e949bda · text · git.iko.soy

git.iko.soy/text

Merge branch 'master' into text-jsstring

commite949bdad6bff1f2f1fe393f5fa2624f4b01ea68b authorMatthew Bauer <mjbauer95@gmail.com> dateMon 18 Jun 2018, 12:20 (8 years ago) parentd1253ab 9fac5db

38 files changed+1136−237

compare with master

unifiedsplit

.gitignore+10−3 .hgtags+1 .travis.yml+111 Data/Text.hs+174−45 Data/Text/Array.hs+4−1 Data/Text/Encoding.hs+67−11 Data/Text/Encoding/Error.hs−1 Data/Text/Foreign.hs+5−2 Data/Text/IO.hs+4−1 Data/Text/Internal/Fusion/CaseMapping.hs+164−4 Data/Text/Internal/Fusion/Size.hs+33−7 Data/Text/Internal/Fusion/Types.hs+4−4 Data/Text/Lazy.hs+54−20 Data/Text/Lazy/Builder.hs−1 Data/Text/Lazy/Builder/Int.hs+46−2 Data/Text/Lazy/Builder/RealFloat.hs+3 Data/Text/Lazy/Encoding.hs+13−3 Data/Text/Lazy/IO.hs−1 Data/Text/Lazy/Read.hs+3−4 Data/Text/Read.hs+3−4 Data/Text/Show.hs+2 Data/Text/Unsafe.hs+2−1 README.markdown+5−16 benchmarks/.gitignore+4−1 benchmarks/cabal.project+2 benchmarks/haskell/Benchmarks.hs+2 benchmarks/haskell/Benchmarks/Concat.hs+25 benchmarks/text-benchmarks.cabal+82−8 cabal.project+3 cbits/cbits.c+6−11 changelog.md+61−19 tests-and-benchmarks.markdown+13−8 tests/Makefile+6−1 tests/Tests/Properties.hs+94−21 tests/Tests/QuickCheckUtils.hs+1−1 tests/Tests/Regressions.hs+11 tests/text-tests.cabal+18−3 text.cabal+100−33

.gitignore+10−3

@@ -1,3 +1,10 @@

1−dist

2−cabal-dev

3−cabal.sandbox.config

1+/dist/

2+/dist-boot/

3+/dist-install/

4+/dist-newstyle/

5+/cabal-dev/

6+/cabal.sandbox.config

7+/ghc.mk

8+/GNUmakefile

9+/.ghc.environment.*

10+/cabal.project.local

.hgtags+1

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

6464631fd055e27fa10983e17e9ad36d1b285b85c661 1.2.1.3

65658221a68ed5ec37028357df9c6c25f52320e54ad9 1.2.2.0

6666ce1c5ac65eb9f5580114b4e380092b05a485e95d 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+174−45

@@ -15,7 +15,6 @@

1515--

1616-- License : BSD-style

1717-- Maintainer : bos@serpentine.com

18−-- Stability : experimental

1918-- Portability : GHC

2019--

2120-- A time and space-efficient implementation of Unicode text.

@@ -32,8 +31,9 @@

3231--

3332-- To use an extended and very rich family of functions for working

3433-- 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 >.

34+-- non-standard encodings, text breaking, and locales), see the

35+-- <http://hackage.haskell.org/package/text-icu text-icu package >.

36+--

3737

3838module Data.Text

3939 (

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

4343 -- * Acceptable data

4444 -- $replacement

4545

46+ -- * Definition of character

47+ -- $character_definition

48+

4649 -- * Fusion

4750 -- $fusion

4851

@@ -60,6 +63,7 @@ module Data.Text

6063 , snoc

6164 , append

6265 , uncons

66+ , unsnoc

6367 , head

6468 , last

6569 , tail

@@ -298,6 +302,17 @@ unpackCString# addr# = Text $ SJS.unstream (SJS.streamCString# addr#)

298302 = singleton a #-}

299303

300304---------------------------------------------------------------------

305+-- $character_definition

306+--

307+-- This package uses the term /character/ to denote Unicode /code points/.

308+--

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

310+-- composition of code points that form one visual symbol). For

311+-- instance, consider the grapheme \"ä\". This symbol has two

312+-- Unicode representations: a single code-point representation

313+-- @U+00E4@ (the @LATIN SMALL LETTER A WITH DIAERESIS@ code point),

314+-- and a two code point representation @U+0061@ (the \"@A@\" code

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

301316

302317-- $strict

303318--

@@ -392,9 +407,13 @@ instance Read Text where

392407 readsPrec p str = [(pack x,y) | (x,y) <- readsPrec p str]

393408

394409#if MIN_VERSION_base(4,9,0)

395−-- Semigroup orphan instances for older GHCs are provided by

396−-- 'semigroups` package

397−

410+-- | Non-orphan 'Semigroup' instance only defined for

411+-- @base-4.9.0.0@ and later; orphan instances for older GHCs are

412+-- provided by

413+-- the [semigroups](http://hackage.haskell.org/package/semigroups)

414+-- package

415+--

416+-- @since 1.2.2.0

398417instance Semigroup Text where

399418 (<>) = append

400419#endif

@@ -412,6 +431,7 @@ instance IsString Text where

412431 fromString = pack

413432

414433#if __GLASGOW_HASKELL__ >= 708

434+-- | @since 1.2.0.0

415435instance Exts.IsList Text where

416436 type Item Text = Char

417437 fromList = pack

@@ -422,6 +442,7 @@ instance Exts.IsList Text where

422442instance NFData Text where rnf !_ = ()

423443#endif

424444

445+-- | @since 1.2.1.0

425446instance Binary Text where

426447 put t = put (encodeUtf8 t)

427448 get = do

@@ -455,6 +476,8 @@ instance Data Text where

455476

456477#if MIN_VERSION_base(4,7,0)

457478-- | Only defined for @base-4.7.0.0@ and later

479+--

480+-- @since 1.2.2.0

458481instance PrintfArg Text where

459482 formatArg t = formatString $ unpack t

460483#endif

@@ -612,6 +635,19 @@ init = coerce JSS.init

612635 unstream (S.init (stream t)) = init t

613636 #-}

614637

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

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

640+--

641+-- @since 1.2.3.0

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

643+unsnoc (Text arr off len)

644+ | len <= 0 = Nothing

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

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

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

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

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

650+

615651-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

616652-- fusion.

617653null :: Text -> Bool

@@ -693,8 +729,15 @@ compareLength t n = S.compareLengthI (stream t) n

693729-- -----------------------------------------------------------------------------

694730-- * Transformations

695731-- | /O(n)/ 'map' @f@ @t@ is the 'Text' obtained by applying @f@ to

696−-- each element of @t@. Subject to fusion. Performs replacement on

697−-- invalid scalar values.

732+-- each element of @t@.

733+--

734+-- Example:

735+--

736+-- >>> let message = pack "I am not angry. Not at all."

737+-- >>> T.map (\c -> if c == '.' then '!' else c) message

738+-- "I am not angry! Not at all!"

739+--

740+-- Subject to fusion. Performs replacement on invalid scalar values.

698741map :: (Char -> Char) -> Text -> Text

699742map = coerce JSS.map

700743-- map f t = unstream (S.map (safe . f) (stream t))

@@ -703,20 +746,38 @@ map = coerce JSS.map

703746-- | /O(n)/ The 'intercalate' function takes a 'Text' and a list of

704747-- 'Text's and concatenates the list after interspersing the first

705748-- argument between each element of the list.

749+--

750+-- Example:

751+--

752+-- >>> T.intercalate "NI!" ["We", "seek", "the", "Holy", "Grail"]

753+-- "WeNI!seekNI!theNI!HolyNI!Grail"

706754intercalate :: Text -> [Text] -> Text

707755intercalate = coerce JSS.intercalate

708756-- intercalate t = concat . (F.intersperse t)

709757{-# INLINE intercalate #-}

710758

711759-- | /O(n)/ The 'intersperse' function takes a character and places it

712−-- between the characters of a 'Text'. Subject to fusion. Performs

713−-- replacement on invalid scalar values.

760+-- between the characters of a 'Text'.

761+--

762+-- Example:

763+--

764+-- >>> T.intersperse '.' "SHIELD"

765+-- "S.H.I.E.L.D"

766+--

767+-- Subject to fusion. Performs replacement on invalid scalar values.

714768intersperse :: Char -> Text -> Text

715769intersperse = coerce JSS.intersperse

716770-- intersperse c t = unstream (S.intersperse (safe c) (stream t))

717771{-# INLINE intersperse #-}

718772

719−-- | /O(n)/ Reverse the characters of a string. Subject to fusion.

773+-- | /O(n)/ Reverse the characters of a string.

774+--

775+-- Example:

776+--

777+-- >>> T.reverse "desrever"

778+-- "reversed"

779+--

780+-- Subject to fusion.

720781reverse :: Text -> Text

721782reverse = coerce JSS.reverse

722783-- reverse t = S.reverse (stream t)

@@ -736,12 +797,14 @@ reverse = coerce JSS.reverse

736797-- @needle@ occurs in @replacement@, that occurrence will /not/ itself

737798-- be replaced recursively:

738799--

739−-- > replace "oo" "foo" "oo" == "foo"

800+-- >>> replace "oo" "foo" "oo"

801+-- "foo"

740802--

741803-- In cases where several instances of @needle@ overlap, only the

742804-- first one will be replaced:

743805--

744−-- > replace "ofo" "bar" "ofofo" == "barfo"

806+-- >>> replace "ofo" "bar" "ofofo"

807+-- "barfo"

745808--

746809-- In (unlikely) bad cases, this function's time complexity degrades

747810-- towards /O(n*m)/.

@@ -859,6 +922,8 @@ toUpper = coerce JSS.toUpper

859922-- guides disagree on whether the book name \"The Hill of the Red

860923-- Fox\" is correctly title cased—but this function will

861924-- capitalize /every/ word.

925+--

926+-- @since 1.0.0.0

862927toTitle :: Text -> Text

863928toTitle = coerce JSS.toTitle

864929-- toTitle t = unstream (S.toTitle (stream t))

@@ -870,8 +935,11 @@ toTitle = coerce JSS.toTitle

870935--

871936-- Examples:

872937--

873−-- > justifyLeft 7 'x' "foo" == "fooxxxx"

874−-- > justifyLeft 3 'x' "foobar" == "foobar"

938+-- >>> justifyLeft 7 'x' "foo"

939+-- "fooxxxx"

940+--

941+-- >>> justifyLeft 3 'x' "foobar"

942+-- "foobar"

875943justifyLeft :: Int -> Char -> Text -> Text

876944justifyLeft = coerce JSS.justifyLeft

877945-- justifyLeft k c t

@@ -893,8 +961,11 @@ justifyLeft = coerce JSS.justifyLeft

893961--

894962-- Examples:

895963--

896−-- > justifyRight 7 'x' "bar" == "xxxxbar"

897−-- > justifyRight 3 'x' "foobar" == "foobar"

964+-- >>> justifyRight 7 'x' "bar"

965+-- "xxxxbar"

966+--

967+-- >>> justifyRight 3 'x' "foobar"

968+-- "foobar"

898969justifyRight :: Int -> Char -> Text -> Text

899970justifyRight = coerce JSS.justifyRight

900971-- justifyRight k c t

@@ -909,7 +980,8 @@ justifyRight = coerce JSS.justifyRight

909980--

910981-- Examples:

911982--

912−-- > center 8 'x' "HS" = "xxxHSxxx"

983+-- >>> center 8 'x' "HS"

984+-- "xxxHSxxx"

913985center :: Int -> Char -> Text -> Text

914986center = coerce JSS.center

915987-- center k c t

@@ -925,6 +997,14 @@ center = coerce JSS.center

925997-- of its 'Text' argument. Note that this function uses 'pack',

926998-- 'unpack', and the list version of transpose, and is thus not very

927999-- efficient.

