Rails: Making sure two tables are "sync'd" - ruby-on-rails

Relationships:
Team has many Rosters
Team has many Announcements
Announcement has one AnnouncementRoster
When creating an Announcement, the user must assign a Roster to it (whose data will be used to create its new AnnouncementRoster).
My models:
class User < ActiveRecord::Base
has_many :roster_players
has_many :rosters, -> { uniq } , :through => :roster_players
has_many :announcement_rosters, -> { uniq } , :through => :announcement_players
end
class Roster < ActiveRecord::Base
has_many :roster_players
has_many :users, -> { uniq }, through: :roster_players
end
class RosterPlayer < ActiveRecord::Base
belongs_to :roster
belongs_to :user
validates :user_id, :uniqueness => { :scope => :roster_id }
end
class Announcement < ActiveRecord::Base
has_one :announcement_roster
end
class AnnouncementRoster < ActiveRecord::Base
has_many :announcement_players
has_many :users, -> { uniq }, through: :announcement_players
end
class AnnouncementPlayer < ActiveRecord::Base
belongs_to :announcement
belongs_to :user
end
I believe I need different relationship models (RosterPlayer, AnnouncementPlayer, TaskPlayer, etc.) because different relationships require different attributes. For example: an AnnouncementPlayer has a read? column, TaskPlayer has a progress column, and RosterPlayer has a captain? column.
Question:
What is the best way to make sure that the AnnouncementRoster is "sync'd" with the Roster it was copied from? In other words, if the copied Roster's data is changed (a user is added or removed from it) than the corresponding AnnouncementRoster should be updated accordingly.

The strategies you want to decide between, are:
use database-level idioms to cascade those updates (invisibly to rails)
Use after_save and after_update, having Rails push the changes
Create non-automatic methods (rake tasks) in Rails to push the updates on a cron
Push the changes with SQL scripts living in bash and cron-controlled (or similarly, have the changes sent by pigeon)
Which of these you choose is really dependent on your environment, application metrics, resiliency, etc.
Sorry this is an overview answer. Someone else can probably talk through the pro-con sides of this and help you make that decision based on yet-unstated info.

Related

Rails model associations - has_one or single table inheritance?

I'm having trouble deciding between Single Table Inheritance and a simple has_one relationship for my two models.
Background: I'm creating a betting website with a "Wager" model. Users may create a wager, at which point it is displayed to all users who may accept the wager if they choose. The wager model has an enum with three statuses: created, accepted, and finished.
Now, I want to add the feature of a "Favorite Wager". The point of this is to make it more convenient for users to create a wager, if they have ones they commonly create. One click instead of ten.
FavoriteWagers exist only as a saved blueprint. They are simply the details of a wager -- when the User wants to create a Wager, they may view FavoriteWagers and click "create", which will take all the fields of the FavoriteWager and create a Wager with them. So the difference is that FavoriteWagers acts as only as a storage for Wager, and also includes a name specified by the user.
I read up on STI, and it seems that a lot of examples have multiple subclassing - eg. Car, Motorcycle, Boat for a "Vehicle" class. Whereas I won't have multiple subclasses, just one (FavoriteWager < Wager). People have also said to defer STI until I can have more classes. I can't see myself subclassing the Wagers class again anytime soon, so that's why I'm hesitant to do STI.
On the other hand, has_one doesn't seem to capture the relationship correctly. Here is an example:
Class User < ApplicationRecord
has_many :favorite_wagers, dependent: :destroy
has_many :wagers, dependent: destroy
end
Class FavoriteWager < ApplicationRecord
has_one :wager
belongs_to: user, index: true, foreign_key: true
end
Class Wager < ApplicationRecord
belongs_to :favorite_wager, optional: true
belongs_to :user
end
I've also thought about just copying the fields directly, but that's not very DRY. Adding an enum with a "draft" option seems too little, because I might need to add more fields in the future (eg. time to auto-create), at which point it starts to evolve into something different. Thoughts on how to approach this?
Why not just do a join table like:
Class User < ApplicationRecord
has_many :favorite_wagers, dependent: :destroy
has_many :wagers, through: :favorite_wagers
end
Class FavoriteWager < ApplicationRecord
belongs_to :wager, index: true, foreign_key: true
belongs_to :user, index: true, foreign_key: true
end
Class Wager < ApplicationRecord
has_one :favorite_wager, dependent: destroy
has_one :user, through: :favorite_wager
end
Your FavoriteWager would have the following fields:
|user_id|wager_id|name|
That way you can access it like:
some_user.favorite_wagers
=> [#<FavoriteWager:0x00007f9adb0fa2f8...
some_user.favorite_wagers.first.name
=> 'some name'
some_user.wagers.first.amount
=> '$10'
some_user.wagers.first.favorite_wager.name
=> 'some name'
which returns an array of favorite wagers. If you only want to have ONE favorite wager per user you can tweak it to limit that. But this gives you the ability to have wagers and users tied together as favorites with a name attribute. I don't quite understand your use case of 'a live wager never has a favorite' but that doesn't matter, you can tweak this to suit your needs.

