Are there any arguments against adding a post_run callback to Test / TestBench?
Similar to post_check, but called regardless of whether test passed or failed.

My use case: Processing log files written to output_path, even for failing tests. For example, archiving log files from automated simulation runs.
I know I could do this without a callback by parsing the mapping text file to get to the output directories, but the callback would be a much easier solution.
Or is there another easier way that I'm not seeing?

And I believe there could be other use cases where this callback would be beneficial.

If there is general consensus that this suggestion is useful, I could work on implementing it.