master..3ca519b · text

git.iko.soy/text

on 3ca519b help sign in

.gitignore+4

since splitsnapshots

unifiedsplit

@@ -1,3 +1,7 @@

1dist

2+dist-boot

3+dist-install

2cabal-dev

4cabal-dev

3cabal.sandbox.config

5cabal.sandbox.config

6+ghc.mk

7+GNUmakefile

.hgtags+1

@@ -64,3 +64,4 @@ f8a60279237c1c28e094868740d9188dd94fcc9e 1.2.0.0

64631fd055e27fa10983e17e9ad36d1b285b85c661 1.2.1.3

658221a68ed5ec37028357df9c6c25f52320e54ad9 1.2.2.0

66ce1c5ac65eb9f5580114b4e380092b05a485e95d 1.2.2.1

67+d72c634a91ffe27ec5cd03921dd53efc8129d8f6 1.2.2.2

.travis.ymlnew file+111

@@ -0,0 +1,111 @@

1+# This Travis job script has been generated by a script via

2+#

3+# make_travis_yml_2.hs '-o' '.travis.yml' 'text.cabal'

4+#

5+# For more information, see https://github.com/hvr/multi-ghc-travis

6+#

7+language: c

8+sudo: false

10+# Bryan threatens to turn off Travis-CI if he gets any email :-)

11+notifications:

12+ email: false

13+

14+cache:

15+ directories:

16+ - $HOME/.cabal/packages

17+ - $HOME/.cabal/store

18+

19+before_cache:

20+ - rm -fv $HOME/.cabal/packages/hackage.haskell.org/build-reports.log

21+ # remove files that are regenerated by 'cabal update'

22+ - rm -fv $HOME/.cabal/packages/hackage.haskell.org/00-index.*

23+ - rm -fv $HOME/.cabal/packages/hackage.haskell.org/*.json

24+ - rm -fv $HOME/.cabal/packages/hackage.haskell.org/01-index.cache

25+ - rm -fv $HOME/.cabal/packages/hackage.haskell.org/01-index.tar

26+ - rm -fv $HOME/.cabal/packages/hackage.haskell.org/01-index.tar.idx

27+

28+matrix:

29+ include:

30+ - compiler: "ghc-7.0.4"

31+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

32+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.0.4], sources: [hvr-ghc]}}

33+ - compiler: "ghc-7.2.2"

34+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

35+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.2.2], sources: [hvr-ghc]}}

36+ - compiler: "ghc-7.4.2"

37+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

38+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.4.2], sources: [hvr-ghc]}}

39+ - compiler: "ghc-7.6.3"

40+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

41+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.6.3], sources: [hvr-ghc]}}

42+ - compiler: "ghc-7.8.4"

43+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

44+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.8.4], sources: [hvr-ghc]}}

45+ - compiler: "ghc-7.10.3"

46+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

47+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.10.3], sources: [hvr-ghc]}}

48+ - compiler: "ghc-8.0.2"

49+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

50+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-8.0.2], sources: [hvr-ghc]}}

51+ - compiler: "ghc-8.2.2"

52+ # env: TEST=--disable-tests BENCH=--disable-benchmarks

53+ addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-8.2.2], sources: [hvr-ghc]}}

54+

55+before_install:

56+ - HC=${CC}

57+ - HCPKG=${HC/ghc/ghc-pkg}

58+ - unset CC

59+ - PATH=/opt/ghc/bin:/opt/ghc-ppa-tools/bin:$PATH

60+ - PKGNAME='text'

61+

62+install:

63+ - cabal --version

64+ - echo "$(${HC} --version) [$(${HC} --print-project-git-commit-id 2> /dev/null || echo '?')]"

65+ - BENCH=${BENCH---enable-benchmarks}

66+ - TEST=${TEST---enable-tests}

67+ - HADDOCK=${HADDOCK-true}

68+ - INSTALLED=${INSTALLED-true}

69+ - travis_retry cabal update -v

70+ - sed -i.bak 's/^jobs:/-- jobs:/' ${HOME}/.cabal/config

71+ - rm -fv cabal.project.local

72+ - "echo 'packages: .' > cabal.project"

73+ - rm -f cabal.project.freeze

74+ - cabal new-build -w ${HC} ${TEST} ${BENCH} --dep -j2 all

75+ - cabal new-build -w ${HC} --disable-tests --disable-benchmarks --dep -j2 all

76+

77+# Here starts the actual work to be performed for the package under test;

78+# any command which exits with a non-zero exit code causes the build to fail.

79+script:

80+ - if [ -f configure.ac ]; then autoreconf -i; fi

81+ - rm -rf .ghc.environment.* dist/

82+ - cabal sdist # test that a source-distribution can be generated

83+ - cd dist/

84+ - SRCTAR=(${PKGNAME}-*.tar.gz)

85+ - SRC_BASENAME="${SRCTAR/%.tar.gz}"

86+ - tar -xvf "./$SRC_BASENAME.tar.gz"

87+ - cd "$SRC_BASENAME/"

88+## from here on, CWD is inside the extracted source-tarball

89+ - rm -fv cabal.project.local

90+ - "echo 'packages: .' > cabal.project"

91+ # this builds all libraries and executables (without tests/benchmarks)

92+ - rm -f cabal.project.freeze

93+ - cabal new-build -w ${HC} --disable-tests --disable-benchmarks all

94+ # this builds all libraries and executables (including tests/benchmarks)

95+ # - rm -rf ./dist-newstyle

96+

97+ # Build with installed constraints for packages in global-db

98+ - if $INSTALLED; then

99+ echo cabal new-build -w ${HC} --disable-tests --disable-benchmarks $(${HCPKG} list --global --simple-output --names-only | sed 's/$[a-zA-Z0-9-]\{1,\}$ */--constraint="\1 installed" /g') all | sh;

100+ else echo "Not building with installed constraints"; fi

101+

102+ # build & run tests, build benchmarks

103+ - cabal new-build -w ${HC} ${TEST} ${BENCH} all

104+ - if [ "x$TEST" = "x--enable-tests" ]; then cabal new-test -w ${HC} ${TEST} all; fi

105+

106+ # haddock

107+ - rm -rf ./dist-newstyle

108+ - if $HADDOCK; then cabal new-haddock -w ${HC} --disable-tests --disable-benchmarks all; else echo "Skipping haddock generation";fi

109+

110+# REGENDATA ["-o",".travis.yml","text.cabal"]

111+# EOF

Data/Text.hs+32−1

@@ -34,6 +34,7 @@

34-- with Unicode text (including normalization, regular expressions,

35-- non-standard encodings, text breaking, and locales), see

36-- <http://hackage.haskell.org/package/text-icu the text-icu package >.

37+--

38module Data.Text

39module Data.Text

39 (

40 (

@@ -43,6 +44,9 @@ module Data.Text

43 -- * Acceptable data

44 -- * Acceptable data

44 -- $replacement

45 -- $replacement

47+ -- * Definition of character

48+ -- $character_definition

49+

46 -- * Fusion

50 -- * Fusion

47 -- $fusion

51 -- $fusion

@@ -60,6 +64,7 @@ module Data.Text

60 , snoc

64 , snoc

61 , append

65 , append

62 , uncons

66 , uncons

67+ , unsnoc

63 , head

68 , head

64 , last

69 , last

65 , tail

70 , tail

@@ -246,6 +251,17 @@ import qualified GHC.Exts as Exts

246import Text.Printf (PrintfArg, formatArg, formatString)

251import Text.Printf (PrintfArg, formatArg, formatString)

247#endif

252#endif

248

253

254+-- $character-definition

255+--

256+-- This package uses the word /character/ to denote Unicode code points.

257+--

258+-- Note that this is not the same thing as a grapheme (e.g. a composition of

259+-- code points that form one visual symbol). For instance, consider the grapheme

260+-- @ä@. This symbol has two Unicode representations: a single code-point

261+-- representation @U+00E4@ (the @LATIN SMALL LETTER A WITH DIAERESIS@ code

262+-- point), and a two code point representation @U+0061@ (the @A@ code point) and

263+-- @U+0308@ (the @COMBINING DIAERESIS@ code point).

264+

249-- $strict

265-- $strict

250--

266--

251-- This package provides both strict and lazy 'Text' types. The

267-- This package provides both strict and lazy 'Text' types. The

@@ -545,6 +561,19 @@ init (Text arr off len) | len <= 0 = emptyError "init"

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

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

546 #-}

562 #-}

547

563

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

565+-- 'Text', or 'Nothing' if empty.

566+--

567+-- @since 1.2.3.0

568+unsnoc :: Text -> Maybe (Text, Char)

569+unsnoc (Text arr off len)

570+ | len <= 0 = Nothing

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

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

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

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

575+{-# INLINE [1] unsnoc #-}

576+

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

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

549-- fusion.

578-- fusion.

550null :: Text -> Bool

579null :: Text -> Bool

@@ -782,6 +811,8 @@ toUpper t = unstream (S.toUpper (stream t))

782-- guides disagree on whether the book name \"The Hill of the Red

811-- guides disagree on whether the book name \"The Hill of the Red

783-- Fox\" is correctly title cased—but this function will

812-- Fox\" is correctly title cased—but this function will

784-- capitalize /every/ word.

813-- capitalize /every/ word.

814+--

815+-- @since 1.0.0.0

785toTitle :: Text -> Text

816toTitle :: Text -> Text

786toTitle t = unstream (S.toTitle (stream t))

817toTitle t = unstream (S.toTitle (stream t))

787{-# INLINE toTitle #-}

818{-# INLINE toTitle #-}

@@ -1099,8 +1130,8 @@ takeEnd n t@(Text arr off len)

1099iterNEnd :: Int -> Text -> Int

1130iterNEnd :: Int -> Text -> Int

1100iterNEnd n t@(Text _arr _off len) = loop (len-1) n

1131iterNEnd n t@(Text _arr _off len) = loop (len-1) n

1101 where loop i !m

1132 where loop i !m

1133+ | m <= 0 = i+1

1102 | i <= 0 = 0

1134 | i <= 0 = 0

1103− | m <= 1 = i

1104 | otherwise = loop (i+d) (m-1)

1135 | otherwise = loop (i+d) (m-1)

1105 where d = reverseIter_ t i

1136 where d = reverseIter_ t i

1106

1137

Data/Text/Encoding.hs+12−2

@@ -79,7 +79,11 @@ import Data.Text.Internal.Unsafe.Shift (shiftR)

79import Data.Text.Show ()

80import Data.Text.Unsafe (unsafeDupablePerformIO)

81import Data.Word (Word8, Word32)

82+#if __GLASGOW_HASKELL__ >= 703

82−import Foreign.C.Types (CSize(..))

83+import Foreign.C.Types (CSize(CSize))

84+#else

85+import Foreign.C.Types (CSize)

86+#endif

83import Foreign.ForeignPtr (withForeignPtr)

87import Foreign.ForeignPtr (withForeignPtr)

84import Foreign.Marshal.Utils (with)

88import Foreign.Marshal.Utils (with)

85import Foreign.Ptr (Ptr, minusPtr, nullPtr, plusPtr)

89import Foreign.Ptr (Ptr, minusPtr, nullPtr, plusPtr)

@@ -209,6 +213,8 @@ decodeUtf8With onErr (PS fp off len) = runText $ \done -> do

209-- or continuation where it is encountered.

213-- or continuation where it is encountered.

210

214

211-- | A stream oriented decoding result.

215-- | A stream oriented decoding result.

