Book Image

Learning Mongoid

By : Gautam Rege
Book Image

Learning Mongoid

By: Gautam Rege

Overview of this book

Mongoid helps you to leverage the power of schema-less and efficient document-based design, dynamic queries, and atomic modifier operations. Mongoid eases the work of Ruby developers while they are working on complex frameworks. Starting with why and how you should use Mongoid, this book covers the various components of Mongoid. It then delves deeper into the detail of queries and relations, and you will learn some tips and tricks on improving performance. With this book, you will be able to build robust and large-scale web applications with Mongoid and Rails. Starting with the basics, this book introduces you to components such as moped and origin, and how information is managed, learn about the various datatypes, embedded documents, arrays, and hashes. You will learn how a document is stored and manipulated with callbacks, validations, and even atomic updates. This book will then show you the querying mechanism in detail, right from simple to complex queries, and even explains eager loading, lazy evaluation, and chaining of queries. Finally, this book will explain the importance of performance tuning and how to use the right indexes. It also explains MapReduce and the Aggregation Framework.
Table of Contents (14 chapters)
Learning Mongoid
About the Author
About the Reviewers

A practical approach using the Sodibee library system

Sodibee (pronounced saw-di-bee) is a library-management system that can manage books, reviews, authors, and bookings. Here are some of the functions that are supported in Sodibee:

  • An author has many books and a book belongs to an author

  • A book has many reviews and has one booking

  • A review belongs to a user and is about a book

  • A booking belongs to a user and a book


In the course of this book we will be working with the latest versions of Ruby 2.0, Mongoid 4, Rails 4, and MongoDB 2.4.

Checking prerequisites

First and foremost, we need to ensure that we have our development environment set up. It's common to use multiple versions of Ruby for development; I use RVM to manage these versions. As we can have multiple versions of the same gem installed on our machines, we use RVM gemsets to manage the gems we need for our work.

Ruby version

To check the Ruby version, check the version that is installed using the following command:

$ rvm list

rvm rubies

   jruby-1.7.4 [ x86_64 ]
   ruby-1.9.3-p385 [ x86_64 ]
=* ruby-2.0.0—p247 [ x86_64 ]

# => - current
# =* - current && default
#  * - default

$ ruby –v
ruby 2.0.0p247 (2013-06-27 revision 41674) [x86_64-darwin12.4.0]


Ruby 2.1 is due to be released in December, 2013. Everything in this book will be fully compatible with Ruby 2.1 too.

MongoDB version

We are currently using MongoDB v2.4.6—verify that, using the following command:

$ mongo
MongoDB shell version: 2.4.6
connecting to: test

If you don't see this, it's quite likely that you have not installed MongoDB or it isn't running. Get going!

Setting up Sodibee

First and foremost, let's install the Rails gem.

$ gem install rails

This installs Rails 4.0.0.


At the time of writing this book, the Rails version was 4.0.0. All commands would be fully compatible with the latest Rails version.

Now, let's create the Sodibee project.

$ rails new sodibee -O -T

This creates a new Rails project. The –O option tells Rails to skip ActiveRecord (we don't need it), and –T tells Rails to skip test unit. (We plan to use rspec later).


When you run the preceding command, it initiates a bundle install and updates our bundle with the default gems. If you are as impatient as I am, you may interrupt the process and press Ctrl + C to stop it, as we need to modify Gemfile to add other gems anyway.

Now, open Gemfile and configure for Mongoid.

gem 'mongoid', git: "git://"


Mongoid master is currently in sync with Rails 4. So, if we install using the released gem, it will install Version 3.x.

We're almost done. Issue the following command to update the bundle with our Mongoid gem:

$ bundle

If you did not use the –O option, you can run the following instructions to remove ActiveRecord from the application as we don't need it. Check and remove database.yml under config, if it has been generated. Next ensure that application.rb under config has the following lines:

require File.expand_path('../boot', __FILE__)

require "action_controller/railtie"
require "action_mailer/railtie"
require "sprockets/railtie"

# Assets should be precompiled for production (so we don't need the gems loaded then)
Bundler.require(*Rails.groups(assets: %w(development test)))


Downloading the example code

You can download the example code files for all Packt books you have purchased from your account at If you purchased this book elsewhere, you can visit and register to have the files e-mailed directly to you.

