Advancing Spark - Json Schema Drift With Databricks Autoloader