(ref) PL/SQL Optimization

[ Oracle PL/SQL ]

Improving comunication: BULK SQL (FORALL and BULK COLLECT)

Topics

BULK SQL:
  • Minimizes overhead in PL/SQL <=> SQL comunication.
  • FORALL: Send statements in batches.
  • FORALL: faster than equivalent FOR LOOP.
  • FORALL: Can contain ONLY one DML statement.
  • SQL%BULK_ROWCOUNT: Is like an AA. keeps the # of rows affected by each DML in the FORALL statement.
  • SQL%BULK_EXCEPTIONS.[COUNT | ERROR_INDEX | ERROR_CODE]: Is like an AA. keeps information about EXCEPTIONS that occurred during the FORALL statement.

  • BULK COLLECT: Receive results in batches
  • Good when query/DML affects +4 database rows



FOR LOOP X FORALL
SQL> set serveroutput on
SQL> DECLARE
 TYPE NumList IS VARRAY(20) of number;
  -- depts to be deleted
 depts NumList := NumList(10,30,70); 
BEGIN
  FOR i IN depts.FIRST..depts.LAST LOOP
   dbms_output.put_line('deleting dept '||i);
   delete from emp_temp
   where department_id = depts(i);
   END LOOP;
END;
/
deleting dept 1
deleting dept 2
deleting dept 3 

PL/SQL procedure successfully completed.
SQL>

  • FORALL: Only one DML Statement.
  • Use SQL%BULK_ROWCOUNT to check # of affected rows.
SQL> set serveroutput on
SQL> DECLARE
 TYPE NumList IS VARRAY(20) of number;
 depts NumList := NumList(10,30,70);
BEGIN
  FORALL i IN depts.FIRST..depts.LAST
    delete from emp_temp
    where department_id = depts(i);

  -- How many rows were affected by
  -- each DELETE statement?
  FOR i IN depts.FIRST..depts.LAST LOOP
   DBMS_OUTPUT.PUT_LINE('Iteration #' 
     || i || ' deleted ' ||
     SQL%BULK_ROWCOUNT(i) ||'rows.');
  END LOOP;
END;
/
Iteration #1 deleted 1 rows.
Iteration #2 deleted 2 rows.
Iteration #3 deleted 5 rows.

PL/SQL procedure successfully completed.
SQL>


Comparing INSERT performance: FOR LOOP X FORALL
set serveroutput on
drop table parts1;
create table parts1 (pnum integer, pname varchar2(15));
drop table parts2;
create table parts2 (pnum integer, pname varchar2(15));

DECLARE
  TYPE NumTab IS TABLE OF parts1.pnum%type INDEX BY PLS_INTEGER;
  TYPE NameTab IS TABLE OF parts1.pname%type INDEX BY PLS_INTEGER;

  pnums NumTab;
  pnames NameTab;
  iterations CONSTANT PLS_INTEGER := 50000;
  t1 INTEGER;
  t2 INTEGER;
  t3 INTEGER;

BEGIN
 --populate collection
 FOR j IN 1..iterations LOOP
   pnums(j)  := j;
   pnames(j) := 'Part No. ' || to_char(j);
 END LOOP;

 -- get t1 before start FOR LOOP
 t1 := dbms_utility.get_time;
 
 FOR i IN 1..iterations LOOP
   insert into parts1 (pnum, pname)
   values (pnums(i), pnames(i));
 END LOOP;

 -- get t2 before start FORALL 
 t2 := dbms_utility.get_time;

 FORALL i IN 1..iterations
   insert into parts2(pnum, pname)
   values (pnums(i), pnames(i));

 t3 := dbms_utility.get_time;

 dbms_output.put_line('Execution Time (secs)');
 dbms_output.put_line('---------------------');
 dbms_output.put_line('FOR LOOP: ' || to_char((t2-t1)/100));
 dbms_output.put_line('FORALL:  ' || to_char((t3-t2)/100));
END;
/
anonymous block completed
Execution Time (secs)
---------------------
FOR LOOP: 1.74
FORALL:  .06

PLS-00436: Implementation restriction - Cannot reference RECORD fields within a FORALL
(...)
DECLARE
  TYPE PartsRec IS RECORD (pnum parts1.pnum%type, name parts1.pname%type);
  TYPE PartsRecTab IS TABLE OF PartsRec INDEX BY PLS_INTEGER;
  precs PartsRecTab;
  (...)
BEGIN
 --populate collection
 FOR j IN 1..iterations LOOP
   precs(j).pnum  := j;
   precs(j).pname := 'Part No. ' || to_char(j);
 END LOOP;
 
 (...)

