+{-# LANGUAGE
+ UnicodeSyntax
+ #-}
+-- |A set of auxiliary parser utilities.
module Network.HTTP.Lucu.Parser
- ( Parser(..)
- , ParserResult(..)
-
- , parse -- Parser a -> ByteString -> (ParserResult a, ByteString)
- , parseStr -- Parser a -> String -> (ParserResult a, ByteString)
-
- , anyChar -- Parser Char
- , eof -- Parser ()
- , allowEOF -- Parser a -> Parser a
- , satisfy -- (Char -> Bool) -> Parser Char
- , char -- Char -> Parser Char
- , string -- String -> Parser String
- , (<|>) -- Parser a -> Parser a -> Parser a
- , oneOf -- [Char] -> Parser Char
- , digit -- Parser Char
- , hexDigit -- Parser Char
- , notFollowedBy -- Parser a -> Parser ()
- , many -- Parser a -> Parser [a]
- , many1 -- Parser a -> Parser [a]
- , manyTill -- Parser a -> Parser end -> Parser [a]
- , many1Till -- Parser a -> Parser end -> Parser [a]
- , count -- Int -> Parser a -> Parser [a]
- , option -- a -> Parser a -> Parser a
- , sepBy -- Parser a -> Parser sep -> Parser [a]
- , sepBy1 -- Parser a -> Parser sep -> Parser [a]
-
- , sp -- Parser Char
- , ht -- Parser Char
- , crlf -- Parser String
+ ( atMost
+ , finishOff
+ , skipManyTill
+ , skipWhile1
+ , skipSpace1
+ , isAlphaNum
)
where
-
-import Control.Monad
-import Control.Monad.State
-import qualified Data.ByteString.Lazy.Char8 as B
-import Data.ByteString.Lazy.Char8 (ByteString)
-
-data Parser a = Parser {
- runParser :: State ParserState (ParserResult a)
- }
-
-type ParserState = (ByteString, IsEOFFatal)
-
-type IsEOFFatal = Bool
-
-data ParserResult a = Success a
- | IllegalInput -- 受理出來ない入力があった
- | ReachedEOF -- 限界を越えて讀まうとした
- deriving (Eq, Show)
-
-
--- (>>=) :: Parser a -> (a -> Parser b) -> Parser b
-instance Monad Parser where
- p >>= f = Parser $ do saved@(_, isEOFFatal) <- get -- 失敗した時の爲に状態を保存
- result <- runParser p
- case result of
- Success a -> runParser (f a)
- IllegalInput -> do put saved -- 状態を復歸
- return IllegalInput
- ReachedEOF -> if isEOFFatal then
- return ReachedEOF
- else
- do put saved
- return IllegalInput
- return = Parser . return . Success
- fail _ = Parser $ return IllegalInput
-
-
-parse :: Parser a -> ByteString -> (ParserResult a, ByteString)
-parse p input = let (result, (input', _)) = runState (runParser p) (input, True)
- in
- (result, input')
-
-
-parseStr :: Parser a -> String -> (ParserResult a, ByteString)
-parseStr p input = parse p $ B.pack input
-
-
-anyChar :: Parser Char
-anyChar = Parser $ do (input, isEOFFatal) <- get
- if B.null input then
- return ReachedEOF
- else
- do let c = B.head input
- put (B.tail input, isEOFFatal)
- return (Success c)
-
-
-eof :: Parser ()
-eof = Parser $ do (input, _) <- get
- if B.null input then
- return $ Success ()
- else
- return IllegalInput
-
-
-allowEOF :: Parser a -> Parser a
-allowEOF f = Parser $ do (input, isEOFFatal) <- get
- put (input, False)
-
- result <- runParser f
-
- (input', _) <- get
- put (input', isEOFFatal)
-
- return result
-
-
-satisfy :: (Char -> Bool) -> Parser Char
-satisfy f = do c <- anyChar
- unless (f c) (fail "")
- return c
-
-
-char :: Char -> Parser Char
-char c = satisfy (== c)
-
-
-string :: String -> Parser String
