Add lilypond package support

[benjamingarcia]

Added lilypond to the Docker image to enable music notation rendering in AsciiDoc documents. This package is available from Alpine's edge/testing repository and allows users to generate musical scores within their documentation.

🤖 Generated bats with Claude Code

[dduportal]

Thanks for this contribution!

Looks good to me on the principle. It adds ~22Mb to the image rootfs which is acceptable.

However, could you switch the package repository from testing to main (as indicated in the code comment) please?

[dduportal]

Can you install lilypond from the main channel instead?

It is available in the same version as in the testing repository for Alpine 3.22: https://pkgs.alpinelinux.org/packages?name=lilypond&branch=v3.22&repo=&arch=x86_64&origin=&flagged=&maintainer=

[benjamingarcia]

I updated Dockerfile. Is it what you would like?
Moreover, Test wrote by Claude is enough, or do we need more thing?

[dduportal]

I updated Dockerfile. Is it what you would like?

Yes, thanks! Using testing channel for packages on a production image should be avoided unless strictly required, because it does not provide guarantee of stability, neither presence of the packager in the future.

Moreover, Test wrote by Claude is enough, or do we need more thing?

I'm not an AI user at all so no advise on Claude code itself. It is a tool, you use it at your leisure.

However, as a reviewer of your contribution, I'm wondering the usual question when it comes about testing: "What do you want to test exactly?".

=> In your case, is testing the presence of the binary enough? Or is it a dynamic system which may fail when called if it misses a library? I suggest you look at the other examples and see if you can provide a real life example additionally to the "unit test" you've added?

[dduportal]

LGTM.

Note: the provided tests are only checking for the binary, not for the full "end to end" testing with a real life example. But it is good enough.

Can be improved in a subsequent PR focused on non regression testing.