[Lake] DuckDB - PersistentDataStore

[idiom-bytes]

Motivation

All ETL compute will be handled on top of persistent data files duck.db. A lot of the logic to operate here, will be similar to PartitionedDataStore.

PersistentDataStore deliver the requirements we need for [Process 1 + Process 2] of the ETL Pipeline. We'll then create a similar middleware to abstract common logic, such that we can update our ETL pipelines to use this.

DoD:

• [x] Implement a base class for DuckDB
• [x] Implement a child class called as "PersistentDataStore" for wrapping duckdb persistent databases
• [x] Implement core methods: create_view, insert_to_table, query_data
• [x] Create test coverage

All middleware commands have been implemented and tested

• [x] _create_and_fill_table
• [x] insert_to_table(self, df: pl.DataFrame, dataset_identifier: str)
• [x] query_data
• [x] fill_from_the_csv_destination
• [x] update_data
• [x] drop_table

[idiom-bytes]

Code merged into PR #734 DuckDB E2E Integration