Sie sind auf Seite 1von 207

Features of Oracle 9i

&
PL/SQL

. 1
Features of Oracle 9i & PL/SQL

Section 1: Oracle Architecture


Section 2: Oracle Structures
Section 3: PL/SQL
Section 4: ORDBMS
Section 5: New Features of Oracle 9i

. 2
Features of Oracle 9i &
PL/SQL

Section 1: Oracle Architecture

. 3
ORACLE RDBMS

Database (information)
Physical Layer
Logical Layer

Server/Instance
Memory
Processes

. 4
Database Physical Layer

Data files – To store DB (end-user) data

Redo Log files – To log of all changes made to the DB

Control files – To store info used to start an instance

Trace files – To log info of background processes

Alert Log files – To log major events in the life of the DB

. 5
Database Logical Layer
Tablespaces – Logically related group of data
Schema Objects – Logical DB structures/abstractions
Segments – Set of extents
Data – Stores user data
Index – Stores indexes
Rollback – Stores info required for rollback of transactions
Temporary – Work area for SQL statements
Extents – Set of data blocks
Data Blocks – Smallest unit of info stored in DB

. 6
Instance Memory – System Global Area (SGA)

DB Buffer Cache – Most recently used data blocks


Dirty Block
Clean Block
Redo Log Buffer – Log of changes made to the DB
Shared Pool
Library Cache – Shared SQL ( parse tree, execution plan)
DD Cache – Logical/Physical structure of DB

. 7
Instance Memory – Program Global Area (PGA)

Stack Space – Session variables, arrays

Session Information

Private SQL Area – Binding variables, runtime buffers

. 8
Instance Processes
Server (Shadow) – Interacts with Oracle to carry out user request
Background – Task within Oracle RDBMS
DBWR – Reads/Writes DB/DD Cache to files
LGWR – Writes Redo Log buffer to files
CKPT – Signals DBWR to write modified buffers to files
PMON – Keeps track of DB processes, cleanup, lock management
RECO – Commit/Rollback pending transactions in a distributed DB
ARCH – Copies filled Redo Logs to archival storage
LCKn – Number of lock processes in Oracle Parallel Server option
Dnnn – Dispatches/routes requests to available server processes (MTS)

. 9
Features of Oracle 9i &
PL/SQL

Section 2: Oracle Structures

. 10
Tables
Table creation:
create table EMPLOYEE
( employee_id NUMBER,
date_of_joining DATE,
department_id VARCHAR2(2),
sex CHAR(1),
passport_no VARCHAR2(20),
address VARCHAR2(50),
salary NUMBER(12, 2),
constraint EMPLOYEE_PK PRIMARY KEY(employee_id)
using index tablespace INDEX_LARGE
)
tablespace DATA_LARGE PCTFREE 10 PCTUSED 40
storage( INITIAL 1M NEXT 1M MINEXTENTS 1 MAXEXTENTS unlimited PCTINCREASE 0 );

. 11
Constraints
Checks/Validations/Referential Integrity imposed on data

Primary Key:
constraint EMPLOYEE_PK PRIMARY KEY(employee_id)

Foreign Key:
constraint DEPT_FK FOREIGN KEY(department_id) references DEPARTMENT(department_id)

Check:
CHECK (salary > 0)
CHECK (sex in (‘M’, ‘F’))

Unique:
constraint PASSPORT_UQ UNIQUE(passport_no)

. 12
Views
Structure showing data from one or more tables when invoked
Data Hiding
Data Modification
Reporting

create view EMPLOYEE_VIEW as


select employee_id, department_id, address
from employee
with read only;

. 13
Indexes
Similar to Index of a book (Listing of keywords accompanied by location of information)
Used to improve query performance or enforce uniqueness
Internally stored using Binary Search (B-Tree) structure
Types of indexes:
Unique
Non-unique
Bitmapped
Function based

create unique index PASSPORT_IDX on EMPLOYEE(passport) tablespace INDEX_LARGE;


create bitmap index PASSPORT_IDX on EMPLOYEE(sex) tablespace INDEX_LARGE;
create index BOOKSHELF_IDX on BOOKSHELF(UPPER(title));

. 14
Indexes
Oracle Optimizer:

2 primary modes:
Rule-based (RBO)
Cost-based (CBO)
Set using OPTIMIZER_MODE parameter
To use CBO, analyze tables and indexes using:
Compute statistics – full object scan
Estimate statistics – part object scan
analyze table BOOKSHELF estimate statistics;
From Oracle 9i use DBMS_STATS package:
execute DBMS_STATS.GATHER_TABLE_STATS (‘TRNG’, ‘BOOKSHELF’);

. 15
Indexes
Operations that access Tables:
TABLE ACCESS FULL
select * from BOOKSHELF;

TABLE ACCESS BY ROWID


select * from BOOKSHELF where title like ‘T%’;

Operations that use Indexes:


INDEX UNIQUE SCAN
select * from BOOKSHELF where title = ‘ORACLE 9i’;

INDEX RANGE SCAN


select * from BOOKSHELF where title like ‘T%’;

. 16
Indexes
When Indexes are used:
When an indexed column is set equal to a value
select title from BOOKSHELF where categoryName = ‘ADULTNF’;
When a range of values is specified for an indexed column
select title from BOOKSHELF where title > ‘M’;
When no functions are performed on the column in the where clause
select title from BOOKSHELF where UPPER(categoryName) = ‘ADULTNF’;
When no IS NULL or IS NOT NULL checks are used on the indexed column
select title from BOOKSHELF where categoryName IS NOT NULL;
When equivalence conditions are used
select title from BOOKSHELF where categoryName != ‘ADULTNF’;
When MAX or MIN functions are used
select MAX(categoryName) from BOOKSHELF;
When the leading column of a multi-column index is set equal to a value
select title from BOOKSHELF where categoryName = ‘ADULTNF’;

. 17
Partitions

Splitting rows of a large table into smaller parts based on certain criteria
Partitions based on 3 criteria:

 Range

 Hash

 List

Indexes can be created on the entire table (Global) or on partitions (Local)

Partitions of partitions (subpartitions) can be created

. 18
Partitions
Splitting rows of a large table into smaller parts based on certain criteria
Benefits of Partitions–
Separation of data as per user choice
Smaller data segments, smaller table scans
Smaller indexes, faster access
Parallel DML on partition basis
Parallel backup and recovery
Parallel loading by partition
Export/Import by partition
Local/Global indexes
Parallel ANALYZE command for cost-based optimization more quick
Better data availability

. 19
Partitions
Range Partitions:

create table BOOKSHELF_RANGE


( title VARCHAR2(100) PRIMARY KEY,
publisher VARCHAR2(20),
categoryName VARCHAR2(20),
rating VARCHAR2(2)
)

partition by range(categoryName)

(partition PART1 values less than (‘B’) tablespace PART1_TS,


partition PART2 values less than (MAXVALUE) tablespace PART2_TS);

. 20
Partitions
Hash Partitions:

create table BOOKSHELF_HASH


( title VARCHAR2(100) PRIMARY KEY,
publisher VARCHAR2(20),
categoryName VARCHAR2(20),
rating VARCHAR2(2)
)
partition by hash(categoryName)
(partition PART1 tablespace PART1_TS,
partition PART2 tablespace PART2_TS);

. 21
Partitions
List Partitions:

create table BOOKSHELF_HASH


( title VARCHAR2(100) PRIMARY KEY,
publisher VARCHAR2(20),
categoryName VARCHAR2(20),
rating VARCHAR2(2)
)
partition by list(categoryName)
(partition PART1 values(‘ADULTFIC’, ‘ADULTNF’, ‘ADULTREF’) tablespace PART1_TS,
partition PART2 values(‘CHILDFIC’, ‘CHILDNF’, ‘CHILDREF’) tablespace PART2_TS);

. 22
Sequences

Unique number generator

Used to ensure uniqueness for primary key/unique columns

Can be cached in memory for faster access

Can be made to cycle beyond maximum value

Accessed using nextval and currval

create sequence employee_seq increment by 1 start with 1000;

select employee_seq.nextval from dual;

. 23
Features of Oracle 9i &
PL/SQL

Section 3: PL/SQL

. 24
PL/SQL

Introduction

. 25
PL/SQL Execution Environments - The PL/SQL Engine

PL/SQL BLOCK PL/SQL BLOCK


DECLARE DECLARE
Procedural Procedural PROCEDURAL
Procedural Procedural STATEMENT
BEGIN BEGIN
Procedural Procedural EXECUTOR
SQL SQL
Procedural Procedural
SQL SQL
END; END;

To the SQL Statement Executors in the ORACLE RDBMS

. 26
PL/SQL BLOCK STRUCTURE

DECLARE

BEGIN

EXCEPTION

END

. 27
PL/SQL BLOCK STRUCTURE

Declaration Section

Executable Section

Exception Handler Section


Quick Note - Block structuring
1. Any block may contain sub-block. Sub-blocks may appear
anywhere an executable statement may legally appear.
2. Statements end with a semi-colon (;)
3. Comments are preceded by -- or surrounded by /* */
4. Declared objects exist within a certain scope

. 28
PL/SQL

Variable Declaration

. 29
Variable Declarations Overview

Syntax of Declaration
identifier [constant ] datatype [not null ] [:= plsql_expression ] ;

Quick Notes - Variable Declaration


1. The rules for identifiers are same as for SQL objects.
2. NOT NULL/CONSTANT may be optionally used
3. Only one identifier per line is allowed .

DECLARE
firstname lastname CHAR(20) ; - illegal
DECLARE
firstname CHAR(20) ; -legal
lastname CHAR(20) ; - legal

. 30
Variable Declarations Overview
NUMBER
Count NUMBER;
revenue NUMBER (9,2);
second_per_day CONSTANT NUMBER := 60 * 60* 24 ;
running _total NUMBER (10,0) := 0;

VARCHAR2
mid_initial VARCHAR2 := ‘K’;
last_name VARCHAR2(10) NOT NULL;
company_name CONSTANT VARCHAR2(12);

DATE
anniversary DATE := ‘05-NOV-78’;
project_complexion DATE;
next_checkup DATE NOT NULL := ‘28-JUN-90’;

BOOLEAN
over_budget BOOLEAN NOT NULL := FALSE;
available BOOLEAN := NULL ;

. 31
Attribute Declaration

PL/SQL objects (such as variables and constants) and database objects


(such as columns and tables ) are associated with certain attributes.

