Handling of datasets and samples

[felixzinn]

Hi,

the handling of datasets and samples is currently handled in multiple python files (sample.py, xsection.py, plot_utils.py) which is a bit confusing IMO.

What do you think about introducing a dataclass for each dataset that holds all the necessary information. The workflow then only needs to specify the name of the dataclass and all fetching, merging and plotting is handled correctly in one place instead of multiple files?

@dataclass
class Dataset:
    name: ...
    sample: ....
    xsection: ...
    label: ...

something like this