I'm trying to use databricks-sql-python in a statically-typed python project, and I'm running into some issues. I'm usingmypy --strict.

1. Despite this project having type annotations, my typechecker (mypy) will not check them (this causes problems for the type implications of the rest of my code). I'm pretty sure that, as perhttps://peps.python.org/pep-0561/#packaging-type-information, you can just add a py.typed file to the sql/ and sqlalchemy/ folders. I don't use the sqlalchemy side so I haven't tested that, but the sql one seems to work perfectly when I modify the copy of databricks-sql-python on my machine.

2. I geterror: Call to untyped function "connect" in typed context [no-untyped-call]. I'm pretty sure you can just change the type signature ofdef connect in the sql folder's__init.py__ todef connect(server_hostname, http_path, access_token=None, **kwargs) -> "Connection". This works perfectly when I test it on my machine.

3. I geterror: Returning Any from function declared to return "List[Row]" [no-any-return] when I have the line return cursor.execute(query, sql_params_dict).fetchall(). Honestly, I have no idea how to fix this.fetchall looks like it has the right type signature to me! Maybe this is a bug in mypy somehow? Or maybe this is a symptom of the previous issue not getting completely fixed...

I can make a pull request about this, at least the first two, but maybe these ideas are bad and you haven't done them yet because it would break something else; let me know!