summaryrefslogtreecommitdiff
path: root/src/Text/Pandoc/Readers/MediaWiki.hs
blob: 136701bd096e8322a9607b7a257db5ff5604ea7f (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
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
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
{-# LANGUAGE RelaxedPolyRec #-} -- needed for inlinesBetween on GHC < 7
{-
  Copyright (C) 2012 John MacFarlane <jgm@berkeley.edu>

This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
-}

{- |
   Module      : Text.Pandoc.Readers.MediaWiki
   Copyright   : Copyright (C) 2012 John MacFarlane
   License     : GNU GPL, version 2 or above

   Maintainer  : John MacFarlane <jgm@berkeley.edu>
   Stability   : alpha
   Portability : portable

Conversion of mediawiki text to 'Pandoc' document.
-}
{-
TODO:
_ correctly handle tables within tables
_ parse templates?
-}
module Text.Pandoc.Readers.MediaWiki ( readMediaWiki ) where

import Text.Pandoc.Definition
import qualified Text.Pandoc.Builder as B
import Text.Pandoc.Builder (Inlines, Blocks, trimInlines, (<>))
import Text.Pandoc.Options
import Text.Pandoc.Readers.HTML ( htmlTag, isBlockTag, isCommentTag )
import Text.Pandoc.XML ( fromEntities )
import Text.Pandoc.Parsing hiding ( nested )
import Text.Pandoc.Walk ( walk )
import Text.Pandoc.Shared ( stripTrailingNewlines, safeRead, stringify, trim )
import Data.Monoid (mconcat, mempty)
import Control.Applicative ((<$>), (<*), (*>), (<$))
import Control.Monad
import Data.List (intersperse, intercalate, isPrefixOf )
import Text.HTML.TagSoup
import Data.Sequence (viewl, ViewL(..), (<|))
import qualified Data.Foldable as F
import Data.Char (isDigit, isSpace)

-- | Read mediawiki from an input string and return a Pandoc document.
readMediaWiki :: ReaderOptions -- ^ Reader options
              -> String        -- ^ String to parse (assuming @'\n'@ line endings)
              -> Pandoc
readMediaWiki opts s =
  case runParser parseMediaWiki MWState{ mwOptions = opts
                                       , mwMaxNestingLevel = 4
                                       , mwNextLinkNumber  = 1
                                       , mwCategoryLinks = []
                                       }
       "source" (s ++ "\n") of
          Left err'    -> error $ "\nError:\n" ++ show err'
          Right result -> result

data MWState = MWState { mwOptions         :: ReaderOptions
                       , mwMaxNestingLevel :: Int
                       , mwNextLinkNumber  :: Int
                       , mwCategoryLinks   :: [Inlines]
                       }

type MWParser = Parser [Char] MWState

--
-- auxiliary functions
--

-- This is used to prevent exponential blowups for things like:
-- ''a'''a''a'''a''a'''a''a'''a
nested :: MWParser a -> MWParser a
nested p = do
  nestlevel <- mwMaxNestingLevel `fmap` getState
  guard $ nestlevel > 0
  updateState $ \st -> st{ mwMaxNestingLevel = mwMaxNestingLevel st - 1 }
  res <- p
  updateState $ \st -> st{ mwMaxNestingLevel = nestlevel }
  return res

specialChars :: [Char]
specialChars = "'[]<=&*{}|\":\\"

spaceChars :: [Char]
spaceChars = " \n\t"

sym :: String -> MWParser ()
sym s = () <$ try (string s)

newBlockTags :: [String]
newBlockTags = ["haskell","syntaxhighlight","source","gallery","references"]

isBlockTag' :: Tag String -> Bool
isBlockTag' tag@(TagOpen t _) = (isBlockTag tag || t `elem` newBlockTags) &&
  t `notElem` eitherBlockOrInline
isBlockTag' tag@(TagClose t) = (isBlockTag tag || t `elem` newBlockTags) &&
  t `notElem` eitherBlockOrInline
isBlockTag' tag = isBlockTag tag

