aboutsummaryrefslogtreecommitdiff
path: root/src/Text/Pandoc/Readers/Muse.hs
blob: e8e309115eade93c1adf09641474806e2732e42c (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
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
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
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
{-# LANGUAGE NoImplicitPrelude #-}
{-# LANGUAGE FlexibleContexts #-}
{-# LANGUAGE TupleSections #-}
{-
  Copyright (C) 2017-2018 Alexander Krotov <ilabdsf@gmail.com>

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.Muse
   Copyright   : Copyright (C) 2017-2018 Alexander Krotov
   License     : GNU GPL, version 2 or above

   Maintainer  : Alexander Krotov <ilabdsf@gmail.com>
   Stability   : alpha
   Portability : portable

Conversion of Muse text to 'Pandoc' document.
-}
{-
TODO:
- Page breaks (five "*")
- Org tables
- table.el tables
- <cite> tag
-}
module Text.Pandoc.Readers.Muse (readMuse) where

import Prelude
import Control.Monad
import Control.Monad.Except (throwError)
import Data.Bifunctor
import Data.Char (isAlphaNum)
import Data.Default
import Data.List (intercalate)
import Data.List.Split (splitOn)
import qualified Data.Map as M
import qualified Data.Set as Set
import Data.Maybe (fromMaybe, isNothing, maybeToList)
import Data.Text (Text, unpack)
import Text.HTML.TagSoup
import Text.Pandoc.Builder (Blocks, Inlines)
import qualified Text.Pandoc.Builder as B
import Text.Pandoc.Class (PandocMonad (..))
import Text.Pandoc.Definition
import Text.Pandoc.Error (PandocError (PandocParsecError))
import Text.Pandoc.Logging
import Text.Pandoc.Options
import Text.Pandoc.Parsing hiding (F, enclosed)
import Text.Pandoc.Shared (crFilter, underlineSpan, mapLeft)

-- | Read Muse from an input string and return a Pandoc document.
readMuse :: PandocMonad m
         => ReaderOptions
         -> Text
         -> m Pandoc
readMuse opts s = do
  let input = crFilter s
  res <- mapLeft (PandocParsecError $ unpack input) `liftM` runParserT parseMuse def{ museOptions = opts } "source" input
  case res of
       Left e  -> throwError e
       Right d -> return d

type F = Future MuseState

data MuseState = MuseState { museMeta :: F Meta -- ^ Document metadata
                           , museOptions :: ReaderOptions
                           , museHeaders :: M.Map Inlines String -- ^ List of headers and ids (used for implicit ref links)
                           , museIdentifierList :: Set.Set String
                           , museLastStrPos :: Maybe SourcePos -- ^ Position after last str parsed
                           , museLogMessages :: [LogMessage]
                           , museNotes :: M.Map String (SourcePos, F Blocks)
                           , museInLink :: Bool -- ^ True when parsing a link description to avoid nested links
                           , museInPara :: Bool -- ^ True when looking for a paragraph terminator
                           }

instance Default MuseState where
  def = MuseState { museMeta = return nullMeta
                  , museOptions = def
                  , museHeaders = M.empty
                  , museIdentifierList = Set.empty
                  , museLastStrPos = Nothing
                  , museLogMessages = []
                  , museNotes = M.empty
                  , museInLink = False
                  , museInPara = False
                  }

type MuseParser = ParserT Text MuseState

instance HasReaderOptions MuseState where
  extractReaderOptions = museOptions

instance HasHeaderMap MuseState where
  extractHeaderMap     = museHeaders
  updateHeaderMap f st = st{ museHeaders = f $ museHeaders st }

instance HasIdentifierList MuseState where
  extractIdentifierList     = museIdentifierList
  updateIdentifierList f st = st{ museIdentifierList = f $ museIdentifierList st }

instance HasLastStrPosition MuseState where
  setLastStrPos pos st = st{ museLastStrPos = Just pos }
  getLastStrPos st     = museLastStrPos st

instance HasLogMessages MuseState where
  addLogMessage m s = s{ museLogMessages = m : museLogMessages s }
  getLogMessages = reverse . museLogMessages

-- | Parse Muse document
parseMuse :: PandocMonad m => MuseParser m Pandoc
parseMuse = do
  many directive
  firstSection <- parseBlocks
  rest <- many parseSection
  let blocks = mconcat (firstSection : rest)
  st <- getState
  let doc = runF (do Pandoc _ bs <- B.doc <$> blocks
                     meta <- museMeta st
                     return $ Pandoc meta bs) st
  reportLogMessages
  return doc

-- * Utility functions

commonPrefix :: String -> String -> String
commonPrefix _ [] = []
commonPrefix [] _ = []
commonPrefix (x:xs) (y:ys)
  | x == y    = x : commonPrefix xs ys
  | otherwise = []

-- | Trim up to one newline from the beginning of the string.
lchop :: String -> String
lchop ('\n':xs) = xs
lchop s = s

