@mistercrunch This PR was rewritten from scratch based on your comments and is now ready for review. I have updated the description above to describe the logic. In short this PR forces quotes around column aliases using native SQLAlchemy constructs for affected databases, which aligns the column names in the query and result. I was able to contain all oracle-like logic within `db_engine_specs.py` and `connectors/sqla/models.py`, i.e. there is no longer any need for dataframe column renaming downstream e.g. `viz.py`.
FYI @HectorLira @minh5 @mmuru: This PR attempts (yet again) to improve stability for Redshift, Oracle, Snowflake (beware, you must use `snowflake-sqlalchemy==1.1.0`). Any testing help to verify whether or not it works would be valuable, especially Redhshift, as this was mainly validated against Snowflake during development. [ Full content available at: https://github.com/apache/incubator-superset/pull/5686 ] This message was relayed via gitbox.apache.org for [email protected]
