aboutsummaryrefslogtreecommitdiff
path: root/src/Text/Pandoc/Readers/TWiki.hs
blob: ebd87359a1dcf6c97ad5d8fd097390214345f689 (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
{-# LANGUAGE FlexibleContexts  #-}
{-# LANGUAGE OverloadedStrings #-}
{- |
   Module      : Text.Pandoc.Readers.TWiki
   Copyright   : Copyright (C) 2014 Alexander Sulfrian
   License     : GNU GPL, version 2 or above

   Maintainer  : Alexander Sulfrian <alexander.sulfrian@fu-berlin.de>
   Stability   : alpha
   Portability : portable

Conversion of twiki text to 'Pandoc' document.
-}
module Text.Pandoc.Readers.TWiki ( readTWiki
                                 ) where

import Control.Monad
import Control.Monad.Except (throwError)
import Data.Char (isAlphaNum)
import qualified Data.Foldable as F
import Data.Maybe (fromMaybe)
import Data.Text (Text)
import qualified Data.Text as T
import Text.HTML.TagSoup
import qualified Text.Pandoc.Builder as B
import Text.Pandoc.Class.PandocMonad (PandocMonad (..))
import Text.Pandoc.Definition
import Text.Pandoc.Options
import Text.Pandoc.Parsing hiding (enclosed, nested)
import Text.Pandoc.Readers.HTML (htmlTag, isCommentTag)
import Text.Pandoc.Shared (crFilter, tshow)
import Text.Pandoc.XML (fromEntities)

-- | Read twiki from an input string and return a Pandoc document.
readTWiki :: PandocMonad m
          => ReaderOptions
          -> Text
          -> m Pandoc
readTWiki opts s = do
  res <- readWithM parseTWiki def{ stateOptions = opts }
             (crFilter s <> "\n\n")
  case res of
       Left e  -> throwError e
       Right d -> return d

type TWParser = ParserT Text ParserState

--
-- utility functions
--

tryMsg :: Text -> TWParser m a -> TWParser m a
tryMsg msg p = try p <?> T.unpack msg

nested :: PandocMonad m => TWParser m a -> TWParser m a
nested p = do
  nestlevel <- stateMaxNestingLevel <$>  getState
  guard $ nestlevel > 0
  updateState $ \st -> st{ stateMaxNestingLevel = stateMaxNestingLevel st - 1 }
  res <- p
  updateState $ \st -> st{ stateMaxNestingLevel = nestlevel }
  return res

htmlElement :: PandocMonad m => Text -> TWParser m (Attr, Text)
htmlElement tag = tryMsg tag $ do
  (TagOpen _ attr, _) <- htmlTag (~== TagOpen tag [])
  content <- T.pack <$> manyTill anyChar (endtag <|> endofinput)
  return (htmlAttrToPandoc attr, trim content)
  where
    endtag     = void $ htmlTag (~== TagClose tag)
    endofinput = lookAhead $ try $ skipMany blankline >> skipSpaces >> eof
    trim       = T.dropAround (=='\n')

htmlAttrToPandoc :: [Attribute Text] -> Attr
htmlAttrToPandoc attrs = (ident, classes, keyvals)
  where
    ident   = fromMaybe "" $ lookup "id" attrs
    classes = maybe [] T.words $ lookup "class" attrs
    keyvals = [(k,v) | (k,v) <- attrs, k /= "id" && k /= "class"]

parseHtmlContentWithAttrs :: PandocMonad m
                          => Text -> TWParser m a -> TWParser m (Attr, [a])
parseHtmlContentWithAttrs tag parser = do
  (attr, content) <- htmlElement tag
  parsedContent <- try $ parseContent content
  return (attr, parsedContent)
  where
    parseContent = parseFromString' $ nested $ manyTill parser endOfContent
    endOfContent = try $ skipMany blankline >> skipSpaces >> eof

parseCharHtmlContentWithAttrs :: PandocMonad m
                          => Text -> TWParser m Char -> TWParser m (Attr, Text)
parseCharHtmlContentWithAttrs tag = fmap go . parseHtmlContentWithAttrs tag
  where
    go (x, y) = (x, T.pack y)

parseHtmlContent :: PandocMonad m => Text -> TWParser m a -> TWParser m [a]
parseHtmlContent tag p = snd <$> parseHtmlContentWithAttrs tag p

