通过 API 将Deepseek响应流式内容输出到前端

要实现通过 API 将流式内容输出到前端，可以采用以下技术方案（以 Python 后端 + 前端 JavaScript 为例）：

方案一：使用 Server-Sent Events (SSE)

这是浏览器原生支持的流式传输方案，推荐首选

# Flask 示例
from flask import Response, stream_with_context@app.route('/stream')
def stream_data():def generate():response = client.chat.completions.create(model="deepseek-chat",messages=messages,stream=True)for chunk in response:if chunk.choices:content = chunk.choices[0].delta.content or ""# SSE 格式要求 data: 前缀和双换行符yield f"data: {json.dumps({'content': content})}\n\n"return Response(stream_with_context(generate()), mimetype='text/event-stream')

// 前端 JavaScript
const eventSource = new EventSource('/stream');eventSource.onmessage = (event) => {const data = JSON.parse(event.data);document.getElementById('output').innerHTML += data.content;
};eventSource.onerror = (err) => {console.error('EventSource failed:', err);eventSource.close();
};

方案二：使用流式 HTTP 响应（NDJSON）

更通用的流式传输方案，适合非浏览器客户端

# FastAPI 示例
from fastapi import APIRouter
from fastapi.responses import StreamingResponse
import json@app.get("/stream")
async def stream_data():async def generate():response = client.chat.completions.create(model="deepseek-chat",messages=messages,stream=True)async for chunk in response:if chunk.choices:content = chunk.choices[0].delta.content or ""yield json.dumps({"content": content}) + "\n"  # NDJSON 格式return StreamingResponse(generate(), media_type='application/x-ndjson')

// 前端 JavaScript 使用 Fetch API
async function streamData() {const response = await fetch('/stream');const reader = response.body.getReader();const decoder = new TextDecoder();while(true) {const { done, value } = await reader.read();if(done) break;const chunk = decoder.decode(value);const data = JSON.parse(chunk);document.getElementById('output').innerHTML += data.content;}
}

关键配置说明

响应头设置：

# Flask
headers = {'Cache-Control': 'no-cache','Connection': 'keep-alive'
}# FastAPI 自动处理

数据格式选择：
- SSE (text/event-stream)：浏览器原生支持，自动重连
- NDJSON (application/x-ndjson)：更通用的流式 JSON 格式
- 纯文本流：简单但结构化能力弱

前端处理建议：

// 更健壮的读取方式（处理分块不完整情况）
let buffer = '';async function processChunk(chunk) {buffer += chunk;while(buffer.includes('\n')) {const lineEnd = buffer.indexOf('\n');const line = buffer.slice(0, lineEnd);buffer = buffer.slice(lineEnd + 1);try {const data = JSON.parse(line);// 处理数据...} catch(e) {console.error('解析错误:', e);}}
}

完整工作流程示例（FastAPI + React）

后端：

# main.py
from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddlewareapp = FastAPI()app.add_middleware(CORSMiddleware,allow_origins=["*"],allow_methods=["*"],allow_headers=["*"],
)@app.get("/chat")
async def chat_stream(prompt: str):async def generate():response = client.chat.completions.create(model="deepseek-chat",messages=[{"role": "user", "content": prompt}],stream=True)async for chunk in response:if content := chunk.choices[0].delta.content:yield json.dumps({"content": content})return StreamingResponse(generate(), media_type="application/x-ndjson")

前端 React 组件：

// ChatComponent.jsx
import { useState } from 'react';export default function ChatComponent() {const [output, setOutput] = useState('');const startStream = async () => {const response = await fetch('http://api/chat?prompt=你好');const reader = response.body.getReader();const decoder = new TextDecoder();let buffer = '';while(true) {const { done, value } = await reader.read();if(done) break;buffer += decoder.decode(value);while(buffer.includes('}')) {const endIndex = buffer.indexOf('}') + 1;const chunk = buffer.slice(0, endIndex);buffer = buffer.slice(endIndex);try {const data = JSON.parse(chunk);setOutput(prev => prev + data.content);} catch(e) {console.error('解析错误:', e);}}}};return (<div><button onClick={startStream}>开始对话</button><div id="output">{output}</div></div>);
}

注意事项

连接管理：

设置合理的超时时间（通常 30-60 秒）
处理客户端提前断开连接的情况

# FastAPI 示例
try:async for chunk in response:# ...处理数据if await request.is_disconnected():break
finally:await client.close()  # 清理资源

性能优化：

使用异步框架（FastAPI 性能优于 Flask）
启用响应压缩

app = FastAPI()
@app.middleware("http")
async def add_compression(request, call_next):response = await call_next(request)response.headers["Content-Encoding"] = "gzip"return response

安全考虑：

限制最大并发连接数
实施速率限制

from fastapi import Request
from fastapi.middleware import Middleware
from slowapi import Limiter
from slowapi.util import get_remote_addresslimiter = Limiter(key_func=get_remote_address)
app.state.limiter = limiter@app.get("/chat")
@limiter.limit("10/minute")
async def chat_stream(request: Request):# ...

错误处理增强：

async def generate():try:response = client.chat.completions.create(...)async for chunk in response:# 处理数据...except Exception as e:yield json.dumps({"error": str(e)})finally:await client.close()  # 确保释放资源