ruff/crates/ty_python_semantic/src
David Peter 42f152108a
[ty] Generic types aliases (implicit and PEP 613) (#21553)
## Summary

Add support for generic PEP 613 type aliases and generic implicit type
aliases:
```py
from typing import TypeVar

T = TypeVar("T")
ListOrSet = list[T] | set[T]

def _(xs: ListOrSet[int]):
    reveal_type(xs)  # list[int] | set[int]
```

closes https://github.com/astral-sh/ty/issues/1643
closes https://github.com/astral-sh/ty/issues/1629
closes https://github.com/astral-sh/ty/issues/1596
closes https://github.com/astral-sh/ty/issues/573
closes https://github.com/astral-sh/ty/issues/221

## Typing conformance

```diff
-aliases_explicit.py:52:5: error[type-assertion-failure] Type `list[int]` does not match asserted type `@Todo(specialized generic alias in type expression)`
-aliases_explicit.py:53:5: error[type-assertion-failure] Type `tuple[str, ...] | list[str]` does not match asserted type `@Todo(Generic specialization of types.UnionType)`
-aliases_explicit.py:54:5: error[type-assertion-failure] Type `tuple[int, int, int, str]` does not match asserted type `@Todo(specialized generic alias in type expression)`
-aliases_explicit.py:56:5: error[type-assertion-failure] Type `(int, str, /) -> str` does not match asserted type `@Todo(Generic specialization of typing.Callable)`
-aliases_explicit.py:59:5: error[type-assertion-failure] Type `int | str | None | list[list[int]]` does not match asserted type `int | str | None | list[@Todo(specialized generic alias in type expression)]`
```

New true negatives ✔️ 

```diff
+aliases_explicit.py:41:36: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
-aliases_explicit.py:57:5: error[type-assertion-failure] Type `(int, str, str, /) -> None` does not match asserted type `@Todo(Generic specialization of typing.Callable)`
+aliases_explicit.py:57:5: error[type-assertion-failure] Type `(int, str, str, /) -> None` does not match asserted type `(...) -> Unknown`
```

These require `ParamSpec`

```diff
+aliases_explicit.py:67:24: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
+aliases_explicit.py:68:24: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
+aliases_explicit.py:69:29: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
+aliases_explicit.py:70:29: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
+aliases_explicit.py:71:29: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
+aliases_explicit.py:102:20: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
```

New true positives ✔️ 

```diff
-aliases_implicit.py:63:5: error[type-assertion-failure] Type `list[int]` does not match asserted type `@Todo(specialized generic alias in type expression)`
-aliases_implicit.py:64:5: error[type-assertion-failure] Type `tuple[str, ...] | list[str]` does not match asserted type `@Todo(Generic specialization of types.UnionType)`
-aliases_implicit.py:65:5: error[type-assertion-failure] Type `tuple[int, int, int, str]` does not match asserted type `@Todo(specialized generic alias in type expression)`
-aliases_implicit.py:67:5: error[type-assertion-failure] Type `(int, str, /) -> str` does not match asserted type `@Todo(Generic specialization of typing.Callable)`
-aliases_implicit.py:70:5: error[type-assertion-failure] Type `int | str | None | list[list[int]]` does not match asserted type `int | str | None | list[@Todo(specialized generic alias in type expression)]`
-aliases_implicit.py:71:5: error[type-assertion-failure] Type `list[bool]` does not match asserted type `@Todo(specialized generic alias in type expression)`
```

New true negatives ✔️ 

```diff
+aliases_implicit.py:54:36: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
-aliases_implicit.py:68:5: error[type-assertion-failure] Type `(int, str, str, /) -> None` does not match asserted type `@Todo(Generic specialization of typing.Callable)`
+aliases_implicit.py:68:5: error[type-assertion-failure] Type `(int, str, str, /) -> None` does not match asserted type `(...) -> Unknown`
```

These require `ParamSpec`

```diff
+aliases_implicit.py:76:24: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
+aliases_implicit.py:77:24: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
+aliases_implicit.py:78:29: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
+aliases_implicit.py:79:29: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
+aliases_implicit.py:80:29: error[invalid-type-arguments] Too many type arguments: expected 1, got 2
+aliases_implicit.py:81:25: error[invalid-type-arguments] Type `str` is not assignable to upper bound `int | float` of type variable `TFloat@GoodTypeAlias12`
+aliases_implicit.py:135:20: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
```

New true positives ✔️ 

```diff
+callables_annotation.py:172:19: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
+callables_annotation.py:175:19: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
+callables_annotation.py:188:25: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
+callables_annotation.py:189:25: error[invalid-type-arguments] Too many type arguments: expected 0, got 1
```

These require `ParamSpec` and `Concatenate`.

```diff
-generics_defaults_specialization.py:26:5: error[type-assertion-failure] Type `SomethingWithNoDefaults[int, str]` does not match asserted type `SomethingWithNoDefaults[int, typing.TypeVar]`
+generics_defaults_specialization.py:26:5: error[type-assertion-failure] Type `SomethingWithNoDefaults[int, str]` does not match asserted type `SomethingWithNoDefaults[int, DefaultStrT]`
```

Favorable diagnostic change ✔️ 

```diff
-generics_defaults_specialization.py:27:5: error[type-assertion-failure] Type `SomethingWithNoDefaults[int, bool]` does not match asserted type `@Todo(specialized generic alias in type expression)`
```

