Hi, usings Drools 4.0.4 and MVEL 1.4, this simple rule: --- package test; import java.util.Collections; dialect "mvel" rule "Rule #1" when then insert(Collections.singletonMap("content", "hello")); end -- produces: java.lang.IllegalAccessError: class org.drools.shadow.java.util.Collections$SingletonMapShadowProxy cannot access its superclass java.util.Collections$SingletonMap at java.lang.ClassLoader.defineClass1(Native Method) at java.lang.ClassLoader.defineClass(ClassLoader.java:620) at org.drools.rule.MapBackedClassLoader.fastFindClass( MapBackedClassLoader.java:60) at org.drools.rule.MapBackedClassLoader.loadClass( MapBackedClassLoader.java:79) at java.lang.ClassLoader.loadClass(ClassLoader.java:251) at org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy( Rete.java:547) at org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData(Rete.java :494) at org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java :461) at org.drools.reteoo.Rete.assertObject(Rete.java:152) at org.drools.reteoo.ReteooRuleBase.assertObject( ReteooRuleBase.java:192) at org.drools.reteoo.ReteooWorkingMemory.doInsert( ReteooWorkingMemory.java:71) at org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:909) at org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:881) at org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:67) at org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:61) It's not clear to me why Drools creates Proxies for such classes as java.util.Collections, or does MVEL do it? - Godmar _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Forgot to say that we are doing an effort to completely remove shadow proxies in the next major version.

2008/2/19, Edson Tirelli <tirelli(a)post.com <mailto:tirelli@post.com>>: Drools tries to create the ShadowProxy. The reason is that it does not know about the implementation... it just knows it is a Map and as so, it must be shadowed. Problem is that SingletonMap is either final or does not have a default constructor. My recommendation, besides opening a JIRA for this, is avoid inserting collections/maps directly as facts. The semantic for such facts is not clearly defined and it may cause undesired behavior. []s Edson 2008/2/19, Godmar Back <godmar(a)gmail.com <mailto:godmar@gmail.com>>: Hi, usings Drools 4.0.4 and MVEL 1.4, this simple rule: --- package test; import java.util.Collections; dialect "mvel" rule "Rule #1" when then insert(Collections.singletonMap("content", "hello")); end -- produces: java.lang.IllegalAccessError: class org.drools.shadow.java.util.Collections$SingletonMapShadowProxy cannot access its superclass java.util.Collections$SingletonMap at java.lang.ClassLoader.defineClass1(Native Method) at java.lang.ClassLoader.defineClass(ClassLoader.java:620) at org.drools.rule.MapBackedClassLoader.fastFindClass(MapBackedClassLoader.java:60) at org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java:79) at java.lang.ClassLoader.loadClass(ClassLoader.java:251) at org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java:547) at org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData(Rete.java:494) at org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) at org.drools.reteoo.Rete.assertObject(Rete.java:152) at org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java:192) at org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java:71) at org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:909) at org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:881) at org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:67) at org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:61) It's not clear to me why Drools creates Proxies for such classes as java.util.Collections, or does MVEL do it? - Godmar _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org <mailto:rules-users@lists.jboss.org> https://lists.jboss.org/mailman/listinfo/rules-users -- Edson Tirelli JBoss Drools Core Development Office: +55 11 3529-6000 Mobile: +55 11 9287-5646 JBoss, a division of Red Hat @ www.jboss.com <http://www.jboss.com> -- Edson Tirelli JBoss Drools Core Development Office: +55 11 3529-6000 Mobile: +55 11 9287-5646 JBoss, a division of Red Hat @ www.jboss.com <http://www.jboss.com> ------------------------------------------------------------------------ _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

As a general comment, the examples for which I find Drools failing are not the actual examples for which my application is failing. It's just the smallest test case I was able to eliminate. I'm now a bit concerned about your comment that Maps and Collections aren't well-defined as Facts. I am planning to make extensive use of them (that's also why I'd prefer to use the MVEL dialect, because in Java I cannot do this without creating Bean wrappers.) Could you elaborate what makes the semantics not "well-defined". I'm specifically concerned with immutable maps (such as the one that would have been returned by Collections.singletonMap), and with collections of maps (such as those obtained via a "from"..." clause). I need to insert immutable maps as facts; I understand that the items returned by "from" aren't inserted as facts. - Godmar On Feb 19, 2008 3:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > Drools tries to create the ShadowProxy. The reason is that it does not > know about the implementation... it just knows it is a Map and as so, it > must be shadowed. Problem is that SingletonMap is either final or does not > have a default constructor. > My recommendation, besides opening a JIRA for this, is avoid inserting > collections/maps directly as facts. The semantic for such facts is not > clearly defined and it may cause undesired behavior. > > []s > Edson > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > > > > > > Hi, > > > > usings Drools 4.0.4 and MVEL 1.4, this simple rule: > > --- > > package test; > > > > import java.util.Collections; > > > > dialect "mvel" > > > > rule "Rule #1" > > when > > then > > insert(Collections.singletonMap("content", "hello")); > > end > > -- > > > > produces: > > java.lang.IllegalAccessError: class > > org.drools.shadow.java.util.Collections$SingletonMapShadowProxy cannot > > access its superclass java.util.Collections$SingletonMap > > at java.lang.ClassLoader.defineClass1(Native Method) > > at java.lang.ClassLoader.defineClass(ClassLoader.java:620) > > at > org.drools.rule.MapBackedClassLoader.fastFindClass( MapBackedClassLoader.java:60) > > at > org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java :79) > > at java.lang.ClassLoader.loadClass(ClassLoader.java:251) > > at > org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java :547) > > at > org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData( Rete.java:494) > > at > org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) > > at org.drools.reteoo.Rete.assertObject(Rete.java:152) > > at > org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java:192) > > at > org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java :71) > > at > org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:909) > > at > org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:881) > > at > org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:67) > > at > org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:61) > > > > > > It's not clear to me why Drools creates Proxies for such classes as > > java.util.Collections, or does MVEL do it? > > > > - Godmar > > _______________________________________________ > > rules-users mailing list > > rules-users(a)lists.jboss.org > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

On Feb 20, 2008 9:23 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > Godmar, > > Short answer: collection/maps objects are not javabeans. > Explain why this is a problem. What is it about JavaBeans that your algorithm relies upon? Is it the fact that the set of properties remains fixed and can be determined at (fact) insertion time via reflection? Otherwise, I do not see any conceptual difference between a map and a bean. If that is the difference, then please allow maps with an immutable key set. - Godmar > Long answer: collection/maps must be shadowed to ensure consistency > during execution, but how can we shadow the data if it is not exposed in a > default, spec manner, as in javabeans? The algorithm we have in place right > now is bellow. As you can see, it is a weak algo, but was the best I could > come up at that time. If you have any suggestions on how to improve that, I > appreciate. > > public Object getShadow(final Object fact) throws RuntimeDroolsException > { > ShadowProxy proxy = null; > if ( isShadowEnabled() ) { > try { > if ( Collection.class.isAssignableFrom( this.shadowClass) > || Map.class.isAssignableFrom( this.shadowClass ) ) { > // if it is a collection, try to instantiate using > constructor > try { > proxy = (ShadowProxy) > this.shadowClass.getConstructor( new Class[]{cls} ).newInstance( new > Object[]{fact} ); > } catch ( Exception e ) { > // not possible to instantiate using constructor > } > } > if ( proxy == null ) { > if ( this.instantiator == null ) { > this.setInstantiator(); > } > proxy = (ShadowProxy) this.instantiator.newInstance (); > } > > proxy.setShadowedObject( fact ); > } catch ( final Exception e ) { > System.out.println( "shadow: " +proxy.getClass() + ":" + > fact.getClass() ); > throw new RuntimeDroolsException( "Error creating shadow > fact for object: " + fact, > e ); > } > } > return proxy; > > > } > > []s > Edson > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > As a general comment, the examples for which I find Drools failing are > > not the actual examples for which my application is failing. It's just > > the smallest test case I was able to eliminate. > > > > I'm now a bit concerned about your comment that Maps and Collections > > aren't well-defined as Facts. I am planning to make extensive use of > > them (that's also why I'd prefer to use the MVEL dialect, because in > > Java I cannot do this without creating Bean wrappers.) > > > > Could you elaborate what makes the semantics not "well-defined". > > > > I'm specifically concerned with immutable maps (such as the one that > > would have been returned by Collections.singletonMap), and with > > collections of maps (such as those obtained via a "from"..." clause). > > I need to insert immutable maps as facts; I understand that the items > > returned by "from" aren't inserted as facts. > > > > - Godmar > > > > On Feb 19, 2008 3:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > > > > Drools tries to create the ShadowProxy. The reason is that it does > not > > > know about the implementation... it just knows it is a Map and as so, it > > > must be shadowed. Problem is that SingletonMap is either final or does > not > > > have a default constructor. > > > My recommendation, besides opening a JIRA for this, is avoid > inserting > > > collections/maps directly as facts. The semantic for such facts is not > > > clearly defined and it may cause undesired behavior. > > > > > > []s > > > Edson > > > > > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > > > > > > > > > > > > > > Hi, > > > > > > > > usings Drools 4.0.4 and MVEL 1.4, this simple rule: > > > > --- > > > > package test; > > > > > > > > import java.util.Collections; > > > > > > > > dialect "mvel" > > > > > > > > rule "Rule #1" > > > > when > > > > then > > > > insert(Collections.singletonMap("content", "hello")); > > > > end > > > > -- > > > > > > > > produces: > > > > java.lang.IllegalAccessError: class > > > > org.drools.shadow.java.util.Collections$SingletonMapShadowProxycannot > > > > access its superclass java.util.Collections$SingletonMap > > > > at java.lang.ClassLoader.defineClass1(Native Method) > > > > at java.lang.ClassLoader.defineClass(ClassLoader.java:620) > > > > at > > > > org.drools.rule.MapBackedClassLoader.fastFindClass( MapBackedClassLoader.java:60) > > > > at > > > > org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java :79) > > > > at java.lang.ClassLoader.loadClass(ClassLoader.java:251) > > > > at > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java :547) > > > > at > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData( Rete.java:494) > > > > at > > > org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) > > > > at org.drools.reteoo.Rete.assertObject(Rete.java:152) > > > > at > > > org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java :192) > > > > at > > > > org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java :71) > > > > at > > > > org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:909) > > > > at > > > > org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:881) > > > > at > > > > org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:67) > > > > at > > > > org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:61) > > > > > > > > > > > > It's not clear to me why Drools creates Proxies for such classes as > > > > java.util.Collections, or does MVEL do it? > > > > > > > > - Godmar > > > > _______________________________________________ > > > > rules-users mailing list > > > > rules-users(a)lists.jboss.org > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > > > -- > > > Edson Tirelli > > > JBoss Drools Core Development > > > Office: +55 11 3529-6000 > > > Mobile: +55 11 9287-5646 > > > JBoss, a division of Red Hat @ www.jboss.com > > > _______________________________________________ > > > rules-users mailing list > > > rules-users(a)lists.jboss.org > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > _______________________________________________ > > rules-users mailing list > > rules-users(a)lists.jboss.org > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

