File size: 15,259 Bytes
f1e6b80 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 |
"""Chain that makes API calls and summarizes the responses to answer a question."""
from __future__ import annotations
from typing import Any, Dict, List, Optional, Sequence, Tuple
from urllib.parse import urlparse
from langchain_core._api import deprecated
from langchain_core.callbacks import (
AsyncCallbackManagerForChainRun,
CallbackManagerForChainRun,
)
from langchain_core.language_models import BaseLanguageModel
from langchain_core.prompts import BasePromptTemplate
from pydantic import Field, model_validator
from typing_extensions import Self
from langchain.chains.api.prompt import API_RESPONSE_PROMPT, API_URL_PROMPT
from langchain.chains.base import Chain
from langchain.chains.llm import LLMChain
def _extract_scheme_and_domain(url: str) -> Tuple[str, str]:
"""Extract the scheme + domain from a given URL.
Args:
url (str): The input URL.
Returns:
return a 2-tuple of scheme and domain
"""
parsed_uri = urlparse(url)
return parsed_uri.scheme, parsed_uri.netloc
def _check_in_allowed_domain(url: str, limit_to_domains: Sequence[str]) -> bool:
"""Check if a URL is in the allowed domains.
Args:
url (str): The input URL.
limit_to_domains (Sequence[str]): The allowed domains.
Returns:
bool: True if the URL is in the allowed domains, False otherwise.
"""
scheme, domain = _extract_scheme_and_domain(url)
for allowed_domain in limit_to_domains:
allowed_scheme, allowed_domain = _extract_scheme_and_domain(allowed_domain)
if scheme == allowed_scheme and domain == allowed_domain:
return True
return False
try:
from langchain_community.utilities.requests import TextRequestsWrapper
@deprecated(
since="0.2.13",
message=(
"This class is deprecated and will be removed in langchain 1.0. "
"See API reference for replacement: "
"https://api.python.langchain.com/en/latest/chains/langchain.chains.api.base.APIChain.html" # noqa: E501
),
removal="1.0",
)
class APIChain(Chain):
"""Chain that makes API calls and summarizes the responses to answer a question.
*Security Note*: This API chain uses the requests toolkit
to make GET, POST, PATCH, PUT, and DELETE requests to an API.
Exercise care in who is allowed to use this chain. If exposing
to end users, consider that users will be able to make arbitrary
requests on behalf of the server hosting the code. For example,
users could ask the server to make a request to a private API
that is only accessible from the server.
Control access to who can submit issue requests using this toolkit and
what network access it has.
See https://python.langchain.com/docs/security for more information.
Note: this class is deprecated. See below for a replacement implementation
using LangGraph. The benefits of this implementation are:
- Uses LLM tool calling features to encourage properly-formatted API requests;
- Support for both token-by-token and step-by-step streaming;
- Support for checkpointing and memory of chat history;
- Easier to modify or extend (e.g., with additional tools, structured responses, etc.)
Install LangGraph with:
