As one of the first tools developed in the OHDSI Open Source tools suite, WhiteRabbit has existed for over 10 years. Since WhiteRabbit is positioned at the very start of an OHDSI OMOP targeting project, it supports many data sources. Not only (SQL) databases, but also other file formats, including CSV.
Adding support for a new database was somewhat challenging, as WhiteRabbit has grown support for more databases over time in an organic way. The “one size fits all” approach, which worked well for quite some time, started to show its limits. For this reason, before Snowflake support was added, two important changes were made first:
Securing the state of WhiteRabbit as a working solution by implementing a set of automated tests (both unit- and integration tests)
Redesign the database support in a more modular fashion
These two steps ensure that adding support for Snowflake would not disrupt the already existing functionality, while also making adding support for new databases more streamlined.
In addition, there is an opportunity to solve some long standing issues with already supported databases. Especially when configuring the connection to a source database, the “one size fits all” approach showed limitations in clarity and feedback, leading to some frustration among users. For databases that use the redesigned database support, users get more targeted feedback, leading to quicker troubleshooting solutions in connection configurations.
Additionally, configuration has been added to the project to ensure that WhiteRabbit (and Rabbit In A Hat) are up to date with regards to:
Removing vulnerabilities in dependencies
Compliance with Open Source licenses of dependencies