Working with sets in snapshots

[Genfood]

I have some objects of complex nested types which I would like to compare using snapshots. But these types also use sets which are not sorted, which leads to wrong assertions, but the values inside the sets are the same. How could I prevent the assertion from failing in such cases?

[Genfood]

I wrote a custom matcher to sort the sets before parsing them. Here is a simple example:

@dataclass
class TestA:
    name: str
    b: TestB

@dataclass
class TestB:
    test_set: set[str]

def sort_sets_matcher(data: Any, path: tuple[Any, ...]) -> Any:
    if isinstance(data, set):
        return sorted(list(data))

    if dataclasses.is_dataclass(data) and not isinstance(data, type):
        # The serializer will walk this dict
        return dataclasses.asdict(data)

    return data

def test_my_dataclass(snapshot: SnapshotAssertion):
    my_object = TestA(
        name="Test Name",
        b=TestB(
            test_set={"c", "a", "z", "b"} 
        )
    )

    assert my_object == snapshot(matcher=custom_data_matcher)

I think the problem wasnt the set, but the set inside the dataclass, which will be serialized using the __repr__ method.