9f6a43f · text · git.iko.soy

git.iko.soy/text

on master help sign in

Make the split* functions undefined over empty delimiters as inputs.

9f6a43fBryan O'Sullivan17 years ago

Data/Text.hs+28−6

compare with master

unifiedsplit

@@ -156,6 +156,7 @@ module Data.Text

156 , findIndices

157 , elemIndex

158 , elemIndices

159+ , count

159 , countChar

160 , countChar

160

161

161 -- * Zipping and unzipping

162 -- * Zipping and unzipping

@@ -899,7 +900,10 @@ tails t | null t = [empty]

899-- are slices of the original.

900-- are slices of the original.

900

901

901-- | /O(m)*O(n)/ Break a 'Text' into pieces separated by the first

902-- | /O(m)*O(n)/ Break a 'Text' into pieces separated by the first

902−-- 'Text' argument, consuming the delimiter. Examples:

903+-- 'Text' argument, consuming the delimiter. An empty delimiter is

904+-- invalid, and will cause an error to be raised.

905+--

906+-- Examples:

903--

907--

904-- > split "\r\n" "a\r\nb\r\nd\r\ne" == ["a","b","d","e"]

908-- > split "\r\n" "a\r\nb\r\nd\r\ne" == ["a","b","d","e"]

905-- > split "aaa" "aaaXaaaXaaaXaaa" == ["","X","X","X",""]

909-- > split "aaa" "aaaXaaaXaaaXaaa" == ["","X","X","X",""]

@@ -913,8 +917,8 @@ split :: Text -- ^ Text to split on

913 -> Text -- ^ Input text

917 -> Text -- ^ Input text

914 -> [Text]

918 -> [Text]

915split pat src0

919split pat src0

916− | l == 0 = [src0]

920+ | l == 0 = emptyError "split"

917− | l == 1 = splitWith (== (head pat)) src0

921+ | l == 1 = splitWith (== (unsafeHead pat)) src0

918 | otherwise = go src0

922 | otherwise = go src0

919 where

923 where

920 l = length pat

924 l = length pat

@@ -936,7 +940,8 @@ split pat src0

936-- and consuming the delimiter. The last element of the list contains

940-- and consuming the delimiter. The last element of the list contains

937-- the remaining text after the number of times to split has been

941-- the remaining text after the number of times to split has been

938-- reached. A value of zero or less for @k@ causes no splitting to

942-- reached. A value of zero or less for @k@ causes no splitting to

939−-- occur.

943+-- occur. An empty delimiter is invalid, and will cause an error to be

944+-- raised.

940--

945--

941-- Examples:

946-- Examples:

942--

947--

@@ -952,8 +957,9 @@ splitTimes :: Int -- ^ Maximum number of times to split

952 -> Text -- ^ Input text

957 -> Text -- ^ Input text

953 -> [Text]

958 -> [Text]

954splitTimes k pat src0

959splitTimes k pat src0

955− | k <= 0 || l == 0 = [src0]

960+ | k <= 0 = [src0]

961+ | l == 0 = emptyError "splitTimes"

956− | otherwise = go k src0

962+ | otherwise = go k src0

957 where

963 where

958 l = length pat

964 l = length pat

959 go !i src = search 0 src

965 go !i src = search 0 src

@@ -1100,6 +1106,22 @@ elemIndices :: Char -> Text -> [Int]

1100elemIndices c t = S.elemIndices c (stream t)

1106elemIndices c t = S.elemIndices c (stream t)

1101{-# INLINE elemIndices #-}

1107{-# INLINE elemIndices #-}

1102

1108

1109+-- | /O(n*m)/ The 'count' function returns the number of times the

1110+-- query string appears in the given 'Text'.

1111+count :: Text -> Text -> Int

1112+count pat src0

1113+ | l == 0 = length src0 + 1

1114+ | l == 1 = countChar (unsafeHead pat) src0

1115+ | otherwise = go 0 src0

1116+ where

1117+ l = length pat

1118+ go !n src = search src

1119+ where

1120+ search s | null s = n

1121+ | pat `isPrefixOf` s = go (n+1) (drop l s)

1122+ | otherwise = search (unsafeTail s)

1123+{-# INLINE [1] count #-}

1124+

1103-- | /O(n)/ The 'countChar' function returns the number of times the

1125-- | /O(n)/ The 'countChar' function returns the number of times the

1104-- query element appears in the given 'Text'. Subject to fusion.

1126-- query element appears in the given 'Text'. Subject to fusion.

1105countChar :: Char -> Text -> Int

1127countChar :: Char -> Text -> Int