Bump codecov-action version to v5

[glatterf42]

dependabot opened #548 automatically, but for some reason, the codecov status is not reported (and the check continually 'expected'). This happens infrequently and usually, I just un-require the check, merge the PR, and re-require it if I now the PR is not making changes to coverage anyway.
This time, though, I'm not too sure that everything's alright since we are changing the version of the codecov action that our pytest workflow is using. I tried running all checks again, but that didn't trigger the codecov check. Next, I has dependabot 'recreate' #548, which triggered some kind of rebase, but the checks still didn't run. In the meantime, I looked at the codecov page for the PR, which seems to say that no reports for the PR's only commit were uploaded. Unfortunately, there doesn't seem to be any option to trigger the check manually, either. That's why I'm now creating this brand new PR with exactly the same changes as #548 in the hopes of triggering the codecov checks.
If this doesn't work, I might have to explore in detail what changed with the version bump and how we relied on that.

How to review

• Read the diff and note that the CI checks all pass (in particular the codecov checks).

PR checklist

• Continuous integration checks all ✅
• Add or expand tests; coverage checks both ✅
• [ ] Add, expand, or update documentation. Just CI version bump.
• [ ] Update release notes. Just CI version bump.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.9%. Comparing base (c49a115) to head (9bdbc6c).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##            main    #549     +/-   ##
=======================================
- Coverage   98.9%   98.9%   -0.1%     
=======================================
  Files         44      44             
  Lines       4819    4819             
=======================================
- Hits        4770    4769      -1     
- Misses        49      50      +1     

see 1 file with indirect coverage changes

[khaeru]

Faster than I could write my comment!

I wonder if another reason may be that Codecov, or something in their or our configuration, disables creating such checks on bot PRs.

[glatterf42]

I don't think so as this doesn't always happen and works for e.g. message_ix, where I'm not aware of differences in the config. Do you want me to investigate more or are we happy that it seems to work and just squash and merge this PR?

[khaeru]

I don't think so as this doesn't always happen and works for e.g. message_ix, where I'm not aware of differences in the config. Do you want me to investigate more or are we happy that it seems to work and just squash and merge this PR?

My 0th-order impression, from a cursory skim of the upstream issues, is that there is/was some flakiness in how Codecov's own bots/services operate. They might also be tweaking things on their back-end after hearing of such issues during the roll out. So let's regard #548 as transient/flaky, and interpret this and the message_ix/message-ix-models PRs as the actual/stable behaviour, and carry on 👍🏾