## Summary
Closes: https://github.com/astral-sh/ty/issues/157
This PR adds support for the following capabilities involving a
`ParamSpec` type variable:
- Representing `P.args` and `P.kwargs` in the type system
- Matching against a callable containing `P` to create a type mapping
- Specializing `P` against the stored parameters
The value of a `ParamSpec` type variable is being represented using
`CallableType` with a `CallableTypeKind::ParamSpecValue` variant. This
`CallableTypeKind` is expanded from the existing `is_function_like`
boolean flag. An `enum` is used as these variants are mutually
exclusive.
For context, an initial iteration made an attempt to expand the
`Specialization` to use `TypeOrParameters` enum that represents that a
type variable can specialize into either a `Type` or `Parameters` but
that increased the complexity of the code as all downstream usages would
need to handle both the variants appropriately. Additionally, we'd have
also need to establish an invariant that a regular type variable always
maps to a `Type` while a paramspec type variable always maps to a
`Parameters`.
I've intentionally left out checking and raising diagnostics when the
`ParamSpec` type variable and it's components are not being used
correctly to avoid scope increase and it can easily be done as a
follow-up. This would also include the scoping rules which I don't think
a regular type variable implements either.
## Test Plan
Add new mdtest cases and update existing test cases.
Ran this branch on pyx, no new diagnostics.
### Ecosystem analysis
There's a case where in an annotated assignment like:
```py
type CustomType[P] = Callable[...]
def value[**P](...): ...
def another[**P](...):
target: CustomType[P] = value
```
The type of `value` is a callable and it has a paramspec that's bound to
`value`, `CustomType` is a type alias that's a callable and `P` that's
used in it's specialization is bound to `another`. Now, ty infers the
type of `target` same as `value` and does not use the declared type
`CustomType[P]`. [This is the
assignment](0980b9d9ab/src/async_utils/gen_transform.py (L108))
that I'm referring to which then leads to error in downstream usage.
Pyright and mypy does seem to use the declared type.
There are multiple diagnostics in `dd-trace-py` that requires support
for `cls`.
I'm seeing `Divergent` type for an example like which ~~I'm not sure
why, I'll look into it tomorrow~~ is because of a cycle as mentioned in
https://github.com/astral-sh/ty/issues/1729#issuecomment-3612279974:
```py
from typing import Callable
def decorator[**P](c: Callable[P, int]) -> Callable[P, str]: ...
@decorator
def func(a: int) -> int: ...
# ((a: int) -> str) | ((a: Divergent) -> str)
reveal_type(func)
```
I ~~need to look into why are the parameters not being specialized
through multiple decorators in the following code~~ think this is also
because of the cycle mentioned in
https://github.com/astral-sh/ty/issues/1729#issuecomment-3612279974 and
the fact that we don't support `staticmethod` properly:
```py
from contextlib import contextmanager
class Foo:
@staticmethod
@contextmanager
def method(x: int):
yield
foo = Foo()
# ty: Revealed type: `() -> _GeneratorContextManager[Unknown, None, None]` [revealed-type]
reveal_type(foo.method)
```
There's some issue related to `Protocol` that are generic over a
`ParamSpec` in `starlette` which might be related to
https://github.com/astral-sh/ty/issues/1635 but I'm not sure. Here's a
minimal example to reproduce:
<details><summary>Code snippet:</summary>
<p>
```py
from collections.abc import Awaitable, Callable, MutableMapping
from typing import Any, Callable, ParamSpec, Protocol
P = ParamSpec("P")
Scope = MutableMapping[str, Any]
Message = MutableMapping[str, Any]
Receive = Callable[[], Awaitable[Message]]
Send = Callable[[Message], Awaitable[None]]
ASGIApp = Callable[[Scope, Receive, Send], Awaitable[None]]
_Scope = Any
_Receive = Callable[[], Awaitable[Any]]
_Send = Callable[[Any], Awaitable[None]]
# Since `starlette.types.ASGIApp` type differs from `ASGIApplication` from `asgiref`
# we need to define a more permissive version of ASGIApp that doesn't cause type errors.
_ASGIApp = Callable[[_Scope, _Receive, _Send], Awaitable[None]]
class _MiddlewareFactory(Protocol[P]):
def __call__(
self, app: _ASGIApp, *args: P.args, **kwargs: P.kwargs
) -> _ASGIApp: ...
class Middleware:
def __init__(
self, factory: _MiddlewareFactory[P], *args: P.args, **kwargs: P.kwargs
) -> None:
self.factory = factory
self.args = args
self.kwargs = kwargs
class ServerErrorMiddleware:
def __init__(
self,
app: ASGIApp,
value: int | None = None,
flag: bool = False,
) -> None:
self.app = app
self.value = value
self.flag = flag
async def __call__(self, scope: Scope, receive: Receive, send: Send) -> None: ...
# ty: Argument to bound method `__init__` is incorrect: Expected `_MiddlewareFactory[(...)]`, found `<class 'ServerErrorMiddleware'>` [invalid-argument-type]
Middleware(ServerErrorMiddleware, value=500, flag=True)
```
</p>
</details>
### Conformance analysis
> ```diff
> -constructors_callable.py:36:13: info[revealed-type] Revealed type:
`(...) -> Unknown`
> +constructors_callable.py:36:13: info[revealed-type] Revealed type:
`(x: int) -> Unknown`
> ```
Requires return type inference i.e.,
https://github.com/astral-sh/ruff/pull/21551
> ```diff
> +constructors_callable.py:194:16: error[invalid-argument-type]
Argument is incorrect: Expected `list[T@__init__]`, found `list[Unknown
| str]`
> +constructors_callable.py:194:22: error[invalid-argument-type]
Argument is incorrect: Expected `list[T@__init__]`, found `list[Unknown
| str]`
> +constructors_callable.py:195:4: error[invalid-argument-type] Argument
is incorrect: Expected `list[T@__init__]`, found `list[Unknown | int]`
> +constructors_callable.py:195:9: error[invalid-argument-type] Argument
is incorrect: Expected `list[T@__init__]`, found `list[Unknown | str]`
> ```
I might need to look into why this is happening...
> ```diff
> +generics_defaults.py:79:1: error[type-assertion-failure] Type
`type[Class_ParamSpec[(str, int, /)]]` does not match asserted type
`<class 'Class_ParamSpec'>`
> ```
which is on the following code
```py
DefaultP = ParamSpec("DefaultP", default=[str, int])
class Class_ParamSpec(Generic[DefaultP]): ...
assert_type(Class_ParamSpec, type[Class_ParamSpec[str, int]])
```
It's occurring because there's no equivalence relationship defined
between `ClassLiteral` and `KnownInstanceType::TypeGenericAlias` which
is what these types are.
