Since github is being unhelpful with running the CI, I mirrored this PR here bdraco#2 to get a run

Codecov Report

Merging #341 (0c3cea3) into master (fe94810) will increase coverage by 0.40%.
The diff coverage is 88.23%.

@@            Coverage Diff             @@
##           master     #341      +/-   ##
==========================================
+ Coverage   91.08%   91.49%   +0.40%     
==========================================
  Files           2        2              
  Lines        2671     2681      +10     
  Branches      360      362       +2     
==========================================
+ Hits         2433     2453      +20     
+ Misses        161      149      -12     
- Partials       77       79       +2     

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fe94810...0c3cea3. Read the comment docs.

These lines are uncovered now so not a new issue, but given this regressed, I could create tests that mock the socket.error failure conditions with patch.object on listen_socket being passed to add_multicast_member if that seems like a good idea?

Will pick this back up tomorrow as its late late here.

These lines are uncovered now so not a new issue, but given this regressed, I could create tests that mock the socket.error failure conditions with patch.object on listen_socket being passed to add_multicast_member if that seems like a good idea?

I mean if you're felling like doing that I think it'll be nice to have. On a general note I think the whole socket-selecting layer needs to be separated from the rest of the code so that it can be unit tested nicely without any monkey patching etc. but that's more of a long-term plan.

Version 0.30.0 has just been relased with this change.