9:36 a.m.
Expert 5.2.Final
My data comes to me as all strings, but I sometimes need to convert some
fields into integers in order to test their value. I wanted to avoid
conversion errors, so I combined rules into an activation-group and used
salience to control the order. However, at runtime I am getting a number
format error in method underWritingClassAsInt. I don't understand why - can
you explain?
rule "XmLifeParticipant underWritingClass Validation"
activation-group "XmLifeParticipant-underWritingClass"
salience 100
when
$lp : XmLifeParticipant(fieldWasModified("underWritingClass")
&& underWritingClass not matches "^\\d$" )
then
$lp.addError(drools.getRule().getName());
end
rule "XmLifeParticipant underWritingClass Value Check"
activation-group "XmLifeParticipant-underWritingClass"
salience 50
when
$lp : XmLifeParticipant(fieldWasModified("underWritingClass")
&& underWritingClassAsInt not in (1, 2, 4, 6))
then
$lp.addError(drools.getRule().getName());
end
--
View this message in context:
http://drools.46999.n3.nabble.com/Salience-activation-group-tp3564649p356...
Sent from the Drools: User forum mailing list archive at Nabble.com.
9:57 a.m.
We had this earlier today ;-)
"Rules are not if statements - salience doesn't change this."
Left hand sides of both rules are fully evaluated for any
XmLifeParticipant. If that value isn't a number, you'll get an error
(unless you write that method very carefully).
-W
On 06/12/2011, ronalbury <ronalbury(a)gmail.com> wrote:
Expert 5.2.Final
My data comes to me as all strings, but I sometimes need to convert some
fields into integers in order to test their value. I wanted to avoid
conversion errors, so I combined rules into an activation-group and used
salience to control the order. However, at runtime I am getting a number
format error in method underWritingClassAsInt. I don't understand why - can
you explain?
rule "XmLifeParticipant underWritingClass Validation"
activation-group "XmLifeParticipant-underWritingClass"
salience 100
when
$lp : XmLifeParticipant(fieldWasModified("underWritingClass")
&& underWritingClass not matches "^\\d$" )
then
$lp.addError(drools.getRule().getName());
end
rule "XmLifeParticipant underWritingClass Value Check"
activation-group "XmLifeParticipant-underWritingClass"
salience 50
when
$lp : XmLifeParticipant(fieldWasModified("underWritingClass")
&& underWritingClassAsInt not in (1, 2, 4, 6))
then
$lp.addError(drools.getRule().getName());
end
--
View this message in context:
http://drools.46999.n3.nabble.com/Salience-activation-group-tp3564649p356...
Sent from the Drools: User forum mailing list archive at Nabble.com.
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users
10:07 a.m.
Thanks for your quick response, but I am still confused.
An activation-group assures that only one rule in the group will fire (an
XOR group)
Salience controls the order of evaluation
The regular expression test has the higher salience and will trigger if it
is a bad number ... so why is the other rule even being tested?
In the mean time I will re-write the underWritingClassAsInt method to deal
with a bad number ... but this seems wrong to me.
--
View this message in context:
http://drools.46999.n3.nabble.com/Salience-activation-group-tp3564649p356...
Sent from the Drools: User forum mailing list archive at Nabble.com.
10:15 a.m.
Thanks for your quick response, but I am still confused.
An activation-group assures that only one rule in the group will fire (an XOR
group)
Salience controls the order of evaluation
No salience controls the order of the
activations firing.
There is no ordering of evaluating rules because rules aren't evaluated.
Instead the rules clauses are transposed into a RETE graph which evaluates each clause as
the necessary data is available.
If two rules share a clause then they may share a node in the RETE graph.
The regular expression test has the higher salience and will trigger
if it is a bad
number ... so why is the other rule even being tested?
The RETE graph is evaluated
as deeply as possible in order to produce the current list of possible activations.
The activations are then sorted by salience and filtered by agenda-group. The top of the
list is taken and fired.
If any other pending activations have the same activation-group those pending activations
are then removed.
(It's a lot more complicated than that as you have things such as lock on active,
no-loop, etc but it hopefully
makes it make a little more sense.
In the mean time I will re-write the underWritingClassAsInt method to deal
with a bad number ... but this seems wrong to me.
Or do the check for both parts in
both rules , or use inheritance, but the easiest way would probably be to just have it
return null if the number isn't valid - at least if you are just doing in
comparisons.
Thomas
**************************************************************************************
This message is confidential and intended only for the addressee. If you have received
this message in error, please immediately notify the postmaster(a)nds.com and delete it from
your system as well as any copies. The content of e-mails as well as traffic data may be
monitored by NDS for employment and security purposes. To protect the environment please
do not print this e-mail unless necessary.
NDS Limited. Registered Office: One London Road, Staines, Middlesex, TW18 4EX, United
Kingdom. A company registered in England and Wales. Registered no. 3080780. VAT no. GB 603
8808 40-00
**************************************************************************************
10:51 a.m.
Thank you Thomas ... I understand what you are saying and appreciate the
explanation (hopefully it will reduce the number of future questions I
have). I guess I was looking for an 'evaluation-group' and confused it with
the 'activation-group'.
Please do not take offense at this - it is important that the development
team understand how Drools is being used, and how it is being received in
the marketplace (and I am not suggesting I could do any better). I was told
that Drools would suck-up my cpu, and I am now gaining some insight into
why.
For all of the theoretical underpinnings of Expert, I am ending up with lots
of rules with essentially the same condition (I am including not <condition>
in that lot). This leads to a less clarity and more difficult maintenance.
I have also read about and received suggestions that I include extra 'dummy'
(e.g. show-stopper) objects to get the behavior I'm seeking. My concern is
that this really screws up the clarity and that our analysts won't really
understand (nor should they be asked to).
Is there any advantage to my considering a rule-flow rather than drl rules?
I am still a newbie and haven't tried rule-flow, but I'm thinking a
rule-flow may improve the clarity. However, I'm also guessing that a rule
flow does the same sort of evaluations you just explained to me. If that is
so, would this approach make a bad cpu situation even worse?
--
View this message in context:
http://drools.46999.n3.nabble.com/Salience-activation-group-tp3564649p356...
Sent from the Drools: User forum mailing list archive at Nabble.com.
4717
days inactive
4717
days old
4 comments
3 participants
participants (3)
-
ronalbury -
Swindells, Thomas -
Wolfgang Laun