I don't really understand what you mean by "shadow". What is the purpose of such shadowing. Mark's email implies that it has to do with concurrency protection; it's not clear what that means. In my view, whatever purpose you pursue with "shadowing", it does not justify treating beans and maps differently. Your example of class Person shows that. If a person has two attributes, name and age, then this is equivalent to a map with two keys 'name' and 'age'. Here's the mapping: p.getName() corresponds to m["name"] p.getAge() corresponds to m["age"] and setName/setAge accordingly. Mathematically, a bean is an associative array with a fixed set of keys (called "properties") that map to values. For all practical purposes, that is the same as a map. There's no reason to treat them differently. Wherever you'd do "getXXX()" with a bean you'd do .get("XXX") with a map. - Godmar On Feb 20, 2008 11:25 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > Ok, let me show one example. Imagine the class Person, with 2 attributes > (name and age) and the corresponding getter/setters. > What are the data for that fact that must be shadowed? easy answer: just > shadow all getXXX() methods (getName and getAge). > > Now, take a Map. What is the data that must be shadowed? > > So, we do our best to work with facts that don't follow the javabean > spec, but collections and maps are a complicated beast. Again, if you have > suggestions on how to improve the current support we provide for them, > please share with us. > > []s > Edson > > 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > > On Feb 20, 2008 9:23 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > > > > Godmar, > > > > > > Short answer: collection/maps objects are not javabeans. > > > > > > > Explain why this is a problem. > > > > What is it about JavaBeans that your algorithm relies upon? Is it the > > fact that the set of properties remains fixed and can be determined at > > (fact) insertion time via reflection? > > > > Otherwise, I do not see any conceptual difference between a map and a > bean. > > If that is the difference, then please allow maps with an immutable key > set. > > > > - Godmar > > > > > > > Long answer: collection/maps must be shadowed to ensure consistency > > > during execution, but how can we shadow the data if it is not exposed in > a > > > default, spec manner, as in javabeans? The algorithm we have in place > right > > > now is bellow. As you can see, it is a weak algo, but was the best I > could > > > come up at that time. If you have any suggestions on how to improve > that, I > > > appreciate. > > > > > > public Object getShadow(final Object fact) throws > RuntimeDroolsException > > > { > > > ShadowProxy proxy = null; > > > if ( isShadowEnabled() ) { > > > try { > > > if ( Collection.class.isAssignableFrom( this.shadowClass > ) > > > || Map.class.isAssignableFrom( this.shadowClass ) ) { > > > // if it is a collection, try to instantiate using > > > constructor > > > try { > > > proxy = (ShadowProxy) > > > this.shadowClass.getConstructor( new Class[]{cls} ).newInstance( new > > > Object[]{fact} ); > > > } catch ( Exception e ) { > > > // not possible to instantiate using constructor > > > } > > > } > > > if ( proxy == null ) { > > > if ( this.instantiator == null ) { > > > this.setInstantiator(); > > > } > > > proxy = (ShadowProxy) > this.instantiator.newInstance(); > > > } > > > > > > proxy.setShadowedObject( fact ); > > > } catch ( final Exception e ) { > > > System.out.println( "shadow: " +proxy.getClass() + ":" + > > > fact.getClass() ); > > > throw new RuntimeDroolsException( "Error creating shadow > > > fact for object: " + fact, > > > e ); > > > } > > > } > > > return proxy; > > > > > > > > > } > > > > > > []s > > > Edson > > > > > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > > As a general comment, the examples for which I find Drools failing are > > > > not the actual examples for which my application is failing. It's just > > > > the smallest test case I was able to eliminate. > > > > > > > > I'm now a bit concerned about your comment that Maps and Collections > > > > aren't well-defined as Facts. I am planning to make extensive use of > > > > them (that's also why I'd prefer to use the MVEL dialect, because in > > > > Java I cannot do this without creating Bean wrappers.) > > > > > > > > Could you elaborate what makes the semantics not "well-defined". > > > > > > > > I'm specifically concerned with immutable maps (such as the one that > > > > would have been returned by Collections.singletonMap), and with > > > > collections of maps (such as those obtained via a "from"..." clause). > > > > I need to insert immutable maps as facts; I understand that the items > > > > returned by "from" aren't inserted as facts. > > > > > > > > - Godmar > > > > > > > > On Feb 19, 2008 3:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > > > > > > > > Drools tries to create the ShadowProxy. The reason is that it > does > > > not > > > > > know about the implementation... it just knows it is a Map and as > so, it > > > > > must be shadowed. Problem is that SingletonMap is either final or > does > > > not > > > > > have a default constructor. > > > > > My recommendation, besides opening a JIRA for this, is avoid > > > inserting > > > > > collections/maps directly as facts. The semantic for such facts is > not > > > > > clearly defined and it may cause undesired behavior. > > > > > > > > > > []s > > > > > Edson > > > > > > > > > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > > > > > > > > > > > > > > > > > > > > > > Hi, > > > > > > > > > > > > usings Drools 4.0.4 and MVEL 1.4, this simple rule: > > > > > > --- > > > > > > package test; > > > > > > > > > > > > import java.util.Collections; > > > > > > > > > > > > dialect "mvel" > > > > > > > > > > > > rule "Rule #1" > > > > > > when > > > > > > then > > > > > > insert(Collections.singletonMap("content", "hello")); > > > > > > end > > > > > > -- > > > > > > > > > > > > produces: > > > > > > java.lang.IllegalAccessError: class > > > > > > org.drools.shadow.java.util.Collections$SingletonMapShadowProxy > cannot > > > > > > access its superclass java.util.Collections$SingletonMap > > > > > > at java.lang.ClassLoader.defineClass1(Native Method) > > > > > > at java.lang.ClassLoader.defineClass(ClassLoader.java :620) > > > > > > at > > > > > > > > > org.drools.rule.MapBackedClassLoader.fastFindClass( MapBackedClassLoader.java:60) > > > > > > at > > > > > > > > > org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java :79) > > > > > > at java.lang.ClassLoader.loadClass(ClassLoader.java :251) > > > > > > at > > > > > > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java :547) > > > > > > at > > > > > > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData( Rete.java:494) > > > > > > at > > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) > > > > > > at org.drools.reteoo.Rete.assertObject(Rete.java:152) > > > > > > at > > > > > > org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java:192) > > > > > > at > > > > > > > > > org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java :71) > > > > > > at > > > > > > > > > org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:909) > > > > > > at > > > > > > > > > org.drools.common.AbstractWorkingMemory.insert( AbstractWorkingMemory.java:881) > > > > > > at > > > > > > > > > org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:67) > > > > > > at > > > > > > > > > org.drools.base.DefaultKnowledgeHelper.insert( DefaultKnowledgeHelper.java:61) > > > > > > > > > > > > > > > > > > It's not clear to me why Drools creates Proxies for such classes > as > > > > > > java.util.Collections, or does MVEL do it? > > > > > > > > > > > > - Godmar > > > > > > _______________________________________________ > > > > > > rules-users mailing list > > > > > > rules-users(a)lists.jboss.org > > > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > Edson Tirelli > > > > > JBoss Drools Core Development > > > > > Office: +55 11 3529-6000 > > > > > Mobile: +55 11 9287-5646 > > > > > JBoss, a division of Red Hat @ www.jboss.com > > > > > _______________________________________________ > > > > > rules-users mailing list > > > > > rules-users(a)lists.jboss.org > > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > _______________________________________________ > > > > rules-users mailing list > > > > rules-users(a)lists.jboss.org > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > > > -- > > > Edson Tirelli > > > JBoss Drools Core Development > > > Office: +55 11 3529-6000 > > > Mobile: +55 11 9287-5646 > > > JBoss, a division of Red Hat @ www.jboss.com > > > _______________________________________________ > > > rules-users mailing list > > > rules-users(a)lists.jboss.org > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > _______________________________________________ > > rules-users mailing list > > rules-users(a)lists.jboss.org > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

