MAINT: Don't set the default language

[michaelosthege]

For Sphinx 5 compatibility.

See release notes and sphinx-doc/sphinx#10474.

Related #1749 and executablebooks/meta#750

---

I know that the CI pipeline isn't running with Sphinx 5 yet, but I tested this locally by forcing a pip install "sphinx==5.0.1" and making this one-line change in the jupyter-book source.
With that my dissertation project built successfully.

[welcome]

Thanks for submitting your first pull request! You are awesome! 🤗

If you haven't done so already, check out EBP's Code of Conduct and our Contributing Guide, as this will greatly help the review process.

Welcome to the EBP community! 🎉

[michaelosthege]

Hi @choldgraf, this is my first PR and I'll need an approval for the CI tests to run..

[michaelosthege]

So the test complained and since Sphinx 5 won't allow for language=None|null I saw essentially two options:

1. Change the tests to expect the language to NOT be set
2. Align the language setting with the Sphinx 5 default ("en")

I situations like this I always prefer not to override one default with another (risking to get them out of sync).

That's why I went with option 1.) here.

[michaelosthege]

ping @chrisjsewell

[codecov]

Codecov Report

Merging #1750 (704d75a) into master (d33a198) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1750   +/-   ##
=======================================
  Coverage   91.37%   91.37%           
=======================================
  Files           7        7           
  Lines         684      684           
=======================================
  Hits          625      625           
  Misses         59       59           

Flag	Coverage Δ
pytests	91.37% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
jupyter_book/config.py	94.21% <ø> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d33a198...704d75a. Read the comment docs.

[choldgraf]

Hmm - wouldn't this result in problems on Sphinx 4 though? I agree we should do this for Sphinx 5 versions. If it wouldn't cause issues on sphinx 4 then I'm +1 in general, let's see what the tests do

[michaelosthege]

The tests look good, but you're the expert.

[michaelosthege]

@choldgraf @chrisjsewell is there anything left I should do to get this over the finish line?

[choldgraf]

I can't think of a reason why this will be a problem, and don't see any documentation for why this needs to be None in the first place, so I'm +1 on it unless anybody else objects.

[pradyunsg]

Thanks for filing this PR @michaelosthege! ^>^

[welcome]

Congrats on your first merged pull request in this project! 🎉


Thank you for contributing, we are very proud of you! ❤️

[choldgraf]

Let's give it a shot!

[choldgraf]

BTW - many thanks @michaelosthege for tracking this and making the fix!

[michaelosthege]

Thank you for accepting my first contribution :)
This way the CI pipeline on my second PR ran on its own. One job failed, but it looks just like a flaky codecov upload.