Everything else looks good to me!
When converting a class (whether specialized or not) into a `Callable`
type, we should carry through any generic context that the constructor
has. This includes both the generic context of the class itself (if it's
generic) and of the constructor methods (if they are separately
generic).
To help test this, this also updates the `generic_context` extension
function to work on `Callable` types and unions; and adds a new
`into_callable` extension function that works just like
`CallableTypeOf`, but on value forms instead of type forms.
Pulled this out of #21551 for separate review.
## Summary
This is another small refactor for
https://github.com/astral-sh/ruff/pull/21445 that splits the single
`paramspec.md` into `generics/legacy/paramspec.md` and
`generics/pep695/paramspec.md`.
## Test Plan
Make sure that all mdtests pass.
## Summary
Previously if an explicit specialization failed (e.g. wrong number of
type arguments or violates an upper bound) we just inferred `Unknown`
for the entire type. This actually caused us to panic on an a case of a
recursive upper bound with invalid specialization; the upper bound would
oscillate indefinitely in fixpoint iteration between `Unknown` and the
given specialization. This could be fixed with a cycle recovery
function, but in this case there's a simpler fix: if we infer
`C[Unknown]` instead of `Unknown` for an invalid attempt to specialize
`C`, that allows fixpoint iteration to quickly converge, as well as
giving a more precise type inference.
Other type checkers actually just go with the attempted specialization
even if it's invalid. So if `C` has a type parameter with upper bound
`int`, and you say `C[str]`, they'll emit a diagnostic but just go with
`C[str]`. Even weirder, if `C` has a single type parameter and you say
`C[str, bytes]`, they'll just go with `C[str]` as the type. I'm not
convinced by this approach; it seems odd to have specializations
floating around that explicitly violate the declared upper bound, or in
the latter case aren't even the specialization the annotation requested.
I prefer `C[Unknown]` for this case.
Fixing this revealed an issue with `collections.namedtuple`, which
returns `type[tuple[Any, ...]]`. Due to
https://github.com/astral-sh/ty/issues/1649 we consider that to be an
invalid specialization. So previously we returned `Unknown`; after this
PR it would be `type[tuple[Unknown]]`, leading to more false positives
from our lack of functional namedtuple support. To avoid that I added an
explicit Todo type for functional namedtuples for now.
## Test Plan
Added and updated mdtests.
The conformance suite changes have to do with `ParamSpec`, so no
meaningful signal there.
The ecosystem changes appear to be the expected effects of having more
precise type information (including occurrences of known issues such as
https://github.com/astral-sh/ty/issues/1495 ). Most effects are just
changes to types in diagnostics.
## Summary
This PR updates the explicit specialization logic to avoid using the
call machinery.
Previously, the logic would use the call machinery by converting the
list of type variables into a `Binding` with a single `Signature` where
all the type variables are positional-only parameters with bounds and
constraints as the annotated type and the default type as the default
parameter value. This has the advantage that it doesn't need to
implement any specific logic but the disadvantages are subpar diagnostic
messages as it would use the ones specific to a function call. But, an
important disadvantage is that the kind of type variable is lost in this
translation which becomes important in #21445 where a `ParamSpec` can
specialize into a list of types which is provided using list literal.
For example,
```py
class Foo[T, **P]: ...
Foo[int, [int, str]]
```
This PR converts the logic to use a simple loop using `zip_longest` as
all type variables and their corresponding type argument maps on a 1-1
basis. They cannot be specified using keyword argument either e.g.,
`dict[_VT=str, _KT=int]` is invalid.
This PR also makes an initial attempt to improve the diagnostic message
to specifically target the specialization part by using words like "type
argument" instead of just "argument" and including information like the
type variable, bounds, and constraints. Further improvements can be made
by highlighting the type variable definition or the bounds / constraints
as a sub-diagnostic but I'm going to leave that as a follow-up.
## Test Plan
Update messages in existing test cases.
## Summary
Derived from #17371Fixesastral-sh/ty#256
Fixes https://github.com/astral-sh/ty/issues/1415
Fixes https://github.com/astral-sh/ty/issues/1433
Fixes https://github.com/astral-sh/ty/issues/1524
Properly handles any kind of recursive inference and prevents panics.
---
Let me explain techniques for converging fixed-point iterations during
recursive type inference.
There are two types of type inference that naively don't converge
(causing salsa to panic): divergent type inference and oscillating type
inference.
### Divergent type inference
Divergent type inference occurs when eagerly expanding a recursive type.
A typical example is this:
```python
class C:
def f(self, other: "C"):
self.x = (other.x, 1)
reveal_type(C().x) # revealed: Unknown | tuple[Unknown | tuple[Unknown | tuple[..., Literal[1]], Literal[1]], Literal[1]]
```
To solve this problem, we have already introduced `Divergent` types
(https://github.com/astral-sh/ruff/pull/20312). `Divergent` types are
treated as a kind of dynamic type [^1].
```python
Unknown | tuple[Unknown | tuple[Unknown | tuple[..., Literal[1]], Literal[1]], Literal[1]]
=> Unknown | tuple[Divergent, Literal[1]]
```
When a query function that returns a type enters a cycle, it sets
`Divergent` as the cycle initial value (instead of `Never`). Then, in
the cycle recovery function, it reduces the nesting of types containing
`Divergent` to converge.
```python
0th: Divergent
1st: Unknown | tuple[Divergent, Literal[1]]
2nd: Unknown | tuple[Unknown | tuple[Divergent, Literal[1]], Literal[1]]
=> Unknown | tuple[Divergent, Literal[1]]
```
Each cycle recovery function for each query should operate only on the
`Divergent` type originating from that query.
For this reason, while `Divergent` appears the same as `Any` to the
user, it internally carries some information: the location where the
cycle occurred. Previously, we roughly identified this by having the
scope where the cycle occurred, but with the update to salsa, functions
that create cycle initial values can now receive a `salsa::Id`
(https://github.com/salsa-rs/salsa/pull/1012). This is an opaque ID that
uniquely identifies the cycle head (the query that is the starting point
for the fixed-point iteration). `Divergent` now has this `salsa::Id`.
### Oscillating type inference
Now, another thing to consider is oscillating type inference.
