langchain 0.3.15__tar.gz → 0.3.16__tar.gz

{langchain-0.3.15 → langchain-0.3.16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: langchain
-Version: 0.3.15
+Version: 0.3.16
 Summary: Building applications with LLMs through composability
 Home-page: https://github.com/langchain-ai/langchain
 License: MIT
@@ -16,7 +16,7 @@ Requires-Dist: PyYAML (>=5.3)
 Requires-Dist: SQLAlchemy (>=1.4,<3)
 Requires-Dist: aiohttp (>=3.8.3,<4.0.0)
 Requires-Dist: async-timeout (>=4.0.0,<5.0.0) ; python_version < "3.11"
-Requires-Dist: langchain-core (>=0.3.31,<0.4.0)
+Requires-Dist: langchain-core (>=0.3.32,<0.4.0)
 Requires-Dist: langchain-text-splitters (>=0.3.3,<0.4.0)
 Requires-Dist: langsmith (>=0.1.17,<0.4)
 Requires-Dist: numpy (>=1.22.4,<2) ; python_version < "3.12"

{langchain-0.3.15 → langchain-0.3.16}/langchain/agents/initialize.py

@@ -84,8 +84,7 @@ def initialize_agent(
             pass
     else:
         raise ValueError(
-            "Somehow both `agent` and `agent_path` are None, "
-            "this should never happen."
+            "Somehow both `agent` and `agent_path` are None, this should never happen."
         )
     return AgentExecutor.from_agent_and_tools(
         agent=agent_obj,

{langchain-0.3.15 → langchain-0.3.16}/langchain/agents/loading.py

@@ -58,8 +58,7 @@ def load_agent_from_config(
     if load_from_tools:
         if llm is None:
             raise ValueError(
-                "If `load_from_llm_and_tools` is set to True, "
-                "then LLM must be provided"
+                "If `load_from_llm_and_tools` is set to True, then LLM must be provided"
             )
         if tools is None:
             raise ValueError(

{langchain-0.3.15 → langchain-0.3.16}/langchain/callbacks/tracers/logging.py

@@ -41,6 +41,6 @@ class LoggingCallbackHandler(FunctionCallbackHandler):
         except TracerException:
             crumbs_str = ""
         self.function_callback(
-            f'{get_colored_text("[text]", color="blue")}'
-            f' {get_bolded_text(f"{crumbs_str}New text:")}\n{text}'
+            f"{get_colored_text('[text]', color='blue')}"
+            f" {get_bolded_text(f'{crumbs_str}New text:')}\n{text}"
         )

{langchain-0.3.15 → langchain-0.3.16}/langchain/chains/api/base.py

@@ -370,6 +370,7 @@ try:
         @property
         def _chain_type(self) -> str:
             return "api_chain"
+
 except ImportError:
 
     class APIChain:  # type: ignore[no-redef]

{langchain-0.3.15 → langchain-0.3.16}/langchain/chains/openai_functions/openapi.py

@@ -233,7 +233,7 @@ class SimpleRequestChain(Chain):
             response = (
                 f"{api_response.status_code}: {api_response.reason}"
                 + f"\nFor {name} "
-                + f"Called with args: {args.get('params','')}"
+                + f"Called with args: {args.get('params', '')}"
             )
         else:
             try:

{langchain-0.3.15 → langchain-0.3.16}/langchain/chains/openai_tools/extraction.py

@@ -68,7 +68,10 @@ def create_extraction_chain_pydantic(
     if not isinstance(pydantic_schemas, list):
         pydantic_schemas = [pydantic_schemas]
     prompt = ChatPromptTemplate.from_messages(
-        [("system", system_message), ("user", "{input}")]
+        [
+            ("system", system_message),
+            ("user", "{input}"),
+        ]
     )
     functions = [convert_pydantic_to_openai_function(p) for p in pydantic_schemas]
     tools = [{"type": "function", "function": d} for d in functions]

{langchain-0.3.15 → langchain-0.3.16}/langchain/chains/question_answering/refine_prompts.py

