15022600357

Committed 14 May 2025 01:54PM UTC coverage: 90.465% (+0.05%) from 90.417%

Build # 15022600357

Build Type

Pull #9342

github

Committed by

web-flow

Commit Message

Merge 2fd50f281 into 42b378950

Pull Request Pull Request #9342: Fix component tool parameters

Run Details

10930 of 12082 relevant lines covered (90.47%)

0.9 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

93.41

haystack/tools/component_tool.py

# SPDX-FileCopyrightText: 2022-present deepset GmbH <info@deepset.ai>
#
# SPDX-License-Identifier: Apache-2.0

from typing import Any, Callable, Dict, Optional, Union, get_args, get_origin

from pydantic import Field, TypeAdapter, create_model

from haystack import logging
from haystack.core.component import Component
from haystack.core.serialization import (
    component_from_dict,
    component_to_dict,
    generate_qualified_class_name,
    import_class_by_name,
)
from haystack.tools import Tool
from haystack.tools.errors import SchemaGenerationError
from haystack.tools.from_function import _remove_title_from_schema
from haystack.tools.parameters_schema_utils import _get_param_descriptions, _resolve_type
from haystack.utils.callable_serialization import deserialize_callable, serialize_callable

logger = logging.getLogger(__name__)


class ComponentTool(Tool):
    """
    A Tool that wraps Haystack components, allowing them to be used as tools by LLMs.

    ComponentTool automatically generates LLM-compatible tool schemas from component input sockets,
    which are derived from the component's `run` method signature and type hints.


    Key features:
    - Automatic LLM tool calling schema generation from component input sockets
    - Type conversion and validation for component inputs
    - Support for types:
        - Dataclasses
        - Lists of dataclasses
        - Basic types (str, int, float, bool, dict)
        - Lists of basic types
    - Automatic name generation from component class name
    - Description extraction from component docstrings

    To use ComponentTool, you first need a Haystack component - either an existing one or a new one you create.
    You can create a ComponentTool from the component by passing the component to the ComponentTool constructor.
    Below is an example of creating a ComponentTool from an existing SerperDevWebSearch component.

    ```python
    from haystack import component, Pipeline
    from haystack.tools import ComponentTool
    from haystack.components.websearch import SerperDevWebSearch
    from haystack.utils import Secret
    from haystack.components.tools.tool_invoker import ToolInvoker
    from haystack.components.generators.chat import OpenAIChatGenerator
    from haystack.dataclasses import ChatMessage

    # Create a SerperDev search component
    search = SerperDevWebSearch(api_key=Secret.from_env_var("SERPERDEV_API_KEY"), top_k=3)

    # Create a tool from the component
    tool = ComponentTool(
        component=search,
        name="web_search",  # Optional: defaults to "serper_dev_web_search"
        description="Search the web for current information on any topic"  # Optional: defaults to component docstring
    )

    # Create pipeline with OpenAIChatGenerator and ToolInvoker
    pipeline = Pipeline()
    pipeline.add_component("llm", OpenAIChatGenerator(model="gpt-4o-mini", tools=[tool]))
    pipeline.add_component("tool_invoker", ToolInvoker(tools=[tool]))

    # Connect components
    pipeline.connect("llm.replies", "tool_invoker.messages")

    message = ChatMessage.from_user("Use the web search tool to find information about Nikola Tesla")

    # Run pipeline
    result = pipeline.run({"llm": {"messages": [message]}})

    print(result)
    ```

    """

    def __init__(
        self,
        component: Component,
        name: Optional[str] = None,
        description: Optional[str] = None,
        parameters: Optional[Dict[str, Any]] = None,
        *,
        outputs_to_string: Optional[Dict[str, Union[str, Callable[[Any], str]]]] = None,
        inputs_from_state: Optional[Dict[str, str]] = None,
        outputs_to_state: Optional[Dict[str, Dict[str, Union[str, Callable]]]] = None,
    ):
        """
        Create a Tool instance from a Haystack component.

