gh-106016: Support customizing of module attributes access with __setattr__/__delattr__ (PEP 726)

[skirpichev]

I'm not sure if this does require a PEP (e.g. Brett Cannon on d.p.o discussion thread told me that it "not necessarily"). If so, I've a draft PEP here: https://github.com/skirpichev/peps/tree/setdelattr (And I'm looking for feedback and a sponsor.)

@vstinner

References

• discussion on d.p.o (the current pr extends the original proposal with the __delattr__ method.)
• Fast attribute access for module subclasses #103951

• Issue: PEP 726: Support defining __setattr__() and __delattr__() in a module #106016

---

📚 Documentation preview 📚: https://cpython-previews--108261.org.readthedocs.build/

[bedevere-bot]

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.

[skirpichev]

@vstinner, I have made the requested changes; please review again. But please see remarks about PyDict_GetItemRef() & PEP 562 test layout above.

[bedevere-bot]

Thanks for making the requested changes!

@vstinner: please review the changes made to this pull request.

[vstinner]

I created PR #108293 to add Lib/test/test_module/ directory. Once it will be merged, you can rebased your PR on top of it.

[vstinner]

Can you please rebase your PR on the main branch to get my #108293 change? (New Lib/test/test_module/ directory.)

[skirpichev]

Can you please rebase your PR on the main branch to get my #108293 change?

Yep. Sorry, first I did merge. (BTW, after 7fe6a06 - setattr/delattr declarations should be moved up again I think.)

PR was updated

[vstinner]

With proposed fixed, ./python -m test test_module -R 3:3 pass (there is no reference/memory leak).

[vstinner]

LGTM!

[skirpichev]

@vstinner, wait! Are you sure this pr doesn't require a PEP?

[AA-Turner]

We have a PEP for adding only one attribute (__call__; PEP-713) -- this is a much larger proposal and as such I believe should go through the PEP process, given 713's precedent.

A

[skirpichev]

With proposed fixed, ./python -m test test_module -R 3:3 pass (there is no reference/memory leak).

This seems to be broken on the main branch as well.

We have a PEP for adding only one attribute (call; PEP-713)

And PEP 562 (two attributes). Fair enough.

I believe should go through the PEP process

I've a draft PEP: https://raw.githubusercontent.com/skirpichev/peps/7cc52e7/pep-9999.rst
It lacks a sponsor so far.

[AA-Turner]

It lacks a sponsor so far.

I'm happy to sponsor if you'd like? Adam Turner <python at quite.org.uk>.

A

[skirpichev]

On Wed, Aug 23, 2023 at 06:21:16PM -0700, Adam Turner wrote: I'm happy to sponsor

Ok. I'll prepare a PR today. UPD: python/peps#3301

[vstinner]

Update: PEP 726 – Module __setattr__ and __delattr__ was written and is being discussed in the PEP section.

[terryjreedy]

Even though Victor closed #106016, I think that this can remain open until the SC decides on the PEP. If it says no, close this. If yes, open a new issue and retarget this by editing the title with the new number. And edit further if required.

[skirpichev]

If yes, open a new issue and retarget this by editing the title with the new number.

Not sure I got idea of this bureaucracy:) If it's required to have an issue for a pr (as it seems) - we already have an associated issue, #106016.

[skirpichev]

Merge conflict was fixed & added the reference to pep in docs.

[skirpichev]

Thanks for review and sponsoring the pep.