aboutsummaryrefslogtreecommitdiff
path: root/src/Text/Pandoc/Shared.hs
blob: b908a0172aff405ac258cd91afe34640dc3cc9ec (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
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
{-# LANGUAGE CPP                   #-}
{-# LANGUAGE FlexibleContexts      #-}
{-# LANGUAGE MultiParamTypeClasses #-}
{-# LANGUAGE ScopedTypeVariables   #-}
{-# LANGUAGE ViewPatterns          #-}
{-# LANGUAGE FlexibleInstances     #-}
{-# LANGUAGE LambdaCase            #-}
{-# LANGUAGE OverloadedStrings     #-}
{- |
   Module      : Text.Pandoc.Shared
   Copyright   : Copyright (C) 2006-2021 John MacFarlane
   License     : GNU GPL, version 2 or above

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

Utility functions and definitions used by the various Pandoc modules.
-}
module Text.Pandoc.Shared (
                     -- * List processing
                     splitBy,
                     splitTextBy,
                     splitByIndices,
                     splitStringByIndices,
                     splitTextByIndices,
                     substitute,
                     ordNub,
                     findM,
                     -- * Text processing
                     ToString (..),
                     ToText (..),
                     tshow,
                     backslashEscapes,
                     escapeStringUsing,
                     elemText,
                     notElemText,
                     stripTrailingNewlines,
                     trim,
                     triml,
                     trimr,
                     trimMath,
                     stripFirstAndLast,
                     camelCaseToHyphenated,
                     camelCaseStrToHyphenated,
                     toRomanNumeral,
                     escapeURI,
                     tabFilter,
                     crFilter,
                     -- * Date/time
                     normalizeDate,
                     -- * Pandoc block and inline list processing
                     orderedListMarkers,
                     extractSpaces,
                     removeFormatting,
                     deNote,
                     deLink,
                     stringify,
                     capitalize,
                     compactify,
                     compactifyDL,
                     linesToPara,
                     makeSections,
                     uniqueIdent,
                     inlineListToIdentifier,
                     isHeaderBlock,
                     headerShift,
                     stripEmptyParagraphs,
                     onlySimpleTableCells,
                     isTightList,
                     taskListItemFromAscii,
                     taskListItemToAscii,
                     addMetaField,
                     makeMeta,
                     eastAsianLineBreakFilter,
                     underlineSpan,
                     htmlSpanLikeElements,
                     splitSentences,
                     filterIpynbOutput,
                     -- * TagSoup HTML handling
                     renderTags',
                     -- * File handling
                     inDirectory,
                     collapseFilePath,
                     uriPathToPath,
                     filteredFilesFromArchive,
                     -- * URI handling
                     schemes,
                     isURI,
                     -- * Error handling
                     mapLeft,
                     -- * for squashing blocks
                     blocksToInlines,
                     blocksToInlines',
                     blocksToInlinesWithSep,
                     defaultBlocksSeparator,
                     -- * Safe read
                     safeRead,
                     safeStrRead,
                     -- * User data directory
                     defaultUserDataDirs,
                     -- * Version
                     pandocVersion
                    ) where

import Codec.Archive.Zip
import qualified Control.Exception as E
import Control.Monad (MonadPlus (..), msum, unless)
import qualified Control.Monad.State.Strict as S
import qualified Data.ByteString.Lazy as BL
import qualified Data.Bifunctor as Bifunctor
import Data.Char (isAlpha, isLower, isSpace, isUpper, toLower, isAlphaNum,
                  generalCategory, GeneralCategory(NonSpacingMark,
                  SpacingCombiningMark, EnclosingMark, ConnectorPunctuation))
import Data.List (find, intercalate, intersperse, stripPrefix, sortOn)
import qualified Data.Map as M
import Data.Maybe (mapMaybe, fromMaybe)
import Data.Monoid (Any (..))
import Data.Sequence (ViewL (..), ViewR (..), viewl, viewr)
import qualified Data.Set as Set
import qualified Data.Text as T
import Data.Version (showVersion)
import Network.URI (URI (uriScheme), escapeURIString, parseURI)
import Paths_pandoc (version)
import System.Directory
import System.FilePath (isPathSeparator, splitDirectories)
import qualified System.FilePath.Posix as Posix
import Text.HTML.TagSoup (RenderOptions (..), Tag (..), renderOptions,
                          renderTagsOptions)
import Text.Pandoc.Builder (Blocks, Inlines, ToMetaValue (..))
import qualified Text.Pandoc.Builder as B
import Data.Time
import Text.Pandoc.Asciify (toAsciiChar)
import Text.Pandoc.Definition
import Text.Pandoc.Extensions (Extensions, Extension(..), extensionEnabled)
import Text.Pandoc.Generic (bottomUp)
import Text.DocLayout (charWidth)
import Text.Pandoc.Walk

-- | Version number of pandoc library.
pandocVersion :: T.Text
pandocVersion = T.pack $ showVersion version

--
-- List processing
--

-- | Split list by groups of one or more sep.
splitBy :: (a -> Bool) -> [a] -> [[a]]
splitBy _ [] = []
splitBy isSep lst =
  let (first, rest) = break isSep lst
      rest'         = dropWhile isSep rest
  in  first:splitBy isSep rest'

splitTextBy :: (Char -> Bool) -> T.Text -> [T.Text]
splitTextBy isSep t
  | T.null t = []
  | otherwise = let (first, rest) = T.break isSep t
                    rest'         = T.dropWhile isSep rest
                in  first : splitTextBy isSep rest'

splitByIndices :: [Int] -> [a] -> [[a]]
splitByIndices [] lst = [lst]
splitByIndices (x:xs) lst = first:splitByIndices (map (\y -> y - x)  xs) rest
  where (first, rest) = splitAt x lst

