Lexer.x 108 KB
Newer Older
1
-----------------------------------------------------------------------------
2
-- (c) The University of Glasgow, 2006
3
--
4
-- GHC's lexer for Haskell 2010 [1].
5
--
6 7
-- This is a combination of an Alex-generated lexer [2] from a regex
-- definition, with some hand-coded bits. [3]
8 9
--
-- Completely accurate information about token-spans within the source
Ian Lynagh's avatar
Ian Lynagh committed
10 11
-- file is maintained.  Every token has a start and end RealSrcLoc
-- attached to it.
12
--
13 14 15 16 17
-- References:
-- [1] https://www.haskell.org/onlinereport/haskell2010/haskellch2.html
-- [2] http://www.haskell.org/alex/
-- [3] https://ghc.haskell.org/trac/ghc/wiki/Commentary/Compiler/Parser
--
18 19 20 21 22 23 24 25 26 27
-----------------------------------------------------------------------------

--   ToDo / known bugs:
--    - parsing integers is a bit slow
--    - readRational is a bit slow
--
--   Known bugs, that were also in the previous version:
--    - M... should be 3 tokens, not 1.
--    - pragma-end should be only valid in a pragma

28
--   qualified operator NOTES.
Ian Lynagh's avatar
Ian Lynagh committed
29
--
30 31 32 33 34 35 36 37 38
--   - If M.(+) is a single lexeme, then..
--     - Probably (+) should be a single lexeme too, for consistency.
--       Otherwise ( + ) would be a prefix operator, but M.( + ) would not be.
--     - But we have to rule out reserved operators, otherwise (..) becomes
--       a different lexeme.
--     - Should we therefore also rule out reserved operators in the qualified
--       form?  This is quite difficult to achieve.  We don't do it for
--       qualified varids.

39 40 41 42

-- -----------------------------------------------------------------------------
-- Alex "Haskell code fragment top"

43
{
44
{-# LANGUAGE BangPatterns #-}
45 46

-- See Note [Warnings in code generated by Alex] in compiler/parser/Lexer.x
47 48
{-# OPTIONS_GHC -fno-warn-unused-matches #-}
{-# OPTIONS_GHC -fno-warn-unused-binds #-}
49
{-# OPTIONS_GHC -fno-warn-unused-imports #-}
50
{-# OPTIONS_GHC -fno-warn-tabs #-}
51
{-# OPTIONS_GHC -fno-warn-missing-signatures #-}
52

53 54
{-# OPTIONS_GHC -funbox-strict-fields #-}

55
module Lexer (
Alan Zimmerman's avatar
Alan Zimmerman committed
56
   Token(..), lexer, pragState, mkPState, PState(..),
Ian Lynagh's avatar
Ian Lynagh committed
57
   P(..), ParseResult(..), getSrcLoc,
58
   getPState, getDynFlags, withThisPackage,
59
   failLocMsgP, failSpanMsgP, srcParseFail,
Ian Lynagh's avatar
Ian Lynagh committed
60
   getMessages,
61
   popContext, pushCurrentContext, setLastToken, setSrcLoc,
vivian's avatar
vivian committed
62
   activeContext, nextIsEOF,
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
63
   getLexState, popLexState, pushLexState,
64
   extension, bangPatEnabled, datatypeContextsEnabled,
65
   traditionalRecordSyntaxEnabled,
66 67
   explicitForallEnabled,
   inRulePrag,
cactus's avatar
cactus committed
68 69 70
   explicitNamespacesEnabled,
   patternSynonymsEnabled,
   sccProfilingOn, hpcEnabled,
Jedai's avatar
Jedai committed
71
   addWarning,
Alan Zimmerman's avatar
Alan Zimmerman committed
72
   lexTokenStream,
73 74
   addAnnotation,AddAnn,addAnnsAt,mkParensApiAnn,
   moveAnnotations
75 76
  ) where

rodlogic's avatar
rodlogic committed
77 78
-- base
import Control.Monad
quchen's avatar
quchen committed
79 80 81
#if __GLASGOW_HASKELL__ > 710
import Control.Monad.Fail
#endif
rodlogic's avatar
rodlogic committed
82
import Data.Bits
83
import Data.Char
rodlogic's avatar
rodlogic committed
84 85 86 87
import Data.List
import Data.Maybe
import Data.Word

88 89 90
-- ghc-boot
import qualified GHC.LanguageExtensions as LangExt

rodlogic's avatar
rodlogic committed
91 92 93 94 95 96 97 98
-- bytestring
import Data.ByteString (ByteString)

-- containers
import Data.Map (Map)
import qualified Data.Map as Map

-- compiler/utils
99
import Bag
100 101 102 103
import Outputable
import StringBuffer
import FastString
import UniqFM
rodlogic's avatar
rodlogic committed
104 105 106 107
import Util             ( readRational )

-- compiler/main
import ErrUtils
108
import DynFlags
rodlogic's avatar
rodlogic committed
109 110 111

-- compiler/basicTypes
import SrcLoc
112
import Module
Alan Zimmerman's avatar
Alan Zimmerman committed
113 114
import BasicTypes     ( InlineSpec(..), RuleMatchInfo(..), FractionalLit(..),
                        SourceText )
115

rodlogic's avatar
rodlogic committed
116 117
-- compiler/parser
import Ctype
Alan Zimmerman's avatar
Alan Zimmerman committed
118 119

import ApiAnnotation
120 121
}

122 123 124
-- -----------------------------------------------------------------------------
-- Alex "Character set macros"

125 126
-- NB: The logic behind these definitions is also reflected in basicTypes/Lexeme.hs
-- Any changes here should likely be reflected there.
127 128 129 130
$unispace    = \x05 -- Trick Alex into handling Unicode. See alexGetByte.
$nl          = [\n\r\f]
$whitechar   = [$nl\v\ $unispace]
$white_no_nl = $whitechar # \n -- TODO #8424
131
$tab         = \t
132 133

$ascdigit  = 0-9
134
$unidigit  = \x03 -- Trick Alex into handling Unicode. See alexGetByte.
135
$decdigit  = $ascdigit -- for now, should really be $digit (ToDo)
136 137 138
$digit     = [$ascdigit $unidigit]

$special   = [\(\)\,\;\[\]\`\{\}]
139 140 141
$ascsymbol = [\!\#\$\%\&\*\+\.\/\<\=\>\?\@\\\^\|\-\~\:]
$unisymbol = \x04 -- Trick Alex into handling Unicode. See alexGetByte.
$symbol    = [$ascsymbol $unisymbol] # [$special \_\"\']
142

143
$unilarge  = \x01 -- Trick Alex into handling Unicode. See alexGetByte.
144
$asclarge  = [A-Z]
145 146
$large     = [$asclarge $unilarge]

