L4 domain integrity

Domains
Understanding Domain Integrity
Version 1.0
(c) Rushdi Shams, CSE@KUET 1

What are Domains
 Domains are pools of values from where
values appear in the column of a table
 Domain of a column is KIND of value it
can represent
 Column: domain- it defines the kind of
values the column represents
 Data type and domain are two different
things
 Data type is a physical thing while
domain is a logical thing

What are domains (continued)
 I said, domain is a all possible collection of
values that a column can contain, in this
example, you have one perspective why I said
so!
 Say, Staffno column has 4 values in it-
01K28, 01B23, 03X98 and 09Y12 (well, I can’t
say what they really mean)! The domain of
Staffno is {01K28, 01B23, 03X98 and 09Y12}.
3(c) Rushdi Shams, CSE@KUET

 I also said that domain represents the
kind of data it represents. The following
example will simplify the statement
 Say, Streetname and Studentname are
two columns of a table. They can both
have the same data type like TEXT or
CHAR or VARCHAR or whatever…can’t
they? But they are different. Aren’t they?

 Again, say, there are two columns in a
table named rollnumber and salary having
the same data type INTEGER, NUMBER or
whatever…but they are different, no?
 So, studentname and streetname have
different domains and rollnumber and
salary have different domains
 In these two example, you can see the
difference between data type and domain
as well. Can’t you?

 Doamins are thus narrower concept than
data types. What is a narrower concept?
Well it means, it is descriptive!
Degreeawarded column can be
represented by simply datatype TEXT (3).
But in domains, it is represented by
{BA, BS, MA, MS, PhD, LLD, MD}
 But remember one thing- you don’t
manipulate domains in database
engine, you work with data type (as data
type is physical and domain is logical thing)

 Sometimes domains are large in
numbers so that they become difficult to
implement. Say, AGE! What if you are
representing the domain AGE? You can
then use the concept of range. The
domain of AGE can be between 1 to
120.

Deep into domains
 Now, we will dig into deep of the domain
and data type concept
 Sometimes, you will need to compare
between columns of two tables.
Then, you will need the concept of
domain.

Deep into domains (continued)
 SALES table has
Salesperson ID which is
eventually the Employee
ID of EMPLOYEE table.
 So, Mr. Michael Suyama
was dealing with Order
ID 10643.
 You may have to
compare them to find out
this information.
 If the domain of
Salesperson ID and
Employee ID is not
same, you could not
compare them!
9(c) Rushdi Shams, CSE@KUET

Deep into domains (continued)
 So, Salesperson ID and Employee ID,
as they are comparable, are called type-
compatible
 Trying to combine the relations on
Employee ID and Order ID, however,
would probably not result in a
meaningful answer, even if the two
domains were defined on the same data
type.

Domain Integrity
 In a database the domain integrity is
defined by
1. Data type and length
2. NULL value acceptance
3. Allowable values through rules
4. Default value
Back 1
Back 2

Domain Constraints
 Domain constraints are rules that define legal
values for columns
 Sometimes more than one rule is needed for a
column to completely define the legal values
 Unfortunately, SQL and other database engine
generally don’t facilitate with powerful tools to
recognize violations on domains.
 If employee ID and Salary are both NUMBERs
then, if you make a query by relating them, they
will certainly give you an answer! Aweful, huh?
 But some tools can still be easily manipulated to
avoid the violations.

Domain Constraints (continued)
DEFAULT
 You can use DEFAULT to set a column value to
some specific value. In this example, the data type
and length, NULL value acceptance and default
value is covered (see Domain Integrity)
CREATE TABLE Course(
Course_ID NUMBER(10) NOT NULL,
Course_Name VARCHAR(20),
Pass_Mark INTEGER(3) DEFAULT 40,
PRIMARY KEY (Course_ID)
);

CHECK
 This is a good tool to check if violation in putting data have
occurred or not. In this example NULL value
acceptance, data types and length and the allowable value
by rules are covered (see Domain Integrity)
CREATE TABLE Subject(
Subject_ID VARCHAR(10) NOT NULL,
Subject_Name VARCHAR(20),
Credit NUMBER(10)
CHECK(Credit>0 AND Credit<=10),
PRIMARY KEY (Subject_ID)
);

 If you use CHECK on more than one
column in a table, that constraint is also
called table constraint.
CREATE TABLE Lecturer(
ID VARCHAR(10) NOT NULL,
Name VARCHAR(20),
Grade CHAR(1),
Salary INTEGER(5),
CHECK (Salary<50000 OR Grade<=‘B’),
PRIMARY KEY (ID)
);

References
 Designing Effective Database Systems
by Rebecca M. Riordan, Addison
Wesley Professional, 2005
 Structured Query Language: A Practical
Introduction by Akil I Din
 Database Systems by Paul Beynon-
Devies, Palgrave Macmillan, 2004
 www.databasedev.co.uk

L4 domain integrity

More Related Content

What's hot (20)

Viewers also liked (14)

Similar to L4 domain integrity (20)

More from Rushdi Shams (20)

Recently uploaded (20)

L4 domain integrity