Lets assume the following situation: I have two entity classes Person and Comment.
//Person class
@Entity
public class Person extends AbstractBusinesObject{
@OneToMany(mappedby ="owner")
private List<Comment> comments;
//Comment class
@Entity
public class Comment extends AbstractBusinesObject{
@ManyToOne
private Person owner;
I would like to know which approach is better (performance, memory-efficiency) from the following two alternatives to determine whether a specific comment belongs to a specific person or not.
public boolean isUsersComment(Person p, Comment c){
return p.getComments().contains(c);
}
or
public boolean isUsersComment(Person p, Comment c){
CriteriaBuilder cb = getEntitymanager().getCriteriaBuilder();
CriteriaQuery<Comment> cq = cb.createQuery(Comment.class);
Root<Comment> root = cq.from(Comment.class);
cq.select(root).where(cb.and(cb.equal(root, c), cb.isTrue(root.in(p.getComments())));
try{
getEntityManager().createQuery(cq).getSingleResult();
return true;
} catch (NoResultException ex){
return false;
} catch (NoUniqueResultException ex){
throw new IllegalStateException(message);
}
}
As far as i know criteria api generates a database query, while the first solution searches in a collection.
Thank you for your answer.
ps: I am new to Criteria API, so I also accept fixes/suggestions.
Edit
Ok, I have found out it was a stupid question; As I declared a bidirectional relationship, the solution is very simple:
return c.getOwner.equals(p);
Anyway, what is the approach in case of unidirectional relationship?
Depends if the original Collection is already loaded, no? If the Collection is loaded then its a straight contains() call, otherwise it likely would be the equivalent of what a query (Criteria or JPQL) would do.
If the Collection was large then I'd go through a query always (since I wouldn't want to load the whole collection). Otherwise I'd use Collection.contains