%TYPE attribute
DECLARE
books_printed NUMBER (6);
books_sold books.sold%TYPE ;
maiden_name emp.ename%TYPE ;

%ROWTYPE attribute
DECLARE
dept_row dept%ROWTYPE ;

. 32
Variable Assignment
PL/SQL Expressions consist of Variables, Constants, Literals, and
Function Calls.
ASSIGNMENT Syntax
plsql_variable := plsql_expression;

Quick notes -Assignment


1. := (ASSIGNMENT ) whereas = (VALUE EQUALITY)
2. The datatype of the left and right hand side of an assignment must be
the same or implicitly convertible to each other.
For ex. , N:=‘7’ is legal because number may be implicitly converted to char.
3. Column or table reference are not allowed on either side of an
assignment operator( : = ).
SCOTT.EMP.EMPNO := 1234;
These are illegal
location := dept.loc.;

. 33
Scoping

SCOPE refers to the visibility of identifiers at different points in the


PL/SQL block.

SCOPING RULES:
1. An identifier is visible in the block in which it is declared and all its
sub-blocks unless rule #2 applies.

2. If an identifier in an enclosing block is redeclared in a sub-block, the


original identifier declared in the enclosing block is no longer visible in
the sub-block .However, the newly declared identifier has the rules of
scope defined in rule #1.

. 34
Scoping Variables and Constants
DECLARE
credit_limit CONSTANT NUMBER (6,2) : =2000;
account NUMBER := 100;
BEGIN
DECLARE
account CHAR(10) := ‘ABC’;
new_balance NUMBER (9,2);
BEGIN
new_balance account credit_limit
END;
DECLARE
account NUMBER := 200;
old_balance NUMBER (9,2);
BEGIN
old_balance account credit_limit
END;

END; account credit_limit

. 35
PL/SQL

SQL in PL/SQL

. 36
SQL & PL/SQL Overview

SQL Data Manipulation Language (DML) statement support


1. INSERT
2. UPDATE
3. DELETE
4. SELECT
QuickNotes - SQL DML Support
1. The full ORACLE syntax is supported for these statements
2. A PL/SQL variable may be placed anywhere a constant may be
legally placed.
3. An identifier is first checked to see if it is a column in the database.
If not, it is assumed to be a PL/SQL identifier.
4. These statements may not appear as part of an expression

. 37
SQL & PL/SQL Overview
INSERT
DECLARE
my_sal NUMBER(7,2) := 3040.22;
my_ename CHAR(25) := ‘WANDA’;
my_hiredate DATE := ‘08-SEP-01’;
BEGIN
INSERT INTO EMP (empno, ename, job, hiredate, sal , deptno)
VALUES (2345, my_ename, ’cab Driver’, my_hiredate, my_sal, 20);
END;
EMPNO ENAME SAL EMPNO ENAME SAL
7644 TURNER 1500 7644 TURNER 1500
7400 ALLEN 1600

INSERT 7400 ALLEN 1600

. 38
SQL & PL/SQL Overview
UPDATE
DECLARE
max_allowed CONSTANT N UMBER := 5000;
good_cust CHAR(8) := ‘VIP’;
BEGIN
UPDATE ACCOUNT SET CREDIT_LIMIT = MAX_ALLOWED
WHERE TYPE = ‘EMPLOYEE ‘ OR TYPE = good_cust;
END;

EMPNO ENAME SAL EMPNO ENAME SAL


7644 TURNER 1500 7644 TURNER 1500
7400 ALLEN 1600 7400 ALLEN 1400

UPDATE

. 39
SQL & PL/SQL Overview
DELETE
DECLARE
bad_child_type CHAR(8) := ‘NAUGHTY’;
BEGIN
DELETE FROM santas_gift_list WHERE kid_rating = bad_child_type ;
END;

EMPNO ENAME SAL EMPNO ENAME SAL


7644 TURNER 1500 7644 TURNER 1500
7400 ALLEN 1600

DELETE

. 40
SQL & PL/SQL Overview

EMPNO ENAME SAL


APPLICATION 7644 TURNER 1500
VAR1 7400 ALLEN 1600
VAR2
VAR3

QuickNotes - SELECT INTO


1. A SELECT statement is the only DML that returns data. You must
provide location for this data to be stored via the INTO clause.
2. A SELECT..INTO statement must return exactly one row. Zero or
multiple returned rows result in an error.
3. For multi-row SELECTs use cursors.
. 41
SQL & PL/SQL Overview
SELECT Syntax
SELECT col1,col2……INTO var1,var2.. FROM table_name WHERE ...
E.g.
DECLARE
part_name parts.name%TYPE;
num_in_stock parts.num%TYPE;
BEGIN
SELECT name, num
INTO part_name, num_in_stock
FROM PARTS
WHERE part_id = 234;
-- manipulate the retrieved data here

. 42
Transaction processing

SAVEPOINT Syntax
SAVEPOINT <marker_name>;

ROLLBACK TO Syntax
ROLLBACK [WORK] TO SAVEPOINT <marker_name>;

SAVEPOINT and ROLLBACK TO Ex.

BEGIN
INSERT INTO temp VALUES (1,1 ‘ROW 1’); SAVEPOINT A;
INSERT INTO temp VALUES (2,2 ‘ROW 2’); SAVEPOINT B ;
….
ROLLBACK TO SAVEPOINT B;
COMMIT ;
END;

. 43
SQL Functions
SQL Functional Support (within a SQL Statement):
1. Numeric (e.g. SQRT, ROUND, POWER)
2. Character (e.g. LENGTH, UPPER)
3. Date (e.g. ADD_MONTHS, MONTH_BETWEEN);
4. Group(e.g. AVG, MAX, COUNT)

INSERT INTO phonebook (lastname) VALUES (UPPER(my_lastname));


Other SQL Functional Support (outside of a SQL Statement):
MOST ORACLE SQL functional are available (except for group
functions).
X := SQRT(y);
lastname := UPPER (lastname);
age_diff := MONTHS_BETWEEN(birthday1,birthday2)/12;

. 44
PL/SQL

Conditional & Iterative Control

. 45
Logical Comparisons
•Logical Comparisons form the basis of conditional control in PL/SQL
•The result of these comparisons are always either TRUE ,FALSE or NULL
•Anything compared with NULL results in a NULL value.
•A NULL in an expression evaluates to NULL (except concatenation)
E.g.
5 + NULL evaluates to NULL
‘PL/’ || NULL || ‘SQL’ evaluates to ‘PL/SQL’

PL /SQL Datatypes Operators


NUMBER <,>
CHAR
=, !=
DATE
BOOLEAN <=, >=

. 46
IF Statements

‘IF’ Statements are used to conditionally execute the statement or


sequence of statements.
‘IF’ Statement Syntax
IF <condition> THEN <sequence of statements >
[ELSEIF <condition> THEN <sequence of statements> ]
-- ELSEIFs may be repeated
[ELSE <sequence of statements>]
END IF;

QuickNotes - IF Statement
1. <condition> must evaluate to a Boolean datatype (TRUE, FALSE,
NULL)
2. If <condition> is TRUE, then the associated <sequence of statements>
is executed; otherwise it is not
3. At most one <sequence of statements > gets executed

. 47
IF Statements
DECLARE
num_jobs NUMBER(7);
BEGIN
SELECT COUNT(*) INTO num_jobs FROM auditions
WHERE actorid=&&actor_id AND called_back =‘YES’;
IF num_jobs> 90 THEN
UPDATE actor SET actor_rating = ‘ OSCAR time’
WHERE actorid = &&actor_id;
ELSE IF num_jobs> 75 THEN
UPDATE actor SET actor_rating = ‘ DAY time soaps’
WHERE actorid = &&actor_id;
ELSE
UPDATE actor SET actor_rating = ‘ Waiter’
WHERE actorid = &&actor_id;
END IF;
COMMIT;
END;

. 48
IF Statements
The NULL Trap
BLOCK 1 BLOCK 2
. .
IF a >= b THEN IF a < b THEN
do_this …..; do_that …..;
ELSE ELSE
do_that….; do_this….;
END IF; END IF;

• Given any pair of non-NULL values for “a” and”b”,


will Block 1 and Block 2 do the same thing?
• What if either “a” or”b” (or both) is NULL?

. 49
Loop Statement Overview

Loops repeat a statement or sequence of statements multiple


times.
Four types of loops:
1. Simple Loops.
2. Numeric FOR Loops.
3. While Loops.
4. Cursor FOR Loops.

