How to use Hibernate SQL projection query? - java

I have a SQL projection query that returns a list, but I want a single result. Here is my projection query:
Criteria criteria = getCurrentSession().createCriteria(TbLoyaltytrans.class)
.add(Restrictions.eq("custid", custId))
.setProjection(Projections.projectionList()
.add(Projections.sum("pointin").as("pointin"))
.add(Projections.sum("pointout").as("pointout"))
.add(Projections.sqlProjection(
"(pointin - pointout) as points",
new String[]{"points"},
new org.hibernate.type.DoubleType[]{
new org.hibernate.type.DoubleType()
}),
"points")
);
I want a single object, i.e.: the count from the above query.
What I did wrong with the above query?

I assume your are using criteria.list() in order to get the list. You can use criteria.uniqueResult() and it will return a single Object or null.

Related

Hibernate projection property whole object

I have a case like this:
Class Foo with two children (A and B), each being Objects.
In Hibernate, if I want to return only a list of the children, I would use projections on my criteria:
criteria.setProjection(Projections.property("A"));
This gives me a list of A objects, but they are all lazy loaded. As soon as I try to access anything other than the id, obviously things go wrong.
My SQL query indeed shows it:
select A from Foo ...
Logically, only my id is filled in, and not the rest of my properties. How do I solve this problem so I get a list of A objects that have everything filled in?
I tried this:
criteria.setResultTransformer(Transformers.aliasToBean(A.class));
but without success.
if you use hql it would be more efficient:
String hql = "SELECT f.A FROM Foo f";
Query query = session.createQuery(hql);
List results = query.list();
Using criteria I would have use this code
Criteria crit = session.createCriteria(Foo.class);
ProjectionList proList = Projections.projectionList();
proList.add(Projections.property("A"));
crit.setProjection(proList);
List As= crit.list();
or this block
Criteria crit = session.createCriteria(Foo.class);
crit.createAlias("A","a").setProjection(Projections.property("a"));
List As= crit.list();

Return single column from multi-group Hibernate projection

What I want to achieve:
Select a susbset of entities that have a property value that exists in a List of values. This list is returned by another Query.
In plain SQL, this can be easily achieved with subqueries. This is the Criteria query that returns the relevant values:
DetachedCriteria subq = DetachedCriteria.forClass(MyClass.class)
.setProjection(
Projections.projectionList()
.add(Projections.alias(Projections.max("interestingVal"), "interestingVal"))
.add(Projections.groupProperty("someval1"))
.add(Projections.groupProperty("someval2"))
.add(Projections.groupProperty("someval3"))
);
I would then like to select the entities that have a value of interesting_val that was returned by the above query. Like so:
List<MyClass> resultList = sessionFactory.getCurrentSession().createCriteria(MyClass.class)
.add(Subqueries.propertyIn("interestingVal", subq))
.list();
Unfortunately, the subq subquery returns a List of Object-arrays with length 4, since I have 4 Projection values. All projection values seem to be automatically added to the SELECT clause. This results in an
SQLSyntaxErrorException: ORA-00913: too many values.
How can I either tell my second Criteria query to only use the first element in the Object array or only retrieve the first column in my subquery?
Instead of propertyIn try propertiesIn.
String[] vals = new String[]{"interestingVal", "someval1", "someval2", "someval3"};
List<MyClass> resultList = sessionFactory.getCurrentSession().createCriteria(MyClass.class)
.add(Subqueries.propertiesIn(vals, subq))
.list();

Hibernate select lowest value greater than

I am trying to get hibernate to return an object that has the lowest ID number that is greater than a certain value.
The code I have is returning an integer though, instead of the object.
Code
Criteria criteria = session.createCriteria(Story.class);
Criterion storyId = Restrictions.ge("id", 2);
ProjectionList projList = Projections.projectionList();
projList.add(Projections.min("id"));
criteria.add(storyId);
criteria.setProjection(projList);
List<?> storyResult = criteria.list();
session.close();
Story story = (Story) storyResult.get(0);
Right now this is returning "3" as an integer. 3 is the next available ID, but why is hibernate giving me back an integer instead of the object?
Thanks
By adding a Projection, you are essentially asking Hibernate to change your query from SELECT * FROM TABLE to a SELECT MIN(ID) FROM TABLE query. That's one of the features of a projection.
To achieve what you actually want to achieve, you need to use the projection to get the ID and then use another query to retrieve the object.
Criteria idQuery = session.createCriteria(Story.class);
idQuery.add(Restrictions.ge("id", 2));
idQuery.setProjection(Projections.min("id"));
int returnedId = criteria.uniqueResult();
Criteria storyQuery = session.createCriteria(Story.class);
storyQuery.add(Restrictions.idEq(returnedId);
Story story = (Story) storyQuery.uniqueResult();
return story;