-- | Trim up to one newline from the end of the string.
rchop :: String -> String
rchop = reverse . lchop . reverse

dropSpacePrefix :: [String] -> [String]
dropSpacePrefix lns =
  map (drop maxIndent) lns
  where flns = filter (not . all (== ' ')) lns
        maxIndent = if null flns then maximum (map length lns) else length $ takeWhile (== ' ') $ foldl1 commonPrefix flns

atStart :: PandocMonad m => MuseParser m a -> MuseParser m a
atStart p = do
  pos <- getPosition
  st <- getState
  guard $ museLastStrPos st /= Just pos
  p

firstColumn :: PandocMonad m => MuseParser m ()
firstColumn = getPosition >>= \pos -> guard (sourceColumn pos == 1)

-- * Parsers

-- | Parse end-of-line, which can be either a newline or end-of-file.
eol :: Stream s m Char => ParserT s st m ()
eol = void newline <|> eof

someUntil :: (Stream s m t)
          => ParserT s u m a
          -> ParserT s u m b
          -> ParserT s u m ([a], b)
someUntil p end = first <$> ((:) <$> p) <*> manyUntil p end

-- ** HTML parsers

openTag :: PandocMonad m => String -> MuseParser m [(String, String)]
openTag tag = try $
  char '<' *> string tag *> manyTill attr (char '>')
  where
    attr = try $ (,)
      <$  many1 spaceChar
      <*> many1 (noneOf "=\n")
      <*  string "=\""
      <*> manyTill (noneOf "\"") (char '"')

closeTag :: PandocMonad m => String -> MuseParser m ()
closeTag tag = try $ string "</" *> string tag *> void (char '>')

-- | Convert HTML attributes to Pandoc 'Attr'
htmlAttrToPandoc :: [Attribute String] -> Attr
htmlAttrToPandoc attrs = (ident, classes, keyvals)
  where
    ident   = fromMaybe "" $ lookup "id" attrs
    classes = maybe [] words $ lookup "class" attrs
    keyvals = [(k,v) | (k,v) <- attrs, k /= "id" && k /= "class"]

parseHtmlContent :: PandocMonad m
                 => String -- ^ Tag name
                 -> MuseParser m (Attr, F Blocks)
parseHtmlContent tag = try $ do
  many spaceChar
  pos <- getPosition
  attr <- openTag tag
  manyTill spaceChar eol
  content <- parseBlocksTill $ try $ count (sourceColumn pos - 1) spaceChar *> closeTag tag
  manyTill spaceChar eol -- closing tag must be followed by optional whitespace and newline
  return (htmlAttrToPandoc attr, content)

-- ** Directive parsers

-- While not documented, Emacs Muse allows "-" in directive name
parseDirectiveKey :: PandocMonad m => MuseParser m String
parseDirectiveKey = char '#' *> many (letter <|> char '-')

parseEmacsDirective :: PandocMonad m => MuseParser m (String, F Inlines)
parseEmacsDirective = (,)
  <$> parseDirectiveKey
  <*  spaceChar
  <*> (trimInlinesF . mconcat <$> manyTill (choice inlineList) eol)

parseAmuseDirective :: PandocMonad m => MuseParser m (String, F Inlines)
parseAmuseDirective = (,)
  <$> parseDirectiveKey
  <*  many1 spaceChar
  <*> (trimInlinesF . mconcat <$> many1Till inline endOfDirective)
  <*  many blankline
  where
    endOfDirective = lookAhead $ eof <|> try (newline *> (void blankline <|> void parseDirectiveKey))

directive :: PandocMonad m => MuseParser m ()
directive = do
  ext <- getOption readerExtensions
  (key, value) <- if extensionEnabled Ext_amuse ext then parseAmuseDirective else parseEmacsDirective
  updateState $ \st -> st { museMeta = B.setMeta (translateKey key) <$> value <*> museMeta st }
  where translateKey "cover" = "cover-image"
        translateKey x = x

-- ** Block parsers

-- | Parse section contents until EOF or next header
parseBlocks :: PandocMonad m
            => MuseParser m (F Blocks)
parseBlocks =
  try (parseEnd <|>
       nextSection <|>
       blockStart <|>
       listStart <|>
       paraStart)
  where
    nextSection = mempty <$ lookAhead headingStart
    parseEnd = mempty <$ eof
    blockStart = (B.<>) <$> (blockElements <|> emacsNoteBlock)
                        <*> parseBlocks
    listStart = do
      updateState (\st -> st { museInPara = False })
      uncurry (B.<>) <$> (anyListUntil parseBlocks <|> amuseNoteBlockUntil parseBlocks)
    paraStart = do
      indent <- length <$> many spaceChar
      uncurry (B.<>) . first (p indent) <$> paraUntil parseBlocks
      where p indent = if indent >= 2 && indent < 6 then fmap B.blockQuote else id

