIO $ withForeignPtr x $ \p -> do
1045 let ptr = p `plusPtr` s
1049 let q = memchr (ptr `plusPtr` n) w (fromIntegral (l-n))
1051 then return [PS x (s+n) (l-n)]
1052 else do let i = q `minusPtr` ptr
1054 return $! PS x (s+n) (i-n) : ls
1056 {-# INLINE split #-}
1059 -- slower. but stays inside Haskell.
1060 split _ (PS _ _ 0) = []
1061 split (W8# w#) (PS fp off len) = splitWith' off len fp
1063 splitWith' off' len' fp' = withPtr fp $ \p ->
1064 splitLoop p 0 off' len' fp'
1066 splitLoop :: Ptr Word8
1067 -> Int -> Int -> Int
1072 splitLoop p idx' off' len' fp'
1073 | p `seq` idx' `seq` off' `seq` len' `seq` fp' `seq` False = undefined
1074 | idx' >= len' = return [PS fp' off' idx']
1076 (W8# x#) <- peekElemOff p (off'+idx')
1077 if word2Int# w# ==# word2Int# x#
1078 then return (PS fp' off' idx' :
1079 splitWith' (off'+idx'+1) (len'-idx'-1) fp')
1080 else splitLoop p (idx'+1) off' len' fp'
1083 -- | Like 'splitWith', except that sequences of adjacent separators are
1084 -- treated as a single separator. eg.
1086 -- > tokens (=='a') "aabbaca" == ["bb","c"]
1088 tokens :: (Word8 -> Bool) -> ByteString -> [ByteString]
1089 tokens f = P.filter (not.null) . splitWith f
1091 -- | The 'group' function takes a ByteString and returns a list of
1092 -- ByteStrings such that the concatenation of the result is equal to the
1093 -- argument. Moreover, each sublist in the result contains only equal
1094 -- elements. For example,
1096 -- > group "Mississippi" = ["M","i","ss","i","ss","i","pp","i"]
1098 -- It is a special case of 'groupBy', which allows the programmer to
1099 -- supply their own equality test. It is about 40% faster than
1101 group :: ByteString -> [ByteString]
1104 | otherwise = ys : group zs
1106 (ys, zs) = spanByte (unsafeHead xs) xs
1108 -- | The 'groupBy' function is the non-overloaded version of 'group'.
1109 groupBy :: (Word8 -> Word8 -> Bool) -> ByteString -> [ByteString]
1112 | otherwise = take n xs : groupBy k (drop n xs)
1114 n = 1 + findIndexOrEnd (not . k (unsafeHead xs)) (unsafeTail xs)
1116 -- | /O(n)/ The 'join' function takes a 'ByteString' and a list of
1117 -- 'ByteString's and concatenates the list after interspersing the first
1118 -- argument between each element of the list.
1119 join :: ByteString -> [ByteString] -> ByteString
1120 join filler pss = concat (splice pss)
1124 splice (x:y:xs) = x:filler:splice (y:xs)
1127 -- | /O(n)/ joinWithByte. An efficient way to join to two ByteStrings
1128 -- with a char. Around 4 times faster than the generalised join.
1130 joinWithByte :: Word8 -> ByteString -> ByteString -> ByteString
1131 joinWithByte c f@(PS ffp s l) g@(PS fgp t m) = create len $ \ptr ->
1132 withForeignPtr ffp $ \fp ->
1133 withForeignPtr fgp $ \gp -> do
1134 memcpy ptr (fp `plusPtr` s) l
1135 poke (ptr `plusPtr` l) c
1136 memcpy (ptr `plusPtr` (l + 1)) (gp `plusPtr` t) m
1138 len = length f + length g + 1
1139 {-# INLINE joinWithByte #-}
1141 -- ---------------------------------------------------------------------
1142 -- Indexing ByteStrings
1144 -- | /O(1)/ 'ByteString' index (subscript) operator, starting from 0.
1145 index :: ByteString -> Int -> Word8
1147 | n < 0 = error $ "ByteString.indexWord8: negative index: " ++ show n
1148 | n >= length ps = error $ "ByteString.indexWord8: index too large: " ++ show n
1149 ++ ", length = " ++ show (length ps)
1150 | otherwise = ps `unsafeIndex` n
1151 {-# INLINE index #-}
1153 -- | /O(n)/ The 'elemIndex' function returns the index of the first
1154 -- element in the given 'ByteString' which is equal to the query
1155 -- element, or 'Nothing' if there is no such element.
1156 -- This implementation uses memchr(3).
1157 elemIndex :: Word8 -> ByteString -> Maybe Int
1158 elemIndex c (PS x s l) = inlinePerformIO $ withForeignPtr x $ \p -> do
1159 let p' = p `plusPtr` s
1160 q = memchr p' c (fromIntegral l)
1161 return $ if q == nullPtr then Nothing else Just $! q `minusPtr` p'
1162 {-# INLINE elemIndex #-}
1164 -- | /O(n)/ The 'elemIndexLast' function returns the last index of the
1165 -- element in the given 'ByteString' which is equal to the query
1166 -- element, or 'Nothing' if there is no such element. The following
1169 -- > elemIndexLast c xs ==
1170 -- > (-) (length xs - 1) `fmap` elemIndex c (reverse xs)
1172 elemIndexLast :: Word8 -> ByteString -> Maybe Int
1173 elemIndexLast ch (PS x s l) = inlinePerformIO $ withForeignPtr x $ \p ->
1174 go (p `plusPtr` s) (l-1)
1177 go p i | i < 0 = return Nothing
1178 | otherwise = do ch' <- peekByteOff p i
1180 then return $ Just i
1182 {-# INLINE elemIndexLast #-}
1184 -- | /O(n)/ The 'elemIndices' function extends 'elemIndex', by returning
1185 -- the indices of all elements equal to the query element, in ascending order.
1186 -- This implementation uses memchr(3).
