Its not entirely true that you can't assert non java beans like string, you
can, you just can't do column constraints on them (but you can do eval).
On 3/29/06, Mark Proctor <[EMAIL PROTECTED]> wrote:
>
> No. I would use XSLT to do the main converstion, that would be a simple
> task. The functions would be more difficult, you would need an ANTL
> parser, drools uses antlr 3.0. But it should be possible to take the
> function body, run it through antlr and get an AST that you can string
> template to turn in functions. I'd start getting standard drools, sans
> functions, to work first with XSLT. Then worry about functions after.
>
> Mark
>
> Ronald R. DiFrango wrote:
> > Mark,
> >
> > Just curious if one were to develop such a thing are there any
> guidelines on
> > the conversion process already documetned some where?
> >
> > Ron
> >
> > On 3/28/06, Mark Proctor <[EMAIL PROTECTED]> wrote:
> >
> >> Simply a lack of time. If no user contributes the conversion utility I
> >> will get round to it eventually.
> >>
> >> Mark
> >> Ronald R. DiFrango wrote:
> >>
> >>> Mark,
> >>>
> >>> Thanks for the update. It is a real bummer that there is not support
> >>>
> >> for
> >>
> >>> the 2.x form in 3.0. Any reason why no backwards compatitbility?
> >>>
> >>> BTW...that Drools NoPrize for a coversion utility sounds outstanding
> :-)
> >>>
> >>> Ron
> >>>
> >>>
> >>> On 3/28/06, Mark Proctor <[EMAIL PROTECTED]> wrote:
> >>>
> >>>
> >>>> Currently there is not. Apart from functions, it should be easy
> >>>>
> >> enough.
> >>
> >>>> Drools 2.x is a subset of drools 3.0. Each 2.x condition maps to a
> 3.0
> >>>> eval - hence why Drools 2.x is not as efficient as 3.0, when rules
> are
> >>>> written properly. Drools 3.0 will now only match againts JavaBeans -
> >>>> that means String, Integer etc cannot be asserted as facts; which we
> >>>> have always claimed is bad practice anyway.
> >>>>
> >>>> <rule name="Hello World">
> >>>> <parameter identifier="person">
> >>>> <class>Person</class>
> >>>> </parameter>
> >>>>
> >>>> <java:condition>person.getName().equals("Stilton")</java:condition>
> >>>>
> >>>> <java:consequence>
> >>>> System.out.println( "Hello Stilton, did you konw you are named
> >>>> after a smelly piece of cheese?" );
> >>>> </java:consequence>
> >>>> </rule>
> >>>>
> >>>> So we bind the parameter and put conditions into evals:
> >>>> rule "hello world"
> >>>> when
> >>>> person : Person // This line is equivalent to the parameter
> >>>> section
> >>>> eval( person.getName().equals("Stilton") )
> >>>> then
> >>>> System.out.println( "Hello Stilton, did you konw you are named
> >>>>
> >> after
> >>
> >>>> a smelly piece of cheese?" )
> >>>> end
> >>>>
> >>>> However I must stress that the above is a very bad way to write
> rules
> >>>>
> >> in
> >>
> >>>> 3.0. Instead whenver possible you should use Field Constraints:
> >>>> rule "hello world"
> >>>> when
> >>>> Person( name == "Stilton" )
> >>>> then
> >>>> System.out.println( "Hello Stilton, did you konw you are named
> >>>>
> >> after
> >>
> >>>> a smelly piece of cheese?" )
> >>>> end
> >>>>
> >>>>
> >>>> In Drools 3.0 it is possible to bind both Facts and the Fact's
> fields:
> >>>> person : Person( personName : name )
> >>>>
> >>>> Please look over the intergration tests to see how Drools 3.0 is used
> >>>>
> >>>>
> >>>>
> >>
> http://anonsvn.labs.jboss.com/trunk/labs/jbossrules/drools-compiler/src/test/java/org/drools/integrationtests/IntegrationCases.java
> >>
> >>>>
> >>
> http://anonsvn.labs.jboss.com/trunk/labs/jbossrules/drools-compiler/src/test/resources/org/drools/integrationtests/
> >>
> >>>> Functions will need some sort of parser/processor. In Drools 2.x it
> was
> >>>> one big block for all functions. In 3.0 we need to pull out each
> >>>> individual function, along with their parameters.
> >>>>
> >>>> There is a Drools NoPrize to the person how creates the first
> complete,
> >>>> including functions, translator for 2.x to 3.0 :)
> >>>>
> >>>> Mark
> >>>>
> >>>> Ronald R. DiFrango wrote:
> >>>>
> >>>>
> >>>>> Mark,
> >>>>>
> >>>>> I may have missed it in previous posts, but is there a translation
> >>>>>
> >>>>>
> >>>> mechanism
> >>>>
> >>>>
> >>>>> from 2.1/2.5 style rules to the new style?
> >>>>>
> >>>>> Ron
> >>>>>
> >>>>>
> >>>>> On 3/28/06, Michael Neale <[EMAIL PROTECTED]> wrote:
> >>>>>
> >>>>>
> >>>>>
> >>>>>> Info and downloads from here:
> >>>>>> http://wiki.jboss.org/wiki/Wiki.jsp?page=JBossRules
> >>>>>>
> >>>>>> Much improved over beta 1, thanks for all the folks testing it, and
> >>>>>>
> >> for
> >>
> >>>>>> the
> >>>>>> feedback.
> >>>>>>
> >>>>>> Michael
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>
> >>>
> >>
> >>
> >
> >
>
>
>
