Thanks for taking the time to evaluate my submission. I've built out my solution as a Rails Applicaiton with Postgres.

This is a pretty standard rails app, but here are some quick links to the most interesting bits:

• app/models/node.rb - Main business logic for node querying
• spec/models/node_spec.rb - Unit tests for Node model
• app/models/bird.rb - Main business logic for bird querying
• spec/models/bird_spec.rb - Unit tests for Bird model
• app/controllers/main_controller.rb Application Controller for both endpoints
• db/seeds.rb - Data loader for getting the provided data into the database

The Application shouldn't need anything more than a relatively recent build of Postgres (I'm on 14.5) and Ruby 3.3.0

Pull down the repo and run these commands to get started:

bundle install
bundle exec rake db:create db:migrate db:seed
bundle exec rails s

NOTE: The db:seed task will truncate both tables to allow for inserting records with specific Ids

I've included two tables, nodes and birds. The db:seed task will truncate both tables, and reset the auto increment fields on them.

To maintain the Id's provided in the data file, I'm inserting records with those id's explicitly. I would consider this an anti-pattern, but I wanted to keep the schema simple, while still allowing you to interact with the actual provided data.

For each Node, I also randomly generate 0-2 birds. I gave birds a name column that I populate with Faker, just for a bit of fun.

To keep to the requirements given, I just made a "Main" controller that surfaces the two requested endpoints. Both endpoints always respond with JSON.

This endpoint takes any two Node Ids. If both nodes exist, and share a common ancestor it will respond with:

{root_id: 130, lowest_common_ancestor: 130, depth: 1}

If either node does not exist, or they do not share a common ancestor, it will respond with:

{root_id: null, lowest_common_ancestor: null, depth: null}

This endpoint takes a comma separated list of node_ids. It finds a unique set of all descendants of the passed nodes, and returns data about all birds related to any of those nodes, eg:

[{"id":1889,"node_id":2426570,"name":"Florinda"},...]

Any node_ids that are not found will be ignored.

If two nodes passed are related by ancestry, it will not result in duplicate birds being returned.

NOTE: I expanded on the bird endpoint a bit to include the id of the node that the bird belongs to, and its name.

• implemented RSpec to cover common use cases at the controller level, as well as more detailed unit coverage in the models. Run rspec with bundle exec rspec spec
• used Rubocop with some custom rule modification to ensure consistent code quality.
• used FactoryBot to build my testing data.
• used the activerecord-import gem to load the seed data, which allowed me to directly load the CSV data, and generated birds into the database in just a second or two. Inserting the rows individually took almost a minute!