1 -----------------------------------------------------------------------------
2 -- (c) The University of Glasgow, 2003
6 -- This is a combination of an Alex-generated lexer from a regex
7 -- definition, with some hand-coded bits.
9 -- Completely accurate information about token-spans within the source
10 -- file is maintained. Every token has a start and end SrcLoc attached to it.
12 -----------------------------------------------------------------------------
16 -- - parsing integers is a bit slow
17 -- - readRational is a bit slow
19 -- Known bugs, that were also in the previous version:
20 -- - M... should be 3 tokens, not 1.
21 -- - pragma-end should be only valid in a pragma
25 Token(..), lexer, mkPState,
26 P(..), ParseResult(..), getSrcLoc,
27 failMsgP, failLocMsgP, failSpanMsgP, srcParseFail,
28 popContext, pushCurrentContext,
31 #include "HsVersions.h"
33 import ErrUtils ( Message )
42 import Util ( maybePrefixMatch )
50 $whitechar = [\ \t\n\r\f\v\xa0]
51 $white_no_nl = $whitechar # \n
55 $digit = [$ascdigit $unidigit]
57 $special = [\(\)\,\;\[\]\`\{\}]
58 $ascsymbol = [\!\#\$\%\&\*\+\.\/\<\=\>\?\@\\\^\|\-\~]
60 $symbol = [$ascsymbol $unisymbol] # [$special \_\:\"\']
63 $asclarge = [A-Z \xc0-\xd6 \xd8-\xde]
64 $large = [$asclarge $unilarge]
67 $ascsmall = [a-z \xdf-\xf6 \xf8-\xff]
68 $small = [$ascsmall $unismall \_]
70 $graphic = [$small $large $symbol $digit $special \:\"\']
73 $hexit = [$digit A-F a-f]
74 $symchar = [$symbol \:]
76 $idchar = [$small $large $digit \']
78 @varid = $small $idchar*
79 @conid = $large $idchar*
81 @varsym = $symbol $symchar*
82 @consym = \: $symchar*
86 @hexadecimal = $hexit+
87 @exponent = [eE] [\-\+]? @decimal
89 -- we support the hierarchical module name extension:
92 @floating_point = @decimal \. @decimal @exponent? | @decimal @exponent
96 -- everywhere: skip whitespace and comments
99 -- Everywhere: deal with nested comments. We explicitly rule out
100 -- pragmas, "{-#", so that we don't accidentally treat them as comments.
101 -- (this can happen even though pragmas will normally take precedence due to
102 -- longest-match, because pragmas aren't valid in every state, but comments
104 "{-" / { notFollowedBy '#' } { nested_comment }
106 -- Single-line comments are a bit tricky. Haskell 98 says that two or
107 -- more dashes followed by a symbol should be parsed as a varsym, so we
108 -- have to exclude those.
109 -- The regex says: "munch all the characters after the dashes, as long as
110 -- the first one is not a symbol".
111 "--"\-* [^$symbol] .* ;
112 "--"\-* / { atEOL } ;
114 -- 'bol' state: beginning of a line. Slurp up all the whitespace (including
115 -- blank lines) until we find a non-whitespace character, then do layout
118 -- One slight wibble here: what if the line begins with {-#? In
119 -- theory, we have to lex the pragma to see if it's one we recognise,
120 -- and if it is, then we backtrack and do_bol, otherwise we treat it
121 -- as a nested comment. We don't bother with this: if the line begins
122 -- with {-#, then we'll assume it's a pragma we know about and go for do_bol.
125 ^\# (line)? { begin line_prag1 }
126 ^\# pragma .* \n ; -- GCC 3.3 CPP generated, apparently
130 -- after a layout keyword (let, where, do, of), we begin a new layout
131 -- context if the curly brace is missing.
132 -- Careful! This stuff is quite delicate.
133 <layout, layout_do> {
134 \{ / { notFollowedBy '-' } { pop_and open_brace }
135 -- we might encounter {-# here, but {- has been handled already
137 ^\# (line)? { begin line_prag1 }
140 -- do is treated in a subtly different way, see new_layout_context
141 <layout> () { new_layout_context True }
142 <layout_do> () { new_layout_context False }
144 -- after a new layout context which was found to be to the left of the
145 -- previous context, we have generated a '{' token, and we now need to
146 -- generate a matching '}' token.
147 <layout_left> () { do_layout_left }
149 <0,glaexts> \n { begin bol }
151 "{-#" $whitechar* (line|LINE) { begin line_prag2 }
153 -- single-line line pragmas, of the form
154 -- # <line> "<file>" <extra-stuff> \n
155 <line_prag1> $digit+ { set_line line_prag1a }
156 <line_prag1a> \" [$graphic \ ]* \" { set_file line_prag1b }
157 <line_prag1b> .* { pop }
159 -- Haskell-style line pragmas, of the form
160 -- {-# LINE <line> "<file>" #-}
161 <line_prag2> $digit+ { set_line line_prag2a }
162 <line_prag2a> \" [$graphic \ ]* \" { set_file line_prag2b }
163 <line_prag2b> "#-}"|"-}" { pop }
164 -- NOTE: accept -} at the end of a LINE pragma, for compatibility
165 -- with older versions of GHC which generated these.
