fix: EncodingWarning

[henryiii]

For reference, this is PEP 597, in Python 3.10+, and happens when PYTHONWARNDEFAULTENCODING is set.

This cherry-picks @methane's #504, but adds two new commits; one that adds testing for this (using my favorite pytest settings), and one that fixes two more cases of missing encodings discovered by adding those tests.

[codecov]

Codecov Report

Patch coverage: 100.00% and project coverage change: -2.65 ⚠️

Comparison is base (e0f18dd) 72.73% compared to head (c099007) 70.09%.

❗ Current head c099007 differs from pull request most recent head c8c99a1. Consider uploading reports for the commit c8c99a1 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #512      +/-   ##
==========================================
- Coverage   72.73%   70.09%   -2.65%     
==========================================
  Files          13       13              
  Lines        1060     1060              
==========================================
- Hits          771      743      -28     
- Misses        289      317      +28     

Impacted Files	Coverage Δ
src/wheel/bdist_wheel.py	51.90% <100.00%> (ø)
src/wheel/cli/pack.py	56.89% <100.00%> (-37.94%)	⬇️
src/wheel/metadata.py	96.66% <100.00%> (ø)

... and 2 files with indirect coverage changes

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[henryiii]

FYI, I'm fine if you select "update by rebase" from the dropdown on the PR for my PRs, you don't need to merge, I don't mind the rebase. If a PR stays open long enough, the merges eventually make it a mess (I spent hours trying to pull the actual changes from pypa/build#361 but couldn't untangle it). Depends on the contributor, but I don't mind resetting.

[agronholm]

FYI, I'm fine if you select "update by rebase" from the dropdown on the PR for my PRs, you don't need to merge, I don't mind the rebase. If a PR stays open long enough, the merges eventually make it a mess (I spent hours trying to pull the actual changes from pypa/build#361 but couldn't untangle it). Depends on the contributor, but I don't mind resetting.

Noted. Personally I prefer merges to the PR branch while I'm working on one.

[agronholm]

Looks like the merge from main I did was somehow wrong, and now tests are failing. Can you fix it?

[henryiii]

Sure. Technically de80240 is now "one more missed encoding", because the other one was fixed when factoring-out into a function in #422. That was probably where the merge messed up.

[agronholm]

Thanks!