-- | Split string into chunks divided at specified indices.
splitStringByIndices :: [Int] -> [Char] -> [[Char]]
splitStringByIndices [] lst = [lst]
splitStringByIndices (x:xs) lst =
  let (first, rest) = splitAt' x lst in
  first : splitStringByIndices (map (\y -> y - x) xs) rest

splitTextByIndices :: [Int] -> T.Text -> [T.Text]
splitTextByIndices ns = fmap T.pack . splitStringByIndices ns . T.unpack

splitAt' :: Int -> [Char] -> ([Char],[Char])
splitAt' _ []          = ([],[])
splitAt' n xs | n <= 0 = ([],xs)
splitAt' n (x:xs)      = (x:ys,zs)
  where (ys,zs) = splitAt' (n - charWidth x) xs

-- | Replace each occurrence of one sublist in a list with another.
substitute :: (Eq a) => [a] -> [a] -> [a] -> [a]
substitute _ _ [] = []
substitute [] _ xs = xs
substitute target replacement lst@(x:xs) =
    case stripPrefix target lst of
      Just lst' -> replacement ++ substitute target replacement lst'
      Nothing   -> x : substitute target replacement xs

ordNub :: (Ord a) => [a] -> [a]
ordNub l = go Set.empty l
  where
    go _ [] = []
    go s (x:xs) = if x `Set.member` s then go s xs
                                      else x : go (Set.insert x s) xs

findM :: forall m t a. (Monad m, Foldable t) => (a -> m Bool) -> t a -> m (Maybe a)
findM p = foldr go (pure Nothing)
  where
    go :: a -> m (Maybe a) -> m (Maybe a)
    go x acc = do
      b <- p x
      if b then pure (Just x) else acc

--
-- Text processing
--

class ToString a where
  toString :: a -> String

instance ToString String where
  toString = id

instance ToString T.Text where
  toString = T.unpack

class ToText a where
  toText :: a -> T.Text

instance ToText String where
  toText = T.pack

instance ToText T.Text where
  toText = id

tshow :: Show a => a -> T.Text
tshow = T.pack . show

-- | Returns an association list of backslash escapes for the
-- designated characters.
backslashEscapes :: [Char]    -- ^ list of special characters to escape
                 -> [(Char, T.Text)]
backslashEscapes = map (\ch -> (ch, T.pack ['\\',ch]))

-- | Escape a string of characters, using an association list of
-- characters and strings.
escapeStringUsing :: [(Char, T.Text)] -> T.Text -> T.Text
escapeStringUsing tbl = T.concatMap $ \c -> fromMaybe (T.singleton c) $ lookup c tbl

-- | @True@ exactly when the @Char@ appears in the @Text@.
elemText :: Char -> T.Text -> Bool
elemText c = T.any (== c)

-- | @True@ exactly when the @Char@ does not appear in the @Text@.
notElemText :: Char -> T.Text -> Bool
notElemText c = T.all (/= c)

-- | Strip trailing newlines from string.
stripTrailingNewlines :: T.Text -> T.Text
stripTrailingNewlines = T.dropWhileEnd (== '\n')

-- | Remove leading and trailing space (including newlines) from string.
trim :: T.Text -> T.Text
trim = T.dropAround (`elemText` " \r\n\t")

-- | Remove leading space (including newlines) from string.
triml :: T.Text -> T.Text
triml = T.dropWhile (`elemText` " \r\n\t")

-- | Remove trailing space (including newlines) from string.
trimr :: T.Text -> T.Text
trimr = T.dropWhileEnd (`elemText` " \r\n\t")

-- | Trim leading space and trailing space unless after \.
trimMath :: T.Text -> T.Text
trimMath = triml . T.reverse . stripBeginSpace . T.reverse -- no Text.spanEnd
  where
    stripBeginSpace t
      | T.null pref = t
      | Just ('\\', _) <- T.uncons suff = T.cons (T.last pref) suff
      | otherwise = suff
      where
        (pref, suff) = T.span (`elemText` " \t\n\r") t

-- | Strip leading and trailing characters from string
stripFirstAndLast :: T.Text -> T.Text
stripFirstAndLast t = case T.uncons t of
  Just (_, t') -> case T.unsnoc t' of
    Just (t'', _) -> t''
    _             -> t'
  _               -> ""

-- | Change CamelCase word to hyphenated lowercase (e.g., camel-case).
camelCaseToHyphenated :: T.Text -> T.Text
camelCaseToHyphenated = T.pack . camelCaseStrToHyphenated . T.unpack

-- This may not work as expected on general Unicode, if it contains
-- letters with a longer lower case form than upper case. I don't know
-- what the camel case practices of affected scripts are, though.
camelCaseStrToHyphenated :: String -> String
camelCaseStrToHyphenated [] = ""
camelCaseStrToHyphenated (a:b:rest)
  | isLower a
  , isUpper b = a:'-':toLower b:camelCaseStrToHyphenated rest
-- handle ABCDef = abc-def
camelCaseStrToHyphenated (a:b:c:rest)
  | isUpper a
  , isUpper b
  , isLower c = toLower a:'-':toLower b:camelCaseStrToHyphenated (c:rest)
camelCaseStrToHyphenated (a:rest) = toLower a:camelCaseStrToHyphenated rest