-string str = do mapM_ char str
- return str
-
-
-infixr 0 <|>
-
-(<|>) :: Parser a -> Parser a -> Parser a
-f <|> g = Parser $ do saved@(_, isEOFFatal) <- get -- 状態を保存
- result <- runParser f
- case result of
- Success a -> return $ Success a
- IllegalInput -> do put saved -- 状態を復歸
- runParser g
- ReachedEOF -> if isEOFFatal then
- return ReachedEOF
- else
- do put saved
- runParser g
-
-
-oneOf :: [Char] -> Parser Char
-oneOf = foldl (<|>) (fail "") . map char
-
-
-notFollowedBy :: Parser a -> Parser ()
-notFollowedBy p = p >>= fail "" <|> return ()
-
-
-digit :: Parser Char
-digit = do c <- anyChar
- if c >= '0' && c <= '9' then
- return c
- else
- fail ""
-
-
-hexDigit :: Parser Char
-hexDigit = do c <- anyChar
- if (c >= '0' && c <= '9') ||
- (c >= 'a' && c <= 'f') ||
- (c >= 'A' && c <= 'F') then
- return c
- else
- fail ""
-
-
-many :: Parser a -> Parser [a]
-many p = do x <- p
- xs <- many p
- return (x:xs)
- <|>
- return []
-
-
-many1 :: Parser a -> Parser [a]
-many1 p = do ret <- many p
- case ret of
- [] -> fail ""
- xs -> return xs
-
-
-manyTill :: Parser a -> Parser end -> Parser [a]
-manyTill p end = many $ do x <- p
- end
- return x
-
-
-many1Till :: Parser a -> Parser end -> Parser [a]
-many1Till p end = many1 $ do x <- p
- end
- return x
-
-
-count :: Int -> Parser a -> Parser [a]
-count 0 _ = return []
-count n p = do x <- p
- xs <- count (n-1) p
- return (x:xs)
-
-
-option :: a -> Parser a -> Parser a
-option def p = p <|> return def
-
-
-sepBy :: Parser a -> Parser sep -> Parser [a]
-sepBy p sep = sepBy1 p sep <|> return []
-
-
-sepBy1 :: Parser a -> Parser sep -> Parser [a]
-sepBy1 p sep = do x <- p
- xs <- many $ sep >> p
- return (x:xs)
-
-
-sp :: Parser Char
-sp = char ' '
-
-
-ht :: Parser Char
-ht = char '\t'
-
-
-crlf :: Parser String
-crlf = string "\x0d\x0a"
+import Control.Applicative
+import Control.Applicative.Unicode
+import Control.Monad.Unicode
+import Data.Attoparsec.Char8
+import Prelude.Unicode
+
+-- |@'atMost' n v@ is like @'P.many' v@ but accumulates @v@ at most
+-- @n@ times.
+atMost ∷ Alternative f ⇒ Int → f α → f [α]
+{-# INLINEABLE atMost #-}
+atMost 0 _ = pure []
+atMost n v = ( (:) <$> v ⊛ atMost (n-1) v )
+ <|>
+ pure []
+
+-- |@'finishOff' p@ is equivalent to @p '>>=' \\a -> 'endOfInput' '>>'
+-- 'return' a@.
+finishOff ∷ Parser α → Parser α
+{-# INLINE finishOff #-}
+finishOff = ((endOfInput *>) ∘ return =≪)
+
+-- |Similar to 'manyTill' but discards the result.
+skipManyTill ∷ Alternative f ⇒ f α → f β → f ()
+{-# INLINEABLE skipManyTill #-}
+skipManyTill p end = go
+ where
+ go = (end *> pure ()) <|> (p *> go)
+
+-- |Similar to 'skipWhile' but consumes at least one character.
+skipWhile1 ∷ (Char → Bool) → Parser ()
+{-# INLINE skipWhile1 #-}
+skipWhile1 p = takeWhile1 p *> pure ()
+
+-- |Similar to 'skipSpace' but consumes at least one whitespace.
+skipSpace1 ∷ Parser ()
+{-# INLINE skipSpace1 #-}
+skipSpace1 = skipMany1 space
+
+-- |@'isAlphaNum' c@ returns 'True' iff @'isDigit' c ||
+-- 'isAlpha_ascii' c@.
+isAlphaNum ∷ Char → Bool
+{-# INLINE isAlphaNum #-}
+isAlphaNum c = isDigit c ∨ isAlpha_ascii c