X-Git-Url: https://git.martlubbers.net/?a=blobdiff_plain;f=lex.icl;h=de200df2f4dcd62930ea1e74bf20f97fb7b9872a;hb=75ac4a13efdc20b444a5eb6cb2d16425c82d333e;hp=cbdcd653ab2f20f4778ffd7f863c721c64a6aadc;hpb=82d6648e3777f5bb74fa1825f18cf41922f3b782;p=cc1516.git

diff --git a/lex.icl b/lex.icl
index cbdcd65..de200df 100644
--- a/lex.icl
+++ b/lex.icl
@@ -1,158 +1,80 @@
 implementation module lex
 
-import Data.Either
-import Data.List
-import StdString
-import System.CommandLine
-import StdFile
-import StdMisc
-from StdFunc import id, const
-import Data.Maybe
-import Control.Applicative
 import Control.Monad
-import Control.Monad.State
-import Data.Functor
-from Data.Func import $
-
-// Misschien moeten we hier continuation style van maken
-instance toString lexerOutput where
-	toString l = "dit is een lexer output, danwel error\n"
-
-lexer :: [Char] -> LexerOutput [Token]
-lexer i = let (out, rest) = runParser (many anyT) i in case rest of
-            []  = out
-            _   = Left "Unexpected input after last token"
-
-//lexer functions
-identT  = alpha >>= \a -> many (char '_' <|> alphaNum) >>= \as -> return $ IdentToken [a:as]
-numberT = optional (char '-') >>= \sign -> (some digit) >>= \n -> case sign of
-            Nothing -> return $ NumberToken $ 5  //fromString n
-            _       -> return $ NumberToken $ -5 //(fromString n) * -1
-charLT  = CharToken <$> (char '\'' *> item <* char '\'')
-char2T  = item >>= \c1 -> case c1 of
-            ':' = char ':' >>| return DoubleColonToken
-            '<' = char '=' >>| return LesserEqToken
-            '>' = char '=' >>| return GreaterEqToken
-            '=' = char '=' >>| return EqualsToken
-            '&' = char '&' >>| return AmpersandsToken
-            '|' = char '|' >>| return PipesToken
-            '-' = char '>' >>| return ArrowToken
-            _   = zero
-char1T  = item >>= \c1 -> findT c1 charTokenMap
-varT    = string (fromString "var") >>| return VarToken
-voidT   = string (fromString "Void") >>| return VoidToken
-returnT = string (fromString "return") >>| return ReturnToken
-ifT     = string (fromString "if") >>| return IfToken
-elseT   = string (fromString "else") >>| return ElseToken
-whileT  = string (fromString "while") >>| return WhileToken
-trueT   = string (fromString "True") >>| return TrueToken
-falseT  = string (fromString "False") >>| return FalseToken
-//note, for the anyToken parser the order matters!
-anyT    = char2T <|> char1T <|> varT <|> voidT <|> returnT <|> ifT <|> elseT <|> whileT <|>
-          trueT <|> falseT <|> numberT <|> identT <|> charLT
-
-Start :: *World -> *World
-Start w
-# (args, w) = getCommandLine w // We lezen nu nog standaard van stdin
-# (out, w) = stdio w
-# (toparse, out) = readEntireFile out
-# out = out <<< toString (lexer toparse)
-# (b, w) = fclose out w
-| not b = setReturnCode 1 w
-= w
+import Data.Either, Data.Maybe, Data.Map, Data.Func
+import StdString
+import StdTuple
+import StdBool
+import StdList
+import StdChar
+
+SingleCharTokens :: Map Char TokenValue
+SingleCharTokens = fromList [
+	('(', BraceOpenToken), (')', BraceCloseToken), ('{', CBraceOpenToken),
+	('}', CBraceCloseToken), ('[', SquareOpenToken), (']', SquareCloseToken),
+	(',', CommaToken), (':', ColonToken), (';', SColonToken),
+	('.', DotToken), ('+', PlusToken), ('*', StarToken), ('/', SlashToken),
+	('%', PercentToken), ('=', AssignmentToken), ('<', LesserToken),
+	('>', BiggerToken), ('!', ExclamationToken)]
+
+EscapeMap :: Map Char Char
+EscapeMap = fromList [('a', toChar 7), ('b', '\b'), ('f', '\f'), ('n', '\n'),
+	('r', '\r'), ('t', '\t'), ('v', '\v')]
+
+lexer :: [Char] -> LexerOutput
+lexer x = lexer` x 1
 	where
-		readEntireFile :: *File -> *([Char], *File)
-		readEntireFile f
-		# (b, c, f) = freadc f
-		| not b = ([], f)
-		# (cs, f) = readEntireFile f
-		= ([c:cs], f)
-
-
-
-charTokenMap =  [('(', BraceOpenToken)
-                ,(')', BraceCloseToken)
-                ,('{', CBraceOpenToken)
-                ,('}', CBraceCloseToken)
-                ,('[', SquareOpenToken)
-                ,(']', SquareCloseToken)
-                ,(',', CommaToken)
-                ,(':', ColonToken)
-                ,(';', SColonToken)
-                ,('.', DotToken)
-                ,('+', PlusToken)
-                ,('-', DashToken)
-                ,('*', StarToken)
-                ,('/', SlashToken)
-                ,('%', PercentToken)