-- | Convert number < 4000 to uppercase roman numeral.
toRomanNumeral :: Int -> T.Text
toRomanNumeral x
  | x >= 4000 || x < 0 = "?"
  | x >= 1000 = "M" <> toRomanNumeral (x - 1000)
  | x >= 900  = "CM" <> toRomanNumeral (x - 900)
  | x >= 500  = "D" <> toRomanNumeral (x - 500)
  | x >= 400  = "CD" <> toRomanNumeral (x - 400)
  | x >= 100  = "C" <> toRomanNumeral (x - 100)
  | x >= 90   = "XC" <> toRomanNumeral (x - 90)
  | x >= 50   = "L"  <> toRomanNumeral (x - 50)
  | x >= 40   = "XL" <> toRomanNumeral (x - 40)
  | x >= 10   = "X" <> toRomanNumeral (x - 10)
  | x == 9    = "IX"
  | x >= 5    = "V" <> toRomanNumeral (x - 5)
  | x == 4    = "IV"
  | x >= 1    = "I" <> toRomanNumeral (x - 1)
  | otherwise = ""

-- | Escape whitespace and some punctuation characters in URI.
escapeURI :: T.Text -> T.Text
escapeURI = T.pack . escapeURIString (not . needsEscaping) . T.unpack
  where needsEscaping c = isSpace c || c `elemText` "<>|\"{}[]^`"


-- | Convert tabs to spaces. Tabs will be preserved if tab stop is set to 0.
tabFilter :: Int       -- ^ Tab stop
          -> T.Text    -- ^ Input
          -> T.Text
tabFilter 0 = id
tabFilter tabStop = T.unlines . map go . T.lines
  where go s =
         let (s1, s2) = T.break (== '\t') s
         in  if T.null s2
                then s1
                else s1 <> T.replicate
                       (tabStop - (T.length s1 `mod` tabStop)) (T.pack " ")
                       <> go (T.drop 1 s2)

-- | Strip out DOS line endings.
crFilter :: T.Text -> T.Text
crFilter = T.filter (/= '\r')

--
-- Date/time
--

-- | Parse a date and convert (if possible) to "YYYY-MM-DD" format. We
-- limit years to the range 1601-9999 (ISO 8601 accepts greater than
-- or equal to 1583, but MS Word only accepts dates starting 1601).
normalizeDate :: T.Text -> Maybe T.Text
normalizeDate = fmap T.pack . normalizeDate' . T.unpack

normalizeDate' :: String -> Maybe String
normalizeDate' s = fmap (formatTime defaultTimeLocale "%F")
  (msum $ map (\fs -> parsetimeWith fs s >>= rejectBadYear) formats :: Maybe Day)
  where rejectBadYear day = case toGregorian day of
          (y, _, _) | y >= 1601 && y <= 9999 -> Just day
          _         -> Nothing
        parsetimeWith = parseTimeM True defaultTimeLocale
        formats = ["%x","%m/%d/%Y", "%D","%F", "%d %b %Y",
                    "%e %B %Y", "%b. %e, %Y", "%B %e, %Y",
                    "%Y%m%d", "%Y%m", "%Y"]

--
-- Pandoc block and inline list processing
--

-- | Generate infinite lazy list of markers for an ordered list,
-- depending on list attributes.
orderedListMarkers :: (Int, ListNumberStyle, ListNumberDelim) -> [T.Text]
orderedListMarkers (start, numstyle, numdelim) =
  let nums = case numstyle of
                     DefaultStyle -> map tshow [start..]
                     Example      -> map tshow [start..]
                     Decimal      -> map tshow [start..]
                     UpperAlpha   -> drop (start - 1) $ cycle $
                                     map T.singleton ['A'..'Z']
                     LowerAlpha   -> drop (start - 1) $ cycle $
                                     map T.singleton ['a'..'z']
                     UpperRoman   -> map toRomanNumeral [start..]
                     LowerRoman   -> map (T.toLower . toRomanNumeral) [start..]
      inDelim str = case numdelim of
                            DefaultDelim -> str <> "."
                            Period       -> str <> "."
                            OneParen     -> str <> ")"
                            TwoParens    -> "(" <> str <> ")"
  in  map inDelim nums


-- | Extract the leading and trailing spaces from inside an inline element
-- and place them outside the element.  SoftBreaks count as Spaces for
-- these purposes.
extractSpaces :: (Inlines -> Inlines) -> Inlines -> Inlines
extractSpaces f is =
  let contents = B.unMany is
      left  = case viewl contents of
                    (Space :< _)     -> B.space
                    (SoftBreak :< _) -> B.softbreak
                    _                -> mempty
      right = case viewr contents of
                    (_ :> Space)     -> B.space
                    (_ :> SoftBreak) -> B.softbreak
                    _                -> mempty in
  (left <> f (B.trimInlines . B.Many $ contents) <> right)

-- | Extract inlines, removing formatting.
removeFormatting :: Walkable Inline a => a -> [Inline]
removeFormatting = query go . walk (deNote . deQuote)
  where go :: Inline -> [Inline]
        go (Str xs)   = [Str xs]
        go Space      = [Space]
        go SoftBreak  = [SoftBreak]
        go (Code _ x) = [Str x]
        go (Math _ x) = [Str x]
        go LineBreak  = [Space]
        go _          = []

deNote :: Inline -> Inline
deNote (Note _) = Str ""
deNote x        = x

deLink :: Inline -> Inline
deLink (Link _ ils _) = Span nullAttr ils
deLink x              = x

deQuote :: Inline -> Inline
deQuote (Quoted SingleQuote xs) =
  Span ("",[],[]) (Str "\8216" : xs ++ [Str "\8217"])
deQuote (Quoted DoubleQuote xs) =
  Span ("",[],[]) (Str "\8220" : xs ++ [Str "\8221"])
deQuote x = x