isInlineTag' :: Tag String -> Bool
isInlineTag' (TagComment _) = True
isInlineTag' t = not (isBlockTag' t)

eitherBlockOrInline :: [String]
eitherBlockOrInline = ["applet", "button", "del", "iframe", "ins",
                               "map", "area", "object"]

htmlComment :: MWParser ()
htmlComment = () <$ htmlTag isCommentTag

inlinesInTags :: String -> MWParser Inlines
inlinesInTags tag = try $ do
  (_,raw) <- htmlTag (~== TagOpen tag [])
  if '/' `elem` raw   -- self-closing tag
     then return mempty
     else trimInlines . mconcat <$>
            manyTill inline (htmlTag (~== TagClose tag))

blocksInTags :: String -> MWParser Blocks
blocksInTags tag = try $ do
  (_,raw) <- htmlTag (~== TagOpen tag [])
  if '/' `elem` raw   -- self-closing tag
     then return mempty
     else mconcat <$> manyTill block (htmlTag (~== TagClose tag))

charsInTags :: String -> MWParser [Char]
charsInTags tag = try $ do
  (_,raw) <- htmlTag (~== TagOpen tag [])
  if '/' `elem` raw   -- self-closing tag
     then return ""
     else manyTill anyChar (htmlTag (~== TagClose tag))

--
-- main parser
--

parseMediaWiki :: MWParser Pandoc
parseMediaWiki = do
  bs <- mconcat <$> many block
  spaces
  eof
  categoryLinks <- reverse . mwCategoryLinks <$> getState
  let categories = if null categoryLinks
                      then mempty
                      else B.para $ mconcat $ intersperse B.space categoryLinks
  return $ B.doc $ bs <> categories

--
-- block parsers
--

block :: MWParser Blocks
block =  mempty <$ skipMany1 blankline
     <|> table
     <|> header
     <|> hrule
     <|> orderedList
     <|> bulletList
     <|> definitionList
     <|> mempty <$ try (spaces *> htmlComment)
     <|> preformatted
     <|> blockTag
     <|> (B.rawBlock "mediawiki" <$> template)
     <|> para

para :: MWParser Blocks
para = do
  contents <- trimInlines . mconcat <$> many1 inline
  if F.all (==Space) contents
     then return mempty
     else return $ B.para contents

table :: MWParser Blocks
table = do
  tableStart
  styles <- option [] parseAttrs <* blankline
  let tableWidth = case lookup "width" styles of
                         Just w  -> maybe 1.0 id $ parseWidth w
                         Nothing -> 1.0
  caption <- option mempty tableCaption
  optional rowsep
  hasheader <- option False $ True <$ (lookAhead (char '!'))
  (cellspecs',hdr) <- unzip <$> tableRow
  let widths = map ((tableWidth *) . snd) cellspecs'
  let restwidth = tableWidth - sum widths
  let zerocols = length $ filter (==0.0) widths
  let defaultwidth = if zerocols == 0 || zerocols == length widths
                        then 0.0
                        else restwidth / fromIntegral zerocols
  let widths' = map (\w -> if w == 0 then defaultwidth else w) widths
  let cellspecs = zip (map fst cellspecs') widths'
  rows' <- many $ try $ rowsep *> (map snd <$> tableRow)
  tableEnd
  let cols = length hdr
  let (headers,rows) = if hasheader
                          then (hdr, rows')
                          else (replicate cols mempty, hdr:rows')
  return $ B.table caption cellspecs headers rows

parseAttrs :: MWParser [(String,String)]
parseAttrs = many1 parseAttr

parseAttr :: MWParser (String, String)
parseAttr = try $ do
  skipMany spaceChar
  k <- many1 letter
  char '='
  char '"'
  v <- many1Till (satisfy (/='\n')) (char '"')
  return (k,v)

tableStart :: MWParser ()
tableStart = try $ guardColumnOne *> skipSpaces *> sym "{|"

tableEnd :: MWParser ()
tableEnd = try $ guardColumnOne *> skipSpaces *> sym "|}"

rowsep :: MWParser ()
rowsep = try $ guardColumnOne *> skipSpaces *> sym "|-" <*
               optional parseAttr <* blanklines

cellsep :: MWParser ()
cellsep = try $
             (guardColumnOne *> skipSpaces <*
                 (  (char '|' <* notFollowedBy (oneOf "-}+"))
                <|> (char '!')
                 )
             )
            <|> (() <$ try (string "||"))
            <|> (() <$ try (string "!!"))

