Use ruff for linting (instead of current flake8/isort/pyflakes)

[jairhenrique]

https://github.com/astral-sh/ruff

[codecov-commenter]

Codecov Report

Merging #727 (b49bea7) into master (34d5384) will increase coverage by 0.05%.
The diff coverage is 90.90%.

❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more.

@@            Coverage Diff             @@
##           master     #727      +/-   ##
==========================================
+ Coverage   90.59%   90.64%   +0.05%     
==========================================
  Files          28       28              
  Lines        1797     1796       -1     
  Branches      270      323      +53     
==========================================
  Hits         1628     1628              
+ Misses        134      133       -1     
  Partials       35       35              

Impacted Files	Coverage Δ
vcr/cassette.py	96.31% <ø> (ø)
vcr/config.py	95.06% <ø> (ø)
vcr/request.py	97.53% <ø> (ø)
vcr/serialize.py	100.00% <ø> (ø)
vcr/stubs/__init__.py	93.87% <ø> (ø)
vcr/stubs/tornado_stubs.py	0.00% <ø> (ø)
vcr/_handle_coroutine.py	33.33% <50.00%> (ø)
vcr/__init__.py	100.00% <100.00%> (ø)
vcr/errors.py	100.00% <100.00%> (ø)
vcr/migration.py	74.48% <100.00%> (ø)
... and 3 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[hartwork]

Hi @jairhenrique I am wondering:

• whether ruff should be replacing flake8 (rather than adding to it, the two living side by side) when https://beta.ruff.rs/docs/faq/#how-does-ruff-compare-to-flake8 says that only a "subset of the E and W rules (which originate from pycodestyle)" are implemented
• whether it's a gain or a loss to drop the noqa markers, in particular those with a number.

What do you think?

[jairhenrique]

@hartwork I will enable default linters "E","F","W" to be more close with flake8.
I think the noqa markers will come back when enabling these linters.

[jairhenrique]

• whether it's a gain or a loss to drop the noqa markers, in particular those with a number.

I usually don't use the numbers, I can't tell if this is a loss or gain.

[hartwork]

@jairhenrique what the numbers allow in theory is to annotate why something was marked noqa semantically (i.e. documentation) but also to limit the scope of the noqa marker to just that aspect so that the QA tool won't silence unrelated issues now or in the future by accident.

I wonder, if we drop the markers, why doesn't ruff complain about the related cases, why doesn't ruff ask for the same issues to be intended and/or tolerable?

[jairhenrique]

@hartwork The ruff dropped noqa because I didn't enable the E, F, W linters.
Enabling these linters I needed to turn noqa back.

[jairhenrique]

@hartwork I returned the numbers in noqa. Removed noqa tags are from "fixed" issues.

[hartwork]

I wonder what to do about the existing block tool.isort further up:

[tool.isort]
line_length = 110
known_first_party = "vcrpy"
multi_line_output = 3
use_parentheses = true
include_trailing_comma = true

In particular:

• does ruff read any of that? (assuming: no)
• should it be dropped? (assuming: yes)
• can any of these options ported to ruff more?

[jairhenrique]

does ruff read any of that?

No 😄

should it be dropped?

Yes, done!

can any of these options ported to ruff more?

That's the default ruff configuration, but some settings configurations was not provided yet.

[hartwork]

@jairhenrique thanks for the adjustments! I have checked locally that flake8 does not report any errors against the latest state (except the now expected "E501 line too long" because flake8 no longer knows we want 110 columns). The Python changes are great and if for some reason ruff should turn out to have serious drawbacks over flake8/isort/pyflakes latter, nothing is blocking a cheap return to these tools, so the risk is minimal. Good work! 👍