-- | Convert pandoc structure to a string with formatting removed.
-- Footnotes are skipped (since we don't want their contents in link
-- labels).
stringify :: Walkable Inline a => a -> T.Text
stringify = query go . walk (deNote . deQuote)
  where go :: Inline -> T.Text
        go Space                                       = " "
        go SoftBreak                                   = " "
        go (Str x)                                     = x
        go (Code _ x)                                  = x
        go (Math _ x)                                  = x
        go (RawInline (Format "html") (T.unpack -> ('<':'b':'r':_)))
                                                       = " " -- see #2105
        go LineBreak                                   = " "
        go _                                           = ""

-- | Bring all regular text in a pandoc structure to uppercase.
--
-- This function correctly handles cases where a lowercase character doesn't
-- match to a single uppercase character – e.g. “Straße” would be converted
-- to “STRASSE”, not “STRAßE”.
capitalize :: Walkable Inline a => a -> a
capitalize = walk go
  where go :: Inline -> Inline
        go (Str s) = Str $ T.toUpper s
        go x       = x

-- | Change final list item from @Para@ to @Plain@ if the list contains
-- no other @Para@ blocks.  Otherwise (if the list items contain @Para@
-- blocks besides possibly at the end), turn any @Plain@s into @Para@s (#5285).
compactify :: [Blocks]  -- ^ List of list items (each a list of blocks)
           -> [Blocks]
compactify [] = []
compactify items =
  let (others, final) = (init items, last items)
  in  case reverse (B.toList final) of
           (Para a:xs)
             | null [Para x | Para x <- xs ++ concatMap B.toList others]
             -> others ++ [B.fromList (reverse (Plain a : xs))]
           _ | null [Para x | Para x <- concatMap B.toList items]
             -> items
           _ -> map (fmap plainToPara) items

plainToPara :: Block -> Block
plainToPara (Plain ils) = Para ils
plainToPara x = x

-- | Like @compactify@, but acts on items of definition lists.
compactifyDL :: [(Inlines, [Blocks])] -> [(Inlines, [Blocks])]
compactifyDL items =
  let defs = concatMap snd items
  in  case reverse (concatMap B.toList defs) of
           (Para x:xs)
             | not (any isPara xs) ->
                   let (t,ds) = last items
                       lastDef = B.toList $ last ds
                       ds' = init ds ++
                             if null lastDef
                                then [B.fromList lastDef]
                                else [B.fromList $ init lastDef ++ [Plain x]]
                    in init items ++ [(t, ds')]
             | otherwise           -> items
           _                       -> items

-- | Combine a list of lines by adding hard linebreaks.
combineLines :: [[Inline]] -> [Inline]
combineLines = intercalate [LineBreak]

-- | Convert a list of lines into a paragraph with hard line breaks. This is
--   useful e.g. for rudimentary support of LineBlock elements in writers.
linesToPara :: [[Inline]] -> Block
linesToPara = Para . combineLines

isPara :: Block -> Bool
isPara (Para _) = True
isPara _        = False

-- | Convert Pandoc inline list to plain text identifier.  HTML
-- identifiers must start with a letter, and may contain only
-- letters, digits, and the characters _-.
inlineListToIdentifier :: Extensions -> [Inline] -> T.Text
inlineListToIdentifier exts =
  dropNonLetter . filterAscii . toIdent . stringify . walk unEmojify
  where
    unEmojify :: [Inline] -> [Inline]
    unEmojify
      | extensionEnabled Ext_gfm_auto_identifiers exts ||
        extensionEnabled Ext_ascii_identifiers exts = walk unEmoji
      | otherwise = id
    unEmoji (Span ("",["emoji"],[("data-emoji",ename)]) _) = Str ename
    unEmoji x = x
    dropNonLetter
      | extensionEnabled Ext_gfm_auto_identifiers exts = id
      | otherwise = T.dropWhile (not . isAlpha)
    filterAscii
      | extensionEnabled Ext_ascii_identifiers exts
        = T.pack . mapMaybe toAsciiChar . T.unpack
      | otherwise = id
    toIdent
      | extensionEnabled Ext_gfm_auto_identifiers exts =
        filterPunct . spaceToDash . T.toLower
      | otherwise = T.intercalate "-" . T.words . filterPunct . T.toLower
    filterPunct = T.filter (\c -> isSpace c || isAlphaNum c || isAllowedPunct c)
    isAllowedPunct c
      | extensionEnabled Ext_gfm_auto_identifiers exts
        = c == '-' || c == '_' ||
          generalCategory c `elem` [NonSpacingMark, SpacingCombiningMark,
                                    EnclosingMark, ConnectorPunctuation]
      | otherwise = c == '_' || c == '-' || c == '.'
    spaceToDash = T.map (\c -> if isSpace c then '-' else c)


