adding new keys to theory template

[giacomomagni]

Adding two missing keys to the theory template: FONLLparts and n3lo_cf_variation

[alecandido]

Even fine, but isn't this definitely breaking the database?

If you don't use the database (and benchmarking infrastructure), there should be no need to use banana.
I don't remember if we're good enough to ignore extra keys. But I wouldn't expect.

[giacomomagni]

I'm adding them to make tests in this PR passing:

https://github.com/NNPDF/yadism/actions/runs/6578472057/job/17872298906?pr=215

I believe this to be needed because I have explicitly modified:

https://github.com/NNPDF/yadism/blob/21f09fa81e33e2762315d21394969dcf262a98ce/src/yadism/input/domains.yaml

If my reasoning is correct, the other option would be to drop the changes in domains.yaml as is currently done in
NNPDF/yadism#195

If you agree, just tell me which option you prefer. (or maybe if you have better one...)

[felixhekhorn]

I'm adding them to make tests in this PR passing:

If you don't use the database (and benchmarking infrastructure), there should be no need to use banana.

just as I need 0.6.9 to make PR xyz pass, so should be fine

Even fine, but isn't this definitely breaking the database?

it will break - just as many of the v0.6.x tags did (if we we serious about this we should be now at at least v0.12)

I don't remember if we're good enough to ignore extra keys. But I wouldn't expect.

we were not - and the problem is we need it (instead of being too much)

[felixhekhorn]

please fix pre-commit and then this should be fine

[alecandido]

If my reasoning is correct, the other option would be to drop the changes in domains.yaml as is currently done in
NNPDF/yadism#195

At the moment I've been quite busy with other commitments, but I will soon get rid of domains.yaml (as promised long ago).

[RoyStegeman]

Just to note that FONLLParts is an optional argument