[rules-users] Unresolved Object Reference
Edson Tirelli
tirelli at post.com
Fri Jul 6 12:20:34 EDT 2007
Ronald,
Shadow facts only role is to ensure the integrity of the working memory.
Basically what they do is to only allow the engine to "see" the changes to
facts at safe "update" points.
Without a shadow, facts that have constrained attributes may present
undetermined behavior... sometimes rules that should fire will not fire or
vice-versa, usually there will be memory leaks, etc. So, to be brief, it
will not work, even if it may appear to be working.
Well, we try to make shadow facts as transparent as possible to the user
and to the engine, because shadow facts obviously will add overhead to the
runtime environment. If they are transparent, advanced user can "tune
performance" by disabling shadow facts for classes that don't need them.
To do that, we chose to go with a "proxy" strategy. So, we create a lazy
loaded cache proxy for each fact asserted into the working memory. To do
that, we need actual asserted fact classes to not be final, and also we are
required to override equals and hashCode methods.
In the future, we may try to pursue other strategies, but for now, our
only recommendation is to not assert facts of classes that are final or have
declared equals/hashcode methods as final.
[]s
Edson
2007/7/6, Ronald R. DiFrango <ron.difrango at gmail.com>:
>
> Edson,
>
> Oh, this could be bad news then. I do have cases where I have code like
> the following:
>
> detailLine: DetailLine(status != "Matched", ....other stuff....)
>
> then
>
> detailLine.setStatus ("Matched")
>
> What might the implications be then?
>
> Ron
>
> On 7/6/07, Edson Tirelli <tirelli at post.com > wrote:
> >
> >
> > Ronald,
> >
> > You can't change constrained values of facts that are not shadowed.
> > Exemple: if you have a rule that does:
> >
> > Person( name == "someValue" )
> >
> > If person is not shadowed, you can't change the name attribute while
> > the object is in the working memory or you may get unexpected behavior from
> > the engine. But for instance, if you don't constrain another attribute, like
> > "age", you can change it without any problems.
> >
> > []s
> > Edson
> >
> > 2007/7/5, Ronald R. DiFrango < ron.difrango at gmail.com>:
> > >
> > > Edson,
> > >
> > > That is indeed the case, XML Beans is generating final equals/hashCdoe
> > > methods. I have no idea why that is case. I may post this over on the XML
> > > Beans board.
> > >
> > > Is disabling the shadowing a problem from the rules execution point of
> > > view?
> > >
> > > Finally, I guess I need to do a source build, can I find the
> > > instructions somewhere on this?
> > >
> > > Ron
> > >
> > > On 7/5/07, Edson Tirelli < tirelli at post.com> wrote:
> > > >
> > > >
> > > > There was a problem with CGLIB proxies much similar to this... so
> > > > probably the same problem.
> > > >
> > > > What happens is that to shadow a fact, drools needs to extend the
> > > > actual class being asserted, and override the equals and hashcode methods.
> > > > So, the class can't be final, nor these methods can be final. Previously I
> > > > wasn't checking for final on these methods and that was the reason for the
> > > > error.
> > > >
> > > > I fixed that in trunk. It means, if XML Beans are generating
> > > > methods equals/hashCode as final, drools won't be able to shadow them (and
> > > > so will automatically disable shadowing for them), but it will not raise the
> > > > error anymore.
> > > >
> > > > Let me know if it is still a problem for you in trunk.
> > > >
> > > > []s
> > > > Edson
> > > >
> > > > 2007/7/5, Ronald R. DiFrango < ron.difrango at gmail.com >:
> > > > >
> > > > > All this sounds eerily similar to the problem I am reporting. In
> > > > > the 4.0 version I am getting the following exception on those
> > > > > inner classes:
> > > > >
> > > > > Exception in thread "main" java.lang.VerifyError: class
> > > > > com.circuitcity.rtvcrms.impl.DetailLineDocumentImpl$DetailLineImplShadowProxyoverrides final method ã Ý*h '+ .û Ý*h '+
> > > > > at java.lang.ClassLoader.defineClass1(Native Method)
> > > > > at java.lang.ClassLoader.defineClass(ClassLoader.java:620)
> > > > > at java.lang.ClassLoader.defineClass (ClassLoader.java:465)
> > > > > at org.drools.rule.MapBackedClassLoader.fastFindClass(
> > > > > MapBackedClassLoader.java:40)
> > > > > at org.drools.rule.MapBackedClassLoader.loadClass(
> > > > > MapBackedClassLoader.java:59)
> > > > > at java.lang.ClassLoader.loadClass (ClassLoader.java:251)
> > > > > at org.drools.reteoo.Rete$ObjectTypeConf.<init>(Rete.java:352)
> > > > > at org.drools.reteoo.Rete.assertObject(Rete.java:152)
> > > > > at org.drools.reteoo.ReteooRuleBase.assertObject(
> > > > > ReteooRuleBase.java :190)
> > > > > at org.drools.reteoo.ReteooWorkingMemory.doInsert(
> > > > > ReteooWorkingMemory.java:70)
> > > > > at org.drools.common.AbstractWorkingMemory.insert(
> > > > > AbstractWorkingMemory.java:772)
> > > > > at org.drools.common.AbstractWorkingMemory.insert (
> > > > > AbstractWorkingMemory.java:584)
> > > > > at
> > > > > com.circuitcity.rtvcrms.rules.RtvDecisionEngine.processDetailLines
> > > > > (RtvDecisionEngine.java:95)
> > > > > at
> > > > > com.circuitcity.rtvcrms.rules.RtvDecisionEngine.executeRules(
> > > > > RtvDecisionEngine.java :64)
> > > > > at com.circuitcity.rtvcrms.test.MainBasedTester.main(
> > > > > MainBasedTester.java:34)
> > > > >
> > > > > The catch is that XML Beans objects are auto generated based upon
> > > > > and XSD and we have limited control over what gets generated.
