fix: db connection string lifecycle issues

[matthew-hagemann]

Fixes #37, #42

The underlying issue is that at times, the charm loses state and reverts to original configuration values. When this happens, we were losing the connection string, as the _waiting_for_database_relation flag and the _env_vars we pass into the container were being reset.

This change now has __init__ check for existing db relations, and if one is found, skip setting the _waiting_for_database_relation flag and go straight to setting up and passing in the connection string.

I also added steps to the lifecycle tests that cold boot the charm and swap out ports while connected to the db to ensure we return to a running state and can reach the container.

In the future I want to rework this to trigger on supplying a db_name as config, but until we need to support metrics, I think we can kick the can down the road.

[codecov]

Codecov Report

Attention: Patch coverage is 50.00000% with 2 lines in your changes missing coverage. Please review.

Project coverage is 81.65%. Comparing base (8e02275) to head (d7ac42a).
Report is 36 commits behind head on main.

Files with missing lines	Patch %	Lines
src/charm.py	50.00%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #45      +/-   ##
==========================================
- Coverage   84.47%   81.65%   -2.82%     
==========================================
  Files           2        2              
  Lines         322      338      +16     
  Branches       25       26       +1     
==========================================
+ Hits          272      276       +4     
- Misses         43       53      +10     
- Partials        7        9       +2     

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