chore(langchain): add ruff rules TC (#31921)

See https://docs.astral.sh/ruff/rules/#flake8-type-checking-tc

libs/langchain/langchain/agents/agent.py

@@ -1185,7 +1185,7 @@ class AgentExecutor(Chain):
         to reflect the changes made in the root_validator.
         """
         if isinstance(self.agent, Runnable):
-            return cast(RunnableAgentType, self.agent)
+            return cast("RunnableAgentType", self.agent)
         return self.agent
 
     def save(self, file_path: Union[Path, str]) -> None:

libs/langchain/langchain/callbacks/streaming_aiter.py

@@ -73,7 +73,7 @@ class AsyncIteratorCallbackHandler(AsyncCallbackHandler):
                 other.pop().cancel()
 
             # Extract the value of the first completed task
-            token_or_done = cast(Union[str, Literal[True]], done.pop().result())
+            token_or_done = cast("Union[str, Literal[True]]", done.pop().result())
 
             # If the extracted value is the boolean True, the done event was set
             if token_or_done is True:

libs/langchain/langchain/chains/base.py

@@ -411,7 +411,7 @@ class Chain(RunnableSerializable[dict[str, Any], dict[str, Any]], ABC):
 
         return self.invoke(
             inputs,
-            cast(RunnableConfig, {k: v for k, v in config.items() if v is not None}),
+            cast("RunnableConfig", {k: v for k, v in config.items() if v is not None}),
             return_only_outputs=return_only_outputs,
             include_run_info=include_run_info,
         )
@@ -461,7 +461,7 @@ class Chain(RunnableSerializable[dict[str, Any], dict[str, Any]], ABC):
         }
         return await self.ainvoke(
             inputs,
-            cast(RunnableConfig, {k: v for k, v in config.items() if k is not None}),
+            cast("RunnableConfig", {k: v for k, v in config.items() if k is not None}),
             return_only_outputs=return_only_outputs,
             include_run_info=include_run_info,
         )

libs/langchain/langchain/chains/combine_documents/map_rerank.py

@@ -229,7 +229,7 @@ class MapRerankDocumentsChain(BaseCombineDocumentsChain):
         docs: list[Document],
         results: Sequence[Union[str, list[str], dict[str, str]]],
     ) -> tuple[str, dict]:
-        typed_results = cast(list[dict], results)
+        typed_results = cast("list[dict]", results)
         sorted_res = sorted(
             zip(typed_results, docs),
             key=lambda x: -int(x[0][self.rank_key]),

libs/langchain/langchain/chains/llm.py

@@ -145,7 +145,7 @@ class LLMChain(Chain):
                 **self.llm_kwargs,
             )
         results = self.llm.bind(stop=stop, **self.llm_kwargs).batch(
-            cast(list, prompts),
+            cast("list", prompts),
             {"callbacks": callbacks},
         )
         generations: list[list[Generation]] = []
@@ -172,7 +172,7 @@ class LLMChain(Chain):
                 **self.llm_kwargs,
             )
         results = await self.llm.bind(stop=stop, **self.llm_kwargs).abatch(
-            cast(list, prompts),
+            cast("list", prompts),
             {"callbacks": callbacks},
         )
         generations: list[list[Generation]] = []

libs/langchain/langchain/chains/openai_functions/qa_with_structure.py

@@ -76,11 +76,11 @@ def create_qa_with_structure_chain(
         raise ValueError(msg)
     if isinstance(schema, type) and is_basemodel_subclass(schema):
         if hasattr(schema, "model_json_schema"):
-            schema_dict = cast(dict, schema.model_json_schema())
+            schema_dict = cast("dict", schema.model_json_schema())
         else:
-            schema_dict = cast(dict, schema.schema())
+            schema_dict = cast("dict", schema.schema())
     else:
-        schema_dict = cast(dict, schema)
+        schema_dict = cast("dict", schema)
     function = {
         "name": schema_dict["title"],
         "description": schema_dict["description"],

libs/langchain/langchain/chains/query_constructor/base.py

@@ -91,7 +91,7 @@ class StructuredQueryOutputParser(BaseOutputParser[StructuredQuery]):
 
