Commit Graph

10 Commits

Author SHA1 Message Date
Alex Waygood 3a8191529c
[ty] Exclude members starting with `_abc_` from a protocol interface (#18467)
## Summary

As well as excluding a hardcoded set of special attributes, CPython at
runtime also excludes any attributes or declarations starting with
`_abc_` from the set of members that make up a protocol interface. I
missed this in my initial implementation.

This is a bit of a CPython implementation detail, but I do think it's
important that we try to model the runtime as best we can here. The
closer we are to the runtime behaviour, the closer we come to sound
behaviour when narrowing types from `isinstance()` checks against
runtime-checkable protocols (for example)

## Test Plan

Extended an existing mdtest
2025-06-04 20:34:09 +01:00
Alex Waygood 0a11baf29c
[ty] Implement implicit inheritance from `Generic[]` for PEP-695 generic classes (#18283) 2025-05-26 20:40:16 +01:00
Alex Waygood e293411679
[ty] `get_protocol_members` returns a frozenset, not a tuple (#18284) 2025-05-23 23:20:34 +00:00
Alex Waygood d02c9ada5d
[ty] Do not carry the generic context of `Protocol` or `Generic` in the `ClassBase` enum (#17989)
## Summary

It doesn't seem to be necessary for our generics implementation to carry
the `GenericContext` in the `ClassBase` variants. Removing it simplifies
the code, fixes many TODOs about `Generic` or `Protocol` appearing
multiple times in MROs when each should only appear at most once, and
allows us to more accurately detect runtime errors that occur due to
`Generic` or `Protocol` appearing multiple times in a class's bases.

In order to remove the `GenericContext` from the `ClassBase` variant, it
turns out to be necessary to emulate
`typing._GenericAlias.__mro_entries__`, or we end up with a large number
of false-positive `inconsistent-mro` errors. This PR therefore also does
that.

Lastly, this PR fixes the inferred MROs of PEP-695 generic classes,
which implicitly inherit from `Generic` even if they have no explicit
bases.

## Test Plan

mdtests
2025-05-22 21:37:03 -04:00
Alex Waygood d1bb10a66b
[ty] Understand classes that inherit from subscripted `Protocol[]` as generic (#17832) 2025-05-09 17:39:15 +01:00
David Peter 642eac452d
[ty] Recursive protocols (#17929)
## Summary

Use a self-reference "marker" ~~and fixpoint iteration~~ to solve the
stack overflow problems with recursive protocols. This is not pretty and
somewhat tedious, but seems to work fine. Much better than all my
fixpoint-iteration attempts anyway.

closes https://github.com/astral-sh/ty/issues/93

## Test Plan

New Markdown tests.
2025-05-09 14:54:02 +02:00
Alex Waygood 9b694ada82
[ty] Report duplicate `Protocol` or `Generic` base classes with `[duplicate-base]`, not `[inconsistent-mro]` (#17971) 2025-05-08 23:41:22 +01:00
David Peter 04457f99b6
[ty] Protocols: Fixpoint iteration for fully-static check (#17880)
## Summary

A recursive protocol like the following would previously lead to stack
overflows when attempting to create the union type for the `P | None`
member, because `UnionBuilder` checks if element types are fully static,
and the fully-static check on `P` would in turn list all members and
check whether all of them were fully static, leading to a cycle.

```py
from __future__ import annotations

from typing import Protocol

class P(Protocol):
    parent: P | None
```

Here, we make the fully-static check on protocols a salsa query and add
fixpoint iteration, starting with `true` as the initial value (assume
that the recursive protocol is fully-static). If the recursive protocol
has any non-fully-static members, we still return `false` when
re-executing the query (see newly added tests).

closes #17861

## Test Plan

Added regression test
2025-05-07 08:55:21 +02:00
Charlie Marsh a2e9a7732a
Update class literal display to use `<class 'Foo'>` style (#17889)
## Summary

Closes https://github.com/astral-sh/ruff/issues/17238.
2025-05-06 20:11:25 -04:00
Micha Reiser b51c4f82ea
Rename Red Knot (#17820) 2025-05-03 19:49:15 +02:00