-                ,('=', AssignmentToken)
-                ,('<', LesserToken)
-                ,('>', BiggerToken)
-                ,('!', ExclamationToken)]
-findT c []           = fail "Unrecognized character"
-findT c [(k,v):xs]   = if (c==k) (return v) (findT c xs)
-
-
-// Clean adaption of Yard, a parsec like parser combinator
-:: Parser a = Parser ([Char] -> (LexerOutput a, [Char]))
-
-runParser :: (Parser a) [Char] -> (LexerOutput a, [Char])
-runParser (Parser p) s = p s
-
-instance Functor Parser where
-    fmap f s    = liftM f s
-
-instance Applicative Parser where
-    pure a      = Parser $ \s -> (Right a, s) 
-    (<*>) sf s  = ap sf s
-
-instance Monad Parser where
-    bind p f    = Parser $ \s -> let (out, rest) = runParser p s in case out of
-                    Left e  = (Left e, rest)
-                    Right t = runParser (f t) rest
-
-//gives us some, many and optional
-instance Alternative Parser where
-    empty       = zero
-    (<|>) p1 p2 = Parser $ \s -> let (out, rest) = runParser p1 s in case out of
-                    Left e  = runParser p2 s
-                    Right t = (Right t, rest)
-
-//parser that fails with error
-fail :: String -> Parser a
-fail e = Parser $ \s -> (Left e, s)
-
-//parser that always fails
-zero :: Parser a
-zero = fail "Zero parser"
-
-//matches exactly one Char
-item :: Parser Char
-item = Parser $ \s -> case s of 
-        []      = (Left "Unexpected empty input", s)
-        [x:xs]  = (Right x, xs)
-
-//matches any char which satisfies f
-satisfy :: (Char -> Bool) -> Parser Char
-satisfy f = item >>= (\r -> if (f r) (return r) zero)
-
-//tries a parser, if it fails returns a default value
-optionalDef :: a (Parser a) -> Parser a
-optionalDef def p = p <|> return def
-
-//matched given char
-char :: Char -> Parser Char
-char c = satisfy (\i -> c==i) //I hate that we can't do: satisfy (==c)
-
-alpha :: Parser Char
-alpha = satisfy isAlpha
-
-digit :: Parser Char
-digit = satisfy isDigit
-
-alphaNum :: Parser Char
-alphaNum = alpha <|> digit
-
-//matches a given String
-string :: [Char] -> Parser [Char]
-string s = mapM_ char s >>| return s
\ No newline at end of file
+		lexer` :: [Char] Int -> LexerOutput
+		lexer` [] i = Right []
+		lexer` x i = case lex x i of
+			(Right t, rest, i) = lexer` rest i >>= \ts.Right [t:ts]
+			(Left e, _, _) = Left e
+
+ret:: TokenValue [Char] Int -> (Either String Token, [Char], Int)
+ret tv rest i = (Right (i, tv), rest, i)
+
+err :: String Int -> (Either String Token, [Char], Int)
+err s i = (Left ("Line " +++ toString i +++ ": " +++ s), [], i)
+
+lex :: [Char] -> (Int -> (Either String Token, [Char], Int))
+lex [] = ret EndOfFileToken []
+lex ['/':'/':xs] = lex (dropWhile ((<>) '\n') xs)
+lex ['/':'*':xs] =  \i.let (c, r) = span ((<>) ('*','/')) $ zip2 [' ':xs] xs in
+	lex (tl $ map snd $ tl $ r) (i + length [x\\(x, _)<-c | x == '\n'])
+lex ['v':'a':'r':xs] = ret VarToken xs
+lex ['V':'o':'i':'d':xs] = ret VoidToken xs
+lex ['r':'e':'t':'u':'r':'n':xs] = ret ReturnToken xs
+lex ['i':'f':xs] = ret IfToken xs
+lex ['e':'l':'s':'e':xs] = ret ElseToken xs
+lex ['w':'h':'i':'l':'e':xs] = ret WhileToken xs
+lex ['T':'r':'u':'e':xs] = ret TrueToken xs
+lex ['F':'a':'l':'s':'e':xs] = ret FalseToken xs
+lex ['I':'n':'t':xs] = ret IntTypeToken xs
+lex ['C':'h':'a':'r':xs] = ret CharTypeToken xs
+lex ['B':'o':'o':'l':xs] = ret BoolTypeToken xs
+lex [':':':':xs] = ret DoubleColonToken xs
+lex ['!':'=':xs] = ret NotEqualToken xs
+lex ['<':'=':xs] = ret LesserEqToken xs
+lex ['>':'=':xs] = ret GreaterEqToken xs
+lex ['=':'=':xs] = ret EqualsToken xs
+lex ['&':'&':xs] = ret AmpersandsToken xs
+lex ['|':'|':xs] = ret PipesToken xs
+lex ['-':'>':xs] = ret ArrowToken xs
+lex ['\'':x:'\'':xs] = ret (CharToken x) xs
+lex ['\'':'\\':x:'\'':xs] = case get x EscapeMap of
+	Just t = ret (CharToken t) xs
+	_ = err ("Unknown escape: \\" +++ toString x)
+lex ['-':t=:[x:xs]]
+| isDigit x = let (n, r) = span isDigit t in
+	ret (NumberToken (toInt $ toString ['-':n])) r
+| otherwise = ret DashToken  xs
+lex t=:[x:xs] = case get x SingleCharTokens of
+	(Just tok) = ret tok xs
+	Nothing
+	| x == '\n' = \i.lex xs (i+1)
+	| isSpace x = lex xs
+	| isDigit x = let (v, r) = span isDigit t in 
+		ret (NumberToken (toInt $ toString v)) r
+	| isAlpha x = let (v, r) = span isIdent t in ret (IdentToken v) r
+		with isIdent c = isAlphanum c || c == '_'
+	| otherwise = err ("Unexpected character: " +++ toString x)