Fix issue with handling </>#4007
Conversation
Markdown recently made a fix for an issue handling `</>` which would get stripped out of content. Reference: Python-Markdown/markdown@64a3c0f MkDocs uses Markdown's monkey-patched HTMLParser, but does not use the HTML extractor that has the logic to handle the things that have been monkey-patched in. Because of this, MkDocs now fails when parsing `</>`. This fixes the issue by adding the same logic to handle the issue that Markdown uses.
|
This fixes #4001 |
|
@tomchristie is this project still being maintained? I'd love to see this bug fix get merged and released. |
|
+1, this broke our team's internal docs build. can this get merged? |
|
It seems I am unable to push this forward. I am willing to provide changes and updates if the Mkdocs team is willing to engage. If not, I guess we are not getting this fix. |
|
I guess we should ping @lovelydinosaur |
|
We also need this fix 🙏🏼 |
|
This bug is breaking our documentation. MkDocs team, would it be possible to merge this PR and release 6.1.2? Much appreciated! |
|
I had to fix some other HTML issues in Python Markdown. Because it broke MkDocs for Python Markdown, I had to resort to hackery. The good news is that the next Python Markdown, will fix this issue for MkDocs. PR is here: Python-Markdown/markdown#1593. |
|
@facelessuser Thanks a lot! I just updated markdown to |
|
Yep, it was one of those things; it hadn't occurred to me that it could be fixed in Python Markdown. I'd argue it shouldn't have had to be fixed there, but necessity is the mother of invention. When Python Markdown, which uses MkDocs as its documentation engine, was now breaking due to this, a solution became necessary. |
themavik
left a comment
themavik
left a comment
There was a problem hiding this comment.
Overriding parse_starttag to push </> through handle_data stops AnchorHtmlValidator from treating that JSX fragment as a real closing tag. nit: this couples you to CPython HTMLParser internals—if the stdlib parser changes, this hook is the first place to re-test.
|
This isn't needed anymore, as a fix that works here is available in the latest Python Markdown directly. |
7 participants
Markdown recently made a fix for an issue handling
</>which would get stripped out of content. Reference:Python-Markdown/markdown@64a3c0f
MkDocs uses Markdown's monkey-patched HTMLParser, but does not use the HTML extractor that has the logic to handle the things that have been monkey-patched in. Because of this, MkDocs now fails when parsing
</>. This fixes the issue by adding the same logic to handle the issue that Markdown uses.Fixes #4001