Oscillating type inference arises from the fact that monotonicity is
broken. Monotonicity here means that for a query function, if it enters
a cycle, the calculation must start from a "bottom value" and progress
towards the final result with each cycle. Monotonicity breaks down in
type systems that have features like overloading and overriding.
```python
class Base:
def flip(self) -> "Sub":
return Sub()
class Sub(Base):
def flip(self) -> "Base":
return Base()
class C:
def __init__(self, x: Sub):
self.x = x
def replace_with(self, other: "C"):
self.x = other.x.flip()
reveal_type(C(Sub()).x)
```
Naive fixed-point iteration results in `Divergent -> Sub -> Base -> Sub
-> ...`, which oscillates forever without diverging or converging. To
address this, the salsa API has been modified so that the cycle recovery
function receives the value of the previous cycle
(https://github.com/salsa-rs/salsa/pull/1012).
The cycle recovery function returns the union type of the current cycle
and the previous cycle. In the above example, the result type for each
cycle is `Divergent -> Sub -> Base (= Sub | Base) -> Base`, which
converges.
The final result of oscillating type inference does not contain
`Divergent` because `Divergent` that appears in a union type can be
removed, as is clear from the expansion. This simplification is
performed at the same time as nesting reduction.
```
T | Divergent = T | (T | (T | ...)) = T
```
[^1]: In theory, it may be possible to strictly treat types containing
`Divergent` types as recursive types, but we probably shouldn't go that
deep yet. (AFAIK, there are no PEPs that specify how to handle
implicitly recursive types that aren't named by type aliases)
## Performance analysis
A happy side effect of this PR is that we've observed widespread
performance improvements!
This is likely due to the removal of the `ITERATIONS_BEFORE_FALLBACK`
and max-specialization depth trick
(https://github.com/astral-sh/ty/issues/1433,
https://github.com/astral-sh/ty/issues/1415), which means we reach a
fixed point much sooner.
## Ecosystem analysis
The changes look good overall.
You may notice changes in the converged values for recursive types,
this is because the way recursive types are normalized has been changed.
Previously, types containing `Divergent` types were normalized by
replacing them with the `Divergent` type itself, but in this PR, types
with a nesting level of 2 or more that contain `Divergent` types are
normalized by replacing them with a type with a nesting level of 1. This
means that information about the non-divergent parts of recursive types
is no longer lost.
```python
# previous
tuple[tuple[Divergent, int], int] => Divergent
# now
tuple[tuple[Divergent, int], int] => tuple[Divergent, int]
```
The false positive error introduced in this PR occurs in class
definitions with self-referential base classes, such as the one below.
```python
from typing_extensions import Generic, TypeVar
T = TypeVar("T")
U = TypeVar("U")
class Base2(Generic[T, U]): ...
# TODO: no error
# error: [unsupported-base] "Unsupported class base with type `<class 'Base2[Sub2, U@Sub2]'> | <class 'Base2[Sub2[Unknown], U@Sub2]'>`"
class Sub2(Base2["Sub2", U]): ...
```
This is due to the lack of support for unions of MROs, or because cyclic
legacy generic types are not inferred as generic types early in the
query cycle.
## Test Plan
All samples listed in astral-sh/ty#256 are tested and passed without any
panic!
## Acknowledgments
Thanks to @MichaReiser for working on bug fixes and improvements to
salsa for this PR. @carljm also contributed early on to the discussion
of the query convergence mechanism proposed in this PR.
---------
Co-authored-by: Carl Meyer <carl@astral.sh>
## Summary
We now use the type context for a lot of things, so re-inferring without
type context actually makes diagnostics more confusing (in most cases).
This patch lets us create specializations from a constraint set. The
constraint encodes the restrictions on which types each typevar can
specialize to. Given a generic context and a constraint set, we iterate
through all of the generic context's typevars. For each typevar, we
abstract the constraint set so that it only mentions the typevar in
question (propagating derived facts if needed). We then find the "best
representative type" for the typevar given the abstracted constraint
set.
When considering the BDD structure of the abstracted constraint set,
each path from the BDD root to the `true` terminal represents one way
that the constraint set can be satisfied. (This is also one of the
clauses in the DNF representation of the constraint set's boolean
formula.) Each of those paths is the conjunction of the individual
constraints of each internal node that we traverse as we walk that path,
giving a single lower/upper bound for the path. We use the upper bound
as the "best" (i.e. "closest to `object`") type for that path.
If there are multiple paths in the BDD, they technically represent
independent possible specializations. If there's a single specialization
that satisfies all of them, we will return that as the specialization.
If not, then the constraint set is ambiguous. (This happens most often
with constrained typevars.) We could in the future turn _each_ of the
paths into separate specializations, but it's not clear what we would do
with that, so instead we just report the ambiguity as a specialization
failure.
This manifested as an error when inferring the type of a PEP-695 generic
class via its constructor parameters:
```py
class D[T, U]:
@overload
def __init__(self: "D[str, U]", u: U) -> None: ...
@overload
def __init__(self, t: T, u: U) -> None: ...
def __init__(self, *args) -> None: ...
# revealed: D[Unknown, str]
# SHOULD BE: D[str, str]
reveal_type(D("string"))
```
This manifested because `D` is inferred to be bivariant in both `T` and
`U`. We weren't seeing this in the equivalent example for legacy
typevars, since those default to invariant. (This issue also showed up
for _covariant_ typevars, so this issue was not limited to bivariance.)
The underlying cause was because of a heuristic that we have in our
current constraint solver, which attempts to handle situations like
this:
```py
def f[T](t: T | None): ...
f(None)
```
Here, the `None` argument matches the non-typevar union element, so this
argument should not add any constraints on what `T` can specialize to.
Our previous heuristic would check for this by seeing if the argument
type is a subtype of the parameter annotation as a whole — even if it
isn't a union! That would cause us to erroneously ignore the `self`
parameter in our constructor call, since bivariant classes are
equivalent to each other, regardless of their specializations.
The quick fix is to move this heuristic "down a level", so that we only
apply it when the parameter annotation is a union. This heuristic should
go away completely 🤞 with the new constraint solver.
## Summary
This PR adds support for understanding the legacy definition and PEP 695
definition for `ParamSpec`.
This is still very initial and doesn't really implement any of the
semantics.
Part of https://github.com/astral-sh/ty/issues/157
## Test Plan
Add mdtest cases.
