SQL Book-1

SQL MATERIAL
Chapter-1
SQL INTRODUCTION
 Oracle is a product from oracle which is used to store data permanently see store device.
 Every organization store same type of data.
 Oracle product of Oracle Corporation.
DATA: It is a collection of “RAW FACTS”.
Ex: 1.Student Marks
2. Customer Names
INFORMATION:
When we are processing data are achieving meaning full results it is called as
“INFORMATION”.
Ex: 1.Marks Sheets
2. Information Of Customer
DATA STORE:
It is a place where we can store data or information.
1. Books & Papers
2. Flat Files
3. Database
FLAT FILES:
Data or Information stored in individual unrelated files these files are also called as
“FLAT FILES”.
G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 1

SQL MATERIAL
Program-1 Program-2 Hard disk
User-1
User-2
File-2
File-1
Fig: Flat file approach
If you want to store data in flat file we are most develop application program in high
level language. In flat file approach every application maintains it own file which is separate
from another application program.
Student
Sno Sname class
Record application program
Flat file
Prof id Prof name Time class
Scheduling application program
DISADVANTAGE:
1. Data Retrieval
2. Data Redundancy
3. Data Integrity
4. Data Security
5. Data Indexing

SQL MATERIAL
1. Data Retrieval:
If you want to retrieval data from flat file we must develop application program in high
level language. Whereas if you want to retrieval data from database then we are using
SEQUEL(Structural English Query Language).
2. Data Redundancy:
Same time we are maintaining multiple copies of the same data in different locations.
This data is also called as “REDUNDANCY DATA”. In flat file mechanism if we are
modifying data in one location. It is not affected in another location. This is called as
“INCONSISTENCY”. Flat file does not maintain consistence data.
Database automatically maintains consistency data. Because in database every transaction

in related having 4 properties. This are also called “ACID PROPERTIES”. These properties
only maintain consistent data. In database we can also reduce redundancy use in normalization
process.
3. Data Integrity:
Integrity means to maintain proper data if we want to maintain proper data when we
define set of rules this set of rules also called “BUSINESS RULES”. In database we are
defining business rules using constrains, trigger. Where as in flat file if we want maintain proper
data. If we most develop application program in high level language.
4. Data Security:
Data stored in flat file cannot be security. Because file does not provides rule based
security.
5. Data Indexing:
If we want to retrieve data very fastly when database provides indexing mechanism but
flat file does not support indexing mechanism.

SQL MATERIAL
What is Database?
It is organized collection of inter related once data is stored in DB. It can be stored
integrated and also simultaneously number of user accesses this data.
 American National Standard Institute (ANSI) as establish three level architecture for DBMS.
 Standards Planning and Requirement Committee (SPARC).
Architecture contains three levels:
1. Internal Data Level
2. Conceptual Data Level
3. External Data Level
DBMS ARCHITECTURE:
Purchase manager
Customer
Item name
External Item name
level Price
Price
Quality
Logical data independence
Create table item(itemname varchar2(10),

Conceptual
level Price number(5),qty number(5));
Physical data independence
Item name type=byte(8) offset=(10);

Internal
Price type=byte(10) offset=(15);
level
Qty type=byte(9) offset=(20);
All dbms s/w implemented based on data independence.db structure is independent on

application program.

SQL MATERIAL
Logical data independence:
When ever we modifying conceptual level it is not affected external level.
Physical data independence:
When ever we modifying internal level it is not affected conceptual level. i.e. if we are
moving db from one location to another location then conceptual level does not affected.
External Conceptual Internal
Physical data
EMP stored
view-1
view-2
DEPT
1.
Fig: ANSI/SPARC Architecture (or) DBMS Architecture
1. Internal Data Level:
This level describes how data internally stored in database. Physical representation of the
database on the computer database administrators only works in internal level. This level
maintain by database administrator (DBA’s).
2. Conceptual Data Level:
This level describes organization view of the database. In this level only of database
develop creating database and maintain data properly.
3. External Data Level:
This level users access or view of the database. Describe that part of database that is
relevant to a particular user. In this level database developer create views and giving to the
number of members.

SQL MATERIAL
DATA MODEL:
Data model describe how data layout store in conceptual level.
They are three level model used by organization:
1. Hierarchical Data Model
2. Network Data Model
3. Relational Data Model
1. Hierarchical Data Model:
Hierarchical data model organizers data in tree like structure. This hierarchy is also called
parent, child hierarchy. This structure employee’s child a record can have repeating information
(generally child segments) in this model data is represented by collection of records (record
type).
Hierarchical model Relational model
Root Eid Ename Deptname

1000 Smith Admin
1001 Allen Hr
1000 Smith admin 1001 Allen Hr 1002 King sw
2005 50000 2005 50000
Eid Ename Deptname

1002 King S/w 1000 2005 50000
1001 2005 50000
1002 2013 70000
1002 2013 60000
2013 70000 2013 60000

SQL MATERIAL
In this a record type corresponding to the table in relational model and also recording
same as row in a table in relational model. This model is implemented based on ONE-TO-
MANY relationship. Based on this restriction number of child on single parent they way in this
model more number of time child segment repeated that weighs more redundant data and also
this model retrieve data very slowly from the database because this model is search data from
root node an wards if want operate hierarchical data model product that we are using procedural
language.
In 1960 IBM introduced IMS (information management system) product based on

hierarchical model.
2. Network Data Model:
In 1970 CODASYL (conference on data system languages) implemented network data

model. This model implemented based on MANY-TO-MANY relationships. This model reduces
duplicate data. In this model also data stored in collection of record and also record type is same
table in relational model and also record are represented also graphs used on this model.
In 1970 IBM implemented IDMS (integrated database management system) products. If

want operate network data model product also then we are using procedural language.
1000 Smith Admin
2005 50000
1001 Allen hr
2013 70000
1002 king sw
2013 60000

SQL MATERIAL
3. Relational Data Model:
In 1970 E.F.CODD introduced relational data model. This model also implement based
on MANY-TO-MANY relationship in this model. We are storing data in two-dimensional table.
Relational model is three components:
1. Collection of objects
2. Set of operations
3. Set of integrity rules

SQL MATERIAL
Chapter-2
ORACLE HISTORY
In 1970 E.F.CODD write a paper “Relational model of data for large stored data banks”.
IBM INGRESS
System/R
INGRESS Larry Ellison (1977)
(Square) (1982)
RSI (1979) Oracle Corporation
Oracle Oracle
Relational Software, Incorporated (RSI)
Oracle
SQL PL/SQL DYNAMIC SQL
What is Oracle?
In 1970 E.F CODD described Oracle. Oracle is a database. Oracle database is an Object-
Relational Database Management System produced and Marketed by Oracle Corporation. Oracle
is a fourth generation Relational Database Management Systems. Oracle is a collection of data
treated as a unit. The purpose of a database is to store and retrieve related information.

SQL MATERIAL
ORACLE VERSIONS
Oracle 2.0 ---------- 1979
 First public release

 Basic sql functionality, Joins
Oracle 3.0 --------- 1983
 Commit, Rollback
 Virtual c++ (window program) (VC++)
 Rewritten in ‘c’ language
Oracle 4.0 ---------- 1984
 Read consistency
Oracle 5.0 ----------- 1985
 Client server architecture
Oracle 6.0 ----------- 1988
 Introduced pl/sql
 Row level locks
Oracle 7.0 ----------- 1992
 Integrity
 Stored procedures, Stored functions
 Triggers
 Truncate table
 Datatype varchar changed into varchar2
 View compilation
Oracle 7.1 ---------- 1992
 Dynamic SQL
 ANSI/ISO SQL92
SQL MATERIAL
Oracle 7.2 ---------- 1992
 Ref cursor
 Inline views
 Dbms_job package
Oracle 7.3 ----------- 1992
 Bitmap indexes
 Utl_file package
Oracle 8.0 ----------- 1997
 Object technology
 Nested table,Varrays
 Lobs (large objects)
 Column increased per a table upto 1000
Oracle 8i ------------- 1999 (I-internet)
 Analytical functions
 Materialized views
 Autonomous transaction
 Bulk collection clause
 Case conditional statement
 Instead of triggers
 Enable, disable trigger
 Function based indexes
Oracle 9i ------------ 2001
 Renaming a column
 ANSI joins
 Flashback query

SQL MATERIAL
 Multi table inserts

 Merge statement
Oracle 10g ------------ 2003 (G-grid technology)
 Recycle bin concept

 Flashback table
 Wm_concat()
 Indices of clause
 Rename table space
 Regular expressions
Oracle 11g ------------ 2007
 Continuous stmts introduced in pl/sql loops

 Read only tables
 Virtual columns
 Simple_integer datatype in pl/sql
 Sequences used in pl/sql without dual table
 Follows clause in triggers
 Compound triggers
 Enable, Disable clause used in trigger specification.
 Named, Mixed notations used in a sub program executed
using select statement.

SQL MATERIAL
SQL (Structured Query Language)
 Sql is a non-procedural language to operate relational database.

 In 1970 E.F.CODD introduced DSL/ALPHA language and operate relational database.
 IBM System/R team developed simplified version of DSL/ALPHA language this is called
“SQUARE”.
 IBM modified DSL SQUARE into SEQUEL.SEQUAL modified into SQL.
 In 1986 ANSI standard are defined in SQL-ANSI SQL.In 1987 ISO statement designed in
SQL.
 ANSI/ISO SQL89 ---1989 (Sql I)
 ANSI/ISO SQL92 ---1992 (Sql II)
 ANSI/ISO SQL99 ---1999
 ANSI/ISO SQL03 ---2003 (Sql III)
End User
Oracle db
SQL
QUERY
TOOLS 1.sql plus *loading data
2.isqlplus *retrieving data
3.toad *modifying data

Collection of objects
4.pl/sql developer *deleting data
5.sql navigator

SQL MATERIAL
Oracle 10g (or) 11g (enterprise edition):
Username: SCOTT
Password: TIGER
Error: Account locked
Unlocked user:
Username: \ SYS AS SYSDBA <┘
Password: SYS <┘
Sql>Alter user scott account unlock;
Sql>conn scott/tiger; <┘
Password: TIGER
Confirm password: TIGER
To view all table:
Sql>select * from tab;
DEPT ------- Master table EMP -------- Child table
To clear screen and set line proper:
Sql>cl scr; Sql>set line 100;
Go to editor:
Sql>ed <┘
Get back into sqlplus:
- Alt+f+x <┘
Sql> / <┘

SQL MATERIAL
Chapter-3
TYPES OF SQL STATEMENT

Data Definition Language (DDL):
- Create
- Alter
- Drop
- Truncate
- Rename (Oracle 9i)
Data Manipulation Language (DML):
- Insert
- Update
- Delete
- Merge (Oracle 9i)
Data Query Language (DQL) or Data Retrieval Language (DRL):
- Select
Transaction Control Language (TCL):
- Commit (Save)
- Rollback (Like undo)
- Savepoint
Data Control Language (DCL):
- Grant (Giving Permissions)
- Revoke (Cancel permissions)

SQL MATERIAL
DATA TYPE
Data type identifier type of data within a table column.
1. Number (P,S)
2. Char ------------- Varchar2 (maxsize)
3. Date
1. Number:
It is used to store fixed, floating point number.
Syntax: columnname number (P, S) P-precision S-scale
Maximum limit of precision is upto 38.
2. Char:
It is used to store fixed length ALPHA numeric data in bytes. Maximum limit is
2000bytes.By default char data type having 1 bytes.
Syntax: columnname char (size)
Long (2 GB) clob (4 GB) blob (image)
Varchar2 (maximum size):
It is used to store variable length alpha numeric data maxsize is 4000 bytes.
Syntax: columnname varchar2(maxsize)
Sql>select dump(name) from test;
3. Date:
It is used to store date in oracle data format. By default in oracle date format “DD-MON-
YY”.
Syntax: columnname date;

SQL MATERIAL
DATA DEFINITION LANGUAGE (DDL):
These commands are used to define structure of the table.
- Create
- Alter
- Drop
- Truncate
- Rename
Create:
This command is used to create database objects like tables, views, procedures, indexes,
sequences………………………
Syntax: Create table tablename(col1 datatype(size),col2 datatype(size),…………);
Ex: Sql>create table india(sno number(5),name varchar2(10));
To view structure of the table:
Sql>desc tablename;
Ex: sql>desc india;
Alter:
It is used to change existing table structure.
Alter
Add Modify Drop
Add: It is used to add number of columns into the existing table.

SQL MATERIAL
Syntax: alter table tablename add(col1 datatype(size),col2 datatype(size),……….);
Ex: sql>alter table india add sal number(10);
Modify: It is used to change column datatype (or) column datatype size only.
Syntax: alter table tablename modify(col1datatype(size),col2 datatype(size),……);
Ex: sql>alter table india modify sno date;
Drop: It is used to remove column from the table.
Method-1:
If we want drop a single column at a time without using parenthesis then we are using
following syntax.
Syntax: alter table tablename drop column columnname;
Method-2:
If we want drop a single (or) multiple column with using parentheses then we are using
following syntax.
Syntax: alter table india drop(name);
NOTE: we cannot drop all columns in a table.
Drop: It is used to remove database object from the table.
Syntax: drop objecttype objectname;
Drop table tablename;
Ex: sql>drop table india;
Sql>drop procedure procedurename;

SQL MATERIAL
Dropping a table:
Before oracle 10g After oracle 10g
Sql>drop table tablename; Sql>drop table tablename;
Oracle db
Oracle db
Table
Table
Recyclebin
Table
Permanently Remove
Get it back from recycle bin:
Syntax:
Flashback table tablename to before drop;
To drop permanently:
Sql>drop table tablename purge;
Get it back from recycle bin:
Sql>flashback table india to before drop;
To drop permanently:
Sql>drop table india purge;
Testing:
Sql>flashback table india to before drop;
Error: object not in recycle bin.

SQL MATERIAL
Recycle bin:
Recycle bin is a read only which stores dropped store. Whenever we are installation
oracle server then automatically so many read only table there. These read only tables are also
called as Data Dictionary.
Oracle db
Sql>create table india(sno number(5));
Recycle bin
Sql>drop table india; Object_name Original_name
Testing:
India
Sql>desc recyclebin;
Sql>select original_name from recyclebin;
Purge:
We can also drop object from recycle bin purge command.
To drop single table from recycle bin:
Syntax: purge table tablename;
To drop all table from recycle bin:
Syntax: purge recyclebin;
Truncate: This command is used to delete all rows permanently from the table.
Syntax: Truncate table tablename;
Sql>create table india as select * from emp;
Sql>select * from india;
Sql>truncate table india;
Testing: sql>select * from india;
Sql>desc india;
SQL MATERIAL
Rename: It is used to renaming a table and renaming a column also.
Renaming a table:
Sql>rename oldtablename to newtablename;
Sql>rename india to andhra;
Renaming a column (Oracle 9i):
Syntax: alter table tablename rename column oldcolname to newcolname;
Sql>alter table emp rename column empno to sno;
Sql>select * from emp;
NOTE: In all db system by default all DDL command is automatically commit (save).
DATA MANIPULATION LANGUAGE (DML):
DML command is used manipulation data in a table.
- insert
- update
- delete
- merge (Oracle 9i)
Insert:
This command is used insert data into a table.
Method-1:
Syntax: insert into tablename values(val1,val2,………….);
Sql>insert into india values(1,’SIMHA’);
Sql>insert into india values(2,’REDDY’);

SQL MATERIAL
Method-2 (using substitutional operator (&)):
Syntax: insert into tablename values(&col1,&col2,………………);
Sql>insert into india values(&sno,’&name’);
Enter value for sno: 3
Enter value for name: abc
Sql>/
Enter value for name: xyz
Method-3 (skipping column):
Syntax: insert into tablename(col1,col2,………..) values(val1,val2,……….)
Sql>insert into india(name)value(‘zzz’);
Update: It is used to modify (change) data in a table.
Syntax: update tablename set columnname=newvalue where columnname=

oldvalue;
Sql>update emp set sal=3000 where ename=’SMITH’;
Sql>alter table india add address varchar2(10);
Sql>update emp set sal=NULL where sal=3000;
Remove the sal 3000
NOTE: If we want to insert particular cell values also then we are using update stmts.
Delete: It is used to delete all rows (or) particular row from the table.

SQL MATERIAL
Syntax: Delete from tablename; (To delete all rows)
Delete from tablename where condition; (To delete particular rows)
Sql>delete from india;
Get it back data:
Sql>rollback;
Difference b/w Delete and Truncate:
We are using delete from tablename. Then automatically total data delete from table and
also deleted a data automatically store in a buffer. We can get it back data using rollback
(without commit).
When we are using truncate from tablename all row delete permanently. It cannot get it
back this data. Because truncate is a DDL command automatically commits.
DATA RETRIEVAL LANGUAGE (DRL) (OR) DATA QUERY LANGUAGE (DQL):
This command is used to retrieval data from the table.
Syntax: select col1,col2,……….
From tablename
Where condition
Group by columnname
Having condition
Order by [asc/desc];

SQL MATERIAL
1.select all cols & all rows.
2.select all cols & particular rows.
3.select particular cols & all rows.
4.select particular cols & particular rows.
Creating a new table from existing table (or) Coping a table from another table:
Syntax: create table newtablename as select * from existingtablename;
Coping table without coping a table:
Syntax: create table newtablename as select * from tablename where falsecondition;
Sql>create table kadapa as select * from emp where 1=2;
Sql>select * from kadapa;
No rows selected.
Sql>desc kadapa;
NOTE: Where we are copping a table from another table constraints are not copied.
OPERATORS USED IN A SELECT STATEMENT:
1. Arithmetic operators (*, +,-,/)

2. Comparison (or) Relational operator (=, <, <=, >, >=, <> (or) != )
Where
3. Logical operator (AND, OR, NOT)
Clause 4. Special operator
Arithmetic operator is used in number, date datatype columns.
Q) Write a query to display ename, sal, annsal from emp table?
Ans: sql>select ename, sal, sal*12 “annsal” from emp;

SQL MATERIAL
Q) Write a query to display the emp except job has clerk from emp table?
Ans: sql>select * from emp where job<>’CLERK’;
Q) Write a query to display EMP’s we are getting more than 2000 salary from emp table?
Ans: sql>select * from emp where sal>2000;
Sql>select * from emp where job=’CLERK’ AND sal>1100;
Sql>select * from emp where job=’CLERK’ OR sal>2000;
Q) Write query to display the emp who are belongs to 20,30,50,70 deptno from emp table?
Ans: sql>select * from emp where deptno=20 or deptno=30 or deptno=50 or deptno=90;
Sql>select * from emp where sal>2000 and sal<5000;
SPECIAL OPERATOR:
In not null
Between not between
Is null is not null
Like not like
In:
This operator is used to pick the values one by one from list of values. Generally IN
operator is used in place of OR operator. Because IN operator performance very high compare to
OR operator.
Syntax: select * from tablename where columnname in (list of value);
NOTE: Generally in all database system we are using multiple rows subqueries then we
must use IN operator.
Sql>select * from emp where deptno in(20,30,50,70,90);
Sql>select * from emp where deptno not in(20,30,50,70,90);

SQL MATERIAL
Ex: sql>select * from emp where ename in(‘SMITH’,’KING’);
Sql>select * from emp where not in(20,30,NULL);
NOTE: NOT IN operator does not work with null values in all database systems.
null:
NULL is an undefined, unknown, unavailable value it is not same as zero. Any arithmetic
operation performed with null value again it will becomes null.
Null+50-----------------Null
Q) Write a query to display ename,sal,comm,sal+comm of the emp SMITH from emp

table?
Ans: sql>select ename,sal,comm,sal+comm From emp where ename=’SMITH’;
Ename sal comm sal+comm
SMITH 1000 --------- --------------
To overcome this problem oracle introduced NVL() function.
nvl():
It is a predefined function which is used to replace/substitute userdefined value (0) in

place of null.
Syntax: nvl(exp-1,exp-2)
-here exp-1,exp-2 must belongs to same datatype.
-exp-1 is null then it returns exp-2 otherwise it returns exp-1.
Ex: nvl(null,10) ----------10
Nvl(40,10) -----------40
Sql>select ename,sal,comm,sal+nvl(comm,0) from emp where ename=’SMITH’;

SQL MATERIAL
Ename sal comm sal+comm =>sal+nvl(comm,0)
SMITH 1000 -------- 1000 =>1300+nvl(null,0)
=>1300+0
=>1300
Between:
This operator is used to retrieve range of values this operator is also called as BETWEEN
………AND operator.
Syntax: select * from tablename where columnname between lowvalue AND

highvalue;
Sql>select * from emp where sal between 2000 and 5000;
Is null, is not null:
These two operator are used in where clause only. This used to whether a column having
null value (or) not.
Syntax: select * from tablename where columnname is null;
Syntax: select * from tablename where columnname is not null;
Q) Write a query to display emp who are not getting commission from emp table?
Ans: sql>select * from emp where comm is null;
Q) Write a query display emp does not have manager from table?
Ans: sql>select * from emp where mgr is null;
Sql>select * from emp where comm is not null;
Like: This operator is used to search data based on character data. This operator performance is
high compare to searching function. Along with LIKE operator we are using 2 special characters.

