Wraps an Iterator so that it can act as a distributed
iterator. A distributed iterator is an iterator where the collection
is stored on a server and elements are transmitted across the network
one element at a time on demand. This contrasts with serialization of
the collection, where the entire collection is transmitted across the
network at once.
If you have a collection whose elements you want to make available
across the network using the distributed iterator paradigm, you
retrieve an iterator for the collection and wrap it with a
DistributedIterator implementation. You then pass the
distributed iterator to a ClientIterator and pass
that across the network. Consider the following RMI method that
returns a distributed iterator for its remote method cats():
private ArrayList cats;
public Iterator cats() throws RemoteException {
DistributedIterator dist = new DistributedIteratorImpl(cats.iterator());
ClientIterator it = new ClientIterator(dist);
return it;
}
The result of this method is that an empty iterator is sent across
the network to the client. That empty iterator knows how to retrieve
each cat from the cats ArrayList from the server on demand
as the client application calls for them. If the client only asks for
the first cat, only the first cat is ever sent across the network.
If the collection of cats contains 1 million cats, the client does
not need to wait on that entire collection to be transmitted across
the network before it can access the first cat.
Last modified $Date: 1999/11/06 18:38:04 $ |