        :param component: The Haystack component to wrap as a tool.
        :param name: Optional name for the tool (defaults to snake_case of component class name).
        :param description: Optional description (defaults to component's docstring).
        :param parameters:
            A JSON schema defining the parameters expected by the Tool.
            Will fall back to the parameters defined in the component's run method signature if not provided.
        :param outputs_to_string:
            Optional dictionary defining how a tool outputs should be converted into a string.
            If the source is provided only the specified output key is sent to the handler.
            If the source is omitted the whole tool result is sent to the handler.
            Example: {
                "source": "docs", "handler": format_documents
            }
        :param inputs_from_state:
            Optional dictionary mapping state keys to tool parameter names.
            Example: {"repository": "repo"} maps state's "repository" to tool's "repo" parameter.
        :param outputs_to_state:
            Optional dictionary defining how tool outputs map to keys within state as well as optional handlers.
            If the source is provided only the specified output key is sent to the handler.
            Example: {
                "documents": {"source": "docs", "handler": custom_handler}
            }
            If the source is omitted the whole tool result is sent to the handler.
            Example: {
                "documents": {"handler": custom_handler}
            }
        :raises ValueError: If the component is invalid or schema generation fails.
        """
        if not isinstance(component, Component):
            message = (
                f"Object {component!r} is not a Haystack component. "
                "Use ComponentTool only with Haystack component instances."
            )
            raise ValueError(message)

        if getattr(component, "__haystack_added_to_pipeline__", None):
            msg = (
                "Component has been added to a pipeline and can't be used to create a ComponentTool. "
                "Create ComponentTool from a non-pipeline component instead."
            )
            raise ValueError(msg)

        self._unresolved_parameters = parameters
        # Create the tools schema from the component run method parameters
        tool_schema = parameters or self._create_tool_parameters_schema(component, inputs_from_state or {})

        def component_invoker(**kwargs):
            """
            Invokes the component using keyword arguments provided by the LLM function calling/tool-generated response.

            :param kwargs: The keyword arguments to invoke the component with.
            :returns: The result of the component invocation.
            """
            converted_kwargs = {}
            input_sockets = component.__haystack_input__._sockets_dict
            for param_name, param_value in kwargs.items():
                param_type = input_sockets[param_name].type

                # Check if the type (or list element type) has from_dict
                target_type = get_args(param_type)[0] if get_origin(param_type) is list else param_type
                if hasattr(target_type, "from_dict"):
                    if isinstance(param_value, list):
                        param_value = [target_type.from_dict(item) for item in param_value if isinstance(item, dict)]
                    elif isinstance(param_value, dict):
                        param_value = target_type.from_dict(param_value)
                else:
                    # Let TypeAdapter handle both single values and lists
                    type_adapter = TypeAdapter(param_type)
                    param_value = type_adapter.validate_python(param_value)

                converted_kwargs[param_name] = param_value
            logger.debug(f"Invoking component {type(component)} with kwargs: {converted_kwargs}")
            return component.run(**converted_kwargs)

        # Generate a name for the tool if not provided
        if not name:
            class_name = component.__class__.__name__
            # Convert camelCase/PascalCase to snake_case
            name = "".join(
                [
                    "_" + c.lower() if c.isupper() and i > 0 and not class_name[i - 1].isupper() else c.lower()
                    for i, c in enumerate(class_name)
                ]
            ).lstrip("_")

        description = description or component.__doc__ or name

        # Create the Tool instance with the component invoker as the function to be called and the schema
        super().__init__(
            name=name,
            description=description,
            parameters=tool_schema,
            function=component_invoker,
            inputs_from_state=inputs_from_state,
            outputs_to_state=outputs_to_state,
            outputs_to_string=outputs_to_string,
        )
        self._component = component

    def to_dict(self) -> Dict[str, Any]:
        """
        Serializes the ComponentTool to a dictionary.
        """
        serialized_component = component_to_dict(obj=self._component, name=self.name)

        serialized = {
            "component": serialized_component,
            "name": self.name,
            "description": self.description,
            "parameters": self._unresolved_parameters,
            "outputs_to_string": self.outputs_to_string,
            "inputs_from_state": self.inputs_from_state,
            "outputs_to_state": self.outputs_to_state,
        }

        if self.outputs_to_state is not None:
            serialized_outputs = {}
            for key, config in self.outputs_to_state.items():
                serialized_config = config.copy()
                if "handler" in config:
                    serialized_config["handler"] = serialize_callable(config["handler"])
                serialized_outputs[key] = serialized_config
            serialized["outputs_to_state"] = serialized_outputs

