主题
Stream 插件
¥Stream Plugin
警告
此插件处于维护模式，不会添加新功能。我们建议使用 改为使用 Generator Stream
¥This plugin is in maintenance mode and will not receive new features. We recommend using the Generator Stream instead
此插件添加了对流式响应或将服务器发送事件发送回客户端的支持。
¥This plugin adds support for streaming response or sending Server-Sent Event back to the client.
使用以下工具安装：
¥Install with:
bash
bun add @elysiajs/stream然后使用它：
¥Then use it:
typescript
import { Elysia } from 'elysia'
import { Stream } from '@elysiajs/stream'
new Elysia()
.get('/', () => new Stream(async (stream) => {
stream.send('hello')
await stream.wait(1000)
stream.send('world')
stream.close()
}))
.listen(3000)默认情况下，Stream 将返回 Response 和 content-type 或 text/event-stream; charset=utf8。
¥By default, Stream will return Response with content-type of text/event-stream; charset=utf8.
构造函数
¥Constructor
以下是 Stream 接受的构造函数参数：
¥Below is the constructor parameter accepted by Stream:
Stream：
自动：根据提供的值自动流式响应
可迭代对象
AsyncIterable
ReadableStream
响应
手动：
(stream: this) => unknown或undefined的回调
选项：
StreamOptions
方法
¥Method
以下是 Stream 提供的方法：
¥Below is the method provided by Stream:
send
将数据排队到流中以发送回客户端
¥Enqueue data to stream to send back to the client
close
关闭流
¥Close the stream
wait
返回一个以毫秒为单位解析提供的值的 Promise。
¥Return a promise that resolves in the provided value in ms
value
ReadableStream 的内部值
¥Inner value of the ReadableStream
模式
¥Pattern
以下是使用该插件的常见模式。
¥Below you can find the common patterns to use the plugin.
OpenAI
当参数为 Iterable 或 AsyncIterable 时，将触发自动模式，并自动将响应流式传输回客户端。
¥Automatic mode is triggered when the parameter is either Iterable or AsyncIterable streaming the response back to the client automatically.
以下是将 ChatGPT 集成到 Elysia 的示例。
¥Below is an example of integrating ChatGPT into Elysia.
ts
new Elysia()
.get(
'/ai',
({ query: { prompt } }) =>
new Stream(
openai.chat.completions.create({
model: 'gpt-3.5-turbo',
stream: true,
messages: [{
role: 'user',
content: prompt
}]
})
)
)默认情况下，openai chatGPT 完成返回 AsyncIterable，因此你应该能够将 OpenAI 封装在 Stream 中。
¥By default openai chatGPT completion returns AsyncIterable so you should be able to wrap the OpenAI in Stream.
获取流
¥Fetch Stream
你可以将从返回流的端点的获取传递给代理流。
¥You can pass a fetch from an endpoint that returns the stream to proxy a stream.
这对于使用 AI 文本生成的端点非常有用，因为你可以直接代理它，例如。 Cloudflare AI。
¥This is useful for those endpoints that use AI text generation since you can proxy it directly, eg. Cloudflare AI.
ts
const model = '@cf/meta/llama-2-7b-chat-int8'
const endpoint = `https://api.cloudflare.com/client/v4/accounts/${process.env.ACCOUNT_ID}/ai/run/${model}`
new Elysia()
.get('/ai', ({ query: { prompt } }) =>
fetch(endpoint, {
method: 'POST',
headers: {
authorization: `Bearer ${API_TOKEN}`,
'content-type': 'application/json'
},
body: JSON.stringify({
messages: [
{ role: 'system', content: 'You are a friendly assistant' },
{ role: 'user', content: prompt }
]
})
})
)服务器发送事件
¥Server Sent Event
当参数为 callback 或 undefined 时，将触发手动模式，让你可以控制流。
¥Manual mode is triggered when the parameter is either callback or undefined, allowing you to control the stream.
callback-based
以下是使用构造函数回调创建服务器发送事件端点的示例。
¥Below is an example of creating a Server-Sent Event endpoint using a constructor callback
ts
new Elysia()
.get('/source', () =>
new Stream((stream) => {
const interval = setInterval(() => {
stream.send('hello world')
}, 500)
setTimeout(() => {
clearInterval(interval)
stream.close()
}, 3000)
})
)value-based
以下是使用基于值的构造函数创建服务器发送事件端点的示例。
¥Below is an example of creating a Server-Sent Event endpoint using a value-based
ts
new Elysia()
.get('/source', () => {
const stream = new Stream()
const interval = setInterval(() => {
stream.send('hello world')
}, 500)
setTimeout(() => {
clearInterval(interval)
stream.close()
}, 3000)
return stream
})基于回调和基于值的流的工作方式相同，但语法有所不同，以根据你的偏好设置。
¥Both callback-based and value-based streams work in the same way but with different syntax for your preference.