CRUD Operations#

Saving Documents#

Mongoid supports all expected CRUD operations for those familiar with other Ruby mappers like Active Record or Data Mapper. What distinguishes Mongoid from other mappers for MongoDB is that the general persistence operations perform atomic updates on only the fields that have changed instead of writing the entire document to the database each time.

The persistence sections will provide examples on what database operation is performed when executing the documented command.

Standard#

Mongoid’s standard persistence methods come in the form of common methods you would find in other mapping frameworks. The following table shows all standard operations with examples.

Operation

Example

Model.create!

Insert a document or multiple documents into the database, raising an error if a validation or server error occurs.

Pass a hash of attributes to create one document with the specified attributes, or an array of hashes to create multiple documents. If a single hash is passed, the corresponding document is returned. If an array of hashes is passed, an array of documents corresponding to the hashes is returned.

If a block is given to create! , it will be invoked with each document as the argument in turn prior to attempting to save that document.

If there is a problem saving any of the documents, such as a validation error or a server error, an exception is raised and, consequently, none of the documents are returned. However, if an array of hashes was passed and previous documents were successfully saved, those documents will remain in the database.

Person.create!(
  first_name: "Heinrich",
  last_name: "Heine"
) # => Person instance

Person.create!([
  { first_name: "Heinrich", last_name: "Heine" },
  { first_name: "Willy", last_name: "Brandt" }
]) # => Array of two Person instances

Person.create!(first_name: "Heinrich") do |doc|
  doc.last_name = "Heine"
end # => Person instance

Model.create

Instantiate a document or multiple documents and, if validations pass, insert them into the database.

create is similar to create! but does not raise exceptions on validation errors. It still raises errors on server errors, such as trying to insert a document with an _id that already exists in the collection.

If any validation errors are encountered, the respective document is not inserted but is returned along with documents that were inserted. Use persisted? , new_record? or errors methods to check which of the returned documents were inserted into the database.

Person.create(
  first_name: "Heinrich",
  last_name: "Heine"
) # => Person instance

Person.create([
  { first_name: "Heinrich", last_name: "Heine" },
  { first_name: "Willy", last_name: "Brandt" }
]) # => Array of two Person instances

Person.create(first_name: "Heinrich") do |doc|
  doc.last_name = "Heine"
end # => Person instance

class Post
  include Mongoid::Document

  validates_uniqueness_of :title
end

posts = Post.create([{title: "test"}, {title: "test"}])
# => array of two Post instances
posts.map { |post| post.persisted? } # => [true, false]

Model#save!

Save the changed attributes to the database atomically, or insert the document if new. Raises an exception if validations fail or there is a server error.

Returns true if the changed attributes were saved, raises an exception otherwise.

person = Person.new(
  first_name: "Heinrich",
  last_name: "Heine"
)
person.save!

person.first_name = "Christian Johan"
person.save!

Model#save

Save the changed attributes to the database atomically, or insert the document if new.

Returns true if the changed attributes were saved. Returns false if there were any validation errors. Raises an exception if the document passed validation but there was a server error during the save.

Pass validate: false option to bypass validations.

Pass touch: false option to ignore the updates to the updated_at field. If the document being save has not been previously persisted, this option is ignored and the created_at and updated_at fields will be updated with the current time.

person = Person.new(
  first_name: "Heinrich",
  last_name: "Heine"
)
person.save
person.save(validate: false)
person.save(touch: false)

person.first_name = "Christian Johan"
person.save

Model#update_attributes

Update the document attributes in the database. Will return true if validation passed, false if not.

person.update_attributes(
  first_name: "Jean",
  last_name: "Zorg"
)

Model#update_attributes!

