What Wolfgang was hinting at (assuming I understood correctly the thread of
the discussion) is the difference between a set of rules like this:
...
rule "n-th test"
when
$b : Bean( field == "x_n" )
then
$b.setSomething( "y_n" )
end
...
and rule-base like this:
declare Holder
selector : Object
target : Object
sal : int
end
rule "meta"
salience $sal
when
$b : Bean( $f : field )
Holder( selector == field, $tgt : target, $sal : sal )
then
$b.setSomething( $tgt );
end
rule "init"
when
then
// for j ...
insert( new Holder( x[j], y[j], j ) );
end
Of course, which approach is better would depend on the actual problem. If
you could share a few rules and highlight any patterns, that would be
helpful.
For example, consider that one of the advantages of RETE-based systems is
the possibility to share evaluations between rules.
So if your rules are all checking different fields for different values,
then maybe you should consider the latter approach.
Best
Davide
--
View this message in context:
http://drools.46999.n3.nabble.com/Large-number-of-rules-in-Drools-tp36303...
Sent from the Drools: User forum mailing list archive at
Nabble.com.