SQLAlchemy 2 update, and bump to 2.0.0

[RudolfCardinal]

No description provided.

[RudolfCardinal]

Also adding some support for Databricks. There is no generic way to convert SQL textual column types to dialect-specific SQLAlchemy column types (sqlalchemy/sqlalchemy#12230). This relates to cardinal_pythonlib.sqlalchemy.schema.get_sqla_coltype_from_dialect_str() and its sub-function, _get_sqla_coltype_class_from_str().

[martinburchell]

Looks good to me. I've added some pre-commit hooks for blanket and unnecessary #noqas.

There's a pattern we use in the CamCOPS and CRATE tests for database level testing with engines, sessions etc created in conftest.py and injected into a common database test class. We might want to adopt this if these tests start getting slow.

Would it be worth asking the Databricks people about dialect.ischema_names ?

[RudolfCardinal]

Have asked at databricks/databricks-sqlalchemy#11