168 "{-#" $whitechar* (SPECIALI[SZ]E|speciali[sz]e)
169 { token ITspecialise_prag }
170 "{-#" $whitechar* (SOURCE|source) { token ITsource_prag }
171 "{-#" $whitechar* (INLINE|inline) { token ITinline_prag }
172 "{-#" $whitechar* (NO(T?)INLINE|no(t?)inline)
173 { token ITnoinline_prag }
174 "{-#" $whitechar* (RULES|rules) { token ITrules_prag }
175 "{-#" $whitechar* (DEPRECATED|deprecated)
176 { token ITdeprecated_prag }
177 "{-#" $whitechar* (SCC|scc) { token ITscc_prag }
178 "{-#" $whitechar* (CORE|core) { token ITcore_prag }
179 "{-#" $whitechar* (UNPACK|unpack) { token ITunpack_prag }
181 "{-#" { nested_comment }
183 -- ToDo: should only be valid inside a pragma:
184 "#-}" { token ITclose_prag}
188 -- '0' state: ordinary lexemes
189 -- 'glaexts' state: glasgow extensions (postfix '#', etc.)
194 "[:" / { ifExtension parrEnabled } { token ITopabrack }
195 ":]" / { ifExtension parrEnabled } { token ITcpabrack }
199 "[|" / { ifExtension thEnabled } { token ITopenExpQuote }
200 "[e|" / { ifExtension thEnabled } { token ITopenExpQuote }
201 "[p|" / { ifExtension thEnabled } { token ITopenPatQuote }
202 "[d|" / { ifExtension thEnabled } { layout_token ITopenDecQuote }
203 "[t|" / { ifExtension thEnabled } { token ITopenTypQuote }
204 "|]" / { ifExtension thEnabled } { token ITcloseQuote }
205 \$ @varid / { ifExtension thEnabled } { skip_one_varid ITidEscape }
206 "$(" / { ifExtension thEnabled } { token ITparenEscape }
210 "(|" / { ifExtension arrowsEnabled `alexAndPred` notFollowedBySymbol }
211 { special IToparenbar }
212 "|)" / { ifExtension arrowsEnabled } { special ITcparenbar }
216 \? @varid / { ifExtension ipEnabled } { skip_one_varid ITdupipvarid }
217 \% @varid / { ifExtension ipEnabled } { skip_one_varid ITsplitipvarid }
221 "(#" / { notFollowedBySymbol } { token IToubxparen }
222 "#)" { token ITcubxparen }
223 "{|" { token ITocurlybar }
224 "|}" { token ITccurlybar }
228 \( { special IToparen }
229 \) { special ITcparen }
230 \[ { special ITobrack }
231 \] { special ITcbrack }
232 \, { special ITcomma }
233 \; { special ITsemi }
234 \` { special ITbackquote }
241 @qual @varid { check_qvarid }
242 @qual @conid { idtoken qconid }
244 @conid { idtoken conid }
247 -- after an illegal qvarid, such as 'M.let',
248 -- we back up and try again in the bad_qvarid state:
250 @conid { pop_and (idtoken conid) }
251 @qual @conid { pop_and (idtoken qconid) }
255 @qual @varid "#"+ { idtoken qvarid }
256 @qual @conid "#"+ { idtoken qconid }
257 @varid "#"+ { varid }
258 @conid "#"+ { idtoken conid }
264 @qual @varsym { idtoken qvarsym }
265 @qual @consym { idtoken qconsym }
271 @decimal { tok_decimal }
272 0[oO] @octal { tok_octal }
273 0[xX] @hexadecimal { tok_hexadecimal }
277 @decimal \# { prim_decimal }
278 0[oO] @octal \# { prim_octal }
279 0[xX] @hexadecimal \# { prim_hexadecimal }
282 <0,glaexts> @floating_point { strtoken tok_float }
283 <glaexts> @floating_point \# { init_strtoken 1 prim_float }
284 <glaexts> @floating_point \# \# { init_strtoken 2 prim_double }
286 -- Strings and chars are lexed by hand-written code. The reason is
287 -- that even if we recognise the string or char here in the regex
288 -- lexer, we would still have to parse the string afterward in order
289 -- to convert it to a String.
