SQL question (easy? hard? dunno!)

Question

Not sure if this is really easy and I'm being dense but....
Given two tables, A & B, how (using SQL only) can I tell whether they contain identical data?
Given that it's not possible to get boolean result from SQL, I don't mind (within reason!) what format the result takes
Cheers
RT

Answer 1

Andrew's query does not work as expected **if there are duplicate rows in each table**. No offence to you Andrew, I am just pointing it out as the matter of the subject, read it to the end and you will see how I found out about this, I did not mean to fail your code in any way.SQL> select * from inv;
       SEQ INV                         AMT       AMT2
         1 AB12                       2000      702.4
         2 ABC123                     2000       48.6
         3 1234                        100         10
-- duplicate all the rows in inv table
SQL> insert into inv select * from inv;
3 rows created.
SQL> select * from inv;
       SEQ INV                         AMT       AMT2
         1 AB12                       2000      702.4
         2 ABC123                     2000       48.6
         3 1234                        100         10
         1 AB12                       2000      702.4
         2 ABC123                     2000       48.6
         3 1234                        100         10
6 rows selected.
SQL> select * from inv2;
no rows selected
-- Make both tables indentical
SQL> insert into inv2 select * from inv;
6 rows created.
SQL> SELECT decode(count(*) , 1, 'SAME', 'DIFFERENT') whatever
  2  FROM   ( SELECT count(*) t FROM inv
  3           INTERSECT
  4           SELECT count(*) FROM inv2 ) tot_rec
  5         , (SELECT count(*) t FROM ( SELECT * FROM inv
  6                                     INTERSECT
  7                                     SELECT *  FROM inv2 ) ) tot_match
  8  WHERE tot_rec.t = tot_match.t
  9  /
WHATEVER
DIFFERENT
-- In fact they are identical
SQL> select * from inv
  2  minus
  3  select * from inv2;
no rows selected
SQL> select * from inv2
  2  minus
  3  select * from inv;
no rows selectedFollowing version has the same problem. I am just posting it for the relevance.
In case where where we can assume that there would not be any duplicate
rows in each table, then here is another simple solution, SELECT decode(cnt1, cnt2, decode(cnt1, cnt3, 'TRUE', 'FALSE'), 'FALSE') yes_no
FROM
(select count(*) cnt1 from inv),
(select count(*) cnt2 from inv2),
(select count(*) cnt3 from (select * from inv union select * from inv2))I was thinking about duplicates problem with UNION alternative and that is when
I found out that Andrew's SQL has the same caveat.
Thx,
SriDHAR