Tests for aggregate and indicators

[x] This PR addresses an already opened issue (for bug fixes / features)
- This PR fixes #xyz
[x] (If applicable) Documentation has been added / updated (for bug fixes / features).
[x] (If applicable) Tests have been added.
[x] This PR does not seem to break the templates.
[x] HISTORY.rst has been updated (with summary of main changes).
- [x] Link to issue (:issue:number) and pull request (:pull:number) has been added.

Tests for multiple functions in aggregate.py and indicators.py.
Minor bugfixes when detected by the tests (yay!).
Changed a couple PosixPath to just Path
Fixed a bug in unstack_dates where it failed for anything other than seasons.
Fixed a bug in unstack_dates related to xr.coding.times.contains_cftime_datetimes in the newest xarray: https://github.com/pydata/xarray/issues/7966
The logic to manage indicators returning multiple outputs (such as rain_season), was changed a little and was completely broken anyway (oops). We now keep all variables produced and move from a DataArray (or a tuple of DataArray) to a Dataset much earlier, which makes everything else simpler to manage.

I got rid of out_dict[key] = xr.merge(o for o in out if o.name in indicators) in compute_indicators, in favor of returning all variables. Previously, xs.compute_indicators(ds, indicators=[("rs", xclim.atmos.rain_season)]) would return an empty Dataset, because rs doesn't correspond to any of the outputs. Similarly, xs.compute_indicators(ds, indicators=[("rain_season_start", xclim.atmos.rain_season)]) would only return 1 out of 3 outputs produced by xclim. This is more a bugfix than a breaking change, but might still break some workflows.

I did not add tests to aggregate.spatial_mean, since changes are coming soon to xESMF, including some bugfixes that will affect that function.

Ouranosinc / xscen