FastString.lhs 17.2 KB
Newer Older
1
%
2
% (c) The University of Glasgow, 1997-2006
3
%
4
\begin{code}
5
{-# OPTIONS -w #-}
6 7 8
-- The above warning supression flag is a temporary kludge.
-- While working on this module you are encouraged to remove it and fix
-- any warnings in the module. See
Ian Lynagh's avatar
Ian Lynagh committed
9
--     http://hackage.haskell.org/trac/ghc/wiki/Commentary/CodingStyle#Warnings
10 11
-- for details

12
{-
Ian Lynagh's avatar
Ian Lynagh committed
13 14 15 16
FastString:     A compact, hash-consed, representation of character strings.
                Comparison is O(1), and you can get a Unique from them.
                Generated by the FSLIT macro
                Turn into SDoc with Outputable.ftext
17

Ian Lynagh's avatar
Ian Lynagh committed
18 19 20 21 22
LitString:      Just a wrapper for the Addr# of a C string (Ptr CChar).
                Practically no operations
                Outputing them is fast
                Generated by the SLIT macro
                Turn into SDoc with Outputable.ptext
23 24

Use LitString unless you want the facilities of FastString
25
-}
26 27
module FastString
       (
Ian Lynagh's avatar
Ian Lynagh committed
28 29
        -- * FastStrings
        FastString(..),     -- not abstract, for now.
30

Ian Lynagh's avatar
Ian Lynagh committed
31
        -- ** Construction
32
        mkFastString,
Ian Lynagh's avatar
Ian Lynagh committed
33
        mkFastStringBytes,
34
        mkFastStringByteList,
Ian Lynagh's avatar
Ian Lynagh committed
35
        mkFastStringForeignPtr,
36
#if defined(__GLASGOW_HASKELL__)
Ian Lynagh's avatar
Ian Lynagh committed
37
        mkFastString#,
38
#endif
Ian Lynagh's avatar
Ian Lynagh committed
39 40
        mkZFastString,
        mkZFastStringBytes,
41

Ian Lynagh's avatar
Ian Lynagh committed
42 43 44
        -- ** Deconstruction
        unpackFS,           -- :: FastString -> String
        bytesFS,            -- :: FastString -> [Word8]
45

Ian Lynagh's avatar
Ian Lynagh committed
46 47 48
        -- ** Encoding
        isZEncoded,
        zEncodeFS,
49

Ian Lynagh's avatar
Ian Lynagh committed
50
        -- ** Operations
51
        uniqueOfFS,
Ian Lynagh's avatar
Ian Lynagh committed
52 53 54
        lengthFS,
        nullFS,
        appendFS,
55 56
        headFS,
        tailFS,
Ian Lynagh's avatar
Ian Lynagh committed
57
        concatFS,
58
        consFS,
Ian Lynagh's avatar
Ian Lynagh committed
59
        nilFS,
60

Ian Lynagh's avatar
Ian Lynagh committed
61
        -- ** Outputing
62 63
        hPutFS,

Ian Lynagh's avatar
Ian Lynagh committed
64 65 66
        -- ** Internal
        getFastStringTable,
        hasZEncoding,
67

Ian Lynagh's avatar
Ian Lynagh committed
68 69
        -- * LitStrings
        LitString,
70
#if defined(__GLASGOW_HASKELL__)
Ian Lynagh's avatar
Ian Lynagh committed
71
        mkLitString#,
72
#else
Ian Lynagh's avatar
Ian Lynagh committed
73
        mkLitString,
74
#endif
Ian Lynagh's avatar
Ian Lynagh committed
75 76
        unpackLitString,
        strLength,
77

Ian Lynagh's avatar
Ian Lynagh committed
78
        ptrStrLength
79 80
       ) where

81 82 83 84
-- This #define suppresses the "import FastString" that
-- HsVersions otherwise produces
#define COMPILING_FAST_STRING
#include "HsVersions.h"
85

