The conventional way of providing this is in an RDBMS through three tables: two value tables and an intersect table. In your example, you would have t1 and t2:
t1.c1 t2.c2
1 5
2 9
3
and the intersect table which would look like your current
many-to-many. Of course, this assumes that you have more attributes associated with t1 and t2. One of Oracle's classic examples of this would be customers and sales people (employees). In this case, you can have one employee serving many customers and one customer being serviced by many employees. Rather than replicate the customers' name, address, shoe size etc and the employee's name, hire date, and so forth, your intersect table might consist of the customer's id number, the employee's id number, the date of service, and probably the sales order id number.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.