@@ -33,7 +33,11 @@ refine_template = (
     "If the context isn't useful, return the original answer."
 )
 CHAT_REFINE_PROMPT = ChatPromptTemplate.from_messages(
-    [("human", "{question}"), ("ai", "{existing_answer}"), ("human", refine_template)]
+    [
+        ("human", "{question}"),
+        ("ai", "{existing_answer}"),
+        ("human", refine_template),
+    ]
 )
 REFINE_PROMPT_SELECTOR = ConditionalPromptSelector(
     default_prompt=DEFAULT_REFINE_PROMPT,
@@ -60,7 +64,10 @@ chat_qa_prompt_template = (
     "answer any questions"
 )
 CHAT_QUESTION_PROMPT = ChatPromptTemplate.from_messages(
-    [("system", chat_qa_prompt_template), ("human", "{question}")]
+    [
+        ("system", chat_qa_prompt_template),
+        ("human", "{question}"),
+    ]
 )
 QUESTION_PROMPT_SELECTOR = ConditionalPromptSelector(
     default_prompt=DEFAULT_TEXT_QA_PROMPT,

{langchain-0.3.15 → langchain-0.3.16}/langchain/chains/sequential.py

@@ -178,7 +178,9 @@ class SimpleSequentialChain(Chain):
         _input = inputs[self.input_key]
         color_mapping = get_color_mapping([str(i) for i in range(len(self.chains))])
         for i, chain in enumerate(self.chains):
-            _input = chain.run(_input, callbacks=_run_manager.get_child(f"step_{i+1}"))
+            _input = chain.run(
+                _input, callbacks=_run_manager.get_child(f"step_{i + 1}")
+            )
             if self.strip_outputs:
                 _input = _input.strip()
             _run_manager.on_text(
@@ -196,7 +198,7 @@ class SimpleSequentialChain(Chain):
         color_mapping = get_color_mapping([str(i) for i in range(len(self.chains))])
         for i, chain in enumerate(self.chains):
             _input = await chain.arun(
-                _input, callbacks=_run_manager.get_child(f"step_{i+1}")
+                _input, callbacks=_run_manager.get_child(f"step_{i + 1}")
             )
             if self.strip_outputs:
                 _input = _input.strip()

{langchain-0.3.15 → langchain-0.3.16}/langchain/chat_models/base.py

@@ -416,6 +416,11 @@ def _init_chat_model_helper(
         from langchain_google_vertexai.model_garden import ChatAnthropicVertex
 
         return ChatAnthropicVertex(model=model, **kwargs)
+    elif model_provider == "deepseek":
+        _check_pkg("langchain_deepseek", pkg_kebab="langchain-deepseek-official")
+        from langchain_deepseek import ChatDeepSeek
+
+        return ChatDeepSeek(model=model, **kwargs)
     else:
         supported = ", ".join(_SUPPORTED_PROVIDERS)
         raise ValueError(
@@ -440,6 +445,7 @@ _SUPPORTED_PROVIDERS = {
     "bedrock",
     "bedrock_converse",
     "google_anthropic_vertex",
+    "deepseek",
 }
 
 
@@ -480,12 +486,11 @@ def _parse_model(model: str, model_provider: Optional[str]) -> Tuple[str, str]:
     return model, model_provider
 
 
-def _check_pkg(pkg: str) -> None:
+def _check_pkg(pkg: str, *, pkg_kebab: Optional[str] = None) -> None:
     if not util.find_spec(pkg):
-        pkg_kebab = pkg.replace("_", "-")
+        pkg_kebab = pkg_kebab if pkg_kebab is not None else pkg.replace("_", "-")
         raise ImportError(
-            f"Unable to import {pkg_kebab}. Please install with "
-            f"`pip install -U {pkg_kebab}`"
+            f"Unable to import {pkg}. Please install with `pip install -U {pkg_kebab}`"
         )
 
 
@@ -590,7 +595,11 @@ class _ConfigurableModel(Runnable[LanguageModelInput, Any]):
         queued_declarative_operations = list(self._queued_declarative_operations)
         if remaining_config:
             queued_declarative_operations.append(
-                ("with_config", (), {"config": remaining_config})
+                (
+                    "with_config",
+                    (),
+                    {"config": remaining_config},
+                )
             )
         return _ConfigurableModel(
             default_config={**self._default_config, **model_params},

{langchain-0.3.15 → langchain-0.3.16}/langchain/embeddings/base.py

@@ -13,6 +13,7 @@ _SUPPORTED_PROVIDERS = {
     "google_vertexai": "langchain_google_vertexai",
     "huggingface": "langchain_huggingface",
     "mistralai": "langchain_mistralai",
+    "ollama": "langchain_ollama",
     "openai": "langchain_openai",
 }
 
