mqtt: resubscribe inbox topic on reconnect

[maurerle]

unify call to _send_external_message by making it async as the other containers

catch RuntimeError when send_external_message fails - this does not resubmit the message to the broker yet

related to #97

This does not yet fix all issues when the broker goes down, as messages are lost and not sent again.
Would be good to handle this in the agent framework and not in the simulation itself.

[codecov]

Codecov Report

Attention: Patch coverage is 60.00000% with 6 lines in your changes missing coverage. Please review.

Project coverage is 88.99%. Comparing base (f14d832) to head (af8228f).
Report is 3 commits behind head on development.

Files with missing lines	Patch %	Lines
mango/container/mqtt.py	60.00%	6 Missing ⚠️

Additional details and impacted files

@@               Coverage Diff               @@
##           development     #169      +/-   ##
===============================================
- Coverage        89.15%   88.99%   -0.16%     
===============================================
  Files               22       22              
  Lines             2507     2516       +9     
===============================================
+ Hits              2235     2239       +4     
- Misses             272      277       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[maurerle]

Lost messages and not sending again could be handled in the AgentContext.send_message, that way would also be beneficial for tcp with packet loss.

For now, I'd like to add this change as is, as it fixes the access to an unitialized self.inbox at the start of a container