## Ecosystem analysis
Most of the diagnostics in `starlette` are due to the fact that ty now
understands `ParamSpec` is not a `Todo` type, so the assignability check
fails. The code looks something like:
```py
class _MiddlewareFactory(Protocol[P]):
def __call__(self, app: ASGIApp, /, *args: P.args, **kwargs: P.kwargs) -> ASGIApp: ... # pragma: no cover
class Middleware:
def __init__(
self,
cls: _MiddlewareFactory[P],
*args: P.args,
**kwargs: P.kwargs,
) -> None:
self.cls = cls
self.args = args
self.kwargs = kwargs
# ty complains that `ServerErrorMiddleware` is not assignable to `_MiddlewareFactory[P]`
Middleware(ServerErrorMiddleware, handler=error_handler, debug=debug)
```
There are multiple diagnostics where there's an attribute access on the
`Wrapped` object of `functools` which Pyright also raises:
```py
from functools import wraps
def my_decorator(f):
@wraps(f)
def wrapper(*args, **kwds):
return f(*args, **kwds)
# Pyright: Cannot access attribute "__signature__" for class "_Wrapped[..., Unknown, ..., Unknown]"
Attribute "__signature__" is unknown [reportAttributeAccessIssue]
# ty: Object of type `_Wrapped[Unknown, Unknown, Unknown, Unknown]` has no attribute `__signature__` [unresolved-attribute]
wrapper.__signature__
return wrapper
```
There are additional diagnostics that is due to the assignability checks
failing because ty now infers the `ParamSpec` instead of using the
`Todo` type which would always succeed. This results in a few
`no-matching-overload` diagnostics because the assignability checks
fail.
There are a few diagnostics related to
https://github.com/astral-sh/ty/issues/491 where there's a variable
which is either a bound method or a variable that's annotated with
`Callable` that doesn't contain the instance as the first parameter.
Another set of (valid) diagnostics are where the code hasn't provided
all the type variables. ty is now raising diagnostics for these because
we include `ParamSpec` type variable in the signature. For example,
`staticmethod[Any]` which contains two type variables.
## Summary
The solver is currently order-dependent, and will choose a supertype
over the exact type if it appears earlier in the list of constraints. We
could be smarter and try to choose the most precise subtype, but I
imagine this is something the new constraint solver will fix anyways,
and this fixes the issue showing up on
https://github.com/astral-sh/ruff/pull/21070.
Generic classes are not allowed to bind or reference a typevar from an
enclosing scope:
```py
def f[T](x: T, y: T) -> None:
class Ok[S]: ...
# error: [invalid-generic-class]
class Bad1[T]: ...
# error: [invalid-generic-class]
class Bad2(Iterable[T]): ...
class C[T]:
class Ok1[S]: ...
# error: [invalid-generic-class]
class Bad1[T]: ...
# error: [invalid-generic-class]
class Bad2(Iterable[T]): ...
```
It does not matter if the class uses PEP 695 or legacy syntax. It does
not matter if the enclosing scope is a generic class or function. The
generic class cannot even _reference_ an enclosing typevar in its base
class list.
This PR adds diagnostics for these cases.
In addition, the PR adds better fallback behavior for generic classes
that violate this rule: any enclosing typevars are not included in the
class's generic context. (That ensures that we don't inadvertently try
to infer specializations for those typevars in places where we
shouldn't.) The `dulwich` ecosystem project has [examples of
this](d912eaaffd/dulwich/config.py (L251))
that were causing new false positives on #20677.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
Implements bidirectional type inference using function return type
annotations.
This PR was originally proposed to solve astral-sh/ty#1167, but this
does not fully resolve it on its own.
Additionally, I believe we need to allow dataclasses to generate their
own `__new__` methods, [use constructor return types for
inference](5844c0103d/crates/ty_python_semantic/src/types.rs (L5326-L5328)),
and a mechanism to discard type narrowing like `& ~AlwaysFalsy` if
necessary (at a more general level than this PR).
## Test Plan
`mdtest/bidirectional.md` is added.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Co-authored-by: Ibraheem Ahmed <ibraheem@ibraheem.ca>
## Summary
This allows us to handle self-referential bounds/constraints/defaults
without panicking.
Handles more cases from https://github.com/astral-sh/ty/issues/256
This also changes the way we infer the types of legacy TypeVars. Rather
than understanding a constructor call to `typing[_extension].TypeVar`
inside of any (arbitrarily nested) expression, and having to use a
special `assigned_to` field of the semantic index to try to best-effort
figure out what name the typevar was assigned to, we instead understand
the creation of a legacy `TypeVar` only in the supported syntactic
position (RHS of a simple un-annotated assignment with one target). In
any other position, we just infer it as creating an opaque instance of
`typing.TypeVar`. (This behavior matches all other type checkers.)
So we now special-case TypeVar creation in `TypeInferenceBuilder`, as a
special case of an assignment definition, rather than deeper inside call
binding. This does mean we re-implement slightly more of
argument-parsing, but in practice this is minimal and easy to handle
correctly.
This is easier to implement if we also make the RHS of a simple (no
unpacking) one-target assignment statement no longer a standalone
expression. Which is fine to do, because simple one-target assignments
don't need to infer the RHS more than once. This is a bonus performance
(0-3% across various projects) and significant memory-usage win, since
most assignment statements are simple one-target assignment statements,
meaning we now create many fewer standalone-expression salsa
ingredients.
This change does mean that inference of manually-constructed
`TypeAliasType` instances can no longer find its Definition in
`assigned_to`, which regresses go-to-definition for these aliases. In a
future PR, `TypeAliasType` will receive the same treatment that
`TypeVar` did in this PR (moving its special-case inference into
`TypeInferenceBuilder` and supporting it only in the correct syntactic
position, and lazily inferring its value type to support recursion),
which will also fix the go-to-definition regression. (I decided a
temporary edge-case regression is better in this case than doubling the
size of this PR.)
This PR also tightens up and fixes various aspects of the validation of
`TypeVar` creation, as seen in the tests.
We still (for now) treat all typevars as instances of `typing.TypeVar`,
even if they were created using `typing_extensions.TypeVar`. This means
we'll wrongly error on e.g. `T.__default__` on Python 3.11, even if `T`
is a `typing_extensions.TypeVar` instance at runtime. We share this
wrong behavior with both mypy and pyrefly. It will be easier to fix
after we pull in https://github.com/python/typeshed/pull/14840.
There are some issues that showed up here with typevar identity and
`MarkTypeVarsInferable`; the fix here (using the new `original` field
and `is_identical_to` methods on `BoundTypeVarInstance` and
`TypeVarInstance`) is a bit kludgy, but it can go away when we eliminate
`MarkTypeVarsInferable`.
