123456789_123456789_123456789_123456789_123456789_

Module: ActiveRecord::Persistence

Relationships & Source Files
Namespace Children
Modules:
Extension / Inclusion / Inheritance Descendants
Included In:
Super Chains via Extension / Inclusion / Inheritance
Class Chain:
Defined in: activerecord/lib/active_record/persistence.rb

Class Method Summary

::ActiveSupport::Concern - Extended

class_methods

Define class methods from given block.

included

Evaluate given block in context of base class, so that you can write class macros here.

prepended

Evaluate given block in context of base class, so that you can write class macros here.

append_features, prepend_features

Instance Attribute Summary

  • #destroyed? ⇒ Boolean readonly

    Returns true if this object has been destroyed, otherwise returns false.

  • #new_record? ⇒ Boolean readonly

    Returns true if this object hasn’t been saved yet – that is, a record for the object doesn’t exist in the database yet; otherwise, returns false.

  • #persisted? ⇒ Boolean readonly

    Returns true if the record is persisted, i.e.

  • #previously_new_record? ⇒ Boolean readonly

    Returns true if this object was just created – that is, prior to the last update or delete, the object didn’t exist in the database and new_record? would have returned true.

  • #previously_persisted? ⇒ Boolean readonly

    Returns true if this object was previously persisted but now it has been deleted.

Instance Method Summary

Instance Attribute Details

#destroyed?Boolean (readonly)

Returns true if this object has been destroyed, otherwise returns false.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 355

def destroyed?
  @destroyed
end

#new_record?Boolean (readonly)

Returns true if this object hasn’t been saved yet – that is, a record for the object doesn’t exist in the database yet; otherwise, returns false.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 338

def new_record?
  @new_record
end

#persisted?Boolean (readonly)

Returns true if the record is persisted, i.e. it’s not a new record and it was not destroyed, otherwise returns false.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 361

def persisted?
  !(@new_record || @destroyed)
end

#previously_new_record?Boolean (readonly)

Returns true if this object was just created – that is, prior to the last update or delete, the object didn’t exist in the database and new_record? would have returned true.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 345

def previously_new_record?
  @previously_new_record
end

#previously_persisted?Boolean (readonly)

Returns true if this object was previously persisted but now it has been deleted.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 350

def previously_persisted?
  !new_record? && destroyed?
end

Instance Method Details

#_create_record(attribute_names = self.attribute_names) {|_self| ... } (private)

Creates a record with values matching those of the instance attributes and returns its id.

Yields:

  • (_self)

Yield Parameters:

  • _self (Persistence)

    the object that the method was called on

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 920

def _create_record(attribute_names = self.attribute_names)
  attribute_names = attributes_for_create(attribute_names)

  self.class.with_connection do |connection|
    returning_columns = self.class._returning_columns_for_insert(connection)

    returning_values = self.class._insert_record(
      connection,
      attributes_with_values(attribute_names),
      returning_columns
    )

    returning_columns.zip(returning_values).each do |column, value|
      _write_attribute(column, type_for_attribute(column).deserialize(value)) if !_read_attribute(column)
    end if returning_values
  end

  @new_record = false
  @previously_new_record = true

  yield(self) if block_given?

  id
end

#_delete_row (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 870

def _delete_row
  self.class._delete_record(_query_constraints_hash)
end

#_find_record(options) (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 826

def _find_record(options)
  all_queries = options ? options[:all_queries] : nil
  base = self.class.all(all_queries: all_queries).preload(strict_loaded_associations)

  if options && options[:lock]
    base.lock(options[:lock]).find_by!(_in_memory_query_constraints_hash)
  else
    base.find_by!(_in_memory_query_constraints_hash)
  end
end

#_in_memory_query_constraints_hash (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 837

def _in_memory_query_constraints_hash
  if self.class.query_constraints_list.nil?
    { @primary_key => id }
  else
    self.class.query_constraints_list.index_with do |column_name|
      attribute(column_name)
    end
  end
end

#_query_constraints_hash (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 852

def _query_constraints_hash
  if self.class.query_constraints_list.nil?
    { @primary_key => id_in_database }
  else
    self.class.query_constraints_list.index_with do |column_name|
      attribute_in_database(column_name)
    end
  end
end