        if self.outputs_to_string is not None and self.outputs_to_string.get("handler") is not None:
            serialized["outputs_to_string"] = serialize_callable(self.outputs_to_string["handler"])

        return {"type": generate_qualified_class_name(type(self)), "data": serialized}

    @classmethod
    def from_dict(cls, data: Dict[str, Any]) -> "Tool":
        """
        Deserializes the ComponentTool from a dictionary.
        """
        inner_data = data["data"]
        component_class = import_class_by_name(inner_data["component"]["type"])
        component = component_from_dict(cls=component_class, data=inner_data["component"], name=inner_data["name"])

        if "outputs_to_state" in inner_data and inner_data["outputs_to_state"]:
            deserialized_outputs = {}
            for key, config in inner_data["outputs_to_state"].items():
                deserialized_config = config.copy()
                if "handler" in config:
                    deserialized_config["handler"] = deserialize_callable(config["handler"])
                deserialized_outputs[key] = deserialized_config
            inner_data["outputs_to_state"] = deserialized_outputs

        if (
            inner_data.get("outputs_to_string") is not None
            and inner_data["outputs_to_string"].get("handler") is not None
        ):
            inner_data["outputs_to_string"]["handler"] = deserialize_callable(
                inner_data["outputs_to_string"]["handler"]
            )

        return cls(
            component=component,
            name=inner_data["name"],
            description=inner_data["description"],
            parameters=inner_data.get("parameters", None),
            outputs_to_string=inner_data.get("outputs_to_string", None),
            inputs_from_state=inner_data.get("inputs_from_state", None),
            outputs_to_state=inner_data.get("outputs_to_state", None),
        )

    def _create_tool_parameters_schema(self, component: Component, inputs_from_state: Dict[str, Any]) -> Dict[str, Any]:
        """
        Creates an OpenAI tools schema from a component's run method parameters.

        :param component: The component to create the schema from.
        :raises SchemaGenerationError: If schema generation fails
        :returns: OpenAI tools schema for the component's run method parameters.
        """
        component_run_description, param_descriptions = _get_param_descriptions(component.run)

        # collect fields (types and defaults) and descriptions from function parameters
        fields: Dict[str, Any] = {}

        for input_name, socket in component.__haystack_input__._sockets_dict.items():  # type: ignore[attr-defined]
            if inputs_from_state is not None and input_name in inputs_from_state:
                continue
            input_type = socket.type
            description = param_descriptions.get(input_name, f"Input '{input_name}' for the component.")

            # if the parameter has not a default value, Pydantic requires an Ellipsis (...)
            # to explicitly indicate that the parameter is required
            default = ... if socket.is_mandatory else socket.default_value
            resolved_type = _resolve_type(input_type)
            fields[input_name] = (resolved_type, Field(default=default, description=description))

        try:
            model = create_model(component.run.__name__, __doc__=component_run_description, **fields)
            parameters_schema = model.model_json_schema()
        except Exception as e:
            raise SchemaGenerationError(
                f"Failed to create JSON schema for the run method of Component '{component.__class__.__name__}'"
            ) from e

        # we don't want to include title keywords in the schema, as they contain redundant information
        # there is no programmatic way to prevent Pydantic from adding them, so we remove them later
        # see https://github.com/pydantic/pydantic/discussions/8504
        _remove_title_from_schema(parameters_schema)