SQL MATERIAL
LIKE
% _ (underscore)
(string or group of character) Single character
Syntax: select * from tablename where columnname like ‘character’;
Q) Write a query to display emp whose ename start with ‘M’ from emp table using like
operator?
Ans: sql>select * from emp where ename like ‘M%’;
Q) Write a query to display the emp whose joining in the ‘month dec’ emp table using like
operator?
Ans: sql>select * from emp where hiredate like ‘%dec%’;
Q) Write a query to display the emp whose joining in the ‘year 81’ emp table using like
operator?
Ans: sql>select * from emp where hiredate like ‘%81’;
Q) Write a query to display the emp whose ename second letter would ‘L’ from emp table
using like?
Ans: sql>select * from emp where ename like ‘_L%’;
If third letter:
Sql>select * from emp where ename like ‘__L%’;
Concatenation operator (||):
This operator is used to concatenation data column values with literal strings.
Sql>select ‘my employee names are’||ename from emp;
Sql>select ename||’ ‘||sal from emp;

SQL MATERIAL
Chapter-4
FUNCTIONS
Functions are used to solve particular talk in a problem and also functions must return a
value oracle having two types of functions.
1. Predefined Functions
2. Userdefined Functions
1. Predefined functions:
1. Number Functions
2. Character Functions
3. Date Functions
4. Group of functions (or) Aggregate functions
Number Functions: These functions operator an number data.
1. abs(): It is used to convert negative value into positive value.
Ex: sql>select abs(-50) from dual;

Sql>select abs(sal-comm) from emp where comm is not null;
Sql>select ename,sal,comm,abs(sal-comm) from emp where comm is not null;
Dual:
Dual is a virtual table which contains one row and one column by default dual table
column data type is varchar2.
Sql>select * from dual;

Sql>desc dual;
Generally dual table is used test predefined userdefined functions functionality. Dual
table is also used to generate sequences values and also used to mathematical operations.
Ex: sql>select 50+70 from dual;

SQL MATERIAL
2.mod(m,n):
It returns remaining after M divided by N.
Sql>select mod(30,2) from dual; o/p: 0
Sql>select mod(sal,3) from emp;
3.round(m,n): It rounds given floating point value number ‘M’ based on ‘N’.
Sql>select round(1.8) from dual; o/p: 2
Sql>select round(1.23456,3) from dual; 1.234/56 56 is more than 50%
o/p: 1.235 1.235 hence 1 is added.
Sql>select round(1285.786,-1) from dual; 1280 5 replace with 0
o/p: 1290 1290 50% above then 1 is added
sql>select ename ,sal,sal/22,round(sal/22),round(sal/22,1),round(sal/22,2) from emp;
NOTE:
Round always checks remaining number it means remaining number is above 50% then 1
is added to the rounded number.
4.trunc(m,n): It truncate given floating value ‘M’ based on ’N’.
Sql>select trunc(1.9) from dual; o/p: 1
Sql>select trunc(1.23456,3) from dual; o/p: 1.234
5.ceil(),floor():
Ceil() returns nearest greatest integer where are floor() returns nearest lowest integer.
Sql>select ceil(1.4) from dual; o/p: 2
Sql>select floor(1.9) from dual; o/p: 1

SQL MATERIAL
6.greatest(exp1,exp2………),least(exp1,exp2…………….):
Greatest() returns maximum value within given expression where as least returns
minimum value among given expression.
Sql>select greatest(3,5,9) from dual; o/p: 9
Sql>select ename,sal,comm,greatest(sal,comm) from emp where comm is not null;
Sql>select greatest(sal) from emp; o/p: 800,1220, 5000,3000 1350…………..
We not show maximum value so there it is the problem. To overcome this value we use
max(sal).
Max(): This function comparing single column only.
Sql>select max(sal) from emp;
Character Function:
1.upper():
It is used to convert a string into uppercase or column value into uppercase.
Sql>select upper(‘abc’) from dual;
Sql>select upper(ename) from emp;
2.lower():
It is used to convert string into lowercase or column value into lowercase.
Sql>select lower(ename) from emp;
Sql>update emp set ename=lower(ename);
3.initcap():
It returns initial letters capital and all remaining letters are small.
Sql>select initcap(‘abcdef’) from dual; o/p: Abcdef

SQL MATERIAL
Sql>select initcap(‘ab cd ef’) from dual; o/p: Ab Cd Ef
Sql>select initcap(ename) from emp; o/p: Smith,Allen,Blake,King…………….
4.length():
This function always returns number datatype it returns total length of the string
including spaces.
Sql>select length(‘abcdef’) from dual; o/p: 6
Sql>select length(‘abc def’) from dual; o/p: 7
5.substr():
It will extract position of the string within the given string based on last two parameters.
Sql>select substr(‘ABCDEF’,2,3) from dual; o/p: BCD
Sql>select substr(‘ABCDEF’,-2,3) from dual; o/p: EF
Syntax: substr(columnname (or) ‘stringname’,string position, number of char from

position);
Q) Write a query to display the emp’s whose ename second letter would ‘LA’ from emp
table using substr() function?
Ans: sql>select * from emp where substr(ename,2,2)=’LA’; o/p: BLAKE,CLARK
NOTE:
In all database systems we are not allowed to use group function in where clause but we
are allowed to use number, character, date function in where clauses.
6.instr():
It returns number datatype this function returns position of the delimiter (or) position of
the character (or) position of the string within given string.
Instr(columnname (or) ‘stringname’,’string’,searching position,no.of occurace from position)

SQL MATERIAL
Sql>select instr(‘ABCD’,’C’) from dual; o/p: 3
Sql>select instr(‘ABCDEFGHICDJKMLCDHJF’,’CD’,-5,2) from dual; o/p: 10
Sql> select instr(‘ABCDEFGHICDJKMLCDHJF’,’CD’,-6,2) from dual; o/p: 3
NOTE: Always instr() returns position based on last two parameter but oracle server
always counts number of characters left side first position on words.
7.Lpad():
It will fills specified character on the left side of the given string. Here always second
parameters returns total length of the string this parameter must be a number.
Syntax: Lpad(columnname (or) ‘stringname’,total length of the string,’filled char);
Ex: sql>select lpad(‘ABCD’,10,’&’) from dual; o/p: &&&&&&ABCD
8.Rpad():
Sql>select rpad(‘ABCD’,10,’&’) from dual; o/p: ABCD&&&&&&
Sql>select rpad(ename,20,’-‘)||sal from emp;
Ans: rpad(ename,20,’-‘)||sal
smith----------------------------------------1800
allen------------------------------------------1250
king-------------------------------------------5000
9.Ltrim():
It is used to remove left side specified characters.
Syntax: Ltrim(columnname (or) ‘stringname’,{set of characters});
Sql>select Ltrim(‘SSMISSTHSS’,’S’) from dual; o/p: MISSTHSS
Sql>select job,Ltrim(job,’lsm’) from emp;

SQL MATERIAL
Ans: job ltrim(job,’CSM’)
CLERK LERK
SALESMAN ALESMAN
MANAGER ANAGER
10.Rtrim():
Sql>select Rtrim(‘SSMISSTHSS’,’S’) from dual; o/p: SSMISSTH
Sql>select job,Rtrim(job,’KNR’) from emp;
Ans: job ltrim(job,’KNR’)
CLERK CLE
SALESMAN SALESMA
MANAGER MANAGE
11.Trim(): (Oracle 8i introduced)
It is used to remove left and right side specified character and also leading and trailing
spaces.
Syntax: trim(‘character’ from columnname (or) ‘stringname’)
Sql>select trim(‘S’ from ‘SSMISSTHSS’) from dual; o/p: MISSTH
Sql>select trim(leading ‘S’ from ‘SSMISSTHSS’) from dual; o/p: MISSTHSS
Sql>select trim(trailing ‘S’ from ‘SSMISSTHSS’) from dual; o/p: SSMISSTH
NOTE: This function also used remove first letter last spaces. Trim set should have only
one character.
Sql>select length(trim(‘abcd’)) from dual;
Ans: 4

SQL MATERIAL
12.Translate(), Replace():
Translate() is used to replaces character by character where as Replace() is used to

replaces character by string (or) string by string.
Syntax: TRANSLATE( string1, string_to_replace, replacement_string )
Sql>select translate(‘ABCD’,’ACBD’,’@$#&’) from dual; o/p: @#$&
Sql>select translate(‘ABCDEF’,’FEDCBA’,’123456’) from dual; o/p: 654321
Ex: Sql>select replace(‘A B C’,’ ‘,’ ORACLE ’) from dual;
Ans: A ORACLE B ORACLE C
Sql>select job,replace(job,’SALESMAN’,’MARKETING’) from emp;
Ans: job replace(‘salesman’,’marketing’)
Clerk clerk
Salesman marketing
King king
Salesman marketing
Sql>select replace(‘SSMISSTHSS’,’S’) from dual; o/p: MITH
Q) Write a query to count number of repeating character from a string?
Sql>select length(‘ABBBC’)-length(replace(‘ABBBC’,’B’)) from dual; o/p: 3

SQL MATERIAL
Date Functions:
In oracle by default date format is ‘DD-MON-YY’. Oracle having following date

functions.
1.sysdate
2.add_months()
3.last_day()
4.next_day()
5.months_between()
1.sysdate: It returns current system date in oracle date format.
Sql>select sysdate from dual; o/p: 31-jul-14
2.add_months():
It is used to add (or) subtract number of months to specified date based on second
parameter.
Syntax: add_months(date,number)
Sql>select add_months(sysdate,1) from dual; o/p: 31-AUG-14
Sql>select add_months(sysdate,-1) from dual; o/p: 30-JUN-14
3.last_day(): It returns last day of the specified months.
Syntax: last_day(data)
Sql>select last_day(sysdate) from dual; o/p: 31-JUL-14
4.next_day():
It returns next day occurrence date from the specified date based on second parameter.
Sql>select next_day(sysdate,’SUN’) from dual; o/p: 03-AUG-14

SQL MATERIAL
5.mon_between():
This function returns number datatype it return number of months between two specified
dates. It date1>date2 then this function returns negative value.
Sql>select months_between(sysdate,hiredate) from emp;
months_between(sysdate,hiredate)
403.481436
401.386775
401.328987
Sql>select ename,round(months_between(sysdate,hiredate)) from emp;
Ename round(months_between(sysdate,hiredate))
SMITH 403
ALLEN 401
KING 409
Date Arithmetic:
1.date+number
2.date-number
3.date1+date2 (not possible)
4.date1-date2
Sql>select sysdate+1 from dual; o/p: 01-AUG-14
Sql>select sysdate-1 from dual; o/p: 30-JUL-14
Sql>select sysdate-sysdate from dual; o/p: 0

SQL MATERIAL
DATE CONVERSION FUNCTION: DD YY DAY
1.TO_CHAR() D YYYY DY
DDD HH Number MONTH Character

2.TO_DATE()
MM MI MON
1.TO_CHAR(): SS YEAR
It is used to convert date type into char type i.e it is used to convert date type into date
string.
Sql>select to_char(sysdate,’DD/MM/YY’) from dual;
o/p: 31/07/14
NOTE: to_char() character is an case sensitive . day-------thursday
Sql>select to_char(sysdate,’DAY’) from dual; dy---------thu
o/p: THURSDAY DY---------THU
Sql>select to_char(sysdate,’DD’) from dual; o/p: 31
D----6 (Day of the week sun-1,mon-2,tue-3…………)
DDD------212 (Day of the year) DD--------31 (Day of the month)
Sql>select to_char(sysdate,’DDSPTH’) from dual; o/p: THIRTY-FIRST
DDSPTH-----SP-spellout (thirty-first) DDTH--(31ST)
Sql>select to_char(sysdate,’HH:MI:SS’) from dual; o/p: 01:20:45
HH24:MI:SS----------13:20:45 --by default 12 hours display
Sql>select to_char(’23-jun-05’,’DD/MONTH/YY’) from dual;
ERROR: Invalid Number
NOTE: Whenever we are using to_char always first parameter must be date type.
Otherwise oracle server returns an error.

SQL MATERIAL
Sql>select to_char(sysdate,’DD/MONTH/YY’) from dual; o/p: 31/JULY/14
2.TO_DATE():
It is used to convert date string into date type (Oracle date format).
Sql>select to_date(‘15/june/04’) from dual;
o/p: 15-JUN-04 DD-MON-YY
Sql>select to_date(‘15/06/04’) from dual; num-char-num
ERROR: not a valid month 15/06/04
Sql>select to_date(‘15/06/04’,’DD/MM/YY’) from dual; o/p: 15-JUN-04
NOTE:
Whenever we are using to_date passed parameter return values match with default date
format return date type otherwise oracle side returns error. To overcame this problem use a
second parameter as same as first parameter format. Then oracle server only automatically
converts date string into date type.
Sql>select to_char(to_date(’15-JUN-05’),’DAY/FMMONTH/YY’) from dual;
o/p: WEDNESDAY/JUNE/05
Sql>select to_date(’08-jul-05’)+6 from dual; o/p: 14-JUL-05
Sql>select to_date(’08-07-05’,’DD-MM-YY’)+6 from dual; o/p: 14-JUL-05
NOTE: When to_char function if we are specifying full length format then oracle server
returns gaps if we passed string return values occupy less size than the default date format return
values to overcome this problem we must use fill mode(FM).
Q) Write a query to display emp who are joining in month DECEMBER from emp table
using to_char function?
Sql>select * from emp where to_char(hiredate,’MM’)=’12’;

SQL MATERIAL
MON=DEC MM=12 (working) MONTH=DECEMBER (not working)
Q) Write a query to display the emp who are joining in the year ‘81’ from emp table using
to_char function?
Sql>select * from emp where to_char(hiredate,’yy’)=’81’;
Sql>select hiredate,to_char(hiredate,’yyyy’) from emp;
NOTE:
Whenever we are passing date string into oracle date functions then oracle server
internally convert date string into date type that why we are not allowed to_date explicitly. In
this case passed parameter must be default date format otherwise oracle server returns an error.
Sql>select last_day(’12-06-05’) from dual;
ERROR: not a valid month
Sql>select last_day(to_date(’12-06-05’,’DD-MM-YY’) from dual; o/p: 30-JUN-05
Round,Truncate function used in date:
Whenever we are using round,truncate function date part can be changed based on time
portion. Whenever we are using round function oracle server internally check time portion≥12 or
noon. It time portion≥12 then one day added to the given date and also automatically time
portion set to zero.
Wherever we are using truncate also oracle server internally time portion set zero.
TESTING:
Sql> select to_char(round(sysdate),’DD/MM/YY HH24:MI:SS’) from dual;
01/08/2014 00:00:00
Sql> select to_char(trunc(sysdate),’DD/MM/YYYY HH24:MI:SS’) from dual;
31/07/2014 00:00:00

SQL MATERIAL
Q) Write a query to display the emp who are joining today from emp table?
Sql>insert into emp(empno,ename,hiredate)values(1,’narasimha’,sysdate);
Sql>select * from emp where hiredate=sysdate;
No Rows Selected
Sql>select * from emp where trunc(hiredate)=trunc(sysdate);
GROUP FUNCTION (or) AGGREGETING:
All this group function operator over group of data and return a single value.
1. max()
2. min()
3. avg()
4. sum()
5. count(*)
6. count(columnname)
1. max(): It returns maximum values within a column.
Sql>select max(sal) from emp; Ans: 5100
Sql>select max(hiredate) from emp; Ans: 23-MAY-87
Sql>select min(ename) from emp; Ans: WARD
NOTE: We are not allowed to use group function in where clause.
Sql>select * from emp where sal=max(sal);
2. min(): Sql>select min(sal) from emp; Ans: 850
Sql>select min(hiredate) from emp; Ans: 17-DEC-80

SQL MATERIAL
3. avg(): sql>select avg(sal) from emp; Ans: 2188.48114
Sql>select avg(comm) from emp; Ans: 550
NOTE: In oracle by default all group function ignores null values except count(*)
function.
Sql>select avg(nvl(comm)) from emp; Ans: 157.142857
4. Sum(): It returns total within a column.
Sql>select sum(sal) from emp; Ans: 29025
Sql>select sum(sal) from emp where deptno=10; Ans: 8750
5. Count (*): It count number of row in a table.
Sql>select count(*) from emp; Ans: 14
6. Count (columnname): In count number of not null values within a column.
Sql>select count(comm) from emp; Ans: 4
Sql>select count(distinct(deptno) from emp; Ans: 3
GROUP BY: This clause is used to divide similar data items into set of logical group.
This clause is used to select statement.
Syntax: select columnname from tablename group by columnname;
Q) Write a query to display no.of emp’s department wise from emp table using group by?
Sql>select deptno,count(*) from emp group by deptno;
DEPTNO COUNT(*)
10 3
20 5
30 6

SQL MATERIAL
Q) Write a query to display number of emp’s job wise from emp table using group by?
Sql>select job,count(*) from emp group by job;
JOB COUNT(*)
CLERK 4
MANAGER 3
ANALYST 2
SALESMAN 4
PRESIDENT 1
Q) Write a query to display no.of emp’s dept wise max(),min() of sal from emp table using
group by?
Sql>select deptno,max(sal),min(sal) from emp group by deptno;
DEPTNO MAX(SAL) MIN(SAL)
10 5000 1300
20 3000 800
30 2850 950
NOTE: We can also use group by clause without using group function.
Sql>select deptno from emp group by deptno; Ans: 10,20,30
RULES:
Sql>select deptno,max(sal),ename from emp group by deptno;
ERROR: Not a GROUP FUNCTION Expression
Other than group function columns specification after select those all columns must be
use after group by. Otherwise oracle server returns ERROR: not a group function expression.

SQL MATERIAL
Sql>select deptno,max(sal),ename from emp group by deptno,ename;
Sql>select deptno,job from emp group by deptno,job;
Ex: sql>select deptno from emp group by deptno,ename;
NOTE:
Whenever we are using group function with another column with using after select then
database server returns an error. To overcome this problems column use group by clause.
Step-1: sql>select sum(sal) from emp; Ans: 29025
Step-2: sql>select job,sum(sal) from emp;
ERROR: Not a SINGLE-GROUP Group Function
Solution: sql>select job,sum(sal) from emp group by job;
JOB SUM(SAL)
CLERK 4150
SALESMAN 5600
MANAGER 8275
ANALYST 6000
PRESIDENT 5000
Q) Write a query to display deptno,count(*) and also those dept having more than 4 emp’s
from emp table using group by clause?
Sql>select deptno,count(*) from emp group by deptno where count(*)>4;
ERROR: Sql Command not properly
Sql> select deptno,count(*) from emp group by deptno having count(*)>4;
Ans: Deptno- 20 30 count(*)- 5 6

SQL MATERIAL
Having:
In all database systems after GROUP BY clause we are not allowed to WHERE clause.
In place of this ANSI/ISO sql introduced using HAVING clause. Generally if you want to
restrict the rows in a table then we are using WHERE clause. Where as if we want to restrict the
groups then after we are using HAVING clause. Generally we are not allowed to use GROUP
function in WHERE clause. Where has in HAVING clause we can also use GROUP function.
Sql>select deptno,sum(sal) from emp group by deptno having sum(sal)>10000;
DEPTNO SUM(SAL)
20 10875
Q) Write a query to display year, number of emp’s in that year which more than one emp
was hide from emp table using group by clause?
Sql>select to_char(hiredate,’YYYY’) “YEAR”, count(*) from emp
group by to_char(hiredate,’YYYY’)having count(*)>1;
YEAR COUNT(*)
1982 2
1981 10
ORDER BY:
This clause is used to arrange data in shorting order along with ORDER BY clause oracle
provided two keywords ASC/DESC. By default ORDER BY clause having ASC order.
ORDER BY
ASC DESC
Syntax: select * from tablename order by columnname[asc/desc];

SQL MATERIAL
Sql>select sal from emp order by sal asc; [optional]
Sql>select sal from emp order by sal desc; [mandatory]
NOTE: we can also use column position on ORDER BY clause to improve performance.
Sql>select * from emp order by 5 desc; [columnnumber]
NOTE:
We can also use more than one column ORDER BY clause. In this based ORDER BY
clause arrange data in first column then value are same in first column those column second
column.
Sql>select ename,deptno,hiredate from emp order by deptno asc,hiredate desc;
Sql>select deptno,count(*) from emp where sal>1000 group by deptno having

count(*)>3 order by deptno desc;
DEPTNO COUNT(*)
30 5
20 4
Rollup,Cube:
Oracle 8i introduced rollup,cube. These clauses are used along with GROUP BY clause
only. This clauses are used to calculate sub total,grand total automatically.
Syntax: select col1,col2........ from emp tablename group by rollup(col1,col2……);
Select col1,col2........ from emp tablename group by cube(col1,col2……);
Sql>select deptno,job,sum(sal),count(*) from emp group by rollup(deptno,job);
Sql> select deptno,job,sum(sal),count(*) from emp group by cube(deptno,job);
ROLLUP is used to calculate subtotal based on a single column. If we want to calculate

subtotal.Grandtotal based on number of columns than we are using CUBE.

SQL MATERIAL
Chapter-5
JOINS
Joins are used to retrieve data from more than one table. If we are joining ‘N’ tables then
we are using N-1 joining condition oracle server having following types of joining. This 4 joins
are also called as 8i joins.
1. Equi join (or) Inner join
2. non-equi join
3. Self join
4. Outer join
ANSI joins (or) 9i joins:
1. Inner join
2. Left outer join
3. Right outer join
4. Full outer join
5. Natural outer join
We can also retrieve data from multiple tables without using joins. In this case server
internally uses cross join. Cross joins must default data. Cross join internal uses catigear product.
If we want to display accurate data from multiple tables then we must use explicit join.
Sql>select ename,sal,dname,loc from emp,dept;
1. Equi join (or) Inner join:
Based on equality condition we are retrieving data from multiple tables. Hare joining
conditional column must belongs to same datatype. When table having common columnname
then only we are using this join.

SQL MATERIAL
Syntax: select col1,col2……………..
From tablename1,tablename2
Where tabblename1.commom columnname=tablename2.common columnname;
Sql>select ename,sal,deptno,daname,loc
From emp,dept
Where emp.deptno=dept.deptno;
ERROR: column ambiguously defined
Sql>select ename,sal,dept.deptno,daname,loc
From emp,dept
NOTE:
To avoid feature ambiguously generally we are specified columnname along with

tablename using DOT operator after SELECT.
Using Alias name:
We can also create deferent names for the table in FROM clause of the joining condition.
Syntax: from tablename1 aliasname1,tablename2 aliasname2;
Sql>select ename,sal,d.deptno,dname,loc
From emp e,dept d
Where e.deptno=d.deptno;
NOTE:
Equi join always return matching rows only (here deptno 40 does not display if we are
using deptno also).

SQL MATERIAL
Q) Write a query to display we are working Chicago from emp,dept tables using joins?
Sql>select ename,loc from emp,dept

Only Chicago display:

Sql>select ename,loc from emp,dept
Where emp.deptno=dept.deptno
And loc=’CHICAGO’;
NOTE:
If we want to retrieve the rows after joining condition then we are using AND operator in
8i joins where has in 9i join we are using either ‘AND’ (or) ‘WHERE’.
Q) Write a query display dname,sum(sal) from emp,dept tables using equi joins?
Dname sum(sal)
Sql>select dname,sum(sal) from emp,dept
ACCOUNTING 8750
Where emp.deptno=dept.deptno group by dname;
RESEARCH 10875
Sql>select d.deptno,dname,sum(sal) from emp e,dept d SALES 9400
Where e.deptno=d.deptno group by d.deptno,dname;
DEPTNO DNAME SUM(SAL)

10 ACCOUNTING 8750
20 RESEARCH 10875
30 SALES 9400
SQL> select dname,sum(sal) from emp,dept
where emp.deptno=dept.deptno group by dname having sum(sal)>10000;
DNAME SUM(SAL)
RESEARCH 10875
where emp.deptno=dept.deptno group by dname having count(*)>3;

SQL MATERIAL
DNAME SUM(SAL)
RESEARCH 10875
SALES 9400
where emp.deptno=dept.deptno group by rollup(dname);
DNAME SUM(SAL)
ACCOUNTING 8750
RESEARCH 10875
SALES 9400
29025
2. Non-equi join:
Based on other than equality condition (<>,>,>=,<,<=,in,between……….) we can

retrieving multiple tables. This join is also called as “between……and” join. Generally table
does not have same common columns then only using this join.
SQL> select ename,sal,losal,hisal from emp,salgrade where sal between losal and hisal;
SQL> select ename,sal,losal,hisal from emp,salgrade where sal>=losal and sal<=hisal;
3. Self join:
Join in a table to itself is called as “self join”. Here joining conditional columns must
belong to same datatype. Generally if we want to compare different column in a single table (or)
if we want to single column values then we are using self join.
Syntax: from tablename aliasname1,tablename aliasname2
Whenever we are using self join we must create an alias name in from clause this alias
names must by different name. Otherwise data oracle server returns ambiguously error. This
reference name used by data based on server in query execution time only.
Q) Write a query to display the working has same job as SCOTT from emp table using self
join?