147
$unismall  = \x02 -- Trick Alex into handling Unicode. See alexGetByte.
148
$ascsmall  = [a-z]
149 150
$small     = [$ascsmall $unismall \_]

151 152
$unigraphic = \x06 -- Trick Alex into handling Unicode. See alexGetByte.
$graphic   = [$small $large $symbol $digit $special $unigraphic \"\']
153

154
$binit     = 0-1
Ian Lynagh's avatar
Ian Lynagh committed
155
$octit     = 0-7
156
$hexit     = [$decdigit A-F a-f]
157

158 159
$uniidchar = \x07 -- Trick Alex into handling Unicode. See alexGetByte.
$idchar    = [$small $large $digit $uniidchar \']
160

161 162
$pragmachar = [$small $large $digit]

163 164
$docsym    = [\| \^ \* \$]

165

166 167 168 169 170 171 172 173
-- -----------------------------------------------------------------------------
-- Alex "Regular expression macros"

@varid     = $small $idchar*          -- variable identifiers
@conid     = $large $idchar*          -- constructor identifiers

@varsym    = ($symbol # \:) $symbol*  -- variable (operator) symbol
@consym    = \: $symbol*              -- constructor (operator) symbol
174

175
@decimal     = $decdigit+
176
@binary      = $binit+
177 178 179 180 181
@octal       = $octit+
@hexadecimal = $hexit+
@exponent    = [eE] [\-\+]? @decimal

@qual = (@conid \.)+
182 183 184 185
@qvarid = @qual @varid
@qconid = @qual @conid
@qvarsym = @qual @varsym
@qconsym = @qual @consym
186 187 188

@floating_point = @decimal \. @decimal @exponent? | @decimal @exponent

189 190 191 192 193
-- normal signed numerical literals can only be explicitly negative,
-- not explicitly positive (contrast @exponent)
@negative = \-
@signed = @negative ?

194 195 196 197

-- -----------------------------------------------------------------------------
-- Alex "Identifier"

198 199
haskell :-

200 201 202 203 204

-- -----------------------------------------------------------------------------
-- Alex "Rules"

-- everywhere: skip whitespace
Ian Lynagh's avatar
Ian Lynagh committed
205
$white_no_nl+ ;
206
$tab          { warnTab }
207 208 209 210 211

-- Everywhere: deal with nested comments.  We explicitly rule out
-- pragmas, "{-#", so that we don't accidentally treat them as comments.
-- (this can happen even though pragmas will normally take precedence due to
-- longest-match, because pragmas aren't valid in every state, but comments
212 213 214 215
-- are). We also rule out nested Haddock comments, if the -haddock flag is
-- set.

"{-" / { isNormalComment } { nested_comment lexToken }
216 217 218 219

-- Single-line comments are a bit tricky.  Haskell 98 says that two or
-- more dashes followed by a symbol should be parsed as a varsym, so we
-- have to exclude those.
220 221

-- Since Haddock comments aren't valid in every state, we need to rule them
Ian Lynagh's avatar
Ian Lynagh committed
222
-- out here.
223 224 225 226 227 228 229

-- The following two rules match comments that begin with two dashes, but
-- continue with a different character. The rules test that this character
-- is not a symbol (in which case we'd have a varsym), and that it's not a
-- space followed by a Haddock comment symbol (docsym) (in which case we'd
-- have a Haddock comment). The rules then munch the rest of the line.

230
"-- " ~$docsym .* { lineCommentToken }
231
"--" [^$symbol \ ] .* { lineCommentToken }
232 233 234

-- Next, match Haddock comments if no -haddock flag

235
"-- " $docsym .* / { ifExtension (not . haddockEnabled) } { lineCommentToken }
236 237 238 239 240 241 242

-- Now, when we've matched comments that begin with 2 dashes and continue
-- with a different character, we need to match comments that begin with three
-- or more dashes (which clearly can't be Haddock comments). We only need to
-- make sure that the first non-dash character isn't a symbol, and munch the
-- rest of the line.

243
"---"\-* ~$symbol .* { lineCommentToken }
244 245 246 247

-- Since the previous rules all match dashes followed by at least one
-- character, we also need to match a whole line filled with just dashes.

Jedai's avatar
Jedai committed
248
"--"\-* / { atEOL } { lineCommentToken }
249 250 251 252

-- We need this rule since none of the other single line comment rules
-- actually match this case.

Jedai's avatar
Jedai committed
253
"-- " / { atEOL } { lineCommentToken }
254 255 256 257 258 259 260 261 262 263 264

-- 'bol' state: beginning of a line.  Slurp up all the whitespace (including
-- blank lines) until we find a non-whitespace character, then do layout
-- processing.
--
-- One slight wibble here: what if the line begins with {-#? In
-- theory, we have to lex the pragma to see if it's one we recognise,
-- and if it is, then we backtrack and do_bol, otherwise we treat it
-- as a nested comment.  We don't bother with this: if the line begins
-- with {-#, then we'll assume it's a pragma we know about and go for do_bol.
<bol> {
Ian Lynagh's avatar
Ian Lynagh committed
265
  \n                                    ;
Adam Gundry's avatar
Adam Gundry committed
266 267
  ^\# line                              { begin line_prag1 }
  ^\# / { followedByDigit }             { begin line_prag1 }
Ian Lynagh's avatar
Ian Lynagh committed
268 269 270
  ^\# pragma .* \n                      ; -- GCC 3.3 CPP generated, apparently
  ^\# \! .* \n                          ; -- #!, for scripts
  ()                                    { do_bol }
271 272 273 274 275
}

-- after a layout keyword (let, where, do, of), we begin a new layout
-- context if the curly brace is missing.
-- Careful! This stuff is quite delicate.
276
<layout, layout_do, layout_if> {
Ian Lynagh's avatar
Ian Lynagh committed
277 278 279 280
  \{ / { notFollowedBy '-' }            { hopefully_open_brace }
        -- we might encounter {-# here, but {- has been handled already
  \n                                    ;
  ^\# (line)?                           { begin line_prag1 }
281 282
}

283 284 285 286 287 288
-- after an 'if', a vertical bar starts a layout context for MultiWayIf
<layout_if> {
  \| / { notFollowedBySymbol }          { new_layout_context True ITvbar }
  ()                                    { pop }
}

289
-- do is treated in a subtly different way, see new_layout_context
290 291
<layout>    ()                          { new_layout_context True  ITvocurly }
<layout_do> ()                          { new_layout_context False ITvocurly }
292 293 294 295

-- after a new layout context which was found to be to the left of the
-- previous context, we have generated a '{' token, and we now need to
-- generate a matching '}' token.
Ian Lynagh's avatar
Ian Lynagh committed
296
<layout_left>  ()                       { do_layout_left }
297

