The constraint representation that we added in #19997 was subtly wrong,
in that it didn't correctly model that type assignability is a _partial_
order — it's possible for two types to be incomparable, with neither a
subtype of the other. That means the negation of a constraint like `T ≤
t` (typevar `T` must be a subtype of `t`) is **_not_** `t < T`, but
rather `t < T ∨ T ≁ t` (using ≁ to mean "not comparable to").
That means we need to update our constraint representation to be an
enum, so that we can track both _range_ constraints (upper/lower bound
on the typevar), and these new _incomparable_ constraints.
Since we need an enum now, that also lets us simplify how we were
modeling range constraints. Before, we let the lower/upper bounds be
either open (<) or closed (≤). Now, range constraints are always closed,
and we add a third kind of constraint for _not equivalent_ (≠). We can
translate an open upper bound `T < t` into `T ≤ t ∧ T ≠ t`.
We already had the logic for doing adding _clauses_ to a _set_ by doing
a pairwise simplification. We copy that over to where we add
_constraints_ to a _clause_. To calculate the intersection or union of
two constraints, the new enum representation makes it easy to break down
all of the possibilities into a small number of cases: intersect range
with range, intersect range with not-equivalent, etc. I've done the math
[here](https://dcreager.net/theory/constraints/) to show that the
simplifications for each of these cases is correct.
## Summary
This is a follow-up to https://github.com/astral-sh/ruff/pull/19321.
Now lazy snapshots are updated to take into account new bindings on
every symbol reassignment.
```python
def outer(x: A | None):
if x is None:
x = A()
reveal_type(x) # revealed: A
def inner() -> None:
# lazy snapshot: {x: A}
reveal_type(x) # revealed: A
inner()
def outer() -> None:
x = None
x = 1
def inner() -> None:
# lazy snapshot: {x: Literal[1]} -> {x: Literal[1, 2]}
reveal_type(x) # revealed: Literal[1, 2]
inner()
x = 2
```
Closesastral-sh/ty#559.
## Test Plan
Some TODOs in `public_types.md` now work properly.
---------
Co-authored-by: Carl Meyer <carl@astral.sh>
## Summary
Adds support for generic PEP695 type aliases, e.g.,
```python
type A[T] = T
reveal_type(A[int]) # A[int]
```
Resolves https://github.com/astral-sh/ty/issues/677.
## Summary
Support cases like the following, where we need the generic context to
include both `Self` and `T` (not just `T`):
```py
from typing import Self
class C:
def method[T](self: Self, arg: T): ...
C().method(1)
```
closes https://github.com/astral-sh/ty/issues/1131
## Test Plan
Added regression test
## Summary
The sub-checks for assignability and subtyping of materializations
performed in `has_relation_in_invariant_position` and
`is_subtype_in_invariant_position` need to propagate the
`HasRelationToVisitor`, or we can stack overflow.
A side effect of this change is that we also propagate the
`ConstraintSet` through, rather than using `C::from_bool`, which I think
may also become important for correctness in cases involving type
variables (though it isn't testable yet, since we aren't yet actually
creating constraints other than always-true and always-false.)
## Test Plan
Added mdtest (derived from code found in pydantic) which
stack-overflowed before this PR.
With this change incorporated, pydantic now checks successfully on my
draft PR for PEP 613 TypeAlias support.
Now that https://github.com/astral-sh/ruff/pull/20263 is merged, we can
update mypy_primer and add the new `egglog-python` project to
`good.txt`. The ecosystem-analyzer run shows that we now add 1,356
diagnostics (where we had over 5,000 previously, due to the unsupported
project layout).
## Summary
Add backreferences to the original item declaration in TypedDict
diagnostics.
Thanks to @AlexWaygood for the suggestion.
## Test Plan
Updated snapshots
## Summary
An annotated assignment `name: annotation` without a right-hand side was
previously not covered by the range returned from
`DefinitionKind::full_range`, because we did expand the range to include
the right-hand side (if there was one), but failed to include the
annotation.
## Test Plan
Updated snapshot tests
## Summary
Add support for `typing.ReadOnly` as a type qualifier to mark
`TypedDict` fields as being read-only. If you try to mutate them, you
get a new diagnostic:
<img width="787" height="234" alt="image"
src="https://github.com/user-attachments/assets/f62fddf9-4961-4bcd-ad1c-747043ebe5ff"
/>
## Test Plan
* New Markdown tests
* The typing conformance changes are all correct. There are some false
negatives, but those are related to the missing support for the
functional form of `TypedDict`, or to overriding of fields via
inheritance. Both of these topics are tracked in
https://github.com/astral-sh/ty/issues/154
Closesastral-sh/ty#456. Part of astral-sh/ty#994.
After all the foundational work, this is only a small change, but let's
see if it exposes any unresolved issues.
## Summary
Part of astral-sh/ty#994. The goal of this PR was to add correct
behavior for attribute access on the top and bottom materializations.
This is necessary for the end goal of using the top materialization for
narrowing generics (`isinstance(x, list)`): we want methods like
`x.append` to work correctly in that case.
It turned out to be convenient to represent materialization as a
TypeMapping, so it can be stashed in the `type_mappings` list of a
function object. This also allowed me to remove most concrete
`materialize` methods, since they usually just delegate to the subparts
of the type, the same as other type mappings. That is why the net effect
of this PR is to remove a few hundred lines.
## Test Plan
I added a few more tests. Much of this PR is refactoring and covered by
existing tests.
## Followups
Assigning to attributes of top materializations is not yet covered. This
seems less important so I'd like to defer it.
I noticed that the `materialize` implementation of `Parameters` was
wrong; it did the same for the top and bottom materializations. This PR
makes the bottom materialization slightly more reasonable, but
implementing this correctly will require extending the struct.
## Summary
Two minor cleanups:
- Return `Option<ClassType>` rather than `Option<ClassLiteral>` from
`TypeInferenceBuilder::class_context_of_current_method`. Now that
`ClassType::is_protocol` exists as a method as well as
`ClassLiteral::is_protocol`, this simplifies most of the call-sites of
the `class_context_of_current_method()` method.
- Make more use of the `MethodDecorator::try_from_fn_type` method in
`class.rs`. Under the hood, this method uses the new methods
`FunctionType::is_classmethod()` and `FunctionType::is_staticmethod()`
that @sharkdp recently added, so it gets the semantics more precisely
correct than the code it's replacing in `infer.rs` (by accounting for
implicit staticmethods/classmethods as well as explicit ones). By using
these methods we can delete some code elsewhere (the
`FunctionDecorators::from_decorator_types()` constructor)
## Test Plan
Existing tests
## Summary
A small set of additional tests for `TypedDict` that I wrote while going
through the spec. Note that this certainly doesn't make the test suite
exhaustive (see remaining open points in the updated list here:
https://github.com/astral-sh/ty/issues/154).
This PR adds two new `ty_extensions` functions,
`reveal_when_assignable_to` and `reveal_when_subtype_of`. These are
closely related to the existing `is_assignable_to` and `is_subtype_of`,
but instead of returning when the property (always) holds, it produces a
diagnostic that describes _when_ the property holds. (This will let us
construct mdtests that print out constraints that are not always true or
always false — though we don't currently have any instances of those.)
I did not replace _every_ occurrence of the `is_property` variants in
the mdtest suite, instead focusing on the generics-related tests where
it will be important to see the full detail of the constraint sets.
As part of this, I also updated the mdtest harness to accept the shorter
`# revealed:` assertion format for more than just `reveal_type`, and
updated the existing uses of `reveal_protocol_interface` to take
advantage of this.
