The alternative and more flexible approach is to use a software package with the Data Mapper pattern. It differs in that one is no longer tightly coupled to th

The alternative and more flexible approach is to use a software package with the Data Mapper pattern. It differs in that one is no longer tightly coupled to the actual database tables. Instead, you can create whatever views of the data make sense to your application. Even though it seems like a higher level of abstraction, you actually have greater access to all the features SQL provides, almost in the form of an SQL builder. One example of the Data Mapper pattern is in python's SQLAlchemy. Another is in perl's DBIx::ObjectMapper.