master...a02c2da · text

git.iko.soy/text

39 files changed+1075−251

since splitsnapshots

unifiedsplit

.gitignore+10−3 .hgtags+1 .travis.yml+111 Data/Text.hs+177−46 Data/Text/Array.hs+4−1 Data/Text/Encoding.hs+17−4 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/Common.hs+44−37 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+3−1 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+56−19 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+90−17

.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+177−46

@@ -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

@@ -246,6 +250,18 @@ import qualified GHC.Exts as Exts

246250import Text.Printf (PrintfArg, formatArg, formatString)

247251#endif

248252

253+-- $character_definition

254+--

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

256+--

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

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

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

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

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

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

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

264+

249265-- $strict

250266--

251267-- This package provides both strict and lazy 'Text' types. The

@@ -334,9 +350,13 @@ instance Read Text where

334350 readsPrec p str = [(pack x,y) | (x,y) <- readsPrec p str]

335351

336352#if MIN_VERSION_base(4,9,0)

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

338−-- 'semigroups` package

339−

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

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

355+-- provided by

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

357+-- package

358+--

359+-- @since 1.2.2.0

340360instance Semigroup Text where

341361 (<>) = append

342362#endif

@@ -354,6 +374,7 @@ instance IsString Text where

354374 fromString = pack

355375

356376#if __GLASGOW_HASKELL__ >= 708

377+-- | @since 1.2.0.0

357378instance Exts.IsList Text where

358379 type Item Text = Char

359380 fromList = pack

@@ -364,6 +385,7 @@ instance Exts.IsList Text where

364385instance NFData Text where rnf !_ = ()

365386#endif

366387

388+-- | @since 1.2.1.0

367389instance Binary Text where

368390 put t = put (encodeUtf8 t)

369391 get = do

@@ -397,6 +419,8 @@ instance Data Text where

397419

398420#if MIN_VERSION_base(4,7,0)

399421-- | Only defined for @base-4.7.0.0@ and later

422+--

423+-- @since 1.2.2.0

400424instance PrintfArg Text where

401425 formatArg txt = formatString $ unpack txt

402426#endif

@@ -545,6 +569,19 @@ init (Text arr off len) | len <= 0 = emptyError "init"

545569 unstream (S.init (stream t)) = init t

546570 #-}

547571

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

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

574+--

575+-- @since 1.2.3.0

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

577+unsnoc (Text arr off len)

578+ | len <= 0 = Nothing

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

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

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

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

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

584+

548585-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

549586-- fusion.

550587null :: Text -> Bool

@@ -624,8 +661,15 @@ compareLength t n = S.compareLengthI (stream t) n

624661-- -----------------------------------------------------------------------------

625662-- * Transformations

626663-- | /O(n)/ 'map' @f@ @t@ is the 'Text' obtained by applying @f@ to

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

628−-- invalid scalar values.

664+-- each element of @t@.

665+--

666+-- Example:

667+--

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

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

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

671+--

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

629673map :: (Char -> Char) -> Text -> Text

630674map f t = unstream (S.map (safe . f) (stream t))

631675{-# INLINE [1] map #-}

@@ -633,18 +677,36 @@ map f t = unstream (S.map (safe . f) (stream t))

633677-- | /O(n)/ The 'intercalate' function takes a 'Text' and a list of

634678-- 'Text's and concatenates the list after interspersing the first

635679-- argument between each element of the list.

680+--

681+-- Example:

682+--

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

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

636685intercalate :: Text -> [Text] -> Text

637686intercalate t = concat . (F.intersperse t)

638687{-# INLINE intercalate #-}

639688

640689-- | /O(n)/ The 'intersperse' function takes a character and places it

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

642−-- replacement on invalid scalar values.

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

691+--

692+-- Example:

693+--

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

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

696+--

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

643698intersperse :: Char -> Text -> Text

644699intersperse c t = unstream (S.intersperse (safe c) (stream t))

645700{-# INLINE intersperse #-}

646701

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

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

703+--

704+-- Example:

705+--

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

707+-- "reversed"

708+--

709+-- Subject to fusion.

648710reverse :: Text -> Text

649711reverse t = S.reverse (stream t)

650712{-# INLINE reverse #-}

@@ -663,12 +725,14 @@ reverse t = S.reverse (stream t)

663725-- @needle@ occurs in @replacement@, that occurrence will /not/ itself

664726-- be replaced recursively:

665727--

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

728+-- >>> replace "oo" "foo" "oo"

729+-- "foo"

667730--

668731-- In cases where several instances of @needle@ overlap, only the

669732-- first one will be replaced:

670733--

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

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

735+-- "barfo"

672736--

673737-- In (unlikely) bad cases, this function's time complexity degrades

674738-- towards /O(n*m)/.

@@ -782,6 +846,8 @@ toUpper t = unstream (S.toUpper (stream t))

782846-- guides disagree on whether the book name \"The Hill of the Red

783847-- Fox\" is correctly title cased—but this function will

784848-- capitalize /every/ word.

