详解JavaScript 中的批处理和缓存

import { wait } from '../async/wait'

/
  将多个并发异步调用合并为一次批处理
  @param handle 批处理的函数
  @param ms 等待的时长（时间越长则可能合并的调用越多，否则将使用微任务只合并一次同步执行的所有调用）
 /
export function batch<P extends any[], R extends any>(
 handle: (list: P[]) => Promise<Map<P, R | Error>>,
 ms: number = 0,
): (...args: P) => Promise<R> {
 //参数 => 结果 映射
 const resultCache = new Map<string, R | Error>()
 //参数 => 次数的映射
 const paramCache = new Map<string, number>()
 //当前是否被锁定
 let lock = false
 return async function (...args: P) {
  const key = JSON.stringify(args)
  paramCache.set(key, (paramCache.get(key) || 0) + 1)
  await Promise.all([wait(() => resultCache.has(key) || !lock), wait(ms)])
  if (!resultCache.has(key)) {
   try {
    lock = true
    Array.from(
     await handle(Array.from(paramCache.keys()).map((v) => JSON.parse(v))),
    ).forEach(([k, v]) => {
     resultCache.set(JSON.stringify(k), v)
    })
   } finally {
    lock = false
   }
  }
  const value = resultCache.get(key)!
  paramCache.set(key, paramCache.get(key)! - 1)
  if ((paramCache.get(key) || 0) <= 0) {
   paramCache.delete(key)
   resultCache.delete(key)
  }
  if (value instanceof Error) {
   resultCache.delete(key)
   throw value
  }
  return value as R
 }
}

const getById = onceOfSameParam(
 batch<[number], Dict>(async (idList) => {
  if (idList.length === 0) {
   return new Map()
  }
  // 一次批量处理多个 id
  const list = await this.getByIdList(uniqueBy(idList.flat()))
  return arrayToMap(
   list,
   (dict) => [dict.id],
   (dict) => dict,
  )
 }, 100),
)