chore(langchain): remove generic from FakeToolCallingModel (#34572)

* Making `FakeToolCallingModel` generic on its `structured_response`
doesn't help anywhere in typing.
* There are more than 120 references of `FakeToolCallingModel` in the
code where you get ` error: Need type annotation for "model"
[var-annotated]` because mypy can't resolve the generic type (we don't
see them atm because they are in files temporarily excluded from mypy
checking). We would need to explicitly type them to
`FakeToolCallingModel[Any]`

Co-authored-by: Mason Daugherty <mason@langchain.dev>

libs/langchain_v1/tests/unit_tests/agents/model.py

@@ -3,9 +3,7 @@ from collections.abc import Callable, Sequence
 from dataclasses import asdict, is_dataclass
 from typing import (
     Any,
-    Generic,
     Literal,
-    TypeVar,
 )
 
 from langchain_core.callbacks import CallbackManagerForLLMRun
@@ -19,13 +17,12 @@ from langchain_core.outputs import ChatGeneration, ChatResult
 from langchain_core.runnables import Runnable
 from langchain_core.tools import BaseTool
 from pydantic import BaseModel
-
-StructuredResponseT = TypeVar("StructuredResponseT")
+from typing_extensions import override
 
 
-class FakeToolCallingModel(BaseChatModel, Generic[StructuredResponseT]):
-    tool_calls: list[list[ToolCall]] | list[list[dict]] | None = None
-    structured_response: StructuredResponseT | None = None
+class FakeToolCallingModel(BaseChatModel):
+    tool_calls: list[list[ToolCall]] | list[list[dict[str, Any]]] | None = None
+    structured_response: Any | None = None
     index: int = 0
     tool_style: Literal["openai", "anthropic"] = "openai"
 
@@ -52,7 +49,9 @@ class FakeToolCallingModel(BaseChatModel, Generic[StructuredResponseT]):
         if is_native and not tool_calls:
             if isinstance(self.structured_response, BaseModel):
                 content_obj = self.structured_response.model_dump()
-            elif is_dataclass(self.structured_response):
+            elif is_dataclass(self.structured_response) and not isinstance(
+                self.structured_response, type
+            ):
                 content_obj = asdict(self.structured_response)
             elif isinstance(self.structured_response, dict):
                 content_obj = self.structured_response
@@ -71,11 +70,14 @@ class FakeToolCallingModel(BaseChatModel, Generic[StructuredResponseT]):
     def _llm_type(self) -> str:
         return "fake-tool-call-model"
 
+    @override
     def bind_tools(
         self,
-        tools: Sequence[dict[str, Any] | type[BaseModel] | Callable | BaseTool],
+        tools: Sequence[dict[str, Any] | type | Callable[..., Any] | BaseTool],
+        *,
+        tool_choice: str | None = None,
         **kwargs: Any,
-    ) -> Runnable[LanguageModelInput, BaseMessage]:
+    ) -> Runnable[LanguageModelInput, AIMessage]:
         if len(tools) == 0:
             msg = "Must provide at least one tool"
             raise ValueError(msg)

libs/langchain_v1/tests/unit_tests/agents/test_react_agent.py

@@ -362,7 +362,7 @@
 #         return "The weather is sunny and 75°F."
 
 #     expected_structured_response = WeatherResponse(temperature=75)
-#     model = FakeToolCallingModel[WeatherResponse](
+#     model = FakeToolCallingModel(
 #         tool_calls=tool_calls, structured_response=expected_structured_response
 #     )
 #     agent = create_agent(

libs/langchain_v1/tests/unit_tests/agents/test_response_format.py

@@ -343,7 +343,7 @@ class TestResponseFormatAsToolStrategy:
             ],
         ]
 
-        model = FakeToolCallingModel[WeatherBaseModel | LocationResponse](tool_calls=tool_calls)
+        model = FakeToolCallingModel(tool_calls=tool_calls)
 
         agent = create_agent(
             model,
@@ -655,7 +655,7 @@ class TestResponseFormatAsProviderStrategy:
             [{"args": {}, "id": "1", "name": "get_weather"}],
         ]
 
-        model = FakeToolCallingModel[WeatherBaseModel](
+        model = FakeToolCallingModel(
             tool_calls=tool_calls, structured_response=EXPECTED_WEATHER_PYDANTIC
         )
 
@@ -678,7 +678,7 @@ class TestResponseFormatAsProviderStrategy:
         ]
 
         # But we're using WeatherBaseModel which has different field requirements
-        model = FakeToolCallingModel[dict](
+        model = FakeToolCallingModel(
             tool_calls=tool_calls,
             structured_response={"invalid": "data"},  # Wrong structure
         )
@@ -699,7 +699,7 @@ class TestResponseFormatAsProviderStrategy:
             [{"args": {}, "id": "1", "name": "get_weather"}],
         ]
 
-        model = FakeToolCallingModel[WeatherDataclass](
+        model = FakeToolCallingModel(
             tool_calls=tool_calls, structured_response=EXPECTED_WEATHER_DATACLASS
         )
 
@@ -719,7 +719,7 @@ class TestResponseFormatAsProviderStrategy:
             [{"args": {}, "id": "1", "name": "get_weather"}],
         ]
 
-        model = FakeToolCallingModel[WeatherTypedDict](
+        model = FakeToolCallingModel(
             tool_calls=tool_calls, structured_response=EXPECTED_WEATHER_DICT
         )
 
@@ -737,7 +737,7 @@ class TestResponseFormatAsProviderStrategy:
             [{"args": {}, "id": "1", "name": "get_weather"}],
         ]
 
-        model = FakeToolCallingModel[dict](
+        model = FakeToolCallingModel(
             tool_calls=tool_calls, structured_response=EXPECTED_WEATHER_DICT
         )
 
@@ -858,7 +858,7 @@ def test_union_of_types() -> None:
         ],
     ]
 
-    model = FakeToolCallingModel[WeatherBaseModel | LocationResponse](
+    model = FakeToolCallingModel(
         tool_calls=tool_calls, structured_response=EXPECTED_WEATHER_PYDANTIC
     )