Lexer.x 102 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
-- XXX The above flags turn off warnings in the generated code:
45
{-# LANGUAGE BangPatterns #-}
Alan Zimmerman's avatar
Alan Zimmerman committed
46
{-# LANGUAGE DeriveDataTypeable #-}
47 48 49 50 51 52 53 54
{-# OPTIONS_GHC -fno-warn-unused-matches #-}
{-# OPTIONS_GHC -fno-warn-unused-binds #-}
{-# OPTIONS_GHC -fno-warn-unused-imports #-}
{-# OPTIONS_GHC -fno-warn-missing-signatures #-}
-- But alex still generates some code that causes the "lazy unlifted bindings"
-- warning, and old compilers don't know about it so we can't easily turn
-- it off, so for now we use the sledge hammer:
{-# OPTIONS_GHC -w #-}
55

56 57
{-# OPTIONS_GHC -funbox-strict-fields #-}

58
module Lexer (
Alan Zimmerman's avatar
Alan Zimmerman committed
59
   Token(..), lexer, pragState, mkPState, PState(..),
Ian Lynagh's avatar
Ian Lynagh committed
60
   P(..), ParseResult(..), getSrcLoc,
61
   getPState, getDynFlags, withThisPackage,
62
   failLocMsgP, failSpanMsgP, srcParseFail,
Ian Lynagh's avatar
Ian Lynagh committed
63
   getMessages,
64
   popContext, pushCurrentContext, setLastToken, setSrcLoc,
vivian's avatar
vivian committed
65
   activeContext, nextIsEOF,
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
66
   getLexState, popLexState, pushLexState,
67
   extension, bangPatEnabled, datatypeContextsEnabled,
68
   traditionalRecordSyntaxEnabled,
69 70
   explicitForallEnabled,
   inRulePrag,
cactus's avatar
cactus committed
71 72 73
   explicitNamespacesEnabled,
   patternSynonymsEnabled,
   sccProfilingOn, hpcEnabled,
Jedai's avatar
Jedai committed
74
   addWarning,
Alan Zimmerman's avatar
Alan Zimmerman committed
75
   lexTokenStream,
Alan Zimmerman's avatar
Alan Zimmerman committed
76
   addAnnotation,AddAnn,mkParensApiAnn
77 78
  ) where

rodlogic's avatar
rodlogic committed
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95
-- base
import Control.Applicative
import Control.Monad
import Data.Bits
import Data.Char
import Data.List
import Data.Maybe
import Data.Ratio
import Data.Word

-- bytestring
import Data.ByteString (ByteString)

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

Alan Zimmerman's avatar
Alan Zimmerman committed
96 97 98 99
-- data/typeable
import Data.Data
import Data.Typeable

rodlogic's avatar
rodlogic committed
100
-- compiler/utils
101
import Bag
102 103 104 105
import Outputable
import StringBuffer
import FastString
import UniqFM
rodlogic's avatar
rodlogic committed
106 107 108 109
import Util             ( readRational )

-- compiler/main
import ErrUtils
110
import DynFlags
rodlogic's avatar
rodlogic committed
111 112 113

-- compiler/basicTypes
import SrcLoc
114
import Module
Alan Zimmerman's avatar
Alan Zimmerman committed
115 116
import BasicTypes     ( InlineSpec(..), RuleMatchInfo(..), FractionalLit(..),
                        SourceText )
117

rodlogic's avatar
rodlogic committed
118 119
-- compiler/parser
import Ctype
Alan Zimmerman's avatar
Alan Zimmerman committed
120 121

import ApiAnnotation
122 123
}

124 125 126
-- -----------------------------------------------------------------------------
-- Alex "Character set macros"

127 128
-- NB: The logic behind these definitions is also reflected in basicTypes/Lexeme.hs
-- Any changes here should likely be reflected there.
129 130 131 132
$unispace    = \x05 -- Trick Alex into handling Unicode. See alexGetByte.
$nl          = [\n\r\f]
$whitechar   = [$nl\v\ $unispace]
$white_no_nl = $whitechar # \n -- TODO #8424
133
$tab         = \t
134 135

$ascdigit  = 0-9
136
$unidigit  = \x03 -- Trick Alex into handling Unicode. See alexGetByte.
137
$decdigit  = $ascdigit -- for now, should really be $digit (ToDo)
138 139 140
$digit     = [$ascdigit $unidigit]

$special   = [\(\)\,\;\[\]\`\{\}]
141 142 143
$ascsymbol = [\!\#\$\%\&\*\+\.\/\<\=\>\?\@\\\^\|\-\~\:]
$unisymbol = \x04 -- Trick Alex into handling Unicode. See alexGetByte.
$symbol    = [$ascsymbol $unisymbol] # [$special \_\"\']
144

145
$unilarge  = \x01 -- Trick Alex into handling Unicode. See alexGetByte.
146
$asclarge  = [A-Z]
147 148
$large     = [$asclarge $unilarge]

149
$unismall  = \x02 -- Trick Alex into handling Unicode. See alexGetByte.
150
$ascsmall  = [a-z]
151 152
$small     = [$ascsmall $unismall \_]

153 154
$unigraphic = \x06 -- Trick Alex into handling Unicode. See alexGetByte.
$graphic   = [$small $large $symbol $digit $special $unigraphic \"\']
155

156
$binit     = 0-1
Ian Lynagh's avatar
Ian Lynagh committed
157
$octit     = 0-7
158
$hexit     = [$decdigit A-F a-f]
159 160
$idchar    = [$small $large $digit \']

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+         { warn Opt_WarnTabs (text "Tab character") }
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.