216+--

217+-- @since 1.0.0.0

212data Decoding = Some Text ByteString (ByteString -> Decoding)

218data Decoding = Some Text ByteString (ByteString -> Decoding)

213

219

214instance Show Decoding where

220instance Show Decoding where

@@ -228,11 +234,15 @@ newtype DecoderState = DecoderState Word32 deriving (Eq, Show, Num, Storable)

228-- thrown (either by this function or a continuation) that cannot be

234-- thrown (either by this function or a continuation) that cannot be

229-- caught in pure code. For more control over the handling of invalid

235-- caught in pure code. For more control over the handling of invalid

230-- data, use 'streamDecodeUtf8With'.

236-- data, use 'streamDecodeUtf8With'.

237+--

238+-- @since 1.0.0.0

231streamDecodeUtf8 :: ByteString -> Decoding

239streamDecodeUtf8 :: ByteString -> Decoding

232streamDecodeUtf8 = streamDecodeUtf8With strictDecode

240streamDecodeUtf8 = streamDecodeUtf8With strictDecode

233

241

234-- | Decode, in a stream oriented way, a 'ByteString' containing UTF-8

242-- | Decode, in a stream oriented way, a 'ByteString' containing UTF-8

235-- encoded text.

243-- encoded text.

244+--

245+-- @since 1.0.0.0

236streamDecodeUtf8With :: OnDecodeError -> ByteString -> Decoding

246streamDecodeUtf8With :: OnDecodeError -> ByteString -> Decoding

237streamDecodeUtf8With onErr = decodeChunk B.empty 0 0

247streamDecodeUtf8With onErr = decodeChunk B.empty 0 0

238 where

248 where

@@ -373,7 +383,7 @@ encodeUtf8 :: Text -> ByteString

373encodeUtf8 (Text arr off len)

383encodeUtf8 (Text arr off len)

374 | len == 0 = B.empty

384 | len == 0 = B.empty

375 | otherwise = unsafeDupablePerformIO $ do

385 | otherwise = unsafeDupablePerformIO $ do

376− fp <- mallocByteString (len*4)

386+ fp <- mallocByteString (len*3) -- see https://github.com/haskell/text/issues/194 for why len*3 is enough

377 withForeignPtr fp $ \ptr ->

387 withForeignPtr fp $ \ptr ->

378 with ptr $ \destPtr -> do

388 with ptr $ \destPtr -> do

379 c_encode_utf8 destPtr (A.aBA arr) (fromIntegral off) (fromIntegral len)

389 c_encode_utf8 destPtr (A.aBA arr) (fromIntegral off) (fromIntegral len)

Data/Text/Foreign.hs+5−1

@@ -108,7 +108,7 @@ takeWord16 (I16 n) t@(Text arr off len)

108 | n >= len || m >= len = t

109 | otherwise = Text arr off m

110 where

111− m | w < 0xDB00 || w > 0xD8FF = n

111+ m | w < 0xD800 || w > 0xDBFF = n

112 | otherwise = n+1

113 w = A.unsafeIndex arr (off+n-1)

114

@@ -157,6 +157,8 @@ asForeignPtr t@(Text _arr _off len) = do

157-- | /O(n)/ Decode a C string with explicit length, which is assumed

158-- to have been encoded as UTF-8. If decoding fails, a

159-- 'UnicodeException' is thrown.

160+--

161+-- @since 1.0.0.0

160peekCStringLen :: CStringLen -> IO Text

162peekCStringLen :: CStringLen -> IO Text

161peekCStringLen cs = do

163peekCStringLen cs = do

162 bs <- unsafePackCStringLen cs

164 bs <- unsafePackCStringLen cs

@@ -169,5 +171,7 @@ peekCStringLen cs = do

169-- The temporary storage is freed when the subcomputation terminates

171-- The temporary storage is freed when the subcomputation terminates

170-- (either normally or via an exception), so the pointer to the

172-- (either normally or via an exception), so the pointer to the

171-- temporary storage must /not/ be used after this function returns.

173-- temporary storage must /not/ be used after this function returns.

174+--

175+-- @since 1.0.0.0

172withCStringLen :: Text -> (CStringLen -> IO a) -> IO a

176withCStringLen :: Text -> (CStringLen -> IO a) -> IO a

173withCStringLen t act = unsafeUseAsCStringLen (encodeUtf8 t) act

177withCStringLen t act = unsafeUseAsCStringLen (encodeUtf8 t) act

Data/Text/Internal/Fusion/CaseMapping.hs+164−4

@@ -1,10 +1,10 @@

1{-# LANGUAGE Rank2Types #-}

2-- AUTOMATICALLY GENERATED - DO NOT EDIT

3-- Generated by scripts/CaseMapping.hs

4−-- CaseFolding-8.0.0.txt

4+-- CaseFolding-9.0.0.txt

5−-- Date: 2015-01-13, 18:16:36 GMT [MD]

5+-- Date: 2016-03-02, 18:54:54 GMT

6−-- SpecialCasing-8.0.0.txt

6+-- SpecialCasing-9.0.0.txt

7−-- Date: 2014-12-16, 23:08:04 GMT [MD]

7+-- Date: 2016-03-02, 18:55:13 GMT

9module Data.Text.Internal.Fusion.CaseMapping where

10import Data.Char

@@ -371,6 +371,24 @@ foldMapping '\x13fb' s = Yield '\x13f3' (CC s '\x0000' '\x0000')

371foldMapping '\x13fc' s = Yield '\x13f4' (CC s '\x0000' '\x0000')

372-- CHEROKEE SMALL LETTER MV

373foldMapping '\x13fd' s = Yield '\x13f5' (CC s '\x0000' '\x0000')

374+-- CYRILLIC SMALL LETTER ROUNDED VE

375+foldMapping '\x1c80' s = Yield '\x0432' (CC s '\x0000' '\x0000')

376+-- CYRILLIC SMALL LETTER LONG-LEGGED DE

377+foldMapping '\x1c81' s = Yield '\x0434' (CC s '\x0000' '\x0000')

378+-- CYRILLIC SMALL LETTER NARROW O

379+foldMapping '\x1c82' s = Yield '\x043e' (CC s '\x0000' '\x0000')

380+-- CYRILLIC SMALL LETTER WIDE ES

381+foldMapping '\x1c83' s = Yield '\x0441' (CC s '\x0000' '\x0000')

382+-- CYRILLIC SMALL LETTER TALL TE

383+foldMapping '\x1c84' s = Yield '\x0442' (CC s '\x0000' '\x0000')

384+-- CYRILLIC SMALL LETTER THREE-LEGGED TE

385+foldMapping '\x1c85' s = Yield '\x0442' (CC s '\x0000' '\x0000')

386+-- CYRILLIC SMALL LETTER TALL HARD SIGN

387+foldMapping '\x1c86' s = Yield '\x044a' (CC s '\x0000' '\x0000')

388+-- CYRILLIC SMALL LETTER TALL YAT

389+foldMapping '\x1c87' s = Yield '\x0463' (CC s '\x0000' '\x0000')

390+-- CYRILLIC SMALL LETTER UNBLENDED UK

391+foldMapping '\x1c88' s = Yield '\xa64b' (CC s '\x0000' '\x0000')

374-- LATIN SMALL LETTER H WITH LINE BELOW

392-- LATIN SMALL LETTER H WITH LINE BELOW

375foldMapping '\x1e96' s = Yield '\x0068' (CC s '\x0331' '\x0000')

393foldMapping '\x1e96' s = Yield '\x0068' (CC s '\x0331' '\x0000')

376-- LATIN SMALL LETTER T WITH DIAERESIS

394-- LATIN SMALL LETTER T WITH DIAERESIS

@@ -545,6 +563,8 @@ foldMapping '\x1ff6' s = Yield '\x03c9' (CC s '\x0342' '\x0000')

545foldMapping '\x1ff7' s = Yield '\x03c9' (CC s '\x0342' '\x03b9')

563foldMapping '\x1ff7' s = Yield '\x03c9' (CC s '\x0342' '\x03b9')

546-- GREEK CAPITAL LETTER OMEGA WITH PROSGEGRAMMENI

564-- GREEK CAPITAL LETTER OMEGA WITH PROSGEGRAMMENI

547foldMapping '\x1ffc' s = Yield '\x03c9' (CC s '\x03b9' '\x0000')

565foldMapping '\x1ffc' s = Yield '\x03c9' (CC s '\x03b9' '\x0000')

566+-- LATIN CAPITAL LETTER SMALL CAPITAL I

567+foldMapping '\xa7ae' s = Yield '\x026a' (CC s '\x0000' '\x0000')

548-- LATIN CAPITAL LETTER J WITH CROSSED-TAIL

568-- LATIN CAPITAL LETTER J WITH CROSSED-TAIL

549foldMapping '\xa7b2' s = Yield '\x029d' (CC s '\x0000' '\x0000')

569foldMapping '\xa7b2' s = Yield '\x029d' (CC s '\x0000' '\x0000')

550-- LATIN CAPITAL LETTER CHI

570-- LATIN CAPITAL LETTER CHI

@@ -737,6 +757,78 @@ foldMapping '\xfb15' s = Yield '\x0574' (CC s '\x056b' '\x0000')

737foldMapping '\xfb16' s = Yield '\x057e' (CC s '\x0576' '\x0000')

757foldMapping '\xfb16' s = Yield '\x057e' (CC s '\x0576' '\x0000')

738-- ARMENIAN SMALL LIGATURE MEN XEH

758-- ARMENIAN SMALL LIGATURE MEN XEH

739foldMapping '\xfb17' s = Yield '\x0574' (CC s '\x056d' '\x0000')

759foldMapping '\xfb17' s = Yield '\x0574' (CC s '\x056d' '\x0000')

760+-- OSAGE CAPITAL LETTER A

761+foldMapping '\x104b0' s = Yield '\x104d8' (CC s '\x0000' '\x0000')

762+-- OSAGE CAPITAL LETTER AI

763+foldMapping '\x104b1' s = Yield '\x104d9' (CC s '\x0000' '\x0000')

764+-- OSAGE CAPITAL LETTER AIN

765+foldMapping '\x104b2' s = Yield '\x104da' (CC s '\x0000' '\x0000')

766+-- OSAGE CAPITAL LETTER AH

767+foldMapping '\x104b3' s = Yield '\x104db' (CC s '\x0000' '\x0000')

768+-- OSAGE CAPITAL LETTER BRA

769+foldMapping '\x104b4' s = Yield '\x104dc' (CC s '\x0000' '\x0000')

770+-- OSAGE CAPITAL LETTER CHA

771+foldMapping '\x104b5' s = Yield '\x104dd' (CC s '\x0000' '\x0000')

772+-- OSAGE CAPITAL LETTER EHCHA

773+foldMapping '\x104b6' s = Yield '\x104de' (CC s '\x0000' '\x0000')

774+-- OSAGE CAPITAL LETTER E

775+foldMapping '\x104b7' s = Yield '\x104df' (CC s '\x0000' '\x0000')

776+-- OSAGE CAPITAL LETTER EIN

777+foldMapping '\x104b8' s = Yield '\x104e0' (CC s '\x0000' '\x0000')

778+-- OSAGE CAPITAL LETTER HA

779+foldMapping '\x104b9' s = Yield '\x104e1' (CC s '\x0000' '\x0000')

780+-- OSAGE CAPITAL LETTER HYA

781+foldMapping '\x104ba' s = Yield '\x104e2' (CC s '\x0000' '\x0000')

782+-- OSAGE CAPITAL LETTER I

783+foldMapping '\x104bb' s = Yield '\x104e3' (CC s '\x0000' '\x0000')

784+-- OSAGE CAPITAL LETTER KA

785+foldMapping '\x104bc' s = Yield '\x104e4' (CC s '\x0000' '\x0000')

786+-- OSAGE CAPITAL LETTER EHKA

787+foldMapping '\x104bd' s = Yield '\x104e5' (CC s '\x0000' '\x0000')