 FORALL i IN 1..iterations
   insert into parts2(pnum, pname)
   values (precs(i).pnum, precs(i).pname);
 (...)
END;
/

Error report:
ORA-06550: line 31, column 12:
PLS-00436: implementation restriction: cannot reference fields of BULK In-BIND 
                                                              table of records





Example 12-10 FORALL Statement for Subset of Collection


Exception Handling


Bulk Statements: BULK COLLECT


  • BULK COLLECT: use the BULK COLLECT: statement with SELECT statements
    • Can be used (a) inside a SQL statement (implicit cursor) or (b) as part of a FETCH statement (explicit cursor).
    • With FETCH: you can append the LIMIT clause to set the maximum number of rows read from the cursor at a time.
    BULK COLLECT Syntax:
    DECLARE
    -- on a SQL Statement:
    SELECT column1 [, column2 [,...]]
    COLLECT BULK INTO collection1 [, collection2 [,...]]
    FROM
    table_name
    [WHERE where_clause_statements];
    
    -- with explicit cursor:
    FETCH cursor_name [(parameter1 [, parameter2 [,...]])]
    BULK COLLECT INTO collection1 [, collection2 [,...]]
    [LIMIT rows_to_return];
    
  • The BULK COLLECT INTO statement is much faster than a standard cursor because it has one parse, execute, and fetch cycle.
  • Scalar collections are the only supported SQL collection datatypes: No composite collections (record collections) (??).
  • FORALL: Use the FORALL: statement to INSERT, UPDATE or DELETE large data sets.
  • SQL%BULK_ROWCOUNT(i): Keeps the # of rows affected by each DML in the FORALL statement.
  • SQL%BULK_EXCEPTIONS(i).[COUNT | ERROR_INDEX | ERROR_CODE]: Keeps information about EXCEPTIONS that occurred during the FORALL statement.


set serveroutput on  
declare
  -- Using BULK COLLECT with scalar collections
  -- Here you will have to ensure that the discrete collections 
  -- remain synchronized
  type empid_c is table of employees.employee_id%type;
  type empdep_c is table of employees.department_id%type;
  type empsal_c is table of employees.salary%type;
  
  empid_ empid_c;
  empdep_ empdep_c;
  empsal_ empsal_c;
begin 
  -- selecting into the three collections, in parallel.
  -- (Q) why would you do such a thing???
  -- (A) The typical reason to opt for parallel collections is 
  --           to move the data from PL/SQL to external
  --           programming languages or web applications.
  
  select employee_id, department_id, salary
  bulk collect into empid_, empdep_, empsal_
  from employees;
  
  for i in 1..empid_.count loop
    dbms_output.put_line('id: ' || empid_(i) ||
                         ' earns: '|| empsal_(i));
  end loop;
end;

anonymous block completed
id: 100 earns: 24000
id: 101 earns: 17000
...


-- Alternativelly, you may use structured collections. 
-- However, these cannot be shared with external programs.
set serveroutput on  
declare
  -- define one list of records. Synchronization is given.
  type emprec is record(empid employees.employee_id%type, 
                        dept employees.department_id%type,
                        sal  employees.salary%type);                     
  type emptab is table of emprec;
  emptab_ emptab;

begin 
  -- collect directly into the structured table
  select employee_id, department_id, salary
  bulk collect into emptab_
  from employees;
  
  for i in 1..emptab_.count loop
    dbms_output.put_line('id: ' || emptab_(i).empid ||
                         ' earns: '|| emptab_(i).sal);
  end loop;
end;
anonymous block completed
id: 100 earns: 24000
id: 101 earns: 17000
...



-- Now using an explicit cursor instead of SQL Stmt
-- Here the BULK COLLECT is used on the FETCH statement.
set serveroutput on  
declare
  -- As before, (why?), using parallel scalar collections.
  -- remember they will need to be pretty well synchronized..
  type empid_c is table of employees.employee_id%type;
  type empdep_c is table of employees.department_id%type;
  type empsal_c is table of employees.salary%type;
  
  empid_ empid_c;
  empdep_ empdep_c;
  empsal_ empsal_c;

  -- here define the explicit cursor c1
  cursor c1 is 
   select employee_id, department_id, salary
   from employees;
  j pls_integer := 0; 
   
begin 
  open c1;
  loop
    -- now when you fetch the cursor records,
    -- use the BULK COLLECT clause and you may LIMIT the
    -- number of rows fetched...
    fetch c1 bulk collect into empid_, empdep_, empsal_ limit 10;
    exit when empid_.count = 0;
    j := j + 1;
    dbms_output.put_line('Fetch # ' || j);
    for i in 1..empid_.count loop
       dbms_output.put_line('id: ' || empid_(i) ||
                         ' earns: '|| empsal_(i));
    end loop;
  end loop;
end;

anonymous block completed
Fetch # 1
id: 100 earns: 24000
...
Fetch # 2
id: 110 earns: 8200
...
Fetch # 3
id: 120 earns: 8000
...









