Run tests without tox

[jairhenrique]

This change remove flaky errors on ci.
Tests are faster with dependency caching.

[codecov-commenter]

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (f1e0241) 90.13% compared to head (cba2b2d) 90.83%.
Report is 5 commits behind head on master.

Files	Patch %	Lines
vcr/stubs/httpx_stubs.py	95.83%	1 Missing ⚠️

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

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #794      +/-   ##
==========================================
+ Coverage   90.13%   90.83%   +0.69%     
==========================================
  Files          27       27              
  Lines        1815     1822       +7     
  Branches      336      338       +2     
==========================================
+ Hits         1636     1655      +19     
+ Misses        134      122      -12     
  Partials       45       45              

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

[jairhenrique]

@hartwork What do you think about this approach? I didn't have any strange errors when running the tests, the suite was executed without the need to run the pipeline to force the tests to turn green.

[hartwork]

@jairhenrique this reduces coverage with regard to covering with urllib3 1.x and 2.x for the same version of Python, e.g. for Python 3.10 we cover:

• py310-requests-urllib3-1
• py310-urllib3-1
• py310-requests-urllib3-2
• py310-urllib3-2

Too keep the same level of coverage and be protected against regressions this would probably need the include variant of the matrix feature of GitHub Actions. And if we take this route we probably want to git rid of tox altogether to not have to maintain it and not have it uncovered. (Dropping tox will affect which tests contributors are running locally one way or another also.)

[hartwork]

@jairhenrique the general direction is good, some details need fixing:

• git grep tox shows more places that need adjustment:

  # git --no-pager grep tox
  .gitignore:.tox
  runtests.sh:# https://blog.ionelmc.ro/2015/04/14/tox-tricks-and-patterns/#when-it-inevitably-leads-to-shell-scripts
  runtests.sh:# You can and should use WSL for running tox on Windows when it calls bash scripts.

• The pull request does 5+ different things all in one commit. It would be great to split this up into semantic units using git rebase -i.

• Plus everything I commented below:

[hartwork]

@jairhenrique exclude is a nice idea — it's more future-proof than include — but there is at least one key showstopper left below:

[hartwork]

@jairhenrique the general direction is good, some details need fixing:

• git grep tox shows more places that need adjustment:

  # git --no-pager grep tox
  .gitignore:.tox
  runtests.sh:# https://blog.ionelmc.ro/2015/04/14/tox-tricks-and-patterns/#when-it-inevitably-leads-to-shell-scripts
  runtests.sh:# You can and should use WSL for running tox on Windows when it calls bash scripts.
  

• The pull request does 5+ different things all in one commit. It would be great to split this up into semantic units using git rebase -i.

• Plus everything I commented below:

@jairhenrique did you see this^^?

[hartwork]

@jairhenrique thanks for the adjustments! Is keeping .tox in .gitignore intentional?

[jairhenrique]

@jairhenrique thanks for the adjustments! Is keeping .tox in .gitignore intentional?

@hartwork yes, to prevent someone from uploading this folder to the repository.

[hartwork]

@jairhenrique thanks for the adjustments! Is keeping .tox in .gitignore intentional?

@hartwork yes, to prevent someone from uploading this folder to the repository.

@jairhenrique I'm not sure I understand — wouldn't we see an addition like that in the pull request when it's happening? Please help me understand.

[jairhenrique]

@jairhenrique I'm not sure I understand — wouldn't we see an addition like that in the pull request when it's happening? Please help me understand.

Yes, we would see. I just left it so I would have to avoid this work. I can remove it if you understand that this is a problem.

[hartwork]

@jairhenrique I'm not sure I understand — wouldn't we see an addition like that in the pull request when it's happening? Please help me understand.

Yes, we would see. I just left it so I would have to avoid this work. I can remove it if you understand that this is a problem.

@jairhenrique I'll leave it to you, we'll probably be good either way.

Thanks for the other adjustments, the nice split up into multiple commits — this pull request has come a long way! 👍

[jairhenrique]

@kevin1024 do you agree with these changes?

[graingert]

This broke the tornado tests: https://github.com/kevin1024/vcrpy/actions/runs/7620100960/job/20754258427?pr=794#step:5:38 note all the warnings

[graingert]

actually it looks like it was broken before

[jairhenrique]

actually it looks like it was broken before

@graingert yes 😞