292 \" { lex_string_tok }
296 -- work around bug in Alex 2.0
297 #if __GLASGOW_HASKELL__ < 503
298 unsafeAt arr i = arr ! i
301 -- -----------------------------------------------------------------------------
305 = ITas -- Haskell keywords
329 | ITscc -- ToDo: remove (we use {-# SCC "..." #-} now)
331 | ITforall -- GHC extension keywords
344 | ITspecialise_prag -- Pragmas
352 | ITcore_prag -- hdaume: core annotations
356 | ITdotdot -- reserved symbols
372 | ITbiglam -- GHC-extension symbols
374 | ITocurly -- special symbols
376 | ITocurlybar -- {|, for type applications
377 | ITccurlybar -- |}, for type applications
381 | ITopabrack -- [:, for parallel arrays with -fparr
382 | ITcpabrack -- :], for parallel arrays with -fparr
393 | ITvarid FastString -- identifiers
395 | ITvarsym FastString
396 | ITconsym FastString
397 | ITqvarid (FastString,FastString)
398 | ITqconid (FastString,FastString)
399 | ITqvarsym (FastString,FastString)
400 | ITqconsym (FastString,FastString)
402 | ITdupipvarid FastString -- GHC extension: implicit param: ?x
403 | ITsplitipvarid FastString -- GHC extension: implicit param: %x
405 | ITpragma StringBuffer
408 | ITstring FastString
410 | ITrational Rational
413 | ITprimstring FastString
415 | ITprimfloat Rational
416 | ITprimdouble Rational
418 -- MetaHaskell extension tokens
419 | ITopenExpQuote -- [| or [e|
420 | ITopenPatQuote -- [p|
421 | ITopenDecQuote -- [d|
422 | ITopenTypQuote -- [t|
424 | ITidEscape FastString -- $x
425 | ITparenEscape -- $(
429 -- Arrow notation extension
436 | ITLarrowtail -- -<<
437 | ITRarrowtail -- >>-
439 | ITunknown String -- Used when the lexer can't make sense of it
440 | ITeof -- end of file token
442 deriving Show -- debugging
445 isSpecial :: Token -> Bool
446 -- If we see M.x, where x is a keyword, but
447 -- is special, we treat is as just plain M.x,
449 isSpecial ITas = True
450 isSpecial IThiding = True
451 isSpecial ITqualified = True
452 isSpecial ITforall = True
453 isSpecial ITexport = True
454 isSpecial ITlabel = True
455 isSpecial ITdynamic = True
456 isSpecial ITsafe = True
457 isSpecial ITthreadsafe = True
458 isSpecial ITunsafe = True
459 isSpecial ITccallconv = True
460 isSpecial ITstdcallconv = True
461 isSpecial ITmdo = True
464 -- the bitmap provided as the third component indicates whether the
465 -- corresponding extension keyword is valid under the extension options
466 -- provided to the compiler; if the extension corresponding to *any* of the
467 -- bits set in the bitmap is enabled, the keyword is valid (this setup
468 -- facilitates using a keyword in two different extensions that can be
469 -- activated independently)
471 reservedWordsFM = listToUFM $
472 map (\(x, y, z) -> (mkFastString x, (y, z)))
473 [( "_", ITunderscore, 0 ),
475 ( "case", ITcase, 0 ),
476 ( "class", ITclass, 0 ),
477 ( "data", ITdata, 0 ),
478 ( "default", ITdefault, 0 ),
479 ( "deriving", ITderiving, 0 ),
481 ( "else", ITelse, 0 ),
482 ( "hiding", IThiding, 0 ),
484 ( "import", ITimport, 0 ),
486 ( "infix", ITinfix, 0 ),
487 ( "infixl", ITinfixl, 0 ),
488 ( "infixr", ITinfixr, 0 ),
489 ( "instance", ITinstance, 0 ),
491 ( "module", ITmodule, 0 ),
492 ( "newtype", ITnewtype, 0 ),
494 ( "qualified", ITqualified, 0 ),
495 ( "then", ITthen, 0 ),
496 ( "type", ITtype, 0 ),
497 ( "where", ITwhere, 0 ),
498 ( "_scc_", ITscc, 0 ), -- ToDo: remove
500 ( "forall", ITforall, bit glaExtsBit),
501 ( "mdo", ITmdo, bit glaExtsBit),
503 ( "foreign", ITforeign, bit ffiBit),
504 ( "export", ITexport, bit ffiBit),
505 ( "label", ITlabel, bit ffiBit),
506 ( "dynamic", ITdynamic, bit ffiBit),
507 ( "safe", ITsafe, bit ffiBit),
508 ( "threadsafe", ITthreadsafe, bit ffiBit),
509 ( "unsafe", ITunsafe, bit ffiBit),
510 ( "stdcall", ITstdcallconv, bit ffiBit),
511 ( "ccall", ITccallconv, bit ffiBit),
512 ( "dotnet", ITdotnet, bit ffiBit),
514 ( "rec", ITrec, bit arrowsBit),
515 ( "proc", ITproc, bit arrowsBit)
518 reservedSymsFM = listToUFM $