849+--

850+-- @since 1.0.0.0

785851toTitle :: Text -> Text

786852toTitle t = unstream (S.toTitle (stream t))

787853{-# INLINE toTitle #-}

@@ -792,8 +858,11 @@ toTitle t = unstream (S.toTitle (stream t))

792858--

793859-- Examples:

794860--

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

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

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

862+-- "fooxxxx"

863+--

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

865+-- "foobar"

797866justifyLeft :: Int -> Char -> Text -> Text

798867justifyLeft k c t

799868 | len >= k = t

@@ -814,8 +883,11 @@ justifyLeft k c t

814883--

815884-- Examples:

816885--

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

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

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

887+-- "xxxxbar"

888+--

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

890+-- "foobar"

819891justifyRight :: Int -> Char -> Text -> Text

820892justifyRight k c t

821893 | len >= k = t

@@ -829,7 +901,8 @@ justifyRight k c t

829901--

830902-- Examples:

831903--

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

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

905+-- "xxxHSxxx"

833906center :: Int -> Char -> Text -> Text

834907center k c t

835908 | len >= k = t

@@ -844,6 +917,14 @@ center k c t

844917-- of its 'Text' argument. Note that this function uses 'pack',

845918-- 'unpack', and the list version of transpose, and is thus not very

846919-- efficient.

920+--

921+-- Examples:

922+--

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

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

925+--

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

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

847928transpose :: [Text] -> [Text]

848929transpose ts = P.map pack (L.transpose (P.map unpack ts))

849930

@@ -1088,7 +1169,10 @@ iterN n t@(Text _arr _off len) = loop 0 0

10881169--

10891170-- Examples:

10901171--

1091−-- > takeEnd 3 "foobar" == "bar"

1172+-- >>> takeEnd 3 "foobar"

1173+-- "bar"

1174+--

1175+-- @since 1.1.1.0

10921176takeEnd :: Int -> Text -> Text

10931177takeEnd n t@(Text arr off len)

10941178 | n <= 0 = empty

@@ -1099,8 +1183,8 @@ takeEnd n t@(Text arr off len)

10991183iterNEnd :: Int -> Text -> Int

11001184iterNEnd n t@(Text _arr _off len) = loop (len-1) n

11011185 where loop i !m

1186+ | m <= 0 = i+1

11021187 | i <= 0 = 0

1103− | m <= 1 = i

11041188 | otherwise = loop (i+d) (m-1)

11051189 where d = reverseIter_ t i

11061190

@@ -1127,7 +1211,10 @@ drop n t@(Text arr off len)

11271211--

11281212-- Examples:

11291213--

1130−-- > dropEnd 3 "foobar" == "foo"

1214+-- >>> dropEnd 3 "foobar"

1215+-- "foo"

1216+--

1217+-- @since 1.1.1.0

11311218dropEnd :: Int -> Text -> Text

11321219dropEnd n t@(Text arr off len)

11331220 | n <= 0 = t

@@ -1157,7 +1244,10 @@ takeWhile p t@(Text arr off len) = loop 0

11571244-- satisfy @p@. Subject to fusion.

11581245-- Examples:

11591246--

1160−-- > takeWhileEnd (=='o') "foo" == "oo"

1247+-- >>> takeWhileEnd (=='o') "foo"

1248+-- "oo"

1249+--

1250+-- @since 1.2.2.0

11611251takeWhileEnd :: (Char -> Bool) -> Text -> Text

11621252takeWhileEnd p t@(Text arr off len) = loop (len-1) len

11631253 where loop !i !l | l <= 0 = t

@@ -1196,7 +1286,8 @@ dropWhile p t@(Text arr off len) = loop 0 0

11961286--

11971287-- Examples:

11981288--

1199−-- > dropWhileEnd (=='.') "foo..." == "foo"

1289+-- >>> dropWhileEnd (=='.') "foo..."

1290+-- "foo"

12001291dropWhileEnd :: (Char -> Bool) -> Text -> Text

12011292dropWhileEnd p t@(Text arr off len) = loop (len-1) len

12021293 where loop !i !l | l <= 0 = empty

@@ -1314,9 +1405,14 @@ tails t | null t = [empty]

13141405--

13151406-- Examples:

13161407--

1317−-- > splitOn "\r\n" "a\r\nb\r\nd\r\ne" == ["a","b","d","e"]

1318−-- > splitOn "aaa" "aaaXaaaXaaaXaaa" == ["","X","X","X",""]

1319−-- > splitOn "x" "x" == ["",""]

1408+-- >>> splitOn "\r\n" "a\r\nb\r\nd\r\ne"

1409+-- ["a","b","d","e"]

1410+--

1411+-- >>> splitOn "aaa" "aaaXaaaXaaaXaaa"

1412+-- ["","X","X","X",""]

1413+--

1414+-- >>> splitOn "x" "x"

1415+-- ["",""]

13201416--

13211417-- and

13221418--

@@ -1352,8 +1448,11 @@ splitOn pat@(Text _ _ l) src@(Text arr off len)

13521448-- resulting components do not contain the separators. Two adjacent

13531449-- separators result in an empty component in the output. eg.

13541450--

1355−-- > split (=='a') "aabbaca" == ["","","bb","c",""]

1356−-- > split (=='a') "" == [""]

1451+-- >>> split (=='a') "aabbaca"

1452+-- ["","","bb","c",""]

1453+--

1454+-- >>> split (=='a') ""

1455+-- [""]

13571456split :: (Char -> Bool) -> Text -> [Text]

13581457split _ t@(Text _off _arr 0) = [t]

13591458split p t = loop t

@@ -1366,8 +1465,11 @@ split p t = loop t

13661465-- element may be shorter than the other chunks, depending on the

13671466-- length of the input. Examples:

13681467--

1369−-- > chunksOf 3 "foobarbaz" == ["foo","bar","baz"]

1370−-- > chunksOf 4 "haskell.org" == ["hask","ell.","org"]

1468+-- >>> chunksOf 3 "foobarbaz"

1469+-- ["foo","bar","baz"]

1470+--

1471+-- >>> chunksOf 4 "haskell.org"

1472+-- ["hask","ell.","org"]

13711473chunksOf :: Int -> Text -> [Text]

13721474chunksOf k = go

13731475 where

@@ -1412,8 +1514,11 @@ filter p t = unstream (S.filter p (stream t))

14121514--

14131515-- Examples:

14141516--

1415−-- > breakOn "::" "a::b::c" ==> ("a", "::b::c")

1416−-- > breakOn "/" "foobar" ==> ("foobar", "")

1517+-- >>> breakOn "::" "a::b::c"

1518+-- ("a","::b::c")

1519+--

1520+-- >>> breakOn "/" "foobar"

1521+-- ("foobar","")

14171522--

14181523-- Laws:

14191524--

@@ -1441,7 +1546,8 @@ breakOn pat src@(Text arr off len)

14411546-- up to and including the last match of @needle@. The second is the

14421547-- remainder of @haystack@, following the match.

