HTML reader: parse `<script type="math/tex` tags as math.

These are used by MathJax.

Closes #4877.

2 files changed, 25 insertions, 0 deletions

diff --git a/src/Text/Pandoc/Readers/HTML.hs b/src/Text/Pandoc/Readers/HTML.hs
index ea62bfcbf..b06e07a80 100644
--- a/src/Text/Pandoc/Readers/HTML.hs
+++ b/src/Text/Pandoc/Readers/HTML.hs
@@ -663,6 +663,7 @@ inline = choice
            , pCode
            , pSpan
            , pMath False
+           , pScriptMath
            , pRawHtmlInline
            ]
 
@@ -822,6 +823,17 @@ toStringAttr :: [(Text, Text)] -> [(String, String)]
 toStringAttr = map go
   where go (x,y) = (T.unpack x, T.unpack y)
 
+pScriptMath :: PandocMonad m => TagParser m Inlines
+pScriptMath = try $ do
+  TagOpen _ attr' <- pSatisfy $ tagOpen (=="script") (const True)
+  isdisplay <- case lookup "type" attr' of
+                    Just x | "math/tex" `T.isPrefixOf` x
+                      -> return $ "display" `T.isSuffixOf` x
+                    _ -> mzero
+  contents <- T.unpack . innerText <$>
+                manyTill pAnyTag (pSatisfy (matchTagClose "script"))
+  return $ (if isdisplay then B.displayMath else B.math) contents
+
 pMath :: PandocMonad m => Bool -> TagParser m Inlines
 pMath inCase = try $ do
   open@(TagOpen _ attr') <- pSatisfy $ tagOpen (=="math") (const True)
diff --git a/test/command/4877.md b/test/command/4877.md
new file mode 100644
index 000000000..070687345
--- /dev/null
+++ b/test/command/4877.md
@@ -0,0 +1,13 @@
+```
+% pandoc -f html -t native
+My <script type="math/tex">\mathcal{D}</script>
+^D
+[Plain [Str "My",Space,Math InlineMath "\\mathcal{D}"]]
+```
+
+```
+% pandoc -f html -t native
+<script type="math/tex; mode=display">\mathcal{D}</script>
+^D
+[Plain [Math DisplayMath "\\mathcal{D}"]]
+```