Discrepancy in database's run_sql() method

[utkarsharma2]

Please describe the feature you'd like to see
Currently, the handler param in the run_sql() method is only used in data bricks all other databases are not using it, which leads to different response types since the handler is processed within run_sql() for data bricks and not for other databases. Also, the signature of run_sql() in data bricks deviates from the base. We need to standardize and when we do, we can remove the below check as well.

Describe the solution you'd like
Ideally, we should have the same signatures for all the run_sql() database implementations. This will prevent adding explicit conditions for data bricks in other parts of the code that should ideally be database agnostic.
example:

astro-sdk/python-sdk/src/astro/sql/operators/raw_sql.py

Line 47 in 2f8c8d0

if self.handler and self.database_impl.sql_type == "delta":

Acceptance Criteria

• Where ever we have used run_sql(), we shouldn't be having any database-specific checks.

[sunank200]

This PR is still a work in progress. This requires a return statement to return the cursor object to the generated result so that its consistent across all the databases. @rajaths010494 mentioned he would pick this up after other tickets are done for 1.5.1. He needs to sync with @dimberman for this as well @rajaths010494 to provide more details

[sunank200]

@rajaths010494 has set up the meeting today. He will provide more details on this ticket.

[rajaths010494]

Connected with @dimberman regarding run_sql we would be using get_wrapped_handler to get results for other databases and make necessary other changes for tests