1 -----------------------------------------------------------------------------
2 -- (c) The University of Glasgow, 2006
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 -----------------------------------------------------------------------------
15 -- - parsing integers is a bit slow
16 -- - readRational is a bit slow
18 -- Known bugs, that were also in the previous version:
19 -- - M... should be 3 tokens, not 1.
20 -- - pragma-end should be only valid in a pragma
22 -- qualified operator NOTES.
24 -- - If M.(+) is a single lexeme, then..
25 -- - Probably (+) should be a single lexeme too, for consistency.
26 -- Otherwise ( + ) would be a prefix operator, but M.( + ) would not be.
27 -- - But we have to rule out reserved operators, otherwise (..) becomes
28 -- a different lexeme.
29 -- - Should we therefore also rule out reserved operators in the qualified
30 -- form? This is quite difficult to achieve. We don't do it for
34 {-# OPTIONS -Wwarn -w #-}
35 -- The above -Wwarn supression flag is a temporary kludge.
36 -- While working on this module you are encouraged to remove it and fix
37 -- any warnings in the module. See
38 -- http://hackage.haskell.org/trac/ghc/wiki/Commentary/CodingStyle#Warnings
41 -- Note that Alex itself generates code with with some unused bindings and
42 -- without type signatures, so removing the flag might not be possible.
44 {-# OPTIONS_GHC -funbox-strict-fields #-}
47 Token(..), lexer, pragState, mkPState, PState(..),
48 P(..), ParseResult(..), getSrcLoc,
50 failLocMsgP, failSpanMsgP, srcParseFail,
52 popContext, pushCurrentContext, setLastToken, setSrcLoc,
53 getLexState, popLexState, pushLexState,
54 extension, standaloneDerivingEnabled, bangPatEnabled,
68 import Util ( readRational )
76 import qualified Data.Map as Map
80 $unispace = \x05 -- Trick Alex into handling Unicode. See alexGetChar.
81 $whitechar = [\ \n\r\f\v $unispace]
82 $white_no_nl = $whitechar # \n
86 $unidigit = \x03 -- Trick Alex into handling Unicode. See alexGetChar.
87 $decdigit = $ascdigit -- for now, should really be $digit (ToDo)
88 $digit = [$ascdigit $unidigit]
90 $special = [\(\)\,\;\[\]\`\{\}]
91 $ascsymbol = [\!\#\$\%\&\*\+\.\/\<\=\>\?\@\\\^\|\-\~]
92 $unisymbol = \x04 -- Trick Alex into handling Unicode. See alexGetChar.
93 $symbol = [$ascsymbol $unisymbol] # [$special \_\:\"\']
95 $unilarge = \x01 -- Trick Alex into handling Unicode. See alexGetChar.
97 $large = [$asclarge $unilarge]
99 $unismall = \x02 -- Trick Alex into handling Unicode. See alexGetChar.
101 $small = [$ascsmall $unismall \_]
103 $unigraphic = \x06 -- Trick Alex into handling Unicode. See alexGetChar.
104 $graphic = [$small $large $symbol $digit $special $unigraphic \:\"\']
107 $hexit = [$decdigit A-F a-f]
108 $symchar = [$symbol \:]
110 $idchar = [$small $large $digit \']
112 $pragmachar = [$small $large $digit]
114 $docsym = [\| \^ \* \$]
116 @varid = $small $idchar*
117 @conid = $large $idchar*
119 @varsym = $symbol $symchar*
120 @consym = \: $symchar*
122 @decimal = $decdigit+
124 @hexadecimal = $hexit+
125 @exponent = [eE] [\-\+]? @decimal
127 -- we support the hierarchical module name extension:
130 @floating_point = @decimal \. @decimal @exponent? | @decimal @exponent
132 -- normal signed numerical literals can only be explicitly negative,
133 -- not explicitly positive (contrast @exponent)
135 @signed = @negative ?
139 -- everywhere: skip whitespace and comments
141 $tab+ { warn Opt_WarnTabs (text "Tab character") }
143 -- Everywhere: deal with nested comments. We explicitly rule out
144 -- pragmas, "{-#", so that we don't accidentally treat them as comments.
145 -- (this can happen even though pragmas will normally take precedence due to
146 -- longest-match, because pragmas aren't valid in every state, but comments
147 -- are). We also rule out nested Haddock comments, if the -haddock flag is
150 "{-" / { isNormalComment } { nested_comment lexToken }
152 -- Single-line comments are a bit tricky. Haskell 98 says that two or
153 -- more dashes followed by a symbol should be parsed as a varsym, so we
154 -- have to exclude those.
156 -- Since Haddock comments aren't valid in every state, we need to rule them
159 -- The following two rules match comments that begin with two dashes, but
160 -- continue with a different character. The rules test that this character
161 -- is not a symbol (in which case we'd have a varsym), and that it's not a
162 -- space followed by a Haddock comment symbol (docsym) (in which case we'd
163 -- have a Haddock comment). The rules then munch the rest of the line.
165 "-- " ~[$docsym \#] .* { lineCommentToken }
166 "--" [^$symbol : \ ] .* { lineCommentToken }
168 -- Next, match Haddock comments if no -haddock flag
170 "-- " [$docsym \#] .* / { ifExtension (not . haddockEnabled) } { lineCommentToken }
172 -- Now, when we've matched comments that begin with 2 dashes and continue
173 -- with a different character, we need to match comments that begin with three
174 -- or more dashes (which clearly can't be Haddock comments). We only need to
175 -- make sure that the first non-dash character isn't a symbol, and munch the
178 "---"\-* [^$symbol :] .* { lineCommentToken }
180 -- Since the previous rules all match dashes followed by at least one
181 -- character, we also need to match a whole line filled with just dashes.
183 "--"\-* / { atEOL } { lineCommentToken }
185 -- We need this rule since none of the other single line comment rules
186 -- actually match this case.
188 "-- " / { atEOL } { lineCommentToken }
190 -- 'bol' state: beginning of a line. Slurp up all the whitespace (including
191 -- blank lines) until we find a non-whitespace character, then do layout
194 -- One slight wibble here: what if the line begins with {-#? In
195 -- theory, we have to lex the pragma to see if it's one we recognise,
196 -- and if it is, then we backtrack and do_bol, otherwise we treat it
197 -- as a nested comment. We don't bother with this: if the line begins
198 -- with {-#, then we'll assume it's a pragma we know about and go for do_bol.
201 ^\# (line)? { begin line_prag1 }
202 ^\# pragma .* \n ; -- GCC 3.3 CPP generated, apparently
203 ^\# \! .* \n ; -- #!, for scripts
207 -- after a layout keyword (let, where, do, of), we begin a new layout
208 -- context if the curly brace is missing.
209 -- Careful! This stuff is quite delicate.
210 <layout, layout_do> {
211 \{ / { notFollowedBy '-' } { pop_and open_brace }
212 -- we might encounter {-# here, but {- has been handled already
214 ^\# (line)? { begin line_prag1 }
217 -- do is treated in a subtly different way, see new_layout_context
218 <layout> () { new_layout_context True }
219 <layout_do> () { new_layout_context False }
221 -- after a new layout context which was found to be to the left of the
222 -- previous context, we have generated a '{' token, and we now need to
223 -- generate a matching '}' token.
224 <layout_left> () { do_layout_left }
226 <0,option_prags> \n { begin bol }
228 "{-#" $whitechar* $pragmachar+ / { known_pragma linePrags }
229 { dispatch_pragmas linePrags }
231 -- single-line line pragmas, of the form
232 -- # <line> "<file>" <extra-stuff> \n
233 <line_prag1> $decdigit+ { setLine line_prag1a }
234 <line_prag1a> \" [$graphic \ ]* \" { setFile line_prag1b }
235 <line_prag1b> .* { pop }
237 -- Haskell-style line pragmas, of the form
238 -- {-# LINE <line> "<file>" #-}
239 <line_prag2> $decdigit+ { setLine line_prag2a }
240 <line_prag2a> \" [$graphic \ ]* \" { setFile line_prag2b }
241 <line_prag2b> "#-}"|"-}" { pop }
242 -- NOTE: accept -} at the end of a LINE pragma, for compatibility
243 -- with older versions of GHC which generated these.