#_raise_readonly_record_error (private)

Raises:

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 957

def _raise_readonly_record_error
  raise ReadOnlyRecord, "#{self.class} is marked as readonly"
end

#_raise_record_not_destroyed (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 949

def _raise_record_not_destroyed
  @_association_destroy_exception ||= nil
  key = self.class.primary_key
  raise @_association_destroy_exception || RecordNotDestroyed.new("Failed to destroy #{self.class} with #{key}=#{id}", self)
ensure
  @_association_destroy_exception = nil
end

#_raise_record_not_touched_error (private)

Raises:

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 961

def _raise_record_not_touched_error
  raise ActiveRecordError, <<~MSG.squish
    Cannot touch on a new or destroyed record object. Consider using
    persisted?, new_record?, or destroyed? before touching.
  MSG
end

#_touch_row(attribute_names, time) (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 874

def _touch_row(attribute_names, time)
  time ||= current_time_from_proper_timezone

  attribute_names.each do |attr_name|
    _write_attribute(attr_name, time)
  end

  _update_row(attribute_names, "touch")
end

#_update_record(attribute_names = self.attribute_names) {|_self| ... } (private)

Updates the associated record with values matching those of the instance attributes. Returns the number of affected rows.

Yields:

  • (_self)

Yield Parameters:

  • _self (Persistence)

    the object that the method was called on

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 900

def _update_record(attribute_names = self.attribute_names)
  attribute_names = attributes_for_update(attribute_names)

  if attribute_names.empty?
    affected_rows = 0
    @_trigger_update_callback = true
  else
    affected_rows = _update_row(attribute_names)
    @_trigger_update_callback = affected_rows == 1
  end

  @previously_new_record = false

  yield(self) if block_given?

  affected_rows
end

#_update_row(attribute_names, attempted_action = "update") (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 884

def _update_row(attribute_names, attempted_action = "update")
  self.class._update_record(
    attributes_with_values(attribute_names),
    _query_constraints_hash
  )
end

#apply_scoping?(options) ⇒ Boolean (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 847

def apply_scoping?(options)
  !(options && options[:unscoped]) &&
    (self.class.default_scopes?(all_queries: true) || self.class.global_current_scope)
end

#becomes(klass)

Returns an instance of the specified klass with the attributes of the current record. This is mostly useful in relation to single table inheritance (STI) structures where you want a subclass to appear as the superclass. This can be used along with record identification in Action Pack to allow, say, Client < Company to do something like render partial: @client.becomes(Company) to render that instance using the companies/company partial instead of clients/client.

Note: The new instance will share a link to the same attributes as the original class. Therefore the STI column value will still be the same. Any change to the attributes on either instance will affect both instances. This includes any attribute initialization done by the new instance.

If you want to change the STI column as well, use #becomes! instead.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 487

def becomes(klass)
  became = klass.allocate

  became.send(:initialize) do |becoming|
    @attributes.reverse_merge!(becoming.instance_variable_get(:@attributes))
    becoming.instance_variable_set(:@attributes, @attributes)
    becoming.instance_variable_set(:@mutations_from_database, @mutations_from_database ||= nil)
    becoming.instance_variable_set(:@new_record, new_record?)
    becoming.instance_variable_set(:@destroyed, destroyed?)
    becoming.errors.copy!(errors)
  end

  became
end

#becomes!(klass)

Wrapper around #becomes that also changes the instance’s STI column value. This is especially useful if you want to persist the changed class in your database.

