Addressing a few issues with #run_ping! (Fixes #231)

[RodneyU215]

Summary

This PR addresses several issues with the initial implementation (#226) of Slack::RealTime::Client#run_ping!; a method which is responsible for ensuring the Slack::RealTime::Client stays connected until Slack::RealTime::Client#stop! is called.

• Slack::RealTime::Client#close will no longer assign nil to @socket because the reactor would still be running in the background if websocket_ping is set.
• Slack::RealTime::Concurrency::Async::Socket#disconnect! now cancels all timers (via @reactor.cancel). This shuts down the websocket_ping timer.
• I've removed the loop in Slack::RealTime::Client#run_ping! and now use on the builtin function Async::Reactor#every to run the ping on an interval.
• I've added Slack::RealTime::Client#restart_sync which is responsible for obtaining a new websocket url and attempting to connect to it.
• If the connection is closed and we've not called Slack::RealTime::Client#stop!; on the next Slack::RealTime::Client#run_ping! we'll attempt to restart the connection (via Slack::RealTime::Client#restart_sync)
• If we're still not connected, after attempting to restart the connection (via Slack::RealTime::Client#restart_async), we reraise Slack::RealTime::Client::ClientNotStartedError
• I've also added a new integration test to ensure it works as expected.

Requirements

• I've read and understood the Contributing Guidelines and have done my best effort to follow them.

[RodneyU215]

#231

[dblock]

Looks good.