Dave Groat has pointed out that one of Jess's experimental features, queries, are very useful in doing the first part of HW9. See the manual for the details.

A query, defined with defquery, is like a rule without a RHS. When you call it, you specify values for any variables declared as *input* arguments and it finds all matches for the conditions in the query. Thus, it is run only when you call it via the function run-query or count-query-results. If the first case, it returns a data structure which you can access from Java to find the sets of facts which satisfy the query. In the second, it just returns the number of answers.

Consider the following facts and the query Q1, which takes a major and finds students with that major:

  (defquery Q1
     (declare (variables ?major))
     (student ?s ?ssn)
     (major ?ssn ?major))
  (deffacts query-test
     (student s1 123-44-5678)    (student s2 223-44-5678)
     (student s3 323-44-5678)    (student s4 423-44-5678)
     (student s5 523-44-5678)
     (major 123-44-5678 cmsc)    (major 223-44-5678 ifsm)
     (major 323-44-5678 cmpe)    (major 423-44-5678 cmsc)
     (major 523-44-5678 cmsc))

    
  Jess> (count-query-results Q1 cmsc)
  3
  Jess> (count-query-results Q1 ifsm)
  1
  Jess> (count-query-results Q1 history)
  0
  Jess>

  (defquery possible-adjacent-pits
    "Finds the caves adjacent to a given (X,Y) which might contain a pit"
    (declare (variable ?x ?y))
    (adj ?x ?y ?x2 ?y2)
    ...)