Hibernate version:
3.0.5

I run into the problem where a database table (legacy system) defines a secondary unique key on a table which hibernate do not take into account during "reordering" of its database operations.

Mainly I delete all records and reimport new records out of a csv file.

Maybe there are the same records in this file maybe not.

Using hibernate "in its correct manner" would force me to write a complex diff tool to archive this. So delete and reimport is much easier.

Is there a chance to define a mapping so hibernate deletes the record before inserting the new one?