langchain-core 1.0.0rc1__py3-none-any.whl → 1.0.0rc3__py3-none-any.whl

langchain_core/runnables/graph_mermaid.py

@@ -60,10 +60,10 @@ def draw_mermaid(
         edges: List of edges, object with a source, target and data.
         first_node: Id of the first node.
         last_node: Id of the last node.
-        with_styles: Whether to include styles in the graph. Defaults to `True`.
-        curve_style: Curve style for the edges. Defaults to CurveStyle.LINEAR.
-        node_styles: Node colors for different types. Defaults to NodeStyles().
-        wrap_label_n_words: Words to wrap the edge labels. Defaults to 9.
+        with_styles: Whether to include styles in the graph.
+        curve_style: Curve style for the edges.
+        node_styles: Node colors for different types.
+        wrap_label_n_words: Words to wrap the edge labels.
         frontmatter_config: Mermaid frontmatter config.
             Can be used to customize theme and styles. Will be converted to YAML and
             added to the beginning of the mermaid graph.
@@ -287,11 +287,11 @@ def draw_mermaid_png(
     Args:
         mermaid_syntax: Mermaid graph syntax.
         output_file_path: Path to save the PNG image.
-        draw_method: Method to draw the graph. Defaults to MermaidDrawMethod.API.
-        background_color: Background color of the image. Defaults to "white".
-        padding: Padding around the image. Defaults to 10.
-        max_retries: Maximum number of retries (MermaidDrawMethod.API). Defaults to 1.
-        retry_delay: Delay between retries (MermaidDrawMethod.API). Defaults to 1.0.
+        draw_method: Method to draw the graph.
+        background_color: Background color of the image.
+        padding: Padding around the image.
+        max_retries: Maximum number of retries (MermaidDrawMethod.API).
+        retry_delay: Delay between retries (MermaidDrawMethod.API).
         base_url: Base URL for the Mermaid.ink API.
 
     Returns:

langchain_core/runnables/graph_png.py

@@ -105,7 +105,7 @@ class PngDrawer:
             source: The source node.
             target: The target node.
             label: The label for the edge.
-            conditional: Whether the edge is conditional. Defaults to `False`.
+            conditional: Whether the edge is conditional.
         """
         viz.add_edge(
             source,

langchain_core/runnables/history.py

@@ -296,9 +296,9 @@ class RunnableWithMessageHistory(RunnableBindingBase):  # type: ignore[no-redef]
                 ```
 
             input_messages_key: Must be specified if the base runnable accepts a dict
-                as input. Default is None.
+                as input.
             output_messages_key: Must be specified if the base runnable returns a dict
-                as output. Default is None.
+                as output.
             history_messages_key: Must be specified if the base runnable accepts a dict
                 as input and expects a separate key for historical messages.
             history_factory_config: Configure fields that should be passed to the

langchain_core/runnables/passthrough.py

@@ -185,7 +185,7 @@ class RunnablePassthrough(RunnableSerializable[Other, Other]):
 
     @classmethod
     def get_lc_namespace(cls) -> list[str]:
-        """Get the namespace of the langchain object.
+        """Get the namespace of the LangChain object.
 
         Returns:
             `["langchain", "schema", "runnable"]`
@@ -409,7 +409,7 @@ class RunnableAssign(RunnableSerializable[dict[str, Any], dict[str, Any]]):
     @classmethod
     @override
     def get_lc_namespace(cls) -> list[str]:
-        """Get the namespace of the langchain object.
+        """Get the namespace of the LangChain object.
 
         Returns:
             `["langchain", "schema", "runnable"]`
@@ -714,7 +714,7 @@ class RunnablePick(RunnableSerializable[dict[str, Any], dict[str, Any]]):
     @classmethod
     @override
     def get_lc_namespace(cls) -> list[str]:
-        """Get the namespace of the langchain object.
+        """Get the namespace of the LangChain object.
 
         Returns:
             `["langchain", "schema", "runnable"]`

langchain_core/runnables/router.py

@@ -96,7 +96,7 @@ class RouterRunnable(RunnableSerializable[RouterInput, Output]):
     @classmethod
     @override
     def get_lc_namespace(cls) -> list[str]:
-        """Get the namespace of the langchain object.
+        """Get the namespace of the LangChain object.
 
