40a3168..master · text

Data/Text.hs−12

since splitsnapshots

unifiedsplit

@@ -60,7 +60,6 @@ module Data.Text

60 , snoc

61 , append

62 , uncons

63− , unsnoc

64 , head

63 , head

65 , last

64 , last

66 , tail

65 , tail

@@ -546,17 +545,6 @@ init (Text arr off len) | len <= 0 = emptyError "init"

546 unstream (S.init (stream t)) = init t

545 unstream (S.init (stream t)) = init t

547 #-}

546 #-}

548

547

549−-- | /O(1)/ Returns all but the last character and the last character of a

550−-- 'Text', or 'Nothing' if empty.

551−unsnoc :: Text -> Maybe (Text, Char)

552−unsnoc (Text arr off len)

553− | len <= 0 = Nothing

554− | n < 0xDC00 || n > 0xDFFF = Just (text arr off (len-1), unsafeChr n)

555− | otherwise = Just (text arr off (len-2), U16.chr2 n0 n)

556− where n = A.unsafeIndex arr (off+len-1)

557− n0 = A.unsafeIndex arr (off+len-2)

558−{-# INLINE [1] unsnoc #-}

559−

560-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

548-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

561-- fusion.

549-- fusion.

562null :: Text -> Bool

550null :: Text -> Bool

Data/Text/Internal/Fusion/Common.hs+9−10

@@ -107,7 +107,7 @@ import Prelude (Bool(..), Char, Eq(..), Int, Integral, Maybe(..),

107import qualified Data.List as L

108import qualified Prelude as P

109import Data.Bits (shiftL)

110−import Data.Char (isLetter, isSpace)

110+import Data.Char (isLetter)

111import Data.Int (Int64)

112import Data.Text.Internal.Fusion.Types

113import Data.Text.Internal.Fusion.CaseMapping (foldMapping, lowerMapping, titleMapping,

@@ -462,16 +462,15 @@ toTitle (Stream next0 s0 len) = Stream next (CC (False :*: s0) '\0' '\0') len

462 where

463 next (CC (letter :*: s) '\0' _) =

464 case next0 s of

465− Done -> Done

465+ Done -> Done

466− Skip s' -> Skip (CC (letter :*: s') '\0' '\0')

466+ Skip s' -> Skip (CC (letter :*: s') '\0' '\0')

467 Yield c s'

468− | nonSpace -> if letter

468+ | letter' -> if letter

469− then lowerMapping c (nonSpace :*: s')

469+ then lowerMapping c (letter' :*: s')

470− else titleMapping c (letter' :*: s')

470+ else titleMapping c (letter' :*: s')

471− | otherwise -> Yield c (CC (letter' :*: s') '\0' '\0')

471+ | otherwise -> Yield c (CC (letter' :*: s') '\0' '\0')

472− where nonSpace = P.not (isSpace c)

473− letter' = isLetter c

472+ where letter' = isLetter c

474− next (CC s a b) = Yield a (CC s b '\0')

473+ next (CC s a b) = Yield a (CC s b '\0')

475{-# INLINE [0] toTitle #-}

474{-# INLINE [0] toTitle #-}

476

475

477data Justify i s = Just1 !i !s

476data Justify i s = Just1 !i !s

Data/Text/Lazy.hs−10

@@ -68,7 +68,6 @@ module Data.Text.Lazy

68 , snoc

69 , append

70 , uncons

71− , unsnoc

72 , head

71 , head

73 , last

72 , last

74 , tail

73 , tail

@@ -562,15 +561,6 @@ init Empty = emptyError "init"

562 unstream (S.init (stream t)) = init t

561 unstream (S.init (stream t)) = init t

563 #-}

562 #-}

564

563

565−-- | /O(n\/c)/ Returns the 'init' and 'last' of a 'Text', or 'Nothing' if

566−-- empty.

567−--

568−-- * It is no faster than using 'init' and 'last'.

569−unsnoc :: Text -> Maybe (Text, Char)

570−unsnoc Empty = Nothing

571−unsnoc ts@(Chunk _ _) = Just (init ts, last ts)

572−{-# INLINE unsnoc #-}

573−

574-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

564-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

575-- fusion.

565-- fusion.

576null :: Text -> Bool

566null :: Text -> Bool

Data/Text/Lazy/Read.hs+3−3

@@ -68,9 +68,9 @@ decimal txt

68 go n d = (n * 10 + fromIntegral (digitToInt d))

69

70-- | Read a hexadecimal integer, consisting of an optional leading

71−-- @\"0x\"@ followed by at least one hexadecimal digit. Input is

71+-- @\"0x\"@ followed by at least one decimal digit. Input is consumed

72−-- consumed until a non-hex-digit or end of string is reached.

