Hi all.
I admit defeat on my preivous attempt at getting an answer to a question here on the Hibernate Users forum (
viewtopic.php?f=1&t=999915&start=0).
I can only assume that the reason why I did not get any answers to my troubles was that I did not clearly explain my question, so I shall try to do better job now.
______________________________________
I have a
J2SE application where I use JPA with Hibernate as persistence provider. The versions of the required hibernate jars that I use are:
hibernate3.jar (3.3.2.GA)
hibernate-annotations.jar (3.4.0.GA)
hibernate-commons-annotations.jar (3.1.0.GA)
hibernate-entitymanager.jar (3.4.0.GA)
I have an entity where i have a bidirectional One-To-Many mapping. This is initialized to an empty list, something that clearly is allowed according to the JPA 1.0 spec.
i.e. I have:
Code:
in class Parent:
@OneToMany(mappedBy="parent", cascade = CascadeType.ALL)
private List<Child> children = ArrayList<Child>();
class Child:
@ManyToOne
@JoinColumn(name="PARENT_ID")
private Parent parent;
A certain sequence of actions result in a HibernateException being thrown (wrapped by JPA in a RollbackException):
- Obtain a new EntityManager
- Begin a transaction
- Fetch a Parent entity from the entity manager (this exists in the db), call this e.
- Apply an EntityManager.remove() on e
- Commit the transaction
- Begin a transaction
- Apply an EntityManager.persist() on e
- Commit the transaction
The exception is thrown on the last commit():
Code:
Caused by: org.hibernate.HibernateException: Found two representations of same collection: Parent.children
at org.hibernate.engine.Collections.processReachableCollection(Collections.java:176)
at org.hibernate.event.def.FlushVisitor.processCollection(FlushVisitor.java:60)
at org.hibernate.event.def.AbstractVisitor.processValue(AbstractVisitor.java:124)
...
Does anyone know why this happens? I have tried to flush(), clear(), and pretty much everything on the entity manager (event fetching the Hibernate Sesssion and clearing this) - the only thing that helps is creating a new one before attempting the last transaction.
Moreover, after some investigation, the persisted entity actually seems to be persisted correctly - despite the exception (this also seems to hold true for any cascaded persists), perhaps the only erroneous thing here is the fact that an exception is thrown when it really should not?
I challenge anyone who knows anything about this issue to please share their thoughts.
-Horat
Login
FAQ
Search