         Returns:
             `["langchain", "schema", "runnable"]`

langchain_core/runnables/utils.py

@@ -136,7 +136,7 @@ def coro_with_context(
     Args:
         coro: The coroutine to await.
         context: The context to use.
-        create_task: Whether to create a task. Defaults to `False`.
+        create_task: Whether to create a task.
 
     Returns:
         The coroutine with the context.
@@ -558,7 +558,7 @@ class ConfigurableField(NamedTuple):
     annotation: Any | None = None
     """The annotation of the field. """
     is_shared: bool = False
-    """Whether the field is shared. Defaults to `False`."""
+    """Whether the field is shared."""
 
     @override
     def __hash__(self) -> int:
@@ -579,7 +579,7 @@ class ConfigurableFieldSingleOption(NamedTuple):
     description: str | None = None
     """The description of the field. """
     is_shared: bool = False
-    """Whether the field is shared. Defaults to `False`."""
+    """Whether the field is shared."""
 
     @override
     def __hash__(self) -> int:
@@ -600,7 +600,7 @@ class ConfigurableFieldMultiOption(NamedTuple):
     description: str | None = None
     """The description of the field. """
     is_shared: bool = False
-    """Whether the field is shared. Defaults to `False`."""
+    """Whether the field is shared."""
 
     @override
     def __hash__(self) -> int:
@@ -626,7 +626,7 @@ class ConfigurableFieldSpec(NamedTuple):
     default: Any = None
     """The default value for the field. """
     is_shared: bool = False
-    """Whether the field is shared. Defaults to `False`."""
+    """Whether the field is shared."""
     dependencies: list[str] | None = None
     """The dependencies of the field. """
 

langchain_core/tools/base.py

@@ -293,10 +293,9 @@ def create_schema_from_function(
         filter_args: Optional list of arguments to exclude from the schema.
             Defaults to `FILTERED_ARGS`.
         parse_docstring: Whether to parse the function's docstring for descriptions
-            for each argument. Defaults to `False`.
+            for each argument.
         error_on_invalid_docstring: if `parse_docstring` is provided, configure
             whether to raise `ValueError` on invalid Google Style docstrings.
-            Defaults to `False`.
         include_injected: Whether to include injected arguments in the schema.
             Defaults to `True`, since we want to include them in the schema
             when *validating* tool inputs.
@@ -481,11 +480,11 @@ class ChildTool(BaseTool):
     """Handle the content of the ValidationError thrown."""
 
     response_format: Literal["content", "content_and_artifact"] = "content"
-    """The tool response format. Defaults to 'content'.
+    """The tool response format.
 
-    If "content" then the output of the tool is interpreted as the contents of a
-    ToolMessage. If "content_and_artifact" then the output is expected to be a
-    two-tuple corresponding to the (content, artifact) of a ToolMessage.
+    If `"content"` then the output of the tool is interpreted as the contents of a
+    ToolMessage. If `"content_and_artifact"` then the output is expected to be a
+    two-tuple corresponding to the (content, artifact) of a `ToolMessage`.
     """
 
     def __init__(self, **kwargs: Any) -> None:
@@ -768,8 +767,8 @@ class ChildTool(BaseTool):
         Args:
             tool_input: The input to the tool.
             verbose: Whether to log the tool's progress.
-            start_color: The color to use when starting the tool. Defaults to 'green'.
-            color: The color to use when ending the tool. Defaults to 'green'.
+            start_color: The color to use when starting the tool.
+            color: The color to use when ending the tool.
             callbacks: Callbacks to be called during tool execution.
             tags: Optional list of tags associated with the tool.
             metadata: Optional metadata associated with the tool.
@@ -880,8 +879,8 @@ class ChildTool(BaseTool):
         Args:
             tool_input: The input to the tool.
             verbose: Whether to log the tool's progress.
-            start_color: The color to use when starting the tool. Defaults to 'green'.
-            color: The color to use when ending the tool. Defaults to 'green'.
+            start_color: The color to use when starting the tool.
+            color: The color to use when ending the tool.
             callbacks: Callbacks to be called during tool execution.
             tags: Optional list of tags associated with the tool.
             metadata: Optional metadata associated with the tool.
@@ -1211,6 +1210,26 @@ class InjectedToolArg:
     """
 