             def ast_parse(raw_filter: str) -> Optional[FilterDirective]:
                 filter_directive = cast(
-                    Optional[FilterDirective],
+                    "Optional[FilterDirective]",
                     get_parser().parse(raw_filter),
                 )
                 return fix_filter_directive(
@@ -144,7 +144,7 @@ def fix_filter_directive(
             return None
         args = [
             cast(
-                FilterDirective,
+                "FilterDirective",
                 fix_filter_directive(
                     arg,
                     allowed_comparators=allowed_comparators,

libs/langchain/langchain/chains/router/llm_router.py

@@ -137,7 +137,7 @@ class LLMRouterChain(RouterChain):
 
         prediction = self.llm_chain.predict(callbacks=callbacks, **inputs)
         return cast(
-            dict[str, Any],
+            "dict[str, Any]",
             self.llm_chain.prompt.output_parser.parse(prediction),
         )
 
@@ -149,7 +149,7 @@ class LLMRouterChain(RouterChain):
         _run_manager = run_manager or CallbackManagerForChainRun.get_noop_manager()
         callbacks = _run_manager.get_child()
         return cast(
-            dict[str, Any],
+            "dict[str, Any]",
             await self.llm_chain.apredict_and_parse(callbacks=callbacks, **inputs),
         )
 

libs/langchain/langchain/chat_models/base.py

@@ -322,7 +322,7 @@ def init_chat_model(
 
     if not configurable_fields:
         return _init_chat_model_helper(
-            cast(str, model),
+            cast("str", model),
             model_provider=model_provider,
             **kwargs,
         )
@@ -632,7 +632,7 @@ class _ConfigurableModel(Runnable[LanguageModelInput, Any]):
         **kwargs: Any,
     ) -> _ConfigurableModel:
         """Bind config to a Runnable, returning a new Runnable."""
-        config = RunnableConfig(**(config or {}), **cast(RunnableConfig, kwargs))
+        config = RunnableConfig(**(config or {}), **cast("RunnableConfig", kwargs))
         model_params = self._model_params(config)
         remaining_config = {k: v for k, v in config.items() if k != "configurable"}
         remaining_config["configurable"] = {
@@ -781,7 +781,7 @@ class _ConfigurableModel(Runnable[LanguageModelInput, Any]):
         if config is None or isinstance(config, dict) or len(config) <= 1:
             if isinstance(config, list):
                 config = config[0]
-            yield from self._model(cast(RunnableConfig, config)).batch_as_completed(  # type: ignore[call-overload]
+            yield from self._model(cast("RunnableConfig", config)).batch_as_completed(  # type: ignore[call-overload]
                 inputs,
                 config=config,
                 return_exceptions=return_exceptions,
@@ -811,7 +811,7 @@ class _ConfigurableModel(Runnable[LanguageModelInput, Any]):
             if isinstance(config, list):
                 config = config[0]
             async for x in self._model(
-                cast(RunnableConfig, config),
+                cast("RunnableConfig", config),
             ).abatch_as_completed(  # type: ignore[call-overload]
                 inputs,
                 config=config,

libs/langchain/langchain/document_loaders/blob_loaders/schema.py

@@ -1,12 +1,9 @@
-from typing import TYPE_CHECKING, Any
+from typing import Any
 
 from langchain_core.document_loaders import Blob, BlobLoader
 
 from langchain._api import create_importer
 
-if TYPE_CHECKING:
-    pass
-
 # Create a way to dynamically look up deprecated imports.
 # Used to consolidate logic for raising deprecation warnings and
 # handling optional imports.

libs/langchain/langchain/embeddings/cache.py

@@ -80,7 +80,7 @@ def _value_serializer(value: Sequence[float]) -> bytes:
 
 def _value_deserializer(serialized_value: bytes) -> list[float]:
     """Deserialize a value."""
-    return cast(list[float], json.loads(serialized_value.decode()))
+    return cast("list[float]", json.loads(serialized_value.decode()))
 
 
 # The warning is global; track emission, so it appears only once.
@@ -192,7 +192,7 @@ class CacheBackedEmbeddings(Embeddings):
                 vectors[index] = updated_vector
 
         return cast(
-            list[list[float]],
+            "list[list[float]]",
             vectors,
         )  # Nones should have been resolved by now
 