86
import Encoding
87 88
import FastTypes
import FastFunctions
89

90 91
import Foreign
import Foreign.C
92 93
import GHC.Exts
import System.IO.Unsafe ( unsafePerformIO )
Ian Lynagh's avatar
Ian Lynagh committed
94 95 96 97 98
import Control.Monad.ST ( stToIO )
import Data.IORef       ( IORef, newIORef, readIORef, writeIORef )
import System.IO        ( hPutBuf )
import Data.Maybe       ( isJust )
import Data.Char        ( ord )
99

100
import GHC.ST
Ian Lynagh's avatar
Ian Lynagh committed
101 102
import GHC.IOBase       ( IO(..) )
import GHC.Ptr          ( Ptr(..) )
103

104 105
#define hASH_TBL_SIZE          4091
#define hASH_TBL_SIZE_UNBOXED  4091#
106 107


108 109 110 111 112
{-|
A 'FastString' is an array of bytes, hashed to support fast O(1)
comparison.  It is also associated with a character encoding, so that
we know how to convert a 'FastString' to the local encoding, or to the
Z-encoding used by the compiler internally.
113

114 115 116 117
'FastString's support a memoized conversion to the Z-encoding via zEncodeFS.
-}

data FastString = FastString {
Ian Lynagh's avatar
Ian Lynagh committed
118 119 120
      uniq    :: {-# UNPACK #-} !Int, -- unique id
      n_bytes :: {-# UNPACK #-} !Int, -- number of bytes
      n_chars :: {-# UNPACK #-} !Int, -- number of chars
121 122 123 124 125
      buf     :: {-# UNPACK #-} !(ForeignPtr Word8),
      enc     :: FSEncoding
  }

data FSEncoding
Ian Lynagh's avatar
Ian Lynagh committed
126
    -- including strings that don't need any encoding
127
  = ZEncoded
Ian Lynagh's avatar
Ian Lynagh committed
128
    -- A UTF-8 string with a memoized Z-encoding
129
  | UTF8Encoded {-# UNPACK #-} !(IORef (Maybe FastString))
130

131 132
instance Eq FastString where
  f1 == f2  =  uniq f1 == uniq f2
133

134
instance Ord FastString where
Ian Lynagh's avatar
Ian Lynagh committed
135
    -- Compares lexicographically, not by unique
136
    a <= b = case cmpFS a b of { LT -> True;  EQ -> True;  GT -> False }
Ian Lynagh's avatar
Ian Lynagh committed
137
    a <  b = case cmpFS a b of { LT -> True;  EQ -> False; GT -> False }
138
    a >= b = case cmpFS a b of { LT -> False; EQ -> True;  GT -> True  }
Ian Lynagh's avatar
Ian Lynagh committed
139 140 141 142 143
    a >  b = case cmpFS a b of { LT -> False; EQ -> False; GT -> True  }
    max x y | x >= y    =  x
            | otherwise =  y
    min x y | x <= y    =  x
            | otherwise =  y
144
    compare a b = cmpFS a b
145

146 147 148
instance Show FastString where
   show fs = show (unpackFS fs)

149 150 151
cmpFS :: FastString -> FastString -> Ordering
cmpFS (FastString u1 l1 _ buf1 _) (FastString u2 l2 _ buf2 _) =
  if u1 == u2 then EQ else
Isaac Dupree's avatar
Isaac Dupree committed
152 153 154 155 156 157 158 159 160 161 162
  case unsafeMemcmp buf1 buf2 (min l1 l2) `compare` 0 of
     LT -> LT
     EQ -> compare l1 l2
     GT -> GT

unsafeMemcmp :: ForeignPtr a -> ForeignPtr b -> Int -> Int
unsafeMemcmp buf1 buf2 l =
      inlinePerformIO $
        withForeignPtr buf1 $ \p1 ->
        withForeignPtr buf2 $ \p2 ->
          memcmp p1 p2 l
163

