gh-122982: Prohibit bitwise inversion on bool type

[Eclips4]

• Issue: Extend deprecation warning period for bitwise inversion on bool type #122982

---

📚 Documentation preview 📚: https://cpython-previews--122983.org.readthedocs.build/

[sobolevn]

Thanks!

[picnixz]

Apart from Nikita's comment, this looks good to me now (TIL I cannot do class A(bool): ...).

[encukou]

If you do this, you should also want to edit Doc/deprecations/pending-removal-in-future.rst.
But note that it's not in Doc/deprecations/pending-removal-in-3.14.rst; removing this feature now would be a surprise for some.

[bedevere-app]

A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated.

Once you have made the requested changes, please leave a comment on this pull request containing the phrase I have made the requested changes; please review again. I will then notify any core developers who have left a review that you're ready for them to take another look at this pull request.

[Eclips4]

I have made the requested changes; please review again

[bedevere-app]

Thanks for making the requested changes!

@encukou, @sobolevn: please review the changes made to this pull request.

[AA-Turner]

If this gets merged, you should open a new PR to move this note to the 3.14 pending removals & backport that.

[serhiy-storchaka]

Just wondering: what if simply set PyBool_Type.tp_as_number->nb_invert = NULL in _PyTypes_InitTypes()?

[Eclips4]

Just wondering: what if simply set PyBool_Type.tp_as_number->nb_invert = NULL in _PyTypes_InitTypes()?

Yeah, this works. Do you think it's worth doing or should we continue with the current approach?

[serhiy-storchaka]

I do not know. How does True.__inv__() work in this case?

On one hand, using the same code to generate error messages guarantees that there is no difference in error messages. On other hand, we may want to have a different error message, for example add a suggestion about not.

[Eclips4]

I do not know. How does True.__inv__() work in this case?

bool.__invert__ would work in this case:

Python 3.14.0a0 (heads/remove_bool_invert-dirty:4760de8ecbb, Aug 16 2024, 09:24:28) [Clang 15.0.0 (clang-1500.3.9.4)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> True.__invert__()
-2

So, this is a wrong approach.

On other hand, we may want to have a different error message, for example add a suggestion about not.

That's sounds good. I'll add a suggestion and tests for it.

pending-removal-in-future.rst is fixed

[sobolevn]

Thank you! I don't click merge, because @encukou's comment here: #122983 (review) seems like it is better to wait for resolution.

[AA-Turner]

Wordsmithing: We seem to use "Maybe you meant" for SyntaxErrors, with "Did you mean" used more generally. Non-blocking suggestion though.

A

[hauntsaninja]

I mentioned this on the issue, but I think in this case we should extend the deprecation period beyond the minimum #122982 (comment)

[hauntsaninja]

I think the text in the warning was helpful, especially since the previous behaviour has been around forever.

Maybe something like:

bad operand type for unary ~: 'bool'. Did you mean 'not' instead of '~'? Use 'not x' for boolean negation and '~int(x)` in the rare case you want bitwise inversion of the underlying int"

[Eclips4]

Closing per #122982 (comment)