 
+class _DirectlyInjectedToolArg:
+    """Annotation for tool arguments that are injected at runtime.
+
+    Injected via direct type annotation, rather than annotated metadata.
+
+    For example, ToolRuntime is a directly injected argument.
+    Note the direct annotation rather than the verbose alternative:
+    Annotated[ToolRuntime, InjectedRuntime]
+    ```python
+    from langchain_core.tools import tool, ToolRuntime
+
+
+    @tool
+    def foo(x: int, runtime: ToolRuntime) -> str:
+        # use runtime.state, runtime.context, runtime.store, etc.
+        ...
+    ```
+    """
+
+
 class InjectedToolCallId(InjectedToolArg):
     """Annotation for injecting the tool call ID.
 
@@ -1238,6 +1257,24 @@ class InjectedToolCallId(InjectedToolArg):
     """
 
 
+def _is_directly_injected_arg_type(type_: Any) -> bool:
+    """Check if a type annotation indicates a directly injected argument.
+
+    This is currently only used for ToolRuntime.
+    Checks if either the annotation itself is a subclass of _DirectlyInjectedToolArg
+    or the origin of the annotation is a subclass of _DirectlyInjectedToolArg.
+
+    Ex: ToolRuntime or ToolRuntime[ContextT, StateT] would both return True.
+    """
+    return (
+        isinstance(type_, type) and issubclass(type_, _DirectlyInjectedToolArg)
+    ) or (
+        (origin := get_origin(type_)) is not None
+        and isinstance(origin, type)
+        and issubclass(origin, _DirectlyInjectedToolArg)
+    )
+
+
 def _is_injected_arg_type(
     type_: type | TypeVar, injected_type: type[InjectedToolArg] | None = None
 ) -> bool:
@@ -1250,7 +1287,15 @@ def _is_injected_arg_type(
     Returns:
         `True` if the type is an injected argument, `False` otherwise.
     """
-    injected_type = injected_type or InjectedToolArg
+    if injected_type is None:
+        # if no injected type is specified,
+        # check if the type is a directly injected argument
+        if _is_directly_injected_arg_type(type_):
+            return True
+        injected_type = InjectedToolArg
+
+    # if the type is an Annotated type, check if annotated metadata
+    # is an intance or subclass of the injected type
     return any(
         isinstance(arg, injected_type)
         or (isinstance(arg, type) and issubclass(arg, injected_type))

langchain_core/tools/convert.py

@@ -81,7 +81,7 @@ def tool(
     parse_docstring: bool = False,
     error_on_invalid_docstring: bool = True,
 ) -> BaseTool | Callable[[Callable | Runnable], BaseTool]:
-    """Make tools out of functions, can be used with or without arguments.
+    """Make tools out of Python functions, can be used with or without arguments.
 
     Args:
         name_or_callable: Optional name of the tool or the callable to be
@@ -93,30 +93,26 @@ def tool(
 
             - `description` argument
                 (used even if docstring and/or `args_schema` are provided)
-            - tool function docstring
+            - Tool function docstring
                 (used even if `args_schema` is provided)
             - `args_schema` description
                 (used only if `description` / docstring are not provided)
         *args: Extra positional arguments. Must be empty.
         return_direct: Whether to return directly from the tool rather
-            than continuing the agent loop. Defaults to `False`.
-        args_schema: optional argument schema for user to specify.
+            than continuing the agent loop.
+        args_schema: Optional argument schema for user to specify.
 
         infer_schema: Whether to infer the schema of the arguments from
             the function's signature. This also makes the resultant tool
             accept a dictionary input to its `run()` function.
-            Defaults to `True`.
-        response_format: The tool response format. If "content" then the output of
-            the tool is interpreted as the contents of a ToolMessage. If
-            "content_and_artifact" then the output is expected to be a two-tuple
-            corresponding to the (content, artifact) of a ToolMessage.
-            Defaults to "content".
+        response_format: The tool response format. If `"content"` then the output of
+            the tool is interpreted as the contents of a `ToolMessage`. If
+            `"content_and_artifact"` then the output is expected to be a two-tuple
+            corresponding to the `(content, artifact)` of a `ToolMessage`.
         parse_docstring: if `infer_schema` and `parse_docstring`, will attempt to
             parse parameter descriptions from Google Style function docstrings.
-            Defaults to `False`.
         error_on_invalid_docstring: if `parse_docstring` is provided, configure
-            whether to raise ValueError on invalid Google Style docstrings.
-            Defaults to `True`.
+            whether to raise `ValueError` on invalid Google Style docstrings.
 
