/pairguru

Primary LanguageRuby

Welcome to Pairguru App!

Place where we examine our candidates

For recruiters

For candidates

🔨 Setup

Before pairing:

  • Clone repo git clone git@github.com:netguru-training/pairguru.git
  • run bundle install within app directory
  • run rake db:setup This will create database, run migrations and fill db with seeds
  • run rails s 💥

Pre-interview tasks

  • You don't have to do any tasks that you weren't asked to solve
  • Do not fork the repository - clone it
  • Do not forget to write tests for your solutions
  • Do not remove contents of this README
  • Put solutions for tasks in separate pull requests with descriptive commits, merge them to master branch after all
  • Do not forget to check if all tests pass by running command RAILS_ENV=test bundle exec rspec from you app folder

Task 1 - consume api

We would like to have more information about the movie in movie show and index page. The information should come from pairguru api (e.g. https://pairguru-api.herokuapp.com/api/v1/movies/Godfather). We should display movie poster (e.g. https://pairguru-api.herokuapp.com/godfather.jpg), rating and plot overview. We can't store this data in database.

Comments

I created separate repository to mock external API for local testing - just run it on port 3001: https://github.com/szymon-przybyl/pairguru_api. It's returning the same data for every movie with ID in range 1-100.

In Movie#api I'm using id as identifier of movies in external API - in case they should be different, we can use there attribute like api_id.

Task 2 - build api

We would like to share our movies via api. We want to return list of all out movies (id and title). Besides that we would like to for user to get specific movie by id.

After a while developers started asking us to extend our results by adding genre details. They want genre id, name and number of movies in this genre to be returned along with the movies. We don't want to break api for existing users or make them to fetch more data.

Comments

Since we need to deal with dynamic scope of query I chosen GraphQL to implement this API. To test it locally you can go to /graphiql playground.

Want all the movies? Execute:

{
  movies {
    id,
    title
  }
}

Just one movie:

{
  movie(id: 1) {
    id,
    title
  }
}

All the movies with genre details:

{
  movies {
    id,
    title,
    genre {
      id,
      name,
      moviesCount
    }
  }
}

One movie with genre:

{
  movie(id: 1) {
    id,
    title,
    genre {
      id,
      name,
      moviesCount
    }
  }
}

Task 3 - long running tasks

We have two tasks that run too long and users leave our page frustrated: emailing details about movie (button in movie show) and emailing movies csv export (in navigation). We should make it that user doesn't have to wait for page load and still get emails.

Comments

I used sidekiq for running background jobs. To test it locally you need to start it along with redis:

sidekiq
redis-server

To monitor jobs you can go to /sidekiq.

Task 4 - brackets validation

Our moderators are adding parentheses and brackets to the titles.

For example The Fellowship of the Ring [Lord of The Rings {Peter Jackson}] (2012).

This is ok, but we need to make sure that all brackets are closed and not empty. They can be nested as in the example above.

Uncomment tests in spec/validators/title_brackets_validator_spec.rb and add validations for movie title that match all the criteria. With all those tests passing you will know that solution is correct.

Hint: It can't be solved via regular expressions.

Task 5 - comments

Our users want to comment on movies. Let's add feature that allows them to do that.

We should have new comment form under the movie description and all the comments listed below that.

Additionally - one user should be allowed to leave only one comment to the movie. He can create another only if he deletes previous one.

Task 6 - sql query (dependent on 5)

We want to reward our users for commenting movies. The idea is to create page with top commenters this week.

It should have 10 users that have the most comments in last week (from 7 days ago to now).

This page will be very popular. We don't want to spam database with many queries in one pageview.

Hint: It can be solved by using ActiveRecord.

For recruiters

There is private pairguru repository. In pull requests you can find example solutions for this tasks.

To contribute fixes/updates

  • Fork repository
  • Create pull request (from forked repository, not pairguru itself)
  • Ask for review one of the contributors
  • After PR is merged branch will be automatically propagated to public repo for candidates

To add new tasks (please do!)

  • Add task description in this README
  • Create example solution on a branch with solution in branch name (eg. solution-task-7-description) - it won't be propagated to public repo this way
  • You can create specific setup for your task (eg. some prepared bug, configured tool to be used)
    • Create new branch with this setup - it will be propagated to public repo
    • Push branch to pairguru. It will be propagated to public repo
    • Specify in this task description that it should be solved on this branch
  • If you change something in the repo please try to update ruby/rails/gems versions to the latest.