PR: Added support for sphinx_math_dollar in help text generation

[oscargus]

Description of Changes

Added support for the sphinx_math_dollar plugin that e.g. SymPy uses.

https://www.sympy.org/sphinx-math-dollar/

The idea is that math equations can be marked using the LaTeX-style $...$ or $$...$$ instead of the more tedious Sphinz marking.

• Wrote at least one-line docstrings (for any new functions)
• Added unit test(s) covering the changes (if testable)

• Included a screenshot or animation (if affecting the UI, see Licecap)

With:

Without:

This checks if sphinx_math_dollar exists and if so adds it to the sphinx configuration (assuming that MathJax is used).

Not clear how to add a test for it. It seems like there are no tests for math rendering in the help plugin at all.

If it makes sense, I can add an additional check box for it in the preferences. But I think that if someone has installed sphinx_math_dollar (it is not a mandatory dependency of SymPy), they probably would like to use it by default.

Issue(s) Resolved

Affirmation

By submitting this Pull Request or typing my (user)name below,
I affirm the Developer Certificate of Origin
with respect to all commits and content included in this PR,
and understand I am releasing the same under Spyder's MIT (Expat) license.

I certify the above statement is true and correct:

oscargus, Oscar Gustafsson

[pep8speaks]

Hello @oscargus! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

• In the file installers/Windows/installer.py:

Line 234:80: E501 line too long (86 > 79 characters)

Comment last updated at 2021-12-19 13:22:18 UTC

[ccordoba12]

Hey @oscargus, thanks for your contribution! I think this is a very nice addition.

I only have one question for you: should we add this extension as a new Spyder dependency? Otherwise, it'd be hard for people to discover this functionality.

What do you think?

[oscargus]

I was also considering adding it as a dependency. It is a quite small package, so from that perspective it may not be a big deal. On the other hand, it seems to primarily be a benefit for SymPy (right now, not sure how many other packages use it, yet), so that is why I didn't add it. Also, it will not break anything, just looks a bit worse compared to what it can.

But I'd be happy to add it. I definitely see the benefit of it as many people will probably not realize that there is a way to get rid of those $.

(I primarily contribute to SymPy, where there is a single dependency. So I guess I may be a bit biased regarding adding dependencies...)

[oscargus]

Finally got around to updating this. Took a bit of time as sphinx-math-dollar does not have a wheel and although the solution wasn't that tricky in the end it wasn't straightforward for me...

I'd say that the logic is a bit more correct now since earlier on packages in SKIP_PACKAGES was not correctly handled as modules (as the set subtraction was on strings with version numbers). Although it doesn't really makes any difference, it should be better now.

Also, I added a feature so that all packages with - in them are replaced with _ when the modules are asked for. This is probably the best guess one can make.

The MacOS-app failure seems unrelated and maybe I should have merged against 5.x instead?