@@ -230,7 +230,7 @@ class CacheBackedEmbeddings(Embeddings):
                 vectors[index] = updated_vector
 
         return cast(
-            list[list[float]],
+            "list[list[float]]",
             vectors,
         )  # Nones should have been resolved by now
 

libs/langchain/langchain/evaluation/agents/trajectory_eval_chain.py

@@ -301,7 +301,7 @@ The following is the expected answer. Use this to measure correctness:
             chain_input,
             callbacks=_run_manager.get_child(),
         )
-        return cast(dict, self.output_parser.parse(raw_output))
+        return cast("dict", self.output_parser.parse(raw_output))
 
     async def _acall(
         self,
@@ -326,7 +326,7 @@ The following is the expected answer. Use this to measure correctness:
             chain_input,
             callbacks=_run_manager.get_child(),
         )
-        return cast(dict, self.output_parser.parse(raw_output))
+        return cast("dict", self.output_parser.parse(raw_output))
 
     @override
     def _evaluate_agent_trajectory(

libs/langchain/langchain/evaluation/parsing/base.py

@@ -166,7 +166,7 @@ class JsonEqualityEvaluator(StringEvaluator):
             dict: A dictionary containing the evaluation score.
         """
         parsed = self._parse_json(prediction)
-        label = self._parse_json(cast(str, reference))
+        label = self._parse_json(cast("str", reference))
         if isinstance(label, list):
             if not isinstance(parsed, list):
                 return {"score": 0}

libs/langchain/langchain/retrievers/document_compressors/chain_extract.py

@@ -82,7 +82,7 @@ class LLMChainExtractor(BaseDocumentCompressor):
             if len(output) == 0:
                 continue
             compressed_docs.append(
-                Document(page_content=cast(str, output), metadata=doc.metadata),
+                Document(page_content=cast("str", output), metadata=doc.metadata),
             )
         return compressed_docs
 

libs/langchain/langchain/retrievers/ensemble.py

@@ -236,7 +236,7 @@ class EnsembleRetriever(BaseRetriever):
         # Enforce that retrieved docs are Documents for each list in retriever_docs
         for i in range(len(retriever_docs)):
             retriever_docs[i] = [
-                Document(page_content=cast(str, doc)) if isinstance(doc, str) else doc
+                Document(page_content=cast("str", doc)) if isinstance(doc, str) else doc
                 for doc in retriever_docs[i]
             ]
 

libs/langchain/langchain/smith/evaluation/runner_utils.py

@@ -214,24 +214,24 @@ def _wrap_in_chain_factory(
         return lambda: lcf
     if callable(llm_or_chain_factory):
         if is_traceable_function(llm_or_chain_factory):
-            runnable_ = as_runnable(cast(Callable, llm_or_chain_factory))
+            runnable_ = as_runnable(cast("Callable", llm_or_chain_factory))
             return lambda: runnable_
         try:
             _model = llm_or_chain_factory()  # type: ignore[call-arg]
         except TypeError:
             # It's an arbitrary function, wrap it in a RunnableLambda
-            user_func = cast(Callable, llm_or_chain_factory)
+            user_func = cast("Callable", llm_or_chain_factory)
             sig = inspect.signature(user_func)
             logger.info("Wrapping function %s as RunnableLambda.", sig)
             wrapped = RunnableLambda(user_func)
             return lambda: wrapped
-        constructor = cast(Callable, llm_or_chain_factory)
+        constructor = cast("Callable", llm_or_chain_factory)
         if isinstance(_model, BaseLanguageModel):
             # It's not uncommon to do an LLM constructor instead of raw LLM,
             # so we'll unpack it for the user.
             return _model
-        if is_traceable_function(cast(Callable, _model)):
-            runnable_ = as_runnable(cast(Callable, _model))
+        if is_traceable_function(cast("Callable", _model)):
+            runnable_ = as_runnable(cast("Callable", _model))
             return lambda: runnable_
         if not isinstance(_model, Runnable):
             # This is unlikely to happen - a constructor for a model function
@@ -1104,7 +1104,7 @@ class _DatasetRunContainer:
     ) -> dict:
         results: dict = {}
         for example, output in zip(self.examples, batch_results):
-            row_result = cast(_RowResult, all_eval_results.get(str(example.id), {}))
+            row_result = cast("_RowResult", all_eval_results.get(str(example.id), {}))
             results[str(example.id)] = {
                 "input": example.inputs,
                 "feedback": row_result.get("feedback", []),
@@ -1131,7 +1131,7 @@ class _DatasetRunContainer:
                     result = evaluator(runs_list, self.examples)
                     if isinstance(result, EvaluationResult):
                         result = result.dict()
-                    aggregate_feedback.append(cast(dict, result))
+                    aggregate_feedback.append(cast("dict", result))
                     executor.submit(
                         self.client.create_feedback,
                         **result,
@@ -1148,7 +1148,7 @@ class _DatasetRunContainer:
         all_eval_results: dict = {}
         all_runs: dict = {}
         for c in self.configs:
-            for callback in cast(list, c["callbacks"]):
+            for callback in cast("list", c["callbacks"]):
                 if isinstance(callback, EvaluatorCallbackHandler):
                     eval_results = callback.logged_eval_results
                     for (_, example_id), v in eval_results.items():
@@ -1171,7 +1171,7 @@ class _DatasetRunContainer:
                         },
                     )
                     all_runs[str(callback.example_id)] = run
-        return cast(dict[str, _RowResult], all_eval_results), all_runs
+        return cast("dict[str, _RowResult]", all_eval_results), all_runs
 
