跨多个列的唯一实体框架

Question

Given the following table

Year - Int
EmployeeOne - String
EmployeeTwo - String

How can we prevent duplicate values across different columns.

Take the following scenario

Year | EmployeeOne | EmployeeTwo
2020 | 0001        | 0002
2020 | 0003        | 0004

This is fine as EmployeeOne and EmployeeTwo don't contain the same values.

However given this scenario

Year | EmployeeOne | EmployeeTwo
2020 | 0001        | 0002
2020 | 0003        | 0001

I want this to fail. As EmployeeOne column has 0001 with a year of 2020 and EmployeeTwo column also has 0001 with a year of 2020

I've tried putting in the place the following composite keys

• Year & EmployeeOne
• Year & EmployeeTwo
• Year & EmployeeOne & EmployeeTwo

But having no joy

Answer 1

If its not possible with Table Constraints then you can do validation in trigger.

Answer 2

This is trivial if your data model has (Year, Employee, Role) instead of (Year, EmployeeOne, EmployeeTwo). Then a key of (Year,Employee) ensures this constraint.