Issues when using databricks-sql-connector==4.6.0

[brain246]

Hello there,

I am using databricks-sqlalchemy (2.0.10) in combination with alembic (1.8.4) in a CI/CD pipeline to automate schema migrations.

When our pipeline suddenly stopped working, i found out after some debugging, that alembic fails with an error message (see below) when using databricks-sql-connector==4.6.0. After pinning the sql connector to databricks-sql-connector==4.2.6 (the version it works with from my local machine), the migrations ran through again.

The corresponding part of my CI/CD pipeline log:

  File "/agent/_work/1/s/.venv/lib/python3.14/site-packages/databricks/sqlalchemy/base.py", line 382, in has_table
    self._describe_table_extended(
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
        connection=connection,
        ^^^^^^^^^^^^^^^^^^^^^^
    ...<2 lines>...
        schema_name=schema,
        ^^^^^^^^^^^^^^^^^^^
    )
    ^
  File "/agent/_work/1/s/.venv/lib/python3.14/site-packages/databricks/sqlalchemy/base.py", line 208, in _describe_table_extended
    fmt_result = _describe_table_extended_result_to_dict_list(result)
  File "/agent/_work/1/s/.venv/lib/python3.14/site-packages/databricks/sqlalchemy/_parse.py", line 40, in _describe_table_extended_result_to_dict_list
    for row in result.all():
               ~~~~~~~~~~^^
  File "/agent/_work/1/s/.venv/lib/python3.14/site-packages/sqlalchemy/engine/result.py", line 987, in all
    return self._allrows()
           ~~~~~~~~~~~~~^^
  File "lib/sqlalchemy/engine/_result_cy.py", line 366, in sqlalchemy.engine._result_cy.BaseResultInternal._allrows
  File "lib/sqlalchemy/engine/_result_cy.py", line 201, in sqlalchemy.engine._result_cy.BaseResultInternal._row_getter.many_rows_simple
  File "lib/sqlalchemy/engine/_row_cy.py", line 72, in sqlalchemy.engine._row_cy.BaseRow.__init__
TypeError: Expected tuple, got Row

Cheers,
Thomas