2022 update: Use new data from 2019+2020

[rugk]

Updating that data here was overdue.. so here we go.. fresh new data from 2019+2020! 🎉

Reasoning

• For my future self: Yes, again for SC we use area, which kinda makes sense, because that is what is mapped.
• I (again?) used the average of the two latest available years of the FAO data, i.e. 2019 and 2020. Given that 2020 was the first year with Covid, and I guess (without having looked) that this could have an influence of the production of crops etc., I guess this is a good idea.

Changes

• The FAO now combines both crops and lifestock (C + L) in the same dataset, so one needs to take care of that when exporting the data from them.
  • other changes from them are not noteworthy for us as far as I read that
• I switched to FAOs internal/own database mapping for countries and only modified the DB (finally now documented... uff, how I hate maintaining legacy software :upside_down_face:) and made that somewhat creepy grep thing for converting the country codes more robust
• I sanity-checked some results...
• Added license stuff and such things.

Sanity-check

Here a sanity check for Germany: The current result is: https://github.com/rugk/crops-parser/blob/bd04f16773fd5c93a4964476a102c2cd400d31c9/result/OsmOnly/mostAreaHarvest_2019%2B2020.yml#L49

And the tonnes metric yields a very similar result: https://github.com/rugk/crops-parser/blob/bd04f16773fd5c93a4964476a102c2cd400d31c9/result/OsmOnly/mostTonnesHarvest_2019%2B2020.yml#L49

(E.g. strawberry and tomato yield more tonnes aka "production quantity" per ha than hop or so, which also sounds very reasonable.)

2017 there was a quite similar result: https://github.com/rugk/crops-parser/blob/bd04f16773fd5c93a4964476a102c2cd400d31c9/result/old/OsmOnly/mostAreaHarvest_2017.yml#L53

https://github.com/rugk/crops-parser/blob/bd04f16773fd5c93a4964476a102c2cd400d31c9/result/old/OsmOnly/mostTonnesHarvest_2017.yml#L53

[rugk]

/cc @westnordost