> > > > >
> > > > > On 7/5/07, Narendra Valada < narendra.valada at gmail.com> wrote:
> > > > > >
> > > > > > Thank you for your quick response.
> > > > > >
> > > > > > I cannot make the Outer class static and public since it is not
> > > > > > an inner class.
> > > > > >
> > > > > > In my example, I have two inner classes : Inner and InnerInner.
> > > > > >
> > > > > > Class Inner is enclosed by Outer and class InnerInner is
> > > > > > enclosed by Inner. I get the "Unable to resolve object type" error only on
> > > > > > the "InnerInner" class and not on the "Inner" class.
> > > > > >
> > > > > > Both the inner classes are defined as non-static.
> > > > > >
> > > > > > Thanks,
> > > > > >
> > > > > > Narendra
> > > > > >
> > > > > >
> > > > > >
> > > > > > On 7/5/07, Mark Proctor <mproctor at codehaus.org > wrote:
> > > > > > >
> > > > > > > you need to make your Outer class static and public. Look at
> > > > > > > the sample generated with a new drools project, that uses a static public
> > > > > > > class.
> > > > > > >
> > > > > > > Mark
> > > > > > > Narendra Valada wrote:
> > > > > > >
> > > > > > > * **Hello,*
> > > > > > > * *
> > > > > > > *I am getting an "Unable to resolve object type" error on
> > > > > > > certain XML Bean generated inner classes. I get this error only on inner
> > > > > > > classes that have been defined within other inner classes. Please see
> > > > > > > example below. *
> > > > > > > * *
> > > > > > > *Is this a known limitation of JBoss Rules (I am using JBOSS
> > > > > > > Rules 3.0.6) or the JANINO compiler?*
> > > > > > > * *
> > > > > > > *Thanks.*
> > > > > > > *
> > > > > > >
> > > > > > > package
> > > > > > > *
> > > > > > > *com.sample;
> > > > > > >
> > > > > > > public
> > > > > > > class Outer{ public class Inner { public boolean inner
> > > > > > > =false; public class InnerInner {
> > > > > > >
> > > > > > > }
> > > > > > >
> > > > > > > }
> > > > > > >
> > > > > > > }
> > > > > > > **package*
> > > > > > > com.sample *
> > > > > > >
> > > > > > > import
> > > > > > > *com.sample.Outer1.Inner.InnerInner; *
> > > > > > >
> > > > > > > rule
> > > > > > > *"Hello World 1" *when*
> > > > > > >
> > > > > > > InnerInner()
> > > > > > > *then*
> > > > > > >
> > > > > > > System.out.println(
> > > > > > > "");* *
> > > > > > >
> > > > > > > *end*
> > > > > > >
> > > > > > > ------------------------------
> > > > > > >
> > > > > > > _______________________________________________
> > > > > > > rules-users mailing list
> > > > > > > rules-users at lists.jboss.org
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > https://lists.jboss.org/mailman/listinfo/rules-users
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > _______________________________________________
> > > > > > > rules-users mailing list
> > > > > > > rules-users at lists.jboss.org
> > > > > > > https://lists.jboss.org/mailman/listinfo/rules-users
> > > > > > >
> > > > > > >
> > > > > >
> > > > > > _______________________________________________
> > > > > > rules-users mailing list
> > > > > > rules-users at lists.jboss.org
> > > > > > https://lists.jboss.org/mailman/listinfo/rules-users
> > > > > >
> > > > > >
> > > > >
> > > > > _______________________________________________
> > > > > rules-users mailing list
> > > > > rules-users at lists.jboss.org
> > > > > https://lists.jboss.org/mailman/listinfo/rules-users
> > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > > Edson Tirelli
> > > > Software Engineer - JBoss Rules Core Developer
> > > > 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 at lists.jboss.org
> > > > https://lists.jboss.org/mailman/listinfo/rules-users
> > > >
> > > >
> > >
> > > _______________________________________________
> > > rules-users mailing list
> > > rules-users at lists.jboss.org
> > > https://lists.jboss.org/mailman/listinfo/rules-users
> > >
> > >
> >
> >
> > --
> > Edson Tirelli
> > Software Engineer - JBoss Rules Core Developer
> > Office: +55 11 3529-6000
> > Mobile: +55 11 9287-5646
> > JBoss, a division of Red Hat @ www.jboss.com
> >
>
>
--
Edson Tirelli
Software Engineer - JBoss Rules Core Developer
Office: +55 11 3529-6000
Mobile: +55 11 9287-5646
JBoss, a division of Red Hat @ www.jboss.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/rules-users/attachments/20070706/ef4869f3/attachment.html
More information about the rules-users
mailing list