788+-- OSAGE CAPITAL LETTER KYA

789+foldMapping '\x104be' s = Yield '\x104e6' (CC s '\x0000' '\x0000')

790+-- OSAGE CAPITAL LETTER LA

791+foldMapping '\x104bf' s = Yield '\x104e7' (CC s '\x0000' '\x0000')

792+-- OSAGE CAPITAL LETTER MA

793+foldMapping '\x104c0' s = Yield '\x104e8' (CC s '\x0000' '\x0000')

794+-- OSAGE CAPITAL LETTER NA

795+foldMapping '\x104c1' s = Yield '\x104e9' (CC s '\x0000' '\x0000')

796+-- OSAGE CAPITAL LETTER O

797+foldMapping '\x104c2' s = Yield '\x104ea' (CC s '\x0000' '\x0000')

798+-- OSAGE CAPITAL LETTER OIN

799+foldMapping '\x104c3' s = Yield '\x104eb' (CC s '\x0000' '\x0000')

800+-- OSAGE CAPITAL LETTER PA

801+foldMapping '\x104c4' s = Yield '\x104ec' (CC s '\x0000' '\x0000')

802+-- OSAGE CAPITAL LETTER EHPA

803+foldMapping '\x104c5' s = Yield '\x104ed' (CC s '\x0000' '\x0000')

804+-- OSAGE CAPITAL LETTER SA

805+foldMapping '\x104c6' s = Yield '\x104ee' (CC s '\x0000' '\x0000')

806+-- OSAGE CAPITAL LETTER SHA

807+foldMapping '\x104c7' s = Yield '\x104ef' (CC s '\x0000' '\x0000')

808+-- OSAGE CAPITAL LETTER TA

809+foldMapping '\x104c8' s = Yield '\x104f0' (CC s '\x0000' '\x0000')

810+-- OSAGE CAPITAL LETTER EHTA

811+foldMapping '\x104c9' s = Yield '\x104f1' (CC s '\x0000' '\x0000')

812+-- OSAGE CAPITAL LETTER TSA

813+foldMapping '\x104ca' s = Yield '\x104f2' (CC s '\x0000' '\x0000')

814+-- OSAGE CAPITAL LETTER EHTSA

815+foldMapping '\x104cb' s = Yield '\x104f3' (CC s '\x0000' '\x0000')

816+-- OSAGE CAPITAL LETTER TSHA

817+foldMapping '\x104cc' s = Yield '\x104f4' (CC s '\x0000' '\x0000')

818+-- OSAGE CAPITAL LETTER DHA

819+foldMapping '\x104cd' s = Yield '\x104f5' (CC s '\x0000' '\x0000')

820+-- OSAGE CAPITAL LETTER U

821+foldMapping '\x104ce' s = Yield '\x104f6' (CC s '\x0000' '\x0000')

822+-- OSAGE CAPITAL LETTER WA

823+foldMapping '\x104cf' s = Yield '\x104f7' (CC s '\x0000' '\x0000')

824+-- OSAGE CAPITAL LETTER KHA

825+foldMapping '\x104d0' s = Yield '\x104f8' (CC s '\x0000' '\x0000')

826+-- OSAGE CAPITAL LETTER GHA

827+foldMapping '\x104d1' s = Yield '\x104f9' (CC s '\x0000' '\x0000')

828+-- OSAGE CAPITAL LETTER ZA

829+foldMapping '\x104d2' s = Yield '\x104fa' (CC s '\x0000' '\x0000')

830+-- OSAGE CAPITAL LETTER ZHA

831+foldMapping '\x104d3' s = Yield '\x104fb' (CC s '\x0000' '\x0000')

740-- OLD HUNGARIAN CAPITAL LETTER A

832-- OLD HUNGARIAN CAPITAL LETTER A

741foldMapping '\x10c80' s = Yield '\x10cc0' (CC s '\x0000' '\x0000')

833foldMapping '\x10c80' s = Yield '\x10cc0' (CC s '\x0000' '\x0000')

742-- OLD HUNGARIAN CAPITAL LETTER AA

834-- OLD HUNGARIAN CAPITAL LETTER AA

@@ -839,4 +931,72 @@ foldMapping '\x10cb0' s = Yield '\x10cf0' (CC s '\x0000' '\x0000')

839foldMapping '\x10cb1' s = Yield '\x10cf1' (CC s '\x0000' '\x0000')

931foldMapping '\x10cb1' s = Yield '\x10cf1' (CC s '\x0000' '\x0000')

840-- OLD HUNGARIAN CAPITAL LETTER US

932-- OLD HUNGARIAN CAPITAL LETTER US

841foldMapping '\x10cb2' s = Yield '\x10cf2' (CC s '\x0000' '\x0000')

933foldMapping '\x10cb2' s = Yield '\x10cf2' (CC s '\x0000' '\x0000')

934+-- ADLAM CAPITAL LETTER ALIF

935+foldMapping '\x1e900' s = Yield '\x1e922' (CC s '\x0000' '\x0000')

936+-- ADLAM CAPITAL LETTER DAALI

937+foldMapping '\x1e901' s = Yield '\x1e923' (CC s '\x0000' '\x0000')

938+-- ADLAM CAPITAL LETTER LAAM

939+foldMapping '\x1e902' s = Yield '\x1e924' (CC s '\x0000' '\x0000')

940+-- ADLAM CAPITAL LETTER MIIM

941+foldMapping '\x1e903' s = Yield '\x1e925' (CC s '\x0000' '\x0000')

942+-- ADLAM CAPITAL LETTER BA

943+foldMapping '\x1e904' s = Yield '\x1e926' (CC s '\x0000' '\x0000')

944+-- ADLAM CAPITAL LETTER SINNYIIYHE

945+foldMapping '\x1e905' s = Yield '\x1e927' (CC s '\x0000' '\x0000')

946+-- ADLAM CAPITAL LETTER PE

947+foldMapping '\x1e906' s = Yield '\x1e928' (CC s '\x0000' '\x0000')

948+-- ADLAM CAPITAL LETTER BHE

949+foldMapping '\x1e907' s = Yield '\x1e929' (CC s '\x0000' '\x0000')

950+-- ADLAM CAPITAL LETTER RA

951+foldMapping '\x1e908' s = Yield '\x1e92a' (CC s '\x0000' '\x0000')

952+-- ADLAM CAPITAL LETTER E

953+foldMapping '\x1e909' s = Yield '\x1e92b' (CC s '\x0000' '\x0000')

954+-- ADLAM CAPITAL LETTER FA

955+foldMapping '\x1e90a' s = Yield '\x1e92c' (CC s '\x0000' '\x0000')

956+-- ADLAM CAPITAL LETTER I

957+foldMapping '\x1e90b' s = Yield '\x1e92d' (CC s '\x0000' '\x0000')

958+-- ADLAM CAPITAL LETTER O

959+foldMapping '\x1e90c' s = Yield '\x1e92e' (CC s '\x0000' '\x0000')

960+-- ADLAM CAPITAL LETTER DHA

961+foldMapping '\x1e90d' s = Yield '\x1e92f' (CC s '\x0000' '\x0000')

962+-- ADLAM CAPITAL LETTER YHE

963+foldMapping '\x1e90e' s = Yield '\x1e930' (CC s '\x0000' '\x0000')

964+-- ADLAM CAPITAL LETTER WAW

965+foldMapping '\x1e90f' s = Yield '\x1e931' (CC s '\x0000' '\x0000')

966+-- ADLAM CAPITAL LETTER NUN

967+foldMapping '\x1e910' s = Yield '\x1e932' (CC s '\x0000' '\x0000')

968+-- ADLAM CAPITAL LETTER KAF

969+foldMapping '\x1e911' s = Yield '\x1e933' (CC s '\x0000' '\x0000')

970+-- ADLAM CAPITAL LETTER YA

971+foldMapping '\x1e912' s = Yield '\x1e934' (CC s '\x0000' '\x0000')

972+-- ADLAM CAPITAL LETTER U

973+foldMapping '\x1e913' s = Yield '\x1e935' (CC s '\x0000' '\x0000')

974+-- ADLAM CAPITAL LETTER JIIM

975+foldMapping '\x1e914' s = Yield '\x1e936' (CC s '\x0000' '\x0000')

976+-- ADLAM CAPITAL LETTER CHI

977+foldMapping '\x1e915' s = Yield '\x1e937' (CC s '\x0000' '\x0000')

978+-- ADLAM CAPITAL LETTER HA

979+foldMapping '\x1e916' s = Yield '\x1e938' (CC s '\x0000' '\x0000')

980+-- ADLAM CAPITAL LETTER QAAF

981+foldMapping '\x1e917' s = Yield '\x1e939' (CC s '\x0000' '\x0000')

982+-- ADLAM CAPITAL LETTER GA

983+foldMapping '\x1e918' s = Yield '\x1e93a' (CC s '\x0000' '\x0000')

984+-- ADLAM CAPITAL LETTER NYA

985+foldMapping '\x1e919' s = Yield '\x1e93b' (CC s '\x0000' '\x0000')

986+-- ADLAM CAPITAL LETTER TU

987+foldMapping '\x1e91a' s = Yield '\x1e93c' (CC s '\x0000' '\x0000')

988+-- ADLAM CAPITAL LETTER NHA

989+foldMapping '\x1e91b' s = Yield '\x1e93d' (CC s '\x0000' '\x0000')

990+-- ADLAM CAPITAL LETTER VA

991+foldMapping '\x1e91c' s = Yield '\x1e93e' (CC s '\x0000' '\x0000')

992+-- ADLAM CAPITAL LETTER KHA

993+foldMapping '\x1e91d' s = Yield '\x1e93f' (CC s '\x0000' '\x0000')

994+-- ADLAM CAPITAL LETTER GBE

995+foldMapping '\x1e91e' s = Yield '\x1e940' (CC s '\x0000' '\x0000')

996+-- ADLAM CAPITAL LETTER ZAL

997+foldMapping '\x1e91f' s = Yield '\x1e941' (CC s '\x0000' '\x0000')

998+-- ADLAM CAPITAL LETTER KPO

999+foldMapping '\x1e920' s = Yield '\x1e942' (CC s '\x0000' '\x0000')

1000+-- ADLAM CAPITAL LETTER SHA

1001+foldMapping '\x1e921' s = Yield '\x1e943' (CC s '\x0000' '\x0000')

842foldMapping c s = Yield (toLower c) (CC s '\0' '\0')

1002foldMapping c s = Yield (toLower c) (CC s '\0' '\0')

Data/Text/Internal/Fusion/Common.hs+44−37

@@ -1,4 +1,4 @@

1−{-# LANGUAGE BangPatterns, MagicHash, Rank2Types #-}

1+{-# LANGUAGE PatternGuards, BangPatterns, MagicHash, Rank2Types #-}

2-- |

3-- Module : Data.Text.Internal.Fusion.Common

4-- Copyright : (c) Bryan O'Sullivan 2009, 2012

@@ -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)

110+import Data.Char (isLetter, isSpace)

111import Data.Int (Int64)

112import Data.Text.Internal.Fusion.Types

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

@@ -117,7 +117,7 @@ import GHC.Prim (Addr#, chr#, indexCharOffAddr#, ord#)

117import GHC.Types (Char(..), Int(..))

118

119singleton :: Char -> Stream Char

120−singleton c = Stream next False 1

120+singleton c = Stream next False (codePointsSize 1)

121 where next False = Yield c True

122 next True = Done

123{-# INLINE [0] singleton #-}

@@ -175,7 +175,7 @@ data C s = C0 !s

175

176-- | /O(n)/ Adds a character to the front of a Stream Char.

