Question

Here’s what I’m trying do to in a single SQL Server procedure:

@ID1 int

select ID2 from TableA where ID1 = @ID1

(for each selected @ID2)
    select * from TableB where ID2 = @ID2

Any ideas?

Answer 1

That can be done in a single statement:

SELECT b.*
  FROM TABLE_B b
  JOIN TABLE_A a ON a.id2 = b.id2
 WHERE a.id1 = @ID1

But this means that there will be duplicates if more than one record in TABLE_A relates to a TABLE_B record. In that situation, use EXISTS rather than adding DISTINCT to the previous query:

SELECT b.*
  FROM TABLE_B b
 WHERE EXISTS(SELECT NULL
                FROM TABLE_A a
               WHERE a.id2 = b.id2
                 AND a.id1 = @ID1)

The IN clause is equivalent, but EXISTS will be faster if there are duplicates:

SELECT b.*
  FROM TABLE_B b
 WHERE b.id2 IN (SELECT a.id2
                   FROM TABLE_A a
                  WHERE a.id1 = @ID1)