Database: Fields: Data Integrity
The integrity of a database depends on the correctness
of the data in the database. There are different types of errors
and several methods to handle them in SQL. The most important
method is to specify constraints when defining a field in a table.
Most database systems will enforce these constraints with no additional
programming. Many systems also all defining triggers which are statements
to be performed when a particular condition occurs.
Required Data
Problem: Some fields must not be NULL.
Example:
SQL Solution: In the CREATE TABLE statement specify the NOT NULL
constraint for that field.
Unique Values
Problem: Certain values in a column must be unique.
Example: A primary key for something like an employee number must be unique.
SQL Solution: Specify the PRIMARY KEY or UNIQUE constraints
for that field.
Valid Data Values
Problem: There may be constraints on certain data values.
Example: A person's age may not be a negative number.
SQL Solution: Specify the CHECK constraint when creating
that field in a table. The field which is the result
Consistency
Problem: Updates to a database often involve changes to several tables.
No access should be allowed when the table is in an intermediate, inconsistent
state.
Example:
SQL Solution: Use transactions.
Other Conditions, Business Rules, etc
Problem: A particular database may have complex conditions that
apply to their situation.
Example: [TO DO]
SQL Solution: These need to be enforced
by some combination of constraints, triggers, and programs.