Jedai's avatar
Jedai committed
230
"-- " ~[$docsym \#] .* { lineCommentToken }
231
"--" [^$symbol \ ] .* { lineCommentToken }
232 233 234

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

Jedai's avatar
Jedai committed
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
265 266 267 268 269
  \n                                    ;
  ^\# (line)?                           { begin line_prag1 }
  ^\# pragma .* \n                      ; -- GCC 3.3 CPP generated, apparently
  ^\# \! .* \n                          ; -- #!, for scripts
  ()                                    { do_bol }
270 271 272 273 274
}

-- 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.
275
<layout, layout_do, layout_if> {
Ian Lynagh's avatar
Ian Lynagh committed
276 277 278 279
  \{ / { notFollowedBy '-' }            { hopefully_open_brace }
        -- we might encounter {-# here, but {- has been handled already
  \n                                    ;
  ^\# (line)?                           { begin line_prag1 }
280 281
}

282 283 284 285 286 287
-- after an 'if', a vertical bar starts a layout context for MultiWayIf
<layout_if> {
  \| / { notFollowedBySymbol }          { new_layout_context True ITvbar }
  ()                                    { pop }
}

288
-- do is treated in a subtly different way, see new_layout_context
289 290
<layout>    ()                          { new_layout_context True  ITvocurly }
<layout_do> ()                          { new_layout_context False ITvocurly }
291 292 293 294

-- 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
295
<layout_left>  ()                       { do_layout_left }
296

Ian Lynagh's avatar
Ian Lynagh committed
297
<0,option_prags> \n                     { begin bol }
298

299 300
"{-#" $whitechar* $pragmachar+ / { known_pragma linePrags }
                                { dispatch_pragmas linePrags }
301 302 303

-- single-line line pragmas, of the form
--    # <line> "<file>" <extra-stuff> \n
304
<line_prag1> @decimal                   { setLine line_prag1a }
Ian Lynagh's avatar
Ian Lynagh committed
305 306
<line_prag1a> \" [$graphic \ ]* \"      { setFile line_prag1b }
<line_prag1b> .*                        { pop }
307 308 309

-- Haskell-style line pragmas, of the form
--    {-# LINE <line> "<file>" #-}
310
<line_prag2> @decimal                   { setLine line_prag2a }
Ian Lynagh's avatar
Ian Lynagh committed
311 312
<line_prag2a> \" [$graphic \ ]* \"      { setFile line_prag2b }
<line_prag2b> "#-}"|"-}"                { pop }
313 314
   -- NOTE: accept -} at the end of a LINE pragma, for compatibility
   -- with older versions of GHC which generated these.
315

316
<0,option_prags> {
Ian Lynagh's avatar
Ian Lynagh committed
317
  "{-#" $whitechar* $pragmachar+
318 319 320 321 322
        $whitechar+ $pragmachar+ / { known_pragma twoWordPrags }
                                 { dispatch_pragmas twoWordPrags }

  "{-#" $whitechar* $pragmachar+ / { known_pragma oneWordPrags }
                                 { dispatch_pragmas oneWordPrags }
323

324
  -- We ignore all these pragmas, but don't generate a warning for them
325 326
  "{-#" $whitechar* $pragmachar+ / { known_pragma ignoredPrags }
                                 { dispatch_pragmas ignoredPrags }
327 328

  -- ToDo: should only be valid inside a pragma:
Ian Lynagh's avatar
Ian Lynagh committed
329
  "#-}"                          { endPrag }
330 331
}

332
<option_prags> {
333 334 335
  "{-#"  $whitechar* $pragmachar+ / { known_pragma fileHeaderPrags }
                                   { dispatch_pragmas fileHeaderPrags }

Ian Lynagh's avatar
Ian Lynagh committed
336
  "-- #"                           { multiline_doc_comment }
David Waern's avatar
David Waern committed
337 338
}

339 340
<0> {
  -- In the "0" mode we ignore these pragmas
341
  "{-#"  $whitechar* $pragmachar+ / { known_pragma fileHeaderPrags }
342 343 344
                     { nested_comment lexToken }
}

David Waern's avatar
David Waern committed
345
<0> {
Jedai's avatar
Jedai committed
346
  "-- #" .* { lineCommentToken }
347
}
348

