PLSQL

PLSQL Satya Manchiganti’s

Page 1 of 54

PLSQL Programming Language of Structure Query Language


Page 2 of 54

Features:

Supports to execute a block of commands as a unit.

Supports variables and constants.

Supports conditional constructs.

Supports iteration control statements.

Supports error handling.

Supports to define composite data types.

Supports to execute the code automatically based on the event threw data base triggers.

Supports to store and shared the code using subprograms.

PLSQL Block:

It is an collection of sql & program language statements.

There are 2 types.

1. Anonymous block

Name less block

Temporary block

Syntax:- Declare

<variable decl>

Begin

<exec stmts>;

End;

2. Named Block

Block with fixed name.

Stored permanently in database

Ex:- Subprograms.

Variable declaration:

All sql data types are supported.

Boolean is supported.

Declare

Veno number(4):=7900;

Vename varchar2(90);

Vjob varchar2(90) NOTNULL :=’CLERK’;

Doj date default sysdate;

Flag Boolean:= TRUE;

Pin constant number(6):=500038;

:= Assignment operator

= Comparission operator

Executable Statements:

DML and TCL are not valid.

DDL and DCL are not allowed.


Page 3 of 54

Select ----- into statement.{ Used to retrieve data into Plsql variables.}

Select <column list> into <variables> from <table name>

Where <condition>;

Used to print message & variables contents onto screen.

DBMS_OUTPUT.PUT_LINE(‘message Text’ || Variables);

Comments

--Comment line

/*Multi

Line

Comment*/

Activates the DBMS stmt output buffer.

>set server output on;

Ex: Plsql Program retrieves the employ details and calculates the net salary and prints.

Declare

Veno number(4):=3400; or &employ;

Vname varchar2(90);

Vsal number(12,2);

Vcomm number(12,2);

Net number(12,2);

Begin

--Retrieve data from table.

Select ename, sal, comm. Into vname, vsal, vcomm from emp

Where empno=veno;

--Calculated net salary

Net:= vsal+nvl(comm., 0);

Dbms_output.put_line(‘employ details are:’||veno||’’||vename||’’||net);

End;

/

To save to local system(.sql file)

Save <filename>

To display the file

Get filename

To open the file in editor

Ed filename

To execute the file

Start filename or @filename


Page 4 of 54

Attribute Declaration:

Used to define the Plsql variables dynamically according to the table structure.

1. %Type:

Column type declaration.

Used to define the variables according to the specific column structure.

Syntax: variable <tablename>.<columnname>%type;

Ex: Declare

--using %type declaration

Veno emp.empno%type := &employ;

Vsal emp.sal%type;

Vcomm emp.comm%type;

Net emp.sal%type;

2. %Rowtype:

Record type declaration

Used to define the variables according to the complete table structure.

Syntax: variable <tablename>%rowtype;

Ex: declare

--using %Rowtype declaration.

I emp%rowtype;

Net number(19,9);

Begin

i.empno :=&empno;

select ename, sal, comm. Into i.ename, i.sal, i.comm from emp

where empno = i.empno;

net:= sal+nvl(comm., 0);

dbms_output.put_line(‘Employ details are:’||i.empno||’’||i.ename||’’||net);

end;

/


Page 5 of 54

Conditional Constructs:

Used to check for multiple conditions while manipulating data in plasql

1. If

2. case(8i)

1. Simple If

If <condition 1> then

<exec stmts>;

Elsif <cond 2> then

<exec stmts>;

--

--

Else < exec stmts>;

End if;

Ex: Plsql block checks for existing new comm. And assign new comm..

Declare

Veno emp.empno%type := &employ;

Vname emp.ename%type;

Vcomm emp.comm%type;

Begin

Select ename, comm. Into vname, vcomm from emp

Where empno = veno;

If vcomm is null then

Vcomm :=3000;

Elsif vcomm = 0 then

Vcomm :=2500;

Else vcomm := vcomm+vcomm*.25;

Endif;

Update emp set comm.= vcomm wehre empno = veno;

Dbms_output.put_line(veno||’’||vname|’’||vcom);

Commit;

End;

/

2. Case(8i) :

o Used to check for multiple conditions easily.

o It will check for equality condition.

o It can be used in “select” statement

Syntax:

1. Case <variable>

When <value1> then

<exec stmts>;

When <value2> then

<exec stmts>;

--

--


Page 6 of 54

--

Else <exec stmts>

End case;

2. Case

When <cond1> then <value1>

When <cond2> then <value2>

Else <value>

End case;

Using case construct:

Ex:

Declare

Grade char(1):=’&grade’;

Begin

Case grade

When ‘A’ then

Dbms_output.put_line(‘Grade is A’);

When ‘B’ then

Dbms_output.put_line(‘Grade is B’);

When ‘C’ then

Dbms_output.put_line(‘Grade is C’);

Else

Dbms_output.put_line(‘Grade is C’);

End case;

End;

/

Case in select:

select empno, ename, sal, job,

case

when job = ‘CLERK’ then ‘C’

when job = ‘salesman’ then ‘B’

when job in (‘Manager’, ‘Analyst’) then ‘B+’

when job = ‘President’ then ‘A’

else ‘D’

end “Grade”

from emp;

--It is used to generate reports.


Page 7 of 54

Iteration control statements: (LOOPS)

Supports to execute a block of stmts until conditions are true.

4 types:

1. simple loop

2. while loop

3. numeric for loop

4. cursor for loop

1. Simple Loop:

It is an infinite loop block.

Syntax: Loop

<exec stmts<;

End loop;

To break simple loop :

o Exit when(cond);

o If(cond) then

Exit;

End if;

Ex: --Prints first 10 numbers onto screen

Declare

A number(3):= 1;

Begin

Dbms_output.put_line(‘The numbers are:’);

Loop

Dbms_output.put_line(‘a’);

A := a+1;

Exit when (a > 10);

End loop;

Dbms_output.put_line(‘End of numbers’);

End;

/

2. While Loop:

It is an pre-tested loop.

Syntax: While (cond)

Loop

<exec stmts>;

End loop;

Ex: Declare

A number(3):=1;

Begin


While (a<=10) loop

Dbms_output.put_line(‘a’);


Page 8 of 54

A := a+1;

End loop;

Dbms_output.put_line(‘End of numbers’);

End;

/

3. Numeric For Loop:

It increments the variable by 1 always.

Variable can’t be assigned with a value in for loop.

Syntax: For <var> in [reverse] <value1>..<value2>

Loop

<Exec stmts>;

End loop;

<var> –Automatically defined by for loop.

Reverse(optional) –Accepts values in reverse order

.. –Range operator

Ex: Begin


For n in 1..10

Loop

Dbms_output.put_line(n);

End loop;

Dbms_output.put_line(‘end of numbers);

End;

Dbms_output.put : prints the results in same line.

Dms_output.put_line : prints the result in new line.

Put_line must be followed with “put” stmt to activate output

buffer.

4. Cursor For loop:

Used to perform cursor operations automatically

Improves performances

Syntax:

For <var> in <cur name>

Loop

<exec stmts>;

End loop;

Advantages:

No need to open cursor

No need to fetch rows

No need to check for end of rows

No need to close cursor

No need to declare variables

Ex: Declare

Cursor c2 is select * from dept;

Begin

For I in c2


Page 9 of 54

Loop

Dbms_output.put_line(c2%rowcount||’’||i.deptno||’’||i.dname||’’||i.loc);

End loop; --close cursor

End;

For I in c2 loop

Declares I variable

Open cursor

Fetch 1 row into i

Check end of rows

Declare

Cursor ecur

Is

Select empno, ename, sal, job, emp.deptno, dname, loc

From

Emp, dept

Where emp.deptno = dept.deptno;


Page 10 of 54

CURSORS:

It is a temporary buffer used to hold the transaction data for manipulation purpose.