Update the document attributes in the database and raise an error if validation failed.

person.update_attributes!(
  first_name: "Leo",
  last_name: "Tolstoy"
)

Model#update_attribute

Update a single attribute, bypassing validations.

person.update_attribute(:first_name, "Jean")

Model#upsert

Performs a MongoDB replace with upsert on the document. If the document exists in the database, it will get overwritten with the current document in the application (any attributes present in the database but not in the application’s document instance will be lost). If the document does not exist in the database, it will be inserted. Note that this only runs the {before|after|around}_upsert callbacks.

person = Person.new(
  first_name: "Heinrich",
  last_name: "Heine"
)
person.upsert

Model#touch

Update the document’s updated_at timestamp, optionally with one extra provided time field. This will cascade the touch to all belongs_to associations of the document with the option set. This operation skips validations and callbacks.

Attempting to touch a destroyed document will raise FrozenError, same as if attempting to update an attribute on a destroyed document.

person.touch
person.touch(:audited_at)

Model#delete

Deletes the document from the database without running callbacks.

If the document is not persisted, Mongoid will attempt to delete from the database any document with the same _id.

person.delete

person = Person.create!(...)
unsaved_person = Person.new(id: person.id)
unsaved_person.delete
person.reload
# raises Mongoid::Errors::DocumentNotFound because the person was deleted

Model#destroy

Deletes the document from the database while running destroy callbacks.

If the document is not persisted, Mongoid will attempt to delete from the database any document with the same _id.

person.destroy

person = Person.create!(...)
unsaved_person = Person.new(id: person.id)
unsaved_person.destroy
person.reload
# raises Mongoid::Errors::DocumentNotFound because the person was deleted

Model.delete_all

Deletes all documents from the database without running any callbacks.

Person.delete_all

Model.destroy_all

Deletes all documents from the database while running callbacks. This is a potentially expensive operation since all documents will be loaded into memory.

Person.destroy_all

Mongoid provides the following persistence-related attributes:

Attribute

Example

Model#new_record?

Returns true if the model instance has not yet been saved to the database. Opposite of persisted?

person = Person.new(
  first_name: "Heinrich",
  last_name: "Heine"
)
person.new_record? # => true
person.save!
person.new_record? # => false

Model#persisted?

Returns true if the model instance has been saved to the database. Opposite of new_record?

person = Person.new(
  first_name: "Heinrich",
  last_name: "Heine"
)
person.persisted? # => false
person.save!
person.persisted? # => true

Atomic#

Mongoid exposes :manual:`MongoDB update operators </reference/operator/update/>` as methods on Mongoid documents. When these methods are used, callbacks are not invoked and validations are not performed. The supported update operators are:

Operation

Example

Model#add_to_set

Performs an atomic $addToSet on the field.

person.add_to_set(aliases: "Bond")

Model#bit

Performs an atomic $bit on the field.

person.bit(age: { and: 10, or: 12 })

Model#inc

Performs an atomic $inc on the field.

person.inc(age: 1)

Model#pop

Performs an atomic $pop on the field.

person.pop(aliases: 1)

Model#pull

Performs an atomic $pull on the field.

person.pull(aliases: "Bond")

Model#pull_all

Performs an atomic $pullAll on the field.

person.pull_all(aliases: [ "Bond", "James" ])

Model#push

Performs an atomic $push on the field.

person.push(aliases: ["007","008"])

Model#rename

Performs an atomic $rename on the field.

person.rename(bday: :dob)

Model#set

Updates an attribute on the model instance and, if the instance is already persisted, performs an atomic $set on the field, bypassing validations.

set can also deeply set values on Hash fields.

set can also deeply set values on embeds_one associations. If such an association’s document is nil, one will be created prior to the update.

set should not be used with has_one associations, as it does not correctly work in such cases.

person = Person.create!(name: "Ricky Bobby")
person.set(name: "Tyler Durden") # updates name in the database


person = Person.new
person.set(name: "Tyler Durden") # does not write to database
person.name # => "Tyler Durden"
person.persisted? # => true


class Post
  include Mongoid::Document

  field :metadata, type: Hash
end

post = Post.create!
post.set('metadata.published_at' => Time.now)
post.metadata['published_at'] # => Time instance

post.set('metadata.approved.today' => true)
post.metadata['approved'] # => {'today' => true}


class Flight
  include Mongoid::Document

  embeds_one :plan
end

class Plan
  include Mongoid::Document

  embedded_in :flight

  field :route, type: String
end

flight = Flight.create!
flight.plan # => nil
flight.set('plan.route', 'test route')
flight.plan # => Plan instance
flight.plan.route # => "test route"