349
<0,option_prags> {
350 351
  "{-#"  { warnThen Opt_WarnUnrecognisedPragmas (text "Unrecognised pragma")
                    (nested_comment lexToken) }
352 353
}

354 355
-- '0' state: ordinary lexemes

356 357
-- Haddock comments

358
<0,option_prags> {
David Waern's avatar
David Waern committed
359 360
  "-- " $docsym      / { ifExtension haddockEnabled } { multiline_doc_comment }
  "{-" \ ? $docsym   / { ifExtension haddockEnabled } { nested_doc_comment }
361 362
}

363 364
-- "special" symbols

365
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
366 367
  "[:" / { ifExtension parrEnabled }    { token ITopabrack }
  ":]" / { ifExtension parrEnabled }    { token ITcpabrack }
368
}
Ian Lynagh's avatar
Ian Lynagh committed
369

370
<0> {
371 372 373
  "[|"        / { ifExtension thEnabled } { token ITopenExpQuote }
  "[||"       / { ifExtension thEnabled } { token ITopenTExpQuote }
  "[e|"       / { ifExtension thEnabled } { token ITopenExpQuote }
374
  "[e||"      / { ifExtension thEnabled } { token ITopenTExpQuote }
375 376 377 378 379 380 381 382 383
  "[p|"       / { ifExtension thEnabled } { token ITopenPatQuote }
  "[d|"       / { ifExtension thEnabled } { layout_token ITopenDecQuote }
  "[t|"       / { ifExtension thEnabled } { token ITopenTypQuote }
  "|]"        / { ifExtension thEnabled } { token ITcloseQuote }
  "||]"       / { ifExtension thEnabled } { token ITcloseTExpQuote }
  \$ @varid   / { ifExtension thEnabled } { skip_one_varid ITidEscape }
  "$$" @varid / { ifExtension thEnabled } { skip_two_varid ITidTyEscape }
  "$("        / { ifExtension thEnabled } { token ITparenEscape }
  "$$("       / { ifExtension thEnabled } { token ITparenTyEscape }
384

385 386 387 388
-- For backward compatibility, accept the old dollar syntax
  "[$" @varid "|"  / { ifExtension qqEnabled }
                     { lex_quasiquote_tok }

389
  "[" @varid "|"  / { ifExtension qqEnabled }
390
                     { lex_quasiquote_tok }
391 392

  -- qualified quasi-quote (#5555)
393 394
  "[" @qvarid "|"  / { ifExtension qqEnabled }
                     { lex_qquasiquote_tok }
395 396
}

397
<0> {
ross's avatar
ross committed
398
  "(|" / { ifExtension arrowsEnabled `alexAndPred` notFollowedBySymbol }
Ian Lynagh's avatar
Ian Lynagh committed
399
                                        { special IToparenbar }
400 401 402
  "|)" / { ifExtension arrowsEnabled }  { special ITcparenbar }
}

403
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
404
  \? @varid / { ifExtension ipEnabled } { skip_one_varid ITdupipvarid }
405 406
}

407
<0> {
408
  "(#" / { ifExtension unboxedTuplesEnabled }
Ian Lynagh's avatar
Ian Lynagh committed
409 410 411 412 413
         { token IToubxparen }
  "#)" / { ifExtension unboxedTuplesEnabled }
         { token ITcubxparen }
}

414
<0,option_prags> {
Ian Lynagh's avatar
Ian Lynagh committed
415 416 417 418 419 420 421 422 423 424
  \(                                    { special IToparen }
  \)                                    { special ITcparen }
  \[                                    { special ITobrack }
  \]                                    { special ITcbrack }
  \,                                    { special ITcomma }
  \;                                    { special ITsemi }
  \`                                    { special ITbackquote }

  \{                                    { open_brace }
  \}                                    { close_brace }
425 426
}

427
<0,option_prags> {
428 429
  @qvarid                       { idtoken qvarid }
  @qconid                       { idtoken qconid }
Ian Lynagh's avatar
Ian Lynagh committed
430
  @varid                        { varid }
431
  @conid                        { idtoken conid }
432 433
}

434
<0> {
435 436
  @qvarid "#"+      / { ifExtension magicHashEnabled } { idtoken qvarid }
  @qconid "#"+      / { ifExtension magicHashEnabled } { idtoken qconid }
437
  @varid "#"+       / { ifExtension magicHashEnabled } { varid }
438
  @conid "#"+       / { ifExtension magicHashEnabled } { idtoken conid }
439 440
}

441 442
-- ToDo: - move `var` and (sym) into lexical syntax?
--       - remove backquote from $special?
443
<0> {
444 445
  @qvarsym                                         { idtoken qvarsym }
  @qconsym                                         { idtoken qconsym }
446 447
  @varsym                                          { varsym }
  @consym                                          { consym }
448 449
}