It is not stored in database.

It is not re-usable.

It is valid in Plsql block only.

It is created in logical memory only.

2 types

A. implicit cursors:

Automatically defined by oracle whenever “DML” operations are performed by user.

It has a fixed name “SQL”

It gives the status of DML statements in Plsql block.

It will not support cursor operations.

It supports %found, %nofound, %rowcount, attributes.

Ex: TO update only 5 employees

Declare

Vdno number(2) := &dno;

Begin

Update emp set sal = sal+1000 where deptno = vdno;

If sql%notfound then

Dbms_output.put_line(‘Unable to update rows –No such dept exists’);

Elsif sql%rowcount > 5 then

Dbms_output.put_line(‘ Invalid operation –Can’t update more than 5 employees’);

Rollback;

Else

Dbms_output.put_line(sql%rowcount||’ employees are updated successfully’);

Commit;

End if;

End;

/

B. Explicit cursors:

Created by user in Plsql block used to retrieve multiple rows into Plsql block for

manipulation purpose.

1. Declaring cursor:

Syntax :

cursor <cur name> is <select stmt>;

2. Cursor Operations:

i. Open cursor:

Open <cur name>;

Used to open the cursor.

Memory will be allocated to cursor after opening it.


Page 11 of 54

ii.Fetch Cursor:

Fetch <cur name> into <var’s>;

Used to retrieve data from cursor to Plsql variables.

At a time it can retrieve only one row into variables

Generally fetch will be placed in loop

iii.Close Cursor:

close <cursor name>;

Used to close the cursor.

Memory allocated will be de-allocated.

3. Cursor Attributes:

Gives the status of cursor

<cur name>%<attribute>

i. %isopen (Returns true/false)

Returns True if cursor is opened successfully

ii. %found (Returns true/false)

Returns True if fetch stmt successfully retrieves the into Plsql var’s.

iii. %notfound (Returns true/false)

Returns true if fetch stmt fails to retrieve the row into Plsql var’s.

iv. %rowcount (Returns true/false)

Returns the no of rows successfully retrieved from cursor so far. Initially it holds

0.

After every successful fetch it is incr by 1.

Ex-1: Declaring cursor var with %Type

Declare

Cursor c1

Is

Select empno, ename, comm. From emp

Order by deptno;

Veno emp.empno%type;


Vcomm emp.sal%type;

Begin

Open c1;

If c%isopen then

Dbms_output.put_line(‘EMP NEW COMM’);

LOOP

Fetch c1 into veno, vname, vcomm;

Exit when c1%notfound;

If vcomm is null then

Vcomm := 3000;

Elsif vcomm = 0 then

Vcomm := 2500;

Else vcomm := vcomm+vcomm*.25;

Endif;

Update emp set comm.= vcomm where empno = veno;


Page 12 of 54

Dbms_output.put_line(veno||vname||vcomm);

End loop;

Dbms_output.put_line(c1%rowcount||’Employees are updated with new comm’);

Close c1;

Commit;

Else dbms_output.put_line(‘Unable to open cursor’);

End if;

End;

/

Declaring cursor var with rowtype.

Ex-2: Declare

Cursor c2

Is select * from dept;

I c2%rowtype;

Begin

Open c2;

Loop

Fetch c2 into I;

Exit when c2%notfound;

Dbms_output.put_line(c2%rowcount||’’||i.deptno||’’||i.dname||’’||i.loc);

End loop;

Close c2;

End;

Cursor with Scalar query:

Ex1: Declare

Cursor c2

is

select eno, ename, sal ,job,

(select min(sal) from emp where job = e.job) lopay,

(select max(sal) from emp where job = e.job) hipay,

from emp E order by job;

Begin

For k in c2

Loop

Dbms_output.put_line(k.eno||’’||k.job||’’||k.ename||’’||k.sal||’’||k.lopay||k.hipay);

End loop;

End;

/

Ex2: Calculate bonus for first 10 lowest employees.

Declare

Cursor c1

Is

Select eno, ename, sal, sal+nvl(comm., 0) net, job from emp

Order by sal;

I c1%rowtype;


Page 13 of 54

Bonus emp.sal%type;

Begin

For I in c1

Loop

If i.job = ‘CLERK’

Then bonus:= round(i.net*.25);

Elsif i.job = ‘EXECUTIVE’

Then bonus := round(i.net*.50);

Else then

Bonus := round(i.net*.1);

End if;

Dbms_output.put_line( i.eno||’’||i.ename||’’||i.job||’’||bonus);

Exit when(c1%rowcount >= 10);

End loop;

End;

/

Cursor with Parameters(8.0):

Used to accept the input dynamically while opening the cursor.

Supports to define generalized cursor.

Maximum 32 parameters can be passed to cursor.

Ex-1: Calculates incr for employees depending on experience for particular dept only.

Declare

Cursor c1(dno number)

Is

Select * from emp

Where deptno = dno;

I c1%rowtype;

Incr number(12);

Expr number(2);

Begin

Open c1(&dno);

Loop

Fetch c1 into i;

Exit when c%notfound;

Net := i.sal + nvl(comm., 0);

Expr := round(months_between(sysdate, i.hiredate))/12);

If expr <=2

Then incr := net*.25;

Elsif expr<=5

Then incr := net*.5;

Else incr := net*.1;

End if;

Update emp set sal := sal+incr where eno = i.eno;


Page 14 of 54

Dbms_output.put_line(i.eno||’’||e.job||’’||expr||incr||i.deptno);

End loop;

Close c1;

End;

/

EX-2 :

Cursor c2(dno number, vjob varchar2)

Is select * from emp

Where deptno = dno and job = vjob;

Open c2(30, ‘clerk’);

For I in c2(20, ‘salesman’) loop

Ex-3:

Cursor c3(vcourse varchar2, vtime varchar2)

Is select * from student

where course = vcourse and timing = vtime;

open c3(‘oracle 9i’, ’10AM’);

for I in c3(‘oracle 9i’, ‘6PM’);

EX-4:

Cursor c4

Is select * from student

Where course = &course’;

It is valid in oracle sql environment.

Cursor parameters are used to pass values from other

application s/w tools.


Page 15 of 54

EXCEPTIONS:

Errors in Plsql block are termed as exceptions.

Error handling solutions are provided in exception block.

3 types…

1. Pre defined exceptions

Defined by oracle

Activated by oracle automatically.

Solution provided by user.

No_data_found:

Automatically activated whenever select stmt fails to retrieve data into

variables.

Too_many_rows:

Automatically activated whenever select stmt retrieves more than one row into

variables.

Value_Error:

Automatically activated whenever data types are not matching.

Zero_divide:

Automatically activated whenever invalid arithmetic operations is performed

(no/zero).

Dup_val_on_index:

Automatically activated whenever duplicate values are inserted into unique

index column.

Storage_error:

Automatically activated whenever there is lack of memory in server.

Invalid_cursor:

Automatically activated whenever user fetch the rows from cursor without

opening the cursor.

Time_out_on_resoursces:

Automatically activated whenever user performs an infinite loo process.

Transaction_backed_out:

Automatically activated whenever there is an communication problem with

server.

Login_denied:

Automatically activated whenever user name or pwd are not valid.