1187 elemIndices :: Word8 -> ByteString -> [Int]
1188 elemIndices w (PS x s l) = inlinePerformIO $ withForeignPtr x $ \p -> do
1189 let ptr = p `plusPtr` s
1193 let q = memchr (ptr `plusPtr` n) w (fromIntegral (l - n))
1196 else do let i = q `minusPtr` ptr
1203 elemIndices :: Word8 -> ByteString -> [Int]
1204 elemIndices c ps = loop 0 ps
1206 loop _ ps' | null ps' = []
1207 loop n ps' | c == unsafeHead ps' = n : loop (n+1) (unsafeTail ps')
1208 | otherwise = loop (n+1) (unsafeTail ps')
1211 -- | count returns the number of times its argument appears in the ByteString
1213 -- > count = length . elemIndices
1215 -- But more efficiently than using length on the intermediate list.
1216 count :: Word8 -> ByteString -> Int
1217 count w (PS x s m) = inlinePerformIO $ withForeignPtr x $ \p ->
1218 return $ c_count (p `plusPtr` s) (fromIntegral m) w
1219 {-# INLINE count #-}
1223 -- around 30% slower
1225 count w (PS x s m) = inlinePerformIO $ withForeignPtr x $ \p ->
1226 go (p `plusPtr` s) (fromIntegral m) 0
1228 go :: Ptr Word8 -> CSize -> Int -> IO Int
1231 let q = memchr p w l
1234 else do let k = fromIntegral $ q `minusPtr` p
1235 go (q `plusPtr` 1) (l-k-1) (i+1)
1238 -- | The 'findIndex' function takes a predicate and a 'ByteString' and
1239 -- returns the index of the first element in the ByteString
1240 -- satisfying the predicate.
1241 findIndex :: (Word8 -> Bool) -> ByteString -> Maybe Int
1242 findIndex k ps@(PS x s l)
1244 | otherwise = inlinePerformIO $ withForeignPtr x $ \f -> go (f `plusPtr` s) 0
1247 go ptr n | n >= l = return Nothing
1248 | otherwise = do w <- peek ptr
1250 then return (Just n)
1251 else go (ptr `plusPtr` 1) (n+1)
1252 {-# INLINE findIndex #-}
1254 -- | The 'findIndices' function extends 'findIndex', by returning the
1255 -- indices of all elements satisfying the predicate, in ascending order.
1256 findIndices :: (Word8 -> Bool) -> ByteString -> [Int]
1257 findIndices p ps = loop 0 ps
1260 loop n qs | null qs = []
1261 | p (unsafeHead qs) = n : loop (n+1) (unsafeTail qs)
1262 | otherwise = loop (n+1) (unsafeTail qs)
1264 -- ---------------------------------------------------------------------
1265 -- Searching ByteStrings
1267 -- | /O(n)/ 'elem' is the 'ByteString' membership predicate.
1268 elem :: Word8 -> ByteString -> Bool
1269 elem c ps = case elemIndex c ps of Nothing -> False ; _ -> True
1272 -- | /O(n)/ 'notElem' is the inverse of 'elem'
1273 notElem :: Word8 -> ByteString -> Bool
1274 notElem c ps = not (elem c ps)
1275 {-# INLINE notElem #-}
1277 -- | /O(n)/ 'filter', applied to a predicate and a ByteString,
1278 -- returns a ByteString containing those characters that satisfy the
1279 -- predicate. This function is subject to array fusion.
1280 filter :: (Word8 -> Bool) -> ByteString -> ByteString
1281 filter p = loopArr . loopU (filterEFL p) noAL
1282 {-# INLINE filter #-}
1284 -- | /O(n)/ 'filterF' is a non-fuseable version of filter, that may be
1285 -- around 2x faster for some one-shot applications.
1286 filterF :: (Word8 -> Bool) -> ByteString -> ByteString
1287 filterF k ps@(PS x s l)
1289 | otherwise = inlinePerformIO $ generate l $ \p -> withForeignPtr x $ \f -> do
1290 t <- go (f `plusPtr` s) p (f `plusPtr` (s + l))
1291 return (t `minusPtr` p) -- actual length
1294 go f t end | f == end = return t
1298 then poke t w >> go (f `plusPtr` 1) (t `plusPtr` 1) end
1299 else go (f `plusPtr` 1) t end
1300 {-# INLINE filterF #-}
1303 -- | /O(n)/ A first order equivalent of /filter . (==)/, for the common
1304 -- case of filtering a single byte. It is more efficient to use
1305 -- /filterByte/ in this case.
1307 -- > filterByte == filter . (==)
1309 -- filterByte is around 10x faster, and uses much less space, than its
1310 -- filter equivalent
1311 filterByte :: Word8 -> ByteString -> ByteString
1312 filterByte w ps = replicate (count w ps) w
1313 {-# INLINE filterByte #-}
1316 -- | /O(n)/ A first order equivalent of /filter . (\/=)/, for the common
1317 -- case of filtering a single byte out of a list. It is more efficient
1318 -- to use /filterNotByte/ in this case.
1320 -- > filterNotByte == filter . (/=)
1322 -- filterNotByte is around 2x faster than its filter equivalent.
1323 filterNotByte :: Word8 -> ByteString -> ByteString
1324 filterNotByte w = filterF (/= w)
1325 {-# INLINE filterNotByte #-}
1327 -- | /O(n)/ The 'find' function takes a predicate and a ByteString,
1328 -- and returns the first element in matching the predicate, or 'Nothing'
1329 -- if there is no such element.
1331 -- > find f p = case findIndex f p of Just n -> Just (p ! n) ; _ -> Nothing
1333 find :: (Word8 -> Bool) -> ByteString -> Maybe Word8
1334 find f p = case findIndex f p of
1335 Just n -> Just (p `unsafeIndex` n)
1341 -- fuseable, but we don't want to walk the whole array.
1343 find k = foldl findEFL Nothing
1344 where findEFL a@(Just _) _ = a
1345 findEFL _ c | k c = Just c
1346 | otherwise = Nothing
1349 -- ---------------------------------------------------------------------
1350 -- Searching for substrings
1352 -- | /O(n)/ The 'isPrefixOf' function takes two ByteStrings and returns 'True'
1353 -- iff the first is a prefix of the second.