## Summary
Pull this out of https://github.com/astral-sh/ruff/pull/18473 as an
isolated change to make sure it has no adverse effects.
The wrong behavior is observable on `main` for something like
```py
class C:
def __new__(cls) -> "C":
cls.x = 1
C.x # previously: Attribute `x` can only be accessed on instances
# now: Type `<class 'C'>` has no attribute `x`
```
where we currently treat `x` as an *instance* attribute (because we
consider `__new__` to be a normal function and `cls` to be the "self"
attribute). With this PR, we do not consider `x` to be an attribute,
neither on the class nor on instances of `C`. If this turns out to be an
important feature, we should add it intentionally, instead of
accidentally.
## Test Plan
Ecosystem checks.
<!--
Thank you for contributing to Ruff/ty! To help us out with reviewing,
please consider the following:
- Does this pull request include a summary of the change? (See below.)
- Does this pull request include a descriptive title? (Please prefix
with `[ty]` for ty pull
requests.)
- Does this pull request include references to any relevant issues?
-->
## Summary
This PR implements
https://docs.astral.sh/ruff/rules/yield-from-in-async-function/ as a
syntax semantic error
## Test Plan
<!-- How was it tested? -->
I have written a simple inline test as directed in
[https://github.com/astral-sh/ruff/issues/17412](https://github.com/astral-sh/ruff/issues/17412)
---------
Signed-off-by: 11happy <soni5happy@gmail.com>
Co-authored-by: Alex Waygood <alex.waygood@gmail.com>
Co-authored-by: Brent Westbrook <36778786+ntBre@users.noreply.github.com>
This re-works the `all_symbols` based added previously to work across
all modules available, and not just what is directly in the workspace.
Note that we always pass an empty string as a query, which makes the
results always empty. We'll fix this in a subsequent commit.
This is to facilitate recursive traversal of all modules in an
environment. This way, we can keep asking for submodules.
This also simplifies how this is used in completions, and probably makes
it faster. Namely, since we return the `Module` itself, callers don't
need to invoke the full module resolver just to get the module type.
Note that this doesn't include namespace packages. (Which were
previously not supported in `Module::all_submodules`.) Given how they
can be spread out across multiple search paths, they will likely require
special consideration here.
This is similar to a change made in the "list top-level modules"
implementation that had been masked by poor Salsa failure modes.
Basically, if we can't find a root here, it *must* be a bug. And if we
just silently skip over it, we risk voiding Salsa's purity contract,
leading to more difficult to debug panics.
This did cause one test to fail, but only because the test wasn't
properly setting up roots.
## Summary
Thread visitors through the rest of `apply_type_mapping`: callable and
protocol types.
## Test Plan
Added mdtest that previously stack overflowed.
## Summary
We have the ability to defer type inference of some parts of
definitions, so as to allow us to create a type that may need to be
recursively referenced in those other parts of the definition.
We also have the ability to do type inference in a context where all
name resolution should be deferred (that is, names should be looked up
from all-reachable-definitions rather than from the location of use.)
This is used for all annotations in stubs, or if `from __future__ import
annotations` is active.
Previous to this PR, these two concepts were linked: deferred-inference
always implied deferred-name-resolution, though we also supported
deferred-name-resolution without deferred-inference, via
`DeferredExpressionState`.
For the upcoming `typing.TypeAlias` support, I will defer inference of
the entire RHS of the alias (so as to support cycles), but that doesn't
imply deferred name resolution; at runtime, the RHS of a name annotated
as `typing.TypeAlias` is executed eagerly.
So this PR fully de-couples the two concepts, instead explicitly setting
the `DeferredExpressionState` in those cases where we should defer name
resolution.
It also fixes a long-standing related bug, where we were deferring name
resolution of all names in class bases, if any of the class bases
contained a stringified annotation.
## Test Plan
Added test that failed before this PR.
## Summary
Fuzzer seed 208 seems to be timing out all fuzzer runs on PRs today.
This has happened on multiple unrelated PRs, as well as on an initial
version of this PR that made a comment-only change in ty and didn't skip
any seeds, so the timeout appears to be consistent in CI, on ty main
branch, as of today, but it started happening due to some change in a
factor outside ty; not sure what.
I checked the code generated for seed 208 locally, and it takes about
30s to check on current ty main branch. This is slow for a fuzzer seed,
but shouldn't be slow enough to make it time out after 20min in CI (even
accounting for GH runners being slower than my laptop.)
I tried to bisect the slowness of checking that code locally, but I
didn't go back far enough to find the change that made it slow. In fact
it seems like it became significantly faster in the last few days (on an
older checkout I had to stop it after several minutes.) So whatever the
cause of the slowness, it's not a recent change in ty.
I don't want to rabbit-hole on this right now (fuzzer-discovered issues
are lower-priority than real-world-code issues), and need a working CI,
so skip this seed for now until we can investigate it.
## Test Plan
CI. This PR contains a no-op (comment) change in ty, so that the fuzz
test is triggered in CI and we can verify it now works (as well as
verify, on the previous commit, that the fuzzer job is timing out on
that seed, even with just a no-op change in ty.)
Reverts astral-sh/ruff#20156. As @sharkdp noted in his post-merge
review, there were several issues with that PR that I didn't spot before
merging — but I'm out for four days now, and would rather not leave
things in an inconsistent state for that long. I'll revisit this on
Wednesday.
## Summary
These projects all check successfully now.
(Pandas still takes 9s, as the comment in `bad.txt` said, but I don't
think this is slow enough to exclude it; mypy-primer overall still runs
in 4 minutes, faster than e.g. the test suite on Windows.)
## Test Plan
mypy-primer CI.
## Summary
This error is about assigning to attributes rather than reading
attributes, so I think `invalid-assignment` makes more sense than
`invalid-attribute-access`
## Test Plan
existing mdtests updated
## Summary
In `is_disjoint_from_impl`, we should unpack type aliases before we
check `TypedDict`. This change probably doesn't have any visible effect
until we have a more discriminating implementation of disjointness for
`TypedDict`, but making the change now can avoid some confusion/bugs in
future.
In `type_ordering.rs`, we should order `TypedDict` near more similar
types, and leave Union/Intersection together at the end of the list.
This is not necessary for correctness, but it's more consistent and it
could have saved me some confusion trying to figure out why I was only
getting an unreachable panic when my code example included a `TypedDict`
type.
## Test Plan
None besides existing tests.
## Summary
Now that we have `Type::TypeAlias`, which can wrap a union, and the
possibility of unions including non-unpacked type aliases (which is
necessary to support recursive type aliases), we can no longer assume in
`UnionType::normalized_impl` that normalizing each element of an
existing union will result in a set of elements that we can order and
then place raw into `UnionType` to create a normalized union. It's now
possible for those elements to themselves include union types (unpacked
from an alias). So instead, we need to feed those elements into the full
`UnionBuilder` (with alias-unpacking turned on) to flatten/normalize
them, and then order them.
## Test Plan
Added mdtest.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
This PR fixes various TODOs around overload call when a variadic
argument is used.
The reason this bug existed is because the specialization wouldn't
account for unpacking the type of the variadic argument.
This is fixed by expanding `MatchedArgument` to contain the type of that
argument _only_ when it is a variadic argument. The reason is that
there's a split for when the argument type is inferred -- the
non-variadic arguments are inferred using `infer_argument_types` _after_
parameter matching while the variadic argument type is inferred _during_
the parameter matching. And, the `MatchedArgument` is populated _during_
parameter matching which means the unpacking would need to happen during
parameter matching.
