BusinessProcessScoped beans in conversation scope cause NPE

[naag]

Hi all,

I've got one more question regarding CDI behaviour in Activiti 5.10. We've got a @BusinessProcessScoped bean that we use in JSF forms. Especially this bean is used on the process start form before the process is started, so we want the bean to be associated to the conversation scope. So via

<f:event type="preRenderView" listener="#{startProcessBean.preStart()}" />
‍‍‍

we start a new conversation on the JSF form that uses our @BusinessProcessScoped bean and then we start the process via BusinessProcess.startProcessByKey(). This used to work in Activiti 5.9, but in 5.10, we get a NPE. This is the stacktrace:

Caused by: java.lang.NullPointerException
   at org.activiti.cdi.impl.context.DefaultContextAssociationManager.getExecutionId(DefaultContextAssociationManager.java:184) [activiti-cdi-5.10.jar:]
   at org.activiti.cdi.BusinessProcess.isAssociated(BusinessProcess.java:260) [activiti-cdi-5.10.jar:]
   at org.activiti.cdi.impl.context.BusinessProcessContext.get(BusinessProcessContext.java:99) [activiti-cdi-5.10.jar:]
   at org.jboss.weld.bean.proxy.ContextBeanInstance.getInstance(ContextBeanInstance.java:90) [weld-core-1.1.5.AS71.Final.jar:2012-02-10 15:31]
   at org.jboss.weld.bean.proxy.ProxyMethodHandler.invoke(ProxyMethodHandler.java:79) [weld-core-1.1.5.AS71.Final.jar:2012-02-10 15:31]
   at foo.bar.SomeBusinessBean$Proxy$_$$_WeldClientProxy.getCostCenter(SomeBusinessBean$Proxy$_$$_WeldClientProxy.java) [business-bean-0.0.2.jar:]
   at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) [rt.jar:1.6.0_31]
   at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) [rt.jar:1.6.0_31]
   at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) [rt.jar:1.6.0_31]
   at java.lang.reflect.Method.invoke(Method.java:597) [rt.jar:1.6.0_31]
   at javax.el.BeanELResolver.getValue(BeanELResolver.java:302) [jboss-el-api_2.2_spec-1.0.0.Final.jar:1.0.0.Final]
   at com.sun.faces.el.DemuxCompositeELResolver._getValue(DemuxCompositeELResolver.java:176) [jsf-impl-2.1.7-jbossorg-2.jar:]
   at com.sun.faces.el.DemuxCompositeELResolver.getValue(DemuxCompositeELResolver.java:203) [jsf-impl-2.1.7-jbossorg-2.jar:]
   at org.apache.el.parser.AstValue.getValue(AstValue.java:169) [jbossweb-7.0.13.Final.jar:]
   at org.apache.el.ValueExpressionImpl.getValue(ValueExpressionImpl.java:189) [jbossweb-7.0.13.Final.jar:]
   at org.jboss.weld.el.WeldValueExpression.getValue(WeldValueExpression.java:50) [weld-core-1.1.5.AS71.Final.jar:2012-02-10 15:31]
   at com.sun.faces.facelets.el.TagValueExpression.getValue(TagValueExpression.java:109) [jsf-impl-2.1.7-jbossorg-2.jar:]
   … 83 more
‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

The failing code is this:

public class DefaultContextAssociationManager implements ContextAssociationManager, Serializable {
  // …
  @Override
  public String getExecutionId() {
    return getExecution().getId();
  }
  // …
}
‍‍‍‍‍‍‍‍‍‍

And this change fixed it for me, but I'm not 100% sure if it's the right solution to this issue.

public class DefaultContextAssociationManager implements ContextAssociationManager, Serializable {
  // …
  @Override
  public String getExecutionId() {
    Execution execution = getExecution();
   if (execution != null) {
     return execution.getId();
   } else {
     return null;
   }
  }
  // …
}
‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

What do you think?

Thanks a lot!
Peter

[bernd_ruecker]

Hi Peter.

Same here, I will have a look. Strange enough, since we use that in camunda fox and I don't see this kind of problems in our QA, but it might be, that we haven't merged all changes into the EE branch yet. I keep you posted.

Cheers
Bernd

[bernd_ruecker]

Well spotted, I created http://jira.codehaus.org/browse/ACT-1340 and fixed it. It will be available in Activiti 5.11 or in the fox-engine 1.28 released end of the week (we release community versions every sprint end / two weeks): https://app.camunda.com/confluence/display/foxUserGuide/fox+engine.

[naag]

OK in the future I will create JIRA's myself, don't want to seem lazy here 😉 Thanks!