1354 isPrefixOf :: ByteString -> ByteString -> Bool
1355 isPrefixOf (PS x1 s1 l1) (PS x2 s2 l2)
1358 | otherwise = inlinePerformIO $ withForeignPtr x1 $ \p1 ->
1359 withForeignPtr x2 $ \p2 -> do
1360 i <- memcmp (p1 `plusPtr` s1) (p2 `plusPtr` s2) l1
1363 -- | /O(n)/ The 'isSuffixOf' function takes two ByteStrings and returns 'True'
1364 -- iff the first is a suffix of the second.
1366 -- The following holds:
1368 -- > isSuffixOf x y == reverse x `isPrefixOf` reverse y
1370 -- However, the real implemenation uses memcmp to compare the end of the
1371 -- string only, with no reverse required..
1372 isSuffixOf :: ByteString -> ByteString -> Bool
1373 isSuffixOf (PS x1 s1 l1) (PS x2 s2 l2)
1376 | otherwise = inlinePerformIO $ withForeignPtr x1 $ \p1 ->
1377 withForeignPtr x2 $ \p2 -> do
1378 i <- memcmp (p1 `plusPtr` s1) (p2 `plusPtr` s2 `plusPtr` (l2 - l1)) l1
1381 -- | Check whether one string is a substring of another. @isSubstringOf
1382 -- p s@ is equivalent to @not (null (findSubstrings p s))@.
1383 isSubstringOf :: ByteString -- ^ String to search for.
1384 -> ByteString -- ^ String to search in.
1386 isSubstringOf p s = not $ P.null $ findSubstrings p s
1388 -- | Get the first index of a substring in another string,
1389 -- or 'Nothing' if the string is not found.
1390 -- @findSubstring p s@ is equivalent to @listToMaybe (findSubstrings p s)@.
1391 findSubstring :: ByteString -- ^ String to search for.
1392 -> ByteString -- ^ String to seach in.
1394 findSubstring = (listToMaybe .) . findSubstrings
1396 -- | Find the indexes of all (possibly overlapping) occurances of a
1397 -- substring in a string. This function uses the Knuth-Morris-Pratt
1398 -- string matching algorithm.
1399 findSubstrings :: ByteString -- ^ String to search for.
1400 -> ByteString -- ^ String to seach in.
1403 findSubstrings pat@(PS _ _ m) str@(PS _ _ n) = search 0 0
1405 patc x = pat `unsafeIndex` x
1406 strc x = str `unsafeIndex` x
1408 -- maybe we should make kmpNext a UArray before using it in search?
1409 kmpNext = listArray (0,m) (-1:kmpNextL pat (-1))
1410 kmpNextL p _ | null p = []
1411 kmpNextL p j = let j' = next (unsafeHead p) j + 1
1413 x = if not (null ps) && unsafeHead ps == patc j'
1414 then kmpNext Array.! j' else j'
1416 search i j = match ++ rest -- i: position in string, j: position in pattern
1417 where match = if j == m then [(i - j)] else []
1418 rest = if i == n then [] else search (i+1) (next (strc i) j + 1)
1419 next c j | j >= 0 && (j == m || c /= patc j) = next c (kmpNext Array.! j)
1422 -- ---------------------------------------------------------------------
1425 -- | /O(n)/ 'zip' takes two ByteStrings and returns a list of
1426 -- corresponding pairs of bytes. If one input ByteString is short,
1427 -- excess elements of the longer ByteString are discarded. This is
1428 -- equivalent to a pair of 'unpack' operations.
1429 zip :: ByteString -> ByteString -> [(Word8,Word8)]
1431 | null ps || null qs = []
1432 | otherwise = (unsafeHead ps, unsafeHead qs) : zip (unsafeTail ps) (unsafeTail qs)
1434 -- | 'zipWith' generalises 'zip' by zipping with the function given as
1435 -- the first argument, instead of a tupling function. For example,
1436 -- @'zipWith' (+)@ is applied to two ByteStrings to produce the list of
1437 -- corresponding sums.
1438 zipWith :: (Word8 -> Word8 -> a) -> ByteString -> ByteString -> [a]
1440 | null ps || null qs = []
1441 | otherwise = f (unsafeHead ps) (unsafeHead qs) : zipWith f (unsafeTail ps) (unsafeTail qs)
1443 -- | /O(n)/ 'unzip' transforms a list of pairs of bytes into a pair of
1444 -- ByteStrings. Note that this performs two 'pack' operations.
1445 unzip :: [(Word8,Word8)] -> (ByteString,ByteString)
1446 unzip ls = (pack (P.map fst ls), pack (P.map snd ls))
1447 {-# INLINE unzip #-}
1449 -- ---------------------------------------------------------------------
1452 -- | /O(n)/ Return all initial segments of the given 'ByteString', shortest first.
1453 inits :: ByteString -> [ByteString]
1454 inits (PS x s l) = [PS x s n | n <- [0..l]]
1456 -- | /O(n)/ Return all final segments of the given 'ByteString', longest first.
1457 tails :: ByteString -> [ByteString]
1458 tails p | null p = [empty]
1459 | otherwise = p : tails (unsafeTail p)
1461 -- less efficent spacewise: tails (PS x s l) = [PS x (s+n) (l-n) | n <- [0..l]]
1463 -- | /O(n)/ breaks a ByteString to a list of ByteStrings, one byte each.
1464 elems :: ByteString -> [ByteString]
1465 elems (PS _ _ 0) = []
1466 elems (PS x s l) = (PS x s 1:elems (PS x (s+1) (l-1)))
1467 {-# INLINE elems #-}
1469 -- ---------------------------------------------------------------------
1470 -- ** Ordered 'ByteString's
1472 -- | /O(n)/ Sort a ByteString efficiently, using counting sort.
1473 sort :: ByteString -> ByteString
1474 sort (PS input s l) = create l $ \p -> allocaArray 256 $ \arr -> do
1476 memset (castPtr arr) 0 (256 * fromIntegral (sizeOf (undefined :: CSize)))
1477 withForeignPtr input (\x -> countEach arr (x `plusPtr` s) l)
1480 go 256 _ = return ()
1481 go i ptr = do n <- peekElemOff arr i
1482 when (n /= 0) $ memset ptr (fromIntegral i) n >> return ()
1483 go (i + 1) (ptr `plusPtr` (fromIntegral n))
1486 -- "countEach counts str l" counts the number of occurences of each Word8 in
1487 -- str, and stores the result in counts.