tableCaption :: MWParser Inlines
tableCaption = try $ do
  guardColumnOne
  skipSpaces
  sym "|+"
  optional (try $ parseAttr *> skipSpaces *> char '|' *> skipSpaces)
  (trimInlines . mconcat) <$> many (notFollowedBy (cellsep <|> rowsep) *> inline)

tableRow :: MWParser [((Alignment, Double), Blocks)]
tableRow = try $ many tableCell

tableCell :: MWParser ((Alignment, Double), Blocks)
tableCell = try $ do
  cellsep
  skipMany spaceChar
  attrs <- option [] $ try $ parseAttrs <* skipSpaces <* char '|' <*
                                 notFollowedBy (char '|')
  skipMany spaceChar
  ls <- concat <$> many (notFollowedBy (cellsep <|> rowsep <|> tableEnd) *>
                     ((snd <$> withRaw table) <|> count 1 anyChar))
  bs <- parseFromString (mconcat <$> many block) ls
  let align = case lookup "align" attrs of
                    Just "left"   -> AlignLeft
                    Just "right"  -> AlignRight
                    Just "center" -> AlignCenter
                    _             -> AlignDefault
  let width = case lookup "width" attrs of
                    Just xs -> maybe 0.0 id $ parseWidth xs
                    Nothing -> 0.0
  return ((align, width), bs)

parseWidth :: String -> Maybe Double
parseWidth s =
  case reverse s of
       ('%':ds) | all isDigit ds -> safeRead ('0':'.':reverse ds)
       _ -> Nothing

template :: MWParser String
template = try $ do
  string "{{"
  notFollowedBy (char '{')
  let chunk = template <|> variable <|> many1 (noneOf "{}") <|> count 1 anyChar
  contents <- manyTill chunk (try $ string "}}")
  return $ "{{" ++ concat contents ++ "}}"

blockTag :: MWParser Blocks
blockTag = do
  (tag, _) <- lookAhead $ htmlTag isBlockTag'
  case tag of
      TagOpen "blockquote" _ -> B.blockQuote <$> blocksInTags "blockquote"
      TagOpen "pre" _ -> B.codeBlock . trimCode <$> charsInTags "pre"
      TagOpen "syntaxhighlight" attrs -> syntaxhighlight "syntaxhighlight" attrs
      TagOpen "source" attrs -> syntaxhighlight "source" attrs
      TagOpen "haskell" _ -> B.codeBlockWith ("",["haskell"],[]) . trimCode <$>
                                charsInTags "haskell"
      TagOpen "gallery" _ -> blocksInTags "gallery"
      TagOpen "p" _ -> mempty <$ htmlTag (~== tag)
      TagClose "p"  -> mempty <$ htmlTag (~== tag)
      _ -> B.rawBlock "html" . snd <$> htmlTag (~== tag)

trimCode :: String -> String
trimCode ('\n':xs) = stripTrailingNewlines xs
trimCode xs        = stripTrailingNewlines xs

syntaxhighlight :: String -> [Attribute String] -> MWParser Blocks
syntaxhighlight tag attrs = try $ do
  let mblang = lookup "lang" attrs
  let mbstart = lookup "start" attrs
  let mbline = lookup "line" attrs
  let classes = maybe [] (:[]) mblang ++ maybe [] (const ["numberLines"]) mbline
  let kvs = maybe [] (\x -> [("startFrom",x)]) mbstart
  contents <- charsInTags tag
  return $ B.codeBlockWith ("",classes,kvs) $ trimCode contents

hrule :: MWParser Blocks
hrule = B.horizontalRule <$ try (string "----" *> many (char '-') *> newline)

