-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi there
I''ve run into a bit of a problem that I can''t seem to figure
out. I
have a model class User which includes our acts_as_role_context
library. This lib acts like most other acts_as libs, evaluating
class and instance methods at runtime. When I try to stub methods
that are defined in acts_as_role_context, I get strange behavior.
Perhaps it''s just me. I am a bit under the weather, and the Nyquil
could be clouding my thinking. :)
Here is my spec. It''s pretty simple, I just want to ensure that the
correct redirect happens when it should. For the record, the
controller works as expected outside of rspec.
- ---------
require File.dirname(__FILE__) + ''/../spec_helper''
describe MemberNetworkController do
before(:each) do
controller.should_receive(:beta_lockdown).and_return true
login
end
after(:each) do
logout
end
describe "on GET to add_as_friend" do
it "should redirect to /network/friends" do
logged_in_user.stub!(:add_user_to_role)
controller.stub!(:make_friend_activity)
get :add_as_friend, :key => logged_in_user.key
response.should redirect_to("/members/#{logged_in_user.id}/
network/friends")
end
end
end
- --------
The line in question is logged_in_user.stub!(:add_user_to_role).
User#add_user_to_role is defined in acts_as_role_context.rb. I''ll
not paste the entire file unless requested. It''s fairly large. the
instance method #add_user_to_role looks like this.
- --------
def add_user_to_role(user, role_or_key)
role = self.class.sanitize_role( role_or_key )
binding = role_bindings.find( :first, :conditions=>
[ ''role_id=? AND user_id=?'', role.id, user.id ] )
if ( binding.nil? )
role_bindings.create
( :context=>self, :user=>user, :role=>role )
end
end
- --------
When #add_user_to_role is called in the controller during my spec,
the method is run, rather than being a stub as I would expect. Is
this expected and/or known behavior? Any comments and direction are
very much appreciated.
Thanks
Lance!
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (Darwin)
iD8DBQFHp2AG+EgDjU0WSaoRArYDAJ92Wpp1Nkybqk7lb8fY+of+tcIawACaAyRk
y6Lo0iCvWS8/afBDTx42XiY=dfEK
-----END PGP SIGNATURE-----
On Feb 4, 2008 12:57 PM, Lance Ball <lance at electricsheepcompany.com> wrote:> -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi there > > I''ve run into a bit of a problem that I can''t seem to figure out. I > have a model class User which includes our acts_as_role_context > library. This lib acts like most other acts_as libs, evaluating > class and instance methods at runtime. When I try to stub methods > that are defined in acts_as_role_context, I get strange behavior. > Perhaps it''s just me. I am a bit under the weather, and the Nyquil > could be clouding my thinking. :) > > Here is my spec. It''s pretty simple, I just want to ensure that the > correct redirect happens when it should. For the record, the > controller works as expected outside of rspec. > > - --------- > require File.dirname(__FILE__) + ''/../spec_helper'' > > describe MemberNetworkController do > before(:each) do > controller.should_receive(:beta_lockdown).and_return true > login > end > > after(:each) do > logout > end > > describe "on GET to add_as_friend" do > it "should redirect to /network/friends" do > logged_in_user.stub!(:add_user_to_role) > controller.stub!(:make_friend_activity) > get :add_as_friend, :key => logged_in_user.key > response.should redirect_to("/members/#{logged_in_user.id}/ > network/friends") > end > end > end > - -------- > > The line in question is logged_in_user.stub!