1488 countEach :: Ptr CSize -> Ptr Word8 -> Int -> IO ()
1490 countEach counts str l = go 0
1493 go i | i == l = return ()
1494 | otherwise = do k <- fromIntegral `fmap` peekElemOff str i
1495 x <- peekElemOff counts k
1496 pokeElemOff counts k (x + 1)
1500 sort :: ByteString -> ByteString
1501 sort (PS x s l) = create l $ \p -> withForeignPtr x $ \f -> do
1502 memcpy p (f `plusPtr` s) l
1503 c_qsort p l -- inplace
1507 sort = pack . List.sort . unpack
1510 -- | The 'sortBy' function is the non-overloaded version of 'sort'.
1512 -- Try some linear sorts: radix, counting
1515 -- sortBy :: (Word8 -> Word8 -> Ordering) -> ByteString -> ByteString
1516 -- sortBy f ps = undefined
1518 -- ---------------------------------------------------------------------
1520 -- Extensions to the basic interface
1523 -- | A variety of 'head' for non-empty ByteStrings. 'unsafeHead' omits the
1524 -- check for the empty case, so there is an obligation on the programmer
1525 -- to provide a proof that the ByteString is non-empty.
1526 unsafeHead :: ByteString -> Word8
1527 unsafeHead (PS x s _) = inlinePerformIO $ withForeignPtr x $ \p -> peekByteOff p s
1528 {-# INLINE unsafeHead #-}
1530 -- | A variety of 'tail' for non-empty ByteStrings. 'unsafeTail' omits the
1531 -- check for the empty case. As with 'unsafeHead', the programmer must
1532 -- provide a separate proof that the ByteString is non-empty.
1533 unsafeTail :: ByteString -> ByteString
1534 unsafeTail (PS ps s l) = PS ps (s+1) (l-1)
1535 {-# INLINE unsafeTail #-}
1537 -- | Unsafe 'ByteString' index (subscript) operator, starting from 0, returning a 'Word8'
1538 -- This omits the bounds check, which means there is an accompanying
1539 -- obligation on the programmer to ensure the bounds are checked in some
1541 unsafeIndex :: ByteString -> Int -> Word8
1542 unsafeIndex (PS x s _) i = inlinePerformIO $ withForeignPtr x $ \p -> peekByteOff p (s+i)
1543 {-# INLINE unsafeIndex #-}
1545 -- ---------------------------------------------------------------------
1546 -- Low level constructors
1548 #if defined(__GLASGOW_HASKELL__)
1549 -- | /O(n)/ Pack a null-terminated sequence of bytes, pointed to by an
1550 -- Addr\# (an arbitrary machine address assumed to point outside the
1551 -- garbage-collected heap) into a @ByteString@. A much faster way to
1552 -- create an Addr\# is with an unboxed string literal, than to pack a
1553 -- boxed string. A unboxed string literal is compiled to a static @char
1554 -- []@ by GHC. Establishing the length of the string requires a call to
1555 -- @strlen(3)@, so the Addr# must point to a null-terminated buffer (as
1556 -- is the case with "string"# literals in GHC). Use 'unsafePackAddress'
1557 -- if you know the length of the string statically.
1561 -- > literalFS = packAddress "literal"#
1563 packAddress :: Addr# -> ByteString
1564 packAddress addr# = inlinePerformIO $ do
1565 p <- newForeignPtr_ cstr
1566 return $ PS p 0 (fromIntegral $ c_strlen cstr)
1569 {-# INLINE packAddress #-}
1571 -- | /O(1)/ 'unsafePackAddress' provides constant-time construction of
1572 -- 'ByteStrings' -- which is ideal for string literals. It packs a
1573 -- null-terminated sequence of bytes into a 'ByteString', given a raw
1574 -- 'Addr\#' to the string, and the length of the string. Make sure the
1575 -- length is correct, otherwise use the safer 'packAddress' (where the
1576 -- length will be calculated once at runtime).
1577 unsafePackAddress :: Int -> Addr# -> ByteString
1578 unsafePackAddress len addr# = inlinePerformIO $ do
1579 p <- newForeignPtr_ cstr
1581 where cstr = Ptr addr#
1585 -- | /O(1)/ Build a ByteString from a ForeignPtr
1586 fromForeignPtr :: ForeignPtr Word8 -> Int -> ByteString
1587 fromForeignPtr fp l = PS fp 0 l
1589 -- | /O(1)/ Deconstruct a ForeignPtr from a ByteString
1590 toForeignPtr :: ByteString -> (ForeignPtr Word8, Int, Int)
1591 toForeignPtr (PS ps s l) = (ps, s, l)
1593 -- | /O(1)/ 'skipIndex' returns the internal skipped index of the
1594 -- current 'ByteString' from any larger string it was created from, as
1596 skipIndex :: ByteString -> Int
1597 skipIndex (PS _ s _) = s
1598 {-# INLINE skipIndex #-}
1600 -- | /O(n)/ Build a @ByteString@ from a @CString@. This value will have /no/
1601 -- finalizer associated to it. The ByteString length is calculated using
1602 -- /strlen(3)/, and thus the complexity is a /O(n)/.
1603 packCString :: CString -> ByteString
1604 packCString cstr = inlinePerformIO $ do
1605 fp <- newForeignPtr_ (castPtr cstr)
1606 return $ PS fp 0 (fromIntegral $ c_strlen cstr)
1608 -- | /O(1)/ Build a @ByteString@ from a @CStringLen@. This value will
1609 -- have /no/ finalizer associated with it. This operation has /O(1)/
1610 -- complexity as we already know the final size, so no /strlen(3)/ is
1612 packCStringLen :: CStringLen -> ByteString
1613 packCStringLen (ptr,len) = inlinePerformIO $ do
1614 fp <- newForeignPtr_ (castPtr ptr)
1615 return $ PS fp 0 (fromIntegral len)
1617 -- | /O(n)/ Build a @ByteString@ from a malloced @CString@. This value will
1618 -- have a @free(3)@ finalizer associated to it.
