Prevent infinite loop with self-referencial records - ruby-on-rails-3

I've got a model that has a self-referencial join through a join table, as defined below:
class Task < ActiveRecord::Base
has_many :dependency_dependents, foreign_key: :dependency_id, class_name: 'TaskDependency', dependent: :destroy
has_many :dependency_dependencies, foreign_key: :task_id, class_name: 'TaskDependency', dependent: :destroy, autosave: true
has_many :dependencies, through: :dependency_dependencies
has_many :dependents, through: :dependency_dependents, source: :task
end
class TaskDependency < ActiveRecord::Base
belongs_to :task
belongs_to :dependency, class_name: 'Task'
end
Everything with the joins works well. When the record is saved, it performs a bunch of calculations based on the records it depends on, and it then moves on the updating the dependent records which in turn perform the same calculations for those records and so on.
The issue is that if somewhere in the chain of dependent tasks, there is a dependency on a task somewhere else up the tree, firstly the calculations will fail, but most importantly, it'll cause an infinite loop of calculations and updates.
Is there a nice way I can check for this infinite loop before I save a record, (ideally before creating the dependency in the first place).
I'm happy to do this either using pure SQL or ruby, just wondering if anyone has a clean solution for this.

Right, I think I've got it figured out. When the record validates, it calls a new validation method which runs a recursive query which pulls out all the ID's in the hierarchy below it, and if the ID of the task that you are trying to make a dependency is found in the list of dependents, then it fails validation. The query for this is below:
WITH RECURSIVE dependencies(task_id, path) AS (
SELECT task_id, ARRAY[task_id]
FROM task_dependencies
WHERE dependency_id = #{self.id}
UNION ALL
SELECT task_dependencies.task_id, path || task_dependencies.task_id
FROM dependencies
JOIN task_dependencies ON dependencies.task_id = task_dependencies.dependency_id
WHERE NOT task_dependencies.task_id = ANY(path)
)
SELECT task_id id FROM dependencies ORDER BY path
Never know, maybe this will help someone along the way.
As mentioned by #craig-ringer though, there is still the possibility of two concurrent inserts occurring leading to a race condition, but this query could also in turn lock the table to prevent this happening if necessary.

Related

Add Arbitrary Attribute to SQL Query from Joins Record without WHERE clause (Active record)

I'm trying to create an attribute in my select statement that depends on whether or not an association exists. I'm not sure if it's possible with a single query, and the goal is to not have to iterate a list afterward.
Here is the structure.
class Project < ApplicationRecord
has_many :subscriptions
has_many :users, through: :subscriptions
end
class User < ApplicationRecord
has_many :subscriptions
has_many :projects, through: :subscriptions
end
class Subscription < ApplicationRecord
belongs_to :project
belongs_to :user
end
Knowing a project, the goal of the query is to return ALL users and include on them a new attribute call subscribed - denoting whether or not they are subscribed.
non-working code (pseudo code):
project = Project.find_by(name: 'has_subscribers')
query = 'users.*, (subscriptions.project_id = ?) AS subscribed'
users = User.includes(:subscriptions).select(query, project.id)
user.first.subscribed
# => true or false
I'm open to whether or not there is a better way of going about this. However, the information is:
You know the project record.
You query a list of ALL users
Each user record has a subscribed attribute, denoting whether its
subscribed to the given project
Solution:
I was able to figure out a straight forward solution using the bool_or aggregate method. Coalesce ensures that the value returned is false instead of nil, should no subscriptions exists.
query = "users.*, COALESCE(bool_or(subscriptions.project_id = '#{project_id}'::uuid), false) as subscribed"
User.left_outer_joins(:subscriptions)
.select(query)
.group('users.id')
Yep, you can do this:
User.joins(:projects).select(Arel.star, Subscription.arel_table[:project_id])
Which will result in a SQL query like this:
SELECT *, "subscriptions"."project_id" FROM "users" INNER JOIN "subscriptions" ON "subscriptions"."user_ud" = "users"."id";
If you want to specify a specific project (i.e. use an expression), you can do it with Arel like this:
User.joins(:projects).select(Arel.star, Subscription.arel_table[:project_id].eq(42))
Unfortunately, you won't have a column name alias, and you can't call as on an Arel::Nodes::Equality instance. I don't know enough about the internals of Arel to have a way out of that box. But you can do this if you want the composability of Arel (e.g. if this is going to be something that needs to work with multiple models or columns):
User.joins(:projects).select(Arel.star, Subscription.arel_table[:project_id].eq(42).to_sql + " as has_project")
This is a bit clunky, but it works and provides a user.has_project method that returns a boolean. You can pretty it up like so:
class User
scope :with_project_status, lambda do |project_id|
has_project =
Subscription.arel_table[:project_id].
eq(project_id).to_sql + " as has_project"
joins(:projects).select(Arel.star, has_project)
end
end
User.with_project_status(42).where(active: true)

Is there a better way to structure these database tables?