519 map (\ (x,y,z) -> (mkFastString x,(y,z)))
520 [ ("..", ITdotdot, 0)
521 ,(":", ITcolon, 0) -- (:) is a reserved op,
522 -- meaning only list cons
535 ,("*", ITstar, bit glaExtsBit) -- For data T (a::*) = MkT
536 ,(".", ITdot, bit glaExtsBit) -- For 'forall a . t'
538 ,("-<", ITlarrowtail, bit arrowsBit)
539 ,(">-", ITrarrowtail, bit arrowsBit)
540 ,("-<<", ITLarrowtail, bit arrowsBit)
541 ,(">>-", ITRarrowtail, bit arrowsBit)
544 -- -----------------------------------------------------------------------------
547 type Action = SrcSpan -> StringBuffer -> Int -> P (Located Token)
549 special :: Token -> Action
550 special tok span _buf len = return (L span tok)
552 token, layout_token :: Token -> Action
553 token t span buf len = return (L span t)
554 layout_token t span buf len = pushLexState layout >> return (L span t)
556 idtoken :: (StringBuffer -> Int -> Token) -> Action
557 idtoken f span buf len = return (L span $! (f buf len))
559 skip_one_varid :: (FastString -> Token) -> Action
560 skip_one_varid f span buf len
561 = return (L span $! f (lexemeToFastString (stepOn buf) (len-1)))
563 strtoken :: (String -> Token) -> Action
564 strtoken f span buf len =
565 return (L span $! (f $! lexemeToString buf len))
567 init_strtoken :: Int -> (String -> Token) -> Action
568 -- like strtoken, but drops the last N character(s)
569 init_strtoken drop f span buf len =
570 return (L span $! (f $! lexemeToString buf (len-drop)))
572 begin :: Int -> Action
573 begin code _span _str _len = do pushLexState code; lexToken
576 pop _span _buf _len = do popLexState; lexToken
578 pop_and :: Action -> Action
579 pop_and act span buf len = do popLexState; act span buf len
581 notFollowedBy char _ _ _ (_,buf) = atEnd buf || currentChar buf /= char
583 notFollowedBySymbol _ _ _ (_,buf)
584 = atEnd buf || currentChar buf `notElem` "!#$%&*+./<=>?@\\^|-~"
586 atEOL _ _ _ (_,buf) = atEnd buf || currentChar buf == '\n'
588 ifExtension pred bits _ _ _ = pred bits
591 nested comments require traversing by hand, they can't be parsed
592 using regular expressions.
594 nested_comment :: Action
595 nested_comment span _str _len = do
598 where go 0 input = do setInput input; lexToken
600 case alexGetChar input of
605 case alexGetChar input of
607 Just ('\125',input) -> go (n-1) input
608 Just (c,_) -> go n input
610 case alexGetChar input of
612 Just ('-',input') -> go (n+1) input'
613 Just (c,input) -> go n input
616 err input = do failLocMsgP (srcSpanStart span) (fst input)
619 open_brace, close_brace :: Action
620 open_brace span _str _len = do
622 setContext (NoLayout:ctx)
623 return (L span ITocurly)
624 close_brace span _str _len = do
626 return (L span ITccurly)
628 -- We have to be careful not to count M.<varid> as a qualified name
629 -- when <varid> is a keyword. We hack around this by catching
630 -- the offending tokens afterward, and re-lexing in a different state.
631 check_qvarid span buf len = do
632 case lookupUFM reservedWordsFM var of
634 | not (isSpecial keyword) ->
638 b <- extension (\i -> exts .&. i /= 0)
641 _other -> return token
643 (mod,var) = splitQualName buf len
644 token = L span (ITqvarid (mod,var))
647 setInput (srcSpanStart span,buf)
648 pushLexState bad_qvarid
651 qvarid buf len = ITqvarid $! splitQualName buf len
652 qconid buf len = ITqconid $! splitQualName buf len
654 splitQualName :: StringBuffer -> Int -> (FastString,FastString)
655 -- takes a StringBuffer and a length, and returns the module name
656 -- and identifier parts of a qualified name. Splits at the *last* dot,
657 -- because of hierarchical module names.
658 splitQualName orig_buf len = split orig_buf 0 0
661 | n == len = done dot_off
662 | lookAhead buf n == '.' = split2 buf n (n+1)
663 | otherwise = split buf dot_off (n+1)
665 -- careful, we might get names like M....
666 -- so, if the character after the dot is not upper-case, this is
667 -- the end of the qualifier part.
