Don't store a RichHandler object in the argparse namespace (#3394)

[cpitclaudel]

Resolves: #3394

• Ensured tests pass and (if applicable) updated functional test output
• Conformed to code style guidelines by running appropriate linting tools
• Added tests for changed code
• Updated documentation for changed code

Unfortunately I can't test locally, but this patch should be harmless, at least.

[snh]

I too was getting the error described in #3394 and can confirm this patch fixes the issue for me.

[justinmayer]

Thanks, Clément! Do you think it's worth adding a test that fails for MacOS images in CI when this change is not present?

[cpitclaudel]

A regression test would be nice, yes. I suppose the best way to go about it would be to start the file watcher and then stop it a few seconds later somehow, asserting that it didn't crash in the meantime? I don't know when I'll have the time to do that.

[justinmayer]

Many thanks for the fix, Clément. I'm going to merge this now so that we can release a fix for this and thereby eliminate the problem for end users as soon as possible. If you think of one or more unit tests that might prevent future regressions, submitting them in a follow-up pull request would be most appreciated. Thanks again!