Doiiars
442 lines
17 KiB
Python
442 lines
17 KiB
Python
"""
|
||
Deep Search Agent主类
|
||
整合所有模块，实现完整的深度搜索流程
|
||
"""
|
||
|
||
import json
|
||
import os
|
||
import re
|
||
from datetime import datetime
|
||
from typing import Optional, Dict, Any, List
|
||
from loguru import logger
|
||
from .llms import LLMClient
|
||
from .nodes import (
|
||
ReportStructureNode,
|
||
FirstSearchNode,
|
||
ReflectionNode,
|
||
FirstSummaryNode,
|
||
ReflectionSummaryNode,
|
||
ReportFormattingNode
|
||
)
|
||
from .state import State
|
||
from .tools import BochaMultimodalSearch, BochaResponse
|
||
from .utils import settings, Settings, format_search_results_for_prompt
|
||
|
||
|
||
class DeepSearchAgent:
|
||
"""Deep Search Agent主类"""
|
||
|
||
def __init__(self, config: Optional[Settings] = None):
|
||
"""
|
||
初始化Deep Search Agent
|
||
|
||
Args:
|
||
config: 配置对象，如果不提供则自动加载
|
||
"""
|
||
self.config = config or settings
|
||
|
||
# 初始化LLM客户端
|
||
self.llm_client = self._initialize_llm()
|
||
|
||
# 初始化搜索工具集
|
||
self.search_agency = BochaMultimodalSearch(api_key=(self.config.BOCHA_API_KEY or self.config.BOCHA_WEB_SEARCH_API_KEY))
|
||
|
||
# 初始化节点
|
||
self._initialize_nodes()
|
||
|
||
# 状态
|
||
self.state = State()
|
||
|
||
# 确保输出目录存在
|
||
os.makedirs(self.config.OUTPUT_DIR, exist_ok=True)
|
||
|
||
logger.info(f"Meida Agent已初始化")
|
||
logger.info(f"使用LLM: {self.llm_client.get_model_info()}")
|
||
logger.info(f"搜索工具集: BochaMultimodalSearch (支持5种多模态搜索工具)")
|
||
|
||
def _initialize_llm(self) -> LLMClient:
|
||
"""初始化LLM客户端"""
|
||
return LLMClient(
|
||
api_key=(self.config.MEDIA_ENGINE_API_KEY or self.config.MINDSPIDER_API_KEY),
|
||
model_name=(self.config.MEDIA_ENGINE_MODEL_NAME or self.config.MINDSPIDER_MODEL_NAME),
|
||
base_url=(self.config.MEDIA_ENGINE_BASE_URL or self.config.MINDSPIDER_BASE_URL),
|
||
)
|
||
|
||
def _initialize_nodes(self):
|
||
"""初始化处理节点"""
|
||
self.first_search_node = FirstSearchNode(self.llm_client)
|
||
self.reflection_node = ReflectionNode(self.llm_client)
|
||
self.first_summary_node = FirstSummaryNode(self.llm_client)
|
||
self.reflection_summary_node = ReflectionSummaryNode(self.llm_client)
|
||
self.report_formatting_node = ReportFormattingNode(self.llm_client)
|
||
|
||
def _validate_date_format(self, date_str: str) -> bool:
|
||
"""
|
||
验证日期格式是否为YYYY-MM-DD
|
||
|
||
Args:
|
||
date_str: 日期字符串
|
||
|
||
Returns:
|
||
是否为有效格式
|
||
"""
|
||
if not date_str:
|
||
return False
|
||
|
||
# 检查格式
|
||
pattern = r'^\d{4}-\d{2}-\d{2}$'
|
||
if not re.match(pattern, date_str):
|
||
return False
|
||
|
||
# 检查日期是否有效
|
||
try:
|
||
datetime.strptime(date_str, '%Y-%m-%d')
|
||
return True
|
||
except ValueError:
|
||
return False
|
||
|
||
def execute_search_tool(self, tool_name: str, query: str, **kwargs) -> BochaResponse:
|
||
"""
|
||
执行指定的搜索工具
|
||
|
||
Args:
|
||
tool_name: 工具名称，可选值：
|
||
- "comprehensive_search": 全面综合搜索（默认）
|
||
- "web_search_only": 纯网页搜索
|
||
- "search_for_structured_data": 结构化数据查询
|
||
- "search_last_24_hours": 24小时内最新信息
|
||
- "search_last_week": 本周信息
|
||
query: 搜索查询
|
||
**kwargs: 额外参数（如max_results）
|
||
|
||
Returns:
