Store updated regression tests computed by runner

[rkunnema]

The regression tests are not kept up to date very well. Moreover, the timings can be very different when run on gitlab's CI runner versus some user's computer. See #505.

My proposal: when the runner fails due to a difference between
case-studies-regression
and
case-studies,
it puts case-studies somewhere (e.g., using this mechanism https://docs.github.com/en/actions/using-workflows/storing-workflow-data-as-artifacts) and outputs a command for the user to obtain this data and overwrite their local case-studies-regression with it. The user confirms that the difference is okay but committing this artefact to the same PR.

[rsasse]

That sounds like a good idea if the runner data can be stored and retrieved easily that way! Great.

[jdreier]

Sound great, will definitely look into that. We just need to make the process public somewhere so that people preparing a PR will know about it.

[rkunnema]

Ideally, whenever regressionTests.py returns an error code, the runner outputs:

The regression tests returned an error code. The analysis results are stored in

$PLACE

Please inspect the regression test's output carefully (CONTRIBUTING.md lists 
the criteria). If your modifications are justified, update and commit the directory 
case-studies-regression in this PRs branch with the following commands:

$COMMANDS