14431548--

1444−-- > breakOnEnd "::" "a::b::c" ==> ("a::b::", "c")

1549+-- >>> breakOnEnd "::" "a::b::c"

1550+-- ("a::b::","c")

14451551breakOnEnd :: Text -> Text -> (Text, Text)

14461552breakOnEnd pat src = (reverse b, reverse a)

14471553 where (a,b) = breakOn (reverse pat) (reverse src)

@@ -1456,10 +1562,11 @@ breakOnEnd pat src = (reverse b, reverse a)

14561562--

14571563-- Examples:

14581564--

1459−-- > breakOnAll "::" ""

1460−-- > ==> []

1461−-- > breakOnAll "/" "a/b/c/"

1462−-- > ==> [("a", "/b/c/"), ("a/b", "/c/"), ("a/b/c", "/")]

1565+-- >>> breakOnAll "::" ""

1566+-- []

1567+--

1568+-- >>> breakOnAll "/" "a/b/c/"

1569+-- [("a","/b/c/"),("a/b","/c/"),("a/b/c","/")]

14631570--

14641571-- In (unlikely) bad cases, this function's time complexity degrades

14651572-- towards /O(n*m)/.

@@ -1658,9 +1765,14 @@ isInfixOf needle haystack

16581765--

16591766-- Examples:

16601767--

1661−-- > stripPrefix "foo" "foobar" == Just "bar"

1662−-- > stripPrefix "" "baz" == Just "baz"

1663−-- > stripPrefix "foo" "quux" == Nothing

1768+-- >>> stripPrefix "foo" "foobar"

1769+-- Just "bar"

1770+--

1771+-- >>> stripPrefix "" "baz"

1772+-- Just "baz"

1773+--

1774+-- >>> stripPrefix "foo" "quux"

1775+-- Nothing

16641776--

16651777-- This is particularly useful with the @ViewPatterns@ extension to

16661778-- GHC, as follows:

@@ -1685,9 +1797,14 @@ stripPrefix p@(Text _arr _off plen) t@(Text arr off len)

16851797--

16861798-- Examples:

16871799--

1688−-- > commonPrefixes "foobar" "fooquux" == Just ("foo","bar","quux")

1689−-- > commonPrefixes "veeble" "fetzer" == Nothing

1690−-- > commonPrefixes "" "baz" == Nothing

1800+-- >>> commonPrefixes "foobar" "fooquux"

1801+-- Just ("foo","bar","quux")

1802+--

1803+-- >>> commonPrefixes "veeble" "fetzer"

1804+-- Nothing

1805+--

1806+-- >>> commonPrefixes "" "baz"

1807+-- Nothing

16911808commonPrefixes :: Text -> Text -> Maybe (Text,Text,Text)

16921809commonPrefixes t0@(Text arr0 off0 len0) t1@(Text arr1 off1 len1) = go 0 0

16931810 where

@@ -1704,9 +1821,14 @@ commonPrefixes t0@(Text arr0 off0 len0) t1@(Text arr1 off1 len1) = go 0 0

17041821--

17051822-- Examples:

17061823--

1707−-- > stripSuffix "bar" "foobar" == Just "foo"

1708−-- > stripSuffix "" "baz" == Just "baz"

1709−-- > stripSuffix "foo" "quux" == Nothing

1824+-- >>> stripSuffix "bar" "foobar"

1825+-- Just "foo"

1826+--

1827+-- >>> stripSuffix "" "baz"

1828+-- Just "baz"

1829+--

1830+-- >>> stripSuffix "foo" "quux"

1831+-- Nothing

17101832--

17111833-- This is particularly useful with the @ViewPatterns@ extension to

17121834-- GHC, as follows:

@@ -1753,3 +1875,12 @@ copy (Text arr off len) = Text (A.run go) 0 len

17531875 marr <- A.new len

17541876 A.copyI marr 0 arr off len

17551877 return marr

1878+

1879+

1880+-------------------------------------------------

1881+-- NOTE: the named chunk below used by doctest;

1882+-- verify the doctests via `doctest -fobject-code Data/Text.hs`

1883+

1884+-- $setup

1885+-- >>> :set -XOverloadedStrings

1886+-- >>> 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+17−4

@@ -11,14 +11,13 @@

1111--

1212-- License : BSD-style

1313-- Maintainer : bos@serpentine.com

14−-- Stability : experimental

1514-- Portability : portable

1615--

1716-- Functions for converting 'Text' values to and from 'ByteString',

1817-- using several standard encodings.

1918--

2019-- To gain access to a much larger family of encodings, use the

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

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

2221

2322module Data.Text.Encoding

2423 (

@@ -79,7 +78,11 @@ import Data.Text.Internal.Unsafe.Shift (shiftR)

7978import Data.Text.Show ()

8079import Data.Text.Unsafe (unsafeDupablePerformIO)

8180import Data.Word (Word8, Word32)

82−import Foreign.C.Types (CSize(..))

