Dynamically Select User Very URGENT !!!!!!!!!!!!!

Question

Dear Friends it is too urgent,
I have to build a query where i want to select records from the tables dynamically from different users. i.e Suppose I have many users 'USER1' , 'USER2' , 'USER3'... AND SO ON. All users have same tables. suppose 'Employee' is the table reside in all users. Now suppose if I supply owner name in the query, it must retrieve rows from Employee table from that particular user.
I'll be very thankful for the kind favour
Regards

Answer 1

Srinivas , thanks for the follow up , actually the problem is the referencing of object , whenever u create object in a schema, it is stored with refernce of schema name. LIKE SCHEMA.OBJECT , when u call the object u don't need to define the schema because u are actually working on that schema but if u try to call the object from other schema u need to require the schema name Now i must describe the workeability of object with examples. Now suppose i have a table EMPLOYEE which has the object Address_Type AND IT IS DEFINED IN 2 USERS
'USER1' AND 'USER2'
EMPLOYEE
EMPNO NUMBER(10),
ENAME VARCHAR2(50),
HIREDATE DATE,
ADDRESS ADDRESS_TYPE
To get value of the address u have to select like that way from 2 different users USER1 and USER2
select a.empno,a.ename, a.hiredate,
address_type(a.address.colname...)
from user1.employee
or u can also select like
select a.empno,a.ename, a.hiredate,a.address.colname
from user2.employee a
Now coming to the problem
if u try to create a dummy table in user1 from user2 it will give u error
CREATE TABLE USER1.DUMMY
AS
SELECT * FROM USER2.EMPLOYEE
ERROR at line 1:
ORA-00902: invalid datatype
This error is because object is created with the name of the schema (USER2.ADDRESS) and the schema where u are creating this object also contain the object type with refernce to the object
(USER1.ADDRESS).
The problem occurs when I try to define a variable of %ROWTYPE
it needs a schema name which i can not set it dynamically in the declaration part like for example
CREATE OR REPLACE PROCEDURE EMP_PROC (V_USER IN VARCHAR2)
IS
U_NAME VARCHA2(100);
type curref is REF CURSOR;
vcur curref;
REC [SCHEMA].EMPLOYEE%ROWTYPE [ HERE THE SCHEMA IS PASSING AS AN ARGUMENT WHICH IS NOT ALLOWED IN THE DECLARATION PART ]
------ IF I TRY TO REPLACE SCHEMA WITH THE VARIABLE
------ U_NAME IT IS NOT ACCEPTED
I think I wrote a lot to tell u about the actual problem.
Please dear friends, specially Barbera, i u look closely to my problem I'll be highly thankful. If suppose it is not possible I need other work around.
Regards
Shah I think the problem comes down to the fact that object address_typ in schema1 will have a different global object identifier to address_typ in schema2. They are therefore seen as different datatypes.
If you had a third schema which owned the object, and referenced this is the other two schemas, then you could quite happily insert the contents from schema1 into schema2.
e.g.
Schema3
CREATE OBJECT address_typ AS OBJECT (line_1 VARCHAR2(100), line_2 VARCHAR2(100));
GRANT ALL ON address_typ TO PUBLIC WITH GRANT OPTION;
Schema1
CREATE TABLE emp(empno NUMBER, ename VARCHAR2(100), address schema3.address_typ);
GRANT SELECT ON emp TO PUBLIC;
INSERT INTO emp VALUES (1, 'Smith', schema3.address_typ('Street','Town'));
Schema2
CREATE TABLE emp(empno NUMBER, ename VARCHAR2(100), address schema3.address_typ);
INSERT INTO emp SELECT * FROM schema1.emp;
In this example, the address column in both schema's is of the same datatype (schema3.address_typ).
I hope this sheds some light on your last post.