Polars Generic Dataset

[wmoreiraa]

Signed-off-by: wmoreiraa walber3@gmail.com

Description

Add polars.GenericDataSet as discussed in #95
Ive probably closed the old PR doing wrong git things #116

Development notes

Similar to pandas.GenericDataSet, changes made:

1. Add entry "write_mode" to be able to use formats that polars doesnt provide a write method (e.g.: delta, excel)
2. Write file permission changed to "wb" from "w".
3. Added more tests cases for different file formats.

Checklist

• Updated the documentation to reflect the code changes
• Added a description of this change in the relevant RELEASE.md file
• Added tests to cover my changes

[wmoreiraa]

Notes: the comments/review from @merelcht on #116 are on point and were not addressed (those not related to spark changes that are not on this PR anymore).

"Maybe I'm misunderstanding this, but if you're using a read-only data format, you simply shouldn't call save right? It kind of sounds that if you change the write mode to ignore you wouldn't get an error, but you would get raise DataSetError(f"Write mode '{self._write_mode}' is read-only.") as set in line 219." . I think this one could go two different ways:

1. Removing the read_only formats and then removing the write_mode option : Easiest way, it would lose the functionality but I do think its very minor and it would allow more general usage and further improvement on this by community.
2. Refactoring and fixing this implementation mistake that Ive made.

@datajoely @astrojuanlu , I might not have any time soon to do this. Think you could finish this fix?

@merelcht all other review/comments are addressed if removing the write_mode, this is why I dont specifically copied here. Also, thank you for the review.

[astrojuanlu]

Thanks a lot for getting this far @wmoreiraa. We'll take care of the final push 💪🏽

[astrojuanlu]

Closing in favor of gh-170.