177cons :: Char -> Stream Char -> Stream Char

178−cons !w (Stream next0 s0 len) = Stream next (C1 s0) (len+1)

178+cons !w (Stream next0 s0 len) = Stream next (C1 s0) (len + codePointsSize 1)

179 where

180 next (C1 s) = Yield w (C0 s)

181 next (C0 s) = case next0 s of

@@ -189,7 +189,7 @@ data Snoc a = N

189

190-- | /O(n)/ Adds a character to the end of a stream.

191snoc :: Stream Char -> Char -> Stream Char

192−snoc (Stream next0 xs0 len) w = Stream next (J xs0) (len+1)

192+snoc (Stream next0 xs0 len) w = Stream next (J xs0) (len + codePointsSize 1)

193 where

194 next (J xs) = case next0 xs of

195 Done -> Yield w N

@@ -237,7 +237,7 @@ uncons :: Stream Char -> Maybe (Char, Stream Char)

237uncons (Stream next s0 len) = loop_uncons s0

238 where

239 loop_uncons !s = case next s of

240− Yield x s1 -> Just (x, Stream next s1 (len-1))

240+ Yield x s1 -> Just (x, Stream next s1 (len - codePointsSize 1))

241 Skip s' -> loop_uncons s'

242 Done -> Nothing

243{-# INLINE [0] uncons #-}

@@ -260,7 +260,7 @@ last (Stream next s0 _len) = loop0_last s0

260-- | /O(1)/ Returns all characters after the head of a Stream Char, which must

261-- be non-empty.

262tail :: Stream Char -> Stream Char

263−tail (Stream next0 s0 len) = Stream next (C0 s0) (len-1)

263+tail (Stream next0 s0 len) = Stream next (C0 s0) (len - codePointsSize 1)

264 where

265 next (C0 s) = case next0 s of

266 Done -> emptyError "tail"

@@ -278,7 +278,7 @@ data Init s = Init0 !s

278-- | /O(1)/ Returns all but the last character of a Stream Char, which

279-- must be non-empty.

280init :: Stream Char -> Stream Char

281−init (Stream next0 s0 len) = Stream next (Init0 s0) (len-1)

281+init (Stream next0 s0 len) = Stream next (Init0 s0) (len - codePointsSize 1)

282 where

283 next (Init0 s) = case next0 s of

284 Done -> emptyError "init"

@@ -318,11 +318,14 @@ lengthI (Stream next s0 _len) = loop_length 0 s0

318-- greater than the number or if the stream can't possibly be as long

319-- as the number supplied, and hence be more efficient.

320compareLengthI :: Integral a => Stream Char -> a -> Ordering

321−compareLengthI (Stream next s0 len) n =

321+compareLengthI (Stream next s0 len) n

322− case compareSize len (fromIntegral n) of

322+ -- Note that @len@ tracks code units whereas we want to compare the length

323− Just o -> o

323+ -- in code points. Specifically, a stream with hint @len@ may consist of

324+ -- anywhere from @len/2@ to @len@ code points.

325+ | Just r <- compareSize len n' = r

324− Nothing -> loop_cmp 0 s0

326+ | otherwise = loop_cmp 0 s0

325 where

327 where

328+ n' = codePointsSize $ fromIntegral n

326 loop_cmp !z s = case next s of

329 loop_cmp !z s = case next s of

327 Done -> compare z n

330 Done -> compare z n

328 Skip s' -> loop_cmp z s'

331 Skip s' -> loop_cmp z s'

@@ -368,7 +371,7 @@ data I s = I1 !s

368-- | /O(n)/ Take a character and place it between each of the

371-- | /O(n)/ Take a character and place it between each of the

369-- characters of a 'Stream Char'.

372-- characters of a 'Stream Char'.

370intersperse :: Char -> Stream Char -> Stream Char

373intersperse :: Char -> Stream Char -> Stream Char

371−intersperse c (Stream next0 s0 len) = Stream next (I1 s0) len

374+intersperse c (Stream next0 s0 len) = Stream next (I1 s0) (len + unknownSize)

372 where

375 where

373 next (I1 s) = case next0 s of

376 next (I1 s) = case next0 s of

374 Done -> Done

377 Done -> Done

@@ -393,9 +396,11 @@ intersperse c (Stream next0 s0 len) = Stream next (I1 s0) len

393-- functions may map one input character to two or three output

396-- functions may map one input character to two or three output

394-- characters.

397-- characters.

395

398

399+-- | Map a 'Stream' through the given case-mapping function.

396caseConvert :: (forall s. Char -> s -> Step (CC s) Char)

400caseConvert :: (forall s. Char -> s -> Step (CC s) Char)

397 -> Stream Char -> Stream Char

401 -> Stream Char -> Stream Char

398−caseConvert remap (Stream next0 s0 len) = Stream next (CC s0 '\0' '\0') len

402+caseConvert remap (Stream next0 s0 len) =

403+ Stream next (CC s0 '\0' '\0') (len `unionSize` 3*len)

399 where

404 where

400 next (CC s '\0' _) =

405 next (CC s '\0' _) =

401 case next0 s of

406 case next0 s of

@@ -458,19 +463,20 @@ toLower = caseConvert lowerMapping

458-- Fox\" is correctly title cased—but this function will

463-- Fox\" is correctly title cased—but this function will

459-- capitalize /every/ word.

464-- capitalize /every/ word.

460toTitle :: Stream Char -> Stream Char

465toTitle :: Stream Char -> Stream Char

461−toTitle (Stream next0 s0 len) = Stream next (CC (False :*: s0) '\0' '\0') len

466+toTitle (Stream next0 s0 len) = Stream next (CC (False :*: s0) '\0' '\0') (len + unknownSize)

462 where

467 where

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

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

464 case next0 s of

469 case next0 s of

465− Done -> Done

470+ Done -> Done

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

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

467 Yield c s'

472 Yield c s'

468− | letter' -> if letter

473+ | nonSpace -> if letter

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

474+ then lowerMapping c (nonSpace :*: s')

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

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

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

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

477+ where nonSpace = P.not (isSpace c)

472− where letter' = isLetter c

478+ letter' = isLetter c

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

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

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

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

475

481

476data Justify i s = Just1 !i !s

482data Justify i s = Just1 !i !s

@@ -478,7 +484,7 @@ data Justify i s = Just1 !i !s

478

484

479justifyLeftI :: Integral a => a -> Char -> Stream Char -> Stream Char

485justifyLeftI :: Integral a => a -> Char -> Stream Char -> Stream Char

480justifyLeftI k c (Stream next0 s0 len) =

486justifyLeftI k c (Stream next0 s0 len) =

481− Stream next (Just1 0 s0) (larger (fromIntegral k) len)

487+ Stream next (Just1 0 s0) (larger (fromIntegral k * charSize c + len) len)

482 where

488 where

483 next (Just1 n s) =

489 next (Just1 n s) =

484 case next0 s of

490 case next0 s of

@@ -698,7 +704,7 @@ replicateI n (Stream next0 s0 len) =

698-- (a,b), in which case, a is the next Char in the string, and b is

704-- (a,b), in which case, a is the next Char in the string, and b is

699-- the seed value for further production.

705-- the seed value for further production.

700unfoldr :: (a -> Maybe (Char,a)) -> a -> Stream Char

706unfoldr :: (a -> Maybe (Char,a)) -> a -> Stream Char

701−unfoldr f s0 = Stream next s0 1 -- HINT maybe too low

707+unfoldr f s0 = Stream next s0 unknownSize

702 where

708 where

703 {-# INLINE next #-}

709 {-# INLINE next #-}

704 next !s = case f s of

710 next !s = case f s of

@@ -712,7 +718,7 @@ unfoldr f s0 = Stream next s0 1 -- HINT maybe too low

712-- 'unfoldr' when the length of the result is known.

718-- 'unfoldr' when the length of the result is known.

713unfoldrNI :: Integral a => a -> (b -> Maybe (Char,b)) -> b -> Stream Char

719unfoldrNI :: Integral a => a -> (b -> Maybe (Char,b)) -> b -> Stream Char

714unfoldrNI n f s0 | n < 0 = empty

720unfoldrNI n f s0 | n < 0 = empty

715− | otherwise = Stream next (0 :*: s0) (fromIntegral (n*2)) -- HINT maybe too high

721+ | otherwise = Stream next (0 :*: s0) (maxSize $ fromIntegral (n*2))

716 where

722 where

717 {-# INLINE next #-}

723 {-# INLINE next #-}

718 next (z :*: s) = case f s of

724 next (z :*: s) = case f s of

@@ -724,12 +730,12 @@ unfoldrNI n f s0 | n < 0 = empty

724-------------------------------------------------------------------------------

730-------------------------------------------------------------------------------

725-- * Substreams

731-- * Substreams

726

732

727−-- | /O(n)/ take n, applied to a stream, returns the prefix of the

733+-- | /O(n)/ @'take' n@, applied to a stream, returns the prefix of the

728-- stream of length @n@, or the stream itself if @n@ is greater than the

734-- stream of length @n@, or the stream itself if @n@ is greater than the

729-- length of the stream.

735-- length of the stream.

730take :: Integral a => a -> Stream Char -> Stream Char

736take :: Integral a => a -> Stream Char -> Stream Char

731take n0 (Stream next0 s0 len) =

737take n0 (Stream next0 s0 len) =

732− Stream next (n0 :*: s0) (smaller len (fromIntegral (max 0 n0)))

738+ Stream next (n0 :*: s0) (smaller len (codePointsSize $ fromIntegral n0))

733 where

739 where

734 {-# INLINE next #-}

740 {-# INLINE next #-}

735 next (n :*: s) | n <= 0 = Done

741 next (n :*: s) | n <= 0 = Done

@@ -742,12 +748,12 @@ take n0 (Stream next0 s0 len) =

742data Drop a s = NS !s

748data Drop a s = NS !s

743 | JS !a !s

749 | JS !a !s

744

750

745−-- | /O(n)/ drop n, applied to a stream, returns the suffix of the

751+-- | /O(n)/ @'drop' n@, applied to a stream, returns the suffix of the

746-- stream after the first @n@ characters, or the empty stream if @n@

752-- stream after the first @n@ characters, or the empty stream if @n@

747-- is greater than the length of the stream.

753-- is greater than the length of the stream.

748drop :: Integral a => a -> Stream Char -> Stream Char

754drop :: Integral a => a -> Stream Char -> Stream Char

749drop n0 (Stream next0 s0 len) =

755drop n0 (Stream next0 s0 len) =

750− Stream next (JS n0 s0) (len - fromIntegral (max 0 n0))

756+ Stream next (JS n0 s0) (len - codePointsSize (fromIntegral n0))

751 where

757 where

752 {-# INLINE next #-}

758 {-# INLINE next #-}

753 next (JS n s)

759 next (JS n s)

@@ -762,10 +768,10 @@ drop n0 (Stream next0 s0 len) =

762 Yield x s' -> Yield x (NS s')

768 Yield x s' -> Yield x (NS s')

763{-# INLINE [0] drop #-}

769{-# INLINE [0] drop #-}

764

770

765−-- | takeWhile, applied to a predicate @p@ and a stream, returns the

771+-- | 'takeWhile', applied to a predicate @p@ and a stream, returns the

766−-- longest prefix (possibly empty) of elements that satisfy p.

772+-- longest prefix (possibly empty) of elements that satisfy @p@.

767takeWhile :: (Char -> Bool) -> Stream Char -> Stream Char

773takeWhile :: (Char -> Bool) -> Stream Char -> Stream Char

768−takeWhile p (Stream next0 s0 len) = Stream next s0 len -- HINT maybe too high

774+takeWhile p (Stream next0 s0 len) = Stream next s0 (len - unknownSize)

769 where

775 where

770 {-# INLINE next #-}

776 {-# INLINE next #-}

771 next !s = case next0 s of

777 next !s = case next0 s of

@@ -775,9 +781,9 @@ takeWhile p (Stream next0 s0 len) = Stream next s0 len -- HINT maybe too high

775 | otherwise -> Done

781 | otherwise -> Done

776{-# INLINE [0] takeWhile #-}

782{-# INLINE [0] takeWhile #-}

777

783

778−-- | dropWhile @p @xs returns the suffix remaining after takeWhile @p @xs.