-- | Parse section that starts with a header
parseSection :: PandocMonad m
             => MuseParser m (F Blocks)
parseSection =
  ((B.<>) <$> emacsHeading <*> parseBlocks) <|>
  (uncurry (B.<>) <$> amuseHeadingUntil parseBlocks)

parseBlocksTill :: PandocMonad m
                => MuseParser m a
                -> MuseParser m (F Blocks)
parseBlocksTill end =
  try (parseEnd <|>
       blockStart <|>
       listStart <|>
       paraStart)
  where
    parseEnd = mempty <$ end
    blockStart = (B.<>) <$> blockElements <*> continuation
    listStart = do
      updateState (\st -> st { museInPara = False })
      uncurry (B.<>) <$> anyListUntil (parseEnd <|> continuation)
    paraStart = uncurry (B.<>) <$> paraUntil (parseEnd <|> continuation)
    continuation = parseBlocksTill end

listItemContentsUntil :: PandocMonad m
                      => Int
                      -> MuseParser m a
                      -> MuseParser m a
                      -> MuseParser m (F Blocks, a)
listItemContentsUntil col pre end =
  try blockStart <|>
  try listStart <|>
  try paraStart
  where
    parsePre = (mempty,) <$> pre
    parseEnd = (mempty,) <$> end
    paraStart = do
      (f, (r, e)) <- paraUntil (parsePre <|> continuation <|> parseEnd)
      return (f B.<> r, e)
    blockStart = first <$> ((B.<>) <$> blockElements)
                       <*> (parsePre <|> continuation <|> parseEnd)
    listStart = do
      updateState (\st -> st { museInPara = False })
      (f, (r, e)) <- anyListUntil (parsePre <|> continuation <|> parseEnd)
      return (f B.<> r, e)
    continuation = try $ do blank <- optionMaybe blankline
                            skipMany blankline
                            indentWith col
                            updateState (\st -> st { museInPara = museInPara st && isNothing blank })
                            listItemContentsUntil col pre end

parseBlock :: PandocMonad m => MuseParser m (F Blocks)
parseBlock = do
  res <- blockElements <|> para
  trace (take 60 $ show $ B.toList $ runF res def)
  return res
  where para = fst <$> paraUntil (try (eof <|> void (lookAhead blockElements)))

blockElements :: PandocMonad m => MuseParser m (F Blocks)
blockElements = do
  updateState (\st -> st { museInPara = False })
  choice [ mempty <$ blankline
         , comment
         , separator
         , example
         , exampleTag
         , literalTag
         , centerTag
         , rightTag
         , quoteTag
         , divTag
         , biblioTag
         , playTag
         , verseTag
         , lineBlock
         , table
         , commentTag
         ]

-- | Parse a line comment, starting with @;@ in the first column.
comment :: PandocMonad m => MuseParser m (F Blocks)
comment = try $ mempty
  <$ firstColumn
  <* char ';'
  <* optional (spaceChar *> many (noneOf "\n"))
  <* eol

-- | Parse a horizontal rule, consisting of 4 or more @\'-\'@ characters.
separator :: PandocMonad m => MuseParser m (F Blocks)
separator = try $ pure B.horizontalRule
  <$ string "----"
  <* many (char '-')
  <* many spaceChar
  <* eol

headingStart :: PandocMonad m => MuseParser m (String, Int)
headingStart = try $ (,)
  <$> option "" (try (parseAnchor <* manyTill spaceChar eol))
  <*  firstColumn
  <*> fmap length (many1 $ char '*')
  <*  spaceChar

-- | Parse a single-line heading.
emacsHeading :: PandocMonad m => MuseParser m (F Blocks)
emacsHeading = try $ do
  guardDisabled Ext_amuse
  (anchorId, level) <- headingStart
  content <- trimInlinesF . mconcat <$> manyTill inline eol
  attr <- registerHeader (anchorId, [], []) (runF content def)
  return $ B.headerWith attr level <$> content

-- | Parse a multi-line heading.
-- It is a Text::Amuse extension, Emacs Muse does not allow heading to span multiple lines.
amuseHeadingUntil :: PandocMonad m
                  => MuseParser m a -- ^ Terminator parser
                  -> MuseParser m (F Blocks, a)
amuseHeadingUntil end = try $ do
  guardEnabled Ext_amuse
  (anchorId, level) <- headingStart
  (content, e) <- paraContentsUntil end
  attr <- registerHeader (anchorId, [], []) (runF content def)
  return (B.headerWith attr level <$> content, e)

-- | Parse an example between @{{{@ and @}}}@.
-- It is an Amusewiki extension influenced by Creole wiki, as described in @Text::Amuse@ documentation.
example :: PandocMonad m => MuseParser m (F Blocks)
example = try $ pure . B.codeBlock
  <$  string "{{{"
  <*  optional blankline
  <*> manyTill anyChar (try (optional blankline *> string "}}}"))

-- | Parse an @\<example>@ tag.
exampleTag :: PandocMonad m => MuseParser m (F Blocks)
exampleTag = try $ fmap pure $ B.codeBlockWith
  <$  many spaceChar
  <*> (htmlAttrToPandoc <$> openTag "example")
  <*> (rchop . intercalate "\n" . dropSpacePrefix . splitOn "\n" . lchop <$> manyTill anyChar (closeTag "example"))
  <*  manyTill spaceChar eol

