implementation module lex
-import Control.Monad
-import Data.Either
-import Data.Maybe
-import Data.Map
-import Data.Func
-import StdString
-import StdTuple
+import Control.Monad, Control.Applicative
+import Data.Either, Data.Func, Data.Void
+from StdFunc import o
import StdBool
import StdList
import StdChar
-import qualified Text
+import StdString
+import StdTuple
-SingleCharTokens :: Map Char Token
-SingleCharTokens = fromList [
- ('(', BraceOpenToken), (')', BraceCloseToken), ('{', CBraceOpenToken),
- ('}', CBraceCloseToken), ('[', SquareOpenToken), (']', SquareCloseToken),
- (',', CommaToken), (':', ColonToken), (';', SColonToken),
- ('.', DotToken), ('+', PlusToken), ('*', StarToken), ('/', SlashToken),
- ('%', PercentToken), ('=', AssignmentToken), ('<', LesserToken),
- ('>', BiggerToken), ('!', ExclamationToken)]
+import yard
+import AST
-EscapeMap :: Map Char Char
-EscapeMap = fromList [('a', toChar 7), ('b', '\b'), ('f', '\f'), ('n', '\n'),
- ('r', '\r'), ('t', '\t'), ('v', '\v')]
+:: LexItem
+ = LexToken Int TokenValue
+ | LexSpace Int Int
+ | LexNL
+ | LexEOF
+ | LexItemError String
lexer :: [Char] -> LexerOutput
-lexer [] = Right []
-lexer x = case lex x of
- (Right t, rest) = lexer rest >>= \ts.Right [t:ts]
- (Left e, _) = Left e
+lexer r = fst $ runParser (lexProgram 1 1) r
+
+lexProgram :: Int Int -> Parser Char [Token]
+lexProgram line column = lexToken >>= \t->case t of
+ LexEOF = pure []
+ LexNL = lexProgram (line+1) 1
+ (LexSpace l c) = lexProgram (line+l) (column+c)
+ (LexItemError e) = fail <?>
+ PositionalError line column ("LexerError: " +++ e)
+ (LexToken c t) = lexProgram line (column+c)
+ >>= \rest->pure [({line=line,col=column}, t):rest]
+
+lexToken :: Parser Char LexItem
+lexToken =
+ //Comments
+ lexBlockComment <|> lexComment <|>
+ //Keyword tokens
+ lexKw "var" VarToken <|> lexKw "Void" VoidToken <|>
+ lexKw "return" ReturnToken <|> lexKw "if" IfToken <|>
+ lexKw "else" ElseToken <|> lexKw "while" WhileToken <|>
+ lexKw "True" TrueToken <|> lexKw "False" FalseToken <|>
+ lexKw "Int" IntTypeToken <|> lexKw "Bool" BoolTypeToken <|>
+ lexKw "Char" CharTypeToken <|>
+ //Character tokens
+ lexEscape <|> lexCharacter <|>
+ //Two char ops tokens
+ lexWord "::" DoubleColonToken <|> lexWord "!=" NotEqualToken <|>
+ lexWord "<=" LesserEqToken <|> lexWord ">=" GreaterEqToken <|>
+ lexWord "==" EqualsToken <|> lexWord "&&" AmpersandsToken <|>
+ lexWord "||" PipesToken <|> lexWord "[]" EmptyListToken <|>
+ lexWord "->" ArrowToken <|>
+ //One char ops tokens
+ lexWord "(" BraceOpenToken <|> lexWord ")" BraceCloseToken <|>
+ lexWord "{" CBraceOpenToken <|> lexWord "}" CBraceCloseToken <|>
+ lexWord "[" SquareOpenToken <|> lexWord "]" SquareCloseToken <|>
+ lexWord "," CommaToken <|> lexWord ":" ColonToken <|>
+ lexWord ";" SColonToken <|> lexWord "." DotToken <|>
+ lexWord "+" PlusToken <|> lexWord "*" StarToken <|>
+ lexWord "/" SlashToken <|> lexWord "%" PercentToken <|>
+ lexWord "=" AssignmentToken <|> lexWord "<" LesserToken <|>
+ lexWord ">" BiggerToken <|> lexWord "!" ExclamationToken <|>
+ lexWord "-" DashToken <|>
+ //Number and identifier tokens
+ lexNumber <|> lexIdentifier <|>
+ (item '\n' >>| pure LexNL) <|>
+ //Whitespace
+ (satisfy isSpace >>| (pure $ LexSpace 0 1)) <|>
+ (eof >>| pure LexEOF)
+ where
+ lexWord :: String TokenValue -> Parser Char LexItem
+ lexWord s tv = list ls >>| pure (LexToken (length ls) tv)
+ where ls = fromString s
+
