java experts can you please help me write detached queries as a part of the criteria query for the following SQL statement.
select A.*
FROM AETABLE A
where not exists
(
select entryid
FROM AETABLE B
where B.classpk = A.classpk
and B.userid = A.userid
and B.modifiedDate > A.modifiedDate
)
and userid = 10146
You need to write a correlated subquery. Assuming property / class names match column / table names above:
DetachedCriteria subquery = DetachedCriteria.forClass(AETable.class, "b")
.add(Property.forName("b.classpk").eqProperty("a.classpk"))
.add(Property.forName("b.userid").eqProperty("a.userid"))
.add(Property.forName("b.modifiedDate").gtProperty("a.modifiedDate"));
Criteria criteria = session.createCriteria(AETable.class, "a")
.add(Property.forName("userid").eq(new Integer(10146)))
.add(Subqueries.notExists(subquery);
Just one addition to the above query. If the entryid is not the primary key, then you'll need to add projection.
DetachedCriteria subquery = DetachedCriteria.forClass(AETable.class, "b")
.add(Property.forName("b.classpk").eqProperty("a.classpk"))
.add(Property.forName("b.userid").eqProperty("a.userid"))
.add(Property.forName("b.modifiedDate").gtProperty("a.modifiedDate"))
.add(setProjection(Projections.property("entryId")); // Additional projection property
Criteria criteria = session.createCriteria(AETable.class, "a")
.add(Property.forName("userid").eq(new Integer(10146)))
.add(Subqueries.notExists(subquery);
Related
I'm trying to write a CriteriaQuery which will query latest observation for each city. City is defined by city_code field, while latest record is defined by observation_time field.
I can easily write it in a plain SQL, but I cant understand how to do it with jpa criteria api.
select distinct m.* from
(select city_code cc, max(observation_time) mo
from observations group by city_code) mx, observations m
where m.city_code = mx.cc and m.observation_time = mx.mo`
It is possible when You are open for loose efficiency.
So first let's transform our query to logical equivalent one:
select distinct m.* from observations m where
m.observation_time = (select max(inn. observation_time) from observations inn
where inn.city_code = m.city_code);
then let's translate it to JPA CriteriaQuery:
public List<Observation> maxForEveryWithSubquery() {
CriteriaBuilder builder = entityManager.getCriteriaBuilder();
CriteriaQuery<Observation> query = builder.createQuery(Observation.class);
Root<Observation> observation = query.from(Observation.class);
query.select(observation);
Subquery<LocalDateTime> subQuery = query.subquery(LocalDateTime.class);
Root<Observation> observationInner = subQuery.from(Observation.class);
subQuery.where(
builder.equal(
observation.get(Observation_.cityCode),
observationInner.get(Observation_.cityCode)
)
);
Subquery<LocalDateTime> subSelect = subQuery.select(builder.greatest(observationInner.get(Observation_.observationTime)));
query.where(
builder.equal(subSelect.getSelection(), observation.get(Observation_.observationTime))
);
TypedQuery<Observation> typedQuery = entityManager.createQuery(query);
return typedQuery.getResultList();
}
Unfortunately JPA does not support sub queries in FROM clause. You need to write a native query or use framework like FluentJPA.
I have a table of users, with user_id.
I'm fetching users like this:
Criteria c = session.createCriteria(User.class);
now I have a table tickets with a foreign key to the users table. there is a column user_id
i want in my criteria to get all the users with ticket type X.
i tried something like this:
c.createAlias("ticket", "ticket");
c.add(Restrictions.ne("ticket.type", "X"));
In the createAlias method you can specify which join you need.
Criteria c = session.createCriteria(User.class)
.createAlias("ticket", JoinType.LEFT_OUTER_JOIN);
or
Criteria c = session.createCriteria(User.class, "user")
.createAlias("user.ticket", "ticket", JoinType.LEFT_OUTER_JOIN);
How can I write this SQL query as a Hibernate JPA Criteria (with Restrictions, etc) in Java ?
SELECT q.*
FROM queue AS q
WHERE q.executed = false AND
q.queued_on = (SELECT min(queued_on) FROM queue WHERE item_id = q.item_id);
I only managed to write the first part like this:
getBaseCriteria()
.add(Restrictions.eq("executed", false))
// Missing Second Where Filter Here
.addOrder(Order.asc("queuedOn"))
.list();
Try to create a separate criteria instance for the subquery and simply add as another restriction as follows:
DetachedCriteria subCriteria = DetachedCriteria.forClass(Queue.class, "sub")
.add(Restrictions.eqProperty("sub.itemId","main.itemId"))
.setProjection(Projections.projectionList().add(Projections.min("sub.queuedOn")));
session.createCriteria(Queue.class, "main")
.add(Subqueries.propertyEq("main.queuedOn", subCriteria ));
.add(Restrictions.eq("main.executed", false));
.addOrder(Order.asc("main.queuedOn"))
.list();
How can I execute this SQL query using the Hibernate Criteria API?
SELECT c.*, count(r.id) FROM COURSE c left join REFERRAL r on c.id = r.course_id group by c.id
Something like this seems close enough:
Criteria criteria = currentSession().createCriteria(Referral.class, "r");
criteria.createAlias("r.course", "course");
ProjectionList projections = Projections.projectionList();
projections.add(Projections.groupProperty("course.id"));
projections.add(Projections.count("r.id"));
return list(criteria
.setProjection(projections));
I need to do this query with Java + Hibernate.
SELECT
table2.id,
COUNT(table2.id) AS count
FROM
table1
JOIN table2
ON table1.fk_tb2 = table2.id --many2one
GROUP BY
table2.id
I would use DetachedCriteria class.....how can i do this ?
Try using projections like this:
Criteria table1Crit = session.createCriteria("table1");
Criteria table2Crit = table1Crit.createCriteria("table2", Criteria.INNER_JOIN);
table2Crit.setProjection( Property.forName("id").count() );
use your query with createNativeQuery method