Reputation: 21
How can we pass a parameter that is declared as an Oracle Object type to a Procedure having a parameter as PLSQL Table type?
Eg:
Procedure A(p_obj_emp t_obj_emp)
Procedure B(p_emp_tab t_emp_tab)
Where t_obj_emp = Oracle Object and t_emp_tab is a PLSQL Table of binary_integer
How can we pass a parameter that is declared as an Oracle Object type to a Procedure having a parameter as PLSQL Record type?
Eg:
Procedure C(p_obj_dept t_obj_dept)
Procedure D(p_dept_rec t_dept_rec)
Where t_obj_dept = Oracle Object having 2 fields (deptid, deptname) and t_dept_rec is declared in package specification as t_dept_rec with 2 fields (deptid, deptname).
Kindly help with some example. Thanks in advance
Upvotes: 2
Views: 9428
Reputation: 64949
The following compiles for me and appears to do what you want:
CREATE OR REPLACE TYPE t_obj_emp AS OBJECT (
emp_id INTEGER,
emp_name VARCHAR2(100)
);
/
CREATE OR REPLACE TYPE t_obj_dept AS OBJECT (
dept_id INTEGER,
dept_name VARCHAR2(100)
);
/
CREATE OR REPLACE PACKAGE my_pkg AS
TYPE t_emp_tab IS TABLE OF t_obj_emp INDEX BY BINARY_INTEGER;
TYPE t_dept_rec IS RECORD (
dept_id INTEGER,
dept_name VARCHAR2(100)
);
PROCEDURE A(p_obj_emp t_obj_emp);
PROCEDURE B(p_emp_tab t_emp_tab);
PROCEDURE C(p_obj_dept t_obj_dept);
PROCEDURE D(p_dept_rec t_dept_rec);
END;
/
CREATE OR REPLACE PACKAGE BODY my_pkg AS
PROCEDURE A(p_obj_emp t_obj_emp)
IS
v_emp_tab t_emp_tab;
BEGIN
v_emp_tab(1) := p_obj_emp;
B(v_emp_tab);
END;
PROCEDURE B(p_emp_tab t_emp_tab) IS BEGIN NULL; END;
PROCEDURE C(p_obj_dept t_obj_dept)
IS
v_dept_rec t_dept_rec;
BEGIN
v_dept_rec.dept_id := p_obj_dept.dept_id;
v_dept_rec.dept_name := p_obj_dept.dept_name;
D(v_dept_rec);
END;
PROCEDURE D(p_dept_rec t_dept_rec) IS BEGIN NULL; END;
END;
/
Note that there isn't any 'convenient' way to create one object/record from another, even if they have identical members (such a t_obj_dept
and t_dept_rec
in the example above). You must copy across the values of all the fields individually.
You say that t_emp_tab
is a "PLSQL Table of binary_integer". I'm guessing you mean that it's a PL/SQL Table of some type indexed by binary_integer
, as it's far more common to index these tables by binary_integer
than it is to store binary_integer
s in them. For the example above I've assumed that it's a table of t_obj_emp
s. If it's not, you'll have to map the t_obj_emp
object to whatever type of object or record t_emp_tab
is a table of.
Upvotes: 2