New true negative ✔️ 

```diff
-generics_defaults_specialization.py:30:1: error[non-subscriptable] Cannot subscript object of type `<class 'SomethingWithNoDefaults[int, typing.TypeVar]'>` with no `__class_getitem__` method
+generics_defaults_specialization.py:30:15: error[invalid-type-arguments] Too many type arguments: expected between 0 and 1, got 2
```

Correct new diagnostic ✔️ 


```diff
-generics_variance.py:175:25: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:175:35: error[non-subscriptable] Cannot subscript object of type `<class 'Co[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:179:29: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:179:39: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:183:21: error[non-subscriptable] Cannot subscript object of type `<class 'Co[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:183:27: error[non-subscriptable] Cannot subscript object of type `<class 'Co[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:187:25: error[non-subscriptable] Cannot subscript object of type `<class 'Co[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:187:31: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:191:33: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:191:43: error[non-subscriptable] Cannot subscript object of type `<class 'Co[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:191:49: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:196:5: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:196:15: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
-generics_variance.py:196:25: error[non-subscriptable] Cannot subscript object of type `<class 'Contra[typing.TypeVar]'>` with no `__class_getitem__` method
```

One of these should apparently be an error, but not of this kind, so
this is good ✔️

```diff
-specialtypes_type.py:152:16: error[invalid-type-form] `typing.TypeVar` is not a generic class
-specialtypes_type.py:156:16: error[invalid-type-form] `typing.TypeVar` is not a generic class
```

Good, those were false positives. ✔️ 

I skipped the analysis for everything involving `TypeVarTuple`.

## Ecosystem impact

**[Full report with detailed
diff](https://david-generic-implicit-alias.ecosystem-663.pages.dev/diff)**

Previous iterations of this PR showed all kinds of problems. In it's
current state, I do not see any large systematic problems, but it is
hard to tell with 5k diagnostic changes.

## Performance

* There is a huge 4x regression in `colour-science/colour`, related to
[this large
file](https://github.com/colour-science/colour/blob/develop/colour/io/luts/tests/test_lut.py)
with [many assignments of hard-coded arrays (lists of lists) to
`np.NDArray`
types](83e754c8b6/colour/io/luts/tests/test_lut.py (L701-L781))
that we now understand. We now take ~2 seconds to check this file, so
definitely not great, but maybe acceptable for now.

## Test Plan

Updated and new Markdown tests
2025-11-28 20:38:24 +01:00
..
module_resolver [ty] Don't suggest things that aren't subclasses of `BaseException` after `raise` 2025-11-24 12:55:30 -05:00
semantic_index [ty] Generic types aliases (implicit and PEP 613) (#21553) 2025-11-28 20:38:24 +01:00
types [ty] Generic types aliases (implicit and PEP 613) (#21553) 2025-11-28 20:38:24 +01:00
ast_node_ref.rs [ty] Improve semantic token classification for names (#21399) 2025-11-12 16:34:26 +00:00
db.rs [ty] Limit shown import paths to at most 5 unless ty runs with `-v` (#20912) 2025-10-16 13:18:09 +02:00
diagnostic.rs [ty] Refactor `CheckSuppressionContext` to use `DiagnosticGuard` (#21587) 2025-11-25 10:54:42 +00:00
dunder_all.rs [ty] follow the breaking API changes made in salsa-rs/salsa#1015 (#21117) 2025-10-29 14:56:12 +00:00
lib.rs [ty] Don't suggest things that aren't subclasses of `BaseException` after `raise` 2025-11-24 12:55:30 -05:00
lint.rs [ty] Add hyperlinks to rule codes in CLI (#21502) 2025-11-18 16:36:59 +01:00
list.rs Update Rust toolchain to 1.89 (#19807) 2025-08-07 18:21:50 +02:00
module_name.rs [ty] Add support for relative import completions 2025-11-21 08:01:02 -05:00
node_key.rs [ty] Implement the legacy PEP-484 convention for indicating positional-only parameters (#20248) 2025-09-05 17:56:06 +01:00
place.rs [ty] Add `db` parameter to `Parameters::new` method (#21674) 2025-11-28 12:29:58 +00:00
program.rs [ty] Remove mentions of VS Code from server logs (#21155) 2025-11-03 14:49:58 +00:00
pull_types.rs Update Rust toolchain to 1.88 and MSRV to 1.86 (#19011) 2025-06-28 20:24:00 +02:00
python_platform.rs Update Rust toolchain to 1.91 (#21179) 2025-11-01 01:50:58 +00:00
rank.rs [ty] Garbage-collect reachability constraints (#19414) 2025-07-21 14:16:27 -04:00
semantic_index.rs [ty] Skip eagerly evaluated scopes for attribute storing (#20856) 2025-11-11 14:45:34 -08:00
semantic_model.rs [ty] handle recursive type inference properly (#20566) 2025-11-26 08:50:26 -08:00
site_packages.rs [ty] Discover site-packages from the environment that ty is installed in (#21286) 2025-11-06 09:27:49 -05:00
subscript.rs [ty] Use "cannot" consistently over "can not" (#21255) 2025-11-03 10:38:20 -05:00
suppression.rs [ty] Add 'remove unused ignore comment' code action (#21582) 2025-11-25 08:08:21 -05:00
types.rs [ty] Generic types aliases (implicit and PEP 613) (#21553) 2025-11-28 20:38:24 +01:00
unpack.rs [ty] Track heap usage of salsa structs (#19790) 2025-08-12 13:28:44 +02:00