How do I write a Rails finder method through a chain of belongs_to associations?

I'm using Rails 5.1. How do I write a finder method when there is a chain of "belongs_to" associations? I have the following models ...
class Plan < ApplicationRecord
...
has_many :plan_items, :dependent => :destroy
class PlanItem < ApplicationRecord
...
belongs_to :offer, :optional => false
class Offer < ApplicationRecord
belongs_to :package, :optional => false
class Package < ApplicationRecord
has_and_belongs_to_many :items
I want to write a finder that gets all Plans with an Item with id = "blah". But the below is failing ...
[19] pry(main)> Plan.joins(plan_items: :offer).joins(packages: :item).where(:item => {:id => "abac"}).count
ActiveRecord::ConfigurationError: Can't join 'Plan' to association named 'packages'; perhaps you misspelled it?
from /Users/davea/.rvm/gems/ruby-2.5.1/gems/activerecord-5.2.2.1/lib/active_record/associations/join_dependency.rb:188:in `find_reflection'
How do I write a finder when there is a chain of belongs_to associations?
First, maybe your table name is wrong. Second, to pass method between belong_to association, you can use delegate
I'm assuming PlanItem is a join table between Plan and Item (that would be inline with the Rails naming convention). This could be done neatly with through associations and scopes. I would do it like this...
class Plan < ApplicationRecord
has_many :plan_items, dependent: :destroy
has_many :items, through: :plan_items
scope :blah_items { items.id_of_blah }
class PlanItem < ApplicationRecord
belongs_to :offer, optional: false
belongs_to :item
class Item < ApplicationRecord
scope :id_of_blah { where(id: 'blah') }
Then you can call it like so... Plan.with_blah_items or if you had an active record collection of plans you could use the scope to narrow it down plans.with_blah_items.
Since ActiveRecord associations will return ActiveRecord relations, you can chain them with any other active record methods (e.g. Plan.first.items.where(item: { id: 'blah' }) Scopes just make it nice and neat. : )
If PlanItem is not a join table between Plan and Item, first thing you should do is rename it. This is not just a best practice, rails spends a lot of time assuming what things are named, and it could cause a bug. After you rename it you should create a join table between Plan and Item called PlanItem. If a join between these tables doesn't make sense with your application architecture, you could always string through associations together, but that would be a code smell.
If you didn't want to mess with scopes, you could always just do a query like this plan.items.where(items: { id: 'blah' }).
Hope that helps!