Cursor_already_open:

Automatically activated whenever user opends the cursor which is already in

open state.

Others:

General exception

Used to handle any pre-defined exception.

It will not support non-predefined/user defined exceptions.

Pseudo columns activated by other exception.

It must be last exception in exception block.

SQLCODE: Holds the currently raised error no.

SQLERRM: Holds the currently raised error message.


Page 16 of 54

2. User defined exceptions

Defined by user.

Activated by using “RAISE” stmt.


Ex: Plsql block handles diff types of errors.

Declare

Veno emp.eno%type := &employ;


Vsal emp.sal%type;

Salary_missing exception; --Declaring user defined excep

Begin

Select ename, sal into vname, vsal from emp

Where eno = veno;

If vsal is null then

Raise salary_missing; --Activating user defined excep

Else

vSal := vsal + vsal * .5

Update emp set sal = vsal where eno = veno;

Dbms(‘salary updated for’||veno||’is’||vsal);

Commit;

End if;

When no_data_found then

Dbms_output.put_line(‘No such emp found’||veno);

when too_many_rows then

dbms_output.put_line(‘More than one emp found for’||veno);

when salary_missing then

dbms_output.put_line(‘Emp has null salary –providing default pay’);

update emp set sal=5000 where eno= veno;

commit;

when others then

dbms_output.put_line(‘Error occurred’);

dbms_output.put_line(sqlcode||’’||sqlerrm);

end;

/

3. Non pre-defined exceptions (undefined exceptions)

Defined by user.

Activated automatically by constraints.


Used to handle constraint violation of errors in Plsql.

Ex: Handling foreign key constraint error.

Declare

Vdno number(2) := &dno;

Employ_exists exception;


Page 17 of 54

Pragma exception_init(employ_exists, -2292);

Begin

Delete from dept where dno = vdno;

Commit;

Exception

When employ_exists then

Dbms_output.put_line(‘ Emp are still working –can’t remove dept’);

End;

/

EXCEPTION_INIT(Exception_name, Error_no);

Built-in sub-pgm used to provide exception name to the

constraint violation error

EXCEPTION PROPAGATION:

Declare

-------

-------

Begin

-------

-------

Begin

------- ERR

-------

-------

Exception

-------

-------

-------

Solution End; No Solution

-------

-------

Exception

------- Sol No sol

------- Quit with error

------- Quit with success

End;

/


Page 18 of 54

PRAGMA:

It is an special instruction to Plsql to execute sub program in declaration section.

DEPT EMP

-- --

-- --

Deptno(PK) eno

Deptno(FK)

Constraint violation error no’s:

-00001 Unique

-1400 Not null

-2290 Check

-2291 Parent not exists(ref)

-2292 Child records exists(ref)

EX-1: create table std_inf(roll number(4), name varchar2(90), course varchar2(20) check( course in

(‘oracle 9i’, ‘unix’, ‘D2k’)), fee number);

Insert into std_inf values( 101, ‘Sai’, ‘java’, 2300);

PLSQL block handling check constraint error

Declare

Invalid_course exception;

Pragma exception_init(invalid_course, -2290);

Begin

Insert into std_inf values(101, ‘Sai’, ‘java’, 2300);

Commit;

Exception

When invalid_course then

Dbms_output.put_line(‘ student course must be oracle 9i or unix or d2k’);

End;

/

EX-2: Using nested blocks, cursors with lock and exceptions.

Create table itemmast(itno number(2), name varchar2(90), qoh number(4), rol number(5));

Insert into itemmast values(&1, ‘&2’, &3, &4);

Create table ittran(itno number(4), tran_type char(1) [I-Issue(-), R-Receipts(+)], qty

number(5), updt char(1));

Insert inot ittran values(&1, ‘&2’, &3, ‘&4’);

Declare

Cursor tran is select into, tran_type, qty from ittran where upper(updt) = ‘N’

for update;

--cursor with lock

Cursor item is select * from itemmast;

Vname varchar2(20);

Vqoh number(4);


Page 19 of 54

Begin

For I in tran loop

Begin --nested loop

Select name, qoh into vname, vqoh from itemmast

Where itno = i.itno;

Exception


Dbms_output.put_line(‘no such item exists with itno’||itno);

End;

If i.tran_type = ‘R’ then

Vqoh := vqoh + i.qty;

Elsif i.tran_type = ‘I’ then

Vqoh := vqoh – i.qty;

End if;

Update itemmast set qoh = vqoh where itno = i.itno;

Update ittran set updt= ‘Y’ where CURRENT OF TRAN;

End loop;

Commit;

Dbms_output.put_line(‘Stock Report’);

For k in item loop

Dbms_output.put_line(k.itno||’’||k.name||’’||k.qty);

End loop;

Exception

When others than

Dbms_output.put_line(sqlcode||’’||sqlerrm);

End;

FOR UPDATE:

Locks the rows retrieved into cursor

WHERE CURRENT OF: Clause (8.0)

Used to locate the current row manipulated by

cursor

To use this cursor must be locked


Page 20 of 54

Composite data types: [User defined data types]

Defined by the user.

Valid in Plsql block only.

They are not stored in the database permanently.

They will not hold data.

They are not re-usable.

This will improve performance of oracle while retrieving or manipulating huge loads of data

into Plsql blocks.

2 types…

1. PLSQL Records:

It is an collection of elements of “different data types” stored at one location.

It is similar to “C Prog” structures.

Syntax:

Type <record name> is record

(element1 <datatype>,

Element2 <datatype>,

…..

Elementn <datatype>);

EX:

Declare --declaring Plsql record.

Type erec is record(eid number(4), name emp.ename%type, basic emp.sal%type, da

emp.sal%type, hra emp.sal%type, e.pf emp.sal%type, gross emp.sal%type);

E erec; --record type variable

Begin

e.eid:= &employ;

select ename, sal into e.name, e.basic from emp

where empno = e.eid;

e.da := e.basic*.24;

e.hra:= e.basic*.15;

e.pf:= e.basic*.15;

e.gross := e.basic+e.da+e.hra-e.pf;

dbms_output.put_line(e.eid||’’||e.name||’’||e.basic||’’||e.pf||’||e.gross);

end;

2. PLSQL Tables:

It is an collection of elements of same datatype stored in continous memory locations.

It is similar to “C Prog” arrays.

It is supported with an index automatically.

Syntax:

Type <table_name> is table of

<datatype> index by binary_integer;


Page 21 of 54

EX:

Declare --declaring Plsql tables.

TYPE names IS TABLE OF emp.ename%type

INDEX BY BINARY_INTEGER;

TYPE pays IS TABLE OF emp.sal%type

INDEX BY BINARY_INTEGER;

N names;

P pays;

Totsal number(12) := 0;

Ctr number(9):=1;

Begin --filling tables

For I in (select ename, sal from emp) loop

N(ctr):= i.ename;

P(ctr):=i.sal;

Ctr:= ctr+1;

End loop;

--printing table contents

For k in 1..n.count loop

Dbms_output.put_line(n(k)||’’||p(k));

Totsal:= totsal + p(k);

End loop;

Dbms_output.put_line(‘Total Salary is :’||totsal);

End;

N.COUNT : Gives no of elements in Plsql table


Page 22 of 54

Nested records and Plsql tables using records:

Declare

Type pf_rec is record(pfno number(9), amt number(12,2));

--Nested record

Type erec is record(eid number(9), name emp.ename%type, basic emp.sal%type, job

emp.job%types, pf pf_rec);

--plsql table using nested records.

