[rules-users] Problem with salience
Wolfgang Laun
wolfgang.laun at gmail.com
Thu Jul 7 12:57:04 EDT 2011
OK, this is a little more to the point.
What you should review critically is the use of paywayManager (in "Pay
complete") and emTicket (in the other rules). Apparently, these objects are
not WM items - are they globals? Globals must not be used in conditions the
way they are apparently used here.
It's impossible to tell from these snippets why the last rule never fires -
there's that ElectromechanicalEvent...
-W
2011/7/7 Aitor González Serrano <jmolly at gmail.com>
> Ok.
> That's my code:
>
> /************************************** Payment Complete
>> **************************************/
>> rule "Pay complete"
>> salience 10
>> no-loop true
>> when
>> event : PaywayEvent(id == PaywayEventType.OPERATION_IN)
>> paymentData : PaymentData(credit >= saleImport)
>> transaction : CurrentTransaction (state ==
>> TransactionState.CHARGE_PRODUCT)
>> eval (!paywayManager.isStopped())
>> then
>> retract(event);
>>
>> updateDebt(languageManager, hmiManager,paymentData.getDebt());
>> log.info("Machine.drl","Charge complete. Stopping devices!. Debt is
>> "+paymentData.getDebt());
>> for(Enumeration<String> e = paywayManager.getPaywayNames();
>> e.hasMoreElements(); ){
>> String payway = e.nextElement();
>> if(paywayManager.getPayway(payway).isLoaded()
>> &&
>> !paywayManager.getPayway(payway).getPaywayApi().isStopped()) {
>> PaywayEvent paywayEvent = new
>> PaywayEvent(PaywayEventType.STOP_PAYMENT_DEVICES);
>> paywayEvent.setDevice(payway);
>> insert(paywayEvent);
>> }
>> }
>> end
>>
>>
>> This is the stop rule of the device
>>
>> /******************************************STOP PAYMENT DEVICES
>> ****************************/
>> rule "Stop EmTicket Device"
>> salience 10
>> no-loop true
>> when
>> event : PaywayEvent(id == PaywayEventType.STOP_PAYMENT_DEVICES &&
>> device == "discountticket")
>> transaction : CurrentTransaction (state ==
>> TransactionState.CHARGE_PRODUCT)
>> eval (!emTicket.isStopped())
>> then
>> retract(event);
>> if (emTicket.stop()){
>> electromechanical.disableTakeTicket();
>> electromechanical.blinkingLedOff();
>> emTicket.setState(EmTicketState.OUT_OF_SERVICE);
>> }
>> else{
>> log.info("EmTicket","EmTicket cannot be stoped!");
>> }
>>
>> insert (new PaywayEvent(PaywayEventType.OPERATION_IN));
>> end
>>
>
>
> So, there are bidirectional calls between STOP_PAYMENT_DEVICES and
> OPERATION_IN (because emTicket can't be stopped, so an event should be fired
> soon)
>
> While this 2 rules are firing the other, the hardware sends an
> DISCOUNT_TICKET_ACCEPTED, that updates the debt, and the state of EmTicket,
> so the in next STOP_PAYMENT_DEVICES emTicket will be stopped.
>
> rule "Discount Ticket discarded when accepted"
>> salience 1000
>> no-loop true
>> when
>> event : ElectromechanicalEvent (eventType ==
>> ElectromechanicalEventType.CONSUM_EVENT
>> && description == "articleDiscarded")
>> transaction : CurrentTransaction(state ==
>> TransactionState.CHARGE_PRODUCT)
>> eval (emTicket.getState() == EmTicketState.ARTICLE_TAKEN)
>> paymentData : PaymentData()
>> then
>> lot of code here..
>> end
>>
>
>
> The problem I'm facing is that rule "Discount Ticket discarded when
> accepted" is never fired (and has a lot more of salience than the other
> two)! The other two rules are in a constant bucle. The event of the
> discarded is sent via enginer.insertEvent() && fireAllRules. Maybe insert
> command in drl files has more priority than facts invoked from java classes?
> Thanks in advance!
>
>
>
>
>
>
>
>
>
>
>
> 2011/7/7 Wolfgang Laun <wolfgang.laun at gmail.com>
>
>> Sorry, but I become more confused with every re-read. I think the original
>> code is required.
>>
>> I think that this is nothing but a FSM. Generally, this sort of problem
>> should be doable without resorting to salience (except for one or two low
>> priority ones acting as the "safety net").
>>
>> -W
>>
>>
>>
>>
>> On 7 July 2011 16:35, jmolly <jmolly at gmail.com> wrote:
>>
>>> Hello.
>>> I'm facing some problems in drools with salience. (maybe I don't
>>> understand
>>> how it works)
>>>
>>> We have a rule that check if an import has been satisfied by the client.
>>> But
>>> we have a lot of pay methods so, when the import has been satisfied, we
>>> have
>>> to stop all the payment devices. If a payment device is in use, we can't
>>> stop it, so we send stop commands until device can stop.
>>> The problem is that if a device is making a pay process, and consumes an
>>> event, that event is never fired.
>>>
>>> Maybe I can explain better with code:
>>>
>>>
>>> rule "Pay device stop"
>>> Salience 10
>>> Try to stop the device and sends an OPERATION_IN event if can't stop
>>> (because there is a ticket inside)
>>>
>>> Machine rule
>>> rule "Operation in received witch devices not stopped"
>>> salience 10
>>> Sent stop events to all devices
>>>
>>> Machine rule
>>> rule "Operation in received with devices stopped"
>>> salience 10
>>> All ok. Finish operation
>>>
>>>
>>> rule "Ticket data read"
>>> salience 1000
>>> updates the debt for the user and sends an OPERATION_IN
>>>
>>> The pay hardware, when is reading a ticket, sent a TICKET_DATA_READ
>>> this rule has salience 100 and should be fired inmediatly but Drools is
>>> entering all time in a bucle from Operation in received and Pay stop
>>> device.
>>>
>>>
>>> Can someone give a clue about what are we doing wrong?
>>>
>>>
>>> Sorry for the english. I can post the original code if needed.
>>> Thanks in advance
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> --
>>> View this message in context:
>>> http://drools.46999.n3.nabble.com/Problem-with-salience-tp3148382p3148382.html
>>> Sent from the Drools: User forum mailing list archive at Nabble.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
>>
>>
>
> _______________________________________________
> rules-users mailing list
> rules-users at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/rules-users/attachments/20110707/2837e7dd/attachment.html
More information about the rules-users
mailing list