Ian Lynagh's avatar
Ian Lynagh committed
298
<0,option_prags> \n                     { begin bol }
299

300 301
"{-#" $whitechar* $pragmachar+ / { known_pragma linePrags }
                                { dispatch_pragmas linePrags }
302 303 304

-- single-line line pragmas, of the form
--    # <line> "<file>" <extra-stuff> \n
305
<line_prag1> @decimal                   { setLine line_prag1a }
Ian Lynagh's avatar
Ian Lynagh committed
306 307
<line_prag1a> \" [$graphic \ ]* \"      { setFile line_prag1b }
<line_prag1b> .*                        { pop }
308 309 310

-- Haskell-style line pragmas, of the form
--    {-# LINE <line> "<file>" #-}
311
<line_prag2> @decimal                   { setLine line_prag2a }
Ian Lynagh's avatar
Ian Lynagh committed
312 313
<line_prag2a> \" [$graphic \ ]* \"      { setFile line_prag2b }
<line_prag2b> "#-}"|"-}"                { pop }
314 315
   -- NOTE: accept -} at the end of a LINE pragma, for compatibility
   -- with older versions of GHC which generated these.
316

317
<0,option_prags> {
Ian Lynagh's avatar
Ian Lynagh committed
318
  "{-#" $whitechar* $pragmachar+
319 320 321 322 323
        $whitechar+ $pragmachar+ / { known_pragma twoWordPrags }
                                 { dispatch_pragmas twoWordPrags }

  "{-#" $whitechar* $pragmachar+ / { known_pragma oneWordPrags }
                                 { dispatch_pragmas oneWordPrags }
324

325
  -- We ignore all these pragmas, but don't generate a warning for them
326 327
  "{-#" $whitechar* $pragmachar+ / { known_pragma ignoredPrags }
                                 { dispatch_pragmas ignoredPrags }
328 329

  -- ToDo: should only be valid inside a pragma:
Ian Lynagh's avatar
Ian Lynagh committed
330
  "#-}"                          { endPrag }
331 332
}

333
<option_prags> {
334 335
  "{-#"  $whitechar* $pragmachar+ / { known_pragma fileHeaderPrags }
                                   { dispatch_pragmas fileHeaderPrags }
David Waern's avatar
David Waern committed
336 337
}

338 339
<0> {
  -- In the "0" mode we ignore these pragmas
340
  "{-#"  $whitechar* $pragmachar+ / { known_pragma fileHeaderPrags }
341 342 343
                     { nested_comment lexToken }
}

344
<0,option_prags> {
345 346
  "{-#"  { warnThen Opt_WarnUnrecognisedPragmas (text "Unrecognised pragma")
                    (nested_comment lexToken) }
347 348
}

349 350
-- '0' state: ordinary lexemes

351 352
-- Haddock comments

353
<0,option_prags> {
David Waern's avatar
David Waern committed
354 355
  "-- " $docsym      / { ifExtension haddockEnabled } { multiline_doc_comment }
  "{-" \ ? $docsym   / { ifExtension haddockEnabled } { nested_doc_comment }
356 357
}

358 359
-- "special" symbols

360
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
361 362
  "[:" / { ifExtension parrEnabled }    { token ITopabrack }
  ":]" / { ifExtension parrEnabled }    { token ITcpabrack }
363
}
Ian Lynagh's avatar
Ian Lynagh committed
364

365
<0> {
366 367 368 369 370 371 372 373 374
  "[|"        / { ifExtension thQuotesEnabled } { token (ITopenExpQuote NoE) }
  "[||"       / { ifExtension thQuotesEnabled } { token (ITopenTExpQuote NoE) }
  "[e|"       / { ifExtension thQuotesEnabled } { token (ITopenExpQuote HasE) }
  "[e||"      / { ifExtension thQuotesEnabled } { token (ITopenTExpQuote HasE) }
  "[p|"       / { ifExtension thQuotesEnabled } { token ITopenPatQuote }
  "[d|"       / { ifExtension thQuotesEnabled } { layout_token ITopenDecQuote }
  "[t|"       / { ifExtension thQuotesEnabled } { token ITopenTypQuote }
  "|]"        / { ifExtension thQuotesEnabled } { token ITcloseQuote }
  "||]"       / { ifExtension thQuotesEnabled } { token ITcloseTExpQuote }
375 376 377 378
  \$ @varid   / { ifExtension thEnabled } { skip_one_varid ITidEscape }
  "$$" @varid / { ifExtension thEnabled } { skip_two_varid ITidTyEscape }
  "$("        / { ifExtension thEnabled } { token ITparenEscape }
  "$$("       / { ifExtension thEnabled } { token ITparenTyEscape }
379

380
  "[" @varid "|"  / { ifExtension qqEnabled }
381
                     { lex_quasiquote_tok }
382 383

  -- qualified quasi-quote (#5555)
384 385
  "[" @qvarid "|"  / { ifExtension qqEnabled }
                     { lex_qquasiquote_tok }
386 387
}

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
388 389 390 391 392 393 394 395
  -- See Note [Lexing type applications]
<0> {
    [^ $idchar \) ] ^
  "@"
    / { ifExtension typeApplicationEnabled `alexAndPred` notFollowedBySymbol }
    { token ITtypeApp }
}

396
<0> {
ross's avatar
ross committed
397
  "(|" / { ifExtension arrowsEnabled `alexAndPred` notFollowedBySymbol }
398 399 400 401 402 403 404 405 406 407 408
                                        { special (IToparenbar NormalSyntax) }
  "|)" / { ifExtension arrowsEnabled }  { special (ITcparenbar NormalSyntax) }

  $unigraphic -- ⦇
    / { ifCurrentChar '⦇' `alexAndPred`
        ifExtension (\i -> unicodeSyntaxEnabled i && arrowsEnabled i) }
    { special (IToparenbar UnicodeSyntax) }
  $unigraphic -- ⦈
    / { ifCurrentChar '⦈' `alexAndPred`
        ifExtension (\i -> unicodeSyntaxEnabled i && arrowsEnabled i) }
    { special (ITcparenbar UnicodeSyntax) }
409 410
}

411
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
412
  \? @varid / { ifExtension ipEnabled } { skip_one_varid ITdupipvarid }
413 414
}

Adam Gundry's avatar
Adam Gundry committed
415 416 417 418 419
<0> {
  "#" @varid / { ifExtension overloadedLabelsEnabled }
               { skip_one_varid ITlabelvarid }
}

420
<0> {
421
  "(#" / { ifExtension unboxedTuplesEnabled }
Ian Lynagh's avatar
Ian Lynagh committed
422 423 424 425 426
         { token IToubxparen }
  "#)" / { ifExtension unboxedTuplesEnabled }
         { token ITcubxparen }
}