. 50
Loop Statements
Simple Loops repeat sequence of statements multiple times.
Simple Loop Syntax
Loop
<Sequence of Statements>
END LOOP ; -- sometimes called an ‘infinite’ loop

Exit statements exit any type of loop immediately


Exit Syntax
EXIT [WHEN <condition >]; -- ‘infinite’ loop insurance

. 51
Loop Statements ……Example
DECLARE
ctr NUMBER(3) := 0;
BEGIN
LOOP
INSERT INTO LOG VALUES (ctr, ’ITERATION COMPLETE’);
ctr := ctr +1;
IF ctr = 1500 THEN EXIT;
END IF;
END LOOP;
END;
DECLARE
ctr NUMBER(3) := 0;
BEGIN
LOOP
UPDATE TABLE 1 SET COMMIT = ‘UPDATES’ WHERE COUNT_COL = ctr;
ctr := ctr +1;
IF ctr = 1500 THEN EXIT;
END IF;
END LOOP;
END;

. 52
Loop Statements
Numeric FOR Loops repeat sequence of statements fixed number
of times.
Numeric FOR Loop Syntax
FOR <index> IN [REVERSE ] <integer>..<integer> LOOP <sequence of
statements>
The Loop Index takes on each value in range , one of a time , either in
forward or reverse order.
E.g.
BEGIN
FOR I IN 1..500 LOOP
INSERT INTO temp(message)VALUES (‘I will not sleep in class.’);
END LOOP;
END;

. 53
Loop Statements
QuickNotes - Index
1. It is implicitly of type NUMBER
2. It is only defined within the loop
3. Value may be referenced in an expression, but a new value may not be
assigned to it within the loop
E.g.
DECLARE
my_index CHAR(20) := ‘Fettuccini Alfredo’;
BEGIN
FOR my index IN REVERSE 21…30 LOOP /* redeclare s my_index*/
INSERT INTO temp(coll.)VALUES (my_index); /* insert the numbers 30 through 21*/
END LOOP;
END;
FOR i I N 1…256 LOOP
x := x + i ; -- legal
i := I + 5; -- illegal
END LOOP;

. 54
Loop Statements
WHILE Loops repeat a sequence of statements until a specific condition is no longer
TRUE.
While Loop Syntax
WHILE <condition > LOOP <sequence of statements > END LOOP;

QuickNotes - WHILE Loops


1. The term <condition> may be any legal PL/SQL condition (i.e. it must return a
Boolean value of TRUE, FALSE or NULL)
2. The sequence of statements will be repeated as long as <condition> evaluates to
TRUE

DECLARE
ctr NUMBER (3) := 0;
BEGIN
WHILE ctr < 500 LOOP
INSERT INTO temp(message) VALUES (‘Well,I might sleep just a little’);
ctr := ctr +1 ;
END LOOP;
END;

. 55
“ GO TO ” Statement Overview

“ GO TO “ Statements jump to a different place in the PL/SQL


block.
“GO TO” Statements have 2 parts
1. The GOTO statement itself.
2. A statement ‘label’
“GO TO “ Statement Syntax
<<label_name >> X :=X+1 ; - - statement label
GOTO LABEL_NAME - - JUMPS TO x := x +1

. 56
“ GO TO ” Statements
NOT ALL GOTOs are Legal !
You can legally a GOTO a statement that is either:
1.in the same sequence of statements as the GOTO STATEMENT
2. In the sequence of statements that encloses the GOTO statement (I.e. an
outer block)

<<dinner>> GOTO your_brothers;


x := x + 1 ; IF a > b THEN
y := y + 1; b := b - a;
IF a >= b THEN <<your_brothers>>
b : = b + c;
x := x - 1;
GOTO dinner;
END IF;
END IF;

. 57
Other Uses for Statement Labels
Labels may label any statement
In addition to their use as targets for GOTO statements, labels may be used for :
1. Blocks
2. Loops
Labeling a block allows referencing of DECLARED objects that would
otherwise not be visible because of Scoping rules.
Syntax
<< label_name>>
[ DECLARE
-- declarations go here ]
BEGIN
-- executable statements go here
[ EXCEPTION
-- exception handlers go here ]
END label_name ; -- must include the label_name

. 58
Other Uses for Statement Labels
E.g.
<< outer_block >>
DECLARE
n NUMBER;
BEGIN
n := 5;
/* Start a sub block */
DECLARE
x NUMBER := 10;
n CHAR (10) := ‘Fifteen’;
BEGIN
INSERT INTO TEMP VALUES (outer_block.n , x , n );
COMMIT;
END ; /* End of the sub block */
END outer_block;

. 59
Other Uses for Statement Labels

Labeling a Block allows you to reference a variable that might be hidden by


a column name

E.g.

<< sample >>


DECLARE
deptno NUMBER := 20;
BEGIN
UPDATE emp SET sal = sal * 1.1
WHERE deptno = sample.deptno;
COMMIT;
END sample;

. 60
Other Uses for Statement Labels

Labeling Loops allows you to reference objects that would otherwise not be
visible because of scoping rules

E.g.

<< compute_loop >>


For i IN 1…10 LOOP
< statements …. >
DECLARE
i NUMBER := 0 ;
BEGIN
INSERT INTO temp VALUES
(i, compute_loop.i, ‘COMPLETE’ );
END;
END LOOP compute_loop; - must include loop name here

. 61
Other Uses for Statement Labels

Labeling EXITs is a convenient way to specify exits from outer loops


E.g.
<< outer_loop >> WHILE a > b LOOP
b := b + 1;
<< inner_loop >> WHILE b > c LOOP
c := c + 2 ;
EXIT outer_loop WHEN c > 200 ;
END LOOP inner_loop;
END LOOP outer_loop;

. 62
PL/SQL

Cursors

. 63
Cursor Overview

Every SQL DML statement processed by PL/SQL has an associated


CURSOR.
Two Types of CURSORS
1. EXPLICIT
Multiple row SELECT STATEMENTS
2. IMPLICIT
All INSERT statements
All UPDATE statements
All DELETE statements
Single row SELECT….INTO Statements

. 64
Using Explicit Cursors
STEP 1 . Declare the cursor
DECLARE
CURSOR <cursor name> IS <regular select statement> ;

QuickNotes - CURSOR Declaration


1. The < regular select statement > must NOT include the INTO clause required in a
single-row SELECT….INTO statement
2. Declared cursors are scoped just like variables
Cursor Declaration Example
DECLARE
X NUMBER ( 7, 2 ) ;
total NUMBER ( 5 )
lower_sal_limit CONSTANT NUMBER ( 4 ) := 1200 ;
CURSOR c1 IS SELECT ename FROM emp WHERE sal > lower_sal_limit ;
BEGIN ...

. 65
Using Explicit Cursors

STEP 2 . Open the cursor


OPEN < cursor name > ;
STEP 3 . Fetch data from the cursor
FETCH < cursor name > INTO < var1 ,var2 > ;
Quick Notes - FETCH
1. Retrieves one row of data from the cursor and stores it in the specified variables
(similar to how a single-row select works)
2. There must be exactly one INTO variable for each column selected by the
SELECT statement
3. The first column gets assigned to var1 , the second to var2 , etc .
STEP 4 . Close the cursor
CLOSE < cursor name > ;

. 66
Explicit Cursors Attributes
%NOTFOUND
E.g.
LOOP
FETCH my_cursor INTO my_ename , my_sal ;
EXIT WHEN my_cursor%NOTFOUND ;
-- process data here
END LOOP ;
%FOUND
E.g.
FETCH my_cursor INTO my_ename ,my_sal ;
WHILE my_cursor%FOUND LOOP
-- process data here
FETCH my_cursor INTO my_ename ,my_sal ;
END LOOP ;

. 67
Explicit Cursor Attributes
%ROWCOUNT
E.g.
LOOP
FETCH my_cursor INTO my_ename , my_sal ;
EXIT WHEN (my_cursor%NOTFOUND)
OR (my_cursor%ROWCOUNT > 10) ;
-- process data here
END LOOP
%ISOPEN
E.g.
IF my_cursor%ISOPEN THEN
FETCH my_cursor INTO my_ename , my_sal ;
ELSE
OPEN my_cursor ;
END IF ;

. 68
Using Explicit Cursors
E.g.
DECLARE
sal_limit NUMBER ( 4 ) := 0 ;
my_ename emp.ename%TYPE ;
my_sal emp.sal%TYPE ;
CURSOR my_cursor IS SELECT ename , sal FROM emp WHERE sal > sal_limit ;
BEGIN
sal_limit := 1200 ;
OPEN my_cursor INTO my_ename , my_sal ;
LOOP
FETCH my_cursor INTO my_ename , my_sal ;
EXIT WHEN my_cursor%NOTFOUND ; -- nothing returned
INSERT INTO new_table VALUES ( my_ename , my_sal ) ;
END LOOP ;
CLOSE my_cursor ;
COMMIT ;
END ;

. 69
Explicit Cursors -FOR Loops
Cursor FOR Loops specify a sequence of statements to be repeated once for
each row that is returned by the cursor.
Cursor FOR Loop Syntax
FOR <record _name> IN <cursor_name> LOOP
--- statements to be repeated go here
END LOOP;
Numeric FOR Loop Similarities
1. Specify a set of rows from a table by using the cursor’s name vs. specifying a
set of integers (i.e. 1…10)
2. Index takes on the values of each row vs. index taking on integer values (i.e. 1
through 10)
Implicitly Declared <record_name>
record_name cursor _name%ROWTYPE;
To reference an element of the record, use the record_name.column_name
notation.

. 70
Explicit Cursors -FOR Loops
Conceptual Cursor Loop Model

Loops When a cursor loop is initiated, an implicit


OPEN cursor_name is executed.

For each row that satisfies the query associated


Loops with the cursor an, implicit FETCH is executed
into the components of record_name.

When there are no more rows left to FETCH,


Loops an implicit CLOSE cursor_name is executed
and the loop is exited.

. 71
Explicit Cursors - FOR Loops
E.g.
DECLARE
sal_limit NUMBER ( 4 ) := 0 ;
total_sal NUMBER (9,2 ) := 0;
CURSOR my_cursor IS SELECT ename , sal FROM emp
WHERE sal > sal_limit ;
BEGIN
sal_limit := 1200 ;
-- implicitly OPEN done next
FOR cursor_row IN my_cursor LOOP
-- an implicit fetch done here
INSERT INTO new_table VALUES (cursor_row.ename ,cursor_row.sal ) ;
total_sal := total_sal + cursor_row.sal;
END LOOP ; --an implicit close done here.
COMMIT ;
END ;

. 72
Implicit Cursors - FOR Loops
An Implicit Cursor is automatically associated with any SQL DML
statement that does not have an explicit cursor associated with it.

This includes :
1. ALL INSERT statements
2. ALL UPDATE statements
3. ALL DELETE statements
4. ALL SELECT…INTO statements
QuickNotes - Implicit Cursors
1. Implicit cursor is called the “SQL” cursor --it stores
information concerning the processing of the last SQL
statement not associated with an explicit cursor.
2.OPEN, FETCH, AND CLOSE don’t apply.
3. All cursor attributes apply.

. 73
Implicit Cursors

SQL %NOTFOUND
E.g.

