Justin Bich

Now the unit-tests are added. A possible next step would be to improve the comparison of dictionaries in the unit-tests.

Now the comparisons in the unit-tests (especially of dictionaries) should have been improved.

A next step should be to enable support for different Hamiltonians in [hsdinput.py](https://github.com/jubich/dftbplus/blob/525c84d11f78f53e7d38fbaf8a8991a679f6c80d/tools/python/ptools/src/dftbplus_ptools/hsdinput.py). This generalizes this file and will be needed anyway in view of the upcoming [new ase calculator](https://gitlab.com/ase/ase/-/merge_requests/2682#note_943586440)...

There will be another commit in the next few days to implement the caseinsenitivity of the hsd dict's. Then I can resolve the conflicts and remove the draft status.

This should reduce the computation time for specific calculations where only the forces in certain directions are important. Also, it can be used for development. The point with the specific...

I have investigated the problem now and I think I found the bug. But when I created the PR in the ase repository, I saw that [the problem was fixed...

Sure, but before that I would like to add the 'usage-message' in convolve.py as suggested by @vanderhe. But I'm not sure how to formulate a good one. So I'm waiting...

The idea behind `change_header_year` is to change the year in any file whose header matches the template, while the second file `set_new_source_headers` is used to create a header in imported...

Both files are now formatted with _black_. I have also added `encoding="utf-8"` to `with open()`. Now `change_header_year` has a pylint score of 9.52 and `set_new_source_headers` has a score of 9.30.

Should I squash these commits now or is there any further feedback?