427
<0,option_prags> {
Ian Lynagh's avatar
Ian Lynagh committed
428 429 430 431 432 433 434 435 436 437
  \(                                    { special IToparen }
  \)                                    { special ITcparen }
  \[                                    { special ITobrack }
  \]                                    { special ITcbrack }
  \,                                    { special ITcomma }
  \;                                    { special ITsemi }
  \`                                    { special ITbackquote }

  \{                                    { open_brace }
  \}                                    { close_brace }
438 439
}

440
<0,option_prags> {
441 442
  @qvarid                       { idtoken qvarid }
  @qconid                       { idtoken qconid }
Ian Lynagh's avatar
Ian Lynagh committed
443
  @varid                        { varid }
444
  @conid                        { idtoken conid }
445 446
}

447
<0> {
448 449
  @qvarid "#"+      / { ifExtension magicHashEnabled } { idtoken qvarid }
  @qconid "#"+      / { ifExtension magicHashEnabled } { idtoken qconid }
450
  @varid "#"+       / { ifExtension magicHashEnabled } { varid }
451
  @conid "#"+       / { ifExtension magicHashEnabled } { idtoken conid }
452 453
}

454 455
-- ToDo: - move `var` and (sym) into lexical syntax?
--       - remove backquote from $special?
456
<0> {
457 458
  @qvarsym                                         { idtoken qvarsym }
  @qconsym                                         { idtoken qconsym }
459 460
  @varsym                                          { varsym }
  @consym                                          { consym }
461 462
}

463 464
-- For the normal boxed literals we need to be careful
-- when trying to be close to Haskell98
465
<0> {
466
  -- Normal integral literals (:: Num a => a, from Integer)
467
  @decimal                                                               { tok_num positive 0 0 decimal }
468
  0[bB] @binary                / { ifExtension binaryLiteralsEnabled }   { tok_num positive 2 2 binary }
469 470 471
  0[oO] @octal                                                           { tok_num positive 2 2 octal }
  0[xX] @hexadecimal                                                     { tok_num positive 2 2 hexadecimal }
  @negative @decimal           / { ifExtension negativeLiteralsEnabled } { tok_num negative 1 1 decimal }
472 473
  @negative 0[bB] @binary      / { ifExtension negativeLiteralsEnabled `alexAndPred`
                                   ifExtension binaryLiteralsEnabled }   { tok_num negative 3 3 binary }
474 475
  @negative 0[oO] @octal       / { ifExtension negativeLiteralsEnabled } { tok_num negative 3 3 octal }
  @negative 0[xX] @hexadecimal / { ifExtension negativeLiteralsEnabled } { tok_num negative 3 3 hexadecimal }
476 477

  -- Normal rational literals (:: Fractional a => a, from Rational)
478 479
  @floating_point                                                        { strtoken tok_float }
  @negative @floating_point    / { ifExtension negativeLiteralsEnabled } { strtoken tok_float }
480 481
}

482
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
483
  -- Unboxed ints (:: Int#) and words (:: Word#)
484 485
  -- It's simpler (and faster?) to give separate cases to the negatives,
  -- especially considering octal/hexadecimal prefixes.
Ian Lynagh's avatar
Ian Lynagh committed
486
  @decimal                     \# / { ifExtension magicHashEnabled } { tok_primint positive 0 1 decimal }
487 488
  0[bB] @binary                \# / { ifExtension magicHashEnabled `alexAndPred`
                                      ifExtension binaryLiteralsEnabled } { tok_primint positive 2 3 binary }
Ian Lynagh's avatar
Ian Lynagh committed
489 490 491
  0[oO] @octal                 \# / { ifExtension magicHashEnabled } { tok_primint positive 2 3 octal }
  0[xX] @hexadecimal           \# / { ifExtension magicHashEnabled } { tok_primint positive 2 3 hexadecimal }
  @negative @decimal           \# / { ifExtension magicHashEnabled } { tok_primint negative 1 2 decimal }
492 493
  @negative 0[bB] @binary      \# / { ifExtension magicHashEnabled `alexAndPred`
                                      ifExtension binaryLiteralsEnabled } { tok_primint negative 3 4 binary }
Ian Lynagh's avatar
Ian Lynagh committed
494
  @negative 0[oO] @octal       \# / { ifExtension magicHashEnabled } { tok_primint negative 3 4 octal }
495
  @negative 0[xX] @hexadecimal \# / { ifExtension magicHashEnabled } { tok_primint negative 3 4 hexadecimal }
496

Ian Lynagh's avatar
Ian Lynagh committed
497
  @decimal                     \# \# / { ifExtension magicHashEnabled } { tok_primword 0 2 decimal }
498 499
  0[bB] @binary                \# \# / { ifExtension magicHashEnabled `alexAndPred`
                                         ifExtension binaryLiteralsEnabled } { tok_primword 2 4 binary }
Ian Lynagh's avatar
Ian Lynagh committed
500 501 502
  0[oO] @octal                 \# \# / { ifExtension magicHashEnabled } { tok_primword 2 4 octal }
  0[xX] @hexadecimal           \# \# / { ifExtension magicHashEnabled } { tok_primword 2 4 hexadecimal }

503 504
  -- Unboxed floats and doubles (:: Float#, :: Double#)
  -- prim_{float,double} work with signed literals
505 506
  @signed @floating_point \# / { ifExtension magicHashEnabled } { init_strtoken 1 tok_primfloat }
  @signed @floating_point \# \# / { ifExtension magicHashEnabled } { init_strtoken 2 tok_primdouble }
507 508 509 510 511 512
}

-- Strings and chars are lexed by hand-written code.  The reason is
-- that even if we recognise the string or char here in the regex
-- lexer, we would still have to parse the string afterward in order
-- to convert it to a String.
513
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
514 515
  \'                            { lex_char_tok }
  \"                            { lex_string_tok }
516 517
}

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543
-- Note [Lexing type applications]
-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-- The desired syntax for type applications is to prefix the type application
-- with '@', like this:
--
--   foo @Int @Bool baz bum
--
-- This, of course, conflicts with as-patterns. The conflict arises because
-- expressions and patterns use the same parser, and also because we want
-- to allow type patterns within expression patterns.
--
-- Disambiguation is accomplished by requiring *something* to appear betwen
-- type application and the preceding token. This something must end with
-- a character that cannot be the end of the variable bound in an as-pattern.
-- Currently (June 2015), this means that the something cannot end with a
-- $idchar or a close-paren. (The close-paren is necessary if the as-bound
-- identifier is symbolic.)
--
-- Note that looking for whitespace before the '@' is insufficient, because
-- of this pathological case:
--
--   foo {- hi -}@Int
--
-- This design is predicated on the fact that as-patterns are generally
-- whitespace-free, and also that this whole thing is opt-in, with the
-- TypeApplications extension.
544 545 546 547

