BUG: relocate to_offset and other pyrefly-inspired changes (#1878)

* pyrefly

* bump ruff

Author: cmp0xff

.pre-commit-config.yaml

@@ -3,12 +3,12 @@ ci:
     autofix_prs: false
 repos:
 -   repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.15.22
+    rev: v0.16.1
     hooks:
     -   id: ruff-check
         args: [--exit-non-zero-on-fix]
 -   repo: https://github.com/codespell-project/codespell
-    rev: v2.4.2
+    rev: v2.4.3
     hooks:
     - id: codespell
       args: [-L, "THIRDPARTY"]

pandas-stubs/_libs/tslibs/offsets.pyi

@@ -18,6 +18,7 @@ import numpy as np
 from pandas import Timestamp
 
 from pandas._typing import (
+    Frequency,
     ShapeT,
     np_ndarray_object,
 )
@@ -104,6 +105,11 @@ class BaseOffset:
 
 class SingleConstructorOffset(BaseOffset): ...
 
+@overload
+def to_offset(freq: None, is_period: bool = False) -> None: ...
+@overload
+def to_offset(freq: Frequency | timedelta, is_period: bool = False) -> BaseOffset: ...
+
 class Tick(SingleConstructorOffset):
     def __init__(self, n: int = ..., normalize: bool = ...) -> None: ...
     @property

pandas-stubs/_libs/tslibs/timestamps.pyi

@@ -48,7 +48,7 @@ _Ambiguous: TypeAlias = bool | Literal["raise", "NaT"]
 
 # Repeated from `_typing.pyi` so as to satisfy mixed strict / non-strict paths.
 # https://github.com/pandas-dev/pandas-stubs/pull/1151#issuecomment-2715130190
-TimeZones: TypeAlias = str | _tzinfo | None | int
+TimeZones: TypeAlias = str | _tzinfo | int | None
 
 class Timestamp(datetime, SupportsIndex):
     min: ClassVar[Timestamp]  # pyright: ignore[reportIncompatibleVariableOverride]

pandas-stubs/_typing.pyi

@@ -703,7 +703,7 @@ StorageOptions: TypeAlias = dict[str, Any] | None
 # compression keywords and compression
 CompressionDict: TypeAlias = dict[str, Any]
 CompressionOptions: TypeAlias = (
-    None | Literal["infer", "gzip", "bz2", "zip", "xz", "zstd", "tar"] | CompressionDict
+    Literal["infer", "gzip", "bz2", "zip", "xz", "zstd", "tar"] | CompressionDict | None
 )
 ParquetCompressionOptions: TypeAlias = (
     Literal["snappy", "gzip", "brotli", "lz4", "zstd"] | None
@@ -1203,7 +1203,7 @@ ExcelWriteEngine: TypeAlias = Literal["openpyxl", "odf", "xlsxwriter"]
 
 # Repeated in `timestamps.pyi` so as to satisfy mixed strict / non-strict paths.
 # https://github.com/pandas-dev/pandas-stubs/pull/1151#issuecomment-2715130190
-TimeZones: TypeAlias = str | tzinfo | None | int
+TimeZones: TypeAlias = str | tzinfo | int | None
 
 ColumnValue: TypeAlias = AnyArrayLike | Scalar | Sequence[Scalar] | range | None
 # Evaluates to a DataFrame column in DataFrame.assign context.

pandas-stubs/core/construction.pyi

@@ -136,7 +136,7 @@ def array(  # type: ignore[overload-overlap]
 @overload
 def array(
     data: (
-        Sequence[IntervalT | None | float]
+        Sequence[IntervalT | float | None]
         | IntervalArray
         | IntervalIndex
         | Series[Interval]