     def _collect_test_results(
         self,

libs/langchain/pyproject.toml

@@ -145,8 +145,8 @@ ignore-words-list = "momento,collison,ned,foor,reworkd,parth,whats,aapply,mysogy
 [tool.ruff.lint]
 select = [
     "A",      # flake8-builtins
-    "B",      # flake8-bugbear
     "ASYNC",  # flake8-async
+    "B",      # flake8-bugbear
     "C4",     # flake8-comprehensions
     "COM",    # flake8-commas
     "D1",     # pydocstyle: missing docstring
@@ -174,11 +174,12 @@ select = [
     "RSE",    # flake8-rst-docstrings
     "RUF",    # ruff
     "S",      # flake8-bandit
-    "SLOT",   # flake8-slots
     "SIM",    # flake8-simplify
     "SLF",    # flake8-self
+    "SLOT",   # flake8-slots
     "T10",    # flake8-debugger
     "T20",    # flake8-print
+    "TC",     # flake8-type-checking
     "TID",    # flake8-tidy-imports
     "TRY",    # tryceratops
     "UP",     # pyupgrade
@@ -192,6 +193,9 @@ ignore = [
     "PLR09",   # Too many something (args, statements, etc)
     "S112",    # Rarely useful
     "RUF012",  # Doesn't play well with Pydantic
+    "TC001",   # Doesn't play well with Pydantic
+    "TC002",   # Doesn't play well with Pydantic
+    "TC003",   # Doesn't play well with Pydantic
     "UP007",   # pyupgrade: non-pep604-annotation-union
 
     # TODO rules

libs/langchain/tests/integration_tests/chat_models/test_base.py

@@ -40,7 +40,7 @@ async def test_init_chat_model_chain() -> None:
 class TestStandard(ChatModelIntegrationTests):
     @property
     def chat_model_class(self) -> type[BaseChatModel]:
-        return cast(type[BaseChatModel], init_chat_model)
+        return cast("type[BaseChatModel]", init_chat_model)
 
     @property
     def chat_model_params(self) -> dict:

libs/langchain/tests/unit_tests/agents/test_agent.py

@@ -571,7 +571,7 @@ async def test_runnable_agent_with_function_calls() -> None:
 
     def fake_parse(inputs: dict) -> Union[AgentFinish, AgentAction]:
         """A parser."""
-        return cast(Union[AgentFinish, AgentAction], next(parser_responses))
+        return cast("Union[AgentFinish, AgentAction]", next(parser_responses))
 
     @tool
     def find_pet(pet: str) -> str:
@@ -683,7 +683,7 @@ async def test_runnable_with_multi_action_per_step() -> None:
 
     def fake_parse(inputs: dict) -> Union[AgentFinish, AgentAction]:
         """A parser."""
-        return cast(Union[AgentFinish, AgentAction], next(parser_responses))
+        return cast("Union[AgentFinish, AgentAction]", next(parser_responses))
 
     @tool
     def find_pet(pet: str) -> str:

libs/langchain/tests/unit_tests/chains/query_constructor/test_parser.py

@@ -82,7 +82,7 @@ def test_parse_disallowed_operator() -> None:
 
 
 def _test_parse_value(x: Any) -> None:
-    parsed = cast(Comparison, (DEFAULT_PARSER.parse(f'eq("x", {x})')))
+    parsed = cast("Comparison", (DEFAULT_PARSER.parse(f'eq("x", {x})')))
     actual = parsed.value
     assert actual == x
 
