aboutsummaryrefslogtreecommitdiff
path: root/src/Text/Pandoc/Readers/HTML/Parsing.hs
blob: a8cdf1de2dcec74ad00fbafa92062570bcd624cf (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
{-# LANGUAGE LambdaCase            #-}
{-# LANGUAGE OverloadedStrings     #-}
{- |
   Module      : Text.Pandoc.Readers.HTML.Parsing
   Copyright   : Copyright (C) 2006-2021 John MacFarlane
   License     : GNU GPL, version 2 or above

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

Parsing functions and utilities.
-}
module Text.Pandoc.Readers.HTML.Parsing
  ( TagOmission (..)
  , pInTags
  , pInTags'
  , pInTag
  , pInTagWithAttribs
  , pAny
  , pCloses
  , pSatisfy
  , pBlank
  , matchTagClose
  , matchTagOpen
  , isSpace
  , maybeFromAttrib
  , toAttr
  , toStringAttr
  )
where

import Control.Monad (void, mzero, mplus)
import Data.Maybe (fromMaybe)
import Data.Text (Text)
import Text.HTML.TagSoup
  ( Attribute, Tag (..), isTagPosition, isTagOpen, isTagClose, (~==) )
import Text.Pandoc.Class.PandocMonad (PandocMonad (..))
import Text.Pandoc.Definition (Attr)
import Text.Pandoc.Parsing
  ( (<|>), eof, getPosition, lookAhead, manyTill, newPos, option, optional
  , skipMany, setPosition, token, try)
import Text.Pandoc.Readers.HTML.TagCategories
import Text.Pandoc.Readers.HTML.Types
import Text.Pandoc.Shared (tshow)
import Text.Pandoc.XML (html5Attributes, html4Attributes, rdfaAttributes)
import qualified Data.Set as Set
import qualified Data.Text as T

-- | Whether no tag, the closing tag, or both tags can be omitted.
data TagOmission
  = TagsRequired       -- ^ Opening and closing tags are both required
  | ClosingTagOptional -- ^ The closing tag can be omitted
  | TagsOmittable      -- ^ Both tags, opening and closing, can be omitted.
  deriving (Eq)

pInTags :: (PandocMonad m, Monoid a) => Text -> TagParser m a -> TagParser m a
pInTags tagtype parser = pInTags' tagtype (const True) parser

pInTags' :: (PandocMonad m, Monoid a)
         => Text
         -> (Tag Text -> Bool)
         -> TagParser m a
         -> TagParser m a
pInTags' tagtype tagtest parser = try $ do
  pSatisfy $ \t -> matchTagOpen tagtype [] t && tagtest t
  mconcat <$> manyTill parser (pCloses tagtype <|> eof)

pInTag :: PandocMonad m
       => TagOmission    -- ^ Whether some tags can be omitted
       -> Text           -- ^ @tagtype@ Tag name
       -> TagParser m a  -- ^ @p@ Content parser
       -> TagParser m a
pInTag tagOmission tagtype = fmap snd . pInTagWithAttribs tagOmission tagtype

-- | Returns the contents of a tag together with its attributes; parses
-- @p@, preceded by an opening tag (optional if TagsOmittable) and
-- followed by a closing tag (optional unless TagsRequired).
pInTagWithAttribs :: PandocMonad m
                  => TagOmission    -- ^ Whether some tags can be omitted
                  -> Text           -- ^ @tagtype@ Tag name
                  -> TagParser m a  -- ^ @p@ Content parser
                  -> TagParser m ([Attribute Text], a)
pInTagWithAttribs tagOmission tagtype p = try $ do
  let openingOptional = tagOmission == TagsOmittable
  let closingOptional = tagOmission /= TagsRequired
  skipMany pBlank
  attribs <- (if openingOptional then option [] else id)
             (getAttribs <$> pSatisfy (matchTagOpen tagtype []))
  skipMany pBlank
  x <- p
  skipMany pBlank
  (if closingOptional then optional else void) $
    pSatisfy (matchTagClose tagtype)
  skipMany pBlank
  return (attribs, x)
  where
    getAttribs = \case
      TagOpen _ attribs -> attribs
      _                 -> []

pCloses :: PandocMonad m => Text -> TagParser m ()
pCloses tagtype = try $ do
  t <- lookAhead $ pSatisfy $ \tag -> isTagClose tag || isTagOpen tag
  case t of
       (TagClose t') | t' == tagtype -> void pAny
       (TagOpen t' _) | t' `closes` tagtype -> return ()
       (TagClose "ul") | tagtype == "li" -> return ()
       (TagClose "ol") | tagtype == "li" -> return ()
       (TagClose "dl") | tagtype == "dd" -> return ()
       (TagClose "table") | tagtype == "td" -> return ()
       (TagClose "table") | tagtype == "th" -> return ()
       (TagClose "table") | tagtype == "tr" -> return ()
       (TagClose "td") | tagtype `Set.member` blockHtmlTags -> return ()
       (TagClose "th") | tagtype `Set.member` blockHtmlTags -> return ()
       (TagClose t') | tagtype == "p" && t' `Set.member` blockHtmlTags
                                            -> return () -- see #3794
       _ -> mzero

