Add support for BodoSQL context in PyDough

[knassre-bodo]

Goals of this issue:

• Add the ability to create a PyDough database context that is attatched to a BodoSQL context (either local, Snowflake, or Iceberg catalog). E.g. pydough.active_session.load_database_context("BodoSQL", context=bc)
  • This may require creating a DB API 2.0 compliant wrapper around the BodoSQL context, which can simply raise an exception for everything except for cursor which fails for eveyrthing except execute and fetchall for the time being.
  • Alternatively, a refactor can be done to the DatabaseConnection API to work around this by allowing either a DB API 2.0 connection or the BodoSQL context, and all methods of the context switch on whether it is the connection or the BodoSQL context.
• Add any necessary changes to SQLGlot transformation to work with a BodoSQL context (the Snowflake dialect should work?)
• Add the ability to use the Pydough BodoSQL context to execute a query in BodoSQL and return the answer as a pandas DataFrame.
• Add any necessary documentation
• Add e2e testing of BodoSQL with local catalog & filesystem-based Iceberg catalog (will need PyDough metadata).
• Add non-CI testing of BodoSQL with a Snowflake catalog & non-filesystem-based Iceberg catalog (requires using secrets to configure the catalogs). If it is possible to have a CI that accesses these secrets without exposing them to users, create a special workflow that runs these tests that triggers nightly or with special permissions. Otherwise, mark these tests so they are always skipped unless the secrets are present, which can only be done in local development.