1619 packMallocCString :: CString -> ByteString
1620 packMallocCString cstr = inlinePerformIO $ do
1621 fp <- newForeignFreePtr (castPtr cstr)
1622 return $ PS fp 0 (fromIntegral $ c_strlen cstr)
1624 #if defined(__GLASGOW_HASKELL__)
1625 -- | /O(1)/ Construct a 'ByteString' given a C Ptr Word8 buffer, a
1626 -- length, and an IO action representing a finalizer. This function is
1627 -- not available on Hugs.
1629 packCStringFinalizer :: Ptr Word8 -> Int -> IO () -> IO ByteString
1630 packCStringFinalizer p l f = do
1631 fp <- FC.newForeignPtr p f
1634 -- | Explicitly run the finaliser associated with a 'ByteString'.
1635 -- Further references to this value may generate invalid memory
1636 -- references. This operation is unsafe, as there may be other
1637 -- 'ByteStrings' referring to the same underlying pages. If you use
1638 -- this, you need to have a proof of some kind that all 'ByteString's
1639 -- ever generated from the underlying byte array are no longer live.
1640 unsafeFinalize :: ByteString -> IO ()
1641 unsafeFinalize (PS p _ _) = finalizeForeignPtr p
1645 -- | /O(n) construction/ Use a @ByteString@ with a function requiring a null-terminated @CString@.
1646 -- The @CString@ should not be freed afterwards. This is a memcpy(3).
1647 useAsCString :: ByteString -> (CString -> IO a) -> IO a
1648 useAsCString (PS ps s l) = bracket alloc (c_free.castPtr)
1650 alloc = withForeignPtr ps $ \p -> do
1651 buf <- c_malloc (fromIntegral l+1)
1652 memcpy (castPtr buf) (castPtr p `plusPtr` s) (fromIntegral l)
1653 poke (buf `plusPtr` l) (0::Word8)
1654 return $ castPtr buf
1656 -- | /O(1) construction/ Use a @ByteString@ with a function requiring a @CString@.
1657 -- Warning: modifying the @CString@ will affect the @ByteString@.
1658 -- Why is this function unsafe? It relies on the null byte at the end of
1659 -- the ByteString to be there. This is /not/ the case if your ByteString
1660 -- has been spliced from a larger string (i.e. with take or drop).
1661 -- Unless you can guarantee the null byte, you should use the safe
1662 -- version, which will copy the string first.
1664 unsafeUseAsCString :: ByteString -> (CString -> IO a) -> IO a
1665 unsafeUseAsCString (PS ps s _) ac = withForeignPtr ps $ \p -> ac (castPtr p `plusPtr` s)
1667 -- | /O(n)/ Make a copy of the 'ByteString' with its own storage.
1668 -- This is mainly useful to allow the rest of the data pointed
1669 -- to by the 'ByteString' to be garbage collected, for example
1670 -- if a large string has been read in, and only a small part of it
1671 -- is needed in the rest of the program.
1672 copy :: ByteString -> ByteString
1673 copy (PS x s l) = create l $ \p -> withForeignPtr x $ \f -> memcpy p (f `plusPtr` s) l
1675 -- | /O(n)/ Duplicate a CString as a ByteString. Useful if you know the
1676 -- CString is going to be deallocated from C land.
1677 copyCString :: CString -> ByteString
1678 copyCString cstr = copyCStringLen (cstr, (fromIntegral $ c_strlen cstr))
1680 -- | /O(n)/ Same as copyCString, but saves a strlen call when the length is known.
1681 copyCStringLen :: CStringLen -> ByteString
1682 copyCStringLen (cstr, len) = inlinePerformIO $ do
1683 fp <- mallocForeignPtrArray (len+1)
1684 withForeignPtr fp $ \p -> do
1685 memcpy p (castPtr cstr) len
1686 poke (p `plusPtr` len) (0 :: Word8)
1687 return $! PS fp 0 len
1689 -- | /O(1) construction/ Use a @ByteString@ with a function requiring a @CStringLen@.
1690 -- Warning: modifying the @CStringLen@ will affect the @ByteString@.
1691 -- This is analogous to unsafeUseAsCString, and comes with the same
1692 -- safety requirements.
1694 unsafeUseAsCStringLen :: ByteString -> (CStringLen -> IO a) -> IO a
1695 unsafeUseAsCStringLen (PS ps s l) ac = withForeignPtr ps $ \p -> ac (castPtr p `plusPtr` s,l)
1697 -- | Given the maximum size needed and a function to make the contents
1698 -- of a ByteString, generate makes the 'ByteString'. The generating
1699 -- function is required to return the actual final size (<= the maximum
1700 -- size), and the resulting byte array is realloced to this size. The
1701 -- string is padded at the end with a null byte.
1703 -- generate is the main mechanism for creating custom, efficient
1704 -- ByteString functions, using Haskell or C functions to fill the space.
1706 generate :: Int -> (Ptr Word8 -> IO Int) -> IO ByteString
1708 p <- mallocArray (i+1)
1710 p' <- reallocArray p (i'+1)
1711 poke (p' `plusPtr` i') (0::Word8) -- XXX so CStrings work
1712 fp <- newForeignFreePtr p'
1715 -- ---------------------------------------------------------------------
1718 #if defined(__GLASGOW_HASKELL__)
1720 -- | getLine, read a line from stdin.
1721 getLine :: IO ByteString
1722 getLine = hGetLine stdin
1724 -- | hGetLine. read a ByteString from a handle
1725 hGetLine :: Handle -> IO ByteString
1726 hGetLine h = wantReadableHandle "Data.ByteString.hGetLine" h $ \ handle_ -> do
1727 case haBufferMode handle_ of
1728 NoBuffering -> error "no buffering"
1729 _other -> hGetLineBuffered handle_
1732 hGetLineBuffered handle_ = do
1733 let ref = haBuffer handle_
1734 buf <- readIORef ref
1735 hGetLineBufferedLoop handle_ ref buf 0 []
1737 hGetLineBufferedLoop handle_ ref
1738 buf@Buffer{ bufRPtr=r, bufWPtr=w, bufBuf=raw } len xss =
1740 off <- findEOL r w raw
1741 let new_len = len + off - r
1742 xs <- mkPS raw r off
1744 -- if eol == True, then off is the offset of the '\n'
1745 -- otherwise off == w and the buffer is now empty.