On Feb 20, 2008 12:30 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > Godmar, > > Shadow Facts are a necessary evil in current version. Basically what they > do is keep the working memory consistent in face of attribute changes on the > facts, that may happen both internally and externally to the working memory. > Our implementation to shadow facts is a lazy proxy that caches the > values until a safe point to synchronize the actual attribute value with the > one seen by the engine. > > So, the question is: given an object: > > (Map) fact > > How can we create an identical copy of it (shadow), if there is no > "clone" operation? Can you explain why you require the use of "clone()"? Cloning a map is otherwise easy - it's also referred to as a "shallow copy" -- HashMap's HashMap(Map) constructor will do it. http://java.sun.com/j2se/1.4.2/docs/api/java/util/HashMap.html#HashMap(ja... I know you know that, so explain what necessitates the use of clone(). > More than that, the shadow must be a subclass of it. java.util.HashMap is a subclass of Map. > > Most Collection and Map implementations have a single parameter > constructor that allows us to do: > > > proxy = (ShadowProxy) this.shadowClass.getConstructor( new Class[]{cls} > ).newInstance( new Object[]{fact} ); > > But the SingletonMap you were using does not accept that constructor. > So, one way is to explicit check if the fact is a SingletonMap and handle it > accordingly, but that is a specific class hack... is there any general > solution we can use? > Forget about the SingletonMap. That was just one of the many things I tried and failed. Fundamentally, I would like Drools to process facts that were obtained from real-world sources, and these facts have properties I do not know in advance. Therefore, I cannot use beans (or using beans would be highly inconvenient since it will require changes to Java code whenever I'm referring to a new property, something I'd rather avoid.) - Godmar > > []s > Edson > > 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > I don't really understand what you mean by "shadow". What is the > > purpose of such shadowing. Mark's email implies that it has to do with > > concurrency protection; it's not clear what that means. > > > > In my view, whatever purpose you pursue with "shadowing", it does not > > justify treating beans and maps differently. > > > > Your example of class Person shows that. If a person has two > > attributes, name and age, then this is equivalent to a map with two > > keys 'name' and 'age'. > > > > Here's the mapping: > > > > p.getName() corresponds to m["name"] > > p.getAge() corresponds to m["age"] > > > > and setName/setAge accordingly. > > > > Mathematically, a bean is an associative array with a fixed set of > > keys (called "properties") that map to values. For all practical > > purposes, that is the same as a map. There's no reason to treat them > > differently. Wherever you'd do "getXXX()" with a bean you'd do > > .get("XXX") with a map. > > > > - Godmar > > > > On Feb 20, 2008 11:25 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > > > > Ok, let me show one example. Imagine the class Person, with 2 > attributes > > > (name and age) and the corresponding getter/setters. > > > What are the data for that fact that must be shadowed? easy answer: > just > > > shadow all getXXX() methods (getName and getAge). > > > > > > Now, take a Map. What is the data that must be shadowed? > > > > > > So, we do our best to work with facts that don't follow the javabean > > > spec, but collections and maps are a complicated beast. Again, if you > have > > > suggestions on how to improve the current support we provide for them, > > > please share with us. > > > > > > []s > > > Edson > > > > > > 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > > > > > > On Feb 20, 2008 9:23 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > > > > > > > > Godmar, > > > > > > > > > > Short answer: collection/maps objects are not javabeans. > > > > > > > > > > > > > Explain why this is a problem. > > > > > > > > What is it about JavaBeans that your algorithm relies upon? Is it the > > > > fact that the set of properties remains fixed and can be determined at > > > > (fact) insertion time via reflection? > > > > > > > > Otherwise, I do not see any conceptual difference between a map and a > > > bean. > > > > If that is the difference, then please allow maps with an immutable > key > > > set. > > > > > > > > - Godmar > > > > > > > > > > > > > Long answer: collection/maps must be shadowed to ensure > consistency > > > > > during execution, but how can we shadow the data if it is not > exposed in > > > a > > > > > default, spec manner, as in javabeans? The algorithm we have in > place > > > right > > > > > now is bellow. As you can see, it is a weak algo, but was the best I > > > could > > > > > come up at that time. If you have any suggestions on how to improve > > > that, I > > > > > appreciate. > > > > > > > > > > public Object getShadow(final Object fact) throws > > > RuntimeDroolsException > > > > > { > > > > > ShadowProxy proxy = null; > > > > > if ( isShadowEnabled() ) { > > > > > try { > > > > > if ( Collection.class.isAssignableFrom( > this.shadowClass > > > ) > > > > > || Map.class.isAssignableFrom( this.shadowClass ) ) { > > > > > // if it is a collection, try to instantiate > using > > > > > constructor > > > > > try { > > > > > proxy = (ShadowProxy) > > > > > this.shadowClass.getConstructor( new Class[]{cls} ).newInstance( new > > > > > Object[]{fact} ); > > > > > } catch ( Exception e ) { > > > > > // not possible to instantiate using > constructor > > > > > } > > > > > } > > > > > if ( proxy == null ) { > > > > > if ( this.instantiator == null ) { > > > > > this.setInstantiator(); > > > > > } > > > > > proxy = (ShadowProxy) > > > this.instantiator.newInstance(); > > > > > } > > > > > > > > > > proxy.setShadowedObject( fact ); > > > > > } catch ( final Exception e ) { > > > > > System.out.println( "shadow: " +proxy.getClass() + > ":" + > > > > > fact.getClass() ); > > > > > throw new RuntimeDroolsException( "Error creating > shadow > > > > > fact for object: " + fact, > > > > > e ); > > > > > } > > > > > } > > > > > return proxy; > > > > > > > > > > > > > > > } > > > > > > > > > > []s > > > > > Edson > > > > > > > > > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > > > > As a general comment, the examples for which I find Drools failing > are > > > > > > not the actual examples for which my application is failing. It's > just > > > > > > the smallest test case I was able to eliminate. > > > > > > > > > > > > I'm now a bit concerned about your comment that Maps and > Collections > > > > > > aren't well-defined as Facts. I am planning to make extensive use > of > > > > > > them (that's also why I'd prefer to use the MVEL dialect, because > in > > > > > > Java I cannot do this without creating Bean wrappers.) > > > > > > > > > > > > Could you elaborate what makes the semantics not "well-defined". > > > > > > > > > > > > I'm specifically concerned with immutable maps (such as the one > that > > > > > > would have been returned by Collections.singletonMap), and with > > > > > > collections of maps (such as those obtained via a "from"..." > clause). > > > > > > I need to insert immutable maps as facts; I understand that the > items > > > > > > returned by "from" aren't inserted as facts. > > > > > > > > > > > > - Godmar > > > > > > > > > > > > On Feb 19, 2008 3:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > > > > > > > > > > > > Drools tries to create the ShadowProxy. The reason is that it > > > does > > > > > not > > > > > > > know about the implementation... it just knows it is a Map and > as > > > so, it > > > > > > > must be shadowed. Problem is that SingletonMap is either final > or > > > does > > > > > not > > > > > > > have a default constructor. > > > > > > > My recommendation, besides opening a JIRA for this, is avoid > > > > > inserting > > > > > > > collections/maps directly as facts. The semantic for such facts > is > > > not > > > > > > > clearly defined and it may cause undesired behavior. > > > > > > > > > > > > > > []s > > > > > > > Edson > > > > > > > > > > > > > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Hi, > > > > > > > > > > > > > > > > usings Drools 4.0.4 and MVEL 1.4, this simple rule: > > > > > > > > --- > > > > > > > > package test; > > > > > > > > > > > > > > > > import java.util.Collections; > > > > > > > > > > > > > > > > dialect "mvel" > > > > > > > > > > > > > > > > rule "Rule #1" > > > > > > > > when > > > > > > > > then > > > > > > > > insert(Collections.singletonMap("content", "hello")); > > > > > > > > end > > > > > > > > -- > > > > > > > > > > > > > > > > produces: > > > > > > > > java.lang.IllegalAccessError: class > > > > > > > > > org.drools.shadow.java.util.Collections$SingletonMapShadowProxy > > > cannot > > > > > > > > access its superclass java.util.Collections$SingletonMap > > > > > > > > at java.lang.ClassLoader.defineClass1(Native Method) > > > > > > > > at > java.lang.ClassLoader.defineClass(ClassLoader.java:620) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.rule.MapBackedClassLoader.fastFindClass(MapBackedClassLoader.java:60) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java:79) > > > > > > > > at > java.lang.ClassLoader.loadClass(ClassLoader.java:251) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java:547) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData(Rete.java:494) > > > > > > > > at > > > > > > > org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) > > > > > > > > at org.drools.reteoo.Rete.assertObject(Rete.java:152) > > > > > > > > at > > > > > > > > > > org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java:192) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java:71) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:909) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:881) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:67) > > > > > > > > at > > > > > > > > > > > > > > > > org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:61) > > > > > > > > > > > > > > > > > > > > > > > > It's not clear to me why Drools creates Proxies for such > classes > > > as > > > > > > > > java.util.Collections, or does MVEL do it? > > > > > > > > > > > > > > > > - Godmar > > > > > > > > _______________________________________________ > > > > > > > > rules-users mailing list > > > > > > > > rules-users(a)lists.jboss.org > > > > > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > Edson Tirelli > > > > > > > JBoss Drools Core Development > > > > > > > Office: +55 11 3529-6000 > > > > > > > Mobile: +55 11 9287-5646 > > > > > > > JBoss, a division of Red Hat @ www.jboss.com > > > > > > > _______________________________________________ > > > > > > > rules-users mailing list > > > > > > > rules-users(a)lists.jboss.org > > > > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > > > rules-users mailing list > > > > > > rules-users(a)lists.jboss.org > > > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > Edson Tirelli > > > > > JBoss Drools Core Development > > > > > Office: +55 11 3529-6000 > > > > > Mobile: +55 11 9287-5646 > > > > > JBoss, a division of Red Hat @ www.jboss.com > > > > > _______________________________________________ > > > > > rules-users mailing list > > > > > rules-users(a)lists.jboss.org > > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > _______________________________________________ > > > > rules-users mailing list > > > > rules-users(a)lists.jboss.org > > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > > > > > > > > > > > > -- > > > Edson Tirelli > > > JBoss Drools Core Development > > > Office: +55 11 3529-6000 > > > Mobile: +55 11 9287-5646 > > > JBoss, a division of Red Hat @ www.jboss.com > > > _______________________________________________ > > > rules-users mailing list > > > rules-users(a)lists.jboss.org > > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > > > _______________________________________________ > > rules-users mailing list > > rules-users(a)lists.jboss.org > > https://lists.jboss.org/mailman/listinfo/rules-users > > > > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > >

A blog I wrote a long time ago about dynamically generated beans: http://blog.athico.com/2006/12/dynamically-generated-class-beans-as.html

Godmar, Ok, now I understood what you mean. This ability to use Maps as a "special" type of facts requires the implementation of what we call "pluggable fact types". It is something we wanted to have for ages, but it is not implemented yet. Unfortunately it is not as "easy" as it may appear, but definitively it is a must have for the future, specially as we move to the ontology space. You are right to think that this will give us another level of expressive power. Among several other advantages, it would allow the the user instead of writing patterns like: Map( this.factType == "SomeFactType", ... ) Where "factType" is a key in the Map, to simply write rules: SomeFactType( ... ) More than that, once we have such feature, we can directly use any structure mappable into a Java class as a fact, like Grrovy/Jython/Whatever beans, CSV files, XML documents, etc. So, it is REALLY powerful and really cool! Our only limitation is indeed "(wo)man power"... :) Seriously speaking, we already have pluggable dialects, pluggable evaluators and an "almost pluggable" extractor framework. All of them are pre-reqs for the pluggable fact types, but we did not reached that point yet. Talking about the project, the team is now committed to deliver support to CEP applications (including temporal reasoning and stream management), support to higher levels of ruleflow and process modeling, improving the whole BRMS and repository tools, plus we have Master Degree and PhD students working on Machine Learning, Decision Trees, Uncertainty Reasoning, Temporal Reasoning and Rulebase Static Analysis. As you can see there is a LOT on our plate, so if anyone wants to step up and help us accelerate the development of pluggable fact types, it is more than welcome. ANYONE up to the task? :) Regarding your question about the removal of shadow fact, yes, it is scheduled for the next major release in a few months time. Although, it is import to remember that you can already disable shadow facts, just by following some requirements. I will write another e-mail about that. []s Edson 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > > > I think that you *should* treat facts that implement java.util.Map > differently from other facts. > Ignore their concrete class and don't worry about applying your > shadowing algorithm. > Then, treat them as if they were beans with getXYZ() methods for each > key "XYZ" they contain. > > Your reply indicates that you haven't even considered this design. I > wonder why not? (It seemed so natural to me that I assumed it's what > Drools *must* do. Especially considering the fact that Drools's chosen > scripting language, MVEL, supports accesses to maps using a special, > javascript-like syntax that allows you to verify that accesses are > side-effect free.) > > > On Wed, Feb 20, 2008 at 4:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > > A blog I wrote a long time ago about dynamically generated beans: > > > > http://blog.athico.com/2006/12/dynamically-generated-class-beans-as.html > > > > > I'm aware that I can generate beans - dynamically or statically, but > that is exactly the hassle I had hoped to avoid. (And, quite frankly, > it's not something I should have to go through when using a framework > such as Drools.) > > Will the issue disappear in a future, shadowless version of your > engine? To what degree will this version depend on facts being > conforming Java beans? > > > - Godmar > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > -- Edson Tirelli JBoss Drools Core Development Office: +55 11 3529-6000 Mobile: +55 11 9287-5646 JBoss, a division of Red Hat @ www.jboss.com _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

I think that you *should* treat facts that implement java.util.Map differently from other facts. Ignore their concrete class and don't worry about applying your shadowing algorithm. Then, treat them as if they were beans with getXYZ() methods for each key "XYZ" they contain. Your reply indicates that you haven't even considered this design. I wonder why not? (It seemed so natural to me that I assumed it's what Drools *must* do. Especially considering the fact that Drools's chosen scripting language, MVEL, supports accesses to maps using a special, javascript-like syntax that allows you to verify that accesses are side-effect free.)

