TypeAdapter fails to validate against generic dataclass when using different type parameters

[timwie]

Initial Checks

• I confirm that I'm using Pydantic V2

Description

I've just started using pydantic, and I'm using TypeAdapters to turn JSON strings into dataclass instances. This works really well until I try to validate the same generic class with a different type parameter. It seems to me that in my example, once you create a TypeAdapter for Outer[int], you can only ever validate that exact type, and not types that use a different type parameter like Outer[str].

Example Code

from pydantic import TypeAdapter
from pydantic.dataclasses import dataclass
from typing import Generic, TypeVar

T = TypeVar("T")

@dataclass
class Outer(Generic[T]):
    data: T

json1 = '{"data": 42}'
json2 = '{"data": "something else"}'

adapter1 = TypeAdapter(Outer[int])
adapter2 = TypeAdapter(Outer[str])

assert Outer(42) == adapter1.validate_json(json1, strict=True)  # works
assert Outer("something else") == adapter2.validate_json(json2, strict=True)  # crashes

# pydantic_core._pydantic_core.ValidationError: 1 validation error for Outer
# data
#   Input should be a valid integer [type=int_type, input_value='something else', input_type=str]
#     For further information visit https://errors.pydantic.dev/2.3/v/int_type

Python, Pydantic & OS Version

pydantic version: 2.3.0
        pydantic-core version: 2.6.3
          pydantic-core build: profile=release pgo=false
                 install path: /Users/timwie/.pyenv/versions/3.9.17/envs/aio-taginfo39/lib/python3.9/site-packages/pydantic
               python version: 3.9.17 (main, Jun 26 2023, 11:24:22)  [Clang 14.0.3 (clang-1403.0.22.14.1)]
                     platform: macOS-13.5.2-arm64-arm-64bit
     optional deps. installed: ['typing-extensions']

[sydney-runkle]

Hey @timwie, thanks for bringing this to our attention. This is a bug. I'm working on a fix now and will get a PR up shortly.