Note: The old instance’s STI column value will be changed too, as both objects share the same set of attributes.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 508

def becomes!(klass)
  became = becomes(klass)
  sti_type = nil
  if !klass.descends_from_active_record?
    sti_type = klass.sti_name
  end
  became.public_send("#{klass.inheritance_column}=", sti_type)
  became
end

#create_or_update(&block) (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 891

def create_or_update(**, &block)
  _raise_readonly_record_error if readonly?
  return false if destroyed?
  result = new_record? ? _create_record(&block) : _update_record(&block)
  result != false
end

#decrement(attribute, by = 1)

Initializes attribute to zero if nil and subtracts the value passed as by (default is 1). The decrement is performed directly on the underlying attribute, no setter is invoked. Only makes sense for number-based attributes. Returns self.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 655

def decrement(attribute, by = 1)
  increment(attribute, -by)
end

#decrement!(attribute, by = 1, touch: nil)

Wrapper around #decrement that writes the update to the database. Only attribute is updated; the record itself is not saved. This means that any other modified attributes will still be dirty. Validations and callbacks are skipped. Supports the #touch option from update_counters, see that for more. Returns self.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 665

def decrement!(attribute, by = 1, touch: nil)
  increment!(attribute, -by, touch: touch)
end

#delete

Deletes the record in the database and freezes this instance to reflect that no changes should be made (since they can’t be persisted). Returns the frozen instance.

The row is simply removed with an SQL DELETE statement on the record’s primary key, and no callbacks are executed.

Note that this will also delete records marked as #readonly?.

To enforce the object’s before_destroy and after_destroy callbacks or any :dependent association options, use #destroy.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 439

def delete
  _delete_row if persisted?
  @destroyed = true
  @previously_new_record = false
  freeze
end

#destroy

Deletes the record in the database and freezes this instance to reflect that no changes should be made (since they can’t be persisted).

There’s a series of callbacks associated with #destroy. If the before_destroy callback throws :abort the action is cancelled and #destroy returns false. See Callbacks for further details.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 453

def destroy
  _raise_readonly_record_error if readonly?
  destroy_associations
  @_trigger_destroy_callback ||= persisted? && destroy_row > 0
  @destroyed = true
  @previously_new_record = false
  freeze
end

#destroy!

Deletes the record in the database and freezes this instance to reflect that no changes should be made (since they can’t be persisted).

There’s a series of callbacks associated with #destroy!. If the before_destroy callback throws :abort the action is cancelled and #destroy! raises RecordNotDestroyed. See Callbacks for further details.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 469

def destroy!
  destroy || _raise_record_not_destroyed
end

#destroy_associations (private)

A hook to be overridden by association modules.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 863

def destroy_associations
end

#destroy_row (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 866

def destroy_row
  _delete_row
end

#increment(attribute, by = 1)

Initializes attribute to zero if nil and adds the value passed as by (default is 1). The increment is performed directly on the underlying attribute, no setter is invoked. Only makes sense for number-based attributes. Returns self.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 632

def increment(attribute, by = 1)
  self[attribute] ||= 0
  self[attribute] += by
  self
end

#increment!(attribute, by = 1, touch: nil)

Wrapper around #increment that writes the update to the database. Only attribute is updated; the record itself is not saved. This means that any other modified attributes will still be dirty. Validations and callbacks are skipped. Supports the #touch option from update_counters, see that for more. Returns self.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 644

def increment!(attribute, by = 1, touch: nil)
  increment(attribute, by)
  change = public_send(attribute) - (public_send(:"#{attribute}_in_database") || 0)
  self.class.update_counters(id, attribute => change, touch: touch)
  public_send(:"clear_#{attribute}_change")
  self
end

#init_internals (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 814

def init_internals
  super
  @_trigger_destroy_callback = @_trigger_update_callback = nil
  @previously_new_record = false
end

#reload(options = nil)

Reloads the record from the database.

This method finds the record by its primary key (which could be assigned manually) and modifies the receiver in-place:

 = Account.new
# => #<Account id: nil, email: nil>
.id = 1
.reload
# Account Load (1.2ms)  SELECT "accounts".* FROM "accounts" WHERE "accounts"."id" = $1 LIMIT 1  [["id", 1]]
# => #<Account id: 1, email: 'account@example.com'>

Attributes are reloaded from the database, and caches busted, in particular the associations cache and the QueryCache.

If the record no longer exists in the database RecordNotFound is raised. Otherwise, in addition to the in-place modification the method returns self for convenience.

