neureka.devices.opencl.utility.WeakTensorReference Maven / Gradle / Ivy
Show all versions of neureka Show documentation
package neureka.devices.opencl.utility;
import java.lang.ref.ReferenceQueue;
import java.lang.ref.WeakReference;
public class WeakTensorReference extends WeakReference
{
/**
* Creates a new phantom reference that refers to the given object and
* is registered with the given queue.
*
* It is possible to create a phantom reference with a {@code null}
* queue, but such a reference is completely useless: Its {@code get}
* method will always return {@code null} and, since it does not have a queue,
* it will never be enqueued.
*
* @param referent the object the new phantom reference will refer to
* @param q the queue with which the reference is to be registered,
* or {@code null} if registration is not required
*/
private final int _hash;
public WeakTensorReference( T referent, ReferenceQueue q ) {
super(referent, q);
_hash = referent.hashCode();
}
@Override
public int hashCode() { return _hash; }
@Override
public boolean equals( Object o ) {
if ( !(o instanceof WeakTensorReference) ) return false;
if ( o.hashCode() == _hash ) return true;
else {
T value = get();
return value != null && o == value;
}
}
}