.. code-block:: bash
pip install -U langgraph
.. code-block:: python
from typing import Annotated, Sequence
from typing_extensions import TypedDict
from langchain.chains.api.prompt import API_URL_PROMPT
from langchain_community.agent_toolkits.openapi.toolkit import RequestsToolkit
from langchain_community.utilities.requests import TextRequestsWrapper
from langchain_core.messages import BaseMessage
from langchain_core.prompts import ChatPromptTemplate
from langchain_openai import ChatOpenAI
from langchain_core.runnables import RunnableConfig
from langgraph.graph import END, StateGraph
from langgraph.graph.message import add_messages
from langgraph.prebuilt.tool_node import ToolNode
# NOTE: There are inherent risks in giving models discretion
# to execute real-world actions. We must "opt-in" to these
# risks by setting allow_dangerous_request=True to use these tools.
# This can be dangerous for calling unwanted requests. Please make
# sure your custom OpenAPI spec (yaml) is safe and that permissions
# associated with the tools are narrowly-scoped.
ALLOW_DANGEROUS_REQUESTS = True
# Subset of spec for https://jsonplaceholder.typicode.com
api_spec = \"\"\"
openapi: 3.0.0
info:
title: JSONPlaceholder API
version: 1.0.0
servers:
- url: https://jsonplaceholder.typicode.com
paths:
/posts:
get:
summary: Get posts
parameters: &id001
- name: _limit
in: query
required: false
schema:
type: integer
example: 2
description: Limit the number of results
\"\"\"
llm = ChatOpenAI(model="gpt-4o-mini", temperature=0)
toolkit = RequestsToolkit(
requests_wrapper=TextRequestsWrapper(headers={}), # no auth required
allow_dangerous_requests=ALLOW_DANGEROUS_REQUESTS,
)
tools = toolkit.get_tools()
api_request_chain = (
API_URL_PROMPT.partial(api_docs=api_spec)
| llm.bind_tools(tools, tool_choice="any")
)
class ChainState(TypedDict):
\"\"\"LangGraph state.\"\"\"
messages: Annotated[Sequence[BaseMessage], add_messages]
async def acall_request_chain(state: ChainState, config: RunnableConfig):
last_message = state["messages"][-1]
response = await api_request_chain.ainvoke(
{"question": last_message.content}, config
)
return {"messages": [response]}
async def acall_model(state: ChainState, config: RunnableConfig):
response = await llm.ainvoke(state["messages"], config)
return {"messages": [response]}
graph_builder = StateGraph(ChainState)
graph_builder.add_node("call_tool", acall_request_chain)
graph_builder.add_node("execute_tool", ToolNode(tools))
graph_builder.add_node("call_model", acall_model)
graph_builder.set_entry_point("call_tool")
graph_builder.add_edge("call_tool", "execute_tool")
graph_builder.add_edge("execute_tool", "call_model")
graph_builder.add_edge("call_model", END)
chain = graph_builder.compile()
.. code-block:: python
example_query = "Fetch the top two posts. What are their titles?"
events = chain.astream(
{"messages": [("user", example_query)]},
stream_mode="values",
)
async for event in events:
event["messages"][-1].pretty_print()
""" # noqa: E501
api_request_chain: LLMChain
api_answer_chain: LLMChain
requests_wrapper: TextRequestsWrapper = Field(exclude=True)
api_docs: str
question_key: str = "question" #: :meta private:
output_key: str = "output" #: :meta private:
limit_to_domains: Optional[Sequence[str]] = Field(
default_factory=list # type: ignore
)
"""Use to limit the domains that can be accessed by the API chain.
* For example, to limit to just the domain `https://www.example.com`, set
`limit_to_domains=["https://www.example.com"]`.
* The default value is an empty tuple, which means that no domains are
allowed by default. By design this will raise an error on instantiation.
* Use a None if you want to allow all domains by default -- this is not
recommended for security reasons, as it would allow malicious users to
make requests to arbitrary URLS including internal APIs accessible from
the server.
"""
@property
def input_keys(self) -> List[str]:
"""Expect input key.
:meta private:
"""
return [self.question_key]
@property
def output_keys(self) -> List[str]:
"""Expect output key.
:meta private:
"""
return [self.output_key]
@model_validator(mode="after")
def validate_api_request_prompt(self) -> Self:
"""Check that api request prompt expects the right variables."""
input_vars = self.api_request_chain.prompt.input_variables
expected_vars = {"question", "api_docs"}
if set(input_vars) != expected_vars:
raise ValueError(
f"Input variables should be {expected_vars}, got {input_vars}"
)
return self
@model_validator(mode="before")
@classmethod
def validate_limit_to_domains(cls, values: Dict) -> Any:
"""Check that allowed domains are valid."""
# This check must be a pre=True check, so that a default of None
# won't be set to limit_to_domains if it's not provided.
if "limit_to_domains" not in values:
raise ValueError(
"You must specify a list of domains to limit access using "
"`limit_to_domains`"
)
if (
not values["limit_to_domains"]
and values["limit_to_domains"] is not None
):
raise ValueError(
"Please provide a list of domains to limit access using "
"`limit_to_domains`."