SQL MATERIAL
SQL> select e2.ename,e1.job from emp e1,emp e2

where e1.job=e2.job and e1.ename='SCOTT';
ENAME JOB e1 e2
Ename Job Ename Job
SCOTT ANALYST
FORD ANALYST SCOTT Z ? Z
Q) Write a query to display ename,manager from emp table using self join?
Sql>select e1.ename “Emaployee”,e2.ename “Manager” from emp e1,emp e2
Where e1.mgr=e2.empno;
Sql>select e1.empno,e1.ename,e1.mgr,e2.ename “Manager” from emp e1,emp e2
Where e1.mgr=e2.empno; Empno Ename Mgr Empno Ename Mgr
7369 SMITH 7902 7902
Q) Write a query to display the emp and their manager in 10th department from emp table
using self join?
SQL>select e1.empno,e1.ename,e1.mgr,e1.deptno,e2.ename "Manager" from emp e1,

emp e2 Where e1.mgr=e2.empno AND e1.deptno=10;
EMPNO ENAME MGR DEPTNO Manager

7934 MILLER 7782 10 CLARK
7782 CLARK 7839 10 KING
Sql>select e1.ename "Employee",e1.sal,e2.ename "Manager",e2.sal
from emp e1,emp e2
where e1.mgr=e2.empno and e1.sal>e2.sal;
Employee SAL Manager SAL
FORD 3000 JONES 2975
SCOTT 3000 JONES 2975

SQL MATERIAL
Q) Display the emp who are before joining their MGR from emp using self join?
SQL> select e1.ename "Employee",e1.hiredate,e2.ename "Manager",e2.hiredate
from emp e1,emp e2 where e1.mgr=e2.empno and e1.hiredate<e2.hiredate;
Q) Write a query to display emp and their mgr who are working in the sales emp,dept
tables using self join?
Sql>select e1.ename "Employee",e1.sal,e2.ename "Manager",e2.sal from emp e1,dept

d,emp e2 Where e1.mgr=e2.empno AND e1.deptno=d.deptno and d.dname='SALES';
Q) Write a query to display the emp who are getting same salary in different dept from
emp table using self join?
Sql>select e1.ename,e1.sal,e1.deptno,e2.sal,e2.ename,e2.deptno from emp e1,emp e2
Where e1.sal=e2.sal and e1.deptno<>e1.deptno;
Q) Write a query to display emp who joining same month from emp table using self join?
Sql> select e1.ename,e1.hiredate,e2.ename,e2.hiredate from emp e1,emp e2

where to_char(e1.hiredate,'MON')=to_char(e2.hiredate,'MON')
and e1.empno<e2.empno;
4. Outer join:
This join is used to retrieve all rows from one table and matching rows for another table.
Generally when we are using equi join database server returns matching rows only. In oracle if
you want retrieve non-matching rows also then we are using outer join (+) operator in equi join
this is called “8i outer join”.
NOTE: In this case join operator can be used only one side at a time in joining condition
of the equi join.
Sql>select ename,sal,d.deptno,dname,loc from emp e,dept d

+ Not allowed
Where e.deptno(+)=d.deptno;
(+) Allowed

SQL MATERIAL
Sql>select ename,sal,d.deptno,dname,loc from emp e,dept d

+ Allowed
Where e.deptno=+d.deptno;
(+) Not Allowed
Olden days both side using (+) operator:
Sql> select ename,sal,d.deptno,dname,loc from emp e,dept d
Where e.deptno(+)=d.deptno; Left only use (+)
UNION Right only use +
select ename,sal,d.deptno,dname,loc from emp e,dept d
Where e.deptno=+d.deptno;
9i joins (or) ANSI joins:
1. Inner join
2. Left outer join
3. Right outer join
4. Full outer join
5. Natural join
1. Inner join:
This joins returns matching rows only. When table having common columnname then
only we are using this join this join performance very high compare to 8i equi join. Here joining
conditional column same datatype.
Sql>select ename,sal,d.deptno,dname,loc from emp e join dept d
On e.deptno=d.deptno;
Q) Write a query to display the emp who are working in the loc Chicago from emp,dept
using 9i equi join?

SQL MATERIAL
Sql>select ename,loc from emp join dept on emp.deptno=dept.deptno where

loc='CHICAGO';
WHERE in this place AND also using.
Using clause:
ANSI join we are also use USING clause in place of ON clause. USING clause
performance very high compare to ON clause and also when we are using USING clause data we
returns database server returns one time only.
Syntax: select * from tablename1 join tablename2 USING(common columnname);
Sql>create table z1(a varchar2(5),b varchar2(5),c varchar2(5));
Sql>insert into z1 values(‘x’,’y’,’z’);
SQL> select * from z1;
A B C
x y z
Sql>create table z2(a varchar2(5),b varchar2(5));
Sql>insert into z1 values(‘x’,’y’);

A B
x y
Sql>select * from z1,z2;
A B C A B
x y z x y
Sql>select * from z1,z2 where z1.a=z2.a and z1.b=z2.b;
Sql> select * from z1 join z2 on(z1.a=z2.a and z1.b=z2.b);
A B C A B
x y z x y
SQL MATERIAL
SQL> select * from z1 join z2 using(a,b);

A B C
x y z
NOTE:
We are using USING clause we are not allowed to use aliasname and joining condition.
Sql>select ename,sal,deptno,dname,loc from emp e join dept d using(deptno);
ENAME SAL DEPTNO DNAME LOC

SMITH 800 20 RESEARCH DALLAS
ALLEN 1600 30 SALES CHICAGO
WARD 1250 30 SALES CHICAGO
---------- ------- ---- --------------- ----------------
2. Left outer join:
This join all rows from left side table and matching rows from right side table and also
return null values in place of non-matching rows display another table.
SQL> insert into z1 values('p','q','r');
A B C
x y z
p q r
SQL> insert into z2 values('s','t');
A B
x y
s t

SQL MATERIAL
SQL> select * from z1 left outer join z2 on(z1.a=z2.a and z1.b=z2.b);

A B C A B
x y z x y
p q r
3. Right outer join:
This join always returns all rows from right side table and matching rows from left side
table and also returns null values in place of non-matching rows.
Sql>select * from z1 right outer join z2 on z1.a=z2.a and z1.b=z2.b;
A B C A B
x y z x y
s t
4. Full outer join:
This join always returns all rows from right side table and matching rows left side table
and also returns null values in place of non-matching rows.
SQL> select * from z1 full outer join z2 on z1.a=z2.a and z1.b=z2.b;
A B C A B
x y z x y
p q r
s t
Sql> select ename,sal,d.deptno,loc,dname from emp e natural join dept d;
ERROR: Column used in NATURAL join cannot have qualifier
5. Natural join:
Natural join return matching rows only in natural we are not allowed to joining condition
When ever table having common column then only we are using this join. This join internally
uses USING clauses. That why this join always returns common column one time only and also
we are not allowed using aliasname and joining conditional column.

SQL MATERIAL
Sql> select ename,sal,deptno,loc,dname from emp e natural join dept d;

SQL> select * from z1 natural join z2;
A B C
x y z
Cross join:
Sql>select ename,sal,dname,loc from emp cross join dept;
Joining more than two tables:
8i joins 9i joins/ANSI joins
Syntax: Syntax:
Select col1,col2……………… Select col1,col2……………..
From table1,table2,table3 From table1 join table2
Where table1.commoncol=table2.commoncol On table1.commoncol=table2.commoncol
And table2.commoncol=table3.commoncol; Join table3
On table2.commoncol=table3.commoncol;

SQL MATERIAL
Chapter-6
CONSTRAINTS
Constraints are used prevents (or) invalid data entry into out table. Generally constraints
are created on table columns oracle server having following type of constraints.
1. Not null
2. Unique
3. Primary key
4. Foreign key
5. Check
All the above constraints are defining two levels
1. Column Level
2. Table Level
1. Column Level:
In this method we are defining constraining on individual columns that is whenever we

defining column then only we are using constraint type.
Syntax: create table tablename(col1 datatype(size) constraints type,col2

datatype(size) constraint type……………);
2. Table Level:
In this method we are defining constraints on group of column that is first we are defining
all the columns and last only we are specifying constraint type along with group of column.
Syntax: create table tablename(col1 datatype(size),col2 datatype(size)…………….
constraint type(col1,col2………………));

SQL MATERIAL
1. Not null:
Not null constraints does not support table level this constrains does not accept null
values but it accept duplicate value.
Sql>Create table c1(sno number(5) not null,name varchar2(15));
SQL> insert into c1 values(null,'a');
ERROR: ORA-01400: cannot insert NULL into SNO
2. Unique:
This constrains defining two level that is does not accept duplicate values but if accept
null values. Whenever we are creating unique constrains oracle server automatically create on
btree index on those column.
Column level: Sql>create table c2(sno number(5) unique,name varchar2(15));

SQL> insert into c2 values(1,'abc');
SQL> insert into c2 values(2,'xyz');
SQL> select * from c2;
SNO NAME
1 abc
2 xyz
Table level: sql> create table c2(sno number(5),name varchar2(15),unique(sno,name));
SQL> select * from c3;
SNO NAME
1 abc
2 xyz
SQL> insert into c3 values(2,'xyz');
ORA-00001: unique constraint violated
1 abc, 2 xyz this is not allowed.

SQL MATERIAL
3. Primary key:
Primary key uniquely identifier record in a table and also where can be only primary in a
table and also primary key does not accept null values and duplicate values. Whenever we are
creating primary key internally oracle server create and btree index on those column.
Column level: create table c4(sno number(5) primary key,name varchar2(15));
Table level: create table c5(sno number(5),name varchar2(15),primary key(sno,name));
This is a composite primary key i.e it is combination column as single primary key.
4. Foreign key:
We want to establish relationship between tables then we are using referential integrity
constraints foreign key. One table foreign key must belong to another table primary key. Here
these two columns must belong to same datatype. Always foreign key values based on primary
key only.
Generally primary key does not excepts duplicate,null values where has foreign key
accepts duplicates,null value.
Column level (References):
Syntax: create table tablename(col1 datatype(size),references

mastertablename(primary key columnname),col2 datatype(size)……);
Sql>create table f1(sno number(5) primary key);
Sql>create table ff1(sno number(5) references f1);
Sql>create table b1(a number(5) references f1(sno));
Table level (Foreign key,References):
Syntax: create table tablename(col1 datatype(size),col2 datatype(size)…….
Foreign key(col1,col2………),reference mastertablename(pk colname);
Sql>create table f2(sno number(5),name varchar2(15) primary key(sno,name));

SQL MATERIAL
Sql>create table ff2(sno number(5),name varchar2(15)col3 number(5),foreign

key(sno,name),references f2);
SNO SNO
Primary key 1 Foreign key

1
2
On delete cascade
2
On delete set null
3
Rule:
Whenever we are establishing relationship between two tables oracle server violates
following two rules.
1. Deletion in master table
2. Insertion in child table
1. Deletion in master table:
if we are try to delete a record in master table than oracle server returns a error ORA-
2292: to reference this problem first we are Child deleting children table records
Master
in a table. Than only we are deleting those record master table otherwise
used on delete cascade clause.
On delete cascade: References
Whenever we are using this clause in child table record oracle server automatically delete
appropriate master,child table record on those table.
Syntax: create table tablename(col1 datatype(size) references

mastertablename(primary key columnname) on delete cascade………);
Ex: sql>create table mas(sno number(5) primary key);
Sql>insert into mas(1);

SQL MATERIAL
Sql>select * from mas; o/p: sno:1 2 3 4……………..

Sql>create table child(sno number(5) references mas on delete cascade);
Sql>insert into mas(1);
Sql>select * from mas; o/p: sno:1 1 1 2 2 2 3 3 3 4……………..
Testing (deletion in master):
Sql>delete from mas where sno=1;
Sql>select * from child;
On delete set null:
Whenever we are using this clause in child table if we are deleting primary key value in
master table then automatically foreign key value set to null in child table.
Syntax: create table tablename(col1 datatype(size) references

mastertablename(primary colname) on delete set null………);
2. Insertion in child table:
If we try to insert other than primary key value into foreign key then oracle server returns
on ERROR: ora-2291: Because always foreign key values primary key values only.
5. Check:
This constraint is used to define logical condition according to our business rules.
Syntax: create table tablename(col1 datatype(size) check(logical condition ),col2

datatype(size),………………..);
Sql>create table nr(name varchar2(15) check(name=upper(name));

Sql>insert into nr values(‘NARASIMHA’);
Name
NARASIMHA
SQL> insert into nr values('narasimha');
ERROR: ORA-02290: check constraint violated

SQL MATERIAL
Assign user defined names to constraints:

In all database systems whenever we are creating constraint internally database server
automatically generates unique identification number in oracle. Whenever we are creating
constraint then oracle server automatically generates an unique identification number in the
format of “SYS_CN”. This is called predefined constraint name we can also create our own
name in place of SYS_CN using constraint keyword at the time of table creation. This is called
user defined constraint name.
Syntax: constraint userconstraintname constrainttype;
Ex: sql>create table test(sno number(10) primary key);

SQL> insert into test values(1);
SQL> insert into test values(1);
User_Constraints User_Constraints_columns
ERROR: ORA-
00001:unique constraint(SYSTEM.SYS_C004274)violated
Sql>create table test1(sno number(10) constraint p_abc primary key);
Sql> insert into test1 values(1);
Sql>insert into test1 values(1);
ERROR:ORA-00001: unique constraint (SYSTEM.P_ABC) violated
Whenever we are installing oracle server automatically creating read only table those
read only table are also called as data dictionary.
Oracle Database
Owner constraint_name constraint_type Constraint_name column_name

SQL MATERIAL
All constraints information stored under in user_constraints data dictionary.
Sql>desc user_constraints;
Sql>select constraint_name,constraint_type from user_constraints
where table_name=’EMP;
cons_name con_type
pk_emp p
fk_deptno R
NOTE: If we want to view column names along with constraint name then we are using
user_constraint_columns.
Sql>select constraint_name,column_name from user_constraints_column
where table_name=’EMP;
cons_name col_name
pk_dept deptno
fk_emp empno
NOTE: If we want view logical condition of the check constraints then we are using
search_condition property user_constraints data dictionary.
Sql>select search_condition from user_constraints where table_name=’g5’;
search_condition
name=upper(name)
NOTE: All column information stored under user_tab_column data dictionary.
Ex: sql>desc user_tab_columns
Sql>select column_name from user_tab_col where table_name=’EMP’;
Column_name
Empno
Ename
-----------

SQL MATERIAL
Q) Write a query to display no.of columns numbers from emp using data dictionary?
Sql>select count(*) from user_tab_column where table_name=’EMP’;
Count(*)
Adding (or) dropping constructions on existing table:
Using ALTER we can also add (or) remove existing table.
NOTE: If we want add constraints into existing table existing columns then we are using
table level syntax method.
X1
SNO
Sql>alter table x1 add primary key(sno);
NOTE: If we want to add a new column along with constraints then we are using column
level syntax method.
X2
N
S
A
N
M
O
E
Sql>alter table x1 add name varchar(15) unique;
Sql>alter table x2 add foreign key(sno) reference x1;
NOTE: If we want add not null constraints existing table columns then we are using alter
with modify.
Syntax: alter table tablename modify colname not null;

SQL MATERIAL
Sql>alter table x3 modify sno not null;
Sql>alter table x3 add name varchar2(15) not null;
Dropping constraints:
Method-1: sql>alter table tablename drop constraintname;
Method-2: sql>alter table tablename drop primary key;
Sql>alter table tablename drop unique(col1,col2,……….);
Ex: sql>create table test(sno number(5) primary key);
Sql>alter table test drop primary key;
Sql>alter table x1 drop primary key;
ERROR: This unique/primary key is reference by some foreign key’s.
NOTE: If we want drop primary key along with foreign key then we are using cascade
clause along with alter drop.
Syntax: alter table tablename drop primary key cascade;
Sql>alter table x1 drop primary key cascade;
Ex: sql>alter table x3 drop constraint sys_coo5358
NOTE: Oracle also support default constraints for a column
Syntax: columnname datatype(size) default actual value
Sql>create table test(sno number(5) default 10,name varchar(15));
SQL> insert into test values(1,'A');
SQL> insert into test values(2,'B');
SQL> insert into test(name) values('C');

SQL MATERIAL
SQL> select * from test;
SNO NAME
1 A
2 B
10 C
Ex: SQL> create table dept1 as select * from dept;
SQL> create table emp1 as select * from emp;
SQL> alter table dept add primary key(deptno);
SQL> alter table emp add primary key(empno);
SQL> alter table emp add foreign key(deptno) references dept(deptno) on delete cascade;

SQL MATERIAL
Chapter-7
SUBQUERY
Query within another query is called nested query (or) subquery. Generally child query
are called subquery. Subqueries are used to retrieve data from single and multiple tables by using
more than one step process. All database systems having two type of subqueries.
1. Non-correlated
2. Correlated
In non-correlated subquery child query is executes first then only parent query executed.
Where as in correlated subquery parent query executed first then only child query executed.
1. Non-correlated:
1. Single rows subquery
2. Multiple rows subquery
3. Multiple column subquery
4. Inline view (or) subquery used in from clause.
Q) Write a query to display the emp’s who are getting more than the average salary from
emp table?
Sql>select * from emp where sal>(select avg(sal) from emp);
This are single row subquery because there child query returns single value. In single row
subquery we are using =,<,<=,>,>=,<> operator.
Execution:
Step-1:- sql>select avg(sal) from emp;
AVG(SAL)
2073.21429
SQL MATERIAL
Step-2:- Sql>select * from emp where sal>2073.21429

Child query: A query which provides value to anther query is called “CHILD QUERY”.
Parent query: A query which receives values from another query is called “PARENT QUERY”.
Q) Write a query to display the emp’s who are working salesman department from
dept,emp table using subquery?
Sql>select * from emp where deptno=(select deptno from dept where dname=’SALES’);
NOTE:
Generally using subquery we are not allowed to display child query column into parent
query to overcome this problem we must use join concept.
SQL> select ename,dname from emp e,dept d where e.deptno=d.deptno and

d.deptno=(select deptno from dept where dname='SALES');
ENAME DNAME
ALLEN SALES
WARD SALES
----------- -----------
Q) Write a query to display the emp’s who are working SMITH dept table.
Sql>select * from emp where deptno=(select deptno from emp where ename=’SMITH’);
Q) Write a query to display senior most emp details from emp table?
Sql>select * from emp where hiredate=(select min(hiredate) from emp);
EMPNO ENAME JOB MGR HIREDATE SAL COMM DEPTNO
7369 SMITH CLERK 7902 17-DEC-80 800 20
Q) Write a query to display the emp who are working under BLAKE from emp table using
empno,mgr?
Sql>select * from emp where mgr=(select empno from emp where ename=’BLAKE’);
7499 ALLEN SALESMAN 7698 20-FEB-81 1600 300 30

SQL MATERIAL
7521 WARD SALESMAN 7698 22-FEB-81 1250 500 30
7654 MARTIN SALESMAN 7698 28-SEP-81 1250 1400 30
7844 TURNER SALESMAN 7698 08-SEP-81 1500 0 30
7900 JAMES CLERK 7698 03-DEC-81 950 30
Q) Write a query to display lowest average salary job from emp table?
Sql>select job,avg(sal) from emp group by job having avg(sal)=(select min(avg(sal))

from emp);
ERROR: ORA-00978: nested group function without GROUP BY
NOTE: Whenever child query contains nested group function then we must using group
by clause.
Sql>select job,avg(sal) from emp group by job having avg(sal)=(select min(avg(sal))

from emp group by job);
JOB AVG(SAL)
CLERK 1037.5
Sql>select deptno,min(sal) from emp group by deptno having min(sal)>(select min(sal)
from emp where deptno=20);
DEPTNO MIN(SAL)
30 950
10 1300
Q) Write a query to display the emp delete who are getting maximum sal in each dept from
emp table?
Sql>select * from emp where sal=(select max(sal) from emp group by deptno);
ERROR: ORA-01427: single-row subquery returns more than one row
This are multiple row subquery because here child query returns multiple value in
multiple rows subquery we are using IN,ALL,ANY operator.
NOTE:

SQL MATERIAL
We can also use IN operator in single row subquery.
Sql>select * from emp where sal in(select max(sal) from emp group by deptno);
Sql> select deptno,sal from emp where sal in(select max(sal) from emp group by deptno);
Q) Write a query to display the emp’s who are working in ACCORDING,SALES

department using emp,dept table?
Sql>select * from emp where deptno in(select deptno from dept where
dname=’ACCORDING’ or dname=’SALES’);
TON-N Analysis:
1. Inline view
2. Rownum
1. Inline view: (Oracle 7.2 introduced)
Generally we are not allowed to use order by clause in child query to overcome this
problem oracle 7.2 introduced subqueries in from clause this is called as “INLINE IEW”.
Syntax: select * from (subquery)
In inline views we are using subquery in place tablename in parent query this subquery
execute to prier executed.
2. Rownum:
Rownum is a pseudo column which behavior table column. This pseudo used in all table.
It cannot perform DML operators. Rownum automatically assign numbers to each row in a table
at the time of selection. By default rownum having temporarily values.
Sql>select rownum,ename from emp;
ROWNUM ENAME
1 SMITH
2 ALLEN

SQL MATERIAL
--- -----------
Sql>select rownum,ename from emp where deptno=10;
ROWNUM ENAME
1 CLARK
2 KING
3 MILLER
Q) Write a query to display first row from emp table using rownum?
Sql>select * from emp where rownum=1;

7369 SMITH CLERK 7902 17-DEC-80 800 20
Q) Write a query to display second row from emp table using rownum?
Sql>select * from emp where rownum=2;
No Row Selected.
NOTE: Rownum does not work with more than one positive integer. Then is it work’s
with <,<= operators.
Q) Write a query to display first 5 row from emp table using rownum?
Sql>select * from emp where rownum<=5;

7369 SMITH CLERK 7902 17-DEC-80 800 20
7499 ALLEN SALES 7698 20-FEB-81 1600 300 30
------- ----------- --------- ------- -------------- ------- ------- ----
Q) Write a query to display first 5 highest salary from emp table using rownum?
SQL> select ename,job,sal,deptno from (select ename,job,sal,deptno from emp order by
sal desc) where rownum<=5;
ENAME JOB SAL DEPTNO
KING PRESIDENT 5000 10
SCOTT ANALYST 3000 20

SQL MATERIAL
---------- ------------------ -------- ----
Q) Write a query to display 5th highest salary employee from emp table using rownum?
SQL> select ename,job,sal,deptno from (select ename,job,sal,deptno from emp order by

sal desc) where rownum<=5
minus
select ename,job,sal,deptno from (select ename,job,sal,deptno from emp order by
sal desc) where rownum<=4;
BLAKE MANAGER 2850 30
Q) Write a query to display second row from emp table assigning rownum?
Sql>select * from emp where rownum<=2
Minus
select * from emp where rownum<=1;
7499 ALLEN SALES 7698 20-FEB-81 1600 300 30
Q) Write a query to display last two rows from emp table using rownum?
Sql>select * from emp
Minus
Select * from emp where rownum<=(select count(*) -2 from emp);
7902 FORD ANALYST 7566 03-DEC-81 3000 20
7934 MILLER CLERK 7782 23-JAN-82 1300 10
Q) Write a query to display the rows between 1 and 5 from emp table using rownum?
Sql>select * from emp where rownum between 1 and 5;
7369 SMITH CLERK 7902 17-DEC-80 800 20
------ ----------- ------------ ------- -------------- ------- ---------- ----

SQL MATERIAL
Q) Write a query to display the rows between 4 and 7 from emp table using rownum?
Sql>select * from emp where rownum<=7
Minus
Select * from emp where rownum<=4;
7369 MARTIN CLERK 7902 17-DEC-80 800 20
7499 CLARK SALESMAN 7698 20-FEB-81 1600 300 30
7654 BLAKE SALESMAN 7698 28-SEP-81 1250 1400 30
Sql>select * from emp where rownum>=1;

Rows selected
Sql>select * from emp where rownum>1;
No rows selected
NOTE:
Wherever we are using aliasname for rownum in inline view that aliasname works with
all sql operators.
Q) Write a query to display 5th row from emp table using rownum aliasname?
Sql> select * from (select rownum r,ename,sal from emp) where r=5;
R ENAME SAL
5 MARTIN 1250
Sql> select * from (select rownum r,ename,sal from emp) where r in(2,3,5,6,10,14);
R ENAME SAL
2 ALLEN 1600
-- ---------- -------
14 MILLER 1300
Sql>select * from (select rownum r,emp.* from emp) where r=5;
R EMPNO ENAME JOB MGR HIREDATE SAL COMM DEPTNO