450 451
-- For the normal boxed literals we need to be careful
-- when trying to be close to Haskell98
452
<0> {
453
  -- Normal integral literals (:: Num a => a, from Integer)
454
  @decimal                                                               { tok_num positive 0 0 decimal }
455
  0[bB] @binary                / { ifExtension binaryLiteralsEnabled }   { tok_num positive 2 2 binary }
456 457 458
  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 }
459 460
  @negative 0[bB] @binary      / { ifExtension negativeLiteralsEnabled `alexAndPred`
                                   ifExtension binaryLiteralsEnabled }   { tok_num negative 3 3 binary }
461 462
  @negative 0[oO] @octal       / { ifExtension negativeLiteralsEnabled } { tok_num negative 3 3 octal }
  @negative 0[xX] @hexadecimal / { ifExtension negativeLiteralsEnabled } { tok_num negative 3 3 hexadecimal }
463 464

  -- Normal rational literals (:: Fractional a => a, from Rational)
465 466
  @floating_point                                                        { strtoken tok_float }
  @negative @floating_point    / { ifExtension negativeLiteralsEnabled } { strtoken tok_float }
467 468
}

469
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
470
  -- Unboxed ints (:: Int#) and words (:: Word#)
471 472
  -- It's simpler (and faster?) to give separate cases to the negatives,
  -- especially considering octal/hexadecimal prefixes.
Ian Lynagh's avatar
Ian Lynagh committed
473
  @decimal                     \# / { ifExtension magicHashEnabled } { tok_primint positive 0 1 decimal }
474 475
  0[bB] @binary                \# / { ifExtension magicHashEnabled `alexAndPred`
                                      ifExtension binaryLiteralsEnabled } { tok_primint positive 2 3 binary }
Ian Lynagh's avatar
Ian Lynagh committed
476 477 478
  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 }
479 480
  @negative 0[bB] @binary      \# / { ifExtension magicHashEnabled `alexAndPred`
                                      ifExtension binaryLiteralsEnabled } { tok_primint negative 3 4 binary }
Ian Lynagh's avatar
Ian Lynagh committed
481
  @negative 0[oO] @octal       \# / { ifExtension magicHashEnabled } { tok_primint negative 3 4 octal }
482
  @negative 0[xX] @hexadecimal \# / { ifExtension magicHashEnabled } { tok_primint negative 3 4 hexadecimal }
483

Ian Lynagh's avatar
Ian Lynagh committed
484
  @decimal                     \# \# / { ifExtension magicHashEnabled } { tok_primword 0 2 decimal }
485 486
  0[bB] @binary                \# \# / { ifExtension magicHashEnabled `alexAndPred`
                                         ifExtension binaryLiteralsEnabled } { tok_primword 2 4 binary }
Ian Lynagh's avatar
Ian Lynagh committed
487 488 489
  0[oO] @octal                 \# \# / { ifExtension magicHashEnabled } { tok_primword 2 4 octal }
  0[xX] @hexadecimal           \# \# / { ifExtension magicHashEnabled } { tok_primword 2 4 hexadecimal }

490 491
  -- Unboxed floats and doubles (:: Float#, :: Double#)
  -- prim_{float,double} work with signed literals
492 493
  @signed @floating_point \# / { ifExtension magicHashEnabled } { init_strtoken 1 tok_primfloat }
  @signed @floating_point \# \# / { ifExtension magicHashEnabled } { init_strtoken 2 tok_primdouble }
494 495 496 497 498 499
}

-- 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.
500
<0> {
Ian Lynagh's avatar
Ian Lynagh committed
501 502
  \'                            { lex_char_tok }
  \"                            { lex_string_tok }
503 504
}

505 506 507 508

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

509
{
510

511 512 513
-- -----------------------------------------------------------------------------
-- The token type

514
data Token
Ian Lynagh's avatar
Ian Lynagh committed
515
  = ITas                        -- Haskell keywords
516 517 518 519 520 521 522 523
  | ITcase
  | ITclass
  | ITdata
  | ITdefault
  | ITderiving
  | ITdo
  | ITelse
  | IThiding
524
  | ITforeign
525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540
  | ITif
  | ITimport
  | ITin
  | ITinfix
  | ITinfixl
  | ITinfixr
  | ITinstance
  | ITlet
  | ITmodule
  | ITnewtype
  | ITof
  | ITqualified
  | ITthen
  | ITtype
  | ITwhere

Ian Lynagh's avatar
Ian Lynagh committed
541
  | ITforall                    -- GHC extension keywords
542 543 544 545
  | ITexport
  | ITlabel
  | ITdynamic
  | ITsafe
546
  | ITinterruptible
547 548 549
  | ITunsafe
  | ITstdcallconv
  | ITccallconv
550
  | ITcapiconv
551
  | ITprimcallconv
thoughtpolice's avatar
thoughtpolice committed
552
  | ITjavascriptcallconv
553
  | ITmdo
554
  | ITfamily
555
  | ITrole
556 557 558
  | ITgroup
  | ITby
  | ITusing
cactus's avatar
cactus committed
559
  | ITpattern
Facundo Domínguez's avatar
Facundo Domínguez committed
560
  | ITstatic
561