|
||
BochaResponse对象
|
||
"""
|
||
logger.info(f" → 执行搜索工具: {tool_name}")
|
||
|
||
if tool_name == "comprehensive_search":
|
||
max_results = kwargs.get("max_results", 10)
|
||
return self.search_agency.comprehensive_search(query, max_results)
|
||
elif tool_name == "web_search_only":
|
||
max_results = kwargs.get("max_results", 15)
|
||
return self.search_agency.web_search_only(query, max_results)
|
||
elif tool_name == "search_for_structured_data":
|
||
return self.search_agency.search_for_structured_data(query)
|
||
elif tool_name == "search_last_24_hours":
|
||
return self.search_agency.search_last_24_hours(query)
|
||
elif tool_name == "search_last_week":
|
||
return self.search_agency.search_last_week(query)
|
||
else:
|
||
logger.info(f" ⚠️ 未知的搜索工具: {tool_name}，使用默认综合搜索")
|
||
return self.search_agency.comprehensive_search(query)
|
||
|
||
def research(self, query: str, save_report: bool = True) -> str:
|
||
"""
|
||
执行深度研究
|
||
|
||
Args:
|
||
query: 研究查询
|
||
save_report: 是否保存报告到文件
|
||
|
||
Returns:
|
||
最终报告内容
|
||
"""
|
||
logger.info(f"\n{'='*60}")
|
||
logger.info(f"开始深度研究: {query}")
|
||
logger.info(f"{'='*60}")
|
||
|
||
try:
|
||
# Step 1: 生成报告结构
|
||
self._generate_report_structure(query)
|
||
|
||
# Step 2: 处理每个段落
|
||
self._process_paragraphs()
|
||
|
||
# Step 3: 生成最终报告
|
||
final_report = self._generate_final_report()
|
||
|
||
# Step 4: 保存报告
|
||
if save_report:
|
||
self._save_report(final_report)
|
||
|
||
logger.info(f"\n{'='*60}")
|
||
logger.info("深度研究完成！")
|
||
logger.info(f"{'='*60}")
|
||
|
||
return final_report
|
||
|
||
except Exception as e:
|
||
import traceback
|
||
error_traceback = traceback.format_exc()
|
||
logger.error(f"研究过程中发生错误: {str(e)} \n错误堆栈: {error_traceback}")
|
||
raise e
|
||
|
||
def _generate_report_structure(self, query: str):
|
||
"""生成报告结构"""
|
||
logger.info(f"\n[步骤 1] 生成报告结构...")
|
||
|
||
# 创建报告结构节点
|
||
report_structure_node = ReportStructureNode(self.llm_client, query)
|
||
|
||
# 生成结构并更新状态
|
||
self.state = report_structure_node.mutate_state(state=self.state)
|
||
|
||
_message = f"报告结构已生成，共 {len(self.state.paragraphs)} 个段落:"
|
||
for i, paragraph in enumerate(self.state.paragraphs, 1):
|
||
_message += f"\n {i}. {paragraph.title}"
|
||
logger.info(_message)
|
||
|
||
def _process_paragraphs(self):
|
||
"""处理所有段落"""
|
||
total_paragraphs = len(self.state.paragraphs)
|
||
|
||
for i in range(total_paragraphs):
|
||
logger.info(f"\n[步骤 2.{i+1}] 处理段落: {self.state.paragraphs[i].title}")
|
||
logger.info("-" * 50)
|
||
|
||
# 初始搜索和总结
|
||
self._initial_search_and_summary(i)
|
||
|
||
# 反思循环
|
||
self._reflection_loop(i)
|
||
|
||
# 标记段落完成
|
||
self.state.paragraphs[i].research.mark_completed()
|
||
|
||
progress = (i + 1) / total_paragraphs * 100
|
||
logger.info(f"段落处理完成 ({progress:.1f}%)")
|
||
|
||
def _initial_search_and_summary(self, paragraph_index: int):
|
||
"""执行初始搜索和总结"""
|
||
paragraph = self.state.paragraphs[paragraph_index]
|
||
|
||
# 准备搜索输入
|
||
search_input = {
|
||
"title": paragraph.title,
|
||
"content": paragraph.content
|
||
}
|
||
|
||
# 生成搜索查询和工具选择
|
||
logger.info(" - 生成搜索查询...")