-- | Parse a @\<literal>@ tag as a raw block.
-- For 'RawInline' @\<literal>@ parser, see 'inlineLiteralTag'.
literalTag :: PandocMonad m => MuseParser m (F Blocks)
literalTag = try $ fmap pure $ B.rawBlock
  <$  many spaceChar
  <*> (fromMaybe "html" . lookup "style" <$> openTag "literal") -- FIXME: Emacs Muse inserts <literal> without style into all output formats, but we assume HTML
  <*  manyTill spaceChar eol
  <*> (rchop . intercalate "\n" . dropSpacePrefix . splitOn "\n" . lchop <$> manyTill anyChar (closeTag "literal"))
  <*  manyTill spaceChar eol

-- | Parse @\<center>@ tag.
-- Currently it is ignored as Pandoc cannot represent centered blocks.
centerTag :: PandocMonad m => MuseParser m (F Blocks)
centerTag = snd <$> parseHtmlContent "center"

-- | Parse @\<right>@ tag.
-- Currently it is ignored as Pandoc cannot represent centered blocks.
rightTag :: PandocMonad m => MuseParser m (F Blocks)
rightTag = snd <$> parseHtmlContent "right"

-- | Parse @\<quote>@ tag.
quoteTag :: PandocMonad m => MuseParser m (F Blocks)
quoteTag = fmap B.blockQuote . snd <$> parseHtmlContent "quote"

-- | Parse @\<div>@ tag.
-- @\<div>@ tag is supported by Emacs Muse, but not Amusewiki 2.025.
divTag :: PandocMonad m => MuseParser m (F Blocks)
divTag = do
  (attrs, content) <- parseHtmlContent "div"
  return $ B.divWith attrs <$> content

-- | Parse @\<biblio>@ tag, the result is the same as @\<div class="biblio">@.
-- @\<biblio>@ tag is supported only in Text::Amuse mode.
biblioTag :: PandocMonad m => MuseParser m (F Blocks)
biblioTag = do
  guardEnabled Ext_amuse
  fmap (B.divWith ("", ["biblio"], [])) . snd <$> parseHtmlContent "biblio"

-- | Parse @\<play>@ tag, the result is the same as @\<div class="play">@.
-- @\<play>@ tag is supported only in Text::Amuse mode.
playTag :: PandocMonad m => MuseParser m (F Blocks)
playTag = do
  guardEnabled Ext_amuse
  fmap (B.divWith ("", ["play"], [])) . snd <$> parseHtmlContent "play"

verseLine :: PandocMonad m => MuseParser m (F Inlines)
verseLine = do
  indent <- (B.str <$> many1 ('\160' <$ char ' ')) <|> pure mempty
  rest <- manyTill (choice inlineList) newline
  return $ trimInlinesF $ mconcat (pure indent : rest)

-- | Parse @\<verse>@ tag.
verseTag :: PandocMonad m => MuseParser m (F Blocks)
verseTag = try $ do
  many spaceChar
  pos <- getPosition
  openTag "verse"
  manyTill spaceChar eol
  let indent = count (sourceColumn pos - 1) spaceChar
  content <- sequence <$> manyTill (indent *> verseLine) (try $ indent *> closeTag "verse")
  manyTill spaceChar eol
  return $ B.lineBlock <$> content

-- | Parse @\<comment>@ tag.
commentTag :: PandocMonad m => MuseParser m (F Blocks)
commentTag = try $ mempty
  <$ many spaceChar
  <* openTag "comment"
  <* manyTill anyChar (closeTag "comment")
  <* manyTill spaceChar eol

-- | Parse paragraph contents.
paraContentsUntil :: PandocMonad m
                  => MuseParser m a -- ^ Terminator parser
                  -> MuseParser m (F Inlines, a)
paraContentsUntil end = do
  updateState (\st -> st { museInPara = True })
  (l, e) <- someUntil inline $ try (manyTill spaceChar eol *> end)
  updateState (\st -> st { museInPara = False })
  return (trimInlinesF $ mconcat l, e)

-- | Parse a paragraph.
paraUntil :: PandocMonad m
          => MuseParser m a -- ^ Terminator parser
          -> MuseParser m (F Blocks, a)
paraUntil end = do
  state <- getState
  guard $ not $ museInPara state
  first (fmap B.para) <$> paraContentsUntil end

noteMarker :: PandocMonad m => MuseParser m String
noteMarker = try $ (:)
  <$  char '['
  <*> oneOf "123456789"
  <*> manyTill digit (char ']')

-- Amusewiki version of note
-- Parsing is similar to list item, except that note marker is used instead of list marker
amuseNoteBlockUntil :: PandocMonad m
                    => MuseParser m a
                    -> MuseParser m (F Blocks, a)
amuseNoteBlockUntil end = try $ do
  guardEnabled Ext_amuse
  ref <- noteMarker <* spaceChar
  pos <- getPosition
  updateState (\st -> st { museInPara = False })
  (content, e) <- listItemContentsUntil (sourceColumn pos - 1) (fail "x") end
  oldnotes <- museNotes <$> getState
  when (M.member ref oldnotes)
    (logMessage $ DuplicateNoteReference ref pos)
  updateState $ \s -> s{ museNotes = M.insert ref (pos, content) oldnotes }
  return (mempty, e)