1747 then do if (w == off + 1)
1748 then writeIORef ref buf{ bufRPtr=0, bufWPtr=0 }
1749 else writeIORef ref buf{ bufRPtr = off + 1 }
1750 mkBigPS new_len (xs:xss)
1752 maybe_buf <- maybeFillReadBuffer (haFD handle_) True (haIsStream handle_)
1753 buf{ bufWPtr=0, bufRPtr=0 }
1755 -- Nothing indicates we caught an EOF, and we may have a
1756 -- partial line to return.
1758 writeIORef ref buf{ bufRPtr=0, bufWPtr=0 }
1760 then mkBigPS new_len (xs:xss)
1763 hGetLineBufferedLoop handle_ ref new_buf new_len (xs:xss)
1765 -- find the end-of-line character, if there is one
1769 (c,r') <- readCharFromBuffer raw r
1771 then return r -- NB. not r': don't include the '\n'
1772 else findEOL r' w raw
1774 maybeFillReadBuffer fd is_line is_stream buf = catch
1775 (do buf' <- fillReadBuffer fd is_line is_stream buf
1777 (\e -> if isEOFError e then return Nothing else ioError e)
1779 -- TODO, rewrite to use normal memcpy
1780 mkPS :: RawBuffer -> Int -> Int -> IO ByteString
1781 mkPS buf start end = do
1782 let len = end - start
1783 fp <- mallocByteString len
1784 withForeignPtr fp $ \p -> do
1785 memcpy_ptr_baoff p buf start (fromIntegral len)
1786 return (PS fp 0 len)
1788 mkBigPS :: Int -> [ByteString] -> IO ByteString
1789 mkBigPS _ [ps] = return ps
1790 mkBigPS _ pss = return $! concat (P.reverse pss)
1794 -- ---------------------------------------------------------------------
1797 -- | Outputs a 'ByteString' to the specified 'Handle'.
1798 hPut :: Handle -> ByteString -> IO ()
1799 hPut _ (PS _ _ 0) = return ()
1800 hPut h (PS ps 0 l) = withForeignPtr ps $ \p-> hPutBuf h p l
1801 hPut h (PS ps s l) = withForeignPtr ps $ \p-> hPutBuf h (p `plusPtr` s) l
1803 -- | Write a ByteString to stdout
1804 putStr :: ByteString -> IO ()
1805 putStr = hPut stdout
1807 -- | Write a ByteString to stdout, appending a newline byte
1808 putStrLn :: ByteString -> IO ()
1809 putStrLn ps = hPut stdout ps >> hPut stdout nl
1810 where nl = packByte 0x0a
1812 -- | Read a 'ByteString' directly from the specified 'Handle'. This
1813 -- is far more efficient than reading the characters into a 'String'
1814 -- and then using 'pack'.
1815 hGet :: Handle -> Int -> IO ByteString
1816 hGet _ 0 = return empty
1817 hGet h i = do fp <- mallocByteString i
1818 l <- withForeignPtr fp $ \p-> hGetBuf h p i
1821 #if defined(__GLASGOW_HASKELL__)
1822 -- | hGetNonBlocking is identical to 'hGet', except that it will never block
1823 -- waiting for data to become available, instead it returns only whatever data
1825 hGetNonBlocking :: Handle -> Int -> IO ByteString
1826 hGetNonBlocking _ 0 = return empty
1827 hGetNonBlocking h i = do
1828 fp <- mallocByteString i
1829 l <- withForeignPtr fp $ \p -> hGetBufNonBlocking h p i
1833 -- | Read entire handle contents into a 'ByteString'.
1835 -- As with 'hGet', the string representation in the file is assumed to
1838 hGetContents :: Handle -> IO ByteString
1840 let start_size = 1024
1841 p <- mallocArray start_size
1842 i <- hGetBuf h p start_size
1844 then do p' <- reallocArray p i
1845 fp <- newForeignFreePtr p'
1851 p' <- reallocArray p s'
1852 i <- hGetBuf h (p' `plusPtr` s) s
1854 then do let i' = s + i
1855 p'' <- reallocArray p' i'
1856 fp <- newForeignFreePtr p''
1860 -- | getContents. Equivalent to hGetContents stdin
1861 getContents :: IO ByteString
1862 getContents = hGetContents stdin
1864 -- | Read an entire file directly into a 'ByteString'. This is far more
1865 -- efficient than reading the characters into a 'String' and then using
1866 -- 'pack'. It also may be more efficient than opening the file and
1867 -- reading it using hGet.
1868 readFile :: FilePath -> IO ByteString
1870 h <- openBinaryFile f ReadMode
1872 s <- hGet h $ fromIntegral l
1876 -- | Write a 'ByteString' to a file.
1877 writeFile :: FilePath -> ByteString -> IO ()
1879 h <- openBinaryFile f WriteMode
1885 -- Disable until we can move it into a portable .hsc file
1888 -- | Like readFile, this reads an entire file directly into a
1889 -- 'ByteString', but it is even more efficient. It involves directly
1890 -- mapping the file to memory. This has the advantage that the contents
1891 -- of the file never need to be copied. Also, under memory pressure the
1892 -- page may simply be discarded, while in the case of readFile it would
1893 -- need to be written to swap. If you read many small files, mmapFile
1894 -- will be less memory-efficient than readFile, since each mmapFile
1895 -- takes up a separate page of memory. Also, you can run into bus
1896 -- errors if the file is modified. As with 'readFile', the string
1897 -- representation in the file is assumed to be ISO-8859-1.
1899 -- On systems without mmap, this is the same as a readFile.
1901 mmapFile :: FilePath -> IO ByteString
1902 mmapFile f = mmap f >>= \(fp,l) -> return $ PS fp 0 l
1904 mmap :: FilePath -> IO (ForeignPtr Word8, Int)
1906 h <- openBinaryFile f ReadMode
1907 l <- fromIntegral `fmap` hFileSize h
1908 -- Don't bother mmaping small files because each mmapped file takes up
1909 -- at least one full VM block.