## Test Plan
Added and updated mdtests.
### Conformance suite impact
The impact here is all positive:
* We now correctly error on a legacy TypeVar with exactly one constraint
type given.
* We now correctly error on a legacy TypeVar with both an upper bound
and constraints specified.
### Ecosystem impact
Basically none; in the setuptools case we just issue slightly different
errors on an invalid TypeVar definition, due to the modified validation
code.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
Currently we do not emit an error on this code:
```py
from ty_extensions import Not
def f[T](x: T, y: Not[T]) -> T:
x = y
return x
```
But we should do! `~T` should never be assignable to `T`.
This fixes a small regression introduced in
14fe1228e7 (diff-8049ab5af787dba29daa389bbe2b691560c15461ef536f122b1beab112a4b48aR1443-R1446),
where a branch that previously returned `false` was replaced with a
branch that returns `C::always_satisfiable` -- the opposite of what it
used to be! The regression occurred because we didn't have any tests for
this -- so I added some tests in this PR that fail on `main`. I only
spotted the problem because I was going through the code of
`has_relation_to_impl` with a fine toothcomb for
https://github.com/astral-sh/ruff/pull/20602😄
## Summary
Add two simple tests that we recently discussed with @dcreager. They
demonstrate that the `TypeMapping::MarkTypeVarsInferable` operation
really does need to keep track of the binding context.
## Test Plan
Made sure that those tests fail if we create
`TypeMapping::MarkTypeVarsInferable(None)`s everywhere.
## Summary
Modify the (external) signature of instance methods such that the first
parameter uses `Self` unless it is explicitly annotated. This allows us
to correctly type-check more code, and allows us to infer correct return
types for many functions that return `Self`. For example:
```py
from pathlib import Path
from datetime import datetime, timedelta
reveal_type(Path(".config") / ".ty") # now Path, previously Unknown
def _(dt: datetime, delta: timedelta):
reveal_type(dt - delta) # now datetime, previously Unknown
```
part of https://github.com/astral-sh/ty/issues/159
## Performance
I ran benchmarks locally on `attrs`, `freqtrade` and `colour`, the
projects with the largest regressions on CodSpeed. I see much smaller
effects locally, but can definitely reproduce the regression on `attrs`.
From looking at the profiling results (on Codspeed), it seems that we
simply do more type inference work, which seems plausible, given that we
now understand much more return types (of many stdlib functions). In
particular, whenever a function uses an implicit `self` and returns
`Self` (without mentioning `Self` anywhere else in its signature), we
will now infer the correct type, whereas we would previously return
`Unknown`. This also means that we need to invoke the generics solver in
more cases. Comparing half a million lines of log output on attrs, I can
see that we do 5% more "work" (number of lines in the log), and have a
lot more `apply_specialization` events (7108 vs 4304). On freqtrade, I
see similar numbers for `apply_specialization` (11360 vs 5138 calls).
Given these results, I'm not sure if it's generally worth doing more
performance work, especially since none of the code modifications
themselves seem to be likely candidates for regressions.
| Command | Mean [ms] | Min [ms] | Max [ms] | Relative |
|:---|---:|---:|---:|---:|
| `./ty_main check /home/shark/ecosystem/attrs` | 92.6 ± 3.6 | 85.9 |
102.6 | 1.00 |
| `./ty_self check /home/shark/ecosystem/attrs` | 101.7 ± 3.5 | 96.9 |
113.8 | 1.10 ± 0.06 |
| Command | Mean [ms] | Min [ms] | Max [ms] | Relative |
|:---|---:|---:|---:|---:|
| `./ty_main check /home/shark/ecosystem/freqtrade` | 599.0 ± 20.2 |
568.2 | 627.5 | 1.00 |
| `./ty_self check /home/shark/ecosystem/freqtrade` | 607.9 ± 11.5 |
594.9 | 626.4 | 1.01 ± 0.04 |
| Command | Mean [ms] | Min [ms] | Max [ms] | Relative |
|:---|---:|---:|---:|---:|
| `./ty_main check /home/shark/ecosystem/colour` | 423.9 ± 17.9 | 394.6
| 447.4 | 1.00 |
| `./ty_self check /home/shark/ecosystem/colour` | 426.9 ± 24.9 | 373.8
| 456.6 | 1.01 ± 0.07 |
## Test Plan
New Markdown tests
## Ecosystem report
* apprise: ~300 new diagnostics related to problematic stubs in apprise
😩
* attrs: a new true positive, since [this
function](4e2c89c823/tests/test_make.py (L2135))
is missing a `@staticmethod`?
* Some legitimate true positives
* sympy: lots of new `invalid-operator` false positives in [matrix
multiplication](cf9f4b6805/sympy/matrices/matrixbase.py (L3267-L3269))
due to our limited understanding of [generic `Callable[[Callable[[T1,
T2], T3]], Callable[[T1, T2], T3]]` "identity"
types](cf9f4b6805/sympy/core/decorators.py (L83-L84))
of decorators. This is not related to type-of-self.
## Typing conformance results
The changes are all correct, except for
```diff
+generics_self_usage.py:50:5: error[invalid-assignment] Object of type `def foo(self) -> int` is not assignable to `(typing.Self, /) -> int`
```
which is related to an assignability problem involving type variables on
both sides:
```py
class CallableAttribute:
def foo(self) -> int:
return 0
bar: Callable[[Self], int] = foo # <- we currently error on this assignment
```
---------
Co-authored-by: Shaygan Hooshyari <sh.hooshyari@gmail.com>
### Summary
This PR includes two changes, both of which are necessary to resolve
https://github.com/astral-sh/ty/issues/1196:
* For a generic class `C[T]`, we previously used `C[Unknown]` as the
upper bound of the `Self` type variable. There were two problems with
this. For one, when `Self` appeared in contravariant position, we would
materialize its upper bound to `Bottom[C[Unknown]]` (which might
simplify to `C[Never]` if `C` is covariant in `T`) when accessing
methods on `Top[C[Unknown]]`. This would result in `invalid-argument`
errors on the `self` parameter. Also, using an upper bound of
`C[Unknown]` would mean that inside methods, references to `T` would be
treated as `Unknown`. This could lead to false negatives. To fix this,
we now use `C[T]` (with a "nested" typevar) as the upper bound for
`Self` on `C[T]`.