1000+--

1001+-- Examples:

1002+--

1003+-- >>> transpose ["green","orange"]

1004+-- ["go","rr","ea","en","ng","e"]

1005+--

1006+-- >>> transpose ["blue","red"]

1007+-- ["br","le","ud","e"]

9281008transpose :: [Text] -> [Text]

9291009transpose = coerce JSS.transpose

9301010-- transpose ts = P.map pack (L.transpose (P.map unpack ts))

@@ -1207,7 +1287,10 @@ take = coerce JSS.take

12071287--

12081288-- Examples:

12091289--

1210−-- > takeEnd 3 "foobar" == "bar"

1290+-- >>> takeEnd 3 "foobar"

1291+-- "bar"

1292+--

1293+-- @since 1.1.1.0

12111294takeEnd :: Int -> Text -> Text

12121295takeEnd = coerce JSS.takeEnd

12131296-- takeEnd n t@(Text arr off len)

@@ -1248,7 +1331,10 @@ drop = coerce JSS.drop

12481331--

12491332-- Examples:

12501333--

1251−-- > dropEnd 3 "foobar" == "foo"

1334+-- >>> dropEnd 3 "foobar"

1335+-- "foo"

1336+--

1337+-- @since 1.1.1.0

12521338dropEnd :: Int -> Text -> Text

12531339dropEnd = coerce JSS.dropEnd

12541340-- dropEnd n t@(Text arr off len)

@@ -1280,7 +1366,10 @@ takeWhile = coerce JSS.takeWhile

12801366-- satisfy @p@. Subject to fusion.

12811367-- Examples:

12821368--

1283−-- > takeWhileEnd (=='o') "foo" == "oo"

1369+-- >>> takeWhileEnd (=='o') "foo"

1370+-- "oo"

1371+--

1372+-- @since 1.2.2.0

12841373takeWhileEnd :: (Char -> Bool) -> Text -> Text

12851374#ifndef __GHCJS__

12861375takeWhileEnd p t@(Text arr off len) = loop (len-1) len

@@ -1329,7 +1418,8 @@ dropWhile = coerce JSS.dropWhile

13291418--

13301419-- Examples:

13311420--

1332−-- > dropWhileEnd (=='.') "foo..." == "foo"

1421+-- >>> dropWhileEnd (=='.') "foo..."

1422+-- "foo"

13331423dropWhileEnd :: (Char -> Bool) -> Text -> Text

13341424dropWhileEnd = coerce JSS.dropWhileEnd

13351425-- dropWhileEnd p t@(Text arr off len) = loop (len-1) len

@@ -1454,9 +1544,14 @@ tails = coerce JSS.tails

14541544--

14551545-- Examples:

14561546--

1457−-- > splitOn "\r\n" "a\r\nb\r\nd\r\ne" == ["a","b","d","e"]

1458−-- > splitOn "aaa" "aaaXaaaXaaaXaaa" == ["","X","X","X",""]

1459−-- > splitOn "x" "x" == ["",""]

1547+-- >>> splitOn "\r\n" "a\r\nb\r\nd\r\ne"

1548+-- ["a","b","d","e"]

1549+--

1550+-- >>> splitOn "aaa" "aaaXaaaXaaaXaaa"

1551+-- ["","X","X","X",""]

1552+--

1553+-- >>> splitOn "x" "x"

1554+-- ["",""]

14601555--

14611556-- and

14621557--

@@ -1493,8 +1588,11 @@ splitOn = coerce JSS.splitOn

14931588-- resulting components do not contain the separators. Two adjacent

14941589-- separators result in an empty component in the output. eg.

14951590--

1496−-- > split (=='a') "aabbaca" == ["","","bb","c",""]

1497−-- > split (=='a') "" == [""]

1591+-- >>> split (=='a') "aabbaca"

1592+-- ["","","bb","c",""]

1593+--

1594+-- >>> split (=='a') ""

1595+-- [""]

14981596split :: (Char -> Bool) -> Text -> [Text]

14991597split = coerce JSS.split

15001598-- split _ t@(Text _off _arr 0) = [t]

@@ -1508,8 +1606,11 @@ split = coerce JSS.split

15081606-- element may be shorter than the other chunks, depending on the

15091607-- length of the input. Examples:

15101608--

1511−-- > chunksOf 3 "foobarbaz" == ["foo","bar","baz"]

1512−-- > chunksOf 4 "haskell.org" == ["hask","ell.","org"]

1609+-- >>> chunksOf 3 "foobarbaz"

1610+-- ["foo","bar","baz"]

1611+--

1612+-- >>> chunksOf 4 "haskell.org"

1613+-- ["hask","ell.","org"]

15131614chunksOf :: Int -> Text -> [Text]

15141615chunksOf = coerce JSS.chunksOf

15151616-- chunksOf k = go

@@ -1558,8 +1659,11 @@ filter = coerce JSS.filter

15581659--

15591660-- Examples:

15601661--

1561−-- > breakOn "::" "a::b::c" ==> ("a", "::b::c")

1562−-- > breakOn "/" "foobar" ==> ("foobar", "")

1662+-- >>> breakOn "::" "a::b::c"

1663+-- ("a","::b::c")

1664+--

1665+-- >>> breakOn "/" "foobar"

1666+-- ("foobar","")

15631667--

15641668-- Laws:

15651669--

@@ -1588,7 +1692,8 @@ breakOn = coerce JSS.breakOn

15881692-- up to and including the last match of @needle@. The second is the

15891693-- remainder of @haystack@, following the match.

15901694--

1591−-- > breakOnEnd "::" "a::b::c" ==> ("a::b::", "c")

1695+-- >>> breakOnEnd "::" "a::b::c"

1696+-- ("a::b::","c")

15921697breakOnEnd :: Text -> Text -> (Text, Text)

15931698breakOnEnd = coerce JSS.breakOnEnd

15941699-- breakOnEnd pat src = (reverse b, reverse a)

@@ -1604,10 +1709,11 @@ breakOnEnd = coerce JSS.breakOnEnd

16041709--

16051710-- Examples:

16061711--

1607−-- > breakOnAll "::" ""

1608−-- > ==> []

1609−-- > breakOnAll "/" "a/b/c/"

1610−-- > ==> [("a", "/b/c/"), ("a/b", "/c/"), ("a/b/c", "/")]

1712+-- >>> breakOnAll "::" ""

1713+-- []

1714+--

1715+-- >>> breakOnAll "/" "a/b/c/"

1716+-- [("a","/b/c/"),("a/b","/c/"),("a/b/c","/")]

16111717--

16121718-- In (unlikely) bad cases, this function's time complexity degrades

16131719-- towards /O(n*m)/.

@@ -1821,9 +1927,14 @@ isInfixOf = coerce JSS.isInfixOf

18211927--

18221928-- Examples:

18231929--

1824−-- > stripPrefix "foo" "foobar" == Just "bar"

1825−-- > stripPrefix "" "baz" == Just "baz"

1826−-- > stripPrefix "foo" "quux" == Nothing

1930+-- >>> stripPrefix "foo" "foobar"

1931+-- Just "bar"

1932+--

1933+-- >>> stripPrefix "" "baz"

1934+-- Just "baz"

1935+--

1936+-- >>> stripPrefix "foo" "quux"

1937+-- Nothing

18271938--

18281939-- This is particularly useful with the @ViewPatterns@ extension to

18291940-- GHC, as follows:

@@ -1849,9 +1960,14 @@ stripPrefix = coerce JSS.stripPrefix

18491960--

18501961-- Examples:

18511962--

1852−-- > commonPrefixes "foobar" "fooquux" == Just ("foo","bar","quux")

1853−-- > commonPrefixes "veeble" "fetzer" == Nothing

1854−-- > commonPrefixes "" "baz" == Nothing

1963+-- >>> commonPrefixes "foobar" "fooquux"

1964+-- Just ("foo","bar","quux")

1965+--

1966+-- >>> commonPrefixes "veeble" "fetzer"

1967+-- Nothing

1968+--

1969+-- >>> commonPrefixes "" "baz"

1970+-- Nothing

18551971commonPrefixes :: Text -> Text -> Maybe (Text,Text,Text)

18561972commonPrefixes = coerce JSS.commonPrefixes

18571973-- commonPrefixes t0@(Text arr0 off0 len0) t1@(Text arr1 off1 len1) = go 0 0

@@ -1869,9 +1985,14 @@ commonPrefixes = coerce JSS.commonPrefixes

18691985--

18701986-- Examples:

18711987--

1872−-- > stripSuffix "bar" "foobar" == Just "foo"

1873−-- > stripSuffix "" "baz" == Just "baz"

1874−-- > stripSuffix "foo" "quux" == Nothing

1988+-- >>> stripSuffix "bar" "foobar"

1989+-- Just "foo"

1990+--

1991+-- >>> stripSuffix "" "baz"

1992+-- Just "baz"

1993+--

1994+-- >>> stripSuffix "foo" "quux"

1995+-- Nothing

18751996--

18761997-- This is particularly useful with the @ViewPatterns@ extension to

18771998-- GHC, as follows:

@@ -1923,3 +2044,11 @@ foreign import javascript unsafe

19232044 "h$jsstringReplicateChar" js_replicateChar :: Int -> Char -> JSString

19242045foreign import javascript unsafe

19252046 "$2.substr($1)" js_substr1 :: Int -> JSString -> JSString

2047+

2048+-------------------------------------------------

2049+-- NOTE: the named chunk below used by doctest;

2050+-- verify the doctests via `doctest -fobject-code Data/Text.hs`

2051+

2052+-- $setup

2053+-- >>> :set -XOverloadedStrings

2054+-- >>> import qualified Data.Text as T

Data/Text/Array.hs+4−1

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

77--

88-- License : BSD-style

99-- Maintainer : bos@serpentine.com

10−-- Stability : experimental

1110-- Portability : portable

1211--

1312-- Packed, unboxed, heap-resident arrays. Suitable for performance

@@ -80,6 +79,8 @@ import GHC.Word (Word16(..))

8079import Prelude hiding (length, read)

8180

8281-- | Immutable array type.

82+--

83+-- The 'Array' constructor is exposed since @text-1.1.1.3@

8384data Array = Array {

8485 aBA :: ByteArray#

8586#if defined(ASSERTS)

@@ -88,6 +89,8 @@ data Array = Array {

8889 }

8990

9091-- | Mutable array type, for use in the ST monad.

