Add support for "jf uv publish" to support Astral uv Python usage

[yoitsmikeho]

Is your feature request related to a problem? Please describe.

Yes, today we are using uv with Artifactory Python repositories and would like to have support extended to the JFrog CLI to expand that usage into our build processes.

Describe the solution you'd like to see

Please add support for jf uv publish similar to the support added for jf twine publish.

Describe alternatives you've considered

We have used twine and pip in the past, but it hasn't met our performance needs like uv does.

Additional context

No response

[goodtune]

Thanks for opening this request for us @yoitsmikeho.

The way we currently get jf to help gather context for our Python packages is like so:

python build --wheel  # generate our package.whl
jf config add company --url https://company.jfrog.io/ --access-token "${CI_SYSTEM_SECRET_LOOKUP}"
jf rt build-clean
jf pip-config --repo-resolve pypi --server-id-resolve company
python -m venv .venv && source .venv/bin/activate  # necessary to get 'pip' on the PATH for jf
jf pip install --module package-name wheel  # installs all the transitive dependencies to add to build info
jf rt upload --module package-name wheel repo/package-name/package-version/  # we did this before 'jf twine' was added which helps keep the hierarchy
jf rt build-publish

The workflow that uv provides is really neat, and Just Works™, but does not result in the build info.

uv build
uv publish

As a goal, this should be the ergonomics of a workflow... given a repository checkout:

jf config add company --url https://company.jfrog.io/ --access-token "${CI_SYSTEM_SECRET_LOOKUP}"
jf rt build-clean
jf uv-config --repo-resolve pypi --server-id-resolve company
jf uv build
jf uv publish

The uv tool can easily be influenced with environment variables, so jf uv-config would be able to generate a config file in a hidden directory and influence the underlying uv tool to consume it.

[curt-arcee]

This feature would be very helpful for our team. More and more of our research team is using uv instead of pip. So, having jf uv support would simplify our pipelines a great deal.