246 "{-#" $whitechar* $pragmachar+
247 $whitechar+ $pragmachar+ / { known_pragma twoWordPrags }
248 { dispatch_pragmas twoWordPrags }
250 "{-#" $whitechar* $pragmachar+ / { known_pragma oneWordPrags }
251 { dispatch_pragmas oneWordPrags }
253 -- We ignore all these pragmas, but don't generate a warning for them
254 "{-#" $whitechar* $pragmachar+ / { known_pragma ignoredPrags }
255 { dispatch_pragmas ignoredPrags }
257 -- ToDo: should only be valid inside a pragma:
262 "{-#" $whitechar* $pragmachar+ / { known_pragma fileHeaderPrags }
263 { dispatch_pragmas fileHeaderPrags }
265 "-- #" { multiline_doc_comment }
269 -- In the "0" mode we ignore these pragmas
270 "{-#" $whitechar* $pragmachar+ / { known_pragma fileHeaderPrags }
271 { nested_comment lexToken }
275 "-- #" .* { lineCommentToken }
279 "{-#" { warnThen Opt_WarnUnrecognisedPragmas (text "Unrecognised pragma")
280 (nested_comment lexToken) }
283 -- '0' state: ordinary lexemes
288 "-- " $docsym / { ifExtension haddockEnabled } { multiline_doc_comment }
289 "{-" \ ? $docsym / { ifExtension haddockEnabled } { nested_doc_comment }
295 "[:" / { ifExtension parrEnabled } { token ITopabrack }
296 ":]" / { ifExtension parrEnabled } { token ITcpabrack }
300 "[|" / { ifExtension thEnabled } { token ITopenExpQuote }
301 "[e|" / { ifExtension thEnabled } { token ITopenExpQuote }
302 "[p|" / { ifExtension thEnabled } { token ITopenPatQuote }
303 "[d|" / { ifExtension thEnabled } { layout_token ITopenDecQuote }
304 "[t|" / { ifExtension thEnabled } { token ITopenTypQuote }
305 "|]" / { ifExtension thEnabled } { token ITcloseQuote }
306 \$ @varid / { ifExtension thEnabled } { skip_one_varid ITidEscape }
307 "$(" / { ifExtension thEnabled } { token ITparenEscape }
309 "[$" @varid "|" / { ifExtension qqEnabled }
310 { lex_quasiquote_tok }
314 "(|" / { ifExtension arrowsEnabled `alexAndPred` notFollowedBySymbol }
315 { special IToparenbar }
316 "|)" / { ifExtension arrowsEnabled } { special ITcparenbar }
320 \? @varid / { ifExtension ipEnabled } { skip_one_varid ITdupipvarid }
324 "(#" / { ifExtension unboxedTuplesEnabled `alexAndPred` notFollowedBySymbol }
325 { token IToubxparen }
326 "#)" / { ifExtension unboxedTuplesEnabled }
327 { token ITcubxparen }
331 "{|" / { ifExtension genericsEnabled } { token ITocurlybar }
332 "|}" / { ifExtension genericsEnabled } { token ITccurlybar }
336 \( { special IToparen }
337 \) { special ITcparen }
338 \[ { special ITobrack }
339 \] { special ITcbrack }
340 \, { special ITcomma }
341 \; { special ITsemi }
342 \` { special ITbackquote }
349 @qual @varid { idtoken qvarid }
350 @qual @conid { idtoken qconid }
352 @conid { idtoken conid }
356 @qual @varid "#"+ / { ifExtension magicHashEnabled } { idtoken qvarid }
357 @qual @conid "#"+ / { ifExtension magicHashEnabled } { idtoken qconid }
358 @varid "#"+ / { ifExtension magicHashEnabled } { varid }
359 @conid "#"+ / { ifExtension magicHashEnabled } { idtoken conid }
362 -- ToDo: - move `var` and (sym) into lexical syntax?
363 -- - remove backquote from $special?
365 @qual @varsym / { ifExtension oldQualOps } { idtoken qvarsym }
366 @qual @consym / { ifExtension oldQualOps } { idtoken qconsym }
367 @qual \( @varsym \) / { ifExtension newQualOps } { idtoken prefixqvarsym }
368 @qual \( @consym \) / { ifExtension newQualOps } { idtoken prefixqconsym }
373 -- For the normal boxed literals we need to be careful
374 -- when trying to be close to Haskell98
376 -- Normal integral literals (:: Num a => a, from Integer)
377 @decimal { tok_num positive 0 0 decimal }
378 0[oO] @octal { tok_num positive 2 2 octal }
379 0[xX] @hexadecimal { tok_num positive 2 2 hexadecimal }
381 -- Normal rational literals (:: Fractional a => a, from Rational)
382 @floating_point { strtoken tok_float }
386 -- Unboxed ints (:: Int#) and words (:: Word#)
387 -- It's simpler (and faster?) to give separate cases to the negatives,
388 -- especially considering octal/hexadecimal prefixes.
389 @decimal \# / { ifExtension magicHashEnabled } { tok_primint positive 0 1 decimal }
390 0[oO] @octal \# / { ifExtension magicHashEnabled } { tok_primint positive 2 3 octal }
391 0[xX] @hexadecimal \# / { ifExtension magicHashEnabled } { tok_primint positive 2 3 hexadecimal }
392 @negative @decimal \# / { ifExtension magicHashEnabled } { tok_primint negative 1 2 decimal }
393 @negative 0[oO] @octal \# / { ifExtension magicHashEnabled } { tok_primint negative 3 4 octal }
394 @negative 0[xX] @hexadecimal \# / { ifExtension magicHashEnabled } { tok_primint negative 3 4 hexadecimal }
396 @decimal \# \# / { ifExtension magicHashEnabled } { tok_primword 0 2 decimal }
397 0[oO] @octal \# \# / { ifExtension magicHashEnabled } { tok_primword 2 4 octal }
398 0[xX] @hexadecimal \# \# / { ifExtension magicHashEnabled } { tok_primword 2 4 hexadecimal }
400 -- Unboxed floats and doubles (:: Float#, :: Double#)
401 -- prim_{float,double} work with signed literals
402 @signed @floating_point \# / { ifExtension magicHashEnabled } { init_strtoken 1 tok_primfloat }
403 @signed @floating_point \# \# / { ifExtension magicHashEnabled } { init_strtoken 2 tok_primdouble }
406 -- Strings and chars are lexed by hand-written code. The reason is
407 -- that even if we recognise the string or char here in the regex
408 -- lexer, we would still have to parse the string afterward in order
409 -- to convert it to a String.
