SQL's Data Definition Language

(DDL) View, Sequence, Index

DDL: View, Index, Sequence,

Synonym

Objects are defined using the CREATE

statement
Some objects can be modified using the
ALTER statement
Some objects can be modified using the
ALTER statement
Some objects cannot be modified - they must
be dropped and then re-created
Objects are removed using the DROP
statement

View

A view is a virtual table no data is actually

maintained for a view
A view is defined based on existing tables by
using a SELECT statement:
CREATE VIEW viewname
AS SELECT
When a view is referenced the view's defining
query is executed and data is retrieved from
tables referenced by the defining query
The defining query for a view is stored in the
data dictionary

View Example
CREATE VIEW management
AS SELECT Emp_no, ename
FROM employees
WHERE emp_no IN
(SELECT mgr
FROM employees
WHERE mgr IS NOT NULL);
Using a view in a query causes the SELECT
statement that defines the view to be executed
and data values retrieved
eg SELECT ename
FROM management;

View Example 2

Views are often used to simplify writing of

complex queries which involve multiple tables
or group functions
Eg: CREATE VIEW deptsummary
AS SELECT d.deptno, dept_name, sum(sal) Salaries
FROM departments d, employees e
WHERE d.dept_no = e.dept_no(+)
GROUP BY d.dept_no, dept_name;
SELECT * FROM deptsummary
WHERE Salaries > 5000;

View Example 3
Views are also used to permit access to only
certain rows and/or columns in a table to some
user(s) and not permit access to the remaining
columns and/or rows
Eg: CREATE VIEW emplist
AS SELECT emp_no, ename, fname, job
FROM employees
WHERE LOWER(job)!='president'
ORDER BY ename, fname;
SELECT * FROM emplist;

Simple and Complex Views

A simple view is based upon a query that

involves only a single table and uses no
grouping; any other view is referred to a
complex view
Data can be modified using INSERT, UPDATE
and DELETE statements applied to a simple
view but not to a complex view
eg: UPDATE emplist
SET job = 'Assistant'
WHERE job = 'Clerk';

Modifying Data using a View

Data can be modified using INSERT, UPDATE
and DELETE statements on a simple view but
this is not normally done
A simple view defined using a WITH READ
ONLY clause will not permit data to be modified
eg CREATE VIEW emplist2
AS SELECT emp_no, ename, fname, job
FROM employees
WHERE LOWER(job)!='president'
ORDER BY ename, fname
WITH READ ONLY;

Change View
Change definition of a view by using the OR
REPLACE clause in the CREATE VIEW
statement
eg CREATE OR REPLACE VIEW emplist
AS SELECT emp_no, ename, fname, job
FROM employees
ORDER BY ename, fname;

Indexes

Indexes store the value of a column(s) being

indexed as well as pointer(s) to the physical
location of the row(s) having this value
Most are implemented using a B*Tree
Indexes are used for 2 purposes: improving
performance of queries and enforcing
uniqueness of column values
CREATE INDEX indexname
ON tablename(col1name,col2name,)

Indexing and Performance

Indexes are often created to improve

performance of queries
However indexes slow performance of INSERT,
UPDATE and DELETE operations because the
index must be updated also
Be aware of trade-off between overhead
involved and usefulness of index
DBMS decides when an index should be used
to retrieve data programmer cannot reference
an index in a query

Indexes

Index can be used to enforce uniqueness on

column(s) in a table
PRIMARY KEY and UNIQUE constraints are
actually implemented by Oracle DBMS defining
a UNIQUE index
CREATE UNIQUE INDEX indexname
ON tablename(col1name,)
eg CREATE UNIQUE INDEX SINIndex
ON STUDENT(SIN)

Composite Indexes

If defining an index on a composite field, the

order of the columns is significant : list the most
commonly queried column first so that the index
can be used for this single column or for the
combined columns

eg CREATE INDEX SectionIndex

ON SECTION (subjcode,sectcode)

SectionIndex can be used to search on subjcode

alone or subjcode and sectcode combined but
is not used to search on sectcode alone

Common Reasons to Define an

Index on Large Tables

Column(s) used frequently in joins therefore

index defined on most foreign key fields
Column(s) used frequently in clauses (WHERE,
GROUP BY, ORDER BY)
Columns with a wide distribution of values so
index entries correspond to a limited number of
rows
Column with large occurrence of NULL values
since NULL values are not indexed

Sequence

A sequence is an object which is used to

generate sequential values which are often
assigned to primary keys
A sequence is not physically linked with a table
or column
It is the responsibility of the developer to only
use a sequence with the primary key or column
that it was defined for
CREATE SEQUENCE seqname
INCREMENT BY n
START WITH n

Sequence(ctd)
Eg CREATE SEQUENCE emp_no_seq
START AT 1000
INCREMENT BY 1
A sequence has two pseudo-columns:
- NEXTVAL : gets the next sequential value of
the sequence
- CURRVAL : gets the current value of the
sequence
Eg INSERT INTO employees
VALUES(emp_no_seq.nextval, )
Eg SELECT emp_no_seq.currval FROM dual

Synonym

A synonym allows you to associate an alternate

name with an object such as a table or view
Eg CREATE SYNONYM emp
FOR employees;
SELECT * FROM emp;

Remove Object from Data

Dictionary

Remove object from data dictionary

DROP VIEW viewname
DROP SEQUENCE sequencename
DROP INDEX indexname
DROP SYNONYM synonymname