Type etab is table of erec

Index by binary_integer;

Ctr number(3):=1;

E etab;

Begin

For I in (select empno, ename, sal, job from emp) pf.rec

Loop

E(ctr).eid := i.empno; erec

E(ctr).name := i.ename;

E(ctr).job := i.job; etab

E(ctr).pf.pfno := i.empno+1000;

E(ctr).pf.amt := round(i.sal * .15);

Ctr := ctr + 1;

End loop;


Dbms_output.put_line(‘employee pay details are:’);

For k in 1..e.count

Loop

Dbms_output.put_line(e(k).eid||’’|| e(k).name||’’|| e(k).basic||’’|| e(k).job||’’||

e(k).pf.pfno||’’|| e(k).pf.amt);

End loop;

End;

/


Page 23 of 54

DATABASE TRIGGERS:

A set of Plsql stmts stored permanently in database and automatically activated whenever

an event raising stmt (DML) is preformed.

They are stored in user_triggers system table.

They are use to impose business rules/user defined restrictions on table.

They are also activated when tables are manipulated by either users or by other

application s/w tools.

They provide high security on tables.

TCL commands are not allowed in triggers.

Trigger parts:

1. Triggering even:

Indicates when to activate the trigger.

Before – insert/update/delete

After – insert/update/delete

2. Trigger types:

a. Row trigger:

Activates the trigger for every row manipulated by DML stmt.

b. Stmt trigger:

Activates the trigger for only once for 1 DML stmt (Default type).

3. Trigger restriction:

Used to stop the activation of trigger based on condition. If condition is true trigger is active.

4. Trigger Body:

A set of Plsql stmts.

Syntax:

Create or replace TRIGGER <trigger_name>

Before/after insert or update or delete (1)

[of <columns>] on <table_name>

[for each row (2)

When (<condition>) (3)

Declare

<var declaration>; ]

Begin

<exec stmts>;

[Exception

<exec stmts>;]

End;

If condition is

true trigger is

activated


Page 24 of 54

EX: Trigger converts student name to upper case automatically.

Create ore replace trigger up_con before insert on stu_info

For each row

Begin

:new.sname := upper(:new.sname);

End;

Triggering Events:

12 Events per table are allowed.

Before insert row before update row before delete row

Before insert stmt before update stmt before delete stmt

After insert row After update row After delete row

After insert stmt After update stmt After delete stmt

EX: Trigger checks fro valid increment.

Create or replace trigger valid_incr

Before update of sal on emp

For each row

Begin

If :new.sal <= :old.sal then

Raise_application_error(-20300, ‘Incr must be more than existing salary’;

End if;

End;

/

NEW & OLD (pseudo columns):

Used to retrieve data from DML stmt

tempory buffer.

Valid with “ROW” triggers only.

INSERT UPDATE DELETE

NEW valid valid invalid

OLD invalid valid valid

Column specification is supported with

update event only. Not valid with insert

and delete. operations(Row level

operations)

Raise_application_error(error no, error message);

Built in subprogram stops DML stmt execution and

displays the error message.

Error_NO range: -20002 to -20990

Oracle Reserved error no’s: -20000 to +20000


Page 25 of 54

Constraints vs Triggers:

Constraint provide standard error message.

Triggers provide user friendly error message.

Constraints will for existing errors.

Triggers will not check for existing errors.

If constraint & trigger are defined on table, one time only trigger will be activated.

Ex: Trigger checks for valid salary according to the job.

Create table job_pays(job varchar2(20), losal number(10), hisal number(10));

Create or replace trigger chk_sal

Before insert on emp for each row

When(new.job<>’PRESIDENT’) --If true trigger is activated

Declare

Minsal number(10);

Maxsal number(10);

Begin

Select losal, hisal into minsal, maxsal from job_pays

Where job =:new.job;

If :new.sal>maxsal and :new.sal<minsal then

Raise_application_error(-20201, ‘salary must be between’||maxsal||’and’||minsal||’for

job’||:new.job);

End if;

Exception


Raise_application_error(-20202, ‘No data found for dept’||:new.job);

End;

/

EX: Trigger adds the fee installment amt automatically into fee installments table

Create fee_installments(roll number(9), prev_fee number(9), curr_fee number(9), dop

date);

Create or replace trigger add_fee

Before update or insert on student

For each row

Begin

If inserting then

Insert into fee_installments values(:new.roll, :new.fee, :new.fee, sysdate);

Elsif updating then

Insert into fee_installments values(:old.rol, :old.fee, :new.fee - :old.fee, sysdate);

End if;

End;

/

Keywords: Inserting, updating, deleting

Represents insert, update, delete operations.

Valid in triggers only


Page 26 of 54

Statement level triggers:

Trigger locks the table for DML operations to provide security.

Trigger checks for

o Valid timings

o Weekends

o Public holidays

Ex:

Create or replace trigger security_chk

Before insert or update or delete on bankmaster

Declare

A number;

Begin

--check for valid timings.

If to_char(sysdate, ‘hh24’) not in (10,11,12,13,14,15,16,17,18) then

Raise_application_error(-20111, ‘No operations allowed –invalid timings’);

End if;

--check for week ends.

If to_char(sysdate, ‘dy’) in (‘sat’,’sun’) then

Raise_applicaiton_error(-20112, ‘No operations allowed –week ends’);

End if;

--check for public holidays

Select count(*) into a from holiday where to_date(hdate) = to_date(sysdate);

If a>0 then

Raise_application_error(-20113, ‘No operations allowed –public holidays’);

End if;

End;

/

Instead of triggers:

Triggers supported only on “views”

Used to perform DML operations on “join views”

Instead of ----insert, update, delete

EX:

Create view edept

As

Select empno, ename, job, sal, emp.dtno, dept.dtno dno, dname, loc from emp, dept

Where emp.dno = dept.dno;

Create trigger T1 instead of insert on edept

For each row

Begin

Insert into dept values(:new.dno, :new.dname, :new.loc);

Insert into emp(empno, ename, sal, job, dtno) values(:new.empno, :new.ename, :new.sal, :new.job,

:new.dtno);

End;


Page 27 of 54

Disabling triggers:

Alter table <table_name> disable all triggers;

Enabling triggers:

Alter table <table_name> enable all triggers;

Removing triggers:

Drop trigger <trigger_name>;

Checking for existing triggers:

Desc user_triggers;

Select trigger_name, trigger_event from user_triggers

Where table_name = ‘emp’;

Select * from user_triggers

Where trigger_name = ‘chk_sal’

Create trigger T2

Before insert on temp for each row

Begin

Update temp set c2 = ‘HARI’;

End;

--It will not perform update in trigger body because 2 DML operations can’t

be performed on table at once

--It leads to a trigger mutation error while working in other s/w tools.


Page 28 of 54

SUBPROGRAMS:

A set of Plsql stmts stored permanently database and used to perform a task.

They can accept input from user as arguments dynamically.

They are stored in “user_source” system table.

They are reusable components.

They can be shared with other users.

They can be used n other application s/2 tools.

They are faster in execution stored n compiled formed.

They support modularity.

2 types…

1. Procedures

2. Functions

1. Procedures:

A sub prg type performs a task and will return a value.

Generally procedures are used to perform DML operations on database.

They can’t be called in SELECT stmt.

Procedure can be used in sql, other application s/w tools, Plsql triggers

procedures, functions, packages, D6i.