Alan Zimmerman's avatar
Alan Zimmerman committed
562 563 564 565 566 567 568 569
  -- 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
570
  | ITline_prag
Alan Zimmerman's avatar
Alan Zimmerman committed
571 572 573 574 575 576
  | ITscc_prag          SourceText
  | ITgenerated_prag    SourceText
  | ITcore_prag         SourceText         -- hdaume: core annotations
  | ITunpack_prag       SourceText
  | ITnounpack_prag     SourceText
  | ITann_prag          SourceText
577
  | ITclose_prag
578 579 580
  | IToptions_prag String
  | ITinclude_prag String
  | ITlanguage_prag
Alan Zimmerman's avatar
Alan Zimmerman committed
581 582 583 584 585 586 587 588 589
  | 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
590

Ian Lynagh's avatar
Ian Lynagh committed
591
  | ITdotdot                    -- reserved symbols
592 593 594 595
  | ITcolon
  | ITdcolon
  | ITequal
  | ITlam
596
  | ITlcase
597 598 599 600 601
  | ITvbar
  | ITlarrow
  | ITrarrow
  | ITat
  | ITtilde
batterseapower's avatar
batterseapower committed
602
  | ITtildehsh
603 604 605 606 607 608
  | ITdarrow
  | ITminus
  | ITbang
  | ITstar
  | ITdot

Ian Lynagh's avatar
Ian Lynagh committed
609
  | ITbiglam                    -- GHC-extension symbols
610

Ian Lynagh's avatar
Ian Lynagh committed
611
  | ITocurly                    -- special symbols
612 613 614 615
  | ITccurly
  | ITvocurly
  | ITvccurly
  | ITobrack
Ian Lynagh's avatar
Ian Lynagh committed
616 617
  | ITopabrack                  -- [:, for parallel arrays with -XParallelArrays
  | ITcpabrack                  -- :], for parallel arrays with -XParallelArrays
618 619 620 621 622 623 624 625 626
  | ITcbrack
  | IToparen
  | ITcparen
  | IToubxparen
  | ITcubxparen
  | ITsemi
  | ITcomma
  | ITunderscore
  | ITbackquote
dreixel's avatar
dreixel committed
627
  | ITsimpleQuote               --  '
628

Ian Lynagh's avatar
Ian Lynagh committed
629
  | ITvarid   FastString        -- identifiers
630 631 632 633 634 635 636
  | ITconid   FastString
  | ITvarsym  FastString
  | ITconsym  FastString
  | ITqvarid  (FastString,FastString)
  | ITqconid  (FastString,FastString)
  | ITqvarsym (FastString,FastString)
  | ITqconsym (FastString,FastString)
637 638
  | ITprefixqvarsym (FastString,FastString)
  | ITprefixqconsym (FastString,FastString)
639

Ian Lynagh's avatar
Ian Lynagh committed
640
  | ITdupipvarid   FastString   -- GHC extension: implicit param: ?x
641

Alan Zimmerman's avatar
Alan Zimmerman committed
642 643 644 645
  | 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
  | ITrational FractionalLit
646

Alan Zimmerman's avatar
Alan Zimmerman committed
647 648 649 650
  | 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
651 652
  | ITprimfloat  FractionalLit
  | ITprimdouble FractionalLit
653

654
  -- Template Haskell extension tokens
Ian Lynagh's avatar
Ian Lynagh committed
655 656 657 658 659
  | ITopenExpQuote              --  [| or [e|
  | ITopenPatQuote              --  [p|
  | ITopenDecQuote              --  [d|
  | ITopenTypQuote              --  [t|
  | ITcloseQuote                --  |]
660 661
  | ITopenTExpQuote             --  [||
  | ITcloseTExpQuote            --  ||]
Ian Lynagh's avatar
Ian Lynagh committed
662 663
  | ITidEscape   FastString     --  $x
  | ITparenEscape               --  $(
664 665
  | ITidTyEscape   FastString   --  $$x
  | ITparenTyEscape             --  $$(
Ian Lynagh's avatar
Ian Lynagh committed
666
  | ITtyQuote                   --  ''
667 668 669 670 671 672 673 674
  | 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 |]
675 676 677 678

  -- Arrow notation extension
  | ITproc
  | ITrec
Ian Lynagh's avatar
Ian Lynagh committed
679 680 681 682 683 684
  | IToparenbar                 --  (|
  | ITcparenbar                 --  |)
  | ITlarrowtail                --  -<
  | ITrarrowtail                --  >-
  | ITLarrowtail                --  -<<
  | ITRarrowtail                --  >>-
685