(7) Composite Data Types

[ Oracle PL/SQL ]
Composite data types: COllections and Records


Composite data type:
  • Stores values that have internal components (scalar or composite).
  • Collections can be built of any SQL or PL/SQL datatype.
  • Fall into two categories: arrays(size allocated at definition) and lists.
  • Lists are called associative arrays when they can be indexed by non-sequential numbers or unique strings.
  • Multidimensional collections can be built as PL/SQL datatypes. In this case, collection elements are record structures.

Collection:
  • Internal components always have the same data type (elements).
  • Access syntax: variable_name(index)
  • To create a collection variable:
    1. Define a collection type
    2. Create a variable of that type
    OR use %TYPE.

Collection Types:
  • Associative Arrays (former PL/SQL table or index-by table)
  • Variable-size Arrays (VARRAY)
  • Nested Table



Number of Elements: { Specified | Unspecified }
  • Whether you need to define the maximum size of the collection at declaration time.
  • Specified: VARRAY; Unspecified: Associative array; Nested table

Uninitialized status: { Empty | Null }
  • Empty: Exists, but has no elements. (Use EXTEND method to add).
  • Null: Does not exist. Must be initialized (Made empty or assigned a value).
  • Null: VARRAY, Nested Table. Empty: Associative Array

Scope: { Local | Public | Standalone (Schema) }
  • Local: Defined and available only within a PL/SQL block.
  • Public: Defined in a package specification.(ref: package_name.type_name)
  • Standalone: Defined at schema level.(ref: CREATE TYPE)


Collections: Associative Array (former PL/SQL TABLE)

  • Name introduced in Oracle 10g
  • Oracle 8 - 9i: index-by tables
  • Oracle 7: Pl/SQL tables
  • Can be indexed by Integer or String (no order here)
  • Can be passed as parameter and returned by function
  • Can be declared as a CONSTANT
  • The AA is only a PL/SQL datatype: can only be referred in a PL/SQL scope.
  • AAs are typically defined in package specifications when they are to be used externally from an anonymous or named block program.
  • AAs Numeric indexes do not need to be sequential.
  • These are sparsely populated structures: can have gaps in index sequences.
  • AAs are dynamically sized (like the NESTED TABLE datatype).

Who has the highest salary in each department?

(a) Define a CURSOR to extract:
[ Dept Name, Employee Name, Max Salary ]
(b) Define a RECORD TYPE to hold:
[ Employee Name, Max Salary ]
(c) Define an ASSOCIATIVE ARRAY TYPE (IS TABLE OF) indexed by String.
(d) Declares a variable of the ASSOCIATIVE ARRAY TYPE
(e) Open the CURSOR and populate the variable with data from.
(f) Print the data.

SQL> Set serveroutput on

SQL> DECLARE
   -- Declare cursor ccur to extract data (dname, ename, maxsal)
 Cursor ccur IS
  select d.department_name dname, 
         e.first_name ||' '|| e.last_name ename, e.salary maxsal
  from employees e, 
       ( select department_id, max(salary) msal
        from employees
        group by department_id) s, 
       departments d
  where e.department_id = d.department_id
   and e.department_id = s.department_id
   and e.salary = s.msal
  order by e.department_id;
  
    -- Define a Record type to hold (ename, maxsal)
  TYPE EmpSalRec IS RECORD (ename varchar2(60), maxsal employees.salary%TYPE);

   -- Define an associative array indexed by department name. The array stores the
   -- the composite structure (ename, maxsal)
  TYPE MaxSalDep IS TABLE OF EmpSalRec INDEX BY departments.department_name%TYPE;
  
   -- Declare a variable of the associative array type.
  v_MaxSalDep MaxSalDep;
  
BEGIN
 FOR deps IN ccur LOOP
    v_MaxSalDep(deps.dname).ename := deps.ename;
    v_MaxSalDep(deps.dname).maxsal := deps.maxsal;
    dbms_output.put_line('MaxSal Dept '      || deps.dname || ': ' || 
                          ' (' || v_MaxSalDep(deps.dname).ename || 
                          ', ' || v_MaxSalDep(deps.dname).maxsal || ');'
                         );
 END LOOP;
END;
/

MaxSal Dept Administration:  (Mary Fansom, 25000);
MaxSal Dept Marketing:  (Michael Hartstein, 13000);
MaxSal Dept Purchasing:  (Den Raphaely, 11000);
MaxSal Dept Human Resources:  (Kimberely Grant, 7000);
MaxSal Dept Shipping:  (Adam Fripp, 8200);
MaxSal Dept IT:  (Alexander Hunold, 9000);
MaxSal Dept Public Relations:  (John Fergunson, 26000);
MaxSal Dept Sales:  (John Russell, 14000);
MaxSal Dept Executive:  (Steven King, 24000);
MaxSal Dept Finance:  (Nancy Greenberg, 12000);
MaxSal Dept Accounting:  (Shelley Higgins, 12000);

Using Associative Array as a Constant
  • The example below loads an Associative Array Constant with the names of all departments.
  • Using the AA Constant, it then lists the number of employees in each department.
  • The AA Constant will be defined with the scope of a package.

