Issue #1434 filter functionality for has many association.

[barsukov]

This commit has few changes :

1. I implement filter functionality for has_many association with issue has_many :through associations filter problem #1434
2. I refactor one spec with rspec best practices and factory girl too.

[bbenezech]

Does it really make sense to filter on a HM relation ?
Maybe it would make more sense to do has_one first. Also nested models with Mongoid aren't covered. Maybe @mshibuya wants to give it a try ?
I can't add HM without HMT and HABTM, this would be awkward, and I am not sure it's a good idea anyway.

[barsukov]

Hm.. Can you explain me what i need to do for compleate this request , I will try to improve my pull request.

[barsukov]

@bbenezech this implementation work if you set type of association of field to :has_many_association type => HABTM and HMT does not implemented for field but I could be wrong.

[coveralls]

Coverage decreased (-5.23%) when pulling 2185234 on barsukov:master into 05a029d on sferik:master.

View Details

[coveralls]

Coverage decreased (-0.25%) when pulling 2185234 on barsukov:master into 05a029d on sferik:master.

View Details

[coveralls]

Coverage decreased (-0.15%) when pulling 2185234 on barsukov:master into 05a029d on sferik:master.

View Details

[coveralls]

Coverage decreased (-0.26%) when pulling cddeb94 on barsukov:master into 0be9e22 on sferik:master.

[rafaelgonzalez]

@bbenezech

It makes sense to me to have a HM or HABTM filter.

I have a Project model with many Task and want to be able to filter out projects that have tasks with a specific name, or specific content in the name. (in truth it is really a HABTM relationship in my case, so I would like to filter all the projects that have the same task assigned to them but I use has_many :through with an intermediate model)

Please see #1434 were a partial solution has been discussed.

[barsukov]

I think this is not actual anymore?

[rafaelgonzalez]

Well the solution I use in #1434 only partially solves my problem (see my comment above), so to me this is still actual.

[mshibuya]

The same as #1363.