Microsoft® SQL Server 2012 Unleashed

Abstracting Data

Data abstraction, in its simplest form, isolates the client code from changes to the underlying structure. A view can be used to implement data abstraction within your database schema. If, for example, you have client code that will retrieve data from a database table that is likely to change, you can implement a view that retrieves data from the underlying table. The client code will then reference the view and never access the underlying table directly. If the underlying tables change or the source of the data for the view changes, these changes can be isolated from the referencing client code.

To demonstrate this scenario, let’s look at the following SELECT statement, which retrieves data directly from the Sales.SalesOrderHeader ...