This one is a little thrown together, I'm willing to rework it and do it differently (e.g. we could make it possible to convert a KeyRingIdentifier to a KeySerialId more explicitly than KeySerialId::new(KeyRingIdentifier as i32) and that might be sufficient for these use cases with a little adjustment to the docs to make clear how to do this.

I do think there does need to be a supported way to link keyrings.

Also willing to add tests and improve the comments if you can do a cursory review and say what's needed.

@afontenot Thanks for the PR!

I think your explicit methods are better than having users convert to a raw ID and using existing Key methods to workaround it.

It looks clean, some tests would be great though.

I'll try to get to these PRs this weekend.

@landhb tests have been added.

Implementing set_perms on KeyRing in the future would be useful, it would also allow testing some of the claims made about e.g. write / linking access (throughout the KeyRing implementation, not specific to this commit).

Sorry for the delay. I'll try to review the 3 open PRs and cut a release soon.

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 69.09%. Comparing base (2e81f72) to head (93224f3).
Report is 3 commits behind head on main.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

@@            Coverage Diff             @@
##             main      #16      +/-   ##
==========================================
- Coverage   78.33%   69.09%   -9.24%     
==========================================
  Files           9        9              
  Lines         420      343      -77     
==========================================
- Hits          329      237      -92     
- Misses         91      106      +15     

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

Thanks @afontenot !