`LUR` cache strategy can relieve the pressure of IO, we can apply this feature to methods or attributes. refet to : - [lru_cache](https://docs.python.org/3/library/functools.html#functools.lru_cache) - [lru cache in wechaty](https://github.com/wechaty/wechaty-puppet/blob/master/src/puppet.ts#L175)