92+--

93+-- The 'MArray' constructor is exposed since @text-1.1.1.3@

9194data MArray s = MArray {

9295 maBA :: MutableByteArray# s

9396#if defined(ASSERTS)

Data/Text/Encoding.hs+67−11

@@ -12,14 +12,13 @@

1212--

1313-- License : BSD-style

1414-- Maintainer : bos@serpentine.com

15−-- Stability : experimental

1615-- Portability : portable

1716--

1817-- Functions for converting 'Text' values to and from 'ByteString',

1918-- using several standard encodings.

2019--

2120-- To gain access to a much larger family of encodings, use the

22−-- @text-icu@ package: <http://hackage.haskell.org/package/text-icu>

21+-- <http://hackage.haskell.org/package/text-icu text-icu package>.

2322

2423module Data.Text.Encoding

2524 (

@@ -67,7 +66,7 @@ import Control.Monad.ST.Unsafe (unsafeIOToST, unsafeSTToIO)

6766import Control.Monad.ST (unsafeIOToST, unsafeSTToIO)

6867#endif

6968

70−import Control.Exception (evaluate, try)

69+import Control.Exception (evaluate, try, throwIO, ErrorCall(ErrorCall))

7170import Control.Monad.ST (runST)

7271import Data.Bits ((.&.))

7372import Data.ByteString as B

@@ -82,7 +81,11 @@ import Data.Text.Show ()

8281#endif

8382import Data.Text.Unsafe (unsafeDupablePerformIO)

8483import Data.Word (Word8, Word32)

85−import Foreign.C.Types (CSize(..))

84+#if __GLASGOW_HASKELL__ >= 703

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

86+#else

87+import Foreign.C.Types (CSize)

88+#endif

8689import Foreign.ForeignPtr (withForeignPtr)

8790import Foreign.Marshal.Utils (with)

8891import Foreign.Ptr (Ptr, minusPtr, nullPtr, plusPtr)

@@ -137,6 +140,13 @@ decodeLatin1 (PS fp off len) = text a 0 len

137140 return dest

138141

139142-- | Decode a 'ByteString' containing UTF-8 encoded text.

143+--

144+-- __NOTE__: The replacement character returned by 'OnDecodeError'

145+-- MUST be within the BMP plane; surrogate code points will

146+-- automatically be remapped to the replacement char @U+FFFD@

147+-- (/since 0.11.3.0/), whereas code points beyond the BMP will throw an

148+-- 'error' (/since 1.2.3.1/); For earlier versions of @text@ using

149+-- those unsupported code points would result in undefined behavior.

140150decodeUtf8With :: OnDecodeError -> ByteString -> Text

141151decodeUtf8With onErr (PS fp off len) = runText $ \done -> do

142152 let go dest = withForeignPtr fp $ \ptr ->

@@ -152,16 +162,52 @@ decodeUtf8With onErr (PS fp off len) = runText $ \done -> do

152162 x <- peek curPtr'

153163 case onErr desc (Just x) of

154164 Nothing -> loop $ curPtr' `plusPtr` 1

155− Just c -> do

156− destOff <- peek destOffPtr

157− w <- unsafeSTToIO $

158− unsafeWrite dest (fromIntegral destOff) (safe c)

159− poke destOffPtr (destOff + fromIntegral w)

160− loop $ curPtr' `plusPtr` 1

165+ Just c

166+ | c > '\xFFFF' -> throwUnsupportedReplChar

167+ | otherwise -> do

168+ destOff <- peek destOffPtr

169+ w <- unsafeSTToIO $

170+ unsafeWrite dest (fromIntegral destOff)

171+ (safe c)

172+ poke destOffPtr (destOff + fromIntegral w)

173+ loop $ curPtr' `plusPtr` 1

161174 loop (ptr `plusPtr` off)

162175 (unsafeIOToST . go) =<< A.new len

163176 where

164177 desc = "Data.Text.Internal.Encoding.decodeUtf8: Invalid UTF-8 stream"

178+

179+ throwUnsupportedReplChar = throwIO $

180+ ErrorCall "decodeUtf8With: non-BMP replacement characters not supported"

181+ -- TODO: The code currently assumes that the transcoded UTF-16

182+ -- stream is at most twice as long (in bytes) as the input UTF-8

183+ -- stream. To justify this assumption one has to assume that the

184+ -- error handler replacement character also satisfies this

185+ -- invariant, by emitting at most one UTF16 code unit.

186+ --

187+ -- One easy way to support the full range of code-points for

188+ -- replacement characters in the error handler is to simply change

189+ -- the (over-)allocation to `A.new (2*len)` and then shrink back the

190+ -- `ByteArray#` to the real size (recent GHCs have a cheap

191+ -- `ByteArray#` resize-primop for that which allow the GC to reclaim

192+ -- the overallocation). However, this would require 4 times as much

193+ -- (temporary) storage as the original UTF-8 required.

194+ --

195+ -- Another strategy would be to optimistically assume that

196+ -- replacement characters are within the BMP, and if the case of a

197+ -- non-BMP replacement occurs reallocate the target buffer (or throw

198+ -- an exception, and fallback to a pessimistic codepath, like e.g.

199+ -- `decodeUtf8With onErr bs = F.unstream (E.streamUtf8 onErr bs)`)

200+ --

201+ -- Alternatively, `OnDecodeError` could become a datastructure which

202+ -- statically encodes the replacement-character range,

203+ -- e.g. something isomorphic to

204+ --

205+ -- Either (... -> Maybe Word16) (... -> Maybe Char)

206+ --

207+ -- And allow to statically switch between the BMP/non-BMP

208+ -- replacement-character codepaths. There's multiple ways to address

209+ -- this with different tradeoffs; but ideally we should optimise for

210+ -- the optimistic/error-free case.

165211{- INLINE[0] decodeUtf8With #-}

166212

167213-- $stream

@@ -218,6 +264,8 @@ decodeUtf8With onErr (PS fp off len) = runText $ \done -> do

218264-- or continuation where it is encountered.

219265

220266-- | A stream oriented decoding result.

267+--

268+-- @since 1.0.0.0

221269data Decoding = Some Text ByteString (ByteString -> Decoding)

222270

223271instance Show Decoding where

@@ -237,11 +285,15 @@ newtype DecoderState = DecoderState Word32 deriving (Eq, Show, Num, Storable)

237285-- thrown (either by this function or a continuation) that cannot be

238286-- caught in pure code. For more control over the handling of invalid

239287-- data, use 'streamDecodeUtf8With'.

288+--

289+-- @since 1.0.0.0

240290streamDecodeUtf8 :: ByteString -> Decoding

241291streamDecodeUtf8 = streamDecodeUtf8With strictDecode

242292

243293-- | Decode, in a stream oriented way, a 'ByteString' containing UTF-8

244294-- encoded text.

295+--

296+-- @since 1.0.0.0

245297streamDecodeUtf8With :: OnDecodeError -> ByteString -> Decoding

246298streamDecodeUtf8With onErr = decodeChunk B.empty 0 0

247299 where

@@ -317,6 +369,8 @@ decodeUtf8' = unsafeDupablePerformIO . try . evaluate . decodeUtf8With strictDec

317369{-# INLINE decodeUtf8' #-}

318370

319371-- | Encode text to a ByteString 'B.Builder' using UTF-8 encoding.

372+--

373+-- @since 1.1.0.0

320374encodeUtf8Builder :: Text -> B.Builder

321375encodeUtf8Builder = encodeUtf8BuilderEscaped (BP.liftFixedToBounded BP.word8)

322376

@@ -325,6 +379,8 @@ encodeUtf8Builder = encodeUtf8BuilderEscaped (BP.liftFixedToBounded BP.word8)

325379--

326380-- Use this function is to implement efficient encoders for text-based formats

327381-- like JSON or HTML.

382+--

383+-- @since 1.1.0.0

328384{-# INLINE encodeUtf8BuilderEscaped #-}

329385-- TODO: Extend documentation with references to source code in @blaze-html@

330386-- or @aeson@ that uses this function.

@@ -392,7 +448,7 @@ encodeUtf8 :: Text -> ByteString

392448encodeUtf8 (Text arr off len)

393449 | len == 0 = B.empty

394450 | otherwise = unsafeDupablePerformIO $ do

395− fp <- mallocByteString (len*4)

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

396452 withForeignPtr fp $ \ptr ->

397453 with ptr $ \destPtr -> do

398454 c_encode_utf8 destPtr (A.aBA arr) (fromIntegral off) (fromIntegral len)

Data/Text/Encoding/Error.hs−1

@@ -10,7 +10,6 @@

1010--

1111-- License : BSD-style

1212-- Maintainer : bos@serpentine.com

13−-- Stability : experimental

1413-- Portability : GHC

1514--

1615-- Types and functions for dealing with encoding and decoding errors

Data/Text/Foreign.hs+5−2

@@ -5,7 +5,6 @@

55--

66-- License : BSD-style

77-- Maintainer : bos@serpentine.com

8−-- Stability : experimental

98-- Portability : GHC

109--

1110-- Support for using 'Text' data with native code via the Haskell

@@ -108,7 +107,7 @@ takeWord16 (I16 n) t@(Text arr off len)

108107 | n >= len || m >= len = t

109108 | otherwise = Text arr off m

110109 where

111− m | w < 0xDB00 || w > 0xD8FF = n

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

112111 | otherwise = n+1

113112 w = A.unsafeIndex arr (off+n-1)

114113

@@ -157,6 +156,8 @@ asForeignPtr t@(Text _arr _off len) = do

157156-- | /O(n)/ Decode a C string with explicit length, which is assumed

158157-- to have been encoded as UTF-8. If decoding fails, a

159158-- 'UnicodeException' is thrown.

159+--

160+-- @since 1.0.0.0

160161peekCStringLen :: CStringLen -> IO Text

161162peekCStringLen cs = do

162163 bs <- unsafePackCStringLen cs

@@ -169,5 +170,7 @@ peekCStringLen cs = do

169170-- The temporary storage is freed when the subcomputation terminates

170171-- (either normally or via an exception), so the pointer to the

171172-- temporary storage must /not/ be used after this function returns.

173+--

174+-- @since 1.0.0.0

172175withCStringLen :: Text -> (CStringLen -> IO a) -> IO a

173176withCStringLen t act = unsafeUseAsCStringLen (encodeUtf8 t) act

Data/Text/IO.hs+4−1

@@ -8,7 +8,6 @@

99-- License : BSD-style

1010-- Maintainer : bos@serpentine.com

11−-- Stability : experimental

1211-- Portability : GHC

1312--

1413-- Efficient locale-sensitive support for text I\/O.

@@ -352,3 +351,7 @@ putStrLn = hPutStrLn stdout

352351-- >

353352-- > putStr_Utf16LE :: Text -> IO ()

354353-- > putStr_Utf16LE t = B.putStr (encodeUtf16LE t)

354+--

355+-- On transcoding errors, an 'IOError' exception is thrown. You can

356+-- use the API in "Data.Text.Encoding" if you need more control over

357+-- error handling or transcoding.

