feat: stopping the crawlers gracefully with BasicCrawler.stop()

[barjin]

Allows users to call crawler.stop() to gracefully stop the crawler.

Currently, the crawlers are "stateless", i.e. calling:

async requestHandler({ crawler, enqueueLinks }) {
   await enqueueLinks();
   crawler.stop();
}

...

await crawler.run(['example.com']);
await crawler.run();

Will only crawl example.com once, then stop and purge the RQ / dataset, so the second crawler.run() call will yield no results.

I suppose this is expected, but we could easily add a crawler.pause() method, which would keep the inner state for resuming with crawler.run().

Closes #2777