通过 API 将Deepseek响应流式内容输出到前端

一个灵活的死胖子

于 2025-05-03 17:56:08 发布

阅读量81

点赞数

文章标签：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 CORSMiddleware

app = 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_address

limiter = 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