164
#ifndef __HADDOCK__
Ian Lynagh's avatar
Ian Lynagh committed
165
foreign import ccall unsafe "ghc_memcmp"
166 167
  memcmp :: Ptr a -> Ptr b -> Int -> IO Int
#endif
168

169 170
-- -----------------------------------------------------------------------------
-- Construction
171

172
{-
173 174 175 176
Internally, the compiler will maintain a fast string symbol
table, providing sharing and fast comparison. Creation of
new @FastString@s then covertly does a lookup, re-using the
@FastString@ if there was a hit.
177
-}
178

Ian Lynagh's avatar
Ian Lynagh committed
179
data FastStringTable =
180
 FastStringTable
181
    {-# UNPACK #-} !Int
182
    (MutableArray# RealWorld [FastString])
183

184
{-# NOINLINE string_table #-}
185
string_table :: IORef FastStringTable
Ian Lynagh's avatar
Ian Lynagh committed
186
string_table =
187
 unsafePerformIO $ do
188 189 190 191
   tab <- IO $ \s1# -> case newArray# hASH_TBL_SIZE_UNBOXED [] s1# of
                           (# s2#, arr# #) ->
                               (# s2#, FastStringTable 0 arr# #)
   newIORef tab
192 193 194 195 196 197 198 199 200

lookupTbl :: FastStringTable -> Int -> IO [FastString]
lookupTbl (FastStringTable _ arr#) (I# i#) =
  IO $ \ s# -> readArray# arr# i# s#

updTbl :: IORef FastStringTable -> FastStringTable -> Int -> [FastString] -> IO ()
updTbl fs_table_var (FastStringTable uid arr#) (I# i#) ls = do
  (IO $ \ s# -> case writeArray# arr# i# ls s# of { s2# -> (# s2#, () #) })
  writeIORef fs_table_var (FastStringTable (uid+1) arr#)
201

202
mkFastString# :: Addr# -> FastString
203
mkFastString# a# = mkFastStringBytes ptr (ptrStrLength ptr)
204
  where ptr = Ptr a#
205

206 207 208
mkFastStringBytes :: Ptr Word8 -> Int -> FastString
mkFastStringBytes ptr len = unsafePerformIO $ do
  ft@(FastStringTable uid tbl#) <- readIORef string_table
209
  let
210 211
   h = hashStr ptr len
   add_it ls = do
Ian Lynagh's avatar
Ian Lynagh committed
212 213 214 215
        fs <- copyNewFastString uid ptr len
        updTbl string_table ft h (fs:ls)
        {- _trace ("new: " ++ show f_str)   $ -}
        return fs
216 217
  --
  lookup_result <- lookupTbl ft h
sof's avatar
sof committed
218
  case lookup_result of
219 220 221 222
    [] -> add_it []
    ls -> do
       b <- bucket_match ls len ptr
       case b of
Ian Lynagh's avatar
Ian Lynagh committed
223 224
         Nothing -> add_it ls
         Just v  -> {- _trace ("re-use: "++show v) $ -} return v
225 226 227 228

mkZFastStringBytes :: Ptr Word8 -> Int -> FastString
mkZFastStringBytes ptr len = unsafePerformIO $ do
  ft@(FastStringTable uid tbl#) <- readIORef string_table
sof's avatar
sof committed
229
  let
230 231
   h = hashStr ptr len
   add_it ls = do
Ian Lynagh's avatar
Ian Lynagh committed
232 233 234 235
        fs <- copyNewZFastString uid ptr len
        updTbl string_table ft h (fs:ls)
        {- _trace ("new: " ++ show f_str)   $ -}
        return fs
236 237
  --
  lookup_result <- lookupTbl ft h
sof's avatar
sof committed
238
  case lookup_result of
239 240 241 242
    [] -> add_it []
    ls -> do
       b <- bucket_match ls len ptr
       case b of
Ian Lynagh's avatar
Ian Lynagh committed
243 244
         Nothing -> add_it ls
         Just v  -> {- _trace ("re-use: "++show v) $ -} return v
245 246 247 248 249 250 251 252

