Add postgres DB handler

[Veritogen]

Describe the solution you'd like
We'd like to keep track of different pipeline results in a dedicated database.

Describe alternatives you've considered

• Implementing a server to collect the pipeline results seemed like a bigger overhead as it also includes more infrastructure to maintain compared to just pushing the results into a database.
• Moving the sqlite database into the CI environment would be another option. This may lead to race conditions, if we have multiple pipelines running in parallel.

Additional context
I got a working prototype (see #78) but would like to discuss the following points:

• How do we want to test the postgres implementation?
  • Proposal: add postgres image to pipeline to be able to run the tests against a database in docker
• What version of psychopg (2, 3 or both) do we want to support?
  • I'd propose to support both to be future proof
  • How do we add this to the requirements?
• Does the implementation proposed match the architecture you envisioned?
• How to ensure people don't need to install uneccessary dependencies? Maybe we could enable people installing pytest-monitor like this pip install pytest-monitor[postgres]

[lhpt2]

For anyone interested in using this feature: As long as it hasn't been merged upstream yet, there is a soft fork available here: https://github.com/einhundert/pytest-monitor/ (We are only providing support for features and changes made by us, don't request anything not related to these)