INTRODUCTION
CHARACTERSTICS
Highly structured, readable and accessible language.
Standard and Protable language.
Embedded language.
Improved execution authority.
10g FEATURES
Optimized compiler.
To change the optimizer settings for the entire database, set the database parameter
PLSQL_OPTIMIZE_LEVEL. Valid settings are as follows
0
-
No optimization
1
-
Moderate optimization
2
-
Aggressive optimization
These settings are also modifiable for the current session.
SQL> alter session set plsql_optimze_level=2;
Oracle retains optimizer settings on a module-by-module basis. When you recompile a particular module with nondefault settings, the settings will stick allowing you to recompile later on using REUSE SETTINGS.
SQL> Alter procedure proc compile plsql_optimize_level=1;
SQL> Alter procedure proc compile reuse settings;
Compile-time warnings.
Starting with oracle database 10g release 1 you can enable additional compile-time warnings to help make your programs more robust. The compiler can detect potential runtime problems with your code, such as identifying lines of code that will never be run. This process, also known as lint checking.
To enable these warnings fo the entire database, set the database parameter PLSQL_WARNINGS.
These settings are also modifiable for the current session.
SQL> alter session set plsql_warnings = ‘enable:all’;
The above can be achieved using the built-in package DBMS_WARNING.
Conditional compilation.
Conditional compilation allows the compiler to allow to compile selected parts of a program based on conditions you provide with the $IF directive.
Support for non-sequential collections in FORALL.
Improved datatype support.
Backtrace an exception to its line number.
When handling an error, how can you find the line number on which the error was originally raised?
In earlier release, the only way to do this was allow you exception to go unhandled and then view the full error trace stack.
Now you can call DBMS_UTILITY.FORMAT_ERROR_BACKTRACE function to obtain that stack and manipulate it programmatically within your program.
Set operators for nested tables.
Support for regular expressions.
Oracle database 10g supports the use of regular expressions inside PL/SQL code via four new built-in functions.
REGEXP_LIKE
REGEXP_INSTR
REGEXP_SUBSTR
REGEXP_REPLACE
Programmer-defined quoting mechanism.
Starting with oracle database 10g release 1, you can define your own quoting mechanism for string literals in both SQL and PL/SQL.
Use the characters q’(q followed by a single quote) to note the programmer-defined deliemeter for you string literal.
Ex:
DECLARE
v varchar(10) := 'computer';
BEGIN
dbms_output.put_line(q'*v = *' || v);
dbms_output.put_line(q'$v = $' || v);
END;
Output:
v = computer
v = computer
Many new built-in packages.
DBMS_SCHEDULER
Represents a major update to DBMS_JOB. DBMS_SCHEDULER provides much improved functionality for scheduling and executing jobs defined via stored procedures.
DBMS_CRYPTO
Offers the ability to encrypt and decrypt common oracle datatype, including RAWs, BLOBs, and CLOBs. It also provides globalization support for encrypting data across different charactersets.
DBMS_MONITOR
Provides an API to control additional tracing and statistics gathering of sessions.
DBMS_WARNING
Provides an API into the PL/SQL compiler warnings module, allowing you to read and change settings that control which warnings are suppressed, displayed, or treated as errors.
STANDARD PACKAGE
Oracle has defined in this special package. Oracle defines quite a few identifiers in this package, including built-in exceptions, functions and subtypes.
You can reference the built-in form by prefixing it with STANDARD.
The basic unit in any PL/SQL program is block. All PL/SQL programs are composed of blocks which can occur sequentially or nested.
BLOCK STRUCTURE
Declare
-- declarative section
Begin
-- executable section
Exception
-- exception section
End;
In the above declarative and exceptiona sections are optional.
BLOCK TYPES
Anonymous blocks
Named blocks
Labeled blocks
Subprograms
Triggers
ANONYMOUS BLOCKS
Anonymous blocks implies basic block structure.
Ex:
BEGIN
Dbms_output.put_line(‘My first program’):
END;
LABELED BLOCKS
Labeled blocks are anonymous blocks with a label which gives a name to the block.
Ex:
<<my_bloock>>
BEGIN
Dbms_output.put_line(‘My first program’):
END;
SUBPROGRAMS
Subprograms are procedures and functions. They can be stored in the database as stand-alone objects, as part of package or as methods of an object type.
TRIGGERS
Triggers consists of a PL/SQL block that is associated with an event that occur in the database.
NESTED BLOCKS
A block can be nested within the executable or exception section of an outer block.
IDENTIFIERS
Identifiers are used to name PL/SQL objects, such as variables, cursors, types and subprograms. Identifiers consists of a letter, optionally followed by any sequence of characters, including letters, numbers, dollar signs, underscores, and pound signs only. The maximum length for an identifier is 30 characters.
QUOTED IDENTIFIERS
If you want to make an identifier case sensitive, include characters such as spaces or use a reserved word, you can enclose the identifier in double quotation marks.
Ex:
DECLARE
"a" number := 5;
"A" number := 6;
BEGIN
dbms_output.put_line('a = ' || a);
dbms_output.put_line('A = ' || A);
END;
Output:
a = 6
A = 6
COMMENTS
Comments improve readability and make your program more understandable. They are ignored by the PL/SQL compiler. There are two types of comments available.
Single line comments
Multiline comments
SINGLE LINE COMMENTS
A single-line comment can start any point on a line with two dashes and continues until the end of the line.
Ex:
BEGIN
Dbms_output.put_line(‘hello’);
-- sample program
END;
MULTILINE COMMENTS
Multiline comments start with the /* delimiter and ends with */ delimiter.
Ex:
BEGIN
Dbms_output.put_line(‘hello’);
/* sample program */
END;
VARIABLE DECLERATIONS
Variables can be declared in declarative section of the block;
Ex:
DECLARE
a number;
b number := 5;
c number default 6;
CONSTANT DECLERATIONS
To declare a constant, you include the CONSTANT keyword, and you must supply a default value.
Ex:
DECLARE
b constant number := 5;
c constant number default 6;
NOT NULL CLAUSE
You can also specify that the variable must be not null.
Ex:
DECLARE
b constant number not null:= 5;
c number not null default 6;
ANCHORED DECLERATIONS
PL/SQL offers two kinds of achoring.
Scalar anchoring
Record anchoring
SCALAR ANCHORING
Use the %TYPE attribute to define your variable based on table’s column of some other PL/SQL scalar variable.
Ex:
DECLARE
dno dept.deptno%type;
Subtype t_number is number;
a t_number;
Subtype t_sno is student.sno%type;
V_sno t_sno;
RECORD ANCHORING
Use the %ROWTYPE attribute to define your record structure based on a table.
Ex:
DECLARE
V_dept dept%rowtype;
BENEFITS OF ANCHORED DECLARATIONS
Synchronization with database columns.
Normalization of local variables.
PROGRAMMER-DEFINED TYPES
With the SUBTYPE statement, PL/SQL allows you to define your own subtypes or aliases of predefined datatypes, sometimes referred to as abstract datatypes.
There are two kinds of subtypes.
Constrained
Unconstrained
CONSTRAINED SUBTYPE
A subtype that restricts or constrains the values normally allowd by the datatype itself.
Ex:
Subtype positive is binary_integer range 1..2147483647;
In the above declaration a variable that is declared as positive can store only ingeger greater than zero even though binary_integer ranges from -2147483647..+2147483647.
UNCONSTRAINED SUBTYPE
A subtype that does not restrict the values of the original datatype in variables declared with the subtype.
Ex:
Subtype float is number;
DATATYPE CONVERSIONS
PL/SQL can handle conversions between different families among the datatypes.
Conversion can be done in two ways.
Explicit conversion
Implicit conversion
EXPLICIT CONVERSION
This can be done using the built-in functions available.
IMPLICIT CONVERSION
PL/SQL will automatically convert between datatype families when possible.
Ex:
DECLARE
a varchar(10);
BEGIN
select deptno into a from dept where dname='ACCOUNTING';
END;
In the above variable a is char type and deptno is number type even though, oracle will automatically converts the numeric data into char type assigns to the variable.
PL/SQL can automatically convert between
Characters and numbers
Characters and dates
VARIABLE SCOPE AND VISIBILITY
The scope of a variable is the portion of the program in which the variable can be accessed. For PL/SQL variables, this is from the variable declaration until the end of the block. When a variable goes out of scope, the PL/SQL engine will free the memory used to store the variable.
The visibility of a variable is the portion of the program where the variable can be accessed without having to qualify the reference. The visibility is always within the scope. If it is out of scope, it is not visible.
Ex1:
DECLARE
a number;
-- scope of a
BEGIN
--------
DECLARE
b number;
-- scope of b
BEGIN
-----
END;
------
END;
Ex2:
DECLARE
a number;
b number;
BEGIN
-- a , b available here
DECLARE
b char(10);
BEGIN
-- a and char type b is available here
END;
-----
END;
Ex3:
<<my_block>>
DECLARE
a number;
b number;
BEGIN
-- a , b available here
DECLARE
b char(10);
BEGIN
-- a and char type b is available here
-- number type b is available using <<my_block>>.b
END;
------
END;
PL/SQL CONTROL STRUCTURES
PL/SQL has a variety of control structures that allow you to control the behaviour of the block as it runs. These structures include conditional statements and loops.
If-then-else
Case
Case with no else
Labeled case
Searched case
Simple loop
While loop
For loop
Goto and Labels
IF-THEN-ELSE
Syntax:
If <condition1> then
Sequence of statements;
Elsif <condition1> then
Sequence of statements;
……
Else
Sequence of statements;
End if;
Ex:
DECLARE
dno number(2);
BEGIN
select deptno into dno from dept where dname = 'ACCOUNTING';
if dno = 10 then
dbms_output.put_line('Location is NEW YORK');
elsif dno = 20 then
dbms_output.put_line('Location is DALLAS');
elsif dno = 30 then
dbms_output.put_line('Location is CHICAGO');
else
dbms_output.put_line('Location is BOSTON');
end if;
END;
Output:
Location is NEW YORK
CASE
Syntax:
Case test-variable
When value1 then sequence of statements;
When value2 then sequence of statements;
……
When valuen then sequence of statements;
Else sequence of statements;
End case;
Ex:
DECLARE
dno number(2);
BEGIN
select deptno into dno from dept where dname = 'ACCOUNTING';
case dno
when 10 then
dbms_output.put_line('Location is NEW YORK');
when 20 then
dbms_output.put_line('Location is DALLAS');
when 30 then
dbms_output.put_line('Location is CHICAGO');
else
dbms_output.put_line('Location is BOSTON');
end case;
END;
Output:
Location is NEW YORK
CASE WITHOUT ELSE
Syntax:
Case test-variable
When value1 then sequence of statements;
When value2 then sequence of statements;
……
When valuen then sequence of statements;
End case;
Ex:
DECLARE
dno number(2);
BEGIN
select deptno into dno from dept where dname = 'ACCOUNTING';
case dno
when 10 then
dbms_output.put_line('Location is NEW YORK');
when 20 then
dbms_output.put_line('Location is DALLAS');
when 30 then
dbms_output.put_line('Location is CHICAGO');
when 40 then
dbms_output.put_line('Location is BOSTON');
end case;
END;
Output:
Location is NEW YORK
LABELED CASE
Syntax:
<<label>>
Case test-variable
When value1 then sequence of statements;
When value2 then sequence of statements;
……
When valuen then sequence of statements;
End case;
Ex:
DECLARE
dno number(2);
BEGIN
select deptno into dno from dept where dname = 'ACCOUNTING';
<<my_case>>
case dno
when 10 then
dbms_output.put_line('Location is NEW YORK');
when 20 then
dbms_output.put_line('Location is DALLAS');
when 30 then
dbms_output.put_line('Location is CHICAGO');
when 40 then
dbms_output.put_line('Location is BOSTON');
end case my_case;
END;
Output:
Location is NEW YORK
SEARCHED CASE
Syntax:
Case
When <condition1> then sequence of statements;
When <condition2> then sequence of statements;
……
When <conditionn> then sequence of statements;
End case;
Ex:
DECLARE
dno number(2);
BEGIN
select deptno into dno from dept where dname = 'ACCOUNTING';
case dno
when dno = 10 then
dbms_output.put_line('Location is NEW YORK');
when dno = 20 then
dbms_output.put_line('Location is DALLAS');
when dno = 30 then
dbms_output.put_line('Location is CHICAGO');
when dno = 40 then
dbms_output.put_line('Location is BOSTON');
end case;
END;
Output:
Location is NEW YORK
SIMPLE LOOP
Syntax:
Loop
Sequence of statements;
Exit when <condition>;
End loop;
In the syntax exit when <condition> is equivalent to
If <condition> then
Exit;
End if;
Ex:
DECLARE
i number := 1;
BEGIN
loop
dbms_output.put_line('i = ' || i);
i := i + 1;
exit when i > 5;
end loop;
END;
Output:
i = 1
i = 2
i = 3
i = 4
i = 5
WHILE LOOP
Syntax:
While <condition> loop
Sequence of statements;
End loop;
Ex:
DECLARE
i number := 1;
BEGIN
While i <= 5 loop
dbms_output.put_line('i = ' || i);
i := i + 1;
end loop;
END;
Output:
i = 1
i = 2
i = 3
i = 4
i = 5
FOR LOOP
Syntax:
For <loop_counter_variable> in low_bound..high_bound loop
Sequence of statements;
End loop;
Ex1:
BEGIN
For i in 1..5 loop
dbms_output.put_line('i = ' || i);
end loop;
END;
Output:
i = 1
i = 2
i = 3
i = 4
i = 5
Ex2:
BEGIN
For i in reverse 1..5 loop
dbms_output.put_line('i = ' || i);
end loop;
END;
Output:
i = 5
i = 4
i = 3
i = 2
i = 1
NULL STATEMENT
Usually when you write a statement in a program, you want it to do something. There are cases, however, when you want to tell PL/SQL to do absolutely nothing, and that is where the NULL comes.
The NULL statement deos nothing except pass control to the next executable statement.
You can use NULL statement in the following situations.
Improving program readability.
Sometimes, it is helpful to avoid any ambiguity inherent in an IF statement that doesn’t cover all possible cases. For example, when you write an IF statement, you do not have to include an ELSE clause.
Nullifying a raised exception.
When you don’t want to write any special code to handle an exception, you can use the NULL statement to make sure that a raised exception halts execution of the current PL/SQL block but does not propagate any exceptions to enclosing blocks.
Using null after a label.
In some cases, you can pair NULL with GOTO to avoid having to execute additional statements. For example, I use a GOTO statement to quickly move to the end of my program if the state of my data indicates that no further processing is required. Because I do not have to do anything at the termination of the program, I place a NULL statement after the label because at least one executable statement is required there. Even though NULL deos nothing, it is still an executable statement.
GOTO AND LABELS
Syntax:
Goto label;
Where label is a label defined in the PL/SQL block. Labels are enclosed in double angle brackets. When a goto statement is evaluated, control immediately passes to the statement identified by the label.