SQL MATERIAL
5 7654 MARTIN SALESMAN 7698 28-SEP-81 1250 1400 30

Q) Write a query to display first and last row from emp table using rownum aliasname?
Sql> select * from (select rownum r,ename,sal from emp) where r=1 or r=(select count(*)
from emp);
R ENAME SAL
1 SMITH 800
14 MILLER 1300
Q) Write a query to display even no.of rows from emp table using rownum aliasname?
Sql> select * from (select rownum r,ename,sal from emp) where mod(r,2)=0;
R ENAME SAL odd number mod(r,2)=1

2 ALLEN 1600
4 JONES 2975
--- ----------- --------
12 JAMES 950
14 MILLER 1300
Q) Write a query to display 5th highest salary from emp table using rownum aliasname?
Sql>select * from (select rownum r,sal,ename from (select * from emp order by sal desc))
where r=5;
R SAL ENAME
5 2850 BLAKE
Analytical Function used in inline views: (Oracle 8i introduced)
Analytical functions also similar to GROUP functions and also these functions display
each row in a group because these functions are calculate results number of rows wise.
1. row_number()
2. rank()

SQL MATERIAL
3. dense_rank()
These three analytical functions automatically assigns numbers to each row in table either
group wise (or) rows wise in a table.
Syntax: analyticalfunctionname() optional clause
over(partition by colname order by colname[asc/desc];
ROW_NUMBER() analytical function assigns different rank numbers when values same
where as RANK(),DENSE_RANK() analytical function assign same rank number then value are
same and also RANK() scrip next consecutive rank number where has DENSE_RANK() does
not scripts next consecutive number.
ROW_NUMBER():
DEPTNO ENAME SAL R

20 SMITH 3000 1
20 SCOTT 3000 2
20 ADAMS 1200 3
RANK():
DEPTNO ENAME SAL R
20 SMITH 3000 1
20 SCOTT 3000 1
20 ADAMS 1200 3
DENSE_RANK():
DEPTNO ENAME SAL R
20 SMITH 3000 1
20 SCOTT 3000 1
20 ADAMS 1200 2
Sql>select * from (select deptno,ename,sal,row_number() over(partition by deptno
order by sal desc)r from emp) where r<=5; RANK()

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280
DENSE_RANK() alsoPage
use 76
SQL MATERIAL
DEPTNO ENAME SAL R
10 KING 5000 1
10 CLARK 2450 2
10 MILLER 1300 3
20 SCOTT 3000 1
----- ------------ ------- ----
20 SMITH 800 5
----- ----------- ------ ----
30 MARTIN 1250 5
Q) Write a query to display 2nd highest salary employee in each department from emp table
using analytical function?
Sql> select * from (select deptno,ename,sal,dense_rank() over(partition by deptno order

by sal desc)r from emp) where r=2;
DEPTNO ENAME SAL R
10 CLARK 2450 2
20 JONES 2975 2
30 ALLEN 1600 2
Q) Write a query to display 5th highest salary employee from emp table using analytical
function?
Sql> select * from (select deptno,ename,sal,dense_rank() over(order by sal desc)r

from emp) where r=5;
DEPTNO ENAME SAL R
10 CLARK 2450 5
Q) Write a query to display 2nd row from emp using analytical function?
SQL> select * from (select ename,sal,dense_rank() over(order by rowid)r from emp)

where r=2;

SQL MATERIAL
ENAME SAL R
ALLEN 1600 2
NOTE:
We are also use analytical function without use inline view always analytical function
returns row wise group data.
LEAD(),LAG():
These two analytical functions compare next row values and returns into the current row.
Syntax: lead(expression,offset,default value) over(partition by colname order by

colname [asc/desc]);
Here offset returns positive integer if we are specifying 1 it returns next row immediate.
Sql>select deptno,ename,sal,lead(sal,1,0) over(partition by deptno

order by sal desc) lead_sal from emp;
DEPTNO ENAME SAL LEAD_SAL
10 KING 5000 2450
10 CLARK 2450 1300
10 MILLER 1300 0
NOTE:
Lag() analytical function returns previous value into it is also same syntax as lead
function but offset returns previous row.
Sql>select deptno,ename,sal,lag(sal,1,0) over(partition by deptno order by sal desc)

lag_sal from emp;
DEPTNO ENAME SAL LAG_SAL
10 KING 5000 0
10 CLARK 2450 5000
10 MILLER 1300 2450
FIRST_VALUE:

SQL MATERIAL
First value analytical function returns first value in each group. This function excepts
only one parameter.
Sql> select deptno,ename,sal,first_value(sal) over(partition by deptno order by sal desc)

first_sal from emp;
DEPTNO ENAME SAL FIRST_SAL
10 KING 5000 5000
10 CLARK 2450 5000
10 MILLER 1300 5000
Q) Write query to display year number of employees in that year and in which year more
than one employee hide?
SQL>select to_char(hiredate,'yyyy') "YEAR",count(*) from emp group by

to_char(hiredate,'yyyy');
YEAR COUNT(*)
1980 1
1983 1
1982 2
1981 10
SQL>select to_char(hiredate,'yyyy') "YEAR",count(*) from emp group by

to_char(hiredate,'yyyy') having count(*)>1;
YEAR COUNT(*)
1982 2
1981 10
Q) Write a query to display duplicate data from a table?
SQL> select deptno,count(*) from emp group by deptno having count(*)>1;
DEPTNO COUNT(*)
30 6

SQL MATERIAL
20 5
10 3
ROWID:
Rowid is a pseudo column it’s behaviors like a table column pseudo column are not
stored in tables. Whenever we are inserting data oracle server automatically generates unique
identifier number identifying each row in a table in hexadecimal format this is called as
“ROWID” (or) “ROWADDRESS”.
Generally rownum having temporarily values where as rowid having fixed value. Using
rowid we are identifying a record uniquely in a table. Using rowid we can also data retrieve very
fastly from the database.
Sql>select rownum,rowid,ename from emp;
ROWNUM ROWID ENAME
1 AAADVOAABAAAIGCAAA SMITH
2 AAADVOAABAAAIGCAAB ALLEN
--- --------------------------------------- -----------
14 AAADVOAABAAAIGCAAN MILLER
SQL> select rownum,rowid,ename from emp where deptno=10;

ROWNUM ROWID ENAME
1 AAADVOAABAAAIGCAAG CLARK
2 AAADVOAABAAAIGCAAI KING
3 AAADVOAABAAAIGCAAN MILLER
NOTE: We can also use max(),min() function in rowid.

Sql>select min(rowid) from emp;
Sql>select max(rowid) from emp;
In subquery we can also use rowid to delete to subquery duplicate rows in a table. i.e if
you want to delete rows except one row each.

SQL MATERIAL
Q) Write a query to display 1st record from the emp using rowid?
SQL> select * from emp where rowid in(select min(rowid) from emp);
7369 SMITH CLERK 7902 17-DEC-80 800 20
Q) Write a query to display last record from the emp using rowid?
SQL> select * from emp where rowid in(select max(rowid) from emp);
7934 MILLER CLERK 7782 23-JAN-82 1300 10
Q) Write a query to display 2nd record from emp table using analytical function?
Sql>select * from (select empno,ename,sal,dense_rank() over(order by rowid) r
from emp) where r=2;

EMPNO ENAME SAL R
7499 ALLEN 1600 2
Q) Write a query to display each dept from the 2nd record from emp table using analytical
function?
SQL>select * from (select deptno,ename,sal,dense_rank() over(partition by deptno

order by rowid) r from emp) where r=2;
DEPTNO ENAME SAL R
10 KING 5000 2
20 JONES 2975 2
30 WARD 1250 2
Q) Write a query to display duplicate data from the table?
SQL> select deptno,count(*) from emp group by deptno having count(*)>1;
Q) Write a query to delete duplicate rows except one row in each group from table using
rowid?

SQL MATERIAL
SQL> delete from test where rowid not in(select max(rowid) from test group by sno);
4 rows deleted.
Sql>select * from test;
MULTIPLE COLUMN SUBQUERY:
We can also compare multiple column of the child query with the multiple column of the
parent query. This type of query is “MULTIPLE COLUMN SUBQUERY”. In multiple column
subqueries we must specify parent query when conditionals with in parenthesis.
Syntax: select * from tablename where(col1,col2……….)
In(select col1,col2…….. from tablename where condition);
Q) Write a query to display the emp’s who are job,mgr match with the job,mgr of the
emp’s SCOTT from emp table?
Sql>select * from emp where (job,mgr) in(select job,mgr from emp

where ename='SCOTT');
7902 FORD ANALYST 7566 03-DEC-81 3000 20
7788 SCOTT ANALYST 7566 09-DEC-82 3000 20
Q) Write a query to display senior most emp’s from each job from emp table using
multiple column subquery?
Sql>select deptno,sal,ename,job,hiredate from emp where(job,hiredate) in(select

job,min(hiredate) from emp group by job);
DEPTNO SAL ENAME JOB HIREDATE
20 800 SMITH CLERK 17-DEC-80
30 1600 ALLEN SALESMAN 20-FEB-81
10 5000 KING PRESIDENT 17-NOV-81
20 2975 JONES MANAGER 02-APR-81
20 3000 FORD ANALYST 03-DEC-81

SQL MATERIAL
Q) Write a query to display the emp’s who are getting maximum sal in each dept from emp
table using multiple row subquery?
Sql>select deptno,ename,sal from emp where sal in(select max(sal) from emp
group by deptno);
DEPTNO ENAME SAL
30 BLAKE 2850
20 FORD 3000
20 SCOTT 3000
10 KING 5000
Q) Write a query to display the emp’s who are getting maximum sal in each dept from emp
table using multiple row subquery?
Sql>select deptno,ename,sal from emp where (deptno,sal) in(select deptno,max(sal)

from emp group by deptno);
DEPTNO ENAME SAL
30 BLAKE 2850
20 SCOTT 3000
10 KING 5000
Q) Write a query to display ename,dname,sal of the emp who sal,comm match with
sal,comm of the emp’s working in location “DALLAS” from emp,dept tables?
SQL>select dname,ename,sal from emp e,dept d where e.deptno=d.deptno
and (sal,nvl(comm,0)) in(select sal,nvl(comm,0) from emp e,dept d
where e.deptno=d.deptno and loc='DALLAS' );
DNAME ENAME SAL
RESEARCH SMITH 800
RESEARCH JONES 2975
RESEARCH FORD 3000

SQL MATERIAL
RESEARCH SCOTT 3000
RESEARCH ADAMS 1100
Q) Write a query to display emp’s who are getting more than the highest paid emp in 20th
dept from emp table?
Sql>select * from emp where sal>(select max(sal) from emp where deptno=20);
7839 KING PRESIDENT 17-NOV-81 5000 10
Q) Write a query to display emp’s who are getting more than the lowest paid emp in 20th
dept from emp table?
Sql>select * from emp where sal>(select min(sal) from emp where deptno=20);
ALL, ANY OPERATOR:
When ever resource table having child amount of data and also child query contain
max(),min() function and if we are comparison more number of value using relational operator.
These type of queries degrades performance of the applications to overcome these problems
ANSI/ISO sql provided subquery special operators these are “ALL,ANY”. This operator is used
along with relational operator. These types of operators improve performance.
Sql>select * from emp where sal>all(select sal from emp where deptno=20);
SQL> select * from emp where sal>any(select sal from emp where deptno=10);
------ -------- ----------------- ------- --------------- ------- -----
ALL---------max(sal) ANY-------min(sal)

SQL MATERIAL
NOTE:
When ever subquery special operator internally optimizer uses when we are using ANY
operator internally optimizers uses ‘OR’ logical operator.
ALL, ANY operator used in multiple row subqueries:
IN ----- It returns same value in the list
ALL ----- It satisfies all values in the list
ANY----- It satisfies any value in the list
Q) Write a query to display the emp’s who are getting more than the salaries of the
‘CLARK’ from emp table using subquery special operator?
Sql>select * from emp where sal>all(select sal from emp where job='CLERK')
order by sal desc;
7902 FORD ANALYST 7566 03-DEC-81 3000 20
------ --------- --------------- ------- --------------- ------ --------- -----
NOTE:
‘OR’ operator internally uses logical operator for ‘AND’ where has ‘ANY’ operator
logical use operator ‘OR’.
Sql>select empno,ename,sal,deptno from emp where deptno>any(10,20);
EMPNO ENAME SAL DEPTNO
7369 SMITH 800 20
7499 ALLEN 1600 30
------- ----------- -------- -----
Sql>select empno,ename,sal,deptno from emp where deptno>all(10,20);

7499 ALLEN 1600 30

SQL MATERIAL
------ ----------- ------- -----
NOTE: ‘=any’ is same as ‘IN’ operator ‘=any’ performance very high compare to IN
operator. Whereas ‘<>any’ not same as ‘NOT IN’ operator.
SQL>select empno,ename,sal,deptno from emp where deptno <>any(10,20);
7369 SMITH 800 20
7499 ALLEN 1600 30
7934 MILLER 1300 10
-------- ---------- ------ ----
SQL> select empno,ename,sal,deptno from emp where deptno not in(10,20);
7499 ALLEN 1600 30
-------- ----------- ------- ----
SQL> select empno,ename,sal,deptno from emp where deptno in(10,20);
7369 SMITH 800 20
------- ---------- ------ ----
7934 MILLER 1300 10
SQL> select empno,ename,sal,deptno from emp where deptno=any(10,20);
7369 SMITH 800 20
------- ---------- ------ ----
7934 MILLER 1300 10
2. CORRELATED SUBQUERY:
Generally in non-correlated subquery child query is execute first then only parent query
executed. Where as in correlated subquery parent query is executed first then only child query
executed.

SQL MATERIAL
In correlated subquery we must create on aliasname for the parent query table in parent
queries use the aliasname in child query where condition
Syntax: select * from tablename aliasname where tablename operator
(select * from tablename where tablename=aliasname.colname);
Execution process:
Whenever we are subquery having correlated subquery database server get a condition
row from parent query table and then control passed into child query where condition based on
evaluation value of where clause it compares query value with parent query.
Generally non-correlated subquery child query executed only once for parent table.
Where has in correlated subqueries child query is executed for each row for a parent table.
Generally when resource table having large amount of data then we have not allowed to
subqueries because in this case correlated is low performance. Generally correlated subquery
used in denormalization process in this process we are modify one table column values based on
another table columns in this use we are using correlated update.
Syntax: update tablename1 aliasname1 set columnname=(select columnname from

tablename2 aliasname2 where aliasname1.commoncolname =
aliasname2.commoncolname);
Ex: sql>alter table emp add dname varchar2(15);

Sql>update emp e set dname=(select dname from dept d
where e.deptno=d.deptno);
Outer query
Difference b/w non-correlated and correlated:
Inner query Inner query
Non-correlated

SQL MATERIAL
Correlated
Outer query
Q) Write a query to display 1st highest salary from emp table?
Sql> select * from emp e1 where 1=(select count(*) from emp e2 where e2.sal>=e1.sal);
Q) Write a query to display 2nd highest salary from emp following table using correlated
subquery?
EMP1
Sql>create table emp1(ename varchar2(5),sal number(5));
Sql>insert into emp1 values(………………………………..);
SQL> select * from emp1;
ENAME SAL
Ename Sal
A 100 Candidate row
A 100
B 150
B 150
C 200 C 200
D 300 D 300
SQL>select * from emp1 e1 where 2=(select count(*) from emp1 e2

where e2.sal>=e1.sal);
ENAME SAL
C 200
Phase-1:
Step-1: get a candidate row (first row) --------- (A 100)
Step-2: select count(*) from emp1 e2 where e2.sal>=100;

Result
Result
Step-3: select * from emp1 e1 where 2=4 (False)
Phase-1:
Step-1: get a candidate row (second row) --------- (B 150)

SQL MATERIAL
Step-3: select * from emp1 e1 where 2=3 (False)
Phase-1:
Step-1: get a candidate row (third row) --------- (C 200)
Step-3: select * from emp1 e1 where 2=2 (True)
Q) Write a query to display 2nd highest sal employee from emp above using correlated
subquery N-1 method?
Sql>select * from emp1 e1 where (2-1)=(select count(*) from emp1 e2
where e2.sal>e1.sal);
ENAME SAL
C 200
NOTE: EMP1
Whenever resource table having duplicate data then we are using N-1 method in case of
N method it will be returns No Row Selected as a message to overcame this problem we must
use DISTINCT clause.
Ename Sal
Sql> insert into emp1(ename,sal)values('c',200); A 100
Sql>select * from emp1 e1 where 2=(select count(*) B 150

C 200
from emp1 e2 where e2.sal>=e1.sal); D 300
No Rows Selected (not current results) C 200
Sql>select * from emp1 e1 where 2=(select count(distinct(sal)) from emp1 e2
ENAME SAL
C 200
C 200

SQL MATERIAL
Q) Write a query to display Nth sal employee from emp table using correlated subquery?
Sql>select * from emp e1 where &n=(select count(distinct(sal)) from emp e2
Enter value for N: 1


Enter value for n: 2
7788 SCOTT ANALYST 7566 19-APR-87 3000 20
7902 FORD ANALYST 7566 03-DEC-81 3000 20
Q) Write a query to display the emp’s who are getting more than the avg salary of the job’s
from emp the using correlated subquery?
Sql>select * from emp e where sal>(select avg(sal) from emp where job=e.job);
7566 JONES MANAGER 7839 02-APR-81 2975 20
7698 BLAKE MANAGER 7839 01-MAY-81 2850 30
7876 ADAMS CLERK 7788 23-MAY-87 1100 20
7934 MILLER CLERK 7782 23-JAN-82 1300 10
EXISTS OPERATOR:
EXISTS operator used in correlated subquery. EXISTS operator performance very high
compare to IN operator. EXISTS operator always returns Boolean value either True (or) False.
EXISTS operator is used to test whether a given set is Empty (or) Non-empty.

SQL MATERIAL
EXISTS operator returns true if set having value non-empty and also this operator returns
false if set is empty. These operator use WHERE clause only. Whenever we are use EXISTS
operator then we are not allowed to use colname along with these operator.
Syntax: select * from tablename aliasname where exists(select * from tablename

where colname=aliasname.colname);
Ex: Exists{1,2,3}=True Exists{ }=False
NOTE: Generally if we want CLERK one table column value into another table column
values then we for EXISTS operator.
Q) Write a query to display those dept details from dept table having emp’s in emp table
assign correlated subquery from emp,dept tables?
Sql> select * from dept d where exists(select * from emp where deptno=d.deptno);
DEPTNO DNAME LOC
10 ACCOUNTING NEW YORK
20 RESEARCH DALLAS
30 SALES CHICAGO
NOTE: Whenever we are using EXISTS operator we can also use constant after select in
child query because EXISTS operator check child query returns Value (or) Not.
Sql> select * from dept d where exists(select 1 from emp where deptno=d.deptno);
DEPTNO DNAME LOC
20 RESEARCH DALLAS
30 SALES CHICAGO
Q) Write a query to display dept table does not having emp’s in emp table from emp,dept
table using correlated?
Sql>select * from dept d where not exists(select * from emp

where deptno=d.deptno);

SQL MATERIAL
DEPTNO DNAME LOC
40 OPERATIONS BOSTON
Q) Write a query to display dept table does not having emp’s in emp table from emp,dept
table using correlated?
Sql>select * from dept where deptno not in(select deptno from emp);
DEPTNO DNAME LOC
NOTE: Generally not in operator does not work with null values to overcome this
problem we must use not EXISTS operator along with correlated subquery.
Sql>insert into emp(empno,ename,deptno)values(1,’murali’,null);
Sql> select * from dept where deptno not in(select deptno from emp);
No Rows Selected
Sql> select * from dept d where not exists(select * from emp where deptno=d.deptno);
DEPTNO DNAME LOC
Q) Write a query to display who are getting same sal as emp’s SCOTT sal from emp table
using correlated subquery in exists operator?
Sql>select empno,ename,job,sal,deptno from emp e1 where exists(select * from emp e2

where e2.ename='SCOTT' and e1.sal=e2.sal);
EMPNO ENAME JOB SAL DEPTNO
7902 FORD ANALYST 3000 20
7788 SCOTT ANALYST 3000 20
MERGE STATEMENT:
Oracle 9i introduced MERGE statement. MERGE is a DML statement which is used to

transfer data from source table into target table if those two table’s structures are same. MERGE
statement are also called as “UPSERT”. Because in MERGE we are using UPDATE, INSERT

SQL MATERIAL
statements. Generally MERGE is used data warehousing applications. When we are using
MERGE statement then we must create aliasname.
Syntax: merge into targettablename using sourcetablename
On (join condition)
When matched then
Update set targettablecolname=sourcetablecolname1, ………………
When not matched then
Insert (targettablecolnames)values(sourcecolvalues);
In MERGE statement we are using UPDATE, INSERT statement but through MERGE
statement we cannot modify ON clause.
Sql> select * from dept; (Target Table)
DEPTNO DNAME LOC
20 RESEARCH DALLAS
30 SALES CHICAGO
Sql> create table dept1 as select * from dept;
Sql> insert into dept1 values(1,'NARASIMHA','YSR');
Sql> select * from dept1; (Source Table)
DEPTNO DNAME LOC
20 RESEARCH DALLAS
30 SALES CHICAGO
1 NARASIMHA YSR

SQL MATERIAL
Sql> merge into dept D using dept1 N

on(D.deptno=N.deptno)
when matched then
update set D.dname=N.dname,D.loc=N.loc
when not matched then
insert(D.deptno,D.dname,D.loc)values(N.deptno,N.dname,N.loc);
sql>select * from dept;
Chapter-8
VIEWS
View is a database object which is used to provide authority level of security. View does
not store data. View is also called as “VIRTUAL TABLE” (or) “WINDOW OF THE TABLE”.
Generally views are created from base table’s database administrator. Generally in all
database we want to restrict table columns from one user into another user then also we are using
views. Generally views created base tables based on base tables. Views are categories into TWO
views.
1. Simple View
2. Complex View (or) Join View
SIMPLE VIEW is a view which is created from only one base table. Where as
COMPLEX VIEW is a view which is created from number of base tables.
1. Simple View:
Syntax: create or replace view viewname

As
Select statement;
Sql> create or replace view v1
as

SQL MATERIAL
select * from emp where deptno=10;

Error: insufficient privileges
Sql>conn sys as sysdba;
Enter password: sys
Sql>grant create any view to SCOTT;
Sql>conn scott/tiger;

as
sql>select * from v1;
7782 CLARK MANAGER 7839 09-JUN-81 2450 10
7934 MILLER CLERK 7782 23-JAN-82 1300 10
DML operations through simple view to base table:
1. When a simple view contains Group functions, Group by clause, Distinct, Rownum, Set
operator, Joins then we cannot perform DML operations through simple view to base table.
2. We must include base table not null column into the view then only we can perform INSERT
operation through simple view to base table.
Ex-1: Sql> create or replace view v1

as
Sql>select * from v1;
Sql>insert into v1(empno,ename,deptno)values(1,’NARASIMHA’,30);
Ex-2: Sql> create or replace view v2

SQL MATERIAL
as
select ename,job,sal,deptno from emp where deptno=10;
sql>select * from v2;
CLARK MANAGER 2450 10
KING PRESIDENT 5000 10
MILLER CLERK 1300 10
SQL> insert into v2(ename,job,sal,deptno)values('SIMHA','MANAGER',20000,20);
ERROR: Cannot insert NULL into ("SCOTT"."EMP"."EMPNO")
NOTE: Primary key include in views then insert otherwise not inserted data.
Sql> update v2 set ename=upper(ename);
Generally when we are creating views then automatically database server storage VIEW
definition in database. In oracle if we want to VIEW definition then we are using USER_VIEWS
data dictionary.
Sql> desc USER_VIEWS;
Sql> select TEXT from USER_VIEWS where VIEW_NAME='V1';
TEXT
select "EMPNO","ENAME","JOB","MGR","HIREDATE","SAL","COMM",
"DEPTNO" from emp where deptno=10;
NOTE:
Whenever we are using group functions or expressions in simple view then we must
create column aliasname for those expressions otherwise oracle server given error.
Sql>create or replace view v3

As

SQL MATERIAL
Select deptno,max(sal),min(sal) from emp group by deptno;

ERROR: must name this expression with a column alias
As
Select deptno,max(sal) maxsal,min(sal) minsal from emp group by deptno;
Sql> select * from v3;
DEPTNO MAXSAL MINSAL
30 2850 950
20 3000 800
10 5000 1300
Sql> create or replace view v3 (deptno,maxsal,minsal)

As
Select deptno,max(sal),min(sal) from emp group by deptno;
DEPTNO MAXSAL MINSAL
30 2850 950
20 3000 800
10 5000 1300

as
select rownum,deptno from emp where deptno=10;
ERROR: must name this expression with a column alias
as
select rownum “rownum”,deptno from emp where deptno=10;

SQL MATERIAL

rownum DEPTNO
1 10
2 10
3 10
In all database systems whenever we are creating a view then automatically view
definition are stored in database. If we want view definition then we must use USER_VIEWS
data dictionary.
With check option:
If we want to create constraint type mechanism on views then only we are using WITH
CHECK OPTION clause. Then we are using this clause only INSERT. This clause visibility
values through base table.