        return parameters_schema

1	# SPDX-FileCopyrightText: 2022-present deepset GmbH <info@deepset.ai>
2	#
3	# SPDX-License-Identifier: Apache-2.0
4
5	from typing import Any, Callable, Dict, Optional, Union, get_args, get_origin	1✔
6
7	from pydantic import Field, TypeAdapter, create_model	1✔
8
9	from haystack import logging	1✔
10	from haystack.core.component import Component	1✔
11	from haystack.core.serialization import (	1✔
12	component_from_dict,
13	component_to_dict,
14	generate_qualified_class_name,
15	import_class_by_name,
16	)
17	from haystack.tools import Tool	1✔
18	from haystack.tools.errors import SchemaGenerationError	1✔
19	from haystack.tools.from_function import _remove_title_from_schema	1✔
20	from haystack.tools.parameters_schema_utils import _get_param_descriptions, _resolve_type	1✔
21	from haystack.utils.callable_serialization import deserialize_callable, serialize_callable	1✔
22
23	logger = logging.getLogger(__name__)	1✔
24
25
26	class ComponentTool(Tool):	1✔
27	"""
28	A Tool that wraps Haystack components, allowing them to be used as tools by LLMs.
29
30	ComponentTool automatically generates LLM-compatible tool schemas from component input sockets,
31	which are derived from the component's `run` method signature and type hints.
32
33
34	Key features:
35	- Automatic LLM tool calling schema generation from component input sockets
36	- Type conversion and validation for component inputs
37	- Support for types:
38	- Dataclasses
39	- Lists of dataclasses
40	- Basic types (str, int, float, bool, dict)
41	- Lists of basic types
42	- Automatic name generation from component class name
43	- Description extraction from component docstrings
44
45	To use ComponentTool, you first need a Haystack component - either an existing one or a new one you create.
46	You can create a ComponentTool from the component by passing the component to the ComponentTool constructor.
47	Below is an example of creating a ComponentTool from an existing SerperDevWebSearch component.
48
49	```python
50	from haystack import component, Pipeline
51	from haystack.tools import ComponentTool
52	from haystack.components.websearch import SerperDevWebSearch
53	from haystack.utils import Secret
54	from haystack.components.tools.tool_invoker import ToolInvoker
55	from haystack.components.generators.chat import OpenAIChatGenerator
56	from haystack.dataclasses import ChatMessage
57
58	# Create a SerperDev search component
59	search = SerperDevWebSearch(api_key=Secret.from_env_var("SERPERDEV_API_KEY"), top_k=3)
60
61	# Create a tool from the component
62	tool = ComponentTool(
63	component=search,
64	name="web_search", # Optional: defaults to "serper_dev_web_search"
65	description="Search the web for current information on any topic" # Optional: defaults to component docstring
66	)
67
68	# Create pipeline with OpenAIChatGenerator and ToolInvoker
69	pipeline = Pipeline()
70	pipeline.add_component("llm", OpenAIChatGenerator(model="gpt-4o-mini", tools=[tool]))
71	pipeline.add_component("tool_invoker", ToolInvoker(tools=[tool]))
72
73	# Connect components
74	pipeline.connect("llm.replies", "tool_invoker.messages")
75
76	message = ChatMessage.from_user("Use the web search tool to find information about Nikola Tesla")
77
78	# Run pipeline
79	result = pipeline.run({"llm": {"messages": [message]}})
80
81	print(result)
82	```
83
84	"""
85
86	def __init__(	1✔
87	self,
88	component: Component,
89	name: Optional[str] = None,
90	description: Optional[str] = None,
91	parameters: Optional[Dict[str, Any]] = None,
92	*,
93	outputs_to_string: Optional[Dict[str, Union[str, Callable[[Any], str]]]] = None,
94	inputs_from_state: Optional[Dict[str, str]] = None,
95	outputs_to_state: Optional[Dict[str, Dict[str, Union[str, Callable]]]] = None,
96	):
97	"""
98	Create a Tool instance from a Haystack component.
99
100	:param component: The Haystack component to wrap as a tool.
101	:param name: Optional name for the tool (defaults to snake_case of component class name).
102	:param description: Optional description (defaults to component's docstring).
103	:param parameters:
104	A JSON schema defining the parameters expected by the Tool.
105	Will fall back to the parameters defined in the component's run method signature if not provided.
106	:param outputs_to_string:
107	Optional dictionary defining how a tool outputs should be converted into a string.
