Join 2 tables which are not related with any primary-foriegn key constraint

Question

Hello,
How to join 2 tables which are not related with any primary key foreign key constraint.
Ex.Consider Table A has 5 columns->A_ID,A_Name,A_Address,A_City,A_Pin(Total 10 rows)
Table B has 5 columns->B_ID,B_Name,B_Adress,B_City,B_City(Total 30 rows)
From both the table i want the data,which i need to use in curosr to display finally as "Address Label".
Both the table are entirely different,but there can be some names of Table A which may come in Table B also,if the name,address and city is also same.That means same person in both the table.
So finally i want total number of distinct records(distinct data of Table A which is not in B+all the data of table B) to come under Address Label.
How can i write the select query for this condition?
Thanks
Swapna

Answer 1

Hi, Swapna,
user11018268 wrote:
Hello,
How to join 2 tables which are not related with any primary key foreign key constraint.
Ex.Consider Table A has 5 columns->A_ID,A_Name,A_Address,A_City,A_Pin(Total 10 rows)
Table B has 5 columns->B_ID,B_Name,B_Adress,B_City,B_City(Total 30 rows)
Both the table are entirely different,but there can be some names of Table A which may come in Table B also,if the name,address and city is also same.That means same person in both the table.I think you want a UNION, not a join.
If you have 10 rows in table_a, and 30 rows in table_b, then
SELECT       A_ID,          A_Name,     A_Address,     A_City,     A_Pin
FROM      table_a
UNION
SELECT       B_ID,          B_Name,     B_Adress,     B_City,     B_City
FROM      table_b
ORDER BY            2,               4          
;will produce up to 40 rows; 40 if there are no duplicates. (UNION implies DISTINCT.)
The corresponding column types should be similar, if not exactly the same.
There's no problem if A_Name is VARCHAR2 (50) and B_Name is CHAR (30).
However, there will be a problem if A_ID is a NUMBER and B_ID is a TIMESTAMP.
You must have the same number of columns in all branches of the UNION.
If you want an ORDER BY clause, put it at the very end, after the last branch of the UNION.
You can use positional notation (where 2 means the 2nd column) like I did, or you can use names from the first prong of the UNION (such as A_Name).
From both the table i want the data,which i need to use in curosr to display finally as "Address Label".
So finally i want total number of distinct records(distinct data of Table A which is not in B+all the data of table B) to come under Address Label.I'm not sure what your mean about "Address Label".
Whenever you have a question, post a little sample data (CREATE TABLE and INSERT statements, relevant columns only) from all tables, and the exact results (formatted) that you want from that data.
You can concatenate all 5 columns into one VARCHAR2 column, if you want to.
You'll probably want to use RPAD (or simillar functions) to make the output line up nicely.
If any of the columns are not strings, use TO_CHAR to make string versions of them.
For example:
SELECT  TO_CHAR (A_ID, '9999999')
     || '   '          -- leave a little space between the left-justified a_id and the right-justified a_name
     || RPAD (A_Name, 25)
     || RPAD (A_Address, 50)
     ...Edited by: Frank Kulash on Nov 14, 2009 10:11 AM