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

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

11{-# LANGUAGE Rank2Types #-}

22-- AUTOMATICALLY GENERATED - DO NOT EDIT

33-- Generated by scripts/CaseMapping.hs

4−-- CaseFolding-8.0.0.txt

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

6−-- SpecialCasing-8.0.0.txt

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

4+-- CaseFolding-9.0.0.txt

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

6+-- SpecialCasing-9.0.0.txt

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

99module Data.Text.Internal.Fusion.CaseMapping where

1010import Data.Char

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

371371foldMapping '\x13fc' s = Yield '\x13f4' (CC s '\x0000' '\x0000')

372372-- CHEROKEE SMALL LETTER MV

373373foldMapping '\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')

374392-- LATIN SMALL LETTER H WITH LINE BELOW

375393foldMapping '\x1e96' s = Yield '\x0068' (CC s '\x0331' '\x0000')

376394-- LATIN SMALL LETTER T WITH DIAERESIS

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

545563foldMapping '\x1ff7' s = Yield '\x03c9' (CC s '\x0342' '\x03b9')

546564-- GREEK CAPITAL LETTER OMEGA WITH PROSGEGRAMMENI

547565foldMapping '\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')

548568-- LATIN CAPITAL LETTER J WITH CROSSED-TAIL

549569foldMapping '\xa7b2' s = Yield '\x029d' (CC s '\x0000' '\x0000')

550570-- LATIN CAPITAL LETTER CHI

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

737757foldMapping '\xfb16' s = Yield '\x057e' (CC s '\x0576' '\x0000')

738758-- ARMENIAN SMALL LIGATURE MEN XEH

739759foldMapping '\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')

740832-- OLD HUNGARIAN CAPITAL LETTER A

741833foldMapping '\x10c80' s = Yield '\x10cc0' (CC s '\x0000' '\x0000')

742834-- OLD HUNGARIAN CAPITAL LETTER AA

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

839931foldMapping '\x10cb1' s = Yield '\x10cf1' (CC s '\x0000' '\x0000')

840932-- OLD HUNGARIAN CAPITAL LETTER US

841933foldMapping '\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')

8421002foldMapping c s = Yield (toLower c) (CC s '\0' '\0')

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

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

1−{-# LANGUAGE CPP #-}

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

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

33-- |

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

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

1919module Data.Text.Internal.Fusion.Size

2020 (

2121 Size

22− , exactly

22+ -- * Sizes

2323 , exactSize

2424 , maxSize

2525 , betweenSize

2626 , unknownSize

27+ , unionSize

28+ , charSize

29+ , codePointsSize

30+ -- * Querying sizes

31+ , exactly

2732 , smaller

2833 , larger

2934 , upperBound

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

3237 , isEmpty

3338 ) where

3439

40+import Data.Char (ord)

3541import Data.Text.Internal (mul)

3642#if defined(ASSERTS)

3743import Control.Exception (assert)

3844#endif

3945

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

4047data Size = Between {-# UNPACK #-} !Int {-# UNPACK #-} !Int -- ^ Lower and upper bounds on size.

4148 | Unknown -- ^ Unknown size.

4249 deriving (Eq, Show)

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

4653exactly _ = Nothing

4754{-# INLINE exactly #-}

4855

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+

4967exactSize :: Int -> Size

5068exactSize n =

5169#if defined(ASSERTS)

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

7189 Between m n

7290{-# INLINE betweenSize #-}

7391

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

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

94+unionSize _ _ = Unknown

95+

7496unknownSize :: Size

7597unknownSize = Unknown

7698{-# INLINE unknownSize #-}

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

140162lowerBound k _ = k

141163{-# INLINE lowerBound #-}

142164

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

144−compareSize (Between ma mb) n

145− | mb < n = Just LT

146− | ma > n = Just GT

147− | ma == n && mb == n = Just EQ

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

166+-- the indeterminate case.

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

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

169+ | mb < na = Just LT

170+ | ma > nb = Just GT

171+ | ma == mb

172+ , ma == na

173+ , ma == nb = Just EQ

148174compareSize _ _ = Nothing

149175

150176

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

@@ -78,16 +78,16 @@ import Data.JSString.Internal.Fusion.Types

7878-- -- unstreaming functions must be able to cope with the hint being too

7979-- -- small or too large.

8080-- --

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

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

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

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

8383-- -- undercount if, for instance, a transformed stream contains astral

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

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

8585

8686-- data Stream a =

8787-- forall s. Stream

8888-- (s -> Step s a) -- stepper function

8989-- !s -- current state

90−-- !Size -- size hint

90+-- !Size -- size hint in code units

9191

9292-- -- | /O(n)/ Determines if two streams are equal.

9393-- eq :: (Eq a) => Stream a -> Stream a -> Bool

Data/Text/Lazy.hs+54−20

@@ -15,7 +15,6 @@

1515--

1616-- License : BSD-style

1717-- Maintainer : bos@serpentine.com

18−-- Stability : experimental

1918-- Portability : GHC

2019--

2120-- A time and space-efficient implementation of Unicode text using

@@ -70,6 +69,7 @@ module Data.Text.Lazy

7069 , snoc

7170 , append

7271 , uncons

72+ , unsnoc

7373 , head

7474 , last

7575 , tail

@@ -278,14 +278,16 @@ import Data.JSString (JSString)

278278-- $replacement

279279--

280280-- A 'Text' value is a sequence of Unicode scalar values, as defined

281−-- in §3.9, definition D76 of the Unicode 5.2 standard:

282−-- <http://www.unicode.org/versions/Unicode5.2.0/ch03.pdf#page=35>. As

283−-- such, a 'Text' cannot contain values in the range U+D800 to U+DFFF

284−-- inclusive. Haskell implementations admit all Unicode code points

285−-- (§3.4, definition D10) as 'Char' values, including code points

286−-- from this invalid range. This means that there are some 'Char'

287−-- values that are not valid Unicode scalar values, and the functions

288−-- in this module must handle those cases.

281+-- in

282+-- <http://www.unicode.org/versions/Unicode5.2.0/ch03.pdf#page=35 §3.9, definition D76 of the Unicode 5.2 standard >.

283+-- As such, a 'Text' cannot contain values in the range U+D800 to

284+-- U+DFFF inclusive. Haskell implementations admit all Unicode code

285+-- points

286+-- (<http://www.unicode.org/versions/Unicode5.2.0/ch03.pdf#page=13 §3.4, definition D10 >)

287+-- as 'Char' values, including code points from this invalid range.

288+-- This means that there are some 'Char' values that are not valid

289+-- Unicode scalar values, and the functions in this module must handle

290+-- those cases.

289291--

290292-- Within this module, many functions construct a 'Text' from one or

291293-- more 'Char' values. Those functions will substitute 'Char' values

@@ -299,8 +301,8 @@ import Data.JSString (JSString)

299301-- range U+D800 through U+DFFF are used by UTF-16 to denote surrogate

300302-- code points, and so cannot be represented. The functions replace

301303-- invalid scalar values, instead of dropping them, as a security

302−-- measure. For details, see Unicode Technical Report 36, §3.5:

303−-- <http://unicode.org/reports/tr36#Deletion_of_Noncharacters>)

304+-- measure. For details, see

305+-- <http://unicode.org/reports/tr36/#Deletion_of_Noncharacters Unicode Technical Report 36, §3.5 >.)

304306

305307equal :: Text -> Text -> Bool

306308equal Empty Empty = True

@@ -362,9 +364,13 @@ instance Read Text where

362364 readsPrec p str = [(pack x,y) | (x,y) <- readsPrec p str]

363365

364366#if MIN_VERSION_base(4,9,0)

365−-- Semigroup orphan instances for older GHCs are provided by

366−-- 'semigroups` package

367−

367+-- | Non-orphan 'Semigroup' instance only defined for

368+-- @base-4.9.0.0@ and later; orphan instances for older GHCs are

369+-- provided by

370+-- the [semigroups](http://hackage.haskell.org/package/semigroups)

371+-- package

372+--

373+-- @since 1.2.2.0

368374instance Semigroup Text where

369375 (<>) = append

370376#endif

@@ -382,6 +388,7 @@ instance IsString Text where

382388 fromString = pack

383389

384390#if __GLASGOW_HASKELL__ >= 708

391+-- | @since 1.2.0.0

385392instance Exts.IsList Text where

386393 type Item Text = Char

387394 fromList = pack

@@ -394,6 +401,7 @@ instance NFData Text where

394401 rnf (Chunk _ ts) = rnf ts

395402#endif

396403

404+-- | @since 1.2.1.0

397405instance Binary Text where

398406 put t = put (encodeUtf8 t)

399407 get = do

@@ -417,6 +425,8 @@ instance Data Text where

417425

418426#if MIN_VERSION_base(4,7,0)

419427-- | Only defined for @base-4.7.0.0@ and later

428+--

429+-- @since 1.2.2.0

420430instance PrintfArg Text where

421431 formatArg txt = formatString $ unpack txt

422432#endif

@@ -495,9 +505,7 @@ fromStrict t = chunk t Empty

495505-- -----------------------------------------------------------------------------

496506-- * Basic functions

497507

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

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

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

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

501509cons :: Char -> Text -> Text

502510cons c t = Chunk (T.singleton c) t

503511{-# INLINE [1] cons #-}

@@ -581,6 +589,17 @@ init Empty = emptyError "init"

581589 unstream (S.init (stream t)) = init t

582590 #-}

583591

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

593+-- empty.

594+--

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

596+--

597+-- @since 1.2.3.0

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

599+unsnoc Empty = Nothing

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

601+{-# INLINE unsnoc #-}

602+

584603-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

585604-- fusion.

586605null :: Text -> Bool

@@ -840,6 +859,8 @@ toUpper t = unstream (S.toUpper (stream t))

840859-- guides disagree on whether the book name \"The Hill of the Red

841860-- Fox\" is correctly title cased—but this function will

842861-- capitalize /every/ word.

862+--

863+-- @since 1.0.0.0

843864toTitle :: Text -> Text

844865toTitle t = unstream (S.toTitle (stream t))

845866{-# INLINE toTitle #-}

@@ -992,6 +1013,8 @@ mapAccumR f = go

9921013

9931014-- | @'repeat' x@ is an infinite 'Text', with @x@ the value of every

9941015-- element.

1016+--

1017+-- @since 1.2.0.5

9951018repeat :: Char -> Text

9961019repeat c = let t = Chunk (T.replicate smallChunkSize (T.singleton c)) t

9971020 in t

@@ -1009,6 +1032,8 @@ replicate n t

10091032

10101033-- | 'cycle' ties a finite, non-empty 'Text' into a circular one, or

10111034-- equivalently, the infinite repetition of the original 'Text'.

1035+--

1036+-- @since 1.2.0.5

10121037cycle :: Text -> Text

10131038cycle Empty = emptyError "cycle"

10141039cycle t = let t' = foldrChunks Chunk t' t

@@ -1018,6 +1043,8 @@ cycle t = let t' = foldrChunks Chunk t' t

10181043-- of @f@ to @x@:

10191044--

10201045-- > iterate f x == [x, f x, f (f x), ...]

