Files
dsProject/dsLightRag/Routes/XueBanRoute.py
2025-08-22 09:36:29 +08:00

124 lines
3.8 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

import logging
import os
import tempfile
from datetime import datetime
from fastapi import APIRouter, Request, File, UploadFile
from fastapi.responses import JSONResponse
# 创建路由路由器
router = APIRouter(prefix="/api", tags=["学伴"])
# 配置日志
logger = logging.getLogger(__name__)
# 导入学伴工具函数
from Util.XueBanUtil import get_xueban_response_async
@router.post("/xueban/upload-audio")
async def upload_audio(file: UploadFile = File(...)):
"""
上传音频文件并进行ASR处理
- 参数: file - 音频文件
- 返回: JSON包含识别结果
"""
try:
# 记录日志
logger.info(f"接收到音频文件: {file.filename}")
# 保存临时文件
timestamp = datetime.now().strftime("%Y%m%d%H%M%S")
file_ext = os.path.splitext(file.filename)[1]
temp_file_name = f"temp_audio_{timestamp}{file_ext}"
temp_file_path = os.path.join(tempfile.gettempdir(), temp_file_name)
with open(temp_file_path, "wb") as f:
content = await file.read()
f.write(content)
logger.info(f"音频文件已保存至临时目录: {temp_file_path}")
# 这里应该调用ASR服务进行处理
# 示例asr_result = await process_asr(temp_file_path)
# 为了演示,这里返回模拟结果
asr_result = {
"text": "这是一段测试音频的识别结果",
"confidence": 0.95,
"audio_duration": len(content) / 1024 / 16 # 估算音频时长
}
# 删除临时文件
os.remove(temp_file_path)
logger.info(f"临时文件已删除: {temp_file_path}")
# 返回识别结果
return JSONResponse(content={
"success": True,
"message": "音频识别成功",
"data": asr_result
})
except Exception as e:
logger.error(f"音频处理失败: {str(e)}")
return JSONResponse(content={
"success": False,
"message": f"音频处理失败: {str(e)}"
}, status_code=500)
# 实际应用中您需要实现ASR处理函数
async def process_asr(audio_path: str) -> dict:
"""
调用ASR服务处理音频文件
:param audio_path: 音频文件路径
:return: 识别结果字典
"""
# 这里应该集成实际的ASR服务
# 例如百度AI、阿里云、讯飞等ASR服务
# 或者本地的ASR模型
pass
@router.post("/xueban/chat")
async def chat_with_xueban(request: Request):
"""
与学伴大模型聊天的接口
- 参数: request body 中的 query_text (用户查询文本)
- 返回: JSON包含聊天响应
"""
try:
# 获取请求体数据
data = await request.json()
query_text = data.get("query_text", "")
if not query_text.strip():
return JSONResponse(content={
"success": False,
"message": "查询文本不能为空"
}, status_code=400)
# 记录日志
logger.info(f"接收到学伴聊天请求: {query_text}")
# 调用异步接口获取学伴响应
response_content = []
async for chunk in get_xueban_response_async(query_text, stream=True):
response_content.append(chunk)
full_response = "".join(response_content)
# 返回响应
return JSONResponse(content={
"success": True,
"message": "聊天成功",
"data": {
"response": full_response
}
})
except Exception as e:
logger.error(f"学伴聊天失败: {str(e)}")
return JSONResponse(content={
"success": False,
"message": f"聊天处理失败: {str(e)}"
}, status_code=500)