784+-- | @'dropWhile' p xs@ returns the suffix remaining after @'takeWhile' p xs@.

779dropWhile :: (Char -> Bool) -> Stream Char -> Stream Char

785dropWhile :: (Char -> Bool) -> Stream Char -> Stream Char

780−dropWhile p (Stream next0 s0 len) = Stream next (L s0) len -- HINT maybe too high

786+dropWhile p (Stream next0 s0 len) = Stream next (L s0) (len - unknownSize)

781 where

787 where

782 {-# INLINE next #-}

788 {-# INLINE next #-}

783 next (L s) = case next0 s of

789 next (L s) = case next0 s of

@@ -811,7 +817,7 @@ isPrefixOf (Stream next1 s1 _) (Stream next2 s2 _) = loop (next1 s1) (next2 s2)

811-------------------------------------------------------------------------------

817-------------------------------------------------------------------------------

812-- ** Searching by equality

818-- ** Searching by equality

813

819

814−-- | /O(n)/ elem is the stream membership predicate.

820+-- | /O(n)/ 'elem' is the stream membership predicate.

815elem :: Char -> Stream Char -> Bool

821elem :: Char -> Stream Char -> Bool

816elem w (Stream next s0 _len) = loop_elem s0

822elem w (Stream next s0 _len) = loop_elem s0

817 where

823 where

@@ -856,7 +862,8 @@ indexI (Stream next s0 _len) n0

856-- returns a stream containing those characters that satisfy the

862-- returns a stream containing those characters that satisfy the

857-- predicate.

863-- predicate.

858filter :: (Char -> Bool) -> Stream Char -> Stream Char

864filter :: (Char -> Bool) -> Stream Char -> Stream Char

865+filter p (Stream next0 s0 len) =

859−filter p (Stream next0 s0 len) = Stream next s0 len -- HINT maybe too high

866+ Stream next s0 (len - unknownSize) -- HINT maybe too high

860 where

867 where

861 next !s = case next0 s of

868 next !s = case next0 s of

862 Done -> Done

869 Done -> Done

Data/Text/Internal/Fusion/Size.hs+33−7

@@ -1,4 +1,4 @@

1−{-# LANGUAGE CPP #-}

1+{-# LANGUAGE CPP, PatternGuards #-}

2{-# OPTIONS_GHC -fno-warn-missing-methods #-}

3-- |

4-- Module : Data.Text.Internal.Fusion.Internal

@@ -19,11 +19,16 @@

19module Data.Text.Internal.Fusion.Size

20 (

21 Size

22− , exactly

22+ -- * Sizes

23 , exactSize

24 , maxSize

25 , betweenSize

26 , unknownSize

27+ , unionSize

28+ , charSize

29+ , codePointsSize

30+ -- * Querying sizes

31+ , exactly

27 , smaller

32 , smaller

28 , larger

33 , larger

29 , upperBound

34 , upperBound

@@ -32,11 +37,13 @@ module Data.Text.Internal.Fusion.Size

32 , isEmpty

37 , isEmpty

33 ) where

38 ) where

40+import Data.Char (ord)

35import Data.Text.Internal (mul)

41import Data.Text.Internal (mul)

36#if defined(ASSERTS)

42#if defined(ASSERTS)

37import Control.Exception (assert)

43import Control.Exception (assert)

38#endif

44#endif

46+-- | A size in UTF-16 code units.

40data Size = Between {-# UNPACK #-} !Int {-# UNPACK #-} !Int -- ^ Lower and upper bounds on size.

47data Size = Between {-# UNPACK #-} !Int {-# UNPACK #-} !Int -- ^ Lower and upper bounds on size.

41 | Unknown -- ^ Unknown size.

48 | Unknown -- ^ Unknown size.

42 deriving (Eq, Show)

49 deriving (Eq, Show)

@@ -46,6 +53,17 @@ exactly (Between na nb) | na == nb = Just na

46exactly _ = Nothing

53exactly _ = Nothing

47{-# INLINE exactly #-}

54{-# INLINE exactly #-}

56+-- | The 'Size' of the given code point.

57+charSize :: Char -> Size

58+charSize c

59+ | ord c < 0x10000 = exactSize 1

60+ | otherwise = exactSize 2

61+

62+-- | The 'Size' of @n@ code points.

63+codePointsSize :: Int -> Size

64+codePointsSize n = Between n (2*n)

65+{-# INLINE codePointsSize #-}

66+

49exactSize :: Int -> Size

67exactSize :: Int -> Size

50exactSize n =

68exactSize n =

51#if defined(ASSERTS)

69#if defined(ASSERTS)

@@ -71,6 +89,10 @@ betweenSize m n =

71 Between m n

89 Between m n

72{-# INLINE betweenSize #-}

90{-# INLINE betweenSize #-}

92+unionSize :: Size -> Size -> Size

93+unionSize (Between a b) (Between c d) = Between (min a c) (max b d)

94+unionSize _ _ = Unknown

95+

74unknownSize :: Size

96unknownSize :: Size

75unknownSize = Unknown

97unknownSize = Unknown

76{-# INLINE unknownSize #-}

98{-# INLINE unknownSize #-}

@@ -140,11 +162,15 @@ lowerBound _ (Between n _) = n

140lowerBound k _ = k

162lowerBound k _ = k

141{-# INLINE lowerBound #-}

163{-# INLINE lowerBound #-}

142

164

165+-- | Determine the ordering relationship between two 'Size's, or 'Nothing' in

166+-- the indeterminate case.

143−compareSize :: Size -> Int -> Maybe Ordering

167+compareSize :: Size -> Size -> Maybe Ordering

144−compareSize (Between ma mb) n

168+compareSize (Between ma mb) (Between na nb)

145− | mb < n = Just LT

169+ | mb < na = Just LT

146− | ma > n = Just GT

170+ | ma > nb = Just GT

171+ | ma == mb

172+ , ma == na

147− | ma == n && mb == n = Just EQ

173+ , ma == nb = Just EQ

148compareSize _ _ = Nothing

174compareSize _ _ = Nothing

149

175

150

176

Data/Text/Internal/Fusion/Types.hs+4−4

@@ -75,16 +75,16 @@ instance (Ord a) => Ord (Stream a) where

75-- unstreaming functions must be able to cope with the hint being too

76-- small or too large.

77--

78−-- The size hint tries to track the UTF-16 code points in a stream,

78+-- The size hint tries to track the UTF-16 code units in a stream,

79−-- but often counts the number of characters instead. It can easily

79+-- but often counts the number of code points instead. It can easily

80-- undercount if, for instance, a transformed stream contains astral

81−-- plane characters (those above 0x10000).

81+-- plane code points (those above 0x10000).

83data Stream a =

84 forall s. Stream

85 (s -> Step s a) -- stepper function

86 !s -- current state

87− !Size -- size hint

87+ !Size -- size hint in code units

89-- | /O(n)/ Determines if two streams are equal.

90eq :: (Eq a) => Stream a -> Stream a -> Bool

Data/Text/Lazy.hs+19−6

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

68 , snoc

69 , append

70 , uncons

71+ , unsnoc

71 , head

72 , head

72 , last

73 , last

73 , tail

74 , tail

@@ -475,9 +476,7 @@ fromStrict t = chunk t Empty

475-- -----------------------------------------------------------------------------

476-- -----------------------------------------------------------------------------

476-- * Basic functions

477-- * Basic functions

477

478

478−-- | /O(n)/ Adds a character to the front of a 'Text'. This function

479+-- | /O(1)/ Adds a character to the front of a 'Text'. Subject to fusion.

479−-- is more costly than its 'List' counterpart because it requires

480−-- copying a new array. Subject to fusion.

481cons :: Char -> Text -> Text

480cons :: Char -> Text -> Text

482cons c t = Chunk (T.singleton c) t

481cons c t = Chunk (T.singleton c) t

483{-# INLINE [1] cons #-}

482{-# INLINE [1] cons #-}

@@ -561,6 +560,17 @@ init Empty = emptyError "init"

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

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

562 #-}

561 #-}

563

562

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

564+-- empty.

565+--

566+-- * It is no faster than using 'init' and 'last'.

567+--

568+-- @since 1.2.3.0

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

570+unsnoc Empty = Nothing

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

572+{-# INLINE unsnoc #-}

573+

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

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

565-- fusion.

575-- fusion.

566null :: Text -> Bool

576null :: Text -> Bool

@@ -820,6 +830,8 @@ toUpper t = unstream (S.toUpper (stream t))

820-- guides disagree on whether the book name \"The Hill of the Red

830-- guides disagree on whether the book name \"The Hill of the Red

821-- Fox\" is correctly title cased—but this function will

831-- Fox\" is correctly title cased—but this function will

822-- capitalize /every/ word.

832-- capitalize /every/ word.

833+--

834+-- @since 1.0.0.0

823toTitle :: Text -> Text

835toTitle :: Text -> Text

824toTitle t = unstream (S.toTitle (stream t))

836toTitle t = unstream (S.toTitle (stream t))

825{-# INLINE toTitle #-}

837{-# INLINE toTitle #-}

@@ -1153,9 +1165,10 @@ takeWhile p t0 = takeWhile' t0

1153takeWhileEnd :: (Char -> Bool) -> Text -> Text

1165takeWhileEnd :: (Char -> Bool) -> Text -> Text

1154takeWhileEnd p = takeChunk empty . L.reverse . toChunks

1166takeWhileEnd p = takeChunk empty . L.reverse . toChunks

1155 where takeChunk acc [] = acc

1167 where takeChunk acc [] = acc

1156− takeChunk acc (t:ts) = if T.length t' < T.length t

1168+ takeChunk acc (t:ts)

1157− then (Chunk t' acc)

1169+ | T.lengthWord16 t' < T.lengthWord16 t

1170+ = chunk t' acc

1158− else takeChunk (Chunk t' acc) ts

1171+ | otherwise = takeChunk (Chunk t' acc) ts

1159 where t' = T.takeWhileEnd p t

1172 where t' = T.takeWhileEnd p t

1160{-# INLINE takeWhileEnd #-}

1173{-# INLINE takeWhileEnd #-}

1161

1174

Data/Text/Lazy/Builder/Int.hs+3

@@ -32,6 +32,9 @@ import GHC.Base (quotInt, remInt)

32import GHC.Num (quotRemInteger)

33import GHC.Types (Int(..))

34import Control.Monad.ST

35+#if MIN_VERSION_base(4,11,0)

36+import Prelude hiding ((<>))

37+#endif

36#ifdef __GLASGOW_HASKELL__

39#ifdef __GLASGOW_HASKELL__

37# if defined(INTEGER_GMP)

40# if defined(INTEGER_GMP)

Data/Text/Lazy/Builder/RealFloat.hs+3

@@ -24,6 +24,9 @@ import Data.Text.Lazy.Builder.Int (decimal)

24import Data.Text.Internal.Builder.RealFloat.Functions (roundTo)

25import Data.Text.Lazy.Builder

26import qualified Data.Text as T

27+#if MIN_VERSION_base(4,11,0)

28+import Prelude hiding ((<>))

29+#endif

28-- | Control the rendering of floating point numbers.

31-- | Control the rendering of floating point numbers.

29data FPFormat = Exponent

32data FPFormat = Exponent

Data/Text/Lazy/Read.hs+3−3

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

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

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

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

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

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

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

73−-- is case insensitive.