As
Select * from tablename
Where condition with check option;
as
select * from emp where deptno=10 with check option;
7934 MILLER CLERK 7782 23-JAN-82 1300 10
Sql> insert into v5(empno,ename,deptno)values(11,'SIMHA',30);

ERROR: View WITH CHECK OPTION where-clause violation

SQL MATERIAL
Sql>insert into v5(empno,ename,deptno)values(11,’SIMHA’,10);

1 Row Inserted
(Here 30 not in v5 that’s why not inserted 10 are in v5 that’s why inserting)
FORCE VIEW: We can also create view without base table these types of view are called
as “FORCE VIEW”.
Syntax: create or replace force view viewname

As
Select * from tablename;
Ex: Sql>create or replace force view v6
as
select * from fan;
Warning: View created with compilation errors.
Sql> create table fan(sno number(5));
Sql> alter view v6 compile;
Sql> desc v6;
Name Null? Type

SNO NUMBER(5)
2. Complex View (or) Join View:

Sql> create or replace force view v7
as
select ename,sal,dname,loc from emp,dept
where emp.deptno=dept.deptno;
ENAME SAL DNAME LOC
SMITH 800 RESEARCH DALLAS
ALLEN 1600 SALES CHICAGO
JONES 2975 RESEARCH DALLAS

SQL MATERIAL
---------- ------ ----------------- -------------
Whenever we are try to performance DML operator through compare view table some
tables column are affected some table column are not affected. Generally we can not modify
non-key preserved table columns. Complex views are created multiple base tables.
DML operations on complex views:
Sql>update v1 set ename=’abc’ where ename=’SMITH’;

Sql>update v1 set dname=’xyz’ where dname=’SALES’;
ERROR: Can not modify a column which map to a non-key preserved table.
In oracle if we want to view modify table, non modify table then we using
USER_UPDATABLE_COLUMNS data dictionary.
Ex: Sql>desc
Sql>select column_name,updatable from user_updatable_columns
where table_name='V7';
COLUMN_NAME UPD
ENAME YES
SAL YES
DNAME NO
LOC NO
Whenever we are performing DML operation through complex view to base table some
table column are not affected to overcome this problem oracle introduced instead of triggers in
pl/sql. Instead of trigger are created on views by default instead of triggers are row level trigger.
TRIGGERS (PL/SQL):
Triggers are also same as stored procedure and it will automatically invoked whenever
DML operation performed on Table (or) View. All database system having two types of triggers.
1. Statement Level Trigger

2. Row Level Trigger

SQL MATERIAL
In statement level triggers trigger body is executed once for DML statement where as in
row level triggers trigger body executed for each row for DML statement.
Syntax: create or replace trigger triggername

Before/after insert/delete/update on tablename
[for each row]
Begin
--------
--------
End;
Difference b/w Statement Level, Row Level Triggers:
Sql>create test(col1 date);
Statement Level Trigger:
Sql> create or replace trigger sl

after update on emp
begin
insert into test values(sysdate);
end;
Testing: Sql> update emp set sal=sal+100 where deptno=10;
Sql> select * from test;
COL1
Trigger 05-SEP-14
Specification Row Level Trigger:
Sql> create or replace trigger rl

Trigger
after update on emp
Body
for each row
begin
SQL MATERIAL
insert into test values(sysdate);

end;
Testing: Sql> update emp set sal=sal+100 where deptno=10;
COL1
05-SEP-14
05-SEP-14
05-SEP-14
insert update delete
NEW
05-SEP-14
OLD
Row Level Trigger:
In row level trigger body executed for each row for done. That way we must use for each
row clause in trigger specification and also data internally stored in to rollback segment
qualifiers. These are qualifiers in trigger body when we are using this qualifier trigger body then
we must use [:] then we must qualifier trigger.
Syntax: :old.colname
:new.colname
  
  
Q) Write a pl/sql row level trigger on emp table whenever user deleting data those deleted
that data stored in another?
Sql>create table backup as select * from emp where 1=2;
Sql>desc backup; here any false condition
Sql>select * from backup;
Sql> delete from emp where sal>2000; 6 rows deleted

SQL MATERIAL
Sql> select * from backup;
7566 JONES MANAGER 7839 02-APR-81 2975 20
7698 BLAKE MANAGER 7839 01-MAY-81 2850 30
7788 SCOTT ANALYST 7566 19-APR-87 3000 20
7902 FORD ANALYST 7566 03-DEC-81 3000 20
6 rows selected
Sql> select * from emp;
7369 SMITH CLERK 7902 17-DEC-80 800 20
7876 ADAMS CLERK 7788 23-MAY-87 1100 20
7900 JAMES CLERK 7698 03-DEC-81 950 30
7934 MILLER CLERK 7782 23-JAN-82 1300 10
8 rows selected
INSTEAD OF TRIGGER:
Oracle 8i introduced instead of triggers. Instead of triggers are created on views. By

default instead of triggers are row level triggers.
Generally we cannot performance DML operations through complex view to base table to
overcome this problem oracle introduced instead of triggers.
Syntax: create or replace trigger triggername

Instead of insert/delete/update on viewname

SQL MATERIAL
For each row

Begin
---------
---------
End;
Solution (using instead of trigger):
Sql> create or replace trigger t1

instead of update on v7
for each row
begin
update dept set dname=:new.dname where dname=:old.dname;
update dept set loc=:new.loc where loc=:old.loc;
end;
Testing: Sql>update v7 set dname='GNSR' where dname='SALES';
Sql> select column_name,updatable
from user_updatable_columns where table_name='V7';
COLUMN_NAME UPD
ENAME YES
SAL YES
DNAME YES
LOC YES
MATERIALIZED VIEW:
Oracle 8i introduced materialized views. Generally views are does not store data where
as Materialized Views are store data. Generally MV is handling by database administrator. MV
is used to improve performance of the Joins (or) Aggregatable queries. MV stores result of the
query. MV is store replication of the remote database into local server (local node).

SQL MATERIAL
Materialized view also store’s data same like a table. But whenever we are refreshing
MV if synchronous of data based on base table.
Syntax: create materialized view viewname

As
Select statement;
Before we are creating materialized view database administrator must give create any
privileges otherwise oracle server returns an Error: Insufficient Privileges.
Syntax: Sql>grant create any materialized view to username;
Ex: Sql> create materialized view mv1

as
select * from emp;
ERROR: Insufficient Privileges
Sql> conn sys as sysdba;
Enter password: SYS
Sql> grant create any materialized view to SCOTT;
Sql> conn scott/tiger;
Sql> create materialized view mv1
as
select * from emp;
NOTE: In oracle one of the materialized view base table must have primary key.
Otherwise oracle server returns on error.
Sql> create table test(sno number(10),name varchar2(15));

as
select * from test;
ERROR: Table 'TEST' does not contain a primary key constraint

SQL MATERIAL
Difference b/w Views, Materialized Views:

Sql>create table base(sno number(5) primary key,varchar2(15));
Sql>insert into base values(…………………….);
Sql>select * from base;
SNO NAME
1 a
2 b
3 c
4 d
as
select * from base;
as
select * from base;
Here materialized view also behaves likes a view. Whenever we are creating materialized
view automatically MV definition permanently stores in a database same like a view definitions.
In oracle if we want MV definition then we are using USER_MVIEWS data dictionary.
Sql>desc USER_MVIEWS;
Sql>select query from user_mviews where mview_name='MV3';
QUERY
SELECT "BASE"."SNO" "SNO","BASE"."NAME" "NAME" FROM "BASE" "BASE"
Sql> select rowid,sno,name from base;
ROWID SNO NAME
AAADwEAAEAAAAFPAAA 1 a

SQL MATERIAL
AAADwEAAEAAAAFPAAB 2 b
AAADwEAAEAAAAFPAAC 3 c
AAADwEAAEAAAAFPAAD 4 d
Sql> select rowid,sno,name from v8;
ROWID SNO NAME
AAADwEAAEAAAAFPAAA 1 a
AAADwEAAEAAAAFPAAB 2 b
AAADwEAAEAAAAFPAAC 3 c
AAADwEAAEAAAAFPAAD 4 d
Here view rowid’s are same as base table rowid’s. That why view does not store data.
Through the view we are viewing base table that’s why view are called as “VIRTUAL TABLE”.
Sql>select rowid,sno,name from mv3;
ROWID SNO NAME
AAADwHAAEAAAAFcAAA 1 a
AAADwHAAEAAAAFcAAB 2 b
AAADwHAAEAAAAFcAAC 3 c
AAADwHAAEAAAAFcAAD 4 d
Here MV rowid’s are called different from base table rowid’s. that’s why materialized
view store data.
Sql> update base set name=upper(name);

SNO NAME
1 A
2 B here affected that’s why capital letters

3 C
4 D
Sql> select * from mv3;

SNO NAME

SQL MATERIAL
1 a
2 b here not affected that’s why small letters

3 c
4 d
Sql> select * from base;

SNO NAME
1 A
2 B
3 C
4 D
When we are refreshing MV it synchronous data based on base table. In oracle if we want
to refresh MV then we are using refresh method from DBMS_MVIEW package.
Syntax: dbms_mview.refresh(‘materialized viewname’);
Sql> EXEC DBMS_MVIEW.REFRESH('MV3');
Sql>select * from mv3;
SNO NAME
1 A
2 B
3 C
4 D
Oracle having two types of Materialized Views
1. Complete Refresh Materialized View
2. Fast Refresh Materialized View
1. Complete Refresh Materialized View:
In oracle by default MV’s are complete materialized view. CRMV when we are
refreshing internally rowid’s are Re-Created. This process does not improve more performance.

SQL MATERIAL
If we are called modifying data in base table also that why when we are refreshing MV
frequently then this type MV are ‘D’ grade performance.

Refresh complete
As
Select statement;
Sql>create materialized view mv4
refresh complete
as
select ename,job,sal,deptno from emp where deptno=10;
Sql>EXEC DBMS_MVIEW.REFRESH('MV4');
Sql> select rowid,sal,deptno from mv4;
ROWID SAL DEPTNO
AAADwUAAEAAAAFsAAD 2450 10
AAADwUAAEAAAAFsAAE 5000 10
AAADwUAAEAAAAFsAAF 1300 10
Sql> select rowid,sal,deptno from mv4;
ROWID SAL DEPTNO
AAADwUAAEAAAAFsAAA 2450 10
AAADwUAAEAAAAFsAAB 5000 10
AAADwUAAEAAAAFsAAC 1300 10 (here rowid’s are changed)
2. Fast Refresh Materialized View:
Fast refresh materialized view are also called as “Incremental Refresh Materialized
View”. When we are refreshing FRMV it will improve performance of the MV’s. In FRMV

SQL MATERIAL
rowid are Not Changed if you are refreshing MV no.of times also. In FRMV data only
synchronous.

Refresh fast
As
Select statement;
Before we are creating FRMV database administrator must create MV log on base table.
These MV create log stores states of the base table into appropriate data dictionary.
Syntax: create materialized view log on basetablename;
Ex: sql> create materialized view log on base;

refresh fast
as
select * from base;
Sql> select * from mv5;
SNO NAME
1 A
2 B
3 C
4 D

Sql> select rowid,sno,name from mv5;
ROWID SNO NAME
AAADwYAAEAAAAF8AAA 1 A
AAADwYAAEAAAAF8AAB 2 B
AAADwYAAEAAAAF8AAC 3 C
AAADwYAAEAAAAF8AAD 4 D
SQL MATERIAL

ROWID SNO NAME
AAADwYAAEAAAAF8AAA 1 A
AAADwYAAEAAAAF8AAB 2 B
AAADwYAAEAAAAF8AAC 3 C
AAADwYAAEAAAAF8AAD 4 D (Here rowid’s are not changed)
ON DEMAND/ON COMMIT:
Generally refresh materialized views are two ways
1. Manually
2. Automatically
1. Manually:
We can also refresh materialized views either Manually (or) Automatically. In manually
method we are using dbms_mview package to refresh MV these method is called as “On
Demand”. By default method is On Demand method.
2. Automatically:
If you want to refresh MV without using dbms_mview package this is called as “On
Commit”. When we are using On Commit method in this case explicitly we must specify On
Commit key word clause in MV.

Refresh complete/refresh fast on demand/on commit
As default
Select statement;
Sql> select * from base;
SNO NAME
1 A

SQL MATERIAL
2 B
3 C
4 D

refresh fast on commit
as
select * from base;
ROWID SNO NAME
AAADwbAAEAAAAGMAAA 1 A
AAADwbAAEAAAAGMAAB 2 B
AAADwbAAEAAAAGMAAC 3 C
AAADwbAAEAAAAGMAAD 4 D
Sql> update base set name='XYZ' where sno=2;

Sql> select rowid,sno,name from base;
ROWID SNO NAME
AAADwEAAEAAAAFPAAA 1 A
AAADwEAAEAAAAFPAAB 2 XYZ
AAADwEAAEAAAAFPAAC 3 C
AAADwEAAEAAAAFPAAD 4 D

ROWID SNO NAME
AAADwbAAEAAAAGMAAB 2 B
Sql> commit; (Here explicitly give commit after affected data)

ROWID SNO NAME

SQL MATERIAL
AAADwbAAEAAAAGMAAB 2 XYZ
View Materialized View
1. It does not store data. 1. It store data.
2. Security purpose 2. Improve performance purpose.
3. Hiding some information from the user i.e it 3. It stores result of the query i.e it is called as
is also called “Window Of The Table”. “Copy Of The Table”.
4. When we are modifying data in base table 4. When we are modifying data in base table
then view is immediately affected. then MV not affected immediately because
here rowid’s are different.
Chapter-9
DCL (Data Control Language)
DCL are two types these are
1. GRANT
2. REVOKE
1. Grant:
Sql GRANT is a command used to provide access or privileges on the database objects to
the users.
Data security point of view oracle provided two types of admin user’s these are
1. SYS
2. SYSTEM
Creating a user:

SQL MATERIAL
Syntax: create user username identified by password;
Syntax: grant connect,resource to username;
Syntax: conn username/password;
Ex: sql>conn sys as sysdba;

Enter password: SYS
Sql> create user NARASIMHA identified by NARASIMHA;
Sql>grant connect,resource to NARASIMHA; ---connect,resource(dba)
Sql> conn NARASIMHA/NARASIMHA;
Error: table or view does not exit
Sql> conn scott/tiger;
Sql> grant all on emp to NARASIMHA;
Error: table or view does not exit
Sql> select * from scott.emp;
Sql> conn narasimha/narasimha;
Sql> create synonym gnsr for scott.emp;
ERROR: insufficient privileges
Enter password: SYS
Sql> grant create any synonym to NARASIMHA;
Sql> create synonym gnsr for scott.emp;
Sql> select * from gnsr;
PRIVILEGES:
Privilege is an write given to the user to perform action to the database.

SQL MATERIAL
Ex: Grant table,create procedure

Data security point of view two types of privileges
1. System Privileges
2. Object Privileges
1. System Privilege:
System privileges are given by database administrator oracle having to system privilege.
System privileges are used to create object, alter table, create procedure, create trigger, create
any materialized, create session …………………etc.
Syntax: grant system privileges to User-1

username1,username2……………; User-2
Ex: Sql> conn sys as sysdba; User-3
Enter password: SYS User-4
Sql> grant create table,create procedure,create any materialized User-5
view,create trigger,
create session,create synonym to scott,narasimha;
ROLE:
Role is nothing but collection of system privileges (or) collection of object privileges.
Userdefined role are create by database administrator only.
Generally in multi user environment number of user’s works on same project in this case
some user require common set of privileges in this case only database administrator creating a
userdefined role and assigns common set of privileges in to role and then only that role given to
the number of user’s.
2 (SYSTEM PRIVILEGES) 1 (ROLE) 3 (USER’S)

SQL MATERIAL
Creating a userdefined ROLE:

Connect, Resource, Dba ----ROLES
Step-1:- (create a role)
Syntax: create role rolename;
Step-2:- (assign system privileges to role)
Syntax: grant system privileges to rolename;
Step-3:- (assign role to no.of users)
Syntax: grant rolename to username1,username2 ……………..;
Enter password: SYS
Sql> create role r1;
Sql>grant create table,create procedure,create any materialized view,create trigger,create

session,create synonym to r1;
In oracle system privileges related to role stored under ROLE_SYS_PRIVS data

dictionary.
Sql> desc role_sys_privs;
Name Null? Type
ROLE NOT NULL VARCHAR2(30)
PRIVILEGE NOT NULL VARCHAR2(40)
ADMIN_OPTION VARCHAR2(3)
Sql> select role,privilege from role_sys_privs where role='R1';
ROLE PRIVILEGE
R1 CREATE SESSION
R1 CREATE TABLE
R1 CREATE SYNONYM

SQL MATERIAL
R1 CREATE TRIGGER
R1 CREATE ANY MATERIALIZED VIEW
R1 CREATE PROCEDURE
PREDEFINED ROLE:
Whenever we are installing server then automatically three ROLES are created. This are
called as “Predefined Roles”.
1. CONNECT ------------- End User
2. RESOURCE ------------- Developer
3. DBA ------------- Database Administrator
Sql>conn sys as sysdba;

Enter password: SYS
Sql>desc role_sys_privs;
Sql>select role,privilege from role_sys_privs where role in('connect','resource');
Sql>select role,privilege from role_sys_privs where role in('dba');
2. Object Privileges:
Object privileges are used to performance. Some privileges object privileges is given by
either developers (or) dba. In oracle insert, update, delete, select, execute, read, write,……….
this are also called as “ALL” (or) “OBJECT PRIVILEGES”.
Syntax: grant object privilege on objectname to username/rolename/public;
Sql> grant all on emp to narasimha,r1;
Sql> grant all on emp to public;
Sql> desc user_tab_privs;
WITH GRANT OPTION:

SQL MATERIAL
Who receive with grant option clause those users given same object privileges to another
user’s.
Syntax: grant object privilege on objectname to username/public with grant option;
Sql>grant all on emp to narasimha with grant option;
Sql>grant all on emp to r1 with grant option;
Error: Cannot GRANT to a role WITH GRANT OPTION
Sql>grant all on emp to public with grant option;
Sql>grant all on emp to narasimha,public with grant option;
NOTE:
In all database systems ROLES does not work WITH GRANT OPTION clause. All
object privileges related to user stored under USER_TAB_PRIVS data dictionary.
Sql>desc user_tab_privs;
2. REVOKE:
REVOKE command is used to cancel either system (or) object privileges.
Syntax: revoke system privilege from username1,username2,……………..;
Revoke object privilege on objectname from username/rolename/public;
NOTE:
Generally if we want to restrict table columns from one user into another user then we
creating view with required column’s then only that views given to the no.of users.
Syntax: grant all on viewname to username1,username2,………………..;

as
select empno,ename,sal,deptno from emp where deptno=10;
SQL MATERIAL
7782 CLARK 2450 10
7839 KING 5000 10
7934 MILLER 1300 10
Sql> grant all on v9 to narasimha;
Sql> select * from scott.v9;
7782 CLARK 2450 10
7839 KING 5000 10
7934 MILLER 1300 10
NOTE:
We can also drop view using drop view viewname. Whenever we are dropping base table
then view is invalid and also view definition are available in database in oracle. If you want view
definition then we are using USER_VIEWS data dictionary.
Sql> drop table base;
ERROR: view "SCOTT.V8" has errors
Sql> desc user_views;
Sql> select text from user_views where view_name='V8';
TEXT
select "SNO","NAME" from base
READ ONLY VIEW:
We can also create read only view using with read only option clause. Whenever we are
using this clause we cannot perform DML operation view to base table.

SQL MATERIAL

As
Select * from tablename where condition with read only;
as
select * from emp where deptno=10 with read only;
7934 MILLER CLERK 7782 23-JAN-82 1300 10
Sql> delete from v10 where deptno=10;
ERROR: cannot delete from view without exactly one key-preserved table
Chapter-10
INDEX
Oracle 8i introduced indexes. INDEX is a database object which is used retrieves data
fast from the database. This process automatically improves performance of the application.
Become INDEX columns retrieve data very fastly from the table. Generally INDEX created on
table columns. INDEX is handling by database administrator.
Index created two ways:
1. Automatically
2. Manually
1. Automatically:
In oracle whenever we are creating unique key (or) primary key then oracle server
automatically creates BTREE indexes on those columns.
SQL MATERIAL
2. Manually:
We can also creates INDEX explicitly by using create INDEX method.
Syntax: create index indexname on tablename(colname);
Whenever we are using select query contain WHERE clause (or) ORDER BY clause then
only oracle server searching for INDEX in the database. If INDEX are available oracle server
retrieve data very fastly the table by using INDEX SCAN() method. If INDEX not available then
oracle server uses full table scan method to retrieve data from the table.
NOTE:
Whenever WHERE clause contain <> (or) is null (or) is not null operators then oracle
server does not search for INDEX if those column already having INDEX also.
Oracle having two types of index: Oracle database

1. Btree Index
2. Bitmap Index
Data Dictionary
no. of distinct value
carnality =
total no. of row
btree = =1 High cardinality bitmap= = 0.357 low cardinality
1. Btree Index:
Btree index are created on high cardinality columns. In oracle by default index are btree
index whenever we are creating btree index automatically oracle server stores actual data along
with rowid’s in leaf blocks of the btree structure. Whenever WHERE condition value available
in leaf blocks is create some retrieve data very fastly from the table along index scan method.
DEVALOPER DBA
Sql> select * from emp where ename=’KING’; Sql> select * from emp where ename=’KIMG’;
A ---------Z Root
A-----------J K----------Z Branch

SQL MATERIAL
Rowid’s Allen Rowid’s King
Rowid’s Blake Rowid’s Scott
------ ----- --------- ----------
------ ------ --------- ---------
Leaf blocks
If we want view performance of the query we must create plan table using explain plan
clause.
Syntax: explain plan for select statement;
If we want view plan table then we must use display method from DBMS_XPLAN
method.
Sql> select * from table(dbms_xplan.display());
Without using Index:
Sql> select * from emp where ename=’KING’;
EMPNO ENAME JOB MGR

Ename index1 normal
HIREDATE SAL COMM DEPTNO
Sql> explain plan for select * from emp where ename=’KING’;
PLAN_TABLE_OUTPUT
| Id | Operation | Name | Rows | Bytes | Cost (%CPU) |Time |
| 0 | SELECT STATEMENT | | 1 | 37 | 3 (0) | 00:00:01 |
|* 1 | TABLE ACCESS FULL | EMP | 1 | 37 | 3 (0) | 00:00:01 |
Using index:
Sql> create index index1 on emp(ename);
Sql> select * from emp where ename=’KING’;

SQL MATERIAL
Sql> explain plan for select * from emp where ename=’KING’;
PLAN_TABLE_OUTPUT
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
| 1 | TABLE ACCESS BY INDEX ROWID | EMP | 1 | 37 | 2 (0) | 00:00:01 |
|* 2 | INDEX RANGE SCAN | INDEX1 | 1 | | 1 (0) | 00:00:01 |
All index information stored under USER_INDEXES data dictionary.