* In order to make this work, we needed to allow assignability/subtyping
of inferable typevars to other types, since we now check assignability
of e.g. `C[int]` to `C[T]` (when checking assignability to the upper
bound of `Self`) when calling an instance-method on `C[int]` whose
`self` parameter is annotated as `self: Self` (or implicitly `Self`,
following https://github.com/astral-sh/ruff/pull/18007).
closes https://github.com/astral-sh/ty/issues/1196
closes https://github.com/astral-sh/ty/issues/1208
### Test Plan
Regression tests for both issues.
`Type::TypeVar` now distinguishes whether the typevar in question is
inferable or not.
A typevar is _not inferable_ inside the body of the generic class or
function that binds it:
```py
def f[T](t: T) -> T:
return t
```
The infered type of `t` in the function body is `TypeVar(T,
NotInferable)`. This represents how e.g. assignability checks need to be
valid for all possible specializations of the typevar. Most of the
existing assignability/etc logic only applies to non-inferable typevars.
Outside of the function body, the typevar is _inferable_:
```py
f(4)
```
Here, the parameter type of `f` is `TypeVar(T, Inferable)`. This
represents how e.g. assignability doesn't need to hold for _all_
specializations; instead, we need to find the constraints under which
this specific assignability check holds.
This is in support of starting to perform specialization inference _as
part of_ performing the assignability check at the call site.
In the [[POPL2015][]] paper, this concept is called _monomorphic_ /
_polymorphic_, but I thought _non-inferable_ / _inferable_ would be
clearer for us.
Depends on #19784
[POPL2015]: https://doi.org/10.1145/2676726.2676991
---------
Co-authored-by: Carl Meyer <carl@astral.sh>
## Summary
For PEP 695 generic functions and classes, there is an extra "type
params scope" (a child of the outer scope, and wrapping the body scope)
in which the type parameters are defined; class bases and function
parameter/return annotations are resolved in that type-params scope.
This PR fixes some longstanding bugs in how we resolve name loads from
inside these PEP 695 type parameter scopes, and also defers type
inference of PEP 695 typevar bounds/constraints/default, so we can
handle cycles without panicking.
We were previously treating these type-param scopes as lazy nested
scopes, which is wrong. In fact they are eager nested scopes; the class
`C` here inherits `int`, not `str`, and previously we got that wrong:
```py
Base = int
class C[T](Base): ...
Base = str
```
But certain syntactic positions within type param scopes (typevar
bounds/constraints/defaults) are lazy at runtime, and we should use
deferred name resolution for them. This also means they can have cycles;
in order to handle that without panicking in type inference, we need to
actually defer their type inference until after we have constructed the
`TypeVarInstance`.
PEP 695 does specify that typevar bounds and constraints cannot be
generic, and that typevar defaults can only reference prior typevars,
not later ones. This reduces the scope of (valid from the type-system
perspective) cycles somewhat, although cycles are still possible (e.g.
`class C[T: list[C]]`). And this is a type-system-only restriction; from
the runtime perspective an "invalid" case like `class C[T: T]` actually
works fine.
I debated whether to implement the PEP 695 restrictions as a way to
avoid some cycles up-front, but I ended up deciding against that; I'd
rather model the runtime name-resolution semantics accurately, and
implement the PEP 695 restrictions as a separate diagnostic on top.
(This PR doesn't yet implement those diagnostics, thus some `# TODO:
error` in the added tests.)
Introducing the possibility of cyclic typevars made typevar display
potentially stack overflow. For now I've handled this by simply removing
typevar details (bounds/constraints/default) from typevar display. This
impacts display of two kinds of types. If you `reveal_type(T)` on an
unbound `T` you now get just `typing.TypeVar` instead of
`typing.TypeVar("T", ...)` where `...` is the bound/constraints/default.
This matches pyright and mypy; pyrefly uses `type[TypeVar[T]]` which
seems a bit confusing, but does include the name. (We could easily
include the name without cycle issues, if there's a syntax we like for
that.)
It also means that displaying a generic function type like `def f[T:
int](x: T) -> T: ...` now displays as `f[T](x: T) -> T` instead of `f[T:
int](x: T) -> T`. This matches pyright and pyrefly; mypy does include
bound/constraints/defaults of typevars in function/callable type
display. If we wanted to add this, we would either need to thread a
visitor through all the type display code, or add a `decycle` type
transformation that replaced recursive reoccurrence of a type with a
marker.
## Test Plan
Added mdtests and modified existing tests to improve their correctness.
After this PR, there's only a single remaining py-fuzzer seed in the
0-500 range that panics! (Before this PR, there were 10; the fuzzer
likes to generate cyclic PEP 695 syntax.)
## Ecosystem report
It's all just the changes to `TypeVar` display.
This fixes our logic for binding a legacy typevar with its binding
context. (To recap, a legacy typevar starts out "unbound" when it is
first created, and each time it's used in a generic class or function,
we "bind" it with the corresponding `Definition`.)
We treat `typing.Self` the same as a legacy typevar, and so we apply
this binding logic to it too. Before, we were using the enclosing class
as its binding context. But that's not correct — it's the method where
`typing.Self` is used that binds the typevar. (Each invocation of the
method will find a new specialization of `Self` based on the specific
instance type containing the invoked method.)
This required plumbing through some additional state to the
`in_type_expression` method.
This also revealed that we weren't handling `Self`-typed instance
attributes correctly (but were coincidentally not getting the expected
false positive diagnostics).
This is subtle, and the root cause became more apparent with #19604,
since we now have many more cases of superclasses and subclasses using
different typevars. The issue is easiest to see in the following:
```py
class C[T]:
def __init__(self, t: T) -> None: ...
class D[U](C[T]):
pass
reveal_type(C(1)) # revealed: C[int]
reveal_type(D(1)) # should be: D[int]
```
When instantiating a generic class, the `__init__` method inherits the
generic context of that class. This lets our call binding machinery
infer a specialization for that context.
Prior to this PR, the instantiation of `C` worked just fine. Its
`__init__` method would inherit the `[T]` generic context, and we would
infer `{T = int}` as the specialization based on the argument
parameters.
It didn't work for `D`. The issue is that the `__init__` method was
inheriting the generic context of the class where `__init__` was defined
(here, `C` and `[T]`). At the call site, we would then infer `{T = int}`
as the specialization — but that wouldn't help us specialize `D[U]`,
since `D` does not have `T` in its generic context!
Instead, the `__init__` method should inherit the generic context of the
class that we are performing the lookup on (here, `D` and `[U]`). That
lets us correctly infer `{U = int}` as the specialization, which we can
successfully apply to `D[U]`.