81+#if __GLASGOW_HASKELL__ >= 703

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

83+#else

84+import Foreign.C.Types (CSize)

85+#endif

8386import Foreign.ForeignPtr (withForeignPtr)

8487import Foreign.Marshal.Utils (with)

8588import Foreign.Ptr (Ptr, minusPtr, nullPtr, plusPtr)

@@ -209,6 +212,8 @@ decodeUtf8With onErr (PS fp off len) = runText $ \done -> do

209212-- or continuation where it is encountered.

210213

211214-- | A stream oriented decoding result.

215+--

216+-- @since 1.0.0.0

212217data Decoding = Some Text ByteString (ByteString -> Decoding)

213218

214219instance Show Decoding where

@@ -228,11 +233,15 @@ newtype DecoderState = DecoderState Word32 deriving (Eq, Show, Num, Storable)

228233-- thrown (either by this function or a continuation) that cannot be

229234-- caught in pure code. For more control over the handling of invalid

230235-- data, use 'streamDecodeUtf8With'.

236+--

237+-- @since 1.0.0.0

231238streamDecodeUtf8 :: ByteString -> Decoding

232239streamDecodeUtf8 = streamDecodeUtf8With strictDecode

233240

234241-- | Decode, in a stream oriented way, a 'ByteString' containing UTF-8

235242-- encoded text.

243+--

244+-- @since 1.0.0.0

236245streamDecodeUtf8With :: OnDecodeError -> ByteString -> Decoding

237246streamDecodeUtf8With onErr = decodeChunk B.empty 0 0

238247 where

@@ -308,6 +317,8 @@ decodeUtf8' = unsafeDupablePerformIO . try . evaluate . decodeUtf8With strictDec

308317{-# INLINE decodeUtf8' #-}

309318

310319-- | Encode text to a ByteString 'B.Builder' using UTF-8 encoding.

320+--

321+-- @since 1.1.0.0

311322encodeUtf8Builder :: Text -> B.Builder

312323encodeUtf8Builder = encodeUtf8BuilderEscaped (BP.liftFixedToBounded BP.word8)

313324

@@ -316,6 +327,8 @@ encodeUtf8Builder = encodeUtf8BuilderEscaped (BP.liftFixedToBounded BP.word8)

316327--

317328-- Use this function is to implement efficient encoders for text-based formats

318329-- like JSON or HTML.

330+--

331+-- @since 1.1.0.0

319332{-# INLINE encodeUtf8BuilderEscaped #-}

320333-- TODO: Extend documentation with references to source code in @blaze-html@

321334-- or @aeson@ that uses this function.

@@ -373,7 +386,7 @@ encodeUtf8 :: Text -> ByteString

373386encodeUtf8 (Text arr off len)

374387 | len == 0 = B.empty

375388 | otherwise = unsafeDupablePerformIO $ do

376− fp <- mallocByteString (len*4)

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

377390 withForeignPtr fp $ \ptr ->

378391 with ptr $ \destPtr -> do

379392 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.

@@ -336,3 +335,7 @@ putStrLn = hPutStrLn stdout

336335-- >

337336-- > putStr_Utf16LE :: Text -> IO ()

338337-- > putStr_Utf16LE t = B.putStr (encodeUtf16LE t)

338+--

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

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

341+-- 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/Common.hs+44−37

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

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

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

22-- |

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

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

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

107107import qualified Data.List as L

108108import qualified Prelude as P

109109import Data.Bits (shiftL)

110−import Data.Char (isLetter)

110+import Data.Char (isLetter, isSpace)

111111import Data.Int (Int64)

112112import Data.Text.Internal.Fusion.Types

113113import Data.Text.Internal.Fusion.CaseMapping (foldMapping, lowerMapping, titleMapping,

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

117117import GHC.Types (Char(..), Int(..))

118118

119119singleton :: Char -> Stream Char

120−singleton c = Stream next False 1

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

121121 where next False = Yield c True

122122 next True = Done

123123{-# INLINE [0] singleton #-}

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

175175

176176-- | /O(n)/ Adds a character to the front of a Stream Char.

177177cons :: Char -> Stream Char -> Stream Char

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

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

179179 where

180180 next (C1 s) = Yield w (C0 s)

181181 next (C0 s) = case next0 s of

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

189189

190190-- | /O(n)/ Adds a character to the end of a stream.

191191snoc :: Stream Char -> Char -> Stream Char

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

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

193193 where

194194 next (J xs) = case next0 xs of

195195 Done -> Yield w N

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

237237uncons (Stream next s0 len) = loop_uncons s0

238238 where

239239 loop_uncons !s = case next s of

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

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

241241 Skip s' -> loop_uncons s'

242242 Done -> Nothing

243243{-# INLINE [0] uncons #-}

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

260260-- | /O(1)/ Returns all characters after the head of a Stream Char, which must

261261-- be non-empty.

262262tail :: Stream Char -> Stream Char

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

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

264264 where

265265 next (C0 s) = case next0 s of

266266 Done -> emptyError "tail"

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

278278-- | /O(1)/ Returns all but the last character of a Stream Char, which

279279-- must be non-empty.

280280init :: Stream Char -> Stream Char

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

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

282282 where

283283 next (Init0 s) = case next0 s of

284284 Done -> emptyError "init"

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

318318-- greater than the number or if the stream can't possibly be as long

319319-- as the number supplied, and hence be more efficient.

