extract more structured data from ESAC registry

[maxheld83]

currently, a lot of the ESAC registry data appears to be somewhat unstructured.
To better make use of this data in #240 #243 et al, it would be good to clean this data and expose it as an R object in {hoad}.

In addition:

• we should get the data programmatically from ESAC scrape esac data #244.

• clean some fields (ie. yes, no should be logical, etc.).

• publisher is an open text field, and would have to be checked against some definitive list

• agreement_url is an open text field

• consortia/institution is an open text field

• access_costs should be ordinal

• worfklow_assessment are 3 separate vars!

• article_types can be parsed down to a few types (that need not be an open field)

• ...