Находки в опенсорсе: Python
1.06K subscribers
5 photos
348 links
Легкие задачки в опенсорсе из мира Python

Чат: @opensource_findings_chat
Download Telegram
🚀 New issue to wemake-services/wemake-python-styleguide by @luminoso
📝 False positive for WPS457 when using asyncio to control loops (#3573)


What's wrong

For the code:

import asyncio


async def infinite_loop() -> None:
try:
while True:
await asyncio.sleep(1)
print("I'm alive. And doing work.")
except asyncio.CancelledError:
print("Loop cancelled")


t = asyncio.create_task(infinite_loop())
await asyncio.sleep(5)
t.cancel()

WPS457: Found an infinite while loop is raised. But infinite loop is being handled. Just not within the while loop.

How it should be

Not to raise WPS457.

Not 100% sure here if it is the best practice or not.
Also probably the solution is too complex and is just easier to add a noqa in the code.

Flake8 version and plugins

{
"platform": {
"python_implementation": "CPython",
"python_version": "3.13.9",
"system": "Linux"
},
"plugins": [
{
"plugin": "mccabe",
"version": "0.7.0"
},
{
"plugin": "pycodestyle",
"version": "2.14.0"
},
{
"plugin": "pyflakes",
"version": "3.4.0"
},
{
"plugin": "wemake-python-styleguide",
"version": "1.4.0"
}
],
"version": "7.3.0"
}

pip information

pip 25.3 from /var/home/luminoso/.local/lib/python3.14/site-packages/pip (python 3.14)
anyio==4.11.0
archspec==0.2.5
argcomplete==3.6.3
asttokens==3.0.0
attrs==25.4.0
boto3==1.42.4
botocore==1.42.4
Brlapi==0.8.7
Brotli==1.1.0
certifi==2025.7.9
charset-normalizer==3.4.3
click==8.1.7
conda==25.11.0
conda-package-handling==2.4.0
conda_package_streaming==0.11.0
cupshelpers==1.0
dasbus==1.7
dbus-python==1.4.0
decorator==5.2.1
distro==1.9.0
executing==2.2.1
fedora-third-party==0.10
file-magic==0.4.0
frozendict==2.4.6
gbinder-python==1.1.2
h11==0.16.0
httpcore==1.0.9
httpx==0.28.1
idna==3.10
ipython==8.37.0
jedi==0.19.2
jmespath==1.0.1
jsonpatch==1.33
jsonpointer==2.4
jsonschema==4.23.0
jsonschema-specifications==2024.10.1
langtable==0.0.69
louis==3.33.0
matplotlib-inline==0.1.7
menuinst==2.3.1
mercurial==7.1.1
mutagen==1.47.0
nftables==0.1
olefile==0.47
packaging==25.0
parso==0.8.5
pexpect==4.9.0
pillow==11.3.0
platformdirs==4.2.2
pluggy==1.6.0
progressbar2==4.5.0
prompt_toolkit==3.0.41
psutil==7.0.0
ptyprocess==0.7.0
pure_eval==0.2.3
PyAudio==0.2.13
pycairo==1.28.0
pyclip==0.7.0
pycosat==0.6.6
pycryptodomex==3.23.0
pycups==2.0.4
pyenchant==3.2.2
pygdbmi==0.11.0.0
Pygments==2.19.1
PyGObject==3.54.5
pyinotify==0.9.6
PySocks==1.7.1
python-dateutil==2.9.0.post0
python-linux-procfs==0.7.3
python-utils==3.9.1
pyudev==0.24.3
pyxdg==0.27
PyYAML==6.0.2
pyynl @ file:///builddir/build/BUILD/kernel-6.17.10-build/kernel-6.17.10/linux-6.17.10-300.fc43.x86_64/tools/net/ynl
RapidFuzz==3.12.2
referencing==0.36.2
regex==2025.10.23
requests==2.32.5
rpds-py==0.27.0
rpm==6.0.0
rpmautospec==0.8.3
rpmautospec-core==0.1.5
ruamel.yaml==0.18.16
ruamel.yaml.clib==0.2.12
s3transfer==0.16.0
selinux @ file:///builddir/build/BUILD/libselinux-3.9-build/libselinux-3.9/src
sentry-sdk==2.35.0
sepolicy @ file:///builddir/build/BUILD/policycoreutils-3.9-build/selinux-3.9/python/sepolicy
setools==4.6.0
setuptools==78.1.1
six==1.17.0
sniffio==1.3.1
sos==4.10.1
stack_data==0.6.3
tqdm==4.67.1
traitlets==5.14.3
typing_extensions==4.15.0
urllib3==2.5.0
wcwidth==0.2.13
websockets==15.0.1
yt-dlp==2025.10.22
zstandard==0.25.0

OS information

$ lsb_release -a
LSB Version: n/a
Distributor ID: Fedora
Description: Fedora Linux 43.20251209.0 (Kinoite)
Release: 43
Codename: n/a


#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
❤1
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 False positive for `WPS430` with whitelisted name (#3589)


def wrap_handler(
method: _MethodSyncHandler | _MethodAsyncHandler,
) -> SyncErrorHandlerT | AsyncErrorHandlerT:
if inspect.iscoroutinefunction(method):

@wraps(method)
async def decorator(
endpoint: 'Endpoint',
controller: 'Controller[BaseSerializer]',
exc: Exception,
) -> HttpResponse:
return await method( # type: ignore[no-any-return]
controller.active_blueprint,
endpoint,
controller,
exc,
)

else:

@wraps(method)
def decorator(
endpoint: 'Endpoint',
controller: 'Controller[BaseSerializer]',
exc: Exception,
) -> HttpResponse:
return method( # type: ignore[return-value]
controller.active_blueprint,
endpoint,
controller,
exc,
)

return decorator

This code falsy raises two WPS430 violations.