1046+--

1047+-- @since 1.2.0.5

10211048iterate :: (Char -> Char) -> Char -> Text

10221049iterate f c = let t c' = Chunk (T.singleton c') (t (f c'))

10231050 in t c

@@ -1081,6 +1108,8 @@ take i t0 = take' i t0

10811108-- Examples:

10821109--

10831110-- > takeEnd 3 "foobar" == "bar"

1111+--

1112+-- @since 1.1.1.0

10841113takeEnd :: Int64 -> Text -> Text

10851114takeEnd n t0

10861115 | n <= 0 = empty

@@ -1119,6 +1148,8 @@ drop i t0

11191148-- Examples:

11201149--

11211150-- > dropEnd 3 "foobar" == "foo"

1151+--

1152+-- @since 1.1.1.0

11221153dropEnd :: Int64 -> Text -> Text

11231154dropEnd n t0

11241155 | n <= 0 = t0

@@ -1179,12 +1210,15 @@ takeWhile p t0 = takeWhile' t0

11791210-- Examples:

11801211--

11811212-- > takeWhileEnd (=='o') "foo" == "oo"

1213+--

1214+-- @since 1.2.2.0

11821215takeWhileEnd :: (Char -> Bool) -> Text -> Text

11831216takeWhileEnd p = takeChunk empty . L.reverse . toChunks

11841217 where takeChunk acc [] = acc

1185− takeChunk acc (t:ts) = if T.length t' < T.length t

1186− then (Chunk t' acc)

1187− else takeChunk (Chunk t' acc) ts

1218+ takeChunk acc (t:ts)

1219+ | T.lengthWord16 t' < T.lengthWord16 t

1220+ = chunk t' acc

1221+ | otherwise = takeChunk (Chunk t' acc) ts

11881222 where t' = T.takeWhileEnd p t

11891223{-# INLINE takeWhileEnd #-}

11901224

Data/Text/Lazy/Builder.hs−1

@@ -11,7 +11,6 @@

1111-- License : BSD-style (see LICENSE)

1212--

1313-- Maintainer : Johan Tibell <johan.tibell@gmail.com>

14−-- Stability : experimental

1514-- Portability : portable to Hugs and GHC

1615--

1716-- Efficient construction of lazy @Text@ values. The principal

Data/Text/Lazy/Builder/Int.hs+46−2

@@ -1,4 +1,17 @@

1−module Data.Text.Lazy.Builder.Int where

1+-- Module: Data.Text.Lazy.Builder.Int

4+-- License: BSD-style

5+-- Maintainer: Bryan O'Sullivan <bos@serpentine.com>

6+-- Portability: portable

7+--

8+-- Efficiently write an integral value to a 'Builder'.

10+module Data.Text.Lazy.Builder.Int

11+ (

12+ decimal

13+ , hexadecimal

14+ ) where

215

316-- import Data.Text.Lazy (Text)

417import qualified Data.Text.Lazy as L

@@ -6,7 +19,38 @@ import qualified Data.Text as T

619import qualified Data.JSString.Int as JSS

720import Data.Text.Internal (Text (..))

821import Data.Text.Lazy.Builder

9−import Data.Monoid

22+

23+import Data.Int (Int8, Int16, Int32, Int64)

24+import Data.Monoid (mempty)

25+import qualified Data.ByteString.Unsafe as B

26+import Data.Text.Internal.Builder.Functions ((<>), i2d)

27+import Data.Text.Internal.Builder

28+import Data.Text.Internal.Builder.Int.Digits (digits)

29+import Data.Text.Array

30+import Data.Word (Word, Word8, Word16, Word32, Word64)

31+import GHC.Base (quotInt, remInt)

32+import GHC.Num (quotRemInteger)

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

34+import Control.Monad.ST

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

36+import Prelude hiding ((<>))

37+#endif

38+

39+#ifdef __GLASGOW_HASKELL__

40+# if defined(INTEGER_GMP)

41+import GHC.Integer.GMP.Internals (Integer(S#))

42+# elif defined(INTEGER_SIMPLE)

43+import GHC.Integer

44+# else

45+# error "You need to use either GMP or integer-simple."

46+# endif

47+#endif

48+

49+#if defined(INTEGER_GMP) || defined(INTEGER_SIMPLE)

50+# define PAIR(a,b) (# a,b #)

51+#else

52+# define PAIR(a,b) (a,b)

53+#endif

1054

1155decimal :: Integral a => a -> Builder

1256decimal n = fromText $ Text $ JSS.decimal n

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

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

2424import Data.Text.Internal.Builder.RealFloat.Functions (roundTo)

2525import Data.Text.Lazy.Builder

2626import qualified Data.Text as T

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

28+import Prelude hiding ((<>))

29+#endif

2730

2831

2932-- | Control the rendering of floating point numbers.

Data/Text/Lazy/Encoding.hs+13−3

@@ -8,14 +8,13 @@

88--

99-- License : BSD-style

1010-- Maintainer : bos@serpentine.com

11−-- Stability : experimental

1211-- Portability : portable

1312--

1413-- Functions for converting lazy 'Text' values to and from lazy

1514-- 'ByteString', using several standard encodings.

1615--

17−-- To gain access to a much larger variety of encodings, use the

18−-- @text-icu@ package: <http://hackage.haskell.org/package/text-icu>

16+-- To gain access to a much larger family of encodings, use the

17+-- <http://hackage.haskell.org/package/text-icu text-icu package>.

1918

2019module Data.Text.Lazy.Encoding

2120 (

@@ -141,6 +140,7 @@ decodeUtf8' bs = unsafeDupablePerformIO $ do

141140 rnf (Chunk _ ts) = rnf ts

142141{-# INLINE decodeUtf8' #-}

143142

143+-- | Encode text using UTF-8 encoding.

144144encodeUtf8 :: Text -> B.ByteString

145145encodeUtf8 Empty = B.empty

146146encodeUtf8 lt@(Chunk t _) =

@@ -154,10 +154,20 @@ encodeUtf8 lt@(Chunk t _) =

154154 firstChunkSize = min B.smallChunkSize (4 * (T.length t + 1))

155155 strategy = B.safeStrategy firstChunkSize B.defaultChunkSize

156156

157+-- | Encode text to a ByteString 'B.Builder' using UTF-8 encoding.

158+--

159+-- @since 1.1.0.0

157160encodeUtf8Builder :: Text -> B.Builder

158161encodeUtf8Builder =

159162 foldrChunks (\c b -> TE.encodeUtf8Builder c `mappend` b) Data.Monoid.mempty

160163

164+-- | Encode text using UTF-8 encoding and escape the ASCII characters using

165+-- a 'BP.BoundedPrim'.

166+--

167+-- Use this function is to implement efficient encoders for text-based formats

168+-- like JSON or HTML.

169+--

170+-- @since 1.1.0.0

161171{-# INLINE encodeUtf8BuilderEscaped #-}

162172encodeUtf8BuilderEscaped :: BP.BoundedPrim Word8 -> Text -> B.Builder

163173encodeUtf8BuilderEscaped prim =

Data/Text/Lazy/IO.hs−1

@@ -8,7 +8,6 @@

99-- License : BSD-style

1010-- Maintainer : bos@serpentine.com

11−-- Stability : experimental

1211-- Portability : GHC

1312--

1413-- Efficient locale-sensitive support for lazy text I\/O.

Data/Text/Lazy/Read.hs+3−4

@@ -11,7 +11,6 @@

1111--

1212-- License : BSD-style

1313-- Maintainer : bos@serpentine.com

14−-- Stability : experimental

1514-- Portability : GHC

1615--

1716-- Functions used frequently when reading textual data.

@@ -68,9 +67,9 @@ decimal txt

6867 go n d = (n * 10 + fromIntegral (digitToInt d))

6968

7069-- | Read a hexadecimal integer, consisting of an optional leading

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

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

73−-- is case insensitive.

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

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

72+-- This function is case insensitive.

7473--

7574-- This function does not handle leading sign characters. If you need

7675-- to handle signed input, use @'signed' 'hexadecimal'@.

Data/Text/Read.hs+3−4

@@ -9,7 +9,6 @@

99--

1010-- License : BSD-style

1111-- Maintainer : bos@serpentine.com

12−-- Stability : experimental

1312-- Portability : GHC

1413--

1514-- Functions used frequently when reading textual data.

@@ -67,9 +66,9 @@ decimal txt

6766 go n d = (n * 10 + fromIntegral (digitToInt d))

6867

6968-- | Read a hexadecimal integer, consisting of an optional leading

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

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

72−-- is case insensitive.

69+-- @\"0x\"@ followed by at least one hexadecimal digit. Input is

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

71+-- This function is case insensitive.

7372--

7473-- This function does not handle leading sign characters. If you need

7574-- to handle signed input, use @'signed' 'hexadecimal'@.

Data/Text/Show.hs+2

@@ -46,6 +46,8 @@ unpack = S.unstreamList . stream

4646-- fusion.

4747--

4848-- This is exposed solely for people writing GHC rewrite rules.

49+--

50+-- @since 1.2.1.1

4951unpackCString# :: Addr# -> Text

5052unpackCString# addr# = unstream (S.streamCString# addr#)

5153{-# NOINLINE unpackCString# #-}

Data/Text/Unsafe.hs+2−1

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

88-- License : BSD-style

99-- Maintainer : bos@serpentine.com

10−-- Stability : experimental

1110-- Portability : portable

1211--

1312-- A module containing unsafe 'Text' operations, for very very careful

@@ -150,6 +149,8 @@ reverseIter (Text txt) i

150149-- | /O(1)/ Iterate one step backwards through a UTF-16 array,

151150-- returning the delta to add (i.e. a negative number) to give the

152151-- next offset to iterate at.

152+--

153+-- @since 1.1.1.0

153154reverseIter_ :: Text -> Int -> Int

154155#ifndef __GHCJS__

155156reverseIter_ (Text arr off _len) i

README.markdown+5−16

@@ -1,29 +1,18 @@

1−# Text: Fast, packed Unicode strings, using stream fusion

1+# `text`: Fast, packed Unicode strings, using stream fusion

33This package provides the Data.Text library, a library for the space-

44and time-efficient manipulation of Unicode text in Haskell.

6−

7−# Normalization, conversion, and collation, oh my!

8−

9−This library intentionally provides a simple API based on the

10−Haskell prelude's list manipulation functions. For more complicated

11−real-world tasks, such as Unicode normalization, conversion to and

12−from a larger variety of encodings, and collation, use the [text-icu

13−package](http://hackage.haskell.org/cgi-bin/hackage-scripts/package/text-icu).

14−

15−That library uses the well-respected and liberally licensed ICU

16−library to provide these facilities.

17−

6+**Please refer to the [package description on Hackage](https://hackage.haskell.org/package/text#description) for more information.**

187

198# Get involved!