Ex:
BEGIN
For i in 1..5 loop
dbms_output.put_line('i = ' || i);
if i = 4 then
goto exit_loop;
end if;
end loop;
<<exit_loop>>
Null;
END;
Output:
i = 1
i = 2
i = 3
i = 4
RESTRICTIONS ON GOTO
It is illegal to branch into an inner block, loop.
At least one executable statement must follow.
It is illegal to branch into an if statement.
It is illegal to branch from one if statement to another if statement.
It is illegal to branch from exception block to the current block.
PRAGMAS
Pragmas are compiler directives. They serve as instructions to the PL/SQL compiler. The compiler will act on the pragma during the compilation of the block.
Syntax:
PRGAMA instruction_to_compiler.
PL/SQL offers several pragmas:
AUTONOMOUS_TRANSACTION
EXCEPTION_INIT
RESTRICT_REFERENCES
SERIALLY_REUSABLE
SUBPROGRAMS
PROCEDURES
A procedure is a module that performs one or more actions.
Syntax:
Procedure [schema.]name [(parameter1 [,parameter2 …])]
[authid definer | current_user] is
-- [declarations]
Begin
-- executable statements
[Exception
-- exception handlers]
End [name];
In the above authid clause defines whether the procedure will execute under the authority of the definer of the procedure or under the authority of the current user.
FUNCTIONS
A function is a module that returns a value.
Syntax:
Function [schema.]name [(parameter1 [,parameter2 …])]
Return return_datatype
[authid definer | current_user]
[deterministic]
[parallel_enable] is
-- [declarations]
Begin
-- executable statements
[Exception
-- exception handlers]
End [name];
In the above authid clause defines whether the procedure will execute under the authority of the definer of the procedure or under the authority of the current user.
Deterministic clause defines, an optimization hint that lets the system use a saved copy of the function’s return result, if available. The quety optimizer can choose whether to use the saved copy or re-call the function.
Parallel_enable clause defines, an optimization hint that enables the function to be executed in parallel when called from within SELECT statement.
PARAMETER MODES
In (Default)
Out
In out
IN
In parameter will act as pl/sql constant.
OUT
Out parameter will act as unintialized variable.
You cannot provide a default value to an out parameter.
Any assignments made to out parameter are rolled back when an exception is raised in the program.
An actual parameter corresponding to an out formal parameter must be a variable.
IN OUT
In out parameter will act as initialized variable.
An actual parameter corresponding to an in out formal parameter must be a variable.
DEFAULT PARAMETERS
Default Parameters will not allow in the beginning and middle.
Out and In Out parameters can not have default values.
Ex:
procedure p(a in number default 5, b in number default 6, c in number default 7) – valid
procedure p(a in number, b in number default 6, c in number default 7) – valild
procedure p(a in number, b in number, c in number default 7) – valild
procedure p(a in number, b in number default 6, c in number) – invalild
procedure p(a in number default 5, b in number default 6, c in number) – invalild
procedure p(a in number default 5, b in number, c in number) – invalild
NOTATIONS
Notations are of two types.
Positional notation
Name notation
We can combine positional and name notation but positional notation can not be followed by the name notation.
Ex:
Suppose we have a procedure proc(a number,b number,c number) and we have one
anonymous block which contains v1,v2, and v3;
SQL> exec proc (v1,v2,v3)
-- Positional notation
SQL> exec proc (a=>v1,b=>v2,c=>v3)
-- Named notation
FORMAL AND ACTUAL PARAMETERS
Parametes which are in calling subprogram are actual parameters.
Parametes which are in called subprogram are formal parameters.
If any subprogram was called, once the call was completed then the values of formal
parameters are copied to the actual parameters.
Ex1:
CREATE OR REPLACE PROCEDURE SAMPLE(a in number,b out number,c in out number) is
BEGIN
dbms_output.put_line('After call');
dbms_output.put_line('a = ' || a ||' b = ' || b || ' c = ' || c);
b := 10;
c := 20;
dbms_output.put_line('After assignment');
dbms_output.put_line('a = ' || a ||' b = ' || b || ' c = ' || c);
END SAMPLE;
DECLARE
v1 number := 4;
v2 number := 5;
v3 number := 6;
BEGIN
dbms_output.put_line('Before call');
dbms_output.put_line('v1 = ' || v1 || ' v2 = ' || v2 || ' v3 = ' || v3);
sample(v1,v2,v3);
dbms_output.put_line('After completion of call');
dbms_output.put_line('v1 = ' || v1 || ' v2 = ' || v2 || ' v3 = ' || v3);
END;
Output:
Before call
v1 = 4 v2 = 5 v3 = 6
After call
a = 4 b = c = 6
After assignment
a = 4 b = 10 c = 20
After completion of call
v1 = 4 v2 = 10 v3 = 20
Ex2:
CREATE OR REPLACE FUN(a in number,b out number,c in out number) return number IS
BEGIN
dbms_output.put_line('After call');
dbms_output.put_line('a = ' || a || ' b = ' || b || ' c = ' || c);
dbms_output.put_line('Before assignement Result = ' || (a*nvl(b,1)*c));
b := 5;
c := 7;
dbms_output.put_line('After assignment');
dbms_output.put_line('a = ' || a || ' b = ' || b || ' c = ' || c);
return (a*b*c);
END FUN;
DECLARE
v1 number := 1;
v2 number := 2;
v3 number := 3;
v number;
BEGIN
dbms_output.put_line('Before call');
dbms_output.put_line('v1 = ' || v1 || ' v2 = ' || v2 || ' v3 = ' || v3);
v := fun(v1,v2,v3);
dbms_output.put_line('After call completed');
dbms_output.put_line('v1 = ' || v1 || ' v2 = ' || v2 || ' v3 = ' || v3);
dbms_output.put_line('Result = ' || v);
END;
Output:
Before call
v1 = 1 v2 = 2 v3 = 3
After call
a = 1 b = c = 3
Before assignement Result = 3
After assignment
a = 1 b = 5 c = 7
After call completed
v1 = 1 v2 = 5 v3 = 7
Result = 35
RESTRICTIONS ON FORMAL PARAMETERS
By declaring with specified size in actual parameters.
By declaring formal parameters with %type specifier.
USING NOCOPY
Nocopy is a hint, not a command. This means that the compiler might silently decide that it can’t fulfill your request for a nocopy parameter.
The copying from formal to actual can be restricted by issuing nocopy qualifier.
To pass the out and in out parameters by reference use nocopy qualifier.
Ex:
CREATE OR REPLACE PROCEDURE PROC(a in out nocopy number) IS
BEGIN
----
END PROC;
CALL AND EXEC
Call is a SQL statement, which can be used to execute subprograms like exec.
Syntax:
Call subprogram_name([argument_list]) [into host_variable];
The parantheses are always required, even if the subprogram takes no arguments.
We can not use call with out and in out parameters.
Call is a SQL statement, it is not valid inside a PL/SQL block;
The INTO clause is used for the output variables of functions only.
We can not use ‘exec’ with out or in out parameters.
Exec is not valid inside a PL/SQL block;
Ex1:
CREATE OR REPLACE PROC IS
BEGIN
dbms_output.put_line('hello world');
END PROC;
Output:
SQL> call proc();
hello world
Ex2:
CREATE OR REPLACE PROC(a in number,b in number) IS
BEGIN
dbms_output.put_line('a = ' || a || ' b = ' || b);
END PROC;
Output:
SQL> call proc(5,6);
a = 5 b = 6
Ex3:
CREATE OR REPLACE FUNCTION FUN RETURN VARCHAR IS
BEGIN
return 'hello world';
END FUN;
Output:
SQL> variable v varchar(20)
SQL> call fun() into :v;
SQL> print v
hello world
CALL BY REFERENCE AND CALL BY VALUE
In parameters by default call by reference where as out and in out call by value.
When parameter passed by reference, a pointer to the actual parameter is passed to the
corresponding formal parameter.
When parameter passed by value it copies the value of the actual parameter to the formal parameter.
Call by reference is faster than the call by value because it avoids the copying.
SUBPROGRAMS OVERLOADING
Possible with different number of parameters.
Possible with different types of data.
Possible with same type with objects.
Can not be possible with different types of modes.
We can overload local subprograms also.
Ex:
SQL> create or replace type t1 as object(a number);/
SQL> create or replace type t1 as object(a number);/
DECLARE
i t1 := t1(5);
j t2 := t2(5);
PROCEDURE P(m t1) IS
BEGIN
dbms_output.put_line('a = ' || m.a);
END P;
PROCEDURE P(n t2) IS
BEGIN
dbms_output.put_line('b = ' || n.b);
END P;
PROCEDURE PRODUCT(a number,b number) IS
BEGIN
dbms_output.put_line('Product of a,b = ' || a * b);
END PRODUCT;
PROCEDURE PRODUCT(a number,b number,c number) IS
BEGIN
dbms_output.put_line('Product of a,b = ' || a * b * c);
END PRODUCT;
BEGIN
p(i);
p(j);
product(4,5);
product(4,5,6);
END;
Output:
a = 5
b = 5
Product of a,b = 20
Product of a,b = 120
BENEFITS OF OVERLOADING
Supporting many data combinations
Fitting the program to the user.
RESTRICTIONS ON OVERLOADING
Overloaded programs with parameter lists that differ only by name must be called using named notation.
The parameter list of overloaded programs must differ by more than parameter mode.
All of the overloaded programs must be defined within the same PL/SQL scope or block.
Overloaded functions must differ by more than their return type.
IMPORTANT POINTS ABOUT SUBPROGRAMS
When a stored subprogram is created, it is stored in the data dictionary.
The subprogram is stored in compile form which is known as p-code in addition to the source text.
The p-code has all of the references in the subprogram evaluated, and the source code is translated into a form that is easily readable by PL/SQL engine.
When the subprogram is called, the p-code is read from the disk, if necessary, and executed.
Once it reads from the disk, the p-code is stored in the shared pool portion of the system global area (SGA), where it can be accessed by multiple users as needed.
Like all of the contents of the shared pool, p-code is aged out of the shared pool according to a least recently used (LRU) algorithm.
Subprograms can be local.
Local subprograms must be declared in the declarative section of PL/SQL block and called from the executable section.
Subprograms can not have the declarative section separately.
Stored subprograms can have local subprograms;
Local subprograms also can have local subprograms.
If the subprogram contains a variable with the same name as the column name of the table then use the dot method to differentiate (subprogram_name.sal).
Subprograms can be invalidated.
PROCEDURES V FUNCTIONS
Procedures may return through out and in out parameters where as function must return.
Procedures can not have return clause where as functions must.
We can use call statement directly for executing procedure where as we need to declare a variable in case of functions.
Functions can use in select statements where as procedures can not.
Functions can call from reports environment where as procedures can not.
We can use exec for executing procedures where as functions can not.
Function can be used in dbms_output where as procedure can not.
Procedure call is a standalone executable statement where as function call is a part of an executable statement.
STORED V LOCAL SUBPROGRAMS
The stored subprogram is stored in compiled p-code in the database, when the procedure is called it does not have to be compiled.
The local subprogram is compiled as part of its containing block. If the containing block
is anonymous and is run multiple times, the subprogram has to be compiled each time.
Stored subprograms can be called from any block submitted by a user who has execute privileges on the subprogram.
Local subprograms can be called only from the block containing the subprogram.
By keeping the stored subprogram code separate from the calling block, the calling block is shorter and easier to understand.
The local subprogram and the calling block are one and the same, which can lead to part
confusion. If a change to the calling block is made, the subprogram will be recompiled as
of the recompilation of the containing block.
The compiled p-code can be pinned in the shared pool using the DBMS_SHARED_POOL
Package. This can improve performance.
Local subprograms cannot be pinned in the shared pool by themselves.
Stand alone stored subprograms can not be overloaded, but packaged subprograms can
be overloaded within the same package.
Local subprograms can be overloaded within the same block.
Ex1:
CREATE OR REPLACE PROCEDURE P IS
BEGIN
dbms_output.put_line('Stored subprogram');
END;
Output:
SQL> exec p
Stored subprogram
Ex2:
DECLARE
PROCEDURE P IS
BEGIN
dbms_output.put_line('Local subprogram');
END;
BEGIN
p;
END;
Output:
Local subprogram
COMPILING SUBPROGRAMS
SQL> Alter procedure P1 compile;
SQL> Alter function F1 compile;
SUBPROGRAMS DEPENDECIES
A stored subprogram is marked as invalid in the data dictionary if it has compile errors.
A stored subprogram can also become invalid if a DDL operation is performed on one of its dependent objects.
If a subprogram is invalidated, the PL/SQL engine will automatically attempt to recompile in the next time it is called.
If we have two procedures like P1 and P2 in which P1 depends on P2. If we compile P2 then P1 is invalidated.
SUBPROGRAMS DEPENDENCIES IN REMOTE DATABASES
We will call remote subprogram using connect string like P1@ORACLE;
If we have two procedures like P1 and P2 in which P1 depends on P2 but P2 was in remote database. If we compile P2 it will not invalidate P1 immediately because the data dictionary does not track remote dependencies.
Instead the validity of remote objects is checked at runtime. When P1 is called, the remote data dictionary is queried to determine the status of P2.
P1 and P2 are compared to see it P1 needs to be recompiled, there are two different methods of comparision
Timestamp Model
Signature Model
TIMESTAMP MODEL
This is the default model used by oracle.
With this model, the timestamps of the last modifications of the two objects are
compared.
The last_ddl_time field of user_objects contains the timestamp.
If the base object has a newer timestamp than the dependent object, the dependent
object will be recompiled.
ISSUES WITH THIS MODEL
If the objects are in different time zones, the comparison is invalid.
When P1 is in a client side PL/SQL engine such as oracle forms, in this case it may not possible to recompile P1, because the source for it may not be included with the forms.
SIGNATURE MODEL
When a procedure is created, a signature is stored in the data dictionary in addition to the p-code.
The signature encodes the types and order of the parametes.
When P1 is compiled the first time, the signature of P2 is included. Thus, P1 only needs to recompiled when the signature of P2 changes.
In order to use the signature model, the parameter REMOTE_DEPENDENCIES_MODE must be set to SIGNATURE. This is a parameter in the database initialization file.
THREE WAYS OF SETTING THIS MODE
Add the line REMOTE_DEPENDENCIES_MODE=SIGNATURE to the database initialization file. The next time the database is started, the mode will be set to SIGNATURE for all sessions.
Alter system set remote_dependencies_mode = signature;
This will affect the entire database (all sessions) from the time the statement is issued.
You must have the ALTER SYSTEM privilege to issue this command.
Alter session set remote_dependencies_mode = signature;
This will only affect your session
ISSUES WITH THIS MODEL
Signatures don’t get modified if the default values of formal parameters are changed.