412 \" { lex_string_tok }
416 -- -----------------------------------------------------------------------------
420 = ITas -- Haskell keywords
444 | ITscc -- ToDo: remove (we use {-# SCC "..." #-} now)
446 | ITforall -- GHC extension keywords
464 | ITinline_prag Bool -- True <=> INLINE, False <=> NOINLINE
465 | ITinline_conlike_prag Bool -- same
466 | ITspec_prag -- SPECIALISE
467 | ITspec_inline_prag Bool -- SPECIALISE INLINE (or NOINLINE)
475 | ITcore_prag -- hdaume: core annotations
479 | IToptions_prag String
480 | ITinclude_prag String
483 | ITdotdot -- reserved symbols
499 | ITbiglam -- GHC-extension symbols
501 | ITocurly -- special symbols
503 | ITocurlybar -- {|, for type applications
504 | ITccurlybar -- |}, for type applications
508 | ITopabrack -- [:, for parallel arrays with -XParr
509 | ITcpabrack -- :], for parallel arrays with -XParr
520 | ITvarid FastString -- identifiers
522 | ITvarsym FastString
523 | ITconsym FastString
524 | ITqvarid (FastString,FastString)
525 | ITqconid (FastString,FastString)
526 | ITqvarsym (FastString,FastString)
527 | ITqconsym (FastString,FastString)
528 | ITprefixqvarsym (FastString,FastString)
529 | ITprefixqconsym (FastString,FastString)
531 | ITdupipvarid FastString -- GHC extension: implicit param: ?x
534 | ITstring FastString
536 | ITrational Rational
539 | ITprimstring FastString
542 | ITprimfloat Rational
543 | ITprimdouble Rational
545 -- Template Haskell extension tokens
546 | ITopenExpQuote -- [| or [e|
547 | ITopenPatQuote -- [p|
548 | ITopenDecQuote -- [d|
549 | ITopenTypQuote -- [t|
551 | ITidEscape FastString -- $x
552 | ITparenEscape -- $(
555 | ITquasiQuote (FastString,FastString,SrcSpan) -- [:...|...|]
557 -- Arrow notation extension
564 | ITLarrowtail -- -<<
565 | ITRarrowtail -- >>-
567 | ITunknown String -- Used when the lexer can't make sense of it
568 | ITeof -- end of file token
570 -- Documentation annotations
571 | ITdocCommentNext String -- something beginning '-- |'
572 | ITdocCommentPrev String -- something beginning '-- ^'
573 | ITdocCommentNamed String -- something beginning '-- $'
574 | ITdocSection Int String -- a section heading
575 | ITdocOptions String -- doc options (prune, ignore-exports, etc)
576 | ITdocOptionsOld String -- doc options declared "-- # ..."-style
577 | ITlineComment String -- comment starting by "--"
578 | ITblockComment String -- comment in {- -}
581 deriving Show -- debugging
585 isSpecial :: Token -> Bool
586 -- If we see M.x, where x is a keyword, but
587 -- is special, we treat is as just plain M.x,
589 isSpecial ITas = True
590 isSpecial IThiding = True
591 isSpecial ITqualified = True
592 isSpecial ITforall = True
593 isSpecial ITexport = True
594 isSpecial ITlabel = True
595 isSpecial ITdynamic = True
596 isSpecial ITsafe = True
597 isSpecial ITthreadsafe = True
598 isSpecial ITunsafe = True
599 isSpecial ITccallconv = True
600 isSpecial ITstdcallconv = True
601 isSpecial ITprimcallconv = True
602 isSpecial ITmdo = True
603 isSpecial ITfamily = True
604 isSpecial ITgroup = True
605 isSpecial ITby = True
606 isSpecial ITusing = True
610 -- the bitmap provided as the third component indicates whether the
611 -- corresponding extension keyword is valid under the extension options
612 -- provided to the compiler; if the extension corresponding to *any* of the
613 -- bits set in the bitmap is enabled, the keyword is valid (this setup
614 -- facilitates using a keyword in two different extensions that can be
615 -- activated independently)
617 reservedWordsFM :: UniqFM (Token, Int)
618 reservedWordsFM = listToUFM $
619 map (\(x, y, z) -> (mkFastString x, (y, z)))
620 [( "_", ITunderscore, 0 ),
622 ( "case", ITcase, 0 ),
623 ( "class", ITclass, 0 ),
624 ( "data", ITdata, 0 ),
625 ( "default", ITdefault, 0 ),
626 ( "deriving", ITderiving, 0 ),
628 ( "else", ITelse, 0 ),
629 ( "hiding", IThiding, 0 ),
631 ( "import", ITimport, 0 ),
633 ( "infix", ITinfix, 0 ),
634 ( "infixl", ITinfixl, 0 ),
635 ( "infixr", ITinfixr, 0 ),
636 ( "instance", ITinstance, 0 ),
638 ( "module", ITmodule, 0 ),
639 ( "newtype", ITnewtype, 0 ),
641 ( "qualified", ITqualified, 0 ),
642 ( "then", ITthen, 0 ),
643 ( "type", ITtype, 0 ),
644 ( "where", ITwhere, 0 ),
645 ( "_scc_", ITscc, 0 ), -- ToDo: remove
647 ( "forall", ITforall, bit explicitForallBit .|. bit inRulePragBit),
648 ( "mdo", ITmdo, bit recursiveDoBit),
649 ( "family", ITfamily, bit tyFamBit),
650 ( "group", ITgroup, bit transformComprehensionsBit),
651 ( "by", ITby, bit transformComprehensionsBit),
652 ( "using", ITusing, bit transformComprehensionsBit),
654 ( "foreign", ITforeign, bit ffiBit),
655 ( "export", ITexport, bit ffiBit),
656 ( "label", ITlabel, bit ffiBit),
657 ( "dynamic", ITdynamic, bit ffiBit),
658 ( "safe", ITsafe, bit ffiBit),
659 ( "threadsafe", ITthreadsafe, bit ffiBit), -- ToDo: remove
660 ( "unsafe", ITunsafe, bit ffiBit),
661 ( "stdcall", ITstdcallconv, bit ffiBit),
662 ( "ccall", ITccallconv, bit ffiBit),
663 ( "prim", ITprimcallconv, bit ffiBit),
665 ( "rec", ITrec, bit recBit),
666 ( "proc", ITproc, bit arrowsBit)
669 reservedSymsFM :: UniqFM (Token, Int -> Bool)
670 reservedSymsFM = listToUFM $
671 map (\ (x,y,z) -> (mkFastString x,(y,z)))
672 [ ("..", ITdotdot, always)
673 -- (:) is a reserved op, meaning only list cons
674 ,(":", ITcolon, always)
675 ,("::", ITdcolon, always)
676 ,("=", ITequal, always)
677 ,("\\", ITlam, always)
678 ,("|", ITvbar, always)
679 ,("<-", ITlarrow, always)
680 ,("->", ITrarrow, always)
682 ,("~", ITtilde, always)
683 ,("=>", ITdarrow, always)
684 ,("-", ITminus, always)
685 ,("!", ITbang, always)
687 -- For data T (a::*) = MkT
688 ,("*", ITstar, always) -- \i -> kindSigsEnabled i || tyFamEnabled i)
689 -- For 'forall a . t'
690 ,(".", ITdot, always) -- \i -> explicitForallEnabled i || inRulePrag i)
692 ,("-<", ITlarrowtail, arrowsEnabled)
693 ,(">-", ITrarrowtail, arrowsEnabled)
694 ,("-<<", ITLarrowtail, arrowsEnabled)
695 ,(">>-", ITRarrowtail, arrowsEnabled)
697 ,("∷", ITdcolon, unicodeSyntaxEnabled)
698 ,("⇒", ITdarrow, unicodeSyntaxEnabled)
699 ,("∀", ITforall, \i -> unicodeSyntaxEnabled i &&
700 explicitForallEnabled i)
701 ,("→", ITrarrow, unicodeSyntaxEnabled)
702 ,("←", ITlarrow, unicodeSyntaxEnabled)
703 ,("⋯", ITdotdot, unicodeSyntaxEnabled)
704 -- ToDo: ideally, → and ∷ should be "specials", so that they cannot
705 -- form part of a large operator. This would let us have a better
706 -- syntax for kinds: ɑ∷*→* would be a legal kind signature. (maybe).
709 -- -----------------------------------------------------------------------------
712 type Action = SrcSpan -> StringBuffer -> Int -> P (Located Token)
714 special :: Token -> Action
715 special tok span _buf _len = return (L span tok)
717 token, layout_token :: Token -> Action
718 token t span _buf _len = return (L span t)
719 layout_token t span _buf _len = pushLexState layout >> return (L span t)
721 idtoken :: (StringBuffer -> Int -> Token) -> Action
722 idtoken f span buf len = return (L span $! (f buf len))
724 skip_one_varid :: (FastString -> Token) -> Action
725 skip_one_varid f span buf len
726 = return (L span $! f (lexemeToFastString (stepOn buf) (len-1)))
728 strtoken :: (String -> Token) -> Action
729 strtoken f span buf len =
730 return (L span $! (f $! lexemeToString buf len))
732 init_strtoken :: Int -> (String -> Token) -> Action
733 -- like strtoken, but drops the last N character(s)
734 init_strtoken drop f span buf len =
735 return (L span $! (f $! lexemeToString buf (len-drop)))
737 begin :: Int -> Action
738 begin code _span _str _len = do pushLexState code; lexToken
741 pop _span _buf _len = do popLexState; lexToken
743 pop_and :: Action -> Action
744 pop_and act span buf len = do popLexState; act span buf len
746 {-# INLINE nextCharIs #-}
747 nextCharIs :: StringBuffer -> (Char -> Bool) -> Bool
748 nextCharIs buf p = not (atEnd buf) && p (currentChar buf)
750 notFollowedBy :: Char -> AlexAccPred Int
751 notFollowedBy char _ _ _ (AI _ _ buf)
752 = nextCharIs buf (/=char)
754 notFollowedBySymbol :: AlexAccPred Int
755 notFollowedBySymbol _ _ _ (AI _ _ buf)
756 = nextCharIs buf (`notElem` "!#$%&*+./<=>?@\\^|-~")
758 -- We must reject doc comments as being ordinary comments everywhere.
759 -- In some cases the doc comment will be selected as the lexeme due to
760 -- maximal munch, but not always, because the nested comment rule is
761 -- valid in all states, but the doc-comment rules are only valid in
762 -- the non-layout states.
