Skip to content

TYP: QuadPrecision <: numpy.floating #218

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Nov 1, 2025
Merged

TYP: QuadPrecision <: numpy.floating #218

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e5c49a0
TYP: `QuadPrecision <: numpy.floating`
jorenham Oct 27, 2025
f12e82c
TYP: fix typos in stub comments
jorenham Oct 28, 2025
29c379d
Merge branch 'main' into typing/QuadPrecision-floating-subtype
SwayamInSync Oct 30, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
115 changes: 75 additions & 40 deletions quaddtype/numpy_quaddtype/_quaddtype_main.pyi
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,7 @@
from typing import Any, Literal, TypeAlias, final, overload

import numpy as np
from numpy._typing import _128Bit # pyright: ignore[reportPrivateUsage]
from typing_extensions import Never, Self, override

_Backend: TypeAlias = Literal["sleef", "longdouble"]
Expand All @@ -12,9 +13,10 @@ _IntoQuad: TypeAlias = (
| np.integer[Any]
| np.bool_
) # fmt: skip
_ScalarItemArg: TypeAlias = Literal[0, -1] | tuple[Literal[0, -1]] | tuple[()]

@final
class QuadPrecDType(np.dtype[QuadPrecision]): # type: ignore[misc, type-var] # pyright: ignore[reportGeneralTypeIssues, reportInvalidTypeArguments]
class QuadPrecDType(np.dtype[QuadPrecision]): # type: ignore[misc] # pyright: ignore[reportGeneralTypeIssues]
def __new__(cls, /, backend: _Backend = "sleef") -> Self: ...

# `numpy.dtype` overrides
Expand Down Expand Up @@ -70,61 +72,94 @@ class QuadPrecDType(np.dtype[QuadPrecision]): # type: ignore[misc, type-var] #
@override
def __getitem__(self, key: Never, /) -> Self: ... # type: ignore[override]

# NOTE: Until `QuadPrecision` will become a subclass of `np.generic`, this class cannot
# be considered "type-safe".
@final
class QuadPrecision:
class QuadPrecision(np.floating[_128Bit]):
# NOTE: At runtime this constructor also accepts array-likes, for which it returns
# `np.ndarray` instances with `dtype=QuadPrecDType()`.
# But because of mypy limitations, it is currently impossible to annotate
# constructors that do no return instances of their class (or a subclass thereof).
# constructors that do not return instances of their class (or a subclass thereof).
# See https://github.com/python/mypy/issues/18343#issuecomment-2571784915
@override
def __new__(cls, /, value: _IntoQuad, backend: _Backend = "sleef") -> Self: ...

# Attributes
# numpy.floating property overrides

@property
@override
def dtype(self) -> QuadPrecDType: ...
@property
@override
def real(self) -> Self: ...
@property
@override
def imag(self) -> Self: ...

# Rich comparison operators
# NOTE: Unlike other numpy scalars, these return `builtins.bool`, not `np.bool`.
# numpy.floating method overrides

@override
def item(self, arg0: _ScalarItemArg = ..., /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def tolist(self, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]

# Equality operators
@override
def __eq__(self, other: object, /) -> bool: ...
@override
def __ne__(self, other: object, /) -> bool: ...
def __lt__(self, other: _IntoQuad, /) -> bool: ...
def __le__(self, other: _IntoQuad, /) -> bool: ...
def __gt__(self, other: _IntoQuad, /) -> bool: ...
def __ge__(self, other: _IntoQuad, /) -> bool: ...

# Binary operators
def __add__(self, other: _IntoQuad, /) -> Self: ...
def __radd__(self, other: _IntoQuad, /) -> Self: ...
def __sub__(self, other: _IntoQuad, /) -> Self: ...
def __rsub__(self, other: _IntoQuad, /) -> Self: ...
def __mul__(self, other: _IntoQuad, /) -> Self: ...
def __rmul__(self, other: _IntoQuad, /) -> Self: ...
def __pow__(self, other: _IntoQuad, mod: None = None, /) -> Self: ...
def __rpow__(self, other: _IntoQuad, mod: None = None, /) -> Self: ...
def __truediv__(self, other: _IntoQuad, /) -> Self: ...
def __rtruediv__(self, other: _IntoQuad, /) -> Self: ...

# Unary operators
def __neg__(self, /) -> Self: ...
def __pos__(self, /) -> Self: ...
def __abs__(self, /) -> Self: ...

# Conversion methods
def __bool__(self, /) -> bool: ...
def __int__(self, /) -> int: ...
def __float__(self, /) -> float: ...

# String representation
@override
def __repr__(self, /) -> str: ...
@override
def __str__(self, /) -> str: ...

# Rich comparison operators
# NOTE: Unlike other numpy scalars, these return `builtins.bool`, not `np.bool`.
@override
def __lt__(self, other: _IntoQuad, /) -> bool: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __le__(self, other: _IntoQuad, /) -> bool: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __gt__(self, other: _IntoQuad, /) -> bool: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __ge__(self, other: _IntoQuad, /) -> bool: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]

# Binary arithmetic operators
@override
def __add__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __radd__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __sub__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __rsub__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __mul__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __rmul__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __pow__(self, other: _IntoQuad, mod: None = None, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __rpow__(self, other: _IntoQuad, mod: None = None, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __truediv__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]
@override
def __rtruediv__(self, other: _IntoQuad, /) -> Self: ... # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride]

# Binary modulo operators
@override
def __floordiv__(self, other: _IntoQuad, /) -> Self: ...
@override
def __rfloordiv__(self, other: _IntoQuad, /) -> Self: ...
@override
def __mod__(self, other: _IntoQuad, /) -> Self: ...
@override
def __rmod__(self, other: _IntoQuad, /) -> Self: ...
@override
def __divmod__(self, other: _IntoQuad, /) -> tuple[Self, Self]: ...
@override
def __rdivmod__(self, other: _IntoQuad, /) -> tuple[Self, Self]: ...

# NOTE: is_integer() and as_integer_ratio() are defined on numpy.floating in the
# stubs, but don't exist at runtime. And because QuadPrecision does not implement
# them, we use this hacky workaround to emulate their absence.
# TODO: Remove after https://github.com/numpy/numpy-user-dtypes/issues/216
is_integer: Never # pyright: ignore[reportIncompatibleMethodOverride]
as_integer_ratio: Never # pyright: ignore[reportIncompatibleMethodOverride]

#
def is_longdouble_128() -> bool: ...
Expand Down