feat: complete ExtensionArrays

pandas-stubs/_libs/arrays.pyi

@@ -0,0 +1,38 @@
+from collections.abc import Sequence
+from typing import Any
+
+import numpy as np
+from typing_extensions import Self
+
+from pandas._typing import (
+    AnyArrayLikeInt,
+    AxisInt,
+    DtypeObj,
+    Shape,
+)
+
+class NDArrayBacked:
+    _dtype: DtypeObj
+    _ndarray: np.ndarray
+    def __setstate__(self, state: Any) -> None: ...
+    def __len__(self) -> int: ...
+    @property
+    def shape(self) -> Shape: ...
+    @property
+    def ndim(self) -> int: ...
+    @property
+    def size(self) -> int: ...
+    @property
+    def nbytes(self) -> int: ...
+    def copy(self, order=...) -> Self: ...
+    def delete(self, loc, axis=...) -> Self: ...
+    def swapaxes(self, axis1, axis2) -> Self: ...
+    def repeat(
+        self,
+        repeats: int | Sequence[int] | AnyArrayLikeInt,
+        axis: AxisInt | None = None,
+    ) -> Self: ...
+    def reshape(self, *args: Any, **kwargs: Any) -> Self: ...
+    def ravel(self, order=...) -> Self: ...
+    @property
+    def T(self) -> Self: ...

pandas-stubs/_typing.pyi

@@ -224,6 +224,8 @@ Dtype: TypeAlias = ExtensionDtype | NpDtype
 # m
 # datetime64
 