Ian Lynagh's avatar
Ian Lynagh committed
686 687
  | ITunknown String            -- Used when the lexer can't make sense of it
  | ITeof                       -- end of file token
688 689 690 691 692 693 694

  -- 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)
David Waern's avatar
David Waern committed
695
  | ITdocOptionsOld   String     -- doc options declared "-- # ..."-style
Jedai's avatar
Jedai committed
696 697
  | ITlineComment     String     -- comment starting by "--"
  | ITblockComment    String     -- comment in {- -}
698

699
  deriving Show
700

Alan Zimmerman's avatar
Alan Zimmerman committed
701 702 703
instance Outputable Token where
  ppr x = text (show x)

Alan Zimmerman's avatar
Alan Zimmerman committed
704

705 706 707 708 709 710 711
-- 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)
--
712
reservedWordsFM :: UniqFM (Token, ExtsBitmap)
713
reservedWordsFM = listToUFM $
Ian Lynagh's avatar
Ian Lynagh committed
714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740
    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 ),

741 742 743
         ( "forall",         ITforall,        xbit ExplicitForallBit .|.
                                              xbit InRulePragBit),
         ( "mdo",            ITmdo,           xbit RecursiveDoBit),
744 745 746
             -- See Note [Lexing type pseudo-keywords]
         ( "family",         ITfamily,        0 ),
         ( "role",           ITrole,          0 ),
747
         ( "pattern",        ITpattern,       xbit PatternSynonymsBit),
Facundo Domínguez's avatar
Facundo Domínguez committed
748
         ( "static",         ITstatic,        0 ),
749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769
         ( "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)
770 771
     ]

772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787
{-----------------------------------
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.
-------------------------------------}
788

789
reservedSymsFM :: UniqFM (Token, ExtsBitmap -> Bool)
790
reservedSymsFM = listToUFM $
Ian Lynagh's avatar
Ian Lynagh committed
791 792 793 794 795 796 797 798 799 800 801 802
    map (\ (x,y,z) -> (mkFastString x,(y,z)))
      [ ("..",  ITdotdot,   always)
        -- (:) is a reserved op, meaning only list cons
       ,(":",   ITcolon,    always)
       ,("::",  ITdcolon,   always)
       ,("=",   ITequal,    always)
       ,("\\",  ITlam,      always)
       ,("|",   ITvbar,     always)
       ,("<-",  ITlarrow,   always)
       ,("->",  ITrarrow,   always)
       ,("@",   ITat,       always)
       ,("~",   ITtilde,    always)
803
       ,("~#",  ITtildehsh, magicHashEnabled)
Ian Lynagh's avatar
Ian Lynagh committed
804 805 806 807 808
       ,("=>",  ITdarrow,   always)
       ,("-",   ITminus,    always)
       ,("!",   ITbang,     always)

        -- For data T (a::*) = MkT
809
       ,("*", ITstar, always) -- \i -> kindSigsEnabled i || tyFamEnabled i)
Ian Lynagh's avatar
Ian Lynagh committed
810
        -- For 'forall a . t'
811
       ,(".", ITdot,  always) -- \i -> explicitForallEnabled i || inRulePrag i)
Ian Lynagh's avatar
Ian Lynagh committed
812 813 814 815 816

       ,("-<",  ITlarrowtail, arrowsEnabled)
       ,(">-",  ITrarrowtail, arrowsEnabled)
       ,("-<<", ITLarrowtail, arrowsEnabled)
       ,(">>-", ITRarrowtail, arrowsEnabled)
817

Ian Lynagh's avatar
Ian Lynagh committed
818 819
       ,("∷",   ITdcolon, unicodeSyntaxEnabled)
       ,("⇒",   ITdarrow, unicodeSyntaxEnabled)
820
       ,("∀",   ITforall, unicodeSyntaxEnabled)
Ian Lynagh's avatar
Ian Lynagh committed
821 822
       ,("→",   ITrarrow, unicodeSyntaxEnabled)
       ,("←",   ITlarrow, unicodeSyntaxEnabled)
823 824 825 826 827 828 829 830

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

       ,("★", ITstar, unicodeSyntaxEnabled)

Simon Marlow's avatar
Simon Marlow committed
831 832 833
        -- 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).
834 835 836 837 838
       ]

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

Ian Lynagh's avatar
Ian Lynagh committed
839
type Action = RealSrcSpan -> StringBuffer -> Int -> P (RealLocated Token)
840

841
special :: Token -> Action
twanvl's avatar
twanvl committed
842
special tok span _buf _len = return (L span tok)
843

844
token, layout_token :: Token -> Action
twanvl's avatar
twanvl committed
845 846
token t span _buf _len = return (L span t)
layout_token t span _buf _len = pushLexState layout >> return (L span t)
847

848 849
idtoken :: (StringBuffer -> Int -> Token) -> Action
idtoken f span buf len = return (L span $! (f buf len))
850