Model#unset

Performs an atomic $unset on the field.

person.unset(:name)

Note that, because these methods skip validations, it is possible to both save invalid documents into the database and end up with invalid documents in the application (which would subsequently fail to save via a save call due to the failing validations).

Atomic Operation Grouping#

Atomic operations may be grouped together using the #atomically method on a document. All operations inside the block given to #atomically are sent to the cluster in a single atomic command. For example:

person.atomically do
  person.inc(age: 1)
  person.set(name: 'Jake')
end

#atomically blocks may be nested. The default behavior is to write changes performed by each block as soon as the block ends:

person.atomically do
  person.atomically do
    person.inc(age: 1)
    person.set(name: 'Jake')
  end
  raise 'An exception'
  # name and age changes are still persisted
end

This behavior can be changed by specifying the join_context: true option to #atomically, or globally by setting the join_contexts configuration option to true. When context joining is enabled, nested #atomically blocks are joined with the outer blocks, and only the outermost block (or the first block where join_contexts is false) actually writes changes to the cluster. For example:

person.atomically do
  person.atomically(join_context: true) do
    person.inc(age: 1)
    person.set(name: 'Jake')
  end
  raise 'An exception'
  # name and age changes are not persisted
end

The context joining behavior can be enabled globally by default by setting join_context option in Mongoid configuration. In this case specifying join_context: false on an #atomically block can be used to obtain the independent persistence context behavior.

If an exception is raised in an #atomically block which has not yet persisted its changes to the cluster, any pending attribute changes on Mongoid models are reverted. For example:

person = Person.new(name: 'Tom')
begin
  person.atomically do
    person.inc(age: 1)
    person.set(name: 'Jake')
    person.name # => 'Jake'
    raise 'An exception'
  end
rescue Exception
  person.name # => 'Tom'
end

Atomic operations described in this section apply to one document at a time, therefore nesting #atomically blocks invoked on multiple documents does not make changes to the different documents be persisted atomically together. However, MongoDB offers multi-document transactions as of server version 4.0 which provide atomic persistence across multiple documents.

Reloading#

Use the reload method to fetch the most recent version of a document from the database. Any unsaved modifications to the document’s attributes are lost:

band = Band.create!(name: 'foo')
# => #<Band _id: 6206d06de1b8324561f179c9, name: "foo", description: nil, likes: nil>

band.name = 'bar'
band
# => #<Band _id: 6206d06de1b8324561f179c9, name: "bar", description: nil, likes: nil>

band.reload
# => #<Band _id: 6206d06de1b8324561f179c9, name: "foo", description: nil, likes: nil>

When a document is reloaded, all of its embedded associations are also reloaded in the same query (since embedded documents are stored in the parent document on the server). If a document has referenced associations, the loaded associations’ are not reloaded but their values are cleared, such that these associations would be loaded from the database at the next access.

Note

Some operations on associations, for example assignment, persists the new document. In these cases there may not be any unsaved modifications to revert by reloading. In the following example, the assignment of the empty array to the association is immediately persisted and reloading does not make any changes to the document:

# Assuming band has many tours, which could be referenced:
band = Band.create!(tours: [Tour.create!])
# ... or embedded:
band = Band.create!(tours: [Tour.new])

# This writes the empty tour list into the database.
band.tours = []

# There are no unsaved modifications in band at this point to be reverted.
band.reload

# Returns the empty array since this is what is in the database.
band.tours
# => []

If the model has a shard key defined, the shard key value is included in the reloading query.

