[RubyAMF] RubyAMF - developers wanted

[RubyAMF]

hey..
http://www.4thtemple.com/stuff/active_record_adapter.rb
this seems to fix the array wrapper problem.. but theres another bug .. lets say you do something like render :amf => addr.user and in as3, if you do re.return.name gets confused there.. anyway i'll work on that tomorrow 

On Thu, 4 Oct 2007 02:48:17 -0400
RubyAMF <rubyamf-discussion at rubyforge.org> wrote:

> On Wed, 3 Oct 2007 19:21:41 -0700
> RubyAMF <rubyamf-discussion at rubyforge.org> wrote:
> 
> yeah.. when stepping through run_multiple/single .. it's doing !ar.empty? .. but ar isn't an array .. so i fixed it so it checks both scenerios..  
> 
>         ok = false;
>           if (ar.instance_of? Array)
>                 if (!ar.empty? && !ar.nil?)
>                    ok=true;
>                 end
>           end
> 
>           if ((!ar.instance_of? Array && !ar.nil?) || ok)
> 
> 
> 
> > U probably are getting a fault. Trace out fe.fault.faultString
> > 
> > 
> > The box said requires Windows XP or greater. So I got a Mac.
> > 
> > On Oct 3, 2007, at 4:15 PM, RubyAMF <rubyamf-discussion at rubyforge.org>  
> > wrote:
> > 
> > > it's funny, ssr actually interprets this situation as a faultEvent.   
> > > Not sure why .. but I'll worry about that later.
> > >
> > > On Mon, 1 Oct 2007 20:49:02 -0700
> > > RubyAMF <rubyamf-discussion at rubyforge.org> wrote:
> > >
> > >> Yeah, so let's say you have these models defined:
> > >>
> > >> class User < ActiveRecord::Base
> > >> has_many :addresses
> > >> end
> > >>
> > >> class Address < ActiveRecord::Base
> > >> belongs_to :user
> > >> end
> > >>
> > >> Now in a controller you do this:
> > >> class MyController < ...
> > >>  def myaction
> > >>    myadd = Address.find(1, :include => :user)
> > >>    render :amf => myadd
> > >>  end
> > >> end
> > >>
> > >>
> > >> The result in RubyAMF will be a VO, and the "user" property will be  
> > >> an
> > >> array, and the 0 index will be the "user" object. This is incorrect  
> > >> as the
> > >> "user" property should be the user object as expected. this happens  
> > >> in the
> > >> active_record_adapter..
> > >>
> > >> -Aaron
> > >>
> > >>
> > >> On 10/1/07, RubyAMF <rubyamf-discussion at rubyforge.org> wrote:
> > >>>
> > >>> I apologize for not being familiar with much of RAMF, my project  
> > >>> hasn't
> > >>> progressed to backend developement yet....
> > >>>
> > >>> But to clarify what is happening, is if a call is made to a model  
> > >>> that
> > >>> uses foreign key constraints, the resulting object will be  
> > >>> returned to the
> > >>> swf as an array wrapped in another array, where the real model  
> > >>> array/hash is
> > >>> inside index 0?
> > >>>
> > >>> (sorry.. i havent actually gotten to the backend portion of my  
> > >>> project
> > >>> yet..)
> > >>>
> > >>>> -Update the active_record_adapter to support a model association of
> > >>>> :belongs_to. Currently if a model has a :belongs_to relationship,  
> > >>>> that
> > >>>> object get's returned wrapped in an array (
> > >>>> re.result.myModelVO.myBelongsToProperty [0]. Which is not  
> > >>>> correct. It
> > >>> should
> > >>>> be re.result.myModelVO.myBelognsToProperty. This happens in the
> > >>>> active_record_adapter#run_single method, because the payload  
> > >>>> object is
> > >>>> initially created as an array (initla_data = []). the object get's
> > >>> pushed
> > >>>> into the array. Some time needs to be put into this to correctly  
> > >>>> alter
> > >>> the
> > >>>> adapter so that it won't break situations where models have  
> > >>>> a :has_many
> > >>>> relationship as well.
> > >>> _______________________________________________
> > >>> Rubyamf-discussion mailing list
> > >>> Rubyamf-discussion at rubyforge.org
> > >>> http://rubyforge.org/mailman/listinfo/rubyamf-discussion
> > >>>
> > >>
> > >
> > >
> > > -- 
> > >
> > > _______________________________________________
> > > Rubyamf-discussion mailing list
> > > Rubyamf-discussion at rubyforge.org
> > > http://rubyforge.org/mailman/listinfo/rubyamf-discussion
> > _______________________________________________
> > Rubyamf-discussion mailing list
> > Rubyamf-discussion at rubyforge.org
> > http://rubyforge.org/mailman/listinfo/rubyamf-discussion
> 
> 
> -- 
> 
> _______________________________________________
> Rubyamf-discussion mailing list
> Rubyamf-discussion at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rubyamf-discussion


--