-- | Create a 'FastString' from an existing 'ForeignPtr'; the difference
-- between this and 'mkFastStringBytes' is that we don't have to copy
-- the bytes if the string is new to the table.
mkFastStringForeignPtr :: Ptr Word8 -> ForeignPtr Word8 -> Int -> IO FastString
mkFastStringForeignPtr ptr fp len = do
  ft@(FastStringTable uid tbl#) <- readIORef string_table
--  _trace ("hashed: "++show (I# h)) $
253
  let
254 255
    h = hashStr ptr len
    add_it ls = do
Ian Lynagh's avatar
Ian Lynagh committed
256 257 258 259
        fs <- mkNewFastString uid ptr fp len
        updTbl string_table ft h (fs:ls)
        {- _trace ("new: " ++ show f_str)   $ -}
        return fs
260 261 262 263 264 265 266
  --
  lookup_result <- lookupTbl ft h
  case lookup_result of
    [] -> add_it []
    ls -> do
       b <- bucket_match ls len ptr
       case b of
Ian Lynagh's avatar
Ian Lynagh committed
267 268
         Nothing -> add_it ls
         Just v  -> {- _trace ("re-use: "++show v) $ -} return v
269 270 271 272 273 274 275 276

mkZFastStringForeignPtr :: Ptr Word8 -> ForeignPtr Word8 -> Int -> IO FastString
mkZFastStringForeignPtr ptr fp len = do
  ft@(FastStringTable uid tbl#) <- readIORef string_table
--  _trace ("hashed: "++show (I# h)) $
  let
    h = hashStr ptr len
    add_it ls = do
Ian Lynagh's avatar
Ian Lynagh committed
277 278 279 280
        fs <- mkNewZFastString uid ptr fp len
        updTbl string_table ft h (fs:ls)
        {- _trace ("new: " ++ show f_str)   $ -}
        return fs
281 282
  --
  lookup_result <- lookupTbl ft h
283
  case lookup_result of
284 285 286 287
    [] -> add_it []
    ls -> do
       b <- bucket_match ls len ptr
       case b of
Ian Lynagh's avatar
Ian Lynagh committed
288 289
         Nothing -> add_it ls
         Just v  -> {- _trace ("re-use: "++show v) $ -} return v
290 291


292 293
-- | Creates a UTF-8 encoded 'FastString' from a 'String'
mkFastString :: String -> FastString
Ian Lynagh's avatar
Ian Lynagh committed
294
mkFastString str =
295 296 297 298 299
  inlinePerformIO $ do
    let l = utf8EncodedLength str
    buf <- mallocForeignPtrBytes l
    withForeignPtr buf $ \ptr -> do
      utf8EncodeString ptr str
Ian Lynagh's avatar
Ian Lynagh committed
300
      mkFastStringForeignPtr ptr buf l
301

302 303
-- | Creates a 'FastString' from a UTF-8 encoded @[Word8]@
mkFastStringByteList :: [Word8] -> FastString
Ian Lynagh's avatar
Ian Lynagh committed
304
mkFastStringByteList str =
305 306 307 308 309
  inlinePerformIO $ do
    let l = Prelude.length str
    buf <- mallocForeignPtrBytes l
    withForeignPtr buf $ \ptr -> do
      pokeArray (castPtr ptr) str
Ian Lynagh's avatar
Ian Lynagh committed
310
      mkFastStringForeignPtr ptr buf l
311 312 313

