By default, Rails' validates_uniqueness_of does not consider "username" and "USERNAME" to be a collision. If you use MySQL this will lead to issues, since string comparisons are case-insensitive in MySQL.

(If you use PostgreSQL, read this instead.)

Say you have a user model

class User < ActiveRecord::Base
  validates_uniqueness_of :name
end

with a unique index in the database.

If you try to create the users "user" and "USER", this will not trigger a validation error, but may fail with an SQL error due ...

When you call a method on an object, Ruby looks for the implementation of that method. It looks in the following places and uses the first implementation it finds:

1. Methods from the object's singleton class (an unnamed class that only exists for that object)
2. Methods from prepended modules (Ruby 2.0+ feature)
3. Methods from the object's class
4. Methods from included modules
5. Methods from the class hierarchy (superclass and its an...

Tests are about 100% control over UI interaction and your test scenario. Randomness makes writing tests hard. You will also push tests that are green for you today, but red for a colleague tomorrow.

That said, please don't do something like this:

Factory(:document) do |document|
  document.category { ['foo', 'bar', 'baz'].sample }
end

Instead do this:

Factory(:document) do |document|
  document.category 'foo'
end

The case against Faker

I even recommend to not use libraries like [Faker]...

Say you have a User with a Carrierwave attribute #avatar:

class User < ActiveRecord::Base
  mount_uploader :avatar, AvatarUploader
end

When whitelisting the avatar field in the controller, you might do this:

params[:user].permit(:avatar)

But you probably want this:

params[:user].permit(:avatar, :avatar_cache, :remove_avatar)

In this example:

• :avatar_cache allows a newly upload image to persist through form roundtrips in the case of validation errors (something that isn't possibl...

Whenever you create a table from a database migration, remember to add updated_at and created_at timestamps to that table. Without those timestamps, investigating future bug reports will be hell. Always have timestamps.

Adding timestamps to new tables

When you create a table using create_table, you can add timestamps by using the timestamps shortcut:

class CreateEpisode < ActiveRecord::Migration
  def change
    create_table :episodes do |t|
      t.string :name
      t.timestam...

When two classes implement the same behavior (methods, callbacks, etc.), you should extract that behavior into a trait or module. This card describes how to test that extracted behavior without repeating yourself.

Note that the examples below use Modularity traits to extract shared behavior. This is simply because we like to do it that way at makandra. The same techniques apply for modules and overriding self.included.

Example
---...

Custom matchers are a useful RSpec feature which you can use to DRY up repetitive expectations in your specs. Unfortunately the default directory structure generated by rspec-rails has no obvious place to put custom matchers or other support code.

I recommend storing them like this:

spec/support/database_cleaner.rb
spec/support/devise.rb
spec/support/factory_bot.rb
spec/support/vcr.rb
spec/support/matchers/be_allowed_access.rb
s...

1. Saving files to a directory that is not shared between deploys or servers

If you save your uploads to a made up directory like "RAILS_ROOT/uploads", this directory goes away after every deploy (since every release gets a new). Also this directory is not shared between multiple application servers, so your uploads are randomly saved to one local filesystem or another. Fixing this afterwards is a lot of fun.

Only two folders are, by default, shared between our application servers and deployments: "RAILS_ROOT/storage" and `"RAILS...

Merge requests are often rejected for similar reasons.

To avoid this, before you send a merge request, please confirm that your code ...

• has been reviewed by yourself beforehand
• fulfills every requirement defined as an acceptance criterion
• does not have any log or debugging statements like console.log(...), byebug etc.
• has green tests
• has tests...

Sometimes you want git to ignore certain files that appear on your machine. You can do this in 3 ways:

• Per project, in the project's .gitignore file
• Per project, in a local exclude file
• Globally on your machine

Downsides of per-project .gitignore entries

While it might be tempting to set it per project (other devs might benefit from it), you

• need to do it each time for every project
• "pollute" a project's .gitignore file with stuff...

When HTTP clients make an request they can define which response formats they can process. They do it by adding a header to the HTTP request like this:

Accept: application/json

This means the client will only understand JSON responses.

When a Rails action is done, it will try to render a template for a format that the client understand. This means when all you are HTML templates, a request that only accepts application/json will raise an error:

An ActionView::MissingTemplate occurred in pages#foo:
  Missing templa...

In the following example the method update_offices_people_count won't be called when office_id changes, because it gets overwritten by the second line:

after_save :update_offices_people_count, :if => :office_id_changed? # is overwritten …
after_save :update_offices_people_count, :if => :trashed_changed? # … by this line

Instead write:

after_save :update_offices_people_count, :if => :office_people_count_needs_update?

private

def office_people_count_needs_update?
  office_id_changed? || trashed_changed?
end

Or...

