chore: remove pyrfc3339 and change to datetime.datetime.fromisoformat…

[EdmilsonRodrigues]

…() and datetime.datetime.isoformat()

Description

I tried using backports.datetime_fromisoformat but than the tox tests broke, probably because it required an import in the unit tests. I tried than just using the usual builtin datetime.datetime.fromisoformat and it worked just well.

<Fixes: >

QA Steps

<Commands / tests / steps to run to verify that the change works:>

tox -e py3 -- tests/unit/...

tox -e integration -- tests/integration/...

All CI tests need to pass.

<Please note that most likely an additional test will be required by the reviewers for any change that's not a one liner to land.>

Notes & Discussion

*This is a secondary branch from the Pull Request #1243 *

[james-garner-canonical]

I tried using backports.datetime_fromisoformat but than the tox tests broke, probably because it required an import in the unit tests

Yeah unfortunately we currently need to specify the dependencies in both pyproject.toml and tox.ini we can add backports-datetime-fromisoformat -- you can replace the pyRFC339 dependency in both

I tried than just using the usual builtin datetime.datetime.fromisoformat and it worked just well

I'm guessing you're using python 3.11+ locally -- datetime.fromisoformat was extended in 3.11 to properly support the format rather than just the subset output by datetime. We need to support python 3.8+, and as you can (now) see from the CI run, unit tests for the cookie handling fail on python > 3.11. Hence the need for the backports dependency

BTW as an external contributor your PRs against this repo don't have our CI run automatically, but I believe that you should be able to open a PR against your own fork's main first and have the tests run there -- this will take care of running linting and running tests against multiple python and juju versions

[james-garner-canonical]

Btw @EdmilsonRodrigues we most likely won't be looking at PRs over the next couple of weeks due to the holiday season but happy to continue after that -- thanks for your contributions so far

[dimaqq]

I'm sorry about our CI:

   input: chore: remove pyrfc3339 and change to datetime.datetime.fromisoformat() and datetime.datetime.isoformat()
✖   header must not be longer than 100 characters, current length is 105 [header-max-length]

Please amend the commit message 🙇🏻

[EdmilsonRodrigues]

I tried to open a PR to my own main branch but no test was created. I think I need to configure the github actions. I'll learn how to do it and do it for the next time.

Also merry holidays for you all!!!

[dimaqq]

Please rebase 🙇🏻

[james-garner-canonical]

Thanks for this, @EdmilsonRodrigues. If you can rebase from main and fix your commit messages (or squash them) to use conventional commits (using commit types from the example at that page), then I think we can go ahead and merge this

[EdmilsonRodrigues]

I'm really sorry if I'm not doing this right. I'm still learning how to do rebasing and working in a project like this.

I think it worked now.

[EdmilsonRodrigues]

Thank you for all the guidance during this process.
I think it's done now.
But, for later moments, how can I see that view that you just showed me to see if I'm doing it right?

[dimaqq]

The command I used was git log --all --decorate --oneline --graph.

rebase: is not a conventional commit type, pls mark it as chore or something.

[dimaqq]

input: refacor: change monkeypatch to use local imports
✖ type must be one of [build, chore, ci, docs, feat, fix, perf, refactor, revert, style, test] [type-enum]

[dimaqq]

A simple fix would be to squash all your commits into one and update the commit message.

got rebase -i main and follow the editor prompt

[EdmilsonRodrigues]

When I used this, it let me rename all my commits, because I still haven't learned how to squash, but I'll study more git this month so the problems I faced with this PR won't happen again in the next one, sorry again.

[dimaqq]

No worries, you'll get a hang of it as you contribute to open source more.

[dimaqq]

⧗   input: chore: remove pyrfc3339 and change to datetime.datetime.fromisoformat() and datetime.datetime.isoformat()
✖   header must not be longer than 100 characters, current length is 105 [header-max-length]

I'm sorry for silly rules that this repo imposes. I'm helping maintain it, but I didn't start it or invent these rules.

[dimaqq]

I think when you've rebased last, your main was not up to date with upstream (juju) main.

* 3aa4221 (edmilson/backports) chore: remove deps backports-datetime-fromisoformat
* 59e3cd4 chore: rebase with main
* 62fcdb2 refactor: change monkeypatch to use local imports
* 6814f06 chore: remove pyrfc3339 and change to datetime.datetime.fromisoformat() and datetime.datetime.isoformat()
* e5d384d refactor: change monkeypatch to use local imports
* 172590a chore: change setup.py to use backports instead of pyrfc
* 1d9d6ba chore: added backports to tox.ini and pyproject.toml
* d1abfa8 chore: remove pyrfc3339 and change to datetime.datetime.fromisoformat() and datetime.datetime.isoformat()
| *   1907f7a (upstream/main, origin/main, main) Merge pull request #1250 from dimaqq/upstream-type-fixes
| |\
| |/
|/|
* | e7bcdc9 (edmilson/main, upstream-type-fixes) chore: don't allow 0.25.2 either, still broken
* | 2423ef7 chore: block broken(?) pytest-asyncio version
* | 35ba68b chore: type hint improvements from the helper thread branch
|/
*   a58645e Merge pull request #1244 from tmihoc/add-discourse-docs

[dimaqq]

Or maybe you've made the mistake of merging upstream main into your main or cherry-picked specific commits.

Typically what I would do in this case is git checkout main && git reset --hard upstream/main.

Or you can rebase your branch on top of upstream/main directly. Up to you.

[EdmilsonRodrigues]

Thank you for your pacience and guidance. I am really thankful for that.
I think I learned how to squash, Let's see if ti works now, but I rebase, and I think it worked this time.

[dimaqq]

Is the output exactly same?

[EdmilsonRodrigues]

Not exactly, these are the outputs:
pyrfc3339: 2025-01-30T14:49:41Z
datetime.isoformat(): 2025-01-30T14:49:41.562687

But this was already discussed, I think in the other PR about the same issue, but as both are different formats that are part of the RFC3339 specification, we decided to proceed.

But if it's not possible, I can find another solution. There was also that manual parse I did to make them equal.

[dimaqq]

It's just something I need to be clear about and run by juju (server) folks.

@SimonRichardson wdyt? or nominate someone, pls.

[dimaqq]

/build