763 isNormalComment :: AlexAccPred Int
764 isNormalComment bits _ _ (AI _ _ buf)
765 | haddockEnabled bits = notFollowedByDocOrPragma
766 | otherwise = nextCharIs buf (/='#')
768 notFollowedByDocOrPragma
769 = not $ spaceAndP buf (`nextCharIs` (`elem` "|^*$#"))
771 spaceAndP :: StringBuffer -> (StringBuffer -> Bool) -> Bool
772 spaceAndP buf p = p buf || nextCharIs buf (==' ') && p (snd (nextChar buf))
775 haddockDisabledAnd p bits _ _ (AI _ _ buf)
776 = if haddockEnabled bits then False else (p buf)
779 atEOL :: AlexAccPred Int
780 atEOL _ _ _ (AI _ _ buf) = atEnd buf || currentChar buf == '\n'
782 ifExtension :: (Int -> Bool) -> AlexAccPred Int
783 ifExtension pred bits _ _ _ = pred bits
785 multiline_doc_comment :: Action
786 multiline_doc_comment span buf _len = withLexedDocType (worker "")
788 worker commentAcc input docType oneLine = case alexGetChar input of
790 | oneLine -> docCommentEnd input commentAcc docType buf span
791 | otherwise -> case checkIfCommentLine input' of
792 Just input -> worker ('\n':commentAcc) input docType False
793 Nothing -> docCommentEnd input commentAcc docType buf span
794 Just (c, input) -> worker (c:commentAcc) input docType oneLine
795 Nothing -> docCommentEnd input commentAcc docType buf span
797 checkIfCommentLine input = check (dropNonNewlineSpace input)
799 check input = case alexGetChar input of
800 Just ('-', input) -> case alexGetChar input of
801 Just ('-', input) -> case alexGetChar input of
802 Just (c, _) | c /= '-' -> Just input
807 dropNonNewlineSpace input = case alexGetChar input of
809 | isSpace c && c /= '\n' -> dropNonNewlineSpace input'
813 lineCommentToken :: Action
814 lineCommentToken span buf len = do
815 b <- extension rawTokenStreamEnabled
816 if b then strtoken ITlineComment span buf len else lexToken
819 nested comments require traversing by hand, they can't be parsed
820 using regular expressions.
822 nested_comment :: P (Located Token) -> Action
823 nested_comment cont span _str _len = do
827 go commentAcc 0 input = do setInput input
828 b <- extension rawTokenStreamEnabled
830 then docCommentEnd input commentAcc ITblockComment _str span
832 go commentAcc n input = case alexGetChar input of
833 Nothing -> errBrace input span
834 Just ('-',input) -> case alexGetChar input of
835 Nothing -> errBrace input span
836 Just ('\125',input) -> go commentAcc (n-1) input
837 Just (_,_) -> go ('-':commentAcc) n input
838 Just ('\123',input) -> case alexGetChar input of
839 Nothing -> errBrace input span
840 Just ('-',input) -> go ('-':'\123':commentAcc) (n+1) input
841 Just (_,_) -> go ('\123':commentAcc) n input
842 Just (c,input) -> go (c:commentAcc) n input
844 nested_doc_comment :: Action
845 nested_doc_comment span buf _len = withLexedDocType (go "")
847 go commentAcc input docType _ = case alexGetChar input of
848 Nothing -> errBrace input span
849 Just ('-',input) -> case alexGetChar input of
850 Nothing -> errBrace input span
851 Just ('\125',input) ->
852 docCommentEnd input commentAcc docType buf span
853 Just (_,_) -> go ('-':commentAcc) input docType False
854 Just ('\123', input) -> case alexGetChar input of
855 Nothing -> errBrace input span
856 Just ('-',input) -> do
858 let cont = do input <- getInput; go commentAcc input docType False
859 nested_comment cont span buf _len
860 Just (_,_) -> go ('\123':commentAcc) input docType False
861 Just (c,input) -> go (c:commentAcc) input docType False
863 withLexedDocType :: (AlexInput -> (String -> Token) -> Bool -> P (Located Token))
865 withLexedDocType lexDocComment = do
866 input@(AI _ _ buf) <- getInput
867 case prevChar buf ' ' of
868 '|' -> lexDocComment input ITdocCommentNext False
869 '^' -> lexDocComment input ITdocCommentPrev False
870 '$' -> lexDocComment input ITdocCommentNamed False
871 '*' -> lexDocSection 1 input
872 '#' -> lexDocComment input ITdocOptionsOld False
873 _ -> panic "withLexedDocType: Bad doc type"
875 lexDocSection n input = case alexGetChar input of
876 Just ('*', input) -> lexDocSection (n+1) input
877 Just (_, _) -> lexDocComment input (ITdocSection n) True
878 Nothing -> do setInput input; lexToken -- eof reached, lex it normally
880 -- RULES pragmas turn on the forall and '.' keywords, and we turn them
881 -- off again at the end of the pragma.
883 rulePrag span _buf _len = do
884 setExts (.|. bit inRulePragBit)
885 return (L span ITrules_prag)
888 endPrag span _buf _len = do
889 setExts (.&. complement (bit inRulePragBit))
890 return (L span ITclose_prag)
893 -------------------------------------------------------------------------------
894 -- This function is quite tricky. We can't just return a new token, we also
895 -- need to update the state of the parser. Why? Because the token is longer
896 -- than what was lexed by Alex, and the lexToken function doesn't know this, so
897 -- it writes the wrong token length to the parser state. This function is
898 -- called afterwards, so it can just update the state.
900 -- This is complicated by the fact that Haddock tokens can span multiple lines,
901 -- which is something that the original lexer didn't account for.
902 -- I have added last_line_len in the parser state which represents the length
903 -- of the part of the token that is on the last line. It is now used for layout
904 -- calculation in pushCurrentContext instead of last_len. last_len is, like it
905 -- was before, the full length of the token, and it is now only used for error
908 docCommentEnd :: AlexInput -> String -> (String -> Token) -> StringBuffer ->
909 SrcSpan -> P (Located Token)
910 docCommentEnd input commentAcc docType buf span = do
912 let (AI loc last_offs nextBuf) = input
913 comment = reverse commentAcc
914 span' = mkSrcSpan (srcSpanStart span) loc
915 last_len = byteDiff buf nextBuf
917 last_line_len = if (last_offs - last_len < 0)
921 span `seq` setLastToken span' last_len last_line_len
922 return (L span' (docType comment))
924 errBrace :: AlexInput -> SrcSpan -> P a
925 errBrace (AI end _ _) span = failLocMsgP (srcSpanStart span) end "unterminated `{-'"
927 open_brace, close_brace :: Action
928 open_brace span _str _len = do
930 setContext (NoLayout:ctx)
931 return (L span ITocurly)
932 close_brace span _str _len = do
934 return (L span ITccurly)
936 qvarid, qconid :: StringBuffer -> Int -> Token
937 qvarid buf len = ITqvarid $! splitQualName buf len False
938 qconid buf len = ITqconid $! splitQualName buf len False
940 splitQualName :: StringBuffer -> Int -> Bool -> (FastString,FastString)
941 -- takes a StringBuffer and a length, and returns the module name
942 -- and identifier parts of a qualified name. Splits at the *last* dot,
943 -- because of hierarchical module names.
944 splitQualName orig_buf len parens = split orig_buf orig_buf
947 | orig_buf `byteDiff` buf >= len = done dot_buf
948 | c == '.' = found_dot buf'
949 | otherwise = split buf' dot_buf
951 (c,buf') = nextChar buf
953 -- careful, we might get names like M....
954 -- so, if the character after the dot is not upper-case, this is
955 -- the end of the qualifier part.
956 found_dot buf -- buf points after the '.'
957 | isUpper c = split buf' buf
958 | otherwise = done buf
960 (c,buf') = nextChar buf
963 (lexemeToFastString orig_buf (qual_size - 1),
964 if parens -- Prelude.(+)
965 then lexemeToFastString (stepOn dot_buf) (len - qual_size - 2)
966 else lexemeToFastString dot_buf (len - qual_size))
968 qual_size = orig_buf `byteDiff` dot_buf
973 case lookupUFM reservedWordsFM fs of
974 Just (keyword,0) -> do
976 return (L span keyword)
977 Just (keyword,exts) -> do
978 b <- extension (\i -> exts .&. i /= 0)
979 if b then do maybe_layout keyword
980 return (L span keyword)
981 else return (L span (ITvarid fs))
982 _other -> return (L span (ITvarid fs))
984 fs = lexemeToFastString buf len
986 conid :: StringBuffer -> Int -> Token
987 conid buf len = ITconid fs
988 where fs = lexemeToFastString buf len
990 qvarsym, qconsym, prefixqvarsym, prefixqconsym :: StringBuffer -> Int -> Token
991 qvarsym buf len = ITqvarsym $! splitQualName buf len False
992 qconsym buf len = ITqconsym $! splitQualName buf len False
993 prefixqvarsym buf len = ITprefixqvarsym $! splitQualName buf len True
994 prefixqconsym buf len = ITprefixqconsym $! splitQualName buf len True
996 varsym, consym :: Action
997 varsym = sym ITvarsym
998 consym = sym ITconsym
1000 sym :: (FastString -> Token) -> SrcSpan -> StringBuffer -> Int
1001 -> P (Located Token)
1002 sym con span buf len =
1003 case lookupUFM reservedSymsFM fs of
1004 Just (keyword,exts) -> do
1006 if b then return (L span keyword)
1007 else return (L span $! con fs)
1008 _other -> return (L span $! con fs)
1010 fs = lexemeToFastString buf len
1012 -- Variations on the integral numeric literal.
