SQL efficient way to join a table where all values exist

Question

table 1
item    group_id
123     aaaa
432     bbbb
534     aaaa
765     cccc
656     aaaa
656     cccc
111     cccc

group_id    group_name
aaaa        groupA
bbbb           


        

Answer 1

You can use the INTERSECT operator. INTERSECT returns the shared distinct values from both tables. In practice I've found this faster than other methods.

In this example I put intersect in-between the queries below:

SELECT T2.groupName 
  FROM table1 T1 INNER JOIN table2 T2 ON T1.groupID = T2.groupID
 WHERE T1.itemID = 765

INTERSECT 

SELECT T2.groupName 
  FROM table1 T1 INNER JOIN table2 T2 ON T1.groupID = T2.groupID
 WHERE T1.itemID = 656

Provides output:

groupName
---------
groupC

Answer 2

The simplest thing would be to use group by and having:

SELECT group_name
FROM table2 g
JOIN table1 t on g.group_id = t.group_id
WHERE t.item in (765,656)
GROUP BY group_name
HAVING COUNT(DISTINCT t.item) = 2

Answer 3

SELECT DISTINCT t1.group_name 
                    FROM table1 t1 join table1 t2 join groups g
                        ON t1.group_id = t2.group_id 
                           and t1.item = 765 
                           and t2.item = 656
                           and g.group_id = t1.group_id