Rails gives you migrations to change your database schema with simple commands like add_column or update.
Unfortunately these commands are simply not expressive enough to handle complex cases.

This card outlines three different techniques you can use to describe nontrivial migrations in Rails / ActiveRecord.

Note that the techniques below should serve you well for tables with many thousand rows. Once your database tables grows to millions of rows, migration performance becomes an iss...

Caution when using .where to exclude records from a scope like this:

# Fragile - avoid
User.where("id NOT IN (?)", excluded_ids)

When the exclusion list is empty, you would expect this to return all records. However, this is not what happens:

# Broken example
User.where("id NOT IN (?)", []).to_sql
=>  SELECT `users`.* FROM `users` WHERE (id NOT IN (NULL))

Passing an empty exclusion list returns no records at all! See below for better implementations.

Rails 4+

Use the .not method to let Rails do the logic

`...

Do not pass times to date attributes. Always convert times to dates when your application uses time zones.

Background

A time-zoned Time attribute on a Rails record is converted to UTC using to_s(:db) to be stored, and converted back into the correct time zone when the record is loaded from the database. So when you are not on UTC, time objects will be converted as follows.

>> Time.current
=> Fri, 15 Mar 2013 11:56:03 CET +01:00
>> Time.current.to_s(:db)
=> "2013-03-15 10:56:03" # This is now UTC

Problem

That will...

JavaScript has no built-in functions to compare two objects or arrays for equality of their contained values.

If your project uses Lodash or Underscore.js, you can use _.isEqual():

_.isEqual([1, 2], [2, 3]) // => false
_.isEqual([1, 2], [1, 2]) // => true

If your project already uses Unpoly you may also use up.util.isEqual() in the same way:

up.util.isEqual([1, 2], [2, 3]) // => false
up.util.isEqual([1, 2], [1, 2]) // => true

If you are wri...

The Edge Rider gem gives your relations a method #traverse_association which
returns a new relation by "pivoting" around a named association.

Say we have a Post model and each Post belongs to an author:

class Post < ActiveRecord::Base
  belongs_to :author
end

To turn a relation of posts into a relation of its authors:

posts = Post.where(:archived => false)
authors = posts.traverse_association(:author)

You can traverse multiple associations in a single call.
E....

TLDR: Ruby class variables (@@foo) are dangerous in many ways. You should avoid them at all cost. See bottom of this card for alternatives.

Class variables are shared between a class hierarchy

When you declare a class variable, it is shared between this and all descending (inheriting) classes. This is rarely what you want.

Class variables are bound at compile-time

Like unqualified constants, class variables are bound to your current scope *whe...

Ruby has two different ways to match the start and the end of a text:

• ^ (Start of line) and $ (End of line)
• \A (Start of string) and \z (End of string)

Most often you want to use \A and \z.

Here is a short example in which we want to validate the content type of a file attachment. Normally we would not expect content_type_1 to be a valid content type with the used regular expression image\/(jpeg|png). But as ^ and $ will match lines, it matches both content_type_1 and content_type_2. Using \A and \z will wo...

validates_uniqueness_of is not sufficient to ensure the uniqueness of a value. The reason for this is that in production, multiple worker processes can cause race conditions:

1. Two concurrent requests try to create a user with the same name (and we want user names to be unique)
2. The requests are accepted on the server by two worker processes who will now process them in parallel
3. Both requests scan the users table and see that the name is available
4. Both requests pass validation and create a user with the seemingly available name...

The problem

Calling bundle update GEMNAME will update a lot more gems than you think. E.g. when you do this:

bundle update cucumber-rails

... you might think this will only update cucumber-rails. But it actually updates cucumber-rails and all of its dependencies. This will explode in your face when one of these dependencies release a new version with breaking API changes. Which is all the time.

In the example above updating cucumber-rails will give you Capybara 2.0 (because capybara is a dependency of `cucumber-rail...

Calling bundle update (without arguments) updates all your gems at once. Given that many gems don't care about stable APIs, this might break your application in a million ways.

To stay sane, update your gems using the applicable way below:

Projects in active development

Update the entire bundle regularily (e.g. once a week). This ensures that your libraries are up-to-date while it's easy to spot major version bumps which may break the app.

Projects that have not been updated in a while

1. [Update a single gem conservatively](htt...

Basic error pages

To add a few basic styles to the default error pages in Rails, just edit the default templates in public, e.g. public/404.html.

A limitation to these default templates is that they're just static files. You cannot use Haml, Rails helpers or your application layout here. If you need Rails to render your error pages, you need the approach below.

Advanced error pages

1. Register your own app as the applicatio...