-- Emacs version of note
-- Notes are allowed only at the end of text, no indentation is required.
emacsNoteBlock :: PandocMonad m => MuseParser m (F Blocks)
emacsNoteBlock = try $ do
  guardDisabled Ext_amuse
  pos <- getPosition
  ref <- noteMarker <* skipSpaces
  content <- mconcat <$> blocksTillNote
  oldnotes <- museNotes <$> getState
  when (M.member ref oldnotes)
    (logMessage $ DuplicateNoteReference ref pos)
  updateState $ \s -> s{ museNotes = M.insert ref (pos, content) oldnotes }
  return mempty
  where
    blocksTillNote =
      many1Till parseBlock (eof <|> () <$ lookAhead noteMarker)

--
-- Verse markup
--

lineVerseLine :: PandocMonad m => MuseParser m (F Inlines)
lineVerseLine = try $ do
  string "> "
  indent <- many ('\160' <$ char ' ')
  let indentEl = if null indent then mempty else B.str indent
  rest <- manyTill (choice inlineList) eol
  return $ trimInlinesF $ mconcat (pure indentEl : rest)

blanklineVerseLine :: PandocMonad m => MuseParser m (F Inlines)
blanklineVerseLine = try $ mempty
  <$ char '>'
  <* blankline

-- | Parse a line block indicated by @\'>\'@ characters.
lineBlock :: PandocMonad m => MuseParser m (F Blocks)
lineBlock = try $ do
  many spaceChar
  col <- sourceColumn <$> getPosition
  lns <- (blanklineVerseLine <|> lineVerseLine) `sepBy1'` try (indentWith (col - 1))
  return $ B.lineBlock <$> sequence lns

-- *** List parsers

bulletListItemsUntil :: PandocMonad m
                     => Int -- ^ Indentation
                     -> MuseParser m a -- ^ Terminator parser
                     -> MuseParser m ([F Blocks], a)
bulletListItemsUntil indent end = try $ do
  char '-'
  void spaceChar <|> lookAhead eol
  updateState (\st -> st { museInPara = False })
  (x, (xs, e)) <- listItemContentsUntil (indent + 2) (try (optional blankline *> indentWith indent *> bulletListItemsUntil indent end)) (([],) <$> end)
  return (x:xs, e)

-- | Parse a bullet list.
bulletListUntil :: PandocMonad m
                => MuseParser m a
                -> MuseParser m (F Blocks, a)
bulletListUntil end = try $ do
  many spaceChar
  pos <- getPosition
  let indent = sourceColumn pos - 1
  guard $ indent /= 0
  (items, e) <- bulletListItemsUntil indent end
  return (B.bulletList <$> sequence items, e)

museOrderedListMarker :: PandocMonad m
                      => ListNumberStyle
                      -> MuseParser m Int
museOrderedListMarker style =
  snd <$> p <* char '.'
  where p = case style of
              Decimal    -> decimal
              UpperRoman -> upperRoman
              LowerRoman -> lowerRoman
              UpperAlpha -> upperAlpha
              LowerAlpha -> lowerAlpha
              _          -> fail "Unhandled case"

orderedListItemsUntil :: PandocMonad m
                      => Int
                      -> ListNumberStyle
                      -> MuseParser m a
                      -> MuseParser m ([F Blocks], a)
orderedListItemsUntil indent style end =
  continuation
  where
    continuation = try $ do
      pos <- getPosition
      void spaceChar <|> lookAhead eol
      updateState (\st -> st { museInPara = False })
      (x, (xs, e)) <- listItemContentsUntil (sourceColumn pos) (try (optional blankline *> indentWith indent *> museOrderedListMarker style *> continuation)) (([],) <$> end)
      return (x:xs, e)

-- | Parse an ordered list.
orderedListUntil :: PandocMonad m
                 => MuseParser m a
                 -> MuseParser m (F Blocks, a)
orderedListUntil end = try $ do
  many spaceChar
  pos <- getPosition
  let indent = sourceColumn pos - 1
  guard $ indent /= 0
  (style, start) <- decimal <|> lowerRoman <|> upperRoman <|> lowerAlpha <|> upperAlpha
  char '.'
  first (fmap (B.orderedListWith (start, style, Period)) . sequence)
    <$> orderedListItemsUntil indent style end

descriptionsUntil :: PandocMonad m
                  => Int
                  -> MuseParser m a
                  -> MuseParser m ([F Blocks], a)
descriptionsUntil indent end = do
  void spaceChar <|> lookAhead eol
  updateState (\st -> st { museInPara = False })
  (x, (xs, e)) <- listItemContentsUntil indent (try (optional blankline *> indentWith indent *> manyTill spaceChar (string "::") *> descriptionsUntil indent end)) (([],) <$> end)
  return (x:xs, e)

definitionListItemsUntil :: PandocMonad m
                         => Int
                         -> MuseParser m a
                         -> MuseParser m ([F (Inlines, [Blocks])], a)
definitionListItemsUntil indent end =
  continuation
  where
    continuation = try $ do
      pos <- getPosition
      term <- trimInlinesF . mconcat <$> manyTill (choice inlineList) (try $ string "::")
      (x, (xs, e)) <- descriptionsUntil (sourceColumn pos) (try (optional blankline *> indentWith indent *> continuation) <|> (([],) <$> end))
      let xx = (,) <$> term <*> sequence x
      return (xx:xs, e)