)
return values
@model_validator(mode="after")
def validate_api_answer_prompt(self) -> Self:
"""Check that api answer prompt expects the right variables."""
input_vars = self.api_answer_chain.prompt.input_variables
expected_vars = {"question", "api_docs", "api_url", "api_response"}
if set(input_vars) != expected_vars:
raise ValueError(
f"Input variables should be {expected_vars}, got {input_vars}"
)
return self
def _call(
self,
inputs: Dict[str, Any],
run_manager: Optional[CallbackManagerForChainRun] = None,
) -> Dict[str, str]:
_run_manager = run_manager or CallbackManagerForChainRun.get_noop_manager()
question = inputs[self.question_key]
api_url = self.api_request_chain.predict(
question=question,
api_docs=self.api_docs,
callbacks=_run_manager.get_child(),
)
_run_manager.on_text(api_url, color="green", end="\n", verbose=self.verbose)
api_url = api_url.strip()
if self.limit_to_domains and not _check_in_allowed_domain(
api_url, self.limit_to_domains
):
raise ValueError(
f"{api_url} is not in the allowed domains: {self.limit_to_domains}"
)
api_response = self.requests_wrapper.get(api_url)
_run_manager.on_text(
str(api_response), color="yellow", end="\n", verbose=self.verbose
)
answer = self.api_answer_chain.predict(
question=question,
api_docs=self.api_docs,
api_url=api_url,
api_response=api_response,
callbacks=_run_manager.get_child(),
)
return {self.output_key: answer}
async def _acall(
self,
inputs: Dict[str, Any],
run_manager: Optional[AsyncCallbackManagerForChainRun] = None,
) -> Dict[str, str]:
_run_manager = (
run_manager or AsyncCallbackManagerForChainRun.get_noop_manager()
)
question = inputs[self.question_key]
api_url = await self.api_request_chain.apredict(
question=question,
api_docs=self.api_docs,
callbacks=_run_manager.get_child(),
)
await _run_manager.on_text(
api_url, color="green", end="\n", verbose=self.verbose
)
api_url = api_url.strip()
if self.limit_to_domains and not _check_in_allowed_domain(
api_url, self.limit_to_domains
):
raise ValueError(
f"{api_url} is not in the allowed domains: {self.limit_to_domains}"
)
api_response = await self.requests_wrapper.aget(api_url)
await _run_manager.on_text(
str(api_response), color="yellow", end="\n", verbose=self.verbose
)
answer = await self.api_answer_chain.apredict(
question=question,
api_docs=self.api_docs,
api_url=api_url,
api_response=api_response,
callbacks=_run_manager.get_child(),
)
return {self.output_key: answer}
@classmethod
def from_llm_and_api_docs(
cls,
llm: BaseLanguageModel,
api_docs: str,
headers: Optional[dict] = None,
api_url_prompt: BasePromptTemplate = API_URL_PROMPT,
api_response_prompt: BasePromptTemplate = API_RESPONSE_PROMPT,
limit_to_domains: Optional[Sequence[str]] = tuple(),
**kwargs: Any,
) -> APIChain:
"""Load chain from just an LLM and the api docs."""
get_request_chain = LLMChain(llm=llm, prompt=api_url_prompt)
requests_wrapper = TextRequestsWrapper(headers=headers)
get_answer_chain = LLMChain(llm=llm, prompt=api_response_prompt)
return cls(
api_request_chain=get_request_chain,
api_answer_chain=get_answer_chain,
requests_wrapper=requests_wrapper,
api_docs=api_docs,
limit_to_domains=limit_to_domains,
**kwargs,
)
@property
def _chain_type(self) -> str:
return "api_chain"
except ImportError:
class APIChain: # type: ignore[no-redef]
def __init__(self, *args: Any, **kwargs: Any) -> None:
raise ImportError(
"To use the APIChain, you must install the langchain_community package."
"pip install langchain_community"
)
|