On Wed, Feb 20, 2008 at 4:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > A blog I wrote a long time ago about dynamically generated beans: > > http://blog.athico.com/2006/12/dynamically-generated-class-beans-as.html > > I'm aware that I can generate beans - dynamically or statically, but that is exactly the hassle I had hoped to avoid. (And, quite frankly, it's not something I should have to go through when using a framework such as Drools.) Will the issue disappear in a future, shadowless version of your engine? To what degree will this version depend on facts being conforming Java beans? - Godmar _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Personally I don't like the idea of a Map as a model as it has no type, so straight away you lose object level descrimination. Also a Map is not declarative in defining what it is you are reasoning over. It's a hack to get over the limitations of the current environment. In the engine we have something called FactTemplats, which we do not currently document - it's a hidden feature. These work much less Jess Deftemplates, and where infact made so that we could support a Jess/clips. The implementation is basically an array and uses name tokenising to get access. i.e. you write person.name == "Godmar" and we rewrite it as person.setField( 0, "Godmar" ). Although we haven't yet got the rewritting part done so currently you have to manually do the above, or make it lookup the position each time with person.setField( "name", "Godmar" ). These FactTemplates can be reasoned over in the LHS just like pojos. However I'm not currently happy with the solution and thinking instead of going down the route of runtime bean generation. This would allow you to define models at runtime, without caring about the underlying implementation, and still give us pojos to work with and also provides more performance. further to this we really want to do our model implementation with ontology support. So that users can supply static and dynamic constraints to the properties they define on a class. It's currently touch and go if either of these will make it into 5.0, I'm really hoping that we can do the later solution, but we currently have many more priorities :( Mark Godmar Back wrote: > On Feb 20, 2008 12:30 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > >> Godmar, >> >> Shadow Facts are a necessary evil in current version. Basically what they >> do is keep the working memory consistent in face of attribute changes on the >> facts, that may happen both internally and externally to the working memory. >> Our implementation to shadow facts is a lazy proxy that caches the >> values until a safe point to synchronize the actual attribute value with the >> one seen by the engine. >> >> So, the question is: given an object: >> >> (Map) fact >> >> How can we create an identical copy of it (shadow), if there is no >> "clone" operation? >> > > Can you explain why you require the use of "clone()"? > > Cloning a map is otherwise easy - it's also referred to as a "shallow > copy" -- HashMap's HashMap(Map) constructor will do it. > http://java.sun.com/j2se/1.4.2/docs/api/java/util/HashMap.html#HashMap(ja... > I know you know that, so explain what necessitates the use of clone(). > > >> More than that, the shadow must be a subclass of it. >> > > java.util.HashMap is a subclass of Map. > > >> Most Collection and Map implementations have a single parameter >> constructor that allows us to do: >> >> >> proxy = (ShadowProxy) this.shadowClass.getConstructor( new Class[]{cls} >> ).newInstance( new Object[]{fact} ); >> >> But the SingletonMap you were using does not accept that constructor. >> So, one way is to explicit check if the fact is a SingletonMap and handle it >> accordingly, but that is a specific class hack... is there any general >> solution we can use? >> >> > > Forget about the SingletonMap. That was just one of the many things I > tried and failed. > > Fundamentally, I would like Drools to process facts that were obtained > from real-world sources, and these facts have properties I do not know > in advance. Therefore, I cannot use beans (or using beans would be > highly inconvenient since it will require changes to Java code > whenever I'm referring to a new property, something I'd rather avoid.) > > - Godmar > > >> []s >> Edson >> >> 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: >> >>> I don't really understand what you mean by "shadow". What is the >>> purpose of such shadowing. Mark's email implies that it has to do with >>> concurrency protection; it's not clear what that means. >>> >>> In my view, whatever purpose you pursue with "shadowing", it does not >>> justify treating beans and maps differently. >>> >>> Your example of class Person shows that. If a person has two >>> attributes, name and age, then this is equivalent to a map with two >>> keys 'name' and 'age'. >>> >>> Here's the mapping: >>> >>> p.getName() corresponds to m["name"] >>> p.getAge() corresponds to m["age"] >>> >>> and setName/setAge accordingly. >>> >>> Mathematically, a bean is an associative array with a fixed set of >>> keys (called "properties") that map to values. For all practical >>> purposes, that is the same as a map. There's no reason to treat them >>> differently. Wherever you'd do "getXXX()" with a bean you'd do >>> .get("XXX") with a map. >>> >>> - Godmar >>> >>> On Feb 20, 2008 11:25 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: >>> >>>> Ok, let me show one example. Imagine the class Person, with 2 >>>> >> attributes >> >>>> (name and age) and the corresponding getter/setters. >>>> What are the data for that fact that must be shadowed? easy answer: >>>> >> just >> >>>> shadow all getXXX() methods (getName and getAge). >>>> >>>> Now, take a Map. What is the data that must be shadowed? >>>> >>>> So, we do our best to work with facts that don't follow the javabean >>>> spec, but collections and maps are a complicated beast. Again, if you >>>> >> have >> >>>> suggestions on how to improve the current support we provide for them, >>>> please share with us. >>>> >>>> []s >>>> Edson >>>> >>>> 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: >>>> >>>> >>>>> On Feb 20, 2008 9:23 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: >>>>> >>>>>> Godmar, >>>>>> >>>>>> Short answer: collection/maps objects are not javabeans. >>>>>> >>>>>> >>>>> Explain why this is a problem. >>>>> >>>>> What is it about JavaBeans that your algorithm relies upon? Is it the >>>>> fact that the set of properties remains fixed and can be determined at >>>>> (fact) insertion time via reflection? >>>>> >>>>> Otherwise, I do not see any conceptual difference between a map and a >>>>> >>>> bean. >>>> >>>>> If that is the difference, then please allow maps with an immutable >>>>> >> key >> >>>> set. >>>> >>>>> - Godmar >>>>> >>>>> >>>>> >>>>>> Long answer: collection/maps must be shadowed to ensure >>>>>> >> consistency >> >>>>>> during execution, but how can we shadow the data if it is not >>>>>> >> exposed in >> >>>> a >>>> >>>>>> default, spec manner, as in javabeans? The algorithm we have in >>>>>> >> place >> >>>> right >>>> >>>>>> now is bellow. As you can see, it is a weak algo, but was the best I >>>>>> >>>> could >>>> >>>>>> come up at that time. If you have any suggestions on how to improve >>>>>> >>>> that, I >>>> >>>>>> appreciate. >>>>>> >>>>>> public Object getShadow(final Object fact) throws >>>>>> >>>> RuntimeDroolsException >>>> >>>>>> { >>>>>> ShadowProxy proxy = null; >>>>>> if ( isShadowEnabled() ) { >>>>>> try { >>>>>> if ( Collection.class.isAssignableFrom( >>>>>> >> this.shadowClass >> >>>> ) >>>> >>>>>> || Map.class.isAssignableFrom( this.shadowClass ) ) { >>>>>> // if it is a collection, try to instantiate >>>>>> >> using >> >>>>>> constructor >>>>>> try { >>>>>> proxy = (ShadowProxy) >>>>>> this.shadowClass.getConstructor( new