-- | Put a list of Pandoc blocks into a hierarchical structure:
-- a list of sections (each a Div with class "section" and first
-- element a Header).  If the 'numbering' parameter is True, Header
-- numbers are added via the number attribute on the header.
-- If the baseLevel parameter is Just n, Header levels are
-- adjusted to be gapless starting at level n.
makeSections :: Bool -> Maybe Int -> [Block] -> [Block]
makeSections numbering mbBaseLevel bs =
  S.evalState (go bs) (mbBaseLevel, [])
 where
  go :: [Block] -> S.State (Maybe Int, [Int]) [Block]
  go (Header level (ident,classes,kvs) title':xs) = do
    (mbLevel, lastnum) <- S.get
    let level' = fromMaybe level mbLevel
    let lastnum' = take level' lastnum
    let newnum =
          if level' > 0
             then case length lastnum' of
                      x | "unnumbered" `elem` classes -> []
                        | x >= level' -> init lastnum' ++ [last lastnum' + 1]
                        | otherwise -> lastnum ++
                             replicate (level' - length lastnum - 1) 0 ++ [1]
             else []
    unless (null newnum) $ S.modify $ \(mbl, _) -> (mbl, newnum)
    let (sectionContents, rest) = break (headerLtEq level) xs
    S.modify $ \(_, ln) -> (fmap (+ 1) mbLevel, ln)
    sectionContents' <- go sectionContents
    S.modify $ \(_, ln) -> (mbLevel, ln)
    rest' <- go rest
    let kvs' = -- don't touch number if already present
               case lookup "number" kvs of
                  Nothing | numbering
                          , "unnumbered" `notElem` classes ->
                        ("number", T.intercalate "." (map tshow newnum)) : kvs
                  _ -> kvs
    let divattr = (ident, "section":classes, kvs')
    let attr = ("",classes,kvs')
    return $
      Div divattr (Header level' attr title' : sectionContents') : rest'
  go (Div divattr@(dident,dclasses,_) (Header level hattr title':ys) : xs)
      | all (\case
               Header level' _ _ -> level' > level
               _                 -> True) ys
      , "column" `notElem` dclasses
      , "columns" `notElem` dclasses = do
    inner <- go (Header level hattr title':ys)
    rest <- go xs
    return $
      case inner of
            [Div divattr'@(dident',_,_) zs]
              | T.null dident || T.null dident' || dident == dident'
              -> Div (combineAttr divattr' divattr) zs : rest
            _ -> Div divattr inner : rest
  go (Div attr xs : rest) = do
    xs' <- go xs
    rest' <- go rest
    return $ Div attr xs' : rest'
  go (x:xs) = (x :) <$> go xs
  go [] = return []

  combineAttr :: Attr -> Attr -> Attr
  combineAttr (id1, classes1, kvs1) (id2, classes2, kvs2) =
    (if T.null id1 then id2 else id1,
     ordNub (classes1 ++ classes2),
     foldr (\(k,v) kvs -> case lookup k kvs of
                             Nothing -> (k,v):kvs
                             Just _  -> kvs) mempty (kvs1 ++ kvs2))

headerLtEq :: Int -> Block -> Bool
headerLtEq level (Header l _ _)  = l <= level
headerLtEq level (Div _ (b:_))   = headerLtEq level b
headerLtEq _ _                   = False

-- | Generate a unique identifier from a list of inlines.
-- Second argument is a list of already used identifiers.
uniqueIdent :: Extensions -> [Inline] -> Set.Set T.Text -> T.Text
uniqueIdent exts title' usedIdents =
  if baseIdent `Set.member` usedIdents
     then maybe baseIdent numIdent
          $ find (\x -> numIdent x `Set.notMember` usedIdents) ([1..60000] :: [Int])
          -- if we have more than 60,000, allow repeats
     else baseIdent
  where
    baseIdent = case inlineListToIdentifier exts title' of
                     "" -> "section"
                     x  -> x
    numIdent n = baseIdent <> "-" <> tshow n

-- | True if block is a Header block.
isHeaderBlock :: Block -> Bool
isHeaderBlock Header{} = True
isHeaderBlock _        = False

-- | Shift header levels up or down.
headerShift :: Int -> Pandoc -> Pandoc
headerShift n (Pandoc meta (Header m _ ils : bs))
  | n < 0
  , m + n == 0 = headerShift n $
                 B.setTitle (B.fromList ils) $ Pandoc meta bs
headerShift n (Pandoc meta bs) = Pandoc meta (walk shift bs)

 where
   shift :: Block -> Block
   shift (Header level attr inner)
     | level + n > 0  = Header (level + n) attr inner
     | otherwise      = Para inner
   shift x            = x

-- | Remove empty paragraphs.
stripEmptyParagraphs :: Pandoc -> Pandoc
stripEmptyParagraphs = walk go
  where go :: [Block] -> [Block]
        go = filter (not . isEmptyParagraph)
        isEmptyParagraph (Para []) = True
        isEmptyParagraph _         = False

-- | Detect if table rows contain only cells consisting of a single
-- paragraph that has no @LineBreak@.
onlySimpleTableCells :: [[[Block]]] -> Bool
onlySimpleTableCells = all isSimpleCell . concat
  where
    isSimpleCell [Plain ils] = not (hasLineBreak ils)
    isSimpleCell [Para ils ] = not (hasLineBreak ils)
    isSimpleCell []          = True
    isSimpleCell _           = False
    hasLineBreak = getAny . query isLineBreak
    isLineBreak LineBreak = Any True
    isLineBreak _         = Any False

-- | Detect if a list is tight.
isTightList :: [[Block]] -> Bool
isTightList = all (\item -> firstIsPlain item || null item)
  where firstIsPlain (Plain _ : _) = True
        firstIsPlain _             = False

-- | Convert a list item containing tasklist syntax (e.g. @[x]@)
-- to using @U+2610 BALLOT BOX@ or @U+2612 BALLOT BOX WITH X@.
taskListItemFromAscii :: Extensions -> [Block] -> [Block]
taskListItemFromAscii = handleTaskListItem fromMd
  where
    fromMd (Str "[" : Space : Str "]" : Space : is) = Str "☐" : Space : is
    fromMd (Str "[x]"                 : Space : is) = Str "☒" : Space : is
    fromMd (Str "[X]"                 : Space : is) = Str "☒" : Space : is
    fromMd is = is

