multiple foreign keys on same column

[mrpants]

i've been googling this for hours and can't seem to find an answer.
i'm a student that's new to oracle and i'm designing a database for a construction company. i have a contacts table that stores contacts for their clients and subcontractors. i originally designed the database in access and have an employer_id field that references the client_id in the client table and the subc_id in the subcontractor table. in oracle, i have two foreign key constraints. the client_id starts at 10,000, the subc_id starts at one and goes up to 9999. when i enter a row in the contacts table, it causes a problem since the record only satisfies one of the constraints.

so is it possible to have a column in an oracle database that references two seperate columns but only has to satisfy one of the FK constraints?


btw, you can see the ERD here: http://www.auburn.edu/~shanncb/db/relationships.bmp

[pragatiswain]

i've been googling this for hours and can't seem to find an answer.
i'm a student that's new to oracle and i'm designing a database for a construction company. i have a contacts table that stores contacts for their clients and subcontractors. i originally designed the database in access and have an employer_id field that references the client_id in the client table and the subc_id in the subcontractor table. in oracle, i have two foreign key constraints. the client_id starts at 10,000, the subc_id starts at one and goes up to 9999. when i enter a row in the contacts table, it causes a problem since the record only satisfies one of the constraints.

so is it possible to have a column in an oracle database that references two seperate columns but only has to satisfy one of the FK constraints?


btw, you can see the ERD here: http://www.auburn.edu/~shanncb/db/relationships.bmp

Either you create 2 different contact tables for client and contractor.
OR
Create a single table for client and contractor.
OR
When you enter a row in Client table, enter a "false row" in contractor table and vice-versa. To identify a false row, keep some flagging. (Else as you have already specified based on ID)

Hope this helps