I think we may have found the cause of an OutOfMemoryError that has been troubling us for some time.

Today we tracked it down to a one-to-many mapping using a List with an index. It would appear that the error was triggered because the value of the index was very large.

Consider the following code in net.sf.hibernate.collection.List:

/**
* @see PersistentCollection#readFrom(ResultSet, CollectionPersister)
*/
public Object readFrom(ResultSet rs, CollectionPersister persister, Object owner)
throws HibernateException, SQLException {

Object element = persister.readElement(rs, owner, session) ;
int index = ( (Integer) persister.readIndex(rs, session) ).intValue();
for ( int i = list.size(); i<=index; i++) {
list.add(i, null);
}
list.set(index, element);
return element;
}

If index is a very large value, say 10,000,000 and the size of the list is small, then a very large number of elements will be added to the list.

Now I am wondering what the purpose of this code is? What is the intent? Is this a bug?

I our case we were able to get around the problem by changing the mapping to not use an index. However, it still looks like a bug to me.

no, it is not a bug. obviously.

Gavin,

Sorry, but it is not obvious to me. What is the intention of adding the null objects to the list in the for loop?

Regards,
Keith

P.S. As I'm sure you know, I like Hibernate. I'm just trying to get to the bottom of the problem we experienced. It it's not a bug, fine. I'm simply trying to understand and make a contribution.

The value of the index column is interpreted as the actual list index, not just an "ordering". Perhaps what you really want is a Map.

Ah. Obviously we were using the index inappropriately.