-- -----------------------------------------------------------------------------
-- Alex "Haskell code fragment bottom"

548
{
549

550 551 552
-- -----------------------------------------------------------------------------
-- The token type

553
data Token
Ian Lynagh's avatar
Ian Lynagh committed
554
  = ITas                        -- Haskell keywords
555 556 557 558 559 560 561 562
  | ITcase
  | ITclass
  | ITdata
  | ITdefault
  | ITderiving
  | ITdo
  | ITelse
  | IThiding
563
  | ITforeign
564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579
  | ITif
  | ITimport
  | ITin
  | ITinfix
  | ITinfixl
  | ITinfixr
  | ITinstance
  | ITlet
  | ITmodule
  | ITnewtype
  | ITof
  | ITqualified
  | ITthen
  | ITtype
  | ITwhere

580
  | ITforall            IsUnicodeSyntax -- GHC extension keywords
581 582 583 584
  | ITexport
  | ITlabel
  | ITdynamic
  | ITsafe
585
  | ITinterruptible
586 587 588
  | ITunsafe
  | ITstdcallconv
  | ITccallconv
589
  | ITcapiconv
590
  | ITprimcallconv
thoughtpolice's avatar
thoughtpolice committed
591
  | ITjavascriptcallconv
592
  | ITmdo
593
  | ITfamily
594
  | ITrole
595 596 597
  | ITgroup
  | ITby
  | ITusing
cactus's avatar
cactus committed
598
  | ITpattern
Facundo Domínguez's avatar
Facundo Domínguez committed
599
  | ITstatic
600

Alan Zimmerman's avatar
Alan Zimmerman committed
601 602 603 604 605 606 607 608
  -- Pragmas, see  note [Pragma source text] in BasicTypes
  | ITinline_prag       SourceText InlineSpec RuleMatchInfo
  | ITspec_prag         SourceText                -- SPECIALISE
  | ITspec_inline_prag  SourceText Bool    -- SPECIALISE INLINE (or NOINLINE)
  | ITsource_prag       SourceText
  | ITrules_prag        SourceText
  | ITwarning_prag      SourceText
  | ITdeprecated_prag   SourceText
609
  | ITline_prag
Alan Zimmerman's avatar
Alan Zimmerman committed
610 611 612 613 614 615
  | ITscc_prag          SourceText
  | ITgenerated_prag    SourceText
  | ITcore_prag         SourceText         -- hdaume: core annotations
  | ITunpack_prag       SourceText
  | ITnounpack_prag     SourceText
  | ITann_prag          SourceText
616
  | ITclose_prag
617 618 619
  | IToptions_prag String
  | ITinclude_prag String
  | ITlanguage_prag
Alan Zimmerman's avatar
Alan Zimmerman committed
620 621 622 623 624 625 626 627 628
  | ITvect_prag         SourceText
  | ITvect_scalar_prag  SourceText
  | ITnovect_prag       SourceText
  | ITminimal_prag      SourceText
  | IToverlappable_prag SourceText  -- instance overlap mode
  | IToverlapping_prag  SourceText  -- instance overlap mode
  | IToverlaps_prag     SourceText  -- instance overlap mode
  | ITincoherent_prag   SourceText  -- instance overlap mode
  | ITctype             SourceText
629

Ian Lynagh's avatar
Ian Lynagh committed
630
  | ITdotdot                    -- reserved symbols
631
  | ITcolon
632
  | ITdcolon            IsUnicodeSyntax
633 634
  | ITequal
  | ITlam
635
  | ITlcase
636
  | ITvbar
637 638
  | ITlarrow            IsUnicodeSyntax
  | ITrarrow            IsUnicodeSyntax
639 640
  | ITat
  | ITtilde
batterseapower's avatar
batterseapower committed
641
  | ITtildehsh
642
  | ITdarrow            IsUnicodeSyntax
643 644 645 646
  | ITminus
  | ITbang
  | ITdot

Ian Lynagh's avatar
Ian Lynagh committed
647
  | ITbiglam                    -- GHC-extension symbols
648

Ian Lynagh's avatar
Ian Lynagh committed
649
  | ITocurly                    -- special symbols
650 651 652 653
  | ITccurly
  | ITvocurly
  | ITvccurly
  | ITobrack
Ian Lynagh's avatar
Ian Lynagh committed
654 655
  | ITopabrack                  -- [:, for parallel arrays with -XParallelArrays
  | ITcpabrack                  -- :], for parallel arrays with -XParallelArrays
656 657 658 659 660 661 662 663 664
  | ITcbrack
  | IToparen
  | ITcparen
  | IToubxparen
  | ITcubxparen
  | ITsemi
  | ITcomma
  | ITunderscore
  | ITbackquote
dreixel's avatar
dreixel committed
665
  | ITsimpleQuote               --  '
666

Ian Lynagh's avatar
Ian Lynagh committed
667
  | ITvarid   FastString        -- identifiers
668 669 670 671 672 673 674 675
  | ITconid   FastString
  | ITvarsym  FastString
  | ITconsym  FastString
  | ITqvarid  (FastString,FastString)
  | ITqconid  (FastString,FastString)
  | ITqvarsym (FastString,FastString)
  | ITqconsym (FastString,FastString)

Ian Lynagh's avatar
Ian Lynagh committed
676
  | ITdupipvarid   FastString   -- GHC extension: implicit param: ?x
Adam Gundry's avatar
Adam Gundry committed
677
  | ITlabelvarid   FastString   -- Overloaded label: #x
678

679 680 681
  | ITchar     SourceText Char       -- Note [Literal source text] in BasicTypes
  | ITstring   SourceText FastString -- Note [Literal source text] in BasicTypes
  | ITinteger  SourceText Integer    -- Note [Literal source text] in BasicTypes
Alan Zimmerman's avatar
Alan Zimmerman committed
682
  | ITrational FractionalLit
683

684 685 686 687
  | ITprimchar   SourceText Char     -- Note [Literal source text] in BasicTypes
  | ITprimstring SourceText ByteString -- Note [Literal source text] @BasicTypes
  | ITprimint    SourceText Integer  -- Note [Literal source text] in BasicTypes
  | ITprimword   SourceText Integer  -- Note [Literal source text] in BasicTypes
688 689
  | ITprimfloat  FractionalLit
  | ITprimdouble FractionalLit
690

