8042750d41
* stage academic conversation * stage document conversation * fix buggy gradio version * file dynamic load * merge more academic plugins * accelerate nltk * feat: 为predict函数添加文件和URL读取功能 - 添加URL检测和网页内容提取功能，支持自动提取网页文本 - 添加文件路径识别和文件内容读取功能，支持private_upload路径格式 - 集成WebTextExtractor处理网页内容提取 - 集成TextContentLoader处理本地文件读取 - 支持文件路径与问题组合的智能处理 * back * block unstable --------- Co-authored-by: XiaoBoAI <liuboyin2019@ia.ac.cn>
64 lines
2.2 KiB
Python
64 lines
2.2 KiB
Python
from typing import List, Dict, Any
|
|
from .query_analyzer import QueryAnalyzer, SearchCriteria
|
|
from .data_sources.arxiv_source import ArxivSource
|
|
from .data_sources.semantic_source import SemanticScholarSource
|
|
from .handlers.review_handler import 文献综述功能
|
|
from .handlers.recommend_handler import 论文推荐功能
|
|
from .handlers.qa_handler import 学术问答功能
|
|
from .handlers.paper_handler import 单篇论文分析功能
|
|
|
|
class QueryProcessor:
|
|
"""查询处理器"""
|
|
|
|
def __init__(self):
|
|
self.analyzer = QueryAnalyzer()
|
|
self.arxiv = ArxivSource()
|
|
self.semantic = SemanticScholarSource()
|
|
|
|
# 初始化各种处理器
|
|
self.handlers = {
|
|
"review": 文献综述功能(self.arxiv, self.semantic),
|
|
"recommend": 论文推荐功能(self.arxiv, self.semantic),
|
|
"qa": 学术问答功能(self.arxiv, self.semantic),
|
|
"paper": 单篇论文分析功能(self.arxiv, self.semantic)
|
|
}
|
|
|
|
async def process_query(
|
|
self,
|
|
query: str,
|
|
chatbot: List[List[str]],
|
|
history: List[List[str]],
|
|
system_prompt: str,
|
|
llm_kwargs: Dict[str, Any],
|
|
plugin_kwargs: Dict[str, Any],
|
|
) -> List[List[str]]:
|
|
"""处理用户查询"""
|
|
|
|
# 设置默认的插件参数
|
|
default_plugin_kwargs = {
|
|
'max_papers': 20, # 最大论文数量
|
|
'min_year': 2015, # 最早年份
|
|
'search_multiplier': 3, # 检索倍数
|
|
}
|
|
# 更新插件参数
|
|
plugin_kwargs.update({k: v for k, v in default_plugin_kwargs.items() if k not in plugin_kwargs})
|
|
|
|
# 1. 分析查询意图
|
|
criteria = self.analyzer.analyze_query(query, chatbot, llm_kwargs)
|
|
|
|
# 2. 根据查询类型选择处理器
|
|
handler = self.handlers.get(criteria.query_type)
|
|
if not handler:
|
|
handler = self.handlers["qa"] # 默认使用QA处理器
|
|
|
|
# 3. 处理查询
|
|
response = await handler.handle(
|
|
criteria,
|
|
chatbot,
|
|
history,
|
|
system_prompt,
|
|
llm_kwargs,
|
|
plugin_kwargs
|
|
)
|
|
|
|
return response |