Best pattern to follow for server validation with SQLDataSource

[jay.l.fisher]

Our server side code consists mostly of subclasses of SQLDataSource. What is the best pattern to follow when adding validation rules in this scenario? Should we override the validate() method, call super.validate() first and then do our own validations and add to the ErrorReport we return?

[Isomorphic]

Validations you can do with Velocity you would generally put inline, and any validators that are re-usable across DataSources are usually called via DMI. DataSource-specific logic would go in validate() as you mention.