I got the following exception :

java.util.ConcurrentModificationException
at java.util.HashMap$HashIterator.nextEntry(HashMap.java:793)
at java.util.HashMap$KeyIterator.next(HashMap.java:828)
at org.hibernate.collection.AbstractPersistentCollection$IteratorProxy.next(AbstractPersistentCollection.java:580)
at org.xxxxx.yyyyyy.NetGroupPeerTest.updateGroupP(NetGroupPTest.java:265)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)

This error occurs during JUNIT test for a class that uses many-to-many relationship implemented by an association class (table)
We are inside the scope of the session under JUNIT framework.

Involved code is :

@Test
public void updateGroupP() {
try {

Query q = session.createQuery("from NetGroup gpp where gpp.netId = :netid");
NetGroup group1 = (NetGroup) q.setString("netid", "uri:000001").uniqueResult();

Set<NetGroupP> gpp = group1.getGroupP();

for(NetGroupP elt : gpp) {
gpp.remove(elt);
}
}

The error occurs just after removing the elt from the collection. What is strange : removing the elt by implementing a method inside the owner class of the collection does not trigger the exception!?


group1.removeP(elt);

public NetGroup {

public void removeP(NetP p) {
for (NetGroupP gpp : groupPs) {
if( gpp.getP().getPId() == p.getPId()) {
groupP.remove(gpp);
}
}
}
}

So, to be brief, getting the collection via a getter and calling its method remove() triggers an ConcurrentModificationException but doing the "same thing" inside the class that holds the collection does not ! Note that the hascode() for the Set of the collection is changing after deleting an object from the collection. According to the JavaDoc, this behavior is normal :

From : http://download.oracle.com/javase/1.5.0 ... Code%28%29
"Returns the hash code value for this set. The hash code of a set is defined to be the sum of the hash codes of the elements in the set, where the hashcode of a null element is defined to be zero. This ensures that s1.equals(s2) implies that s1.hashCode()==s2.hashCode() for any two sets s1 and s2, as required by the general contract of the Object.hashCode method."

Does anybody has a beginning of explanation ? since that apparently we are on the same thread, we got hopefully the same object and we are inside the same hibernate session.

Regards,

I think I've found an explanation, at least a beginning...! that makes me to laugh about me.... It is because I'm cutting the branch were I'm sitting on... I mean the code is deleting an elt from the collection which is itself include in a loop ! However, it doe not explain why the "same code" runs perfectly inside the owner class ! So, to resolve the problem from the external point of view, the objects to be deleted are collected and an another loop performs the delete operation.