This split seems a bit inconsistent but I don't want to spend a lot of
time on trying to merge them such that all argument type inference
happens in a single place. I might look into it while adding support for
`**kwargs`.
## Test Plan
Update existing tests by resolving the todos.
The ecosystem changes looks correct to me except for the `slice` call
but it seems that it's unrelated to this PR as we infer `slice[Any, Any,
Any]` for a `slice(1, 2, 3)` call on `main` as well
([playground](https://play.ty.dev/9eacce00-c7d5-4dd5-a932-4265cb2bb4f6)).
This PR adds an implementation of constraint sets.
An individual constraint restricts the specialization of a single
typevar to be within a particular lower and upper bound: the typevar can
only specialize to types that are a supertype of the lower bound, and a
subtype of the upper bound. (Note that lower and upper bounds are fully
static; we take the bottom and top materializations of the bounds to
remove any gradual forms if needed.) Either bound can be “closed” (where
the bound is a valid specialization), or “open” (where it is not).
You can then build up more complex constraint sets using union,
intersection, and negation operations. We use a disjunctive normal form
(DNF) representation, just like we do for types: a _constraint set_ is
the union of zero or more _clauses_, each of which is the intersection
of zero or more individual constraints. Note that the constraint set
that contains no clauses is never satisfiable (`⋃ {} = 0`); and the
constraint set that contains a single clause, which contains no
constraints, is always satisfiable (`⋃ {⋂ {}} = 1`).
One thing to note is that this PR does not change the logic of the
actual assignability checks, and in particular, we still aren't ever
trying to create an "individual constraint" that constrains a typevar.
Technically we're still operating only on `bool`s, since we only ever
instantiate `C::always_satisfiable` (i.e., `true`) and
`C::unsatisfiable` (i.e., `false`) in the `has_relation_to` methods. So
if you thought that #19838 introduced an unnecessarily complex stand-in
for `bool`, well here you go, this one is worse! (But still seemingly
not yielding a performance regression!) The next PR in this series,
#20093, is where we will actually create some non-trivial constraint
sets and use them in anger.
That said, the PR does go ahead and update the assignability checks to
use the new `ConstraintSet` type instead of `bool`. That part is fairly
straightforward since we had already updated the assignability checks to
use the `Constraints` trait; we just have to actively choose a different
impl type. (For the `is_whatever` variants, which still return a `bool`,
we have to convert the constraint set, but the explicit
`is_always_satisfiable` calls serve as nice documentation of our
intent.)
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Co-authored-by: Carl Meyer <carl@astral.sh>
This is a variant of #20076 that moves some complexity out of
`apply_type_mapping_impl` in `generics.rs`. The tradeoff is that now
every place that applies `TypeMapping::Specialization` must take care to
call `.materialize()` afterwards. (A previous version of this didn't
work because I had missed a spot where I had to call `.materialize()`.)
@carljm as asked in
https://github.com/astral-sh/ruff/pull/20076#discussion_r2305385298 .
## Summary
Decrease the maximum number of literals in a union before we collapse to
the supertype. The better fix for this will be
https://github.com/astral-sh/ty/issues/957, but it is very tempting to
solve this for now by simply decreasing the limit by one, to get below
the salsa limit of 200.
closes https://github.com/astral-sh/ty/issues/660
## Test Plan
Added a regression test that would previously lead to a "too many cycle
iterations" panic.
## Summary
With this PR, we stop performing boundness analysis for implicit
instance attributes:
```py
class C:
def __init__(self):
if False:
self.x = 1
C().x # would previously show an error, with this PR we pretend the attribute exists
```
This PR is potentially just a temporary measure until we find a better
fix. But I have already invested a lot of time trying to find the root
cause of https://github.com/astral-sh/ty/issues/758 (and [this
example](https://github.com/astral-sh/ty/issues/758#issuecomment-3206108262),
which I'm not entirely sure is related) and I still don't understand
what is going on. This PR fixes the performance problems in both of
these problems (in a rather crude way).
The impact of the proposed change on the ecosystem is small, and the
three new diagnostics are arguably true positives (previously hidden
because we considered the code unreachable, based on e.g. `assert`ions
that depended on implicit instance attributes). So this seems like a
reasonable fix for now.
Note that we still support cases like these:
```py
class D:
if False: # or any other expression that statically evaluates to `False`
x: int = 1
D().x # still an error
class E:
if False: # or any other expression that statically evaluates to `False`
def f(self):
self.x = 1
E().x # still an error
```
closes https://github.com/astral-sh/ty/issues/758
## Test Plan
Updated tests, benchmark results
## Summary
closes https://github.com/astral-sh/ty/issues/692
If the expression (or any child expressions) is not definitely bound the
reachability constraint evaluation is determined as ambiguous.
This fixes the infinite cycles panic in the following code:
```py
from typing import Literal
class Toggle:
def __init__(self: "Toggle"):
if not self.x:
self.x: Literal[True] = True
```
Credit of this solution is for David.
## Test Plan
- Added a test case with too many cycle iterations panic.
- Previous tests.
---------
Co-authored-by: David Peter <mail@david-peter.de>
Part of #994. This adds a new field to the Specialization struct to
record when we're dealing with the top or bottom materialization of an
invariant generic. It also implements subtyping and assignability for
these objects.
Next planned steps after this is done are to implement other operations
on top/bottom materializations; probably attribute access is an
important one.
---------
Co-authored-by: Carl Meyer <carl@astral.sh>
There are some situations that we have a confusing diagnostics due to
identical class names.
## Class with same name from different modules
```python
import pandas
import polars
df: pandas.DataFrame = polars.DataFrame()
```
This yields the following error:
**Actual:**
error: [invalid-assignment] "Object of type `DataFrame` is not
assignable to `DataFrame`"
**Expected**:
error: [invalid-assignment] "Object of type `polars.DataFrame` is not
assignable to `pandas.DataFrame`"
## Nested classes
```python
from enum import Enum
class A:
class B(Enum):
ACTIVE = "active"
INACTIVE = "inactive"
class C:
class B(Enum):
ACTIVE = "active"
INACTIVE = "inactive"
```
**Actual**:
error: [invalid-assignment] "Object of type `Literal[B.ACTIVE]` is not
assignable to `B`"
**Expected**:
error: [invalid-assignment] "Object of type
`Literal[my_module.C.B.ACTIVE]` is not assignable to `my_module.A.B`"
## Solution
In this MR we added an heuristics to detect when to use a fully
qualified name:
- There is an invalid assignment and;
- They are two different classes and;
- They have the same name
The fully qualified name always includes:
- module name
- nested classes name
- actual class name
There was no `QualifiedDisplay` so I had to implement it from scratch.
I'm very new to the codebase, so I might have done things inefficiently,
so I appreciate feedback.
Should we pre-compute the fully qualified name or do it on demand?
## Not implemented
### Function-local classes
Should we approach this in a different PR?
**Example**:
```python
# t.py
from __future__ import annotations
def function() -> A:
class A:
pass
return A()
class A:
pass
a: A = function()
```
#### mypy
```console
t.py:8: error: Incompatible return value type (got "t.A@5", expected "t.A") [return-value]
```
From my testing the 5 in `A@5` comes from the like number.