--
-- main parser
--

parseTWiki :: PandocMonad m => TWParser m Pandoc
parseTWiki =
  B.doc . mconcat <$> many block <* spaces <* eof


--
-- block parsers
--

block :: PandocMonad m => TWParser m B.Blocks
block = do
  res <- mempty <$ skipMany1 blankline
         <|> blockElements
         <|> para
  skipMany blankline
  trace (T.take 60 $ tshow $ B.toList res)
  return res

blockElements :: PandocMonad m => TWParser m B.Blocks
blockElements = choice [ separator
                       , header
                       , verbatim
                       , literal
                       , list ""
                       , table
                       , blockQuote
                       , noautolink
                       ]

separator :: PandocMonad m => TWParser m B.Blocks
separator = tryMsg "separator" $ string "---" >> newline >> return B.horizontalRule

header :: PandocMonad m => TWParser m B.Blocks
header = tryMsg "header" $ do
  string "---"
  level <- length <$> many1 (char '+')
  guard $ level <= 6
  classes <- option [] $ string "!!" >> return ["unnumbered"]
  skipSpaces
  content <- B.trimInlines . mconcat <$> manyTill inline newline
  attr <- registerHeader ("", classes, []) content
  return $ B.headerWith attr level content

verbatim :: PandocMonad m => TWParser m B.Blocks
verbatim = uncurry B.codeBlockWith <$> (htmlElement "verbatim" <|> htmlElement "pre")

literal :: PandocMonad m => TWParser m B.Blocks
literal = rawBlock <$> htmlElement "literal"
  where
    format (_, _, kvs)        = fromMaybe "html" $ lookup "format" kvs
    rawBlock (attrs, content) = B.rawBlock (format attrs) content

list :: PandocMonad m => Text -> TWParser m B.Blocks
list prefix = choice [ bulletList prefix
                     , orderedList prefix
                     , definitionList prefix]

definitionList :: PandocMonad m => Text -> TWParser m B.Blocks
definitionList prefix = tryMsg "definitionList" $ do
  indent <- lookAhead $ textStr prefix *> many1 (textStr "   ") <* textStr "$ "
  elements <- many $ parseDefinitionListItem (prefix <> T.concat indent)
  return $ B.definitionList elements
  where
    parseDefinitionListItem :: PandocMonad m
                            => Text -> TWParser m (B.Inlines, [B.Blocks])
    parseDefinitionListItem indent = do
      textStr (indent <> "$ ") >> skipSpaces
      term <- many1Till inline $ string ": "
      line <- listItemLine indent $ string "$ "
      return (mconcat term, [line])

bulletList :: PandocMonad m => Text -> TWParser m B.Blocks
bulletList prefix = tryMsg "bulletList" $
                    parseList prefix (char '*') (char ' ')

orderedList :: PandocMonad m => Text -> TWParser m B.Blocks
orderedList prefix = tryMsg "orderedList" $
                     parseList prefix (oneOf "1iIaA") (string ". ")

parseList :: PandocMonad m
          => Text -> TWParser m Char -> TWParser m a -> TWParser m B.Blocks
parseList prefix marker delim = do
  (indent, style) <- lookAhead $ textStr prefix *> listStyle <* delim
  blocks <- many $ parseListItem (prefix <> indent) (char style <* delim)
  return $ case style of
    '1' -> B.orderedListWith (1, DefaultStyle, DefaultDelim) blocks
    'i' -> B.orderedListWith (1, LowerRoman, DefaultDelim) blocks
    'I' -> B.orderedListWith (1, UpperRoman, DefaultDelim) blocks
    'a' -> B.orderedListWith (1, LowerAlpha, DefaultDelim) blocks
    'A' -> B.orderedListWith (1, UpperAlpha, DefaultDelim) blocks
    _   -> B.bulletList blocks
  where
    listStyle = do
      indent <- many1 $ textStr "   "
      style <- marker
      return (T.concat indent, style)

parseListItem :: (PandocMonad m, Show a)
              => Text -> TWParser m a -> TWParser m B.Blocks
parseListItem prefix marker = textStr prefix >> marker >> listItemLine prefix marker