Class[]{cls} ).newInstance( new >>>>>> Object[]{fact} ); >>>>>> } catch ( Exception e ) { >>>>>> // not possible to instantiate using >>>>>> >> constructor >> >>>>>> } >>>>>> } >>>>>> if ( proxy == null ) { >>>>>> if ( this.instantiator == null ) { >>>>>> this.setInstantiator(); >>>>>> } >>>>>> proxy = (ShadowProxy) >>>>>> >>>> this.instantiator.newInstance(); >>>> >>>>>> } >>>>>> >>>>>> proxy.setShadowedObject( fact ); >>>>>> } catch ( final Exception e ) { >>>>>> System.out.println( "shadow: " +proxy.getClass() + >>>>>> >> ":" + >> >>>>>> fact.getClass() ); >>>>>> throw new RuntimeDroolsException( "Error creating >>>>>> >> shadow >> >>>>>> fact for object: " + fact, >>>>>> e ); >>>>>> } >>>>>> } >>>>>> return proxy; >>>>>> >>>>>> >>>>>> } >>>>>> >>>>>> []s >>>>>> Edson >>>>>> >>>>>> 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: >>>>>> >>>>>>> As a general comment, the examples for which I find Drools failing >>>>>>> >> are >> >>>>>>> not the actual examples for which my application is failing. It's >>>>>>> >> just >> >>>>>>> the smallest test case I was able to eliminate. >>>>>>> >>>>>>> I'm now a bit concerned about your comment that Maps and >>>>>>> >> Collections >> >>>>>>> aren't well-defined as Facts. I am planning to make extensive use >>>>>>> >> of >> >>>>>>> them (that's also why I'd prefer to use the MVEL dialect, because >>>>>>> >> in >> >>>>>>> Java I cannot do this without creating Bean wrappers.) >>>>>>> >>>>>>> Could you elaborate what makes the semantics not "well-defined". >>>>>>> >>>>>>> I'm specifically concerned with immutable maps (such as the one >>>>>>> >> that >> >>>>>>> would have been returned by Collections.singletonMap), and with >>>>>>> collections of maps (such as those obtained via a "from"..." >>>>>>> >> clause). >> >>>>>>> I need to insert immutable maps as facts; I understand that the >>>>>>> >> items >> >>>>>>> returned by "from" aren't inserted as facts. >>>>>>> >>>>>>> - Godmar >>>>>>> >>>>>>> On Feb 19, 2008 3:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: >>>>>>> >>>>>>>> Drools tries to create the ShadowProxy. The reason is that it >>>>>>>> >>>> does >>>> >>>>>> not >>>>>> >>>>>>>> know about the implementation... it just knows it is a Map and >>>>>>>> >> as >> >>>> so, it >>>> >>>>>>>> must be shadowed. Problem is that SingletonMap is either final >>>>>>>> >> or >> >>>> does >>>> >>>>>> not >>>>>> >>>>>>>> have a default constructor. >>>>>>>> My recommendation, besides opening a JIRA for this, is avoid >>>>>>>> >>>>>> inserting >>>>>> >>>>>>>> collections/maps directly as facts. The semantic for such facts >>>>>>>> >> is >> >>>> not >>>> >>>>>>>> clearly defined and it may cause undesired behavior. >>>>>>>> >>>>>>>> []s >>>>>>>> Edson >>>>>>>> >>>>>>>> 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: >>>>>>>> >>>>>>>>> Hi, >>>>>>>>> >>>>>>>>> usings Drools 4.0.4 and MVEL 1.4, this simple rule: >>>>>>>>> --- >>>>>>>>> package test; >>>>>>>>> >>>>>>>>> import java.util.Collections; >>>>>>>>> >>>>>>>>> dialect "mvel" >>>>>>>>> >>>>>>>>> rule "Rule #1" >>>>>>>>> when >>>>>>>>> then >>>>>>>>> insert(Collections.singletonMap("content", "hello")); >>>>>>>>> end >>>>>>>>> -- >>>>>>>>> >>>>>>>>> produces: >>>>>>>>> java.lang.IllegalAccessError: class >>>>>>>>> >>>>>>>>> >> org.drools.shadow.java.util.Collections$SingletonMapShadowProxy >> >>>> cannot >>>> >>>>>>>>> access its superclass java.util.Collections$SingletonMap >>>>>>>>> at java.lang.ClassLoader.defineClass1(Native Method) >>>>>>>>> at >>>>>>>>> >> java.lang.ClassLoader.defineClass(ClassLoader.java:620) >> >>>>>>>>> at >>>>>>>>> >> org.drools.rule.MapBackedClassLoader.fastFindClass(MapBackedClassLoader.java:60) >> >>>>>>>>> at >>>>>>>>> >> org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java:79) >> >>>>>>>>> at >>>>>>>>> >> java.lang.ClassLoader.loadClass(ClassLoader.java:251) >> >>>>>>>>> at >>>>>>>>> >> org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java:547) >> >>>>>>>>> at >>>>>>>>> >> org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData(Rete.java:494) >> >>>>>>>>> at >>>>>>>>> >>>>>>>> org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) >>>>>>>> >>>>>>>>> at org.drools.reteoo.Rete.assertObject(Rete.java:152) >>>>>>>>> at >>>>>>>>> >>>> org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java:192) >>>> >>>>>>>>> at >>>>>>>>> >> org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java:71) >> >>>>>>>>> at >>>>>>>>> >> org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:909) >> >>>>>>>>> at >>>>>>>>> >> org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:881) >> >>>>>>>>> at >>>>>>>>> >> org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:67) >> >>>>>>>>> at >>>>>>>>> >> org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:61) >> >>>>>>>>> It's not clear to me why Drools creates Proxies for such >>>>>>>>> >> classes >> >>>> as >>>> >>>>>>>>> java.util.Collections, or does MVEL do it? >>>>>>>>> >>>>>>>>> - Godmar >>>>>>>>> _______________________________________________ >>>>>>>>> rules-users mailing list >>>>>>>>> rules-users(a)lists.jboss.org >>>>>>>>> https://lists.jboss.org/mailman/listinfo/rules-users >>>>>>>>> >>>>>>>>> >>>>>>>> -- >>>>>>>> Edson Tirelli >>>>>>>> JBoss Drools Core Development >>>>>>>> Office: +55 11 3529-6000 >>>>>>>> Mobile: +55 11 9287-5646 >>>>>>>> JBoss, a division of Red Hat @ www.jboss.com >>>>>>>> _______________________________________________ >>>>>>>> rules-users mailing list >>>>>>>> rules-users(a)lists.jboss.org >>>>>>>> https://lists.jboss.org/mailman/listinfo/rules-users >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> _______________________________________________ >>>>>>> rules-users mailing list >>>>>>> rules-users(a)lists.jboss.org >>>>>>> https://lists.jboss.org/mailman/listinfo/rules-users >>>>>>> >>>>>>> >>>>>> -- >>>>>> Edson Tirelli >>>>>> JBoss Drools Core Development >>>>>> Office: +55 11 3529-6000 >>>>>> Mobile: +55 11 9287-5646 >>>>>> JBoss, a division of Red Hat @ www.jboss.com >>>>>> _______________________________________________ >>>>>> rules-users mailing list >>>>>> rules-users(a)lists.jboss.org >>>>>> https://lists.jboss.org/mailman/listinfo/rules-users >>>>>> >>>>>> >>>>>> >>>>> _______________________________________________ >>>>> rules-users mailing list >>>>> rules-users(a)lists.jboss.org >>>>> https://lists.jboss.org/mailman/listinfo/rules-users >>>>> >>>>> >>>> >>>> -- >>>> Edson Tirelli >>>> JBoss Drools Core Development >>>> Office: +55 11 3529-6000 >>>> Mobile: +55 11 9287-5646 >>>> JBoss, a division of Red Hat @ www.jboss.com >>>> _______________________________________________ >>>> rules-users mailing list >>>> rules-users(a)lists.jboss.org >>>> https://lists.jboss.org/mailman/listinfo/rules-users >>>> >>>> >>>> >>> _______________________________________________ >>> rules-users mailing list >>> rules-users(a)lists.jboss.org >>> https://lists.jboss.org/mailman/listinfo/rules-users >>> >>> >> -- >> Edson Tirelli >> JBoss Drools Core Development >> Office: +55 11 3529-6000 >> Mobile: +55 11 9287-5646 >> JBoss, a division of Red Hat @ www.jboss.com >> _______________________________________________ >> rules-users mailing list >> rules-users(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/rules-users >> >> >> > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > ------------------------------------------------------------------------ _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Now that I read Ed's post, I realize I can simply turn shadowing off. I don't use facts whose attributes change (*). - Godmar (*) This probably shows my ignorance here - I freely admit the last time I used expert systems before the current exercise was during an AI class in college in the 90s - but why do systems such as Drools support modifiable facts in the first place? If facts become invalid, they should be retracted and new facts be provided,

