dsProject/dsRagAnything/Start.py

import json
import logging
import os
from logging.handlers import RotatingFileHandler

import fastapi
import uvicorn
from fastapi import FastAPI
from lightrag import LightRAG
from lightrag.kg.shared_storage import initialize_pipeline_status
from raganything import RAGAnything
from sse_starlette import EventSourceResponse
from starlette.staticfiles import StaticFiles

from Util.RagUtil import create_llm_model_func, create_vision_model_func, create_embedding_func

# 初始化日志
logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)

# 配置日志处理器
log_file = os.path.join(os.path.dirname(__file__), 'Logs', 'app.log')
os.makedirs(os.path.dirname(log_file), exist_ok=True)

# 文件处理器
file_handler = RotatingFileHandler(
    log_file, maxBytes=1024 * 1024, backupCount=5, encoding='utf-8')
file_handler.setFormatter(logging.Formatter(
    '%(asctime)s - %(name)s - %(levelname)s - %(message)s'))

# 控制台处理器
console_handler = logging.StreamHandler()
console_handler.setFormatter(logging.Formatter(
    '%(asctime)s - %(name)s - %(levelname)s - %(message)s'))

logger.addHandler(file_handler)
logger.addHandler(console_handler)


async def lifespan(app: FastAPI):
    yield


async def print_stream(stream):
    async for chunk in stream:
        if chunk:
            print(chunk, end="", flush=True)


app = FastAPI(lifespan=lifespan)

# 挂载静态文件目录
app.mount("/static", StaticFiles(directory="Static"), name="static")


@app.post("/api/rag")
async def rag(request: fastapi.Request):
    data = await request.json()
    topic = data.get("topic") # Chinese, Math
    # 拼接路径
    WORKING_PATH= "./Topic/" + topic
    # 查询的问题
    query = data.get("query")
    # 关闭参考资料
    query = query + "\n 1、不要输出参考资料 或者 References ！"
    query = query + "\n 2、如果问题与提供的知识库内容不符，则明确告诉未在知识库范围内提到！"
    #query = query + "\n 3、严格按照知识库内容进行回答，不要进行推理！知识库中条目清晰体现的才算答案。"

    #system_prompt = "注意：如果用户提供了Latex 的行间公式或行内公式，一定要严格按原文输出,不能修改！"

    async def generate_response_stream(query: str):
        try:
            # 初始化RAG组件
            llm_model_func = create_llm_model_func()
            vision_model_func = create_vision_model_func(llm_model_func)
            embedding_func = create_embedding_func()

            lightrag_instance = LightRAG(
                working_dir=WORKING_PATH,
                llm_model_func=llm_model_func,
                embedding_func=embedding_func
            )

            await lightrag_instance.initialize_storages()
            await initialize_pipeline_status()

            # 创建RAG实例并保存到app.state
            app.state.rag = RAGAnything(
                lightrag=lightrag_instance,
                vision_model_func=vision_model_func,
            )
            # 直接使用app.state中已初始化的rag实例
            resp = await app.state.rag.aquery(
                query=query,
                mode="hybrid",
                stream=True
            )

            async for chunk in resp:
                if not chunk:
                    continue
                yield f"data: {json.dumps({'reply': chunk})}\n\n"
                print(chunk, end='', flush=True)
        except Exception as e:
            yield f"data: {json.dumps({'error': str(e)})}\n\n"
            logger.error(f"处理查询时出错: {query}. 错误: {str(e)}")
        finally:
            # 清理资源
            await app.state.rag.lightrag.finalize_storages()

    return EventSourceResponse(generate_response_stream(query=query))


if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=8000)