# Validates\_presence\_of :parent\_id stopping save **URL:** \ **Category:** Rails\ **Created:** [March 13, 2010, 4:54pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094 "2010-03-13T16:54:44Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 13, 2010, 4:54pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/1 "2010-03-13T16:54:44Z")
Hi. I’m using rails 2.3.4 and have this association between two of my models. Question has\_many :gradings Grading belongs\_to :question Grading has this validation: validates\_presence\_of :question\_id And this validation test is stopping me from doingg this, which i would normally expect to work @question = Question.new(attributes) @grading = @question.gradings.build(more\_attributes) @question.save #fails because grading has no question\_id Now, i thought that this would work ok because (in my understanding) the question is saved and then it saves it’s built associated objects. How can i ensure this will work? Do i need to drop the validation? I’d rather not if possible. Grateful for any advice - max ---
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 14, 2010, 12:36pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/2 "2010-03-14T12:36:11Z")
Hi, I’m not sure 100% what you’re trying to do but the presence of question\_id will not be there until the record has been saved. No? And I’m not sure why you’re doing this … @grading = @question.gradings.build(more\_attributes) Shouldn’t you have done a build before presenting your form if you have nested forms? Maybe you could teach me something on what you’re trying to do. On 13 March 2010 16:54, Max W. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > validates\_presence\_of :question\_id > can i ensure this will work? Do i need to drop the validation? I’d > To unsubscribe from this group, send email to > removed\_email\_address@domain.invalid[removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) > . > For more options, visit this group at > [http://groups.google.com/group/rubyonrails-talk?hl=en](http://groups.google.com/group/rubyonrails-talk?hl=en). – 100% naturally selected. 0% designed. ---
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 14, 2010, 2:49pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/3 "2010-03-14T14:49:36Z")
I’d say the best way to deal with this is never validate for a association\_id #validates\_presence\_of :question\_id validates\_presence\_of :question would be better. if grading.question\_id is nil then grading.question will also be nil. So you’re effectively doing the same check. I hope this helps RobL On 13 March 2010 15:54, Max W. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > validates\_presence\_of :question\_id > can i ensure this will work? Do i need to drop the validation? I’d > For more options, visit this group at [http://groups.google.com/group/rubyonrails-talk?hl=en](http://groups.google.com/group/rubyonrails-talk?hl=en). – Rob L. removed\_email\_address@domain.invalid > **[Posts](https://www.robl.me/)** > > I write thems codes ---
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 14, 2010, 10:26pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/4 "2010-03-14T22:26:05Z")
Hi Anthony, thanks. I’m not using the rails-supplied nested attributes methods here, though maybe i should. I’m effectively hand-rolling my own, so i create a question and some associated objects (eg a grading in this case) at the same time, from a single form. When i drop the validation on Grading#question\_id then it works fine. I’m just a little puzzled about the best way to go about validating question\_id, since the default way seems to stop me building a question and gradings at once. ---
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 14, 2010, 10:21pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/5 "2010-03-14T22:21:09Z")
Rob L. wrote: > I’d say the best way to deal with this is never validate for a > association\_id > > #validates\_presence\_of :question\_id > validates\_presence\_of :question > > would be better. Hi Rob! I played with this already, and got a surprising result: > > @question = Question.new(attributes) > > =\> > > > @grading = @question.gradings.build(other\_attributes) > > =\> > > > @grading.question > > =\>nil This surprised me because i thought that the use of build set up the association both ways, ie in both objects, so that even though @grading had no question\_id it ‘knew’ that it was associated with @question. I must admit that this aspect of rails (ie associations that have been built but not saved yet) is something that i’ve always found a bit magical and hence unpredicatable. ---
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 15, 2010, 1:23pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/6 "2010-03-15T13:23:52Z")
I don’t have your original mail and am not sure of the model but question\_id is a foreign key.No? You are checking for the existence of it before it has been created, that is why it’s failing, no? validate methods validate the data before its saved. Unless I’m missing something from your original query. Also, build will only build the object without saving it. So I think you are thinking something should be happening when it’s not. But I’m quite new to ruby/rails and although I’m developing an application and trying to learn at the same time, it’s difficult to grab all the nuances of Rails. There’s a lot to learn to become proficient. On 14 March 2010 22:26, Max W. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > Posted via [http://www.ruby-forum.com/](http://www.ruby-forum.com/). – 100% naturally selected. 0% designed. ---
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 16, 2010, 10:37am UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/7 "2010-03-16T10:37:12Z")
paul h wrote: > Hi Max, > > Have you tried: > > @question = Question.create(attributes) > @grading = @question.gradings.build(more\_attrs) > > Using create instead of new will save the Object to the DB first, and > hence: Hi Paul. Yes, i know that, thanks - in this situation i would do gradings.create rather than gradings.build. One of the advantages, though, of using build is that you can do it before the parent has been saved, and when the parent is saved it will save the child as well. In general, i try to stick to the standard controller code as much as possible, which means setting the child gradings through a set method which will be called by .update\_attributes. This is a nice approach as it means no changes to the standard controller code is required. When i drop the validation on Grading#question\_id (or Grading#question) then it works fine. My question was really about the behaviour of the unsaved child objects, and whether they ‘know’ they have a parent. It seems they don’t. ---
**Author:** ![cmaxvv](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/85f322/32.png) [@cmaxvv](https://www.ruby-forum.com/u/cmaxvv)\ **Post date:** [March 15, 2010, 7:53pm UTC](https://www.ruby-forum.com/t/validates-presence-of-parent-id-stopping-save/185094/8 "2010-03-15T19:53:13Z")
Hi Max, Have you tried: @question = Question.create(attributes) @grading = @question.gradings.build(more\_attrs) Using create instead of new will save the Object to the DB first, and hence: @question[:id] will/should/may not be nil when building your gradings. I’m six months in to Rails, don’t use build, and haven’t needed to use any validation yet - am currently extending the generator for my current and future Apps, so I have held fire on my first app until I get this completed, but if your only issue is that the FK question\_id requires @question[:id] to exist, then unless I’m mistaken(which I could be), using create over new should work… HTH Paul