|
||
search_output = self.first_search_node.run(search_input)
|
||
search_query = search_output["search_query"]
|
||
search_tool = search_output.get("search_tool", "comprehensive_search") # 默认工具
|
||
reasoning = search_output["reasoning"]
|
||
|
||
logger.info(f" - 搜索查询: {search_query}")
|
||
logger.info(f" - 选择的工具: {search_tool}")
|
||
logger.info(f" - 推理: {reasoning}")
|
||
|
||
# 执行搜索
|
||
logger.info(" - 执行网络搜索...")
|
||
|
||
# 处理特殊参数（新的工具集不需要日期参数处理）
|
||
search_kwargs = {}
|
||
if search_tool in ["comprehensive_search", "web_search_only"]:
|
||
# 这些工具支持max_results参数
|
||
search_kwargs["max_results"] = 10
|
||
|
||
search_response = self.execute_search_tool(search_tool, search_query, **search_kwargs)
|
||
|
||
# 转换为兼容格式
|
||
search_results = []
|
||
if search_response and search_response.webpages:
|
||
# 每种搜索工具都有其特定的结果数量，这里取前10个作为上限
|
||
max_results = min(len(search_response.webpages), 10)
|
||
for result in search_response.webpages[:max_results]:
|
||
search_results.append({
|
||
'title': result.name,
|
||
'url': result.url,
|
||
'content': result.snippet,
|
||
'score': None, # Bocha API不提供score
|
||
'raw_content': result.snippet,
|
||
'published_date': result.date_last_crawled # 使用爬取日期
|
||
})
|
||
|
||
if search_results:
|
||
_message = f" - 找到 {len(search_results)} 个搜索结果"
|
||
for j, result in enumerate(search_results, 1):
|
||
date_info = f" (发布于: {result.get('published_date', 'N/A')})" if result.get('published_date') else ""
|
||
_message += f"\n {j}. {result['title'][:50]}...{date_info}"
|
||
logger.info(_message)
|
||
else:
|
||
logger.info(" - 未找到搜索结果")
|
||
|
||
# 更新状态中的搜索历史
|
||
paragraph.research.add_search_results(search_query, search_results)
|
||
|
||
# 生成初始总结
|
||
logger.info(" - 生成初始总结...")
|
||
summary_input = {
|
||
"title": paragraph.title,
|
||
"content": paragraph.content,
|
||
"search_query": search_query,
|
||
"search_results": format_search_results_for_prompt(
|
||
search_results, self.config.SEARCH_CONTENT_MAX_LENGTH
|
||
)
|
||
}
|
||
|
||
# 更新状态
|
||
self.state = self.first_summary_node.mutate_state(
|
||
summary_input, self.state, paragraph_index
|
||
)
|
||
|
||
logger.info(" - 初始总结完成")
|
||
|
||
def _reflection_loop(self, paragraph_index: int):
|
||
"""执行反思循环"""
|
||
paragraph = self.state.paragraphs[paragraph_index]
|
||
|
||
for reflection_i in range(self.config.MAX_REFLECTIONS):
|
||
logger.info(f" - 反思 {reflection_i + 1}/{self.config.MAX_REFLECTIONS}...")
|
||
|
||
# 准备反思输入
|
||
reflection_input = {
|
||
"title": paragraph.title,
|
||
"content": paragraph.content,
|
||
"paragraph_latest_state": paragraph.research.latest_summary
|
||
}
|
||
|
||
# 生成反思搜索查询
|
||
reflection_output = self.reflection_node.run(reflection_input)
|
||
search_query = reflection_output["search_query"]
|
||
search_tool = reflection_output.get("search_tool", "comprehensive_search") # 默认工具
|
||
reasoning = reflection_output["reasoning"]
|
||
|
||
logger.info(f" 反思查询: {search_query}")
|
||
logger.info(f" 选择的工具: {search_tool}")
|
||
logger.info(f" 反思推理: {reasoning}")
|
||
|
||
# 执行反思搜索
|
||
# 处理特殊参数
|
||
search_kwargs = {}
|
||
if search_tool in ["comprehensive_search", "web_search_only"]:
|
||
# 这些工具支持max_results参数
|
||
search_kwargs["max_results"] = 10
|
||
|
||
search_response = self.execute_search_tool(search_tool, search_query, **search_kwargs)
|
||
|
||
# 转换为兼容格式
|
||
search_results = []
|
||
if search_response and search_response.webpages:
|
||
# 每种搜索工具都有其特定的结果数量，这里取前10个作为上限
|
||
max_results = min(len(search_response.webpages), 10)
|
||
for result in search_response.webpages[:max_results]:
|
||
search_results.append({