@@ -174,8 +175,7 @@ def init_embeddings(
     if not model:
         providers = _SUPPORTED_PROVIDERS.keys()
         raise ValueError(
-            "Must specify model name. "
-            f"Supported providers are: {', '.join(providers)}"
+            f"Must specify model name. Supported providers are: {', '.join(providers)}"
         )
 
     provider, model_name = _infer_model_and_provider(model, provider=provider)
@@ -210,6 +210,10 @@ def init_embeddings(
         from langchain_huggingface import HuggingFaceEmbeddings
 
         return HuggingFaceEmbeddings(model_name=model_name, **kwargs)
+    elif provider == "ollama":
+        from langchain_ollama import OllamaEmbeddings
+
+        return OllamaEmbeddings(model=model_name, **kwargs)
     else:
         raise ValueError(
             f"Provider '{provider}' is not supported.\n"

{langchain-0.3.15 → langchain-0.3.16}/langchain/evaluation/embedding_distance/base.py

@@ -310,7 +310,10 @@ class EmbeddingDistanceEvalChain(_EmbeddingDistanceChainMixin, StringEvaluator):
             Dict[str, Any]: The computed score.
         """
         embedded = await self.embeddings.aembed_documents(
-            [inputs["prediction"], inputs["reference"]]
+            [
+                inputs["prediction"],
+                inputs["reference"],
+            ]
         )
         vectors = np.array(embedded)
         score = self._compute_score(vectors)
@@ -427,7 +430,10 @@ class PairwiseEmbeddingDistanceEvalChain(
         """
         vectors = np.array(
             self.embeddings.embed_documents(
-                [inputs["prediction"], inputs["prediction_b"]]
+                [
+                    inputs["prediction"],
+                    inputs["prediction_b"],
+                ]
             )
         )
         score = self._compute_score(vectors)
@@ -449,7 +455,10 @@ class PairwiseEmbeddingDistanceEvalChain(
             Dict[str, Any]: The computed score.
         """
         embedded = await self.embeddings.aembed_documents(
-            [inputs["prediction"], inputs["prediction_b"]]
+            [
+                inputs["prediction"],
+                inputs["prediction_b"],
+            ]
         )
         vectors = np.array(embedded)
         score = self._compute_score(vectors)

{langchain-0.3.15 → langchain-0.3.16}/langchain/indexes/vectorstore.py

@@ -1,3 +1,5 @@
+"""Vectorstore stubs for the indexing api."""
+
 from typing import Any, Dict, List, Optional, Type
 
 from langchain_core.document_loaders import BaseLoader
@@ -13,6 +15,7 @@ from langchain.chains.retrieval_qa.base import RetrievalQA
 
 
 def _get_default_text_splitter() -> TextSplitter:
+    """Return the default text splitter used for chunking documents."""
     return RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
 