(Note that `__init__` refers to `C`'s typevars in its signature, but
that's okay; our member lookup logic already applies the `T = U`
specialization when returning a member of `C` while performing a lookup
on `D`, transforming its signature from `(Self, T) -> None` to `(Self,
U) -> None`.)
Closes https://github.com/astral-sh/ty/issues/588
This PR introduces a few related changes:
- We now keep track of each time a legacy typevar is bound in a
different generic context (e.g. class, function), and internally create
a new `TypeVarInstance` for each usage. This means the rest of the code
can now assume that salsa-equivalent `TypeVarInstance`s refer to the
same typevar, even taking into account that legacy typevars can be used
more than once.
- We also go ahead and track the binding context of PEP 695 typevars.
That's _much_ easier to track since we have the binding context right
there during type inference.
- With that in place, we can now include the name of the binding context
when rendering typevars (e.g. `T@f` instead of `T`)
## Summary
This PR improves our generics solver such that we are able to solve the
`TypeVar` in this snippet to `int | str` (the union of the elements in
the heterogeneous tuple) by upcasting the heterogeneous tuple to its
pure-homogeneous-tuple supertype:
```py
def f[T](x: tuple[T, ...]) -> T:
return x[0]
def g(x: tuple[int, str]):
reveal_type(f(x))
```
## Test Plan
Mdtests. Some TODOs remain in the mdtest regarding solving `TypeVar`s
for mixed tuples, but I think this PR on its own is a significant step
forward for our generics solver when it comes to tuple types.
---------
Co-authored-by: Douglas Creager <dcreager@dcreager.net>
We now correctly exclude legacy typevars from enclosing scopes when
constructing the generic context for a generic function.
more detail:
A function is generic if it refers to legacy typevars in its signature:
```py
from typing import TypeVar
T = TypeVar("T")
def f(t: T) -> T:
return t
```
Generic functions are allowed to appear inside of other generic
contexts. When they do, they can refer to the typevars of those
enclosing generic contexts, and that should not rebind the typevar:
```py
from typing import TypeVar, Generic
T = TypeVar("T")
U = TypeVar("U")
class C(Generic[T]):
@staticmethod
def method(t: T, u: U) -> None: ...
# revealed: def method(t: int, u: U) -> None
reveal_type(C[int].method)
```
This substitution was already being performed correctly, but we were
also still including the enclosing legacy typevars in the method's own
generic context, which can be seen via `ty_extensions.generic_context`
(which has been updated to work on generic functions and methods):
```py
from ty_extensions import generic_context
# before: tuple[T, U]
# after: tuple[U]
reveal_type(generic_context(C[int].method))
```
---------
Co-authored-by: Carl Meyer <carl@astral.sh>
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
Having a recursive type method to check whether a type is fully static
is inefficient, unnecessary, and makes us overly strict about subtyping
relations.
It's inefficient because we end up re-walking the same types many times
to check for fully-static-ness.
It's unnecessary because we can check relations involving the dynamic
type appropriately, depending whether the relation is subtyping or
assignability.
We use the subtyping relation to simplify unions and intersections. We
can usefully consider that `S <: T` for gradual types also, as long as
it remains true that `S | T` is equivalent to `T` and `S & T` is
equivalent to `S`.
One conservative definition (implemented here) that satisfies this
requirement is that we consider `S <: T` if, for every possible pair of
materializations `S'` and `T'`, `S' <: T'`. Or put differently the top
materialization of `S` (`S+` -- the union of all possible
materializations of `S`) is a subtype of the bottom materialization of
`T` (`T-` -- the intersection of all possible materializations of `T`).
In the most basic cases we can usefully say that `Any <: object` and
that `Never <: Any`, and we can handle more complex cases inductively
from there.
This definition of subtyping for gradual subtypes is not reflexive
(`Any` is not a subtype of `Any`).
As a corollary, we also remove `is_gradual_equivalent_to` --
`is_equivalent_to` now has the meaning that `is_gradual_equivalent_to`
used to have. If necessary, we could restore an
`is_fully_static_equivalent_to` or similar (which would not do an
`is_fully_static` pre-check of the types, but would instead pass a
relation-kind enum down through a recursive equivalence check, similar
to `has_relation_to`), but so far this doesn't appear to be necessary.
Credit to @JelleZijlstra for the observation that `is_fully_static` is
unnecessary and overly restrictive on subtyping.
There is another possible definition of gradual subtyping: instead of
requiring that `S+ <: T-`, we could instead require that `S+ <: T+` and
`S- <: T-`. In other words, instead of requiring all materializations of
`S` to be a subtype of every materialization of `T`, we just require
that every materialization of `S` be a subtype of _some_ materialization
of `T`, and that every materialization of `T` be a supertype of some
materialization of `S`. This definition also preserves the core
invariant that `S <: T` implies that `S | T = T` and `S & T = S`, and it
restores reflexivity: under this definition, `Any` is a subtype of
`Any`, and for any equivalent types `S` and `T`, `S <: T` and `T <: S`.
But unfortunately, this definition breaks transitivity of subtyping,
because nominal subclasses in Python use assignability ("consistent
subtyping") to define acceptable overrides. This means that we may have
a class `A` with `def method(self) -> Any` and a subtype `B(A)` with
`def method(self) -> int`, since `int` is assignable to `Any`. This
means that if we have a protocol `P` with `def method(self) -> Any`, we
would have `B <: A` (from nominal subtyping) and `A <: P` (`Any` is a
subtype of `Any`), but not `B <: P` (`int` is not a subtype of `Any`).
Breaking transitivity of subtyping is not tenable, so we don't use this
definition of subtyping.
## Test Plan
Existing tests (modified in some cases to account for updated
semantics.)
Stable property tests pass at a million iterations:
`QUICKCHECK_TESTS=1000000 cargo test -p ty_python_semantic -- --ignored
types::property_tests::stable`
### Changes to property test type generation
Since we no longer have a method of categorizing built types as
fully-static or not-fully-static, I had to add a previously-discussed
feature to the property tests so that some tests can build types that
are known by construction to be fully static, because there are still
properties that only apply to fully-static types (for example,
reflexiveness of subtyping.)
