aboutsummaryrefslogtreecommitdiff
path: root/src/Text/Pandoc/Readers/Org/Blocks.hs
blob: 5e15c2c36465605f014a9ce0e3fb59cf9e31daf8 (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
{-# LANGUAGE FlexibleContexts  #-}
{-# LANGUAGE LambdaCase        #-}
{-# LANGUAGE OverloadedStrings #-}
{-# LANGUAGE RecordWildCards   #-}
{- |
   Module      : Text.Pandoc.Readers.Org.Blocks
   Copyright   : Copyright (C) 2014-2021 Albert Krewinkel
   License     : GNU GPL, version 2 or above

   Maintainer  : Albert Krewinkel <tarleb+pandoc@moltkeplatz.de>

Parsers for Org-mode block elements.
-}
module Text.Pandoc.Readers.Org.Blocks
  ( blockList
  , meta
  ) where

import Text.Pandoc.Readers.Org.BlockStarts
import Text.Pandoc.Readers.Org.DocumentTree (documentTree,
                                             unprunedHeadlineToBlocks)
import Text.Pandoc.Readers.Org.Inlines
import Text.Pandoc.Readers.Org.Meta (metaExport, metaKey, metaLine)
import Text.Pandoc.Readers.Org.ParserState
import Text.Pandoc.Readers.Org.Parsing
import Text.Pandoc.Readers.Org.Shared (cleanLinkText, isImageFilename,
                                       originalLang, translateLang, exportsCode)

import Text.Pandoc.Builder (Blocks, Inlines)
import Text.Pandoc.Class.PandocMonad (PandocMonad)
import Text.Pandoc.Definition
import Text.Pandoc.Options
import Text.Pandoc.Shared (compactify, compactifyDL, safeRead)

import Control.Monad (foldM, guard, mplus, mzero, void)
import Data.Char (isSpace)
import Data.Default (Default)
import Data.Functor (($>))
import Data.List (foldl', intersperse)
import Data.Maybe (fromMaybe, isJust, isNothing)
import Data.Text (Text)
import Data.List.NonEmpty (nonEmpty)

import qualified Data.Text as T
import qualified Text.Pandoc.Builder as B
import qualified Text.Pandoc.Walk as Walk
import Text.Pandoc.Sources (ToSources(..))

--
-- parsing blocks
--

-- | Get a list of blocks.
blockList :: PandocMonad m => OrgParser m [Block]
blockList = do
  fHeadlineTree  <- documentTree blocks inline
  st             <- getState
  let headlineTree = runF fHeadlineTree st
  unprunedHeadlineToBlocks headlineTree st

-- | Get the meta information saved in the state.
meta :: Monad m => OrgParser m Meta
meta = do
  meta' <- metaExport
  runF meta' <$> getState

blocks :: PandocMonad m => OrgParser m (F Blocks)
blocks = mconcat <$> manyTill block (void (lookAhead headerStart) <|> eof)

block :: PandocMonad m => OrgParser m (F Blocks)
block = choice [ mempty <$ blanklines
               , table
               , orgBlock
               , figure
               , example
               , genericDrawer
               , include
               , specialLine
               , horizontalRule
               , list
               , latexFragment
               , noteBlock
               , rawOrgLine
               , paraOrPlain
               ] <?> "block"


-- | Parse a horizontal rule into a block element
horizontalRule :: Monad m => OrgParser m (F Blocks)
horizontalRule = return B.horizontalRule <$ try hline


--
-- Block Attributes
--

-- | Attributes that may be added to figures (like a name or caption).
data BlockAttributes = BlockAttributes
  { blockAttrName      :: Maybe Text
  , blockAttrLabel     :: Maybe Text
  , blockAttrCaption   :: Maybe (F Inlines)
  , blockAttrKeyValues :: [(Text, Text)]
  }

-- | Convert BlockAttributes into pandoc Attr
attrFromBlockAttributes :: BlockAttributes -> Attr
attrFromBlockAttributes BlockAttributes{..} =
  let
    ident   = fromMaybe mempty $ lookup "id" blockAttrKeyValues
    classes = maybe [] T.words $ lookup "class" blockAttrKeyValues
    kv      = filter ((`notElem` ["id", "class"]) . fst) blockAttrKeyValues
  in (ident, classes, kv)

stringyMetaAttribute :: Monad m => OrgParser m (Text, Text)
stringyMetaAttribute = try $ do
  metaLineStart
  attrName <- T.toLower <$> many1TillChar nonspaceChar (char ':')
  skipSpaces
  attrValue <- anyLine <|> ("" <$ newline)
  return (attrName, attrValue)