209

2110Please report bugs via the

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

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

2312

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

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

2514

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

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

2716

2817There's also a [Mercurial mirror](https://bitbucket.org/bos/text):

2918

benchmarks/.gitignore+4−1

@@ -1 +1,4 @@

1−dist

1+/dist/

2+/dist-newstyle/

3+/.ghc.environment.*

4+/cabal.project.local

benchmarks/cabal.projectnew file+2

@@ -0,0 +1,2 @@

1+-- NB: we use a separate project

2+packages: .

benchmarks/haskell/Benchmarks.hs+2

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

1010import System.IO (IOMode (WriteMode), openFile, hSetEncoding, utf8)

1111

1212import qualified Benchmarks.Builder as Builder

13+import qualified Benchmarks.Concat as Concat

1314import qualified Benchmarks.DecodeUtf8 as DecodeUtf8

1415import qualified Benchmarks.EncodeUtf8 as EncodeUtf8

1516import qualified Benchmarks.Equality as Equality

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

4142 -- Traditional benchmarks

4243 bs <- sequence

4344 [ Builder.benchmark

45+ , Concat.benchmark

4446 , DecodeUtf8.benchmark "html" (tf "libya-chinese.html")

4547 , DecodeUtf8.benchmark "xml" (tf "yiwiki.xml")

4648 , 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+82−8

@@ -12,8 +12,7 @@ author: Jasper Van der Jeugt <jaspervdj@gmail.com>,

1212maintainer: jaspervdj@gmail.com

1313category: Text

1414build-type: Simple

15−

16−cabal-version: >=1.2

15+cabal-version: >=1.8

1716

1817flag bytestring-builder

1918 description: Depend on the bytestring-builder package for backwards compatibility.

@@ -26,16 +25,12 @@ flag llvm

2625 manual: True

2726

2827executable text-benchmarks

29− hs-source-dirs: haskell ..

30− c-sources: ../cbits/cbits.c

31− cbits/time_iconv.c

32− include-dirs: ../include

33− main-is: Benchmarks.hs

3428 ghc-options: -Wall -O2 -rtsopts

3529 if flag(llvm)

3630 ghc-options: -fllvm

3731 cpp-options: -DHAVE_DEEPSEQ -DINTEGER_GMP

38− build-depends: base == 4.*,

32+ build-depends: array,

33+ base == 4.*,

3934 binary,

4035 blaze-builder,

4136 bytestring-lexing >= 0.5.0,

@@ -47,6 +42,7 @@ executable text-benchmarks

4742 ghc-prim,

4843 integer-gmp,

4944 stringsearch,

45+ transformers,

5046 utf8-string,

5147 vector

5248

@@ -56,9 +52,87 @@ executable text-benchmarks

5652 else

5753 build-depends: bytestring >= 0.10.4

5854

55+ -- modules for benchmark proper

56+ c-sources: cbits/time_iconv.c

57+ hs-source-dirs: haskell

58+ main-is: Benchmarks.hs

59+ other-modules:

60+ Benchmarks.Builder

61+ Benchmarks.Concat

62+ Benchmarks.DecodeUtf8

63+ Benchmarks.EncodeUtf8

64+ Benchmarks.Equality

65+ Benchmarks.FileRead

66+ Benchmarks.FoldLines

67+ Benchmarks.Mul

68+ Benchmarks.Programs.BigTable

69+ Benchmarks.Programs.Cut

70+ Benchmarks.Programs.Fold

71+ Benchmarks.Programs.Sort

72+ Benchmarks.Programs.StripTags

73+ Benchmarks.Programs.Throughput

74+ Benchmarks.Pure

75+ Benchmarks.ReadNumbers

76+ Benchmarks.Replace

77+ Benchmarks.Search

78+ Benchmarks.Stream

79+ Benchmarks.WordFrequencies

80+

81+ -- Source code for IUT (implementation under test)

82+ -- "borrowed" from parent folder

83+ include-dirs: ../include

84+ c-sources: ../cbits/cbits.c

85+ hs-source-dirs: ..

86+ other-modules:

87+ Data.Text

88+ Data.Text.Array

89+ Data.Text.Encoding

90+ Data.Text.Encoding.Error

91+ Data.Text.Foreign

92+ Data.Text.IO

93+ Data.Text.Internal

94+ Data.Text.Internal.Builder

95+ Data.Text.Internal.Builder.Functions

96+ Data.Text.Internal.Builder.Int.Digits

97+ Data.Text.Internal.Builder.RealFloat.Functions

98+ Data.Text.Internal.Encoding.Fusion

99+ Data.Text.Internal.Encoding.Fusion.Common

100+ Data.Text.Internal.Encoding.Utf16

101+ Data.Text.Internal.Encoding.Utf32

102+ Data.Text.Internal.Encoding.Utf8

103+ Data.Text.Internal.Functions

104+ Data.Text.Internal.Fusion

105+ Data.Text.Internal.Fusion.CaseMapping

106+ Data.Text.Internal.Fusion.Common

107+ Data.Text.Internal.Fusion.Size

108+ Data.Text.Internal.Fusion.Types

109+ Data.Text.Internal.IO

110+ Data.Text.Internal.Lazy

111+ Data.Text.Internal.Lazy.Encoding.Fusion

112+ Data.Text.Internal.Lazy.Fusion

113+ Data.Text.Internal.Lazy.Search

114+ Data.Text.Internal.Private

115+ Data.Text.Internal.Read

116+ Data.Text.Internal.Search

117+ Data.Text.Internal.Unsafe

118+ Data.Text.Internal.Unsafe.Char

119+ Data.Text.Internal.Unsafe.Shift

120+ Data.Text.Lazy

121+ Data.Text.Lazy.Builder

122+ Data.Text.Lazy.Builder.Int

123+ Data.Text.Lazy.Builder.RealFloat

124+ Data.Text.Lazy.Encoding

125+ Data.Text.Lazy.IO

126+ Data.Text.Lazy.Internal

127+ Data.Text.Lazy.Read

128+ Data.Text.Read

129+ Data.Text.Unsafe

130+ Data.Text.Show

131+

59132executable text-multilang

60133 hs-source-dirs: haskell

61134 main-is: Multilang.hs

135+ other-modules: Timer

62136 ghc-options: -Wall -O2

63137 build-depends: base == 4.*,

64138 bytestring,

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,

206206 const uint8_t *srcend,

207207 uint32_t *codepoint0, uint32_t *state0)

208208{

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

210− codepoint0, state0);

211− if (*state0 == UTF8_REJECT)

212− ret -=1;

213− return ret;

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

210+

211+ return *src;

214212}

215213

216214/*

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

222220{

223221 uint32_t codepoint;

224222 uint32_t state = UTF8_ACCEPT;

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

226− &codepoint, &state);

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

228− if (state != UTF8_ACCEPT)

229− ret -= 1;

230− return ret;

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

224+ &codepoint, &state);

225+ return src;

231226}

232227

233228void

changelog.md+61−19

@@ -1,15 +1,57 @@

1−1.2.2.1

1+### 1.2.3.1 TBD

3+* Make `decodeUtf8With` fail explicitly for unsupported non-BMP

4+ replacement characters instead silent undefined behaviour (gh-213)

6+### 1.2.3.0

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

9+ (updated from 8.0).

10+

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

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

13+ (gh-184).

14+

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

16+

17+* New function: `unsnoc` (gh-173).

18+

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

20+

21+* Improve UTF-8 decoder error-recovery (gh-182).

22+

23+* Minor documentation improvements (`@since` annotations, more

24+ examples, clarifications).

25+

26+#### 1.2.2.2

27+

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

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

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

31+

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

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

34+ smaller code generated.

35+

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

37+ used after some rewrite rule tweaking.

38+

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

40+

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

42+

43+#### 1.2.2.1

244

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

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

547 maintainers apologizes for the churn.

648

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

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

850 (updated from 7.0)

951

1052* An STG lint error has been fixed

1153

12−1.2.2.0

54+### 1.2.2.0

1355

1456* The `integer-simple` package, upon which this package optionally

1557 depended, has been replaced with `integer-pure`. The build flag has

@@ -25,7 +67,7 @@

2567 * if `base` >= 4.7: `PrintfArg`

2668 * if `base` >= 4.9: `Semigroup`

2769

28−1.2.1.3

70+#### 1.2.1.3

2971

3072* Bug fix: As it turns out, moving the literal rewrite rules to simplifier

3173 phase 2 does not prevent competition with the `unpack` rule, which is

@@ -33,7 +75,7 @@

3375 test environment mistake. Moving literal rules back to phase 1 finally

3476 fixes GHC Trac #10528 correctly.

3577

36−1.2.1.2

78+#### 1.2.1.2

3779

3880* Bug fix: Run literal rewrite rules in simplifier phase 2.

3981 The behavior of the simplifier changed in GHC 7.10.2,

@@ -41,21 +83,21 @@

4183 and long compilation times. See

4284 [GHC Trac #10528](https://ghc.haskell.org/trac/ghc/ticket/10528).

4385

44−1.2.1.1

86+#### 1.2.1.1

4587

4688* Expose unpackCString#, which you should never use.

4789

48−1.2.1.0

90+### 1.2.1.0

4991

5092* Added Binary instances for both Text types. (If you have previously

5193 been using the text-binary package to get a Binary instance, it is

5294 now obsolete.)

5395

54−1.2.0.6

96+#### 1.2.0.6

5597

5698* Fixed a space leak in UTF-8 decoding

5799

58−1.2.0.5

100+#### 1.2.0.5

59101

60102* Feature parity: repeat, cycle, iterate are now implemented for lazy

61103 Text, and the Data instance is more complete

@@ -75,26 +117,26 @@

75117* Spec compliance: toCaseFold now follows the Unicode 7.0 spec

76118 (updated from 6.3)

77119

78−1.2.0.4

120+#### 1.2.0.4

79121

80122* Fixed an incompatibility with base < 4.5

81123

82−1.2.0.3

124+#### 1.2.0.3

83125

84126* Update formatRealFloat to correspond to the definition in versions

85127 of base newer than 4.5 (https://github.com/bos/text/issues/105)

86128

87−1.2.0.2

129+#### 1.2.0.2

88130

89131* Bumped lower bound on deepseq to 1.4 for compatibility with the

90132 upcoming GHC 7.10

91133

92−1.2.0.1

134+#### 1.2.0.1

93135

94136* Fixed a buffer overflow in rendering of large Integers

95137 (https://github.com/bos/text/issues/99)

96138

97−1.2.0.0

139+## 1.2.0.0

98140

99141* Fixed an integer overflow in the replace function

100142 (https://github.com/bos/text/issues/81)

@@ -107,7 +149,7 @@

107149

108150* Added an instance of IsList for GHC 7.8 and above

109151

110−1.1.1.0

152+### 1.1.1.0

111153

112154* The Data.Data instance now allows gunfold to work, via a virtual

113155 pack constructor

@@ -117,12 +159,12 @@

117159* Comparing the length of a Text against a number can now

118160 short-circuit in more cases

119161

120−1.1.0.1

162+#### 1.1.0.1

121163

122164* streamDecodeUtf8: fixed gh-70, did not return all unconsumed bytes

123165 in single-byte chunks

124166

125−1.1.0.0

167+## 1.1.0.0

126168

127169* encodeUtf8: Performance is improved by up to 4x.