691
  -- Template Haskell extension tokens
692
  | ITopenExpQuote HasE         --  [| or [e|
Ian Lynagh's avatar
Ian Lynagh committed
693 694 695 696
  | ITopenPatQuote              --  [p|
  | ITopenDecQuote              --  [d|
  | ITopenTypQuote              --  [t|
  | ITcloseQuote                --  |]
697
  | ITopenTExpQuote HasE        --  [|| or [e||
698
  | ITcloseTExpQuote            --  ||]
Ian Lynagh's avatar
Ian Lynagh committed
699 700
  | ITidEscape   FastString     --  $x
  | ITparenEscape               --  $(
701 702
  | ITidTyEscape   FastString   --  $$x
  | ITparenTyEscape             --  $$(
Ian Lynagh's avatar
Ian Lynagh committed
703
  | ITtyQuote                   --  ''
704 705 706 707 708 709 710 711
  | ITquasiQuote (FastString,FastString,RealSrcSpan)
    -- ITquasiQuote(quoter, quote, loc)
    -- represents a quasi-quote of the form
    -- [quoter| quote |]
  | ITqQuasiQuote (FastString,FastString,FastString,RealSrcSpan)
    -- ITqQuasiQuote(Qual, quoter, quote, loc)
    -- represents a qualified quasi-quote of the form
    -- [Qual.quoter| quote |]
712 713 714 715

  -- Arrow notation extension
  | ITproc
  | ITrec
716 717
  | IToparenbar  IsUnicodeSyntax --  (|
  | ITcparenbar  IsUnicodeSyntax --  |)
718 719 720 721
  | ITlarrowtail IsUnicodeSyntax --  -<
  | ITrarrowtail IsUnicodeSyntax --  >-
  | ITLarrowtail IsUnicodeSyntax --  -<<
  | ITRarrowtail IsUnicodeSyntax --  >>-
722

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
723 724 725 726 727 728 729
  -- type application '@' (lexed differently than as-pattern '@',
  -- due to checking for preceding whitespace)
  | ITtypeApp


  | ITunknown String            -- Used when the lexer can't make sense of it
  | ITeof                       -- end of file token
730 731 732 733 734 735 736

  -- Documentation annotations
  | ITdocCommentNext  String     -- something beginning '-- |'
  | ITdocCommentPrev  String     -- something beginning '-- ^'
  | ITdocCommentNamed String     -- something beginning '-- $'
  | ITdocSection      Int String -- a section heading
  | ITdocOptions      String     -- doc options (prune, ignore-exports, etc)
Jedai's avatar
Jedai committed
737 738
  | ITlineComment     String     -- comment starting by "--"
  | ITblockComment    String     -- comment in {- -}
739

740
  deriving Show
741

Alan Zimmerman's avatar
Alan Zimmerman committed
742 743 744
instance Outputable Token where
  ppr x = text (show x)

Alan Zimmerman's avatar
Alan Zimmerman committed
745

746 747 748 749 750 751 752
-- the bitmap provided as the third component indicates whether the
-- corresponding extension keyword is valid under the extension options
-- provided to the compiler; if the extension corresponding to *any* of the
-- bits set in the bitmap is enabled, the keyword is valid (this setup
-- facilitates using a keyword in two different extensions that can be
-- activated independently)
--
753
reservedWordsFM :: UniqFM (Token, ExtsBitmap)
754
reservedWordsFM = listToUFM $
Ian Lynagh's avatar
Ian Lynagh committed
755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781
    map (\(x, y, z) -> (mkFastString x, (y, z)))
        [( "_",              ITunderscore,    0 ),
         ( "as",             ITas,            0 ),
         ( "case",           ITcase,          0 ),
         ( "class",          ITclass,         0 ),
         ( "data",           ITdata,          0 ),
         ( "default",        ITdefault,       0 ),
         ( "deriving",       ITderiving,      0 ),
         ( "do",             ITdo,            0 ),
         ( "else",           ITelse,          0 ),
         ( "hiding",         IThiding,        0 ),
         ( "if",             ITif,            0 ),
         ( "import",         ITimport,        0 ),
         ( "in",             ITin,            0 ),
         ( "infix",          ITinfix,         0 ),
         ( "infixl",         ITinfixl,        0 ),
         ( "infixr",         ITinfixr,        0 ),
         ( "instance",       ITinstance,      0 ),
         ( "let",            ITlet,           0 ),
         ( "module",         ITmodule,        0 ),
         ( "newtype",        ITnewtype,       0 ),
         ( "of",             ITof,            0 ),
         ( "qualified",      ITqualified,     0 ),
         ( "then",           ITthen,          0 ),
         ( "type",           ITtype,          0 ),
         ( "where",          ITwhere,         0 ),

782 783
         ( "forall",         ITforall NormalSyntax,
                                              xbit ExplicitForallBit .|.
784 785
                                              xbit InRulePragBit),
         ( "mdo",            ITmdo,           xbit RecursiveDoBit),
786 787 788
             -- See Note [Lexing type pseudo-keywords]
         ( "family",         ITfamily,        0 ),
         ( "role",           ITrole,          0 ),
789
         ( "pattern",        ITpattern,       xbit PatternSynonymsBit),
Facundo Domínguez's avatar
Facundo Domínguez committed
790
         ( "static",         ITstatic,        0 ),
791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811
         ( "group",          ITgroup,         xbit TransformComprehensionsBit),
         ( "by",             ITby,            xbit TransformComprehensionsBit),
         ( "using",          ITusing,         xbit TransformComprehensionsBit),

         ( "foreign",        ITforeign,       xbit FfiBit),
         ( "export",         ITexport,        xbit FfiBit),
         ( "label",          ITlabel,         xbit FfiBit),
         ( "dynamic",        ITdynamic,       xbit FfiBit),
         ( "safe",           ITsafe,          xbit FfiBit .|.
                                              xbit SafeHaskellBit),
         ( "interruptible",  ITinterruptible, xbit InterruptibleFfiBit),
         ( "unsafe",         ITunsafe,        xbit FfiBit),
         ( "stdcall",        ITstdcallconv,   xbit FfiBit),
         ( "ccall",          ITccallconv,     xbit FfiBit),
         ( "capi",           ITcapiconv,      xbit CApiFfiBit),
         ( "prim",           ITprimcallconv,  xbit FfiBit),
         ( "javascript",     ITjavascriptcallconv, xbit FfiBit),

         ( "rec",            ITrec,           xbit ArrowsBit .|.
                                              xbit RecursiveDoBit),
         ( "proc",           ITproc,          xbit ArrowsBit)
812 813
     ]

814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829
{-----------------------------------
Note [Lexing type pseudo-keywords]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

One might think that we wish to treat 'family' and 'role' as regular old
varids whenever -XTypeFamilies and -XRoleAnnotations are off, respectively.
But, there is no need to do so. These pseudo-keywords are not stolen syntax:
they are only used after the keyword 'type' at the top-level, where varids are
not allowed. Furthermore, checks further downstream (TcTyClsDecls) ensure that
type families and role annotations are never declared without their extensions
on. In fact, by unconditionally lexing these pseudo-keywords as special, we
can get better error messages.