UPDATE emp SET sal = sal * 10.0


WHERE ename =“WARD” ;

IF SQL %NOTFOUND THEN


-- WARD wasn’t found
INSERT INTO emp (empno, ename ,sal)
VALUES ( 1234,’WARD’ 99999 );
END IF ;

. 74
Implicit Cursors

SQL%ROWCOUNT
E.g.
DELETE FROM baseball_team
WHERE batting _avg. < .100;

IF SQL%ROWCOUNT > 5 THEN


INSERT INTO temp(message)
VALUES(‘Your team needs helps.’);
END IF;

. 75
PL/SQL

Exception Handling

. 76
Exception Overview

 In PL/SQL error are called exceptions

 When an exception is raised, processing jumps to the


exception handlers

 An exception handler is a sequence of statements to be


processed when a certain exception occurs

 When an exception handler is complete processing of the


block terminates

. 77
Exception Overview

Two Types of Exceptions


1. PREDEFINED INTERNAL EXCEPTIONS
2. USER-DEFINED EXCEPTIONS

PL/SQL’s Exception Handlers


vs.
Conventional Error Handling

. 78
Predefined Internal Exceptions
Any ORACLE error “raises” an exception automatically; some of the
more common ones have names.
TOO_MANY_ROWS ORA-(01427)
- a single row SELECT returned more than one row
NO_DATA_FOUND ORA-(01403)
- a single row SELECT returned no data
INVALID_CURSOR ORA-(01001)
- invalid cursor was specified
VALUE_ERROR ORA-(06502)
- arithmetic ,numeric, string , conversion,or constraint error occurred.
ZERO_DIVIDE ORA-(01476)
- attempted to divide by zero
DUP_VAL_ON_INDEX ORA-(00001)
- attempted to insert a duplicate value into a column that has a unique index
specified.

. 79
Exception Handlers
Syntax
WHEN <exception_name [OR <exception_name…]> then <sequence of statements>
OR
WHEN OTHERS THEN -- if used , must be last handler <sequence of statements>
E.g.
DECLARE
employee_num emp.empno%TYPE;
BEGIN
SELECT empno INTO employee_num FROM emp;
WHERE ename = ‘BLAKE’;
INSERT INTO temp VALUES(NULL, empno,Blake's employee_num’);
DELETE FROM emp WHERE ename =‘BLAKE’;
EXCEPTION
WHEN TOO_MANY_ROWS OR NO_DATA_FOUND THEN
ROLLBACK;
INSERT INTO temp VALUES (NULL,’Blake not found, or more than one Blake’);
COMMIT;

WHEN OTHERS THEN


ROLLBACK;
END;

. 80
User - Defined Exceptions
User - defined Exceptions must be defined and explicitly raised by the user.
E.g.
DECLARE
x NUMBER;
my_exception EXCEPTION; -- a new object type.

Raise your_exception;
RAISE my_exception;

Quick Notes -RAISE <exception_name>


1. Once an exception is RAISED manually, it is treated exactly the same as
if it were a predefined internal exception.
2. Declared exceptions are scoped just like variables.
3. A user-defined exception is checked for manually and then RAISED , if
appropriate.

. 81
User - Defined Exceptions
E.g.
DECLARE
my_ename emp.ename%TYPE :=‘BLAKE’;
assigned_projects NUMBER;
too_few_projects EXCEPTION
BEGIN ---- get no of projects assigned to BLAKE
IF assigned_project < 3 THEN
RAISE too_few_projects;
END IF;
EXCEPTION --begin the exception handlers
WHEN too_few_projects THEN
INSERT INTO temp
VALUES(my_ename,assigned_projects,’LESS THAN 3 PROJECTS!’)
COMMIT;
END;

. 82
Exceptions Propagation
Step# 1 The current block is searched for a handler. If not found, go to step 2.
Step# 2 If an enclosing block is found, it is searched for it handler.
Step# 3 Step #1 and#2 are repeated until either there are no more enclosing
blocks, or a handler is found .
 If there are no more enclosing blocks, the exception is passed back to the
calling environment (SQL *Plus,SQL *Forms, a precompiled program,etc.)
 If the handler is found ,it is executed .when done the block in which the handler
was found is terminated, and control is passed to thee enclosing block (if one
exists), or to environment (if there is no enclosing block)
Quick notes
1. Only one handler per block may be active at a time
2. If an exception is raised in a handler, the search for a handler for the new
exception begins in the enclosing block of the current block

. 83
Exceptions Propagation
Example 1

BEGIN
...
BEGIN
IF X=1 THEN RAISE A:
ELSEIF X=2 THEN RAISE B;
ELSE RAISE C;
Exception A is handled
locally and execution
EXCEPTION resumes in the outer
WHEN A THEN block
...
END;

EXCEPTION
WHEN B THEN
...
END;

. 84
Exceptions Propagation
Example 2

BEGIN
...
BEGIN
IF X=1 THEN RAISE A:
ELSEIF X=2 THEN RAISE B;
ELSE RAISE C;
Exception B
PROPAGATES to the
EXCEPTION first outer block with an
WHEN A THEN appropriate handler
...
END;

EXCEPTION Exception B handled and


WHEN B THEN control is passed back to
... the calling environment

END;

. 85
Exceptions Propagation
Example 3
BEGIN
...
BEGIN
IF X=1 THEN RAISE A:
ELSEIF X=2 THEN RAISE B;
ELSE RAISE C;

EXCEPTION
WHEN A THEN
…..
END;

EXCEPTION Exception C has no


handler and will result
WHEN B THEN in runtime unhandled
… .. exception
END;

. 86
Other uses of RAISE

By itself ,the RAISE statement simply re-raise the


current exception (as if it were being propagated)

Syntax

RAISE;

Quick Notes - RAISE

1. RAISE may only be used in an exception handler

. 87
Error Reporting Functions
SQLCODE and SQLERRM
1. Provided information on the exception currently being handled.
2. Especially useful in the OTHERS handler.
SQLCODE
1. Returns the ORACLE error number of the exception or 1 if it was user-
defined exception
SQLERRM
1. Return the ORACLE error message currently associated with the current
value of SQLCODE
2. May also use any ORACLE error number as an argument.
QuickNotes - Error Reporting
1. If no exception is active …
SQLCODE = 0
SQLERRM = “normal , successful completion”
2. SQLCODE and SQLERRM cannot be used within a SQL statement.

. 88
Error Reporting Functions
E.g.
DECLARE
sqlcode_val NUMBER;
sqlcode_val CHAR(70);
BEGIN

EXCEPTION
WHEN OTHERS THEN
sqlcode _val := SQLCODE; -- can’t insert directly
sqlerrm_val := SQLERRM ; - -- can’t insert directly
INSERT INTO temp VALUES(sqlcode_val, NULL, sqlerrm_val);
END;

. 89
PL/SQL

Procedures & Packages

. 90
Stored Procedures and Functions

 Collections of SQL and PL/SQL statements


 Stored in complied from in the database
 Can call other procedures
 Can be called from all client environments
 Procedures and function (including remote) are the same,
except a function returns a values and a procedure does not.

. 91
Uses for Procedures
 Define central well-known business functions.
 Create an order
 Delete a customer
 Store batch job in the database
 Weekly account rollups
 Encapsulate a transaction
 Gather and process information from remote nodes
 Funnel activity on a table through a single path
 All changes to employee salaries should change department
budgets.

. 92
Procedure Arguments

Argument Modes

IN Data value comes in from the calling


process and is not changed

OUT No data value comes in from the calling


process; on normal exit ,value of argument
is passed back to caller

IN OUT Data value comes in from the calling


process, and another value is returned on
normal exit

. 93
Creating a Procedure

Ο E.g.
CREATE PROCEDURE
fire_employee (empid NUMBER)
AS
BEGIN

DELETE
FROM emp
WHERE empno=
fire_employee.empid;
END

Ο Tip:Write each procedure in a text file, and save

. 94
Creating a Function

E.g.

CREATE FUNCTION
get_bal (acc_no NUMBER(4))
RETURN NUMBER(11,2);
IS
acc_bal NUMBER(11,2);
BEGIN
SELECT balance
INTO acc_bal
FROM accounts
WHERE account_id_no=acc_no;
RETURN (acc_bal);
END;

. 95
Statements in procedures

Ο Valid statements in a procedure or function


• SQL DML or PL/SQL statements
• Calls to other procedures and functions stored in the database
• Calls to other procedures and functions in a remote database

Ο Restricted statements
• DDL
• Dynamic SQL
• In trigger, COMMIT, SAVEPOINT and ROLLBACK

. 96
Executing a stored procedure

Ο From within a PL/SQL block


fire_employee (empno);
scott.fire_employee (empno);
Ο On a remote node
scott.fire_employee@ny (empno);
Ο From SQL*DBA and some ORACLE tools
EXECUTE fire_employee (1043)
Ο Within an anonymous block (when EXECUTE not available)
SQLPLUS>
BEGIN FIRE_EMPLOYEE(1043); END;
Ο From a precompiler application
EXEC SQL
FIRE_EMPLOYEE(:empno);

. 97
Specifying procedure arguments

Ο E.g.
• CREATE PROCEDURE update_sal
(empno NUMBER,
bonus NUMBER,
sal_incr NUMBER) ….;
Ο Positional Method
• List values in the order they are declared
update_sal (7000,20,500);

Ο Named Method
• List argument names and values in any order, using special syntax
update_sal
(bonus => 20,
sal_incr => 500,
empno => 7000);

. 98
Specifying procedure arguments

Ο Combination method
• Use both positional and named methods
• Once you specify a named parameter, must use named method for all remaining
parameters

update_sal Legal
(7000,sal_incr=>500,bonus=>20);

update_sal Illegal
(empno=>7000,
sal_incr=>500,20);

. 99
How procedures are entered into the database

Ο PL/SQL Engine compiles the source code

Ο ORACLE stores a procedure in a database:


• Object name
• Source code
• Parse tree
• Pseudo code(P-code)
• Syntax errors in dictionary table
• Dependency information

Ο SQL in procedure not stored in parsed form


• Uses shared and cached SQL
• Allows SQL to be optimized dynamically (without recompiling
referencing procedures)

. 100
PL/SQL Compilation Errors

Ο Compile done by PL/SQL engine in RDBMS


Ο Error stored in the database
Ο To view errors:

• Use SQL*DBA command SHOW ERRORS


• Select from errors views
• USER_ERRORS
• ALL_ERRORS
• DBA_ERRORS

. 101
PL/SQL Compilation Errors Executing SHOW
ERRORS

SQLDBA>create procedure test1 is begin test2;end;


DBA-00072: Warning: PROCEDURE TEST1
created with compilation errors.

