时间戳

首页 > 解决方案 > Parallel asynchronous iteraor - is it possible?

javascript - Parallel asynchronous iteraor - is it possible?

问题描述

Right now I have the following code:

import axios from 'axios'

const urls = ['https://google.com', 'https://yahoo.com']

async function* requests() {
  for (const url of urls) {
    yield axios.get(url)
  }
}

;(async () => {
  for await (const n of requests()) {
    console.log(n.config.url) // prints https://google.com and then https://yahoo.com
  }
})()

As is, the requests won't block the single thread of node, but they will happen in sequence. I'm wondering if it would be possible to change the code to force parallelism.

标签： javascriptnode.jsasync-awaitasync-iterator

解决方案

“更简单”的 no-deps 方法是对它们进行批处理并使用Promise.all

import axios from 'axios'

const urls = [
  'https://jsonplaceholder.typicode.com/todos/1', 
  'https://jsonplaceholder.typicode.com/posts/1',
  'https://jsonplaceholder.typicode.com/users/1',
  'https://jsonplaceholder.typicode.com/comments/1'
]

async function* requests(batchSize = 1) {
  let batchedRequests = [];
  for (const url of urls) {
    batchedRequests.push(axios.get(url));
    if (batchedRequests.length === batchSize) {
      yield Promise.all(batchedRequests);
      batchedRequests = [];
    }
  }
  if (batchedRequests.length) { //if there are requests left in batch
    yield Promise.all(batchedRequests);
  }
}

;(async () => {
  for await (const batch of requests(2)) {
    batch.forEach(n => console.log(n.config.url)) // prints https://google.com and then https://yahoo.com
  }
})()

您可以使用rxjsobservables 在灵活性方面获得类似的结果，但它是另一个库，如果您不熟悉反应流，它可能会更复杂。这是我在该主题上找到的详细帖子：https ://medium.com/@ravishivt/batch-processing-with-rxjs-6408b0761f39

推荐阅读