     Raises:
         ValueError: If too many positional arguments are provided.
@@ -124,8 +120,8 @@ def tool(
         ValueError: If the first argument is not a string or callable with
             a `__name__` attribute.
         ValueError: If the function does not have a docstring and description
-            is not provided and `infer_schema` is False.
-        ValueError: If `parse_docstring` is True and the function has an invalid
+            is not provided and `infer_schema` is `False`.
+        ValueError: If `parse_docstring` is `True` and the function has an invalid
             Google-style docstring and `error_on_invalid_docstring` is True.
         ValueError: If a Runnable is provided that does not have an object schema.
 
@@ -133,7 +129,7 @@ def tool(
         The tool.
 
     Requires:
-        - Function must be of type (str) -> str
+        - Function must be of type `(str) -> str`
         - Function must have a docstring
 
     Examples:
@@ -197,7 +193,7 @@ def tool(
         Note that parsing by default will raise `ValueError` if the docstring
         is considered invalid. A docstring is considered invalid if it contains
         arguments not in the function signature, or is unable to be parsed into
-        a summary and "Args:" blocks. Examples below:
+        a summary and `"Args:"` blocks. Examples below:
 
         ```python
         # No args section

langchain_core/tools/retriever.py

@@ -82,12 +82,12 @@ def create_retriever_tool(
         description: The description for the tool. This will be passed to the language
             model, so should be descriptive.
         document_prompt: The prompt to use for the document.
-        document_separator: The separator to use between documents. Defaults to "\n\n".
-        response_format: The tool response format. If "content" then the output of
-            the tool is interpreted as the contents of a ToolMessage. If
-            "content_and_artifact" then the output is expected to be a two-tuple
-            corresponding to the (content, artifact) of a ToolMessage (artifact
-            being a list of documents in this case). Defaults to "content".
+        document_separator: The separator to use between documents.
+        response_format: The tool response format. If `"content"` then the output of
+            the tool is interpreted as the contents of a `ToolMessage`. If
+            `"content_and_artifact"` then the output is expected to be a two-tuple
+            corresponding to the `(content, artifact)` of a `ToolMessage` (artifact
+            being a list of documents in this case).
 
     Returns:
         Tool class to pass to an agent.

langchain_core/tools/simple.py

@@ -176,7 +176,7 @@ class Tool(BaseTool):
             func: The function to create the tool from.
             name: The name of the tool.
             description: The description of the tool.
-            return_direct: Whether to return the output directly. Defaults to `False`.
+            return_direct: Whether to return the output directly.
             args_schema: The schema of the tool's input arguments.
             coroutine: The asynchronous version of the function.
             **kwargs: Additional arguments to pass to the tool.

langchain_core/tools/structured.py

@@ -149,21 +149,16 @@ class StructuredTool(BaseTool):
             description: The description of the tool.
                 Defaults to the function docstring.
             return_direct: Whether to return the result directly or as a callback.
-                Defaults to `False`.
             args_schema: The schema of the tool's input arguments.
             infer_schema: Whether to infer the schema from the function's signature.
-                Defaults to `True`.
-            response_format: The tool response format. If "content" then the output of
-                the tool is interpreted as the contents of a ToolMessage. If
-                "content_and_artifact" then the output is expected to be a two-tuple
-                corresponding to the (content, artifact) of a ToolMessage.
-                Defaults to "content".
+            response_format: The tool response format. If `"content"` then the output of
+                the tool is interpreted as the contents of a `ToolMessage`. If
+                `"content_and_artifact"` then the output is expected to be a two-tuple
+                corresponding to the `(content, artifact)` of a `ToolMessage`.
             parse_docstring: if `infer_schema` and `parse_docstring`, will attempt
                 to parse parameter descriptions from Google Style function docstrings.
-                Defaults to `False`.
             error_on_invalid_docstring: if `parse_docstring` is provided, configure
-                whether to raise ValueError on invalid Google Style docstrings.
-                Defaults to `False`.
+                whether to raise `ValueError` on invalid Google Style docstrings.
             **kwargs: Additional arguments to pass to the tool
 
         Returns:

langchain_core/tracers/memory_stream.py

@@ -5,7 +5,7 @@ channel. The writer and reader can be in the same event loop or in different eve
 loops. When they're in different event loops, they will also be in different
 threads.
 