SQLDBA>show errors
ERRORS FOR PROCEDURE TEST1:
LINE/COL ERROR
---------------------------------------------------------------------------------------------
1/0 PL/SQL: Compilation unit
analysis terminated
1/33 PL/SQL-00219:’test2’ is not defined’
2 rows selected

. 102
PL/SQL Compilation Errors
Fields in ERRORS views

Ο NAME:name of the object

Ο TYPE: one of the following:


• PROCEDURE
• FUNCTION
• PACKAGE
• PACKAGE BODY
Ο LINE:line number where error occurs

Ο TEXT:text of error

. 103
USER-DEFINED System Errors

Ο Any procedure can raise an error and return a user defined error message
and error number

Ο Error number range is -20000 to -20999

Ο Range always reserved for user defined errors

Ο Oracle does not check if user defined error numbers are used uniquely

Ο Raise error within PL/SQL block with procedure


raise_application_error
(error_number,’text of the message’)
Ο Full pathname of procedure may be needed in early releases
sys.standard_utilities.
raise_application_error

. 104
USER-DEFINED System Errors
Example
CREATE PROCEDURE
fire_employee (empid NUMBER)
AS
BEGIN
IF empid <=0 THEN
raise_application_error (-20100,’Employee number must be> 0’);
ELSE
DELETE
FROM emp
WHERE EMPNO =EMPID;
END IF;
END;

SQLDBA> EXECUTE FIRE_EMPLOYEE(-1);


ORA=-20100: Employee number must be >0

. 105
Dependencies and Procedures
Ο A procedure is dependent on:

• every database object to which it refers (direct dependency)


procedures, functions, packages, tables, views, synonyms, sequences

• the database objects those objects depend on (indirect dependency)

Ο Two types of dependencies

local: objects are on the same node


remote: objects are on separate nodes

Ο ORACLE automatically checks dependencies

. 106
Recompilation of Dependent procedures
Ο When an object changes, its dependent objects are marked for
recompilation
Ο Any change to definition of referenced object implies new
version of reference object
Ο Dependent objects must be recompiled if referenced object
changes
Ο Recompilation of dependent objects takes place automatically at
runtime
Ο Reasons recompilation could fail
• Changing parameter list in called procedure
• Changing definition of or removing referenced column from referenced table
• Dropping referenced table

. 107
Recompilation

Ο Procedure/function can be recompiled be either


• RDBMS automatically, when next accessed (only if marked for
recompilation)
• Manually by the user, using ALTER PROCEDURE command

Ο If recompilation fails, error information is put to error table

. 108
Manual Recompilation

ALTER PROCEDURE

Procedure COMPILE

schema

Ο Example
ALTER PROCEDURE
add_department COMPILE

. 109
Changing a Procedure

Ο To modify a procedure, replace it:

CREATE OR REPLACE PROCEDURE


fire_employee AS . . . END;

Ο OR REPLACE option:
• Recreates the procedure even if it already exists
• Retains existing grants (need not reissue)
• Creates procedure even if there are syntax errors
• Marks dependent objects for recompilation
• Users executing old procedure finish that call: next invocation gets
new procedure
• Facilitates development (one step)
Ο CREATE without OR REPLACE on existing procedure fails

. 110
Dropping a Procedure

DROP PROCEDURE

Procedure

schema
Example
DROP PROCEDURE fire_employee;

Marks dependent objects for recompilation

. 111
Privileges for Procedures
Ο Example
GRANT EXECUTE
ON scott.hire_fire
TO mkennedy
WITH GRANT OPTION;

Ο Procedure executes under the authority of owner, not user executing


procedure

Ο User of procedure need not have access to objects inside procedure

Ο Can only GRANT privileges on an entire package, not a procedure,


function, or variable defined in the package

. 112
Privileges for Procedures

PROCEDURE system privileges apply to procedures, functions and packages

To do this Need either And


CREATE CREATE Owner must have
PROCEDURE or access to all objects
CREATE ANY referenced in the
PROCEDURE procedure
system privilege
ALTER Own the procedure
or ALTER ANY
PROCEDURE
system privilege

DROP Own the procedure


or DROP ANY
PROCEDURE
system privilege

. 113
Privileges for Procedures

To do this Need either And

Execute a Own the procedure Procedure owner


procedure or access or be granted must be explicitly
a package construct EXECUTE granted access to all
PRIVILEGE or database objects in
EXECUTE ANY the procedure(not
PROCEDURE through roles)
system privilege

. 114
Benefits of Procedures

Ο Security
• Executes under security domain of procedure’s owner
• Provides controlled indirect access to database objects to non-
privileged users
Ο Integrity
• Defines allowed operations on data
• Ensures related actions are performed together

Ο Performance
• Reduces number of calls to thedatabase
• Decreases network traffic
• Pre-parses PL/SQL statements

. 115
Benefits of Procedures
Ο Memory savings
• Takes advantages of shared SQL
• Requires only one copy of the code for multiple users

Ο Productivity
• Avoids redundant code for common procedures in multiple applications
• Reduces coding errors: no redundant code written

. 116
Benefits of Procedures
Ο Maintainability
• Enables system wide changes with one update
• Makes testing easier: duplicate testing not needed
• Dependency tracked by ORACLE

Ο High availability
• Allows changing procedured on-line while users execute
previous version

. 117
Package

Ο A database object that groups related package constructs


 Procedures
 functions
 cursor definitions
 variables and constants
 exception definitions

Ο Package comprises
Specification
Body

. 118
Parts of a Package
Ο Package specification
• Declares (specifies) package constructs, including names and parameters
publicly available procedures and functions
Ο Package body

• May declare additional, private package constructs that are not publicly
available
• Defines all package constructs (public and private)
• May be replaced without affecting package specification (Breaks
dependency chain)
• Each session has own version of state

. 119
Public and Private Package Constructs

Ο Public package constructs

• Declared in the package specification


• Available to anyone who can run the package

Ο Private package constructs


• Only declared in the package body
• Available to other package constructs within the package body
• Allows hiding procedures from programmers referencing the public
constructs

. 120
Public and Private Package Constructs

Public PACKAGE hire_fire IS


declarations PROCEDURE hire_employee (. .);
PROCEDURE fire_employee (. .);
valid CHAR(1);
END;

PACKAGE BODY hire_fire IS


PROCEDURE hire_employee (. .);
Definition of IS
BEGIN. . . END;
public constructs
PROCEDURE fire_employee ( . .)
IS
BEGIN . . . END;

FUNCTION check_num (. . )
Definition of RETURN. .
private IS
function BEGIN . . END;

END;

. 121
Public and Private Package Constructs

Package PK Procedure A
Package
Variable C specification
x :=pk.c;

Variable D
Package
body
Procedure B

y :=d;
z :=c;

. 122
Uses of Packages
Ο Group related constructs
Ο Declare globally accessible variables
Ο Declare variables with persistent state
Ο Organize development activity
• Define modules, collections of procedures known to on team
Ο Minimize name conflicts within a schema
• Personnel.audit  inventory.audit
Ο Simplify security
• GRANT EXECUTE on entire package
Ο limit recompilation
• Change body of procedure or function without changing specification

. 123
Creating a Package Specification
Example
/* Package specification declaring procedures and variables for
hiring and firing employees */
CREATE PACKAGE hire_fire
AS
/*Declare procedures */
PROCEDURE hire_employee
(empno NUMBER, ename CHAR,
mgr NUMBER, sal NUMBER,
comm NUMBER, deptno NUMBER);
PROCEDURE fire_employee (empid NUMBER);
/*Global variable declaration*/
valid CHAR(1);
END hire_fire;

. 124
Creating a package Body
Example
/*package body containing procedures to hire and ire employee */
CREATE PACKAGE BODY hire_fire
AS
/*Procedure to hire a employee*/
PROCEDURE hire_employee
(empno NUMBER, ename CHAR,
mgr NUMBER, sal NUMBER,
comm NUMBER, deptno NUMBER);
IS
BEGIN
/*VALID is declared in the package definition*/
valid :=check_sum(empno);
/*IF valid empno then add user*/
IF valid =‘T’ THEN
INSERT INTO EMP VALUES
(empno,ename,mgr,sal,comm,deptno);
END IF;
END; (continued)

. 125
Creating a package body
Example(continued)
/* Procedure to fire an employee number */
PROCEDURE fire_employee (empid NUMBER)
IS
BEGIN
DELETE FROM emp
WHERE empno =empid;
END;
/*function to check that the employee number>0.Local to the package */
FUNCTION check_sum (empno NUMBER)
RETURN CHAR(1) IS
answer CHAR(1);
BEGIN
answer :=‘T’;
IF empno <0 THEN
answer := ‘F’;
END IF;
RETURN (answer);
END;
END hire_fire; /*End of package*/

. 126
Accessing package constructs

PROCEDURE employee_admin
/* The action to perform and the employee ID have been entered previously*/
IF action =“HIRE”THEN
scott.hire_fire.hire_employee
( employee,ename,mgr,sal,comm,deptno);
IF scott.hire_fire.valid =‘T’ THEN
/*sports_club is another package that handles membership to the company
sports club*/
sports_club.add (employee)
END IF;
ELSIF action =“FIRE” THEN
scott.hire_fire.fire_employee
(employee);
sports_club.remove (employee);
END IF;

. 127
Dropping a Package

DROP PACKAGE

Procedure

schema

Example
DROP PACKAGE hire_fire;

. 128
Benefit Of Package
Performance
•Reduces disk I/O for subsequent calls
- First call to package loads whole package into
memory

Persistence state
•Retain values of package constructs for
an entire session
Security
•Access to the package allows access to
public constructs in the package only.
•No need to issue GRANT for every
procedure in package.
. 129
Benefit Of Package

Ο Productivity
•Stores related procedures and function
together
•Easier to manger changing the specification
or definition of a construct.
•Reduces cascading dependencies

•Dependencies are at package ,not


procedure level
•Can change package body with
changing or affecting package
specification

. 130
PL/SQL

Triggers

. 131
Triggers

Ο Definition
Ο Creating Triggers
Ο Restrictions on Triggers
Ο Dropping and recompiling
Ο Privileges for creating
Ο Applications
Ο Benefits