128170

@@ -141,12 +183,12 @@

141183 use at your own risk - there are no API stability guarantees for

142184 internal modules!

143185

144−1.0.0.1

186+#### 1.0.0.1

145187

146188* decodeUtf8: Fixed a regression that caused us to incorrectly

147189 identify truncated UTF-8 as valid (gh-61)

148190

149−1.0.0.0

191+# 1.0.0.0

150192

151193* Added support for Unicode 6.3.0 to case conversion functions

152194

tests-and-benchmarks.markdown+13−8

@@ -13,10 +13,13 @@ you can clone from one of the following locations:

1313* Git mirror repository:

1414 [github.com/bos/text-test-data](https://github.com/bos/text-test-data)

1515

16−You should clone that repository into the `tests` subdirectory (your

17−clone must be named `text-test-data` locally), then run `make -C

18−tests/text-test-data` to uncompress the test files. Many tests and

19−benchmarks will fail if the test files are missing.

16+You can clone either repository into the `tests` subdirectory using

17+

18+ cd tests/

19+ make text-test-data # to clone from mercurial, OR

20+ make VCS=git text-test-data # to clone from git

21+

22+Many tests and benchmarks will fail if the test files are missing.

2023

2124Functional tests

2225----------------

@@ -54,10 +57,12 @@ To compile the benchmarks, navigate to the `benchmarks` subdirectory and run

5457

5558 ./dist/build/text-benchmarks/text-benchmarks

5659

57−However, since there's quite a lot of benchmarks, you usually don't want to

58−run them all. Instead, use the `-l` flag to get a list of benchmarks:

60+Or if you have a recent enough `cabal`, you can build and run the

61+benchmarks via

5962

60− ./dist/build/text-benchmarks/text-benchmarks

63+ cabal new-run exe:text-benchmarks -- --help

6164

62−And run the ones you want to inspect. If you want to configure the benchmarks

65+However, since there's quite a lot of benchmarks, you usually don't want to

66+run them all. Instead, use the `-l` flag to get a list of benchmarks

67+and run the ones you want to inspect. If you want to configure the benchmarks

6368further, the exact parameters can be changed in `Benchmarks.hs`.

tests/Makefile+6−1

@@ -1,3 +1,4 @@

1+VCS = hg

12count = 1000

34all: coverage literal-rule-test

@@ -12,7 +13,11 @@ build: text-test-data

1213 cabal build

1314

1415text-test-data:

16+ifeq ($(VCS),git)

17+ git clone https://github.com/bos/text-test-data.git

18+else

1519 hg clone https://bitbucket.org/bos/text-test-data

20+endif

1621 $(MAKE) -C text-test-data

1722

1823coverage/text-tests.tix:

@@ -37,4 +42,4 @@ coverage/hpc_index.html: coverage/coverage.tix

3742clean:

3843 rm -rf dist coverage .hpc

3944

40−.PHONY: build coverage all literal-rule-test

45+.PHONY: all build clean coverage literal-rule-test

tests/Tests/Properties.hs+94−21

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

99 ) where

1010

1111import Control.Applicative ((<$>), (<*>))

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

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

1313import 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)

1515import Data.Int (Int8, Int16, Int32, Int64)

1616import Data.Monoid (Monoid(..))

1717import Data.String (IsString(fromString))

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

2323import Data.Text.Lazy.Read as TL

2424import Data.Text.Read as T

2525import Data.Word (Word, Word8, Word16, Word32, Word64)

26+import Data.Maybe (mapMaybe)

2627import Numeric (showEFloat, showFFloat, showGFloat, showHex)

2728import Prelude hiding (replicate)

2829import Test.Framework (Test, testGroup)

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

3031import Test.QuickCheck hiding ((.&.))

3132import Test.QuickCheck.Monadic

3233import Test.QuickCheck.Property (Property(..))

34+import Test.QuickCheck.Unicode (char)

3335import Tests.QuickCheckUtils

3436import Tests.Utils

3537import Text.Show.Functions ()

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

3739import qualified Data.Bits as Bits (shiftL, shiftR)

3840import qualified Data.ByteString as B

3941import qualified Data.ByteString.Lazy as BL

42+import qualified Data.Char as C

4043import qualified Data.List as L

4144import qualified Data.Text as T

4245import qualified Data.Text.Encoding as E

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

6063tl_pack_unpack = (TL.unpack . TL.pack) `eq` id

6164t_stream_unstream = (S.unstream . S.stream) `eq` id

6265tl_stream_unstream = (SL.unstream . SL.stream) `eq` id

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

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

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

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

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

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

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

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

6770 where f n = SL.unstreamChunks n . S.streamList

6871tl_chunk_unchunk = (TL.fromChunks . TL.toChunks) `eq` id

6972tl_from_to_strict = (TL.fromStrict . TL.toStrict) `eq` id

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

7477encodeLazyL1 :: TL.Text -> BL.ByteString

7578encodeLazyL1 = BL.fromChunks . map encodeL1 . TL.toChunks

7679

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

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

7881 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

8083 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

8285 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

8487 where a = TL.map (\c -> chr (ord c `mod` 256)) t

8588t_utf8 = forAll genUnicode $ (E.decodeUtf8 . E.encodeUtf8) `eq` id

8689t_utf8' = forAll genUnicode $ (E.decodeUtf8' . E.encodeUtf8) `eq` (id . Right)

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

111114 let b = E.encodeUtf8 t

112115 ls = concatMap (leftover . E.encodeUtf8 . T.singleton) . T.unpack $ t

113116 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

115118

116119data Badness = Solo | Leading | Trailing

117120 deriving (Eq, Show)

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

177180 k <- choose (0,n)

178181 vectorOf k gen

179182

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+

180210s_Eq s = (s==) `eq` ((S.streamList s==) . S.streamList)

181211 where _types = s :: String

182212sf_Eq p s =

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

201231 mconcat `eq` (unpackS . mconcat . L.map T.pack)

202232tl_mconcat = unsquare $

203233 mconcat `eq` (unpackS . mconcat . L.map TL.pack)

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

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

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

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

206236t_IsString = fromString `eqP` (T.unpack . fromString)

207237tl_IsString = fromString `eqP` (TL.unpack . fromString)

208238

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

229259 (fmap (second unpackS) . S.uncons . S.filter p)

230260t_uncons = uncons `eqP` (fmap (second unpackS) . T.uncons)

231261tl_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+

232269s_head = head `eqP` S.head

233270sf_head p = (head . L.filter p) `eqP` (S.head . S.filter p)

234271t_head = head `eqP` T.head

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

326363 where p = T.length . T.filter isUpper

327364tl_toUpper_upper t = p (TL.toUpper t) >= p t

328365 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'

329379

330380justifyLeft k c xs = xs ++ L.replicate (k - length xs) c

331381justifyRight 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)

514564s_takeWhile_s p = L.takeWhile p `eqP` (unpackS . S.unstream . S.takeWhile p)

515565sf_takeWhile q p = (L.takeWhile p . L.filter q) `eqP`

516566 (unpackS . S.takeWhile p . S.filter q)

567+noMatch = do

568+ c <- char

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

570+ return (c,d)

517571t_takeWhile p = L.takeWhile p `eqP` (unpackS . T.takeWhile p)

518572tl_takeWhile p = L.takeWhile p `eqP` (unpackS . TL.takeWhile p)

519573t_takeWhileEnd p = (L.reverse . L.takeWhile p . L.reverse) `eqP`

520574 (unpackS . T.takeWhileEnd p)

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

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

521577tl_takeWhileEnd p = (L.reverse . L.takeWhile p . L.reverse) `eqP`

522578 (unpackS . TL.takeWhileEnd p)

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

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

523581s_dropWhile p = L.dropWhile p `eqP` (unpackS . S.dropWhile p)

524582s_dropWhile_s p = L.dropWhile p `eqP` (unpackS . S.unstream . S.dropWhile p)

525583sf_dropWhile q p = (L.dropWhile p . L.filter q) `eqP`

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

720778 conc = T.concat . TL.toChunks

721779t_indices_occurs = unsquare $ \(NotEmpty t) ts ->

722780 let s = T.intercalate t ts

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

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

724782

725783-- Bit shifts.

726784shiftL 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

812870-- Reading.

813871

814872t_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)

816874 where t = T.dropWhile isDigit s

817875tl_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)

819877 where t = TL.dropWhile isDigit s

820878t_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)

822880 where t = T.dropWhile isHexDigit s

823881 p = if ox then "0x" else ""

824882 n = getPositive m :: Int

825883tl_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)

827885 where t = TL.dropWhile isHexDigit s

828886 p = if ox then "0x" else ""

829887 n = getPositive m :: Int

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

867925

868926t_dropWord16 m t = dropWord16 m t `T.isSuffixOf` t

869927t_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

871929 where n = small m

872930t_use_from t = monadicIO $ assert . (==t) =<< run (useAsPtr t fromPtr)

873931

874−t_copy t = T.copy t == t

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

875933

876934-- Regression tests.

877935s_filter_eq s = S.filter p t == S.streamList (filter p s)

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

924982 testGroup "errors" [

925983 testProperty "t_utf8_err" t_utf8_err,

926984 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

927994 ]

928995 ],

929996

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

9671034 testProperty "sf_uncons" sf_uncons,

9681035 testProperty "t_uncons" t_uncons,

9691036 testProperty "tl_uncons" tl_uncons,

1037+ testProperty "t_unsnoc" t_unsnoc,

1038+ testProperty "tl_unsnoc" tl_unsnoc,

9701039 testProperty "s_head" s_head,

9711040 testProperty "sf_head" sf_head,

9721041 testProperty "t_head" t_head,

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

10301099 testProperty "tl_toLower_lower" tl_toLower_lower,

10311100 testProperty "t_toUpper_length" t_toUpper_length,

10321101 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

10341105 ],

10351106

10361107 testGroup "justification" [

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

11471218 testProperty "t_takeWhile" t_takeWhile,

11481219 testProperty "tl_takeWhile" tl_takeWhile,

11491220 testProperty "t_takeWhileEnd" t_takeWhileEnd,

1221+ testProperty "t_takeWhileEnd_null" t_takeWhileEnd_null,

11501222 testProperty "tl_takeWhileEnd" tl_takeWhileEnd,

1223+ testProperty "tl_takeWhileEnd_null" tl_takeWhileEnd_null,

11511224 testProperty "sf_dropWhile" sf_dropWhile,

11521225 testProperty "s_dropWhile" s_dropWhile,

11531226 testProperty "s_dropWhile_s" s_dropWhile_s,

tests/Tests/QuickCheckUtils.hs+1−1

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

106106 where smallish = round . (sqrt :: Double -> Double) . fromIntegral . abs

107107

108108instance Arbitrary T.Text where

109− arbitrary = T.pack `fmap` arbitrary

109+ arbitrary = T.pack `fmap` string

110110 shrink = map T.pack . shrink . T.unpack

111111

112112instance Arbitrary TL.Text where

tests/Tests/Regressions.hs+11

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

7272 assertEqual "mapAccumL should correctly size buffers for two-word results"

