James Addison

> > > I'm struggling a bit to see the problem: aside from deserialization, I can't imagine for what reason we'd be reading index entries from Python code? I skimmed...

NodeJS tests on this branch began failing from 9fef43ba3ed3c5cb09a264994b55dc1d124e7810 onwards; that isn't easily visible from the CI results in the GitHub web UI because we weren't running those on each...

Good news, I think. > NodeJS tests on this branch began failing from [9fef43b](https://github.com/sphinx-doc/sphinx/commit/9fef43ba3ed3c5cb09a264994b55dc1d124e7810) onwards; that isn't easily visible from the CI results in the GitHub web UI because we...

This is ready, yep; the `mypy` lint failure on the latest commit (64387c4ae9a10cbd6851a4a693b21b00d369511b) is unrelated (#12510), and my last two messages where I'm talking to myself (not uncommon, sadly) are...

Well done completing the defence @picnixz :) Thank you!

Thank you @picnixz!

Closing this to reduce the number of pull requests open (both for Sphinx, and also in my own backlog). I'll revisit this next year.

It looks like the relevant code here originates from `docutils` ([ref](https://github.com/docutils/docutils/blob/b768e2626088711dec257b0847b563d02700a712/docutils/docutils/writers/_html_base.py#L1516-L1535)), and in fact the assertion has been removed recently (within the past month). If we have a way to...

Ok, hm. This was found using TypeScript 5.2.2 and running a unit test using `mocha`. Trying to remove `mocha` from the equation seems to help: the following code runs with...

Ah, blast. That output was produced using `fraction.js` v4.2.1 - potentially unaffected, as per the issue description. I'll check again with v4.3.5 in a few moments.