73+-- This function 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))

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

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

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

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

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

72−-- is case insensitive.

72+-- This function is case insensitive.

73--

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

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

README.markdown+3−3

@@ -19,11 +19,11 @@ library to provide these facilities.

19# Get involved!

21Please report bugs via the

22−[github issue tracker](https://github.com/bos/text/issues).

22+[github issue tracker](https://github.com/haskell/text/issues).

24−Master [git repository](https://github.com/bos/text):

24+Master [git repository](https://github.com/haskell/text):

26−* `git clone git://github.com/bos/text.git`

26+* `git clone git://github.com/haskell/text.git`

28There's also a [Mercurial mirror](https://bitbucket.org/bos/text):

benchmarks/haskell/Benchmarks.hs+2

@@ -10,6 +10,7 @@ import System.FilePath ((</>))

10import System.IO (IOMode (WriteMode), openFile, hSetEncoding, utf8)

12import qualified Benchmarks.Builder as Builder

13+import qualified Benchmarks.Concat as Concat

13import qualified Benchmarks.DecodeUtf8 as DecodeUtf8

14import qualified Benchmarks.DecodeUtf8 as DecodeUtf8

14import qualified Benchmarks.EncodeUtf8 as EncodeUtf8

15import qualified Benchmarks.EncodeUtf8 as EncodeUtf8

15import qualified Benchmarks.Equality as Equality

16import qualified Benchmarks.Equality as Equality

@@ -41,6 +42,7 @@ benchmarks = do

41 -- Traditional benchmarks

42 -- Traditional benchmarks

42 bs <- sequence

43 bs <- sequence

43 [ Builder.benchmark

44 [ Builder.benchmark

45+ , Concat.benchmark

44 , DecodeUtf8.benchmark "html" (tf "libya-chinese.html")

46 , DecodeUtf8.benchmark "html" (tf "libya-chinese.html")

45 , DecodeUtf8.benchmark "xml" (tf "yiwiki.xml")

47 , DecodeUtf8.benchmark "xml" (tf "yiwiki.xml")

46 , DecodeUtf8.benchmark "ascii" (tf "ascii.txt")

48 , DecodeUtf8.benchmark "ascii" (tf "ascii.txt")

benchmarks/haskell/Benchmarks/Concat.hsnew file+25

@@ -0,0 +1,25 @@

1+{-# LANGUAGE OverloadedStrings #-}

3+module Benchmarks.Concat (benchmark) where

5+import Control.Monad.Trans.Writer

6+import Criterion (Benchmark, bgroup, bench, whnf)

7+import Data.Text as T

9+benchmark :: IO Benchmark

10+benchmark = return $ bgroup "Concat"

11+ [ bench "append" $ whnf (append4 "Text 1" "Text 2" "Text 3") "Text 4"

12+ , bench "concat" $ whnf (concat4 "Text 1" "Text 2" "Text 3") "Text 4"

13+ , bench "write" $ whnf (write4 "Text 1" "Text 2" "Text 3") "Text 4"

14+ ]

15+

16+append4, concat4, write4 :: Text -> Text -> Text -> Text -> Text

17+

18+{-# NOINLINE append4 #-}

19+append4 x1 x2 x3 x4 = x1 `append` x2 `append` x3 `append` x4

20+

21+{-# NOINLINE concat4 #-}

22+concat4 x1 x2 x3 x4 = T.concat [x1, x2, x3, x4]

23+

24+{-# NOINLINE write4 #-}

25+write4 x1 x2 x3 x4 = execWriter $ tell x1 >> tell x2 >> tell x3 >> tell x4

benchmarks/text-benchmarks.cabal+1

@@ -47,6 +47,7 @@ executable text-benchmarks

47 ghc-prim,

48 integer-gmp,

49 stringsearch,

50+ transformers,

50 utf8-string,

51 utf8-string,

51 vector

52 vector

cabal.projectnew file+3

@@ -0,0 +1,3 @@

1+-- See http://cabal.readthedocs.io/en/latest/nix-local-build-overview.html

2+packages: .

3+tests: True

cbits/cbits.c+6−11

@@ -206,11 +206,9 @@ _hs_text_decode_utf8_state(uint16_t *const dest, size_t *destoff,

206 const uint8_t *srcend,

207 uint32_t *codepoint0, uint32_t *state0)

208{

209− uint8_t const *ret = _hs_text_decode_utf8_int(dest, destoff, src, srcend,

209+ _hs_text_decode_utf8_int(dest, destoff, src, srcend, codepoint0, state0);

210− codepoint0, state0);

210+

211− if (*state0 == UTF8_REJECT)

212− ret -=1;

213− return ret;

211+ return *src;

214}

212}

215

213

216/*

214/*

@@ -222,12 +220,9 @@ _hs_text_decode_utf8(uint16_t *const dest, size_t *destoff,

222{

220{

223 uint32_t codepoint;

221 uint32_t codepoint;

224 uint32_t state = UTF8_ACCEPT;

222 uint32_t state = UTF8_ACCEPT;

225− uint8_t const *ret = _hs_text_decode_utf8_int(dest, destoff, &src, srcend,

223+ _hs_text_decode_utf8_int(dest, destoff, &src, srcend,

226− &codepoint, &state);

224+ &codepoint, &state);

227− /* Back up if we have an incomplete or invalid encoding */

228− if (state != UTF8_ACCEPT)

229− ret -= 1;

230− return ret;

225+ return src;

231}

226}

232

227

233void

228void

changelog.md+35−1

@@ -1,10 +1,44 @@

1+1.2.3.0

3+* `unsnoc` implemented

5+* Bug fix: the lazy `takeWhileEnd` function violated the

6+ [lazy text invariant](https://github.com/bos/text/blob/1.2.3.0/Data/Text/Internal/Lazy.hs#L51).

8+* Spec compliance: `toCaseFold` now follows the Unicode 9.0 spec

9+ (updated from 8.0)

10+

11+* Bug fix: Fixed usage of size hints causing incorrect behavior (gh-197).

12+

13+* Improve UTF8 decoder error-recovery (gh-182).

14+

15+* Reduce memory overhead in `encodeUTF8` (gh-194).

16+

17+

18+1.2.2.2

19+

20+* The `toTitle` function now correctly handles letters that

21+ immediately follow punctuation. Before, `"there's"` would turn into

22+ `"There'S"`. Now, it becomes `"There's"`.

23+

24+* The implementation of unstreaming is faster, resulting in operations

25+ such as `map` and `intersperse` speeding up by up to 30%, with

26+ smaller code generated.

27+

28+* The optimised length comparison function is now more likely to be

29+ used after some rewrite rule tweaking.

30+

31+* Bug fix: an off-by-one bug in `takeEnd` is fixed.

32+

33+* Bug fix: a logic error in `takeWord16` is fixed.

34+

11.2.2.1

351.2.2.1

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

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

4 The build flag has been renamed accordingly. Your army of diligent

38 The build flag has been renamed accordingly. Your army of diligent

5 maintainers apologizes for the churn.

39 maintainers apologizes for the churn.

7−* Spec compliance: toCaseFold now follows the Unicode 8.0 spec

41+* Spec compliance: `toCaseFold` now follows the Unicode 8.0 spec

8 (updated from 7.0)

42 (updated from 7.0)

10* An STG lint error has been fixed

44* An STG lint error has been fixed

tests/Tests/Properties.hs+94−21

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

9 ) where

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

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

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

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

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

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

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

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

17import Data.String (IsString(fromString))

@@ -23,6 +23,7 @@ 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)

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

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

27import Prelude hiding (replicate)

28import Prelude hiding (replicate)

28import Test.Framework (Test, testGroup)

29import Test.Framework (Test, testGroup)

@@ -30,6 +31,7 @@ import Test.Framework.Providers.QuickCheck2 (testProperty)

30import Test.QuickCheck hiding ((.&.))

31import Test.QuickCheck hiding ((.&.))

31import Test.QuickCheck.Monadic

32import Test.QuickCheck.Monadic

32import Test.QuickCheck.Property (Property(..))

33import Test.QuickCheck.Property (Property(..))

34+import Test.QuickCheck.Unicode (char)

33import Tests.QuickCheckUtils

35import Tests.QuickCheckUtils

34import Tests.Utils

36import Tests.Utils

35import Text.Show.Functions ()

37import Text.Show.Functions ()

@@ -37,6 +39,7 @@ import qualified Control.Exception as Exception

37import qualified Data.Bits as Bits (shiftL, shiftR)

39import qualified Data.Bits as Bits (shiftL, shiftR)

38import qualified Data.ByteString as B

40import qualified Data.ByteString as B

39import qualified Data.ByteString.Lazy as BL

41import qualified Data.ByteString.Lazy as BL

42+import qualified Data.Char as C

40import qualified Data.List as L

43import qualified Data.List as L

41import qualified Data.Text as T

44import qualified Data.Text as T

42import qualified Data.Text.Encoding as E

45import qualified Data.Text.Encoding as E

@@ -60,10 +63,10 @@ t_pack_unpack = (T.unpack . T.pack) `eq` id

60tl_pack_unpack = (TL.unpack . TL.pack) `eq` id

63tl_pack_unpack = (TL.unpack . TL.pack) `eq` id

61t_stream_unstream = (S.unstream . S.stream) `eq` id

64t_stream_unstream = (S.unstream . S.stream) `eq` id

62tl_stream_unstream = (SL.unstream . SL.stream) `eq` id

65tl_stream_unstream = (SL.unstream . SL.stream) `eq` id

63−t_reverse_stream t = (S.reverse . S.reverseStream) t == t

66+t_reverse_stream t = (S.reverse . S.reverseStream) t === t

64−t_singleton c = [c] == (T.unpack . T.singleton) c

67+t_singleton c = [c] === (T.unpack . T.singleton) c

65−tl_singleton c = [c] == (TL.unpack . TL.singleton) c

68+tl_singleton c = [c] === (TL.unpack . TL.singleton) c

66−tl_unstreamChunks x = f 11 x == f 1000 x

69+tl_unstreamChunks x = f 11 x === f 1000 x

67 where f n = SL.unstreamChunks n . S.streamList

70 where f n = SL.unstreamChunks n . S.streamList

68tl_chunk_unchunk = (TL.fromChunks . TL.toChunks) `eq` id

71tl_chunk_unchunk = (TL.fromChunks . TL.toChunks) `eq` id

69tl_from_to_strict = (TL.fromStrict . TL.toStrict) `eq` id

72tl_from_to_strict = (TL.fromStrict . TL.toStrict) `eq` id

@@ -74,13 +77,13 @@ encodeL1 = B.pack . map (fromIntegral . fromEnum) . T.unpack

74encodeLazyL1 :: TL.Text -> BL.ByteString

77encodeLazyL1 :: TL.Text -> BL.ByteString

75encodeLazyL1 = BL.fromChunks . map encodeL1 . TL.toChunks

78encodeLazyL1 = BL.fromChunks . map encodeL1 . TL.toChunks

77−t_ascii t = E.decodeASCII (E.encodeUtf8 a) == a

80+t_ascii t = E.decodeASCII (E.encodeUtf8 a) === a

78 where a = T.map (\c -> chr (ord c `mod` 128)) t

81 where a = T.map (\c -> chr (ord c `mod` 128)) t

79−tl_ascii t = EL.decodeASCII (EL.encodeUtf8 a) == a

82+tl_ascii t = EL.decodeASCII (EL.encodeUtf8 a) === a

80 where a = TL.map (\c -> chr (ord c `mod` 128)) t

83 where a = TL.map (\c -> chr (ord c `mod` 128)) t

81−t_latin1 t = E.decodeLatin1 (encodeL1 a) == a

84+t_latin1 t = E.decodeLatin1 (encodeL1 a) === a

82 where a = T.map (\c -> chr (ord c `mod` 256)) t

85 where a = T.map (\c -> chr (ord c `mod` 256)) t

83−tl_latin1 t = EL.decodeLatin1 (encodeLazyL1 a) == a

86+tl_latin1 t = EL.decodeLatin1 (encodeLazyL1 a) === a

84 where a = TL.map (\c -> chr (ord c `mod` 256)) t

87 where a = TL.map (\c -> chr (ord c `mod` 256)) t

85t_utf8 = forAll genUnicode $ (E.decodeUtf8 . E.encodeUtf8) `eq` id

88t_utf8 = forAll genUnicode $ (E.decodeUtf8 . E.encodeUtf8) `eq` id

86t_utf8' = forAll genUnicode $ (E.decodeUtf8' . E.encodeUtf8) `eq` (id . Right)