Suppose P2 has a default value for one of its parameters, and P1 is using this default
value. If the default in the specification for P2 is changed, P1 will not be recompiled
by default. The old value for the default parameter will still be used until P1 is manually
recompiled.
If P1 is calling a packaged procedure P2, and a new overloaded version of P2 is added to
the remote package, the signature is not changed. P1 will still use the old version
(not the new overloaded one) until P1 is recompiled manually.
FORWARD DECLERATION
Before going to use the procedure in any other subprogram or other block , you must declare the prototype of the procedure in declarative section.
Ex1:
DECLARE
PROCEDURE P1 IS
BEGIN
dbms_output.put_line('From procedure p1');
p2;
END P1;
PROCEDURE P2 IS
BEGIN
dbms_output.put_line('From procedure p2');
p3;
END P2;
PROCEDURE P3 IS
BEGIN
dbms_output.put_line('From procedure p3');
END P3;
BEGIN
p1;
END;
Output:
p2;
*
ERROR at line 5:
ORA-06550: line 5, column 1:
PLS-00313: 'P2' not declared in this scope
ORA-06550: line 5, column 1:
PL/SQL: Statement ignored
ORA-06550: line 10, column 1:
PLS-00313: 'P3' not declared in this scope
ORA-06550: line 10, column 1:
PL/SQL: Statement ignored
Ex2:
DECLARE
PROCEDURE P2; -- forward declaration
PROCEDURE P3;
PROCEDURE P1 IS
BEGIN
dbms_output.put_line('From procedure p1');
p2;
END P1;
PROCEDURE P2 IS
BEGIN
dbms_output.put_line('From procedure p2');
p3;
END P2;
PROCEDURE P3 IS
BEGIN
dbms_output.put_line('From procedure p3');
END P3;
BEGIN
p1;
END;
Output:
From procedure p1
From procedure p2
From procedure p3
PRIVILEGES AND STORED SUBPROGRAMS
EXECUTE PREVILEGE
For stored subprograms and packages the relevant privilege is EXECUTE.
If user A had the procedure called emp_proc then user A grants execute privilege on procedure to user B with the following command.
SQL> Grant execute on emp_proc to user B.
Then user B can run the procedure by issuing
SQL> Exec user A.emp_proc
userA created the following procedure
CREATE OR REPLACE PROCEDURE P IS
cursor is select *from student1;
BEGIN
for v in c loop
insert into student2 values(v.no,v.name,v.marks);
end loop;
END P;
userA granted execute privilege to userB using
SQL> grant execute on p to userB
Then userB executed the procedure
SQL> Exec userA.p
If suppose userB also having student2 table then which table will populate whether userA’s or userB’s.
The answer is userA’s student2 table only because by default the procedure will execute under the privlige set of its owner.
The above procedure is known as definer’s procedure.
HOW TO POPULATE USER B’s TABLE
Oracle introduces Invoker’s and Definer’s rights.
By default it will use the definer’s rights.
An invoker’s rights routine can be created by using AUTHID clause to populate the
userB’s table.
It is valid for stand-alone subprograms, package specifications, and object type
specifications only.
userA created the following procedure
CREATE OR REPLACE PROCEDURE P
AUTHID CURRENT_USER IS
cursor is select *from student1;
BEGIN
for v in c loop
insert into student2 values(v.no,v.name,v.marks);
end loop;
END P;
Then grant execute privilege on p to userB.
Executing the procedure by userB, which populates userB’s table.
The above procedure is called invoker’s procedure.
Instead of current_user of authid clause, if you use definer then it will be called definer’ procedure.
STORED SUBPROGRAMS AND ROLES
we have two users saketh and sudha in which saketh has student table and sudha does not.
Sudha is going to create a procedure based on student table owned by saketh. Before doing this saketh must grant the permissions on this table to sudha.
SQL> conn saketh/saketh
SQL> grant all on student to sudha;
then sudha can create procedure
SQL> conn sudha/sudha
CREATE OR REPLACE PROCEDURE P IS
cursor c is select *from saketh.student;
BEGIN
for v in c loop
dbms_output.put_line(‘No = ‘ || v.no);
end loop;
END P;
here procedure will be created.
If the same privilege was granted through a role it wont create the procedure.
Examine the following code
SQL> conn saketh/saketh
SQL> create role saketh_role;
SQL> grant all on student to saketh_role;
SQL> grant saketh_role to sudha;
then conn sudha/sudha
CREATE OR REPLACE PROCEDURE P IS
cursor c is select *from saketh.student;
BEGIN
for v in c loop
dbms_output.put_line(‘No = ‘ || v.no);
end loop;
END P;
The above code will raise error instead of creating procedure .
This is because of early binding which PL/SQL uses by default in which references are evaluated in compile time but when you are using a role this will affect immediately.
ISSUES WITH INVOKER’S RIGHTS
In an invoker’s rights routine, external references in SQL statements will be resolved using the caller’s privilege set.
But references in PL/SQL statements are still resolved under the owner’s privilege set.
TRIGGERS, VIEWS AND INVOKER’S RIGHTS
A database trigger will always be executed with definer’s rights and will execute under the privilege set of the schema that owns the triggering table.
This is also true for PL/SQL function that is called from a view. In this case, the function will execute under the privilege set of the view’s owner.
PACKAGES
A package is a container for related objects. It has specification and body. Each of them is stored separately in data dictionary.
PACKAGE SYNTAX
Create or replace package <package_name> is
-- package specification includes subprograms signatures, cursors and global or public
variables.
End <package_name>;
Create or replace package body <package_name> is
-- package body includes body for all the subprograms declared in the spec, private
Variables and cursors.
Begin
-- initialization section
Exception
-- Exception handling seciton
End <package_name>;
IMPORTANT POINGS ABOUT PACKAGES
The first time a packaged subprogram is called or any reference to a packaged variable or type is made, the package is instantiated.
Each session will have its own copy of packaged variables, ensuring that two sessions executing subprograms in the same package use different memory locations.
In many cases initialization needs to be run the first time the package is instantiated within a session. This can be done by adding initialization section to the package body after all the objects.
Packages are stored in the data dictionary and can not be local.
Packaged subprograms has an advantage over stand alone subprogram.
When ever any reference to package, the whole package p-code was stored in shared pool of SGA.
Package may have local subprograms.
You can include authid clause inside the package spec not in the body.
The execution section of a package is know as initialization section.
You can have an exception section at the bottom of a package body.
Packages subprograms are not invalidated.
COMPILING PACKAGES
SQL> Alter package PKG compile;
SQL> Alter package PKG compile specification;
SQL> Alter package PKG compile body;
PACKAGE DEPENDENCIES
The package body depends on the some objects and the package header.
The package header does not depend on the package body, which is an advantage of packages.
We can change the package body with out changing the header.
PACKAGE RUNTIME STATE
Package runtime state is differ for the following packages.
Serially reusable packages
Non serially reusable packages
SERIALLY REUSABLE PACKAGES
To force the oracle to use serially reusable version then include PRAGMA SERIALLY_REUSABLE in both package spec and body, Examine the following package.
CREATE OR REPLACE PACKAGE PKG IS
pragma serially_reusable;
procedure emp_proc;
END PKG;
CREATE OR REPLACE PACKAGE BODY PKG IS
pragma serially_reusable;
cursor c is select ename from emp;
PROCEDURE EMP_PROC IS
v_ename emp.ename%type;
v_flag boolean := true;
v_numrows number := 0;
BEGIN
if not c%isopen then
open c;
end if;
while v_flag loop
fetch c into v_ename;
v_numrows := v_numrows + 1;
if v_numrows = 5 then
v_flag := false;
end if;
dbms_output.put_line('Ename = ' || v_ename);
end loop;
END EMP_PROC;
END PKG;
SQL> exec pkg.emp_proc
Ename = SMITH
Ename = ALLEN
Ename = WARD
Ename = JONES
Ename = MARTIN
SQL> exec pkg.emp_proc
Ename = SMITH
Ename = ALLEN
Ename = WARD
Ename = JONES
Ename = MARTIN
The above package displays the same output for each execution even though the cursor is not closed.
Because the serially reusable version resets the state of the cursor each time it was called.
NON SERIALL Y REUSABLE PACKAGES
This is the default version used by the oracle, examine the following package.
CREATE OR REPLACE PACKAGE PKG IS
procedure emp_proc;
END PKG;
CREATE OR REPLACE PACKAGE BODY IS
cursor c is select ename from emp;
PROCEDURE EMP_PROC IS
v_ename emp.ename%type;
v_flag boolean := true;
v_numrows number := 0;
BEGIN
if not c%isopen then
open c;
end if;
while v_flag loop
fetch c into v_ename;
v_numrows := v_numrows + 1;
if v_numrows = 5 then
v_flag := false;
end if;
dbms_output.put_line('Ename = ' || v_ename);
end loop;
END EMP_PROC;
END PKG;
SQL> exec pkg.emp_proc
Ename = SMITH
Ename = ALLEN
Ename = WARD
Ename = JONES
Ename = MARTIN
SQL> exec pkg.emp_proc
Ename = BLAKE
Ename = CLARK
Ename = SCOTT
Ename = KING
Ename = TURNER
The above package displays the different output for each execution even though the cursor is not closed.
Because the non-serially reusable version remains the state of the cursor over database calls.
DEPENDENCIES OF PACKAGE RUNTIME STATE
Dependencies can exists between package state and anonymous blocks.
Examine the following program
Create this package in first session
CREATE OR REPLACE PACKAGE PKG IS
v number := 5;
procedure p;
END PKG;
CREATE OR REPLACE PACKAGE BODY PKG IS
PROCEDURE P IS
BEGIN
dbms_output.put_line('v = ' || v);
v := 10;
dbms_output.put_line('v = ' || v);
END P;
END PKG;
Connect to second session, run the following code.
BEGIN
pkg.p;
END;
The above code wil work.
Go back to first session and recreate the package using create.
Then connect to second session and run the following code again.
BEGIN
pkg.p;
END;
This above code will not work because of the following.
The anonymous block depends on pkg. This is compile time dependency.
There is also a runtime dependency on the packaged variables, since each session has its own copy of packaged variables.
Thus when pkg is recompiled the runtime dependency is followed, which invalidates the block and raises the oracle error.
Runtime dependencies exist only on package state. This includes variables and cursors declared in a package.
If the package had no global variables, the second execution of the anonymous block would have succeeded.
PURITY LEVELS
In general, calls to subprograms are procedural, they cannot be called from SQL statements. However, if a stand-alone or packaged function meets certain restrictions, it can be called during execution of a SQL statement.
User-defined functions are called the same way as built-in functions but it must meet different restrictions. These restrictions are defined in terms of purity levels.
There are four types of purity levels.
WNDS
--
Writes No Database State
RNDS
--
Reads No Database State
WNPS
--
Writes No Package State
RNPS
--
Reads No Package State
In addition to the preceding restrictions, a user-defined function must also meet the following requirements to be called from a SQL statement.
The function has to be stored in the database, either stand-alone or as part of a package.
The function can take only in parametes.
The formal parameters must use only database types, not PL/SQL types such as boolean
or record.
The return type of the function must also be a database type.
The function must not end the current transaction with commit or rollback, or rollback to
a savepoint prior to the function execution.
It also must not issue any alter session or alter system commands.
RESTRICT_REFERENCES
For packaged functions, however, the RESTRICT_REFERENCES pragma is required to specify the purity level of a given function.
Syntax:
PRAGMA RESTRICT_REFERENCES(subprogram_name or package_name, WNDS [,WNPS] [,RNDS]
[,RNPS]);
Ex:
CREATE OR REPLACE PACKAGE PKG IS
function fun1 return varchar;
pragma restrict_references(fun1,wnds);
function fun2 return varchar;
pragma restrict_references(fun2,wnds);
END PKG;
CREATE OR REPLACE PACKAGE BODY PKG IS
FUNCTION FUN1 return varchar IS
BEGIN
update dept set deptno = 11;
return 'hello';
END FUN1;
FUNCTION FUN2 return varchar IS
BEGIN
update dept set dname ='aa';
return 'hello';
END FUN2;
END PKG;
The above package body will not created, it will give the following erros.
PLS-00452: Subprogram 'FUN1' violates its associated pragma
PLS-00452: Subprogram 'FUN2' violates its associated pragma
CREATE OR REPLACE PACKAGE BODY PKG IS
FUNCTION FUN1 return varchar IS
BEGIN
return 'hello';
END FUN1;
FUNCTION FUN2 return varchar IS
BEGIN
return 'hello';
END FUN2;
END PKG;
Now the package body will be created.
DEFAULT
If there is no RESTRICT_REFERENCES pragma associated with a given packaged function, it will not have any purity level asserted. However, you can change the default purity level for a package. The DEFAULT keyword is used instead of the subprogram name in the pragma.
Ex:
CREATE OR REPLACE PACKAGE PKG IS
pragma restrict_references(default,wnds);
function fun1 return varchar;
function fun2 return varchar;
END PKG;
CREATE OR REPLACE PACKAGE BODY PKG IS
FUNCTION FUN1 return varchar IS
BEGIN
update dept set deptno = 11;
return 'hello';
END FUN1;
FUNCTION FUN2 return varchar IS
BEGIN
update dept set dname ='aa';
return 'hello';
END FUN2;
END PKG;
The above package body will not created, it will give the following erros because the pragma will apply to all the functions.
PLS-00452: Subprogram 'FUN1' violates its associated pragma
PLS-00452: Subprogram 'FUN2' violates its associated pragma
CREATE OR REPLACE PACKAGE BODY PKG IS
FUNCTION FUN1 return varchar IS
BEGIN
return 'hello';
END FUN1;
FUNCTION FUN2 return varchar IS
BEGIN
return 'hello';
END FUN2;
END PKG;
Now the package body will be created.
TRUST
If the TRUST keyword is present, the restrictions listed in the pragma are not enforced. Rather, they are trusted to be true.
Ex:
CREATE OR REPLACE PACKAGE PKG IS
function fun1 return varchar;
pragma restrict_references(fun1,wnds,trust);
function fun2 return varchar;
pragma restrict_references(fun2,wnds,trust);
END PKG;
CREATE OR REPLACE PACKAGE BODY PKG IS
FUNCTION FUN1 return varchar IS
BEGIN
update dept set deptno = 11;
return 'hello';
END FUN1;
FUNCTION FUN2 return varchar IS
BEGIN
update dept set dname ='aa';
return 'hello';
END FUN2;
END PKG;
The above package will be created successfully.
IMPORTANT POINTS ABOUT RESTRICT_REFERENCES
This pragma can appear anywhere in the package specification, after the function
declaration.
It can apply to only one function definition.
For overload functions, the pragma applies to the nearest definition prior to the pragma.
This pragma is required only for packages functions not for stand-alone functions.
The Pragma can be declared only inside the package specification.
The pragma is checked at compile time, not runtime.
It is possible to specify without any purity levels when trust or combination of default
and trust keywords are present.