Sql> desc user_indexes;
Sql> select index_name,index_type from user_indexes
where table_name='EMP';
INDEX_NAME INDEX_TYPE
PK_EMP NORMAL -------------------- Automatically
INDEX1 NORMAL -------------------- Manually
NOTE:
If we want view column names along with index name then we are using
user_ind_columns data dictionary.
Sql> desc user_ind_columns;
Sql> select COLUMN_NAME,INDEX_NAME from user_ind_columns
where table_name='EMP';
COLUMN_NAME INDEX_NAME
EMPNO PK_EMP
ENAME INDEX1

SQL MATERIAL
FUNCTION BASED INDEX:
Oracle 8i introduced function based index. Function based index are by default btree
index. Generally whenever WHERE clause contain function (or) expressions then oracle server
does not search for index if already those columns having also to overcome this problem oracle
8i introduced extension of the btree index this are also called as “function based index”. In
function based column along with function (or) expression index.
Syntax: create index indexname on tablename(function/expression(colname));
Sql> select * from emp where upper(ename)='KING';
Sql> explain plan for select * from emp where upper(ename)='KING';
PLAN_TABLE_OUTPUT
| Id | Operation | Name | Rows | Bytes | Cost (%CPU) |Time |
|* 1 | TABLE ACCESS FULL | EMP | 1 | 37 | 3 (0) | 00:00:01 |
Sql> create index index2 on emp (upper(ename));
Sql> explain plan for select * from emp where upper(ename)='KING';
PLAN_TABLE_OUTPUT
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
| 1 | TABLE ACCESS BY INDEX ROWID | EMP | 1 | 37 | 2 (0) | 00:00:01 |
|* 2 | INDEX RANGE SCAN | INDEX2 | 1 | | 1 (0) | 00:00:01 |
Sql> select index_name,index_type from user_indexes where table_name='EMP';

SQL MATERIAL
PK_EMP NORMAL
INDEX1 NORMAL
INDEX2 FUNCTION-BASED NORMAL
VIRTUAL COLUMN:
Oracle 11g introduced virtual columns using virtual column we can store expressions
directly into database. Generally in oracle if we want to stored expressions then we are using
either function based index (or) view these two methods in directly stores stored expression in
expression to overcome this problem oracle 11g introduced virtual column to stores stored
expression directly into database using generated always as.
Syntax: columnname datatype(size) generated always
as(function/expression(colname);
Sql> create table t1(a number(5),b number(5),c number(10));

Sql> insert into t1(a,b)values(10,20);
Sql> select * from t1;
A B C
10 20 30
NOTE:
If you want view virtual column expression then we are using DATA_DEFAULT
property from USER_TAB_COLUMNS data dictionary.
Sql> desc user_tab_columns;

Sql> select column_name,data_default from user_tab_columns where table_name=’T1’;
COLUMN_NAME DATA_DEFAULT
C “Á+B”
Oracle having two btree indexes:

SQL MATERIAL
1. Unique Btree Index

2. Non-Unique Btree Index
In oracle by default pk,unique key constrained columns are unique btree index and to
give explicitly this are non-unique btree index. We can also create unique btree index explicitly
by using following index.
Syntax: create unique index indexname on tablename(colname);
We are not allowed to create unique index on duplicate value columns.
Sql> create unique index index3 on emp(ename);
Sql> select index_name,index_type from user_indexes where table_name='EMP';
PK_EMP NORMAL
INDEX3 NORMAL
Sql> create unique index index4 on emp(job)
ERROR: cannot CREATE UNIQUE INDEX; duplicate keys found
2. Bitmap:
Oracle 7.3 introduced bitmap index. Generally bitmap index are used in data warehousing
application. Bitmap index are created in low cardinality columns (more duplicate data) low
cardinality columns more duplicate data.
Syntax: create bitmap index indexname on tablename(colname);
Sql> create bitmap index index4 on emp(job);
Whenever we are creating bitmap index oracle server automatically create a bitmap table
on the specified columns and using bits. Whenever user requesting using logical operator. Then
only oracle server directly operates bits within bitmap table and also resultant bitmap
automatically convert into rowid using internal bitmap function.
JOB 1 2 3 4 5 6 7 8 9 10 11 12 13 14

SQL MATERIAL
CLERK 1 0 0 0 0 0 0 0 0 0 1 1 0 0
SALESMAN 0 1 1 0 1 0 0 0 0 1 0 0 0 0
MANAGER 0 0 0 1 0 1 1 0 0 0 0 0 0 0
ANALYST 0 0 0 0 0 0 0 1 0 0 0 0 1 0
PRESIDENT 0 0 0 0 0 0 0 0 1 0 0 0 0 0
Chapter-11
SEQUENCES
Sequences are database object which is used to generate sequence number

automatically. Generally sequence is used to generate primary key values automatically.
Sequence is an independent database object. Generally database administrator creates sequence.
Once sequence access that sequence.
Syntax: create sequence sequencename

Start with n
Increment by n
Minvalue n
Maxvalue n
Cycle/nocycle
Cache/nocache;

SQL MATERIAL
If you want access sequence values then we are using 2 pseudo columns.
1. Currval
2. Nextval
Syntax: sequencename.currval
Syntax: sequencename.nextval
If you want generate sequence values using select statement then we must use DUAL
table.
Syntax: select sequencename.currval from dual;
Syntax: select sequencename.nextval from dual;
If you want generate first sequence numbers then we must nextval pseudo column. In this
case nextval pseudo column returns starting sequence number currval pseudo column return
current value of the sequence session already having a value.
Sql> create sequence s1

start with 5
increment by 2
maxvalue 15;
Sql> select s1.currval from dual;
ERROR: sequence S1.CURRVAL is not yet defined in this session
Sql> select s1.nextval from dual;

NEXTVAL
Sql> /
NEXTVAL
Sql> select s1.currval from dual;

SQL MATERIAL
CURRVAL
NOTE:
We can also return currval along with nextval pseudo then we are generating first
sequence value.
start with 1
increment by 1
maxvalue 10;
NEXTVAL
Sql> /
NEXTVAL
Sql> select s2.nextval,s2.currval from dual;
NEXTVAL CURRVAL
3 3
NOTE:
Sequence are used in insert,update,delete,select statements. We can also modify sequence

property using ALTER but we can not starting with sequence number (start with value).
Syntax: alter sequence sequencename

Propertyname newvalue;
Sql> alter sequence s2
Increment by -1;
SQL MATERIAL
Sequence alter
Sql> alter sequence s2
Start with 4
Maxvalue 10;
ERROR: cannot alter starting sequence number
NOTE:
Whenever we decrement the sequence we cannot instantiate below minimum value and
also start with cannot be less minimum value.
Start with 3
Increment by 1
Minvalue 5
Maxvalue 10;
Error: START WITH cannot be less than MINVALUE

CYCLE/NOCYCLE:
When we are defined cycle whenever reach value max value next values start with
starting value.
Nocycle: Sql> create sequence s3

start with 6
increment by 1
minvalue 6
maxvalue 10
nocycle
nocache;

SQL MATERIAL
NEXTVAL
Sql> /
NEXTVAL
Sql> /
NEXTVAL
Sql> /
NEXTVAL
Sql> /
NEXTVAL
10
Sql> /
ERROR: sequence S3.NEXTVAL exceeds MAXVALUE and cannot be instantiated
Cycle: Sql> create sequence s4

start with 6
increment by 1
minvalue 6
maxvalue 8
cycle
nocache;
NEXTVAL
Sql> /
NEXTVAL

SQL MATERIAL
Sql> /
NEXTVAL
Sql> /
NEXTVAL
Sql> /
NEXTVAL
Sql> /
NEXTVAL
Sql> /
NEXTVAL
Sql> select * from user_sequences;
Sequence_Name Min_Value Max_Value Increment_By C O Cache_Size Last_Number
S2 1 10 -1 N N 20 0
S1 1 15 2 N N 20 17
S3 6 10 1 N N 0 11
S4 6 8 1 Y N 0 7
CACHE/NOCACHE:
Cache is an optional clause used in sequence we access sequence number very fastly i.e
whenever we defined sequence if we want access then we run set of sequence value in cache
memory then only application access sequence values from cache not from the disk that why
cache is memory area which stores set of sequence and improve performance that why cache
value by defined by database administrators.
NOTE:
SQL MATERIAL
In oracle by default cache value is 20 and also cache minimum value is 2. Whenever
system caches then automatically generates because cache is a memory area whenever system
crush cache values are lost.

start with 2
increment by 1
cache 1;
ERROR: the number of values to CACHE must be greater than 1
NOTE:
Whenever we are defining cache value that is less than one cycle when we are defining
cycle option otherwise oracle server returns on error to overcome this problem we can also use
no cache when no cache when we are defining cycle option. We must create primary key column
otherwise not properly inserted data and duplicate value inserting.

start with 5
minvalue 4
maxvalue 6
cycle
cache 40;
ERROR: number to CACHE must be less than one cycle
Sql> create table test(sno number(5) primary key,name varchar2(15));
Start with 1;
Sql> insert into test(sno,name) values(s5.nextval,’&name’);
Enter value for name: ABC
SQL> /
Enter value for name: XYZ
SQL MATERIAL
SQL> select * from test;

SNO NAME
1 ABC
2 DEF
3 GHI
4 JKL
5 XYZ
All sequence information stored under USER_SEQUENCES data dictionary.
Sql> desc USER_SEQUENCES;
To view all sequence table information’s
Sql> select * from USER_SEQUENCES;
Chapter -12
NORMALIZATION
Normalization is a scientific process decompose table into number of tables.

These processes automatically reduce redundancy.
Normalization process automatically avoids insert, delete and update problems. In design
place of the SDLC(Software Development Life Cycle) database designers design logical model
of the database. In these logical model only database designer uses normalization process
through normal forms.
In 1970 E.F CODD had written a paper “Relational Model Of Data For Shared Data
Banks” in this paper only E.F CODD introduced first three normal forms.
Normal Forms:
1. First Normal Form

2. Second Normal Form
SQL MATERIAL
3. Third Normal Form

4. BCNF
5. Fourth Normal Form
6. Fifth Normal Form
1. First Normal Form:

PK
If a table is in 1NF then that table does not contain Order Master Table multi table i.e
every cell contain atomic (single value) and also identifier record uniquely using a key.
Item name Color Price Tax

Item name Color Price Tax
Marker Blake 20 0.2
Black,
Marker 20 0.2 Marker Red 20 0.2
red
Blue, Pen
1NF Blue
Order 30
item details 0.3
Pen 30 0.3
green Pen Green 30 0.3
FK
Process:
Identified repeating groups and putting into separate table in more atomic forms by
default 1NF process table is an child table because here one column having duplicate data.
Electronic Shop
Order no:
Order date:
Order Customer Customer
Order no
date Address phone
Customer name:
1 - - -
Customer Not In 1NF address: Candidate Key 1NF
Customer phone:
Order no Item name Amount

1 Refrigerator 30000

SQL MATERIAL
Item names: 1 Led 40000
Refrigerator 1 A/C 50000
Led
A/C
√ Add another item
Amount:
SUBMIT
2. Second Normal Form:
A table is a 1NF and also all non-key attribute fully functionally dependent total
candidate key.
Generally 1NF deals with atomicity where as 2NF deals with relationship b/w key,non-
key attribute. When a table is in 1NF and also table contain partial non-key attribute then that
table not in 2NF.
Process:
Identified partial non-key attributes which depends on partially key attributes putting into
separate table these are called as “2NF”. In these tables only all attribute fully dependency. By
default this is on master table.
Item name Price Tax

SQL MATERIAL
key attributes 1NF p Non-key attributes Fk

Pk 2NF Master table
Item name color price tax Item name color

Master black 20 0.2
Master red 20 0.2
Pen blue 30 0.3
Pen green 30 0.3
Sno Sname Actvity1 Cost1 Activity2 Cost2

101 ABC CRICKET $10 FOOTBALL $20
102 XYZ GOLF $30 SWIMMING $40
103 ZZZ CRICKET $10 GOLF $30
104 AAA CRICKET $10 SWIMMING $40
105 BBB FOOTBALL $20
Sno Sname Sno Actvity1 Cost1 Activity2 Cost2

101 ABC 101 CRICKET $10 FOOTBALL $20
102 XYZ 102 GOLF $30 SWIMMING $40
103 ZZZ
103 CRICKET $10 GOLF $30
104 AAA
104 CRICKET $10 SWIMMING $40
105 BBB
105 FOOTBALL $20
ACTIVITY TABLE
Sno Act cost

101 Ckt $10
101 Fb $20
102 Golf $30
102 Swim $40

SQL MATERIAL
2NF (Master table) Child table 2NF (Master table) Non-key
Not in 2NFCandidate key attribute
Fk P
Fk
Pk pk Sno Sname
103 Ckt $10
Sno Act 103 Golf $30
Sno Sname Act cost 101 ABC
101 Ckt 104 Ckt $10
101 ABC 101 Fb 104 Swim $40
Ckt $10 102 XYZ
102 Golf 105 Fb $20
102 XYZ 103 ZZZ
102 Swim Fb $20
103 ZZZ 103 Ckt 104 AAA
103 Golf Golf $30
104 AAA 105 BBB
104 Ckt
104 Swim Swim $40
105 BBB
105 Fb
Project table: p p
Candidate key key attributes non-key attributes
Ecode Projcode Dept Depthead Hours F Phase-1:-

E101 P1 Systems D1 5
E102 P1 Sales D2 3
1. Fd
E101 P2 Systems D1 6 Ecode hours ( )
E103 P2 Research D3 4
E102 P3 Sales D2 7 Value more than 1 hour
E101 P3 Systems D1 8
5
E101 6
8
2. fd 3. fd
Projcode hours (  ) ecode+Projcode hours ( √ )
5
P1
3
Phase-2:- fd fd
Ecode dept ( √ ) Projcode dept (  )
System
E101----systems p1 Sales

SQL MATERIAL
Phase-3:-
1. Ecode dept ( √ ) 2. Projcode dept (  )
(Master table) (Child table)

Pk 2NF
Fk
Ecode Dept Depthead Ecode Projcode hours

E101 P1 5
E101 Systems D1 E102 P1 3
E101 P2 6
E102 Sales D2
E103 P2 4
E103 Research D3 E102 P3 7
E101 P3 8
Before normalization process above resource table having insertion, deletion, updation
problems occurred.
Insertion problem:
In the above resource table we cannot insert particular dept emp’s then we must assigning
a projects this is called as “Insertion Problem”.
Updation problem:
In the above resource table Ecode,Dept,Depthead attribute values are repeated. Whenever
on employee transferred one dept to another dept all this three attributes value modified currently
otherwise inconsistency problems occurred this called as “Updation Problem”.
Deletion problem:
In the resource table when we are tried to delete particular emp automatically dept delete
also delete this is called as “Deletion Problem”.
All this problems are automatically inserting,deleting,updating avoided then we are using
normalization process.

SQL MATERIAL
Order form:
Order no: PK
Order date: Order Order Cust Cust

Cust phone
no date name address
Cust name:
Cust address:
Cust phone:
Item name: Item no:
Candidate key (not in 2NF) P P F F
Orde Item Item

Amount Quantity Discount
r no no name
Quantity: Discount:
Amount:
SUBMIT
Repository items
Frontend application
Item no:
Item name:
Amount:
Order details table
OK FK FK
Order Item
Quantity Discount
PK Item no Item name amount no no
1
111 LED 30000 1
112 A/C 20000 1
1

SQL MATERIAL
3. Third Normal Form:
If a table in 2NF and also all non-key attributes only dependent on candidate key (or)
primary key. If non-key attributes which depends on another non-key attributes then that table
not in 3NF.
Process:
Identified non-key attributes which depends on another non-key attributes putting into
separate table this is called as “3NF Table”. By default this is an master table in this table only
all non-key attributes only dependent on Candidate key (or) Primary key.
Not in 3NF
Candidate key (not in 2NF) non-key attribute PK 2NF Master table
1NF P P
Item
Item name color price tax color Item no Price tax
no
Master black 20 0.2
2NF
Master red 20 0.2
Pen blue 30 0.3
Pen green 30 0.3
PK 3NF
PK
3NF Master table

Item no Price Item no color
Price Tax

SQL MATERIAL
Order form:
Order no:
Order date: PK
Cust no: Order Order Cust Cust Cust

Phone
no date name address no
Cust name:
Cust address:
Cust phone:
Item name: Item no:

PK 2NF item mater table
Refrigerator Item no Item name Amount
Led tv
A/C
Quantity Discount
FK Order details table
Amount: Order no Item no quantity discount
SUBMIT
Customer enquiry PK 3NF mater table PK order mater details FK

Cust no:
Cust Cust Cust Cust Order Order Cust
Cust name: no name address phone no date no
Cust address:
Cust phone:
OK

SQL MATERIAL
Functional Dependency:
If any given two tuples in a relation r then if then x(on attribute set) agrees then y(another
attribute set) cannot agrees the x->y is called as “Functional Dependency”.
x--y Here y is functional dependent on x (or) x is functional determines on y
Candidate key (not in 3NF) non-key attribute child table 3NF master table
Student table 2NF table PK FK PK
Curse Grade Curse Grade
Sname Grade Sname Grade Grade
id value id value
Aaa Cs111 A 4.00 Aaa Cs111 A A 4.00
Bbb Cs111 B 3.00 Bbb Cs111 B
B 3.00
Ccc Cs112 C 2.00 Ccc Cs112 C
Ddd Cs112 A 4.00 C 2.00
Ddd Cs112 A
Logical Diagram:
Grade value
Sname
Grade
Course id
1NF-------------- Remove repeating groups
2NF-------------- Remove partial attributes
3NF--------------- Remove attributes which are not dependent on candidate key.
(or)
Remove non-key attribute which are dependent on another non-key

attributes.

SQL MATERIAL
Course id: Course name:
Tutor id: Tutor name:
Sno Sname DOB Gender Last attendance
UNF 1NF 2NF 3NF
Course id √course id √course id √course id
Course name course name not in course name course name
Tutor id tutor id 3NF tutor id tutor id FK
Tutor name tutor name tutor name
Candidate key 3NF Tutor id PK
Sno √course id √sno tutor name
Sname √sno 2NF sname
DOB not in sname P DOB √sno
Gender 2NF DOB P gender sname
Last attendance 1NF gender P 2NF DOB
Last attendance F √course id gender
√Sno
Last attendance √course id
√Sno
Last attendance

SQL MATERIAL
CLUSTER:
Cluster is a database object which contains no.of tables together and also tables share
same data block. Clusters are created by database administrator. Clusters are used to improve
performance of JOINED query.
Cluster tables are created based on a common columnname this common columnname is
called as “Cluster key”. Clusters are created at the time of table creation only table create
complete after cluster creation not possible. In those table are available in cluster then database
server very fastly retrieve data from cluster table. In cluster rowid’s are same.
Step-1(creating a cluster):
Syntax: create cluster clustername (commoncolumnname datatype(size));
Step-2 (creating an index on cluster):

Syntax: create index indexname on cluster clustername;
Step-3 (create actual table)

Syntax: create table tablename (commoncolumnname datatype(size),col1
datatype(size),…………….) cluster clustername (common
columnname);
Sql> create cluster emp_dept(deptno number(5));

Sql> create index index4 on cluster emp_dept;
Sql> create table c1(deptno number(5),empno number(10),ename varchar2(15),
sal number(10)) cluster emp_dept(deptno);
Sql> create table c2(deptno number(5),dname varchar2(15),loc varchar2(15))

cluster emp_dept(deptno);
Sql>insert into c1(deptno,empno,ename,sal )values(&deptno,&empno,'&ename',

&sal);
Enter value for deptno: 10

Enter value for empno: 7000
Enter value for ename: NARASIMHA
Enter value for sal: 90000
SQL MATERIAL
Sql> /
Sql> select * from c1;
DEPTNO EMPNO ENAME SAL
30 7002 REDDY 80000
10 7000 NARASIMHA 90000
20 7001 SIMHA 99000
Sql> insert into c2(deptno,dname,loc)values(&deptno,'&dname','&loc');

Enter value for dname: NARASIMHA
Enter value for loc: YSR
Sql> /
Sql> select * from c2;
DEPTNO DNAME LOC
30 REDDY INDIA
10 NARASIMHA YSR
20 SIMHA AP
Sql> select rowid from c1;
ROWID
AAAD0AAAEAAAAHEAAA
AAAD0AAAEAAAAHHAAA
AAAD0AAAEAAAAHIAAA
Sql> select rowid from c2;
ROWID
AAAD0AAAEAAAAHEAAA
AAAD0AAAEAAAAHHAAA
AAAD0AAAEAAAAHIAAA

SQL MATERIAL
Sql> drop cluster emp_dept;
ERROR: cluster not empty
We cannot drop cluster if cluster having tables to overcome this problem oracle 8i
introduced INCLUDING TABLES clause to drop cluster along with table.
Syntax: drop cluster clustername including tables;
Sql> drop cluster emp_dept including tables;
All cluster information stored under USER_CLUSTER data dictionary.
Sql> desc user_clusters;
Super key:
It is a single (or) combination of column’s uniquely identified record in a table.
Candidate key:
A candidate key is a combination of attributes that can be uniquely used to identify a

database record without any extraneous data. Each table may have one or more candidate keys.
One of these candidate keys is selected as the table primary key.
(or)
Minimal super key is called “candidate key”. That is a super key does not subset of a
super key.
Primary key:
The PRIMARY KEY constraint uniquely identifies each record in a database table.
Primary keys must contain unique values. A primary key column cannot contain NULL values.
If any one of the candidate key which uniquely identified record in a table.
Foreign key:
A foreign key is a key used to link two tables together. This is sometimes called a
referencing key. Foreign Key is a column or a combination of columns whose values match a
Primary Key in a different table. You can create a foreign key by defining a FOREIGN KEY
constraint when you create or modify a table.