  69:9     WPS430 Found nested function: factory
async def factory(
^

84:9 WPS430 Found nested function: factory
def factory(
^

But, the name is correct. decorator name is in whitelist of names. This is a false positive.
We need to fix it and add a test case.


#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 `WPS617` false positive with keyword args (#3596)


self._media_by_precedence = sorted(
(
media_type
for parser in self._parsers.values()
if (media_type := MediaType(parser.content_type)).quality != 0
),
key=lambda media: (media.specificity, media.quality), # noqa: WPS617
reverse=True,
)

It says:

django_modern_rest/negotiation.py

52:17 WPS617 Found lambda assigned as an attribute
key=lambda media: (media.specificity, media.quality),
^

Which is clearly a bug, it is a named param, not an attribute.
PRs are welcome, this is easy to fix.


#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 Remove `WPS354` (#3601)


This rules has more problems that it solves:

1. Two yields and yield from are two different things
2. Async iterators cannot use yield from, but WPS354 still raises even in async def

So, it is time to remove this rule. I was wrong about that :(
See other rules with disabled_since attribute on how to disable a rule.


#feature #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🫡1
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 `WPS457` false positive with `while` in `try` / `except` (#3603)


I have the following code:

def factory() -> Iterator[bytes]:
try:
while True: # noqa: WPS457
yield async_to_sync(async_anext)(iterator)
except StopAsyncIteration:
pass

It should not raise WPS457, because while is in try / except.
It can clearly raise and we handle the error. So, it is not infinite.

PRs are welcome! 👍


#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 Build a MCP extra (#3770)


We need to add an MCP protocol support to wemake-python-styleguide.
It should be installed with wemake-python-styleguide[mcp] and advertised in the readme.

What should it do? Provide the same output as wps explain WPSXXX.
But, tool call is much more expensive than an MCP call.

CC @Khabib73


#feature #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🚀 New issue to wemake-services/wemake-python-styleguide by @Novohudonossor
📝 tstring_same_prefix params contain f-strings — the t-string coverage from #3702 is never exercised (#3793)


What's wrong

Two test constants touched by #3702 ("extend f-string rules to t-strings") do not exercise t-strings at all.

1. tests/test_visitors/test_ast/test_complexity/test_overuses/test_overused_string.py:112-131

tstring_same_prefix1 = pytest.param(
"""
x = f'Hello, {pattern}'
y = f'Hello, {pattern}'
""",
marks=pytest.mark.skipif(not PY314, reason='t-strings are only in Python 3.14+'),
)

The snippet is identical to fstring_same_prefix1 right above it. Below 3.14 the param is skipped by the mark; on 3.14+ it runs and asserts on f-strings. Either way WPS226 is never checked against a t-string, and the suite stays green.

2. tests/test_visitors/test_tokenize/test_comments/test_comment_in_formatted_string312.py:87

PREFIXES is ['f', 't'] and both tests run code.format(prefix), so every constant is exercised twice. rfstring_with_comment_triple_single_quotes hardcodes rf'''…''' instead of r{0}'''…''' like its two siblings above it, so the t round re-tests rf.

How it should be
1. t'Hello, {pattern}' and t'{pattern}-postfix' inside the two tstring_same_prefix* params.
2. r{0}'''test{{a # comment\n}}''', matching rfstring_with_comment_single_quotes and rfstring_with_comment_triple_quotes.

Happy to open a PR for both if that is useful.

Flake8 version and plugins / pip information / OS information

Not applicable — this is not a runtime report. Found by reading the tests, verified against HEAD 8ebc607; there is no flake8 --bug-report output to paste.

Provenance

Both came out of a review run with ReviewGate (https://reviewgate.dev), a local review gate I build, over 76b3ac3~1..c580fc3. The run was local, on my own model key, and nothing left my machine.


#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 Do not count docstrings for `WPS226` (#3803)


dmr/security/jwt/views.py

155:9 WPS226 Found string literal over-use: 'Lazy endpoint spec.' 6 > 3
"""Lazy endpoint spec."""
^

I just got this, where """Lazy endpoint spec.""" is an actual docstring.
This is not right.


#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 Incorrect Jones complexity for formatted strings (#3820)


These two lines:

• tuple(f'/{pref.strip("/")}/' for pref in (prefix, *prefixes))
• tuple(f'/{pref.strip("/")}' for pref in (prefix, *prefixes)) (notice the missing / at the end)

have different Jones complexity.
They must not. It should be counted as:

• 1 per every {}
• As regular for all python expressions inside {}
• All string parts of fstring are counted as total of 1

#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator
🚀 New issue to wemake-services/wemake-python-styleguide by @sobolevn
📝 `WPS235`: ignore imports from `typing` (#3824)


I can't limit the number of types I import from typing, I can't refactor types to use typing. prefix and import import typing

We need to just ignore this module. And typing_extensions as well.


#bug #help_wanted #levelstarter #good_first_issue #wemake_python_styleguide #wps
sent via relator