 
@@ -33,7 +36,17 @@ class VectorStoreIndexWrapper(BaseModel):
         retriever_kwargs: Optional[Dict[str, Any]] = None,
         **kwargs: Any,
     ) -> str:
-        """Query the vectorstore."""
+        """Query the vectorstore using the provided LLM.
+
+        Args:
+            question: The question or prompt to query.
+            llm: The language model to use. Must not be None.
+            retriever_kwargs: Optional keyword arguments for the retriever.
+            **kwargs: Additional keyword arguments forwarded to the chain.
+
+        Returns:
+            The result string from the RetrievalQA chain.
+        """
         if llm is None:
             raise NotImplementedError(
                 "This API has been changed to require an LLM. "
@@ -55,7 +68,17 @@ class VectorStoreIndexWrapper(BaseModel):
         retriever_kwargs: Optional[Dict[str, Any]] = None,
         **kwargs: Any,
     ) -> str:
-        """Query the vectorstore."""
+        """Asynchronously query the vectorstore using the provided LLM.
+
+        Args:
+            question: The question or prompt to query.
+            llm: The language model to use. Must not be None.
+            retriever_kwargs: Optional keyword arguments for the retriever.
+            **kwargs: Additional keyword arguments forwarded to the chain.
+
+        Returns:
+            The asynchronous result string from the RetrievalQA chain.
+        """
         if llm is None:
             raise NotImplementedError(
                 "This API has been changed to require an LLM. "
@@ -77,7 +100,17 @@ class VectorStoreIndexWrapper(BaseModel):
         retriever_kwargs: Optional[Dict[str, Any]] = None,
         **kwargs: Any,
     ) -> dict:
-        """Query the vectorstore and get back sources."""
+        """Query the vectorstore and retrieve the answer along with sources.
+
+        Args:
+            question: The question or prompt to query.
+            llm: The language model to use. Must not be None.
+            retriever_kwargs: Optional keyword arguments for the retriever.
+            **kwargs: Additional keyword arguments forwarded to the chain.
+
+        Returns:
+            A dictionary containing the answer and source documents.
+        """
         if llm is None:
             raise NotImplementedError(
                 "This API has been changed to require an LLM. "
@@ -99,7 +132,17 @@ class VectorStoreIndexWrapper(BaseModel):
         retriever_kwargs: Optional[Dict[str, Any]] = None,
         **kwargs: Any,
     ) -> dict:
-        """Query the vectorstore and get back sources."""
+        """Asynchronously query the vectorstore and retrieve the answer and sources.
+
+        Args:
+            question: The question or prompt to query.
+            llm: The language model to use. Must not be None.
+            retriever_kwargs: Optional keyword arguments for the retriever.
+            **kwargs: Additional keyword arguments forwarded to the chain.
+
+        Returns:
+            A dictionary containing the answer and source documents.
+        """
         if llm is None:
             raise NotImplementedError(
                 "This API has been changed to require an LLM. "
@@ -149,14 +192,28 @@ class VectorstoreIndexCreator(BaseModel):
     )
 
     def from_loaders(self, loaders: List[BaseLoader]) -> VectorStoreIndexWrapper:
-        """Create a vectorstore index from loaders."""
+        """Create a vectorstore index from a list of loaders.
+
+        Args:
+            loaders: A list of `BaseLoader` instances to load documents.
+
+        Returns:
+            A `VectorStoreIndexWrapper` containing the constructed vectorstore.
+        """
         docs = []
         for loader in loaders:
             docs.extend(loader.load())
         return self.from_documents(docs)
 
     async def afrom_loaders(self, loaders: List[BaseLoader]) -> VectorStoreIndexWrapper:
-        """Create a vectorstore index from loaders."""
+        """Asynchronously create a vectorstore index from a list of loaders.
+
+        Args:
+            loaders: A list of `BaseLoader` instances to load documents.
+
+        Returns:
+            A `VectorStoreIndexWrapper` containing the constructed vectorstore.
+        """
         docs = []
         for loader in loaders:
             async for doc in loader.alazy_load():
@@ -164,7 +221,14 @@ class VectorstoreIndexCreator(BaseModel):
         return await self.afrom_documents(docs)
 
     def from_documents(self, documents: List[Document]) -> VectorStoreIndexWrapper:
-        """Create a vectorstore index from documents."""
+        """Create a vectorstore index from a list of documents.
+
+        Args:
+            documents: A list of `Document` objects.
+
+        Returns:
+            A `VectorStoreIndexWrapper` containing the constructed vectorstore.
+        """
         sub_docs = self.text_splitter.split_documents(documents)
         vectorstore = self.vectorstore_cls.from_documents(
             sub_docs, self.embedding, **self.vectorstore_kwargs
@@ -174,7 +238,14 @@ class VectorstoreIndexCreator(BaseModel):
     async def afrom_documents(
         self, documents: List[Document]
     ) -> VectorStoreIndexWrapper:
-        """Create a vectorstore index from documents."""
+        """Asynchronously create a vectorstore index from a list of documents.
+
+        Args:
+            documents: A list of `Document` objects.
+
+        Returns:
+            A `VectorStoreIndexWrapper` containing the constructed vectorstore.
+        """
         sub_docs = self.text_splitter.split_documents(documents)
         vectorstore = await self.vectorstore_cls.afrom_documents(
             sub_docs, self.embedding, **self.vectorstore_kwargs

