return kwargs from adapter.to_file funtions

[savente93]

Issue addressed

Fixes #182

Explanation

The description wasn't super clear about how this was supposed to be used but it was super quick to do, so I just took the liberty. For now the returned source_kwargs in the export function is unused, but hopefully, those (i.e. @DirkEilander) can use it now that it is available. This PR doesn't actually add any functionality, therefore I don't think tests or doc updates are necessary.

Checklist

• Updated tests or added new tests
• Branch is up to date with main
• Tests & pre-commit hooks pass
• Updated documentation if needed
• Updated changelog.rst if needed

Additional Notes (optional)

Add any additional notes or information that may be helpful.

[DirkEilander]

This issue indeed wasn't clear at all. What I think we want to return in the to_file methods are any keyword arguments required to read back the data into an identical object. To read self-describing formats like netcdf and geotiff to a xarray.Dataset this is not necessary, but to properly read a csv back to a pandas.DataFrame you need to know the index column. I guess the required "return kwargs" can be specified per driver, but would need to try to be sure.

[savente93]

For the CSV I've added the index_col to the kwargs that is returned, so I think it should be all good now. If we can think of any attributes needed later we can always add them

[DirkEilander]

I added a small comment, let me know if you have questions.

[savente93]

@DirkEilander Perhaps a tangential question, but I noticed that the export_data in the test_export_global_datasets raises the following warning: UserWarning: using the driver setting is deprecated. Please use vector_table instead. I'm wondering if this is something we can solve here now? We would check if the export we're doing is a csv based one but that feel really inelegant. What do you think?

[savente93]

not to @savente93 :

1. leave warnings, that's other PR
2. update to reader_kwargs in other adapters
3. add index_col = 0 to reader_kwargs in csv in dataframe and geodataframe

[savente93]

@DirkEilander implemented what we discussed. Should be good to go now.

[DirkEilander]

Made a small change to actually use the returned driver_kwargs.
Otherwise good to be merged!