listItemLine :: (PandocMonad m, Show a)
             => Text -> TWParser m a -> TWParser m B.Blocks
listItemLine prefix marker = mconcat <$> (lineContent >>= parseContent)
  where
    lineContent = do
      content <- anyLine
      continuation <- optionMaybe listContinuation
      return $ filterSpaces content <> "\n" <> maybe "" ("   " <>) continuation
    filterSpaces = T.dropWhileEnd (== ' ')
    listContinuation = notFollowedBy (textStr prefix >> marker) >>
                       string "   " >> lineContent
    parseContent = parseFromString' $ many1 $ nestedList <|> parseInline
    parseInline = (B.plain . mconcat) <$> many1Till inline (lastNewline <|> newlineBeforeNestedList)
    nestedList = list prefix
    lastNewline = try $ char '\n' <* eof
    newlineBeforeNestedList = try $ char '\n' <* lookAhead nestedList

table :: PandocMonad m => TWParser m B.Blocks
table = try $ do
  tableHead <- optionMaybe (unzip <$> many1Till tableParseHeader newline)
  rows <- many1 tableParseRow
  return $ buildTable mempty rows $ fromMaybe (align rows, columns rows) tableHead
  where
    buildTable caption rows (aligns, heads)
                    = B.table (B.simpleCaption $ B.plain caption)
                              aligns
                              (TableHead nullAttr $ toHeaderRow heads)
                              [TableBody nullAttr 0 [] $ map toRow rows]
                              (TableFoot nullAttr [])
    align rows      = replicate (columCount rows) (AlignDefault, ColWidthDefault)
    columns rows    = replicate (columCount rows) mempty
    columCount rows = length $ head rows
    toRow           = Row nullAttr . map B.simpleCell
    toHeaderRow l = if null l then [] else [toRow l]

tableParseHeader :: PandocMonad m => TWParser m ((Alignment, ColWidth), B.Blocks)
tableParseHeader = try $ do
  char '|'
  leftSpaces <- length <$> many spaceChar
  char '*'
  content <- tableColumnContent (char '*' >> skipSpaces >> char '|')
  char '*'
  rightSpaces <- length <$> many spaceChar
  optional tableEndOfRow
  return (tableAlign leftSpaces rightSpaces, content)
  where
    tableAlign left right
      | left >= 2 && left == right = (AlignCenter, ColWidthDefault)
      | left > right = (AlignRight, ColWidthDefault)
      | otherwise = (AlignLeft, ColWidthDefault)

tableParseRow :: PandocMonad m => TWParser m [B.Blocks]
tableParseRow = many1Till tableParseColumn newline

tableParseColumn :: PandocMonad m => TWParser m B.Blocks
tableParseColumn = char '|' *> skipSpaces *>
                   tableColumnContent (skipSpaces >> char '|')
                   <* skipSpaces <* optional tableEndOfRow

tableEndOfRow :: PandocMonad m => TWParser m Char
tableEndOfRow = lookAhead (try $ char '|' >> char '\n') >> char '|'

tableColumnContent :: PandocMonad m => TWParser m a -> TWParser m B.Blocks
tableColumnContent end = (B.plain . mconcat) <$> manyTill content (lookAhead $ try end)
  where
    content = continuation <|> inline
    continuation = try $ char '\\' >> newline >> return mempty

blockQuote :: PandocMonad m => TWParser m B.Blocks
blockQuote = (B.blockQuote . mconcat) <$> parseHtmlContent "blockquote" block

noautolink :: PandocMonad m => TWParser m B.Blocks
noautolink = do
  (_, content) <- htmlElement "noautolink"
  st <- getState
  setState $ st{ stateAllowLinks = False }
  blocks <- try $ parseContent content
  setState $ st{ stateAllowLinks = True }
  return $ mconcat blocks
  where
    parseContent = parseFromString' $ many block

para :: PandocMonad m => TWParser m B.Blocks
para = (result . mconcat) <$> many1Till inline endOfParaElement
 where
   endOfParaElement = lookAhead $ endOfInput <|> endOfPara <|> newBlockElement
   endOfInput       = try $ skipMany blankline >> skipSpaces >> eof
   endOfPara        = try $ blankline >> skipMany1 blankline
   newBlockElement  = try $ blankline >> void blockElements
   result content   = if F.all (==Space) content
                      then mempty
                      else B.para $ B.trimInlines content