PINNING IN THE SHARED POOL
The shared pool is the portion of the SGS that contains, among other things, the p-code of compiled subprograms as they are run. The first time a stored a store subprogram is called, the p-code is loaded from disk into the shared pool. Once the object is no longer referenced, it is free to be aged out. Objects are aged out of the shared pool using an LRU(Least Recently Used) algorithm.
The DBMS_SHARED_POOL package allows you to pin objects in the shared pool. When an object is pinned, it will never be aged out until you request it, no matter how full the pool gets or how often the object is accessed. This can improve performance, as it takes time to reload a package from disk.
DBMS_SHARED_POOL has four procedures
KEEP
UNKEEP
SIZES
ABORTED_REQUEST_THRESHOLD
KEEP
The DBMS_SHARED_POOL.KEEP procedure is used to pin objects in the pool.
Syntax:
PROCEDURE KEEP(object_name varchar2,flag char default ‘P’);
Here the flag represents different types of flag values for different types of objects.
P
--
Package, function or procedure
Q
--
Sequence
R
--
Trigger
C
--
SQL Cursor
T
--
Object type
JS
--
Java source
JC
--
Java class
JR
--
Java resource
JD
--
Java shared data
UNKEEP
UNKEEP is the only way to remove a kept object from the shared pool, without restarting the database. Kept objects are never aged out automatically.
Syntax:
PROCEDURE UNKEEP(object_name varchar2, flag char default ‘P’);
SIZES
SIZES will echo the contents of the shared pool to the screen.
Syntax:
PROCEDURE SIZES(minsize number);
Objects with greater than the minsize will be returned. SIZES uses DBMS_OUTPUT to return the data.
ABORTED_REQUEST_THRESHOLD
When the database determines that there is not enough memory in the shared pool to satisfy a given request, it will begin aging objects out until there is enough memory. It enough objects are aged out, this can have a performance impact on other database sessions. The ABORTED_REQUEST_THRESHOLD can be used to remedy this.
Syntax:
PROCEDURE ABORTED_REQUEST_THRESHOLD(threshold_size number);
Once this procedure is called, oracle will not start aging objects from the pool unless at least threshold_size bytes is needed.
DATA MODEL FOR SUBPROGRAMS AND PACKAGES
USER_OBJECTS
USER_SOURCE
USER_ERRORS
CURSORS
Cursor is a pointer to memory location which is called as context area which contains the information necessary for processing, including the number of rows processed by the statement, a pointer to the parsed representation of the statement, and the active set which is the set of rows returned by the query.
Cursor contains two parts
Header
Body
Header includes cursor name, any parameters and the type of data being loaded.
Body includes the select statement.
Ex:
Cursor c(dno in number) return dept%rowtype is select *from dept;
In the above
Header – cursor c(dno in number) return dept%rowtype
Body – select *from dept
CURSOR TYPES
Implicit (SQL)
Explicit
Parameterized cursors
REF cursors
CURSOR STAGES
Open
Fetch
Close
CURSOR ATTRIBUTES
%found
%notfound
%rowcount
%isopen
%bulk_rowcount
%bulk_exceptions
CURSOR DECLERATION
Syntax:
Cursor <cursor_name> is select statement;
Ex:
Cursor c is select *from dept;
CURSOR LOOPS
Simple loop
While loop
For loop
SIMPLE LOOP
Syntax:
Loop
Fetch <cursor_name> into <record_variable>;
Exit when <cursor_name> % notfound;
<statements>;
End loop;
Ex:
DECLARE
cursor c is select * from student;
v_stud student%rowtype;
BEGIN
open c;
loop
fetch c into v_stud;
exit when c%notfound;
dbms_output.put_line('Name = ' || v_stud.name);
end loop;
close c;
END;
Output:
Name = saketh
Name = srinu
Name = satish
Name = sudha
WHILE LOOP
Syntax:
While <cursor_name> % found loop
Fetch <cursor_name> into <record_variable>;
<statements>;
End loop;
Ex:
DECLARE
cursor c is select * from student;
v_stud student%rowtype;
BEGIN
open c;
fetch c into v_stud;
while c%found loop
fetch c into v_stud;
dbms_output.put_line('Name = ' || v_stud.name);
end loop;
close c;
END;
Output:
Name = saketh
Name = srinu
Name = satish
Name = sudha
FOR LOOP
Syntax:
for <record_variable> in <cursor_name> loop
<statements>;
End loop;
Ex:
DECLARE
cursor c is select * from student;
BEGIN
for v_stud in c loop
dbms_output.put_line('Name = ' || v_stud.name);
end loop;
END;
Output:
Name = saketh
Name = srinu
Name = satish
Name = sudha
PARAMETARIZED CURSORS
This was used when you are going to use the cursor in more than one place with different values for the same where clause.
Cursor parameters must be in mode.
Cursor parameters may have default values.
The scope of cursor parameter is within the select statement.
Ex:
DECLARE
cursor c(dno in number) is select * from dept where deptno = dno;
v_dept dept%rowtype;
BEGIN
open c(20);
loop
fetch c into v_dept;
exit when c%notfound;
dbms_output.put_line('Dname = ' || v_dept.dname || ' Loc = ' || v_dept.loc);
end loop;
close c;
END;
Output:
Dname = RESEARCH Loc = DALLAS
PACKAGED CURSORS WITH HEADER IN SPEC AND BODY IN PACKAGE BODY
cursors declared in packages will not close automatically.
In packaged cursors you can modify the select statement without making any changes to the cursor header in the package specification.
Packaged cursors with must be defined in the package body itself, and then use it as global for the package.
You can not define the packaged cursor in any subprograms.
Cursor declaration in package with out body needs the return clause.
Ex:
CREATE OR REPLACE PACKAGE PKG IS
cursor c return dept%rowtype is select * from dept;
procedure proc is
END PKG;
CREATE OR REPLACE PAKCAGE BODY PKG IS
cursor c return dept%rowtype is select * from dept;
PROCEDURE PROC IS
BEGIN
for v in c loop
dbms_output.put_line('Deptno = ' || v.deptno || ' Dname = ' || v.dname || '
Loc = ' || v.loc);
end loop;
END PROC;
END PKG;
Output:
SQL> exec pkg.proc
Deptno = 10 Dname = ACCOUNTING Loc = NEW YORK
Deptno = 20 Dname = RESEARCH Loc = DALLAS
Deptno = 30 Dname = SALES Loc = CHICAGO
Deptno = 40 Dname = OPERATIONS Loc = BOSTON
CREATE OR REPLACE PAKCAGE BODY PKG IS
cursor c return dept%rowtype is select * from dept where deptno > 20;
PROCEDURE PROC IS
BEGIN
for v in c loop
dbms_output.put_line('Deptno = ' || v.deptno || ' Dname = ' || v.dname || '
Loc = ' || v.loc);
end loop;
END PROC;
END PKG;
Output:
SQL> exec pkg.proc
Deptno = 30 Dname = SALES Loc = CHICAGO
Deptno = 40 Dname = OPERATIONS Loc = BOSTON
REF CURSORS AND CURSOR VARIABLES
This is unconstrained cursor which will return different types depends upon the user input.
Ref cursors can not be closed implicitly.
Ref cursor with return type is called strong cursor.
Ref cursor with out return type is called weak cursor.
You can declare ref cursor type in package spec as well as body.
You can declare ref cursor types in local subprograms or anonymous blocks.
Cursor variables can be assigned from one to another.
You can declare a cursor variable in one scope and assign another cursor variable with different scope, then you can use the cursor variable even though the assigned cursor variable goes out of scope.
Cursor variables can be passed as a parameters to the subprograms.
Cursor variables modes are in or out or in out.
Cursor variables can not be declared in package spec and package body (excluding subprograms).
You can not user remote procedure calls to pass cursor variables from one server to another.
Cursor variables can not use for update clause.
You can not assign nulls to cursor variables.
You can not compare cursor variables for equality, inequality and nullity.
Ex:
CREATE OR REPLACE PROCEDURE REF_CURSOR(TABLE_NAME IN VARCHAR) IS
type t is ref cursor;
c t;
v_dept dept%rowtype;
type r is record(ename emp.ename%type,job emp.job%type,sal emp.sal%type);
v_emp r;
v_stud student.name%type;
BEGIN
if table_name = 'DEPT' then
open c for select * from dept;
elsif table_name = 'EMP' then
open c for select ename,job,sal from emp;
elsif table_name = 'STUDENT' then
open c for select name from student;
end if;
loop
if table_name = 'DEPT' then
fetch c into v_dept;
exit when c%notfound;
dbms_output.put_line('Deptno = ' || v_dept.deptno || ' Dname = ' ||
v_dept.dname || ' Loc = ' || v_dept.loc);
elsif table_name = 'EMP' then
fetch c into v_emp;
exit when c%notfound;
dbms_output.put_line('Ename = ' || v_emp.ename || ' Job = ' || v_emp.job || ' Sal
= ' || v_emp.sal);
elsif table_name = 'STUDENT' then
fetch c into v_stud;
exit when c%notfound;
dbms_output.put_line('Name = ' || v_stud);
end if;
end loop;
close c;
END;
Output:
SQL> exec ref_cursor('DEPT')
Deptno = 10 Dname = ACCOUNTING Loc = NEW YORK
Deptno = 20 Dname = RESEARCH Loc = DALLAS
Deptno = 30 Dname = SALES Loc = CHICAGO
Deptno = 40 Dname = OPERATIONS Loc = BOSTON
SQL> exec ref_cursor('EMP')
Ename = SMITH Job = CLERK Sal = 800
Ename = ALLEN Job = SALESMAN Sal = 1600
Ename = WARD Job = SALESMAN Sal = 1250
Ename = JONES Job = MANAGER Sal = 2975
Ename = MARTIN Job = SALESMAN Sal = 1250
Ename = BLAKE Job = MANAGER Sal = 2850
Ename = CLARK Job = MANAGER Sal = 2450
Ename = SCOTT Job = ANALYST Sal = 3000
Ename = KING Job = PRESIDENT Sal = 5000
Ename = TURNER Job = SALESMAN Sal = 1500
Ename = ADAMS Job = CLERK Sal = 1100
Ename = JAMES Job = CLERK Sal = 950
Ename = FORD Job = ANALYST Sal = 3000
Ename = MILLER Job = CLERK Sal = 1300
SQL> exec ref_cursor('STUDENT')
Name = saketh
Name = srinu
Name = satish
Name = sudha
CURSOR EXPRESSIONS
You can use cursor expressions in explicit cursors.
You can use cursor expressions in dynamic SQL.
You can use cursor expressions in REF cursor declarations and variables.
You can not use cursor expressions in implicit cursors.
Oracle opens the nested cursor defined by a cursor expression implicitly as soon as it fetches the data containing the cursor expression from the parent or outer cursor.
Nested cursor closes if you close explicitly.
Nested cursor closes whenever the outer or parent cursor is executed again or closed or canceled.
Nested cursor closes whenever an exception is raised while fetching data from a parent cursor.
Cursor expressions can not be used when declaring a view.
Cursor expressions can be used as an argument to table function.
You can not perform bind and execute operations on cursor expressions when using the cursor expressions in dynamic SQL.
USING NESTED CURSORS OR CURSOR EXPRESSIONS
Ex:
DECLARE
cursor c is select ename,cursor(select dname from dept d where e.empno = d.deptno) from emp e;
type t is ref cursor;
c1 t;
c2 t;
v1 emp.ename%type;
v2 dept.dname%type;
BEGIN
open c;
loop
fetch c1 into v1;
exit when c1%notfound;
fetch c2 into v2;
exit when c2%notfound;
dbms_output.put_line('Ename = ' || v1 || ' Dname = ' || v2);
end loop;
end loop;
close c;
END;
CURSOR CLAUSES
Return
For update
Where current of
Bulk collect
RETURN
Cursor c return dept%rowtype is select *from dept;
Or
Cursor c1 is select *from dept;
Cursor c return c1%rowtype is select *from dept;
Or
Type t is record(deptno dept.deptno%type, dname dept.dname%type);
Cursor c return t is select deptno, dname from dept;
FOR UPDATE AND WHERE CURRENT OF
Normally, a select operation will not take any locks on the rows being accessed. This will allow other sessions connected to the database to change the data being selected. The result set is still consistent. At open time, when the active set is determined, oracle takes a snapshot of the table. Any changes that have been committed prior to this point are reflected in the active set. Any changes made after this point, even if they are committed, are not reflected unless the cursor is reopened, which will evaluate the active set again.
However, if the FOR UPDATE caluse is pesent, exclusive row locks are taken on the rows in the active set before the open returns. These locks prevent other sessions from changing the rows in the active set until the transaction is committed or rolled back. If another session already has locks on the rows in the active set, then SELECT … FOR UPDATE operation will wait for these locks to be released by the other session. There is no time-out for this waiting period. The SELECT…FOR UPDATE will hang until the other session releases the lock. To handle this situation, the NOWAIT clause is available.
Syntax:
Select …from … for update of column_name [wait n];
If the cursor is declared with the FOR UPDATE clause, the WHERE CURRENT OF clause can be used in an update or delete statement.
Syntax:
Where current of cursor;
Ex:
DECLARE
cursor c is select * from dept for update of dname;
BEGIN
for v in c loop
update dept set dname = 'aa' where current of c;
commit;
end loop;
END;
BULK COLLECT
This is used for array fetches
With this you can retrieve multiple rows of data with a single roundtrip.
This reduces the number of context switches between the pl/sql and sql engines.
Reduces the overhead of retrieving data.
You can use bulk collect in both dynamic and static sql.
You can use bulk collect in select, fetch into and returning into clauses.
SQL engine automatically initializes and extends the collections you reference in the bulk collect clause.
Bulk collect operation empties the collection referenced in the into clause before executing the query.
You can use the limit clause of bulk collect to restrict the no of rows retrieved.
You can fetch into multible collections with one column each.
Using the returning clause we can return data to the another collection.