1013 tok_integral :: (Integer -> Token)
1014 -> (Integer -> Integer)
1015 -- -> (StringBuffer -> StringBuffer) -> (Int -> Int)
1017 -> (Integer, (Char->Int)) -> Action
1018 tok_integral itint transint transbuf translen (radix,char_to_int) span buf len =
1019 return $ L span $ itint $! transint $ parseUnsignedInteger
1020 (offsetBytes transbuf buf) (subtract translen len) radix char_to_int
1022 -- some conveniences for use with tok_integral
1023 tok_num :: (Integer -> Integer)
1025 -> (Integer, (Char->Int)) -> Action
1026 tok_num = tok_integral ITinteger
1027 tok_primint :: (Integer -> Integer)
1029 -> (Integer, (Char->Int)) -> Action
1030 tok_primint = tok_integral ITprimint
1031 tok_primword :: Int -> Int
1032 -> (Integer, (Char->Int)) -> Action
1033 tok_primword = tok_integral ITprimword positive
1034 positive, negative :: (Integer -> Integer)
1037 decimal, octal, hexadecimal :: (Integer, Char -> Int)
1038 decimal = (10,octDecDigit)
1039 octal = (8,octDecDigit)
1040 hexadecimal = (16,hexDigit)
1042 -- readRational can understand negative rationals, exponents, everything.
1043 tok_float, tok_primfloat, tok_primdouble :: String -> Token
1044 tok_float str = ITrational $! readRational str
1045 tok_primfloat str = ITprimfloat $! readRational str
1046 tok_primdouble str = ITprimdouble $! readRational str
1048 -- -----------------------------------------------------------------------------
1049 -- Layout processing
1051 -- we're at the first token on a line, insert layout tokens if necessary
1053 do_bol span _str _len = do
1057 --trace "layout: inserting '}'" $ do
1059 -- do NOT pop the lex state, we might have a ';' to insert
1060 return (L span ITvccurly)
1062 --trace "layout: inserting ';'" $ do
1064 return (L span ITsemi)
1069 -- certain keywords put us in the "layout" state, where we might
1070 -- add an opening curly brace.
1071 maybe_layout :: Token -> P ()
1072 maybe_layout ITdo = pushLexState layout_do
1073 maybe_layout ITmdo = pushLexState layout_do
1074 maybe_layout ITof = pushLexState layout
1075 maybe_layout ITlet = pushLexState layout
1076 maybe_layout ITwhere = pushLexState layout
1077 maybe_layout ITrec = pushLexState layout
1078 maybe_layout _ = return ()
1080 -- Pushing a new implicit layout context. If the indentation of the
1081 -- next token is not greater than the previous layout context, then
1082 -- Haskell 98 says that the new layout context should be empty; that is
1083 -- the lexer must generate {}.
1085 -- We are slightly more lenient than this: when the new context is started
1086 -- by a 'do', then we allow the new context to be at the same indentation as
1087 -- the previous context. This is what the 'strict' argument is for.
1089 new_layout_context :: Bool -> Action
1090 new_layout_context strict span _buf _len = do
1092 (AI _ offset _) <- getInput
1095 Layout prev_off : _ |
1096 (strict && prev_off >= offset ||
1097 not strict && prev_off > offset) -> do
1098 -- token is indented to the left of the previous context.
1099 -- we must generate a {} sequence now.
1100 pushLexState layout_left
1101 return (L span ITvocurly)
1103 setContext (Layout offset : ctx)
1104 return (L span ITvocurly)
1106 do_layout_left :: Action
1107 do_layout_left span _buf _len = do
1109 pushLexState bol -- we must be at the start of a line
1110 return (L span ITvccurly)
1112 -- -----------------------------------------------------------------------------
1115 setLine :: Int -> Action
1116 setLine code span buf len = do
1117 let line = parseUnsignedInteger buf len 10 octDecDigit
1118 setSrcLoc (mkSrcLoc (srcSpanFile span) (fromIntegral line - 1) 0)
1119 -- subtract one: the line number refers to the *following* line
1124 setFile :: Int -> Action
1125 setFile code span buf len = do
1126 let file = lexemeToFastString (stepOn buf) (len-2)
1127 setSrcLoc (mkSrcLoc file (srcSpanEndLine span) (srcSpanEndCol span))
1133 -- -----------------------------------------------------------------------------
1134 -- Options, includes and language pragmas.
1136 lex_string_prag :: (String -> Token) -> Action
1137 lex_string_prag mkTok span _buf _len
1138 = do input <- getInput
1142 return (L (mkSrcSpan start end) tok)
1144 = if isString input "#-}"
1145 then do setInput input
1146 return (mkTok (reverse acc))
1147 else case alexGetChar input of
1148 Just (c,i) -> go (c:acc) i
1149 Nothing -> err input
1150 isString _ [] = True
1152 = case alexGetChar i of
1153 Just (c,i') | c == x -> isString i' xs
1155 err (AI end _ _) = failLocMsgP (srcSpanStart span) end "unterminated options pragma"
1158 -- -----------------------------------------------------------------------------
1161 -- This stuff is horrible. I hates it.
1163 lex_string_tok :: Action
1164 lex_string_tok span _buf _len = do
1165 tok <- lex_string ""
1167 return (L (mkSrcSpan (srcSpanStart span) end) tok)
1169 lex_string :: String -> P Token
1172 case alexGetChar' i of
1173 Nothing -> lit_error
1177 magicHash <- extension magicHashEnabled
1181 case alexGetChar' i of
1185 then failMsgP "primitive string literal must contain only characters <= \'\\xFF\'"
1186 else let s' = mkZFastString (reverse s) in
1187 return (ITprimstring s')
1188 -- mkZFastString is a hack to avoid encoding the
1189 -- string in UTF-8. We just want the exact bytes.
1191 return (ITstring (mkFastString (reverse s)))
1193 return (ITstring (mkFastString (reverse s)))
1196 | Just ('&',i) <- next -> do
1197 setInput i; lex_string s
1198 | Just (c,i) <- next, is_space c -> do
1199 setInput i; lex_stringgap s
1200 where next = alexGetChar' i
1206 lex_stringgap :: String -> P Token
1207 lex_stringgap s = do
1210 '\\' -> lex_string s
1211 c | is_space c -> lex_stringgap s
1215 lex_char_tok :: Action
1216 -- Here we are basically parsing character literals, such as 'x' or '\n'
1217 -- but, when Template Haskell is on, we additionally spot
1218 -- 'x and ''T, returning ITvarQuote and ITtyQuote respectively,
1219 -- but WITHOUT CONSUMING the x or T part (the parser does that).