-- | Parse a set of block attributes. Block attributes are given through
-- lines like @#+caption: block caption@ or @#+attr_html: :width 20@.
-- Parsing will fail if any line contains an attribute different from
-- those attributes known to work on blocks.
blockAttributes :: PandocMonad m => OrgParser m BlockAttributes
blockAttributes = try $ do
  kv <- many stringyMetaAttribute
  guard $ all (isBlockAttr . fst) kv
  let caption = foldl' (appendValues "caption") Nothing kv
  let kvAttrs = foldl' (appendValues "attr_html") Nothing kv
  let name    = lookup "name" kv
  let label   = lookup "label" kv
  caption' <- traverse (parseFromString inlines . (<> "\n")) caption
  kvAttrs' <- parseFromString keyValues . (<> "\n") $ fromMaybe mempty kvAttrs
  return BlockAttributes
           { blockAttrName = name
           , blockAttrLabel = label
           , blockAttrCaption = caption'
           , blockAttrKeyValues = kvAttrs'
           }
 where
   isBlockAttr :: Text -> Bool
   isBlockAttr = flip elem
                 [ "name", "label", "caption"
                 , "attr_html", "attr_latex"
                 , "results"
                 ]

   appendValues :: Text -> Maybe Text -> (Text, Text) -> Maybe Text
   appendValues attrName accValue (key, value) =
     if key /= attrName
     then accValue
     else case accValue of
            Just acc -> Just $ acc <> " " <> value
            Nothing  -> Just value

-- | Parse key-value pairs for HTML attributes
keyValues :: Monad m => OrgParser m [(Text, Text)]
keyValues = try $
  manyTill ((,) <$> key <*> value) newline
 where
   key :: Monad m => OrgParser m Text
   key = try $ skipSpaces *> char ':' *> many1Char nonspaceChar

   value :: Monad m => OrgParser m Text
   value = skipSpaces *> manyTillChar anyChar endOfValue

   endOfValue :: Monad m => OrgParser m ()
   endOfValue =
     lookAhead $ (() <$ try (many1 spaceChar <* key))
              <|> () <$ newline


--
-- Org Blocks (#+begin_... / #+end_...)
--

-- | Read an org-mode block delimited by #+begin_type and #+end_type.
orgBlock :: PandocMonad m => OrgParser m (F Blocks)
orgBlock = try $ do
  blockAttrs <- blockAttributes
  blkType <- blockHeaderStart
  ($ blkType) $
    case T.toLower blkType of
      "export"  -> exportBlock
      "comment" -> rawBlockLines (const mempty)
      "html"    -> rawBlockLines (return . B.rawBlock (lowercase blkType))
      "latex"   -> rawBlockLines (return . B.rawBlock (lowercase blkType))
      "ascii"   -> rawBlockLines (return . B.rawBlock (lowercase blkType))
      "example" -> exampleBlock blockAttrs
      "quote"   -> parseBlockLines (fmap B.blockQuote)
      "verse"   -> verseBlock
      "src"     -> codeBlock blockAttrs
      _         -> parseBlockLines $
                   let (ident, classes, kv) = attrFromBlockAttributes blockAttrs
                   in fmap $ B.divWith (ident, classes ++ [blkType], kv)
 where
   blockHeaderStart :: Monad m => OrgParser m Text
   blockHeaderStart = try $ skipSpaces *> stringAnyCase "#+begin_" *> orgArgWord

   lowercase :: Text -> Text
   lowercase = T.toLower

exampleBlock :: PandocMonad m => BlockAttributes -> Text -> OrgParser m (F Blocks)
exampleBlock blockAttrs _label = do
  skipSpaces
  (classes, kv) <- switchesAsAttributes
  newline
  content <- rawBlockContent "example"
  let id' = fromMaybe mempty $ blockAttrName blockAttrs
  let codeBlck = B.codeBlockWith (id', "example":classes, kv) content
  return . return $ codeBlck

rawBlockLines :: Monad m => (Text   -> F Blocks) -> Text -> OrgParser m (F Blocks)
rawBlockLines f blockType = ignHeaders *> (f <$> rawBlockContent blockType)

parseBlockLines :: PandocMonad m => (F Blocks -> F Blocks) -> Text -> OrgParser m (F Blocks)
parseBlockLines f blockType = ignHeaders *> (f <$> parsedBlockContent)
 where
   parsedBlockContent :: PandocMonad m => OrgParser m (F Blocks)
   parsedBlockContent = try $ do
     raw <- rawBlockContent blockType
     parseFromString blocks (raw <> "\n")