Syntax:

Create or replace procedure <p_name> [(arguments MODE <datatype>)]

Is

[<var_declaration>]

Begin

<exec stmts>;

End;

Calling a procedure: procedure_name(arguments);

EX:

Create or replace procedure Sim_ins(p number, n number, r number)

Is

Si number(12);

Begin

Si := (p*n*r)/100;

Dbms_output.put_line(‘simple interest:’||si);

End;

/

SQL: PLSQL:

Desc sim_ins begin

Exec sim_ins(2,3,4); sim_ins(123,12,8.5);

End;

A huge task divided into

N no of sub tasks. Easy to

manage the logic easy to

debug errors.


Page 29 of 54

2. Functions:

A sub prog type performs a task and will return the value (Return only one value).

Generally functions are used for reporting purpose or for calculating purpose.

They can be called in SELECT stmt.

Function can be used in sql, select, other application tools, Plsql, triggers, procedures,

functions, packages.

Syntax:

Create or replace function <function_name> [(arg’s MODE <datatype>)]

Return <datatype>

Is

[<var declaration>];

Begin

<exec stmts>;

Return (variable);

[Exception

<exec stmts>;]

End;

Calling a function: var := fun_name(arg’s);

EX: Create or replace function si(p number, n number, r number)

Return number

Is

Si number(9);

Begin

Si := (p*n*r)/100;

Return(round(si));

End;

SQL: PLSQL:

i. Desc si Declare

Select si(1200,8,7) from dual; result number(12);

begin

ii. Variable result number result:=si(1200,9,7);

exec:result:=si(5400, 10, 9.5); dbms_output.put_line(result);

print result; end;

Create table loan_master(cust_id number(4), cname varchare(9), amount number(12),

duration number(3), irate number(5,2));

Insert into loan_master values(&1, ‘&2’, &3, &4, &5);

Select cust_id, upper(cname), si(amount, duration, irate) interest from loan_master;


Page 30 of 54

EX-1: Function checks for leap year

Create or replace function chk_year(y number)

Return varchar2

Is

Begin

If mod(y,4) = 0 then

Return(‘leap year’);

Else

Return (‘ Not leap year’);

End if;

End;

Select chk_year(2008), chk_year(2010) from dual;

Variable result varchar2(90);

Exec:result:=chk_year(2012);

Print result;

SET AUTOPRINT ON Automatically prints the variable,

No need to use print stmt.

EX-2: Procedure cal the new comm. For all emp of particular dept.

Create or replace procedure call_comm(vdept number)

Is

Cursor c1

Is

Select empno, ename, comm. From emp

Where deptno = vdept;

I c1%rowtype;

Begin

For I in c1

Loop

If i.comm is null

Then

i.comm :=3000;

elseif i.comm = 0

then

i.comm:=2500;

else

i.comm = i.comm+i.comm*.25;

end if;

update emp set comm. = i.comm where empno = i.empno;

dbms_output.put_line(i.empno||’has been updated comm. For’||i.comm);

end loop;

commit;

end;


Page 31 of 54

Desc user_source

Select name from user_source;

To Remove sub prog’s:

Drop procedure <proc_name>;

Drop function <fun_name>;

Create sequence s1 increment by 1 start with 11;

EX-3: Procedure adds a dept into Dept table

Create or replace procedure add_dept

(vname varchar2 default ‘unknown’, vloc varchar2 default ‘hyderabad’)

Is

Begin

Insert into dept values(s1.nextval, vname, loc);

Commit;

End;

Desc add_dept

Exec add_dept;

Exec add_dept(‘sales’, ‘mumbai’);

Exec add_dept(‘export’);

Exec add_dept(vloc => ‘chennai’);

Exec add_dept(vloc=> ‘pune’, vname => ‘software’);

Sharing Sub-Programs:

Scott:

Grant execute on add_dept to user1;

User1:

Desc scott.add_dept

Exec scott.add_dept(‘Testing’);

Insert into scott.dept values(22, ‘HR’, ‘sec’); --Error

Parameter modes in sub-programs:

“MODE” indicates the behaviour of argument in sub program

3 types….

1. IN: (default mode)

IN parameter is used to carry the input into subprogram.

It can’t be assigned with a value in subprogram.

User can manipulate the

table only through procedure

but not directly.

Named notation:

Arguments are passed

based on their name.


Page 32 of 54

2. OUT:

OUT parameter is used to return the results from sub program.

It can’t be assigned with a value in sub program

3. IN OUT:

INOUT parameter is used to carry the input and return the results from subprogram.

It can be assigned with a value in subprogram.

EX-1:

Create or replace procedure P1(a in out number)

Is

Begin

A:= A*A*A;

End;

SQL:

Desc p1

Var x2 number;

Exec :x2 := 5; --intialising variable at sql

Exec p3(:x2);

Print x2

EX-2: Calculate factorial for given number and checks for odd/even number.

Create or replace function call_fact( n in number, f out number);

Return Boolean

Is

Fac number(9) :=1;

Begin

-- cal factorial 5! = 5*4*3*2*1 =120

For I in 1..n

Loop

Fac := fac * I;

End loop;

F:= fac; -- storing result in out parameter

If mod(n,2) = 0

Then

Return(TRUE);

Else

Return(false);

End if;

End;

Using “CALL_FACT” function:

Declare

A number(9) := &num;

Fact number(90);

Begin

Function returning

“Boolean” value (or) with

OUT parameter can’t be

used in sql or select stmt.


Page 33 of 54

If call_fact(a, fact)

Then

Dbms_output.put_line(‘Given number is even’);

Else

Dbms_output.put_line(‘Given number is odd’);

End if;

Dbms_output.put_line(‘Factorial of given number is:’||fact);

End;

Triggers vs Programs:

Triggers are automatically activated by DML statements.

Procedures & functions has to be explicitly activated by user.

Advantages:

Code reusability.

Faster execution of code.

Easy to manage code.

Executed in the oracle server memory even if it is activated from other application software

tools.


Page 34 of 54

PACKAGES:

It is a collection of related var’s, cursors, procedures and functions.

It is stored in “user_source” system table.

It can be shared with other users.

They will improve performance of oracle while accessing sub programs from client

environment.

They support oops features live encapsulation, data hiding, and function over loading.

2 parts…

i. package specification:

It holds the declaration of variables, cursors, procedures and functions.

Syntax:

Create or replace package Package_name

As

<Var dec>;

Procedure <proc_name>(arg’s datatype);

Functions <fun_name>(arg’s datatype) return datatype;

End <pack_name>;

ii. Package body:

it holds the code of sub program.

Syntax:

Create or replace package body Package_name

As

Procedure <proc_name>(arg’s datatype)

Is

<exec stmts>;

End <proc_name>;

Functions <fun_name>(arg’s datatype) return datatype

Is

<exe stmts>;

End <fun_name>;

End <pack_name>;

Ex-1:

Create or replace package pack1

as

result number(12)

procedure add_num( a number, b number);

function mul_num(x number, y number);

end pack1;

create or replace package body pack1

as

procedure add_num(a number, b number)

is

begin

result := a+b;


Page 35 of 54

dbms_output.put_line(‘sum of numbers is:’||result);

end add_num;

function mul_num(x number, y number) return number

is

begin

result := x*y;

return (round(result));

end mul_num;

end pack1;

/

Using package elements:

Exec pack1.add_num(2099, 8908);

Select pack1.mul_num(200,22) from dual;

Select ename, job, pack1.mul_num(sal,.15) pf,

pack1.mul_num(sal, 2) bonus from emp;

EX-2: LOCAL sub program are valid in package body only.

