Vielzeug

Appearance

parallel ​

Problem ​

You need to run async tasks over an array concurrently but want to cap the number of in-flight tasks to avoid overwhelming a server or rate-limit.

Solution ​

Use parallel(array, fn, { limit }) to process items with bounded concurrency. Results are returned in input order.

ts
import { parallel } from '@vielzeug/arsenal';

const ids = [1, 2, 3, 4, 5, 6, 7, 8];

const results = await parallel(
  ids,
  async (id) => fetch(`/api/items/${id}`).then((r) => r.json()),
  { limit: 3 }, // at most 3 in-flight at once
);
// results[0] corresponds to ids[0], etc.

Without a limit ​

ts
import { parallel } from '@vielzeug/arsenal';

// All tasks start simultaneously — equivalent to Promise.all with a map
const all = await parallel([1, 2, 3], async (n) => n * 2);
// [2, 4, 6]

Pitfalls ​

  • Without limit, all items start concurrently — set limit explicitly for large arrays.
  • If any task throws, the whole call rejects. Wrap individual tasks in attempt if you need partial failures.
  • Results are in input order, not completion order.