diff options
| author | John MacFarlane <[email protected]> | 2025-03-10 14:41:46 -0700 |
|---|---|---|
| committer | John MacFarlane <[email protected]> | 2025-03-10 14:41:46 -0700 |
| commit | 7fe8c928212197354196724e9b95b71a854fdca9 (patch) | |
| tree | 4cf7da86a75a84a8dcefa5ba59c7bdb0bc828c60 /src | |
| parent | d73475cec8c7363055c482f5107a2759e1f6a277 (diff) | |
HTML reader: ignore style tags in the body.
They are invalid but do occur in the wild.
Closes #10643.
Diffstat (limited to 'src')
| -rw-r--r-- | src/Text/Pandoc/Readers/HTML.hs | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/src/Text/Pandoc/Readers/HTML.hs b/src/Text/Pandoc/Readers/HTML.hs index c21ab67aa..fcc1147ca 100644 --- a/src/Text/Pandoc/Readers/HTML.hs +++ b/src/Text/Pandoc/Readers/HTML.hs @@ -242,7 +242,7 @@ block = ((do "main" -> pDiv "figure" -> pFigure "iframe" -> pIframe - "style" -> pRawHtmlBlock + "style" -> mempty <$ pHtmlBlock "style" -- see #10643 "textarea" -> pRawHtmlBlock "switch" | epubExts @@ -537,7 +537,7 @@ pIframe = try $ do pRawHtmlBlock :: PandocMonad m => TagParser m Blocks pRawHtmlBlock = do - raw <- pHtmlBlock "script" <|> pHtmlBlock "style" <|> pHtmlBlock "textarea" + raw <- pHtmlBlock "script" <|> pHtmlBlock "textarea" <|> pRawTag exts <- getOption readerExtensions if extensionEnabled Ext_raw_html exts && not (T.null raw) @@ -716,6 +716,7 @@ inline = pTagText <|> do "input" | lookup "type" attr == Just "checkbox" -> asks inListItem >>= guard >> pCheckbox + "style" -> mempty <$ pHtmlBlock "style" -- see #10643 "script" | Just x <- lookup "type" attr , "math/tex" `T.isPrefixOf` x -> pScriptMath @@ -1080,6 +1081,8 @@ isInlineTag :: Tag Text -> Bool isInlineTag t = isCommentTag t || case t of TagOpen "script" _ -> "math/tex" `T.isPrefixOf` fromAttrib "type" t TagClose "script" -> True + TagOpen "style" _ -> True -- see #10643, invalid but it happens + TagClose "style" -> True TagOpen name _ -> isInlineTagName name TagClose name -> isInlineTagName name _ -> False |