#### ty
```console
error[invalid-return-type]: Return type does not match returned value
--> t.py:4:19
|
4 | def function() -> A:
| - Expected `A` because of return type
5 | class A:
6 | pass
7 |
8 | return A()
| ^^^ expected `A`, found `A`
|
info: rule `invalid-return-type` is enabled by default
```
Fixes https://github.com/astral-sh/ty/issues/848
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Co-authored-by: Carl Meyer <carl@astral.sh>
## Summary
While looking at some logging output that I added to
`ReachabilityConstraintBuilder::add_and_constraint` in order to debug
https://github.com/astral-sh/ty/issues/1091, I noticed that it seemed to
suggest that the TDD was built in an imbalanced way for code like the
following, where we have a sequence of non-nested `if` conditions:
```py
def f(t1, t2, t3, t4, …):
x = 0
if t1:
x = 1
if t2:
x = 2
if t3:
x = 3
if t4:
x = 4
…
```
To understand this a bit better, I added some code to the
`ReachabilityConstraintBuilder` to render the resulting TDD. On `main`,
we get a tree that looks like the following, where you can see a pattern
of N sub-trees that grow linearly with N (number of `if` statements).
This results in an overall tree structure that has N² nodes (see graph
below):
<img alt="normal order"
src="https://github.com/user-attachments/assets/aab40ce9-e82a-4fcd-823a-811f05f15f66"
/>
If we zoom in to one of these subgraphs, we can see what the problem is.
When we add new constraints that represent combinations like `t1 AND ~t2
AND ~t3 AND t4 AND …`, they start with the evaluation of "early"
conditions (`t1`, `t2`, …). This means that we have to create new
subgraphs for each new `if` condition because there is little sharing
with the previous structure. We evaluate the Boolean condition in a
right-associative way: `t1 AND (~t2 AND (~t3 AND t4)))`:
<img width="500" align="center"
src="https://github.com/user-attachments/assets/31ea7182-9e00-4975-83df-d980464f545d"
/>
If we change the ordering of TDD atoms, we can change that to a
left-associative evaluation: `(((t1 AND ~t2) AND ~t3) AND t4) …`. This
means that we can re-use previous subgraphs `(t1 AND ~t2)`, which
results in a much more compact graph structure overall (note how "late"
conditions are now at the top, and "early" conditions are further down
in the graph):
<img alt="reverse order"
src="https://github.com/user-attachments/assets/96a6b7c1-3d35-4192-a917-0b2d24c6b144"
/>
If we count the number of TDD nodes for a growing number if `if`
statements, we can see that this change results in a slower growth. It's
worth noting that the growth is still superlinear, though:
<img width="800" height="600" alt="plot"
src="https://github.com/user-attachments/assets/22e8394f-e74e-4a9e-9687-0d41f94f2303"
/>
On the actual code from the referenced ticket (the `t_main.py` file
reduced to its main function, with the main function limited to 2000
lines instead of 11000 to allow the version on `main` to run to
completion), the effect is much more dramatic. Instead of 26 million TDD
nodes (`main`), we now only create 250 thousand (this branch), which is
slightly less than 1%.
The change in this PR allows us to build the semantic index and
type-check the problematic `t_main.py` file in
https://github.com/astral-sh/ty/issues/1091 in 9 seconds. This is still
not great, but an obvious improvement compared to running out of memory
after *minutes* of execution.
An open question remains whether this change is beneficial for all kinds
of code patterns, or just this linear sequence of `if` statements. It
does not seem unreasonable to think that referring to "earlier"
conditions is generally a good idea, but I learned from Doug that it's
generally not possible to find a TDD-construction heuristic that is
non-pathological for all kinds of inputs. Fortunately, it seems like
this change here results in performance improvements across *all of our
benchmarks*, which should increase the confidence in this change:
| Benchmark | Improvement |
|---------------------|-------------------------|
| hydra-zen | +13% |
| DateType | +5% |
| sympy (walltime) | +4% |
| attrs | +4% |
| pydantic (walltime) | +2% |
| pandas (walltime) | +2% |
| altair (walltime) | +2% |
| static-frame | +2% |
| anyio | +1% |
| freqtrade | +1% |
| colour-science | +1% |
| tanjun | +1% |
closes https://github.com/astral-sh/ty/issues/1091
---------
Co-authored-by: Douglas Creager <dcreager@dcreager.net>
## Summary
Properly preserve type qualifiers when accessing attributes on unions
and intersections. This is a prerequisite for
https://github.com/astral-sh/ruff/pull/19579.
Also fix a completely wrong implementation of
`map_with_boundness_and_qualifiers`. It now closely follows
`map_with_boundness` (just above).
## Test Plan
I thought about it, but didn't find any easy way to test this. This only
affected `Type::member`. Things like validation of attribute writes
(where type qualifiers like `ClassVar` and `Final` are important) were
already handling things correctly.
## Summary
Add a subtly different test case for recursive PEP 695 type aliases,
which does require that we relax our union simplification, so we don't
eagerly unpack aliases from user-provided union annotations.
## Test Plan
Added mdtest.
## Summary
This has been here for awhile (since our initial PEP 695 type alias
support) but isn't really correct. The right-hand-side of a PEP 695 type
alias is a distinct scope, and we don't mark it as an "eager" nested
scope, so it automatically gets "deferred" resolution of names from
outer scopes (just like a nested function). Thus it's
redundant/unnecessary for us to use `DeferredExpressionState::Deferred`
for resolving that RHS expression -- that's for deferring resolution of
individual names within a scope. Using it here causes us to wrongly
ignore applicable outer-scope narrowing.
## Test Plan
Added mdtest that failed before this PR (the second snippet -- the first
snippet always passed.)
## Summary
Implement validation for `TypedDict` constructor calls and dictionary
literal assignments, including support for `total=False` and proper
field management.
Also add support for `Required` and `NotRequired` type qualifiers in
`TypedDict` classes, along with proper inheritance behavior and the
`total=` parameter.
Support both constructor calls and dict literal syntax
part of https://github.com/astral-sh/ty/issues/154
### Basic Required Field Validation
```py
class Person(TypedDict):
name: str
age: int | None
# Error: Missing required field 'name' in TypedDict `Person` constructor
incomplete = Person(age=25)
# Error: Invalid argument to key "name" with declared type `str` on TypedDict `Person`
wrong_type = Person(name=123, age=25)
# Error: Invalid key access on TypedDict `Person`: Unknown key "extra"
extra_field = Person(name="Bob", age=25, extra=True)
```
<img width="773" height="191" alt="Screenshot 2025-08-07 at 17 59 22"
src="https://github.com/user-attachments/assets/79076d98-e85f-4495-93d6-a731aa72a5c9"
/>
### Support for `total=False`
```py
class OptionalPerson(TypedDict, total=False):
name: str
age: int | None
# All valid - all fields are optional with total=False
charlie = OptionalPerson()
david = OptionalPerson(name="David")
emily = OptionalPerson(age=30)
frank = OptionalPerson(name="Frank", age=25)
# But type validation and extra fields still apply
invalid_type = OptionalPerson(name=123) # Error: Invalid argument type
invalid_extra = OptionalPerson(extra=True) # Error: Invalid key access
```
### Dictionary Literal Validation
```py
# Type checking works for both constructors and dict literals
person: Person = {"name": "Alice", "age": 30}
reveal_type(person["name"]) # revealed: str
reveal_type(person["age"]) # revealed: int | None
# Error: Invalid key access on TypedDict `Person`: Unknown key "non_existing"
reveal_type(person["non_existing"]) # revealed: Unknown
```
### `Required`, `NotRequired`, `total`
```python
from typing import TypedDict
from typing_extensions import Required, NotRequired
class PartialUser(TypedDict, total=False):
name: Required[str] # Required despite total=False
age: int # Optional due to total=False
email: NotRequired[str] # Explicitly optional (redundant)
class User(TypedDict):
name: Required[str] # Explicitly required (redundant)
age: int # Required due to total=True
bio: NotRequired[str] # Optional despite total=True
# Valid constructions
partial = PartialUser(name="Alice") # name required, age optional
full = User(name="Bob", age=25) # name and age required, bio optional
# Inheritance maintains original field requirements
class Employee(PartialUser):
department: str # Required (new field)
# name: still Required (inherited)
# age: still optional (inherited)
emp = Employee(name="Charlie", department="Engineering") # ✅
Employee(department="Engineering") # ❌
e: Employee = {"age": 1} # ❌
```
<img width="898" height="683" alt="Screenshot 2025-08-11 at 22 02 57"
src="https://github.com/user-attachments/assets/4c1b18cd-cb2e-493a-a948-51589d121738"
/>
## Implementation
The implementation reuses existing validation logic done in
https://github.com/astral-sh/ruff/pull/19782
### ℹ️ Why I did NOT synthesize an `__init__` for `TypedDict`:
`TypedDict` inherits `dict.__init__(self, *args, **kwargs)` that accepts
all arguments.