Create or replace package epack

As

Function gross(vsal number) return number;

Function experience(doj date) return number;

End epack;

Create or replace package epack

As

Function pf(vbasic number) return number is

Begin

Return( round(vbasic*.15));

End if;

Function gross(vsal number) return number is

Epf number(12);

Begin

Epf := pf(vsal);

Return (round(vsal+vsal * .25 + vsal*.35 – epf));

End gross;

Function experience (doj date) return number is

Begin

Return (round(months_between(sysdate, doj))/12);

End experience;

End epack;

/

Select empno, ename, job, round(epack.experience(hiredate) experience,

epack.gross(sal) “Gross pay” from emp;

Select ename, epack.pf(sal) “PF” from emp;

Sub-program not

defined in

specification are local

sub programs. We

can’t call this pgm’s

from outside of

package.


Page 36 of 54

--ERROR, you can’t call local subprogram.

Package with function overloading feature:

Create or replace package load_pack

As

Function add_data(a number, b number) return number;

Function add_data(a varchar2, b varchar2) return varchar2;

End load pack;

/

Create or replace package load_pack body

As

Function add_data(a number, b number) return number

Is

Begin

Return (a+b);

End add_data;

Function add_data(a varchar2, b varchar2) return number

Is

Begin

Return(a||’ ‘||b);

End add_data;

End load_pack;

/

Desc load_pack;

Select load_pack.add_data(5,6) total,

load_pack.add_data(‘Satya’, ‘M’) name from dual;

Removing Pack:

Drop package <pack_name>;

Sharing Pack:

Grant execute on load_pack to user1;

To view existing package body:

Select text from user_source wehre name = ‘pack1’;


Page 37 of 54

8.0 Features:

Returning into clause:

o Used to return the values through “DML” stms.

o Used with update and delete stmts.

Ex:

Var a varchar2(20)

Var b number

Update emp set sal = sal+3000 where empno = 7900

Returning ename, sal, into :a, :b;

Print a b

Delete from emp where mepno= 7920

Returning ename, sal into :a, :b;

Print a b

Bulk Collect:

o Used to return bulk data into Plsql variables

o Variables must be of Plsql table type only

o Improves performance while retrieving data.

o Used with select, update, delete, fetch stmts.

Examples:

Using in select stmt:

Declare

Type names table of emp.ename%type


Type pays table of emp.sal%type

Index by bindary_integer;

N names;

P pays;

Begin

--Retrieving all emp’s in 1 transaction

Select ename, sal bulk collect into N, P from emp;

--Printing table contents

Dbms_output.put_line(‘Employ details are’);

For I in 1..n loop

Dbms_output.put_line(n(i)||’’||p(i));

End loop;

End;

Using in update stmt:

Update emp set sal=sal+3000 where deptno = 30

Returning ename, sal bulk colleck into n,p;

Using in delete stmt:

Delete form emp where job = ‘CLERK’

Returning ename, sal bulk collect into n,p;


Page 38 of 54

Using in fetch stmt:

Declare

Type names is table of emp.ename%type


Type names is table of emp.sal%type


N names;

P pays;

Cursor c1

is

Select ename, sal into n, p from emp;

Begin

Open c1;

Fetch c1 bulk collect into n,p;


For I in 1..n count

Loop

Dbms_output.put_line(n(i)||’’||p(i));

End loop;

End;

/

Dynamic SQL:

o Supports to execute DDL commands in Plsql block.

Execute immediate(‘DDL stmt’);

EX:

Begin

Execute immediate(‘create table employ(ecode number, ename varchar2(90), sal number));

End;

Begin

Execute immediate(‘drop table employ1’);

End;

Using Long and Raw data types:

Long:

o Used to represent numbers or character values

o Maximum limit is 2GB.

o But only once it can be used in table.

Raw(n):

o Used to represent images

o Max limit is 256 bytes(7.0)/2000 bytes (8.0)

Long Raw:

o Used to represent images upto 2 GB.

o Raw and long raw can be manipulated by GUI tools only

NOTE:

Table can’t be manipulated

in same Plsql block.


Page 39 of 54

Ex:

Create table emp(ecode number(3), ename varchar2(20), sal number(12,2), description long,

deptno number(2));

Create table ephoto(ecode number(4), photo long raw);

Create table customer(cust_id number(4), cname varchar2(20), addr varchar2(50), photo

longraw);

File input & output (8.0):

Used to transfer oracle table contents to OS file and OS file content to oracle table.

Standard package holds all functions (arthematic, date, general);

UTL_FILE Built in package supports file operations.

Sub-Programs in UTL_FILE pkg:

Fopen:

Used to open the file in the specified mode and returns the address of the file.

Mode R Read, W Write, A Append

Fclose: Used to close the file

Get_line: Used to read a line from file to Plsql variable.

Put: Used to write a line into file but in New line.

Put_line: Used to write a line into file but in new line.

Fflush: Writes the file content from temporary buffer to physical memory (saving file

into OS)

File_type: Data type

Used to define the variable to hold the address of file.

Invalid_file_handle, Read_error: Pre-defined exception automatically activated while working with files.

Plsql block transfers data from oracle table to OS file.

Ex: Plsql procedure transfer data from OS file to database table.

Create or replace procedure load_student(pdir varchar2, pfile varchar2)

Is

Vfile utl_file.file_type;

Vtext varchar2(200);

Vname varchar2(20);

Vcourse varchar2(20);

Vfee number(5);

Fcomma number(3);

Scomma number(4);

Begin

Vfile:= utl_file.fopen(pdir, pfile, ‘R’);


Page 40 of 54

Loop

Begin

Utl_file.get_line(vfile, vtext);

Exception


Exit;

End;

Fcomma := instr(vtext, ‘,’,1,1);

Scomma := instr(vtext, ‘,’, 1,2);

Vname := substr(vtext,1,fcomma,-1);

Vcourse:=substr(vtext, fcomma+1, soma-fcomma);

Vfee := substr(vtext, scomma+1);

Insert into student values(s1.nextval, vname, vcourse, vfee);

End loop;

Commit;

Exception

When utl_fjile.read_error then

Dbms_output.put_line(‘unable to read the file….’);

End;

Exec load_student(‘c:\’, ‘student.txt’):

Note: Provide the below parameter in “init.ora” file

(open in notepad)

UTL_FILE_DIR = ‘C:\Oracle\ora92\Bin’

Init.ora --system parameter file.

Holds the parameter defined for oracle server.


Page 41 of 54

Object Technology(8.0):

Object is a collection of different datatypes stored at one location.

It will not hold data

It can have methods (functions) defined in it.

They are stored permanently in “user_types” system table

They support inheritance feature.

They can be shared with other users.

They are reusable components.

CDT: (composite data type)

Object ‘C’ structure

Object with methods ‘C++’ class

Nested tables

Varying arrays

EX: Plsql records

Plsql tables

Syntax:

Create type <obj name> as object

(element1 <datatype>,

ele 2 <datatype>,

------

Element N <datatype>);

Object types:

2 types…

1. Column object:

Only few columns of table represent objects.

2. Row Object:

Entire table structure depends on object.