. 132
Trigger Definition
Database
Application
UPDATE t UPDATE(trigger)

SET ….;
Table T
INSERT
INSERT(trigger)
INTO t…..;
DELETE
FROM t…;
DELETE(trigger)

. 133
What is a Trigger

Ο A user-defined PL/SQL block associated with a specific table,


and implicitly fired (executed) when a triggering statement is issued
against the table
Ο Made up of parts
- Triggering event (INSERT/UPDATE/DELETE)
- Trigger type (BEFORE/AFTER, per statement or per row)
- Trigger restriction (optional)
* WHEN clause
- Trigger action
* PL/SQL BLOCK
Ο Not the same as a SQL * Forms trigger

. 134
E.g. of a Trigger - Keeping salary in range for a job

CREATE TRIGGER scott.salary_check


BEFORE INSERT OR UPDATE sal, job ON scott.emp
FOR EACH ROW
WHEN (NEW .job <> ‘PRESIDENT’)
DECLARE
minsal NUMBER;
maxsal NUMBER;
BEGIN /* get min and max salaries for the employee’s job from the SAL_GUIDE*/
SELECT minsal,maxsal INTO minsal,maxsal FROM sal_guide
WHERE job = :NEW.job;
/* If salary below min or above max,generate an error*/
IF (:NEW.sal < minsal.OR :NEW.sal > maxsal)
THEN raise_application_error ( -20500,’salary’ || :NEW.sal|| ‘out of range for job’||
:NEW.job|| ‘for employee’|| :NEW.ENAME);
ENDIF;
END; /* End of Trigger*/

. 135
Triggers and Stored procedures
Similarities

Ο Made up of PL/SQL and SQL statements


Ο Use shared SQL areas
Ο Cannot be Changed (must be dropped and recreated )
Ο Dependencies tracked automatically by ORACLE

. 136
Triggers and Stored procedures
Differences

Ο Triggers invoked implicitly ;procedures invoked


explicitly
Ο Triggers and procedures crated with different SQL
statements
Ο No CREATE OR REPLACE TRIGGER statement
Ο Triggers are associated with tables
Ο COMMIT ,ROLLBACK,SAVEPOINT not allowed in
Triggers (nor in procedures called by Triggers)

. 137
Triggers vs. SQL*Forms Triggers
Database trigger
• Fires while statement executes
• Fires independently of and in addition to SQL *From Triggers
• Fired by SQL statement from any tool or application
• Can prevent completion of SQL statement
• Fire as each statement is executed

SQL * Form trigger


• Associated with particular form
• Only executes when from is run
• Can fire after each field is entered
• Pre/Post INSERT/UPDATE /DELETE Triggers execute when COMMIT
key is pressed

. 138
Types of Triggers

Ο Type of a trigger determines


• The time when the trigger fires
• BEFORE trigger:
before the triggering action
•AFTER trigger:
after the triggering action
• The item the trigger fires on
• Row trigger: once for each row affected by the triggering statement
• Statement trigger: once for the triggering statement, regardless of the number
rows affected

Ο Each table have up to 12 Triggers in all:


INSERT/UPDATE/DELETE)
BEFORE/AFTER)
STATEMENT/ROW)

. 139
Types of Triggers
How to use each type

Ο BEFORE statement trigger


• To initialize global variables used in Triggers
• To prevent update before it occurs
Ο BEFORE row trigger
• To compute derived fields in the new row
• To prevent updates before they occur
Ο AFTER row trigger
• For auditing (by value,by row)
• Used by ORACLE snapshot mechanism
Ο AFTER statement trigger
• For auditing

. 140
Trigger - Firing Sequence
Ο INSERT,UPDATE or DELETE is applied to table statement to execute
Ο Execute BEFORE statement trigger
Ο For each row affected by the SQL statement:
• Execute BEFORE row trigger
• Change row and perform integrity constraint checking
• Execute AFTER row trigger
Ο Execute AFTER statement trigger
Ο Complete deferred integrity constraint checking
Ο Returns to application

. 141
Expressions in Triggers
Referring to values in row Triggers

Ο To refer to the old and new values of a column in row Triggers, use
the :OLD and :NEW prefixes:
• IF :NEW.sal < :OLD.sal. THEN
Ο Notes:

• Values available in row Triggers only


• New and old both available for UPDATE
• The old value in an INSERT is NULL
• The new value in a DELETE is NULL
• BEFORE row trigger can assign values to :NEW if it is not set by UPDATE
SET clause or INSERT VALUES list
• Can replace :NEW and :OLD with other correlation names if desired
• Colon dropped in when clauses

. 142
Expressions in Triggers
Conditional Predicates

Ο If a trigger can fire on more than one type of DML


operation use pre defined PL/SQL boolean variables to
determine which caused the trigger to fire:
IF INSERTING . . .
IF UPDATING . . .
IF DELETING . . .
Ο To detect which column is being updated:

IF UPDATING (‘columnname’)

. 143
Expressions in Triggers

CREATE TRIGGER total_salary


AFTER DELETE OR INSERT OR UPDATE
OF deptno,sal ON emp
FOR EACH ROW
BEGIN
IF (DELETING) OR (UPDATING AND: OLD deptno <> :NEW deptno.)
THEN UPDATE dept
SET total_sal = total_sal -:OLD.sal
WHERE deptno. = :OLD.deptno;
END IF;
END;

. 144
Restrictions on Triggers

Ο Maximum number of 12 Triggers for a table


• Up to three (INSERT/UPDATE/DELETE) Triggers of each type

Ο Prohibited statements in Triggers:


• ROLLBACK
• COMMIT
• SAVEPOINT
N.B. Also applies to procedures called by Triggers(including remote procedures

Ο Cascading trigger limit


• The action of a trigger may cause another trigger to fire and so on (called
“cascading Triggers”)
• Limit the depth of cascading with an INIT.ORA parameter

. 145
Restrictions on Triggers
Mutating tables
Original EMP
ENAME SAL
SMITH 1000
UPDATE emp
JONES 1000
SET sal = sal *1.1;

mutating EMP UPDATE(trigger)


ENAME SAL
SMITH 1100
SELECT sal
JONES 1000 FROM emp
WHERE

Ο A table that is being modified by an UPDATE, DELETE, or INSERT in a single


user statement
Ο A trigger cannot SELECT from or change a mutating table (except current row,
using :NEW and :OLD)

. 146
Restrictions on Triggers
Changes to updated/inserted values
EMPNO DEPTNO
0450 20
UPDATE emp
0407 10
SET deptno=10
WHERE empno =0405;

mutating EMP
EMPNO DEPTNO GRATE TRIGGER bad
0450 10 UPDATE BEGIN
0407 10 NEW.deptno:=30
(trigger) END;

A trigger cannot change values explicitly referenced in the


UPDATE statement SET clause or INSERT statement
. 147
Enabling and disabling Triggers

Ο Possible states for a trigger


• Enabled
• Executes its triggered action if both:
an appropriate statement is issued.
trigger WHEN clause evaluates to TRUE(if present).

• Disabled
•Does not execute its triggered action

Ο Triggers are automatically enabled on creation

. 148
Enabling and disabling Triggers
Reasons for disabling the trigger

Ο Have to load a large amount of data, and want to proceed quickly


without firing Triggers
• Example: SQL*Loader using direct path automatically disables Triggers

Ο Want to INSERT, UPDATE or DELETE on a table whose trigger


references a database object that is not available

. 149
Enabling and disabling Triggers
With ALTER TRIGGER

ALTER TRIGGER ENABLE


trigger

schema DISABLE

Examples

ALTER TRIGGER reorder DISABLE;


ALTER TRIGGER reorder ENABLE;

. 150
Enabling and disabling Triggers
With ALTER TABLE

ALTER TABLE table

schema

ENABLE trigger trigger

DISABLE schema

• Examples
ALTER TABLE INVENTORY
DISABLE TRIGGER REORDER;
ALTER TABLE INVENTORY
ENABLE TRIGGER REORDER;

. 151
Dropping Triggers

DROP TRIGGER trigger

schema

Ο Example
DROP TRIGGER reorder;

Ο Triggers cannot be altered; they must be dropped and recreated

. 152
Recompiling a trigger

COMPILE
ALTER TRIGGER trigger

schema

Ο Manually recompile to resolve dependencies (same as procedures)


Ο Example
ALTER TRIGGER reorder
COMPILE;

. 153
Applications of Triggers

 Maintaining derived fields

 Implementing complex security rules

 Enforcing complex business rules

 Performing value-based auditing

 Making implied changes

 Maintaining table replication

. 154
Examples of using Triggers
Deriving column values

Ο Derive column values for a row based upon a value provided by an


INSERT or UPDATE statement.
Ο Must use BEFORE ROW trigger
• Value must be derived first so the INSERT or UPDATE statement can use it.
• Trigger must fire for each row affected.
Ο Cannot assign to new values set by triggering INSERT or UPDATE.