851
skip_one_varid :: (FastString -> Token) -> Action
Ian Lynagh's avatar
Ian Lynagh committed
852
skip_one_varid f span buf len
853
  = return (L span $! f (lexemeToFastString (stepOn buf) (len-1)))
854

855 856 857 858
skip_two_varid :: (FastString -> Token) -> Action
skip_two_varid f span buf len
  = return (L span $! f (lexemeToFastString (stepOn (stepOn buf)) (len-2)))

859
strtoken :: (String -> Token) -> Action
Ian Lynagh's avatar
Ian Lynagh committed
860
strtoken f span buf len =
861
  return (L span $! (f $! lexemeToString buf len))
862

863
init_strtoken :: Int -> (String -> Token) -> Action
864
-- like strtoken, but drops the last N character(s)
Ian Lynagh's avatar
Ian Lynagh committed
865
init_strtoken drop f span buf len =
866
  return (L span $! (f $! lexemeToString buf (len-drop)))
867 868

begin :: Int -> Action
869
begin code _span _str _len = do pushLexState code; lexToken
870 871

pop :: Action
Ian Lynagh's avatar
Ian Lynagh committed
872 873
pop _span _buf _len = do _ <- popLexState
                         lexToken
874

875 876 877 878 879 880 881 882 883 884 885
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
886
              else failSpanMsgP (RealSrcSpan span) (text "Missing block")
887

888
pop_and :: Action -> Action
Ian Lynagh's avatar
Ian Lynagh committed
889 890
pop_and act span buf len = do _ <- popLexState
                              act span buf len
891

892
{-# INLINE nextCharIs #-}
Ian Lynagh's avatar
Ian Lynagh committed
893
nextCharIs :: StringBuffer -> (Char -> Bool) -> Bool
894 895
nextCharIs buf p = not (atEnd buf) && p (currentChar buf)

896 897 898 899
{-# INLINE nextCharIsNot #-}
nextCharIsNot :: StringBuffer -> (Char -> Bool) -> Bool
nextCharIsNot buf p = not (nextCharIs buf p)

900
notFollowedBy :: Char -> AlexAccPred ExtsBitmap
Ian Lynagh's avatar
Ian Lynagh committed
901
notFollowedBy char _ _ _ (AI _ buf)
902
  = nextCharIsNot buf (== char)
903

904
notFollowedBySymbol :: AlexAccPred ExtsBitmap
905
notFollowedBySymbol _ _ _ (AI _ buf)
906
  = nextCharIsNot buf (`elem` "!#$%&*+./<=>?@\\^|-~")
ross's avatar
ross committed
907

908 909 910 911 912
-- 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.
913
isNormalComment :: AlexAccPred ExtsBitmap
914
isNormalComment bits _ _ (AI _ buf)
915
  | haddockEnabled bits = notFollowedByDocOrPragma
916
  | otherwise           = nextCharIsNot buf (== '#')
917 918
  where
    notFollowedByDocOrPragma
919
       = afterOptionalSpace buf (\b -> nextCharIsNot b (`elem` "|^*$#"))
920

921 922 923 924 925
afterOptionalSpace :: StringBuffer -> (StringBuffer -> Bool) -> Bool
afterOptionalSpace buf p
    = if nextCharIs buf (== ' ')
      then p (snd (nextChar buf))
      else p buf
926

927
atEOL :: AlexAccPred ExtsBitmap
928
atEOL _ _ _ (AI _ buf) = atEnd buf || currentChar buf == '\n'
929

930
ifExtension :: (ExtsBitmap -> Bool) -> AlexAccPred ExtsBitmap
931 932
ifExtension pred bits _ _ _ = pred bits

933 934 935
multiline_doc_comment :: Action
multiline_doc_comment span buf _len = withLexedDocType (worker "")
  where
936
    worker commentAcc input docType oneLine = case alexGetChar' input of
Ian Lynagh's avatar
Ian Lynagh committed
937
      Just ('\n', input')
938 939 940 941 942 943
        | oneLine -> docCommentEnd input commentAcc docType buf span
        | otherwise -> case checkIfCommentLine input' of
          Just input -> worker ('\n':commentAcc) input docType False
          Nothing -> docCommentEnd input commentAcc docType buf span
      Just (c, input) -> worker (c:commentAcc) input docType oneLine
      Nothing -> docCommentEnd input commentAcc docType buf span
Ian Lynagh's avatar
Ian Lynagh committed
944

945 946
    checkIfCommentLine input = check (dropNonNewlineSpace input)
      where
947 948 949
        check input = case alexGetChar' input of
          Just ('-', input) -> case alexGetChar' input of
            Just ('-', input) -> case alexGetChar' input of
950 951 952 953 954
              Just (c, _) | c /= '-' -> Just input
              _ -> Nothing
            _ -> Nothing
          _ -> Nothing

