Synonyms can be PRIVATE or PUBLIC. If you create a private synonym then only you can use that synonym. The default is PRIVATE.

PUBLIC synonyms can be used by all database users.

See below some of the usage of Synonyms:

1.

Create public Synonym to give short names to a table

Create synonym tccg for schema1.t_country_city_group;

This statement will create a short name tccg for the table schema1.t_country_city_group. So you can now query the table using the statement:

Select * from tccg;

2.

Create a synonym where schema name is long

Create public synonym tccg for schema1isalongschema.t_country_city_group;

If you have to specify the name schema1isalongschema again and again, I am sure you will start pulling your hair. So if you create the synonym “tccg” then you can simply refer that table as

Select * from tccg;

3.

Create a synonym for a table in another database.

Create public synonym tccg for schema1.t_country_city_group@remote_db1.world;

The above statement is creating a synonym called “tccg” for an object which is in another database – referenced by database link remote_db1.world.

4.

Create synonym for a procedure with long name.

Create public synonym dispdata for schema1.p_proceduretodisplaydata;

Here you are giving short name dispdata for stored procedure p_proceduretodisplaydata that reside in schema1 schema.

A REF CURSOR is a cursor which does not have a fixed SQL query associated with it. Basically you declare a TYPE of REF CURSOR type. Then you define a variable of that TYPE. That variable is called and cursor variable.

Then you assign a query to that cursor variable. After that you can use that cursor variable like any other cursor.

CURSOR VARIABLE is a variable whose data type is a TYPE of REF CURSOR.

On the other hand, in a normal cursor you write a select query and you give a name to that select query. You can open, fetch data and close that cursor whenever you want.

For example, you declare a normal cursor as:

cursor emp_cur is
select * from emp where deptno =10;

In this example you are giving a name “emp_cur” to the SQL query. The sql query is always static. It will always return you all the employees who are in department id 10. Now you can use open-fetch-close on this cursor as many times as you want in your pl/sql program.

Coming back to REF CURSOR, you can define a REF CURSOR TYPE as

TYPE RefCurTyp IS REF CURSOR;

Here RefCurTyp is a user defined datatype of REF CURSOR. Now any variable that you declare of datatype RefCurTyp will be a cursor variable.

For example here ref_cv is a cursor variable:

ref_cv RefCurTyp;

In the example below the package has two stored procedures. One of the procedure will accept a number as input parameter. Based on that number the other procedure will OPEN a REF cursor and return data to the calling procedure. Then the calling procedure can process the ref cursor as any other normal static query cursor. Using two separate stored procedure makes it clearer to understand and easier for future modifications.

CREATE OR REPLACE PACKAGE pkg_ref_cursor_test IS
TYPE RefCurTyp IS REF CURSOR;
PROCEDURE p_process_data(in_indi number);
PROCEDURE p_open_refcur (ref_cv IN OUT RefCurTyp, in_indi number);
END pkg_ref_cursor_test;
/

CREATE OR REPLACE PACKAGE BODY pkg_ref_cursor_test IS

PROCEDURE p_process_data(in_indi number) IS
ref_cv RefCurTyp;
l_count number;
BEGIN
p_open_refcur(ref_cv, in_indi);
LOOP
FETCH ref_cv INTO l_count;
EXIT WHEN ref_cv%NOTFOUND;
IF in_indi = 10 THEN
dbms_output.put_line(‘Total emp=’||l_count);
ELSIF in_indi =20 THEN
dbms_output.put_line(‘Total dept=’||l_count);
END IF;
END LOOP;
CLOSE ref_cv;
EXCEPTION
WHEN OTHERS THEN
dbms_output.put_line(‘Error = ‘||SQLERRM);
END;

PROCEDURE p_open_refcur (ref_cv IN OUT RefCurTyp, in_indi number) IS
BEGIN
IF in_indi = 10 THEN
open ref_cv FOR SELECT COUNT(*) FROM emp;
ELSIF in_indi =20 THEN
open ref_cv FOR SELECT COUNT(*) FROM dept;
END IF;
END;
END pkg_ref_cursor_test;
/

As you can see from the above example, the cursor variable ref_cv is associated either with emp table or with dept table. So the SQL query for this cursor is not static – not declared when the cursor
is declared. The cursor returns either number of employee or number of departments based on the input parameter

Also note the exception section in p_process_data stored procedure. When a cursor can not be opened (a number other than 10 or 20 is passed) then the program will encounter this error. So handling exception here should trap that as this stored program will try to fetch from the cursor even the cursor was not opened at all.