1220 -- So we have to do two characters of lookahead: when we see 'x we need to
1221 -- see if there's a trailing quote
1222 lex_char_tok span _buf _len = do -- We've seen '
1223 i1 <- getInput -- Look ahead to first character
1224 let loc = srcSpanStart span
1225 case alexGetChar' i1 of
1226 Nothing -> lit_error
1228 Just ('\'', i2@(AI end2 _ _)) -> do -- We've seen ''
1229 th_exts <- extension thEnabled
1232 return (L (mkSrcSpan loc end2) ITtyQuote)
1235 Just ('\\', i2@(AI _end2 _ _)) -> do -- We've seen 'backslash
1237 lit_ch <- lex_escape
1238 mc <- getCharOrFail -- Trailing quote
1239 if mc == '\'' then finish_char_tok loc lit_ch
1240 else do setInput i2; lit_error
1242 Just (c, i2@(AI _end2 _ _))
1243 | not (isAny c) -> lit_error
1246 -- We've seen 'x, where x is a valid character
1247 -- (i.e. not newline etc) but not a quote or backslash
1248 case alexGetChar' i2 of -- Look ahead one more character
1249 Just ('\'', i3) -> do -- We've seen 'x'
1251 finish_char_tok loc c
1252 _other -> do -- We've seen 'x not followed by quote
1253 -- (including the possibility of EOF)
1254 -- If TH is on, just parse the quote only
1255 th_exts <- extension thEnabled
1256 let (AI end _ _) = i1
1257 if th_exts then return (L (mkSrcSpan loc end) ITvarQuote)
1258 else do setInput i2; lit_error
1260 finish_char_tok :: SrcLoc -> Char -> P (Located Token)
1261 finish_char_tok loc ch -- We've already seen the closing quote
1262 -- Just need to check for trailing #
1263 = do magicHash <- extension magicHashEnabled
1264 i@(AI end _ _) <- getInput
1265 if magicHash then do
1266 case alexGetChar' i of
1267 Just ('#',i@(AI end _ _)) -> do
1269 return (L (mkSrcSpan loc end) (ITprimchar ch))
1271 return (L (mkSrcSpan loc end) (ITchar ch))
1273 return (L (mkSrcSpan loc end) (ITchar ch))
1275 lex_char :: Char -> AlexInput -> P Char
1278 '\\' -> do setInput inp; lex_escape
1279 c | isAny c -> do setInput inp; return c
1282 isAny :: Char -> Bool
1283 isAny c | c > '\x7f' = isPrint c
1284 | otherwise = is_any c
1286 lex_escape :: P Char
1300 '^' -> do c <- getCharOrFail
1301 if c >= '@' && c <= '_'
1302 then return (chr (ord c - ord '@'))
1305 'x' -> readNum is_hexdigit 16 hexDigit
1306 'o' -> readNum is_octdigit 8 octDecDigit
1307 x | is_decdigit x -> readNum2 is_decdigit 10 octDecDigit (octDecDigit x)
1311 case alexGetChar' i of
1312 Nothing -> lit_error
1314 case alexGetChar' i2 of
1315 Nothing -> do setInput i2; lit_error
1317 let str = [c1,c2,c3] in
1318 case [ (c,rest) | (p,c) <- silly_escape_chars,
1319 Just rest <- [stripPrefix p str] ] of
1320 (escape_char,[]):_ -> do
1323 (escape_char,_:_):_ -> do
1328 readNum :: (Char -> Bool) -> Int -> (Char -> Int) -> P Char
1329 readNum is_digit base conv = do
1333 then readNum2 is_digit base conv (conv c)
1334 else do setInput i; lit_error
1336 readNum2 :: (Char -> Bool) -> Int -> (Char -> Int) -> Int -> P Char
1337 readNum2 is_digit base conv i = do
1340 where read i input = do
1341 case alexGetChar' input of
1342 Just (c,input') | is_digit c -> do
1343 read (i*base + conv c) input'
1345 if i >= 0 && i <= 0x10FFFF
1346 then do setInput input; return (chr i)
1349 silly_escape_chars :: [(String, Char)]
1350 silly_escape_chars = [
1387 -- before calling lit_error, ensure that the current input is pointing to
1388 -- the position of the error in the buffer. This is so that we can report
1389 -- a correct location to the user, but also so we can detect UTF-8 decoding
1390 -- errors if they occur.
1392 lit_error = lexError "lexical error in string/character literal"
1394 getCharOrFail :: P Char
1397 case alexGetChar' i of
1398 Nothing -> lexError "unexpected end-of-file in string/character literal"
1399 Just (c,i) -> do setInput i; return c
1401 -- -----------------------------------------------------------------------------
1404 lex_quasiquote_tok :: Action
1405 lex_quasiquote_tok span buf len = do
1406 let quoter = reverse $ takeWhile (/= '$')
1407 $ reverse $ lexemeToString buf (len - 1)
1408 quoteStart <- getSrcLoc
1409 quote <- lex_quasiquote ""
1411 return (L (mkSrcSpan (srcSpanStart span) end)
1412 (ITquasiQuote (mkFastString quoter,
1413 mkFastString (reverse quote),
1414 mkSrcSpan quoteStart end)))
1416 lex_quasiquote :: String -> P String
1417 lex_quasiquote s = do
1419 case alexGetChar' i of
1420 Nothing -> lit_error
1423 | Just ('|',i) <- next -> do
1424 setInput i; lex_quasiquote ('|' : s)
1425 | Just (']',i) <- next -> do
1426 setInput i; lex_quasiquote (']' : s)
1427 where next = alexGetChar' i
1430 | Just (']',i) <- next -> do
1431 setInput i; return s
1432 where next = alexGetChar' i
1435 setInput i; lex_quasiquote (c : s)
1437 -- -----------------------------------------------------------------------------
1440 warn :: DynFlag -> SDoc -> Action
1441 warn option warning srcspan _buf _len = do
1442 addWarning option srcspan warning
1445 warnThen :: DynFlag -> SDoc -> Action -> Action
1446 warnThen option warning action srcspan buf len = do
1447 addWarning option srcspan warning
1448 action srcspan buf len
1450 -- -----------------------------------------------------------------------------
1461 SrcSpan -- The start and end of the text span related to
1462 -- the error. Might be used in environments which can
1463 -- show this span, e.g. by highlighting it.
1464 Message -- The error message
1466 data PState = PState {
1467 buffer :: StringBuffer,
1469 messages :: Messages,
1470 last_loc :: SrcSpan, -- pos of previous token
1471 last_offs :: !Int, -- offset of the previous token from the
1472 -- beginning of the current line.
1473 -- \t is equal to 8 spaces.
1474 last_len :: !Int, -- len of previous token
1475 last_line_len :: !Int,
1476 loc :: SrcLoc, -- current loc (end of prev token + 1)
1477 extsBitmap :: !Int, -- bitmap that determines permitted extensions
1478 context :: [LayoutContext],
1481 -- last_loc and last_len are used when generating error messages,
1482 -- and in pushCurrentContext only. Sigh, if only Happy passed the
1483 -- current token to happyError, we could at least get rid of last_len.
1484 -- Getting rid of last_loc would require finding another way to
1485 -- implement pushCurrentContext (which is only called from one place).
1487 newtype P a = P { unP :: PState -> ParseResult a }
1489 instance Monad P where
1495 returnP a = a `seq` (P $ \s -> POk s a)
1497 thenP :: P a -> (a -> P b) -> P b
1498 (P m) `thenP` k = P $ \ s ->
1500 POk s1 a -> (unP (k a)) s1
1501 PFailed span err -> PFailed span err
1503 failP :: String -> P a
1504 failP msg = P $ \s -> PFailed (last_loc s) (text msg)
1506 failMsgP :: String -> P a
1507 failMsgP msg = P $ \s -> PFailed (last_loc s) (text msg)
1509 failLocMsgP :: SrcLoc -> SrcLoc -> String -> P a
1510 failLocMsgP loc1 loc2 str = P $ \_ -> PFailed (mkSrcSpan loc1 loc2) (text str)
1512 failSpanMsgP :: SrcSpan -> SDoc -> P a
1513 failSpanMsgP span msg = P $ \_ -> PFailed span msg
1515 getPState :: P PState
1516 getPState = P $ \s -> POk s s
1518 extension :: (Int -> Bool) -> P Bool
1519 extension p = P $ \s -> POk s (p $! extsBitmap s)
1522 getExts = P $ \s -> POk s (extsBitmap s)
1524 setExts :: (Int -> Int) -> P ()
1525 setExts f = P $ \s -> POk s{ extsBitmap = f (extsBitmap s) } ()
1527 setSrcLoc :: SrcLoc -> P ()
1528 setSrcLoc new_loc = P $ \s -> POk s{loc=new_loc} ()
1530 getSrcLoc :: P SrcLoc
1531 getSrcLoc = P $ \s@(PState{ loc=loc }) -> POk s loc
1533 setLastToken :: SrcSpan -> Int -> Int -> P ()
1534 setLastToken loc len line_len = P $ \s -> POk s {
1537 last_line_len=line_len
1540 data AlexInput = AI SrcLoc {-#UNPACK#-}!Int StringBuffer
1542 alexInputPrevChar :: AlexInput -> Char
1543 alexInputPrevChar (AI _ _ buf) = prevChar buf '\n'
1545 alexGetChar :: AlexInput -> Maybe (Char,AlexInput)
1546 alexGetChar (AI loc ofs s)
1548 | otherwise = adj_c `seq` loc' `seq` ofs' `seq` s' `seq`
1549 --trace (show (ord c)) $
1550 Just (adj_c, (AI loc' ofs' s'))
1551 where (c,s') = nextChar s
1552 loc' = advanceSrcLoc loc c
1553 ofs' = advanceOffs c ofs
1561 other_graphic = '\x6'
1564 | c <= '\x06' = non_graphic
1566 -- Alex doesn't handle Unicode, so when Unicode
1567 -- character is encountered we output these values
1568 -- with the actual character value hidden in the state.
