Relational Database Objects and Abstract Data Types Oracle Database Tips by Donald Burleson

Unlike traditional database management system which only provide for primitive data types such as INTEGER and CHARACTER, the object-oriented programming languages allow for the creation of abstract  data types.   The data types offered in commercial database systems, CHAR INTEGER NUMBER, VARCHAR, BIT, are sufficient for most relational database applications but developers are now beginning to realize that the ability to create user-defined data types can greatly simplify there database design.  While these data types were popular within the programming languages, they have only recently been introduced into the mainstream world of database objects.

Some commercial relations database vendors have committed to incorporating user-defined data type in their future releases.  Oracle, the popular relational database for midrange computers, has announced that Oracle version 8 will support abstract  data typing by extending SQL to allow for a CREATE TYPE definition.

At the most basic level, and abstract data type is nothing more than a collection of smaller, basic data types that can be treated as a single entity.  While this is a simple concept, the changes in database object design will be dramatic.  Some argue that a database must be able to support data types which contain lists rather than finite values, and some of the object/relational databases such as UniSQL allow for a single data type (a column) to contain lists of values, or even another table.

These features are called by several names.  In the C programming language they are called structures, in data structure theory they are called abstract data types (ADT's), while in the marketplace they are referred to as user-defined data types.  Each of these terms mean essentially the same thing, and we can think of these terms as being interchangeable.

While pre-relational databases supported primitive ADT's, the more robust data typing was not introduced until the object/relational hybrid database became popular.  UniSQL, the relational/object-oriented database developed by Dr. Wong Kim, supports the concept of nested tables, whereby a data "field" in a table may be a range of values or an entire table.  This concept is called complex, or unstructured data typing.  With this approach the domain of values for a specific field in a relational database may be defined.  This ability to "nest" data tables, allows for relationship data to be incorporated directly into the table structure.  (Figure 5.1)  For example, the OCCUPATIONS field in the table establishes a one-to-many relationship between an employee and their valid occupations.  Also note that the ability to nest the entire SKILLS table within a single field.   In this example, only valid skills may reside in the SKILLS field, and this implements the relational concept of "domain integrity".

Figure 5.1   An example of embedded complex data types

Now that see understand the basic idea behind abstract data types, let's explore some of the compelling benefits to this approach.

 

---