-This is useful in situations when there's a mix of synchronous and asynchronous
+Useful in situations when there's a mix of synchronous and asynchronous
 used in the code.
 """
 

langchain_core/tracers/root_listeners.py

@@ -24,7 +24,7 @@ class RootListenersTracer(BaseTracer):
     """Tracer that calls listeners on run start, end, and error."""
 
     log_missing_parent = False
-    """Whether to log a warning if the parent is missing. Default is False."""
+    """Whether to log a warning if the parent is missing."""
 
     def __init__(
         self,
@@ -79,7 +79,7 @@ class AsyncRootListenersTracer(AsyncBaseTracer):
     """Async Tracer that calls listeners on run start, end, and error."""
 
     log_missing_parent = False
-    """Whether to log a warning if the parent is missing. Default is False."""
+    """Whether to log a warning if the parent is missing."""
 
     def __init__(
         self,

langchain_core/tracers/stdout.py

@@ -49,8 +49,7 @@ class FunctionCallbackHandler(BaseTracer):
     """Tracer that calls a function with a single str parameter."""
 
     name: str = "function_callback_handler"
-    """The name of the tracer. This is used to identify the tracer in the logs.
-    Default is "function_callback_handler"."""
+    """The name of the tracer. This is used to identify the tracer in the logs."""
 
     def __init__(self, function: Callable[[str], None], **kwargs: Any) -> None:
         """Create a FunctionCallbackHandler.

langchain_core/utils/__init__.py

@@ -1,4 +1,4 @@
-"""**Utility functions** for LangChain.
+"""Utility functions for LangChain.
 
 These functions do not depend on any other LangChain module.
 """

langchain_core/utils/aiter.py

@@ -201,7 +201,7 @@ class Tee(Generic[T]):
 
         Args:
             iterable: The iterable to split.
-            n: The number of iterators to create. Defaults to 2.
+            n: The number of iterators to create.
             lock: The lock to synchronise access to the shared buffers.
 
         """

langchain_core/utils/function_calling.py

@@ -114,7 +114,7 @@ def _convert_json_schema_to_openai_function(
             used.
         description: The description of the function. If not provided, the description
             of the schema will be used.
-        rm_titles: Whether to remove titles from the schema. Defaults to `True`.
+        rm_titles: Whether to remove titles from the schema.
 
     Returns:
         The function description.
@@ -148,7 +148,7 @@ def _convert_pydantic_to_openai_function(
             used.
         description: The description of the function. If not provided, the description
             of the schema will be used.
-        rm_titles: Whether to remove titles from the schema. Defaults to `True`.
+        rm_titles: Whether to remove titles from the schema.
 
     Raises:
         TypeError: If the model is not a Pydantic model.
@@ -334,11 +334,11 @@ def convert_to_openai_function(
 
     Args:
         function:
-            A dictionary, Pydantic BaseModel class, TypedDict class, a LangChain
-            Tool object, or a Python function. If a dictionary is passed in, it is
+            A dictionary, Pydantic `BaseModel` class, `TypedDict` class, a LangChain
+            `Tool` object, or a Python function. If a dictionary is passed in, it is
             assumed to already be a valid OpenAI function, a JSON schema with
-            top-level 'title' key specified, an Anthropic format
-            tool, or an Amazon Bedrock Converse format tool.
+            top-level `title` key specified, an Anthropic format tool, or an Amazon
+            Bedrock Converse format tool.
         strict:
             If `True`, model output is guaranteed to exactly match the JSON Schema
             provided in the function definition. If `None`, `strict` argument will not
@@ -351,17 +351,8 @@ def convert_to_openai_function(
     Raises:
         ValueError: If function is not in a supported format.
 
