[rules-dev] Perhaps an undocumented feature of Backward Chaining?

[Mark Proctor]

On 25/06/2011 08:17, Wolfgang Laun wrote:
> Added  BackwardChainingTest.testInsertionOrderTwo  which will *fail*, 
> so don't be shocked if Hudson turns purple.
Thanks,

But please remember to rebase, or you screw up linear history in Hudson:

In short before pushing always do the following.

git checkout -b b1 #creates a branch to work in
... do work here
git stash
git pull --rebase origin master
git stash pop
git commit -a -m "JBRULES-X blah blah blah"
git checkout master
git merge b1
git push

That will ensure that your commits are applied linearly, which reduces 
the chances of you accidentally removing other people's work.

Mark
>
> -W
>
>
> On 24 June 2011 22:35, Mark Proctor <mproctor at codehaus.org 
> <mailto:mproctor at codehaus.org>> wrote:
> > On 24/06/2011 21:22, Michael Anstis wrote:
> >
> > Your test, Mark, is not the same example to which Wolfgang refers.
> >
> > You'll need to re-test with the DRL at the end of Wolfgang's email 
> (which is
> > what we've included in N&N for 5.2).
> >
> > It's quite possible one of the ?'s is simply wrong... but I don't 
> understand
> > the subtleties of BC syntax yet.
> >
> > If the query has "?" in it, then it's pull only. In which case the 
> insertion
> > order matters, as the query is triggered when Here or in below Person is
> > triggered. Inserting the Location objects after the query is 
> triggered will
> > not return a result, unless you use leave off the "?" for a reactive 
> "open
> > query"
> >
> > Mark
> >
> > With kind regards,
> >
> > Mike
> >
> > On 24 June 2011 21:14, Mark Proctor <mproctor at codehaus.org 
> <mailto:mproctor at codehaus.org>> wrote:
> >>
> >> On 23/06/2011 13:20, Wolfgang Laun wrote:
> >>
> >> Still not resolved...
> >>
> >> I have added a test for all the possible insertion orders for that
> >> particular use case and was not able to reproduce a problem:
> >>
> >> 
> https://github.com/droolsjbpm/drools/commit/501f2196dcef754eed5a28b066b50665a5bc29a3
> >>
> >> MArk
> >>
> >> On 22 June 2011 21:40, Mark Proctor <mproctor at codehaus.org 
> <mailto:mproctor at codehaus.org>> wrote:
> >>>
> >>> heh, I made a mistake with the rule
> >>> ( Location(z, y;) and ?hasFood(x, z;) )
> >>> should be
> >>> ( Location(z, y;) and hasFood(x, z;) )
> >>>
> >>> Otherwise it's not reactive, and then ordering matters.
> >>>
> >>> With that change I can do the insertions in any order and all is fine.
> >>>
> >>
> >> This works for the example you gave. But the one taken from the
> >> "Introduction" still depends on the order the facts are inserted. 
> Below is
> >> the .drl, which does not need any pojos. Rule kickOff contains the 
> order
> >> that produces nothing to see and eat:
> >> You are in the table
> >>   You can see []
> >>   You can eat []
> >>
> >> Insert Here after the others, and it works.
> >>
> >> -W
> >>
> >>  import java.util.List
> >>
> >> declare Thing
> >>     thing : String @key
> >> end
> >>
> >> declare Edible extends Thing
> >> end
> >>
> >> declare Location extends Thing
> >>     location : String  @key
> >> end
> >>
> >> declare Here
> >>     place : String
> >> end
> >>
> >> rule kickOff
> >> when
> >> then
> >>     System.out.println( "insert Here" );
> >>     insert( new Here( "table" ) );
> >>     System.out.println( "insert Edible" );
> >>     insert( new Edible( "peach" ) );
> >>     System.out.println( "insert Location" );
> >>     insert( new Location( "peach", "table" ) );
> >> end
> >>
> >> # 2011-06-22
> >> query isContainedIn( String x, String y )
> >>     Location(x, y;)
> >>     or
> >>     ( Location(z, y;) and /*?*/isContainedIn(x, z;) )
> >> end
> >>
> >> query whereFood( String x, String y )
> >>     ( Location(x, y;) and
> >>     Edible(x;) )
> >>     or
> >>     ( Location(z, y;) and /*?*/whereFood(x, z;) )
> >> end
> >>
> >> query look(String place, List things, List food)
> >>     Here(place;)
> >>     things := List() from accumulate( Location(thing, place;),
> >>                                       collectList( thing ) )
> >>     food := List() from accumulate( /*?*/whereFood(thing, place;),
> >>                                     collectList( thing ) )
> >> end
> >>
> >> rule reactiveLook
> >> when
> >>     Here( $place : place)
> >>     /*?*/look($place, $things; $food := food)
> >> then
> >>     System.out.println( "You are in the " + $place);
> >>     System.out.println( "  You can see " + $things );
> >>     System.out.println( "  You can eat " + $food );
> >> end
> >>
> >>
> >> _______________________________________________
> >> rules-dev mailing list
> >> rules-dev at lists.jboss.org <mailto:rules-dev at lists.jboss.org>
> >> https://lists.jboss.org/mailman/listinfo/rules-dev
> >>
> >>
> >> _______________________________________________
> >> rules-dev mailing list
> >> rules-dev at lists.jboss.org <mailto:rules-dev at lists.jboss.org>
> >> https://lists.jboss.org/mailman/listinfo/rules-dev
> >>
> >
> >
> > _______________________________________________
> > rules-dev mailing list
> > rules-dev at lists.jboss.org <mailto:rules-dev at lists.jboss.org>
> > https://lists.jboss.org/mailman/listinfo/rules-dev
> >
> >
> > _______________________________________________
> > rules-dev mailing list
> > rules-dev at lists.jboss.org <mailto:rules-dev at lists.jboss.org>
> > https://lists.jboss.org/mailman/listinfo/rules-dev
> >
> >
>
>
> _______________________________________________
> rules-dev mailing list
> rules-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/rules-dev/attachments/20110625/d0a3bd22/attachment-0001.html