## 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
Came across this while debugging some ecosystem changes in
https://github.com/astral-sh/ruff/pull/18347. I think the meta-type of a
typevar-annotated variable should be equal to `type`, not `<class
'object'>`.
## Test Plan
New Markdown tests.
## 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.
This is a follow-on to #18155. For the example raised in
https://github.com/astral-sh/ty/issues/370:
```py
import tempfile
with tempfile.TemporaryDirectory() as tmp: ...
```
the new logic would notice that both overloads of `TemporaryDirectory`
match, and combine their specializations, resulting in an inferred type
of `str | bytes`.
This PR updates the logic to match our other handling of other calls,
where we only keep the _first_ matching overload. The result for this
example then becomes `str`, matching the runtime behavior. (We still do
not implement the full [overload resolution
algorithm](https://typing.python.org/en/latest/spec/overload.html#overload-call-evaluation)
from the spec.)
This primarily comes up with annotated `self` parameters in
constructors:
```py
class C[T]:
def __init__(self: C[int]): ...
```
Here, we want infer a specialization of `{T = int}` for a call that hits
this overload.
Normally when inferring a specialization of a function call, typevars
appear in the parameter annotations, and not in the argument types. In
this case, this is reversed: we need to verify that the `self` argument
(`C[T]`, as we have not yet completed specialization inference) is
assignable to the parameter type `C[int]`.
To do this, we simply look for a typevar/type in both directions when
performing inference, and apply the inferred specialization to argument
types as well as parameter types before verifying assignability.
As a wrinkle, this exposed that we were not checking
subtyping/assignability for function literals correctly. Our function
literal representation includes an optional specialization that should
be applied to the signature. Before, function literals were considered
subtypes of (assignable to) each other only if they were identical Salsa
objects. Two function literals with different specializations should
still be considered subtypes of (assignable to) each other if those
specializations result in the same function signature (typically because
the function doesn't use the typevars in the specialization).
Closes https://github.com/astral-sh/ty/issues/370
Closes https://github.com/astral-sh/ty/issues/100
Closes https://github.com/astral-sh/ty/issues/258
---------
Co-authored-by: Carl Meyer <carl@astral.sh>
Follows on from (and depends on)
https://github.com/astral-sh/ruff/pull/18021.
This updates our function specialization inference to infer type
mappings from parameters that are generic protocols.
For now, this only works when the argument _explicitly_ implements the
protocol by listing it as a base class. (We end up using exactly the
same logic as for generic classes in #18021.) For this to work with
classes that _implicitly_ implement the protocol, we will have to check
the types of the protocol members (which we are not currently doing), so
that we can infer the specialization of the protocol that the class
implements.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
This updates our function specialization inference to infer type
mappings from parameters that are generic aliases, e.g.:
```py
def f[T](x: list[T]) -> T: ...
reveal_type(f(["a", "b"])) # revealed: str
```
Though note that we're still inferring the type of list literals as
`list[Unknown]`, so for now we actually need something like the
following in our tests:
```py
def _(x: list[str]):
reveal_type(f(x)) # revealed: str
```
We were not inducting into instance types and subclass-of types when
looking for legacy typevars, nor when apply specializations.
This addresses
https://github.com/astral-sh/ruff/pull/17832#discussion_r2081502056
```py
from __future__ import annotations
from typing import TypeVar, Any, reveal_type
S = TypeVar("S")
class Foo[T]:
def method(self, other: Foo[S]) -> Foo[T | S]: ... # type: ignore[invalid-return-type]
def f(x: Foo[Any], y: Foo[Any]):
reveal_type(x.method(y)) # revealed: `Foo[Any | S]`, but should be `Foo[Any]`
```
We were not detecting that `S` made `method` generic, since we were not
finding it when searching the function signature for legacy typevars.
Function literals have an optional specialization, which is applied to
the parameter/return type annotations lazily when the function's
signature is requested. We were previously only applying this
specialization to the final overload of an overloaded function.
This manifested most visibly for `list.__add__`, which has an overloaded
definition in the typeshed:
b398b83631/crates/ty_vendored/vendor/typeshed/stdlib/builtins.pyi (L1069-L1072)
Closes https://github.com/astral-sh/ty/issues/314
It's possible for a typevar to list another typevar as its default
value:
```py
class C[T, U = T]: ...
```
When specializing this class, if a type isn't provided for `U`, we would
previously use the default as-is, leaving an unspecialized `T` typevar
in the specialization. Instead, we want to use what `T` is mapped to as
the type of `U`.
```py
reveal_type(C()) # revealed: C[Unknown, Unknown]
reveal_type(C[int]()) # revealed: C[int, int]
reveal_type(C[int, str]()) # revealed: C[int, str]
```
This is especially important for the `slice` built-in type.
#17897 added variance handling for legacy typevars — but they were only
being considered when checking generic aliases of the same class:
```py
class A: ...
class B(A): ...
class C[T]: ...
static_assert(is_subtype_of(C[B], C[A]))
```
and not for generic subclasses:
```py
class D[U](C[U]): ...
static_assert(is_subtype_of(D[B], C[A]))
```
Now we check those too!
Closes https://github.com/astral-sh/ty/issues/101
We now track the variance of each typevar, and obey the `covariant` and
`contravariant` parameters to the legacy `TypeVar` constructor. We still
don't yet infer variance for PEP-695 typevars or for the
`infer_variance` legacy constructor parameter.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Co-authored-by: Carl Meyer <carl@astral.sh>
@AlexWaygood discovered that even though we've been propagating
specializations to _parent_ base classes correctly, we haven't been
passing them on to _grandparent_ base classes:
https://github.com/astral-sh/ruff/pull/17832#issuecomment-2854360969
```py
class Bar[T]:
x: T
class Baz[T](Bar[T]): ...
class Spam[T](Baz[T]): ...
reveal_type(Spam[int]().x) # revealed: `T`, but should be `int`
```
This PR updates the MRO machinery to apply the current specialization
when starting to iterate the MRO of each base class.
If a typevar is declared as having a default, we shouldn't require a
type to be specified for that typevar when explicitly specializing a
generic class:
```py
class WithDefault[T, U = int]: ...
reveal_type(WithDefault[str]()) # revealed: WithDefault[str, int]
```
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Fixes
https://github.com/astral-sh/ruff/pull/17832#issuecomment-2851224968. We
had a comment that we did not need to apply specializations to generic
aliases, or to the bound `self` of a bound method, because they were
already specialized. But they might be specialized with a type variable,
which _does_ need to be specialized, in the case of a "multi-step"
specialization, such as:
```py
class LinkedList[T]: ...
class C[U]:
def method(self) -> LinkedList[U]:
return LinkedList[U]()
```
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