Create type etype as object( eid number(9), ename varchar2(90), sal number, deptno

number(9));

Create table etab of etype; Row object

Desc etab = desc etype

Insert into etab values(1001, ‘RAM’, 210,10);

Select * rom etab;

Select eid, ename, sal from etab;

Create table employ(einfo etype, job varchar2(90), hiredate date); Column object

Ex:

Create type addr_type as object (hno varchar2(90), street varchar2(90), city varchar2(90), pin

number(6));

Create type Pf_type as object (pfno number(5), amt number(4));

Desc addr_type;

Desc pf_type;

Select * form user_types;


Page 42 of 54

Naming object:

Create table student(roll number(3), sname varchar2(20), address addr_type, course

varchar2(90), fee number(5));

Desc student;

Insert into student values(101, ‘SAI’, addr_type(’12-A’, ‘ameerpet’, ‘Hyd’, 500038), ‘oracle

9i’, 2500);

Select * from student;

Select sname, address from student;

Select sname, s.address.city, s.address.pin from student s;

Update student s set s.address.hno = ‘308A’ where roll=101;

Delete from student s where s.address.city = ‘hyd’;

Create trigger city_chk before insert on emp

for each row

begin

if :new.address.hno is null then

Raise_aplication_error(-20100, ‘Student hno can’t be null’);

End if;

End;

Sharing object:

Scott: Grant execute on addr_type to user1;

User1:

Create table supplier(sid number, sname varchar2(90), saddr scott.addr_type);

Select sname s.saddr.city from supplier s

Nested objects: [Max 32 levels]

Create type person_type as object(pname vachar2(20), paddr addr_type);

Desc person_type;

Create table customer(cid number, cinfo person_type);

Select cid, c.cinfo.paddr.city, c.cinfo.paddr.pin from customer c;

Desc customer;

Drop type <obj_name>;

Drop type addr_type;

Drop type pf_type;

Purity level test:

Checks for member functions are having DML stmts or not.

DML are not valid in member functions.

Objects will not support constraints.

Trigger check for null values in city.


Page 43 of 54

Restrict_references: Built-in sub program used to check

for DML stmts in methods

WNDS: Write no database state

Pragma: Instruction to Plsql supports to execute sub

program in declaration section

Object with methods:

Create or replace type etype as object(ecode number, name varchar2(90), basic number(90),

job varchar2(4), hiredate date,

member function gross(vbasic number) return number,

pragma restrict_references(gross, WNDS),

member function experience(doj date) return number,

Pragma restrict_references(experience, WNDS));

Create or replace type body etype as

Member function gross(vbasic number) return number

Is

Begin

Return(round(vbasic+vbasic*.25+vbasic*.35-vbasic*.15));

End;

Member function experience(doj date) return number

Is

Begin

Return(round(months_between(sysdate, doj))/12);

End;

End;

Create table etab of etype; Row object

Insert into etab values(1001, ‘RAM’ 3000, ‘Analyst’, ’10-dec-01’);

Select ecode, name, job, basic, e.gross(basic) “gross”, e.experience(hiredate) “exp in years”

from etab e;


Page 44 of 54

Collections(8.0):

A group of similar rows stored at one location.

2 types…

1. Nested table:

Table with in a table

A column holds a table of contents in it.

Ex:

Create type book_type as object (bookid number(4), title varchar2(90), doi date, dor

datae);

Object representing table of books

Create type books is table of book_type;

Create table stu_lib(roll number, name varchar2(90), course varchar2(90), binfo books)

nested table binfo store as book_table;

Desc book_type;

Desc books;

Desc stu_lib;

Insert into stu_lib values(101, ‘RAM’, ‘Oracle 9i’,

books(book_type(11, ‘sql’, sysdate, sysdate+10),

book_type(12, ‘plsql’, sysdate, sysdate+11),

book_type(13, ‘dba’, sysdate, sysdate+15)));

Select * from stu_lib;

Select name, binfo from stu_lib;

Roll Name Course Binfo

101

102

Book id Title Doi Dor

11

12

13

2. Varrying arrays examples:

It is an array of elements stored in a column.

EX:

i. Create type blist is varray(5) of varchar2(25);

--Object representing array of books.

ii. Create table items(itno number, name varchar2(90), brands blist);

iii. Insert into items values(102, ‘Micro-oven’,

blist(‘LG’, ‘ONIDA’, ‘IFB’));

iv. desc blist

desc items


Page 45 of 54

v. update items set brands = blist(‘LG’, ‘ONIDA’, ‘IFB’, ‘KENSTAR’,

‘SAMSUNG’) where itno = 102;

vi. select name, brands from items;

BINFO

Roll Name Course Bid Title Doi dor

101 11

12

13

102

Nested table Varray

Stored outside the table Stored with in the table

DML is allowed DML is not allowed

Stores unlimited data stores limited data.

Note:

Collections will not support constraints

Long, raw, lob data types are not valid in

collections.

Collections will not support dev-6i, VB (GUI

tools)

Java supports collections.

DML on nested tables:

THE: Operator is used to perform DML on nested table.

EX:

Inserting a book(only 1 book allowed)

Insert into

THE(select binfo from stud_lib where roll=101)

Values(book_type(15, ‘DBA’, sysdate, sysdate+10));

Changing a book details

Update

THE(select binfo from stu_lib where roll = 101)

Set dor=sysdate+3 where book_id = 12;

Update

THE(select binfo from stu_lib where roll=101)

Set title=’Tuning’ where book_id = 15;

Removing a book info

Delete from

THE(select binfo from stu_lib where roll=101)

Where book_id =12;


Page 46 of 54

LOBS – Large Objects(8.0):

Used to represent huge loads of fdata

Max limit is 4 GB

N columns allowed per table

4 types…

1. CLOB – Char Large Object

Used to represent char information upto 4 GB.

2. NCLOB – Native char large object

Used to represent other language text.

*Nchar, Nvarchar2 -- Represents other language test.

3. BLOB – Binary large object

Used to represent images upto 4GB.

4. BFILE – Binary File

Used to hold the address of OS file.

Representing Null values:

CLOB - Empty_CLOB( )

NCLOB- Empty_NCLOB( )

BLOB - Empty_BLOB( )

BFILE - Skip the column while inserting

EX: Create table emp_lob(ecode number, name varchar(90), description CLOB, sal number,

photo BLOB, edocs BFILE);

Making logical dir to insert Bfile value:

DBA > Grant create any directory to scott;

SCOTT > Create directory emp_dir as ‘C:\employ\bonds’;

> Insert into emp_lob values (101, ;’RAM’, ‘………’, 2100, empty_BLOB( ),

Bfilename(emp_dir, ‘e101.doc’));

> Select ecode, ename, sal, description from emp_lob;

Built in subprogram

used to represent OS

file details into bfile

column.


Page 47 of 54

Temparary Tables(8i):

Used to hold the information in logical memory but not in physical memory.

They hold the data for a particular period of time but not permanently.

EX-1:

Create global temporary table temp(c1 date) on commit delete rows;

Insert into temp values(sysdate);

Select * from temp; --28-Mar-2012

Commit;

Select * from temp; -- No rows.

EX-2:

Create global temporary table(c1 date) on commit preserve rows;



Commit;


Exit;

Select * from temp; --No rows.

Autonomous Transactions:

Defines the trigger as an independent transaction

Allows TCL commands in triggers.

Syntax:

Create trigger add-incr before insert on incr

For each row

Declare

Pragma autonomous transaction;

Update emp set sal = sal+ :new.amt

Where empno = :new.empno;

Commit;

End;

Select empno, sal from emp; --7900 8000

Insert into incr values(7900, 4000);


Select empno, amt from incr; --7900 4000

Rollback;

Select empno, amt from incr; --No rows



Page 48 of 54

REF CURSORS:

Dynamic Cursors:

Supports to define the cursor without select stmt.

Select stmt will be provided while opening cursor.

Used to send cursor as an parameter in sub programs.

Made easy to transfer huge data through sub programs.

