mkdocs: fix linkage

[p-linnane]

• Have you followed the guidelines for contributing?
• Have you ensured that your commits follow the commit style guide?
• Have you checked that there aren't other open pull requests for the same formula update/change?
• Have you built your formula locally with HOMEBREW_NO_INSTALL_FROM_API=1 brew install --build-from-source <formula>, where <formula> is the name of the formula you're submitting?
• Is your test running fine brew test <formula>, where <formula> is the name of the formula you're submitting?
• Does your build pass brew audit --strict <formula> (after doing HOMEBREW_NO_INSTALL_FROM_API=1 brew install --build-from-source <formula>)? If this is a new formula, does it pass brew audit --new <formula>?

---

[Bo98]

Seems weird to be syntax only to introduce a dependency when the current bottles aren't using it no? The linkage of the current bottles shouldn't be broken, otherwise brew linkage would have caught that. PyYAML has a fast and slow mode depending on whether libyaml was present at build time.

[p-linnane]

Working through a bunch of indirect linkage stuff so we can get Homebrew/brew#16718 merged.

[Bo98]

Ok so if it's already in the dep tree indirectly, doesn't that mean all Python dependents have it indirectly as there's no other dependencies here?

[p-linnane]

Ok this is strange. Compiling locally I'm not seeing any linkage with libyaml like all of the other formulae we've had to push this to recently: https://github.com/Homebrew/homebrew-core/pulls?q=is%3Apr+sort%3Aupdated-desc+is%3Amerged+fix+linkage

Using arxiv_latex_cleaner as an example, our current bottles show this linkage:

╰─ brew linkage arxiv_latex_cleaner
System libraries:
  /usr/lib/libSystem.B.dylib
Homebrew libraries:
  /opt/homebrew/opt/libyaml/lib/libyaml-0.2.dylib (libyaml)
Undeclared dependencies with linkage:
  libyaml
Dependencies with no linkage:
  pillow

This is despite libyaml not being in the tree (other than due to the PR where I added it). Not sure what is going on here now.

[Bo98]

arxiv_latex_cleaner bottles were rebuilt in #163377. Do you have an example where bottles were not rebuilt?

[Bo98]

To be clear, the behaviour I expect is that libyaml linkage is introduced by adding this dependency (and is a good idea to do so, for opportunistic linkage reasons + performance in some cases), but would require a bottle rebuild for that to show.

For example, cookiecutter after #163103:

==> brew linkage --cached cookiecutter
System libraries:
  /usr/lib/libSystem.B.dylib

cookiecutter after #163327:

==> brew linkage --cached cookiecutter
System libraries:
  /usr/lib/libSystem.B.dylib
Homebrew libraries:
  /opt/homebrew/opt/libyaml/lib/libyaml-0.2.dylib (libyaml)
Dependencies with no linkage:
  python-certifi

[p-linnane]

So essentially we need to do a rebuild and publish those bottles when adding libyaml for pyyaml? I can run back through and check everything and rebuild if so.

[Bo98]

Yes

[p-linnane]

Thank you for the clarification. I have a better understanding now. I've kicked off a rebuild in this PR, and will run back through the previous ones to get those handled as well.

[github-actions]

🤖 An automated task has requested bottles to be published to this PR.