Database Reference
In-Depth Information
Every data manipulation request that can be completed through the use of one
language can be done through the use of the other language. Let us say that you
write a relational algebra solution for a complex query involving several source rela-
tions. The solution is likely to contain several relational algelra operations such
SELECT, PROJECT, JOIN, DIVIDE, and so on. Assume that you are able to arrive
at the solution through a commercial language implementing relational algebra.
Relational equivalency means that you can arrive at the solution for the same
complex query in relational calculus and a proper commercial language implemen-
tation of relational calculus will find the exactly same solution.
Relationally Complete
According to the author of the relational model, the true
test for whether a commercial language implementation is truly relational is
straightforward. If a commercial language can produce the same results that are
produced by either of the two generic languages for any query, then the commer-
cial language is said to be relationally complete.
The property of being relationally complete is an important qualification for a
language that can be used in a relational database. In other words, the commercial
language must be able to perform any set of relational algebra operations, however
complex the set may be. Similarly, the commercial language must be able to imple-
ment and produce the result expressed by any relational calculus statement,
however complex the statement may be.
Commercial Language Implementations
Structured Query Language (SQL)
has gained acceptance as the standard for the relational model. The language incor-
porates functions of the generic languages, although the standard SQL statement is
seen to be more like a relational calculus implementation. However, specific rela-
tional algebra operations are also included in the repertoire of SQL.
Commercial database vendors include SQL as part of their database products.
SQL interface is a standard component of the relational database management
systems on the market. However, each vendor tries to augment SQL standards with
its own proprietary enhancements. This results in a kind of incompatibility. SQL
code written for one commercial DBMS may not work for another DBMS unless
you refrain from using any of the proprietary revisions to SQL standards.
RELATIONAL MODEL DESIGN
From the discussion so far, you have captured the significance of the relational data
model. You have understood how it stands on a solid mathematical foundation
and is, therefore, a disciplined approach to perceiving data. The view of data in the
form of the common two-dimensional tables adds to the elegance and simplicity of
the model. At the same time, relational constraints or rules, to which the two-
dimensional tables must conform, ensure data integrity and consistency.
Commercial relational database management systems are implementations of
the relational data model. So, to develop and build a relational database system for
your organization, you need to learn how to design and put together a relational
data model. Although the model appears to be simple, how do you create a rela-
tional data model from the requirements? The previous chapters covered the details
Search WWH ::
Custom Search