7373 (count * 2) (T.lengthWord16 (snd val))

7474

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+

7585tests :: F.Test

7686tests = F.testGroup "Regressions"

7787 [ F.testCase "hGetContents_crash" hGetContents_crash

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

7989 , F.testCase "mapAccumL_resize" mapAccumL_resize

8090 , F.testCase "replicate_crash" replicate_crash

8191 , F.testCase "utf8_decode_unsafe" utf8_decode_unsafe

92+ , F.testCase "t197" t197

8293 ]

tests/text-tests.cabal+18−3

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

2020 default: False

2121 manual: True

2222

23+flag bytestring-builder

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

25+ default: False

26+ manual: False

27+

2328executable text-tests

2429 main-is: Tests.hs

2530

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

4853 HUnit >= 1.2,

4954 QuickCheck >= 2.7,

5055 base == 4.*,

51− bytestring,

5256 deepseq,

5357 directory,

54− quickcheck-unicode,

58+ quickcheck-unicode >= 1.0.1.0,

5559 random,

5660 test-framework >= 0.4,

5761 test-framework-hunit >= 0.2,

5862 test-framework-quickcheck2 >= 0.2,

5963 text-tests

6064

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+

6171executable text-tests-stdio

6272 main-is: Tests/IO.hs

6373

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

137147 array,

138148 base == 4.*,

139149 binary,

140− bytestring,

141150 deepseq,

142151 ghc-prim,

143152 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+100−33

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

11name: text

2−version: 1.2.2.1

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

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

2+version: 1.2.3.0

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

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

55synopsis: An efficient packed Unicode text type.

66description:

77 .

@@ -14,11 +14,13 @@ description:

1414 in terms of large data quantities and high speed.

1515 .

1616 The 'Text' type provides character-encoding, type-safe case

17− conversion via whole-string case conversion functions. It also

18− provides a range of functions for converting 'Text' values to and from

19− 'ByteStrings', using several standard encodings.

17+ conversion via whole-string case conversion functions (see "Data.Text").

18+ It also provides a range of functions for converting 'Text' values to

19+ and from 'ByteStrings', using several standard encodings

20+ (see "Data.Text.Encoding").

2021 .

21− Efficient locale-sensitive support for text IO is also supported.

22+ Efficient locale-sensitive support for text IO is also supported

23+ (see "Data.Text.IO").

2224 .

2325 These modules are intended to be imported qualified, to avoid name

2426 clashes with Prelude functions, e.g.

@@ -28,8 +30,9 @@ description:

2830 To use an extended and very rich family of functions for working

2931 with Unicode text (including normalization, regular expressions,

3032 non-standard encodings, text breaking, and locales), see

31− the @text-icu@ package:

32− <http://hackage.haskell.org/package/text-icu>

33+ the [text-icu package](https://hackage.haskell.org/package/text-icu)

34+ based on the well-respected and liberally

35+ licensed [ICU library](http://site.icu-project.org/).

3336

3437license: BSD2

3538license-file: LICENSE

3942category: Data, Text

4043build-type: Simple

4144cabal-version: >= 1.8

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

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

4247extra-source-files:

4348 -- scripts/CaseFolding.txt

4449 -- scripts/SpecialCasing.txt

@@ -65,7 +70,9 @@ extra-source-files:

6570 tests/text-tests.cabal

6671

6772flag bytestring-builder

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

73+ description:

74+ Depend on the [bytestring-builder](https://hackage.haskell.org/package/bytestring-builder)

75+ package for backwards compatibility.

6976 default: False

7077 manual: False

7178

@@ -75,7 +82,9 @@ flag developer

7582 manual: True

7683

7784flag integer-simple

78− description: Use the simple integer library instead of GMP

85+ description:

86+ Use the [simple integer library](http://hackage.haskell.org/package/integer-simple)

87+ instead of [integer-gmp](http://hackage.haskell.org/package/integer-gmp)

7988 default: False

8089 manual: False

8190

@@ -88,13 +97,13 @@ library

8897 Data.Text.Array

8998 Data.Text.Encoding

9099 Data.Text.Encoding.Error

91− -- Data.Text.Foreign

100+ Data.Text.Foreign

92101 Data.Text.IO

93102 Data.Text.Internal

94103 Data.Text.Internal.Builder

95− -- Data.Text.Internal.Builder.Functions

96− -- Data.Text.Internal.Builder.Int.Digits

97− -- Data.Text.Internal.Builder.RealFloat.Functions

104+ Data.Text.Internal.Builder.Functions

105+ Data.Text.Internal.Builder.Int.Digits

106+ Data.Text.Internal.Builder.RealFloat.Functions

98107 Data.Text.Internal.Encoding.Fusion

99108 Data.Text.Internal.Encoding.Fusion.Common

100109 Data.Text.Internal.Encoding.Utf16

@@ -106,14 +115,14 @@ library

106115 Data.Text.Internal.Fusion.Common

107116 Data.Text.Internal.Fusion.Size

108117 Data.Text.Internal.Fusion.Types

109− -- Data.Text.Internal.IO

118+ Data.Text.Internal.IO

110119 Data.Text.Internal.Lazy

111− -- Data.Text.Internal.Lazy.Encoding.Fusion

120+ Data.Text.Internal.Lazy.Encoding.Fusion

112121 Data.Text.Internal.Lazy.Fusion

113122 Data.Text.Internal.Lazy.Search

114123 Data.Text.Internal.Private

115124 Data.Text.Internal.Read

116− -- Data.Text.Internal.Search

125+ Data.Text.Internal.Search

117126 Data.Text.Internal.Unsafe

118127 Data.Text.Internal.Unsafe.Char

119128 Data.Text.Internal.Unsafe.Shift

@@ -124,23 +133,18 @@ library

124133 Data.Text.Lazy.Encoding

125134 Data.Text.Lazy.IO

126135 Data.Text.Lazy.Internal

127− -- Data.Text.Lazy.Read

136+ Data.Text.Lazy.Read

128137 Data.Text.Read

129138 Data.Text.Unsafe

130139

131− -- moved from ghcjs-base

132− -- reflex-dom depends in it

133− Data.JSString.Text

134− -- other-modules:

135− -- Data.Text.Show

140+ other-modules:

141+ Data.Text.Show

136142

137143 build-depends:

138− -- array >= 0.3,

144+ array >= 0.3,

139145 base >= 4.2 && < 5,

140146 binary,

141147 deepseq >= 1.1.0.0,

142− ghcjs-base,

143− ghcjs-prim,

144148 ghc-prim >= 0.2

145149

146150 if flag(bytestring-builder)

@@ -165,27 +169,90 @@ library

165169

166170-- test-suite tests

167171-- type: exitcode-stdio-1.0

168−-- hs-source-dirs: tests .

169−-- main-is: Tests.hs

170172-- c-sources: cbits/cbits.c

171173-- include-dirs: include

172174

173175-- ghc-options:

174−-- -Wall -threaded -O0 -rtsopts

176+-- -Wall -threaded -rtsopts

175177

176178-- cpp-options:

177179-- -DASSERTS -DHAVE_DEEPSEQ -DTEST_SUITE

178180

181+-- -- modules specific to test-suite

182+-- hs-source-dirs: tests

183+-- main-is: Tests.hs

184+-- other-modules:

185+-- Tests.Properties

186+-- Tests.Properties.Mul

187+-- Tests.QuickCheckUtils

188+-- Tests.Regressions

189+-- Tests.SlowFunctions

190+-- Tests.Utils

191+

192+-- -- Same as in `library` stanza; this is needed by cabal for accurate

193+-- -- file-monitoring as well as to avoid `-Wmissing-home-modules`

194+-- -- warnings We can't use an inter-package library dependency because

195+-- -- of different `ghc-options`/`cpp-options` (as a side-benefitt,

196+-- -- this enables per-component build parallelism in `cabal

197+-- -- new-build`!); We could, however, use cabal-version:2.2's `common`

198+-- -- blocks at some point in the future to reduce the duplication.

199+-- hs-source-dirs: .

200+-- other-modules:

201+-- Data.Text

202+-- Data.Text.Array

203+-- Data.Text.Encoding

204+-- Data.Text.Encoding.Error

205+-- Data.Text.Foreign

206+-- Data.Text.IO

207+-- Data.Text.Internal

208+-- Data.Text.Internal.Builder

209+-- Data.Text.Internal.Builder.Functions

210+-- Data.Text.Internal.Builder.Int.Digits

211+-- Data.Text.Internal.Builder.RealFloat.Functions

212+-- Data.Text.Internal.Encoding.Fusion

213+-- Data.Text.Internal.Encoding.Fusion.Common

214+-- Data.Text.Internal.Encoding.Utf16

215+-- Data.Text.Internal.Encoding.Utf32

216+-- Data.Text.Internal.Encoding.Utf8

217+-- Data.Text.Internal.Functions

218+-- Data.Text.Internal.Fusion

219+-- Data.Text.Internal.Fusion.CaseMapping

220+-- Data.Text.Internal.Fusion.Common

221+-- Data.Text.Internal.Fusion.Size

222+-- Data.Text.Internal.Fusion.Types

223+-- Data.Text.Internal.IO

224+-- Data.Text.Internal.Lazy

225+-- Data.Text.Internal.Lazy.Encoding.Fusion

226+-- Data.Text.Internal.Lazy.Fusion

227+-- Data.Text.Internal.Lazy.Search

228+-- Data.Text.Internal.Private

229+-- Data.Text.Internal.Read

230+-- Data.Text.Internal.Search

231+-- Data.Text.Internal.Unsafe

232+-- Data.Text.Internal.Unsafe.Char

233+-- Data.Text.Internal.Unsafe.Shift

234+-- Data.Text.Lazy

235+-- Data.Text.Lazy.Builder

236+-- Data.Text.Lazy.Builder.Int

237+-- Data.Text.Lazy.Builder.RealFloat

238+-- Data.Text.Lazy.Encoding

239+-- Data.Text.Lazy.IO

240+-- Data.Text.Lazy.Internal

241+-- Data.Text.Lazy.Read

242+-- Data.Text.Read

243+-- Data.Text.Unsafe

244+-- Data.Text.Show

245+

179246-- build-depends:

180247-- HUnit >= 1.2,

181−-- QuickCheck >= 2.7,

248+-- QuickCheck >= 2.7 && < 2.10,

182249-- array,

183250-- base,

184251-- binary,

185252-- deepseq,

186253-- directory,

187254-- ghc-prim,

188−-- quickcheck-unicode,

255+-- quickcheck-unicode >= 1.0.1.0,

189256-- random,

190257-- test-framework >= 0.4,

191258-- test-framework-hunit >= 0.2,

@@ -206,7 +273,7 @@ library

206273

207274-- source-repository head

208275-- type: git

209−-- location: https://github.com/bos/text

276+-- location: https://github.com/haskell/text

210277

211278-- source-repository head

212279-- type: mercurial