Optimization of Nested Queries in a Complex Object Model

Transformation of nested SQL queries into join queries is advantageous because a nested SQL query can be looked upon as a nested-loop join, which is just one of the several join implementations that may be available in a relational DBMS. In join queries, dangling (unmatched) operand tuples are lost, which causes a problem in transforming nested queries having the aggregate function COUNT between query blocks-a problem that has become well-known as the COUNT bug. In the relational context, the outerjoin has been employed to solve the COUNT bug. In complex object models supporting an SQL-like query language, transformation of nested queries into join queries is an important optimization issue as well. The COUNT bug turns out to be a special case of a general problem being revealed in a complex object model. To solve the more general problem, we introduce the nest join operator, which is a generalization of the outerjoin for complex objects.