89t_utf8' = forAll genUnicode $ (E.decodeUtf8' . E.encodeUtf8) `eq` (id . Right)

@@ -111,7 +114,7 @@ t_utf8_undecoded = forAll genUnicode $ \t ->

111 let b = E.encodeUtf8 t

114 let b = E.encodeUtf8 t

112 ls = concatMap (leftover . E.encodeUtf8 . T.singleton) . T.unpack $ t

115 ls = concatMap (leftover . E.encodeUtf8 . T.singleton) . T.unpack $ t

113 leftover = (++ [B.empty]) . init . tail . B.inits

116 leftover = (++ [B.empty]) . init . tail . B.inits

114− in (map snd . feedChunksOf 1 E.streamDecodeUtf8) b == ls

117+ in (map snd . feedChunksOf 1 E.streamDecodeUtf8) b === ls

115

118

116data Badness = Solo | Leading | Trailing

119data Badness = Solo | Leading | Trailing

117 deriving (Eq, Show)

120 deriving (Eq, Show)

@@ -177,6 +180,33 @@ genInvalidUTF8 = B.pack <$> oneof [

177 k <- choose (0,n)

180 k <- choose (0,n)

178 vectorOf k gen

181 vectorOf k gen

179

182

183+-- See http://unicode.org/faq/utf_bom.html#gen8

184+-- A sequence such as <110xxxxx2 0xxxxxxx2> is illegal ...

185+-- When faced with this illegal byte sequence ... a UTF-8 conformant process

186+-- must treat the first byte 110xxxxx2 as an illegal termination error

187+-- (e.g. filter it out or replace by 0xFFFD) ...

188+-- ... and continue processing at the second byte 0xxxxxxx2

189+t_decode_with_error2 =

190+ E.decodeUtf8With (\_ _ -> Just 'x') (B.pack [0xC2, 97]) === "xa"

191+t_decode_with_error3 =

192+ E.decodeUtf8With (\_ _ -> Just 'x') (B.pack [0xE0, 97, 97]) === "xaa"

193+t_decode_with_error4 =

194+ E.decodeUtf8With (\_ _ -> Just 'x') (B.pack [0xF0, 97, 97, 97]) === "xaaa"

195+

196+t_decode_with_error2' =

197+ case E.streamDecodeUtf8With (\_ _ -> Just 'x') (B.pack [0xC2, 97]) of

198+ E.Some x _ _ -> x === "xa"

199+t_decode_with_error3' =

200+ case E.streamDecodeUtf8With (\_ _ -> Just 'x') (B.pack [0xC2, 97, 97]) of

201+ E.Some x _ _ -> x === "xaa"

202+t_decode_with_error4' =

203+ case E.streamDecodeUtf8With (\_ _ -> Just 'x') (B.pack [0xC2, 97, 97, 97]) of

204+ E.Some x _ _ -> x === "xaaa"

205+

206+t_infix_concat bs1 text bs2 rep =

207+ text `T.isInfixOf`

208+ E.decodeUtf8With (\_ _ -> rep) (B.concat [bs1, E.encodeUtf8 text, bs2])

209+

180s_Eq s = (s==) `eq` ((S.streamList s==) . S.streamList)

210s_Eq s = (s==) `eq` ((S.streamList s==) . S.streamList)

181 where _types = s :: String

211 where _types = s :: String

182sf_Eq p s =

212sf_Eq p s =

@@ -201,8 +231,8 @@ t_mconcat = unsquare $

201 mconcat `eq` (unpackS . mconcat . L.map T.pack)

231 mconcat `eq` (unpackS . mconcat . L.map T.pack)

202tl_mconcat = unsquare $

232tl_mconcat = unsquare $

203 mconcat `eq` (unpackS . mconcat . L.map TL.pack)

233 mconcat `eq` (unpackS . mconcat . L.map TL.pack)

204−t_mempty = mempty == (unpackS (mempty :: T.Text))

234+t_mempty = mempty === (unpackS (mempty :: T.Text))

205−tl_mempty = mempty == (unpackS (mempty :: TL.Text))

235+tl_mempty = mempty === (unpackS (mempty :: TL.Text))

206t_IsString = fromString `eqP` (T.unpack . fromString)

236t_IsString = fromString `eqP` (T.unpack . fromString)

207tl_IsString = fromString `eqP` (TL.unpack . fromString)

237tl_IsString = fromString `eqP` (TL.unpack . fromString)

208

238

@@ -229,6 +259,13 @@ sf_uncons p = (uncons . L.filter p) `eqP`

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

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

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

260t_uncons = uncons `eqP` (fmap (second unpackS) . T.uncons)

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

261tl_uncons = uncons `eqP` (fmap (second unpackS) . TL.uncons)

262+

263+unsnoc xs@(_:_) = Just (init xs, last xs)

264+unsnoc [] = Nothing

265+

266+t_unsnoc = unsnoc `eqP` (fmap (first unpackS) . T.unsnoc)

267+tl_unsnoc = unsnoc `eqP` (fmap (first unpackS) . TL.unsnoc)

268+

232s_head = head `eqP` S.head

269s_head = head `eqP` S.head

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

270sf_head p = (head . L.filter p) `eqP` (S.head . S.filter p)

234t_head = head `eqP` T.head

271t_head = head `eqP` T.head

@@ -326,6 +363,19 @@ t_toUpper_upper t = p (T.toUpper t) >= p t

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

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

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

364tl_toUpper_upper t = p (TL.toUpper t) >= p t

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

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

366+t_toTitle_title t = all (<= 1) (caps w)

367+ where caps = fmap (T.length . T.filter isUpper) . T.words . T.toTitle

368+ -- TIL: there exist uppercase-only letters

369+ w = T.filter (\c -> if C.isUpper c then C.toLower c /= c else True) t

370+t_toTitle_1stNotLower = and . notLow . T.toTitle . T.filter stable

371+ where notLow = mapMaybe (fmap (not . isLower) . (T.find isLetter)) . T.words

372+ -- Surprise! The Spanish/Portuguese ordinal indicators changed

373+ -- from category Ll (letter, lowercase) to Lo (letter, other)

374+ -- in Unicode 7.0

375+ -- Oh, and there exist lowercase-only letters (see previous test)

376+ stable c = if isLower c

377+ then C.toUpper c /= c

378+ else c /= '\170' && c /= '\186'

329

379

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

380justifyLeft k c xs = xs ++ L.replicate (k - length xs) c

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

381justifyRight m n xs = L.replicate (m - length xs) n ++ xs

@@ -514,12 +564,20 @@ s_takeWhile p = L.takeWhile p `eqP` (unpackS . S.takeWhile p)

514s_takeWhile_s p = L.takeWhile p `eqP` (unpackS . S.unstream . S.takeWhile p)

564s_takeWhile_s p = L.takeWhile p `eqP` (unpackS . S.unstream . S.takeWhile p)

515sf_takeWhile q p = (L.takeWhile p . L.filter q) `eqP`

565sf_takeWhile q p = (L.takeWhile p . L.filter q) `eqP`

516 (unpackS . S.takeWhile p . S.filter q)

566 (unpackS . S.takeWhile p . S.filter q)

567+noMatch = do

568+ c <- char

569+ d <- suchThat char (/= c)

570+ return (c,d)

517t_takeWhile p = L.takeWhile p `eqP` (unpackS . T.takeWhile p)

571t_takeWhile p = L.takeWhile p `eqP` (unpackS . T.takeWhile p)

518tl_takeWhile p = L.takeWhile p `eqP` (unpackS . TL.takeWhile p)

572tl_takeWhile p = L.takeWhile p `eqP` (unpackS . TL.takeWhile p)

519t_takeWhileEnd p = (L.reverse . L.takeWhile p . L.reverse) `eqP`

573t_takeWhileEnd p = (L.reverse . L.takeWhile p . L.reverse) `eqP`

520 (unpackS . T.takeWhileEnd p)

574 (unpackS . T.takeWhileEnd p)

575+t_takeWhileEnd_null t = forAll noMatch $ \(c,d) -> T.null $

576+ T.takeWhileEnd (==d) (T.snoc t c)

521tl_takeWhileEnd p = (L.reverse . L.takeWhile p . L.reverse) `eqP`

577tl_takeWhileEnd p = (L.reverse . L.takeWhile p . L.reverse) `eqP`

522 (unpackS . TL.takeWhileEnd p)

578 (unpackS . TL.takeWhileEnd p)

579+tl_takeWhileEnd_null t = forAll noMatch $ \(c,d) -> TL.null $

580+ TL.takeWhileEnd (==d) (TL.snoc t c)

523s_dropWhile p = L.dropWhile p `eqP` (unpackS . S.dropWhile p)

581s_dropWhile p = L.dropWhile p `eqP` (unpackS . S.dropWhile p)

524s_dropWhile_s p = L.dropWhile p `eqP` (unpackS . S.unstream . S.dropWhile p)

582s_dropWhile_s p = L.dropWhile p `eqP` (unpackS . S.unstream . S.dropWhile p)

525sf_dropWhile q p = (L.dropWhile p . L.filter q) `eqP`

583sf_dropWhile q p = (L.dropWhile p . L.filter q) `eqP`

@@ -720,7 +778,7 @@ tl_indices (NotEmpty s) = lazyIndices s `eq` S.indices s

720 conc = T.concat . TL.toChunks

778 conc = T.concat . TL.toChunks

721t_indices_occurs = unsquare $ \(NotEmpty t) ts ->

779t_indices_occurs = unsquare $ \(NotEmpty t) ts ->

722 let s = T.intercalate t ts

780 let s = T.intercalate t ts

723− in Slow.indices t s == indices t s

781+ in Slow.indices t s === indices t s

724

782

725-- Bit shifts.

783-- Bit shifts.

726shiftL w = forAll (choose (0,width-1)) $ \k -> Bits.shiftL w k == U.shiftL w k

784shiftL w = forAll (choose (0,width-1)) $ \k -> Bits.shiftL w k == U.shiftL w k

@@ -812,18 +870,18 @@ tb_formatRealFloat_double (a::Double) = tb_formatRealFloat a

812-- Reading.

870-- Reading.