. 155
Examples of using Triggers
Deriving column values
Emp no Ename Uppername Soundexname Job
7329 Smith Smith S530 Clerk
7499 Allen Allen A450 Salesman
7566 Jones Jones J520 Manager

CREATE TRIGGER upper_soundex


BEFORE INSERT OR UPDATE OF
ename,
uppername,
soundexname
ON emp ;
FOR EACH ROW
BEGIN
:NEW.uppername := UPPER (:NEW.ename);
:NEW.soundexname := :SOUNDEX(:NEW.ename);
END;

. 156
Examples of using Triggers
Complex Security Authorization

Ο Allows more complex security checking than provided by ORACLE

• Examples
- Check for time of day,day of week
- Check for terminal being used
- Permit updates of certain amounts by specific users

. 157
Examples of using Triggers
Complex Security Authorization
CREATE TRIGGER emp_permit_changes
BEFORE INSERT OR DELETE OR UPDATE ON emp
DECLARE dummy INTEGER;
BEGIN
IF(TO_CHAR (sysdate,’DY’) IN(‘SAT,’SUN’))
THEN
raise_application_error(-20504,’cannot change emp table during weekend’);
END IF;
SELECT COUNT(* ) INTO dummy
FROM company_holidays
WHERE day = TRUNC(sysdate);
IF dummy>0 THEN
raise_application_error(-20505,’cannot change emp table during holiday’);
END IF;
IF (TO_NUMBER(sysdate,’HH24’)
NOT BETWEEN 8 AND 18) THEN
raise_application_error (-20506,cannot change emp table in of_hours’);
END IF;
END;

. 158
Examples of using Triggers
Enforcing complex Business Rules

Complex check constraints that are not definable using declarative constraints

Can be used to maintain data integrity across a distributed database


(declarative integrity cannot span distributed database)

Note: simple constraints are best handled by declarative constraints features

. 159
Examples of using Triggers
Enforcing complex Business Rules

CREATE TRIGGER increase_chk


BEFORE UPDATING OF sal ON emp
FOR EACH ROW
WHEN (NEW.sal <OLD.sal OR
NEW.sal >1.1 * OLD. Sal)
BEGIN
raise_application_error(-20502,
‘may not decreased salary.
Increase must be <10%’)’
END;

. 160
Examples of using Triggers
Enforcing complex Business Rules
CREATE TRIGGER scott.salary_check
BEFORE INSERT OR UPDATE OR UPDATE OF sal,
ON scott.emp
FOR EACH ROW
WHEN (NEW.job <>’PRESIDENT’)
DECLARE minsal NUMBER;
maxsal NUMBER;
BEGIN
SELECT minsal,maxsal
INTO minsal,maxsal
FROM sal_guide
WHERE job= :NEW.job ;
IF (:NEW.sal <minsal OR
:NEW.sal > maxsal)
THEN raise_application_error ( -20503,’salary’ || :NEW.sal|| ‘out of range for
job’||:NEW.job|| ‘for employee’|| :NEW.ENAME);
END IF;
END;

. 161
Examples of using Triggers
Value based auditing

Ο Auditing that cannot be accomplished with standard RDBMS auditing


features
ΟAllows
• Exclusively DML auditing
• Per row auditing
• Storage of update details
ΟHowever Triggers cannot audit:
• DDL
• SELECTs
• Logons

. 162
Examples of using Triggers
Value based auditing

CREATE TRIGGER audit_employee


AFTER INSERT OR DELETE OR UPDATE ON emp
FOR EACH ROW
BEGIN
IF auditpackage.reason IS NULL THEN
raise_application_error(-20501,’MUST specify reason for update before performing
update;use auditpackage.set_reason()”’);
END IF;
INSERT INTO audit_employee VALUES
( :OLD.ssn,
:OLD.name;
:OLD.classification,
:OLD.sal,
:NEW.ssn;
:NEW.name,
:NEW.classification,
:NEW.sal,
auditpackage.reason,user,sysdate);
END;

. 163
Examples of using Triggers
Implied data changes

PENDING_ORDERS
PART_NO ORD_QTY ORD_DATE
00234 15 15-JAN-92
00342 25 15-JAN-92

INVENTORY
PART_NO ON_HAND REORD_PT REORD_QTY
00234 34 30 15
00342 52 50 25

Ο Transparently perform an action when a triggering is executed


Ο Example: Inventory check generates a new order

. 164
Examples of using Triggers
Implied data changes

CREATE TRIGGER inv_check


AFTER UPDATE of on_hand
ON inventory
FOR EACH ROW
WHEN (NEW.ON_HAND <= NEW.reord_pt)
DECLARE x NUMBER;
BEGIN
SELECT COUNT(*)
INTO x
FROM pending_orders
WHERE pending_orders.part_no = :NEW.part_no;

IF x = 0 THEN
INSERT INTO pending_orders
VALUES
(:NEW.part_no,
:NEW.reord_qty,
SYSDATE);
END IF;
END;

. 165
Examples of using Triggers
Synchronous Table Replication
Link identical tables(replicas) on different nodes so when replica is altered,
changes are synchronously reflected in other replicas
Replicas must contain a flag field that is set when they are updated to stop trigger
cascading

EMP_REP1 EMP_REP2

UPDATE... UPDATE...

. 166
Benefits of Triggers

Ο Security
• Allows sophisticated security checking
• Enables customized auditing mechanism to be built

Ο Integrity
• Ensures related operations on data are performed together
• Can be used to enforce complex business rules

. 167
Benefits of Triggers

Ο Performance
• Reduces number of calls to the RDBMS
• Decreases network traffic

Ο Memory savings
• Takes advantage of shared SQL
• Requires only one copy of the code for multiple users

Ο Productivity
• Requires only a single copy of the code be written and
maintained(not multiple copies in client applications)

. 168
Features of Oracle 9i &
PL/SQL

Section 4: ORDBMS

. 169
Object Oriented Concepts

Relational – Traditional RDBMS


Object-relational – Abstract Datatypes, Varying Arrays, Nested Tables
Object-oriented – Database using object-oriented analysis and design

Benefits:
Reusability
Adherence to Standards
Defined access paths

. 170
Types of Objects

Abstract Datatypes
Object Views
Varying Arrays
Nested Tables
References
Large Objects

. 171
Abstract Datatypes
Grouping related columns (subtypes) into objects
create type ADDRESS_TY as object
( street VARCHAR2(50),
city VARCHAR2(25),
state VARCHAR2(2),
zip NUMBER);

create type PERSON_TY as object


( name VARCHAR2(25),
address ADDRESS_TY);

Owner needs to grant EXECUTE privilege on the Datatype to other users


for them to use it

. 172
Abstract Datatypes
Table creation:
create table CUSTOMER
( customer_id NUMBER,
person PERSON_TY);

Table can be described using set describe depth to 2, 3 etc. in order to


view the attributes for abstract datatypes

Insert operation:
insert into CUSTOMER values
( 1, PERSON_TY(‘Peter’,
ADDRESS_TY(‘Street S’, ‘City C’, ‘ST’, 11111)));

. 173
Abstract Datatypes
Select operation:
select customer_id, c.person.name
from customer c;

select c.person.name, c.person.address.city


from customer c
where c.person.address.city like ‘F%’;

Index creation:
create index I_CUSTOMER_CITY
on CUSTOMER(person.address.city);

. 174
Object Views
The ability to overlay object oriented structures on existing relational tables
create table CUSTOMER
( customer_id NUMBER PRIMARY KEY,
name VARCHAR2(25),
street VARCHAR2(50),
city VARCHAR2(25),
state VARCHAR2(2),
zip NUMBER);

create type ADDRESS_TY as object


( street VARCHAR2(50),
city VARCHAR2(25),
state VARCHAR2(2),
zip NUMBER);

. 175
Object Views
create type PERSON_TY as object
( name VARCHAR2(25),
address ADDRESS_TY);

Object View creation:


create view CUSTOMER_OV (customer_id, person) as
select customer_id,
PERSON_TY(name,
ADDRESS_TY(street, city, state, zip))
from customer
where state = ‘DE’; /* optional */

. 176
Object Views
Relational Insert operation:
insert into CUSTOMER values
(123, ‘Richard’, ’12 Farmside Close’, ‘Lewiston’, ‘NJ’, 4352);

Object View Insert operation:


insert into CUSTOMER_OV values
(123,
PERSON_TY(‘Richard’,
ADDRESS_TY(’12 Farmside Close’, ‘Lewiston’, ‘NJ’, 4352)));

. 177
Object Views
INSTEAD OF Triggers:

BOOKSHELF_AUTHOR (title, author);


BOOKSHELF_PUBLISHER (title, publisher);

create view AUTHOR_PUBLISHER as


select a.author, a.title, p.publisher
from BOOKSHELF_AUTHOR a, BOOKSHELF_PUBLISHER p
where a.title = p.title;

update AUTHOR_PUBLISHER
set publisher = ‘MARINER’
where author = ‘HARDY’;
 Gives error ORA-01779

. 178
Object Views
create trigger AUTHOR_PUBLISHER_UPDATE /* INSTEAD OF Trigger */
instead of UPDATE on AUTHOR_PUBLISHER
for each row
begin
if :old.publisher <> :new.publisher then
update BOOKSHELF_PUBLISHER
set publisher = :new.publisher
where title = :old.title;
end if;
if :old.author <> :new.author then
update BOOKSHELF_AUTHOR
set author = :new.author
where title = :old.title;
end if;
end;

. 179
Object Views
Methods:

create type ANIMAL_TY as object


( breed VARCHARE2(25),
name VARCHAR2(25),
birthdate DATE,
member function age(birthdate IN DATE) return NUMBER);

create type body ANIMAL_TY as


member function age(birthdate DATE) return NUMBER is
begin
return ROUND(sysdate – birthdate);
end;
end;

. 180
Object Views
Data Security:

create type ANIMAL_TY as object


( breed VARCHARE2(25),
name VARCHAR2(25),
birthdate DATE,
member function age(birthdate IN DATE) return NUMBER,
PRAGMA RESTRICT_REFERENCES(age, WNDS));

WNDS = Write No Database State


RNDS = Read No Database State
WNPS = Write No Package State
RNPS = Read No Package State

. 181
Object Views
Using Methods:
create table ANIMAL
( Id NUMBER,
animal ANIMAL_TY);

insert into ANIMAL values


(11, ANIMAL_TY(‘Cow’, ‘Mimi’, TO_DATE(’01-JAN-1998’, ‘DD-MM-YYYY’)));

select a.animal.age(a.animal.birthdate)
from animal;

. 182
Varying Arrays
Allows storage of repeating attributes of a record in a single row

create type TOOL_TY as object


( toolname VARCHAR2(25));

create type TOOLS_VA as varray(5) of TOOL_TY; /* Using abstract datatype */


create type TOOLS_VA as varray(5) of VARCHAR2(25); /* Using base datatype */

create table BORROWER


( name VARCHAR2(25) PRIMARY KEY,
tools TOOLS_VA);

Data Dictionary Tables for more information:


USER_TYPES
USER_COLL_TYPES

. 183
Varying Arrays
Insert Operation:
insert into BORROWER values
(‘AMY’,
TOOLS_VA(‘Hammer’, ‘Sledge’, ‘Axe’));

insert into BORROWER values


(‘DAVID’,
TOOLS_VA(‘Hammer’, ‘Axe’, NULL, NULL, NULL));