955
        dropNonNewlineSpace input = case alexGetChar' input of
Ian Lynagh's avatar
Ian Lynagh committed
956
          Just (c, input')
957 958 959 960
            | isSpace c && c /= '\n' -> dropNonNewlineSpace input'
            | otherwise -> input
          Nothing -> input

Jedai's avatar
Jedai committed
961 962 963 964 965
lineCommentToken :: Action
lineCommentToken span buf len = do
  b <- extension rawTokenStreamEnabled
  if b then strtoken ITlineComment span buf len else lexToken

966 967 968 969
{-
  nested comments require traversing by hand, they can't be parsed
  using regular expressions.
-}
Ian Lynagh's avatar
Ian Lynagh committed
970
nested_comment :: P (RealLocated Token) -> Action
Alan Zimmerman's avatar
Alan Zimmerman committed
971
nested_comment cont span buf len = do
972
  input <- getInput
Alan Zimmerman's avatar
Alan Zimmerman committed
973
  go (reverse $ drop 2 $ lexemeToString buf len) (1::Int) input
974
  where
Alan Zimmerman's avatar
Alan Zimmerman committed
975 976 977 978 979 980
    go commentAcc 0 input = do
      setInput input
      b <- extension rawTokenStreamEnabled
      if b
        then docCommentEnd input commentAcc ITblockComment buf span
        else cont
981
    go commentAcc n input = case alexGetChar' input of
982
      Nothing -> errBrace input span
983
      Just ('-',input) -> case alexGetChar' input of
984
        Nothing  -> errBrace input span
Jedai's avatar
Jedai committed
985 986
        Just ('\125',input) -> go commentAcc (n-1) input
        Just (_,_)          -> go ('-':commentAcc) n input
987
      Just ('\123',input) -> case alexGetChar' input of
988
        Nothing  -> errBrace input span
Jedai's avatar
Jedai committed
989 990 991
        Just ('-',input) -> go ('-':'\123':commentAcc) (n+1) input
        Just (_,_)       -> go ('\123':commentAcc) n input
      Just (c,input) -> go (c:commentAcc) n input
992 993 994 995

nested_doc_comment :: Action
nested_doc_comment span buf _len = withLexedDocType (go "")
  where
996
    go commentAcc input docType _ = case alexGetChar' input of
997
      Nothing -> errBrace input span
998
      Just ('-',input) -> case alexGetChar' input of
999
        Nothing -> errBrace input span
twanvl's avatar
twanvl committed
1000
        Just ('\125',input) ->
1001
          docCommentEnd input commentAcc docType buf span
twanvl's avatar
twanvl committed
1002
        Just (_,_) -> go ('-':commentAcc) input docType False
1003
      Just ('\123', input) -> case alexGetChar' input of
1004 1005 1006 1007 1008
        Nothing  -> errBrace input span
        Just ('-',input) -> do
          setInput input
          let cont = do input <- getInput; go commentAcc input docType False
          nested_comment cont span buf _len
twanvl's avatar
twanvl committed
1009
        Just (_,_) -> go ('\123':commentAcc) input docType False
1010 1011
      Just (c,input) -> go (c:commentAcc) input docType False

Ian Lynagh's avatar
Ian Lynagh committed
1012 1013
withLexedDocType :: (AlexInput -> (String -> Token) -> Bool -> P (RealLocated Token))
                 -> P (RealLocated Token)
1014
withLexedDocType lexDocComment = do
1015
  input@(AI _ buf) <- getInput
1016 1017 1018 1019
  case prevChar buf ' ' of
    '|' -> lexDocComment input ITdocCommentNext False
    '^' -> lexDocComment input ITdocCommentPrev False
    '$' -> lexDocComment input ITdocCommentNamed False
David Waern's avatar
David Waern committed
1020 1021
    '*' -> lexDocSection 1 input
    '#' -> lexDocComment input ITdocOptionsOld False
Ian Lynagh's avatar
Ian Lynagh committed
1022
    _ -> panic "withLexedDocType: Bad doc type"
Ian Lynagh's avatar
Ian Lynagh committed
1023
 where
1024
    lexDocSection n input = case alexGetChar' input of
1025
      Just ('*', input) -> lexDocSection (n+1) input
twanvl's avatar
twanvl committed
1026
      Just (_,   _)     -> lexDocComment input (ITdocSection n) True
1027 1028
      Nothing -> do setInput input; lexToken -- eof reached, lex it normally

1029 1030 1031
-- RULES pragmas turn on the forall and '.' keywords, and we turn them
-- off again at the end of the pragma.
rulePrag :: Action
Alan Zimmerman's avatar
Alan Zimmerman committed
1032
rulePrag span buf len = do
1033
  setExts (.|. xbit InRulePragBit)
Alan Zimmerman's avatar
Alan Zimmerman committed
1034 1035
  let !src = lexemeToString buf len
  return (L span (ITrules_prag src))
1036 1037

endPrag :: Action
1038
endPrag span _buf _len = do