How to check which column data differs from master table and archive table

Question

Hi All,
i have two tables, table a (a1 number,a2 varchar2,a3 varchar2) and table b (b1 number,b2 varchar2,b3 varchar2).
how to check the data in both the table are same( including all columns).
data in a.a1 is same as b.b1 and a.a2 is same as b.b2 like that.
if they not same , i need to know which field differs.
Kindly Share ur ideas.

Answer 1

887268 wrote:
thanks Sven W. ,
above reply clearly shows what my question is.
one column must be primary key, based on that key i need to find out which are the fields having different data..
im strugling with this, i tried the following already, but not able to get.
select the columns from a MINUS select the columns from b.
-- from this i can find whether the difference occurred or not.
but i cant able to get which are the fields value changed.Good. Then you would match the rows using the PK column and need to compare the columns
Instead of a MINUS + UNION ALL + MINUS we can now use a FULL OUTER JOIN
It is a little task to write out all column names, but 40 columns can be handled.
This statement would show you both tables with matching rows on the same line.
select a.*, b.*
from a
FULL OUTER JOIN b on a.id = b.idNow filter/check for mismatches
select case when a.col1 != b.col1 then 'COL1 value changed'
                when a.col2 != b.col2 then 'COL2 value changed'
                when a.col3 != b.col3 then 'COL3 value changed'
         end as compare_result
        ,a.*, b.*
from a
FULL OUTER JOIN b on a.id = b.id
/* return only non matching columns */
where (a.col1,a.col2,a.col3) != (b.col1,b.col2,b.col3) You might need to add nvls to take care of null values. Test this!
Another way could be to group upon the primary key
select *
from (
  select id 
           ,count(distinct col1)-1 cnt_col1
           ,count(distinct col2)-1 cnt_col2
           ,count(distinct col3)-1 cnt_col3
   from
     select 'A' source, a.*
     from a
     UNION ALL
     select 'B' source, b.*
     from b)
   group by ID
/* only records with differences */
where 1 in (cnt_col1, cnt_col2, cnt_col3)
;The count columns will hold either 1 or 0. If it is 1 then this column has a difference.