+ lexKw :: String TokenValue -> Parser Char LexItem
+ lexKw kw tv = lexWord kw tv <* check (not o isIdentChar)
+
+ lexComment :: Parser Char LexItem
+ lexComment = list (fromString "//")
+ >>| top until (eof <|> (item '\n' >>| pure Void)) >>| pure LexNL
+
+ lexBlockComment :: Parser Char LexItem
+ lexBlockComment = list (fromString "/*")
+ >>| (top until (list (fromString "*/")))
+ >>= \chars->pure $ widthHeight chars 0 0
+ where
+ widthHeight :: [Char] Int Int -> LexItem
+ widthHeight [] l c = LexSpace l c
+ widthHeight ['\n':xs] l _ = widthHeight xs (l+1) 0
+ widthHeight [x:xs] l c = widthHeight xs l (c+1)
+
+ lexNumber :: Parser Char LexItem
+ lexNumber = some (satisfy isDigit) >>= \si->pure $
+ LexToken (length si) (NumberToken $ toInt $ toString si)
+
+ lexIdentifier :: Parser Char LexItem
+ lexIdentifier = some (satisfy isIdentChar)
+ >>= \si->pure $ LexToken (length si) (IdentToken $ toString si)
+
+ isIdentChar c = isAlphanum c || c == '_'
+
+ lexCharacter :: Parser Char LexItem
+ lexCharacter = item '\'' *> satisfy ((<>) '\'') <* item '\''
+ >>= \char->pure $ LexToken 3 (CharToken char)
-lex :: [Char] -> (Either String Token, [Char])
-lex [] = (Right EndOfFileToken, [])
-lex ['/':'/':xs] = lex $ dropWhile ((<>) '\n') xs
-lex ['/':'*':xs] =
- lex $ tl $ map snd $ dropWhile ((<>) ('*', '/')) $ zip2 [zero:xs] xs
-lex ['v':'a':'r':xs] = (Right VarToken, xs)
-lex ['V':'o':'i':'d':xs] = (Right VoidToken, xs)
-lex ['r':'e':'t':'u':'r':'n':xs] = (Right ReturnToken, xs)
-lex ['i':'f':xs] = (Right IfToken, xs)
-lex ['e':'l':'s':'e':xs] = (Right ElseToken, xs)
-lex ['w':'h':'i':'l':'e':xs] = (Right WhileToken, xs)
-lex ['T':'r':'u':'e':xs] = (Right TrueToken, xs)
-lex ['F':'a':'l':'s':'e':xs] = (Right FalseToken, xs)
-lex ['I':'n':'t':xs] = (Right IntTypeToken, xs)
-lex ['C':'h':'a':'r':xs] = (Right CharTypeToken, xs)
-lex ['B':'o':'o':'l':xs] = (Right BoolTypeToken, xs)
-lex [':':':':xs] = (Right DoubleColonToken, xs)
-lex ['!':'=':xs] = (Right NotEqualToken, xs)
-lex ['<':'=':xs] = (Right LesserEqToken, xs)
-lex ['>':'=':xs] = (Right GreaterEqToken, xs)
-lex ['=':'=':xs] = (Right EqualsToken, xs)
-lex ['&':'&':xs] = (Right AmpersandsToken, xs)
-lex ['|':'|':xs] = (Right PipesToken, xs)
-lex ['-':'>':xs] = (Right ArrowToken, xs)
-lex ['\'':x:'\'':xs] = (Right (CharToken x), xs)
-lex ['\'':'\\':x:'\'':xs] = case get x EscapeMap of
- Just t = (Right (CharToken t), xs)
- _ = (Left ("Unknown escape: \\" +++ toString x), [])
-lex ['-':t=:[x:xs]]
-| isDigit x = let (i, r) = span isDigit t in (Right (NumberToken ['-':i]), r)
-| otherwise = (Right DashToken, xs)
-lex t=:[x:xs] = case get x SingleCharTokens of
- (Just tok) = (Right tok, xs)
- Nothing
- | isSpace x = lex xs
- | isDigit x = let (i, r) = span isDigit t in (Right (NumberToken i), r)
- | isAlpha x = let (i, r) = span isIdent t in (Right (IdentToken i), r)
- with isIdent c = isAlphanum c || c == '_'
- | otherwise = (Left ("Unexpected character: " +++ toString x), [])
+ lexEscape :: Parser Char LexItem
+ lexEscape = item '\'' *> item '\\' *> top <* item '\''
+ >>= \char->pure case char of
+ 'a' = LexToken 4 (CharToken $ toChar 7)
+ 'b' = LexToken 4 (CharToken '\b')
+ 'b' = LexToken 4 (CharToken '\b')
+ 'f' = LexToken 4 (CharToken '\f')
+ 'n' = LexToken 4 (CharToken '\n')
+ 'r' = LexToken 4 (CharToken '\t')
+ 'v' = LexToken 4 (CharToken '\v')
+ '\'' =LexToken 4 (CharToken '\'')
+ c = (LexItemError $ "Unknown escape: " +++ toString c)