Also, note that these are included in the `varid` production in the parser --
a key detail to make all this work.
-------------------------------------}
830

831
reservedSymsFM :: UniqFM (Token, ExtsBitmap -> Bool)
832
reservedSymsFM = listToUFM $
Ian Lynagh's avatar
Ian Lynagh committed
833
    map (\ (x,y,z) -> (mkFastString x,(y,z)))
834
      [ ("..",  ITdotdot,              always)
Ian Lynagh's avatar
Ian Lynagh committed
835
        -- (:) is a reserved op, meaning only list cons
836 837 838 839 840 841 842 843 844 845 846 847 848
       ,(":",   ITcolon,               always)
       ,("::",  ITdcolon NormalSyntax, always)
       ,("=",   ITequal,               always)
       ,("\\",  ITlam,                 always)
       ,("|",   ITvbar,                always)
       ,("<-",  ITlarrow NormalSyntax, always)
       ,("->",  ITrarrow NormalSyntax, always)
       ,("@",   ITat,                  always)
       ,("~",   ITtilde,               always)
       ,("~#",  ITtildehsh,            magicHashEnabled)
       ,("=>",  ITdarrow NormalSyntax, always)
       ,("-",   ITminus,               always)
       ,("!",   ITbang,                always)
Ian Lynagh's avatar
Ian Lynagh committed
849 850

        -- For 'forall a . t'
851
       ,(".", ITdot,  always) -- \i -> explicitForallEnabled i || inRulePrag i)
Ian Lynagh's avatar
Ian Lynagh committed
852

853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872
       ,("-<",  ITlarrowtail NormalSyntax, arrowsEnabled)
       ,(">-",  ITrarrowtail NormalSyntax, arrowsEnabled)
       ,("-<<", ITLarrowtail NormalSyntax, arrowsEnabled)
       ,(">>-", ITRarrowtail NormalSyntax, arrowsEnabled)

       ,("∷",   ITdcolon UnicodeSyntax, unicodeSyntaxEnabled)
       ,("⇒",   ITdarrow UnicodeSyntax, unicodeSyntaxEnabled)
       ,("∀",   ITforall UnicodeSyntax, unicodeSyntaxEnabled)
       ,("→",   ITrarrow UnicodeSyntax, unicodeSyntaxEnabled)
       ,("←",   ITlarrow UnicodeSyntax, unicodeSyntaxEnabled)

       ,("⤙",   ITlarrowtail UnicodeSyntax,
                                \i -> unicodeSyntaxEnabled i && arrowsEnabled i)
       ,("⤚",   ITrarrowtail UnicodeSyntax,
                                \i -> unicodeSyntaxEnabled i && arrowsEnabled i)
       ,("⤛",   ITLarrowtail UnicodeSyntax,
                                \i -> unicodeSyntaxEnabled i && arrowsEnabled i)
       ,("⤜",   ITRarrowtail UnicodeSyntax,
                                \i -> unicodeSyntaxEnabled i && arrowsEnabled i)

Simon Marlow's avatar
Simon Marlow committed
873 874 875
        -- ToDo: ideally, → and ∷ should be "specials", so that they cannot
        -- form part of a large operator.  This would let us have a better
        -- syntax for kinds: ɑ∷*→* would be a legal kind signature. (maybe).
876 877 878 879 880
       ]

-- -----------------------------------------------------------------------------
-- Lexer actions

Ian Lynagh's avatar
Ian Lynagh committed
881
type Action = RealSrcSpan -> StringBuffer -> Int -> P (RealLocated Token)
882

883
special :: Token -> Action
twanvl's avatar
twanvl committed
884
special tok span _buf _len = return (L span tok)
885

886
token, layout_token :: Token -> Action
twanvl's avatar
twanvl committed
887 888
token t span _buf _len = return (L span t)
layout_token t span _buf _len = pushLexState layout >> return (L span t)
889

890 891
idtoken :: (StringBuffer -> Int -> Token) -> Action
idtoken f span buf len = return (L span $! (f buf len))
892

893
skip_one_varid :: (FastString -> Token) -> Action
Ian Lynagh's avatar
Ian Lynagh committed
894
skip_one_varid f span buf len
895
  = return (L span $! f (lexemeToFastString (stepOn buf) (len-1)))
896

897 898 899 900
skip_two_varid :: (FastString -> Token) -> Action
skip_two_varid f span buf len
  = return (L span $! f (lexemeToFastString (stepOn (stepOn buf)) (len-2)))

901
strtoken :: (String -> Token) -> Action
Ian Lynagh's avatar
Ian Lynagh committed
902
strtoken f span buf len =
903
  return (L span $! (f $! lexemeToString buf len))
904

905
init_strtoken :: Int -> (String -> Token) -> Action
906
-- like strtoken, but drops the last N character(s)
Ian Lynagh's avatar
Ian Lynagh committed
907
init_strtoken drop f span buf len =
908
  return (L span $! (f $! lexemeToString buf (len-drop)))
909 910

begin :: Int -> Action
911
begin code _span _str _len = do pushLexState code; lexToken
912 913

pop :: Action
Ian Lynagh's avatar
Ian Lynagh committed
914 915
pop _span _buf _len = do _ <- popLexState
                         lexToken
916

917 918 919 920 921 922 923 924 925 926 927
hopefully_open_brace :: Action
hopefully_open_brace span buf len
 = do relaxed <- extension relaxedLayout
      ctx <- getContext
      (AI l _) <- getInput
      let offset = srcLocCol l
          isOK = relaxed ||
                 case ctx of
                 Layout prev_off : _ -> prev_off < offset
                 _                   -> True
      if isOK then pop_and open_brace span buf len
Ian Lynagh's avatar
Ian Lynagh committed
928
              else failSpanMsgP (RealSrcSpan span) (text "Missing block")
929

930
pop_and :: Action -> Action
Ian Lynagh's avatar
Ian Lynagh committed
931 932
pop_and act span buf len = do _ <- popLexState
                              act span buf len
933

934
{-# INLINE nextCharIs #-}
Ian Lynagh's avatar
Ian Lynagh committed
935
nextCharIs :: StringBuffer -> (Char -> Bool) -> Bool
936 937
nextCharIs buf p = not (atEnd buf) && p (currentChar buf)

938 939 940 941
{-# INLINE nextCharIsNot #-}
nextCharIsNot :: StringBuffer -> (Char -> Bool) -> Bool
nextCharIsNot buf p = not (nextCharIs buf p)

