Help with distinct column

Question

Hi;
I have a three table TB1, TB2 and TB3 which has 20 columns each.
Table TB1 has columns like x1,x2,x3...x20, table TB2 has columns
like x1,x21,x22...x39, TABLE TB3 has columns like x21,x40...x58.
I want to create a table of all disctinct column like
create TABLE  TB4 as select TB1.*, TB2.*, TB3 from TB1 left outer join TB2
on TB1.x1=TB2.x1 left outer join TB3 on TB2.X21=TB3.x21However it gives me error duplicate column name, is there anyway I can generate table without mentionin specifically all column names.
I want to use basically wild character like TB1.{noformat}*{noformat}, TB2.{noformat}*{noformat} but if any duplicate column just take from table TB1.
Vkm
Edited by: BluShadow on 16-May-2011 14:17
added {noformat}{noformat} tags for at least some readability                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   

Answer 1

Hi,
You can use a =TopN QUery+ to make sure the result set contains only one row for every distinct value on some column(s), for example:
WITH     got_r_num     AS
     SELECT  tb1.*
     ,     tb2.*
     ,     ROW_NUMBER () OVER ( PARTITION BY  tb1.x1
                                ORDER BY         NULL     -- Or whatever you want
                       )          AS r_num
     FROM         tb1                    
     LEFT OUTER JOIN     tb2  ON       tb1.x1  = tb2.x1
SELECT     *
FROM     got_r_num
WHERE     r_num     = 1
;However, this will include r_num itself in the result set. You can list all the columns except r_num in the main SELECT clause, but you said you don't want to do that. (Why?) If you're just displaying this output, you your front end may have a way to hide the r_num column. SQL*Plus, for example, has "COLUMN r_num NOPRINT".
If you're using this in a CREATE TABLE AS statement, you'll need to give unique names to all the columns, anyway.
I hope this answers your question.
If not, post a little sample data (CREATE TABLE and INSERT statements, relevant columns only) for all tables, and also post the results you want from that data.
Explain, using specific examples, how you get those results from that data.
Always say which version of Oracle you're using.
You'll get better answers faster if you always supply this information whenever you post a question.