rtk-0.12: test/golden/block/BlockLexer.x
-- Generated by RTK from grammar 'Block'. Do not edit by hand.
{
{-# LANGUAGE StandaloneDeriving, DeriveDataTypeable #-}
module BlockLexer(scanTokens, alexScanTokens, Token(..), PosToken(..), AlexPosn(..))
where
import Data.Data (Data)
}
%wrapper "monad"
tokens :- "tok_Block_dummy_8" { simple Tk__tok_Block_dummy_8 }
"tok_Exp_dummy_7" { simple Tk__tok_Exp_dummy_7 }
"tok_Function_dummy_6" { simple Tk__tok_Function_dummy_6 }
"tok_Ident_dummy_5" { simple Tk__tok_Ident_dummy_5 }
"tok_ParamList_dummy_4" { simple Tk__tok_ParamList_dummy_4 }
"tok_Program_dummy_9" { simple Tk__tok_Program_dummy_9 }
"tok_Statement_dummy_3" { simple Tk__tok_Statement_dummy_3 }
"tok_StatementList_dummy_2" { simple Tk__tok_StatementList_dummy_2 }
"}" { simple Tk__tok__symbol__5 }
"{" { simple Tk__tok__symbol__4 }
"return" { simple Tk__tok_return_8 }
"fn" { simple Tk__tok_fn_0 }
"=" { simple Tk__tok__eql__6 }
";" { simple Tk__tok__semi__7 }
"," { simple Tk__tok__coma__3 }
")" { simple Tk__tok__rparen__2 }
"(" { simple Tk__tok__lparen__1 }
([\ \t\n\r]+) ;
([0-9]+) { simple1 $ Tk__num . (read) }
([a-zA-Z_] [a-zA-Z0-9_]*) { simple1 $ Tk__id . (id) }
("$" "Ident" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_Ident . ((tail . dropWhile (/= ':'))) }
("$" "Exp" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_Exp . ((tail . dropWhile (/= ':'))) }
("$" "Statement" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_Statement . ((tail . dropWhile (/= ':'))) }
("$" "StatementList" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_StatementList . ((tail . dropWhile (/= ':'))) }
("$" "Block" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_Block . ((tail . dropWhile (/= ':'))) }
("$" "ParamList" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_ParamList . ((tail . dropWhile (/= ':'))) }
("$" "Function" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_Function . ((tail . dropWhile (/= ':'))) }
("$" "Program" ":" [a-zA-Z_] [A-Za-z0-9_]*) { simple1 $ Tk__qq_Program . ((tail . dropWhile (/= ':'))) }
. { rtkError }
{
data Token = EndOfFile |
Tk__tok_Block_dummy_8 |
Tk__tok_Exp_dummy_7 |
Tk__tok_Function_dummy_6 |
Tk__tok_Ident_dummy_5 |
Tk__tok_ParamList_dummy_4 |
Tk__tok_Program_dummy_9 |
Tk__tok_Statement_dummy_3 |
Tk__tok_StatementList_dummy_2 |
Tk__tok__symbol__5 |
Tk__tok__symbol__4 |
Tk__tok_return_8 |
Tk__tok_fn_0 |
Tk__tok__eql__6 |
Tk__tok__semi__7 |
Tk__tok__coma__3 |
Tk__tok__rparen__2 |
Tk__tok__lparen__1 |
Tk__num Int |
Tk__id String |
Tk__qq_Ident String |
Tk__qq_Exp String |
Tk__qq_Statement String |
Tk__qq_StatementList String |
Tk__qq_Block String |
Tk__qq_ParamList String |
Tk__qq_Function String |
Tk__qq_Program String
deriving (Show)
-- AlexPosn is defined by the alex wrapper, so the Data instance (required by
-- the parser's RtkPos position type) can only be attached via standalone
-- deriving - the same solution as the hand-written Lexer.x
deriving instance Data AlexPosn
-- A token together with the source position where it starts
data PosToken = PosToken { ptPos :: AlexPosn, ptToken :: Token }
deriving (Show)
alexEOF = do
(pos, _, _, _) <- alexGetInput
return $ PosToken pos EndOfFile
-- Lex the input into a token stream, returning the positioned error message
-- on a lexical error (encoded as "LINE:COL:message", see rtkError below).
-- The returned list always ends with an EndOfFile token that carries the
-- position of the end of input, so parse errors at end of input can be
-- reported with a position too
scanTokens :: String -> Either String [PosToken]
scanTokens str = runAlex str $ do
let loop toks = do tok <- alexMonadScan
case tok of
PosToken _ EndOfFile -> return $ reverse (tok : toks)
_ -> let toks' = tok : toks
in toks' `seq` loop toks'
loop []
-- Thin compatibility wrapper: callers that have not switched to 'scanTokens'
-- get the error message thrown instead
alexScanTokens :: String -> [PosToken]
alexScanTokens str =
case scanTokens str of
Right toks -> toks
Left err -> errorWithoutStackTrace err
simple1 :: (String -> Token) -> AlexInput -> Int -> Alex PosToken
simple1 t (pos, _, _, str) len = return $ PosToken pos (t (take len str))
simple :: Token -> AlexInput -> Int -> Alex PosToken
simple t (pos, _, _, _) len = return $ PosToken pos t
-- Encode the position as "LINE:COL:message" so callers can split it back out
-- into a structured position - the same encoding the rtk grammar lexer uses
rtkError ((AlexPn _ line column), _, _, str) len = alexError $ (show line) ++ ":" ++ (show column) ++ ":lexical error. Following chars: " ++ (take 10 str)
}