流式传输 - Mock Service Worker 中文网

提供 ReadableStream 实例作为模拟响应主体以启动数据流。

¥Provide a ReadableStream instance as the mocked response body to initate a stream of data.

基本示例

¥Basic example

在这个基本示例中，我们将以三个块的形式将字符串 BrandNewWorld 流式传输到客户端。

¥In this basic example we will stream a string BrandNewWorld to the client in three chunks.

import { http, HttpResponse } from 'msw'
 
const encoder = new TextEncoder()
 
export const handlers = [
  http.get('/video', () => {
    const stream = new ReadableStream({
      start(controller) {
        // Encode the string chunks using "TextEncoder".
        controller.enqueue(encoder.encode('Brand'))
        controller.enqueue(encoder.encode('New'))
        controller.enqueue(encoder.encode('World'))
        controller.close()
      },
    })
 
    // Send the mocked response immediately.
    return new HttpResponse(stream, {
      headers: {
        'Content-Type': 'text/plain',
      },
    })
  }),
]

请注意，客户端还必须将响应作为流读取才能查看传入的块。使用 response.text() 或 response.json() 等方法读取响应将等待流完成，然后再返回响应主体。

¥Note that the client must also read the response as a stream to see the incoming chunks. Reading the response with methods like response.text() or response.json() will await the stream completion before returning the response body.

转换流示例

¥Transform stream example

使用 TransformStream API，我们可以描述更复杂的流式传输场景。

¥Using the TransformStream API, we can describe more complex streaming scenarios.

在下面的例子中，我们将描述一个视频端点，它从原始视频资源建立可读流，并通过自定义转换流将其传输，在每个视频流块之间插入延迟。当我们希望测试我们的 UI 如何在流期间处理延迟时，这很方便。

¥In the example below, we will describe a video endpoint that establishes a readable stream from the original video resource and pipes it through a custom transform stream that inserts latency between each video stream chunk. This is handy when we wish to test how our UI handles lantecy during streams.

import { http, HttpResponse, delay } from 'msw'
 
export const handlers = [
  http.get('/video', async () => {
    // Request the original video stream.
    const videoResponse = await fetch(
      'https://nickdesaulniers.github.io/netfix/demo/frag_bunny.mp4'
    )
    const videoStream = videoResponse.body
 
    // Implement a custom transform stream that
    // takes any stream and inserts a random latency
    // between its chunks.
    const latencyStream = new TransformStream({
      start() {},
      async transform(chunk, controller) {
        await delay()
        controller.enqueue(chunk)
      },
    })
 
    return new HttpResponse(
      // Respond with the original video stream
      // piped through the latency transform stream.
      videoStream.pipeThrough(latencyStream),
      // Inherit the rest of the original video
      // response data, like "headers".
      videoResponse
    )
  }),
]