BULK COLLECT IN FETCH
Ex:
DECLARE
Type t is table of dept%rowtype;
nt t;
Cursor c is select *from dept;
BEGIN
Open c;
Fetch c bulk collect into nt;
Close c;
For i in nt.first..nt.last loop
dbms_output.put_line('Dname = ' || nt(i).dname || ' Loc = ' || nt(i).loc);
end loop;
END;
Output:
Dname = ACCOUNTING Loc = NEW YORK
Dname = RESEARCH Loc = DALLAS
Dname = SALES Loc = CHICAGO
Dname = OPERATIONS Loc = BOSTON
BULK COLLECT IN SELECT
Ex:
DECLARE
Type t is table of dept%rowtype;
Nt t;
BEGIN
Select * bulk collect into nt from dept;
for i in nt.first..nt.last loop
dbms_output.put_line('Dname = ' || nt(i).dname || ' Loc = ' || nt(i).loc);
end loop;
END;
Output:
Dname = ACCOUNTING Loc = NEW YORK
Dname = RESEARCH Loc = DALLAS
Dname = SALES Loc = CHICAGO
Dname = OPERATIONS Loc = BOSTON
LIMIT IN BULK COLLECT
Ex:
DECLARE
Type t is table of dept%rowtype;
nt t;
Cursor c is select *from dept;
BEGIN
Open c;
Fetch c bulk collect into nt;
Close c;
For i in nt.first..nt.last loop
dbms_output.put_line('Dname = ' || nt(i).dname || ' Loc = ' || nt(i).loc);
end loop;
END;
Output:
Dname = ACCOUNTING Loc = NEW YORK
Dname = RESEARCH Loc = DALLAS
MULTIPLE FETCHES IN INTO CLAUSE
Ex1:
DECLARE
Type t is table of dept.dname%type;
nt t;
Type t1 is table of dept.loc%type;
nt1 t;
Cursor c is select dname,loc from dept;
BEGIN
Open c;
Fetch c bulk collect into nt,nt1;
Close c;
For i in nt.first..nt.last loop
dbms_output.put_line('Dname = ' || nt(i));
end loop;
For i in nt1.first..nt1.last loop
dbms_output.put_line('Loc = ' || nt1(i));
end loop;
END;
Output:
Dname = ACCOUNTING
Dname = RESEARCH
Dname = SALES
Dname = OPERATIONS
Loc = NEW YORK
Loc = DALLAS
Loc = CHICAGO
Loc = BOSTON
Ex2:
DECLARE
type t is table of dept.dname%type;
type t1 is table of dept.loc%type;
nt t;
nt1 t1;
BEGIN
Select dname,loc bulk collect into nt,nt1 from dept;
for i in nt.first..nt.last loop
dbms_output.put_line('Dname = ' || nt(i));
end loop;
for i in nt1.first..nt1.last loop
dbms_output.put_line('Loc = ' || nt1(i));
end loop;
END;
Output:
Dname = ACCOUNTING
Dname = RESEARCH
Dname = SALES
Dname = OPERATIONS
Loc = NEW YORK
Loc = DALLAS
Loc = CHICAGO
Loc = BOSTON
RETURNING CLAUSE IN BULK COLLECT
declare
type t is table of number(2);
nt t := t(1,2,3,4);
type t1 is table of varchar(2);
nt1 t1;
type t2 is table of student%rowtype;
nt2 t2;
begin
select name bulk collect into nt1 from student;
forall v in nt1.first..nt1.last
update student set no = nt(v) where name = nt1(v) returning no,name,marks bulk collect into nt2;
for v in nt2.first..nt2.last loop
dbms_output.put_line('Marks = ' || nt2(v));
end loop;
end;
POINTS TO REMEMBER
Cursor name can be up to 30 characters in length.
Cursors declared in anonymous blocks or subprograms closes automatically when that block terminates execution.
%bulk_rowcount and %bulk_exceptions can be used only with forall construct.
Cursor declarations may have expressions with column aliases.
These expressions are called virtual columns or calculated columns.
SQL IN PL/SQL
The only statements allowed directly in pl/sql are DML and TCL.
BINDING
Binding a variable is the process of identifying the storage location associated with an identifier in the program.
Types of binding
Early binding
Late binding
Binding during the compiled phase is early binding.
Binding during the runtime phase is late binding.
In early binding compile phase will take longer because of binding work but the execution
is faster.
In late binding it will shorten the compile phase but lengthens the execution time.
Pl/sql by default uses early binding.
Binding also involves checking the database for permissions to access the object
Referenced.
DYNAMIC SQL
If you use DDL in pl/sql it validates the permissions and existence if requires during compile time which makes invalid.
We can avoid this by using Dynamic SQL.
Dynamic SQL allows you to create a SQL statement dynamically at runtime.
Two techniques are available for Dynamic SQL.
Native Dynamic SQL
DBMS_SQL package
USING NATIVE DYNAMIC SQL
Using execute immediate
Begin
Execute immediate ‘create table student(no number(2),name varchar(10))’;
or
Execute immediate (‘create table student(no number(2),name varchar(10))’);
End;
Using execute immediate with pl/sql variables
declare
v varchar(100);
begin
v := 'create table student(no number(2),name varchar(10))';
execute immediate v;
end;
Using execute immediate with bind variables and using clause
declare
v varchar(100);
begin
v := 'insert into student values(:v1,:v2,:v3)';
execute immediate v using 6,'f',600;
end;
Executing queries with open for and using clause
create or replace procedure p(smarks in number) is
s varchar(100) := 'select *from student where marks > :m';
type t is ref cursor;
c t;
v student%rowtype;
begin
open c for s using smarks;
loop
fetch c into v;
exit when c%notfound;
dbms_output.put_line('Student Marks = ' || v.marks);
end loop;
close c;
end;
Queries with execute immediate
declare
d_name dept.dname%type;
lc dept.loc%type;
v varchar(100);
begin
v := 'select dname from dept where deptno = 10';
execute immediate v into d_name;
dbms_output.put_line('Dname = '|| d_name);
v := 'select loc from dept where dname = :dn';
execute immediate v into lc using d_name;
dbms_output.put_line('Loc = ' || lc);
end;
Bind variables
Declare
V number := 500;
Begin
Update student set marks = v where;
-- here v is bind variable
End;
Variable Names
Declare
Marks number(3) := 100;
Begin
Delete student where marks = marks;
-- this will delete all the rows in the student table
End;
This can be avoided by using the labeled blocks.
<<my_block>>
Declare
Marks number(3) := 100;
Begin
Delete student where marks = my_block.marks;
-- delete rows which has a marks of 100
End;
Getting data into pl/sql variables
Declare
V1 number;
V2 varchar(2);
Begin
Select no,name into v1,v2 from student where marks = 100;
End;
DML and Records
create or replace procedure p(srow in student%rowtype) is
begin
insert into student values srow;
end p;
declare
s student%rowtype;
begin
s.no := 11;
s.name := 'aa';
s.marks := 100;
p(s);
end;
Record based inserts
declare
srow student%rowtype;
begin
srow.no := 7;
srow.name := 'cc';
srow.marks := 500;
insert into student values srow;
end;
Record based updates
declare
srow student%rowtype;
begin
srow.no := 6;
srow.name := 'cc';
srow.marks := 500;
update student set row=srow where no = srow.no;
end;
Using records with returning clause
declare
srow student%rowtype;
sreturn student%rowtype;
begin
srow.no := 8;
srow.name := 'dd';
srow.marks := 500;
insert into student values srow returning no,name,marks into sreturn;
dbms_output.put_line('No = ' || sreturn.no);
dbms_output.put_line('No = ' || sreturn.name);
dbms_output.put_line('No = ' || sreturn.marks);
end;
Forall with non-sequential arrays
declare
type t is table of student.no%type index by binary_integer;
ibt t;
begin
ibt(1) := 1;
ibt(10) := 2;
forall i in ibt.first..ibt.last
update student set marks = 900 where no = ibt(i);
end;
The above program will give error like ‘element at index [2] does not exists.
Usage of indices of to avoid the above error
declare
type t is table of student.no%type index by binary_integer;
ibt t;
type t1 is table of boolean index by binary_integer;
ibt1 t1;
begin
ibt(1) := 1;
ibt(10) := 2;
ibt(100) := 3;
ibt1(1) := true;
ibt1(10) := true;
ibt1(100) := true;
forall i in indices of ibt1
update student set marks = 900 where no = ibt(i);
end;
declare
type t is table of student.no%type index by binary_integer;
ibt t;
type t1 is table of pls_integer index by binary_integer;
ibt1 t1;
begin
ibt(1) := 1;
ibt(10) := 2;
ibt(100) := 3;
ibt1(11) := 1;
ibt1(15) := 10;
ibt1(18) := 100;
forall i in values of ibt1
update student set marks = 567 where no = ibt(i);
end;
Bulk Binds
Passing the entire pl/sql table to the SQL engine in one step is known as bulk bind.
Bulk binds are done using the forall statement.
If there is an error processing one of the rows in bulk DML operation, only that row is rolled back.
Returning clause
This will be used only with DML statements to return data into pl/sql variables.
This will be useful in situations like , when performing insert or update or delete if you want to know the data of the table which has been effected by the DML.
With out going for another SELECT using RETURNING clause we will get the data which will avoid a call to RDBMS kernel.
COLLECTIONS
Collections are also composite types, in that they allow you to treat several variables as a unit. A collection combines variables of the same type.
TYPES
Varrays
Nested tables
Index - by tables
VARRAYS
A varray is datatype very similar to an array. A varray has a fixed limit on its size, specified as part of the declaration. Elements are inserted into varray starting at index 1, up to maximum lenth declared in the varray type. The maximum size of the varray is 2 giga bytes.
Syntax:
Type <type_name> is varray | varying array (<limit>) of <element_type>;
Ex1:
DECLARE
type t is varray(10) of varchar(2);
va t := t('a','b','c','d');
flag boolean;
BEGIN
dbms_output.put_line('Limit = ' || va.limit);
dbms_output.put_line('Count = ' || va.count);
dbms_output.put_line('First Index = ' || va.first);
dbms_output.put_line('Last Index = ' || va.last);
dbms_output.put_line('Next Index = ' || va.next(2));
dbms_output.put_line('Previous Index = ' || va.prior(3));
dbms_output.put_line('VARRAY ELEMENTS');
for i in va.first..va.last loop
dbms_output.put_line('va[' || i || '] = ' || va(i));
end loop;
flag := va.exists(3);
if flag = true then
dbms_output.put_line('Index 3 exists with an element ' || va(3));
else
dbms_output.put_line('Index 3 does not exists');
end if;
va.extend;
dbms_output.put_line('After extend of one index, Count = ' || va.count);
flag := va.exists(5);
if flag = true then
dbms_output.put_line('Index 5 exists with an element ' || va(5));
else
dbms_output.put_line('Index 5 does not exists');
end if;
flag := va.exists(6);
if flag = true then
dbms_output.put_line('Index 6 exists with an element ' || va(6));
else
dbms_output.put_line('Index 6 does not exists');
end if;
va.extend(2);
dbms_output.put_line('After extend of two indexes, Count = ' || va.count);
dbms_output.put_line('VARRAY ELEMENTS');
for i in va.first..va.last loop
dbms_output.put_line('va[' || i || '] = ' || va(i));
end loop;
va(5) := 'e';
va(6) := 'f';
va(7) := 'g';
dbms_output.put_line('AFTER ASSINGNING VALUES TO EXTENDED ELEMENTS,
VARRAY ELEMENTS');
for i in va.first..va.last loop
dbms_output.put_line('va[' || i || '] = ' || va(i));
end loop;
va.extend(3,2);
dbms_output.put_line('After extend of three indexes, Count = ' || va.count);
dbms_output.put_line('VARRAY ELEMENTS');
for i in va.first..va.last loop
dbms_output.put_line('va[' || i || '] = ' || va(i));
end loop;
va.trim;
dbms_output.put_line('After trim of one index, Count = ' || va.count);
va.trim(3);
dbms_output.put_line('After trim of three indexs, Count = ' || va.count);
dbms_output.put_line('AFTER TRIM, VARRAY ELEMENTS');
for i in va.first..va.last loop
dbms_output.put_line('va[' || i || '] = ' || va(i));
end loop;
va.delete;
dbms_output.put_line('After delete of entire varray, Count = ' || va.count);
END;
Output:
Limit = 10
Count = 4
First Index = 1
Last Index = 4
Next Index = 3
Previous Index = 2
VARRAY ELEMENTS
va[1] = a
va[2] = b
va[3] = c
va[4] = d
Index 3 exists with an element c
After extend of one index, Count = 5
Index 5 exists with an element
Index 6 does not exists
After extend of two indexes, Count = 7
VARRAY ELEMENTS
va[1] = a
va[2] = b
va[3] = c
va[4] = d
va[5] =
va[6] =
va[7] =
AFTER ASSINGNING VALUES TO EXTENDED ELEMENTS, VARRAY ELEMENTS
va[1] = a
va[2] = b
va[3] = c
va[4] = d
va[5] = e
va[6] = f
va[7] = g
After extend of three indexes, Count = 10
VARRAY ELEMENTS
va[1] = a
va[2] = b
va[3] = c
va[4] = d
va[5] = e
va[6] = f
va[7] = g
va[8] = b
va[9] = b
va[10] = b
After trim of one index, Count = 9
After trim of three indexs, Count = 6
AFTER TRIM, VARRAY ELEMENTS
va[1] = a
va[2] = b
va[3] = c
va[4] = d
va[5] = e
va[6] = f
After delete of entire varray, Count = 0
Ex2:
DECLARE
type t is varray(4) of student%rowtype;
va t := t(null,null,null,null);
BEGIN
for i in 1..va.count loop
select * into va(i) from student where sno = i;
dbms_output.put_line('Sno = ' || va(i).sno || ' Sname = ' || va(i).sname);
end loop;
END;
Output:
Sno = 1 Sname = saketh
Sno = 2 Sname = srinu
Sno = 3 Sname = divya
Sno = 4 Sname = manogni
Ex3:
DECLARE
type t is varray(4) of student.smarks%type;
va t := t(null,null,null,null);
BEGIN
for i in 1..va.count loop
select smarks into va(i) from student where sno = i;
dbms_output.put_line('Smarks = ' || va(i));
end loop;
END;
Output:
Smarks = 100
Smarks = 200
Smarks = 300
Smarks = 400
Ex4:
DECLARE
type r is record(c1 student.sname%type,c2 student.smarks%type);
type t is varray(4) of r;
va t := t(null,null,null,null);
BEGIN
for i in 1..va.count loop
select sname,smarks into va(i) from student where sno = i;
dbms_output.put_line('Sname = ' || va(i).c1 || ' Smarks = ' || va(i).c2);
end loop;
END;
Output:
Sname = saketh Smarks = 100
Sname = srinu Smarks = 200
Sname = divya Smarks = 300
Sname = manogni Smarks = 400
Ex5:
DECLARE
type t is varray(1) of addr;
va t := t(null);
cursor c is select * from employ;
i number := 1;
BEGIN
for v in c loop
select address into va(i) from employ where ename = v.ename;
dbms_output.put_line('Hno = ' || va(i).hno || ' City = ' || va(i).city);
end loop;
END;
Output:
Hno = 11 City = hyd
Hno = 22 City = bang
Hno = 33 City = kochi
Ex6:
DECLARE
type t is varray(5) of varchar(2);
va1 t;
va2 t := t();
BEGIN
if va1 is null then
dbms_output.put_line('va1 is null');
else
dbms_output.put_line('va1 is not null');
end if;
if va2 is null then
dbms_output.put_line('va2 is null');
else
dbms_output.put_line('va2 is not null');
end if;
END;
Output:
va1 is null
va2 is not null
NESTED TABLES
A nested table is thought of a database table which has no limit on its size. Elements are inserted into nested table starting at index 1. The maximum size of the varray is 2 giga bytes.
