Well, there are no bugs, only features! ;) But we still should change the feature to properly match the spec semantics of "not". :) Can you please open a ticket and attach your test case bellow? Should be an easy fix. Thanks, Edson 2010/1/12 Taras Svirskyi <stware(a)gmail.com&gt; > Hi, > I don't know exactly is it a bug or am I doing something wrong. > Consider following example. > There are nodes that are connected into directed graph. Each node has > special collection of attributes. Under some conditions attributes > should be propagated along the edges in the graph. > > Lets model this with following code. > > /******* propagation rules **********************/ > package com.test.collection; > > rule "move child" > when > c : Child() > s : CNode( children contains c) > l : Link( src == s, > t : target && target.children not contains c ) > then > System.out.println("adding "+c+" from "+s+" to "+t); > t.addChild(c); > update(t); > end > > > /** Child.java (represent special attributes of a node) *******/ > package com.test.collection; > > public class Child { > private String name; > public Child(String name){ > this.name = name; > } > public String getName() { > return name; > } > @Override > public String toString() { > return "child("+name+")"; > } > } > > /********* CNode.java - nodes of a graph *************/ > package com.test.collection; > > import java.util.HashSet; > import java.util.Set; > > public class CNode { > private Set<Child> children; > > public CNode() { > this.children = new HashSet<Child>(); > } > > public Set<Child> getChildren() { > return children; > } > > public void addChild(Child ch){ > children.add(ch); > } > > @Override > public String toString() { > return "Node{children="+children+"}"; > } > } > > /**** Link.java - edges of a graph *******/ > package com.test.collection; > > public class Link { > private CNode src; > private CNode target; > > public Link(CNode src, CNode target) { > this.src = src; > this.target = target; > } > > public CNode getSrc() { > return src; > } > > public CNode getTarget() { > return target; > } > } > > /********************* populate session with facts **********/ > Child c0 = new Child("x"); > Child c1 = new Child("y"); > Child c2 = new Child("z"); > > CNode src = new CNode(); > src.addChild(c0); > src.addChild(c1); > > CNode target = new CNode(); > target.addChild(c0); > > Link link = new Link(src,target); > > // populate with facts > ksession.insert(c0); > ksession.insert(c1); > ksession.insert(c2); > ksession.insert(link); > ksession.insert(src); > ksession.insert(target); > > /****************************************************************/ > So with all that code listed below I've got following output from rules: > -- > adding child(x) from Node{children=[child(x), child(y)]} to > Node{children=[child(x)]} > -- > That means that 'not contains' operator actually works as 'contains' > one. Still when I changed it to older operator 'excludes' everything > goes right printing out. > --- > adding child(y) from Node{children=[child(y), child(x)]} to > Node{children=[child(x)]} > --- > > I've noticed that rewriting rule in form of simple statements also fix > issue. > > rule "move child" > when > c : Child() > s : CNode( children contains c) > t : CNode( children not contains c ) > l : Link( src == s, target == t) > then > System.out.println("adding "+c+" from "+s+" to "+t); > t.addChild(c); > update(t); > end > > > regards, > Taras > _______________________________________________ > rules-dev mailing list > rules-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-dev > -- Edson Tirelli JBoss Drools Core Development JBoss by Red Hat @ www.jboss.com