-- | Read the raw string content of a block
rawBlockContent :: Monad m => Text -> OrgParser m Text
rawBlockContent blockType = try $ do
  blkLines <- manyTill rawLine blockEnder
  tabLen <- getOption readerTabStop
  trimP <- orgStateTrimLeadBlkIndent <$> getState
  let stripIndent strs = if trimP then map (T.drop (shortestIndent strs)) strs else strs
  (T.unlines
   . stripIndent
   . map (tabsToSpaces tabLen . commaEscaped)
   $ blkLines)
   <$ updateState (\s -> s { orgStateTrimLeadBlkIndent = True })
 where
   rawLine :: Monad m => OrgParser m Text
   rawLine = try $ ("" <$ blankline) <|> anyLine

   blockEnder :: Monad m => OrgParser m ()
   blockEnder = try $ skipSpaces <* stringAnyCase ("#+end_" <> blockType)

   shortestIndent :: [Text] -> Int
   shortestIndent = foldr (min . T.length . T.takeWhile isSpace) maxBound
                    . filter (not . T.null)

   tabsToSpaces :: Int -> Text -> Text
   tabsToSpaces tabStop t =
     let (ind, suff) = T.span (\c -> c == ' ' || c == '\t') t
         tabNum      = T.length $ T.filter (== '\n') ind
         spaceNum    = T.length ind - tabNum
     in  T.replicate (spaceNum + tabStop * tabNum) " " <> suff

   commaEscaped t =
     let (ind, suff) = T.span (\c -> c == ' ' || c == '\t') t
     in  case T.uncons suff of
           Just (',', cs)
             | "*"  <- T.take 1 cs -> ind <> cs
             | "#+" <- T.take 2 cs -> ind <> cs
           _                       -> t

-- | Read but ignore all remaining block headers.
ignHeaders :: Monad m => OrgParser m ()
ignHeaders = (() <$ newline) <|> (() <$ anyLine)

-- | Read a block containing code intended for export in specific backends
-- only.
exportBlock :: Monad m => Text -> OrgParser m (F Blocks)
exportBlock blockType = try $ do
  exportType <- skipSpaces *> orgArgWord <* ignHeaders
  contents   <- rawBlockContent blockType
  returnF (B.rawBlock (T.toLower exportType) contents)

verseBlock :: PandocMonad m => Text -> OrgParser m (F Blocks)
verseBlock blockType = try $ do
  ignHeaders
  content <- rawBlockContent blockType
  fmap B.lineBlock . sequence
    <$> mapM parseVerseLine (T.lines content)
 where
   -- replace initial spaces with nonbreaking spaces to preserve
   -- indentation, parse the rest as normal inline
   parseVerseLine :: PandocMonad m => Text -> OrgParser m (F Inlines)
   parseVerseLine cs = do
     let (initialSpaces, indentedLine) = T.span isSpace cs
     let nbspIndent = if T.null initialSpaces
                      then mempty
                      else B.str $ T.map (const '\160') initialSpaces
     line <- parseFromString inlines (indentedLine <> "\n")
     return (trimInlinesF $ pure nbspIndent <> line)

-- | Read a code block and the associated results block if present.  Which of
-- the blocks is included in the output is determined using the "exports"
-- argument in the block header.
codeBlock :: PandocMonad m => BlockAttributes -> Text -> OrgParser m (F Blocks)
codeBlock blockAttrs blockType = do
  skipSpaces
  (classes, kv)  <- codeHeaderArgs <|> (mempty <$ ignHeaders)
  content        <- rawBlockContent blockType
  resultsContent <- option mempty babelResultsBlock
  let identifier = fromMaybe mempty $ blockAttrName blockAttrs
  let codeBlk    = B.codeBlockWith (identifier, classes, kv) content
  let wrap       = maybe pure addCaption (blockAttrCaption blockAttrs)
  return $
    (if exportsCode kv    then wrap codeBlk   else mempty) <>
    (if exportsResults kv then resultsContent else mempty)
 where
   addCaption :: F Inlines -> Blocks -> F Blocks
   addCaption caption blk = B.divWith ("", ["captioned-content"], [])
                         <$> (mkCaptionBlock caption <> pure blk)

   mkCaptionBlock :: F Inlines -> F Blocks
   mkCaptionBlock = fmap (B.divWith ("", ["caption"], []) . B.plain)

   exportsResults :: [(Text, Text)] -> Bool
   exportsResults = maybe False (`elem` ["results", "both"]) . lookup "exports"

-- | Parse the result of an evaluated babel code block.
babelResultsBlock :: PandocMonad m => OrgParser m (F Blocks)
babelResultsBlock = try $ do
  blanklines
  resultsMarker <|>
    (lookAhead . void . try $
      manyTill (metaLineStart *> anyLineNewline) resultsMarker)
  block
 where
  resultsMarker = try . void $ stringAnyCase "#+RESULTS:" *> blankline

