[ty] Homogeneous and mixed tuples (#18600)
Some checks are pending
CI / cargo fuzz build (push) Blocked by required conditions
CI / Determine changes (push) Waiting to run
CI / cargo fmt (push) Waiting to run
CI / cargo clippy (push) Blocked by required conditions
CI / cargo test (linux) (push) Blocked by required conditions
CI / cargo test (linux, release) (push) Blocked by required conditions
CI / cargo test (windows) (push) Blocked by required conditions
CI / cargo test (wasm) (push) Blocked by required conditions
CI / cargo build (release) (push) Waiting to run
CI / cargo build (msrv) (push) Blocked by required conditions
CI / fuzz parser (push) Blocked by required conditions
CI / test scripts (push) Blocked by required conditions
CI / ecosystem (push) Blocked by required conditions
CI / Fuzz for new ty panics (push) Blocked by required conditions
CI / cargo shear (push) Blocked by required conditions
CI / python package (push) Waiting to run
CI / pre-commit (push) Waiting to run
CI / mkdocs (push) Waiting to run
CI / formatter instabilities and black similarity (push) Blocked by required conditions
CI / test ruff-lsp (push) Blocked by required conditions
CI / check playground (push) Blocked by required conditions
CI / benchmarks-instrumented (push) Blocked by required conditions
CI / benchmarks-walltime (push) Blocked by required conditions
[ty Playground] Release / publish (push) Waiting to run

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>
This commit is contained in:
Douglas Creager 2025-06-20 18:23:54 -04:00 committed by GitHub
parent d9266284df
commit ea812d0813
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
32 changed files with 2432 additions and 758 deletions

View file

@ -69,8 +69,64 @@ def _(m: int, n: int):
t[::0] # error: [zero-stepsize-in-slice]
tuple_slice = t[m:n]
# TODO: Should be `tuple[Literal[1, 'a', b"b"] | None, ...]`
reveal_type(tuple_slice) # revealed: tuple[Unknown, ...]
reveal_type(tuple_slice) # revealed: tuple[Literal[1, "a", b"b"] | None, ...]
```
## Slices of homogeneous and mixed tuples
```toml
[environment]
python-version = "3.11"
```
```py
from typing import Literal
def homogeneous(t: tuple[str, ...]) -> None:
reveal_type(t[0]) # revealed: str
reveal_type(t[1]) # revealed: str
reveal_type(t[2]) # revealed: str
reveal_type(t[3]) # revealed: str
reveal_type(t[-1]) # revealed: str
reveal_type(t[-2]) # revealed: str
reveal_type(t[-3]) # revealed: str
reveal_type(t[-4]) # revealed: str
def mixed(s: tuple[str, ...]) -> None:
t = (1, 2, 3) + s + (8, 9, 10)
reveal_type(t[0]) # revealed: Literal[1]
reveal_type(t[1]) # revealed: Literal[2]
reveal_type(t[2]) # revealed: Literal[3]
reveal_type(t[3]) # revealed: str | Literal[8]
reveal_type(t[4]) # revealed: str | Literal[8, 9]
reveal_type(t[5]) # revealed: str | Literal[8, 9, 10]
reveal_type(t[-1]) # revealed: Literal[10]
reveal_type(t[-2]) # revealed: Literal[9]
reveal_type(t[-3]) # revealed: Literal[8]
reveal_type(t[-4]) # revealed: Literal[3] | str
reveal_type(t[-5]) # revealed: Literal[2, 3] | str
reveal_type(t[-6]) # revealed: Literal[1, 2, 3] | str
```
## `tuple` as generic alias
For tuple instances, we can track more detailed information about the length and element types of
the tuple. This information carries over to the generic alias that the tuple is an instance of.
```py
def _(a: tuple, b: tuple[int], c: tuple[int, str], d: tuple[int, ...]) -> None:
reveal_type(a) # revealed: tuple[Unknown, ...]
reveal_type(b) # revealed: tuple[int]
reveal_type(c) # revealed: tuple[int, str]
reveal_type(d) # revealed: tuple[int, ...]
reveal_type(tuple) # revealed: <class 'tuple'>
reveal_type(tuple[int]) # revealed: <class 'tuple[int]'>
reveal_type(tuple[int, str]) # revealed: <class 'tuple[int, str]'>
reveal_type(tuple[int, ...]) # revealed: <class 'tuple[int, ...]'>
```
## Inheritance
@ -83,8 +139,13 @@ python-version = "3.9"
```py
class A(tuple[int, str]): ...
# revealed: tuple[<class 'A'>, <class 'tuple[@Todo(Generic tuple specializations), ...]'>, <class 'Sequence[@Todo(Generic tuple specializations)]'>, <class 'Reversible[@Todo(Generic tuple specializations)]'>, <class 'Collection[@Todo(Generic tuple specializations)]'>, <class 'Iterable[@Todo(Generic tuple specializations)]'>, <class 'Container[@Todo(Generic tuple specializations)]'>, typing.Protocol, typing.Generic, <class 'object'>]
# revealed: tuple[<class 'A'>, <class 'tuple[int, str]'>, <class 'Sequence[int | str]'>, <class 'Reversible[int | str]'>, <class 'Collection[int | str]'>, <class 'Iterable[int | str]'>, <class 'Container[int | str]'>, typing.Protocol, typing.Generic, <class 'object'>]
reveal_type(A.__mro__)
class C(tuple): ...
# revealed: tuple[<class 'C'>, <class 'tuple[Unknown, ...]'>, <class 'Sequence[Unknown]'>, <class 'Reversible[Unknown]'>, <class 'Collection[Unknown]'>, <class 'Iterable[Unknown]'>, <class 'Container[Unknown]'>, typing.Protocol, typing.Generic, <class 'object'>]
reveal_type(C.__mro__)
```
## `typing.Tuple`
@ -109,9 +170,19 @@ def _(c: Tuple, d: Tuple[int, A], e: Tuple[Any, ...]):
Inheriting from `Tuple` results in a MRO with `builtins.tuple` and `typing.Generic`. `Tuple` itself
is not a class.
```toml
[environment]
python-version = "3.9"
```
```py
from typing import Tuple
class A(Tuple[int, str]): ...
# revealed: tuple[<class 'A'>, <class 'tuple[int, str]'>, <class 'Sequence[int | str]'>, <class 'Reversible[int | str]'>, <class 'Collection[int | str]'>, <class 'Iterable[int | str]'>, <class 'Container[int | str]'>, typing.Protocol, typing.Generic, <class 'object'>]
reveal_type(A.__mro__)
class C(Tuple): ...
# revealed: tuple[<class 'C'>, <class 'tuple[Unknown, ...]'>, <class 'Sequence[Unknown]'>, <class 'Reversible[Unknown]'>, <class 'Collection[Unknown]'>, <class 'Iterable[Unknown]'>, <class 'Container[Unknown]'>, typing.Protocol, typing.Generic, <class 'object'>]