-- | Convert a list item containing text starting with @U+2610 BALLOT BOX@
-- or @U+2612 BALLOT BOX WITH X@ to tasklist syntax (e.g. @[x]@).
taskListItemToAscii :: Extensions -> [Block] -> [Block]
taskListItemToAscii = handleTaskListItem toMd
  where
    toMd (Str "☐" : Space : is) = rawMd "[ ]" : Space : is
    toMd (Str "☒" : Space : is) = rawMd "[x]" : Space : is
    toMd is = is
    rawMd = RawInline (Format "markdown")

handleTaskListItem :: ([Inline] -> [Inline]) -> Extensions -> [Block] -> [Block]
handleTaskListItem handleInlines exts bls =
  if Ext_task_lists `extensionEnabled` exts
  then handleItem bls
  else bls
  where
    handleItem (Plain is : bs) = Plain (handleInlines is) : bs
    handleItem (Para is  : bs) = Para  (handleInlines is) : bs
    handleItem bs = bs

-- | Set a field of a 'Meta' object.  If the field already has a value,
-- convert it into a list with the new value appended to the old value(s).
addMetaField :: ToMetaValue a
             => T.Text
             -> a
             -> Meta
             -> Meta
addMetaField key val (Meta meta) =
  Meta $ M.insertWith combine key (toMetaValue val) meta
  where combine newval (MetaList xs) = MetaList (xs ++ tolist newval)
        combine newval x             = MetaList [x, newval]
        tolist (MetaList ys) = ys
        tolist y             = [y]

-- | Create 'Meta' from old-style title, authors, date.  This is
-- provided to ease the transition from the old API.
makeMeta :: [Inline] -> [[Inline]] -> [Inline] -> Meta
makeMeta title authors date =
      addMetaField "title" (B.fromList title)
    $ addMetaField "author" (map B.fromList authors)
    $ addMetaField "date" (B.fromList date) nullMeta

-- | Remove soft breaks between East Asian characters.
eastAsianLineBreakFilter :: Pandoc -> Pandoc
eastAsianLineBreakFilter = bottomUp go
  where go (x:SoftBreak:y:zs)
          | Just (_, b) <- T.unsnoc $ stringify x
          , Just (c, _) <- T.uncons $ stringify y
          , charWidth b == 2
          , charWidth c == 2
          = x:y:zs
          | otherwise
          = x:SoftBreak:y:zs
        go xs
          = xs

{-# DEPRECATED underlineSpan "Use Text.Pandoc.Builder.underline instead" #-}
-- | Builder for underline (deprecated).
-- This probably belongs in Builder.hs in pandoc-types.
-- Will be replaced once Underline is an element.
underlineSpan :: Inlines -> Inlines
underlineSpan = B.underline

-- | Set of HTML elements that are represented as Span with a class equal as
-- the element tag itself.
htmlSpanLikeElements :: Set.Set T.Text
htmlSpanLikeElements = Set.fromList ["kbd", "mark", "dfn"]

-- | Returns the first sentence in a list of inlines, and the rest.
breakSentence :: [Inline] -> ([Inline], [Inline])
breakSentence [] = ([],[])
breakSentence xs =
  let isSentenceEndInline (Str ys)
        | Just (_, c) <- T.unsnoc ys = c == '.' || c == '?'
      isSentenceEndInline LineBreak  = True
      isSentenceEndInline _          = False
      (as, bs) = break isSentenceEndInline xs
  in  case bs of
        []             -> (as, [])
        [c]            -> (as ++ [c], [])
        (c:Space:cs)   -> (as ++ [c], cs)
        (c:SoftBreak:cs) -> (as ++ [c], cs)
        (Str ".":Str s@(T.uncons -> Just (')',_)):cs)
          -> (as ++ [Str ".", Str s], cs)
        (x@(Str (T.stripPrefix ".)" -> Just _)):cs) -> (as ++ [x], cs)
        (LineBreak:x@(Str (T.uncons -> Just ('.',_))):cs) -> (as ++[LineBreak], x:cs)
        (c:cs)         -> (as ++ [c] ++ ds, es)
          where (ds, es) = breakSentence cs

-- | Split a list of inlines into sentences.
splitSentences :: [Inline] -> [[Inline]]
splitSentences xs =
  let (sent, rest) = breakSentence xs
  in  if null rest then [sent] else sent : splitSentences rest

-- | Process ipynb output cells.  If mode is Nothing,
-- remove all output.  If mode is Just format, select
-- best output for the format.  If format is not ipynb,
-- strip out ANSI escape sequences from CodeBlocks (see #5633).
filterIpynbOutput :: Maybe Format -> Pandoc -> Pandoc
filterIpynbOutput mode = walk go
  where go (Div (ident, "output":os, kvs) bs) =
          case mode of
            Nothing  -> Div (ident, "output":os, kvs) []
            -- "best" for ipynb includes all formats:
            Just fmt
              | fmt == Format "ipynb"
                          -> Div (ident, "output":os, kvs) bs
              | otherwise -> Div (ident, "output":os, kvs) $
                              walk removeANSI $
                              take 1 $ sortOn rank bs
                 where
                  rank (RawBlock (Format "html") _)
                    | fmt == Format "html" = 1 :: Int
                    | fmt == Format "markdown" = 2
                    | otherwise = 3
                  rank (RawBlock (Format "latex") _)
                    | fmt == Format "latex" = 1
                    | fmt == Format "markdown" = 2
                    | otherwise = 3
                  rank (RawBlock f _)
                    | fmt == f = 1
                    | otherwise = 3
                  rank (Para [Image{}]) = 1
                  rank _ = 2
                  removeANSI (CodeBlock attr code) =
                    CodeBlock attr (removeANSIEscapes code)
                  removeANSI x = x
                  removeANSIEscapes t
                    | Just cs <- T.stripPrefix "\x1b[" t =
                        removeANSIEscapes $ T.drop 1 $ T.dropWhile (/='m') cs
                    | Just (c, cs) <- T.uncons t = T.cons c $ removeANSIEscapes cs
                    | otherwise = ""
        go x = x