(a) In the Package Spec, define the specification of the AA Constant TYPE and the specification of the Function that will initialize the AA Constant.
(b) In the Package Body, write the initialization function.
(c) In an anonymous block, declare constant, intialize it, and loop through the array.

SQL> Set serveroutput on

SQL> CREATE OR REPLACE PACKAGE My_Types AUTHID DEFINER IS
   -- Declaring an AA in a package spec makes it available for invoked and invoking 
   -- subprograms and to anonymous blocks.
  TYPE AA_FixDept IS TABLE OF departments.department_name%TYPE
                    INDEX BY pls_integer;
  FUNCTION Init_AA_FixDept RETURN AA_FixDept;
END My_Types;
/
Package created.

SQL> CREATE OR REPLACE PACKAGE BODY My_Types IS
    -- Function Init_AA_FixDept will initialize the AA Constant.
  FUNCTION Init_AA_FixDept RETURN AA_FixDept IS    
    CURSOR ccur IS
      SELECT department_name FROM departments;
    Ret AA_FixDept;
    i PLS_INTEGER := 0;
    BEGIN    
     FOR j IN ccur LOOP
      Ret(i) := j.department_name;
       -- Use dbms_output to check the proper loading of the AA Constant.
       -- Comment or remove the line later.
      dbms_output.put_line( 'Loading dept '|| i || ': ' || j.department_name);
      i := i + 1;
     END LOOP;
   RETURN Ret;   
  END Init_AA_FixDept;
END My_Types;
/
Package created.

SQL> Set serveroutput on
SQL> DECLARE
   -- Declare the AA Constan c_AAdept. 
   -- The Constant should be initialized at declaration time.
 c_AAdept CONSTANT My_Types.AA_FixDept := My_Types.Init_AA_FixDept();
 CURSOR cnumep (namedep departments.department_name%TYPE) IS
    SELECT COUNT(*) 
    FROM employees e, departments d
    WHERE e.department_id = d.department_id
    and department_name = namedep;
 v_index integer;
 v_numep integer;
BEGIN
  -- Listing the number of employees in each department;
  -- Loop through each component of the AA Constant and pass the department name
  -- as parameter to the explicit cursor cnumep.
  v_index := c_AAdept.First;
  LOOP
   OPEN cnumep(c_AAdept(v_index));
   FETCH cnumep INTO v_numep;
   dbms_output.put_line('Dept ' || c_AAdept(v_index) || ' has ' || 
                        v_numep || ' employees');
   CLOSE cnumep;
   EXIT WHEN v_index = c_AAdept.Last;
   v_index := c_AAdept.next(v_index);
  END LOOP;
END;
/

(Edited ouput):
Loading dept 0: Administration
Loading dept 1: Marketing
(...)
Loading dept 26: Payroll

Dept Administration has 2 employees
Dept Marketing has 2 employees
Dept Purchasing has 6 employees
(...)
Dept Payroll has 0 employees

When to use what?

Appropriate uses for AAs:
- The physical size is unknown and
- the type will not be used in tables.
- Good for standard solutions (i.e. using maps and sets).
- Good for small lookup tables.

When passing collections App <==> DB server. Use either:
VARRAYs
- You know the maximum number of elements.
- You usually access the elements sequentially.
- The physical size of the collection is static
- The collection may be used in tables.
- Varrays are the closest thing to arrays in other programming languages (Java, C, C++, C#).

NESTED TABLES
- The physical size is unknown and the type may be used in tables.
- Nested tables are like lists and bags in other programming languages.





Collections: VARRAYS
  • Introduced in Oracle 8
  • Can be used in table, record, and object definitions.
  • Fixed max size at creation.
  • Num elements: Between 0 and declared maximum size.
  • May be stored in permanent tables and accessed by SQL.
  • If VARRAY variable is less than 4 KB: it resides inside the table of which it is a column; otherwise, it resides outside the table but in the same tablespace.
  • Must be initialized. Otherwise is a null collection.
  • Densely populated structures: they do not allow gaps in the index values.
  • VARRAY and NESTED TABLE datatypes are structures indexed by sequential integers.