72+-- until a non-hex-digit or end of string is reached. This function

73−-- This function is case insensitive.

73+-- is case insensitive.

74--

75-- This function does not handle leading sign characters. If you need

76-- to handle signed input, use @'signed' 'hexadecimal'@.

Data/Text/Read.hs+3−3

@@ -67,9 +67,9 @@ decimal txt

67 go n d = (n * 10 + fromIntegral (digitToInt d))

68

69-- | Read a hexadecimal integer, consisting of an optional leading

70−-- @\"0x\"@ followed by at least one hexadecimal digit. Input is

70+-- @\"0x\"@ followed by at least one decimal digit. Input is consumed

71−-- consumed until a non-hex-digit or end of string is reached.

71+-- until a non-hex-digit or end of string is reached. This function

72−-- This function is case insensitive.

72+-- is case insensitive.

73--

74-- This function does not handle leading sign characters. If you need

75-- to handle signed input, use @'signed' 'hexadecimal'@.

changelog.md−11

@@ -1,14 +1,3 @@

1−* The `toTitle` function now correctly handles letters that

2− immediately follow puncutation. Before, `"there's"` would turn into

3− `"There'S"`. Now, it becomes `"There's"`.

4−

5−* The implementation of unstreaming is faster, resulting in operations

6− such as `map` and `intersperse` speeding up by up to 30%, with

7− smaller code generated.

8−

9−* The optimised length comparison function is now more likely to be

10− used after some rewrite rule tweaking.

11−

121.2.2.1

11.2.2.1

13

2

14* The switch to `integer-pure` in 1.2.2.0 was apparently mistaken.

3* The switch to `integer-pure` in 1.2.2.0 was apparently mistaken.

tests/Tests/Properties.hs+3−23

@@ -9,9 +9,9 @@ module Tests.Properties

9 ) where

10

11import Control.Applicative ((<$>), (<*>))

12−import Control.Arrow ((***), first, second)

12+import Control.Arrow ((***), second)

13import Data.Bits ((.&.))

14−import Data.Char (chr, isDigit, isHexDigit, isLower, isSpace, isLetter, isUpper, ord)

14+import Data.Char (chr, isDigit, isHexDigit, isLower, isSpace, isUpper, ord)

15import Data.Int (Int8, Int16, Int32, Int64)

16import Data.Monoid (Monoid(..))

17import Data.String (IsString(fromString))

@@ -23,7 +23,6 @@ import Data.Text.Internal.Search (indices)

23import Data.Text.Lazy.Read as TL

24import Data.Text.Read as T

25import Data.Word (Word, Word8, Word16, Word32, Word64)

26−import Data.Maybe (mapMaybe)

27import Numeric (showEFloat, showFFloat, showGFloat, showHex)

26import Numeric (showEFloat, showFFloat, showGFloat, showHex)

28import Prelude hiding (replicate)

27import Prelude hiding (replicate)

29import Test.Framework (Test, testGroup)

28import Test.Framework (Test, testGroup)

@@ -230,13 +229,6 @@ sf_uncons p = (uncons . L.filter p) `eqP`

230 (fmap (second unpackS) . S.uncons . S.filter p)

229 (fmap (second unpackS) . S.uncons . S.filter p)

231t_uncons = uncons `eqP` (fmap (second unpackS) . T.uncons)

230t_uncons = uncons `eqP` (fmap (second unpackS) . T.uncons)

232tl_uncons = uncons `eqP` (fmap (second unpackS) . TL.uncons)

231tl_uncons = uncons `eqP` (fmap (second unpackS) . TL.uncons)

233−

234−unsnoc xs@(_:_) = Just (init xs, last xs)

235−unsnoc [] = Nothing

236−

237−t_unsnoc = unsnoc `eqP` (fmap (first unpackS) . T.unsnoc)

238−tl_unsnoc = unsnoc `eqP` (fmap (first unpackS) . TL.unsnoc)

239−

240s_head = head `eqP` S.head

232s_head = head `eqP` S.head

241sf_head p = (head . L.filter p) `eqP` (S.head . S.filter p)

233sf_head p = (head . L.filter p) `eqP` (S.head . S.filter p)

242t_head = head `eqP` T.head

234t_head = head `eqP` T.head

@@ -334,14 +326,6 @@ t_toUpper_upper t = p (T.toUpper t) >= p t

334 where p = T.length . T.filter isUpper

326 where p = T.length . T.filter isUpper

335tl_toUpper_upper t = p (TL.toUpper t) >= p t

327tl_toUpper_upper t = p (TL.toUpper t) >= p t

336 where p = TL.length . TL.filter isUpper

328 where p = TL.length . TL.filter isUpper

337−t_toTitle_title t = all (<= 1) (caps t)