The type resolution system finds this inherited signature **before**
looking for synthesized members.
So `own_synthesized_member()` is never called because a signature
already exists.
To force synthesis, you'd have to override Python’s inheritance
mechanism, which would break compatibility with the existing ecosystem.
This is why I went with ad-hoc validation. IMO it's the only viable
approach that respects Python’s
inheritance semantics while providing the required validation.
### Refacto of `Field`
**Before:**
```rust
struct Field<'db> {
declared_ty: Type<'db>,
default_ty: Option<Type<'db>>, // NamedTuple and dataclass only
init_only: bool, // dataclass only
init: bool, // dataclass only
is_required: Option<bool>, // TypedDict only
}
```
**After:**
```rust
struct Field<'db> {
declared_ty: Type<'db>,
kind: FieldKind<'db>,
}
enum FieldKind<'db> {
NamedTuple { default_ty: Option<Type<'db>> },
Dataclass { default_ty: Option<Type<'db>>, init_only: bool, init: bool },
TypedDict { is_required: bool },
}
```
## Test Plan
Updated Markdown tests
---------
Co-authored-by: David Peter <mail@david-peter.de>
## Summary
This PR limits the argument type expansion size for an overload call
evaluation to 512.
The limit chosen is arbitrary but I've taken the 256 limit from Pyright
into account and bumped it x2 to start with.
Initially, I actually started out by trying to refactor the entire
argument type expansion to be lazy. Currently, expanding a single
argument at any position eagerly creates the combination (argument
lists) and returns that (`Vec<CallArguments>`) but I thought we could
make it lazier by converting the return type of `expand` from
`Iterator<Item = Vec<CallArguments>>` to `Iterator<Item = Iterator<Item
= CallArguments>>` but that's proving to be difficult to implement
mainly because we **need** to maintain the previous expansion to
generate the next expansion which is the main reason to use
`std::iter::successors` in the first place.
Another approach would be to eagerly expand all the argument types and
then use the `combinations` from `itertools` to generate the
combinations but we would need to find the "boundary" between arguments
lists produced from expanding argument at position 1 and position 2
because that's important for the algorithm.
Closes: https://github.com/astral-sh/ty/issues/868
## Test Plan
Add test case to demonstrate the limit along with the diagnostic
snapshot stating that the limit has been reached.
Part of astral-sh/ty#994
## Summary
Add new special forms to `ty_extensions`, `Top[T]` and `Bottom[T]`.
Remove `ty_extensions.top_materialization` and
`ty_extensions.bottom_materialization`.
## Test Plan
Converted the existing `materialization.md` mdtest to the new syntax.
Added some tests for invalid use of the new special form.
## Summary
Removes the `module_ptr` field from `AstNodeRef` in release mode, and
change `NodeIndex` to a `NonZeroU32` to reduce the size of
`Option<AstNodeRef<_>>` fields.
I believe CI runs in debug mode, so this won't show up in the memory
report, but this reduces memory by ~2% in release mode.
## Summary
Previously we held off from doing this because we weren't sure that it
was worth the added complexity cost. But our code has changed in the
months since we made that initial decision, and I think the structure of
the code is such that it no longer really leads to much added complexity
to add precise inference when unpacking a string literal or a bytes
literal.
The improved inference we gain from this has real benefits to users (see
the mypy_primer report), and this PR doesn't appear to have a
performance impact.
## Test plan
mdtests
## Summary
We use the `System` abstraction in ty to abstract away the host/system
on which ty runs.
This has a few benefits:
* Tests can run in full isolation using a memory system (that uses an
in-memory file system)
* The LSP has a custom implementation where `read_to_string` returns the
content as seen by the editor (e.g. unsaved changes) instead of always
returning the content as it is stored on disk
* We don't require any file system polyfills for wasm in the browser
However, it does require extra care that we don't accidentally use
`std::fs` or `std::env` (etc.) methods in ty's code base (which is very
easy).
This PR sets up Clippy and disallows the most common methods, instead
pointing users towards the corresponding `System` methods.
The setup is a bit awkward because clippy doesn't support inheriting
configurations. That means, a crate can only override the entire
workspace configuration or not at all.
The approach taken in this PR is:
* Configure the disallowed methods at the workspace level
* Allow `disallowed_methods` at the workspace level
* Enable the lint at the crate level using the warn attribute (in code)
The obvious downside is that it won't work if we ever want to disallow
other methods, but we can figure that out once we reach that point.
What about false positives: Just add an `allow` and move on with your
life :) This isn't something that we have to enforce strictly; the goal
is to catch accidental misuse.
## Test Plan
Clippy found a place where we incorrectly used `std::fs::read_to_string`
## Summary
Rename `TypeAliasType::Bare` to `TypeAliasType::ManualPEP695`, and
`BareTypeAliasType` to `ManualPEP695TypeAliasType`.
Why?
Both existing variants of `TypeAliasType` are specific to features added
in PEP 695 (which introduced both the `type` statement and
`types.TypeAliasType`), so it doesn't make sense to name one with the
name `PEP695` and not the other.
A "bare" type alias, in my mind, is a legacy type alias like `IntOrStr =
int | str`, which is "bare" in that there is nothing at all
distinguishing it as a type alias. I will want to use the "bare" name
for this variant, in a future PR.
The renamed variant here describes a type alias created with `IntOrStr =
types.TypeAliasType("IntOrStr", int | str)`, which is not "bare", it's
just "manually" instantiated instead of using the `type` statement
syntax sugar. (This is useful when using the `typing_extensions`
backport of `TypeAliasType` on older Python versions.)
## Test Plan
Pure rename, existing tests pass.
## Summary
This PR fixes https://github.com/astral-sh/ty/issues/1071
The core issue is that `CallableType` is a salsa interned but
`Signature` (which `CallableType` stores) ignores the `Definition` in
its `Eq` and `Hash` implementation.
This PR tries to simplest fix by removing the custom `Eq` and `Hash`
implementation. The main downside of this fix is that it can increase
memory usage because `CallableType`s that are equal except for their
`Definition` are now interned separately.
The alternative is to remove `Definition` from `CallableType` and
instead, call `bindings` directly on the callee (call_expression.func).
However, this would require
addressing the TODO
here
39ee71c2a5/crates/ty_python_semantic/src/types.rs (L4582-L4586)
This might probably be worth addressing anyway, but is the more involved
fix. That's why I opted for removing the custom `Eq` implementation.
