Files
django-vue3-admin-gd/ai_service/api/v1/ai_chat.py
2025-07-18 22:14:37 +08:00

121 lines
4.4 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 os
import asyncio
from fastapi import APIRouter, Depends, Request, Query
from fastapi.responses import StreamingResponse
from sqlalchemy.orm import Session
from typing import List
from pydantic import BaseModel, SecretStr
from langchain.chains import ConversationChain
from api.v1.vo import MessageVO
from deps.auth import get_current_user
from services.chat_service import ChatDBService
from db.session import get_db
from models.ai import ChatConversation, ChatMessage
from utils.resp import resp_success, Response
from langchain_deepseek import ChatDeepSeek
router = APIRouter()
def get_deepseek_llm(api_key: SecretStr, model: str):
# deepseek 兼容 OpenAI API需指定 base_url
return ChatDeepSeek(
api_key=api_key,
model=model,
streaming=True,
)
@router.post('/stream')
async def chat_stream(request: Request, user=Depends(get_current_user), db: Session = Depends(get_db)):
body = await request.json()
content = body.get('content')
conversation_id = body.get('conversation_id')
model = 'deepseek-chat'
api_key = os.getenv("DEEPSEEK_API_KEY")
llm = get_deepseek_llm(SecretStr(api_key), model)
if not content or not isinstance(content, str):
from fastapi.responses import JSONResponse
return JSONResponse({"error": "content不能为空"}, status_code=400)
user_id = user["user_id"]
# 1. 获取对话
try:
conversation = ChatDBService.get_conversation(db, conversation_id)
conversation = db.merge(conversation) # ✅ 防止 DetachedInstanceError
except ValueError as e:
from fastapi.responses import JSONResponse
return JSONResponse({"error": str(e)}, status_code=400)
# 2. 插入当前消息
ChatDBService.add_message(db, conversation, user_id, content)
context = [
("system", "You are a helpful assistant. Answer all questions to the best of your ability in {language}.")
]
# 3. 查询历史消息,组装上下文
history = ChatDBService.get_history(db, conversation.id)
# === 新增:如果只有一条消息,更新 title ===
if len(history) == 1:
ChatDBService.update_conversation_title(db, conversation.id, content[:255])
for msg in history:
# 假设 msg.type 存储的是 'user' 或 'assistant'
# role = msg.type if msg.type in ("user", "assistant") else "user"
context.append((msg.type, msg.content))
ai_reply = ""
async def event_generator():
nonlocal ai_reply
async for chunk in llm.astream(context):
if hasattr(chunk, 'content'):
ai_reply += chunk.content
yield f"data: {chunk.content}\n\n"
else:
ai_reply += chunk
yield f"data: {chunk}\n\n"
await asyncio.sleep(0.01)
# 只保留最新AI回复
if ai_reply:
ChatDBService.insert_ai_message(db, conversation, user_id, ai_reply, model)
return StreamingResponse(event_generator(), media_type='text/event-stream')
@router.post("/conversations")
def create_conversation(db: Session = Depends(get_db), user=Depends(get_current_user),):
user_id = user["user_id"]
model = 'deepseek-chat'
conversation = ChatDBService.get_or_create_conversation(db, None, user_id, model, '新对话')
return resp_success(data=conversation.id)
@router.get('/conversations')
async def get_conversations(
db: Session = Depends(get_db),
user=Depends(get_current_user)
):
"""获取当前用户的聊天对话列表last_message为字符串"""
user_id = user["user_id"]
conversations = db.query(ChatConversation).filter(ChatConversation.user_id == user_id).order_by(ChatConversation.update_time.desc()).all()
return resp_success(data=[
{
'id': c.id,
'title': c.title,
'update_time': c.update_time,
'last_message': c.messages[-1].content if c.messages else None,
}
for c in conversations
])
@router.get('/messages', response_model=Response[List[MessageVO]])
def get_messages(
conversation_id: int = Query(...),
db: Session = Depends(get_db),
user=Depends(get_current_user)
):
"""获取指定会话的消息列表(当前用户)"""
user_id = user["user_id"]
query = db.query(ChatMessage).filter(ChatMessage.conversation_id == conversation_id,
ChatMessage.user_id == user_id).order_by(ChatMessage.id).all()
return resp_success(data=query)