Using Varray: loading values from table
  • Load and print name and hire_date of the three longest employeed workers

(a) Define a Varray of STRING with max 3 rows.
(b) Declare the varray variable. Use constructor to set it to null.
(c) Declare CURSOR to return list of data from database.
(d-f) Print empty varray. Load Varray looping through cursor. Print loaded Varray.

SQL> set serveroutput on
DECLARE
-- Define a varray with a maximum of 3 rows.
TYPE str_varray IS VARRAY(3) OF VARCHAR2(100);

-- Declare the varray with and initialize it with null values.
-- Failure to do so at this point will produce an 
-- "ORA-06531: Reference to uninitialized collection".

-- Memory and proper index values are allocated opon initialization of EACH ELEMENT.
-- If you fail to initialize ALL elements, you need to use the EXTEND method or 
-- you'll face an "ORA-06533: Subscript beyond count" at run-time.
varray_str str_varray := str_varray(NULL,NULL,NULL);

vstr varchar2(100);
i PLS_INTEGER :=1;
CURSOR ccur IS
  select ename, hire_date
  from (select first_name || last_name as ename, hire_date, 
                  row_number() over (order by hire_date) r
          from employees)
  where r between 1 and 3;

BEGIN
-- Print initialized null values.
dbms_output.put_line('Varray initialized as nulls.');
dbms_output.put_line('––––––––––––––--------------');
FOR i IN 1..3 LOOP
 dbms_output.put_line ('String Varray ['||i||'] ' ||
                                '['||varray_str(i)||']';
END LOOP;

-- Assign values to subscripted members of the varray.
FOR j in ccur LOOP
 varray_str(i) := j.ename || ' ' || to_char(j.hire_date);
 i := i + 1;
END LOOP;
-- Print initialized null values.
dbms_output.put (CHR(10));  -- Visual line break.
dbms_output.put_line('Varray initialized as values.');
dbms_output.put_line('-----------------------------');

FOR j IN 1..3 LOOP
 dbms_output.put_line('String Varray ['||j||'] ' ||
                               '['||varray_str(j)||']');
END LOOP;
END;
/

anonymous block completed
Varray initialized as nulls.
––––––––––––––--------------
String Varray [1] []
String Varray [2] []
String Varray [3] []

Varray initialized as values.
-----------------------------
String Varray [1] [StevenKing 17-JUN-87]
String Varray [2] [JenniferWhalen 17-SEP-87]
String Varray [3] [NeenaKochhar 21-SEP-89]

SQL> CREATE OR REPLACE TYPE str_varray AS 
                       VARRAY(3) of VARCHAR2(100);
/

Type created.

DECLARE
-- initialize without allocating space
varray_str str_varray := str_varray(); 
(...)
BEGIN
FOR j in ccur LOOP
 -- increment index and allocate space
 varray_str.EXTEND;
 varray_str(i) := j.ename || ' ' || to_char(j.hire_date);
 i := i + 1;
END LOOP;
(...)
END;

Alternatively:
  1. You could creat a user-defined object type.
  2. Benefits: it may be referenced from any programs that have permission to use it, whereas a PL/SQL varray type structure is limited to the program unit.
  3. You could allocate space for the VARRAY
    at run-time, as you increment the index:




Collections: Nested Tables
  • Introduced in Oracle 8
  • Initially defined as densely populated arrays, but may become sparsely populated as records are deleted.
  • May be stored in permanent tables and accessed by SQL.
  • May be dynamically extended: act more like bags and sets than arrays.
  • May contain a list of one or more compound datatypes (PL/SQL records) when they work exclusively in a PL/SQL scope.
  • In the database, a nested table column type stores an unspecified # of rows in no particular order.
  • When you retrieve a nested table value from the database into a PL/SQL nested table variable, PL/SQL gives the rows consecutive indexes, starting at 1.

(1) Define a composite NESTED TABLE TYPE (IS TABLE OF) of the type of employees table
(2) Create variable of the defined type. Initialize empty.
(3) Read all employees records on Cursor
(4) Iteract through the cursor and
(4.1) Extend the nested table
(4.2) Load three fields of the nested table with data from cursor.

