nf-core download: Ability to add custom tags to revisions in downloaded pipelines

[MatthiasZepper]

This PR introduces a -a / --additional-tag argument to nf-core download --platform. It must be followed by a string in a "key=value" format and can be provided multiple times. The left-hand side must refer to a valid branch, tag or commit SHA. The right-hand side must comply with the naming conventions for Git tags and may not yet exist in the repository.

The new -a / --additional-tag argument allows customizing the downloaded pipeline with additional tags that can be used to select particular revisions in the Seqera Platform interface.

For example, an accredited facility may opt to tag particular revisions according to their structured release management process: -a "3.12.0=testing" -a "3.9.0=validated" so their staff can easily ensure that the correct version of the pipeline is run in production.

I believe this feature will be useful for facilities and those who are setting up pipelines for others to run.

Additionaly, this PR also bumps the dev version from 2.13.2dev to 2.14dev to comply with SemVer. Mind that other PRs already merged on dev, e.g. #2853 by @ewels should have triggered this before because it deprecated the --tower argument:

7. Minor version Y (x.Y.z | x > 0) MUST be incremented if new, backward compatible functionality is introduced to the public API. It MUST be incremented if any public API functionality is marked as deprecated.

PR checklist

• This comment contains a description of changes (with reason)
• CHANGELOG.md is updated
• If you've fixed a bug or added code that should be tested, add tests!
• Documentation in docs is updated

[MatthiasZepper]

Interesting...adding tags depends on a successfully configured git user:

ERROR    nf_core.download:download.py:1741 [red]Additional tag(s) could not be applied:[/]
Cmd('git') failed due to: exit code(128)
  cmdline: git tag -m Synonynmous tag to 3.7; added by 'nf-core download'. a.tad.outdated 3.7
  stderr: 'Committer identity unknown

*** Please tell me who you are.

Run

  git config --global user.email "you@example.com"
  git config --global user.name "Your Name"

to set your account's default identity.
Omit --global to set the identity only in this repository.

fatal: unable to auto-detect email address (got 'runner@fv-az566-7.(none)')'

nf-core download already deletes superfluous tags...which apparently is no problem.

[codecov]

Codecov Report

Attention: Patch coverage is 85.10638% with 7 lines in your changes are missing coverage. Please review.

Project coverage is 75.42%. Comparing base (98abe32) to head (c95880f).
Report is 8 commits behind head on dev.

Files	Patch %	Lines
nf_core/download.py	86.20%	4 Missing ⚠️
nf_core/synced_repo.py	86.66%	2 Missing ⚠️
nf_core/__main__.py	66.66%	1 Missing ⚠️

Additional details and impacted files

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[mirpedrol]

LGTM, but I will let @mashehu have another look before approving :)

[mashehu]

Almost there, would just like to discuss alternatives to -a before giving 👍🏻

[mashehu]

tbh not a big fan to use -a for something other then --all 🙂 how about just going for --tag?

[ewels]

+1 for --tag. I'm not too fussed about always providing short-form flags for every option. I think it's fine to have only long-form for the more unusual ones like this.

[MatthiasZepper]

Notwithstanding, it is so forthcoming of Seqera to rename their web application just to free up the -t short-form flag for the future 😝...

[ewels]

👉🏻 #2938 (comment)

[mashehu]

are we sure we want to use the nf-core-bot as the default user? I would prefer an error I think

[MatthiasZepper]

It doesn't need to be the nf-core bot we are using here as default, but an error for that reason is not a real option in my opinion.

After all, the downloaded bare repositories are pretty ephemeral and one would need to configure git properly to avoid. Admittedly, it is just two git commands that need to be run once and then never again, but I for example use nf-core download exclusively on Rackham, so I have Singularity available. But I am not doing development work on Rackham, so I haven't set up git there properly. The same goes for the func user that we use to prepare the deployments.

Lastly, mind that it was actually the failing CI that prompted me to include this, so if we switch back to error out, we first have to adapt our Pytest CI 🫣.

[ewels]

I agree with @MatthiasZepper on this one - this git action of adding a tag isn't really meaningful as a development step. This is purely an action to make running the pipeline easier, where the user behind the action is not visible. So I don't think that it matters what the account is, so I would prefer not to error.

That being said, I don't think that it needs to be a real account either? So maybe we can set it to automated@tag.com or some other fake email..?

[MatthiasZepper]

Yes, I have changed the default e-mail address. I presumed, it was a sinkhole / do not reply address, but apparently not.

[mashehu]

Suggested change

	self.ensure_git_user_config(f"nf-core download v{nf_core.__version__}", "core@nf-co.re")
	self.ensure_git_user_config(f"nf-core download v{nf_core.__version__}", "test@example.com")

[MatthiasZepper]

Thanks for the suggestion! I had at the same time googled a bit about suitable sinkhole addresses, and now settled on dev-null@example.com. Ultimately, I could not find any information on behalf of ICANN / IETF that one may indeed use any e-mail of that reserved domain for anything else than documentation.