Syntax:
Type <type_name> is table of <table_type>;
Ex1:
DECLARE
type t is table of varchar(2);
nt t := t('a','b','c','d');
flag boolean;
BEGIN
if nt.limit is null then
dbms_output.put_line('No limit to Nested Tables');
else
dbms_output.put_line('Limit = ' || nt.limit);
end if;
dbms_output.put_line('Count = ' || nt.count);
dbms_output.put_line('First Index = ' || nt.first);
dbms_output.put_line('Last Index = ' || nt.last);
dbms_output.put_line('Next Index = ' || nt.next(2));
dbms_output.put_line('Previous Index = ' || nt.prior(3));
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
flag := nt.exists(3);
if flag = true then
dbms_output.put_line('Index 3 exists with an element ' || nt(3));
else
dbms_output.put_line('Index 3 does not exists');
end if;
nt.extend;
dbms_output.put_line('After extend of one index, Count = ' || nt.count);
flag := nt.exists(5);
if flag = true then
dbms_output.put_line('Index 5 exists with an element ' || nt(5));
else
dbms_output.put_line('Index 5 does not exists');
end if;
flag := nt.exists(6);
if flag = true then
dbms_output.put_line('Index 6 exists with an element ' || nt(6));
else
dbms_output.put_line('Index 6 does not exists');
end if;
nt.extend(2);
dbms_output.put_line('After extend of two indexes, Count = ' || nt.count);
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
nt(5) := 'e';
nt(6) := 'f';
nt(7) := 'g';
dbms_output.put_line('AFTER ASSINGNING VALUES TO EXTENDED ELEMENTS, NESTED TABLE
ELEMENTS');
for i in 1..nt.count loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
nt.extend(5,2);
dbms_output.put_line('After extend of five indexes, Count = ' || nt.count);
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
nt.trim;
dbms_output.put_line('After trim of one index, Count = ' || nt.count);
nt.trim(3);
dbms_output.put_line('After trim of three indexs, Count = ' || nt.count);
dbms_output.put_line('AFTER TRIM, NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
nt.delete(1);
dbms_output.put_line('After delete of first index, Count = ' || nt.count);
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 2..nt.count+1 loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
nt.delete(4);
dbms_output.put_line('After delete of fourth index, Count = ' || nt.count);
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 2..3 loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
for i in 5..nt.count+2 loop
dbms_output.put_line('nt[' || i || '] = ' || nt(i));
end loop;
nt.delete;
dbms_output.put_line('After delete of entire nested table, Count = ' || nt.count);
END;
Output:
No limit to Nested Tables
Count = 4
First Index = 1
Last Index = 4
Next Index = 3
Previous Index = 2
NESTED TABLE ELEMENTS
nt[1] = a
nt[2] = b
nt[3] = c
nt[4] = d
Index 3 exists with an element c
After extend of one index, Count = 5
Index 5 exists with an element
Index 6 does not exists
After extend of two indexes, Count = 7
NESTED TABLE ELEMENTS
nt[1] = a
nt[2] = b
nt[3] = c
nt[4] = d
nt[5] =
nt[6] =
nt[7] =
AFTER ASSINGNING VALUES TO EXTENDED ELEMENTS, NESTED TABLE ELEMENTS
nt[1] = a
nt[2] = b
nt[3] = c
nt[4] = d
nt[5] = e
nt[6] = f
nt[7] = g
After extend of five indexes, Count = 12
NESTED TABLE ELEMENTS
nt[1] = a
nt[2] = b
nt[3] = c
nt[4] = d
nt[5] = e
nt[6] = f
nt[7] = g
nt[8] = b
nt[9] = b
nt[10] = b
nt[11] = b
nt[12] = b
After trim of one index, Count = 11
After trim of three indexs, Count = 8
AFTER TRIM, NESTED TABLE ELEMENTS
nt[1] = a
nt[2] = b
nt[3] = c
nt[4] = d
nt[5] = e
nt[6] = f
nt[7] = g
nt[8] = b
After delete of first index, Count = 7
NESTED TABLE ELEMENTS
nt[2] = b
nt[3] = c
nt[4] = d
nt[5] = e
nt[6] = f
nt[7] = g
nt[8] = b
After delete of fourth index, Count = 6
NESTED TABLE ELEMENTS
nt[2] = b
nt[3] = c
nt[5] = e
nt[6] = f
nt[7] = g
nt[8] = b
After delete of entire nested table, Count = 0
Ex2:
DECLARE
type t is table of student%rowtype;
nt t := t(null,null,null,null);
BEGIN
for i in 1..nt.count loop
select * into nt(i) from student where sno = i;
dbms_output.put_line('Sno = ' || nt(i).sno || ' Sname = ' || nt(i).sname);
end loop;
END;
Output:
Sno = 1 Sname = saketh
Sno = 2 Sname = srinu
Sno = 3 Sname = divya
Sno = 4 Sname = manogni
Ex3:
DECLARE
type t is table of student.smarks%type;
nt t := t(null,null,null,null);
BEGIN
for i in 1..nt.count loop
select smarks into nt(i) from student where sno = i;
dbms_output.put_line('Smarks = ' || nt(i));
end loop;
END;
Output:
Smarks = 100
Smarks = 200
Smarks = 300
Smarks = 400
Ex4:
DECLARE
type r is record(c1 student.sname%type,c2 student.smarks%type);
type t is table of r;
nt t := t(null,null,null,null);
BEGIN
for i in 1..nt.count loop
select sname,smarks into nt(i) from student where sno = i;
dbms_output.put_line('Sname = ' || nt(i).c1 || ' Smarks = ' || nt(i).c2);
end loop;
END;
Output:
Sname = saketh Smarks = 100
Sname = srinu Smarks = 200
Sname = divya Smarks = 300
Sname = manogni Smarks = 400
Ex5:
DECLARE
type t is table of addr;
nt t := t(null);
cursor c is select * from employ;
i number := 1;
BEGIN
for v in c loop
select address into nt(i) from employ where ename = v.ename;
dbms_output.put_line('Hno = ' || nt(i).hno || ' City = ' || nt(i).city);
end loop;
END;
Output:
Hno = 11 City = hyd
Hno = 22 City = bang
Hno = 33 City = kochi
Ex6:
DECLARE
type t is varray(5) of varchar(2);
nt1 t;
nt2 t := t();
BEGIN
if nt1 is null then
dbms_output.put_line('nt1 is null');
else
dbms_output.put_line('nt1 is not null');
end if;
if nt2 is null then
dbms_output.put_line('nt2 is null');
else
dbms_output.put_line('nt2 is not null');
end if;
END;
Output:
nt1 is null
nt2 is not null
INDEX-BY TABLES
An index-by table has no limit on its size. Elements are inserted into index-by table whose index may start non-sequentially including negative integers.
Syntax:
Type <type_name> is table of <table_type> index by binary_integer;
Ex:
DECLARE
type t is table of varchar(2) index by binary_integer;
ibt t;
flag boolean;
BEGIN
ibt(1) := 'a';
ibt(-20) := 'b';
ibt(30) := 'c';
ibt(100) := 'd';
if ibt.limit is null then
dbms_output.put_line('No limit to Index by Tables');
else
dbms_output.put_line('Limit = ' || ibt.limit);
end if;
dbms_output.put_line('Count = ' || ibt.count);
dbms_output.put_line('First Index = ' || ibt.first);
dbms_output.put_line('Last Index = ' || ibt.last);
dbms_output.put_line('Next Index = ' || ibt.next(2));
dbms_output.put_line('Previous Index = ' || ibt.prior(3));
dbms_output.put_line('INDEX BY TABLE ELEMENTS');
dbms_output.put_line('ibt[-20] = ' || ibt(-20));
dbms_output.put_line('ibt[1] = ' || ibt(1));
dbms_output.put_line('ibt[30] = ' || ibt(30));
dbms_output.put_line('ibt[100] = ' || ibt(100));
flag := ibt.exists(30);
if flag = true then
dbms_output.put_line('Index 30 exists with an element ' || ibt(30));
else
dbms_output.put_line('Index 30 does not exists');
end if;
flag := ibt.exists(50);
if flag = true then
dbms_output.put_line('Index 50 exists with an element ' || ibt(30));
else
dbms_output.put_line('Index 50 does not exists');
end if;
ibt.delete(1);
dbms_output.put_line('After delete of first index, Count = ' || ibt.count);
ibt.delete(30);
dbms_output.put_line('After delete of index thirty, Count = ' || ibt.count);
dbms_output.put_line('INDEX BY TABLE ELEMENTS');
dbms_output.put_line('ibt[-20] = ' || ibt(-20));
dbms_output.put_line('ibt[100] = ' || ibt(100));
ibt.delete;
dbms_output.put_line('After delete of entire index-by table, Count = ' || ibt.count);
END;
Output:
No limit to Index by Tables
Count = 4
First Index = -20
Last Index = 100
Next Index = 30
Previous Index = 1
INDEX BY TABLE ELEMENTS
ibt[-20] = b
ibt[1] = a
ibt[30] = c
ibt[100] = d
Index 30 exists with an element c
Index 50 does not exists
After delete of first index, Count = 3
After delete of index thirty, Count = 2
INDEX BY TABLE ELEMENTS
ibt[-20] = b
ibt[100] = d
After delete of entire index-by table, Count = 0
DIFFERENCES AMONG COLLECTIONS
Varrays has limit, nested tables and index-by tables has no limit.
Varrays and nested tables must be initialized before assignment of elements, in index-by tables we can directly assign elements.
Varrays and nested tables stored in database, but index-by tables can not.
Nested tables and index-by tables are PL/SQL tables, but varrays can not.
Keys must be positive in case of nested tables and varrays, in case of index-by tables keys can be positive or negative.
Referencing nonexistent elements raises SUBSCRIPT_BEYOND_COUNT in both nested tables and varrays, but in case of index-by tables NO_DATA_FOUND raises.
Keys are sequential in both nested tables and varrays, non-sequential in index-by tables.
Individual indexes can be deleted in both nested tables and index-by tables, but in varrays can not.
Individual indexes can be trimmed in both nested tables and varrays, but in index-by tables can not.
Individual indexes can be extended in both nested tables and varrays, but in index-by tables can not.
MULTILEVEL COLLECTIONS
Collections of more than one dimension which is a collection of collections, known as multilevel collections.
Syntax:
Type <type_name1> is table of <table_type> index by binary_integer;
Type <type_name2> is varray(<limit>) | table | of <type_name1> | index by
binary_integer;
Ex1:
DECLARE
type t1 is table of varchar(2) index by binary_integer;
type t2 is varray(5) of t1;
va t2 := t2();
c number := 97;
flag boolean;
BEGIN
va.extend(4);
dbms_output.put_line('Count = ' || va.count);
dbms_output.put_line('Limit = ' || va.limit);
for i in 1..va.count loop
for j in 1..va.count loop
va(i)(j) := chr(c);
c := c + 1;
end loop;
end loop;
dbms_output.put_line('VARRAY ELEMENTS');
for i in 1..va.count loop
for j in 1..va.count loop
dbms_output.put_line('va[' || i || '][' || j || '] = ' || va(i)(j));
end loop;
end loop;
dbms_output.put_line('First index = ' || va.first);
dbms_output.put_line('Last index = ' || va.last);
dbms_output.put_line('Next index = ' || va.next(2));
dbms_output.put_line('Previous index = ' || va.prior(3));
flag := va.exists(2);
if flag = true then
dbms_output.put_line('Index 2 exists');
else
dbms_output.put_line('Index 2 exists');
end if;
va.extend;
va(1)(5) := 'q';
va(2)(5) := 'r';
va(3)(5) := 's';
va(4)(5) := 't';
va(5)(1) := 'u';
va(5)(2) := 'v';
va(5)(3) := 'w';
va(5)(4) := 'x';
va(5)(5) := 'y';
dbms_output.put_line('After extend of one index, Count = ' || va.count);
dbms_output.put_line('VARRAY ELEMENTS');
for i in 1..va.count loop
for j in 1..va.count loop
dbms_output.put_line('va[' || i || '][' || j || '] = ' || va(i)(j));
end loop;
end loop;
va.trim;
dbms_output.put_line('After trim of one index, Count = ' || va.count);
va.trim(2);
dbms_output.put_line('After trim of two indexes, Count = ' || va.count);
dbms_output.put_line('VARRAY ELEMENTS');
for i in 1..va.count loop
for j in 1..va.count loop
dbms_output.put_line('va[' || i || '][' || j || '] = ' || va(i)(j));
end loop;
end loop;
va.delete;
dbms_output.put_line('After delete of entire varray, Count = ' || va.count);
END;
Output:
Count = 4
Limit = 5
VARRAY ELEMENTS
va[1][1] = a
va[1][2] = b
va[1][3] = c
va[1][4] = d
va[2][1] = e
va[2][2] = f
va[2][3] = g
va[2][4] = h
va[3][1] = i
va[3][2] = j
va[3][3] = k
va[3][4] = l
va[4][1] = m
va[4][2] = n
va[4][3] = o
va[4][4] = p
First index = 1
Last index = 4
Next index = 3
Previous index = 2
Index 2 exists
After extend of one index, Count = 5
VARRAY ELEMENTS
va[1][1] = a
va[1][2] = b
va[1][3] = c
va[1][4] = d
va[1][5] = q
va[2][1] = e
va[2][2] = f
va[2][3] = g
va[2][4] = h
va[2][5] = r
va[3][1] = i
va[3][2] = j
va[3][3] = k
va[3][4] = l
va[3][5] = s
va[4][1] = m
va[4][2] = n
va[4][3] = o
va[4][4] = p
va[4][5] = t
va[5][1] = u
va[5][2] = v
va[5][3] = w
va[5][4] = x
va[5][5] = y
After trim of one index, Count = 4
After trim of two indexes, Count = 2
VARRAY ELEMENTS
va[1][1] = a
va[1][2] = b
va[2][1] = e
va[2][2] = f
After delete of entire varray, Count = 0
Ex2:
DECLARE
type t1 is table of varchar(2) index by binary_integer;
type t2 is table of t1;
nt t2 := t2();
c number := 65;
v number := 1;
flag boolean;
BEGIN
nt.extend(4);
dbms_output.put_line('Count = ' || nt.count);
if nt.limit is null then
dbms_output.put_line('No limit to Nested Tables');
else