669 | isUpper (lookAhead buf n) = split buf dot_off (n+1)
670 | otherwise = done dot_off
673 (lexemeToFastString orig_buf dot_off,
674 lexemeToFastString (stepOnBy (dot_off+1) orig_buf) (len - dot_off -1))
677 case lookupUFM reservedWordsFM fs of
678 Just (keyword,0) -> do
680 return (L span keyword)
681 Just (keyword,exts) -> do
682 b <- extension (\i -> exts .&. i /= 0)
683 if b then do maybe_layout keyword
684 return (L span keyword)
685 else return (L span (ITvarid fs))
686 _other -> return (L span (ITvarid fs))
688 fs = lexemeToFastString buf len
690 conid buf len = ITconid fs
691 where fs = lexemeToFastString buf len
693 qvarsym buf len = ITqvarsym $! splitQualName buf len
694 qconsym buf len = ITqconsym $! splitQualName buf len
696 varsym = sym ITvarsym
697 consym = sym ITconsym
699 sym con span buf len =
700 case lookupUFM reservedSymsFM fs of
701 Just (keyword,0) -> return (L span keyword)
702 Just (keyword,exts) -> do
703 b <- extension (\i -> exts .&. i /= 0)
704 if b then return (L span keyword)
705 else return (L span $! con fs)
706 _other -> return (L span $! con fs)
708 fs = lexemeToFastString buf len
710 tok_decimal span buf len
711 = return (L span (ITinteger $! parseInteger buf len 10 oct_or_dec))
713 tok_octal span buf len
714 = return (L span (ITinteger $! parseInteger (stepOnBy 2 buf) (len-2) 8 oct_or_dec))
716 tok_hexadecimal span buf len
717 = return (L span (ITinteger $! parseInteger (stepOnBy 2 buf) (len-2) 16 hex))
719 prim_decimal span buf len
720 = return (L span (ITprimint $! parseInteger buf (len-1) 10 oct_or_dec))
722 prim_octal span buf len
723 = return (L span (ITprimint $! parseInteger (stepOnBy 2 buf) (len-3) 8 oct_or_dec))
725 prim_hexadecimal span buf len
726 = return (L span (ITprimint $! parseInteger (stepOnBy 2 buf) (len-3) 16 hex))
728 tok_float str = ITrational $! readRational__ str
729 prim_float str = ITprimfloat $! readRational__ str
730 prim_double str = ITprimdouble $! readRational__ str
732 parseInteger :: StringBuffer -> Int -> Integer -> (Char->Int) -> Integer
733 parseInteger buf len radix to_int
735 where go i x | i == len = x
736 | otherwise = go (i+1) (x * radix + toInteger (to_int (lookAhead buf i)))
738 -- -----------------------------------------------------------------------------
741 -- we're at the first token on a line, insert layout tokens if necessary
743 do_bol span _str _len = do
744 pos <- getOffside (srcSpanEnd span)
747 --trace "layout: inserting '}'" $ do
749 -- do NOT pop the lex state, we might have a ';' to insert
750 return (L span ITvccurly)
752 --trace "layout: inserting ';'" $ do
754 return (L span ITsemi)
759 -- certain keywords put us in the "layout" state, where we might
760 -- add an opening curly brace.
761 maybe_layout ITdo = pushLexState layout_do
762 maybe_layout ITmdo = pushLexState layout_do
763 maybe_layout ITof = pushLexState layout
764 maybe_layout ITlet = pushLexState layout
765 maybe_layout ITwhere = pushLexState layout
766 maybe_layout ITrec = pushLexState layout
767 maybe_layout _ = return ()
769 -- Pushing a new implicit layout context. If the indentation of the
770 -- next token is not greater than the previous layout context, then
771 -- Haskell 98 says that the new layout context should be empty; that is
772 -- the lexer must generate {}.
774 -- We are slightly more lenient than this: when the new context is started
775 -- by a 'do', then we allow the new context to be at the same indentation as
776 -- the previous context. This is what the 'strict' argument is for.
778 new_layout_context strict span _buf _len = do
780 let offset = srcSpanStartCol span
783 Layout prev_off : _ |
784 (strict && prev_off >= offset ||
785 not strict && prev_off > offset) -> do
786 -- token is indented to the left of the previous context.
787 -- we must generate a {} sequence now.
788 pushLexState layout_left
789 return (L span ITvocurly)
791 setContext (Layout offset : ctx)
792 return (L span ITvocurly)
794 do_layout_left span _buf _len = do
796 pushLexState bol -- we must be at the start of a line
797 return (L span ITvccurly)
799 -- -----------------------------------------------------------------------------
802 set_line :: Int -> Action
803 set_line code span buf len = do
804 let line = parseInteger buf len 10 oct_or_dec
805 setSrcLoc (mkSrcLoc (srcSpanFile span) (fromIntegral line - 1) 0)
806 -- subtract one: the line number refers to the *following* line
811 set_file :: Int -> Action
812 set_file code span buf len = do
813 let file = lexemeToFastString (stepOn buf) (len-2)
814 setSrcLoc (mkSrcLoc file (srcSpanEndLine span) (srcSpanEndCol span))
819 -- -----------------------------------------------------------------------------
822 -- This stuff is horrible. I hates it.