1570 case generalCategory c of
1571 UppercaseLetter -> upper
1572 LowercaseLetter -> lower
1573 TitlecaseLetter -> upper
1574 ModifierLetter -> other_graphic
1575 OtherLetter -> lower -- see #1103
1576 NonSpacingMark -> other_graphic
1577 SpacingCombiningMark -> other_graphic
1578 EnclosingMark -> other_graphic
1579 DecimalNumber -> digit
1580 LetterNumber -> other_graphic
1581 OtherNumber -> other_graphic
1582 ConnectorPunctuation -> symbol
1583 DashPunctuation -> symbol
1584 OpenPunctuation -> other_graphic
1585 ClosePunctuation -> other_graphic
1586 InitialQuote -> other_graphic
1587 FinalQuote -> other_graphic
1588 OtherPunctuation -> symbol
1589 MathSymbol -> symbol
1590 CurrencySymbol -> symbol
1591 ModifierSymbol -> symbol
1592 OtherSymbol -> symbol
1594 _other -> non_graphic
1596 -- This version does not squash unicode characters, it is used when
1598 alexGetChar' :: AlexInput -> Maybe (Char,AlexInput)
1599 alexGetChar' (AI loc ofs s)
1601 | otherwise = c `seq` loc' `seq` ofs' `seq` s' `seq`
1602 --trace (show (ord c)) $
1603 Just (c, (AI loc' ofs' s'))
1604 where (c,s') = nextChar s
1605 loc' = advanceSrcLoc loc c
1606 ofs' = advanceOffs c ofs
1608 advanceOffs :: Char -> Int -> Int
1609 advanceOffs '\n' _ = 0
1610 advanceOffs '\t' offs = (offs `quot` 8 + 1) * 8
1611 advanceOffs _ offs = offs + 1
1613 getInput :: P AlexInput
1614 getInput = P $ \s@PState{ loc=l, last_offs=o, buffer=b } -> POk s (AI l o b)
1616 setInput :: AlexInput -> P ()
1617 setInput (AI l o b) = P $ \s -> POk s{ loc=l, last_offs=o, buffer=b } ()
1619 pushLexState :: Int -> P ()
1620 pushLexState ls = P $ \s@PState{ lex_state=l } -> POk s{lex_state=ls:l} ()
1622 popLexState :: P Int
1623 popLexState = P $ \s@PState{ lex_state=ls:l } -> POk s{ lex_state=l } ls
1625 getLexState :: P Int
1626 getLexState = P $ \s@PState{ lex_state=ls:_ } -> POk s ls
1628 -- for reasons of efficiency, flags indicating language extensions (eg,
1629 -- -fglasgow-exts or -XParr) are represented by a bitmap stored in an unboxed
1633 genericsBit = 0 -- {| and |}
1644 explicitForallBit :: Int
1645 explicitForallBit = 7 -- the 'forall' keyword and '.' symbol
1647 bangPatBit = 8 -- Tells the parser to understand bang-patterns
1648 -- (doesn't affect the lexer)
1650 tyFamBit = 9 -- indexed type families: 'family' keyword and kind sigs
1652 haddockBit = 10 -- Lex and parse Haddock comments
1654 magicHashBit = 11 -- "#" in both functions and operators
1656 kindSigsBit = 12 -- Kind signatures on type variables
1657 recursiveDoBit :: Int
1658 recursiveDoBit = 13 -- mdo
1659 unicodeSyntaxBit :: Int
1660 unicodeSyntaxBit = 14 -- the forall symbol, arrow symbols, etc
1661 unboxedTuplesBit :: Int
1662 unboxedTuplesBit = 15 -- (# and #)
1663 standaloneDerivingBit :: Int
1664 standaloneDerivingBit = 16 -- standalone instance deriving declarations
1665 transformComprehensionsBit :: Int
1666 transformComprehensionsBit = 17
1668 qqBit = 18 -- enable quasiquoting
1669 inRulePragBit :: Int
1671 rawTokenStreamBit :: Int
1672 rawTokenStreamBit = 20 -- producing a token stream with all comments included
1673 newQualOpsBit :: Int
1674 newQualOpsBit = 21 -- Haskell' qualified operator syntax, e.g. Prelude.(+)
1678 always :: Int -> Bool
1680 genericsEnabled :: Int -> Bool
1681 genericsEnabled flags = testBit flags genericsBit
1682 parrEnabled :: Int -> Bool
1683 parrEnabled flags = testBit flags parrBit
1684 arrowsEnabled :: Int -> Bool
1685 arrowsEnabled flags = testBit flags arrowsBit
1686 thEnabled :: Int -> Bool
1687 thEnabled flags = testBit flags thBit
1688 ipEnabled :: Int -> Bool
1689 ipEnabled flags = testBit flags ipBit
1690 explicitForallEnabled :: Int -> Bool
1691 explicitForallEnabled flags = testBit flags explicitForallBit
1692 bangPatEnabled :: Int -> Bool
1693 bangPatEnabled flags = testBit flags bangPatBit
1694 -- tyFamEnabled :: Int -> Bool
1695 -- tyFamEnabled flags = testBit flags tyFamBit
1696 haddockEnabled :: Int -> Bool
1697 haddockEnabled flags = testBit flags haddockBit
1698 magicHashEnabled :: Int -> Bool
1699 magicHashEnabled flags = testBit flags magicHashBit
1700 -- kindSigsEnabled :: Int -> Bool
1701 -- kindSigsEnabled flags = testBit flags kindSigsBit
1702 unicodeSyntaxEnabled :: Int -> Bool
1703 unicodeSyntaxEnabled flags = testBit flags unicodeSyntaxBit
1704 unboxedTuplesEnabled :: Int -> Bool
1705 unboxedTuplesEnabled flags = testBit flags unboxedTuplesBit
1706 standaloneDerivingEnabled :: Int -> Bool
1707 standaloneDerivingEnabled flags = testBit flags standaloneDerivingBit
1708 qqEnabled :: Int -> Bool
1709 qqEnabled flags = testBit flags qqBit
1710 -- inRulePrag :: Int -> Bool
1711 -- inRulePrag flags = testBit flags inRulePragBit
1712 rawTokenStreamEnabled :: Int -> Bool
1713 rawTokenStreamEnabled flags = testBit flags rawTokenStreamBit
1714 newQualOps :: Int -> Bool
1715 newQualOps flags = testBit flags newQualOpsBit
1716 oldQualOps :: Int -> Bool
1717 oldQualOps flags = not (newQualOps flags)
1719 -- PState for parsing options pragmas
1721 pragState :: DynFlags -> StringBuffer -> SrcLoc -> PState
1722 pragState dynflags buf loc =
1725 messages = emptyMessages,
1727 last_loc = mkSrcSpan loc loc,
1734 lex_state = [bol, option_prags, 0]
1738 -- create a parse state
1740 mkPState :: StringBuffer -> SrcLoc -> DynFlags -> PState
1741 mkPState buf loc flags =
1745 messages = emptyMessages,
1746 last_loc = mkSrcSpan loc loc,
1751 extsBitmap = fromIntegral bitmap,
1753 lex_state = [bol, 0]
1754 -- we begin in the layout state if toplev_layout is set
1757 bitmap = genericsBit `setBitIf` dopt Opt_Generics flags
1758 .|. ffiBit `setBitIf` dopt Opt_ForeignFunctionInterface flags
1759 .|. parrBit `setBitIf` dopt Opt_PArr flags
1760 .|. arrowsBit `setBitIf` dopt Opt_Arrows flags
1761 .|. thBit `setBitIf` dopt Opt_TemplateHaskell flags
1762 .|. qqBit `setBitIf` dopt Opt_QuasiQuotes flags
1763 .|. ipBit `setBitIf` dopt Opt_ImplicitParams flags
1764 .|. explicitForallBit `setBitIf` dopt Opt_ExplicitForAll flags
1765 .|. bangPatBit `setBitIf` dopt Opt_BangPatterns flags
1766 .|. tyFamBit `setBitIf` dopt Opt_TypeFamilies flags
1767 .|. haddockBit `setBitIf` dopt Opt_Haddock flags
1768 .|. magicHashBit `setBitIf` dopt Opt_MagicHash flags
1769 .|. kindSigsBit `setBitIf` dopt Opt_KindSignatures flags
1770 .|. recursiveDoBit `setBitIf` dopt Opt_RecursiveDo flags
1771 .|. recBit `setBitIf` dopt Opt_DoRec flags
1772 .|. recBit `setBitIf` dopt Opt_Arrows flags
1773 .|. unicodeSyntaxBit `setBitIf` dopt Opt_UnicodeSyntax flags
1774 .|. unboxedTuplesBit `setBitIf` dopt Opt_UnboxedTuples flags
1775 .|. standaloneDerivingBit `setBitIf` dopt Opt_StandaloneDeriving flags
1776 .|. transformComprehensionsBit `setBitIf` dopt Opt_TransformListComp flags
1777 .|. rawTokenStreamBit `setBitIf` dopt Opt_KeepRawTokenStream flags
1778 .|. newQualOpsBit `setBitIf` dopt Opt_NewQualifiedOperators flags
1780 setBitIf :: Int -> Bool -> Int
1781 b `setBitIf` cond | cond = bit b
1784 addWarning :: DynFlag -> SrcSpan -> SDoc -> P ()
1785 addWarning option srcspan warning
1786 = P $ \s@PState{messages=(ws,es), dflags=d} ->
1787 let warning' = mkWarnMsg srcspan alwaysQualify warning
1788 ws' = if dopt option d then ws `snocBag` warning' else ws
1789 in POk s{messages=(ws', es)} ()
1791 getMessages :: PState -> Messages
1792 getMessages PState{messages=ms} = ms
1794 getContext :: P [LayoutContext]
1795 getContext = P $ \s@PState{context=ctx} -> POk s ctx
1797 setContext :: [LayoutContext] -> P ()
1798 setContext ctx = P $ \s -> POk s{context=ctx} ()
1801 popContext = P $ \ s@(PState{ buffer = buf, context = ctx,
1802 last_len = len, last_loc = last_loc }) ->
1804 (_:tl) -> POk s{ context = tl } ()
1805 [] -> PFailed last_loc (srcParseErr buf len)
1807 -- Push a new layout context at the indentation of the last token read.