-- | Creates a Z-encoded 'FastString' from a 'String'
mkZFastString :: String -> FastString
Ian Lynagh's avatar
Ian Lynagh committed
314
mkZFastString str =
315 316 317 318 319
  inlinePerformIO $ do
    let l = Prelude.length str
    buf <- mallocForeignPtrBytes l
    withForeignPtr buf $ \ptr -> do
      pokeCAString (castPtr ptr) str
Ian Lynagh's avatar
Ian Lynagh committed
320
      mkZFastStringForeignPtr ptr buf l
321 322 323 324

bucket_match [] _ _ = return Nothing
bucket_match (v@(FastString _ l _ buf _):ls) len ptr
      | len == l  =  do
Ian Lynagh's avatar
Ian Lynagh committed
325 326 327 328 329
         b <- cmpStringPrefix ptr buf len
         if b then return (Just v)
              else bucket_match ls len ptr
      | otherwise =
         bucket_match ls len ptr
330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363

mkNewFastString uid ptr fp len = do
  ref <- newIORef Nothing
  n_chars <- countUTF8Chars ptr len
  return (FastString uid len n_chars fp (UTF8Encoded ref))

mkNewZFastString uid ptr fp len = do
  return (FastString uid len len fp ZEncoded)


copyNewFastString uid ptr len = do
  fp <- copyBytesToForeignPtr ptr len
  ref <- newIORef Nothing
  n_chars <- countUTF8Chars ptr len
  return (FastString uid len n_chars fp (UTF8Encoded ref))

copyNewZFastString uid ptr len = do
  fp <- copyBytesToForeignPtr ptr len
  return (FastString uid len len fp ZEncoded)


copyBytesToForeignPtr ptr len = do
  fp <- mallocForeignPtrBytes len
  withForeignPtr fp $ \ptr' -> copyBytes ptr' ptr len
  return fp

cmpStringPrefix :: Ptr Word8 -> ForeignPtr Word8 -> Int -> IO Bool
cmpStringPrefix ptr fp len =
  withForeignPtr fp $ \ptr' -> do
    r <- memcmp ptr ptr' len
    return (r == 0)


hashStr  :: Ptr Word8 -> Int -> Int
364
 -- use the Addr to produce a hash value between 0 & m (inclusive)
365
hashStr (Ptr a#) (I# len#) = loop 0# 0#
Ian Lynagh's avatar
Ian Lynagh committed
366
   where
367
    loop h n | n GHC.Exts.==# len# = I# h
Ian Lynagh's avatar
Ian Lynagh committed
368 369 370 371
             | otherwise  = loop h2 (n GHC.Exts.+# 1#)
          where c = ord# (indexCharOffAddr# a# n)
                h2 = (c GHC.Exts.+# (h GHC.Exts.*# 128#)) `remInt#`
                     hASH_TBL_SIZE#
372

373 374
-- -----------------------------------------------------------------------------
-- Operations
375

376 377 378
-- | Returns the length of the 'FastString' in characters
lengthFS :: FastString -> Int
lengthFS f = n_chars f
379

380 381 382
-- | Returns 'True' if the 'FastString' is Z-encoded
isZEncoded :: FastString -> Bool
isZEncoded fs | ZEncoded <- enc fs = True
Ian Lynagh's avatar
Ian Lynagh committed
383
              | otherwise          = False
384

385 386 387 388 389 390 391 392 393
-- | Returns 'True' if this 'FastString' is not Z-encoded but already has
-- a Z-encoding cached (used in producing stats).
hasZEncoding :: FastString -> Bool
hasZEncoding fs@(FastString uid n_bytes _ fp enc) =
  case enc of
    ZEncoded -> False
    UTF8Encoded ref ->
      inlinePerformIO $ do
        m <- readIORef ref
Ian Lynagh's avatar
Ian Lynagh committed
394
        return (isJust m)
395

396 397 398 399 400 401
-- | Returns 'True' if the 'FastString' is empty
nullFS :: FastString -> Bool
nullFS f  =  n_bytes f == 0

