Create a nnpdf data cli

[scarlehoff]

This is just a collection of scripts / ideas that I think they might be useful. I've been writing some scripts already while preparing the talk for DIS so might as well make it into a cli for the data package.

• Given a dataset name, return the hepdata / arxiv entry
• Given an arxiv/hepdata entry, return matching datasets.
• Given a runcard / yaml file with dataset_inputs, return a list of all matching hep data / arxiv entry. Perhaps even prepare a nice latex format table and, if inspire/arxiv info is available, through the API, the bib file with the corresponding references.
• Some filtering (e.g., all datasets that match CMS_1JET_13TEV and have a theory)
• Provide a way to read and/or search in all theories