get all the values in where clause for empty stirng using hibernate

i am building a shopping cart using jsp and hibernate.
i am filtering the content by brand size and price using checkboxes
the checked checkboxes are returned to the class where hql query exists.
so i want i single hql query that can handle this.
as like if one of the parameter like size is empty (means user doesnt uses it to filter the content ) than an empty string is passed to the hql query which returns any value...
so is there anything possible that all values can be retrived in where clause for empty string or some other alternative except coding different methods for different parameter...
I typically use the Criteria api for things like this... if the user does not specify a size, do not add it to the criteria query.
Criteria criteria = session.createCriteria(MyClass.class);
if(size != null && !size.isEmpty()){
criteria.add(Restrictions.eq("size", size);
}
To have multiple restrictions via an OR statement, you use Disjunction. For an AND, you use Conjunction.
Criteria criteria = session.createCriteria(MyClass.class);
Disjunction sizeDisjunction = Restrictions.disjunction();
String[] sizes = { "small", "medium", "large" };
for(int i = 0; i < sizes.length; i++){
sizeDisjunction.add(Restrictions.eq("size", sizes[i]);
}
criteria.add(sizeDisjunction );
First, good practices say that instead of passing and empty String to the query, you should pass null instead. That said, this hql should help you:
from Product p
where p.brand = coalesce(:brand, p.brand)
and p.size = coalesce(:size, p.size)
and p.price = coalesce (:price, p.price)

How do I find a value in a column that just have unique values with EclipseLink?

You have the EntityManager.find(Class entityClass, Object primaryKey) method to find a specific row with a primary key.
But how do I find a value in a column that just have unique values and is not a primary key?
You can use appropriate JPQL with TypedQuery.
try {
TypedQuery<Bean> tq = em.createQuery("from Bean WHERE column=?", Bean.class);
Bean result = tq.setParameter(1, "uniqueKey").getSingleResult();
} catch(NoResultException noresult) {
// if there is no result
} catch(NonUniqueResultException notUnique) {
// if more than one result
}
For example, like this:
List<T> results = em.createQuery("SELECT t FROM TABLE t", T.class)
.getResultList();
With parameters:
List<T> results = em.createQuery("SELECT t FROM TABLE t where t.value = :value1")
.setParameter("value1", "some value").getResultList();
For single result replace getResultList() with getSingleResult():
T entity = em.createQuery("SELECT t FROM TABLE t where t.uniqueKey = :value1")
.setParameter("value1", "KEY1").getSingleResult();
One other way is to use Criteria API.
You can use a Query, either JPQL, Criteria, or SQL.
Not sure if your concern is in obtaining cache hits similar to find(). In EclipseLink 2.4 cache indexes were added to allow you to index non-primary key fields and obtain cache hits from JPQL or Criteria.
See,
http://wiki.eclipse.org/EclipseLink/UserGuide/JPA/Basic_JPA_Development/Caching/Indexes
Prior to 2.4 you could use in-memory queries to query the cache on non-id fields.
TL;DR
With in DSL level - JPA no practice mentioned in previous answers
How do I find a value in a column that just have unique values and is not a primary key?
There isn't specification for query with custom field with in root interface of javax.persistence.EntityManager, you need to have criteria base query.
CriteriaBuilder criteriaBuilder = entityManager.getCriteriaBuilder();
CriteriaQuery<R> criteriaQuery = criteriaBuilder.createQuery(EntityType.class)
Root<R> root = criteriaQuery.from(type);
criteriaBuilder.and(criteriaBuilder.equal(root.get(your_field), value));
You can also group your predicates together and pass them all together.
andPredicates.add(criteriaBuilder.and(root.get(field).in(child)));
criteriaBuilder.and(andPredicates.toArray(new Predicate[]{});
And calling result(rather single entity or a list of entities) with
entityManager.createQuery(suitable_criteria_query).getSingleResult();
entityManager.createQuery(suitable_criteria_query).getResultList();

Categories