- Add regression tests
- Add unit tests
- Add support for Python 3.5
- Check to see if FK exists between 2 tables - if it does, append an integer to the end of the constraint_name. Right now we just catch exceptions, and only on a subset of supported RDBMSs (OperationalError - MySQL, ProgrammingError - PostgreSQL)
- Replace column-type guessing process (iterating over every row) with a GROUP BY query to improve performance.
- Add parameter for "quoted_strings_enclosed_by" to ETLAlchemySource to override default .csv quote-character.
- Add parameter for "cleanup_table_csv_files", with default value of True, allowing the user to override default and let files persist after they are loaded into Target DB.