Question

I have this following problem:

I have two tables 1 table for customer bought items and the other for employee sales:

Table 1 (Emp)
EmpId 
ProductId
dateSold
Price
Qty

Table 2 (Cust)
CustomerId
ProductId
dateSold
Price
Qty

do this is the query im tring at the moment:

SELECT  SUM(EmployeeSales.productPrice * EmployeeSales.Qty + 
            userBoughtItems.ProductPrice * userBoughtItems.qty) AS Total, 
        EmployeeSales.dateSold, 
        userBoughtItems.dateSold AS Expr1
FROM    EmployeeSales 
INNER JOIN userBoughtItems ON EmployeeSales.dateSold = userBoughtItems.dateSold
GROUP BY EmployeeSales.dateSold, userBoughtItems.dateSold

I want to get the total of both tables on the same date…

Answer 1

A join repeats the right table for every matching row in the left table. In your example I’d expect that to result in a lot of duplicated rows. Consider using a union instead:

select  cast(dt as date)
,       sum(sales)
from    (
        select  dateSold as dt
        ,       productPrice * Qty as sales
        from    EmployeeSales
        union all
        select  dateSold
        ,       ProductPrice * qty
        from    userBoughtItems
        ) as SubQueryAlias
group by
        cast(dt as date)

The subquery contains a list of all sales, and the outer query sums them per day.

If you’re using SQL Sever 2005 or older, replace cast(dt as date) with dateadd(dd, 0, datediff(dd, 0, dt)).