(:add_user_to_role). > User#add_user_to_role is defined in acts_as_role_context.rb. I''ll > not paste the entire file unless requested. It''s fairly large. the > instance method #add_user_to_role looks like this. > > - -------- > def add_user_to_role(user, role_or_key) > role = self.class.sanitize_role( role_or_key ) > binding = role_bindings.find( :first, :conditions=> > [ ''role_id=? AND user_id=?'', role.id, user.id ] ) > if ( binding.nil? ) > role_bindings.create > ( :context=>self, :user=>user, :role=>role ) > end > end > - -------- > > When #add_user_to_role is called in the controller during my spec, > the method is run, rather than being a stub as I would expect. Is > this expected and/or known behavior? Any comments and direction are > very much appreciated.My best guess is that the instance of logged_in_user is not the same one that the controller action is using. You''re going through the controller and passing it a key, at which point it''s probably using that key to access the logged in user from a source that is not the same as your logged_in_user method in the spec. HTH, David> > Thanks > Lance! > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.7 (Darwin) > > iD8DBQFHp2AG+EgDjU0WSaoRArYDAJ92Wpp1Nkybqk7lb8fY+of+tcIawACaAyRk > y6Lo0iCvWS8/afBDTx42XiY> =dfEK > -----END PGP SIGNATURE----- > _______________________________________________ > rspec-users mailing list > rspec-users at rubyforge.org > http://rubyforge.org/mailman/listinfo/rspec-users >
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Feb 4, 2008, at 2:34 PM, David Chelimsky wrote:> On Feb 4, 2008 12:57 PM, Lance Ball > <lance at electricsheepcompany.com> wrote: >> >> When #add_user_to_role is called in the controller during my spec, >> the method is run, rather than being a stub as I would expect. Is >> this expected and/or known behavior? Any comments and direction are >> very much appreciated. > > My best guess is that the instance of logged_in_user is not the same > one that the controller action is using. You''re going through the > controller and passing it a key, at which point it''s probably using > that key to access the logged in user from a source that is not the > same as your logged_in_user method in the spec.Yes, that was my thought as well. So, I printed the ID of the logged in user in the spec and the ID in the controller. The IDs were the same. After that, I sent this note to the list. But, it turns out I should have been more thorough in my investigation. Instead of printing out the ID, this time I printed the object itself and found: #<User:0x303fc5c> #<User:0x2fa9bbc> So you are correct! Thanks. :) Lance -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (Darwin) iD8DBQFHp20H+EgDjU0WSaoRAnNkAJ9GpF2noGGfbfMyzXFJ3S8WxncnPACfdxHZ ukGJeaOcZ8vnesebC7hRgD0=8cNF -----END PGP SIGNATURE-----
On Feb 4, 2008 1:52 PM, Lance Ball <lance at electricsheepcompany.com> wrote:> -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > > On Feb 4, 2008, at 2:34 PM, David Chelimsky wrote: > > > On Feb 4, 2008 12:57 PM, Lance Ball > > <lance at electricsheepcompany.com> wrote: > >> > >> When #add_user_to_role is called in the controller during my spec, > >> the method is run, rather than being a stub as I would expect. Is > >> this expected and/or known behavior? Any comments and direction are > >> very much appreciated. > > > > My best guess is that the instance of logged_in_user is not the same > > one that the controller action is using. You''re going through the > > controller and passing it a key, at which point it''s probably using > > that key to access the logged in user from a source that is not the > > same as your logged_in_user method in the spec. > > Yes, that was my thought as well. So, I printed the ID of the logged > in user in the spec and the ID in the controller. The IDs were the > same. After that, I sent this note to the list. > > But, it turns out I should have been more thorough in my > investigation. Instead of printing out the ID, this time I printed > the object itself and found: > > #<User:0x303fc5c> > #<User:0x2fa9bbc> > > So you are correct! Thanks. :)Sure. You good on how to solve it?> > Lance > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.7 (Darwin) > > iD8DBQFHp20H+EgDjU0WSaoRAnNkAJ9GpF2noGGfbfMyzXFJ3S8WxncnPACfdxHZ > ukGJeaOcZ8vnesebC7hRgD0> =8cNF > > -----END PGP SIGNATURE----- > _______________________________________________ > rspec-users mailing list > rspec-users at rubyforge.org > http://rubyforge.org/mailman/listinfo/rspec-users >