Notice that there is no require "rails/all". This ensures that the ActiveRecord railtie is not loaded. However, sometimes this causes a conflict with the environment settings. So, in case you face a problem starting the Rails console, remove the following line from development.rb under config/environments (and as required from the other environment files):

# config.active_record.migration_error = :page_load

This should get us going. Now issue the following command to set up Mongoid:

$ bundle exec rails generate mongoid:config

This generates mongoid.yml under config.

Test this basic Rails setup by starting the console.

$ rails c
Loading development environment (Rails 4.0.0)
2.0.0-p247 :001 >

If you see the preceding command prompt, we are set.

Creating models

Now that we have our environment set up, let's create our basic models. In the Author model, we shall now add a field called name, and create a relation between the Author model and the Address model.

$ rails generate model Author
# app/models/author.rb
class Author
  include Mongoid::Document
  include Mongoid::Attributes::Dynamic

  field :name, type: String

  embeds_one :address

Now, let's create the Address model with a number of fields and relations.

$ rails generate model Address
class Address
  include Mongoid::Document

  field :street, type: String
  field :city, type: String
  field :state, type: String
  field :zipcode, type: String
  field :country, type: String

  embedded_in :author

Now, let's test the code that we have written.


Did you notice that we are not using the hash rocket (=>) notation for defining the hash of options for the field method? Instead of :type => String, we are using the JSON notation instead. We shall follow this standard throughout the book.

Testing the models

Now that we have created the models, let's test it out quickly:

irb> a = Author.create(name: "Charles Dickens")
 => #<Author _id: 5143678345db7ca255000001, name: "Charles Dickens">

irb> a.create_address(street: "Picadilly Circus", city: "London", country: "UK")
 => #<Address _id: 514367f445db7ca255000003, street: "Picadilly Circus", city: "London", state: nil, zipcode: nil, country: "UK">

As we can see, this creates an Author object and its corresponding Address object. Mongoid includes ActiveModel and you may notice the similarity in these methods if you have used ActiveRecord.


We have used create_address because an author has only one embedded address. If, an author had multiple addresses, we would have used a.addresses.create.

irb> Author.first
 => #<Author _id: 5143678345db7ca255000001, name: "Charles Dickens">

irb> Author.first.address
 => #<Address _id: 514367f445db7ca255000003, street: "Picadilly Circus", city: "London", state: nil, zipcode: nil, country: "UK">

Here, we have double-checked that the author is indeed persisted to the database. Since this is MongoDB, we can dynamically add attributes to the object!

irb> a['language'] = "English"
 => "English"

 => true

irb> Author.first
 => #<Author _id: 5143678345db7ca255000001, name: "Charles Dickens", language: "English">

Introducing Moped

So, let's see what happened in Mongoid and MongoDB. First, let's see what is in the log file development.log under log.

When we issued the command Author.create(name: "Charles Dickens"), it generated the following output:

MOPED: INSERT       database=sodibee_development collection=authors documents=[{"_id"=>"5143678345db7ca255000001", "name"=>"Charles Dickens"}] flags=[] (0.2460ms)

Now, when we issued the second command a.create_address(street: "Picadilly Circus", city: "London", country: "UK"), it updated the Author object, and created an embedded Address document as seen in the following line:

MOPED: UPDATE       database=sodibee_development collection=authors selector={"_id"=>"5143678345db7ca255000001"} update={"$set"=>{"address"=>{"_id"=>"514367f445db7ca255000003", "street"=>"Picadilly Circus", "city"=>"London", "country"=>"UK"}}} flags=[] (0.1211ms)