320320compareLengthI :: Integral a => Stream Char -> a -> Ordering

321−compareLengthI (Stream next s0 len) n =

322− case compareSize len (fromIntegral n) of

323− Just o -> o

324− Nothing -> loop_cmp 0 s0

321+compareLengthI (Stream next s0 len) n

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

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

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

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

326+ | otherwise = loop_cmp 0 s0

325327 where

328+ n' = codePointsSize $ fromIntegral n

326329 loop_cmp !z s = case next s of

327330 Done -> compare z n

328331 Skip s' -> loop_cmp z s'

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

368371-- | /O(n)/ Take a character and place it between each of the

369372-- characters of a 'Stream Char'.

370373intersperse :: Char -> Stream Char -> Stream Char

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

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

372375 where

373376 next (I1 s) = case next0 s of

374377 Done -> Done

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

393396-- functions may map one input character to two or three output

394397-- characters.

395398

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

396400caseConvert :: (forall s. Char -> s -> Step (CC s) Char)

397401 -> Stream Char -> Stream Char

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

402+caseConvert remap (Stream next0 s0 len) =

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

399404 where

400405 next (CC s '\0' _) =

401406 case next0 s of

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

458463-- Fox\" is correctly title cased—but this function will

459464-- capitalize /every/ word.

460465toTitle :: Stream Char -> Stream Char

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

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

462467 where

463468 next (CC (letter :*: s) '\0' _) =

464469 case next0 s of

465− Done -> Done

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

470+ Done -> Done

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

467472 Yield c s'

468− | letter' -> if letter

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

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

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

472− where letter' = isLetter c

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

473+ | nonSpace -> if letter

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

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

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

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

478+ letter' = isLetter c

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

474480{-# INLINE [0] toTitle #-}

475481

476482data Justify i s = Just1 !i !s

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

478484

479485justifyLeftI :: Integral a => a -> Char -> Stream Char -> Stream Char

480486justifyLeftI k c (Stream next0 s0 len) =

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

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

482488 where

483489 next (Just1 n s) =

484490 case next0 s of

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

698704-- (a,b), in which case, a is the next Char in the string, and b is

699705-- the seed value for further production.

700706unfoldr :: (a -> Maybe (Char,a)) -> a -> Stream Char

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

707+unfoldr f s0 = Stream next s0 unknownSize

702708 where

703709 {-# INLINE next #-}

704710 next !s = case f s of

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

712718-- 'unfoldr' when the length of the result is known.

713719unfoldrNI :: Integral a => a -> (b -> Maybe (Char,b)) -> b -> Stream Char

714720unfoldrNI n f s0 | n < 0 = empty

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

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

716722 where

717723 {-# INLINE next #-}

718724 next (z :*: s) = case f s of

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

724730-------------------------------------------------------------------------------

725731-- * Substreams

726732

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

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

728734-- stream of length @n@, or the stream itself if @n@ is greater than the

729735-- length of the stream.

730736take :: Integral a => a -> Stream Char -> Stream Char

731737take n0 (Stream next0 s0 len) =

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

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

733739 where

734740 {-# INLINE next #-}

735741 next (n :*: s) | n <= 0 = Done

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

742748data Drop a s = NS !s

743749 | JS !a !s

744750

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

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

746752-- stream after the first @n@ characters, or the empty stream if @n@

747753-- is greater than the length of the stream.

748754drop :: Integral a => a -> Stream Char -> Stream Char

749755drop n0 (Stream next0 s0 len) =

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

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

751757 where

752758 {-# INLINE next #-}

753759 next (JS n s)

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

762768 Yield x s' -> Yield x (NS s')

763769{-# INLINE [0] drop #-}

764770

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

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

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

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

767773takeWhile :: (Char -> Bool) -> Stream Char -> Stream Char

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

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

769775 where

770776 {-# INLINE next #-}

771777 next !s = case next0 s of

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

775781 | otherwise -> Done

776782{-# INLINE [0] takeWhile #-}

777783

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

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

779785dropWhile :: (Char -> Bool) -> Stream Char -> Stream Char

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

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

781787 where

782788 {-# INLINE next #-}

783789 next (L s) = case next0 s of

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

811817-------------------------------------------------------------------------------

812818-- ** Searching by equality

813819

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

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

815821elem :: Char -> Stream Char -> Bool

816822elem w (Stream next s0 _len) = loop_elem s0

817823 where

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

856862-- returns a stream containing those characters that satisfy the

857863-- predicate.

858864filter :: (Char -> Bool) -> Stream Char -> Stream Char

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

865+filter p (Stream next0 s0 len) =

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

860867 where

861868 next !s = case next0 s of

862869 Done -> Done

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

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

7575-- unstreaming functions must be able to cope with the hint being too

7676-- small or too large.

7777--

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

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

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

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

8080-- undercount if, for instance, a transformed stream contains astral

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

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

8282

8383data Stream a =

8484 forall s. Stream

8585 (s -> Step s a) -- stepper function

8686 !s -- current state

87− !Size -- size hint

87+ !Size -- size hint in code units

8888

8989-- | /O(n)/ Determines if two streams are equal.

9090eq :: (Eq a) => Stream a -> Stream a -> Bool

Data/Text/Lazy.hs+54−20

@@ -13,7 +13,6 @@

1313--

1414-- License : BSD-style

1515-- Maintainer : bos@serpentine.com

16−-- Stability : experimental

1716-- Portability : GHC

1817--

1918-- A time and space-efficient implementation of Unicode text using

@@ -68,6 +67,7 @@ module Data.Text.Lazy

6867 , snoc

6968 , append

7069 , uncons

70+ , unsnoc

7171 , head

7272 , last

7373 , tail

@@ -273,14 +273,16 @@ import Text.Printf (PrintfArg, formatArg, formatString)

273273-- $replacement

274274--

275275-- A 'Text' value is a sequence of Unicode scalar values, as defined

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

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

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

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

280−-- (§3.4, definition D10) as 'Char' values, including code points

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

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

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

276+-- in

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

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

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

280+-- points

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

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

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

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

285+-- those cases.