If the database does not contain a matching document, Mongoid normally raises Mongoid::Errors::DocumentNotFound. However, if the configuration option raise_not_found_error is set to false, and the database does not contain a matching document, Mongoid replaces the current document with a newly created document whose attributes are set to default values. Importantly, this generally causes the _id of the document to change, as the following example demonstrates:

band = Band.create!
# => #<Band _id: 6206d00de1b8324561f179c7, name: "foo", description: nil, likes: nil>

Mongoid.raise_not_found_error = false
band.destroy

band.reload
# => #<Band _id: 6206d031e1b8324561f179c8, name: nil, description: nil, likes: nil>

For this reason, it is not recommended to use reload when raise_not_found_error is set to false.

Reloading Unsaved Documents#

reload can be called when the document has not yet been persisted. In this case reload performs a find query using the id value specified in the document (and the shard key value, if a shard key is defined):

existing = Band.create!(name: 'Photek')

# Unsaved document
band = Band.new(id: existing.id)
band.reload
band.name
# => "Photek"

Accessing Field Values#

Mongoid provides several ways of accessing field values.

Note

All of the access methods described below raise Mongoid::Errors::AttributeNotLoaded when the field being accessed is projected out, either by virtue of not being included in only or by virtue of being included in without. This applies to both reads and writes.

Getters & Setters#

The recommended way is to use the getter and setter methods generated for each declared field:

class Person
  include Mongoid::Document

  field :first_name
end

person = Person.new

person.first_name = "Artem"
person.first_name
# => "Artem"

To use this mechanism, each field must be explicitly declared, or the model class must enable dynamic fields.

Custom Getters & Setters#

It is possible to explicitly define the getter and setter methods to provide custom behavior when reading or writing fields, for example value transformations or storing values under different field names. In this case read_attribute and write_attribute methods can be used to read and write the values directly into the attributes hash:

class Person
  include Mongoid::Document

  def first_name
    read_attribute(:fn)
  end

  def first_name=(value)
    write_attribute(:fn, value)
  end
end

person = Person.new

person.first_name = "Artem"
person.first_name
# => "Artem"

person.attributes
# => {"_id"=>BSON::ObjectId('606477dc2c97a628cf47075b'), "fn"=>"Artem"}

read_attribute & write_attribute#

The read_attribute and write_attribute methods can be used explicitly as well. Note that if a field specifies its storage field name, both read_attribute and write_attribute accept either the declared field name or the storage field name for operations:

class Person
  include Mongoid::Document

  field :first_name, as: :fn
  field :last_name, as: :ln
end

person = Person.new(first_name: "Artem")
# => #<Person _id: 60647a522c97a6292c195b4b, first_name(fn): "Artem", last_name(ln): nil>

person.read_attribute(:first_name)
# => "Artem"

person.read_attribute(:fn)
# => "Artem"

person.write_attribute(:last_name, "Pushkin")
person
# => #<Person _id: 60647a522c97a6292c195b4b, first_name(fn): "Artem", last_name(ln): "Pushkin">

person.write_attribute(:ln, "Medvedev")
person
# => #<Person _id: 60647a522c97a6292c195b4b, first_name(fn): "Artem", last_name(ln): "Medvedev">

read_attribute and write_attribute do not require that a field with the used name is defined, but writing field values with write_attribute does not cause the respective field to be defined either:

person.write_attribute(:undefined, "Hello")
person
# => #<Person _id: 60647b212c97a6292c195b4c, first_name(fn): "Artem", last_name(ln): "Medvedev">
person.attributes
# => {"_id"=>BSON::ObjectId('60647b212c97a6292c195b4c'), "first_name"=>"Artem", "last_name"=>"Medvedev", "undefined"=>"Hello"}

person.read_attribute(:undefined)
# => "Hello"
person.undefined
# raises NoMethodError

When read_attribute is used to access a missing field, it returns nil.

