T-SQL equivalent to MySQL aggragate from other table

Question

I'm struggling to do something in SQL Server that is super easy in MySQL.

I want to grab some columns (let's just say one for simplicity) from one table and then a count other columns of another table based on a key from the first table and have both values in my result set.

The following query works in MySQL but does not in SQL Server:

SELECT 
    ph.PONum, 
    COUNT(pd.POLine) AS LineCount
FROM 
    POHeader AS ph
INNER JOIN 
    PODetail AS pd ON ph.POnum = pd.PONum
WHERE 
    ph.PONum = 4444

Results:

What is the SQL Server (T-SQL) way to do this?

Answer 1

Unlike MySQL, SQL Server does not assume what columns you want to group by. Every column that is not being aggregated needs to be defined in the GROUP BY list.

For your query it should be:

SELECT 
    ph.PONum, 
    COUNT(pd.POLine) AS LineCount
FROM 
    [Dbo].POHeader AS ph
INNER JOIN 
    [Dbo].PODetail AS pd ON ph.POnum = pd.PONum
WHERE 
    ph.PONum = 4444
GROUP BY 
    ph.PONum

Answer 2

You can also use a correlated subquery. EG

SELECT 
    ph.PONum, 
    ( select COUNT(*) from PODetail where PONum = ph.POnum ) LineCount
FROM 
    POHeader ph
WHERE
    ph.PONum = 4444