From a0c1bdeb8b93e2661132a714668dff5f32ffb167 Mon Sep 17 00:00:00 2001
From: John MacFarlane <jgm@berkeley.edu>
Date: Wed, 10 Apr 2024 17:18:05 -0700
Subject: [PATCH] Markdown reader: auto-close unclosed divs.

This applies to both fenced and HTML-ish varieties.
Otherwise we face an exponential performance problem with
backtracking.

This also accords with the behavior of the `fenced_divs`
extension in commonmark.

A warning is issued when a div is implicitly closed.

Closes #9635.
---
 src/Text/Pandoc/Readers/Markdown.hs | 26 +++++++++++++-------------
 test/Tests/Readers/Markdown.hs      |  2 +-
 test/command/9635.md                | 16 ++++++++++++++++
 3 files changed, 30 insertions(+), 14 deletions(-)
 create mode 100644 test/command/9635.md

diff --git a/src/Text/Pandoc/Readers/Markdown.hs b/src/Text/Pandoc/Readers/Markdown.hs
index 206e5bc3cc12..a78b1062ad3c 100644
--- a/src/Text/Pandoc/Readers/Markdown.hs
+++ b/src/Text/Pandoc/Readers/Markdown.hs
@@ -2102,30 +2102,29 @@ divHtml :: PandocMonad m => MarkdownParser m (F Blocks)
 divHtml = do
   guardEnabled Ext_native_divs
   try $ do
-    (TagOpen _ attrs, rawtag) <- htmlTag (~== TagOpen ("div" :: Text) [])
+    openpos <- getPosition
+    (TagOpen _ attrs, _) <- htmlTag (~== TagOpen ("div" :: Text) [])
     -- we set stateInHtmlBlock so that closing tags that can be either block
     -- or inline will not be parsed as inline tags
     oldInHtmlBlock <- stateInHtmlBlock <$> getState
     updateState $ \st -> st{ stateInHtmlBlock = Just "div" }
-    bls <- option "" (blankline >> option "" blanklines)
+    optional blanklines
     contents <- mconcat <$>
                 many (notFollowedBy' (htmlTag (~== TagClose ("div" :: Text)))
                       >> block)
-    closed <- option False (True <$ htmlTag (~== TagClose ("div" :: Text)))
-    if closed
-       then do
-         updateState $ \st -> st{ stateInHtmlBlock = oldInHtmlBlock }
-         let ident = fromMaybe "" $ lookup "id" attrs
-         let classes = maybe [] T.words $ lookup "class" attrs
-         let keyvals = [(k,v) | (k,v) <- attrs, k /= "id" && k /= "class"]
-         return $ B.divWith (ident, classes, keyvals) <$> contents
-       else -- avoid backtracing
-         return $ return (B.rawBlock "html" (rawtag <> bls)) <> contents
+    void (htmlTag (~== TagClose ("div" :: Text))) <|>
+       (getPosition >>= report . UnclosedDiv openpos)
+    let ident = fromMaybe "" $ lookup "id" attrs
+    let classes = maybe [] T.words $ lookup "class" attrs
+    let keyvals = [(k,v) | (k,v) <- attrs, k /= "id" && k /= "class"]
+    updateState $ \st -> st{ stateInHtmlBlock = oldInHtmlBlock }
+    return $ B.divWith (ident, classes, keyvals) <$> contents
 
 divFenced :: PandocMonad m => MarkdownParser m (F Blocks)
 divFenced = do
   guardEnabled Ext_fenced_divs
   try $ do
+    openpos <- getPosition
     string ":::"
     skipMany (char ':')
     skipMany spaceChar
@@ -2135,7 +2134,8 @@ divFenced = do
     blankline
     updateState $ \st ->
       st{ stateFencedDivLevel = stateFencedDivLevel st + 1 }
-    bs <- mconcat <$> manyTill block divFenceEnd
+    bs <- mconcat <$> many (notFollowedBy divFenceEnd >> block)
+    divFenceEnd <|> (getPosition >>= report . UnclosedDiv openpos)
     updateState $ \st ->
       st{ stateFencedDivLevel = stateFencedDivLevel st - 1 }
     return $ B.divWith attribs <$> bs
diff --git a/test/Tests/Readers/Markdown.hs b/test/Tests/Readers/Markdown.hs
index 001ca02077c7..5f67e111b416 100644
--- a/test/Tests/Readers/Markdown.hs
+++ b/test/Tests/Readers/Markdown.hs
@@ -449,7 +449,7 @@ tests = [ testGroup "inline code"
                   <>
                   codeBlockWith ("",["haskell"],[]) "b"
                   <>
-                  rawBlock "html" "<div>\n\n"
+                  divWith ("",[],[]) mempty
           ]
 -- the round-trip properties frequently fail
 --        , testGroup "round trip"
diff --git a/test/command/9635.md b/test/command/9635.md
new file mode 100644
index 000000000000..088a5b6d294e
--- /dev/null
+++ b/test/command/9635.md
@@ -0,0 +1,16 @@
+```
+% pandoc
+> ::: {.fence}
+> that is
+> not closed
+
+okay
+^D
+2> [WARNING] Div at _chunk line 1 column 1 unclosed at _chunk line 5 column 1, closing implicitly.
+<blockquote>
+<div class="fence">
+<p>that is not closed</p>
+</div>
+</blockquote>
+<p>okay</p>
+```