--
-- inline parsers
--

inline :: PandocMonad m => TWParser m B.Inlines
inline = choice [ whitespace
                , br
                , macro
                , strong
                , strongHtml
                , strongAndEmph
                , emph
                , emphHtml
                , boldCode
                , smart
                , link
                , htmlComment
                , code
                , codeHtml
                , nop
                , autoLink
                , str
                , symbol
                ] <?> "inline"

whitespace :: PandocMonad m => TWParser m B.Inlines
whitespace = lb <|> regsp
  where lb = try $ skipMany spaceChar >> linebreak >> return B.space
        regsp = try $ skipMany1 spaceChar >> return B.space

br :: PandocMonad m => TWParser m B.Inlines
br = try $ string "%BR%" >> return B.linebreak

linebreak :: PandocMonad m => TWParser m B.Inlines
linebreak = newline >> notFollowedBy newline >> (lastNewline <|> innerNewline)
  where lastNewline  = eof >> return mempty
        innerNewline = return B.space

between :: (Monoid c, PandocMonad m, Show b)
        => TWParser m a -> TWParser m b -> (TWParser m b -> TWParser m c)
        -> TWParser m c
between start end p =
  mconcat <$> try (start >> notFollowedBy whitespace >> many1Till (p end) end)

enclosed :: (Monoid b, PandocMonad m, Show a)
         => TWParser m a -> (TWParser m a -> TWParser m b) -> TWParser m b
enclosed sep p = between sep (try $ sep <* endMarker) p
  where
    endMarker   = lookAhead $ void endSpace <|> void (oneOf ".,!?:)|") <|> eof
    endSpace    = (spaceChar <|> newline) >> return B.space

macro :: PandocMonad m => TWParser m B.Inlines
macro = macroWithParameters <|> withoutParameters
  where
    withoutParameters = emptySpan <$> enclosed (char '%') (const macroName)
    emptySpan name = buildSpan name [] mempty

macroWithParameters :: PandocMonad m => TWParser m B.Inlines
macroWithParameters = try $ do
  char '%'
  name <- macroName
  (content, kvs) <- attributes
  char '%'
  return $ buildSpan name kvs $ B.str content

buildSpan :: Text -> [(Text, Text)] -> B.Inlines -> B.Inlines
buildSpan className kvs = B.spanWith attrs
  where
    attrs             = ("", ["twiki-macro", className] ++ additionalClasses, kvsWithoutClasses)
    additionalClasses = maybe [] T.words $ lookup "class" kvs
    kvsWithoutClasses = [(k,v) | (k,v) <- kvs, k /= "class"]

macroName :: PandocMonad m => TWParser m Text
macroName = do
  first <- letter
  rest <- many $ alphaNum <|> char '_'
  return $ T.pack $ first:rest

attributes :: PandocMonad m => TWParser m (Text, [(Text, Text)])
attributes = foldr (either mkContent mkKvs) ("", [])
  <$> (char '{' *> spnl *> many (attribute <* spnl) <* char '}')
  where
    spnl                      = skipMany (spaceChar <|> newline)
    mkContent c  ("", kvs)    = (c, kvs)
    mkContent c  (rest, kvs)  = (c <> " " <> rest, kvs)
    mkKvs     kv (cont, rest) = (cont, kv : rest)

attribute :: PandocMonad m => TWParser m (Either Text (Text, Text))
attribute = withKey <|> withoutKey
  where
    withKey = try $ do
      key <- macroName
      char '='
      curry Right key <$> parseValue False
    withoutKey = try $ Left <$> parseValue True
    parseValue allowSpaces = fromEntities <$> (withQuotes <|> withoutQuotes allowSpaces)
    withQuotes             = between (char '"') (char '"') (\_ -> countChar 1 $ noneOf ['"'])
    withoutQuotes allowSpaces
      | allowSpaces = many1Char $ noneOf "}"
      | otherwise   = many1Char $ noneOf " }"

nestedInlines :: (Show a, PandocMonad m)
              => TWParser m a -> TWParser m B.Inlines
nestedInlines end = innerSpace <|> nestedInline
  where
    innerSpace   = try $ whitespace <* notFollowedBy end
    nestedInline = notFollowedBy whitespace >> nested inline

strong :: PandocMonad m => TWParser m B.Inlines
strong = try $ B.strong <$> enclosed (char '*') nestedInlines

strongHtml :: PandocMonad m => TWParser m B.Inlines
strongHtml = B.strong . mconcat <$> (parseHtmlContent "strong" inline <|> parseHtmlContent "b" inline)