284286--

285287-- Within this module, many functions construct a 'Text' from one or

286288-- more 'Char' values. Those functions will substitute 'Char' values

@@ -294,8 +296,8 @@ import Text.Printf (PrintfArg, formatArg, formatString)

294296-- range U+D800 through U+DFFF are used by UTF-16 to denote surrogate

295297-- code points, and so cannot be represented. The functions replace

296298-- invalid scalar values, instead of dropping them, as a security

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

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

299+-- measure. For details, see

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

299301

300302equal :: Text -> Text -> Bool

301303equal Empty Empty = True

@@ -342,9 +344,13 @@ instance Read Text where

342344 readsPrec p str = [(pack x,y) | (x,y) <- readsPrec p str]

343345

344346#if MIN_VERSION_base(4,9,0)

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

346−-- 'semigroups` package

347−

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

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

349+-- provided by

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

351+-- package

352+--

353+-- @since 1.2.2.0

348354instance Semigroup Text where

349355 (<>) = append

350356#endif

@@ -362,6 +368,7 @@ instance IsString Text where

362368 fromString = pack

363369

364370#if __GLASGOW_HASKELL__ >= 708

371+-- | @since 1.2.0.0

365372instance Exts.IsList Text where

366373 type Item Text = Char

367374 fromList = pack

@@ -374,6 +381,7 @@ instance NFData Text where

374381 rnf (Chunk _ ts) = rnf ts

375382#endif

376383

384+-- | @since 1.2.1.0

377385instance Binary Text where

378386 put t = put (encodeUtf8 t)

379387 get = do

@@ -397,6 +405,8 @@ instance Data Text where

397405

398406#if MIN_VERSION_base(4,7,0)

399407-- | Only defined for @base-4.7.0.0@ and later

408+--

409+-- @since 1.2.2.0

400410instance PrintfArg Text where

401411 formatArg txt = formatString $ unpack txt

402412#endif

@@ -475,9 +485,7 @@ fromStrict t = chunk t Empty

475485-- -----------------------------------------------------------------------------

476486-- * Basic functions

477487

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

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

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

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

481489cons :: Char -> Text -> Text

482490cons c t = Chunk (T.singleton c) t

483491{-# INLINE [1] cons #-}

@@ -561,6 +569,17 @@ init Empty = emptyError "init"

561569 unstream (S.init (stream t)) = init t

562570 #-}

563571

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

573+-- empty.

574+--

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

576+--

577+-- @since 1.2.3.0

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

579+unsnoc Empty = Nothing

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

581+{-# INLINE unsnoc #-}

582+

564583-- | /O(1)/ Tests whether a 'Text' is empty or not. Subject to

565584-- fusion.

566585null :: Text -> Bool

@@ -820,6 +839,8 @@ toUpper t = unstream (S.toUpper (stream t))

820839-- guides disagree on whether the book name \"The Hill of the Red

821840-- Fox\" is correctly title cased—but this function will

822841-- capitalize /every/ word.

842+--

843+-- @since 1.0.0.0

823844toTitle :: Text -> Text

824845toTitle t = unstream (S.toTitle (stream t))

825846{-# INLINE toTitle #-}

@@ -972,6 +993,8 @@ mapAccumR f = go

972993

973994-- | @'repeat' x@ is an infinite 'Text', with @x@ the value of every

974995-- element.

996+--

997+-- @since 1.2.0.5

975998repeat :: Char -> Text

976999repeat c = let t = Chunk (T.replicate smallChunkSize (T.singleton c)) t

9771000 in t

@@ -989,6 +1012,8 @@ replicate n t

9891012

9901013-- | 'cycle' ties a finite, non-empty 'Text' into a circular one, or

9911014-- equivalently, the infinite repetition of the original 'Text'.

1015+--

1016+-- @since 1.2.0.5

9921017cycle :: Text -> Text

9931018cycle Empty = emptyError "cycle"

9941019cycle t = let t' = foldrChunks Chunk t' t

@@ -998,6 +1023,8 @@ cycle t = let t' = foldrChunks Chunk t' t

9981023-- of @f@ to @x@:

9991024--

10001025-- > iterate f x == [x, f x, f (f x), ...]

1026+--

1027+-- @since 1.2.0.5

10011028iterate :: (Char -> Char) -> Char -> Text

10021029iterate f c = let t c' = Chunk (T.singleton c') (t (f c'))

10031030 in t c

@@ -1061,6 +1088,8 @@ take i t0 = take' i t0

10611088-- Examples:

10621089--

10631090-- > takeEnd 3 "foobar" == "bar"

1091+--

1092+-- @since 1.1.1.0

10641093takeEnd :: Int64 -> Text -> Text

10651094takeEnd n t0

10661095 | n <= 0 = empty

@@ -1099,6 +1128,8 @@ drop i t0

10991128-- Examples:

11001129--

11011130-- > dropEnd 3 "foobar" == "foo"

1131+--

1132+-- @since 1.1.1.0

11021133dropEnd :: Int64 -> Text -> Text

11031134dropEnd n t0

11041135 | n <= 0 = t0

@@ -1150,12 +1181,15 @@ takeWhile p t0 = takeWhile' t0

11501181-- Examples:

