fix: node draining now properly completes at stop

[b3by]

Description

This PR addresses a unit test that occasionally fails, depending on the timing of the machine running the test. With the old implementation, any messages still in the queue or the buffer when the update thread exits would be dropped.

In order to implement this mechanism, a couple of missing primitives were added to the transport protocol, namely, wait() for the Signal, empty() and qsize() for the Queue. A timeout argument was also added to the put for the Queue class. Lastly, a JUTURNA_DRAIN_TIMEOUT` constant was introduced, to indicate how long a node should wait to drain its pending messages.

PR type

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Code refactoring
• Performance improvement

Key modifications and changes

• Slightly changed the stop() mechanism, so now the node truly waits to drain all pending messages
• New primitives added to the transport protocol
• New constant JUTUNRA_DRAIN_TIMEOUT added

Affected components

• The Node class
• The signatures for the transport Queue and Signal