108	If the source is provided only the specified output key is sent to the handler.
109	If the source is omitted the whole tool result is sent to the handler.
110	Example: {
111	"source": "docs", "handler": format_documents
112	}
113	:param inputs_from_state:
114	Optional dictionary mapping state keys to tool parameter names.
115	Example: {"repository": "repo"} maps state's "repository" to tool's "repo" parameter.
116	:param outputs_to_state:
117	Optional dictionary defining how tool outputs map to keys within state as well as optional handlers.
118	If the source is provided only the specified output key is sent to the handler.
119	Example: {
120	"documents": {"source": "docs", "handler": custom_handler}
121	}
122	If the source is omitted the whole tool result is sent to the handler.
123	Example: {
124	"documents": {"handler": custom_handler}
125	}
126	:raises ValueError: If the component is invalid or schema generation fails.
127	"""
128	if not isinstance(component, Component):	1✔
129	message = (	1✔
130	f"Object {component!r} is not a Haystack component. "
131	"Use ComponentTool only with Haystack component instances."
132	)
133	raise ValueError(message)	1✔
134
135	if getattr(component, "__haystack_added_to_pipeline__", None):	1✔
136	msg = (	1✔
137	"Component has been added to a pipeline and can't be used to create a ComponentTool. "
138	"Create ComponentTool from a non-pipeline component instead."
139	)
140	raise ValueError(msg)	1✔
141
142	self._unresolved_parameters = parameters	1✔
143	# Create the tools schema from the component run method parameters
144	tool_schema = parameters or self._create_tool_parameters_schema(component, inputs_from_state or {})	1✔
145
146	def component_invoker(**kwargs):	1✔
147	"""
148	Invokes the component using keyword arguments provided by the LLM function calling/tool-generated response.
149
150	:param kwargs: The keyword arguments to invoke the component with.
151	:returns: The result of the component invocation.
152	"""
153	converted_kwargs = {}	1✔
154	input_sockets = component.__haystack_input__._sockets_dict	1✔
155	for param_name, param_value in kwargs.items():	1✔
156	param_type = input_sockets[param_name].type	1✔
157
158	# Check if the type (or list element type) has from_dict
159	target_type = get_args(param_type)[0] if get_origin(param_type) is list else param_type	1✔
160	if hasattr(target_type, "from_dict"):	1✔
161	if isinstance(param_value, list):	1✔
162	param_value = [target_type.from_dict(item) for item in param_value if isinstance(item, dict)]	1✔
163	elif isinstance(param_value, dict):	×
164	param_value = target_type.from_dict(param_value)	×
165	else:
166	# Let TypeAdapter handle both single values and lists
167	type_adapter = TypeAdapter(param_type)	1✔
168	param_value = type_adapter.validate_python(param_value)	1✔
169
170	converted_kwargs[param_name] = param_value	1✔
171	logger.debug(f"Invoking component {type(component)} with kwargs: {converted_kwargs}")	1✔
172	return component.run(**converted_kwargs)	1✔
173
174	# Generate a name for the tool if not provided
175	if not name:	1✔
176	class_name = component.__class__.__name__	1✔
177	# Convert camelCase/PascalCase to snake_case
178	name = "".join(	1✔
179	[
180	"_" + c.lower() if c.isupper() and i > 0 and not class_name[i - 1].isupper() else c.lower()
181	for i, c in enumerate(class_name)
182	]
183	).lstrip("_")
184
185	description = description or component.__doc__ or name	1✔
186
187	# Create the Tool instance with the component invoker as the function to be called and the schema
188	super().__init__(	1✔
189	name=name,
190	description=description,
191	parameters=tool_schema,
192	function=component_invoker,
193	inputs_from_state=inputs_from_state,
194	outputs_to_state=outputs_to_state,
195	outputs_to_string=outputs_to_string,
196	)
197	self._component = component	1✔
198
199	def to_dict(self) -> Dict[str, Any]:	1✔
200	"""
201	Serializes the ComponentTool to a dictionary.
202	"""
203	serialized_component = component_to_dict(obj=self._component, name=self.name)	1✔
204
205	serialized = {	1✔
206	"component": serialized_component,
207	"name": self.name,
208	"description": self.description,
209	"parameters": self._unresolved_parameters,
210	"outputs_to_string": self.outputs_to_string,
211	"inputs_from_state": self.inputs_from_state,