+# PyArrow boolean type and its string alias
+PyArrowBooleanDtypeArg: TypeAlias = Literal["bool[pyarrow]", "boolean[pyarrow]"]
 BooleanDtypeArg: TypeAlias = (
     # Builtin bool type and its string alias
     type[bool]  # noqa: PYI030
@@ -235,9 +237,12 @@ BooleanDtypeArg: TypeAlias = (
     # https://numpy.org/doc/stable/reference/arrays.scalars.html#numpy.bool_
     | type[np.bool_]
     | Literal["?", "b1", "bool_"]
-    # PyArrow boolean type and its string alias
-    | Literal["bool[pyarrow]", "boolean[pyarrow]"]
+    | PyArrowBooleanDtypeArg
 )
+# PyArrow integer types and their string aliases
+PyArrowIntDtypeArg: TypeAlias = Literal[
+    "int8[pyarrow]", "int16[pyarrow]", "int32[pyarrow]", "int64[pyarrow]"
+]
 IntDtypeArg: TypeAlias = (
     # Builtin integer type and its string alias
     type[int]  # noqa: PYI030
@@ -267,9 +272,12 @@ IntDtypeArg: TypeAlias = (
     # https://numpy.org/doc/stable/reference/arrays.scalars.html#numpy.intp
     | type[np.intp]  # signed pointer (=`intptr_t`, platform dependent)
     | Literal["p", "intp"]
-    # PyArrow integer types and their string aliases
-    | Literal["int8[pyarrow]", "int16[pyarrow]", "int32[pyarrow]", "int64[pyarrow]"]
+    | PyArrowIntDtypeArg
 )
+# PyArrow unsigned integer types and their string aliases
+PyArrowUIntDtypeArg: TypeAlias = Literal[
+    "uint8[pyarrow]", "uint16[pyarrow]", "uint32[pyarrow]", "uint64[pyarrow]"
+]
 UIntDtypeArg: TypeAlias = (
     # Pandas nullable unsigned integer types and their string aliases
     pd.UInt8Dtype  # noqa: PYI030
@@ -296,9 +304,16 @@ UIntDtypeArg: TypeAlias = (
     # https://numpy.org/doc/stable/reference/arrays.scalars.html#numpy.uintp
     | type[np.uintp]  # unsigned pointer (=`uintptr_t`, platform dependent)
     | Literal["P", "uintp"]
-    # PyArrow unsigned integer types and their string aliases
-    | Literal["uint8[pyarrow]", "uint16[pyarrow]", "uint32[pyarrow]", "uint64[pyarrow]"]
+    | PyArrowUIntDtypeArg
 )
+# PyArrow floating point types and their string aliases
+PyArrowFloatDtypeArg: TypeAlias = Literal[
+    "float[pyarrow]",
+    "double[pyarrow]",
+    "float16[pyarrow]",
+    "float32[pyarrow]",
+    "float64[pyarrow]",
+]
 FloatDtypeArg: TypeAlias = (
     # Builtin float type and its string alias
     type[float]  # noqa: PYI030
@@ -321,14 +336,7 @@ FloatDtypeArg: TypeAlias = (
     # https://numpy.org/doc/stable/reference/arrays.scalars.html#numpy.longdouble
     | type[np.longdouble]
     | Literal["g", "f16", "float128", "longdouble", "longfloat"]
-    # PyArrow floating point types and their string aliases
-    | Literal[
-        "float[pyarrow]",
-        "double[pyarrow]",
-        "float16[pyarrow]",
-        "float32[pyarrow]",
-        "float64[pyarrow]",
-    ]
+    | PyArrowFloatDtypeArg
 )
 ComplexDtypeArg: TypeAlias = (
     # Builtin complex type and its string alias
@@ -353,112 +361,123 @@ ComplexDtypeArg: TypeAlias = (
         "longcomplex",
     ]
 )
-# Refer to https://numpy.org/doc/stable/reference/arrays.datetime.html#datetime-units
-TimedeltaDtypeArg: TypeAlias = Literal[
-    "timedelta64[Y]",
-    "timedelta64[M]",
-    "timedelta64[W]",
-    "timedelta64[D]",
-    "timedelta64[h]",
-    "timedelta64[m]",
-    "timedelta64[s]",
-    "timedelta64[ms]",
-    "timedelta64[us]",
-    "timedelta64[μs]",
-    "timedelta64[ns]",
-    "timedelta64[ps]",
-    "timedelta64[fs]",
-    "timedelta64[as]",
-    # numpy type codes
-    "m8[Y]",
-    "m8[M]",
-    "m8[W]",
-    "m8[D]",
-    "m8[h]",
-    "m8[m]",
-    "m8[s]",
-    "m8[ms]",
-    "m8[us]",
-    "m8[μs]",
-    "m8[ns]",
-    "m8[ps]",
-    "m8[fs]",
-    "m8[as]",
-    # little endian
-    "<m8[Y]",
-    "<m8[M]",
-    "<m8[W]",
-    "<m8[D]",
-    "<m8[h]",
-    "<m8[m]",
-    "<m8[s]",
-    "<m8[ms]",
-    "<m8[us]",
-    "<m8[μs]",
-    "<m8[ns]",
-    "<m8[ps]",
-    "<m8[fs]",
-    "<m8[as]",
-    # PyArrow duration type and its string alias
+# PyArrow duration type and its string alias
+PyArrowTimedeltaDtypeArg: TypeAlias = Literal[
     "duration[s][pyarrow]",
     "duration[ms][pyarrow]",
     "duration[us][pyarrow]",
     "duration[ns][pyarrow]",
 ]
-TimestampDtypeArg: TypeAlias = Literal[
-    "datetime64[Y]",
-    "datetime64[M]",
-    "datetime64[W]",
-    "datetime64[D]",
-    "datetime64[h]",
-    "datetime64[m]",
-    "datetime64[s]",
-    "datetime64[ms]",
-    "datetime64[us]",
-    "datetime64[μs]",
-    "datetime64[ns]",
-    "datetime64[ps]",
-    "datetime64[fs]",
-    "datetime64[as]",
-    # numpy type codes
-    "M8[Y]",
-    "M8[M]",
-    "M8[W]",
-    "M8[D]",
-    "M8[h]",
-    "M8[m]",
-    "M8[s]",
-    "M8[ms]",
-    "M8[us]",
-    "M8[μs]",
-    "M8[ns]",
-    "M8[ps]",
-    "M8[fs]",
-    "M8[as]",
-    # little endian
-    "<M8[Y]",
-    "<M8[M]",
-    "<M8[W]",
-    "<M8[D]",
-    "<M8[h]",
-    "<M8[m]",
-    "<M8[s]",
-    "<M8[ms]",
-    "<M8[us]",
-    "<M8[μs]",
-    "<M8[ns]",
-    "<M8[ps]",
-    "<M8[fs]",
-    "<M8[as]",
-    # PyArrow timestamp type and its string alias
+# Refer to https://numpy.org/doc/stable/reference/arrays.datetime.html#datetime-units
+TimedeltaDtypeArg: TypeAlias = (
+    Literal[
+        "timedelta64[Y]",
+        "timedelta64[M]",
+        "timedelta64[W]",
+        "timedelta64[D]",
+        "timedelta64[h]",
+        "timedelta64[m]",
+        "timedelta64[s]",
+        "timedelta64[ms]",
+        "timedelta64[us]",
+        "timedelta64[μs]",
+        "timedelta64[ns]",
+        "timedelta64[ps]",
+        "timedelta64[fs]",
+        "timedelta64[as]",
+        # numpy type codes
+        "m8[Y]",
+        "m8[M]",
+        "m8[W]",
+        "m8[D]",
+        "m8[h]",
+        "m8[m]",
+        "m8[s]",
+        "m8[ms]",
+        "m8[us]",
+        "m8[μs]",
+        "m8[ns]",
+        "m8[ps]",
+        "m8[fs]",
+        "m8[as]",
+        # little endian
+        "<m8[Y]",
+        "<m8[M]",
+        "<m8[W]",
+        "<m8[D]",
+        "<m8[h]",
+        "<m8[m]",
+        "<m8[s]",
+        "<m8[ms]",
+        "<m8[us]",
+        "<m8[μs]",
+        "<m8[ns]",
+        "<m8[ps]",
+        "<m8[fs]",
+        "<m8[as]",
+    ]
+    | PyArrowTimedeltaDtypeArg
+)
+# PyArrow timestamp type and its string alias
+PyArrowTimestampDtypeArg: TypeAlias = Literal[
     "date32[pyarrow]",
     "date64[pyarrow]",
     "timestamp[s][pyarrow]",
     "timestamp[ms][pyarrow]",
     "timestamp[us][pyarrow]",
     "timestamp[ns][pyarrow]",
 ]
-
+TimestampDtypeArg: TypeAlias = (
+    Literal[
+        "datetime64[Y]",
+        "datetime64[M]",
+        "datetime64[W]",
+        "datetime64[D]",
+        "datetime64[h]",
+        "datetime64[m]",
+        "datetime64[s]",
+        "datetime64[ms]",
+        "datetime64[us]",
+        "datetime64[μs]",
+        "datetime64[ns]",
+        "datetime64[ps]",
+        "datetime64[fs]",
+        "datetime64[as]",
+        # numpy type codes
+        "M8[Y]",
+        "M8[M]",
+        "M8[W]",
+        "M8[D]",
+        "M8[h]",
+        "M8[m]",
+        "M8[s]",
+        "M8[ms]",
+        "M8[us]",
+        "M8[μs]",
+        "M8[ns]",
+        "M8[ps]",
+        "M8[fs]",
+        "M8[as]",
+        # little endian
+        "<M8[Y]",
+        "<M8[M]",
+        "<M8[W]",
+        "<M8[D]",
+        "<M8[h]",
+        "<M8[m]",
+        "<M8[s]",
+        "<M8[ms]",
+        "<M8[us]",
+        "<M8[μs]",
+        "<M8[ns]",
+        "<M8[ps]",
+        "<M8[fs]",
+        "<M8[as]",
+    ]
+    | PyArrowTimestampDtypeArg
+)
+# PyArrow string type and its string alias
+PyArrowStrDtypeArg: TypeAlias = Literal["string[pyarrow]"]
 StrDtypeArg: TypeAlias = (
     # Builtin str type and its string alias
     type[str]  # noqa: PYI030
@@ -470,9 +489,10 @@ StrDtypeArg: TypeAlias = (
     # https://numpy.org/doc/stable/reference/arrays.scalars.html#numpy.str_
     | type[np.str_]
     | Literal["U", "str_", "str0", "unicode", "unicode_"]
-    # PyArrow string type and its string alias
-    | Literal["string[pyarrow]"]
+    | PyArrowStrDtypeArg
 )
+# PyArrow binary type and its string alias
+PyArrowBytesDtypeArg: TypeAlias = Literal["binary[pyarrow]"]
 BytesDtypeArg: TypeAlias = (
     # Builtin bytes type and its string alias
     type[bytes]  # noqa: PYI030
@@ -481,8 +501,7 @@ BytesDtypeArg: TypeAlias = (
     # https://numpy.org/doc/stable/reference/arrays.scalars.html#numpy.bytes_
     | type[np.bytes_]
     | Literal["S", "bytes_", "bytes0", "string_"]
-    # PyArrow binary type and its string alias
-    | Literal["binary[pyarrow]"]
+    | PyArrowBytesDtypeArg
 )
 CategoryDtypeArg: TypeAlias = CategoricalDtype | Literal["category"]
 