-- | Parse code block arguments
codeHeaderArgs :: Monad m => OrgParser m ([Text], [(Text, Text)])
codeHeaderArgs = try $ do
  language   <- skipSpaces *> orgArgWord
  (switchClasses, switchKv) <- switchesAsAttributes
  parameters <- manyTill blockOption newline
  return ( translateLang language : switchClasses
         , originalLang language <> switchKv <> parameters
         )

switchesAsAttributes :: Monad m => OrgParser m ([Text], [(Text, Text)])
switchesAsAttributes = try $ do
  switches <- skipSpaces *> try (switch `sepBy` many1 spaceChar)
  return $ foldr addToAttr ([], []) switches
 where
  addToAttr :: (Char, Maybe Text, SwitchPolarity)
            -> ([Text], [(Text, Text)])
            -> ([Text], [(Text, Text)])
  addToAttr ('n', lineNum, pol) (cls, kv) =
    let kv' = case lineNum of
                Just num -> ("startFrom", num):kv
                Nothing  -> kv
        cls' = case pol of
                 SwitchPlus  -> "continuedSourceBlock":cls
                 SwitchMinus -> cls
    in ("numberLines":cls', kv')
  addToAttr _ x = x

-- | Whether a switch flag is specified with @+@ or @-@.
data SwitchPolarity = SwitchPlus | SwitchMinus
  deriving (Show, Eq)

-- | Parses a switch's polarity.
switchPolarity :: Monad m => OrgParser m SwitchPolarity
switchPolarity = (SwitchMinus <$ char '-') <|> (SwitchPlus <$ char '+')

-- | Parses a source block switch option.
switch :: Monad m => OrgParser m (Char, Maybe Text, SwitchPolarity)
switch = try $ lineNumberSwitch <|> labelSwitch
               <|> whitespaceSwitch <|> simpleSwitch
 where
   simpleSwitch = (\pol c -> (c, Nothing, pol)) <$> switchPolarity <*> letter
   labelSwitch = genericSwitch 'l' $
     char '"' *> many1TillChar nonspaceChar (char '"')

whitespaceSwitch :: Monad m => OrgParser m (Char, Maybe Text, SwitchPolarity)
whitespaceSwitch = do
  string "-i"
  updateState $ \s -> s { orgStateTrimLeadBlkIndent = False }
  return ('i', Nothing, SwitchMinus)

-- | Generic source block switch-option parser.
genericSwitch :: Monad m
              => Char
              -> OrgParser m Text
              -> OrgParser m (Char, Maybe Text, SwitchPolarity)
genericSwitch c p = try $ do
  polarity <- switchPolarity <* char c <* skipSpaces
  arg <- optionMaybe p
  return (c, arg, polarity)

-- | Reads a line number switch option. The line number switch can be used with
-- example and source blocks.
lineNumberSwitch :: Monad m => OrgParser m (Char, Maybe Text, SwitchPolarity)
lineNumberSwitch = genericSwitch 'n' (manyChar digit)

blockOption :: Monad m => OrgParser m (Text, Text)
blockOption = try $ do
  argKey <- orgArgKey
  paramValue <- option "yes" orgParamValue
  return (argKey, paramValue)

orgParamValue :: Monad m => OrgParser m Text
orgParamValue = try $ fmap T.pack $
  skipSpaces
    *> notFollowedBy orgArgKey
    *> noneOf "\n\r" `many1Till` endOfValue
    <* skipSpaces
 where
  endOfValue = lookAhead $  try (skipSpaces <* oneOf "\n\r")
                        <|> try (skipSpaces1 <* orgArgKey)


--
-- Drawers
--

-- | A generic drawer which has no special meaning for org-mode.
-- Whether or not this drawer is included in the output depends on the drawers
-- export setting.
genericDrawer :: PandocMonad m => OrgParser m (F Blocks)
genericDrawer = try $ do
  name    <- T.toUpper <$> drawerStart
  content <- manyTill drawerLine (try drawerEnd)
  state   <- getState
  -- Include drawer if it is explicitly included in or not explicitly excluded
  -- from the list of drawers that should be exported.  PROPERTIES drawers are
  -- never exported.
  case exportDrawers . orgStateExportSettings $ state of
    _           | name == "PROPERTIES" -> return mempty
    Left  names | name `elem`    names -> return mempty
    Right names | name `notElem` names -> return mempty
    _           -> drawerDiv name <$> parseLines content
 where
  parseLines :: PandocMonad m => [Text] -> OrgParser m (F Blocks)
  parseLines = parseFromString blocks . (<> "\n") . T.unlines

  drawerDiv :: Text -> F Blocks -> F Blocks
  drawerDiv drawerName = fmap $ B.divWith (mempty, [drawerName, "drawer"], mempty)

drawerLine :: Monad m => OrgParser m Text
drawerLine = anyLine

drawerEnd :: Monad m => OrgParser m Text
drawerEnd = try $
  skipSpaces *> stringAnyCase ":END:" <* skipSpaces <* newline


