ASCII-only for PoseidonID and GroupNames

[stschiff]

I just tried to change both the Janno column types for PoseidonID and for GroupNames to ByteString.Char8 within the new VCF-writing PR, but got weird compiler errors around makeInstances. I don't know template Haskell, so can't continue on this. I will for now leave these types as String and Text, respectively, and would need some help to get this done.

On the sequence-formats side, we now use exclusively ByteString.Char8 for these ASCII-restricted fields.

[stschiff]

I would also like to add that I have - for now - decided to not implement checks on the level of the sequence-formats library. I looked into it and I noticed that there are many many data formats in that library where I'm parsing into ByteStrings, so no Unicode support. Setting up checks for all of these is not possible, and feels a bit arbitrary now to implement checks at that level only for individual and group names. It feels then that sequence-formats is not the right place for this.

sequence-formats is a low-level library to parse these Bioinformatic formats and constraints for the Encoding are not really a question of low-level parsing, but of client code. So I would like to implement Unicode-checks at the level of poseidon-hs, not of sequence-formats.