--
-- TagSoup HTML handling
--

-- | Render HTML tags.
renderTags' :: [Tag T.Text] -> T.Text
renderTags' = renderTagsOptions
               renderOptions{ optMinimize = matchTags ["hr", "br", "img",
                                                       "meta", "link", "col"]
                            , optRawTag   = matchTags ["script", "style"] }
              where matchTags tags = flip elem tags . T.toLower

--
-- File handling
--

-- | Perform an IO action in a directory, returning to starting directory.
inDirectory :: FilePath -> IO a -> IO a
inDirectory path action = E.bracket
                             getCurrentDirectory
                             setCurrentDirectory
                             (const $ setCurrentDirectory path >> action)

--
-- Error reporting
--

mapLeft :: (a -> b) -> Either a c -> Either b c
mapLeft = Bifunctor.first

-- | Remove intermediate "." and ".." directories from a path.
--
-- > collapseFilePath "./foo" == "foo"
-- > collapseFilePath "/bar/../baz" == "/baz"
-- > collapseFilePath "/../baz" == "/../baz"
-- > collapseFilePath "parent/foo/baz/../bar" ==  "parent/foo/bar"
-- > collapseFilePath "parent/foo/baz/../../bar" ==  "parent/bar"
-- > collapseFilePath "parent/foo/.." ==  "parent"
-- > collapseFilePath "/parent/foo/../../bar" ==  "/bar"
collapseFilePath :: FilePath -> FilePath
collapseFilePath = Posix.joinPath . reverse . foldl go [] . splitDirectories
  where
    go rs "." = rs
    go r@(p:rs) ".." = case p of
                            ".."                              -> "..":r
                            (checkPathSeperator -> Just True) -> "..":r
                            _                                 -> rs
    go _ (checkPathSeperator -> Just True) = [[Posix.pathSeparator]]
    go rs x = x:rs
    isSingleton []  = Nothing
    isSingleton [x] = Just x
    isSingleton _   = Nothing
    checkPathSeperator = fmap isPathSeparator . isSingleton

-- Convert the path part of a file: URI to a regular path.
-- On windows, @/c:/foo@ should be @c:/foo@.
-- On linux, @/foo@ should be @/foo@.
uriPathToPath :: T.Text -> FilePath
uriPathToPath (T.unpack -> path) =
#ifdef _WINDOWS
  case path of
    '/':ps -> ps
    ps     -> ps
#else
  path
#endif

--
-- File selection from the archive
--
filteredFilesFromArchive :: Archive -> (FilePath -> Bool) -> [(FilePath, BL.ByteString)]
filteredFilesFromArchive zf f =
  mapMaybe (fileAndBinary zf) (filter f (filesInArchive zf))
  where
    fileAndBinary :: Archive -> FilePath -> Maybe (FilePath, BL.ByteString)
    fileAndBinary a fp = findEntryByPath fp a >>= \e -> Just (fp, fromEntry e)


--
-- IANA URIs
--

-- | Schemes from http://www.iana.org/assignments/uri-schemes.html plus
-- the unofficial schemes doi, javascript, isbn, pmid.
schemes :: Set.Set T.Text
schemes = Set.fromList
  -- Official IANA schemes
  [ "aaa", "aaas", "about", "acap", "acct", "acr", "adiumxtra", "afp", "afs"
  , "aim", "appdata", "apt", "attachment", "aw", "barion", "beshare", "bitcoin"
  , "blob", "bolo", "browserext", "callto", "cap", "chrome", "chrome-extension"
  , "cid", "coap", "coaps", "com-eventbrite-attendee", "content", "crid", "cvs"
  , "data", "dav", "dict", "dis", "dlna-playcontainer", "dlna-playsingle"
  , "dns", "dntp", "dtn", "dvb", "ed2k", "example", "facetime", "fax", "feed"
  , "feedready", "file", "filesystem", "finger", "fish", "ftp", "geo", "gg"
  , "git", "gizmoproject", "go", "gopher", "graph", "gtalk", "h323", "ham"
  , "hcp", "http", "https", "hxxp", "hxxps", "hydrazone", "iax", "icap", "icon"
  , "im", "imap", "info", "iotdisco", "ipn", "ipp", "ipps", "irc", "irc6"
  , "ircs", "iris", "iris.beep", "iris.lwz", "iris.xpc", "iris.xpcs"
  , "isostore", "itms", "jabber", "jar", "jms", "keyparc", "lastfm", "ldap"
  , "ldaps", "lvlt", "magnet", "mailserver", "mailto", "maps", "market"
  , "message", "mid", "mms", "modem", "mongodb", "moz", "ms-access"
  , "ms-browser-extension", "ms-drive-to", "ms-enrollment", "ms-excel"
  , "ms-gamebarservices", "ms-getoffice", "ms-help", "ms-infopath"
  , "ms-media-stream-id", "ms-officeapp", "ms-project", "ms-powerpoint"
  , "ms-publisher", "ms-search-repair", "ms-secondary-screen-controller"
  , "ms-secondary-screen-setup", "ms-settings", "ms-settings-airplanemode"
  , "ms-settings-bluetooth", "ms-settings-camera", "ms-settings-cellular"
  , "ms-settings-cloudstorage", "ms-settings-connectabledevices"
  , "ms-settings-displays-topology", "ms-settings-emailandaccounts"
  , "ms-settings-language", "ms-settings-location", "ms-settings-lock"
  , "ms-settings-nfctransactions", "ms-settings-notifications"
  , "ms-settings-power", "ms-settings-privacy", "ms-settings-proximity"
  , "ms-settings-screenrotation", "ms-settings-wifi", "ms-settings-workplace"
  , "ms-spd", "ms-sttoverlay", "ms-transit-to", "ms-virtualtouchpad"
  , "ms-visio", "ms-walk-to", "ms-whiteboard", "ms-whiteboard-cmd", "ms-word"
  , "msnim", "msrp", "msrps", "mtqp", "mumble", "mupdate", "mvn", "news", "nfs"
  , "ni", "nih", "nntp", "notes", "ocf", "oid", "onenote", "onenote-cmd"
  , "opaquelocktoken", "pack", "palm", "paparazzi", "pkcs11", "platform", "pop"
  , "pres", "prospero", "proxy", "pwid", "psyc", "qb", "query", "redis"
  , "rediss", "reload", "res", "resource", "rmi", "rsync", "rtmfp", "rtmp"
  , "rtsp", "rtsps", "rtspu", "secondlife", "service", "session", "sftp", "sgn"
  , "shttp", "sieve", "sip", "sips", "skype", "smb", "sms", "smtp", "snews"
  , "snmp", "soap.beep", "soap.beeps", "soldat", "spotify", "ssh", "steam"
  , "stun", "stuns", "submit", "svn", "tag", "teamspeak", "tel", "teliaeid"
  , "telnet", "tftp", "things", "thismessage", "tip", "tn3270", "tool", "turn"
  , "turns", "tv", "udp", "unreal", "urn", "ut2004", "v-event", "vemmi"
  , "ventrilo", "videotex", "vnc", "view-source", "wais", "webcal", "wpid"
  , "ws", "wss", "wtai", "wyciwyg", "xcon", "xcon-userid", "xfire"
  , "xmlrpc.beep", "xmlrpc.beeps", "xmpp", "xri", "ymsgr", "z39.50", "z39.50r"
  , "z39.50s"
  -- Unofficial schemes
  , "doi", "isbn", "javascript", "pmid"
  ]