-- | Parse a definition list.
definitionListUntil :: PandocMonad m
                    => MuseParser m a -- ^ Terminator parser
                    -> MuseParser m (F Blocks, a)
definitionListUntil end = try $ do
  many spaceChar
  pos <- getPosition
  let indent = sourceColumn pos - 1
  guardDisabled Ext_amuse <|> guard (indent /= 0) -- Initial space is required by Amusewiki, but not Emacs Muse
  first (fmap B.definitionList . sequence) <$> definitionListItemsUntil indent end

anyListUntil :: PandocMonad m
             => MuseParser m a -- ^ Terminator parser
             -> MuseParser m (F Blocks, a)
anyListUntil end =
  bulletListUntil end <|> orderedListUntil end <|> definitionListUntil end

-- *** Table parsers

-- | Internal Muse table representation.
data MuseTable = MuseTable
  { museTableCaption :: Inlines
  , museTableHeaders :: [[Blocks]]
  , museTableRows    :: [[Blocks]]
  , museTableFooters :: [[Blocks]]
  }

data MuseTableElement = MuseHeaderRow [Blocks]
                      | MuseBodyRow [Blocks]
                      | MuseFooterRow [Blocks]
                      | MuseCaption Inlines

museToPandocTable :: MuseTable -> Blocks
museToPandocTable (MuseTable caption headers body footers) =
  B.table caption attrs headRow rows
  where ncol = maximum (0 : map length (headers ++ body ++ footers))
        attrs = replicate ncol (AlignDefault, 0.0)
        headRow = if null headers then [] else head headers
        rows = (if null headers then [] else tail headers) ++ body ++ footers

museAppendElement :: MuseTableElement
                  -> MuseTable
                  -> MuseTable
museAppendElement element tbl =
  case element of
    MuseHeaderRow row -> tbl{ museTableHeaders = row : museTableHeaders tbl }
    MuseBodyRow row -> tbl{ museTableRows = row : museTableRows tbl }
    MuseFooterRow row -> tbl{ museTableFooters = row : museTableFooters tbl }
    MuseCaption inlines -> tbl{ museTableCaption = inlines }

tableCell :: PandocMonad m => MuseParser m (F Blocks)
tableCell = try $ fmap B.plain . trimInlinesF . mconcat <$> manyTill inline (lookAhead cellEnd)
  where cellEnd = try $ void (many1 spaceChar *> char '|') <|> eol

tableElements :: PandocMonad m => MuseParser m (F [MuseTableElement])
tableElements = sequence <$> (tableParseElement `sepEndBy1` eol)

elementsToTable :: [MuseTableElement] -> MuseTable
elementsToTable = foldr museAppendElement emptyTable
  where emptyTable = MuseTable mempty mempty mempty mempty

-- | Parse a table.
table :: PandocMonad m => MuseParser m (F Blocks)
table = try $ fmap (museToPandocTable . elementsToTable) <$> tableElements

tableParseElement :: PandocMonad m => MuseParser m (F MuseTableElement)
tableParseElement = tableParseHeader
                <|> tableParseBody
                <|> tableParseFooter
                <|> tableParseCaption

tableParseRow :: PandocMonad m
              => Int -- ^ Number of separator characters
              -> MuseParser m (F [Blocks])
tableParseRow n = try $
  sequence <$> (tableCell `sepBy2` fieldSep)
    where p `sepBy2` sep = (:) <$> p <*> many1 (sep *> p)
          fieldSep = many1 spaceChar *> count n (char '|') *> (void (many1 spaceChar) <|> void (lookAhead newline))

-- | Parse a table header row.
tableParseHeader :: PandocMonad m => MuseParser m (F MuseTableElement)
tableParseHeader = fmap MuseHeaderRow <$> tableParseRow 2

-- | Parse a table body row.
tableParseBody :: PandocMonad m => MuseParser m (F MuseTableElement)
tableParseBody = fmap MuseBodyRow <$> tableParseRow 1

-- | Parse a table footer row.
tableParseFooter :: PandocMonad m => MuseParser m (F MuseTableElement)
tableParseFooter = fmap MuseFooterRow <$> tableParseRow 3

-- | Parse table caption.
tableParseCaption :: PandocMonad m => MuseParser m (F MuseTableElement)
tableParseCaption = try $ fmap MuseCaption . trimInlinesF . mconcat
  <$  many spaceChar
  <*  string "|+"
  <*> many1Till inline (string "+|")