guardColumnOne :: MWParser ()
guardColumnOne = getPosition >>= \pos -> guard (sourceColumn pos == 1)

preformatted :: MWParser Blocks
preformatted = try $ do
  guardColumnOne
  char ' '
  let endline' = B.linebreak <$ (try $ newline <* char ' ')
  let whitespace' = B.str <$> many1 ('\160' <$ spaceChar)
  let spToNbsp ' ' = '\160'
      spToNbsp x   = x
  let nowiki' = mconcat . intersperse B.linebreak . map B.str .
                lines . fromEntities . map spToNbsp <$> try
                  (htmlTag (~== TagOpen "nowiki" []) *>
                   manyTill anyChar (htmlTag (~== TagClose "nowiki")))
  let inline' = whitespace' <|> endline' <|> nowiki'
                  <|> (try $ notFollowedBy newline *> inline)
  let strToCode (Str s) = Code ("",[],[]) s
      strToCode  x      = x
  contents <- mconcat <$> many1 inline'
  let spacesStr (Str xs) = all isSpace xs
      spacesStr _        = False
  if F.all spacesStr contents
     then return mempty
     else return $ B.para $ walk strToCode contents

header :: MWParser Blocks
header = try $ do
  guardColumnOne
  eqs <- many1 (char '=')
  let lev = length eqs
  guard $ lev <= 6
  contents <- trimInlines . mconcat <$> manyTill inline (count lev $ char '=')
  return $ B.header lev contents

bulletList :: MWParser Blocks
bulletList = B.bulletList <$>
   (   many1 (listItem '*')
   <|> (htmlTag (~== TagOpen "ul" []) *> spaces *> many (listItem '*' <|> li) <*
        optional (htmlTag (~== TagClose "ul"))) )

orderedList :: MWParser Blocks
orderedList =
       (B.orderedList <$> many1 (listItem '#'))
   <|> (B.orderedList <$> (htmlTag (~== TagOpen "ul" []) *> spaces *>
        many (listItem '#' <|> li) <*
        optional (htmlTag (~== TagClose "ul"))))
   <|> do (tag,_) <- htmlTag (~== TagOpen "ol" [])
          spaces
          items <- many (listItem '#' <|> li)
          optional (htmlTag (~== TagClose "ol"))
          let start = maybe 1 id $ safeRead $ fromAttrib "start" tag
          return $ B.orderedListWith (start, DefaultStyle, DefaultDelim) items

definitionList :: MWParser Blocks
definitionList = B.definitionList <$> many1 defListItem

defListItem :: MWParser (Inlines, [Blocks])
defListItem = try $ do
  terms <- mconcat . intersperse B.linebreak <$> many defListTerm
  -- we allow dd with no dt, or dt with no dd
  defs  <- if B.isNull terms
              then notFollowedBy (try $ string ":<math>") *>
                       many1 (listItem ':')
              else many (listItem ':')
  return (terms, defs)

defListTerm  :: MWParser Inlines
defListTerm = char ';' >> skipMany spaceChar >> anyLine >>=
  parseFromString (trimInlines . mconcat <$> many inline)

listStart :: Char -> MWParser ()
listStart c = char c *> notFollowedBy listStartChar

listStartChar :: MWParser Char
listStartChar = oneOf "*#;:"

anyListStart :: MWParser Char
anyListStart =  char '*'
            <|> char '#'
            <|> char ':'
            <|> char ';'

li :: MWParser Blocks
li = lookAhead (htmlTag (~== TagOpen "li" [])) *>
     (firstParaToPlain <$> blocksInTags "li") <* spaces

listItem :: Char -> MWParser Blocks
listItem c = try $ do
  extras <- many (try $ char c <* lookAhead listStartChar)
  if null extras
     then listItem' c
     else do
       skipMany spaceChar
       first <- concat <$> manyTill listChunk newline
       rest <- many
                (try $ string extras *> (concat <$> manyTill listChunk newline))
       contents <- parseFromString (many1 $ listItem' c)
                          (unlines (first : rest))
       case c of
           '*'  -> return $ B.bulletList contents
           '#'  -> return $ B.orderedList contents
           ':'  -> return $ B.definitionList [(mempty, contents)]
           _    -> mzero