--
-- Figures
--

-- | Figures or an image paragraph (i.e. an image on a line by itself). Only
-- images with a caption attribute are interpreted as figures.
figure :: PandocMonad m => OrgParser m (F Blocks)
figure = try $ do
  figAttrs <- blockAttributes
  src <- skipSpaces *> selfTarget <* skipSpaces <* endOfParagraph
  case cleanLinkText src of
    Nothing     -> mzero
    Just imgSrc -> do
      guard (isImageFilename imgSrc)
      let isFigure = isJust $ blockAttrCaption figAttrs
      return $ imageBlock isFigure figAttrs imgSrc
 where
   selfTarget :: PandocMonad m => OrgParser m Text
   selfTarget = try $ char '[' *> linkTarget <* char ']'

   imageBlock :: Bool -> BlockAttributes -> Text -> F Blocks
   imageBlock isFigure figAttrs imgSrc =
     let
       figName    = fromMaybe mempty $ blockAttrName figAttrs
       figLabel   = fromMaybe mempty $ blockAttrLabel figAttrs
       figCaption = fromMaybe mempty $ blockAttrCaption figAttrs
       figKeyVals = blockAttrKeyValues figAttrs
       attr       = (figLabel, mempty, figKeyVals)
       figTitle   = (if isFigure then withFigPrefix else id) figName
     in
       B.para . B.imageWith attr imgSrc figTitle <$> figCaption

   withFigPrefix :: Text -> Text
   withFigPrefix cs =
     if "fig:" `T.isPrefixOf` cs
     then cs
     else "fig:" <> cs

-- | Succeeds if looking at the end of the current paragraph
endOfParagraph :: Monad m => OrgParser m ()
endOfParagraph = try $ skipSpaces *> newline *> endOfBlock


--
-- Examples
--

-- | Example code marked up by a leading colon.
example :: Monad m => OrgParser m (F Blocks)
example = try $ returnF . exampleCode . T.unlines =<< many1 exampleLine
 where
   exampleLine :: Monad m => OrgParser m Text
   exampleLine = try $ exampleLineStart *> anyLine

exampleCode :: Text -> Blocks
exampleCode = B.codeBlockWith ("", ["example"], [])


--
-- Comments, Options and Metadata
--

specialLine :: PandocMonad m => OrgParser m (F Blocks)
specialLine = fmap return . try $ rawExportLine <|> metaLine <|> commentLine

-- | Include the content of a file.
include :: PandocMonad m => OrgParser m (F Blocks)
include = try $ do
  metaLineStart <* stringAnyCase "include:" <* skipSpaces
  filename <- includeTarget
  includeArgs <- many (try $ skipSpaces *> many1Char alphaNum)
  params <- keyValues
  blocksParser <- case includeArgs of
      ("example" : _) -> return $ pure . B.codeBlock <$> parseRaw
      ["export"] -> return . returnF $ B.fromList []
      ["export", format] -> return $ pure . B.rawBlock format <$> parseRaw
      ("src" : rest) -> do
        let attr = case rest of
                     [lang] -> (mempty, [lang], mempty)
                     _ -> nullAttr
        return $ pure . B.codeBlockWith attr <$> parseRaw
      _ -> return $ return . B.fromList . blockFilter params <$> blockList
  insertIncludedFile blocksParser toSources ["."] filename Nothing Nothing
 where
  includeTarget :: PandocMonad m => OrgParser m FilePath
  includeTarget = do
    char '"'
    manyTill (noneOf "\n\r\t") (char '"')

  parseRaw :: PandocMonad m => OrgParser m Text
  parseRaw = manyChar anyChar

  blockFilter :: [(Text, Text)] -> [Block] -> [Block]
  blockFilter params blks =
    let minlvl = lookup "minlevel" params
    in case (minlvl >>= safeRead :: Maybe Int) of
         Nothing -> blks
         Just lvl -> let levels = Walk.query headerLevel blks
                         curMin = maybe 0 minimum $ nonEmpty levels
                     in Walk.walk (shiftHeader (curMin - lvl)) blks

  headerLevel :: Block -> [Int]
  headerLevel (Header lvl _attr _content) = [lvl]
  headerLevel _ = []

  shiftHeader :: Int -> Block -> Block
  shiftHeader shift blk =
    case blk of
      (Header lvl attr content)
       | lvl - shift > 0  -> Header (lvl - shift) attr content
       | otherwise        -> Para content
      _ -> blk