Syntax:

Type <ref cursor name> is ref cursor;

Create Package pack1

as

Type rear is ref cursor;

End pack1;

Procedure returning multiple rows through OUT parameter.

Create or replace procedure get_rows(vdept in number, vcur out pack1.rear)

Is

Begin

Open vcur for select empno, ename, sal, job from emp where deptno = vdept;

End;

Using Procedure:

Declare

scur pack1.rear;

vroll number(9);

vname varchar2(90);

vfee number;

Begin

Open scur for select roll, name, fee from student

Where course = ‘&course’;

Loop

Fetch when scur%not found;

DOPL(vroll||’’||vname||’||vfee);

End loop;

Close scur;

End;


Page 49 of 54

9i Features:

9i Joins:

Supports ANSI/ISO standard sql 1999 syntax.

Modularity for application sw tools to understand sql queries.

1. Natural Join

2. Join with using Equi Join (giving same output)

3. Join with ON

4. Inner Join

5. Left Outer Join

6. Right Outer Join

7. Full Outer Join Left outer join union right outer join

8. Cross Join

Select empno, ename, sal, deptno, dname, loc from emp Natural join dept;

Select empno, ename, sal, job, deptno, dname, loc from emp join dept using (deptno);

Select e.empno, e.ename, e.sal, e.job, e.deptno, d.dname, d.loc from emp e join dept d on

(e.deptno = d.deptno);

Select e.empno, e.ename, e.sal, e.job, e.deptno, d.dname, d.loc from emp e Inner join dept d

on (e.deptno = d.deptno);

Select e.empno, e.ename, e.sal, e.job, e.deptno, d.dname, d.loc from emp e left outer join

dept d on (e.deptno = d.deptno);

Select e.empno, e.ename, e.sal, e.job, e.deptno, d.dname, d.loc from emp e right outer join

dept d on (e.deptno = d.depno);

Select e.empno, e.ename, e.sal, e.job, e.deptno, d.dname, d.loc from emp e full outer join

dept don (e.deptno = d.deptno);

Select empno, ename, sal, job, deptno, dname, loc from emp cross join dept;

Date functions:

Systimestamp:

Gives date and time including rational seconds in SERVER time zone.

Current_timestamp:

Gives date and time including fractional seconds in CLIENT time zone.

Sysdate:

Gives only date in server time zone.

Current_date:

Gives only date in client time zone.

Extract:

Used to retrieve a particular value from the given date (day/month/year)

To_timestamp(d):

Converts given date into date and time info with am/pm.

Dbtimezone:

Gives server time zone value

Timestamp: Datatype

Automatically stores date and time information with am/pm


Page 50 of 54

Select systimestamp, current_timestamp from dual;

Select sysdate, current_date from dual;

Select dbtimezone from dual;

28-Mar-12 6:50:23 PM

Select extract(day from sysdate),

Extract(month from sysdate),

Extract(year from sysdate) from dual

Create table temp(c1 timestamp);


Select * from temp;

28-Mar-12 6:55:29 PM

NVL:

1. NVL(exp1, exp2):

NVL is used to fill Null value with known value.

If exp1 is NULL value then it will take exp2

If(exp1 = null)

Then expr2

Else

Expr1;

Endif;

2. NVL2(exp1, exp2, exp3):

If expr1 is null it manipulates exp3, if exp1 is not null it manipulates expr2

If(exp1 = null)

Then

Expr3

Else

Expr2

Endif;


Page 51 of 54

MULTIPLE INSERTS:

Supports to insert into more than 1 table at a time but input must be retrieved from existing

table.

EX: Make 3 empty tables same as dept table.

create table D1 as select * from dept where rownum is null;



insert all into D1 values(deptno, dname, loc)

into D2 values(deptno, dname, loc)

into D3 values(deptno, dname, loc)

Select * from dept;

Conditional Insert:

EX-1: Insert all

When (deptno <40) then

Into D1 values(deptno, dname, loc)

When (deptno<90) then


Else


Select * from dept;

EX-2: Insert all

When course = ‘oracle 9i’ then

Into Stu_ora(roll, name, fee)

When course = ‘java’ then

Into stu_java(roll, name, fee)

When course = ‘unix’ then

Into stu_unix values(roll, name, fee)

Select * from student;

MERGE:

Used to compare the 2 table contents them equal.

It supports only update & insert operations.

Ex:

Merge

Into temp t

Using emp e

On (t.tno = e.empno)

When matched then

Update set t.sal = e.sal,

t.comm = e.comm,

t.mgr = e.mgr,

t.deptno = e.deptno


Page 52 of 54

when not matched then

insert values(e.empno, e.ename, e.sal, e.com,…

Clauses:

Into --Target

Using --Source

On --join condition

Emp --1 crore rows

Temp --copy of emp -1 crore rows

Emp --5000 inserts & 1000 performed.

Partition Tables:

Supports to divide the huge tables into multiple partitions logically.

This will improve the performance while retrieving or updating the data.

Ex:

Create table emp(empno number(4) primary key,----

---------)

Partition by range(empno)

(partition p1 values less than (101),

Partition p2 values less than (201),

Partition p3 values less than (maxvalue));

Insert into emp values(10, ….); P1 P1 -- 1-100

“ “ (110….); P2 P2 -- 101-200

“ “ (210….); P3 P3 -- 201 – N

“ “ (310,…); P3

Select * from emp;

Update emp partition p2 set sal = sal+3000;

Select * from emp partition p3 where sal > 10000;

Alter table emp merge partitions (p1, p2);

Alter table emp drop partition p3;

Alter table emp add partition p3 values less than (401);

Alter table emp add partition p4 values less than <maxvalue);


Page 53 of 54

Oracle Utilities:

Performed by the DBA on the dos-Prompt of Server.

3 utilitites….

1. Export:

Used to translate information from database format to OS file format.

EX:

CD:\> EXP

Enter username & pwd : scott/tiger

Enter export buffer size(default): Press “enter” key.

Enter export file name: mydata.dmp

Export user(u), table(t) database : T

Export table data : Y

Compress the contents while exporting : Y

Enter table1:emp

Enter table 2:dept

Enter table3: Press “enter” key

2. Import:

Used to translate information from OS file format to database format.

EX:

CD:\> IMP

Enter username & pwd : scott/tiger

Enter import buffer size(default): Press “enter” key.

Enter import file name: mydata.dmp

Export entire file : Y

Export table data : Y

List the contents while exporting : N

Ignore error due to object existence : N

3. Sql Loader:

Used to transfer data from OS file to database table.

Step-1:

Abc.txt

10, sales, Hyd

20, import, Pune

30, export, Chennai

40, researc, Bgr


Page 54 of 54

Step-2:

abc.ctl (control file)

Load data

infile abc.txt

insert into table dept

fields terminated by ‘,’

(deptno, dname, loc)

Step-3:

C:\>SQLLDR Userid = scott/Tiger control = abc.ctl

NOTE:

Insert (Default option): Supports

only if table is empty.

Append: Allows to add the rows

to existing table with data.

To avoid the abc.ctl.txt(.txt). we

need to put file in “”.

Oracle 10g: [ G Grid]

Supports “Grid” technology

It is an DBA version

It supports to store “EJB” components directly into database without translation.

It supports with “regular expr’s” to search character data.

It provides high performance compared to other versions.

<<<<<<<<<<<<<<<<<<<<<<<<<<<<< All the Best >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

Don’t Judge a book, by its cover

PLSQL

Documents

server time zone

client time zone

replace package epack

ename type index

sal type index

sal bulk collect

create table customer

member function gross