I am trying to create a nested SELECT SQL statment. I store all value and id and want to select rows that satisfy multiple values. How can I generate a SELECT statement using Java? For example,
ID VALUE
1 RED
2 BIG
1 SMALL
1 SMOOTH
2 TALL
.....
To select an item that is both red and small the statement would be:
SELECT *
FROM table
WHERE table.value = RED AND
id IN (SELECT *
FROM table
WHERE table.value = SMALL AND id IN (...))
This type of problem is called Relational Division
SELECT ID
FROM tableName
WHERE VALUE IN ('SMALL','SMOOTH')
GROUP BY ID
HAVING COUNT(*) = 2
the query above will result 1 since the ID contains both records.
If no unique constraint was enforce on value for every ID, DISTINCT is required.
SELECT ID
FROM tableName
WHERE VALUE IN ('SMALL','SMOOTH')
GROUP BY ID
HAVING COUNT(DISTINCT VALUE) = 2
OTHER(s)
select ID
from MyTable
where VALUE in ('RED', 'SMALL')
group by ID
having count(distinct VALUE) = 2
Results:
| ID |
------
| 1 |
Here is a general way to approach this:
select id
from t
group by id
having max(case when value = 'Red' then 1 else 0 end) = 1 and
max(case when value = 'Small' then 1 else 0 end) = 1
In other words, membership in the set becomes a clause in the having statement. These can be both inclusion and exclusion (use = 0 instead of = 1) and optional (use or instead of and).
