Fix parsing of defer CLI command --at and --in options

[stinovlas]

Closes #997

I fixed parsing of --at and --in CLI options of defer command and added two integration tests for those.

Successful PR Checklist:

• [x] Tests
  • yes, I added two integration tests covering defer --at and --in options
• [x] Documentation not applicable – just a bugfix

PR label(s):

• [x] https://github.com/procrastinate-org/procrastinate/labels/PR%20type%3A%20bugfix%20%F0%9F%95%B5%EF%B8%8F

[stinovlas]

Congratulations for your first contribution, welcome to the joyful bunch of procrastinators :)

Thanks :⁠-⁠). There are still some tests failing in the CI, I'll fix that tomorrow.

[stinovlas]

Congratulations for your first contribution, welcome to the joyful bunch of procrastinators :)

Thanks :⁠-⁠). There are still some tests failing in the CI, I'll fix that tomorrow.

Fixed! (I hope, pytest passes on my machine now, but PR CI requires approval)

[ewjoachim]

Seems all good. Don't forget to squash your fixups :)

[github-actions[bot]]

Coverage report

This PR does not seem to contain any modification to coverable code.

[stinovlas]

Seems all good. Don't forget to squash your fixups :)

Sure, I wasn't sure what's the convention here (whether you squash it during merge or if it's the PR author responsibility). Squashed :-).

[ewjoachim]

I wasn't sure what's the convention here

It depends. When contributors put some effort on commit names, I try to keep it in the history. When the commit log looks like this, I squash.

[stinovlas]

I wasn't sure what's the convention here

It depends. When contributors put some effort on commit names, I try to keep it in the history. When the commit log looks like this, I squash.

Makes sense :-).

This is weird. All I did was squash the fixup with the previous commit, which produced the exactly same code, but now the CI is failing in a part of the code that shouldn't have anything in common with the changes. And it passed in py3.8 and py3.10, but failed in py3.11. It seems to be some kind of race-condition random error? It doesn't pass for me on main either. But I'm not yet familiar with the procrastinate codebase enough to be sure.

[ewjoachim]

Yup, seems like a test decided to become flaky all of a sudden. I've rerun the CI, if there's a fix to do, we'll do it in another PR.