set serveroutput on
DECLARE
 TYPE emp_table IS TABLE OF employees%ROWTYPE; 
 emps emp_table := emp_table();
 CURSOR ccur IS
  select employee_id, first_name, salary
  from employees;
 j PLS_INTEGER;
BEGIN
-- Print initialized null values.
dbms_output.put_line('Nested table initialized empty.');
dbms_output.put_line('-------------------------------');
j := 1;
dbms_output.put_line('Begin loading...');
FOR i IN ccur LOOP
 emps.extend;
 emps(j).employee_id := i.employee_id;
 emps(j).first_name := i.first_name;
 emps(j).salary := i.salary;
dbms_output.put ('Nested Table Extended ['||j||'] ');
dbms_output.put_line('['||emps(j).employee_id ||', '
      || emps(j).first_name || ', ' || emps(j).salary ||']'); 
j := j + 1;
END LOOP;
END;
/

anonymous block completed
Nested table initialized empty.
-------------------------------
Begin loading...
Nested Table Extended [1] [100, Steven, 24000]
Nested Table Extended [2] [101, Neena, 17000]
Nested Table Extended [3] [102, Lex, 17000]
(...)
Nested Table Extended [108] [210, John, 26000]
Nested Table Extended [109] [211, Mary, 25000]

Collection Methods
Method Type Description
DELETE Procedure Deletes elements from collection.
TRIM Procedure Deletes elements from end of varray or nested table.
EXTEND Procedure Adds elements to end of varray or nested table.
EXISTS Function Returns TRUE if and only if specified element of varray or nested table exists.
FIRST Function Returns first index in collection.
LAST Function Returns last index in collection.
COUNT Function Returns number of elements in collection.
LIMIT Function Returns maximum number of elements that collection can have.
PRIOR Function Returns index that precedes specified index.
NEXT Function Returns index that succeeds specified index.

Using collection methods
set serveroutput on 
declare 
 type nbrlast is table of varchar2(10);
  n nbrlast := nbrlast('paul','john','ringo','george');
begin
 dbms_output.put_line('Num elements held: ' || n.count);
 dbms_output.put_line('The last element is ' || n(n.last));
 n.extend;     -- append a new (null) element
 n(n.last) := 'Yoko';
 dbms_output.put_line('Added Yoko');
 dbms_output.put_line('Now, num elements held: ' || n.count);
 dbms_output.put_line('The last element now is ' || n(n.last));
 n.extend(3);  -- append two null elements to the collection
 dbms_output.put_line('appended three null elements');
 dbms_output.put_line('Now, num elements held: ' || n.count);
 n.delete(6);
 dbms_output.put_line('deleted element 6');
 dbms_output.put_line('Now, num elements held (count): ' || n.count);
 dbms_output.put_line('Now, highest subscription # (last): ' || n.last);
 n(7) := 'Ghost';
 dbms_output.put_line('element 7 is now: ' || n(7));
 dbms_output.put_line('Now, num elements: ' || n.count);

 if n.limit is null then
   dbms_output.put_line('Max num elements: unlimited.');
 end if;
 
 for i in n.first..n.last loop
   if n.exists(i) then 
     if n(i) IS NULL then
        dbms_output.put_line('Element #  '|| i || ' is NULL');
     else
        dbms_output.put_line('Element #  '|| i || ' is '|| n(i));
     end if;
   else
        dbms_output.put_line('Element #  '|| i || ' does not exist');
   end if;
  end loop;
end;
/

anonymous block completed
Num elements held: 4
The last element is george
Added Yoko
Now, num elements held: 5
The last element now is Yoko
appended three null elements
Now, num elements held: 8
deleted element 6
Now, num elements held (count): 7
Now, highest subscription # (last): 8
element 7 is now: Ghost
Now, num elements: 7
Max num elements: unlimited.
Element #  1 is paul
Element #  2 is john
Element #  3 is ringo
Element #  4 is george
Element #  5 is Yoko
Element #  6 does not exist
Element #  7 is Ghost
Element #  8 is NULL

















(ref) INSTEAD OF triggers

NON-UPDATABLE Views contain any:
  • One or more Table Joins.
  • A GROUP BY clause.
  • START WITH ...CONNECT BY.
  • DISTINCT.
  • Aggregate function (i.e. SUM, AVG, MIN, MAX, COUNT).
  • Set operators.
  • Derived columns (i.e. concatenated, pseudocolumns, etc).

INSTEAD OF triggers

INSTEAD OF triggers are created on
  • (usually) a non-updatable view
  • a nested table column of a non-updatable view

What?
  • INSTEAD OF trigger cannot be conditional.
  • Is the only way to update a view that is not inherently updatable
  • Is always a row-level trigger
  • Can read :OLD and :NEW, but cannot change them.
  • The timing events (BEFORE and AFTER) is not applicable for INSTEAD OF triggers.
  • INSTEAD OF trigger can fire for all three DML statements (I/U/D).
  • Prior to Oracle8.1.6 INSTEAD OF triggers functionality was included in the Enterprise Edition only.