Hash Access#

Mongoid model instances define the [] and []= methods to provide Hash style access to the attributes. [] is an alias for read_attribute and []= is an alias for write_attribute; see the section on read_attribute and write_attribute for the detailed description of their behavior.

class Person
  include Mongoid::Document

  field :first_name, as: :fn
  field :last_name, as: :ln
end

person = Person.new(first_name: "Artem")

person["fn"]
# => "Artem"

person[:first_name]
# => "Artem"

person[:ln] = "Medvedev"
person
# => #<Person _id: 606483742c97a629bdde5cfc, first_name(fn): "Artem", last_name(ln): "Medvedev">

person["last_name"] = "Pushkin"
person
# => #<Person _id: 606483742c97a629bdde5cfc, first_name(fn): "Artem", last_name(ln): "Pushkin">

Bulk Attribute Writes#

In cases where you want to set multiple field values at once, there are a few different ways of accomplishing this as well.

# Get the field values as a hash.
person.attributes

# Set the field values in the document.
Person.new(first_name: "Jean-Baptiste", middle_name: "Emmanuel")
person.attributes = { first_name: "Jean-Baptiste", middle_name: "Emmanuel" }
person.write_attributes(
  first_name: "Jean-Baptiste",
  middle_name: "Emmanuel",
)

Dirty Tracking#

Mongoid supports tracking of changed or “dirty” fields with an API that mirrors that of Active Model. If a defined field has been modified in a model the model will be marked as dirty and some additional behavior comes into play.

Viewing Changes#

There are various ways to view what has been altered on a model. Changes are recorded from the time a document is instantiated, either as a new document or via loading from the database up to the time it is saved. Any persistence operation clears the changes.

class Person
  include Mongoid::Document
  field :name, type: String
end

person = Person.first
person.name = "Alan Garner"

# Check to see if the document has changed.
person.changed? # true

# Get an array of the names of the changed fields.
person.changed # [ :name ]

# Get a hash of the old and changed values for each field.
person.changes # { "name" => [ "Alan Parsons", "Alan Garner" ] }

# Check if a specific field has changed.
person.name_changed? # true

# Get the changes for a specific field.
person.name_change # [ "Alan Parsons", "Alan Garner" ]

# Get the previous value for a field.
person.name_was # "Alan Parsons"

Note

Setting the associations on a document does not cause the changes or changed_attributes hashes to be modified. This is true for all associations whether referenced or embedded. Note that changing the _id(s) field on referenced associations does cause the changes to show up in the changes and the changed_attributes hashes.

Resetting Changes#

You can reset changes of a field to its previous value by calling the reset method.

person = Person.first
person.name = "Alan Garner"

# Reset the changed name back to the original
person.reset_name!
person.name # "Alan Parsons"

Persistence#

Mongoid uses dirty tracking as the core of its persistence operations. It looks at the changes on a document and atomically updates only what has changed, unlike other frameworks that write the entire document on each save. If no changes have been made, Mongoid will not hit the database on a call to Model#save.

Viewing Previous Changes#

After a document has been persisted, you can see what the changes were previously by calling Model#previous_changes.

person = Person.first
person.name = "Alan Garner"
person.save # Clears out current changes.

# View the previous changes.
person.previous_changes # { "name" => [ "Alan Parsons", "Alan Garner" ] }

Updating Container Fields#

Be aware that, until MONGOID-2951 is resolved, all fields including container ones must be assigned to for their values to be persisted to the database.

For example, adding to a set like this does not work:

class Band
  include Mongoid::Document

  field :tours, type: Set
end

band = Band.new
band.tours
# => #<Set: {}>

band.tours << 'London'
# => #<Set: {"London"}>
band.tours
# => #<Set: {}>

Instead, the field value must be modified outside of the model and assigned back to the model as follows:

class Band
  include Mongoid::Document

  field :tours, type: Set
end

band = Band.new

tours = band.tours
# => #<Set: {}>

tours << 'London'
# => #<Set: {"London"}>

band.tours = tours
# => #<Set: {"London"}>

band.tours
# => #<Set: {"London"}>