Index impact on used memeory

Question

Hi all,
I'm wondering what impact index has on the memory used by the grid?
I'm to keep some "tree like" objects in the grid. Object structure is quite simple. Object of type A contains a list of objects of type B. The list is accessible through A.getListOfB()
I would like to get all A objects that contain a given B object.
The idea was to use an index to fasten the query. The query itself is based on a filter like this:
B b = new B(...);
myCache().invokeAll( new ContainsFilter("getListOfB", b), entryProcessor);
However, before doing that, I'd like to clarify my understanding of index in Coherence.
What happens when we create an index of this kind : cache.addIndex(new ReflectionExtractor("getListOfB"), false, null);
Is it a kind of "inverted map" created in memory?
Is it efficient doing it that way?
What are the other options?
Rgds,
Harry

Answer 1

Here is the example codes I used to test it.
The filter class
public class ListFilter extends ExtractorFilter implements IndexAwareFilter {
     public ListFilter () {
          super();
     @Override
     public Filter applyIndex(Map mapIndexes, Set setKeys) {
          MapIndex index = (MapIndex)mapIndexes.get(new IdentityExtractor());
          if (index == null) {
               return this;
          Map mapContents = index.getIndexContents();
          for (Iterator iter = mapContents.entrySet().iterator(); iter.hasNext();) {
               Map.Entry entry = (Map.Entry) iter.next();
               /* the Entry we received here is B -> A like Robert mentioned, not {B1,B2,...} -> A */
               System.out.println(entry);
          return null;
     @Override
     public int calculateEffectiveness(Map arg0, Set arg1) {
          return 0;
     @Override
     protected boolean evaluateExtracted(Object arg0) {
          return false;
}The testing code
CacheFactory.ensureCluster();
NamedCache ca = CacheFactory.getCache("ACustomer");
ArrayList<String> list = new ArrayList<String>();
list.add("One");
list.add("Two");
list.add("Three");
ca.put("A", list);
ca.addIndex(new IdentityExtractor(), false, null);
ca.entrySet(new ListFilter());