Handle rate limits

[bocops]

Mastodon instances have rate limits defaulting to 300 requests per 5 minutes both per account and per IP: documentation

The currently closest limit to be hit by a client is returned in the response header of each request. In a recent test to retrieve a list and iterate over all its user accounts, these look like this:

x-ratelimit-limit : 300
x-ratelimit-remaining : 299
x-ratelimit-reset : 2024-11-14T15:45:00.509266Z

x-ratelimit-limit : 300
x-ratelimit-remaining : 298
x-ratelimit-reset : 2024-11-14T15:45:00.616648Z

x-ratelimit-limit : 300
x-ratelimit-remaining : 270
x-ratelimit-reset : 2024-11-14T15:45:00.785389Z

As a library user, it would be nice to be aware of approaching limits somehow, to be able to throttle before a limit is reached and subsequent requests fail. Questions:

1. What exactly should be returned to the llibrary user? We can probably return limit and remaining (allowed requests) as-is, but should probably convert the reset timestamp into something more sensible, like "time remaining until reset". Note that, if we return the remaining time in seconds, a simple comparison between "requests remaining" and "seconds remaining" would suffice to see if our client is currently behaving well, because 300/5 requests/min boils down to an allowed average of 1 request/second.
2. How can we return this information? In my test, I just printed header information that is available in MastodonRequest.execute(), but that already returns the requested data.