but why support changing a fact literally "after the fact," if you pardon the pun?

Ed's example with person.likes changing from cheese to chocolate is particularly telling, because such as change would - in common parlance - be considered a change of opinion rather than a changing of facts.

On Wed, Feb 20, 2008 at 9:40 PM, Mark Proctor <mproctor(a)codehaus.org&gt; wrote: > If you want to use FactTemplates as a temporary, unsupported approach, feel > free. There are unit tests that show how this work. > > Btw when you use MVEL notation to access nested objects, like you do with a > Map, we just re-write it as an eval. This means you won't get any of the > indexing performance. You can still get indexing if you represent your facts > with FactTemplates. > > Mark > Mark Proctor wrote: > > Personally I don't like the idea of a Map as a model as it has no type, so > straight away you lose object level descrimination. Also a Map is not > declarative in defining what it is you are reasoning over. It's a hack to > get over the limitations of the current environment. > > In the engine we have something called FactTemplats, which we do not > currently document - it's a hidden feature. These work much less Jess > Deftemplates, and where infact made so that we could support a Jess/clips. > The implementation is basically an array and uses name tokenising to get > access. i.e. you write person.name == "Godmar" and we rewrite it as > person.setField( 0, "Godmar" ). Although we haven't yet got the rewritting > part done so currently you have to manually do the above, or make it lookup > the position each time with person.setField( "name", "Godmar" ). These > FactTemplates can be reasoned over in the LHS just like pojos. > > However I'm not currently happy with the solution and thinking instead of > going down the route of runtime bean generation. This would allow you to > define models at runtime, without caring about the underlying > implementation, and still give us pojos to work with and also provides more > performance. further to this we really want to do our model implementation > with ontology support. So that users can supply static and dynamic > constraints to the properties they define on a class. > > It's currently touch and go if either of these will make it into 5.0, I'm > really hoping that we can do the later solution, but we currently have many > more priorities :( > > > Mark > Godmar Back wrote: > On Feb 20, 2008 12:30 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Godmar, > > Shadow Facts are a necessary evil in current version. Basically what they > do is keep the working memory consistent in face of attribute changes on the > facts, that may happen both internally and externally to the working memory. > Our implementation to shadow facts is a lazy proxy that caches the > values until a safe point to synchronize the actual attribute value with the > one seen by the engine. > > So, the question is: given an object: > > (Map) fact > > How can we create an identical copy of it (shadow), if there is no > "clone" operation? > > Can you explain why you require the use of "clone()"? > > Cloning a map is otherwise easy - it's also referred to as a "shallow > copy" -- HashMap's HashMap(Map) constructor will do it. > http://java.sun.com/j2se/1.4.2/docs/api/java/util/HashMap.html#HashMap(ja... > I know you know that, so explain what necessitates the use of clone(). > > > > More than that, the shadow must be a subclass of it. > > java.util.HashMap is a subclass of Map. > > > > Most Collection and Map implementations have a single parameter > constructor that allows us to do: > > > proxy = (ShadowProxy) this.shadowClass.getConstructor( new Class[]{cls} > ).newInstance( new Object[]{fact} ); > > But the SingletonMap you were using does not accept that constructor. > So, one way is to explicit check if the fact is a SingletonMap and handle it > accordingly, but that is a specific class hack... is there any general > solution we can use? > > > Forget about the SingletonMap. That was just one of the many things I > tried and failed. > > Fundamentally, I would like Drools to process facts that were obtained > from real-world sources, and these facts have properties I do not know > in advance. Therefore, I cannot use beans (or using beans would be > highly inconvenient since it will require changes to Java code > whenever I'm referring to a new property, something I'd rather avoid.) > > - Godmar > > > > []s > Edson > > 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > > I don't really understand what you mean by "shadow". What is the > purpose of such shadowing. Mark's email implies that it has to do with > concurrency protection; it's not clear what that means. > > In my view, whatever purpose you pursue with "shadowing", it does not > justify treating beans and maps differently. > > Your example of class Person shows that. If a person has two > attributes, name and age, then this is equivalent to a map with two > keys 'name' and 'age'. > > Here's the mapping: > > p.getName() corresponds to m["name"] > p.getAge() corresponds to m["age"] > > and setName/setAge accordingly. > > Mathematically, a bean is an associative array with a fixed set of > keys (called "properties") that map to values. For all practical > purposes, that is the same as a map. There's no reason to treat them > differently. Wherever you'd do "getXXX()" with a bean you'd do > .get("XXX") with a map. > > - Godmar > > On Feb 20, 2008 11:25 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Ok, let me show one example. Imagine the class Person, with 2 > > attributes > > > > (name and age) and the corresponding getter/setters. > What are the data for that fact that must be shadowed? easy answer: > > just > > > > shadow all getXXX() methods (getName and getAge). > > Now, take a Map. What is the data that must be shadowed? > > So, we do our best to work with facts that don't follow the javabean > spec, but collections and maps are a complicated beast. Again, if you > > have > > > > suggestions on how to improve the current support we provide for them, > please share with us. > > []s > Edson > > 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > > > On Feb 20, 2008 9:23 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Godmar, > > Short answer: collection/maps objects are not javabeans. > > > Explain why this is a problem. > > What is it about JavaBeans that your algorithm relies upon? Is it the > fact that the set of properties remains fixed and can be determined at > (fact) insertion time via reflection? > > Otherwise, I do not see any conceptual difference between a map and a > > bean. > > > If that is the difference, then please allow maps with an immutable > > key > > > > set. > > > - Godmar > > > > > Long answer: collection/maps must be shadowed to ensure > > consistency > > > > > > during execution, but how can we shadow the data if it is not > > exposed in > > > > a > > > > default, spec manner, as in javabeans? The algorithm we have in > > place > > > > right > > > > now is bellow. As you can see, it is a weak algo, but was the best I > > could > > > > come up at that time. If you have any suggestions on how to improve > > that, I > > > > appreciate. > > public Object getShadow(final Object fact) throws > > RuntimeDroolsException > > > > { > ShadowProxy proxy = null; > if ( isShadowEnabled() ) { > try { > if ( Collection.class.isAssignableFrom( > > this.shadowClass > > > > ) > > > > || Map.class.isAssignableFrom( this.shadowClass ) ) { > // if it is a collection, try to instantiate > > using > > > > > > constructor > try { > proxy = (ShadowProxy) > this.shadowClass.getConstructor( new Class[]{cls} ).newInstance( new > Object[]{fact} ); > } catch ( Exception e ) { > // not possible to instantiate using > > constructor > > > > > > } > } > if ( proxy == null ) { > if ( this.instantiator == null ) { > this.setInstantiator(); > } > proxy = (ShadowProxy) > > this.instantiator.newInstance(); > > > > } > > proxy.setShadowedObject( fact ); > } catch ( final Exception e ) { > System.out.println( "shadow: " +proxy.getClass() + > > ":" + > > > > > > fact.getClass() ); > throw new RuntimeDroolsException( "Error creating > > shadow > > > > > > fact for object: " + fact, > e ); > } > } > return proxy; > > > } > > []s > Edson > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > As a general comment, the examples for which I find Drools failing > > are > > > > > > > not the actual examples for which my application is failing. It's > > just > > > > > > > the smallest test case I was able to eliminate. > > I'm now a bit concerned about your comment that Maps and > > Collections > > > > > > > aren't well-defined as Facts. I am planning to make extensive use > > of > > > > > > > them (that's also why I'd prefer to use the MVEL dialect, because > > in > > > > > > > Java I cannot do this without creating Bean wrappers.) > > Could you elaborate what makes the semantics not "well-defined". > > I'm specifically concerned with immutable maps (such as the one > > that > > > > > > > would have been returned by Collections.singletonMap), and with > collections of maps (such as those obtained via a "from"..." > > clause). > > > > > > > I need to insert immutable maps as facts; I understand that the > > items > > > > > > > returned by "from" aren't inserted as facts. > > - Godmar > > On Feb 19, 2008 3:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Drools tries to create the ShadowProxy. The reason is that it > > does > > > > not > > > > know about the implementation... it just knows it is a Map and > > as > > > > so, it > > > > > > must be shadowed. Problem is that SingletonMap is either final > > or > > > > does > > > > not > > > > have a default constructor. > My recommendation, besides opening a JIRA for this, is avoid > > inserting > > > > collections/maps directly as facts. The semantic for such facts > > is > > > > not > > > > > > clearly defined and it may cause undesired behavior. > > []s > Edson > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > Hi, > > usings Drools 4.0.4 and MVEL 1.4, this simple rule: > --- > package test; > > import java.util.Collections; > > dialect "mvel" > > rule "Rule #1" > when > then > insert(Collections.singletonMap("content", "hello")); > end > -- > > produces: > java.lang.IllegalAccessError: class > > > org.drools.shadow.java.util.Collections$SingletonMapShadowProxy > > > > cannot > > > > > > > access its superclass java.util.Collections$SingletonMap > at java.lang.ClassLoader.defineClass1(Native Method) > at > > java.lang.ClassLoader.defineClass(ClassLoader.java:620) > > > > > > > > > at > > > org.drools.rule.MapBackedClassLoader.fastFindClass(MapBackedClassLoader.java:60) > > > > > > > > > at > > > org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java:79) > > > > > > > > > at > > java.lang.ClassLoader.loadClass(ClassLoader.java:251) > > > > > > > > > at > > > org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java:547) > > > > > > > > > at > > > org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData(Rete.java:494) > > > > > > > > > at > > org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) > > > at org.drools.reteoo.Rete.assertObject(Rete.java:152) > at > > org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java:192) > > > > > > > at > > org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java:71) > > > > > > > > > at > > > org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:909) > > > > > > > > > at > > > org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:881) > > > > > > > > > at > > > org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:67) > > > > > > > > > at > > > org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:61) > > > > > > > > > It's not clear to me why Drools creates Proxies for such > > classes > > > > as > > > > > > > java.util.Collections, or does MVEL do it? > > - Godmar > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > ________________________________ > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

On Wed, Feb 20, 2008 at 9:40 PM, Mark Proctor <mproctor(a)codehaus.org&gt; wrote: > Btw when you use MVEL notation to access nested objects, like you do with a > Map, we just re-write it as an eval. This means you won't get any of the > indexing performance. You can still get indexing if you represent your facts > with FactTemplates. > > FWIW - I used MVEL notation to bypass the restriction that function calls aren't allowed on LHS, not for indexing reasons, that is, e["prop"] instead of e.get("prop") which to Drools looks like a function call with a potential side effects. I had hoped to avoid turning my maps into POJOs/beans; additionally, one of my goals is to keep my rule base easily readable (thus avoid the explicit use of eval()).

Regarding your other comment about supporting Maps are special fact types - you mention that you'd lose type information - that's not necessarily the case if you postulate that only maps whose key set and value types don't change past insertion. In that case, you can find the type via runtime reflection on the concrete objects being carried in the map.

Overall, though, I agree that it's probably only a stopgap measure and a better design is required. See also my other comment about what uses Drools should provide for IMO.

