extract periods from date range, add new models

[charlienegri]

note on edge cases where start/end dates are the end/beginning of a year:

with make_period(datetime(2018,12,31),datetime(2022,1,1)) this gives
['20181231-20220101', '20181231-20181231', '20190101-20191231', '20200101-20201231', '20210101-20211231', '20220101-20220101']
I assume this is the intended behavior, but if not we can change the function

[avaldebe]

@charlienegri the lint, docs and venv jobs are crashing because Cartopy packages are incompatible with the new dependency resolver used by pip. See #686 for details.

After that PR is merged, merging main-dev into this branch should fix some of the CI jobs.
However, I do not know why the conda jobs are failing.

[charlienegri]

I have no clue, a lot of tests are failing but I don't think this branch ever passed the pipeline

[avaldebe]

I have no clue, a lot of tests are failing but I don't think this branch ever passed the pipeline

mmmm, maybe we should have a bot that closes PRs if they fail all tests, delete the branch and kick the PR author from the project. Or maybe it is time to "go home" and have diner... 💩

[charlienegri]

the tests are failing because of the different way of handling the dates as datetime objects and because of the new definition of periods, I can look into this more and try to fix them

[dulte]

Looks good. will fix failing tests in the other branch