dbms_output.put_line('Limit = ' || nt.limit);
end if;
for i in 1..nt.count loop
for j in 1..nt.count loop
nt(i)(j) := chr(c);
c := c + 1;
if c = 91 then
c := 97;
end if;
end loop;
end loop;
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
for j in 1..nt.count loop
dbms_output.put_line('nt[' || i || '][' || j || '] = ' || nt(i)(j));
end loop;
end loop;
dbms_output.put_line('First index = ' || nt.first);
dbms_output.put_line('Last index = ' || nt.last);
dbms_output.put_line('Next index = ' || nt.next(2));
dbms_output.put_line('Previous index = ' || nt.prior(3));
flag := nt.exists(2);
if flag = true then
dbms_output.put_line('Index 2 exists');
else
dbms_output.put_line('Index 2 exists');
end if;
nt.extend(2);
nt(1)(5) := 'Q';
nt(1)(6) := 'R';
nt(2)(5) := 'S';
nt(2)(6) := 'T';
nt(3)(5) := 'U';
nt(3)(6) := 'V';
nt(4)(5) := 'W';
nt(4)(6) := 'X';
nt(5)(1) := 'Y';
nt(5)(2) := 'Z';
nt(5)(3) := 'a';
nt(5)(4) := 'b';
nt(5)(5) := 'c';
nt(5)(6) := 'd';
nt(6)(1) := 'e';
nt(6)(2) := 'f';
nt(6)(3) := 'g';
nt(6)(4) := 'h';
nt(6)(5) := 'i';
nt(6)(6) := 'j';
dbms_output.put_line('After extend of one index, Count = ' || nt.count);
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
for j in 1..nt.count loop
dbms_output.put_line('nt[' || i || '][' || j || '] = ' || nt(i)(j));
end loop;
end loop;
nt.trim;
dbms_output.put_line('After trim of one indexe, Count = ' || nt.count);
nt.trim(2);
dbms_output.put_line('After trim of two indexes, Count = ' || nt.count);
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
for j in 1..nt.count loop
dbms_output.put_line('nt[' || i || '][' || j || '] = ' || nt(i)(j));
end loop;
end loop;
nt.delete(2);
dbms_output.put_line('After delete of second index, Count = ' || nt.count);
dbms_output.put_line('NESTED TABLE ELEMENTS');
loop
exit when v = 4;
for j in 1..nt.count+1 loop
dbms_output.put_line('nt[' || v || '][' || j || '] = ' || nt(v)(j));
end loop;
v := v + 1;
if v= 2 then
v := 3;
end if;
end loop;
nt.delete;
dbms_output.put_line('After delete of entire nested table, Count = ' || nt.count);
END;
Output:
Count = 4
No limit to Nested Tables
NESTED TABLE ELEMENTS
nt[1][1] = A
nt[1][2] = B
nt[1][3] = C
nt[1][4] = D
nt[2][1] = E
nt[2][2] = F
nt[2][3] = G
nt[2][4] = H
nt[3][1] = I
nt[3][2] = J
nt[3][3] = K
nt[3][4] = L
nt[4][1] = M
nt[4][2] = N
nt[4][3] = O
nt[4][4] = P
First index = 1
Last index = 4
Next index = 3
Previous index = 2
Index 2 exists
After extend of one index, Count = 6
NESTED TABLE ELEMENTS
nt[1][1] = A
nt[1][2] = B
nt[1][3] = C
nt[1][4] = D
nt[1][5] = Q
nt[1][6] = R
nt[2][1] = E
nt[2][2] = F
nt[2][3] = G
nt[2][4] = H
nt[2][5] = S
nt[2][6] = T
nt[3][1] = I
nt[3][2] = J
nt[3][3] = K
nt[3][4] = L
nt[3][5] = U
nt[3][6] = V
nt[4][1] = M
nt[4][2] = N
nt[4][3] = O
nt[4][4] = P
nt[4][5] = W
nt[4][6] = X
nt[5][1] = Y
nt[5][2] = Z
nt[5][3] = a
nt[5][4] = b
nt[5][5] = c
nt[5][6] = d
nt[6][1] = e
nt[6][2] = f
nt[6][3] = g
nt[6][4] = h
nt[6][5] = i
nt[6][6] = j
After trim of one indexe, Count = 5
After trim of two indexes, Count = 3
NESTED TABLE ELEMENTS
nt[1][1] = A
nt[1][2] = B
nt[1][3] = C
nt[2][1] = E
nt[2][2] = F
nt[2][3] = G
nt[3][1] = I
nt[3][2] = J
nt[3][3] = K
After delete of second index, Count = 2
NESTED TABLE ELEMENTS
nt[1][1] = A
nt[1][2] = B
nt[1][3] = C
nt[3][1] = I
nt[3][2] = J
nt[3][3] = K
After delete of entire nested table, Count = 0
Ex3:
DECLARE
type t1 is table of varchar(2) index by binary_integer;
type t2 is table of t1 index by binary_integer;
ibt t2;
flag boolean;
BEGIN
dbms_output.put_line('Count = ' || ibt.count);
if ibt.limit is null then
dbms_output.put_line('No limit to Index-by Tables');
else
dbms_output.put_line('Limit = ' || ibt.limit);
end if;
ibt(1)(1) := 'a';
ibt(4)(5) := 'b';
ibt(5)(1) := 'c';
ibt(6)(2) := 'd';
ibt(8)(3) := 'e';
ibt(3)(4) := 'f';
dbms_output.put_line('INDEX-BY TABLE ELEMENTS');
dbms_output.put_line('ibt([1][1] = ' || ibt(1)(1));
dbms_output.put_line('ibt([4][5] = ' || ibt(4)(5));
dbms_output.put_line('ibt([5][1] = ' || ibt(5)(1));
dbms_output.put_line('ibt([6][2] = ' || ibt(6)(2));
dbms_output.put_line('ibt([8][3] = ' || ibt(8)(3));
dbms_output.put_line('ibt([3][4] = ' || ibt(3)(4));
dbms_output.put_line('First Index = ' || ibt.first);
dbms_output.put_line('Last Index = ' || ibt.last);
dbms_output.put_line('Next Index = ' || ibt.next(3));
dbms_output.put_line('Prior Index = ' || ibt.prior(8));
ibt(1)(2) := 'g';
ibt(1)(3) := 'h';
ibt(1)(4) := 'i';
ibt(1)(5) := 'k';
ibt(1)(6) := 'l';
ibt(1)(7) := 'm';
ibt(1)(8) := 'n';
dbms_output.put_line('Count = ' || ibt.count);
dbms_output.put_line('INDEX-BY TABLE ELEMENTS');
for i in 1..8 loop
dbms_output.put_line('ibt[1][' || i || '] = ' || ibt(1)(i));
end loop;
dbms_output.put_line('ibt([4][5] = ' || ibt(4)(5));
dbms_output.put_line('ibt([5][1] = ' || ibt(5)(1));
dbms_output.put_line('ibt([6][2] = ' || ibt(6)(2));
dbms_output.put_line('ibt([8][3] = ' || ibt(8)(3));
dbms_output.put_line('ibt([3][4] = ' || ibt(3)(4));
flag := ibt.exists(3);
if flag = true then
dbms_output.put_line('Index 3 exists');
else
dbms_output.put_line('Index 3 exists');
end if;
ibt.delete(1);
dbms_output.put_line('After delete of first index, Count = ' || ibt.count);
ibt.delete(4);
dbms_output.put_line('After delete of fourth index, Count = ' || ibt.count);
dbms_output.put_line('INDEX-BY TABLE ELEMENTS');
dbms_output.put_line('ibt([5][1] = ' || ibt(5)(1));
dbms_output.put_line('ibt([6][2] = ' || ibt(6)(2));
dbms_output.put_line('ibt([8][3] = ' || ibt(8)(3));
dbms_output.put_line('ibt([3][4] = ' || ibt(3)(4));
ibt.delete;
dbms_output.put_line('After delete of entire index-by table, Count = ' || ibt.count);
END;
Output:
Count = 0
No limit to Index-by Tables
INDEX-BY TABLE ELEMENTS
ibt([1][1] = a
ibt([4][5] = b
ibt([5][1] = c
ibt([6][2] = d
ibt([8][3] = e
ibt([3][4] = f
First Index = 1
Last Index = 8
Next Index = 4
Prior Index = 6
Count = 6
INDEX-BY TABLE ELEMENTS
ibt[1][1] = a
ibt[1][2] = g
ibt[1][3] = h
ibt[1][4] = i
ibt[1][5] = k
ibt[1][6] = l
ibt[1][7] = m
ibt[1][8] = n
ibt([4][5] = b
ibt([5][1] = c
ibt([6][2] = d
ibt([8][3] = e
ibt([3][4] = f
Index 3 exists
After delete of first index, Count = 5
After delete of fourth index, Count = 4
INDEX-BY TABLE ELEMENTS
ibt([5][1] = c
ibt([6][2] = d
ibt([8][3] = e
ibt([3][4] = f
After delete of entire index-by table, Count = 0
Ex3:
DECLARE
type t1 is table of varchar(2) index by binary_integer;
type t2 is table of t1 index by binary_integer;
type t3 is table of t2;
nt t3 := t3();
c number := 65;
BEGIN
nt.extend(2);
dbms_output.put_line('Count = ' || nt.count);
for i in 1..nt.count loop
for j in 1..nt.count loop
for k in 1..nt.count loop
nt(i)(j)(k) := chr(c);
c := c + 1;
end loop;
end loop;
end loop;
dbms_output.put_line('NESTED TABLE ELEMENTS');
for i in 1..nt.count loop
for j in 1..nt.count loop
for k in 1..nt.count loop
dbms_output.put_line('nt[' || i || '][' || j || '][' || k || '] = ' ||
nt(i)(j)(k));
end loop;
end loop;
end loop;
END;
Output:
Count = 2
NESTED TABLE ELEMENTS
nt[1][1][1] = A
nt[1][1][2] = B
nt[1][2][1] = C
nt[1][2][2] = D
nt[2][1][1] = E
nt[2][1][2] = F
nt[2][2][1] = G
nt[2][2][2] = H
OBJECTS USED IN THE EXAMPLES
SQL> select * from student;
SNO SNAME SMARKS
---------- -------------- ----------
1 saketh 100
2 srinu 200
3 divya 300
4 manogni 400
SQL> create or replace type addr as object(hno number(2),city varchar(10));/
SQL> select * from employ;
ENAME JOB ADDRESS(HNO, CITY)
---------- ---------- -----------------------------
Ranjit clerk ADDR(11, 'hyd')
Satish manager ADDR(22, 'bang')
Srinu engineer ADDR(33, 'kochi')
ERROR HANDLING
PL/SQL implements error handling with exceptions and exception handlers. Exceptions can be associated with oracle errors or with your own user-defined errors. By using exceptions and exception handlers, you can make your PL/SQL programs robust and able to deal with both unexpected and expected errors during execution.
ERROR TYPES
Compile-time errors
Runtime errors
Errors that occur during the compilation phase are detected by the PL/SQL engine and reported back to the user, we have to correct them.
Runtime errors are detected by the PL/SQL runtime engine which can programmatically raise and caught by exception handlers.
Exceptions are designed for run-time error handling, rather than compile-time error handling.
HANDLING EXCEPTIONS
When exception is raised, control passes to the exception section of the block. The exception section consists of handlers for some or all of the exceptions. An exception handler contains the code that is executed when the error associated with the exception occurs, and the exception is raised.
Syntax:
EXCEPTION
When exception_name then
Sequence_of_statements;
When exception_name then
Sequence_of_statements;
When others then
Sequence_of_statements;
END;
EXCEPTION TYPES
Predefined exceptions
User-defined exceptions
PREDEFINED EXCEPTIONS
Oracle has predefined several exceptions that corresponds to the most common oracle errors. Like the predefined types, the identifiers of these exceptions are defined in the STANDARD package. Because of this, they are already available to the program, it is not necessary to declare them in the declarative secion.
Ex1:
DECLARE
a number;
b varchar(2);
v_marks number;
cursor c is select * from student;
type t is varray(3) of varchar(2);
va t := t('a','b');
va1 t;
BEGIN
-- NO_DATA_FOUND
BEGIN
select smarks into v_marks from student where sno = 50;
EXCEPTION
when no_data_found then
dbms_output.put_line('Invalid student number');
END;
-- CURSOR_ALREADY_OPEN
BEGIN
open c;
open c;
EXCEPTION
when cursor_already_open then
dbms_output.put_line('Cursor is already opened');
END;
-- INVALID_CURSOR
BEGIN
close c;
open c;
close c;
close c;
EXCEPTION
when invalid_cursor then
dbms_output.put_line('Cursor is already closed');
END;
-- TOO_MANY_ROWS
BEGIN
select smarks into v_marks from student where sno > 1;
EXCEPTION
when too_many_rows then
dbms_output.put_line('Too many values are coming to marks variable');
END;
-- ZERO_DIVIDE
BEGIN
a := 5/0;
EXCEPTION
when zero_divide then
dbms_output.put_line('Divided by zero - invalid operation');
END;
-- VALUE_ERROR
BEGIN
b := 'saketh';
EXCEPTION
when value_error then
dbms_output.put_line('Invalid string length');
END;
-- INVALID_NUMBER
BEGIN
insert into student values('a','srinu',100);
EXCEPTION
when invalid_number then
dbms_output.put_line('Invalid number');
END;
-- SUBSCRIPT_OUTSIDE_LIMIT
BEGIN
va(4) := 'c';
EXCEPTION
when subscript_outside_limit then
dbms_output.put_line('Index is greater than the limit');
END;
-- SUBSCRIPT_BEYOND_COUNT
BEGIN
va(3) := 'c';
EXCEPTION
when subscript_beyond_count then
dbms_output.put_line('Index is greater than the count');
END;
-- COLLECTION_IS_NULL
BEGIN
va1(1) := 'a';
EXCEPTION
when collection_is_null then
dbms_output.put_line('Collection is empty');
END;
--
END;
Output:
Invalid student number
Cursor is already opened
Cursor is already closed
Too many values are coming to marks variable
Divided by zero - invalid operation
Invalid string length
Invalid number
Index is greater than the limit
Index is greater than the count
Collection is empty
Ex2:
DECLARE
c number;
BEGIN
c := 5/0;
EXCEPTION
when zero_divide then
dbms_output.put_line('Invalid Operation');
when others then
dbms_output.put_line('From OTHERS handler: Invalid Operation');
END;
Output:
Invalid Operation
USER-DEFINED EXCEPTIONS
A user-defined exception is an error that is defined by the programmer. User-defined exceptions are declared in the declarative secion of a PL/SQL block. Just like variables, exeptions have a type EXCEPTION and scope.
RAISING EXCEPTIONS
User-defined exceptions are raised explicitly via the RAISE statement.
Ex:
DECLARE
e exception;
BEGIN
raise e;
EXCEPTION
when e then
dbms_output.put_line('e is raised');
END;
Output:
e is raised
BULIT-IN ERROR FUNCTIONS
SQLCODE AND SQLERRM
SQLCODE returns the current error code, and SQLERRM returns the current error message text;
For user-defined exception SQLCODE returns 1 and SQLERRM returns “user-deifned exception”.
SQLERRM wiil take only negative value except 100. If any positive value other than 100 returns non-oracle exception.