I hope this article and example has given you a good understanding of Oracle REF CURSOR.

The basic form of an Oracle stored procedure is

CREATE OR REPLACE PROCEDURE p_procedure1 IS
BEGIN

null;

END p_procedure1;

The stored procedure above will compile successfully.

Now lets discuss the procedure above.

CREATE OR REPLACE: You must use CREATE command whereas REPLACE is optional. But if you do not put REPLACE then you can not recompile the stored procedure if it already exists. By putting CREATE OR REPLACE you are saying that if it already exists then replace it with the new version.

PROCEDURE: This clause specifies the type of object that you want to create. Here we are saying it to be a procedure.

BEGIN and END: This is the body of the procedure. All code that you want to put in the procedure should live here. In the example above I have put null; It means do nothing. However even if you do nothing and just want to create a stored procedure (as above) you still need to put the Begin and End to complete the frame of the stored procedure. For clarity I have put End p_procedure1 which is not required. Instead you could just put End;

Example of Oracle Store Procedure with parameters:

Most of the time when calling an Oracle procedure you would like to pass on some parameter to the procedure to process it or to do something with the data in the parameter. You can pass parameters to a stored procedure as below:

CREATE OR REPLACE PROCEDURE p_procedure2(p_name varchar2) IS
BEGIN

dbms_output.put_line(p_name);

END p_procedure2;

Here I am passing a parameter called p_name to the stored procedure p_procedure1; Then in the body section I am displaying the value of the parameter.

You can use “dbms_output.put_line()” to display the values of parameters in SQL*Plus window. This is same PRINT/ECHO/PRINTLN used in other programming languages.

Now the parameters that you pass to a stored procedure can be of three types (IN, OUT and IN OUT).

IN - is the default type. So if you do not specify the parameter type then IN is used. This is telling that I am supplying a parameter to the stored procedure.

OUT – is used to get values back from the stored procedure.

IN OUT – is used to pass values to a stored procedure and to get values from the stored procedure using the same parameter.

Unlike Oracle functions you can return multiple values using the out parameter.

For example: if you want to get back the employee name of an employee and you have just the emp_no (employee number) then you can pass the emp_no to a stored procedure using an IN parameter and then get the employee name using an out parameter. This is demonstrated below:

CREATE OR REPLACE PROCEDURE p_procedure3(p_emp_no IN number, p_name OUT varchar2) IS
BEGIN

SELECT ename
INTO p_name
FROM emp
WHERE emp_no = p_emp_no;

END p_procedure3;

Here we are passing the emp_no and using that we are getting ename from table emp. The returned value is put into p_name and the calling program will be able to access this p_name.

Now when we use out parameter it becomes a little bit different while executing a stored procedure.

For example:  to execute the first two stored procedures mentioned above you simple login to Oracle using SQL*Plus and run the commands:

SQL>exec p_procedure1;

Or

SQL>set serverout on;

SQL> exec p_procedure2(‘ITSME’);

But when if use an OUT parameter (as in p_procedure3)  you can not simply run execute command like “exec p_procedure3(parameters here). Because the stored procedure is returning a value and you need to read that value.

In such case you can write a very simple PL/SQL block as below

DECLARE
l_name varchar2(20);
BEGIN

— now call the stored procedure

p_procedure3(100, l_name);

— now display the returned value

dbms_output.put_line(l_name);

END;

In the code above I have created a variable called l_name. While calling p_procedure3 I am passing the name of this variable. The procedure will find the name of the employee with emo_no=100 and return the name of the employee in my variable l_name. I can then display or do whatever I want with the value in the variable.