212	"outputs_to_state": self.outputs_to_state,
213	}
214
215	if self.outputs_to_state is not None:	1✔
216	serialized_outputs = {}	1✔
217	for key, config in self.outputs_to_state.items():	1✔
218	serialized_config = config.copy()	1✔
219	if "handler" in config:	1✔
220	serialized_config["handler"] = serialize_callable(config["handler"])	1✔
221	serialized_outputs[key] = serialized_config	1✔
222	serialized["outputs_to_state"] = serialized_outputs	1✔
223
224	if self.outputs_to_string is not None and self.outputs_to_string.get("handler") is not None:	1✔
225	serialized["outputs_to_string"] = serialize_callable(self.outputs_to_string["handler"])	×
226
227	return {"type": generate_qualified_class_name(type(self)), "data": serialized}	1✔
228
229	@classmethod	1✔
230	def from_dict(cls, data: Dict[str, Any]) -> "Tool":	1✔
231	"""
232	Deserializes the ComponentTool from a dictionary.
233	"""
234	inner_data = data["data"]	1✔
235	component_class = import_class_by_name(inner_data["component"]["type"])	1✔
236	component = component_from_dict(cls=component_class, data=inner_data["component"], name=inner_data["name"])	1✔
237
238	if "outputs_to_state" in inner_data and inner_data["outputs_to_state"]:	1✔
239	deserialized_outputs = {}	1✔
240	for key, config in inner_data["outputs_to_state"].items():	1✔
241	deserialized_config = config.copy()	1✔
242	if "handler" in config:	1✔
243	deserialized_config["handler"] = deserialize_callable(config["handler"])	1✔
244	deserialized_outputs[key] = deserialized_config	1✔
245	inner_data["outputs_to_state"] = deserialized_outputs	1✔
246
247	if (	1✔
248	inner_data.get("outputs_to_string") is not None
249	and inner_data["outputs_to_string"].get("handler") is not None
250	):
251	inner_data["outputs_to_string"]["handler"] = deserialize_callable(	×
252	inner_data["outputs_to_string"]["handler"]
253	)
254
255	return cls(	1✔
256	component=component,
257	name=inner_data["name"],
258	description=inner_data["description"],
259	parameters=inner_data.get("parameters", None),
260	outputs_to_string=inner_data.get("outputs_to_string", None),
261	inputs_from_state=inner_data.get("inputs_from_state", None),
262	outputs_to_state=inner_data.get("outputs_to_state", None),
263	)
264
265	def _create_tool_parameters_schema(self, component: Component, inputs_from_state: Dict[str, Any]) -> Dict[str, Any]:	1✔
266	"""
267	Creates an OpenAI tools schema from a component's run method parameters.
268
269	:param component: The component to create the schema from.
270	:raises SchemaGenerationError: If schema generation fails
271	:returns: OpenAI tools schema for the component's run method parameters.
272	"""
273	component_run_description, param_descriptions = _get_param_descriptions(component.run)	1✔
274
275	# collect fields (types and defaults) and descriptions from function parameters
276	fields: Dict[str, Any] = {}	1✔
277
278	for input_name, socket in component.__haystack_input__._sockets_dict.items(): # type: ignore[attr-defined]	1✔
279	if inputs_from_state is not None and input_name in inputs_from_state:	1✔
280	continue	1✔
281	input_type = socket.type	1✔
282	description = param_descriptions.get(input_name, f"Input '{input_name}' for the component.")	1✔
283
284	# if the parameter has not a default value, Pydantic requires an Ellipsis (...)
285	# to explicitly indicate that the parameter is required
286	default = ... if socket.is_mandatory else socket.default_value	1✔
287	resolved_type = _resolve_type(input_type)	1✔
288	fields[input_name] = (resolved_type, Field(default=default, description=description))	1✔
289
290	try:	1✔
291	model = create_model(component.run.__name__, __doc__=component_run_description, **fields)	1✔
292	parameters_schema = model.model_json_schema()	1✔
293	except Exception as e:	×
294	raise SchemaGenerationError(	×
295	f"Failed to create JSON schema for the run method of Component '{component.__class__.__name__}'"
296	) from e
297
298	# we don't want to include title keywords in the schema, as they contain redundant information
299	# there is no programmatic way to prevent Pydantic from adding them, so we remove them later
300	# see https://github.com/pydantic/pydantic/discussions/8504
301	_remove_title_from_schema(parameters_schema)	1✔
302
303	return parameters_schema	1✔

deepset-ai / haystack / 15022600357

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous