I wondered why this sort of thing didn’t already exist. I had to develop for IBM Db2 (for i Series) on a Linux environment (using MariaDB) due to the licensing

I wondered why this sort of thing didn’t already exist. I had to develop for IBM Db2 (for i Series) on a Linux environment (using MariaDB) due to the licensing that stopped us from having a test env. I wound up writing a very ad-hoc C++ program that would parse a base SQL file and generate the appropriate DB2 and MariaDB dialect versions. Not flexible enough for reuse but it got the job done.