Why use?
  • Used to intercept DML statements and replace them with alternative code.
  • Useful when application does not need (or should not) see the underlying tables (i.e. for security reasons). In this case:
    1. The application issues DMLs against the view
    2. The INSTEAD OF trigger intercepts and rewrite the DMLs to the underlying tables.
  • Useful also when the system uses object-relational functionality.
  • Updates against complex views may result in "ORA-01776-cannot modify more than one base table through a join view" or "ORA-01779-cannot modify a column which maps to a non key-preserved table"
How?
  • The database fires the INSTEAD OF trigger instead of running the triggering DML statement.
  • The trigger should: (a) determine what operation was intended and (b) perform the appropriate DML operations on the underlying tables.
Creation Syntax:
CREATE [OR REPLACE] TRIGGER trigger_name
 INSTEAD OF {dml_statement }
 ON {object_name | database | schema}
 FOR EACH ROW
 [DISABLE]
 [WHEN (logical_expression)]
 [DECLARE]
  declaration_statements;
BEGIN
  execution_statements;
END [trigger_name];
/
In the example below (taken from here) you:

(1) Create the base tables Custormer and Orders.
(2) Create a view order_info.
(3) Create a INSTEAD OF trigger order_info_insert that translates the DML statement into statements acting on the base tables (Orders and Customers).
(4) Insert data on the order_info view. Check the insertion on the base tables.

(1-2) Create base tables Customer, Orders and view.
SQL> CREATE TABLE customers ( 
     customer_id NUMBER (6)  NOT NULL  PRIMARY KEY, 
     cust_first_name VARCHAR2 (20)  NOT NULL , 
     cust_last_name VARCHAR2 (20)  NOT NULL , 
     cust_address VARCHAR2 (200) , 
     phone_numbers VARCHAR2 (12) , 
     nls_language VARCHAR2 (3) , 
     nls_territory VARCHAR2 (30) , 
     credit_limit NUMBER (9,2) , 
     cust_email VARCHAR2 (30) , 
     account_mgr_id NUMBER (6));

Table created.
SQL> CREATE TABLE orders ( 
     order_id NUMBER (12)  NOT NULL Primary key, 
     order_date TIMESTAMP  NOT NULL , 
     order_mode VARCHAR2 (8) , 
     customer_id NUMBER (6)  NOT NULL , 
     order_status NUMBER (2) , 
     order_total NUMBER (8,2) , 
     sales_rep_id NUMBER (6) , 
     promotion_id NUMBER (6));

Table created.
SQL> CREATE OR REPLACE VIEW order_info AS
   SELECT c.customer_id, c.cust_last_name, c.cust_first_name,
          o.order_id, o.order_date, o.order_status
   FROM customers c, orders o
   WHERE c.customer_id = o.customer_id;

View created.
SQL>

Now, if you try to perform an INSERT in the view order_info, it will fail with an ORA-01779 error:

SQL>INSERT INTO order_info 
    values (345, 'Rogers', 'John', 1250, sysdate, 1);

Error: ORA-01779: "cannot modify a column which maps to a non key-preserved table"
*Cause:    An attempt was made to insert or update columns of a join view which
           map to a non-key-preserved table.
*Action:   Modify the underlying base tables directly.


A Key preserved table means that the row from the base table will appear AT MOST ONCE in the output view on that table. Here one customer clearly may have multiple orders, so customer_id will most probably appear multiple times on order_info.

So, if you issue an INSERT against order_info, how can the database decide what to do with it?
Should it insert into orders AND customers? Only into orders? Or only into customers?

Well, you need to write an instead of trigger to intercept the DML against the trigger and issue DMLs against the base table(s) according to your business rules..

In this example, the trigger will split the INSERT coming from the application and insert on each table accordingly:


(3) Create a INSTEAD OF trigger order_info_insert that translates the DML statement into statements acting on the base tables (Orders and Customers).
SQL> CREATE OR REPLACE TRIGGER order_info_insert
  INSTEAD OF INSERT ON order_info
DECLARE
 duplicate_info EXCEPTION;
 PRAGMA EXCEPTION_INIT (duplicate_info, -00001);
BEGIN
 INSERT INTO Customers
   (customer_id, cust_last_name, cust_first_name)
 VALUES (:new.customer_id, :new.cust_last_name, :new.cust_first_name);
 INSERT INTO Orders (order_id, order_date, customer_id, order_status)
 VALUES (:new.order_id, :new.order_date, :new.customer_id, :new.order_status);
EXCEPTION
 WHEN duplicate_info THEN
   raise_application_error (
     num=> -20107,
     msg=> 'Duplicate customer or order ID');
END order_info_insert;
/
Trigger created.
SQL>

(4) Insert data on the order_info view. Check the insertion on the base tables.
SQL> INSERT into order_info 
 (Customer_id, cust_last_name, cust_first_name, order_id, order_date, order_status)
VALUES
 (1, 'Smith', 'John', 2500, '13-AUG-09', 0);
1 rows inserted.