942
notFollowedBy :: Char -> AlexAccPred ExtsBitmap
Ian Lynagh's avatar
Ian Lynagh committed
943
notFollowedBy char _ _ _ (AI _ buf)
944
  = nextCharIsNot buf (== char)
945

946
notFollowedBySymbol :: AlexAccPred ExtsBitmap
947
notFollowedBySymbol _ _ _ (AI _ buf)
948
  = nextCharIsNot buf (`elem` "!#$%&*+./<=>?@\\^|-~")
ross's avatar
ross committed
949

Adam Gundry's avatar
Adam Gundry committed
950 951 952 953
followedByDigit :: AlexAccPred ExtsBitmap
followedByDigit _ _ _ (AI _ buf)
  = afterOptionalSpace buf (\b -> nextCharIs b (`elem` ['0'..'9']))

954 955 956 957
ifCurrentChar :: Char -> AlexAccPred ExtsBitmap
ifCurrentChar char _ (AI _ buf) _ _
  = nextCharIs buf (== char)

958 959 960 961 962
-- We must reject doc comments as being ordinary comments everywhere.
-- In some cases the doc comment will be selected as the lexeme due to
-- maximal munch, but not always, because the nested comment rule is
-- valid in all states, but the doc-comment rules are only valid in
-- the non-layout states.
963
isNormalComment :: AlexAccPred ExtsBitmap
964
isNormalComment bits _ _ (AI _ buf)
965
  | haddockEnabled bits = notFollowedByDocOrPragma
966
  | otherwise           = nextCharIsNot buf (== '#')
967 968
  where
    notFollowedByDocOrPragma
969
       = afterOptionalSpace buf (\b -> nextCharIsNot b (`elem` "|^*$#"))
970

971 972 973 974 975
afterOptionalSpace :: StringBuffer -> (StringBuffer -> Bool) -> Bool
afterOptionalSpace buf p
    = if nextCharIs buf (== ' ')
      then p (snd (nextChar buf))
      else p buf
976

977
atEOL :: AlexAccPred ExtsBitmap
978
atEOL _ _ _ (AI _ buf) = atEnd buf || currentChar buf == '\n'
979

980
ifExtension :: (ExtsBitmap -> Bool) -> AlexAccPred ExtsBitmap
981 982
ifExtension pred bits _ _ _ = pred bits

983 984 985
multiline_doc_comment :: Action
multiline_doc_comment span buf _len = withLexedDocType (worker "")
  where
986
    worker commentAcc input docType checkNextLine = case alexGetChar' input of
Ian Lynagh's avatar
Ian Lynagh committed
987
      Just ('\n', input')
988 989
        | checkNextLine -> case checkIfCommentLine input' of
          Just input -> worker ('\n':commentAcc) input docType checkNextLine
990
          Nothing -> docCommentEnd input commentAcc docType buf span
991 992
        | otherwise -> docCommentEnd input commentAcc docType buf span
      Just (c, input) -> worker (c:commentAcc) input docType checkNextLine
993
      Nothing -> docCommentEnd input commentAcc docType buf span
Ian Lynagh's avatar
Ian Lynagh committed
994

995 996 997 998 999 1000 1001 1002
    -- Check if the next line of input belongs to this doc comment as well.
    -- A doc comment continues onto the next line when the following
    -- conditions are met:
    --   * The line starts with "--"
    --   * The line doesn't start with "---".
    --   * The line doesn't start with "-- $", because that would be the
    --     start of a /new/ named haddock chunk (#10398).
    checkIfCommentLine :: AlexInput -> Maybe AlexInput
1003 1004
    checkIfCommentLine input = check (dropNonNewlineSpace input)
      where
1005 1006 1007 1008 1009 1010 1011 1012 1013 1014
        check input = do
          ('-', input) <- alexGetChar' input
          ('-', input) <- alexGetChar' input
          (c, after_c) <- alexGetChar' input
          case c of
            '-' -> Nothing
            ' ' -> case alexGetChar' after_c of
                     Just ('$', _) -> Nothing
                     _ -> Just input
            _   -> Just input
1015

1016
        dropNonNewlineSpace input = case alexGetChar' input of
Ian Lynagh's avatar
Ian Lynagh committed
1017
          Just (c, input')
1018 1019 1020 1021
            | isSpace c && c /= '\n' -> dropNonNewlineSpace input'
            | otherwise -> input
          Nothing -> input

Jedai's avatar
Jedai committed
1022 1023 1024 1025 1026
lineCommentToken :: Action
lineCommentToken span buf len = do
  b <- extension rawTokenStreamEnabled
  if b then strtoken ITlineComment span buf len else lexToken

1027 1028 1029 1030
{-
  nested comments require traversing by hand, they can't be parsed
  using regular expressions.
-}
Ian Lynagh's avatar
Ian Lynagh committed
1031
nested_comment :: P (RealLocated Token) -> Action
Alan Zimmerman's avatar
Alan Zimmerman committed
1032
nested_comment cont span buf len = do
1033
  input <- getInput
1034
  go (reverse $ lexemeToString buf len) (1::Int) input
1035
  where
Alan Zimmerman's avatar
Alan Zimmerman committed
1036 1037 1038 1039 1040 1041
    go commentAcc 0 input = do
      setInput input
      b <- extension rawTokenStreamEnabled
      if b
        then docCommentEnd input commentAcc ITblockComment buf span
        else cont
1042
    go commentAcc n input = case alexGetChar' input of
1043
      Nothing -> errBrace input span
1044
      Just ('-',input) -> case alexGetChar' input of
1045
        Nothing  -> errBrace input span
1046
        Just ('\125',input) -> go ('\125':'-':commentAcc) (n-1) input -- '}'
Jedai's avatar
Jedai committed
1047
        Just (_,_)          -> go ('-':commentAcc) n input
1048
      Just ('\123',input) -> case alexGetChar' input of  -- '{' char
1049
        Nothing  -> errBrace input span
Jedai's avatar
Jedai committed
1050 1051 1052
        Just ('-',input) -> go ('-':'\123':commentAcc) (n+1) input
        Just (_,_)       -> go ('\123':commentAcc) n input
      Just (c,input) -> go (c:commentAcc) n input
1053 1054 1055 1056

nested_doc_comment :: Action
nested_doc_comment span buf _len = withLexedDocType (go "")
  where
1057
    go commentAcc input docType _ = case alexGetChar' input of
1058
      Nothing -> errBrace input span
1059
      Just ('-',input) -> case alexGetChar' input of
1060
        Nothing -> errBrace input span
twanvl's avatar
twanvl committed
1061
        Just ('\125',input) ->