-- | Check if the string is a valid URL with a IANA or frequently used but
-- unofficial scheme (see @schemes@).
isURI :: T.Text -> Bool
isURI = maybe False hasKnownScheme . parseURI . T.unpack
  where
    hasKnownScheme = (`Set.member` schemes) . T.toLower .
                     T.filter (/= ':') . T.pack . uriScheme

---
--- Squash blocks into inlines
---

blockToInlines :: Block -> Inlines
blockToInlines (Plain ils) = B.fromList ils
blockToInlines (Para ils) = B.fromList ils
blockToInlines (LineBlock lns) = B.fromList $ combineLines lns
blockToInlines (CodeBlock attr str) = B.codeWith attr str
blockToInlines (RawBlock (Format fmt) str) = B.rawInline fmt str
blockToInlines (BlockQuote blks) = blocksToInlines' blks
blockToInlines (OrderedList _ blkslst) =
  mconcat $ map blocksToInlines' blkslst
blockToInlines (BulletList blkslst) =
  mconcat $ map blocksToInlines' blkslst
blockToInlines (DefinitionList pairslst) =
  mconcat $ map f pairslst
  where
    f (ils, blkslst) = B.fromList ils <> B.str ":" <> B.space <>
      mconcat (map blocksToInlines' blkslst)
blockToInlines (Header _ _  ils) = B.fromList ils
blockToInlines HorizontalRule = mempty
blockToInlines (Table _ _ _ (TableHead _ hbd) bodies (TableFoot _ fbd)) =
  mconcat $ intersperse B.linebreak $
    map (mconcat . map blocksToInlines') (plainRowBody <$> hbd <> unTableBodies bodies <> fbd)
  where
    plainRowBody (Row _ body) = cellBody <$> body
    cellBody (Cell _ _ _ _ body) = body
    unTableBody (TableBody _ _ hd bd) = hd <> bd
    unTableBodies = concatMap unTableBody
blockToInlines (Div _ blks) = blocksToInlines' blks
blockToInlines Null = mempty

blocksToInlinesWithSep :: Inlines -> [Block] -> Inlines
blocksToInlinesWithSep sep =
  mconcat . intersperse sep . map blockToInlines

blocksToInlines' :: [Block] -> Inlines
blocksToInlines' = blocksToInlinesWithSep defaultBlocksSeparator

blocksToInlines :: [Block] -> [Inline]
blocksToInlines = B.toList . blocksToInlines'

-- | Inline elements used to separate blocks when squashing blocks into
-- inlines.
defaultBlocksSeparator :: Inlines
defaultBlocksSeparator =
  -- This is used in the pandoc.utils.blocks_to_inlines function. Docs
  -- there should be updated if this is changed.
  B.space <> B.str "¶" <> B.space

--
-- Safe read
--

safeRead :: (MonadPlus m, Read a) => T.Text -> m a
safeRead = safeStrRead . T.unpack

safeStrRead :: (MonadPlus m, Read a) => String -> m a
safeStrRead s = case reads s of
                  (d,x):_
                    | all isSpace x -> return d
                  _                 -> mzero
--
-- User data directory
--

-- | Return appropriate user data directory for platform.  We use
-- XDG_DATA_HOME (or its default value), but fall back to the
-- legacy user data directory ($HOME/.pandoc on *nix) if this is
-- missing.
defaultUserDataDirs :: IO [FilePath]
defaultUserDataDirs = E.catch (do
  xdgDir <- getXdgDirectory XdgData "pandoc"
  legacyDir <- getAppUserDataDirectory "pandoc"
  return $ ordNub [xdgDir, legacyDir])
 (\(_ :: E.SomeException) -> return [])