Replace necessity argument with **kwargs

[JohannesWiesner]

tcy currently requires the user to have a "necessity" column in the .tsv file. This was introduced primarily because the github runners did not have enough RAM in the past to solve .yml files with too many packages. However, RAM for the runners was increased in the last year. Pitching for adapting this argument, so that tcy can accept any other key-value(s) pair where the key represents a column and the value(s) represent rows that should be filtered for. This gives more flexibility to the users.

**kwargs could also be named **filter_kwargs or something

[JohannesWiesner]

Maybe this helps:

https://stackoverflow.com/questions/46106746/how-to-implement-args-and-kwargs-in-argparse-python

[JohannesWiesner]

Should be possible to do something like this

--column_name value1 value2

or

--column_name value1

The first one would mean that for this column we only want rows that have any of these values and the second one would mean that we only want to include rows where this column only has this value

[JohannesWiesner]

Update: See #86 . Actually it makes sense to keep an --include flag. But it also makes sense to add **filter_kwargs