strongAndEmph :: PandocMonad m => TWParser m B.Inlines
strongAndEmph = try $ B.emph . B.strong <$> enclosed (string "__") nestedInlines

emph :: PandocMonad m => TWParser m B.Inlines
emph = try $ B.emph <$> enclosed (char '_')
                        (\p -> notFollowedBy (char '|') >> nestedInlines p)
-- emphasis closers can't cross table cell boundaries, see #3921

emphHtml :: PandocMonad m => TWParser m B.Inlines
emphHtml = B.emph . mconcat <$> (parseHtmlContent "em" inline <|> parseHtmlContent "i" inline)

nestedString :: (Show a, PandocMonad m)
             => TWParser m a -> TWParser m Text
nestedString end = innerSpace <|> countChar 1 nonspaceChar
  where
    innerSpace = try $ many1Char spaceChar <* notFollowedBy end

boldCode :: PandocMonad m => TWParser m B.Inlines
boldCode = try $ (B.strong . B.code . fromEntities) <$> enclosed (string "==") nestedString

htmlComment :: PandocMonad m => TWParser m B.Inlines
htmlComment = htmlTag isCommentTag >> return mempty

code :: PandocMonad m => TWParser m B.Inlines
code = try $ (B.code . fromEntities) <$> enclosed (char '=') nestedString

codeHtml :: PandocMonad m => TWParser m B.Inlines
codeHtml = do
  (attrs, content) <- parseCharHtmlContentWithAttrs "code" anyChar
  return $ B.codeWith attrs $ fromEntities content

autoLink :: PandocMonad m => TWParser m B.Inlines
autoLink = try $ do
  state <- getState
  guard $ stateAllowLinks state
  (text, url) <- parseLink
  guard $ checkLink (T.last url)
  return $ makeLink (text, url)
  where
    parseLink            = notFollowedBy nop >> (uri <|> emailAddress)
    makeLink (text, url) = B.link url "" $ B.str text
    checkLink c
      | c == '/' = True
      | otherwise = isAlphaNum c

str :: PandocMonad m => TWParser m B.Inlines
str = B.str <$> (many1Char alphaNum <|> countChar 1 characterReference)

nop :: PandocMonad m => TWParser m B.Inlines
nop = try $ (void exclamation <|> void nopTag) >> followContent
  where
    exclamation   = char '!'
    nopTag        = stringAnyCase "<nop>"
    followContent = B.str . fromEntities <$> many1Char nonspaceChar

symbol :: PandocMonad m => TWParser m B.Inlines
symbol = B.str <$> countChar 1 nonspaceChar

smart :: PandocMonad m => TWParser m B.Inlines
smart = do
  guardEnabled Ext_smart
  doubleQuoted <|> singleQuoted <|>
    choice [ apostrophe
           , dash
           , ellipses
           ]

singleQuoted :: PandocMonad m => TWParser m B.Inlines
singleQuoted = try $ do
  singleQuoteStart
  withQuoteContext InSingleQuote
    (B.singleQuoted . B.trimInlines . mconcat <$> many1Till inline singleQuoteEnd)

doubleQuoted :: PandocMonad m => TWParser m B.Inlines
doubleQuoted = try $ do
  doubleQuoteStart
  contents <- mconcat <$> many (try $ notFollowedBy doubleQuoteEnd >> inline)
  withQuoteContext InDoubleQuote (doubleQuoteEnd >>
   return (B.doubleQuoted $ B.trimInlines contents))
   <|> return (B.str "\8220" B.<> contents)

link :: PandocMonad m => TWParser m B.Inlines
link = try $ do
  st <- getState
  guard $ stateAllowLinks st
  setState $ st{ stateAllowLinks = False }
  (url, title, content) <- linkText
  setState $ st{ stateAllowLinks = True }
  return $ B.link url title content

linkText :: PandocMonad m => TWParser m (Text, Text, B.Inlines)
linkText = do
  string "[["
  url <- T.pack <$> many1Till anyChar (char ']')
  content <- option (B.str url) (mconcat <$> linkContent)
  char ']'
  return (url, "", content)
  where
    linkContent      = char '[' >> many1Till anyChar (char ']') >>= parseLinkContent . T.pack
    parseLinkContent = parseFromString' $ many1 inline