SQL MATERIAL
Alternate key:
Other than primary key, candidate key which uniquely identified those keys are “alternate
key”.
Empno Ename Skill id Skills Voter id

111 Narasimha 11 Oracle V1
111 Narasimha 22 Java V1
111 Narasimha 33 C++ V1
222 Simha 44 C V2
222 Simha 11 Oracle V2
333 Shobha 55 Pl/sql V3
444 Rani ---- ----- V4
555 Shobharani 66 Sql V5
666 Narasimha Reddy 444 C V6
Super key:
1. empno+skill -------------------------- ck-1

2. empno+ename+skill
3. empno+skill+voter id
4. ename+skill -------------------------- ck-2
5. ename+skill+voter id
6. skill+voter id ------------------------- ck-3
7. empno+ename+skill+voter id
Candidate key:
1. empno+skill --------- ck-1

2. ename+skill --------- ck-2
3. skill+voter id -------- ck-3
Primary key:
1. empno+skill ------------- pk
Alternate key:
1. ename+skill
2. skill+voter id

SQL MATERIAL
4. BCNF (Boyce-Codd Normal Form):
A table is in BCNF the every determinate is a candidate key when a table contain
multiple candidate keys.
When a table contain multiple candidate key and also candidate key are overlap and also
one candidate keys non-key attribute which depends on another candidate key non-key attribute
then only process BCNF process.
If a table contain single candidate key the BCNF is also same as 3NF when a table
contain more than one candidate key then deletion problem occurred to overcome this problem
we are using BCNF process. BCNF specifies relationship between candidate key it self.
Identifies one candidate key non-key attributes which depends on another candidate key
attributes non-key attributes then putted those entire attributes separate table this table is called
as “BCNF”. In this table only every determinate is a candidate key.
NOTE: In this BCNF table does not have non-key attributes.
Project table non-key attributes BCNF
Ecode Ename Projcode Hours Ecode Ename

E1 Narasimha P1 10 E1 Narasimha
E2 Simha P1 7
E2 Simha
E3 Reddy P2 5
E3 Reddy
E4 Shobha P2 6
E4 Shobha P1 8 E4 Shobha
FD
1. ecode+Projcode Hours 1.Ecode Ename

Determinate
2. ename+projectcode Hours ck FD
2.Ename Ecode
Determinate ck

SQL MATERIAL
University table Non-key attribute
Projcode Dept Hod Hours

P1 Physics Narasimha 5
P1 Computers Simha 7
P2 Chemistry Reddy 9
P3 Zoology Shobha 8
P4 Botany Rani 10
BCNF
Dept Hod Projcode Dept Hours
Physics Narasimha P1 Physics 5
Computer Simha P1 Computer 7
chemistry Reddy P2 Chemistry 9
Zoology Shobha P3 Zoology 8
Botany Rani P4 Botany 10
1. Projcode+dept --------- ck-1
2. Projcode+hod ---------- ck-2
FD
1. projcode+dept dept
FD
FD
1. Dept Hod
Determinate
2. projectcode+dept Hours ck FD
2. Hod dept
Determinate ck

SQL MATERIAL
5. Fourth Normal Form:
If a table is in 4NF then that table does not contain more than one independent multi
valued attribute.
If a table contains more than two attribute and also identifying a record uniquely
combination attributes and also that table does contain and also one attribute set of values depend
on another attribute set of value and also same attribute set of value which depend on another
attribute set of values and also some attribute are not logically related then only database
designers uses 4NF process.
Process:
Identifies attributes which dependence on another attribute those attributes putted into
another table this table is called as “4NF table”. This table does not contain more than one
independent multi valued attribute. These tables automatically reduce redundancy.
Movie table
Candidate key
Movie Star Producer Movie Star Movie Producer
M1 S1 P1
M1 S2 P1
M1 S3 P1
M1 S4 P2
M1 S5 P2
M2 S6 P3
M2 S3 P3
M2 S2 P3
Before 4NF process in the down table whenever on get a new project then we have to
supply null values for the skill and also whenever having new skill we have to supply null values
for the project id. Whenever we are using 4NF process this problems are avoided.
Assumption:
Each employee has multiple projects and also each employee has multiple skills.

SQL MATERIAL
Project table
Candidate key
Ecode Projcode Skills Ecode Projcode Ecode Skills

Railway 111 C
111 C
reservation Railway
111 111 Java
Railway reservation
111 Java
reservation Library 111 Db2
Railway 4NF 111
111 Db2 management
reservation 111 Oracle
University
Library 111
111 Oracle management 111 Php
management
Library 111 Pl/sql
111 Php
management
University
111
management
111 Pl/sql
6. Fifth Normal Form:
A table is in 5NF then that table does not contain cyclic dependency. If a table contain s
more than two attribute and also identified record uniquely using combination of all attribute and
also all attribute set of values related to one another then only using 5NF.
Generally in 4NF some attributes are not logical related where as in 5NF all attribute are
related. 5NF is also called as “Projection-join normal form”.
In this normal form we are splitting a table into no.of tables and also when we are joining
these tables resultant record must be available on resource table.
Bayer table Bayer product

Bayer Company
Product company
Bayer Product company
B1 Shirt Levis
B1 Jeans Arrow
B1 Shirts Peeps

SQL MATERIAL
Chapter-13
CONVERSIONS
Set operators:
Set operators are used to retrieve data from single (or) multiple tables. This operator is
also called as “Vertical joins”.
1. Union ------------- It returns values one time

2. Union all --------- unique+duplicate
3. Intersect ---------- Common values
4. Minus ------------- Values are in first query those values are not in
second query.
Sql> select job from emp where deptno=10

union
select dname from dept where deptno=20;
JOB
CLERK
MANAGER
PRESIDENT
RESEARCH
Sql> select job from emp where deptno=10

union
select job from emp where deptno=20;
JOB
ANALYST
CLERK
MANAGER
PRESIDENT

SQL MATERIAL
NOTE:
In all database we are using set operators always corresponding expression must belong
to same datatype and also set operator returns first query columnname as column heading’s.
Sql> select deptno from emp
union
select dname from dept;
ERROR: expression must have same datatype as corresponding exp
Sql> select ename from emp
union
select dname from dept;
ENAME
ACCOUNTING
ADAMS
-----------
WARD
18 rows selected.
NOTE:
In oracle we can also retrieve data using set operators if corresponding expressions not
belongs to same datatype also in this case we must use corresponding type conversion function.
Sql> select deptno "deptno",to_char(null)"deptname" from emp

union
select to_number(null),dname from dept;

SQL MATERIAL
DEPTNO DEPTNAME
10
20
30
ACCOUNTING
GNSR
OPERATIONS
RESEARCH
Sql> select deptno "deptno",to_char('wwwww')"deptname" from emp

union
select to_number(99),dname from dept;
DEPTNO DEPTNAME
10 wwwww
20 wwwww
30 wwwww
99 ACCOUNTING
99 GNSR
99 OPERATIONS
99 RESEARCH
CONVERSIONS:
Converting one datatype into another datatype is called as “Conversions”. Conversions

are two types.
1. Implicit Conversion (or) Automatic Conversion
2. Explicit Conversion
1. Implicit Conversion:
NOTE-1: When a sql expression contains a string representing pure numeric values then
oracle server automatically convert into string type into number type.

SQL MATERIAL
Sql> select sal+’100’ from emp;
SAL+'100'
900
1700
--------
Sql> select sal+’100Z’ from emp;

ERROR: ora-1722: invalid number
NOTE-2:
Whenever we are passing number into character function then oracle server internally
uses implicit conversion to convert number type into string type.
Sql> select length(111) from dual;
LENGTH (111)
Sql> select length('abcd') from dual;
LENGTH ('ABCD')
NOTE-3:
Whenever we are passing date string into data function then oracle server automatically
converts date string into date type but here passed parameter must be default date (number-
character-number) format.
Sql> select last_day('15-aug-05')+4 from dual;
LAST_DAY (
04-SEP-05
Sql> select last_day('15-07-05') from dual;
ERROR: not a valid month

SQL MATERIAL
Implicit conversion
From To Assignment Evaluation of expression

Varchar2 (or) char Number Yes Yes
Varchar2 (or) char Date Yes Yes
Number Varchar2 Yes No
Date Varchar Yes No
2. Explicit Conversion:
To_number () to_date ()
Number char date
To_char () to_char ()
1. To_number()
2. To_char()
3. To_date()
1. To_number():
It is used to converting a string representing a numeric value with format into a numeric
value without format.
Sql> select '$24.6'+3 from dual;
ERROR: ORA-01722: invalid number
Sql> select to_number ('$24.6') +3 from dual;
ERROR: ORA-01722: invalid number
Sql> select to_number ('$24.6','$99.9’) +3 from dual;
TO_NUMBER ('$24.6','$99.9') +3
27.6

SQL MATERIAL
NOTE: Only format models working otherwise your own way given not working.
Sql> select to_number ('d24.6','d99.9') +3 from dual;
ERROR: ORA-01481: invalid number format model
2. To_char():
This is an overloading function which is used to convert number type into char type and
also date type into date string.
Sql> select to_char(20000,'99,999.99') from dual;
TO_CHAR(20
20,000.00
Sql> select to_char(20000,'99g999d99') from dual;
TO_CHAR(20
20,000.00
Sql> select to_char(123456.78,'$99,99,99,999.99') from dual;
TO_CHAR(123456.78
$1,23,456.78
Sql> select ename,sal,to_char(sal,'L99,99,999.99','nLS_currency=Rs')
"INDIAN CURRENCY" from EMP;
ENAME SAL INDIAN CURRENCY
SMITH 800 Rs 800.00
ALLEN 1600 Rs 1,600.00
--------- ------- ----------------
G ------------- Group separator
D ------------- Decimal indicator
L -------------- Local currency

SQL MATERIAL
Sql> select to_char(20000,'9,999') from dual;
TO_CHA
######
Sql> select to_char(20000,'99,999') from dual;
TO_CHAR
20,000
Sql> select ename,to_char(sal,'L99,99,99,999') "CURRENCY" from emp;

ENAME CURRENCY
SMITH $800
ALLEN $1,600
---------- ----------
NOTE:
By default oracle server returns local currency ‘$’ if you want return our own currency
format we are using ‘NLS_CURRENCY’ format in third format of the to_char function.
Sql> select ename,sal,to_char(sal,'L99,99,999.99','nLS_currency=Rs')
"INDIAN CURRENCY" from EMP;
Sql> select ename,sal,nvl(to_char(comm),' No Commission') from EMP;
ENAME SAL NVL(TO_CHAR(COMM),'NOCOMMISSION')
SMITH 800 No Commission
ALLEN 1600 300
WARD 1250 500
JONES 2975 No Commission
NOTE:
If we want add leading zero’s then we must use zero in the format. If we want add last
only zero’s then we are using 9 in the format.

SQL MATERIAL
Sql> select to_char(345.6,'0999.9999') from dual;

TO_CHAR(34
0345.6000
NOTE:
If we want to display number datatype values into char format then we are using to_char
with format value.
Sql> select to_char(sysdate,'dd/mm/yyyy') from dual;
TO_CHAR(SY
18/09/2014
3. To_date():
It is used to convert date string into date type.
Sql> select to_date('12/sep/14') from dual;
TO_DATE('
12-SEP-14
NOTE:
Whenever we are using to_date passed parameter return value match with the default date
datatype return value otherwise oracle server returns an error to overcome this problem use a one
parameter formation.
NOTE:
Whenever we ar5e passing date string into date function then oracle server automatically
convert into date string into date type. But in this case passed parameter format number-
character-number otherwise use in to_date() function explicit.
Sql> select last_day('11/JAN/06') from dual;
LAST_DAY(
31-JAN-06
Sql> select last_day(to_date(‘11/03/06’,’DD/MM/YY’)) from dual;
LAST_DAY(
31-MAR-06
SQL MATERIAL
DECODE:
Decode is a conversion function which is used to convert one type into type. Decode
internally used equality (=) operator. It is also called same as if-----then-----elseif construct.
Syntax: decode(colname,value1,stmt1,value2,stmt2,……………stmtN)
If colname=value1 then stmt;
Elseif colname=value2 then stmts;
Sql> select ename,deptno,decode(deptno,10,'ten',20,'twenty','other') from emp;

ENAME DEPTNO DECODE
SMITH 20 twenty
ALLEN 30 other
WARD 30 other
JONES 20 twenty
MARTIN 30 other
we can also use decode function to display row values into column values.
Sql> select job,sum(decode(deptno,10,sal)) “deptno 10”,sum(decode(deptno,20,sal))
“deptno 20” ,sum(decode(deptno,30,sal)) “deptno 30” from emp group by
job;
JOB deptno 10 deptno 20 deptno 30
CLERK 1300 1900 950
SALESMAN 5600
PRESIDENT 5000
MANAGER 2450 2975 2850
ANALYST 6000
Sql> select dname,sum(decode(job,’CLERK’,1,0)) “CLERK”,
Sum(decode(job,’SALESMAN’,1,0)) “SALESMAN”,
Sum(decode(job,’MANAGER’,1,0)) “MANAGER”,
Sum(decode(job,’ANALYST’,1,0)) “ANALYST”,

SQL MATERIAL
Sum(decode(job,’PRESIDENT’,1,0)) “PRESIDENT”
From emp e,dept d where e.deptno=d.deptno group by dname;
DNAME CLERK SALESMAN MANAGER ANALYST PRESIDENT
ACCOUNTING 1 0 1 0 1
RESEARCH 2 0 1 2 0
SALES 1 4 1 0 0
Sql> select dname,sum(decode(job,’CLERK’,1,0)) “CLERK”,

Sum(decode(job,’SALESMAN’,1,0)) “SALESMAN”,
Sum(decode(job,’MANAGER’,1,0)) “MANAGER”,
Sum(decode(job,’ANALYST’,1,0)) “ANALYST”,
Sum(decode(job,’PRESIDENT’,1,0)) “PRESIDENT”
From emp e,dept d where e.deptno(+)=d.deptno group by dname;
DNAME CLERK SALESMAN MANAGER ANALYST PRESIDENT
ACCOUNTING 1 0 1 0 1
OPERATIONS 0 0 0 0 0
RESEARCH 2 0 1 2 0
SALES 1 4 1 0 0
NOTE:
We can also use decode function in update statement for modifying data conditionally.
Q) Write a query to modifying commission column data of employee from emp table based
on following condition?
1. if job=’CLERK’ then comm ------ 10% of sal
2. if job=’SALESMAN’ then comm ---- 20% of sal and all other job’s
comm ------- 30% of sal
Sql> update emp set comm=decode (job,’CLERK’,sal*0.1, ’SALESMAN’,sal*0.2,

sal*0.3);
SQL MATERIAL
7369 SMITH CLERK 7902 17-DEC-80 800 80 20
7566 JONES MANAGER 7839 02-APR-81 2975 892.5 20
-------- -------- ---------------- ------- --------------- ------ ----- ----
NOTE:
If we want count different values from a columns then we are using decode function
within group by clause. Whenever we are using decode function within group by clause then
automatically rows values convert into column values.
Sql> select job,sum(decode(deptno,10,sal)) "deptno 10",

sum(decode(deptno,20,sal)) "deptno 20",
sum(decode(deptno,30,sal)) "deptno 30"
from emp group by job;
JOB deptno 10 deptno 20 deptno 30
CLERK 1300 1900 950
SALESMAN 5600
PRESIDENT 5000
MANAGER 2450 2975 2850
ANALYST 6000

SQL MATERIAL
CASE STATEMENT:
Oracle 8.0 introduced CASE statement and also 8i introduced case conditional stmt. Case
is a statement which is used to decoding the values case stmt is performance very high compare
to decode function.
Method-1: case columnname

When value1 then stmt1
When value2 then stmt2
-------------------------------
Else stmt end
Sql> select ename,sal,deptno,case deptn
when 10 then 'TEN'
when 20 then 'TWENTY'
else 'OTHERS' end from emp;
ENAME SAL DEPTNO CASEDE
SMITH 800 20 TWENTY
ALLEN 1600 30 OTHERS
CLARK 2450 10 TEN
KING 5000 10 TEN
---------- ------- ---- ----------
Method-2: (case conditional statement – oracle 8i)
Case
When colcondition1 then stmt1
When colcondition2 then stmt2
----------------------------------------
Else stmt end

SQL MATERIAL
Sql> select ename,sal,case
when sal<1000 then 'LOW SALARY'
when sal between 1000 and 2500 then 'MEDIUM SALARY'
when sal in(3150,3000,3500,2850) then 'SPECIAL SALARY'
else 'OTHER SALARY' end from emp;
ENAME SAL CASEWHENSAL<10
SMITH 800 LOW SALARIES
ALLEN 1600 MEdIUM SALARY
WARD 1250 MEdIUM SALARY
JONES 2975 OTHER SALARY
MARTIN 1250 MEdIUM SALARY
BLAKE 2850 SPECIAL SALARY
------------ ------- ---------------------------

SQL MATERIAL
Chapter-14
SYNONYMS
Synonym is a database object. Synonym is an reference (or) aliasname for the original
object. Synonyms is an guide another designer,username,object name,database link synonym
provides security for another object. Synonyms are two types.
1. Public Synonyms
2. Private Synonyms
By default synonyms are private synonyms. If we want to create public synonyms then
we are using “create public synonym privileges”.
Sql> grant create public synonym to username;
Syntax: create public synonym Synonymname

for username.objectname@databaselink;
Scott/tiger Simha/Simha reddy/reddy
Sql>grant all on emp sql>select * from scott.emp; (√) sql>select * from scott.emp; (√)
To Simha,reddy;
Sql>create synonym AAA for
Scott.emp;
Sql>select * from AAA; (√) sql>select * from AAA; (√)
Error: table or view does not

exit
Sql>crate public synonym ABC
For scott.emp;
Error: insufficient privileges

SQL MATERIAL
Sql>conn sys as sysdba
Enter password: SYS
Sql>grant create public synonym
to Simha;
Sql>conn Simha/Simha;
Sql>create public synonym ABC
For scott.emp;
Sql>select * from ABC; (√) sql>select * from ABC; (√)
All synonyms information stored undered USER_SYNONYMS data dictionary.
Sql> desc user_synonyms;

SQL MATERIAL
Chapter-15
TCL (Transaction Control Language)
Transaction:
It is a logical unit of work in between two points Oracle having two transaction
commands.
1. Commit
2. Savepoint
1. Commit:
This command is used to save the transaction permanently hard disk.
Syntax: COMMIT
2. Savepoint:
It is a logical mark between transactions.
Syntax: SAVEPOINT
Rollback to particular Savepoint:

Syntax: rollback to savepointname;
Sql> create table test(sno number(5));
Sql> insert into test values(&sno);
SNO


SQL MATERIAL
Sql> update test set sno=999 where sno=3;

Sql> delete from test where sno=2;
Sql> Savepoint s1;
SNO
999

Sql> Savepoint s2;
Sql>select * from test;
SNO
999
8
SQL MATERIAL

SNO
999
Sql> rollback to s2;

SNO
999
Sql> rollback to s1;

SNO
999

SQL MATERIAL
Chapter-16
Oracle Server Architecture

Oracle server mainly consist of two parts
1. Storage Area
2. Instance
1. Storage Area:
Whenever we are installing oracle server automatically three files are created in a hard
disk. These three files are also called as ‘Physical Database’. This files are stored ‘Physical
Storage Area’.
1. Data files (.dbf)

2. Control files (.ctl)
3. Log files (.log)
Sqlplus client tool RAM Memory Oracle Server Architecture

I
User name: N
SGA
Password: S
Database Library cache Java Large Redo log
OK T
buffer pool pool buffer
A cache Dictionary
cache
N
User
process C
E DBWR SMON PMON CKPTR LGWR
Server Process Redo log files (.log)

Data files (.dbf) Control files (.ctl)
process
PGA
G1 M1 M2
SGA----------- System Global Area (or) Shared Global Area
Database
Storage Area
G2 M1 M2
Hard disk
Groups Member
SQL MATERIAL
PGA----------- Private Global Area LGWR-------- Log Writer

DBWR-------- Database Writer SMON----------- System Monitor
PMON-------- Process Monitor CKPT------------ Checkpoint
SGA----------- System Global Area (or) Shared Global Area
STORED FILES:
Data files stores database objects physically that is file stores index, procedures, tables,
function, …………….etc. Data files also rollback segment in log onwards this segment’s also
called as “Undo Segment”. All data files information stored under DBA_DATA_FILE data
dictionary. If you want view path of the data file then we are using FILE_NAME property from
DBA_DATA_FILE data dictionary;
Sql> desc dba_data_files;

Sql> select file_name from dba_data_files;
FILE_NAME
C:\ORACLEXE\ORADATA\XE\USERS.DBF
C:\ORACLEXE\ORADATA\XE\SYSAUX.DBF
C:\ORACLEXE\ORADATA\XE\UNDO.DBF
C:\ORACLEXE\ORADATA\XE\SYSTEM.DBF
CONTROL FILES:
Control files controls all other files in storage area this file extension is .ctl. Oracle 9i
introduced minimum three control files in storage area control files also stores database. All
control files information stores under V$CONTROLFILE data dictionary. Control files are used
by database administrator in recovery process.
Sql> desc V$CONTROLFILE;

Sql> select name from v$controlfile;
NAME
C:\ORACLEXE\ORADATA\XE\CONTROL1.DBF

SQL MATERIAL
REDO LOG FILE:
Redo log file store committed information redo log buffer. This files also used by
database administrator. In recovery process all files information log file stored under
V$LOGFILE data dictionary.
If we want view path of the log file then we are using MEMBER property from
V$LOGFILE data dictionary.
Sql> desc v$logfile;
Sql> select member from v$logfile;
MEMBER
C:\ORACLEXE\APP\ORACLE\FLASH_RECOVERY_AREA\XE\ONLINELOG\O1_MF_2_9VHP6GDM_.LOG
C:\ORACLEXE\APP\ORACLE\FLASH_RECOVERY_AREA\XE\ONLINELOG\O1_MF_1_9VHP6DQX_.LOG
2. INSTANCE:
Whenever we are connect into the server using a client oracle server automatically on
instance created in memory area this instance having two parts.
1. SGA
2. PGA
SGA:
SGA is also called as System Global Area (or) Shared Global Area. SGA consists of set
buffer these buffers are
1. Database Buffer Cache

2. Shared Pool
3. Java Pool
4. Large Pool
5. Redo Log Buffer

SQL MATERIAL
Whenever we are subheading SQL, PL/SQL codes into the database server that code
store in library cache within shared pool library cache. Always reduces passing that is library
cache checks syntax and semantic checking of the SQL, PL/SQL code. Whenever user
requesting data from the table oracle server process checks request table available in database
buffer cache (or) Not If request object is not available in database buffer cache the DBWR
process checks request table available in data files. When it is available then copy of the table
created in database buffer cache.
Oracle server
SGA
11111111111111
11111111111111 User
11111111111111
11111111111111
DBWR Select * from emp;
Data files
Emp
Whenever we are subheading update, delete stmt those transaction are performing redo
log buffer that is redo log of buffer copy of the table from database buffer cache after
modification whenever we are using commit,rollback those values are stored in data files,redo
log files and also redo log changes effected in database buffer cache.

SQL MATERIAL
Redo
SGA
Database buffer cache Redo log buffer
101010101010 101010101010
101010101010 101010101010
101010101010 101010101010
101010101010 101010101010
User
101010101010 101010101010
101010101010 101010101010
Update
DBWR Log wr
Delete
Sql> commit
Data files Redo log

Emp ----------------
----------------
----------------
----------------
----------------
----------------
Library cache also stores cache value defined in sequence object. Shared pool also
contains dictionary cache which executed DCL related data dictionary property. Java pool
executes java related object and also large pool used by database administrator in recovery
process. Server process also contains PGA which uniquely identifies each client connect to the
server. Whenever we are using DML transaction new values for the transaction also stored in
redo log buffer this values are also called as “ONOFF”. Whenever user using commit (or) 1/3 fill
of redo log buffer then automatically redo log buffer data transfer into redo log files this files are
used by database administrator in recovery process.