We already "ignore" the definition during normalization, thank's to
Alex's work in https://github.com/astral-sh/ruff/pull/19615
## Test Plan
https://github.com/user-attachments/assets/248d1cb1-12fd-4441-adab-b7e0866d23eb
"Why would you do this? This looks like you just replaced `bool` with an
overly complex trait"
Yes that's correct!
This should be a no-op refactoring. It replaces all of the logic in our
assignability, subtyping, equivalence, and disjointness methods to work
over an arbitrary `Constraints` trait instead of only working on `bool`.
The methods that `Constraints` provides looks very much like what we get
from `bool`. But soon we will add a new impl of this trait, and some new
methods, that let us express "fuzzy" constraints that aren't always true
or false. (In particular, a constraint will express the upper and lower
bounds of the allowed specializations of a typevar.)
Even once we have that, most of the operations that we perform on
constraint sets will be the usual boolean operations, just on sets.
(`false` becomes empty/never; `true` becomes universe/always; `or`
becomes union; `and` becomes intersection; `not` becomes negation.) So
it's helpful to have this separate PR to refactor how we invoke those
operations without introducing the new functionality yet.
Note that we also have translations of `Option::is_some_and` and
`is_none_or`, and of `Iterator::any` and `all`, and that the `and`,
`or`, `when_any`, and `when_all` methods are meant to short-circuit,
just like the corresponding boolean operations. For constraint sets,
that depends on being able to implement the `is_always` and `is_never`
trait methods.
---------
Co-authored-by: Carl Meyer <carl@astral.sh>
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
Part of: https://github.com/astral-sh/ty/issues/868
This PR adds a heuristic to avoid argument type expansion if it's going
to eventually lead to no matching overload.
This is done by checking whether the non-expandable argument types are
assignable to the corresponding annotated parameter type. If one of them
is not assignable to all of the remaining overloads, then argument type
expansion isn't going to help.
## Test Plan
Add mdtest that would otherwise take a long time because of the number
of arguments that it would need to expand (30).
This commit corrects the type checker's behavior when handling
`dataclass_transform` decorators that don't explicitly specify
`field_specifiers`. According to [PEP 681 (Data Class
Transforms)](https://peps.python.org/pep-0681/#dataclass-transform-parameters),
when `field_specifiers` is not provided, it defaults to an empty tuple,
meaning no field specifiers are supported and
`dataclasses.field`/`dataclasses.Field` calls should be ignored.
Fixes https://github.com/astral-sh/ty/issues/980
This basically splits `list_modules` into a higher level "aggregation"
routine and a lower level "get modules for one search path" routine.
This permits Salsa to cache the lower level components, e.g., many
search paths refer to directories that rarely change. This saves us
interaction with the system.
This did require a fair bit of surgery in terms of being careful about
adding file roots. Namely, now that we rely even more on file roots
existing for correct handling of cache invalidation, there were several
spots in our code that needed to be updated to add roots (that we
weren't previously doing). This feels Not Great, and it would be better
if we had some kind of abstraction that handled this for us. But it
isn't clear to me at this time what that looks like.
This ensures there is some level of consistency between the APIs.
This did require exposing a couple more things on `Module` for good
error messages. This also motivated a switch to an interned struct
instead of a tracked struct. This ensures that `list_modules` and
`resolve_modules` reuse the same `Module` values when the inputs are the
same.
Ref https://github.com/astral-sh/ruff/pull/19883#discussion_r2272520194
This makes `import <CURSOR>` and `from <CURSOR>` completions work.
This also makes `import os.<CURSOR>` and `from os.<CURSOR>`
completions work. In this case, we are careful to only offer
submodule completions.
The actual implementation wasn't too bad. It's not long
but pretty fiddly. I copied over the tests from the existing
module resolver and adapted them to work with this API. Then
I added a number of my own tests as well.
Previously, if the module was just `foo-stubs`, we'd skip over
stripping the `-stubs` suffix which would lead to us returning
`None`.
This function is now a little convoluted and could be simpler
if we did an intermediate allocation. But I kept the iterative
approach and added a special case to handle `foo-stubs`.
These tests capture existing behavior.
I added these when I stumbled upon what I thought was an
oddity: we prioritize `foo.pyi` over `foo.py`, but
prioritize `foo/__init__.py` over `foo.pyi`.
(I plan to investigate this more closely in follow-up
work. Particularly, to look at other type checkers. It
seems like we may want to change this to always prioritize
stubs.)
This is a port of the logic in https://github.com/astral-sh/uv/pull/7691
The basic idea is we use CONDA_DEFAULT_ENV as a signal for whether
CONDA_PREFIX is just the ambient system conda install, or the user has
explicitly activated a custom one. If the former, then the conda is
treated like a system install (having lowest priority). If the latter,
the conda is treated like an activated venv (having priority over
everything but an Actual activated venv).
Fixes https://github.com/astral-sh/ty/issues/611
## Summary
Closes: https://github.com/astral-sh/ty/issues/669
(This turned out to be simpler that I thought :))
## Test Plan
Update existing test cases.
### Ecosystem report
Most of them are basically because ty has now started inferring more
precise types for the return type to an overloaded call and a lot of the
types are defined using type aliases, here's some examples:
<details><summary>Details</summary>
<p>
> attrs (https://github.com/python-attrs/attrs)
> + tests/test_make.py:146:14: error[unresolved-attribute] Type
`Literal[42]` has no attribute `default`
> - Found 555 diagnostics
> + Found 556 diagnostics
This is accurate now that we infer the type as `Literal[42]` instead of
`Unknown` (Pyright infers it as `int`)
> optuna (https://github.com/optuna/optuna)
> + optuna/_gp/search_space.py:181:53: error[invalid-argument-type]
Argument to function `_round_one_normalized_param` is incorrect:
Expected `tuple[int | float, int | float]`, found `tuple[Unknown |
ndarray[Unknown, <class 'float'>], Unknown | ndarray[Unknown, <class
'float'>]]`
> + optuna/_gp/search_space.py:181:83: error[invalid-argument-type]
Argument to function `_round_one_normalized_param` is incorrect:
Expected `int | float`, found `Unknown | ndarray[Unknown, <class
'float'>]`
> + tests/gp_tests/test_search_space.py:109:13:
error[invalid-argument-type] Argument to function
`_unnormalize_one_param` is incorrect: Expected `tuple[int | float, int
| float]`, found `Unknown | ndarray[Unknown, <class 'float'>]`
> + tests/gp_tests/test_search_space.py:110:13:
error[invalid-argument-type] Argument to function
`_unnormalize_one_param` is incorrect: Expected `int | float`, found
`Unknown | ndarray[Unknown, <class 'float'>]`
> - Found 559 diagnostics
> + Found 563 diagnostics
Same as above where ty is now inferring a more precise type like
`Unknown | ndarray[tuple[int, int], <class 'float'>]` instead of just
`Unknown` as before
> jinja (https://github.com/pallets/jinja)
> + src/jinja2/bccache.py:298:39: error[invalid-argument-type] Argument
to bound method `write_bytecode` is incorrect: Expected `IO[bytes]`,
found `_TemporaryFileWrapper[str]`
> - Found 186 diagnostics
> + Found 187 diagnostics
This requires support for type aliases to match the correct overload.