813

871

814t_decimal (n::Int) s =

872t_decimal (n::Int) s =

815− T.signed T.decimal (T.pack (show n) `T.append` t) == Right (n,t)

873+ T.signed T.decimal (T.pack (show n) `T.append` t) === Right (n,t)

816 where t = T.dropWhile isDigit s

874 where t = T.dropWhile isDigit s

817tl_decimal (n::Int) s =

875tl_decimal (n::Int) s =

818− TL.signed TL.decimal (TL.pack (show n) `TL.append` t) == Right (n,t)

876+ TL.signed TL.decimal (TL.pack (show n) `TL.append` t) === Right (n,t)

819 where t = TL.dropWhile isDigit s

877 where t = TL.dropWhile isDigit s

820t_hexadecimal m s ox =

878t_hexadecimal m s ox =

821− T.hexadecimal (T.concat [p, T.pack (showHex n ""), t]) == Right (n,t)

879+ T.hexadecimal (T.concat [p, T.pack (showHex n ""), t]) === Right (n,t)

822 where t = T.dropWhile isHexDigit s

880 where t = T.dropWhile isHexDigit s

823 p = if ox then "0x" else ""

881 p = if ox then "0x" else ""

824 n = getPositive m :: Int

882 n = getPositive m :: Int

825tl_hexadecimal m s ox =

883tl_hexadecimal m s ox =

826− TL.hexadecimal (TL.concat [p, TL.pack (showHex n ""), t]) == Right (n,t)

884+ TL.hexadecimal (TL.concat [p, TL.pack (showHex n ""), t]) === Right (n,t)

827 where t = TL.dropWhile isHexDigit s

885 where t = TL.dropWhile isHexDigit s

828 p = if ox then "0x" else ""

886 p = if ox then "0x" else ""

829 n = getPositive m :: Int

887 n = getPositive m :: Int

@@ -867,11 +925,11 @@ tl_write_read_line e m b t = write_read head TL.filter TL.hPutStrLn

867

925

868t_dropWord16 m t = dropWord16 m t `T.isSuffixOf` t

926t_dropWord16 m t = dropWord16 m t `T.isSuffixOf` t

869t_takeWord16 m t = takeWord16 m t `T.isPrefixOf` t

927t_takeWord16 m t = takeWord16 m t `T.isPrefixOf` t

870−t_take_drop_16 m t = T.append (takeWord16 n t) (dropWord16 n t) == t

928+t_take_drop_16 m t = T.append (takeWord16 n t) (dropWord16 n t) === t

871 where n = small m

929 where n = small m

872t_use_from t = monadicIO $ assert . (==t) =<< run (useAsPtr t fromPtr)

930t_use_from t = monadicIO $ assert . (==t) =<< run (useAsPtr t fromPtr)

873

931

874−t_copy t = T.copy t == t

932+t_copy t = T.copy t === t

875

933

876-- Regression tests.

934-- Regression tests.

877s_filter_eq s = S.filter p t == S.streamList (filter p s)

935s_filter_eq s = S.filter p t == S.streamList (filter p s)

@@ -924,6 +982,15 @@ tests =

924 testGroup "errors" [

982 testGroup "errors" [

925 testProperty "t_utf8_err" t_utf8_err,

983 testProperty "t_utf8_err" t_utf8_err,

926 testProperty "t_utf8_err'" t_utf8_err'

984 testProperty "t_utf8_err'" t_utf8_err'

985+ ],

986+ testGroup "error recovery" [

987+ testProperty "t_decode_with_error2" t_decode_with_error2,

988+ testProperty "t_decode_with_error3" t_decode_with_error3,

989+ testProperty "t_decode_with_error4" t_decode_with_error4,

990+ testProperty "t_decode_with_error2'" t_decode_with_error2',

991+ testProperty "t_decode_with_error3'" t_decode_with_error3',

992+ testProperty "t_decode_with_error4'" t_decode_with_error4',

993+ testProperty "t_infix_concat" t_infix_concat

927 ]

994 ]

928 ],

995 ],

929

996

@@ -967,6 +1034,8 @@ tests =

967 testProperty "sf_uncons" sf_uncons,

1034 testProperty "sf_uncons" sf_uncons,

968 testProperty "t_uncons" t_uncons,

1035 testProperty "t_uncons" t_uncons,

969 testProperty "tl_uncons" tl_uncons,

1036 testProperty "tl_uncons" tl_uncons,

1037+ testProperty "t_unsnoc" t_unsnoc,

1038+ testProperty "tl_unsnoc" tl_unsnoc,

970 testProperty "s_head" s_head,

1039 testProperty "s_head" s_head,

971 testProperty "sf_head" sf_head,

1040 testProperty "sf_head" sf_head,

972 testProperty "t_head" t_head,

1041 testProperty "t_head" t_head,

@@ -1030,7 +1099,9 @@ tests =

1030 testProperty "tl_toLower_lower" tl_toLower_lower,

1099 testProperty "tl_toLower_lower" tl_toLower_lower,

1031 testProperty "t_toUpper_length" t_toUpper_length,

1100 testProperty "t_toUpper_length" t_toUpper_length,

1032 testProperty "t_toUpper_upper" t_toUpper_upper,

1101 testProperty "t_toUpper_upper" t_toUpper_upper,

1033− testProperty "tl_toUpper_upper" tl_toUpper_upper

1102+ testProperty "tl_toUpper_upper" tl_toUpper_upper,

1103+ testProperty "t_toTitle_title" t_toTitle_title,

1104+ testProperty "t_toTitle_1stNotLower" t_toTitle_1stNotLower

1034 ],

1105 ],

1035

1106

1036 testGroup "justification" [

1107 testGroup "justification" [

@@ -1147,7 +1218,9 @@ tests =

1147 testProperty "t_takeWhile" t_takeWhile,

1218 testProperty "t_takeWhile" t_takeWhile,

1148 testProperty "tl_takeWhile" tl_takeWhile,

1219 testProperty "tl_takeWhile" tl_takeWhile,

1149 testProperty "t_takeWhileEnd" t_takeWhileEnd,

1220 testProperty "t_takeWhileEnd" t_takeWhileEnd,

1221+ testProperty "t_takeWhileEnd_null" t_takeWhileEnd_null,

1150 testProperty "tl_takeWhileEnd" tl_takeWhileEnd,

1222 testProperty "tl_takeWhileEnd" tl_takeWhileEnd,

1223+ testProperty "tl_takeWhileEnd_null" tl_takeWhileEnd_null,

1151 testProperty "sf_dropWhile" sf_dropWhile,

1224 testProperty "sf_dropWhile" sf_dropWhile,

1152 testProperty "s_dropWhile" s_dropWhile,

1225 testProperty "s_dropWhile" s_dropWhile,

1153 testProperty "s_dropWhile_s" s_dropWhile_s,

1226 testProperty "s_dropWhile_s" s_dropWhile_s,

tests/Tests/QuickCheckUtils.hs+1−1

@@ -106,7 +106,7 @@ smallArbitrary = sized $ \n -> resize (smallish n) arbitrary

106 where smallish = round . (sqrt :: Double -> Double) . fromIntegral . abs

107

108instance Arbitrary T.Text where

109− arbitrary = T.pack `fmap` arbitrary

109+ arbitrary = T.pack `fmap` string

110 shrink = map T.pack . shrink . T.unpack

111

112instance Arbitrary TL.Text where

tests/Tests/Regressions.hs+11

@@ -72,6 +72,16 @@ mapAccumL_resize = do

72 assertEqual "mapAccumL should correctly size buffers for two-word results"

73 (count * 2) (T.lengthWord16 (snd val))

75+-- See GitHub #197

76+t197 :: IO ()

77+t197 =

78+ assertBool "length (filter (==',') \"0,00\") should be 1" (currencyParser "0,00")

79+ where

80+ currencyParser x = cond == 1

81+ where

82+ cond = length fltr

83+ fltr = filter (== ',') x

84+

75tests :: F.Test

85tests :: F.Test

76tests = F.testGroup "Regressions"

86tests = F.testGroup "Regressions"

77 [ F.testCase "hGetContents_crash" hGetContents_crash

87 [ F.testCase "hGetContents_crash" hGetContents_crash

@@ -79,4 +89,5 @@ tests = F.testGroup "Regressions"

79 , F.testCase "mapAccumL_resize" mapAccumL_resize

89 , F.testCase "mapAccumL_resize" mapAccumL_resize

80 , F.testCase "replicate_crash" replicate_crash

90 , F.testCase "replicate_crash" replicate_crash

81 , F.testCase "utf8_decode_unsafe" utf8_decode_unsafe

91 , F.testCase "utf8_decode_unsafe" utf8_decode_unsafe

92+ , F.testCase "t197" t197

82 ]

93 ]

tests/text-tests.cabal+18−3

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

20 default: False

21 manual: True

23+flag bytestring-builder

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

25+ default: False

26+ manual: False

27+

23executable text-tests

28executable text-tests

24 main-is: Tests.hs

29 main-is: Tests.hs

@@ -48,16 +53,21 @@ executable text-tests

48 HUnit >= 1.2,

53 HUnit >= 1.2,

49 QuickCheck >= 2.7,

54 QuickCheck >= 2.7,

50 base == 4.*,

55 base == 4.*,

51− bytestring,

52 deepseq,

56 deepseq,

53 directory,

57 directory,

54− quickcheck-unicode,

58+ quickcheck-unicode >= 1.0.1.0,

55 random,

59 random,

56 test-framework >= 0.4,

60 test-framework >= 0.4,

57 test-framework-hunit >= 0.2,

61 test-framework-hunit >= 0.2,

58 test-framework-quickcheck2 >= 0.2,

62 test-framework-quickcheck2 >= 0.2,

59 text-tests

63 text-tests

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+

61executable text-tests-stdio

71executable text-tests-stdio

62 main-is: Tests/IO.hs

72 main-is: Tests/IO.hs

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

137 array,

147 array,

138 base == 4.*,

148 base == 4.*,

139 binary,

149 binary,

140− bytestring,

141 deepseq,

150 deepseq,

142 ghc-prim,

151 ghc-prim,

143 integer-gmp

152 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

text.cabal+9−7

@@ -1,7 +1,7 @@

1name: text

2−version: 1.2.2.1

2+version: 1.2.3.0

3−homepage: https://github.com/bos/text

3+homepage: https://github.com/haskell/text

4−bug-reports: https://github.com/bos/text/issues

4+bug-reports: https://github.com/haskell/text/issues

5synopsis: An efficient packed Unicode text type.

6description:

7 .

39category: Data, Text

40build-type: Simple

41cabal-version: >= 1.8

42+tested-with: GHC==8.2.2, GHC==8.0.2, GHC==7.10.3, GHC==7.8.4,

43+ GHC==7.6.3, GHC==7.4.2, GHC==7.2.2, GHC==7.0.4

42extra-source-files:

44extra-source-files:

43 -- scripts/CaseFolding.txt

45 -- scripts/CaseFolding.txt

44 -- scripts/SpecialCasing.txt

46 -- scripts/SpecialCasing.txt

@@ -166,21 +168,21 @@ test-suite tests

166 include-dirs: include

168 include-dirs: include

167

169

168 ghc-options:

170 ghc-options:

169− -Wall -threaded -O0 -rtsopts

171+ -Wall -threaded -O1 -rtsopts

170

172

171 cpp-options:

173 cpp-options:

172 -DASSERTS -DHAVE_DEEPSEQ -DTEST_SUITE

174 -DASSERTS -DHAVE_DEEPSEQ -DTEST_SUITE

173

175

174 build-depends:

176 build-depends:

175 HUnit >= 1.2,

177 HUnit >= 1.2,

176− QuickCheck >= 2.7,

178+ QuickCheck >= 2.7 && < 2.10,

177 array,

179 array,

178 base,

180 base,

179 binary,

181 binary,

180 deepseq,

182 deepseq,

181 directory,

183 directory,

182 ghc-prim,

184 ghc-prim,

183− quickcheck-unicode,

185+ quickcheck-unicode >= 1.0.1.0,

184 random,

186 random,

185 test-framework >= 0.4,

187 test-framework >= 0.4,

186 test-framework-hunit >= 0.2,

188 test-framework-hunit >= 0.2,

@@ -201,7 +203,7 @@ test-suite tests

201

203

202source-repository head

204source-repository head

203 type: git

205 type: git

204− location: https://github.com/bos/text

206+ location: https://github.com/haskell/text

205

207

206source-repository head

208source-repository head

207 type: mercurial

209 type: mercurial