Add bucketing for file_format parquet

### Describe the feature
Add the possibility to specify bucketing with file_format=parquet to improve performance on reading.

### Describe alternatives you've considered
Nothing that come to mind, bucketing is different than partitioning.

### Additional context
Bucketing can speed up queries when working with high cardinality fields, it's a must have.

### Who will this benefit?
All the use cases where specific WHERE conditions are used on the bucketed columns (e.g. id/user_id), in general bucketing is recommended for high cardinality columns.

### Are you interested in contributing this feature?
Yup.

No files, tests, or entry points are named. Start by locating the adapter's handling of file_format=parquet and compare it with the existing partitioning behavior; done should mean bucketing can be specified for Parquet and improves reads for filters on bucketed columns.

Written by the indexing model from the issue text.