[Lake][ETL] Insert GQL data into PersistentStore

[idiom-bytes]

Motivation

Update gql_data_factory to integrate PersistentStore and push data to raw tables.

Persist SLA Only new records should be fetched & inserted into db.

1. get last timestamp from db.table
2. fetch subgraph from_last_timestamp to_now
3. insert subgraph new_records into db.table

DoD:

• [x] tables in duckdb are created w/ the same schema that we persist in files
• [x] update table schemas & system to use pydantic if required, rather than creating multiple schemas or mappings
• [x] gql_data_factory fetch/update functions insert new records into db raw tables
• [x] tests validate db has new records registered
• [x] etl.py is ready to be updated, as all raw subgraph records are being correctly inserted into the db

[kdetry]

Although Pydantic is a great tool, it's not always necessary since we already have Polars data schemas that validate the data structure. Using Pydantic now would result in double validation.

[idiom-bytes]

The whole etl pipeline is now implemented using CSV + DuckDB.

It has been verified and merged into PR #685, which is a large Epic that includes many other updates to improve Lake/ETL/etc... https://github.com/oceanprotocol/pdr-backend/issues/685