pandas-stubs/arrays/__init__.pyi

@@ -1,12 +1,31 @@
 from pandas.core.arrays import (
-    BooleanArray as BooleanArray,
-    Categorical as Categorical,
-    DatetimeArray as DatetimeArray,
-    IntegerArray as IntegerArray,
-    IntervalArray as IntervalArray,
-    PandasArray as PandasArray,
-    PeriodArray as PeriodArray,
-    SparseArray as SparseArray,
-    StringArray as StringArray,
-    TimedeltaArray as TimedeltaArray,
+    ArrowExtensionArray,
+    ArrowStringArray,
+    BooleanArray,
+    Categorical,
+    DatetimeArray,
+    FloatingArray,
+    IntegerArray,
+    IntervalArray,
+    NumpyExtensionArray,
+    PeriodArray,
+    SparseArray,
+    StringArray,
+    TimedeltaArray,
 )
+
+__all__ = [
+    "ArrowExtensionArray",
+    "ArrowStringArray",
+    "BooleanArray",
+    "Categorical",
+    "DatetimeArray",
+    "FloatingArray",
+    "IntegerArray",
+    "IntervalArray",
+    "NumpyExtensionArray",
+    "PeriodArray",
+    "SparseArray",
+    "StringArray",
+    "TimedeltaArray",
+]