I'm struggling with the best way to design/query my db on a new Rails app. This is what I have in place right now:
documents:
title
has_many :document_sections
document_sections:
belongs_to :document
habtm :resources
resources_document_sections:
belongs_to :resource
belongs_to :document_section
resources:
text
So it's easy to say document_section.resources. But document.resources is giving me trouble
The only way I've found to do it so far is to collect the document section ids, and then run a second query:
d = Document.last
s_ids = d.document_section_ids
Resource.joins(:document_sections)
.where(document_sections: { id: s_ids })
.uniq
So this starts bad, and gets worse as the queries get more complicated. It's becoming quite a headache every time I have to touch this relationship.
I'm wondering if there is a different pattern that I could follow in laying out these tables, such that querying against them is not such a headache? Or is there a better querying strategy that I'm missing?
Your document has no relationships. You need to add something to both models in a relationship, you can't just add it to document_sections and expect documents to have any kind of relationship to anything.
You need to add a has_many ... through: to your Document:
class Document < ActiveRecord::Base
has_many :document_sections
has_many :relationships, through: :document_sections
end

Rails 3 has_many :through accessing attributes

I am working with a has_many through for the first time, and despite a lot of reading here and in the guide I am not understanding the correct way to access attributes on the through table. My tables are the same as this example from another post.
class Product < ActiveRecord::Base
has_many :collaborators
has_many :users, :through => :collaborators
end
class User < ActiveRecord::Base
has_many :collaborators
has_many :products, :through => :collaborators
end
class Collaborator < ActiveRecord::Base
belongs_to :product
belongs_to :user
end
Assuming that the collaborators table has additional attributes, say hours_spent, what is the correct way to find the hours_spent from the collaborator table for a particular user and product?
When I have found my users via the product, and am iterating over them as in
#product.users.each do |user|
This seems to work
user.collaborator[0].hours_spent
I get the correct value, but since there should only be one collaborator record for each User/Product pair, the index is throwing me off, making me think I’m doing something wrong.
Thank you for reading!
EDIT
Perhaps I am not getting the has_many through concept. Maybe a MySQL example would help.
What I was thinking is that if I did
SELECT * FROM collaborators where user_id = 1;
I would expect a set (zero or more) as the result. Similarly
SELECT * FROM collaborators where product_id = 1;
would also give me a set, but
SELECT * FROM collaborators where user_id = 1 and product_id = 1;
would give at most 1 row.
If I am understanding properly, all 3 queries return a set. So I guess I need some kind of uniqueness constraint, but that would have to be a compound key of sorts, on both of the belongs to keys. Is that even possible? Is there a structure that better models this?
Thanks so much for the quick and helpful responses!
There may be a single database row per pair, but when considering a single user, that user can be associated to many products, so a user can have many rows in the collaborators table. Similarly, when considering a single product, that product can be associated to many users, so a product can have many rows in the collaborators table.
Also, instead of using user.collaborators[0].hours_spent, use user.collaborators.first.try(:hours_spent) (which may return null), if you only want the first collaborator's hours spent.
If a single user can only have one single product and a single product can only have a single user, then switch the has_many's to has_one's for everything.
Update: The preceding is the answer to the original question which has since been clarified via comments. See comments for detail and see comments on other answer by Peter.
Perhaps you should use has_and_belongs_to_many. If your Collaborator is used only to make link between User and Product without having more fields.
class Product < ActiveRecord::Base
has_and_belongs_to_many :users
end
class User < ActiveRecord::Base
has_and_belongs_to_many :products
end
The beetween migration would be:
class CreateUsersProducts < ActiveRecord::Migration
def change
create_table "users_products", :id => false do |t|
t.integer :user_id
t.integer :product_id
end
end
end
After implementing this, what I found was that I think I had the correct relationships setup, I had to use the has_many :though as users could have many products, and it needed to be :through because there are additional attributes on the collaborator table. The sticking point was how to get there to only be a single Collaborator record for each user/product pair, and then how do I guarantee I got it. And to this point the answer I've found is it has to be done in code.
To make sure there is only a single record for each pair, I used
class Collaborator < ActiveRecord::Base
validates :product_id, :presence => true, :uniqueness => {:scope => [:user_id], :message => "This is a duplicate join"}
And then to make doubly sure I'm finding the right record, I have a scope
scope :collaboration_instance, lambda {|p_id, u_id| where("collaborations.product_id = ? && collaborations.user_id = ?", p_id, u_id)}
If someone has a more elegant solution, or just wants to improve this one, please post and I will change yours to the selected answer.