338− where caps = fmap (T.length . T.filter isUpper) . T.words . T.toTitle

339−t_toTitle_1stNotLower = and . notLow . T.toTitle . T.filter stable

340− where notLow = mapMaybe (fmap (not . isLower) . (T.find isLetter)) . T.words

341− -- Surprise! The Spanish/Portuguese ordinal indicators changed

342− -- from category Ll (letter, lowercase) to Lo (letter, other)

343− -- in Unicode 7.0

344− stable c = c /= '\170' && c /= '\186'

345

329

346justifyLeft k c xs = xs ++ L.replicate (k - length xs) c

330justifyLeft k c xs = xs ++ L.replicate (k - length xs) c

347justifyRight m n xs = L.replicate (m - length xs) n ++ xs

331justifyRight m n xs = L.replicate (m - length xs) n ++ xs

@@ -983,8 +967,6 @@ tests =

983 testProperty "sf_uncons" sf_uncons,

967 testProperty "sf_uncons" sf_uncons,

984 testProperty "t_uncons" t_uncons,

968 testProperty "t_uncons" t_uncons,

985 testProperty "tl_uncons" tl_uncons,

969 testProperty "tl_uncons" tl_uncons,

986− testProperty "t_unsnoc" t_unsnoc,

987− testProperty "tl_unsnoc" tl_unsnoc,

988 testProperty "s_head" s_head,

970 testProperty "s_head" s_head,

989 testProperty "sf_head" sf_head,

971 testProperty "sf_head" sf_head,

990 testProperty "t_head" t_head,

972 testProperty "t_head" t_head,

@@ -1048,9 +1030,7 @@ tests =

1048 testProperty "tl_toLower_lower" tl_toLower_lower,

1030 testProperty "tl_toLower_lower" tl_toLower_lower,

1049 testProperty "t_toUpper_length" t_toUpper_length,

1031 testProperty "t_toUpper_length" t_toUpper_length,

1050 testProperty "t_toUpper_upper" t_toUpper_upper,

1032 testProperty "t_toUpper_upper" t_toUpper_upper,

1051− testProperty "tl_toUpper_upper" tl_toUpper_upper,

1033+ testProperty "tl_toUpper_upper" tl_toUpper_upper

1052− testProperty "t_toTitle_title" t_toTitle_title,

1053− testProperty "t_toTitle_1stNotLower" t_toTitle_1stNotLower

1054 ],

1034 ],

1055

1035

1056 testGroup "justification" [

1036 testGroup "justification" [

tests/text-tests.cabal+2−17

@@ -20,11 +20,6 @@ flag hpc

20 default: False

21 manual: True

22

23−flag bytestring-builder

24− description: Depend on the bytestring-builder package for backwards compatibility.

25− default: False

26− manual: False

27−

28executable text-tests

23executable text-tests

29 main-is: Tests.hs

24 main-is: Tests.hs

30

25

@@ -53,6 +48,7 @@ executable text-tests

53 HUnit >= 1.2,

48 HUnit >= 1.2,

54 QuickCheck >= 2.7,

49 QuickCheck >= 2.7,

55 base == 4.*,

50 base == 4.*,

51+ bytestring,

56 deepseq,

52 deepseq,

57 directory,

53 directory,

58 quickcheck-unicode,

54 quickcheck-unicode,

@@ -62,12 +58,6 @@ executable text-tests

62 test-framework-quickcheck2 >= 0.2,

58 test-framework-quickcheck2 >= 0.2,

63 text-tests

59 text-tests

64

60

65− if flag(bytestring-builder)

66− build-depends: bytestring >= 0.9 && < 0.10.4,

67− bytestring-builder >= 0.10.4

68− else

69− build-depends: bytestring >= 0.10.4

70−

71executable text-tests-stdio

61executable text-tests-stdio

72 main-is: Tests/IO.hs

62 main-is: Tests/IO.hs

73

63

@@ -147,12 +137,7 @@ library

147 array,

137 array,

148 base == 4.*,

138 base == 4.*,

149 binary,

139 binary,

140+ bytestring,

150 deepseq,

141 deepseq,

151 ghc-prim,

142 ghc-prim,

152 integer-gmp

143 integer-gmp

153−

154− if flag(bytestring-builder)

155− build-depends: bytestring >= 0.9 && < 0.10.4,

156− bytestring-builder >= 0.10.4

157− else

158− build-depends: bytestring >= 0.10.4