1808 -- This is only used at the outer level of a module when the 'module'
1809 -- keyword is missing.
1810 pushCurrentContext :: P ()
1811 pushCurrentContext = P $ \ s@PState{ last_offs=offs, last_line_len=len, context=ctx } ->
1812 POk s{context = Layout (offs-len) : ctx} ()
1813 --trace ("off: " ++ show offs ++ ", len: " ++ show len) $ POk s{context = Layout (offs-len) : ctx} ()
1815 getOffside :: P Ordering
1816 getOffside = P $ \s@PState{last_offs=offs, context=stk} ->
1817 let ord = case stk of
1818 (Layout n:_) -> compare offs n
1822 -- ---------------------------------------------------------------------------
1823 -- Construct a parse error
1826 :: StringBuffer -- current buffer (placed just after the last token)
1827 -> Int -- length of the previous token
1830 = hcat [ if null token
1831 then ptext (sLit "parse error (possibly incorrect indentation)")
1832 else hcat [ptext (sLit "parse error on input "),
1833 char '`', text token, char '\'']
1835 where token = lexemeToString (offsetBytes (-len) buf) len
1837 -- Report a parse failure, giving the span of the previous token as
1838 -- the location of the error. This is the entry point for errors
1839 -- detected during parsing.
1841 srcParseFail = P $ \PState{ buffer = buf, last_len = len,
1842 last_loc = last_loc } ->
1843 PFailed last_loc (srcParseErr buf len)
1845 -- A lexical error is reported at a particular position in the source file,
1846 -- not over a token range.
1847 lexError :: String -> P a
1850 (AI end _ buf) <- getInput
1851 reportLexError loc end buf str
1853 -- -----------------------------------------------------------------------------
1854 -- This is the top-level function: called from the parser each time a
1855 -- new token is to be read from the input.
1857 lexer :: (Located Token -> P a) -> P a
1859 tok@(L _span _tok__) <- lexToken
1860 -- trace ("token: " ++ show tok__) $ do
1863 lexToken :: P (Located Token)
1865 inp@(AI loc1 _ buf) <- getInput
1868 case alexScanUser exts inp sc of
1870 let span = mkSrcSpan loc1 loc1
1871 setLastToken span 0 0
1872 return (L span ITeof)
1873 AlexError (AI loc2 _ buf) ->
1874 reportLexError loc1 loc2 buf "lexical error"
1875 AlexSkip inp2 _ -> do
1878 AlexToken inp2@(AI end _ buf2) _ t -> do
1880 let span = mkSrcSpan loc1 end
1881 let bytes = byteDiff buf buf2
1882 span `seq` setLastToken span bytes bytes
1885 reportLexError :: SrcLoc -> SrcLoc -> StringBuffer -> [Char] -> P a
1886 reportLexError loc1 loc2 buf str
1887 | atEnd buf = failLocMsgP loc1 loc2 (str ++ " at end of input")
1890 c = fst (nextChar buf)
1892 if c == '\0' -- decoding errors are mapped to '\0', see utf8DecodeChar#
1893 then failLocMsgP loc2 loc2 (str ++ " (UTF-8 decoding error)")
1894 else failLocMsgP loc1 loc2 (str ++ " at character " ++ show c)
1896 lexTokenStream :: StringBuffer -> SrcLoc -> DynFlags -> ParseResult [Located Token]
1897 lexTokenStream buf loc dflags = unP go initState
1898 where initState = mkPState buf loc (dopt_set (dopt_unset dflags Opt_Haddock) Opt_KeepRawTokenStream)
1900 ltok <- lexer return
1902 L _ ITeof -> return []
1903 _ -> liftM (ltok:) go
1905 linePrags = Map.singleton "line" (begin line_prag2)
1907 fileHeaderPrags = Map.fromList([("options", lex_string_prag IToptions_prag),
1908 ("options_ghc", lex_string_prag IToptions_prag),
1909 ("options_haddock", lex_string_prag ITdocOptions),
1910 ("language", token ITlanguage_prag),
1911 ("include", lex_string_prag ITinclude_prag)])
1913 ignoredPrags = Map.fromList (map ignored pragmas)
1914 where ignored opt = (opt, nested_comment lexToken)
1915 impls = ["hugs", "nhc98", "jhc", "yhc", "catch", "derive"]
1916 options_pragmas = map ("options_" ++) impls
1917 -- CFILES is a hugs-only thing.
1918 pragmas = options_pragmas ++ ["cfiles", "contract"]
1920 oneWordPrags = Map.fromList([("rules", rulePrag),
1921 ("inline", token (ITinline_prag True)),
1922 ("notinline", token (ITinline_prag False)),
1923 ("specialize", token ITspec_prag),
1924 ("source", token ITsource_prag),
1925 ("warning", token ITwarning_prag),
1926 ("deprecated", token ITdeprecated_prag),
1927 ("scc", token ITscc_prag),
1928 ("generated", token ITgenerated_prag),
1929 ("core", token ITcore_prag),
1930 ("unpack", token ITunpack_prag),
1931 ("ann", token ITann_prag)])
1933 twoWordPrags = Map.fromList([("inline conlike", token (ITinline_conlike_prag True)),
1934 ("notinline conlike", token (ITinline_conlike_prag False)),
1935 ("specialize inline", token (ITspec_inline_prag True)),
1936 ("specialize notinline", token (ITspec_inline_prag False))])
1939 dispatch_pragmas :: Map String Action -> Action
1940 dispatch_pragmas prags span buf len = case Map.lookup (clean_pragma (lexemeToString buf len)) prags of
1941 Just found -> found span buf len
1942 Nothing -> lexError "unknown pragma"
1944 known_pragma :: Map String Action -> AlexAccPred Int
1945 known_pragma prags _ _ len (AI _ _ buf) = (isJust $ Map.lookup (clean_pragma (lexemeToString (offsetBytes (- len) buf) len)) prags)
1946 && (nextCharIs buf (\c -> not (isAlphaNum c || c == '_')))
1948 clean_pragma :: String -> String
1949 clean_pragma prag = canon_ws (map toLower (unprefix prag))
1950 where unprefix prag' = case stripPrefix "{-#" prag' of
1953 canonical prag' = case prag' of
1954 "noinline" -> "notinline"
1955 "specialise" -> "specialize"
1956 "constructorlike" -> "conlike"
1958 canon_ws s = unwords (map canonical (words s))