1911 then do thefp <- mallocByteString l
1912 withForeignPtr thefp $ \p-> hGetBuf h p l
1917 fd <- fromIntegral `fmap` handleToFd h
1919 fp <- if p == nullPtr
1920 then do thefp <- mallocByteString l
1921 withForeignPtr thefp $ \p' -> hGetBuf h p' l
1924 -- The munmap leads to crashes on OpenBSD.
1925 -- maybe there's a use after unmap in there somewhere?
1926 #if !defined(__OpenBSD__)
1927 let unmap = c_munmap p l >> return ()
1929 let unmap = return ()
1931 fp <- FC.newForeignPtr p unmap
1936 where mmap_limit = 16*1024
1939 #if defined(__GLASGOW_HASKELL__)
1941 -- | A ByteString equivalent for getArgs. More efficient for large argument lists
1943 getArgs :: IO [ByteString]
1945 alloca $ \ p_argc ->
1946 alloca $ \ p_argv -> do
1947 getProgArgv p_argc p_argv
1948 p <- fromIntegral `fmap` peek p_argc
1950 P.map packCString `fmap` peekArray (p - 1) (advancePtr argv 1)
1953 -- ---------------------------------------------------------------------
1954 -- Internal utilities
1956 -- Unsafe conversion between 'Word8' and 'Char'. These are nops, and
1957 -- silently truncate to 8 bits Chars > '\255'. They are provided as
1958 -- convenience for ByteString construction.
1959 w2c :: Word8 -> Char
1960 #if !defined(__GLASGOW_HASKELL__)
1961 w2c = chr . fromIntegral
1963 w2c = unsafeChr . fromIntegral
1967 c2w :: Char -> Word8
1968 c2w = fromIntegral . ord
1971 -- Wrapper of mallocForeignPtrArray. Any ByteString allocated this way
1972 -- is padded with a null byte.
1973 mallocByteString :: Int -> IO (ForeignPtr Word8)
1974 mallocByteString l = do
1975 fp <- mallocForeignPtrArray (l+1)
1976 withForeignPtr fp $ \p -> poke (p `plusPtr` l) (0::Word8)
1979 -- | A way of creating ForeignPtrs outside the IO monad. The @Int@
1980 -- argument gives the final size of the ByteString. Unlike 'generate'
1981 -- the ByteString is not reallocated if the final size is less than the
1982 -- estimated size. Also, unlike 'generate' ByteString's created this way
1983 -- are managed on the Haskell heap.
1984 create :: Int -> (Ptr Word8 -> IO ()) -> ByteString
1985 create l write_ptr = inlinePerformIO $ do
1986 fp <- mallocByteString (l+1)
1987 withForeignPtr fp $ \p -> write_ptr p
1989 {-# INLINE create #-}
1991 -- | Perform an operation with a temporary ByteString
1992 withPtr :: ForeignPtr a -> (Ptr a -> IO b) -> b
1993 withPtr fp io = inlinePerformIO (withForeignPtr fp io)
1994 {-# INLINE withPtr #-}
1996 -- Common up near identical calls to `error' to reduce the number
1997 -- constant strings created when compiled:
1998 errorEmptyList :: String -> a
1999 errorEmptyList fun = error ("Data.ByteString." ++ fun ++ ": empty ByteString")
2000 {-# INLINE errorEmptyList #-}
2002 -- 'findIndexOrEnd' is a variant of findIndex, that returns the length
2003 -- of the string if no element is found, rather than Nothing.
2004 findIndexOrEnd :: (Word8 -> Bool) -> ByteString -> Int
2005 STRICT2(findIndexOrEnd)
2008 | f (unsafeHead ps) = 0
2009 | otherwise = 1 + findIndexOrEnd f (unsafeTail ps)
2010 {-# INLINE findIndexOrEnd #-}
2012 -- Find from the end of the string using predicate
2013 findFromEndUntil :: (Word8 -> Bool) -> ByteString -> Int
2014 STRICT2(findFromEndUntil)
2015 findFromEndUntil f ps@(PS x s l) =
2017 else if f (last ps) then l
2018 else findFromEndUntil f (PS x s (l-1))
2020 -- Just like inlinePerformIO, but we inline it. Big performance gains as
2021 -- it exposes lots of things to further inlining
2023 {-# INLINE inlinePerformIO #-}
2024 inlinePerformIO :: IO a -> a
2025 #if defined(__GLASGOW_HASKELL__)
2026 inlinePerformIO (IO m) = case m realWorld# of (# _, r #) -> r
2028 inlinePerformIO = unsafePerformIO
2031 {-# INLINE newForeignFreePtr #-}
2032 newForeignFreePtr :: Ptr Word8 -> IO (ForeignPtr Word8)
2033 #if defined(__GLASGOW_HASKELL__)
2034 newForeignFreePtr p = FC.newForeignPtr p (c_free p)
2036 newForeignFreePtr p = newForeignPtr c_free_finalizer p
2039 -- ---------------------------------------------------------------------
2041 -- Standard C functions
2044 foreign import ccall unsafe "string.h strlen" c_strlen
2047 foreign import ccall unsafe "stdlib.h malloc" c_malloc
2048 :: CInt -> IO (Ptr Word8)
2050 foreign import ccall unsafe "static stdlib.h free" c_free
2051 :: Ptr Word8 -> IO ()
2053 #if !defined(__GLASGOW_HASKELL__)
2054 foreign import ccall unsafe "static stdlib.h &free" c_free_finalizer
2055 :: FunPtr (Ptr Word8 -> IO ())
2058 foreign import ccall unsafe "string.h memset" memset
2059 :: Ptr Word8 -> Word8 -> CSize -> IO (Ptr Word8)
2061 foreign import ccall unsafe "string.h memchr" memchr
2062 :: Ptr Word8 -> Word8 -> CSize -> Ptr Word8
2064 foreign import ccall unsafe "string.h memcmp" memcmp
2065 :: Ptr Word8 -> Ptr Word8 -> Int -> IO Int
2067 foreign import ccall unsafe "string.h memcpy" memcpy
2068 :: Ptr Word8 -> Ptr Word8 -> Int -> IO ()
2070 -- ---------------------------------------------------------------------
2075 foreign import ccall unsafe "static fpstring.h reverse" c_reverse
2076 :: Ptr Word8 -> Ptr Word8 -> Int -> IO ()
2078 foreign import ccall unsafe "static fpstring.h intersperse" c_intersperse
2079 :: Ptr Word8 -> Ptr Word8 -> Int -> Word8 -> IO ()
2081 foreign import ccall unsafe "static fpstring.h maximum" c_maximum
2082 :: Ptr Word8 -> Int -> Word8
2084 foreign import ccall unsafe "static fpstring.h minimum" c_minimum
2085 :: Ptr Word8 -> Int -> Word8
2087 foreign import ccall unsafe "static fpstring.h count" c_count
2088 :: Ptr Word8 -> Int -> Word8 -> Int
2090 -- ---------------------------------------------------------------------
2094 foreign import ccall unsafe "static fpstring.h my_mmap" my_mmap
2095 :: Int -> Int -> IO (Ptr Word8)
2097 foreign import ccall unsafe "static unistd.h close" c_close
2100 # if !defined(__OpenBSD__)
2101 foreign import ccall unsafe "static sys/mman.h munmap" c_munmap
2102 :: Ptr Word8 -> Int -> IO Int
2106 -- ---------------------------------------------------------------------
2107 -- Internal GHC Haskell magic
2109 #if defined(__GLASGOW_HASKELL__)
2110 foreign import ccall unsafe "RtsAPI.h getProgArgv"
2111 getProgArgv :: Ptr CInt -> Ptr (Ptr CString) -> IO ()
2113 foreign import ccall unsafe "__hscore_memcpy_src_off"
2114 memcpy_ptr_baoff :: Ptr a -> RawBuffer -> Int -> CSize -> IO (Ptr ())
2117 -- ---------------------------------------------------------------------
2119 -- Functional array fusion for ByteStrings.