{langchain-0.3.15 → langchain-0.3.16}/langchain/memory/chat_memory.py

@@ -71,7 +71,10 @@ class BaseChatMemory(BaseMemory, ABC):
         """Save context from this conversation to buffer."""
         input_str, output_str = self._get_input_output(inputs, outputs)
         self.chat_memory.add_messages(
-            [HumanMessage(content=input_str), AIMessage(content=output_str)]
+            [
+                HumanMessage(content=input_str),
+                AIMessage(content=output_str),
+            ]
         )
 
     async def asave_context(
@@ -80,7 +83,10 @@ class BaseChatMemory(BaseMemory, ABC):
         """Save context from this conversation to buffer."""
         input_str, output_str = self._get_input_output(inputs, outputs)
         await self.chat_memory.aadd_messages(
-            [HumanMessage(content=input_str), AIMessage(content=output_str)]
+            [
+                HumanMessage(content=input_str),
+                AIMessage(content=output_str),
+            ]
         )
 
     def clear(self) -> None:

{langchain-0.3.15 → langchain-0.3.16}/langchain/retrievers/document_compressors/cohere_rerank.py

@@ -92,7 +92,10 @@ class CohereRerank(BaseDocumentCompressor):
         result_dicts = []
         for res in results:
             result_dicts.append(
-                {"index": res.index, "relevance_score": res.relevance_score}
+                {
+                    "index": res.index,
+                    "relevance_score": res.relevance_score,
+                }
             )
         return result_dicts
 

{langchain-0.3.15 → langchain-0.3.16}/langchain/retrievers/ensemble.py

@@ -223,7 +223,7 @@ class EnsembleRetriever(BaseRetriever):
             retriever.invoke(
                 query,
                 patch_config(
-                    config, callbacks=run_manager.get_child(tag=f"retriever_{i+1}")
+                    config, callbacks=run_manager.get_child(tag=f"retriever_{i + 1}")
                 ),
             )
             for i, retriever in enumerate(self.retrievers)
@@ -265,7 +265,8 @@ class EnsembleRetriever(BaseRetriever):
                 retriever.ainvoke(
                     query,
                     patch_config(
-                        config, callbacks=run_manager.get_child(tag=f"retriever_{i+1}")
+                        config,
+                        callbacks=run_manager.get_child(tag=f"retriever_{i + 1}"),
                     ),
                 )
                 for i, retriever in enumerate(self.retrievers)

{langchain-0.3.15 → langchain-0.3.16}/langchain/smith/evaluation/runner_utils.py

@@ -247,8 +247,7 @@ def _get_prompt(inputs: Dict[str, Any]) -> str:
     if "prompt" in inputs:
         if not isinstance(inputs["prompt"], str):
             raise InputFormatError(
-                "Expected string for 'prompt', got"
-                f" {type(inputs['prompt']).__name__}"
+                f"Expected string for 'prompt', got {type(inputs['prompt']).__name__}"
             )
         prompts = [inputs["prompt"]]
     elif "prompts" in inputs:

{langchain-0.3.15 → langchain-0.3.16}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "langchain"
-version = "0.3.15"
+version = "0.3.16"
 description = "Building applications with LLMs through composability"
 authors = []
 license = "MIT"
@@ -12,6 +12,7 @@ readme = "README.md"
 repository = "https://github.com/langchain-ai/langchain"
 
 [tool.ruff]
+target-version = "py39"
 exclude = [ "tests/integration_tests/examples/non-utf8-encoding.py",]
 
 [tool.mypy]
@@ -33,7 +34,7 @@ langchain-server = "langchain.server:main"
 
 [tool.poetry.dependencies]
 python = ">=3.9,<4.0"
-langchain-core = "^0.3.31"
+langchain-core = "^0.3.32"
 langchain-text-splitters = "^0.3.3"
 langsmith = ">=0.1.17,<0.4"
 pydantic = "^2.7.4"
@@ -51,7 +52,12 @@ version = ">=1.26.2,<3"
 python = ">=3.12"
 
 [tool.ruff.lint]
-select = [ "E", "F", "I", "T201",]
+select = [ "E", "F", "I", "T201", "D",]
+pydocstyle = { convention = "google" }
+
+[tool.ruff.lint.per-file-ignores]
+"tests/*" = ["D"]
+"!langchain/indexes/vectorstore.py" = ["D"]
 
 [tool.coverage.run]
 omit = [ "tests/*",]
@@ -119,7 +125,7 @@ cassio = "^0.1.0"
 langchainhub = "^0.1.16"
 
 [tool.poetry.group.lint.dependencies]
-ruff = "^0.5"
+ruff = "^0.9.2"
 [[tool.poetry.group.lint.dependencies.cffi]]
 version = "<1.17.1"
 python = "<3.10"