-- ** Inline parsers

inlineList :: PandocMonad m => [MuseParser m (F Inlines)]
inlineList = [ whitespace
             , br
             , anchor
             , footnote
             , strong
             , strongTag
             , emph
             , emphTag
             , underlined
             , superscriptTag
             , subscriptTag
             , strikeoutTag
             , verbatimTag
             , classTag
             , nbsp
             , linkOrImage
             , code
             , codeTag
             , mathTag
             , inlineLiteralTag
             , str
             , symbol
             ]

inline :: PandocMonad m => MuseParser m (F Inlines)
inline = endline <|> choice inlineList <?> "inline"

-- | Parse a soft break.
endline :: PandocMonad m => MuseParser m (F Inlines)
endline = try $ pure B.softbreak <$ newline <* notFollowedBy blankline

parseAnchor :: PandocMonad m => MuseParser m String
parseAnchor = try $ (:)
  <$  firstColumn
  <*  char '#'
  <*> letter
  <*> many (letter <|> digit <|> char '-')

anchor :: PandocMonad m => MuseParser m (F Inlines)
anchor = try $ do
  anchorId <- parseAnchor
  skipMany spaceChar <|> void newline
  return $ return $ B.spanWith (anchorId, [], []) mempty

-- | Parse a footnote reference.
footnote :: PandocMonad m => MuseParser m (F Inlines)
footnote = try $ do
  inLink <- museInLink <$> getState
  guard $ not inLink
  ref <- noteMarker
  return $ do
    notes <- asksF museNotes
    case M.lookup ref notes of
      Nothing -> return $ B.str $ "[" ++ ref ++ "]"
      Just (_pos, contents) -> do
        st <- askF
        let contents' = runF contents st { museNotes = M.delete ref (museNotes st) }
        return $ B.note contents'

whitespace :: PandocMonad m => MuseParser m (F Inlines)
whitespace = try $ pure B.space <$ skipMany1 spaceChar

-- | Parse @\<br>@ tag.
br :: PandocMonad m => MuseParser m (F Inlines)
br = try $ pure B.linebreak <$ string "<br>"

emphasisBetween :: (PandocMonad m, Show a) => MuseParser m a -> MuseParser m (F Inlines)
emphasisBetween c = try $ enclosedInlines c c

-- | Parses material enclosed between start and end parsers.
enclosed :: (Show end, Stream s  m Char) => ParserT s st m t   -- ^ start parser
         -> ParserT s st m end  -- ^ end parser
         -> ParserT s st m a    -- ^ content parser (to be used repeatedly)
         -> ParserT s st m [a]
enclosed start end parser = try $
  start *> notFollowedBy spaceChar *> many1Till parser end

enclosedInlines :: (PandocMonad m, Show a, Show b)
                => MuseParser m a
                -> MuseParser m b
                -> MuseParser m (F Inlines)
enclosedInlines start end = try $ trimInlinesF . mconcat
  <$> enclosed (atStart start) end inline
  <*  notFollowedBy (satisfy isAlphaNum)

-- | Parse an inline tag, such as @\<em>@ and @\<strong>@.
inlineTag :: PandocMonad m
          => String -- ^ Tag name
          -> MuseParser m (F Inlines)
inlineTag tag = try $ mconcat
  <$  openTag tag
  <*> manyTill inline (closeTag tag)

-- | Parse strong inline markup, indicated by @**@.
strong :: PandocMonad m => MuseParser m (F Inlines)
strong = fmap B.strong <$> emphasisBetween (string "**")

-- | Parse emphasis inline markup, indicated by @*@.
emph :: PandocMonad m => MuseParser m (F Inlines)
emph = fmap B.emph <$> emphasisBetween (char '*')

-- | Parse underline inline markup, indicated by @_@.
-- Supported only in Emacs Muse mode, not Text::Amuse.
underlined :: PandocMonad m => MuseParser m (F Inlines)
underlined = fmap underlineSpan
  <$  guardDisabled Ext_amuse -- Supported only by Emacs Muse
  <*> emphasisBetween (char '_')

-- | Parse @\<strong>@ tag.
strongTag :: PandocMonad m => MuseParser m (F Inlines)
strongTag = fmap B.strong <$> inlineTag "strong"

-- | Parse @\<em>@ tag.
emphTag :: PandocMonad m => MuseParser m (F Inlines)
emphTag = fmap B.emph <$> inlineTag "em"

-- | Parse @\<sup>@ tag.
superscriptTag :: PandocMonad m => MuseParser m (F Inlines)
superscriptTag = fmap B.superscript <$> inlineTag "sup"

-- | Parse @\<sub>@ tag.
subscriptTag :: PandocMonad m => MuseParser m (F Inlines)
subscriptTag = fmap B.subscript <$> inlineTag "sub"

-- | Parse @\<del>@ tag.
strikeoutTag :: PandocMonad m => MuseParser m (F Inlines)
strikeoutTag = fmap B.strikeout <$> inlineTag "del"

