Stabilize redundant-none-literal (PYI061)

[ntBre]

Tests and docs look good

[github-actions]

ruff-ecosystem results

Linter (stable)

ℹ️ ecosystem check detected linter changes. (+21 -0 violations, +0 -0 fixes in 5 projects; 50 projects unchanged)

apache/airflow (+1 -0 violations, +0 -0 fixes)

ruff check --no-cache --exit-zero --ignore RUF9 --no-fix --output-format concise --no-preview --select ALL

+ airflow-core/src/airflow/serialization/serialized_objects.py:3239:36: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 

langchain-ai/langchain (+6 -0 violations, +0 -0 fixes)

+ libs/langchain/langchain/chat_models/base.py:40:34: PYI061 [*] Use `None` rather than `Literal[None]`
+ libs/langchain/langchain/chat_models/base.py:48:20: PYI061 [*] Use `None` rather than `Literal[None]`
+ libs/langchain/langchain/chat_models/base.py:51:34: PYI061 [*] Use `None` rather than `Literal[None]`
+ libs/langchain_v1/langchain/chat_models/base.py:36:34: PYI061 [*] Use `None` rather than `Literal[None]`
+ libs/langchain_v1/langchain/chat_models/base.py:44:20: PYI061 [*] Use `None` rather than `Literal[None]`
+ libs/langchain_v1/langchain/chat_models/base.py:47:34: PYI061 [*] Use `None` rather than `Literal[None]`

latchbio/latch (+1 -0 violations, +0 -0 fixes)

+ src/latch/types/metadata.py:496:45: PYI061 [*] Use `Optional[Literal[...]]` rather than `Literal[None, ...]` 

pandas-dev/pandas (+4 -0 violations, +0 -0 fixes)

+ pandas/core/groupby/groupby.py:4324:39: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ pandas/core/groupby/indexing.py:299:39: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ pandas/io/html.py:1045:28: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ pandas/io/html.py:223:32: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 

zulip/zulip (+9 -0 violations, +0 -0 fixes)

ruff check --no-cache --exit-zero --ignore RUF9 --no-fix --output-format concise --no-preview --select ALL

+ corporate/views/billing_page.py:329:24: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ corporate/views/remote_billing_page.py:155:26: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ corporate/views/remote_billing_page.py:156:42: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ corporate/views/remote_billing_page.py:157:48: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ corporate/views/remote_billing_page.py:676:26: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ corporate/views/remote_billing_page.py:677:42: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ corporate/views/remote_billing_page.py:678:48: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ corporate/views/remote_billing_page.py:67:5: PYI061 [*] Use `Literal[...] | None` rather than `Literal[None, ...]` 
+ zerver/lib/event_types.py:906:62: PYI061 [*] Use `None` rather than `Literal[None]`

Changes by rule (1 rules affected)

code	total	+ violation	- violation	+ fix	- fix
PYI061	21	21	0	0	0

Linter (preview)

✅ ecosystem check detected no linter changes.

[ntBre]

I can kind of see why people would prefer Literal[1, 2, 3, None] over Literal[1, 2, 3] | None or Optional[Literal[1, 2, 3]], but these all look like true positives to me. If you prefer the other style, I guess you just wouldn't activate the rule. Literal[None] -> None seems less contentious, but it does appear in the ecosystem check, mostly for @overloads.

[ntBre]

Just to record the decision here, we're aware of #20265, but I agree with Alex that it doesn't need to block stabilization.

[user27182]

I can kind of see why people would prefer Literal[1, 2, 3, None] over Literal[1, 2, 3] | None or Optional[Literal[1, 2, 3]], but these all look like true positives to me.

This is more than a style preference, as it modifies the output of typing.get_args, see #20729. This auto-fix caused test failures in the pyvista repo when trying to update the ruff pre-commit hook:
pyvista/pyvista@d908bd7

[ntBre]

Ah that's good to know. Thank you for the report!