-- | Parses a meta line which defines a raw block. Currently recognized:
-- @#+LATEX:@, @#+HTML:@, @#+TEXINFO:@, and @#+BEAMER@.
rawExportLine :: PandocMonad m => OrgParser m Blocks
rawExportLine = try $ do
  metaLineStart
  key <- metaKey
  if key `elem` ["latex", "html", "texinfo", "beamer"]
    then B.rawBlock key <$> anyLine
    else mzero

-- | Parses any meta line, i.e., a line starting with @#+@, into a raw
-- org block. This should be the last resort when trying to parse
-- keywords. Leading spaces are discarded.
rawOrgLine :: PandocMonad m => OrgParser m (F Blocks)
rawOrgLine = do
  line <- metaLineStart *> anyLine
  returnF $ B.rawBlock "org" $ "#+" <> line

commentLine :: Monad m => OrgParser m Blocks
commentLine = commentLineStart *> anyLine $> mempty


--
-- Tables
--
data ColumnProperty = ColumnProperty
  { columnAlignment :: Maybe Alignment
  , columnRelWidth  :: Maybe Int
  } deriving (Show, Eq)

instance Default ColumnProperty where
  def = ColumnProperty Nothing Nothing

data OrgTableRow = OrgContentRow (F [Blocks])
                 | OrgAlignRow [ColumnProperty]
                 | OrgHlineRow

-- OrgTable is strongly related to the pandoc table ADT.  Using the same
-- (i.e. pandoc-global) ADT would mean that the reader would break if the
-- global structure was to be changed, which would be bad.  The final table
-- should be generated using a builder function.
data OrgTable = OrgTable
  { orgTableColumnProperties :: [ColumnProperty]
  , orgTableHeader           :: [Blocks]
  , orgTableRows             :: [[Blocks]]
  }

table :: PandocMonad m => OrgParser m (F Blocks)
table = do
  withTables <- getExportSetting exportWithTables
  tbl <- gridTableWith blocks True <|> orgTable
  return $ if withTables then tbl else mempty

-- | A normal org table
orgTable :: PandocMonad m => OrgParser m (F Blocks)
orgTable = try $ do
  -- don't allow a table on the first line of a list item; org requires that
  -- tables start at first non-space character on the line
  let isFirstInListItem st = orgStateParserContext st == ListItemState &&
                             isNothing (orgStateLastPreCharPos st)
  guard . not . isFirstInListItem =<< getState
  blockAttrs <- blockAttributes
  lookAhead tableStart
  rows <- tableRows

  let caption = fromMaybe mempty (blockAttrCaption blockAttrs)
  let orgTbl = normalizeTable <$> rowsToTable rows
  -- wrap table in div if a name or label is given
  let identMb = blockAttrName blockAttrs `mplus` blockAttrLabel blockAttrs
  let wrap = case identMb of
        Just ident -> B.divWith (ident, mempty, mempty)
        Nothing    -> id
  return . fmap wrap $ (orgToPandocTable <$> orgTbl <*> caption)

orgToPandocTable :: OrgTable
                 -> Inlines
                 -> Blocks
orgToPandocTable (OrgTable colProps heads lns) caption =
  let totalWidth = if any (isJust . columnRelWidth) colProps
                   then Just . sum $ map (fromMaybe 1 . columnRelWidth) colProps
                   else Nothing
  in B.table (B.simpleCaption $ B.plain caption)
             (map (convertColProp totalWidth) colProps)
             (TableHead nullAttr $ toHeaderRow heads)
             [TableBody nullAttr 0 [] $ map toRow lns]
             (TableFoot nullAttr [])
 where
   toRow = Row nullAttr . map B.simpleCell
   toHeaderRow l = [toRow l | not (null l)]
   convertColProp :: Maybe Int -> ColumnProperty -> (Alignment, ColWidth)
   convertColProp totalWidth colProp =
     let
       align' = fromMaybe AlignDefault $ columnAlignment colProp
       width' = (\w t -> fromIntegral w / fromIntegral t)
                <$> columnRelWidth colProp
                <*> totalWidth
     in (align', maybe ColWidthDefault ColWidth width')

tableRows :: PandocMonad m => OrgParser m [OrgTableRow]
tableRows = try $ many (tableAlignRow <|> tableHline <|> tableContentRow)

tableContentRow :: PandocMonad m => OrgParser m OrgTableRow
tableContentRow = try $
  OrgContentRow . sequence <$> (tableStart *> many1Till tableContentCell newline)

tableContentCell :: PandocMonad m => OrgParser m (F Blocks)
tableContentCell = try $
  fmap B.plain . trimInlinesF . mconcat <$> manyTill inline endOfCell

tableAlignRow :: Monad m => OrgParser m OrgTableRow
tableAlignRow = try $ do
  tableStart
  colProps <- many1Till columnPropertyCell newline
  -- Empty rows are regular (i.e. content) rows, not alignment rows.
  guard $ any (/= def) colProps
  return $ OrgAlignRow colProps