|
||
'title': result.name,
|
||
'url': result.url,
|
||
'content': result.snippet,
|
||
'score': None, # Bocha API不提供score
|
||
'raw_content': result.snippet,
|
||
'published_date': result.date_last_crawled
|
||
})
|
||
|
||
if search_results:
|
||
_message = f" 找到 {len(search_results)} 个反思搜索结果"
|
||
for j, result in enumerate(search_results, 1):
|
||
date_info = f" (发布于: {result.get('published_date', 'N/A')})" if result.get('published_date') else ""
|
||
_message += f"\n {j}. {result['title'][:50]}...{date_info}"
|
||
logger.info(_message)
|
||
else:
|
||
logger.info(" 未找到反思搜索结果")
|
||
|
||
# 更新搜索历史
|
||
paragraph.research.add_search_results(search_query, search_results)
|
||
|
||
# 生成反思总结
|
||
reflection_summary_input = {
|
||
"title": paragraph.title,
|
||
"content": paragraph.content,
|
||
"search_query": search_query,
|
||
"search_results": format_search_results_for_prompt(
|
||
search_results, self.config.SEARCH_CONTENT_MAX_LENGTH
|
||
),
|
||
"paragraph_latest_state": paragraph.research.latest_summary
|
||
}
|
||
|
||
# 更新状态
|
||
self.state = self.reflection_summary_node.mutate_state(
|
||
reflection_summary_input, self.state, paragraph_index
|
||
)
|
||
|
||
logger.info(f" 反思 {reflection_i + 1} 完成")
|
||
|
||
def _generate_final_report(self) -> str:
|
||
"""生成最终报告"""
|
||
logger.info(f"\n[步骤 3] 生成最终报告...")
|
||
|
||
# 准备报告数据
|
||
report_data = []
|
||
for paragraph in self.state.paragraphs:
|
||
report_data.append({
|
||
"title": paragraph.title,
|
||
"paragraph_latest_state": paragraph.research.latest_summary
|
||
})
|
||
|
||
# 格式化报告
|
||
try:
|
||
final_report = self.report_formatting_node.run(report_data)
|
||
except Exception as e:
|
||
logger.info(f"LLM格式化失败，使用备用方法: {str(e)}")
|
||
final_report = self.report_formatting_node.format_report_manually(
|
||
report_data, self.state.report_title
|
||
)
|
||
|
||
# 更新状态
|
||
self.state.final_report = final_report
|
||
self.state.mark_completed()
|
||
|
||
logger.info("最终报告生成完成")
|
||
return final_report
|
||
|
||
def _save_report(self, report_content: str):
|
||
"""保存报告到文件"""
|
||
# 生成文件名
|
||
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
|
||
query_safe = "".join(c for c in self.state.query if c.isalnum() or c in (' ', '-', '_')).rstrip()
|
||
query_safe = query_safe.replace(' ', '_')[:30]
|
||
|
||
filename = f"deep_search_report_{query_safe}_{timestamp}.md"
|
||
filepath = os.path.join(self.config.OUTPUT_DIR, filename)
|
||
|
||
# 保存报告
|
||
with open(filepath, 'w', encoding='utf-8') as f:
|
||
f.write(report_content)
|
||
|
||
logger.info(f"报告已保存到: {filepath}")
|
||
|
||
# 保存状态（如果配置允许）
|
||
if self.config.SAVE_INTERMEDIATE_STATES:
|
||
state_filename = f"state_{query_safe}_{timestamp}.json"
|
||
state_filepath = os.path.join(self.config.OUTPUT_DIR, state_filename)
|
||
self.state.save_to_file(state_filepath)
|
||
logger.info(f"状态已保存到: {state_filepath}")
|
||
|
||
def get_progress_summary(self) -> Dict[str, Any]:
|
||
"""获取进度摘要"""
|
||
return self.state.get_progress_summary()
|
||
|
||
def load_state(self, filepath: str):
|
||
"""从文件加载状态"""
|
||
self.state = State.load_from_file(filepath)
|
||
logger.info(f"状态已从 {filepath} 加载")
|
||
|
||
def save_state(self, filepath: str):
|
||
"""保存状态到文件"""
|
||
self.state.save_to_file(filepath)
|
||
logger.info(f"状态已保存到 {filepath}")
|
||
|
||
|
||
def create_agent(config_file: Optional[str] = None) -> DeepSearchAgent:
|
||
"""
|
||
创建Deep Search Agent实例的便捷函数
|
||
|
||
Args:
|
||
config_file: 配置文件路径
|
||
|
||
Returns:
|
||
DeepSearchAgent实例
|
||
"""
|
||
settings = Settings()
|
||
return DeepSearchAgent(settings)
|