pandas-stubs/core/indexes/timedeltas.pyi

@@ -160,7 +160,7 @@ def timedelta_range(
     freq: Frequency | Timedelta | timedelta | None = None,
     name: Hashable | None = None,
     closed: Literal["left", "right"] | None = None,
-    unit: None | str = None,
+    unit: str | None = None,
 ) -> TimedeltaIndex: ...
 @overload
 def timedelta_range(
@@ -170,7 +170,7 @@ def timedelta_range(
     freq: Frequency | Timedelta | timedelta | None = None,
     name: Hashable | None = None,
     closed: Literal["left", "right"] | None = None,
-    unit: None | str = None,
+    unit: str | None = None,
 ) -> TimedeltaIndex: ...
 @overload
 def timedelta_range(
@@ -180,7 +180,7 @@ def timedelta_range(
     freq: Frequency | Timedelta | timedelta | None = None,
     name: Hashable | None = None,
     closed: Literal["left", "right"] | None = None,
-    unit: None | str = None,
+    unit: str | None = None,
 ) -> TimedeltaIndex: ...
 @overload
 def timedelta_range(
@@ -190,5 +190,5 @@ def timedelta_range(
     *,
     name: Hashable | None = None,
     closed: Literal["left", "right"] | None = None,
-    unit: None | str = None,
+    unit: str | None = None,
 ) -> TimedeltaIndex: ...

pandas-stubs/core/series.pyi

@@ -128,6 +128,7 @@ from pandas.core.window.rolling import (
     Rolling,
     Window,
 )
+from typing_extensions import override
 import xarray as xr
 
 from pandas._libs.interval import Interval
@@ -1874,7 +1875,8 @@ class Series(IndexOpsMixin[S1], ElementOpsMixin[S1], NDFrame):
     @final
     def last_valid_index(self) -> Scalar: ...
     @overload
-    def value_counts(  # pyrefly: ignore
+    @override
+    def value_counts(
         self,
         normalize: Literal[False] = False,
         sort: _bool = ...,

pandas-stubs/tseries/frequencies.pyi

@@ -1,18 +1,7 @@
-from datetime import timedelta
-from typing import overload
+from pandas.core.indexes.datetimes import DatetimeIndex
+from pandas.core.indexes.timedeltas import TimedeltaIndex
+from pandas.core.series import Series
 
-from pandas import (
-    DatetimeIndex,
-    Series,
-    TimedeltaIndex,
-)
+from pandas._libs.tslibs.offsets import to_offset as to_offset
 
-from pandas._typing import Frequency
-
-from pandas.tseries.offsets import BaseOffset
-
-@overload
-def to_offset(freq: None, is_period: bool = False) -> None: ...
-@overload
-def to_offset(freq: Frequency | timedelta, is_period: bool = False) -> BaseOffset: ...
 def infer_freq(index: Series | DatetimeIndex | TimedeltaIndex) -> str | None: ...

pyproject.toml

@@ -201,6 +201,7 @@ extend-select = ["ALL"]
 ignore = [
   # The following rules are ignored permanently for good reasons
   "COM",    # https://docs.astral.sh/ruff/rules/#flake8-commas-com
+  "CPY001", # https://docs.astral.sh/ruff/rules/missing-copyright-notice/
   "D",      # https://docs.astral.sh/ruff/rules/#pydocstyle-d
   "E501",   # handled by black https://docs.astral.sh/ruff/rules/line-too-long/
   "FBT",    # https://docs.astral.sh/ruff/rules/#flake8-boolean-trap-fbt
@@ -347,10 +348,12 @@ filterwarnings = [
 project-includes = ["pandas-stubs", "tests"]
 enabled-ignores = ["pyrefly"]
 python-version = "3.11"
+min-severity = "warn"
 
 [tool.pyrefly.errors]
 explicit-any = false  # We do use implicit Any
 implicit-any-type-argument = false  # We do use implicit Any
+implicit-bool = false  # We do use implicit bool
 untyped-import = false  # Import discovery
 
 [tool.ty.environment]

tests/__init__.py

@@ -91,8 +91,10 @@ def check(
     value: Any
     if isinstance(actual, pd.Series):
         # cast is by design microsoft/pyright#11191
+        # pyrefly: ignore[redundant-cast]
         value = cast(pd.Series, actual).iloc[index_to_check_for_type]
     elif isinstance(actual, pd.Index):
+        # pyrefly: ignore[redundant-cast]
         # cast is by design microsoft/pyright#11191
         value = cast(pd.Index, actual)[index_to_check_for_type]
     elif isinstance(actual, BaseGroupBy):