> hydra-zen (https://github.com/mit-ll-responsible-ai/hydra-zen)
> + src/hydra_zen/wrapper/_implementations.py:945:16:
error[invalid-return-type] Return type does not match returned value:
expected `DataClass_ | type[@Todo(type[T] for protocols)] | ListConfig |
DictConfig`, found `@Todo(unsupported type[X] special form) | (((...) ->
Any) & dict[Unknown, Unknown]) | (DataClass_ & dict[Unknown, Unknown]) |
dict[Any, Any] | (ListConfig & dict[Unknown, Unknown]) | (DictConfig &
dict[Unknown, Unknown]) | (((...) -> Any) & list[Unknown]) | (DataClass_
& list[Unknown]) | list[Any] | (ListConfig & list[Unknown]) |
(DictConfig & list[Unknown])`
> + tests/annotations/behaviors.py:60:28: error[call-non-callable]
Object of type `Path` is not callable
> + tests/annotations/behaviors.py:64:21: error[call-non-callable]
Object of type `Path` is not callable
> + tests/annotations/declarations.py:167:17: error[call-non-callable]
Object of type `Path` is not callable
> + tests/annotations/declarations.py:524:17:
error[unresolved-attribute] Type `<class 'int'>` has no attribute
`_target_`
> - Found 561 diagnostics
> + Found 566 diagnostics
Same as above, this requires support for type aliases to match the
correct overload.
> paasta (https://github.com/yelp/paasta)
> + paasta_tools/utils.py:4188:19: warning[redundant-cast] Value is
already of type `list[str]`
> - Found 888 diagnostics
> + Found 889 diagnostics
This is correct.
> colour (https://github.com/colour-science/colour)
> + colour/plotting/diagrams.py:448:13: error[invalid-argument-type]
Argument to bound method `__init__` is incorrect: Expected
`Sequence[@Todo(Support for `typing.TypeAlias`)]`, found
`ndarray[tuple[int, int, int], dtype[Unknown]]`
> + colour/plotting/diagrams.py:462:13: error[invalid-argument-type]
Argument to bound method `__init__` is incorrect: Expected
`Sequence[@Todo(Support for `typing.TypeAlias`)]`, found
`ndarray[tuple[int, int, int], dtype[Unknown]]`
> + colour/plotting/models.py:419:13: error[invalid-argument-type]
Argument to bound method `__init__` is incorrect: Expected
`Sequence[@Todo(Support for `typing.TypeAlias`)]`, found
`ndarray[tuple[int, int, int], dtype[Unknown]]`
> + colour/plotting/temperature.py:230:9: error[invalid-argument-type]
Argument to bound method `__init__` is incorrect: Expected
`Sequence[@Todo(Support for `typing.TypeAlias`)]`, found
`ndarray[tuple[int, int, int], dtype[Unknown]]`
> + colour/plotting/temperature.py:474:13: error[invalid-argument-type]
Argument to bound method `__init__` is incorrect: Expected
`Sequence[@Todo(Support for `typing.TypeAlias`)]`, found
`ndarray[tuple[int, int, int], dtype[Unknown]]`
> + colour/plotting/temperature.py:495:17: error[invalid-argument-type]
Argument to bound method `__init__` is incorrect: Expected
`Sequence[@Todo(Support for `typing.TypeAlias`)]`, found
`ndarray[tuple[int, int, int], dtype[Unknown]]`
> + colour/plotting/temperature.py:513:13: error[invalid-argument-type]
Argument to bound method `text` is incorrect: Expected `int | float`,
found `ndarray[@Todo(Support for `typing.TypeAlias`), dtype[Unknown]]`
> + colour/plotting/temperature.py:514:13: error[invalid-argument-type]
Argument to bound method `text` is incorrect: Expected `int | float`,
found `ndarray[@Todo(Support for `typing.TypeAlias`), dtype[Unknown]]`
> - Found 480 diagnostics
> + Found 488 diagnostics
Most of them are correct except for the last two diagnostics which I'm
not sure
what's happening, it's trying to index into an `np.ndarray` type (which
is
inferred correctly) but I think it might be picking up an incorrect
overload
for the `__getitem__` method.
Scipy's diagnostics also requires support for type alises to pick the
correct overload.
</p>
</details>
In implementing partial stubs I had observed that this continue in the
namespace package code seemed erroneous since the same continue for
partial stubs didn't work. Unfortunately I wasn't confident enough to
push on that hunch. Fortunately I remembered that hunch to make this an
easy fix.
The issue with the continue is that it bails out of the current
search-path without testing any .py files. This breaks when for example
`google` and `google-stubs`/`types-google` are both in the same
site-packages dir -- failing to find a module in `types-google` has us
completely skip over `google`!
Fixes https://github.com/astral-sh/ty/issues/520
fix https://github.com/astral-sh/ty/issues/1047
## Summary
This PR fixes how `KW_ONLY` is applied in dataclasses. Previously, the
sentinel leaked into subclasses and incorrectly marked their fields as
keyword-only; now it only affects fields declared in the same class.
```py
from dataclasses import dataclass, KW_ONLY
@dataclass
class D:
x: int
_: KW_ONLY
y: str
@dataclass
class E(D):
z: bytes
# This should work: x=1 (positional), z=b"foo" (positional), y="foo" (keyword-only)
E(1, b"foo", y="foo")
reveal_type(E.__init__) # revealed: (self: E, x: int, z: bytes, *, y: str) -> None
```
<!-- What's the purpose of the change? What does it do, and why? -->
## Test Plan
<!-- How was it tested? -->
mdtests
Requires some iteration, but this includes the most tedious part --
threading a new concept of DisplaySettings through every type display
impl. Currently it only holds a boolean for multiline, but in the future
it could also take other things like "render to markdown" or "here's
your base indent if you make a newline".
For types which have exposed display functions I've left the old
signature as a compatibility polyfill to avoid having to audit
everywhere that prints types right off the bat (notably I originally
tried doing multiline functions unconditionally and a ton of things
churned that clearly weren't ready for multi-line (diagnostics).
The only real use of this API in this PR is to multiline render function
types in hovers, which is the highest impact (see snapshot changes).
Fixes https://github.com/astral-sh/ty/issues/1000
## Summary
Fixes https://github.com/astral-sh/ty/issues/1046
We special-case iteration of certain types because they may have a more
detailed tuple-spec. Now that type aliases are a distinct type variant,
we need to handle them as well.
I don't love that `Type::TypeAlias` means we have to remember to add a
case for it basically anywhere we are special-casing a certain kind of
type, but at the moment I don't have a better plan. It's another
argument for avoiding fallback cases in `Type` matches, which we usually
prefer; I've updated this match statement to be comprehensive.
## Test Plan
Added mdtest.
`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
This PR adds a new lint, `invalid-await`, for all sorts of reasons why
an object may not be `await`able, as discussed in astral-sh/ty#919.
Precisely, `__await__` is guarded against being missing, possibly
unbound, or improperly defined (expects additional arguments or doesn't
return an iterator).
Of course, diagnostics need to be fine-tuned. If `__await__` cannot be
called with no extra arguments, it indicates an error (or a quirk?) in
the method signature, not at the call site. Without any doubt, such an
object is not `Awaitable`, but I feel like talking about arguments for
an *implicit* call is a bit leaky.
I didn't reference any actual diagnostic messages in the lint
definition, because I want to hear feedback first.
Also, there's no mention of the actual required method signature for
`__await__` anywhere in the docs. The only reference I had is the
`typing` stub. I basically ended up linking `[Awaitable]` to ["must
implement
`__await__`"](https://docs.python.org/3/library/collections.abc.html#collections.abc.Awaitable),
which is insufficient on its own.
