STYLE use TypeAlias and upgrade ruff

[MarcoGorelli]

The latest autoupdate job removes some useful type aliases definitions

As noted by @twoertwein , this is because we haven't annotated them as TypeAlias

So, the task here is:

1. pre-commit autoupdate
2. add

  # Useless statement
  "B018",

under Additional checks that don't pass yet in pyproject.toml

3. check which files were updated in https://github.com/pandas-dev/pandas/pull/52345/files . For any place where the type alias definition was replaced with ..., you should annotate the variable with TypeAlias

Here's an example of the kind of change you'd be making:

diff --git a/pandas/_libs/interval.pyi b/pandas/_libs/interval.pyi
index 4c36246e04..f0412b8397 100644
--- a/pandas/_libs/interval.pyi
+++ b/pandas/_libs/interval.pyi
@@ -3,6 +3,7 @@ from typing import (
     Generic,
     TypeVar,
     overload,
+    TypeAlias,
 )
 
 import numpy as np
@@ -16,9 +17,9 @@ from pandas._typing import (
 
 VALID_CLOSED: frozenset[str]
 
-_OrderableScalarT = TypeVar("_OrderableScalarT", int, float)
-_OrderableTimesT = TypeVar("_OrderableTimesT", Timestamp, Timedelta)
-_OrderableT = TypeVar("_OrderableT", int, float, Timestamp, Timedelta)
+_OrderableScalarT: TypeAlias = TypeVar("_OrderableScalarT", int, float)
+_OrderableTimesT: TypeAlias = TypeVar("_OrderableTimesT", Timestamp, Timedelta)
+_OrderableT: TypeAlias = TypeVar("_OrderableT", int, float, Timestamp, Timedelta)
 
 class _LengthDescriptor:
     @overload

4. run pre-commit run ruff --all-files, and check that no type alias definitions are overwritten by ...
5. run pre-commit run --all-files, check everything passes
6. stage, commit, open pull request

See the contributing guide for how to get started https://pandas.pydata.org/docs/dev/development/contributing.html

[sjdex]

Hi @MarcoGorelli , I was working on this. Came across this no_default: Final = ... . Do I need to replace Final with TypeAlias in such case, not sure we can append it with Final.

[MarcoGorelli]

hey @sjdex - I think that one's fine to become as ..., as it's already defined here

pandas/pandas/_libs/lib.pyx

Line 2758 in 64eed4c

no_default = _NoDefault.no_default # Sentinel indicating the default value.

likewise cache_readonly

it's only the ones which look like they're type aliases (which aren't defined in other files) which we want to TypeAlias to

[twoertwein]

Hi @MarcoGorelli , I was working on this. Came across this no_default: Final = ... . Do I need to replace Final with TypeAlias in such case, not sure we can append it with Final.

You can test it by running pre-commit run -av ruff :) (after upgrading ruff to the latest version).

I think could have something like Final[Typing Alias] but I'm not sure whether that is valid/correct. I believe the value of this enum cannot be replaced with .... If you have a pandas development environment, it might be worth running mypy and pyright: https://pandas.pydata.org/pandas-docs/stable/development/contributing_codebase.html#validating-type-hints after you make changes to no_default.

[yusharth]

@sjdex Hey, is there any scope for improvement or help I can provide you to close this issue altogether? I am new to open source and seek your guidance. Let me know if I can help in anyway.

[aswinj18]

Hi @MarcoGorelli,
Is this issue still open. If so can you assign it to me?

[snorfyang]

@MarcoGorelli, Is anyone still working on this issue? If no I can solve the problem.

[MarcoGorelli]

there's an open PR

[MarcoGorelli]

Hi @snorfyang - looks like the original PR has gone stale, so if you're interested in working on this, please feel free to open a new PR! thanks 🙏

[snorfyang]

take