rspec users - Nov 2007 - Trouble using should_receive

I have this code trying to ensure my reset method works.  I want to  
make sure all the participants have their destroy method called.

# in my spec for Room
     r = Room.new(:name => ''bob'')
     r.save

     p = Participant.new(:login => ''a'', :password =>  
''b'', :password_confirmation => ''b'')
     p.room = r
     p.save!
     p.should_receive(:destroy)

     r.reset

#in room.rb
   def reset
     participants.each { |p| p.destroy }
     save!
   end

Unfortunately it fails.  I print out "p" in both locations and while  
they have all the same data, they have different memory addresses.  I  
don''t know if this is the explanation for the issue but it''s
all I can
find so far.

How do I fix this?  Thanks!

Glenn
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://rubyforge.org/pipermail/rspec-users/attachments/20071115/e5bee540/attachment.html

On Nov 15, 2007 10:50 AM, Glenn Ford <glenn at aldenta.com>
wrote:
>
> I have this code trying to ensure my reset method works. I want to make
sure
> all the participants have their destroy method called.
>
> # in my spec for Room r = Room.new(:name => ''bob'')
r.save p > Participant.new(:login => ''a'', :password
=> ''b'', :password_confirmation =>
> ''b'') p.room = r p.save! p.should_receive(:destroy)
r.reset #in room.rb def
> reset participants.each { |p| p.destroy } save! end
>
> Unfortunately it fails. I print out "p" in both locations and
while they
> have all the same data, they have different memory addresses. I
don''t know
> if this is the explanation for the issue but it''s all I can find
so far.
>
> How do I fix this? Thanks!
>
> Glenn
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-users
>
Hey Glenn,

The problem, as you mentioned, is that the objects loaded by
#participants don''t include the object you created.  Rails does a
"SELECT * FROM participates WHERE room_id=?" and instantiates new
objects.

The first option is to use participants#<< instead of setting the
room.  So instead of
p.room = r

you would have
r.partipants << p

And then it ought to work.

Alternatively, since you''re actually hitting the database, you can just
do

lambda { r.reset }.should change(Participant, :count).by(-1)

And as a side note, AR gives you #destroy_all, so instead of iterating
through the participants you can just do
participants.destroy_all

hth

Pat

On Nov 15, 2007, at 2:03 PM, Pat Maddox wrote:
> On Nov 15, 2007 10:50 AM, Glenn Ford <glenn at aldenta.com> wrote:
>>
>> I have this code trying to ensure my reset method works. I want to  
>> make sure
>> all the participants have their destroy method called.
>>
>> # in my spec for Room r = Room.new(:name => ''bob'')
r.save p >> Participant.new(:login => ''a'', :password
=>
>> ''b'', :password_confirmation =>
>> ''b'') p.room = r p.save! p.should_receive(:destroy)
r.reset #in
>> room.rb def
>> reset participants.each { |p| p.destroy } save! end
>>
>> Unfortunately it fails. I print out "p" in both locations and
while
>> they
>> have all the same data, they have different memory addresses. I  
>> don''t know
>> if this is the explanation for the issue but it''s all I can
find so
>> far.
>>
>> How do I fix this? Thanks!
>>
>> Glenn
>> _______________________________________________
>> rspec-users mailing list
>> rspec-users at rubyforge.org
>> http://rubyforge.org/mailman/listinfo/rspec-users
>>
>
> Hey Glenn,
>
> The problem, as you mentioned, is that the objects loaded by
> #participants don''t include the object you created.  Rails does a
> "SELECT * FROM participates WHERE room_id=?" and instantiates new
> objects.
>
> The first option is to use participants#<< instead of setting the
> room.  So instead of
> p.room = r
>
> you would have
> r.partipants << p
>
> And then it ought to work.
>
> Alternatively, since you''re actually hitting the database, you can
> just do
>
> lambda { r.reset }.should change(Participant, :count).by(-1)
>
> And as a side note, AR gives you #destroy_all, so instead of iterating
> through the participants you can just do
> participants.destroy_all
>
> hth
>
> Pat
r.participants << p didn''t do it, but the "should
change" did and I
like that solution best.  Court3nay also suggested "r.stub! 
(:participants).and_return([p])" which works also.

Thanks for your help, and thanks for the extra destroy_all trick too!

Glenn
>
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-users