Now that we have seen what INSERT and UPDATE look like, querying the Author collection with Author.first generates the following result:

MOPED: QUERY        database=sodibee_development collection=authors selector={"$query"=>{}, "$orderby"=>{:_id=>1}} flags=[:slave_ok] limit=-1 skip=0 batch_size=nil fields=nil (66.7090ms)

And since we want to query the address, we look it up using Author.first.address. This generates the following line:

MOPED: QUERY        database=sodibee_development collection=authors selector={"$query"=>{}, "$orderby"=>{:_id=>1}} flags=[:slave_ok] limit=-1 skip=0 batch_size=nil fields=nil (0.5021ms)

Now there's something interesting about the preceding output—the last two commands on the Author model fired the same query, and look at the difference in the query result! The same query is fired because the address is an embedded document. So, to fetch the address of an author, you fetch the Author object itself. The difference of 66 ms and 0.5 ms in the query response is because for the first lookup MongoDB loads the document from the disk and puts it into its memory-mapped file. The second time, the document is simply looked up in cache (the memory-mapped file) and hence the lookup is faster.

Dynamic attributes

When we issued the command a['language'] = "English", and saved the object using; this is what we see:

  MOPED: UPDATE       database=sodibee_development collection=authors selector={"_id"=>"5143678345db7ca255000001"} update={"$set"=>{"language"=>"English"}} flags=[] (0.1121ms)

This is the result of dynamic attribute update. Even though we did not specify language as a field in the Author model, we can set it as an attribute for the Author object. Did you notice that the update for dynamic attributes is no different from the standard update query in MongoDB?

However, there is a difference when accessing it in Mongoid. The Author.first.language parameter may throw an error sometimes, but Author.first[:language] will always succeed. Let's see an example:

irb> a = Author.create(name: "Gautam")
 => #<Author _id: 515085fd45db7c911e000003, name: "Gautam">

Here we have created a new Author object. However, when we try to update the object using the dot notation a.language, it gives an error. As we can see in the following command lines, method_missing does not dynamically create the accessor method if the dynamic attribute does not already exist.

irb> a.language = "English"
NoMethodError: undefined method `language=' for #<Author _id: 515085fd45db7c911e000003, name: "Gautam">
  from lib/mongoid/attributes.rb:317:in `method_missing'
  from (irb):12
  from lib/rails/commands/console.rb:88:in `start'
  from lib/rails/commands/console.rb:9:in `start'
  from lib/rails/commands.rb:64:in `<top (required)>'
  from bin/rails:4:in `require'
  from bin/rails:4:in `<main>'

Now, if we try to update the dynamic attribute without using the dot notation, it works!

irb> a[:language] = "English"
 => "English"
 => true

Since we have saved it now, when we access the dynamic attribute language again, method_missing creates the accessor method because the dynamic attribute exists. So, now even the dot notation works.

irb> a.language
 => "English"
irb> a[:language]
 => "English"

Introducing Origin

Origin is a gem that provides the DSL for Mongoid queries. Though at first glance, a question may seem to arise as to why we need a DSL for Mongoid queries; If we are finally going to convert the query to a MongoDB-compliant hash, then why do we need a DSL?

Origin was extracted from Mongoid gem and put into a new gem, so that there is a standard for querying. It has no dependency on any other gem and is a standalone pure query builder. The idea was that this could be a generic DSL that can be used even without Mongoid!

So, now we have a very generic and standard querying pattern. For example, in Mongoid 2.x we had the criteria any_in and any_of, and no direct support for the and, or, and nor operations. In Mongoid 2.x, the only way we could fire a $or or a $and query was like this:

Author.where("$or" => {'language' => 'English', '' => 'London

And now in Mongoid 3, we have a cleaner approach.

Author.or(:language => 'English', '' => 'London')

Origin also provides good selectors directly in our models. So, this is now much more readable:

Book.gte(published_at: Date.parse('2012/11/11'))

As we shall see later in the book, Origin has a lot more cool features.