-    !!! warning "Behavior changed in 0.2.29"
-        `strict` arg added.
-
-    !!! warning "Behavior changed in 0.3.13"
-        Support for Anthropic format tools added.
-
-    !!! warning "Behavior changed in 0.3.14"
-        Support for Amazon Bedrock Converse format tools added.
-
     !!! warning "Behavior changed in 0.3.16"
-        'description' and 'parameters' keys are now optional. Only 'name' is
+        `description` and `parameters` keys are now optional. Only `name` is
         required and guaranteed to be part of the output.
     """
     # an Anthropic format tool
@@ -459,16 +450,14 @@ def convert_to_openai_tool(
 ) -> dict[str, Any]:
     """Convert a tool-like object to an OpenAI tool schema.
 
-    OpenAI tool schema reference:
-    https://platform.openai.com/docs/api-reference/chat/create#chat-create-tools
+    [OpenAI tool schema reference](https://platform.openai.com/docs/api-reference/chat/create#chat-create-tools)
 
     Args:
         tool:
-            Either a dictionary, a pydantic.BaseModel class, Python function, or
-            BaseTool. If a dictionary is passed in, it is
-            assumed to already be a valid OpenAI function, a JSON schema with
-            top-level 'title' key specified, an Anthropic format
-            tool, or an Amazon Bedrock Converse format tool.
+            Either a dictionary, a `pydantic.BaseModel` class, Python function, or
+            `BaseTool`. If a dictionary is passed in, it is assumed to already be a
+            valid OpenAI function, a JSON schema with top-level `title` key specified,
+            an Anthropic format tool, or an Amazon Bedrock Converse format tool.
         strict:
             If `True`, model output is guaranteed to exactly match the JSON Schema
             provided in the function definition. If `None`, `strict` argument will not
@@ -478,26 +467,14 @@ def convert_to_openai_tool(
         A dict version of the passed in tool which is compatible with the
         OpenAI tool-calling API.
 
-    !!! warning "Behavior changed in 0.2.29"
-        `strict` arg added.
-
-    !!! warning "Behavior changed in 0.3.13"
-        Support for Anthropic format tools added.
-
-    !!! warning "Behavior changed in 0.3.14"
-        Support for Amazon Bedrock Converse format tools added.
-
     !!! warning "Behavior changed in 0.3.16"
-        'description' and 'parameters' keys are now optional. Only 'name' is
+        `description` and `parameters` keys are now optional. Only `name` is
         required and guaranteed to be part of the output.
 
     !!! warning "Behavior changed in 0.3.44"
         Return OpenAI Responses API-style tools unchanged. This includes
-        any dict with "type" in "file_search", "function", "computer_use_preview",
-        "web_search_preview".
-
-    !!! warning "Behavior changed in 0.3.61"
-        Added support for OpenAI's built-in code interpreter and remote MCP tools.
+        any dict with `"type"` in `"file_search"`, `"function"`,
+        `"computer_use_preview"`, `"web_search_preview"`.
 
     !!! warning "Behavior changed in 0.3.63"
         Added support for OpenAI's image generation built-in tool.

langchain_core/utils/input.py

@@ -66,7 +66,7 @@ def print_text(
     Args:
         text: The text to print.
         color: The color to use.
-        end: The end character to use. Defaults to "".
+        end: The end character to use.
         file: The file to write to.
     """
     text_to_print = get_colored_text(text, color) if color else text

langchain_core/utils/iter.py

@@ -137,7 +137,7 @@ class Tee(Generic[T]):
 
         Args:
             iterable: The iterable to split.
-            n: The number of iterators to create. Defaults to 2.
+            n: The number of iterators to create.
             lock: The lock to synchronise access to the shared buffers.
 
         """

langchain_core/utils/json.py

@@ -51,7 +51,7 @@ def parse_partial_json(s: str, *, strict: bool = False) -> Any:
 
     Args:
         s: The JSON string to parse.
-        strict: Whether to use strict parsing. Defaults to `False`.
+        strict: Whether to use strict parsing.
 
     Returns:
         The parsed JSON object as a Python dictionary.

langchain_core/utils/strings.py

@@ -57,7 +57,7 @@ def sanitize_for_postgres(text: str, replacement: str = "") -> str:
 
     Args:
         text: The text to sanitize.
-        replacement: String to replace NUL bytes with. Defaults to empty string.
+        replacement: String to replace NUL bytes with.
 
     Returns:
         The sanitized text with NUL bytes removed or replaced.