columnPropertyCell :: Monad m => OrgParser m ColumnProperty
columnPropertyCell = emptyOrgCell <|> propCell <?> "alignment info"
 where
   emptyOrgCell = ColumnProperty Nothing Nothing <$ try (skipSpaces *> endOfCell)
   propCell = try $ ColumnProperty
                 <$> (skipSpaces
                      *> char '<'
                      *> optionMaybe tableAlignFromChar)
                 <*> (optionMaybe (many1Char digit >>= safeRead)
                      <* char '>'
                      <* emptyOrgCell)

tableAlignFromChar :: Monad m => OrgParser m Alignment
tableAlignFromChar = try $
  choice [ char 'l' $> AlignLeft
         , char 'c' $> AlignCenter
         , char 'r' $> AlignRight
         ]

tableHline :: Monad m => OrgParser m OrgTableRow
tableHline = try $
  OrgHlineRow <$ (tableStart *> char '-' *> anyLine)

endOfCell :: Monad m => OrgParser m Char
endOfCell = try $ char '|' <|> lookAhead newline

rowsToTable :: [OrgTableRow]
            -> F OrgTable
rowsToTable = foldM rowToContent emptyTable
 where emptyTable = OrgTable mempty mempty mempty

normalizeTable :: OrgTable -> OrgTable
normalizeTable (OrgTable colProps heads rows) =
  OrgTable colProps' heads rows
 where
   refRow = if heads /= mempty
            then heads
            else case rows of
                   (r:_) -> r
                   _     -> mempty
   cols = length refRow
   fillColumns base padding = take cols $ base ++ repeat padding
   colProps' = fillColumns colProps def

-- One or more horizontal rules after the first content line mark the previous
-- line as a header.  All other horizontal lines are discarded.
rowToContent :: OrgTable
             -> OrgTableRow
             -> F OrgTable
rowToContent tbl row =
  case row of
    OrgHlineRow       -> return singleRowPromotedToHeader
    OrgAlignRow props -> return . setProperties $ props
    OrgContentRow cs  -> appendToBody cs
 where
   singleRowPromotedToHeader :: OrgTable
   singleRowPromotedToHeader = case tbl of
     OrgTable{ orgTableHeader = [], orgTableRows = [b] } ->
            tbl{ orgTableHeader = b , orgTableRows = [] }
     _   -> tbl

   setProperties :: [ColumnProperty] -> OrgTable
   setProperties ps = tbl{ orgTableColumnProperties = ps }

   appendToBody :: F [Blocks] -> F OrgTable
   appendToBody frow = do
     newRow <- frow
     let oldRows = orgTableRows tbl
     -- NOTE: This is an inefficient O(n) operation.  This should be changed
     -- if performance ever becomes a problem.
     return tbl{ orgTableRows = oldRows ++ [newRow] }


--
-- LaTeX fragments
--
latexFragment :: PandocMonad m => OrgParser m (F Blocks)
latexFragment = try $ do
  envName <- latexEnvStart
  texOpt  <- getExportSetting exportWithLatex
  let envStart = "\\begin{" <> envName <> "}"
  let envEnd = "\\end{" <> envName <> "}"
  envLines <- do
    content <- manyTill anyLine (latexEnd envName)
    return $ envStart : content ++ [envEnd]
  returnF $ case texOpt of
    TeXExport -> B.rawBlock "latex" . T.unlines $ envLines
    TeXIgnore   -> mempty
    TeXVerbatim -> B.para . mconcat . intersperse B.softbreak $
                   map B.str envLines
 where
  latexEnd :: Monad m => Text -> OrgParser m ()
  latexEnd envName = try . void
     $ skipSpaces
    <* textStr ("\\end{" <> envName <> "}")
    <* blankline


--
-- Footnote definitions
--
noteBlock :: PandocMonad m => OrgParser m (F Blocks)
noteBlock = try $ do
  ref <- noteMarker <* skipSpaces <* updateLastPreCharPos
  content <- mconcat <$> many1Till block endOfFootnote
  addToNotesTable (ref, content)
  return mempty
 where
   endOfFootnote =  eof
                <|> () <$ lookAhead noteMarker
                <|> () <$ lookAhead headerStart
                <|> () <$ lookAhead (try $ blankline *> blankline)

-- Paragraphs or Plain text
paraOrPlain :: PandocMonad m => OrgParser m (F Blocks)
paraOrPlain = try $ do
  -- Make sure we are not looking at a headline
  notFollowedBy' headerStart
  ils <- inlines
  nl <- option False (newline $> True)
  -- Read block as paragraph, except if we are in a list context and the block
  -- is directly followed by a list item, in which case the block is read as
  -- plain text.
  try (guard nl
       *> notFollowedBy (inList *> (orderedListStart <|> bulletListStart))
       $> (B.para <$> ils))
    <|>  return (B.plain <$> ils)


