Allow passing a base branch that doesn't have version info

[Jackenmen]

Resolves #69

I could add enforcement of something here if this is too lax but I'm not sure what kind of check that could be. I suppose I could also add an option to .cherry_picker.toml that needs to be explicitly enabled to allow passing non-version branches if that makes it any better.

I'm aware this issue wasn't accepted but I kinda felt like doing it anyway, even if it doesn't end up merged :P

[Mariatta]

I think this change is reasonable.

[Jackenmen]

Tests are currently failing on main and so this PR is affected, I fixed them in #76.

[Mariatta]

I think allowing it to be configured via the toml file is a good idea as well. Would you be able to add that in a separate PR?

[Jackenmen]

I can do it here or in a separate PR, whatever you prefer. Can probably get to it tomorrow.

I assume you would want this to be a true/false setting determining whether branches without version info are allowed (defaulting to false to keep the current behavior intact)?

[Mariatta]

I think it should support the original behavior if not supplied.

[Jackenmen]

Alright, I added a require_version_in_branch_name option. Quite a mouthful but I couldn't really come up with anything else that was still clear about what the option is. At most I could maybe manage to make the name 4-5 characters shorter but the names I came up with were, in my opinion, a lot more vague than the current name.

---

While looking into this, I also realized that I suggested adding such an option in the PR's description which I didn't really realize when you asked me to add it 😄

[Jackenmen]

I created python/miss-islington#639 to make miss-islington work with this PR since I realized that this is going to be necessary now that this also adds a new configuration option.

[Jackenmen]

I addressed the review and resolved the merge conflicts. Seems like I may have missed the merge window for the release though 😄

If by the time this is looked at by a maintainer new conflicts arise, just let me know and I'll resolve them - maybe I'll finally be able to use upstream directly :)

[webknjaz]

@hugovk mind taking a look?

[hugovk]

@Mariatta @ezio-melotti Any further comments?

[gopidesupavan]

Thank you @Jackenmen for creating this; it's really helpful. In Airflow, we're working on automating backporting. We attempted to use cherry-picker, but due to version requirements in the base branch, it isn't feasible. This is because our Apache Airflow project uses slightly different branch naming conventions (ex: v2-1-test, v3-1-stable etc;) .

@hugovk Is there a plan to release this in the next cycle? If so, could you please share when that might be?

[Mariatta]

Thank you

[Jackenmen]

I'm glad to learn I'm not the only person that finds this feature useful 😄 Have you tried installing from this PR to see that it definitely works for your use case? Might be worth doing considering that cherry-picker doesn't have a very regular release schedule.

[gopidesupavan]

I'm glad to learn I'm not the only person that finds this feature useful 😄 Have you tried installing from this PR to see that it definitely works for your use case? Might be worth doing considering that cherry-picker doesn't have a very regular release schedule.

Yes i have tested it is working :) . cool

[potiuk]

Indeed you are not the only one @Jackenmen it would be great to merge this one, we really would love to use it in Airflow. @uranusjr - one of the PyPI maintainers and our fellow PMC member recommended it to us. Also we checked a number of other solutions and cherry-picker only lacks this particular feature and @gopidesupavan had already tested that it would work for for Apache Airflow. And we are now considering forking the project rather than installing from PyPI. So it would be great to get this one merged and released.

@hugovk @ezio-melotti @ambv - Is there anything we can help with to make it happen :) ? Or maybe there are some concerns with merging that one ?

[hugovk]

I'll merge and release later today :)

[potiuk]

🙇 🙇 🙇 🙇

[webknjaz]

@potiuk I can also offer using my Patchback bot that can work as an automatic companion to this semi-annual CLI tool FYI.
Had I known earlier, I would've shown it to you on Saturday when we were in the same room...

Examples: pytest-dev/pytest#12912 (comment) / aio-libs/aiohttp#9852 (comment).

Feel free to ping me in a separate issue or on slack if interested, so we don't cause unnecessary notifications here..

[potiuk]

@webknjaz our world is small :) . I think we have certain limitations int the Apache Software Foundation and 3rd-party bots that can update content are generally not approved by the infrastructure. Of course details matters and maybe @gopidesupavan can take a look at it :)

[webknjaz]

@potiuk yeah, I figured you might want to run own instance, which is entirely possible 🤷‍♂️

[potiuk]

@potiuk yeah, I figured you might want to run own instance, which is entirely possible 🤷‍♂️

Yep. We might do just that.

[gopidesupavan]

@webknjaz our world is small :) . I think we have certain limitations int the Apache Software Foundation and 3rd-party bots that can update content are generally not approved by the infrastructure. Of course details matters and maybe @gopidesupavan can take a look at it :)

Yeah i will have a look.

[hugovk]

Released as 2.4.0 🍒🚀

[potiuk]

Released as 2.4.0 🍒🚀

Fantastic! Many thanks!

[gopidesupavan]

Released as 2.4.0 🍒🚀

Woohoo thank you. Appreciate your help and support 🙂