Track licenses and collect metrics about them.

Created by: adamreichold

This is a hand-written alternative to #19 as I am not yet convinced that automatically importing the controlled vocabulary is too helpful as it only applies to what is formally open data. (For example, DORIS has a custom version of CC-BY-NC-SA-3.0 available at which we would need to link even though the license will likely never be part of the DCAT-AP.de vocabulary.)

Just this tiny bit of code already handles most datasets where these is a dataset-level license defined:

Of course, it also looks like most datasets only have resource-level license information which is the main problem here. I think a heuristic that infers the license by looking if all resources have the same license attached might be helpful as a next step (and of course, extending the other harvesters to gather license information).

Another useful next step would be to make the licenses searchable, either by their ID or by their name or properly using faceted search.

In any case, the metrics already provide a useful goal to work on improving thing IMHO.