--
-- list blocks
--

list :: PandocMonad m => OrgParser m (F Blocks)
list = choice [ definitionList, bulletList, orderedList ] <?> "list"

definitionList :: PandocMonad m => OrgParser m (F Blocks)
definitionList = try $ do
  indent <- lookAhead bulletListStart
  fmap (B.definitionList . compactifyDL) . sequence
    <$> many1 (definitionListItem (bulletListStart `indented` indent))

bulletList :: PandocMonad m => OrgParser m (F Blocks)
bulletList = try $ do
  indent <- lookAhead bulletListStart
  fmap (B.bulletList . compactify) . sequence
    <$> many1 (listItem (bulletListStart `indented` indent))

indented :: OrgParser m Int -> Int -> OrgParser m Int
indented indentedMarker minIndent = try $ do
  n <- indentedMarker
  guard (minIndent <= n)
  return n

orderedList :: PandocMonad m => OrgParser m (F Blocks)
orderedList = try $ do
  indent <- lookAhead orderedListStart
  fmap (B.orderedList . compactify) . sequence
    <$> many1 (listItem (orderedListStart `indented` indent))

definitionListItem :: PandocMonad m
                   => OrgParser m Int
                   -> OrgParser m (F (Inlines, [Blocks]))
definitionListItem parseIndentedMarker = try $ do
  markerLength <- parseIndentedMarker
  term <- manyTillChar (noneOf "\n\r") (try definitionMarker)
  line1 <- anyLineNewline
  blank <- option "" ("\n" <$ blankline)
  cont <- T.concat <$> many (listContinuation markerLength)
  term' <- parseFromString inlines term
  contents' <- parseFromString blocks $ line1 <> blank <> cont
  return $ (,) <$> term' <*> fmap (:[]) contents'
 where
   definitionMarker =
     spaceChar *> string "::" <* (spaceChar <|> lookAhead newline)

-- | Checkbox for tasks.
data Checkbox
  = UncheckedBox
  | CheckedBox
  | SemicheckedBox

-- | Parses a checkbox in a plain list.
checkbox :: PandocMonad m
         => OrgParser m Checkbox
checkbox = do
  guardEnabled Ext_task_lists
  try (char '[' *> status <* char ']') <?> "checkbox"
  where
    status = choice
      [ UncheckedBox   <$ char ' '
      , CheckedBox     <$ char 'X'
      , SemicheckedBox <$ char '-'
      ]

checkboxToInlines :: Checkbox -> Inline
checkboxToInlines = B.Str . \case
  UncheckedBox   -> "☐"
  SemicheckedBox -> "☐"
  CheckedBox     -> "☒"

-- | parse raw text for one list item
listItem :: PandocMonad m
         => OrgParser m Int
         -> OrgParser m (F Blocks)
listItem parseIndentedMarker = try . withContext ListItemState $ do
  markerLength <- try parseIndentedMarker
  box <- optionMaybe checkbox
  firstLine <- anyLineNewline
  blank <- option "" ("\n" <$ blankline)
  rest <- T.concat <$> many (listContinuation markerLength)
  contents <- parseFromString blocks $ firstLine <> blank <> rest
  return (maybe id (prependInlines . checkboxToInlines) box <$> contents)

-- | Prepend inlines to blocks, adding them to the first paragraph or
-- creating a new Plain element if necessary.
prependInlines :: Inline -> Blocks -> Blocks
prependInlines inlns = B.fromList . prepend . B.toList
  where
    prepend (Plain is : bs) = Plain (inlns : Space : is) : bs
    prepend (Para  is : bs) = Para  (inlns : Space : is) : bs
    prepend bs              = Plain [inlns, Space] : bs

-- continuation of a list item - indented and separated by blankline or endline.
-- Note: nested lists are parsed as continuations.
listContinuation :: PandocMonad m => Int -> OrgParser m Text
listContinuation markerLength = try $ do
  notFollowedBy' blankline
  mappend <$> (T.concat <$> many1 (listContinuation' markerLength))
          <*> manyChar blankline
 where
   listContinuation' indentation =
      blockLines indentation <|> listLine indentation
   listLine indentation = try $ indentWith indentation *> anyLineNewline
  -- The block attributes and start must be appropriately indented,
  -- but the contents, and end do not.
   blockLines indentation =
      try $ lookAhead (indentWith indentation
                       >> blockAttributes
                       >>= (\blockAttrs ->
                              case attrFromBlockAttributes blockAttrs of
                                ("", [], []) -> countChar 1 anyChar
                                _ -> indentWith indentation))
            >> (snd <$> withRaw orgBlock)