-- | Parse @\<verbatim>@ tag.
verbatimTag :: PandocMonad m => MuseParser m (F Inlines)
verbatimTag = return . B.text
  <$  openTag "verbatim"
  <*> manyTill anyChar (closeTag "verbatim")

-- | Parse @\<class>@ tag.
classTag :: PandocMonad m => MuseParser m (F Inlines)
classTag = do
  classes <- maybe [] words . lookup "name" <$> openTag "class"
  res <- manyTill inline $ closeTag "class"
  return $ B.spanWith ("", classes, []) <$> mconcat res

-- | Parse "~~" as nonbreaking space.
nbsp :: PandocMonad m => MuseParser m (F Inlines)
nbsp = try $ pure (B.str "\160") <$ string "~~"

-- | Parse code markup, indicated by @\'=\'@ characters.
code :: PandocMonad m => MuseParser m (F Inlines)
code = try $ do
  atStart $ char '='
  contents <- many1Till (noneOf "\n\r" <|> (newline <* notFollowedBy newline)) $ char '='
  guard $ not $ null contents
  guard $ head contents `notElem` " \t\n"
  guard $ last contents `notElem` " \t\n"
  notFollowedBy $ satisfy isAlphaNum
  return $ return $ B.code contents

-- | Parse @\<code>@ tag.
codeTag :: PandocMonad m => MuseParser m (F Inlines)
codeTag = fmap pure $ B.codeWith
  <$> (htmlAttrToPandoc <$> openTag "code")
  <*> manyTill anyChar (closeTag "code")

-- | Parse @\<math>@ tag.
-- @\<math>@ tag is an Emacs Muse extension enabled by @(require 'muse-latex2png)@
mathTag :: PandocMonad m => MuseParser m (F Inlines)
mathTag = return . B.math
  <$  openTag "math"
  <*> manyTill anyChar (closeTag "math")

-- | Parse inline @\<literal>@ tag as a raw inline.
inlineLiteralTag :: PandocMonad m => MuseParser m (F Inlines)
inlineLiteralTag = try $ fmap pure $ B.rawInline
  <$> (fromMaybe "html" . lookup "style" <$> openTag "literal") -- FIXME: Emacs Muse inserts <literal> without style into all output formats, but we assume HTML
  <*> manyTill anyChar (closeTag "literal")

str :: PandocMonad m => MuseParser m (F Inlines)
str = return . B.str <$> many1 alphaNum <* updateLastStrPos

symbol :: PandocMonad m => MuseParser m (F Inlines)
symbol = return . B.str <$> count 1 nonspaceChar

-- | Parse a link or image.
linkOrImage :: PandocMonad m => MuseParser m (F Inlines)
linkOrImage = try $ do
  st <- getState
  guard $ not $ museInLink st
  setState $ st{ museInLink = True }
  res <- explicitLink <|> image <|> link
  updateState (\state -> state { museInLink = False })
  return res

linkContent :: PandocMonad m => MuseParser m (F Inlines)
linkContent = trimInlinesF . mconcat
  <$  char '['
  <*> manyTill inline (char ']')

-- | Parse a link starting with @URL:@
explicitLink :: PandocMonad m => MuseParser m (F Inlines)
explicitLink = try $ do
  string "[[URL:"
  url <- manyTill anyChar $ char ']'
  content <- option (pure $ B.str url) linkContent
  char ']'
  return $ B.link url "" <$> content

image :: PandocMonad m => MuseParser m (F Inlines)
image = try $ do
  string "[["
  (url, (ext, width, align)) <- manyUntil (noneOf "]") (imageExtensionAndOptions <* char ']')
  content <- optionMaybe linkContent
  char ']'
  let widthAttr = case align of
                    Just 'f' -> [("width", fromMaybe "100" width ++ "%"), ("height", "75%")]
                    _ -> maybeToList (("width",) . (++ "%") <$> width)
  let alignClass = case align of
                     Just 'r' -> ["align-right"]
                     Just 'l' -> ["align-left"]
                     Just 'f' -> []
                     _        -> []
  return $ B.imageWith ("", alignClass, widthAttr) (url ++ ext) mempty <$> fromMaybe (return mempty) content
  where -- Taken from muse-image-regexp defined in Emacs Muse file lisp/muse-regexps.el
        imageExtensions = [".eps", ".gif", ".jpg", ".jpeg", ".pbm", ".png", ".tiff", ".xbm", ".xpm"]
        imageExtension = choice (try . string <$> imageExtensions)
        imageExtensionAndOptions = do
          ext <- imageExtension
          (width, align) <- option (Nothing, Nothing) imageAttrs
          return (ext, width, align)
        imageAttrs = (,)
          <$  many1 spaceChar
          <*> optionMaybe (many1 digit)
          <*  many spaceChar
          <*> optionMaybe (oneOf "rlf")

link :: PandocMonad m => MuseParser m (F Inlines)
link = try $ do
  string "[["
  url <- manyTill anyChar $ char ']'
  content <- optionMaybe linkContent
  char ']'
  return $ B.link url "" <$> fromMaybe (return $ B.str url) content