SQL MATERIAL
Undo area
sqlplus client Ram memory dirty reader UNDO
User name: SGA
5 1
Password:
Database Redo log buffer
Buffer cache
Undo file
T1
Update 1
1 5
Redo log
Delete
2
files
Commit 3
Data files
Hard disk
Whenever we are using DML transaction old data stored in undo area within database
buffer cache and also new data stored in dirty reader within database buffer cache whenever user
using commit automatically new data permanently stored in data files and also old data stored in
undo files within hard disk. This undo file used in FLASHBACK query. Flashback query are
used by database administrator content of a table can be query based on a specific point of time.
Oracle 9i introduced flashback query to retrieve accidental data after commit transaction also.
Oracle server having three structures:
1. Physical Structure
2. Logical Structure
3. Logical Storage Structure
1. Physical Structure:
Physical structure handle database administrator only. Physical structure contain these are
data files, control files, redo log files. This structure is visible in the operating system.
2. Logical structure:
Logical structure not visible to the operating system this structure contains database
object like Tables, Views, Index, Sequence, Procedure, Trigger, Function, Package,…………etc.
These structures handle by either Developer (or) Administrator.

SQL MATERIAL
3. Logical Storage Area:
This structure is handling by database administrator only.
1. Tablespace
2. Segment
3. Extent
4. Block
Tablespace:
Tablespace Tablespace Tablespace

---------- ---------- ----------
Logical storage ---------- ---------- ----------
---------- ---------- ----------
Not visible in OS
Physical Structure Data Data Data
Files Files Files

Visible in OS
Generally without starting database also we are viewing physical structure of the
database. Whereas whenever starting database then only we doing logical storage structure.
Tablespace is nothing but collection of data files. One data file belongs to one tablespace only.
Whenever we are installing oracle server automatically SIX tablespace are created. If you want
run oracle minimum TWO tablespace are required. These are SYSAUX tablespace, SYSTEM
tablespace. System tablespace contains metadata that is system tablespace all data dictionary.
Database administrator also creating tablespace explicitly uses.
Syntax: create tablespace tablespacename data file ‘path of data file’
All tablespace information is stored under DBA_TABLESPACES data dictionary.

SQL MATERIAL
Sql> conn sys as sysdba

Enter password: SYS
Sql> desc dba_tablespaces;
Sql> select tablespace_name from dba_tablespaces;
If we want view data files and related tablespace then we are using DBA_DATA_FILES
data dictionary.
Sql> desc dba_data_files;

Sql> select tablespace_name,file_name from dba_data_files;
TABLESPACE_NAME FILE_NAME
USERS C:\ORACLEXE\ORADATA\XE\USERS.DBF
SYSAUX C:\ORACLEXE\ORADATA\XE\SYSAUX.DBF
UNDO C:\ORACLEXE\ORADATA\XE\UNDO.DBF
SYSTEM C:\ORACLEXE\ORADATA\XE\SYSTEM.DBF
Users---logical storage structure physical storage structure
C:\ORACLEXE\ORADATA\XE\USERS.DBF
System tablespace (metadata) Bits----bytes----block Segment

[all data dictionary] Extent 1
Oracle Database
Extent 2
Segment1 Segment2
------------- --------------
------------- --------------
-------------- -------------
-------------- --------------
Segment1 Segment2
Tablespace data files sysaux tablespace undo tablespace

SQL MATERIAL
Segment, Extent, Block:
Sql> create table T1(sno number(5),varchar2(10));
Sql> insert into T1 values(1,’aaa’);
Sql> insert into T1values(2,’bbb’); Block

Header
Segment 2mb Row address row
. .
Dictionary
Name of table Table

Extent1 1mb Extent2 1mb
1 aaa Dictionary
2 bbb
Data
Tablespace Blocks
Segment is an collection of extents that forms a database object likes tables, views, index,
…………..etc.
LOCKS
Locks are a mechanism which prevents unauthorized access for our resources. All
database systems having two types lock.
1. Row Level Lock
2. Table Level Lock
1. Row Level Lock:
Oracle 6.0 introduced row level lock. In this method we are locking set of rows using for
update clause. This clause used in select statement only using locks only that perform use DML
statement not using another user.

SQL MATERIAL
Syntax: select * from tablespace where condition for update [nowait];
Whenever we are forming lock another user query the data but they cannot perform DML
operation. Whenever we are use COMMIT, ROLLBACK when lock are release.
Scott/tiger Simha/simha
Sql> select * from emp where deptno=10
for update; Sql> select scott.emp set sal=sal+100
where deptno=10;
sql> commit; [for release locks] [we cannot perform DML]
NOWAIT:
Nowait is a optional clause is used in FOR UPDATE clause. Wherever we are using this
clause oracle server automatic get control into current user if another user not releasing lock in
this case oracle server returns an ERROR: ora-0054: Resource busy.
Sql> select * from scott.emp where deptno=10 for

update;
Sql> select * from emp where dept=10
For update nowait;
ERROR: ora-0054: Resource busy
Sql> /
NOTE:
In all database whenever we are using DML statement automatically database server
internally uses exclusive lock. Wherever we are using COMMIT,ROLLBACK then only these
lock are released.

SQL MATERIAL
2. Table Level Lock:
In this method we are locking a table. These are 2 types of table level lock using by
database developer.
1. Share Lock
2. Exclusive Lock
1. Share Lock:
Whenever we are using this lock another uses query the data but they cannot perform
DML operations and also at a time no.of user lock resource.
Syntax: lock table tablename in share mode;
Sql> lock table emp in share mode; Sql> select * from scott.emp; ( √ )
Sql> lock table scott.emp in share mode; ( √ )
Sql> update scott.emp set sal=sal+100;
Sql> commit; (for release lock) [we cannot perform DML]
2. Exclusive Lock:
Whenever we are using this lock another user query the data but we cannot perform DML
operations and also at a time only one user lock the resource.
Sql> lock table emp in exclusive mode; Sql> select * from scott.emp; ( √ )
Sql> lock table scott.emp in share mode; (  )
Sql> commit; (for release lock)
NOTE:
In all database procedure language when we are using cursor locking mechanism
internally database server uses exclusive lock.

SQL MATERIAL
DEADLOCK:
In oracle whenever we are try to perform same operation in a single resource through
different session then automatically deadlock occurred. In this case oracle server returns an
ERROR: ora-0060: Deadlock detected while waiting for resource.
Sql> update emp set sal=sal+100 Sql> update emp set sal=sal+100 where deptno=20;
Where deptno=10; Sql> update emp set sal=sal+100 where deptno=10;
Sql> update emp set sal=sal+100 [deadlock detected]
where deptno=20;
[we cannot perform DML]
ERROR: ora-0060: Deadlock detected
while waiting for resource
Sql> commit; (for release lock)

SQL MATERIAL
Chapter-17
TABLES
Global Temporary Tables (GTT):
Global Temporary Tables stores data temporary that is these table data stored in RAM
memory area. These tables are used by database administrator to process data very fastly
retrieve. Oracle having 2 type of global temporary table
1. Transaction Specific
2. Session Specific
1. Transaction Specific:
In this type of GTT stores data until we are using commit (or) rollback.
Syntax: create global temporary table tablename(col1 datatype(size),………….)
On commit delete rows;
Sql> create global temporary table GTT(sno number(5)) on commit delete rows;
Sql> insert into GTT values(1);
Sql> select * from gtt;
SNO
Sql> commit;
No Rows Selected

SQL MATERIAL
2. Session Specific:
This Global Temporary Tables data available in particular sessions if we are giving
commit also another session cannot view that data.
Syntax: create global temporary table tablename(col1 datatype(size),…………….)

On commit preserve rows;
Session-1:
Sql>create global temporary table GTT(sno number(5)) on commit preserve rows;
SNO
Sql> commit;
SNO
Session-2:
No Rows Selected

SQL MATERIAL
NESTED TABLE:
Oracle 8.0 introduced nested table. Table within another table is called as “Nested
Table”. Nested table is a userdefined table. This is used to multiple data store in a single unit.
Generally in oracle if we want to store no.of data items then we are using index by table in
PL/SQL. But these tables not allowed to stores this table permanently in database. To overcome
this problem oracle 8.0 introduced extension of the index by table called as “Nested Table”. This
is used to store no.of data items permanently in database using SQL.
Step-1: creating an objecttype
Step-2: create an nested table type
Step-3: create an actual table
Creating an object type:
This is a userdefined datatype. Which is used represent using different datatype in a

single unit it is also same as structure in ‘C’ language.
Syntax: create or replace type typename as
object(attributename1 datatype(size),attribute2 datatype(size),…..);
Create a nested table type:
Syntax: create or replace type typename as table of objecttypename;
Create an actual table:
Syntax: create table tablename (col1 datatype(size),………colN nestedtabletype)
Nested table colN store as anyname;
Sql> create or replace type nt1 as
object(bookid number(5),bookname varchar2(15),price number(10));

SQL MATERIAL
Sql> create or replace type FAN as table of nt1;
Sql> create table student(sno number(5),sname varchar2(15),col3 FAN)
nested table col3 store as ABC;
Sql> desc student;
Name Null? Type
SNO NUMBER (5)
SNAME VARCHAR2 (15)
COL3 FAN
Nested table type
Sql> desc fan;
Fan TABLE OF NT1
Name Null? Type
BOOKID NUMBER (5)
BOOKNAME VARCHAR2 (15)
PRICE NUMBER (10)
If we want to insert data into nested table columns then we must using constructor. Here
constructor name is same as typename.
Sql> insert into student values(1,'NARASIMHA',
FAN(nt1(101,'JAVA',500),nt1(102,'ORACLE',3000)));
Sql> select * from student;
SNO SNAME COL3 (BOOKID, BOOKNAME, PRICE)
1 NARASIMHA FAN (NT1 (101, 'JAVA', 500), NT1 (102, 'ORACLE', 3000))
Sql> select col3 from student;

SQL MATERIAL
COL3 (BOOKID, BOOKNAME, PRICE)
FAN (NT1 (101, 'JAVA', 500), NT1 (102, 'ORACLE', 3000))
If we want to view nested table data then we are using table operator.
Number(5) varchar2(15) FAN
Sno Sname Col3 Bookid Bookname Price

1 NARASIMHA 101 JAVA 500
102 ORACLE 3000
Syntax: select * from table (select nestedtablename from tablename);
Sql> select * from table( select col3 from student);
BOOKID BOOKNAME PRICE
101 JAVA 500
102 ORACLE 3000
This table operator used in insert,update,delete,select statement all userdefined types

stored in USER_TYPES data dictionary.
Sql> desc user_types;
Sql> select TYPE_NAME,ATTRIBUTES,METHODS,TYPEID,PREDEFINED
from user_types;
TYPE_NAME ATTRIBUTES METHODS TYPEID PRE
FAN 0 0 NO
NT1 3 0 NO
We can also drop type using drop type typename
Sql> drop type typename;

SQL MATERIAL
PARTITIONS:
Table can be separation into number of tables is called ‘Partition’. Partitions table are
created by database performance of the application. These tables are used in backup and
recovery process. Generally partitions table created when database table having large amount of
data partition are created based on a column this is called ‘Partition Key’.
1. Range Partition
2. List Partition
3. Hash Partition
1. Range Partition:
If we want view partition then we are using following syntax. This method created based
on range of values.
Syntax: select * from tablename partition (partitionname1,partitionname2,……);
Syntax: create table tablename(col1 datatype(size),col2 datatype(size),………..)
Partition by range(keycolname) (partition partitionname values less

than (actual value),
--------------------------------------------------------------------------------
Partition partitionname values less than (maxvalue));
Sql> create table test(sno number(10),ename varchar2(15),sal number(10)) partition by

range(sal)(partition p1 values less than(1000),partition p2 values less
than(2000),partition other values less than(maxvalue));
Sql> insert into test values(……………………………..);
Sql> select * from test partition(p1);
Sql> select * from test partition(p2);

SQL MATERIAL
2. List Partition:
Oracle 9i introduced list partition. Using this partition we can also create partition based
character datatype columns. Database administrator creates on list of values.
Syntax: create table tablename(col1 datatype(size),……………….)

Partition by list(keycolname)
(partition partitionname values(value1,value2,…………..),
---------------------------------------------------------------------------
Partition partitionname values(default));
Sql> create table test(sno number(10),name varchar2(10))
Partition by list(name)
(partition p1 values(‘india’,’pak’),
Partition p2 values(‘usa’,’usk’),
Partition other values(default));
Sql> inert into test values(……………………………);
3. Hash Partition:
In this method partition when we are specified partition number then automatically
created oracle server based on hash partition.
Syntax: create table tablename(col1 datatype(size),………)
Partition by hash(keycolname) partitions anynumber;
Sql> create table test(sno number(10),sal number(10)) partition by hash(sal) partitions 5;
If we want views partition information using USER_TAB_PARTITIONS data dictionary.
Sql> desc user_tab_partitions;
Sql> select partition_name from user_tab_partitions where table_name=’TEST’;

SQL MATERIAL
HIERARCHICAL QUERIES:
In relational database we can also stored hierarchical data. If we want to store

hierarchical data then that table contains minimum three columns and also two columns must
belongs to same datatype and relations. In oracle if we want to retrieve hierarchical data then we
are using following clause. Whenever we are using hierarchical relational one column is null
value mandatory.
1. Start With
2. Connect By
3. Level
1. Start with:
This clause is used to specify searching condition in hierarchy.
Syntax: start with condition
2. Connect by:
This clause is used to specify relationship between hierarchical (Parent, Child) column
using prior operator.
Syntax: connect by prior parentcolname=childcolname
3. Level:
Oracle provided LEVEL pseudo column. In hierarchical is pseudo column which assign
numbers to each level in tree structure.
Syntax: select level,colname1,colname2,………………..
From tablename where condition
Start with condition
Connect by prior parentcolname=childcolname;

SQL MATERIAL
Q) Write hierarchical query to display then employee’s who are working under KING
from emp table using hierarchy?
Sql> select level,ename from emp start with ename=’KING’ connect by prior empno=mgr;
LEVEL ENAME
1 KING
2 JONES
3 SCOTT
4 ADAMS
3 FORD
--- --------
(Or)
Sql> select level,lpad(' ',level*2)||ename from emp start with ename=’KING’
connect by prior empno=mgr;
LEVEL LPAD ('', LEVEL*2)||ENAME
1 KING
2 JONES
3 SCOTT
4 ADAMS
3 FORD
4 SMITH
2 BLAKE
3 ALLEN
3 WARD
3 MARTIN
3 TURNER
3 JAMES
2 CLARK
3 MILLER

SQL MATERIAL
KING
JONES BLAKE CLARK
SCOTT FORD ALLEN WARD MARTIN TURNER JAMES MILLER
ADAMS SMITH
NOTE:
Oracle 9i introduced SYS_CONNECT_BY_PATH() which returns path of the hierarchy

in tree structure.
Syntax: SYS_CONNECT_BY_PATH(columnname,’delimitername’)
Sql> select level,sys_connect_by_path(ename,'--->') "emp names tree structure"

from emp start with mgr is null connect by prior empno=mgr;
LEVEL employee names tree structure
1 --->KING
2 --->KING--->JONES
3 --->KING--->JONES--->SCOTT
4 --->KING--->JONES--->SCOTT--->ADAMS
3 --->KING--->JONES--->FORD
4 --->KING--->JONES--->FORD--->SMITH
2 --->KING--->BLAKE
3 --->KING--->BLAKE--->ALLEN
3 --->KING--->BLAKE--->WARD
3 --->KING--->BLAKE--->MARTIN
3 --->KING--->BLAKE--->TURNER

SQL MATERIAL
3 --->KING--->BLAKE--->JAMES
2 --->KING--->CLARK
3 --->KING--->CLARK--->MILLER
Q) Write a query to display the emp’s who are working under BLAKE from emp table
using hierarchy?
Sql> select level,sys_connect_by_path(ename,'--->') "employee names tree structure"

from emp start with ename='BLAKE' connect by prior empno=mgr;
1 --->BLAKE
2 --->BLAKE--->ALLEN
2 --->BLAKE--->WARD
2 --->BLAKE--->MARTIN
2 --->BLAKE--->TURNER
2 --->BLAKE--->JAMES
SQL> select level,sys_connect_by_path(ename,'--->') "employee names tree structure"

from emp start with ename='MILLER' connect by empno=prior mgr;
1 --->MILLER
2 --->MILLER--->CLARK
3 --->MILLER--->CLARK--->KING
NOTE:
Prior is a unary operator which specifies parent column i.e oracle server searching data.
NOTE:
Oracle 10g introduced CONNECT_BY_ISLEAF pseudo column which returns 1 to the

leaf node and 0 to the all other nodes.
Sql> select level,ename,connect_by_isleaf from emp start with mgr is null connect by
prior empno=mgr;

SQL MATERIAL
LEVEL ENAME CONNECT_BY_ISLEAF
1 KING 0
2 JONES 0
3 SCOTT 0
4 ADAMS 1
3 FORD 0
4 SMITH 1
--- ----------- -----
NOTE:
Oracle 10g introduced CONNECT_BY_ROOT operator which returns root node in the
hierarchy.
Syntax: connect_by_root columnname
Sql> select level,ename,connect_by_root ename "RootnodeENAME" from emp

start with mgr is null connect by prior empno=mgr;
LEVEL ENAME RootnodeENAME
1 KING KING
2 JONES KING
3 SCOTT KING
4 ADAMS KING
3 FORD KING
NVL():
Nvl() is predefined function which is to replace userdefined function in place of null.
Syntax: NVL(exp1,exp2)
If expression1 is null then its return expression2 otherwise its returns expression1.
NVL2():
Oracle 9i introduced NVL2() function this function excepts three parameter.
Syntax: NVL2(exp1,exp2,exp3)

SQL MATERIAL
If exp1 is null then its returns exp3 otherwise its returns exp2.
Sql> select nvl2(null,30,60) from dual;

NVL2 (NULL, 30, 60)
60
Sql> select nvl2(10,30,60) from dual;

NVL2 (10, 30, 60)
30
Q) Write a query to update emp’s based following condition using NVL2() function?
1) if comm is null then update comm ------ 500
2) if comm is not null then update comm -- 200
Sql> update emp set comm=nvl2(comm,comm+200,500);
COALESCE():
Coalesce() is an ASCII function which return first not null value in the given expression
coalesce() no.of expression.
Syntax: coalesce(exp1,exp2,……………..)
Sql> select coalesce(null,null,10,null,30) from dual;
COALESCE(NULL,NULL,10,NULL,30)
10
Sql> select coalesce(null,30,null,10,null,50) from dual;
COALESCE(NULL,30,NULL,10,NULL,50)
30
Sql> select coalesce(40,30,null,10,null,50) from dual;
COALESCE(40,30,NULL,10,NULL,50)
40
Sql> select ename,comm,sal,coalesce(comm,sal) from emp;

SQL MATERIAL
ENAME COMM SAL COALESCE(COMM,SAL)
SMITH 800 800
ALLEN 300 1600 300
WARD 500 1250 500
JONES 2975 2975
MARTIN 1400 1250 1400
BLAKE 2850 2850
Nvl() internally uses implicit conversion. Whereas coalesce() does not uses implicit
conversion and also coalesce() performance very high compare to nvl().
Sql> select nvl('A',sysdate) from dual;
NVL ('A',S
A
Sql> select coalesce('A',sysdate) from dual;
ERROR: ORA-00932: inconsistent datatype: expected CHAR got DATE
ROLLBACK:
Rollback is nothing but undo from query in oracle when ever using rollback oracle server
automatically past undo area into dirty reader within buffer cache.
Syntax: rollback;
FLASHBACK QUERY:
Oracle 9i introduced flashback query. Flashback query allows content of a table to be

retrieved with reference to a specific point of time using “AS OF” clause.
Flashback query are handle by administrator using flashback query then we are retrieving
accidental data flashback query works based on undo file to retrieve data using flashback query
following two methods.
1. Using timestamp datatype

2. Using SCN (system Change Number)
SQL MATERIAL
1. Using Timestamp:
Syntax: select * from tablename AS OF timestamp(specifictime)

Sql> create table test(sno number(10),name varchar2(15));
Sql> insert into test values(&sno,'&name');
SNO NAME
100 simha
101 narasimha
102 reddy
103 narasimha reddy
Sql> commit;
Sql> insert into test values(11,'shobha');
SNO NAME
100 simha
101 narasimha
102 reddy
103 narasimha reddy
104 shobha
Sql> commit;
Sql> delete from test;
No Rows Selected
Sql> select * from test as of timestamp(systimestamp-interval '2' minute);
SNO NAME
100 simha
101 narasimha
102 reddy

SQL MATERIAL
103 narasimha reddy
104 shobha
2. Using SCN:
Whenever we are using transaction oracle server automatically generates a unique

identification number in undo file. This number is called as “SCN”. If we want to view number
then we are using CURRENT_SCN property from V$DATABASE data dictionary.
Syntax: select current_scn from v$database;
To retrieve data:
Syntax: select * from tablename as of scn scnnumber;

Sql> conn sys as sysdba
Enter password: SYS
Sql> create table ABC(sno number(5));
Sql> desc v$database;
Sql> select current_scn from v$database;
CURRENT_SCN
6048901
Sql> insert into ABC values(50);
Sql> select * from ABC;
SNO
50
Sql> select count(*) from ABC;
COUNT (*)
1
Sql> select count(*) from abc as of scn 6048901;
COUNT (*)
0

SQL Book-1

Uploaded by

Copyright:

Available Formats

SQL Book-1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SQL Book-1

Uploaded by

Copyright:

Available Formats

SQL MATERIAL

DATA: It is a collection of “RAW FACTS”.

Ex: 1.Student Marks

Ex: 1.Marks Sheets

It is a place where we can store data or information.

1. Books & Papers

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 1

Program-1 Program-2 Hard disk

Fig: Flat file approach

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 2

Database automatically maintains consistency data. Because in database every transaction

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 3

Architecture contains three levels:

1. Internal Data Level

2. Conceptual Data Level

3. External Data Level

Logical data independence

Create table item(itemname varchar2(10),

Physical data independence

Item name type=byte(8) offset=(10);

All dbms s/w implemented based on data independence.db structure is independent on

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 4

Logical data independence:

When ever we modifying conceptual level it is not affected external level.

Physical data independence:

External Conceptual Internal

Fig: ANSI/SPARC Architecture (or) DBMS Architecture

1. Internal Data Level:

2. Conceptual Data Level:

3. External Data Level:

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 5

Data model describe how data layout store in conceptual level.

They are three level model used by organization:

1. Hierarchical Data Model

2. Network Data Model

3. Relational Data Model

1. Hierarchical Data Model:

Root Eid Ename Deptname

2005 50000 2005 50000

Eid Ename Deptname

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 6

In 1960 IBM introduced IMS (information management system) product based on

2. Network Data Model:

In 1970 CODASYL (conference on data system languages) implemented network data

In 1970 IBM implemented IDMS (integrated database management system) products. If

1000 Smith Admin

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 7

3. Relational Data Model:

Relational model is three components:

3. Set of integrity rules

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 8

RSI (1979) Oracle Corporation

SQL PL/SQL DYNAMIC SQL

G.NARASIMHA REDDY (B.Tech) ph no: 9014909280 Page 9

Oracle 2.0 ---------- 1979

 First public release

Oracle 3.0 --------- 1983

Oracle 4.0 ---------- 1984

Oracle 5.0 ----------- 1985