-- The point of this is to handle stuff like
-- * {{cite book
-- | blah
-- | blah
-- }}
-- * next list item
-- which seems to be valid mediawiki.
listChunk :: MWParser String
listChunk = template <|> count 1 anyChar

listItem' :: Char -> MWParser Blocks
listItem' c = try $ do
  listStart c
  skipMany spaceChar
  first <- concat <$> manyTill listChunk newline
  rest <- many (try $ char c *> lookAhead listStartChar *>
                   (concat <$> manyTill listChunk newline))
  parseFromString (firstParaToPlain . mconcat <$> many1 block)
      $ unlines $ first : rest

firstParaToPlain :: Blocks -> Blocks
firstParaToPlain contents =
  case viewl (B.unMany contents) of
       (Para xs) :< ys -> B.Many $ (Plain xs) <| ys
       _               -> contents

--
-- inline parsers
--

inline :: MWParser Inlines
inline =  whitespace
      <|> url
      <|> str
      <|> doubleQuotes
      <|> strong
      <|> emph
      <|> image
      <|> internalLink
      <|> externalLink
      <|> math
      <|> inlineTag
      <|> B.singleton <$> charRef
      <|> inlineHtml
      <|> (B.rawInline "mediawiki" <$> variable)
      <|> (B.rawInline "mediawiki" <$> template)
      <|> special

str :: MWParser Inlines
str = B.str <$> many1 (noneOf $ specialChars ++ spaceChars)

math :: MWParser Inlines
math = (B.displayMath . trim <$> try (char ':' >> charsInTags "math"))
   <|> (B.math . trim <$> charsInTags "math")
   <|> (B.displayMath . trim <$> try (dmStart *> manyTill anyChar dmEnd))
   <|> (B.math . trim <$> try (mStart *> manyTill (satisfy (/='\n')) mEnd))
 where dmStart = string "\\["
       dmEnd   = try (string "\\]")
       mStart  = string "\\("
       mEnd    = try (string "\\)")

variable :: MWParser String
variable = try $ do
  string "{{{"
  contents <- manyTill anyChar (try $ string "}}}")
  return $ "{{{" ++ contents ++ "}}}"

inlineTag :: MWParser Inlines
inlineTag = do
  (tag, _) <- lookAhead $ htmlTag isInlineTag'
  case tag of
       TagOpen "ref" _ -> B.note . B.plain <$> inlinesInTags "ref"
       TagOpen "nowiki" _ -> try $ do
          (_,raw) <- htmlTag (~== tag)
          if '/' `elem` raw
             then return mempty
             else B.text . fromEntities <$>
                       manyTill anyChar (htmlTag (~== TagClose "nowiki"))
       TagOpen "br" _ -> B.linebreak <$ (htmlTag (~== TagOpen "br" []) -- will get /> too
                            *> optional blankline)
       TagOpen "strike" _ -> B.strikeout <$> inlinesInTags "strike"
       TagOpen "del" _ -> B.strikeout <$> inlinesInTags "del"
       TagOpen "sub" _ -> B.subscript <$> inlinesInTags "sub"
       TagOpen "sup" _ -> B.superscript <$> inlinesInTags "sup"
       TagOpen "code" _ -> B.code <$> charsInTags "code"
       TagOpen "tt" _ -> B.code <$> charsInTags "tt"
       TagOpen "hask" _ -> B.codeWith ("",["haskell"],[]) <$> charsInTags "hask"
       _ -> B.rawInline "html" . snd <$> htmlTag (~== tag)

special :: MWParser Inlines
special = B.str <$> count 1 (notFollowedBy' (htmlTag isBlockTag') *>
                             oneOf specialChars)

inlineHtml :: MWParser Inlines
inlineHtml = B.rawInline "html" . snd <$> htmlTag isInlineTag'