824 lex_string_tok :: Action
825 lex_string_tok span buf len = do
828 return (L (mkSrcSpan (srcSpanStart span) end) tok)
830 lex_string :: String -> P Token
833 case alexGetChar i of
838 glaexts <- extension glaExtsEnabled
842 case alexGetChar i of
846 then failMsgP "primitive string literal must contain only characters <= \'\\xFF\'"
847 else let s' = mkFastStringNarrow (reverse s) in
848 -- always a narrow string/byte array
849 return (ITprimstring s')
851 return (ITstring (mkFastString (reverse s)))
853 return (ITstring (mkFastString (reverse s)))
856 | Just ('&',i) <- next -> do
857 setInput i; lex_string s
858 | Just (c,i) <- next, is_space c -> do
859 setInput i; lex_stringgap s
860 where next = alexGetChar i
870 c | is_space c -> lex_stringgap s
874 lex_char_tok :: Action
875 -- Here we are basically parsing character literals, such as 'x' or '\n'
876 -- but, when Template Haskell is on, we additionally spot
877 -- 'x and ''T, returning ITvarQuote and ITtyQuote respectively,
878 -- but WIHTOUT CONSUMING the x or T part (the parser does that).
879 -- So we have to do two characters of lookahead: when we see 'x we need to
880 -- see if there's a trailing quote
881 lex_char_tok span buf len = do -- We've seen '
882 i1 <- getInput -- Look ahead to first character
883 let loc = srcSpanStart span
884 case alexGetChar i1 of
887 Just ('\'', i2@(end2,_)) -> do -- We've seen ''
888 th_exts <- extension thEnabled
891 return (L (mkSrcSpan loc end2) ITtyQuote)
894 Just ('\\', i2@(end2,_)) -> do -- We've seen 'backslash
897 mc <- getCharOrFail -- Trailing quote
898 if mc == '\'' then finish_char_tok loc lit_ch
901 Just (c, i2@(end2,_)) | not (is_any c) -> lit_error
904 -- We've seen 'x, where x is a valid character
905 -- (i.e. not newline etc) but not a quote or backslash
906 case alexGetChar i2 of -- Look ahead one more character
908 Just ('\'', i3) -> do -- We've seen 'x'
910 finish_char_tok loc c
911 _other -> do -- We've seen 'x not followed by quote
912 -- If TH is on, just parse the quote only
913 th_exts <- extension thEnabled
914 if th_exts then return (L (mkSrcSpan loc (fst i1)) ITvarQuote)
917 finish_char_tok :: SrcLoc -> Char -> P (Located Token)
918 finish_char_tok loc ch -- We've already seen the closing quote
919 -- Just need to check for trailing #
920 = do glaexts <- extension glaExtsEnabled
921 i@(end,_) <- getInput
923 case alexGetChar i of
924 Just ('#',i@(end,_)) -> do
926 return (L (mkSrcSpan loc end) (ITprimchar ch))
928 return (L (mkSrcSpan loc end) (ITchar ch))
930 return (L (mkSrcSpan loc end) (ITchar ch))
937 c | is_any c -> return c
954 '^' -> do c <- getCharOrFail
955 if c >= '@' && c <= '_'
956 then return (chr (ord c - ord '@'))
959 'x' -> readNum is_hexdigit 16 hex
960 'o' -> readNum is_octdigit 8 oct_or_dec
961 x | is_digit x -> readNum2 is_digit 10 oct_or_dec (oct_or_dec x)
965 case alexGetChar i of
968 case alexGetChar i2 of
971 let str = [c1,c2,c3] in
972 case [ (c,rest) | (p,c) <- silly_escape_chars,
973 Just rest <- [maybePrefixMatch p str] ] of
974 (escape_char,[]):_ -> do
977 (escape_char,_:_):_ -> do
982 readNum :: (Char -> Bool) -> Int -> (Char -> Int) -> P Char
983 readNum is_digit base conv = do
986 then readNum2 is_digit base conv (conv c)
989 readNum2 is_digit base conv i = do
992 where read i input = do
993 case alexGetChar input of
994 Just (c,input') | is_digit c -> do
995 read (i*base + conv c) input'
998 if i >= 0 && i <= 0x10FFFF
1004 || (c >= 'a' && c <= 'f')
1005 || (c >= 'A' && c <= 'F')
1007 hex c | is_digit c = ord c - ord '0'
1008 | otherwise = ord (to_lower c) - ord 'a' + 10
1010 oct_or_dec c = ord c - ord '0'
1012 is_octdigit c = c >= '0' && c <= '7'
1015 | c >= 'A' && c <= 'Z' = chr (ord c - (ord 'A' - ord 'a'))
1018 silly_escape_chars = [
1055 lit_error = lexError "lexical error in string/character literal"
1057 getCharOrFail :: P Char
1060 case alexGetChar i of
1061 Nothing -> lexError "unexpected end-of-file in string/character literal"