Ex1:
DECLARE
e exception;
v_dname varchar(10);
BEGIN
-- USER-DEFINED EXCEPTION
BEGIN
raise e;
EXCEPTION
when e then
dbms_output.put_line(SQLCODE || ' ' || SQLERRM);
END;
-- PREDEFINED EXCEPTION
BEGIN
select dname into v_dname from dept where deptno = 50;
EXCEPTION
when no_data_found then
dbms_output.put_line(SQLCODE || ' ' || SQLERRM);
END;
END;
Output:
1 User-Defined Exception
100 ORA-01403: no data found
Ex2:
BEGIN
dbms_output.put_line(SQLERRM(100));
dbms_output.put_line(SQLERRM(0));
dbms_output.put_line(SQLERRM(1));
dbms_output.put_line(SQLERRM(-100));
dbms_output.put_line(SQLERRM(-500));
dbms_output.put_line(SQLERRM(200));
dbms_output.put_line(SQLERRM(-900));
END;
Output:
ORA-01403: no data found
ORA-0000: normal, successful completion
User-Defined Exception
ORA-00100: no data found
ORA-00500: Message 500 not found; product=RDBMS; facility=ORA
-200: non-ORACLE exception
ORA-00900: invalid SQL statement
DBMS_UTILITY.FORMAT_ERROR_STACK
The built-in function, like SQLERRM, returns the message associated with the current error.
It differs from SQLERRM in two ways:
Its length is not restricted; it will return the full error message string.
You can not pass an error code number to this function; it cannot be used to return the message for a random error code.
Ex:
DECLARE
v number := 'ab';
BEGIN
null;
EXCEPTION
when others then
dbms_output.put_line(dbms_utility.format_error_stack);
END;
Output:
declare
*
ERROR at line 1:
ORA-06502: PL/SQL: numeric or value error: character to number conversion error
ORA-06512: at line 2
DBMS_UTILITY.FORMAT_CALL_STACK
This function returns a formatted string showing the execution call stack inside your PL/SQL application. Its usefulness is not restricted to error management; you will also find its handy for tracing the exectution of your code. You may not use this function in exception block.
Ex:
BEGIN
dbms_output.put_line(dbms_utility.format_call_stack);
END;
Output:
----- PL/SQL Call Stack -----
Object_handle line_number object_name
69760478 2 anonymous block
DBMS_UTILITY.FORMAT_ERROR_BACKTRACE
It displays the execution stack at the point where an exception was raised. Thus , you can call this function with an exception section at the top level of your stack and still find out where the error was raised deep within the call stack.
Ex:
CREATE OR REPLACE PROCEDURE P1 IS
BEGIN
dbms_output.put_line('from procedure 1');
raise value_error;
END P1;
CREATE OR REPLACE PROCEDURE P2 IS
BEGIN
dbms_output.put_line('from procedure 2');
p1;
END P2;
CREATE OR REPLACE PROCEDURE P3 IS
BEGIN
dbms_output.put_line('from procedure 3');
p2;
EXCEPTION
when others then
dbms_output.put_line(dbms_utility.format_error_backtrace);
END P3;
Output:
SQL> exec p3
from procedure 3
from procedure 2
from procedure 1
ORA-06512: at "SAKETH.P1", line 4
ORA-06512: at "SAKETH.P2", line 4
ORA-06512: at "SAKETH.P3", line 4
EXCEPTION_INIT PRAGMA
Using this you can associate a named exception with a particular oracle error. This gives you the ability to trap this error specifically, rather than via an OTHERS handler.
Syntax:
PRAGMA EXCEPTION_INIT(exception_name, oracle_error_number);
Ex:
DECLARE
e exception;
pragma exception_init(e,-1476);
c number;
BEGIN
c := 5/0;
EXCEPTION
when e then
dbms_output.put_line('Invalid Operation');
END;
Output:
Invalid Operation
RAISE_APPLICATION_ERROR
You can use this built-in function to create your own error messages, which can be more descriptive than named exceptions.
Syntax:
RAISE_APPLICATION_ERROR(error_number, error_message,, [keep_errors_flag]);
The Boolean parameter keep_errors_flag is optional. If it is TRUE, the new error is added to the list of errors already raised. If it is FALSE, which is default, the new error will replace the current list of errors.
Ex:
DECLARE
c number;
BEGIN
c := 5/0;
EXCEPTION
when zero_divide then
raise_application_error(-20222,'Invalid Operation');
END;
Output:
DECLARE
*
ERROR at line 1:
ORA-20222: Invalid Operation
ORA-06512: at line 7
EXCEPTION PROPAGATION
Exceptions can occur in the declarative, the executable, or the exception section of a PL/SQL block.
EXCEPTION RAISED IN THE EXECUATABLE SECTION
Exceptions raised in execuatable section can be handled in current block or outer block.
Ex1:
DECLARE
e exception;
BEGIN
BEGIN
raise e;
END;
EXCEPTION
when e then
dbms_output.put_line('e is raised');
END;
Output:
e is raised
Ex2:
DECLARE
e exception;
BEGIN
BEGIN
raise e;
END;
END;
Output:
ERROR at line 1:
ORA-06510: PL/SQL: unhandled user-defined exception
ORA-06512: at line 5
EXCEPTION RAISED IN THE DECLARATIVE SECTION
Exceptions raised in the declarative secion must be handled in the outer block.
Ex1:
DECLARE
c number(3) := 'abcd';
BEGIN
dbms_output.put_line('Hello');
EXCEPTION
when others then
dbms_output.put_line('Invalid string length');
END;
Output:
ERROR at line 1:
ORA-06502: PL/SQL: numeric or value error: character to number conversion error
ORA-06512: at line 2
Ex2:
BEGIN
DECLARE
c number(3) := 'abcd';
BEGIN
dbms_output.put_line('Hello');
EXCEPTION
when others then
dbms_output.put_line('Invalid string length');
END;
EXCEPTION
when others then
dbms_output.put_line('From outer block: Invalid string length');
END;
Output:
From outer block: Invalid string length
EXCEPTION RAISED IN THE EXCEPTION SECTION
Exceptions raised in the declarative secion must be handled in the outer block.
Ex1:
DECLARE
e1 exception;
e2 exception;
BEGIN
raise e1;
EXCEPTION
when e1 then
dbms_output.put_line('e1 is raised');
raise e2;
when e2 then
dbms_output.put_line('e2 is raised');
END;
Output:
e1 is raised
DECLARE
*
ERROR at line 1:
ORA-06510: PL/SQL: unhandled user-defined exception
ORA-06512: at line 9
ORA-06510: PL/SQL: unhandled user-defined exception
Ex2:
DECLARE
e1 exception;
e2 exception;
BEGIN
BEGIN
raise e1;
EXCEPTION
when e1 then
dbms_output.put_line('e1 is raised');
raise e2;
when e2 then
dbms_output.put_line('e2 is raised');
END;
EXCEPTION
when e2 then
dbms_output.put_line('From outer block: e2 is raised');
END;
Output:
e1 is raised
From outer block: e2 is raised
Ex3:
DECLARE
e exception;
BEGIN
raise e;
EXCEPTION
when e then
dbms_output.put_line('e is raised');
raise e;
END;
Output:
e is raised
DECLARE
*
ERROR at line 1:
ORA-06510: PL/SQL: unhandled user-defined exception
ORA-06512: at line 8
ORA-06510: PL/SQL: unhandled user-defined exception
RESTRICTIONS
You can not pass exception as an argument to a subprogram.
DATABASE TRIGGERS
Triggers are similar to procedures or functions in that they are named PL/SQL blocks with declarative, executable, and exception handling sections. A trigger is executed implicitly whenever the triggering event happens. The act of executing a trigger is known as firing the trigger.
RESTRICTIONS ON TRIGGERES
Like packages, triggers must be stored as stand-alone objects in the database and cannot be local to a block or package.
A trigger does not accept arguments.
USE OF TRIGGERS
Maintaining complex integrity constraints not possible through declarative constraints enable at table creation.
Auditing information in a table by recording the changes made and who made them.
Automatically signaling other programs that action needs to take place when chages are made to a table.
Perform validation on changes being made to tables.
Automate maintenance of the database.
TYPES OF TRIGGERS
DML Triggers
Instead of Triggers
DDL Triggers
System Triggers
Suspend Triggers
CATEGORIES
Timing
--
Before or After
Level
--
Row or Statement
Row level trigger fires once for each row affected by the triggering statement. Row level trigger is identified by the FOR EACH ROW clause.
Statement level trigger fires once either before or after the statement.
DML TRIGGER SYNTAX
Create or replace trigger <trigger_name>
Before | after on insert or update or delete
[For each row]
Begin
-- trigger body
End <trigger_name>;
DML TRIGGERS
A DML trigger is fired on an INSERT, UPDATE, or DELETE operation on a database table. It can be fired either before or after the statement executes, and can be fired once per affected row, or once per statement.
The combination of these factors determines the types of the triggers. These are a total of 12 possible types (3 statements * 2 timing * 2 levels).
ORDER OF DML TRIGGER FIRING
Before statement level
Before row level
After row level
After statement level
Ex:
Suppose we have a follwing table.
SQL> select * from student;
NO NAME MARKS
----- ------- ----------
1 a 100
2 b 200
3 c 300
4 d 400
Also we have triggering_firing_order table with firing_order as the field.
CREATE OR REPLACE TRIGGER TRIGGER1
before insert on student
BEGIN
insert into trigger_firing_order values('Before Statement Level');
END TRIGGER1;
CREATE OR REPLACE TRIGGER TRIGGER2
before insert on student
for each row
BEGIN
insert into trigger_firing_order values('Before Row Level');
END TRIGGER2;
CREATE OR REPLACE TRIGGER TRIGGER3
after insert on student
BEGIN
insert into trigger_firing_order values('After Statement Level');
END TRIGGER3;
CREATE OR REPLACE TRIGGER TRIGGER4
after insert on student
for each row
BEGIN
insert into trigger_firing_order values('After Row Level');
END TRIGGER4;
Output:
SQL> select * from trigger_firing_order;
no rows selected
SQL> insert into student values(5,'e',500);
1 row created.
SQL> select * from trigger_firing_order;
FIRING_ORDER
--------------------------------------------------
Before Statement Level
Before Row Level
After Row Level
After Statement Level
SQL> select * from student;
NO NAME MARKS
---- -------- ----------
1 a 100
2 b 200
3 c 300
4 d 400
5 e 500
CORRELATION IDENTIFIERS IN ROW-LEVEL TRIGGERS
Inside the trigger, you can access the data in the row that is currently being processed. This is accomplished through two correlation identifiers - :old and :new.
A correlation identifier is a special kind of PL/SQL bind variable. The colon in front of each indicates that they are bind variables, in the sense of host variables used in embedded PL/SQL, and indicates that they are not regular PL/SQL variables. The PL/SQL compiler will treat them as records of type
Triggering_table%ROWTYPE.
Although syntactically they are treated as records, in reality they are not. :old and :new are also known as pseudorecords, for this reason.
TRIGGERING STATEMENT
:OLD
:NEW
-------------------------------------- ---------------------------- -----------------------------------------------
INSERT
all fields are NULL. values that will be inserted
When the statement is completed.
UPDATE
original values for new values that will be updated
the row before the when the statement is completed.
update.
DELETE original values before all fields are NULL.
the row is deleted.
Ex:
Suppose we have a table called marks with fields no, old_marks, new_marks.
CREATE OR REPLACE TRIGGER OLD_NEW
before insert or update or delete on student
for each row
BEGIN
insert into marks values(:old.no,:old.marks,:new.marks);
END OLD_NEW;
Output:
SQL> select * from student;
NO NAME MARKS
----- ------- ----------
1 a 100
2 b 200
3 c 300
4 d 400
5 e 500
SQL> select * from marks;
no rows selected
SQL> insert into student values(6,'f',600);
1 row created.
SQL> select * from student;
NO NAME MARKS
---- -------- ----------
1 a 100
2 b 200
3 c 300
4 d 400
5 e 500
6 f 600
SQL> select * from marks;
NO OLD_MARKS NEW_MARKS
---- --------------- ---------------
600
SQL> update student set marks=555 where no=5;
1 row updated.
SQL> select * from student;
NO NAME MARKS
----- ------- ----------
1 a 100
2 b 200
3 c 300
4 d 400
5 e 555
6 f 600
SQL> select * from marks;
NO OLD_MARKS NEW_MARKS
------ ---------------- ---------------
600
5 500 555
SQL> delete student where no = 2;
1 row deleted.
SQL> select * from student;
NO NAME MARKS
---- -------- ----------
1 a 100
3 c 300
4 d 400
5 e 555
6 f 600
SQL> select * from marks;
NO OLD_MARKS NEW_MARKS
----- -------------- ----------------
600
5 500 555
2 200
REFERENCING CLAUSE
If desired, you can use the REFERENCING clause to specify a different name for :old ane :new. This clause is found after the triggering event, before the WHEN clause.
Syntax:
REFERENCING [old as old_name] [new as new_name]
Ex:
CREATE OR REPLACE TRIGGER REFERENCE_TRIGGER
before insert or update or delete on student
referencing old as old_student new as new_student
for each row
BEGIN
insert into marks
values(:old_student.no,:old_student.marks,:new_student.marks);
END REFERENCE_TRIGGER;
WHEN CLAUSE
WHEN clause is valid for row-level triggers only. If present, the trigger body will be executed only for those rows that meet the condition specified by the WHEN clause.
Syntax:
WHEN trigger_condition;
Where trigger_condition is a Boolean expression. It will be evaluated for each row. The :new and :old records can be referenced inside trigger_condition as well, but like REFERENCING, the colon is not used there. The colon is only valid in the trigger body.
Ex:
CREATE OR REPLACE TRIGGER WHEN_TRIGGER
before insert or update or delete on student
referencing old as old_student new as new_student
for each row
when (new_student.marks > 500)
BEGIN
insert into marks
values(:old_student.no,:old_student.marks,:new_student.marks);
END WHEN_TRIGGER;
TRIGGER PREDICATES
There are three Boolean functions that you can use to determine what the operation is.
The predicates are
INSERTING
UPDATING
DELETING
Ex:
CREATE OR REPLACE TRIGGER PREDICATE_TRIGGER
before insert or update or delete on student
BEGIN
if inserting then
insert into predicates values('I');
elsif updating then
insert into predicates values('U');
elsif deleting then
insert into predicates values('D');
end if;
END PREDICATE_TRIGGER;
Output:
SQL> delete student where no=1;
1 row deleted.
SQL> select * from predicates;
MSG
---------------
D
SQL> insert into student values(7,'g',700);
1 row created.
SQL> select * from predicates;
MSG
---------------
D
I
SQL> update student set marks = 777 where no=7;
1 row updated.
SQL> select * from predicates;
MSG
---------------
D
I
U
INSTEAD-OF TRIGGERS
Instead-of triggers fire instead of a DML operation. Also, instead-of triggers can be defined only on views. Instead-of triggers are used in two cases:
To allow a view that would otherwise not be modifiable to be modified.
To modify the columns of a nested table column in a view.