## Changes to handling of `*args, **kwargs` signatures
This PR "discovered" that, once we allow non-fully-static types to
participate in subtyping under the above definitions, `(*args: Any,
**kwargs: Any) -> Any` is now a subtype of `() -> object`. This is true,
if we take a literal interpretation of the former signature: all
materializations of the parameters `*args: Any, **kwargs: Any` can
accept zero arguments, making the former signature a subtype of the
latter. But the spec actually says that `*args: Any, **kwargs: Any`
should be interpreted as equivalent to `...`, and that makes a
difference here: `(...) -> Any` is not a subtype of `() -> object`,
because (unlike a literal reading of `(*args: Any, **kwargs: Any)`),
`...` can materialize to _any_ signature, including a signature with
required positional arguments.
This matters for this PR because it makes the "any two types are both
assignable to their union" property test fail if we don't implement the
equivalence to `...`. Because `FunctionType.__call__` has the signature
`(*args: Any, **kwargs: Any) -> Any`, and if we take that at face value
it's a subtype of `() -> object`, making `FunctionType` a subtype of `()
-> object)` -- but then a function with a required argument is also a
subtype of `FunctionType`, but not a subtype of `() -> object`. So I
went ahead and implemented the equivalence to `...` in this PR.
## Ecosystem analysis
* Most of the ecosystem report are cases of improved union/intersection
simplification. For example, we can now simplify a union like `bool |
(bool & Unknown) | Unknown` to simply `bool | Unknown`, because we can
now observe that every possible materialization of `bool & Unknown` is
still a subtype of `bool` (whereas before we would set aside `bool &
Unknown` as a not-fully-static type.) This is clearly an improvement.
* The `possibly-unresolved-reference` errors in sockeye, pymongo,
ignite, scrapy and others are true positives for conditional imports
that were formerly silenced by bogus conflicting-declarations (which we
currently don't issue a diagnostic for), because we considered two
different declarations of `Unknown` to be conflicting (we used
`is_equivalent_to` not `is_gradual_equivalent_to`). In this PR that
distinction disappears and all equivalence is gradual, so a declaration
of `Unknown` no longer conflicts with a declaration of `Unknown`, which
then results in us surfacing the possibly-unbound error.
* We will now issue "redundant cast" for casting from a typevar with a
gradual bound to the same typevar (the hydra-zen diagnostic). This seems
like an improvement.
* The new diagnostics in bandersnatch are interesting. For some reason
primer in CI seems to be checking bandersnatch on Python 3.10 (not yet
sure why; this doesn't happen when I run it locally). But bandersnatch
uses `enum.StrEnum`, which doesn't exist on 3.10. That makes the `class
SimpleDigest(StrEnum)` a class that inherits from `Unknown` (and
bypasses our current TODO handling for accessing attributes on enum
classes, since we don't recognize it as an enum class at all). This PR
improves our understanding of assignability to classes that inherit from
`Any` / `Unknown`, and we now recognize that a string literal is not
assignable to a class inheriting `Any` or `Unknown`.
We already had support for homogeneous tuples (`tuple[int, ...]`). This
PR extends this to also support mixed tuples (`tuple[str, str,
*tuple[int, ...], str str]`).
A mixed tuple consists of a fixed-length (possibly empty) prefix and
suffix, and a variable-length portion in the middle. Every element of
the variable-length portion must be of the same type. A homogeneous
tuple is then just a mixed tuple with an empty prefix and suffix.
The new data representation uses different Rust types for a fixed-length
(aka heterogeneous) tuple. Another option would have been to use the
`VariableLengthTuple` representation for all tuples, and to wrap the
"variable + suffix" portion in an `Option`. I don't think that would
simplify the method implementations much, though, since we would still
have a 2×2 case analysis for most of them.
One wrinkle is that the definition of the `tuple` class in the typeshed
has a single typevar, and canonically represents a homogeneous tuple.
When getting the class of a tuple instance, that means that we have to
summarize our detailed mixed tuple type information into its
"homogeneous supertype". (We were already doing this for heterogeneous
types.)
A similar thing happens when concatenating two mixed tuples: the
variable-length portion and suffix of the LHS, and the prefix and
variable-length portion of the RHS, all get unioned into the
variable-length portion of the result. The LHS prefix and RHS suffix
carry through unchanged.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
An issue seen here https://github.com/astral-sh/ty/issues/500
The `__init__` method of dataclasses had no inherited generic context,
so we could not infer the type of an instance from a constructor call
with generics
## Test Plan
Add tests to classes.md` in generics folder
## Summary
Allow a typevar to be callable if it is bound to a callable type, or
constrained to callable types.
I spent some time digging into why this support didn't fall out
naturally, and ultimately the reason is that we look up `__call__` on
the meta type (since its a dunder), and our implementation of
`Type::to_meta_type` for `Type::Callable` does not return a type with
`__call__`.
A more general solution here would be to have `Type::to_meta_type` for
`Type::Callable` synthesize a protocol with `__call__` and return an
intersection with that protocol (since for a type to be callable, we
know its meta-type must have `__call__`). That solution could in
principle also replace the special-case handling of `Type::Callable`
itself, here in `Type::bindings`. But that more general approach would
also be slower, and our protocol support isn't quite ready for that yet,
and handling this directly in `Type::bindings` is really not bad.
Fixes https://github.com/astral-sh/ty/issues/480
## Test Plan
Added mdtests.
## 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
This implements the stopgap approach described in
https://github.com/astral-sh/ty/issues/336#issuecomment-2880532213 for
handling literal types in generic class specializations.
With this approach, we will promote any literal to its instance type,
but _only_ when inferring a generic class specialization from a
constructor call:
```py
class C[T]:
def __init__(self, x: T) -> None: ...
reveal_type(C("string")) # revealed: C[str]
```
If you specialize the class explicitly, we still use whatever type you
provide, even if it's a literal:
```py
from typing import Literal
reveal_type(C[Literal[5]](5)) # revealed: C[Literal[5]]
```
And this doesn't apply at all to generic functions:
```py
def f[T](x: T) -> T:
return x
reveal_type(f(5)) # revealed: Literal[5]
```
---
As part of making this happen, we also generalize the `TypeMapping`
machinery. This provides a way to apply a function to type, returning a
new type. Complicating matters is that for function literals, we have to
apply the mapping lazily, since the function's signature is not created
until (and if) someone calls its `signature` method. That means we have
to stash away the mappings that we want to apply to the signatures
parameter/return annotations once we do create it. This requires some
minor `Cow` shenanigans to continue working for partial specializations.
Dhruv Manilawala
Douglas Creager
Alex Waygood
Ibraheem Ahmed
Carl Meyer
Shunsuke Shibayama
David Peter
Matthew Mckee