Rails STI association with subclasses

I'm getting some strange behaviour when fetching collections from a has_many association with rails 3 when using STI. I have:
class Branch < ActiveRecord::Base
has_many :employees, class_name: 'User::Employee'
has_many :admins, class_name: 'User::BranchAdmin'
end
class User < ActiveRecord::Base
end
class User::Employee < User
belongs_to :branch
end
class User::BranchAdmin < User::Employee
end
The desired behaviour is that branch.employees returns all employees including branch admins. The branch admins only seem to be 'loaded' under this collection when they have been accessed by branch.admins, this is output from the console:
Branch.first.employees.count
=> 2
Branch.first.admins.count
=> 1
Branch.first.employees.count
=> 3
This can be seen in the generated SQL, the first time:
SELECT COUNT(*) FROM "users" WHERE "users"."type" IN ('User::Employee') AND "users"."branch_id" = 1
and the second time:
SELECT COUNT(*) FROM "users" WHERE "users"."type" IN ('User::Employee', 'User::BranchAdmin') AND "users"."branch_id" = 1
I could solve this problem by just specifying:
class Branch < ActiveRecord::Base
has_many :employees, class_name: 'User'
has_many :admins, class_name: 'User::BranchAdmin'
end
since they all be found from their branch_id but this creates problems in the controller if I want to do branch.employees.build then the class will default to User and I have to hack at the type column somewhere. I have got around this for now with:
has_many :employees, class_name: 'User::Employee',
finder_sql: Proc.new{
%Q(SELECT users.* FROM users WHERE users.type IN ('User::Employee','User::BranchAdmin') AND users.branch_id = #{id})
},
counter_sql: Proc.new{
%Q(SELECT COUNT(*) FROM "users" WHERE "users"."type" IN ('User::Employee', 'User::BranchAdmin') AND "users"."branch_id" = #{id})
}
but I would really like to avoid this if possible. Anyone, any ideas?
EDIT:
The finder_sql and counter_sql haven't really solved it for me because it seems that parent associations don't use this and so organisation.employees that has_many :employees, through: :branches will again only include the User::Employee class in the selection.
Basically, the problem only exists in the development environment where classes are loaded as needed. (In production, classes are loaded and kept available.)
The problem comes in due to the interpreter not having seen yet that Admins are a type of Employee when you first run the Employee.find, etc. call.
(Notice that it later uses IN ('User::Employee', 'User::BranchAdmin'))
This happens with every use of model classes that are more than one level deep, but only in dev-mode.
Subclasses always autoload their parent hierarchy. Base classes don't autoload their child hierachies.
Hack-fix:
You can force the correct behaviour in dev-mode by explicitly requiring all your child classes from the base class rb file.
Can you use :conditions?
class Branch < ActiveRecord::Base
has_many :employees, class_name: 'User::Employee', :conditions => {:type => "User::Employee"}
has_many :admins, class_name: 'User::BranchAdmin', :conditions => {:type => "User::BranchAdmin"}
end
This would be my preferred method. One other way to do it might be to add a default scope to the polymorphic models.
class User::BranchAdmin < User::Employee
default_scope where("type = ?", name)
end
A similar problem continues to exist in Rails 6.
This link outlines the issue and workaround. It contains the following explanation and code snippet:
Active Record needs to have STI hierarchies fully loaded in order to generate correct SQL. Preloading in Zeitwerk was designed for this use case:
By preloading the leaves of the tree, autoloading will take care of the entire hierarchy upwards following superclasses.
These files are going to be preloaded on boot, and on each reload.
# config/initializers/preload_vehicle_sti.rb
autoloader = Rails.autoloaders.main
sti_leaves = %w(car motorbike truck)
sti_leaves.each do |leaf|
autoloader.preload("#{Rails.root}/app/models/#{leaf}.rb")
end
You may require a spring stop for the configuration changes to take.
Indeed, that was the plan in the early days of the gem, but it was abandoned soon (in 2019, before Rails 6 was out). Preloading has been deprecated for a long time, and has been deleted in the forthcoming Zeitwerk 2.5.
In a Rails application you can do it this way:
# config/initializers/preload_vehicle_sti.rb
Rails.application.config.to_prepare do
Car
Motorbike
Truck
end
That is, you "preload" just by using the constants in a to_prepare block.

Querying for rows without matching ID in associated table

I have a very standard app backed by an SQL database with a User model, a Problem model, and a CompletedProblem model acting as a join table between the two.
I'm trying to create a method that returns all problems not solved by a particular user. I have run into a wall, however, and I would appreciate pointers on what my method should look like.
Below are the models as well as my latest (incorrect) pass at creating this method.
class User < ActiveRecord::Base
has_many :completed_problems
has_many :problems, :through => :completed_problems
def unsolved_problems
Problem.includes({:wall => :gym}, :completed_problems).
where('completed_problems.user_id != ? OR completed_problems.user_id IS NULL)', self.id)
end
end
class Problem < ActiveRecord::Base
has_many :completed_problems
has_many :users, :through => :completed_problems
end
class CompletedProblem < ActiveRecord::Base
belongs_to :user
belongs_to :problem
end
(For the curious: this method does work so long as there is only one user marking problems as solved. As soon as you add a second, each user starts to return only those problems that have been solved by other users, instead of those not solved by herself.)
Via a friend:
select * from problems where id not in (select problem_id from completed_problems where user_id = USER_ID))
Although I'd still be interested in hearing if there's a way in ActiveRecord to do this.
I think something like this will do it:
Problem.where(["id NOT IN (?)", self.problems.all.map(&:id)])