lex.icl

   1 implementation module lex
   2
   3 import Control.Monad, Control.Applicative
   4 import Data.Either, Data.Func
   5 from StdFunc import o
   6 import StdBool
   7 import StdList
   8 import StdChar
   9 import StdString
  10
  11 import yard
  12 from AST import :: Pos(..)
  13
  14 :: LexItem
  15         = LexToken Int TokenValue
  16         | LexSpace Int Int
  17         | LexNL
  18         | LexEOF
  19         | LexItemError String
  20
  21 lexer :: [Char] -> LexerOutput
  22 lexer r = case runParser (lexProgram 1 1) r of
  23         (Right p, _) = Right p
  24         (Left e, _) = Left e
  25
  26 lexProgram :: Int Int -> Parser Char [Token]
  27 lexProgram line column = lexToken >>= \t->case t of
  28         LexEOF = pure []
  29         LexNL = lexProgram (line+1) 1
  30         (LexSpace l c) = lexProgram (line+l) (column+c)
  31         (LexItemError e) = fail <?>
  32                 PositionalError line column ("LexerError: " +++ e)
  33         (LexToken c t) = lexProgram line (column+c)
  34                 >>= \rest->pure [({line=line,col=column}, t):rest]
  35
  36 lexToken :: Parser Char LexItem
  37 lexToken =
  38         //Comments
  39         lexBlockComment <|> lexComment <|>
  40         //Keyword tokens
  41         lexKw "var" VarToken <|> lexKw "Void" VoidToken <|>
  42         lexKw "return" ReturnToken <|> lexKw "if" IfToken <|>
  43         lexKw "else" ElseToken <|> lexKw "while" WhileToken <|>
  44         lexKw "True" TrueToken <|> lexKw "False" FalseToken <|>
  45         lexKw "Int" IntTypeToken <|> lexKw "Bool" BoolTypeToken <|>
  46         lexKw "Char" CharTypeToken <|>
  47         //Character tokens
  48         lexEscape <|> lexCharacter <|>
  49         //Two char ops tokens
  50         lexWord "::" DoubleColonToken <|> lexWord "!=" NotEqualToken <|>
  51         lexWord "<=" LesserEqToken <|> lexWord ">=" GreaterEqToken <|>
  52         lexWord "==" EqualsToken <|> lexWord "&&" AmpersandsToken <|>
  53         lexWord "||" PipesToken <|> lexWord "[]" EmptyListToken <|>
  54         lexWord "->" ArrowToken <|>
  55         //One char ops tokens
  56     lexWord "(" BraceOpenToken <|> lexWord ")" BraceCloseToken <|>
  57         lexWord "{" CBraceOpenToken <|> lexWord "}" CBraceCloseToken <|>
  58         lexWord "[" SquareOpenToken <|> lexWord "]" SquareCloseToken <|>
  59         lexWord "," CommaToken <|> lexWord ":" ColonToken <|>
  60         lexWord ";" SColonToken <|> lexWord "." DotToken <|>
  61         lexWord "+" PlusToken <|> lexWord "*" StarToken <|>
  62         lexWord "/" SlashToken <|> lexWord "%" PercentToken <|>
  63         lexWord "=" AssignmentToken <|> lexWord "<" LesserToken <|>
  64         lexWord ">" BiggerToken <|> lexWord "!" ExclamationToken <|>
  65         lexWord "-" DashToken <|>
  66         //Number and identifier tokens
  67         lexNumber <|> lexIdentifier <|>
  68         (item '\n' >>| pure LexNL) <|>
  69         //Whitespace
  70         (satisfy isSpace >>| (pure $ LexSpace 0 1)) <|>
  71         (eof >>| pure LexEOF)
  72         where
  73                 lexWord :: String TokenValue -> Parser Char LexItem
  74                 lexWord s tv = list ls >>| pure (LexToken (length ls) tv)
  75                         where ls = fromString s
  76
  77                 lexKw :: String TokenValue -> Parser Char LexItem
  78                 lexKw kw tv = lexWord kw tv <* check (not o isIdentChar)
  79
  80                 lexComment :: Parser Char LexItem
  81                 lexComment = list (fromString "//")
  82                         >>| top until (eof <|> (item '\n' >>| pure Void)) >>| pure LexNL
  83
  84                 lexBlockComment :: Parser Char LexItem
  85                 lexBlockComment = list (fromString "/*")
  86                         >>| (top until (list (fromString "*/")))
  87                         >>= \chars->pure $ widthHeight chars 0 0
  88                         where
  89                                 widthHeight :: [Char] Int Int -> LexItem
  90                                 widthHeight [] l c = LexSpace l c
  91                                 widthHeight ['\n':xs] l _ = widthHeight xs (l+1) 0
  92                                 widthHeight [x:xs] l c = widthHeight xs l (c+1)
  93
  94                 lexNumber :: Parser Char LexItem
  95                 lexNumber = some (satisfy isDigit) >>= \si->pure $
  96                         LexToken (length si) (NumberToken $ toInt $ toString si)
  97
  98                 lexIdentifier :: Parser Char LexItem
  99                 lexIdentifier = some (satisfy isIdentChar)
 100                         >>= \si->pure $ LexToken (length si) (IdentToken $ toString si)
 101
 102                 isIdentChar c = isAlphanum c || c == '_'
 103
 104                 lexCharacter :: Parser Char LexItem
 105                 lexCharacter = item '\'' *> satisfy ((<>) '\'') <* item '\''
 106                         >>= \char->pure $ LexToken 3 (CharToken char)
 107
 108                 lexEscape :: Parser Char LexItem
 109                 lexEscape = item '\'' *> item '\\' *> top <* item '\''
 110                         >>= \char->pure case char of
 111                                 'a' = LexToken 4 (CharToken $ toChar 7)
 112                                 'b' = LexToken 4 (CharToken '\b')
 113                                 'b' = LexToken 4 (CharToken '\b')
 114                                 'f' = LexToken 4 (CharToken '\f')
 115                                 'n' = LexToken 4 (CharToken '\n')
 116                                 'r' = LexToken 4 (CharToken '\t')
 117                                 'v' = LexToken 4 (CharToken '\v')
 118                                 '\'' =LexToken 4 (CharToken '\'')
 119                                 c = (LexItemError $ "Unknown escape: " +++ toString c)