. 184
Varying Arrays
Select Operation
Using PL/SQL:
declare
cursor borrower_cur is select * from BORROWER;
begin
for borrower_rec in borrower_cur
loop
dbms_output.put_line(‘Contact Name: ‘ || borrower_rec.name);
for i in 1..borrower_rec.tools.count
loop
dbms_output.put_line(‘Contact Name: ‘ || borrower_rec.tools(i));
end loop;
end loop;
end;

. 185
Varying Arrays
Select Operation
Using Table Function:

select b.name, n.*


from BORROWER b, TABLE(b.tools) n;

Output:

NAME COLUMN_VALUE
AMY HAMMER
AMY SLEDGE
AMY AXE
DAVID HAMMER
DAVID AXE

. 186
Nested Tables
A Table within a Table
create type ANIMAL_TY as object
( breed VARCHAR2(25),
name VARCHAR2(25),
birthdate DATE);

create type ANIMAL_NT as table of ANIMAL_TY;

create table BREEDER


( breedername VARCHAR2(25),
animals ANIMAL_NT)
nested table ANIMALS store as ANIMALS_NT_TAB;

. 187
Nested Tables
Insert Operation:
insert into BREEDER values
(‘JANE’,
ANIMALS_NT(
ANIMAL_TY(‘DOG’, ‘SCOOBY’, ’31-MAR-01’),
ANIMAL_TY(‘DOG’, ‘BUZO’, ’05-JUN-01’),
ANIMAL_TY(‘DOG’, ‘KATIE’, ’10-JUL-01’)
));

Using Table Function:


insert into TABLE(select animals from BREEDER where breedername = ‘JANE’)
values
(ANIMAL_TY(‘CAT’, ‘JEMIMA’, ’01-AUG-02’));

. 188
Nested Tables
Select Operation:
Using Table Function:
select breedername, n.name, n.birthdate
from BREEDER, TABLE(breeder.animals) n;

select breedername, n.name, n.birthdate


from BREEDER, TABLE(breeder.animals) n
where n.name = ‘SCOOBY’;

. 189
References
Column Objects vs. Row Objects:

Column objects are embedded objects represented as columns in tables


Based on extensions of features already in database
Accessed via main table
E.g. Abstract datatypes, varying arrays, nested tables or LOBs

Row objects are referenced objects accessible via references from other objects
Used to create references between rows of different tables
Not embedded in main table.
Main table contains a reference to another table
E.g. Object tables

. 190
References
Object Tables and OIDs
Each row is an object
Each row has an Object Identifier (OID) Value assigned by Oracle during creation of the row
Rows can be referenced by other objects within the database using OID
OIDs are not re-used by Oracle

create type ANIMAL_TY as object


( breed VARCHAR2(25),
name VARCHAR2(25),
birthdate DATE);

create table ANIMAL of ANIMAL_TY;

. 191
References
Insert Operation:
insert into ANIMAL values

( ANIMAL_TY(‘MULE’, ‘FRANCES’, ’01-APR-02’));

Select Operation:
select name from ANIMAL;

Update Operation:
update ANIMAL

set birthdate = ’01-MAY-01’

where name = ‘FRANCES’;

Delete Operation:
delete from ANIMAL where name = ‘FRANCES’;

. 192
References
REF Function:
Allows reference to existing row objects
Allows selection of OID assigned to each row
Takes as input alias given to the object table and returns reference value (OID)

select REF(a)
from ANIMAL a
where name = ‘FRANCES’;

REF(a)
------------------------------------------------------------------------------------
0000280209FFD8EC317DA111D6B008444456464GF2452525032

. 193
References
DEREF Function:
Opposite of REF
Takes as input a reference value (OID) and returns the value of the row object

create type ANIMAL_TY as object


( breed VARCHAR2(25),
name VARCHAR2(25),
birthdate DATE);

create table ANIMAL of ANIMAL_TY;

create table KEEPER


( keepername VARCHAR2(25),
animalkept REF ANIMAL_TY);

. 194
References
DEREF Function:
insert into KEEPER
select ‘CATHERINE’, REF(a)
from ANIMAL a
where name = ‘FRANCES’;

select * from KEEPER;


Output:
CATHERINE 0000280209FFD8EC317DA111D6B008444456464GF2452525032

select DEREF(k.animalkept)
from KEEPER k
where keepername = ‘CATHERINE’;
Output:
ANIMAL_TY(‘MULE’, ‘FRANCES’, ’01-APR-02’)

. 195
References
VALUE Function:
To see the same structures from a query of the ANIMAL object table

select VALUE(a)
From ANIMAL a
where name = ‘FRANCES’;
Output:
ANIMAL_TY(‘MULE’, ‘FRANCES’, ’01-APR-02’)

Invalid Reference:
delete from ANIMAL
where name = ‘FRANCES’;
=> Produces a dangling REF

. 196
References
FK relationships vs. References:
create table CUSTOMER
( customer_id NUMBER constraint CUSTOMER_PK PRIMARY KEY,
name VARCHAR2(25),
street VARCHAR2(50),
city VARCHAR2(25),
state VARCHAR2(2),
zip NUMBER);

create table CUSTOMER_CALL


( customer_id NUMBER,
call_number NUMBER,
call_date DATE,
constraint CUSTOMER_CALL_PK PRIMARY KEY(customer_id, call_number),
constraint CUSTOMER_CALL_FK FOREIGN KEY(customer_id) REFERENCES
CUSTOMER(customer_id));

. 197
References
FK relationships vs. References:
create type CUSTOMER_TY as object
( customer_id NUMBER,
name VARCHAR2(25),
street VARCHAR2(50),
city VARCHAR2(25),
state VARCHAR2(2),
zip NUMBER);

create view CUSTOMER_OV of CUSTOMER_TY


with object identifier (customer_id) as
select customer_id, name, street, city, state, zip from CUSTOMER;

. 198
References
FK relationships vs. References:
create view CUSTOMER_CALL as
select MAKE_REF(CUSTOMER_OV, customer_id) customer_id, call_number, call_date
from CUSTOMER_CALL;

select DEREF(ccov.customer_id)
from CUSTOMER_CALL_OV ccov
where call_date = trunc(sysdate);

. 199
Object PL/SQL
declare
cust1 CUSTOMER_TY;
begin
select VALUE(cov) into cust1 from CUSTOMER_OV cov where customer_id = 1;
dbms_output.put_line(cust1.name);
dbms_output.put_line(cust1.street);
end;

declare
newCust CUSTOMER_TY;
begin
newCust := CUSTOMER_TY(345, ‘NewCust’, ‘Street’, ‘City’, ‘ST’, 0);
insert into CUSTOMER_OV values(newCust);
end;

. 200
Large Objects
LONG – character data, 2 GB
LONG RAW – binary data, 2 GB
CLOB – character LOB, 4 GB
BLOB – binary LOB, 4 GB
NCLOB – CLOB with multi-byte character set support
BFILE – binary file, read-only binary data stored outside database, size OS dependent

create table PROPOSAL


( proposal_id NUMBER(10) PRIMARY KEY,
recipient_name VARCHAR2(25),
proposal_name VARCHAR2(25),
short_description VARCHAR2(1000),
proposal_text CLOB,
budget BLOB,
cover_letter BFILE)

. 201
Large Objects
storage(INITIAL 50K NEXT 50K PCTINCREASE 0)
tablespace PROPOSALS
LOB(proposal_text, budget) store as
( tablespace PROPOSAL_LOB
storage(INITIAL 100K NEXT 100K PCTINCREASE 0)
CHUNK 16K PCTVERSION 10 NOCACHE LOGGING);

CHUNK – Space allocated for each LOB manipulation. Default 1K, upto 32 K
PCTVERSION – Max % of overall LOB storage space used for creating new version of the LOB
NOCACHE – Not stored in memory
LOGGING – All operations against the LOB will be recorded in redo log files

. 202
Large Objects
Initialization:
BLOB – EMPTY_BLOB( )
CLOB/NCLOB – EMPTY_CLOB( )
BFILE – BFILENAME(<directory name>, <file name>);

create directory proposal_dir as ‘/u01/proposal/letters’;


insert into PROPOSAL values
(2, ‘M/S ABC’, ‘Energy Billing System’, NULL, EMPTY_CLOB(), EMPTY_BLOB(),
BFILENAME(proposal_dir, ‘p1.doc’));

Using Subqueries:
insert into PROPOSAL
select 3, ‘M/S XYZ’, ‘Credit Card Billing System’, NULL, proposal_text, budget, cover_letter
from PROPOSAL
where proposal_id = 1;

. 203
Large Objects
String Functions:
SUBSTR – select SUBSTR(proposal_text, 1, 10) from PROPOSAL where proposal_id = 1;
INSTR – select INSTR(proposal_text, ‘new’, 1, 1) from PROPOSAL where proposal_id = 3;
INITCAP – select INITCAP(proposal_text) from PROPOSAL where proposal_id = 3;
LTRIM/RTRIM – select LTRIM(proposal_text) from PROPOSAL where proposal_id = 1;

DBMS_LOB Package:
READ, WRITE
COPY, APPEND
ERASE
SUBSTR, INSTR, GETLENGTH, TRIM
COMPARE

. 204
Large Objects
READ:
declare
locator_var CLOB;
amount_var number;
offset_var number;
output_var varchar2(10);
begin
amount_var := 10;
offset_var := 1;
select proposal_text into locator_var
from PROPOSAL
where proposal_id = 1;
DBMS_LOB.READ(locator_var, amount_var, offset_var, output_var);
DBMS_OUTPUT.PUT_LINE(‘Start of proposal text: ‘ || output_var);
end;

. 205
Large Objects
WRITE:
declare
locator_var CLOB;
amount_var number;
offset_var number;
buffer_var varchar2(12);
begin
amount_var := 12;
offset_var := 10;
buffer_var := ‘ADD NEW TEXT’;
select proposal_text into locator_var
from PROPOSAL
where proposal_id = 3 for update;
DBMS_LOB.WRITE(locator_var, amount_var, offset_var, buffer_var);
end;

. 206
. 207

Das könnte Ihnen auch gefallen