@@ -104,14 +104,14 @@ def test_parse_list_value(x: list) -> None:
 
 @pytest.mark.parametrize("x", ['""', '" "', '"foo"', "'foo'"])
 def test_parse_string_value(x: str) -> None:
-    parsed = cast(Comparison, DEFAULT_PARSER.parse(f'eq("x", {x})'))
+    parsed = cast("Comparison", DEFAULT_PARSER.parse(f'eq("x", {x})'))
     actual = parsed.value
     assert actual == x[1:-1]
 
 
 @pytest.mark.parametrize("x", ["true", "True", "TRUE", "false", "False", "FALSE"])
 def test_parse_bool_value(x: str) -> None:
-    parsed = cast(Comparison, DEFAULT_PARSER.parse(f'eq("x", {x})'))
+    parsed = cast("Comparison", DEFAULT_PARSER.parse(f'eq("x", {x})'))
     actual = parsed.value
     expected = x.lower() == "true"
     assert actual == expected
@@ -127,7 +127,7 @@ def test_parser_unpack_single_arg_operation(op: str, arg: str) -> None:
 
 @pytest.mark.parametrize("x", ['"2022-10-20"', "'2022-10-20'", "2022-10-20"])
 def test_parse_date_value(x: str) -> None:
-    parsed = cast(Comparison, DEFAULT_PARSER.parse(f'eq("x", {x})'))
+    parsed = cast("Comparison", DEFAULT_PARSER.parse(f'eq("x", {x})'))
     actual = parsed.value["date"]
     assert actual == x.strip("'\"")
 
@@ -152,7 +152,7 @@ def test_parse_date_value(x: str) -> None:
 def test_parse_datetime_value(x: str, expected: dict) -> None:
     """Test parsing of datetime values with ISO 8601 format."""
     try:
-        parsed = cast(Comparison, DEFAULT_PARSER.parse(f'eq("publishedAt", {x})'))
+        parsed = cast("Comparison", DEFAULT_PARSER.parse(f'eq("publishedAt", {x})'))
         actual = parsed.value
         assert actual == expected, f"Expected {expected}, got {actual}"
     except ValueError as e:

libs/langchain/tests/unit_tests/llms/fake_chat_model.py

@@ -122,7 +122,7 @@ class GenericFakeChatModel(BaseChatModel):
             # Use a regular expression to split on whitespace with a capture group
             # so that we can preserve the whitespace in the output.
             assert isinstance(content, str)
-            content_chunks = cast(list[str], re.split(r"(\s)", content))
+            content_chunks = cast("list[str]", re.split(r"(\s)", content))
 
             for token in content_chunks:
                 chunk = ChatGenerationChunk(
@@ -140,7 +140,7 @@ class GenericFakeChatModel(BaseChatModel):
                     for fkey, fvalue in value.items():
                         if isinstance(fvalue, str):
                             # Break function call by `,`
-                            fvalue_chunks = cast(list[str], re.split(r"(,)", fvalue))
+                            fvalue_chunks = cast("list[str]", re.split(r"(,)", fvalue))
                             for fvalue_chunk in fvalue_chunks:
                                 chunk = ChatGenerationChunk(
                                     message=AIMessageChunk(

libs/langchain/tests/unit_tests/llms/fake_llm.py

@@ -53,7 +53,7 @@ class FakeLLM(LLM):
 
     @property
     def _get_next_response_in_sequence(self) -> str:
-        queries = cast(Mapping, self.queries)
+        queries = cast("Mapping", self.queries)
         response = queries[list(queries.keys())[self.response_index]]
         self.response_index = self.response_index + 1
         return response

libs/langchain/tests/unit_tests/storage/test_lc_store.py

@@ -22,7 +22,7 @@ def test_create_lc_store(file_store: LocalFileStore) -> None:
     """Test that a docstore is created from a base store."""
     docstore = create_lc_store(file_store)
     docstore.mset([("key1", Document(page_content="hello", metadata={"key": "value"}))])
-    fetched_doc = cast(Document, docstore.mget(["key1"])[0])
+    fetched_doc = cast("Document", docstore.mget(["key1"])[0])
     assert fetched_doc.page_content == "hello"
     assert fetched_doc.metadata == {"key": "value"}