whitespace :: MWParser Inlines
whitespace = B.space <$ (skipMany1 spaceChar <|> endline <|> htmlComment)

endline :: MWParser ()
endline = () <$ try (newline <*
                     notFollowedBy spaceChar <*
                     notFollowedBy newline <*
                     notFollowedBy' hrule <*
                     notFollowedBy tableStart <*
                     notFollowedBy' header <*
                     notFollowedBy anyListStart)

image :: MWParser Inlines
image = try $ do
  sym "[["
  sym "File:" <|> sym "Image:"
  fname <- many1 (noneOf "|]")
  _ <- many (try $ char '|' *> imageOption)
  caption <-   (B.str fname <$ sym "]]")
           <|> try (char '|' *> (mconcat <$> manyTill inline (sym "]]")))
  return $ B.image fname ("fig:" ++ stringify caption) caption

imageOption :: MWParser String
imageOption =
      try (oneOfStrings [ "border", "thumbnail", "frameless"
                        , "thumb", "upright", "left", "right"
                        , "center", "none", "baseline", "sub"
                        , "super", "top", "text-top", "middle"
                        , "bottom", "text-bottom" ])
  <|> try (string "frame")
  <|> try (many1 (oneOf "x0123456789") <* string "px")
  <|> try (oneOfStrings ["link=","alt=","page=","class="] <* many (noneOf "|]"))

internalLink :: MWParser Inlines
internalLink = try $ do
  sym "[["
  let addUnderscores x = let (pref,suff) = break (=='#') x
                         in  pref ++ intercalate "_" (words suff)
  pagename <- unwords . words <$> many (noneOf "|]")
  label <- option (B.text pagename) $ char '|' *>
             (  (mconcat <$> many1 (notFollowedBy (char ']') *> inline))
             -- the "pipe trick"
             -- [[Help:Contents|] -> "Contents"
             <|> (return $ B.text $ drop 1 $ dropWhile (/=':') pagename) )
  sym "]]"
  linktrail <- B.text <$> many letter
  let link = B.link (addUnderscores pagename) "wikilink" (label <> linktrail)
  if "Category:" `isPrefixOf` pagename
     then do
       updateState $ \st -> st{ mwCategoryLinks = link : mwCategoryLinks st }
       return mempty
     else return link

externalLink :: MWParser Inlines
externalLink = try $ do
  char '['
  (_, src) <- uri
  lab <- try (trimInlines . mconcat <$>
              (skipMany1 spaceChar *> manyTill inline (char ']')))
       <|> do char ']'
              num <- mwNextLinkNumber <$> getState
              updateState $ \st -> st{ mwNextLinkNumber = num + 1 }
              return $ B.str $ show num
  return $ B.link src "" lab

url :: MWParser Inlines
url = do
  (orig, src) <- uri
  return $ B.link src "" (B.str orig)

-- | Parses a list of inlines between start and end delimiters.
inlinesBetween :: (Show b) => MWParser a -> MWParser b -> MWParser Inlines
inlinesBetween start end =
  (trimInlines . mconcat) <$> try (start >> many1Till inner end)
    where inner      = innerSpace <|> (notFollowedBy' (() <$ whitespace) >> inline)
          innerSpace = try $ whitespace >>~ notFollowedBy' end

emph :: MWParser Inlines
emph = B.emph <$> nested (inlinesBetween start end)
    where start = sym "''" >> lookAhead nonspaceChar
          end   = try $ notFollowedBy' (() <$ strong) >> sym "''"

strong :: MWParser Inlines
strong = B.strong <$> nested (inlinesBetween start end)
    where start = sym "'''" >> lookAhead nonspaceChar
          end   = try $ sym "'''"

doubleQuotes :: MWParser Inlines
doubleQuotes = B.doubleQuoted . trimInlines . mconcat <$> try
 ((getState >>= guard . readerSmart . mwOptions) *>
   openDoubleQuote *> manyTill inline closeDoubleQuote )
    where openDoubleQuote = char '"' <* lookAhead alphaNum
          closeDoubleQuote = char '"' <* notFollowedBy alphaNum