Ingomar Otter wrote:
Folks, I have a buch of rather simple rules in a configuration use-case which look like rule "requires-rule" when o1: O(id=2) o2:O(id=4) (not (exists (O(id=7)) then //assert a conflict "O2 and O4 require O7"In subsequent rules I deal with the conflicts detected. The resolution of these holds the "juicy bits". As the above creates quite some text and does not allow me to reason on the rule itself, I was thinking whether I should introduce the "domain rule" (o2 & O4 require O7) as a fact and use 'meta rules' to do the processing. Have you seen similar scenarios? Or is this misuse? For the time being, I have modeled the requires-rule as something like class RequiresDomainRule { Collection<Items> andTermItems; //eg o2, o4 Item requiredItem; O7 } Then I assert o2,o4 and a RequiresRule( (o2,o4),o6) But I have a knot in my brain and can't get the LHS written down, which in prose would be: there is a "requires rule" whose andTermItems are all present in WM and whose requiredItem is not present in WM then... The "andTermItems are all present" got me. I can test for any but all? - I'm stuck, need help.
checkout the forall and not conditional elements
Cheers, Ingomar
_______________________________________________ rules-users mailing list [email protected] https://lists.jboss.org/mailman/listinfo/rules-users