-- | unpacks and decodes the FastString
unpackFS :: FastString -> String
Ian Lynagh's avatar
Ian Lynagh committed
402
unpackFS (FastString _ n_bytes _ buf enc) =
403 404
  inlinePerformIO $ withForeignPtr buf $ \ptr ->
    case enc of
Ian Lynagh's avatar
Ian Lynagh committed
405 406
        ZEncoded      -> peekCAStringLen (castPtr ptr,n_bytes)
        UTF8Encoded _ -> utf8DecodeString ptr n_bytes
407 408

bytesFS :: FastString -> [Word8]
Ian Lynagh's avatar
Ian Lynagh committed
409
bytesFS (FastString _ n_bytes _ buf enc) =
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425
  inlinePerformIO $ withForeignPtr buf $ \ptr ->
    peekArray n_bytes ptr

-- | returns a Z-encoded version of a 'FastString'.  This might be the
-- original, if it was already Z-encoded.  The first time this
-- function is applied to a particular 'FastString', the results are
-- memoized.
--
zEncodeFS :: FastString -> FastString
zEncodeFS fs@(FastString uid n_bytes _ fp enc) =
  case enc of
    ZEncoded -> fs
    UTF8Encoded ref ->
      inlinePerformIO $ do
        m <- readIORef ref
        case m of
Ian Lynagh's avatar
Ian Lynagh committed
426 427
          Just fs -> return fs
          Nothing -> do
428
            let efs = mkZFastString (zEncodeString (unpackFS fs))
Ian Lynagh's avatar
Ian Lynagh committed
429 430
            writeIORef ref (Just efs)
            return efs
431 432 433 434 435 436 437 438

appendFS :: FastString -> FastString -> FastString
appendFS fs1 fs2 = mkFastString (unpackFS fs1 ++ unpackFS fs2)

concatFS :: [FastString] -> FastString
concatFS ls = mkFastString (Prelude.concat (map unpackFS ls)) -- ToDo: do better

headFS :: FastString -> Char
Ian Lynagh's avatar
Ian Lynagh committed
439
headFS (FastString _ n_bytes _ buf enc) =
440 441
  inlinePerformIO $ withForeignPtr buf $ \ptr -> do
    case enc of
Ian Lynagh's avatar
Ian Lynagh committed
442 443 444 445 446
      ZEncoded -> do
         w <- peek (castPtr ptr)
         return (castCCharToChar w)
      UTF8Encoded _ ->
         return (fst (utf8DecodeChar ptr))
447 448

tailFS :: FastString -> FastString
Ian Lynagh's avatar
Ian Lynagh committed
449
tailFS (FastString _ n_bytes _ buf enc) =
450 451 452
  inlinePerformIO $ withForeignPtr buf $ \ptr -> do
    case enc of
      ZEncoded -> do
Ian Lynagh's avatar
Ian Lynagh committed
453
        return $! mkZFastStringBytes (ptr `plusPtr` 1) (n_bytes - 1)
454
      UTF8Encoded _ -> do
Ian Lynagh's avatar
Ian Lynagh committed
455 456 457
         let (_,ptr') = utf8DecodeChar ptr
         let off = ptr' `minusPtr` ptr
         return $! mkFastStringBytes (ptr `plusPtr` off) (n_bytes - off)
458 459 460 461

consFS :: Char -> FastString -> FastString
consFS c fs = mkFastString (c : unpackFS fs)

462 463
uniqueOfFS :: FastString -> FastInt
uniqueOfFS (FastString u _ _ _ _) = iUnbox u
464 465

nilFS = mkFastString ""
466

467 468 469 470 471 472
-- -----------------------------------------------------------------------------
-- Stats

getFastStringTable :: IO [[FastString]]
getFastStringTable = do
  tbl <- readIORef string_table
473
  buckets <- mapM (lookupTbl tbl) [0 .. hASH_TBL_SIZE]
474 475
  return buckets

