Well, I can't explain the exact problem based just on what we can see
here; there's probably a little more involved. Jess should produce
somewhat better error messages, as well: if you're printing out your
own exceptions, be sure to check getCause() for nested exception
objects.
But I think your guess is probably in the right ballpark. If a fact
already exists, then assert returns FALSE. Therefore the code you
have here which binds the result of calling assert to a variable,
then immediately calls methods on that variable is dangerous; you'd
need to check for FALSE, first; i.e.,
(if (bind ?x (assert (y))) then
(whatever))
On Dec 29, 2006, at 1:59 AM, Robert Kildare wrote:
Below are two rules used to record the occurrence of sequences 4
and 5 as
they appear in data sets: facts which are removed and added in
groups as
appropriate to needs. (The system is not yet working, but is
tantalizingly close - as is 2007.)
As it happens, tidy5 fires in the first data set and tidy4 fires
after it
in the second.
;;;;;;;;;;;;;;;;;;;;;;;;;
(defrule tidy4
(declare (salience -25))
(frame (frameNumber ?framenumber))
?done <- (finishPat (patternID 4)(patComplete yes))
?previous1 <- (patPos (parentPattern 4)(position 1)(memberID
?m1)(tool_classificationID ?t1))
?previous2 <- (patPos (parentPattern 4)(position 2)(memberID
?m2)(tool_classificationID ?t2))
=>
(bind ?pat (assert (patInstance (factID nil)(patternID 4))))
(bind ?fid (?pat getFactId))(modify ?pat (factID ?fid))
(retract ?previous1 ?previous2)
(retract ?done)
(printout t "tidy pattern 4 match " crlf))
(defrule tidy5
(declare (salience -25))
(frame (frameNumber ?framenumber))
?done <- (finishPat (patternID 5)(patComplete yes))
?previous1 <- (patPos (parentPattern 5)(position 1)(memberID
?m1)(tool_classificationID ?t1))
?previous2 <- (patPos (parentPattern 5)(position 2)(memberID
?m2)(tool_classificationID ?t2))
=>
(bind ?pat (assert (patInstance (factID nil)(patternID 5))))
(bind ?fid (?pat getFactId))
(modify ?pat (factID ?fid))
(retract ?previous1 ?previous2)
(retract ?done)
(printout t "tidy pattern 5 match " crlf))
FIRE 48 MAIN::tidy5 f-75, f-131, f-127, f-129
==> f-134 (MAIN::patInstance (factID nil) (patternID 5))
==> Activation: MAIN::pattern_11_1 : f-75, f-134, f-67
<== Activation: MAIN::pattern_11_1 : f-75, f-134, f-67
<=> f-134 (MAIN::patInstance (factID 134) (patternID 5))
==> Activation: MAIN::pattern_11_1 : f-75, f-134, f-67
<== f-127 (MAIN::patPos (tool_classificationID 556) (memberID 30)
(parentPattern 5) (position 1) (frame 1))
<== f-129 (MAIN::patPos (tool_classificationID 2162) (memberID 33)
(parentPattern 5) (position 2) (frame 1))
<== f-131 (MAIN::finishPat (patternID 5) (patComplete yes))
tidy pattern 5 match
;;----------------- [old data has been removed and new data added,
with
considerable activation and firing ]----------------------------
FIRE 49 MAIN::tidy4 f-137, f-198, f-132, f-196
==> f-200 (MAIN::patInstance (factID nil) (patternID 4))
An error occurred at line 1which looks like( run )
Message: '134' is an integer, not a Java object
java.lang.NullPointerException
;;;;;;;;;;;;;;;;;;;;
What appears to be happening is that the old (patInstance (factID
nil)(patternID 5)) which was asserted is still in existence.... as ?
pat
with a value of 134 and when I try to assert it a second time I am
asserting an int rather than the object representing the variable
(is it
a ValueVector??)
Is this so ? If not, what can you tell me ?
In either case, any suggested solutions?
Rob
Robert Kildare
PhD candidate
School of Computing
University of Tasmania
--------------------------------------------------------------------
To unsubscribe, send the words 'unsubscribe jess-users
[EMAIL PROTECTED]'
in the BODY of a message to [EMAIL PROTECTED], NOT to the list
(use your own address!) List problems? Notify owner-jess-
[EMAIL PROTECTED]
--------------------------------------------------------------------
---------------------------------------------------------
Ernest Friedman-Hill
Advanced Software Research Phone: (925) 294-2154
Sandia National Labs FAX: (925) 294-2234
PO Box 969, MS 9012 [EMAIL PROTECTED]
Livermore, CA 94550 http://www.jessrules.com
--------------------------------------------------------------------
To unsubscribe, send the words 'unsubscribe jess-users [EMAIL PROTECTED]'
in the BODY of a message to [EMAIL PROTECTED], NOT to the list
(use your own address!) List problems? Notify [EMAIL PROTECTED]
--------------------------------------------------------------------
