当前位置：首页 > news >正文

基于ReAction范式的问答系统实现demo

news 2025/6/7 7:25:46

基于ReAction范式的问答系统实现demo

参考文档

ReAct论文解读：LLM ReAct范式，在大语言模型中结合推理和动作

说明

由于我最近在做一个基于图数据库的问答系统，所以样例就以查询图数据背景，实现过程仅供参考，希望能够大家带来帮助。

源码

import os
import json
from typing import Generator, Optional, Dict, Any
from neo4j import GraphDatabase
from openai import OpenAI
from dotenv import load_dotenvload_dotenv()# ----------------------------
# Neo4j 工具类
# ----------------------------
class Neo4jSearchTool:def __init__(self):self.driver = GraphDatabase.driver(os.getenv("NEO4J_URI"),auth=(os.getenv("NEO4J_USER"),os.getenv("NEO4J_PASSWORD")))def run(self, query: str) -> str:try:with self.driver.session() as session:result = session.run(query)data = [dict(record) for record in result]return json.dumps(data, ensure_ascii=False) if data else "[]"except Exception as e:return f"ERROR: {str(e)}"class Neo4jSchemaTool:def __init__(self, driver):self.driver = driverdef get_node_schema(self, session):q = """CALL db.schema.nodeTypeProperties()YIELD nodeType, propertyName, propertyTypesRETURN nodeType, propertyName, propertyTypes"""schema = {}for rec in session.run(q):label = rec["nodeType"].strip(":`")prop = rec["propertyName"]types = ", ".join(rec["propertyTypes"]) or "Unknown"schema.setdefault(label, {})[prop] = typesreturn schema# ----------------------------------------------------------------------def get_relationship_schema(self, session):"""For each relType: collect property definitions + a sampled (srcLabel, tgtLabel)."""# 1) property mapq_props = """CALL db.schema.relTypeProperties()YIELD relType, propertyName, propertyTypesRETURN relType, propertyName, propertyTypes"""rel_schema = {}for rec in session.run(q_props):rtype = rec["relType"].strip(":`")prop = rec["propertyName"]if prop:types = ", ".join(rec["propertyTypes"]) or "Unknown"rel_schema.setdefault(rtype, {})[prop] = types# 2) sample endpoints for each relationship typefor rtype in rel_schema:q_sample = f"""MATCH (s)-[r:`{rtype}`]->(t)WITH labels(s)[0] AS src, labels(t)[0] AS tgtRETURN src, tgt LIMIT 1"""rec = session.run(q_sample).single()if rec:rel_schema[rtype]["_endpoints"] = [rec["src"], rec["tgt"]]else:  # no relationship instance foundrel_schema[rtype]["_endpoints"] = ["Unknown", "Unknown"]return rel_schemadef get_schema(self) -> dict:"""提取数据库中的所有标签、关系和属性"""with self.driver.session() as session:# 获取所有节点标签labels = self.get_node_schema(session)rel_types = self.get_relationship_schema(session)return {"NodeTypes": labels,"RelationshipTypes": rel_types}def format_schema_prompt(self) -> str:"""将schema转换为自然语言描述"""schema = self.get_schema()prompt = "数据库包含以下结构：\n"# 标签和属性prompt += "## 节点类型\n"prompt += json.dumps(schema["NodeTypes"],ensure_ascii=False)# 关系prompt += "\n## 关系类型\n"prompt += json.dumps(schema["RelationshipTypes"],ensure_ascii=False)return promptclass AnswerValidator:@staticmethoddef is_valid_answer(observation: str) -> bool:"""检查工具返回是否包含有效答案"""if observation.startswith("ERROR") or observation == "[]":return Falsetry:data = json.loads(observation)if isinstance(data, list) and len(data) > 0:first_item = data[0]# 检查是否有非空值return any(v for v in first_item.values() if v not in [None, ""])return Falseexcept:return False@staticmethoddef should_terminate(llm_response: str) -> bool:"""通过LLM判断是否应该终止"""prompt = f"""判断以下模型响应是否包含最终答案：
响应内容：{llm_response}只需返回true或false："""response = OpenAI().chat.completions.create(model="deepseek-chat",messages=[{"role": "user", "content": prompt}],temperature=0)return "true" in response.choices[0].message.content.lower()# ----------------------------
# ReAct 引擎
# ----------------------------
class ReActQASystem:def __init__(self):self.llm = OpenAI()neo4j_driver = GraphDatabase.driver(os.getenv("NEO4J_URI"),auth=(os.getenv("NEO4J_USER"), os.getenv("NEO4J_PASSWORD")))self.tools = {"neo4j_search": Neo4jSearchTool(),"get_schema": Neo4jSchemaTool(neo4j_driver)}self.schema_prompt = self.tools["get_schema"].format_schema_prompt()self.max_steps = 5def _build_prompt(self, query: str, scratchpad: str = "") -> str:base_prompt = f"""你是一个审计专家，需要根据数据库结构编写准确的Cypher查询。{self.schema_prompt}可用工具：- neo4j_search: 执行Cypher查询，输入应为JSON格式的{{"query": "MATCH..."}}当前问题：{query}历史步骤：{scratchpad}严格按格式响应：Thought: 分析问题并确认需要查询的标签和关系Action:```json{{"action": "工具名", "action_input": {{...}}}}```"""return base_promptdef execute(self, query: str) -> Generator[str, None, None]:scratchpad = ""for step in range(self.max_steps):# 调用LLM生成响应prompt = self._build_prompt(query, scratchpad)print(f"LLM prompt: {prompt}")response = self.llm.chat.completions.create(model="deepseek-chat",messages=[{"role": "user", "content": prompt}],temperature=0)content = response.choices[0].message.contentprint(f"LLM Response: {content}")print(f"================================================")# 解析响应thought, action = self._parse_response(content)scratchpad += f"\n{content}\n"if not action:yield f"Final Answer: {thought}"break# 执行工具调用tool_name = action["action"]if AnswerValidator.should_terminate(action["action_input"]):yield f"Final Answer: {action['action_input']}"breakelif tool_name in self.tools:tool_result = self.tools[tool_name].run(action["action_input"]["query"])observation = f"Observation: {tool_result}"scratchpad += observation + "\n"yield observationelse:yield f"ERROR: 未知工具 {tool_name}"def _parse_response(self, text: str) -> tuple[str, Optional[Dict]]:thought = ""action = None# 提取Thought部分thought_start = text.find("Thought:") + len("Thought:")thought_end = text.find("Action:")if thought_start >= 0 and thought_end >= 0:thought = text[thought_start:thought_end].strip()# 提取Action部分action_start = text.find("```json") + len("```json")action_end = text.find("```", action_start)if action_start >= 0 and action_end >= 0:try:action = json.loads(text[action_start:action_end].strip())except json.JSONDecodeError:passreturn thought, action# ----------------------------
# 主程序
# ----------------------------
def main():qa_system = ReActQASystem()print("审计问答系统已启动（输入quit退出）")while True:query = input("\n用户提问: ")if query.lower() == "quit":breakprint("\n系统响应:")for response in qa_system.execute(query):print(response)if __name__ == "__main__":main()