11511182--

11521183-- > takeWhileEnd (=='o') "foo" == "oo"

1184+--

1185+-- @since 1.2.2.0

11531186takeWhileEnd :: (Char -> Bool) -> Text -> Text

11541187takeWhileEnd p = takeChunk empty . L.reverse . toChunks

11551188 where takeChunk acc [] = acc

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

1157− then (Chunk t' acc)

1158− else takeChunk (Chunk t' acc) ts

1189+ takeChunk acc (t:ts)

1190+ | T.lengthWord16 t' < T.lengthWord16 t

1191+ = chunk t' acc

1192+ | otherwise = takeChunk (Chunk t' acc) ts

11591193 where t' = T.takeWhileEnd p t

11601194{-# INLINE takeWhileEnd #-}

11611195

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+3−1

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

1010-- License: BSD-style

1111-- Maintainer: Bryan O'Sullivan <bos@serpentine.com>

12−-- Stability: experimental

1312-- Portability: portable

1413--

1514-- Efficiently write an integral value to a 'Builder'.

@@ -32,6 +31,9 @@ import GHC.Base (quotInt, remInt)

3231import GHC.Num (quotRemInteger)

3332import GHC.Types (Int(..))

3433import Control.Monad.ST

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

35+import Prelude hiding ((<>))

36+#endif

3537

3638#ifdef __GLASGOW_HASKELL__

3739# if defined(INTEGER_GMP)

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-- | Control the rendering of floating point numbers.

2932data FPFormat = Exponent

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

@@ -4,7 +4,6 @@

55-- License : BSD-style

66-- Maintainer : bos@serpentine.com

7−-- Stability : experimental

87-- Portability : portable

98--

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

@@ -98,6 +97,8 @@ reverseIter (Text arr off _len) i

9897-- | /O(1)/ Iterate one step backwards through a UTF-16 array,

9998-- returning the delta to add (i.e. a negative number) to give the

10099-- next offset to iterate at.

100+--

101+-- @since 1.1.1.0

101102reverseIter_ :: Text -> Int -> Int

102103reverseIter_ (Text arr off _len) i

103104 | m < 0xDC00 || m > 0xDFFF = -1

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+56−19

@@ -1,15 +1,52 @@

1−1.2.2.1

1+### 1.2.3.0

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

4+ (updated from 8.0).

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

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

8+ (gh-184).

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

11+

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

13+

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

15+

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

17+

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

19+ examples, clarifications).

20+

21+#### 1.2.2.2

22+

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

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

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

26+

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

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

29+ smaller code generated.

30+

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

32+ used after some rewrite rule tweaking.

33+

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

35+

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

37+

38+#### 1.2.2.1

239

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

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

542 maintainers apologizes for the churn.

643

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

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

845 (updated from 7.0)

946

1047* An STG lint error has been fixed

1148

12−1.2.2.0

49+### 1.2.2.0

1350

1451* The `integer-simple` package, upon which this package optionally

1552 depended, has been replaced with `integer-pure`. The build flag has

@@ -25,7 +62,7 @@

2562 * if `base` >= 4.7: `PrintfArg`

2663 * if `base` >= 4.9: `Semigroup`

2764

28−1.2.1.3

65+#### 1.2.1.3

2966

3067* Bug fix: As it turns out, moving the literal rewrite rules to simplifier

3168 phase 2 does not prevent competition with the `unpack` rule, which is

@@ -33,7 +70,7 @@

3370 test environment mistake. Moving literal rules back to phase 1 finally

3471 fixes GHC Trac #10528 correctly.

3572

36−1.2.1.2

73+#### 1.2.1.2

3774

3875* Bug fix: Run literal rewrite rules in simplifier phase 2.

3976 The behavior of the simplifier changed in GHC 7.10.2,

@@ -41,21 +78,21 @@

4178 and long compilation times. See

4279 [GHC Trac #10528](https://ghc.haskell.org/trac/ghc/ticket/10528).

4380

44−1.2.1.1

81+#### 1.2.1.1

4582

4683* Expose unpackCString#, which you should never use.

4784

48−1.2.1.0

85+### 1.2.1.0

4986

5087* Added Binary instances for both Text types. (If you have previously

5188 been using the text-binary package to get a Binary instance, it is

5289 now obsolete.)

5390

54−1.2.0.6

91+#### 1.2.0.6

5592

5693* Fixed a space leak in UTF-8 decoding

5794

58−1.2.0.5

95+#### 1.2.0.5

5996

6097* Feature parity: repeat, cycle, iterate are now implemented for lazy

6198 Text, and the Data instance is more complete

@@ -75,26 +112,26 @@

75112* Spec compliance: toCaseFold now follows the Unicode 7.0 spec

76113 (updated from 6.3)

77114

78−1.2.0.4

115+#### 1.2.0.4

79116

80117* Fixed an incompatibility with base < 4.5

81118

82−1.2.0.3

119+#### 1.2.0.3

83120

84121* Update formatRealFloat to correspond to the definition in versions

85122 of base newer than 4.5 (https://github.com/bos/text/issues/105)

86123

87−1.2.0.2

124+#### 1.2.0.2

88125

89126* Bumped lower bound on deepseq to 1.4 for compatibility with the

90127 upcoming GHC 7.10

91128

92−1.2.0.1

129+#### 1.2.0.1

93130

94131* Fixed a buffer overflow in rendering of large Integers

95132 (https://github.com/bos/text/issues/99)

96133

97−1.2.0.0

134+## 1.2.0.0

98135

99136* Fixed an integer overflow in the replace function

100137 (https://github.com/bos/text/issues/81)

@@ -107,7 +144,7 @@

107144

108145* Added an instance of IsList for GHC 7.8 and above

109146

110−1.1.1.0

147+### 1.1.1.0

111148

112149* The Data.Data instance now allows gunfold to work, via a virtual

113150 pack constructor

@@ -117,12 +154,12 @@

117154* Comparing the length of a Text against a number can now

118155 short-circuit in more cases

119156

120−1.1.0.1

157+#### 1.1.0.1

121158

122159* streamDecodeUtf8: fixed gh-70, did not return all unconsumed bytes

123160 in single-byte chunks

124161

125−1.1.0.0

162+## 1.1.0.0

126163

127164* encodeUtf8: Performance is improved by up to 4x.