1062 Just (c,i) -> do setInput i; return c
1064 -- -----------------------------------------------------------------------------
1067 readRational :: ReadS Rational -- NB: doesn't handle leading "-"
1069 (n,d,s) <- readFix r
1071 return ((n%1)*10^^(k-d), t)
1074 (ds,s) <- lexDecDigits r
1075 (ds',t) <- lexDotDigits s
1076 return (read (ds++ds'), length ds', t)
1078 readExp (e:s) | e `elem` "eE" = readExp' s
1079 readExp s = return (0,s)
1081 readExp' ('+':s) = readDec s
1082 readExp' ('-':s) = do
1085 readExp' s = readDec s
1088 (ds,r) <- nonnull isDigit s
1089 return (foldl1 (\n d -> n * 10 + d) [ ord d - ord '0' | d <- ds ],
1092 lexDecDigits = nonnull isDigit
1094 lexDotDigits ('.':s) = return (span isDigit s)
1095 lexDotDigits s = return ("",s)
1097 nonnull p s = do (cs@(_:_),t) <- return (span p s)
1100 readRational__ :: String -> Rational -- NB: *does* handle a leading "-"
1101 readRational__ top_s
1103 '-' : xs -> - (read_me xs)
1107 = case (do { (x,"") <- readRational s ; return x }) of
1109 [] -> error ("readRational__: no parse:" ++ top_s)
1110 _ -> error ("readRational__: ambiguous parse:" ++ top_s)
1112 -- -----------------------------------------------------------------------------
1122 SrcSpan -- The start and end of the text span related to
1123 -- the error. Might be used in environments which can
1124 -- show this span, e.g. by highlighting it.
1125 Message -- The error message
1127 data PState = PState {
1128 buffer :: StringBuffer,
1129 last_loc :: SrcSpan, -- pos of previous token
1130 last_len :: !Int, -- len of previous token
1131 loc :: SrcLoc, -- current loc (end of prev token + 1)
1132 extsBitmap :: !Int, -- bitmap that determines permitted extensions
1133 context :: [LayoutContext],
1136 -- last_loc and last_len are used when generating error messages,
1137 -- and in pushCurrentContext only.
1139 newtype P a = P { unP :: PState -> ParseResult a }
1141 instance Monad P where
1147 returnP a = P $ \s -> POk s a
1149 thenP :: P a -> (a -> P b) -> P b
1150 (P m) `thenP` k = P $ \ s ->
1152 POk s1 a -> (unP (k a)) s1
1153 PFailed span err -> PFailed span err
1155 failP :: String -> P a
1156 failP msg = P $ \s -> PFailed (last_loc s) (text msg)
1158 failMsgP :: String -> P a
1159 failMsgP msg = P $ \s -> PFailed (last_loc s) (text msg)
1161 failLocMsgP :: SrcLoc -> SrcLoc -> String -> P a
1162 failLocMsgP loc1 loc2 str = P $ \s -> PFailed (mkSrcSpan loc1 loc2) (text str)
1164 failSpanMsgP :: SrcSpan -> String -> P a
1165 failSpanMsgP span msg = P $ \s -> PFailed span (text msg)
1167 extension :: (Int -> Bool) -> P Bool
1168 extension p = P $ \s -> POk s (p $! extsBitmap s)
1171 getExts = P $ \s -> POk s (extsBitmap s)
1173 setSrcLoc :: SrcLoc -> P ()
1174 setSrcLoc new_loc = P $ \s -> POk s{loc=new_loc} ()
1176 getSrcLoc :: P SrcLoc
1177 getSrcLoc = P $ \s@(PState{ loc=loc }) -> POk s loc
1179 setLastToken :: SrcSpan -> Int -> P ()
1180 setLastToken loc len = P $ \s -> POk s{ last_loc=loc, last_len=len } ()
1182 type AlexInput = (SrcLoc,StringBuffer)
1184 alexInputPrevChar :: AlexInput -> Char
1185 alexInputPrevChar (_,s) = prevChar s '\n'
1187 alexGetChar :: AlexInput -> Maybe (Char,AlexInput)
1190 | otherwise = c `seq` loc' `seq` s' `seq` Just (c, (loc', s'))
1191 where c = currentChar s
1192 loc' = advanceSrcLoc loc c
1195 getInput :: P AlexInput
1196 getInput = P $ \s@PState{ loc=l, buffer=b } -> POk s (l,b)
1198 setInput :: AlexInput -> P ()
1199 setInput (l,b) = P $ \s -> POk s{ loc=l, buffer=b } ()
1201 pushLexState :: Int -> P ()
1202 pushLexState ls = P $ \s@PState{ lex_state=l } -> POk s{lex_state=ls:l} ()
1204 popLexState :: P Int
1205 popLexState = P $ \s@PState{ lex_state=ls:l } -> POk s{ lex_state=l } ls
1207 getLexState :: P Int
1208 getLexState = P $ \s@PState{ lex_state=ls:l } -> POk s ls
1210 -- for reasons of efficiency, flags indicating language extensions (eg,
1211 -- -fglasgow-exts or -fparr) are represented by a bitmap stored in an unboxed
