On Apr 7, 2008, at 4:11 PM, [EMAIL PROTECTED] wrote:
I have been unsuccessful in exactly replicating my issue. I do
believe
that I still got accumulate to behave unexpectedly, however. I
will note
that I tested it in the version my company is running (some old 7.0
version), the 7.0p2 version, and the new 7.1b3 version. This issue
occurs
in 7.0, 7.0p2, but appears to be corrected in 7.1b3.
This time, instead of matching facts that it shouldn't have, the
second
rule appeared to do the opposite this time and wasn't matching
facts that
it should have. If I turn the first rule off, the second will work
fine.
But, if I have both running, the behavior presents itself.
I can send a zip file of my eclipse project j-unit test that
highlights the
issue.
That sounds good. Send it directly to me, not to the list.
"Ernest
Friedman-Hill"
<[EMAIL PROTECTED] To
ov> "jess-users"
Sent by: <[email protected]>
owner-jess-
users@ cc
sandia.gov
Subject
Re: JESS: Odd accumulate
behavior
04/01/2008 08:42 with identical 'result'
variable
PM names
Please respond to
[EMAIL PROTECTED]
.gov
My simple attempts at duplicating this in Jess 7.0p2 or 7.1b2 failed.
Can you check what version you're using? If it's older, can you try
with a current build?
On Apr 1, 2008, at 6:15 PM, [EMAIL PROTECTED] wrote:
I noticed some odd behavior with the accumulate function. I've got
two
very similar rules
1) validates a string based on a fact and asserts an error fact if
the
string is incorrect
2) validates a string based on a fact and asserts a different fact
which is
used to correct the string later
I used the accumulate function to accomplish this, because one of any
possible values for this 'string' are considered to be acceptable.
Since
the two rules had minute differences, I copied the rule text,
changed the
name, and changed the minor details. I left the accumulate
variable the
same in both rules, and it was named ?result.
rule 1 accumulate:
?someList <- (accumulate (bind ?result (create
$)) ;;
initializer
(bind ?result (create$ ?result (...str from ?
fact...))) ;;
action
?
result ;;
result
(SOME-FACT-NAME ?fact
&:(...condition1...)
&:(...condition2...)
)
rule 2 accumulate:
?someList <- (accumulate (bind ?result (create
$)) ;;
initializer
(bind ?result (create$ ?result (...str from ?
fact...))) ;;
action
?
result ;;
result
(SOME-FACT-NAME ?fact
&:(...condition1...)
&:(...condition2...)
&:(...condition3...)
)
The first rule worked fine, and generated its error fact. However, I
noticed that the 2nd accumulate function seemed to be matching on
facts
that it should not have, and firing. The condition 1 and condition
2 were
filtering just fine, but the 3rd condition was not filtering out
what it
should have been. It was funny, because I even placed the same
condition 3
check on the RHS of the rule and they would print out FALSE,
meaning that
the accumulate should have not even matched the fact that made it
through.
After scratching my head for a while and trying a bunch of random
stuff, I
found that the problem resolved itself when I changed these
variable names
to be distinct. IE, instead of two nearly identical rules with
nearly
identical accumulate functions in which the result variables were
named the
same - ?result, I changed the names to the equivalent of ?result_a
and
?result_b and this problem went away. It's like the second
accumulate
function was already finding the result variable that was generated
by the
first accumulate, and automatically plopping the contents into the
second
?someList.
I'm thinking this is a Jess bug, but I am not positive about that
(I am
pretty new to Jess). I am using Jess 7.0 as well (not sure what
build), so
I suppose it also could have been corrected since then. Hopefully
this
example is enough to illustrate what's happening, as I have already
rambled
on for a while now. If more detail is required, I will have to
replicate
the problem differently since the actual code is my company's
property.
.
This message and any attachments contain information from
Union Pacific which may be confidential and/or privileged.
If you are not the intended recipient, be aware that any
disclosure, copying, distribution or use of the contents of this
message is strictly prohibited by law. If you receive this message
in error, please contact the sender immediately and delete the
message and any attachments.
--------------------------------------------------------------------
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
Informatics & Decision Sciences 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 owner-jess-
[EMAIL PROTECTED]
--------------------------------------------------------------------
.
This message and any attachments contain information from
Union Pacific which may be confidential and/or privileged.
If you are not the intended recipient, be aware that any
disclosure, copying, distribution or use of the contents of this
message is strictly prohibited by law. If you receive this message
in error, please contact the sender immediately and delete the
message and any attachments.
--------------------------------------------------------------------
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
Informatics & Decision Sciences 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]
--------------------------------------------------------------------