128165

@@ -141,12 +178,12 @@

141178 use at your own risk - there are no API stability guarantees for

142179 internal modules!

143180

144−1.0.0.1

181+#### 1.0.0.1

145182

146183* decodeUtf8: Fixed a regression that caused us to incorrectly

147184 identify truncated UTF-8 as valid (gh-61)

148185

149−1.0.0.0

186+# 1.0.0.0

150187

151188* Added support for Unicode 6.3.0 to case conversion functions

152189

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+90−17

@@ -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.

@@ -28,17 +30,21 @@ 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

3639author: Bryan O'Sullivan <bos@serpentine.com>

3740maintainer: Bryan O'Sullivan <bos@serpentine.com>

3841copyright: 2009-2011 Bryan O'Sullivan, 2008-2009 Tom Harper

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

3943category: Data, Text

4044build-type: Simple

4145cabal-version: >= 1.8

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

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

4248extra-source-files:

4349 -- scripts/CaseFolding.txt

4450 -- scripts/SpecialCasing.txt

@@ -65,7 +71,9 @@ extra-source-files:

6571 tests/text-tests.cabal

6672

6773flag bytestring-builder

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

74+ description:

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

76+ package for backwards compatibility.

6977 default: False

7078 manual: False

7179

@@ -75,7 +83,9 @@ flag developer

7583 manual: True

7684

7785flag integer-simple

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

86+ description:

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

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

7989 default: False

8090 manual: False

8191

@@ -160,27 +170,90 @@ library

160170

161171test-suite tests

162172 type: exitcode-stdio-1.0

163− hs-source-dirs: tests .

164− main-is: Tests.hs

165173 c-sources: cbits/cbits.c

166174 include-dirs: include

167175

168176 ghc-options:

169− -Wall -threaded -O0 -rtsopts

177+ -Wall -threaded -rtsopts

170178

171179 cpp-options:

172180 -DASSERTS -DHAVE_DEEPSEQ -DTEST_SUITE

173181

182+ -- modules specific to test-suite

183+ hs-source-dirs: tests

184+ main-is: Tests.hs

185+ other-modules:

186+ Tests.Properties

187+ Tests.Properties.Mul

188+ Tests.QuickCheckUtils

189+ Tests.Regressions

190+ Tests.SlowFunctions

191+ Tests.Utils

192+

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

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

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

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

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

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

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

200+ hs-source-dirs: .

201+ other-modules:

202+ Data.Text

203+ Data.Text.Array

204+ Data.Text.Encoding

205+ Data.Text.Encoding.Error

206+ Data.Text.Foreign

207+ Data.Text.IO

208+ Data.Text.Internal

209+ Data.Text.Internal.Builder

210+ Data.Text.Internal.Builder.Functions

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

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

213+ Data.Text.Internal.Encoding.Fusion

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

215+ Data.Text.Internal.Encoding.Utf16

216+ Data.Text.Internal.Encoding.Utf32

217+ Data.Text.Internal.Encoding.Utf8

218+ Data.Text.Internal.Functions

219+ Data.Text.Internal.Fusion

220+ Data.Text.Internal.Fusion.CaseMapping

221+ Data.Text.Internal.Fusion.Common

222+ Data.Text.Internal.Fusion.Size

223+ Data.Text.Internal.Fusion.Types

224+ Data.Text.Internal.IO

225+ Data.Text.Internal.Lazy

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

227+ Data.Text.Internal.Lazy.Fusion

228+ Data.Text.Internal.Lazy.Search

229+ Data.Text.Internal.Private

230+ Data.Text.Internal.Read

231+ Data.Text.Internal.Search

232+ Data.Text.Internal.Unsafe

233+ Data.Text.Internal.Unsafe.Char

234+ Data.Text.Internal.Unsafe.Shift

235+ Data.Text.Lazy

236+ Data.Text.Lazy.Builder

237+ Data.Text.Lazy.Builder.Int

238+ Data.Text.Lazy.Builder.RealFloat

239+ Data.Text.Lazy.Encoding

240+ Data.Text.Lazy.IO

241+ Data.Text.Lazy.Internal

242+ Data.Text.Lazy.Read

243+ Data.Text.Read

244+ Data.Text.Unsafe

245+ Data.Text.Show

246+

174247 build-depends:

175248 HUnit >= 1.2,

176− QuickCheck >= 2.7,

249+ QuickCheck >= 2.7 && < 2.10,

177250 array,

178251 base,

179252 binary,

180253 deepseq,

181254 directory,

182255 ghc-prim,

183− quickcheck-unicode,

256+ quickcheck-unicode >= 1.0.1.0,

184257 random,

185258 test-framework >= 0.4,

186259 test-framework-hunit >= 0.2,

@@ -201,7 +274,7 @@ test-suite tests

201274

202275source-repository head

203276 type: git

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

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

205278

206279source-repository head

207280 type: mercurial