Databricks connector missing catalog

[justin-gerolami]

Describe the bug
The Databricks connector attempts to list the columns using the information schema, however, a catalog is never specified in the configuration. This results in errors because the sql query doesn't find the information_schema.

In wren-engine-ibis/app/model/connector.py, the DatabricksConnector calls dbsql.connect() without a catalog parameter. And in wren-engine-ibis/app/model/metadata/databricks.py, the get_table_list() query references INFORMATION_SCHEMA.COLUMNS without a catalog prefix. In Databricks Unity Catalog environments with multiple catalogs, if no default catalog is set on the SQL warehouse, Spark cannot resolve INFORMATION_SCHEMA without a catalog qualifier.

The DatabricksTokenConnectionInfo model in app/model/__init__.py also has no catalog field, so there is no way for the user to specify one through the UI.

SELECT
  c.TABLE_CATALOG AS TABLE_CATALOG,
  c.TABLE_SCHEMA AS TABLE_SCHEMA,
  c.TABLE_NAME AS TABLE_NAME,
  c.COLUMN_NAME AS COLUMN_NAME,
  c.DATA_TYPE AS DATA_TYPE,
  c.IS_NULLABLE AS IS_NULLABLE,
  c.COMMENT AS COLUMN_COMMENT,
  t.COMMENT AS TABLE_COMMENT
FROM
  INFORMATION_SCHEMA.COLUMNS c
    JOIN INFORMATION_SCHEMA.TABLES t
      ON c.TABLE_SCHEMA = t.TABLE_SCHEMA
      AND c.TABLE_NAME = t.TABLE_NAME
      AND c.TABLE_CATALOG = t.TABLE_CATALOG
WHERE
  c.TABLE_SCHEMA NOT IN ('information_schema')

To Reproduce
Steps to reproduce the behavior:

1. Have a Databricks workspace with Unity Catalog enabled and multiple catalogs
2. Deploy WrenAI via Docker Compose
3. Open the Wren UI and configure a new Databricks data source with server hostname, HTTP path, and access token
4. Submit the connection — the ibis-server calls get_table_list() which runs the above query
5. Error: [TABLE_OR_VIEW_NOT_FOUND] The table or view INFORMATION_SCHEMA.COLUMNS cannot be found. SQLSTATE: 42P01

Expected behavior
The Databricks connection form should include a catalog field. The selected catalog should be:

1. Passed as the catalog parameter to dbsql.connect() in DatabricksConnector.__init__()
2. Or used to qualify the query as <catalog>.INFORMATION_SCHEMA.COLUMNS

This would align with how the Databricks SQL Connector for Python supports catalog selection: dbsql.connect(server_hostname=..., http_path=..., access_token=..., catalog="my_catalog")

Workaround
Manually patching connector.py inside the ibis-server container to add catalog="<name>" to the dbsql.connect() call resolves the issue.

Desktop (please complete the following information):

• OS: macOS (Darwin 25.0.0, Apple Silicon)
• Browser: Chrome

Wren AI Information

• Version: 0.29.1 (latest as of 2025-11-28)
• Wren Engine: 0.22.0
• Wren AI Service: 0.29.0
• Ibis Server: 0.22.0
• Wren UI: 0.32.2
• Bootstrap: 0.1.5

Relevant log output

ibis-server stack trace:

File "/app/app/model/metadata/databricks.py", line 61, in get_table_list
    response = self.connection.query(sql).to_pandas().to_dict(orient="records")
  File "/app/app/model/connector.py", line 627, in query
    cursor.execute(sql)
  ...
databricks.sql.exc.ServerOperationError: [TABLE_OR_VIEW_NOT_FOUND] The table or view `INFORMATION_SCHEMA`.`COLUMNS` cannot be found. Verify the spelling and correctness of the schema and catalog.
If you did not qualify the name with a schema, verify the current_schema() output, or qualify the name with the correct schema and catalog.
To tolerate the error on drop use DROP VIEW IF EXISTS or DROP TABLE IF EXISTS. SQLSTATE: 42P01; line 12 pos 16