Create Rails scope comparing fields on two tables

I have a number of associated tables in an application
class Listing < ActiveRecord::Base
belongs_to :house
belongs_to :multiple_listing_service
end
class House < ActiveRecord::Base
has_one :zip_code
has_one :primary_mls, through: :zip_code
end
I wanted to create a scope that produces all the Listings that are related to the Primary MLS for the associated House. Put another way, the scope should produce all the Listings where the multiple_listing_service_id = primary_mls.id for the associated house.
I've tried dozens of nested joins scopes, and none seem to work. At best they just return all the Listings, and normally they fail out.
Any ideas?
If I understand correctly, I'm not sure a pure scope would be the way to go. Assuming you have:
class MultipleListingService < ActiveRecord::Base
has_many :listings
has_many :zip_codes
end
I would go for something like:
class House < ActiveRecord::Base
...
def associated_listings
primary_mls.listings
end
end
Update 1
If your goal is to just get the primary listing then I would add an is_primary field to the Listing. This would be the most efficient. The alternative is a 3 table join which can work but is hard to optimize well:
class Listing < ActiveRecord::Base
...
scope :primary, -> { joins(:houses => [:zip_codes])
.where('zip_codes.multiple_listing_service_id = listings.multiple_listing_service_id') }

Rails: Validating an array of ids

Application
I am working on a college admissions system where a student can make an application to up to 5 courses. The way I have designed this is to have an Application model and a CourseApplication model. An application can consist of many course_applications:
class Application < ActiveRecord::Base
# Assosciations
belongs_to :user
has_many :course_applications, dependent: :destroy
has_many :courses, through: :course_applications
has_one :reference
# Validations
validates :course_applications, presence: true
end
Course Application
class CourseApplication < ActiveRecord::Base
# Intersection entity between course and application.
# Represents an application to a particular course, has an optional offer
# Associations
belongs_to :application
belongs_to :course
has_one :offer, dependent: :destroy
end
I want to make sure that a student cannot apply to the same course twice. I have already done some research but have had no success. This is the UI for a student making an application:
Screenshot of application form
When a course is selected, the course id is added to an array of course ids:
def application_params
params.require(:application).permit(:user, course_ids: [])
end
Right now a student can select the same course twice, I want to prevent them from doing this. Any help is much appreciated.
For the rails side, I would do on the CourseApplication
validates :course, uniqueness: { scope: :application }
For your reference this can be found at: http://guides.rubyonrails.org/active_record_validations.html#uniqueness
Also suggest on the database side to make a migration
add_index :course_applications, [:course, :application], :unique => true
For the validating on the form you will have to write javascript to make sure two things are selected, this will just return an error when someone tries to do it.

Rails - insert many random items on create with has_many_through relation

I want to create a random pack of 15 cards which should be invoked in the cardpacks_controller on create. I have the following models:
Card:
class Card < ActiveRecord::Base
# relations
has_many :cardpacks, through: :cardpackcards
belongs_to :cardset
end
Cardpack:
class Cardpack < ActiveRecord::Base
#relations
has_many :cards, through: :cardpackcards
belongs_to :cardset
# accept attributes
accepts_nested_attributes_for :cards
end
Cardpackcards:
class Cardpackcard < ActiveRecord::Base
#relations
belongs_to :card
belongs_to :cardpack
end
Cardsets:
class Cardset < ActiveRecord::Base
#relations
has_many :cards
has_many :cardsets
end
How can I create 15 Cardpackcards records with random card_id values and with the same cardpack_id (so they belong to the same pack)
I have watched the complex form series tutorial but it gives me no comprehension as how to tackle this problem.
I hope anyone can help me solve this problem and give me more insight in the rails language.
Thanks,
Erik
Depending on the database system you might be able to use an order random clause to find 15 random records. For example, in Postgres:
Model.order("RANDOM()").limit(15)
Given the random models, you can add a before_create method that will setup the associations.
If the Cardpackcard model doesn't do anything but provide a matching between cards and cardpacks, you could use a has_and_belongs_to_many association instead, which would simplify things a bit.
Without it, the controller code might look something like this:
cardset = Cardset.find(params[:cardset_id])
cardpack = Cardpack.create(:cardset => cardset)
15.times do
cardpack.cardpackcards.create(:card => Card.create(:cardset => cardset))
end

Resources