2121 -- From the Data Parallel Haskell project,
2122 -- http://www.cse.unsw.edu.au/~chak/project/dph/
2125 -- |Data type for accumulators which can be ignored. The rewrite rules rely on
2126 -- the fact that no bottoms of this type are ever constructed; hence, we can
2127 -- assume @(_ :: NoAL) `seq` x = x@.
2131 -- | Special forms of loop arguments
2133 -- * These are common special cases for the three function arguments of gen
2134 -- and loop; we give them special names to make it easier to trigger RULES
2135 -- applying in the special cases represented by these arguments. The
2136 -- "INLINE [1]" makes sure that these functions are only inlined in the last
2137 -- two simplifier phases.
2139 -- * In the case where the accumulator is not needed, it is better to always
2140 -- explicitly return a value `()', rather than just copy the input to the
2141 -- output, as the former gives GHC better local information.
2144 -- | Element function expressing a mapping only
2145 mapEFL :: (Word8 -> Word8) -> (NoAL -> Word8 -> (NoAL, Maybe Word8))
2146 mapEFL f = \_ e -> (noAL, (Just $ f e))
2147 {-# INLINE [1] mapEFL #-}
2149 -- | Element function implementing a filter function only
2150 filterEFL :: (Word8 -> Bool) -> (NoAL -> Word8 -> (NoAL, Maybe Word8))
2151 filterEFL p = \_ e -> if p e then (noAL, Just e) else (noAL, Nothing)
2152 {-# INLINE [1] filterEFL #-}
2154 -- |Element function expressing a reduction only
2155 foldEFL :: (acc -> Word8 -> acc) -> (acc -> Word8 -> (acc, Maybe Word8))
2156 foldEFL f = \a e -> (f a e, Nothing)
2157 {-# INLINE [1] foldEFL #-}
2162 {-# INLINE [1] noAL #-}
2164 -- | Projection functions that are fusion friendly (as in, we determine when
2165 -- they are inlined)
2166 loopArr :: (ByteString, acc) -> ByteString
2167 loopArr (arr, _) = arr
2168 {-# INLINE [1] loopArr #-}
2170 loopAcc :: (ByteString, acc) -> acc
2171 loopAcc (_, acc) = acc
2172 {-# INLINE [1] loopAcc #-}
2174 loopSndAcc :: (ByteString, (acc1, acc2)) -> (ByteString, acc2)
2175 loopSndAcc (arr, (_, acc)) = (arr, acc)
2176 {-# INLINE [1] loopSndAcc #-}
2178 ------------------------------------------------------------------------
2180 -- | Iteration over over ByteStrings
2181 loopU :: (acc -> Word8 -> (acc, Maybe Word8)) -- ^ mapping & folding, once per elem
2182 -> acc -- ^ initial acc value
2183 -> ByteString -- ^ input ByteString
2184 -> (ByteString, acc)
2186 loopU f start (PS z s i) = inlinePerformIO $ withForeignPtr z $ \a -> do
2187 fp <- mallocByteString i
2188 (ptr,n,acc) <- withForeignPtr fp $ \p -> do
2189 (acc, i') <- go (a `plusPtr` s) p start
2191 then return (fp,i,acc) -- no realloc for map
2192 else do fp_ <- mallocByteString (i'+1) -- realloc
2193 withForeignPtr fp_ $ \p' -> do
2195 poke (p' `plusPtr` i') (0::Word8)
2198 return (PS ptr 0 n, acc)
2203 trans a_off ma_off acc
2204 | a_off >= i = return (acc, ma_off)
2206 x <- peekByteOff p a_off
2207 let (acc', oe) = f acc x
2208 ma_off' <- case oe of
2209 Nothing -> return ma_off
2210 Just e -> do pokeByteOff ma ma_off e
2212 trans (a_off+1) ma_off' acc'
2214 {-# INLINE [1] loopU #-}
2218 "array fusion!" forall em1 em2 start1 start2 arr.
2219 loopU em2 start2 (loopArr (loopU em1 start1 arr)) =
2220 let em (acc1, acc2) e =
2222 (acc1', Nothing) -> ((acc1', acc2), Nothing)
2225 (acc2', res) -> ((acc1', acc2'), res)
2226 in loopSndAcc (loopU em (start1, start2) arr)
2228 "loopArr/loopSndAcc" forall x.
2229 loopArr (loopSndAcc x) = loopArr x
2231 "seq/NoAL" forall (u::NoAL) e.