pBlank :: PandocMonad m => TagParser m ()
pBlank = void $ pSatisfy isBlank
 where
  isBlank (TagText t) = T.all isSpace t
  isBlank (TagComment _) = True
  isBlank _ = False

pLocation :: PandocMonad m => TagParser m ()
pLocation = do
  (TagPosition r c) <- pSat isTagPosition
  setPosition $ newPos "input" r c

pSat :: PandocMonad m => (Tag Text -> Bool) -> TagParser m (Tag Text)
pSat f = do
  pos <- getPosition
  token tshow (const pos) (\x -> if f x then Just x else Nothing)

pSatisfy :: PandocMonad m => (Tag Text -> Bool) -> TagParser m (Tag Text)
pSatisfy f = try $ optional pLocation >> pSat f

matchTagClose :: Text -> (Tag Text -> Bool)
matchTagClose t = (~== TagClose t)

matchTagOpen :: Text -> [(Text, Text)] -> (Tag Text -> Bool)
matchTagOpen t as = (~== TagOpen t as)

pAny :: PandocMonad m => TagParser m (Tag Text)
pAny = pSatisfy (const True)

isSpace :: Char -> Bool
isSpace ' '  = True
isSpace '\t' = True
isSpace '\n' = True
isSpace '\r' = True
isSpace _    = False

-- taken from HXT and extended
-- See http://www.w3.org/TR/html5/syntax.html sec 8.1.2.4 optional tags
closes :: Text -> Text -> Bool
_ `closes` "body" = False
_ `closes` "html" = False
"body" `closes` "head" = True
"a" `closes` "a" = True
"li" `closes` "li" = True
"th" `closes` t | t `elem` ["th","td"] = True
"td" `closes` t | t `elem` ["th","td"] = True
"tr" `closes` t | t `elem` ["th","td","tr","colgroup"] = True
"dd" `closes` t | t `elem` ["dt", "dd"] = True
"dt" `closes` t | t `elem` ["dt","dd"] = True
"rt" `closes` t | t `elem` ["rb", "rt", "rtc"] = True
"col" `closes` "col" = True
"colgroup" `closes` "col" = True
"optgroup" `closes` "optgroup" = True
"optgroup" `closes` "option" = True
"option" `closes` "option" = True
-- https://html.spec.whatwg.org/multipage/syntax.html#optional-tags
x `closes` "p" | x `elem` ["address", "article", "aside", "blockquote",
   "dir", "div", "dl", "fieldset", "footer", "form", "h1", "h2", "h3", "h4",
   "h5", "h6", "header", "hr", "main", "menu", "nav", "ol", "p", "pre", "section",
   "table", "ul"] = True
_ `closes` "meta" = True
"form" `closes` "form" = True
"label" `closes` "label" = True
"map" `closes` "map" = True
"object" `closes` "object" = True
_ `closes` t | t `elem` ["option","style","script","textarea","title"] = True
t `closes` "select" | t /= "option" = True
"thead" `closes` "colgroup" = True
"tfoot" `closes` t | t `elem` ["thead","colgroup"] = True
"tbody" `closes` t | t `elem` ["tbody","tfoot","thead","colgroup"] = True
t `closes` t2 |
   t `elem` ["h1","h2","h3","h4","h5","h6","dl","ol","ul","table","div","main","p"] &&
   t2 `elem` ["h1","h2","h3","h4","h5","h6","p" ] = True -- not "div" or "main"
t1 `closes` t2 |
   t1 `Set.member` blockTags &&
   t2 `Set.notMember` blockTags &&
   t2 `Set.notMember` eitherBlockOrInline = True
_ `closes` _ = False

toStringAttr :: [(Text, Text)] -> [(Text, Text)]
toStringAttr = foldr go []
  where
   go :: (Text, Text) -> [(Text, Text)] -> [(Text, Text)]
   -- treat xml:lang as lang
   go ("xml:lang",y) ats = go ("lang",y) ats
   -- prevent duplicate attributes
   go (x,y) ats
     | any (\(x',_) -> x == x') ats = ats
     | otherwise      =
        case T.stripPrefix "data-" x of
          Just x' | x' `Set.notMember` (html5Attributes <>
                                        html4Attributes <> rdfaAttributes)
            -> go (x',y) ats
          _ -> (x,y):ats

-- Unlike fromAttrib from tagsoup, this distinguishes
-- between a missing attribute and an attribute with empty content.
maybeFromAttrib :: Text -> Tag Text -> Maybe Text
maybeFromAttrib name (TagOpen _ attrs) = lookup name attrs
maybeFromAttrib _ _ = Nothing

mkAttr :: [(Text, Text)] -> Attr
mkAttr attr = (attribsId, attribsClasses, attribsKV)
  where attribsId = fromMaybe "" $ lookup "id" attr `mplus` lookup "name" attr
        attribsClasses = T.words (fromMaybe "" $ lookup "class" attr) <> epubTypes
        attribsKV = filter (\(k,_) -> k /= "class" && k /= "id" && k /= "name")
                           attr
        epubTypes = T.words $ fromMaybe "" $ lookup "epub:type" attr

toAttr :: [(Text, Text)] -> Attr
toAttr = mkAttr . toStringAttr