1214 glaExtsBit, ffiBit, parrBit :: Int
1222 glaExtsEnabled, ffiEnabled, parrEnabled :: Int -> Bool
1223 glaExtsEnabled flags = testBit flags glaExtsBit
1224 ffiEnabled flags = testBit flags ffiBit
1225 parrEnabled flags = testBit flags parrBit
1226 arrowsEnabled flags = testBit flags arrowsBit
1227 thEnabled flags = testBit flags thBit
1228 ipEnabled flags = testBit flags ipBit
1230 -- create a parse state
1232 mkPState :: StringBuffer -> SrcLoc -> DynFlags -> PState
1233 mkPState buf loc flags =
1236 last_loc = mkSrcSpan loc loc,
1239 extsBitmap = fromIntegral bitmap,
1241 lex_state = [bol, if glaExtsEnabled bitmap then glaexts else 0]
1242 -- we begin in the layout state if toplev_layout is set
1245 bitmap = glaExtsBit `setBitIf` dopt Opt_GlasgowExts flags
1246 .|. ffiBit `setBitIf` dopt Opt_FFI flags
1247 .|. parrBit `setBitIf` dopt Opt_PArr flags
1248 .|. arrowsBit `setBitIf` dopt Opt_Arrows flags
1249 .|. thBit `setBitIf` dopt Opt_TH flags
1250 .|. ipBit `setBitIf` dopt Opt_ImplicitParams flags
1252 setBitIf :: Int -> Bool -> Int
1253 b `setBitIf` cond | cond = bit b
1256 getContext :: P [LayoutContext]
1257 getContext = P $ \s@PState{context=ctx} -> POk s ctx
1259 setContext :: [LayoutContext] -> P ()
1260 setContext ctx = P $ \s -> POk s{context=ctx} ()
1263 popContext = P $ \ s@(PState{ buffer = buf, context = ctx,
1264 loc = loc, last_len = len, last_loc = last_loc }) ->
1266 (_:tl) -> POk s{ context = tl } ()
1267 [] -> PFailed last_loc (srcParseErr buf len)
1269 -- Push a new layout context at the indentation of the last token read.
1270 -- This is only used at the outer level of a module when the 'module'
1271 -- keyword is missing.
1272 pushCurrentContext :: P ()
1273 pushCurrentContext = P $ \ s@PState{ last_loc=loc, context=ctx } ->
1274 POk s{ context = Layout (srcSpanStartCol loc) : ctx} ()
1276 getOffside :: SrcLoc -> P Ordering
1277 getOffside loc = P $ \s@PState{context=stk} ->
1278 let ord = case stk of
1279 (Layout n:_) -> compare (srcLocCol loc) n
1283 -- ---------------------------------------------------------------------------
1284 -- Construct a parse error
1287 :: StringBuffer -- current buffer (placed just after the last token)
1288 -> Int -- length of the previous token
1291 = hcat [ if null token
1292 then ptext SLIT("parse error (possibly incorrect indentation)")
1293 else hcat [ptext SLIT("parse error on input "),
1294 char '`', text token, char '\'']
1296 where token = lexemeToString (stepOnBy (-len) buf) len
1298 -- Report a parse failure, giving the span of the previous token as
1299 -- the location of the error. This is the entry point for errors
1300 -- detected during parsing.
1302 srcParseFail = P $ \PState{ buffer = buf, last_len = len,
1303 last_loc = last_loc, loc = loc } ->
1304 PFailed last_loc (srcParseErr buf len)
1306 -- A lexical error is reported at a particular position in the source file,
1307 -- not over a token range. TODO: this is slightly wrong, because we record
1308 -- the error at the character position following the one which caused the
1309 -- error. We should somehow back up by one character.
1310 lexError :: String -> P a
1313 i@(end,_) <- getInput
1314 failLocMsgP loc end str
1316 -- -----------------------------------------------------------------------------
1317 -- This is the top-level function: called from the parser each time a
1318 -- new token is to be read from the input.
1320 lexer :: (Located Token -> P a) -> P a
1322 tok@(L _ tok__) <- lexToken
1323 --trace ("token: " ++ show tok__) $ do
1326 lexToken :: P (Located Token)
1328 inp@(loc1,buf) <- getInput
1331 case alexScanUser exts inp sc of
1332 AlexEOF -> do let span = mkSrcSpan loc1 loc1
1334 return (L span ITeof)
1335 AlexError (loc2,_) -> do failLocMsgP loc1 loc2 "lexical error"
1336 AlexSkip inp2 _ -> do
1339 AlexToken inp2@(end,buf2) len t -> do
1341 let span = mkSrcSpan loc1 end
1342 span `seq` setLastToken span len