476 477 478
-- -----------------------------------------------------------------------------
-- Outputting 'FastString's

479 480 481 482 483
-- |Outputs a 'FastString' with /no decoding at all/, that is, you
-- get the actual bytes in the 'FastString' written to the 'Handle'.
hPutFS handle (FastString _ len _ fp _)
  | len == 0  = return ()
  | otherwise = do withForeignPtr fp $ \ptr -> hPutBuf handle ptr len
484

485 486
-- ToDo: we'll probably want an hPutFSLocal, or something, to output
-- in the current locale's encoding (for error messages and suchlike).
487

488 489
-- -----------------------------------------------------------------------------
-- LitStrings, here for convenience only.
490

491 492 493 494 495 496 497
-- hmm, not unboxed (or rather FastPtr), interesting
--a.k.a. Ptr CChar, Ptr Word8, Ptr (), hmph.  We don't
--really care about C types in naming, where we can help it.
type LitString = Ptr Word8
--Why do we recalculate length every time it's requested?
--If it's commonly needed, we should perhaps have
--data LitString = LitString {-#UNPACK#-}!(FastPtr Word8) {-#UNPACK#-}!FastInt
498

499
#if defined(__GLASGOW_HASKELL__)
500
mkLitString# :: Addr# -> LitString
501
mkLitString# a# = Ptr a#
502
#endif
503

504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551
--can/should we use FastTypes here?
--Is this likely to be memory-preserving if only used on constant strings?
--should we inline it? If lucky, that would make a CAF that wouldn't
--be computationally repeated... although admittedly we're not
--really intending to use mkLitString when __GLASGOW_HASKELL__...
--(I wonder, is unicode / multi-byte characters allowed in LitStrings
-- at all?)
{-# INLINE mkLitString #-}
mkLitString :: String -> LitString
mkLitString s =
 unsafePerformIO (do
   p <- mallocBytes (length s + 1)
   let
     loop :: Int -> String -> IO ()
     loop n cs | n `seq` null cs = pokeByteOff p n (0 :: Word8)
     loop n (c:cs) = do
        pokeByteOff p n (fromIntegral (ord c) :: Word8)
        loop (1+n) cs
   loop 0 s
   return p
 )

unpackLitString :: LitString -> String
unpackLitString p_ = case pUnbox p_ of
 p -> unpack (_ILIT(0))
  where
    unpack n = case indexWord8OffFastPtrAsFastChar p n of
      ch -> if ch `eqFastChar` _CLIT('\0')
            then [] else cBox ch : unpack (n +# _ILIT(1))

strLength :: LitString -> Int
strLength = ptrStrLength

-- for now, use a simple String representation
--no, let's not do that right now - it's work in other places
#if 0
type LitString = String

mkLitString :: String -> LitString
mkLitString = id

unpackLitString :: LitString -> String
unpackLitString = id

strLength :: LitString -> Int
strLength = length

#endif
552 553 554 555

-- -----------------------------------------------------------------------------
-- under the carpet

Ian Lynagh's avatar
Ian Lynagh committed
556
foreign import ccall unsafe "ghc_strlen"
557
  ptrStrLength :: Ptr Word8 -> Int
558

559
-- NB. does *not* add a '\0'-terminator.
560 561
-- We only use CChar here to be parallel to the imported
-- peekC(A)StringLen.
562 563 564
pokeCAString :: Ptr CChar -> String -> IO ()
pokeCAString ptr str =
  let
Ian Lynagh's avatar
Ian Lynagh committed
565 566
        go [] n     = return ()
        go (c:cs) n = do pokeElemOff ptr n (castCharToCChar c); go cs (n+1)
567 568 569
  in
  go str 0

570
#if defined(__GLASGOW_HASKELL__) && __GLASGOW_HASKELL__ <= 602
571 572
peekCAStringLen = peekCStringLen
#endif
573
\end{code}