Question

I have a table with 14 columns in Oracle. All the columns are of integer type. i need to check no two columns in one row have same integer value in them. How can i do this using SQL. Or is it only possible to do using PL/SQL?

Answer 1

Check the following examples.
Firstly I group the values into a collect. I’ve used the built in SYS.DBMS_DEBUG_VC2COLL but it is better if you create your won.
Then I take that collection then do a MULTISET UNION DISTINCT with an empty collection of the same type. That will remove duplicated entries.

select * from
 (select sys.dbms_debug_vc2coll(1,2,3) a, 
         sys.dbms_debug_vc2coll(1,2,3) multiset union distinct sys.dbms_debug_vc2coll() b
 from dual)
where a=b;

select * from
 (select sys.dbms_debug_vc2coll(1,2,3,1) a, 
         sys.dbms_debug_vc2coll(1,2,3,1) multiset union distinct sys.dbms_debug_vc2coll() b
 from dual)
where a=b

Finally I compare that DISTINCT’d collection with the original collection. If they match, then the collection already only had unique values.