SQL> Select Customer_id, cust_last_name, cust_first_name from customers;

CUSTOMER_ID  CUST_FIRST_NAME  CUST_LAST_NAME       
------------ ---------------- ----------------- 
           1 John             Smith 

SQL> Select order_id, order_date, order_status from orders;

ORDER_ID    ORDER_DATE                              ORDER_STATUS
---------- ---------------------------------------- ------------------------
     2500  13-AUG-09 12.00.00.000000 AM             0

SQL>

INSTEAD OF triggers on Nested Table Column of View


Nested Table Column of View (?)

You can create a view that has a nested table as one of its columns.
Now if you want to be able to update such a view, you need a instead of trigger.
It will intercept the object-relational INSERT and translate it to one that inserts into the base table.



(1) To create a view with a nested table as one of its columns you can:

(a) create the object type for the elements of the nested table. In this case, the type nte captures attributes of employees.

CREATE OR REPLACE TYPE nte AUTHID DEFINER IS
 OBJECT (
   emp_id   number(6),
   lastname varchar2(25),
   job      varchar2(10),
   sal      number(8,2));


(b) create the object table type. Here it will define a table of objects of the type nte.

CREATE OR REPLACE TYPE emp_list_ IS TABLE OF nte;

(c) Now you define the view containing the nested table in one of its columns (emplist).

CREATE OR REPLACE VIEW dept_view AS
  SELECT d.department_id,
  d.department_name,
  CAST (MULTISET (SELECT e.employee_id, e.last_name, e.job_id, i.salary
                  FROM employees e
                  WHERE e.department_id = d.department_id) 
                  AS emp_list_
                 ) emplist
  FROM departments d; 

(d) You can select the contents of emplist, and compare with a direct select of the base table:

SQL> select emplist from dept_view where department_id=10;

EMPLIST(EMP_ID, LASTNAME, JOB, SAL)
-----------------------------------------------
EMP_LIST_(NTE(200, 'Whalen', 'AD_ASST', 4400)


SQL> SELECT employee_id, last_name, job_id, salary
     FROM employees
     WHERE department_id = 10;

EMPLOYEE_ID LAST_NAME		      JOB_ID	     SALARY
----------- ------------------------- ---------- ----------
200 Whalen		      AD_ASST	       4400

(e) Now, if you try to insert directly into the dept_view, you'll receive an ORA-25015.

SQL> INSERT into table (
                        select d.emplist
                        from dept_view d
                        where department_id = 10
     ) 
     VALUES (1001, 'Glenn', 'AC_MGR', 10000);

SQL Error: ORA-25015: cannot perform DML on this nested table view column
*Cause:    DML cannot be performed on a nested table view column except through
an INSTEAD OF trigger
*Action:   Create an INSTEAD OF trigger over the nested table view column
and then perform the DML.

(f) So you need an INSTEAD OF trigger:

CREATE OR REPLACE TRIGGER dept_emplist_tr
 INSTEAD OF INSERT ON NESTED TABLE emplist OF dept_view
 REFERENCING NEW as Employee
             PARENT as Department
 FOR EACH ROW
BEGIN
  -- Intercept the insert into the nested table.
  -- Translate it to an insert into the base table.
  INSERT INTO employees (
              employee_id, last_name, email, hire_date, job_id, salary, department_id)
  VALUES (
     :Employee.emp_id,
     :Employee.lastname,
     :Employee..lastname || '@company.com', 
     SYSDATE,
     :Employee.job,
     :Employee.sal,
     :Department.department_id);
END;


(g) Now you if repeat the insert statement, the trigger will fire and perform the insert in the base table employees:

SQL> INSERT into table (
                        select d.emplist
                        from dept_view d
                        where department_id = 10
     ) 
     VALUES (1001, 'Glenn', 'AC_MGR', 10000);

1 rows inserted.

(h) You can check the results by querying the dept_view view and employees table:

SQL> select emplist from dept_view where department_id = 10;

EMPLIST(EMP_ID, LASTNAME, JOB, SAL)
--------------------------------------------------------------------------------
EMP_LIST_(NTE(200, 'Whalen', 'AD_ASST', 4400), NTE(1001, 'Glenn', 'AC_MGR', 10000))


SQL> SELECT employee_id, last_name, job_id, salary
     FROM employees
     WHERE department_id = 10;

EMPLOYEE_ID LAST_NAME		      JOB_ID	     SALARY
----------- ------------------------- ---------- ----------
	200 Whalen		      AD_ASST	       4400
       1001 Glenn		      AC_MGR	      10000

2 rows selected.