Revert "Let SupportsDunderLT return SupportsBool instead of `bool…

[srittau]

As @JelleZijlstra pointed out, this is basically equivalent to returning object (since all objects implement __bool__). So I'm reverting this.

[github-actions]

Diff from mypy_primer, showing the effect of this PR on open source code:

colour (https://github.com/colour-science/colour)
+ colour/plotting/colorimetry.py:565: error: Value of type variable "SupportsRichComparisonT" of "min" cannot be "floating[_16Bit] | floating[_32Bit] | float64"  [type-var]
+ colour/plotting/colorimetry.py:566: error: Value of type variable "SupportsRichComparisonT" of "max" cannot be "floating[_16Bit] | floating[_32Bit] | float64"  [type-var]

jax (https://github.com/google/jax)
- jax/_src/pallas/mosaic/pipeline.py:1546: error: Unused "type: ignore" comment  [unused-ignore]

ibis (https://github.com/ibis-project/ibis)
+ ibis/expr/types/relations.py:2986: error: Value of type variable "SupportsRichComparisonT" of "sorted" cannot be "NumericValue | float"  [type-var]

arviz (https://github.com/arviz-devs/arviz)
+ arviz/stats/ecdf_utils.py:72: error: Value of type variable "SupportsRichComparisonT" of "min" cannot be "Any | float64"  [type-var]

[jorenham]

As @JelleZijlstra pointed out, this is basically equivalent to returning object (since all objects implement __bool__).

It isn't though:

from typing import Protocol, final

@final
class NotTruthy:
    __bool__ = None

class SupportsBool(Protocol):
    def __bool__(self, /) -> bool: ...

x: SupportsBool = NotTruthy()  # rejected

basedpyright playground

[srittau]

@jorenham I don't think that's currently an officially supported pattern – opposed to __hash__ = None. Neither is mentioned in the typing spec, though, but typeshed does not contain a single instance of __bool__ = None, yet. That said, I'd definitely be in favor of that pattern, but this should be a wider discussion on discuss.python.org, and then be added to the spec (together with __hash__). (I seem to remember that I mentioned something like this in previous discuss threads, but I'm not sure.) Then we could revisit this PR.

[AlexWaygood]

The official data model states:

Setting a special method to None indicates that the corresponding operation is not available. For example, if a class sets __iter__() to None, the class is not iterable, so calling iter() on its instances will raise a TypeError (without falling back to __getitem__()).

See also https://mail.python.org/archives/list/[email protected]/thread/YGAK34DRWJFSIV2VZ4NC2J24XO37GCMM/

[jorenham]

@jorenham I don't think that's currently an officially supported pattern – opposed to __hash__ = None. Neither is mentioned in the typing spec, though, but typeshed does not contain a single instance of __bool__ = None, yet. That said, I'd definitely be in favor of that pattern, but this should be a wider discussion on discuss.python.org, and then be added to the spec (together with __hash__). (I seem to remember that I mentioned something like this in previous discuss threads, but I'm not sure.) Then we could revisit this PR.

Ok fair enough. But then what about the fact that object.__bool__ doesn't exist?

typeshed/stdlib/builtins.pyi

Lines 106 to 139 in 213e696

	class object:
	__doc__: str | None
	__dict__: dict[str, Any]
	__module__: str
	__annotations__: dict[str, Any]
	@property
	def __class__(self) -> type[Self]: ...
	@__class__.setter
	def __class__(self, type: type[Self], /) -> None: ...
	def __init__(self) -> None: ...
	def __new__(cls) -> Self: ...
	# N.B. `object.__setattr__` and `object.__delattr__` are heavily special-cased by type checkers.
	# Overriding them in subclasses has different semantics, even if the override has an identical signature.
	def __setattr__(self, name: str, value: Any, /) -> None: ...
	def __delattr__(self, name: str, /) -> None: ...
	def __eq__(self, value: object, /) -> bool: ...
	def __ne__(self, value: object, /) -> bool: ...
	def __str__(self) -> str: ... # noqa: Y029
	def __repr__(self) -> str: ... # noqa: Y029
	def __hash__(self) -> int: ...
	def __format__(self, format_spec: str, /) -> str: ...
	def __getattribute__(self, name: str, /) -> Any: ...
	def __sizeof__(self) -> int: ...
	# return type of pickle methods is rather hard to express in the current type system
	# see #6661 and https://docs.python.org/3/library/pickle.html#object.__reduce__
	def __reduce__(self) -> str | tuple[Any, ...]: ...
	def __reduce_ex__(self, protocol: SupportsIndex, /) -> str | tuple[Any, ...]: ...
	if sys.version_info >= (3, 11):
	def __getstate__(self) -> object: ...
	def __dir__(self) -> Iterable[str]: ...
	def __init_subclass__(cls) -> None: ...
	@classmethod
	def __subclasshook__(cls, subclass: type, /) -> bool: ...

>>> object.__bool__
Traceback (most recent call last):
  File "<python-input-1>", line 1, in <module>
    object.__bool__
AttributeError: type object 'object' has no attribute '__bool__'. Did you mean: '__doc__'?

[srittau]

Ok, makes sense to me. So let's re-apply the original PR and start using bool = None in typeshed where applicable?

[jorenham]

and start using bool = None in typeshed where applicable?

That should only be needed if a base class implements __bool__:

from typing import Protocol

class SupportsBool(Protocol):
    def __bool__(self, /) -> bool: ...

x: SupportsBool = object()  # rejected

basedpyright playground

[randolf-scholz]

@srittau I already reopened the PR with an additional unit test that checks that plain object is not accepted at type checking time (despite working at runtime); see #14375.