- Godmar > Mark > Mark Proctor wrote: > > Personally I don't like the idea of a Map as a model as it has no type, so > straight away you lose object level descrimination. Also a Map is not > declarative in defining what it is you are reasoning over. It's a hack to > get over the limitations of the current environment. > > In the engine we have something called FactTemplats, which we do not > currently document - it's a hidden feature. These work much less Jess > Deftemplates, and where infact made so that we could support a Jess/clips. > The implementation is basically an array and uses name tokenising to get > access. i.e. you write person.name == "Godmar" and we rewrite it as > person.setField( 0, "Godmar" ). Although we haven't yet got the rewritting > part done so currently you have to manually do the above, or make it lookup > the position each time with person.setField( "name", "Godmar" ). These > FactTemplates can be reasoned over in the LHS just like pojos. > > However I'm not currently happy with the solution and thinking instead of > going down the route of runtime bean generation. This would allow you to > define models at runtime, without caring about the underlying > implementation, and still give us pojos to work with and also provides more > performance. further to this we really want to do our model implementation > with ontology support. So that users can supply static and dynamic > constraints to the properties they define on a class. > > It's currently touch and go if either of these will make it into 5.0, I'm > really hoping that we can do the later solution, but we currently have many > more priorities :( > > > Mark > Godmar Back wrote: > On Feb 20, 2008 12:30 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Godmar, > > Shadow Facts are a necessary evil in current version. Basically what they > do is keep the working memory consistent in face of attribute changes on the > facts, that may happen both internally and externally to the working memory. > Our implementation to shadow facts is a lazy proxy that caches the > values until a safe point to synchronize the actual attribute value with the > one seen by the engine. > > So, the question is: given an object: > > (Map) fact > > How can we create an identical copy of it (shadow), if there is no > "clone" operation? > > Can you explain why you require the use of "clone()"? > > Cloning a map is otherwise easy - it's also referred to as a "shallow > copy" -- HashMap's HashMap(Map) constructor will do it. > http://java.sun.com/j2se/1.4.2/docs/api/java/util/HashMap.html#HashMap(ja... > I know you know that, so explain what necessitates the use of clone(). > > > > More than that, the shadow must be a subclass of it. > > java.util.HashMap is a subclass of Map. > > > > Most Collection and Map implementations have a single parameter > constructor that allows us to do: > > > proxy = (ShadowProxy) this.shadowClass.getConstructor( new Class[]{cls} > ).newInstance( new Object[]{fact} ); > > But the SingletonMap you were using does not accept that constructor. > So, one way is to explicit check if the fact is a SingletonMap and handle it > accordingly, but that is a specific class hack... is there any general > solution we can use? > > > Forget about the SingletonMap. That was just one of the many things I > tried and failed. > > Fundamentally, I would like Drools to process facts that were obtained > from real-world sources, and these facts have properties I do not know > in advance. Therefore, I cannot use beans (or using beans would be > highly inconvenient since it will require changes to Java code > whenever I'm referring to a new property, something I'd rather avoid.) > > - Godmar > > > > []s > Edson > > 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > > I don't really understand what you mean by "shadow". What is the > purpose of such shadowing. Mark's email implies that it has to do with > concurrency protection; it's not clear what that means. > > In my view, whatever purpose you pursue with "shadowing", it does not > justify treating beans and maps differently. > > Your example of class Person shows that. If a person has two > attributes, name and age, then this is equivalent to a map with two > keys 'name' and 'age'. > > Here's the mapping: > > p.getName() corresponds to m["name"] > p.getAge() corresponds to m["age"] > > and setName/setAge accordingly. > > Mathematically, a bean is an associative array with a fixed set of > keys (called "properties") that map to values. For all practical > purposes, that is the same as a map. There's no reason to treat them > differently. Wherever you'd do "getXXX()" with a bean you'd do > .get("XXX") with a map. > > - Godmar > > On Feb 20, 2008 11:25 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Ok, let me show one example. Imagine the class Person, with 2 > > attributes > > > > (name and age) and the corresponding getter/setters. > What are the data for that fact that must be shadowed? easy answer: > > just > > > > shadow all getXXX() methods (getName and getAge). > > Now, take a Map. What is the data that must be shadowed? > > So, we do our best to work with facts that don't follow the javabean > spec, but collections and maps are a complicated beast. Again, if you > > have > > > > suggestions on how to improve the current support we provide for them, > please share with us. > > []s > Edson > > 2008/2/20, Godmar Back <godmar(a)gmail.com&gt;: > > > > On Feb 20, 2008 9:23 AM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Godmar, > > Short answer: collection/maps objects are not javabeans. > > > Explain why this is a problem. > > What is it about JavaBeans that your algorithm relies upon? Is it the > fact that the set of properties remains fixed and can be determined at > (fact) insertion time via reflection? > > Otherwise, I do not see any conceptual difference between a map and a > > bean. > > > If that is the difference, then please allow maps with an immutable > > key > > > > set. > > > - Godmar > > > > > Long answer: collection/maps must be shadowed to ensure > > consistency > > > > > > during execution, but how can we shadow the data if it is not > > exposed in > > > > a > > > > default, spec manner, as in javabeans? The algorithm we have in > > place > > > > right > > > > now is bellow. As you can see, it is a weak algo, but was the best I > > could > > > > come up at that time. If you have any suggestions on how to improve > > that, I > > > > appreciate. > > public Object getShadow(final Object fact) throws > > RuntimeDroolsException > > > > { > ShadowProxy proxy = null; > if ( isShadowEnabled() ) { > try { > if ( Collection.class.isAssignableFrom( > > this.shadowClass > > > > ) > > > > || Map.class.isAssignableFrom( this.shadowClass ) ) { > // if it is a collection, try to instantiate > > using > > > > > > constructor > try { > proxy = (ShadowProxy) > this.shadowClass.getConstructor( new Class[]{cls} ).newInstance( new > Object[]{fact} ); > } catch ( Exception e ) { > // not possible to instantiate using > > constructor > > > > > > } > } > if ( proxy == null ) { > if ( this.instantiator == null ) { > this.setInstantiator(); > } > proxy = (ShadowProxy) > > this.instantiator.newInstance(); > > > > } > > proxy.setShadowedObject( fact ); > } catch ( final Exception e ) { > System.out.println( "shadow: " +proxy.getClass() + > > ":" + > > > > > > fact.getClass() ); > throw new RuntimeDroolsException( "Error creating > > shadow > > > > > > fact for object: " + fact, > e ); > } > } > return proxy; > > > } > > []s > Edson > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > As a general comment, the examples for which I find Drools failing > > are > > > > > > > not the actual examples for which my application is failing. It's > > just > > > > > > > the smallest test case I was able to eliminate. > > I'm now a bit concerned about your comment that Maps and > > Collections > > > > > > > aren't well-defined as Facts. I am planning to make extensive use > > of > > > > > > > them (that's also why I'd prefer to use the MVEL dialect, because > > in > > > > > > > Java I cannot do this without creating Bean wrappers.) > > Could you elaborate what makes the semantics not "well-defined". > > I'm specifically concerned with immutable maps (such as the one > > that > > > > > > > would have been returned by Collections.singletonMap), and with > collections of maps (such as those obtained via a "from"..." > > clause). > > > > > > > I need to insert immutable maps as facts; I understand that the > > items > > > > > > > returned by "from" aren't inserted as facts. > > - Godmar > > On Feb 19, 2008 3:11 PM, Edson Tirelli <tirelli(a)post.com&gt; wrote: > > > Drools tries to create the ShadowProxy. The reason is that it > > does > > > > not > > > > know about the implementation... it just knows it is a Map and > > as > > > > so, it > > > > > > must be shadowed. Problem is that SingletonMap is either final > > or > > > > does > > > > not > > > > have a default constructor. > My recommendation, besides opening a JIRA for this, is avoid > > inserting > > > > collections/maps directly as facts. The semantic for such facts > > is > > > > not > > > > > > clearly defined and it may cause undesired behavior. > > []s > Edson > > 2008/2/19, Godmar Back <godmar(a)gmail.com&gt;: > > > Hi, > > usings Drools 4.0.4 and MVEL 1.4, this simple rule: > --- > package test; > > import java.util.Collections; > > dialect "mvel" > > rule "Rule #1" > when > then > insert(Collections.singletonMap("content", "hello")); > end > -- > > produces: > java.lang.IllegalAccessError: class > > > org.drools.shadow.java.util.Collections$SingletonMapShadowProxy > > > > cannot > > > > > > > access its superclass java.util.Collections$SingletonMap > at java.lang.ClassLoader.defineClass1(Native Method) > at > > java.lang.ClassLoader.defineClass(ClassLoader.java:620) > > > > > > > > > at > > > org.drools.rule.MapBackedClassLoader.fastFindClass(MapBackedClassLoader.java:60) > > > > > > > > > at > > > org.drools.rule.MapBackedClassLoader.loadClass(MapBackedClassLoader.java:79) > > > > > > > > > at > > java.lang.ClassLoader.loadClass(ClassLoader.java:251) > > > > > > > > > at > > > org.drools.reteoo.Rete$ClassObjectTypeConf.loadOrGenerateProxy(Rete.java:547) > > > > > > > > > at > > > org.drools.reteoo.Rete$ClassObjectTypeConf.defineShadowProxyData(Rete.java:494) > > > > > > > > > at > > org.drools.reteoo.Rete$ClassObjectTypeConf.<init>(Rete.java:461) > > > at org.drools.reteoo.Rete.assertObject(Rete.java:152) > at > > org.drools.reteoo.ReteooRuleBase.assertObject(ReteooRuleBase.java:192) > > > > > > > at > > org.drools.reteoo.ReteooWorkingMemory.doInsert(ReteooWorkingMemory.java:71) > > > > > > > > > at > > > org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:909) > > > > > > > > > at > > > org.drools.common.AbstractWorkingMemory.insert(AbstractWorkingMemory.java:881) > > > > > > > > > at > > > org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:67) > > > > > > > > > at > > > org.drools.base.DefaultKnowledgeHelper.insert(DefaultKnowledgeHelper.java:61) > > > > > > > > > It's not clear to me why Drools creates Proxies for such > > classes > > > > as > > > > > > > java.util.Collections, or does MVEL do it? > > - Godmar > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > -- > Edson Tirelli > JBoss Drools Core Development > Office: +55 11 3529-6000 > Mobile: +55 11 9287-5646 > JBoss, a division of Red Hat @ www.jboss.com > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > ________________________________ > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users > > > _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

In Drools 3.0 we did not need to shadow as we hade a more complicated algorithm that maintained more references allowing for asymetrical behaviour - however I couldn't get the algorithm to perform with Jess like performance and it used a lot more memory. I dropped Drools 4.0 back to how Jess does it with shadow facts and symmetrical assert/retract. For Drools 5.0 I have re-written the algorithm and have managed to get asymettrical behaviour with actual performance increases, not losses. This is now in a branch and I hope to merge into trunk soon.