## Test Plan
The following code was tested:
```python
import asyncio
import typing
class Awaitable:
def __await__(self) -> typing.Generator[typing.Any, None, int]:
yield None
return 5
class NoDunderMethod:
pass
class InvalidAwaitArgs:
def __await__(self, value: int) -> int:
return value
class InvalidAwaitReturn:
def __await__(self) -> int:
return 5
class InvalidAwaitReturnImplicit:
def __await__(self):
pass
async def main() -> None:
result = await Awaitable() # valid
result = await NoDunderMethod() # `__await__` is missing
result = await InvalidAwaitReturn() # `__await__` returns `int`, which is not a valid iterator
result = await InvalidAwaitArgs() # `__await__` expects additional arguments and cannot be called implicitly
result = await InvalidAwaitReturnImplicit() # `__await__` returns `Unknown`, which is not a valid iterator
asyncio.run(main())
```
---------
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 PR adds a type tag to the `CycleDetector` visitor (and its
aliases).
There are some places where we implement e.g. an equivalence check by
making a disjointness check. Both `is_equivalent_to` and
`is_disjoint_from` use a `PairVisitor` to handle cycles, but they should
not use the same visitor. I was finding it tedious to remember when it
was appropriate to pass on a visitor and when not to. This adds a
`PhantomData` type tag to ensure that we can't pass on one method's
visitor to a different method.
For `has_relation` and `apply_type_mapping`, we have an existing type
that we can use as the tag. For the other methods, I've added empty
structs (`Normalized`, `IsDisjointFrom`, `IsEquivalentTo`) to use as
tags.
This also reintroduces the `ResolvedDefinition::Module` variant because
reverse-engineering it in several places is a bit confusing. In an ideal
world we wouldn't have `ResolvedDefinition::FileWithRange` as it kinda
kills the ability to do richer analysis, so I want to chip away at its
scope wherever I can (currently it's used to point at asname parts of
import statements when doing `ImportAliasResolution::PreserveAliases`,
and also keyword arguments).
This also makes a kind of odd change to allow a hover to *only* produce
a docstring. This works around an oddity where hovering over a module
name in an import fails to resolve to a `ty` even though hovering over
uses of that imported name *does*.
The two fixed tests reflect the two interesting cases here.
## Summary
A [passing
comment](https://github.com/astral-sh/ruff/pull/19711#issuecomment-3169312014)
led me to explore why we didn't report a class attribute as possibly
unbound if it was a method and defined in two different conditional
branches.
I found that the reason was because of our handling of "conflicting
declarations" in `place_from_declarations`. It returned a `Result` which
would be `Err` in case of conflicting declarations.
But we only actually care about conflicting declarations when we are
actually doing type inference on that scope and might emit a diagnostic
about it. And in all cases (including that one), we want to otherwise
proceed with the union of the declared types, as if there was no
conflict.
In several cases we were failing to handle the union of declared types
in the same way as a normal declared type if there was a declared-types
conflict. The `Result` return type made this mistake really easy to
make, as we'd match on e.g. `Ok(Place::Type(...))` and do one thing,
then match on `Err(...)` and do another, even though really both of
those cases should be handled the same.
This PR refactors `place_from_declarations` to instead return a struct
which always represents the declared type we should use in the same way,
as well as carrying the conflicting declared types, if any. This struct
has a method to allow us to explicitly ignore the declared-types
conflict (which is what we want in most cases), as well as a method to
get the declared type and the conflict information, in the case where we
want to emit a diagnostic on the conflict.
## Test Plan
Existing CI; added a test showing that we now understand a
multiply-conditionally-defined method as possibly-unbound.
This does trigger issues on a couple new fuzzer seeds, but the issues
are just new instances of an already-known (and rarely occurring)
problem which I already plan to address in a future PR, so I think it's
OK to land as-is.
I happened to build this initially on top of
https://github.com/astral-sh/ruff/pull/19711, which adds invalid-await
diagnostics, so I also updated some invalid-syntax tests to not await on
an invalid type, since the purpose of those tests is to check the
syntactic location of the `await`, not the validity of the awaited type.
## Summary
Support recursive type aliases by adding a `Type::TypeAlias` type
variant, which allows referring to a type alias directly as a type
without eagerly unpacking it to its value.
We still unpack type aliases when they are added to intersections and
unions, so that we can simplify the intersection/union appropriately
based on the unpacked value of the type alias.
This introduces new possible recursive types, and so also requires
expanding our usage of recursion-detecting visitors in Type methods. The
use of these visitors is still not fully comprehensive in this PR, and
will require further expansion to support recursion in more kinds of
types (I already have further work on this locally), but I think it may
be better to do this incrementally in multiple PRs.
## Test Plan
Added some recursive type-alias tests and made them pass.
## Summary
After https://github.com/astral-sh/ruff/pull/19871, I realized that now
that we are passing around shared references to `CycleDetector`
visitors, we can now also simplify the `visit` callback signature; we
don't need to smuggle a single visitor reference through it anymore.
This is a pretty minor simplification, and it doesn't really make
anything shorter since I typically used a very short name (`v`) for the
smuggled reference, but I think it reduces cognitive overhead in reading
these `visit` usages; the extra variable would likely be confusing
otherwise for a reader.
## Test Plan
Existing CI.
## Summary
Type visitors are conceptually immutable, they just internally track the
types they've seen (and some maintain a cache of results.) Passing
around mutable visitors everywhere can get us into borrow-checker
trouble in some cases, where we need to recursively pass along the
visitor inside more than one closure with non-disjoint lifetime.
Use interior mutability (via `RefCell` and `Cell`) inside the visitors
instead, to allow us to pass around shared references.
## Test Plan
Existing tests.
The [minimal
reproduction](https://gist.github.com/dcreager/fc53c59b30d7ce71d478dcb2c1c56444)
of https://github.com/astral-sh/ty/issues/948 is an example of a class
with implicit attributes whose types end up depending on themselves. Our
existing cycle detection for `infer_expression_types` is usually enough
to handle this situation correctly, but when there are very many of
these implicit attributes, we get a combinatorial explosion of running
time and memory usage.
Adding a separate cycle handler for `ClassLiteral::implicit_attribute`
lets us catch and recover from this situation earlier.
Closes https://github.com/astral-sh/ty/issues/948
by using essentially the same logic for system site-packages, on the
assumption that system site-packages are always a subdir of the stdlib
we were looking for.
fix https://github.com/astral-sh/ty/issues/943
## Summary
Add module-level `__getattr__` support for ty's type checker, fixing
issue https://github.com/astral-sh/ty/issues/943.
Module-level `__getattr__` functions ([PEP
562](https://peps.python.org/pep-0562/)) are now respected when
resolving dynamic attributes, matching the behavior of mypy and pyright.
## Implementation
Thanks @sharkdp for the guidance in
https://github.com/astral-sh/ty/issues/943#issuecomment-3157566579
- Adds module-specific `__getattr__` resolution in
`ModuleLiteral.static_member()`
- Maintains proper attribute precedence: explicit attributes >
submodules > `__getattr__`
## Test Plan
- New mdtest covering basic functionality, type annotations, attribute
precedence, and edge cases
(run ```cargo nextest run -p ty_python_semantic
mdtest__import_module_getattr```)
- All new tests pass, verifying `__getattr__` is called correctly and
returns proper types
- Existing test suite passes, ensuring no regressions introduced
Alex Waygood
Douglas Creager
Renkai Ge
Shunsuke Shibayama
Ibraheem Ahmed
David Peter
justin
Carl Meyer
Eric Mark Martin
Jelle Zijlstra
Samuel Rigaud
Bhuminjay Soni
Andrew Gallant
Eric Jolibois
Dhruv Manilawala
Shaygan Hooshyari
Leandro Braga
Renkai Ge
Dylan
github-actions[bot]
Micha Reiser
Aria Desires
Matthew Mckee
Andrii Turov