The optional :lock flag option allows you to lock the reloaded record:

reload(lock: true) # reload with pessimistic locking

Reloading is commonly used in test suites to test something is actually written to the database, or when some action modifies the corresponding row in the database but not the object in memory:

assert .deposit!(25)
assert_equal 25, .credit        # check it is updated in memory
assert_equal 25, .reload.credit # check it is also persisted

Another common use case is optimistic locking handling:

def with_optimistic_retry
  begin
    yield
  rescue ActiveRecord::StaleObjectError
    begin
      # Reload lock_version in particular.
      reload
    rescue ActiveRecord::RecordNotFound
      # If the record is gone there is nothing to do.
    else
      retry
    end
  end
end
[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 742

def reload(options = nil)
  self.class.connection_pool.clear_query_cache

  fresh_object = if apply_scoping?(options)
    _find_record((options || {}).merge(all_queries: true))
  else
    self.class.unscoped { _find_record(options) }
  end

  @association_cache = fresh_object.instance_variable_get(:@association_cache)
  @association_cache.each_value { |association| association.owner = self }
  @attributes = fresh_object.instance_variable_get(:@attributes)
  @new_record = false
  @previously_new_record = false
  self
end

#save(**options)

Saves the model.

If the model is new, a record gets created in the database, otherwise the existing record gets updated.

By default, save always runs validations. If any of them fail the action is cancelled and #save returns false, and the record won’t be saved. However, if you supply validate: false, validations are bypassed altogether. See Validations for more information.

By default, #save also sets the updated_at+/+updated_on attributes to the current time. However, if you supply touch: false, these timestamps will not be updated.

There’s a series of callbacks associated with #save. If any of the before_* callbacks throws :abort the action is cancelled and #save returns false. See Callbacks for further details.

Attributes marked as readonly are silently ignored if the record is being updated.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 390

def save(**options, &block)
  create_or_update(**options, &block)
rescue ActiveRecord::RecordInvalid
  false
end

#save!(**options)

Saves the model.

If the model is new, a record gets created in the database, otherwise the existing record gets updated.

By default, #save! always runs validations. If any of them fail RecordInvalid gets raised, and the record won’t be saved. However, if you supply validate: false, validations are bypassed altogether. See Validations for more information.

By default, #save! also sets the updated_at+/+updated_on attributes to the current time. However, if you supply touch: false, these timestamps will not be updated.

There’s a series of callbacks associated with #save!. If any of the before_* callbacks throws :abort the action is cancelled and #save! raises RecordNotSaved. See Callbacks for further details.

Attributes marked as readonly are silently ignored if the record is being updated.

Unless an error is raised, returns true.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 423

def save!(**options, &block)
  create_or_update(**options, &block) || raise(RecordNotSaved.new("Failed to save the record", self))
end

#strict_loaded_associations (private)

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 820

def strict_loaded_associations
  @association_cache.find_all do |_, assoc|
    assoc.owner.strict_loading? && !assoc.owner.strict_loading_n_plus_one_only?
  end.map(&:first)
end

#toggle(attribute)

Assigns to attribute the boolean opposite of attribute?. So if the predicate returns true the attribute will become false. This method toggles directly the underlying value without calling any setter. Returns self.

Example:

user = User.first
user.banned? # => false
user.toggle(:banned)
user.banned? # => true
[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 681

def toggle(attribute)
  self[attribute] = !public_send("#{attribute}?")
  self
end

#toggle!(attribute)

Wrapper around #toggle that saves the record. This method differs from its non-bang version in the sense that it passes through the attribute setter. Saving is not subjected to validation checks. Returns true if the record could be saved.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 690

def toggle!(attribute)
  toggle(attribute).update_attribute(attribute, self[attribute])
end

#touch(*names, time: nil)

Saves the record with the updated_at/on attributes set to the current time or the time specified. Please note that no validation is performed and only the after_touch, after_commit and after_rollback callbacks are executed.

This method can be passed attribute names and an optional time argument. If attribute names are passed, they are updated along with updated_at/on attributes. If no time argument is passed, the current time is used as default.

product.touch                         # updates updated_at/on with current time
product.touch(time: Time.new(2015, 2, 16, 0, 0, 0)) # updates updated_at/on with specified time
product.touch(:designed_at)           # updates the designed_at attribute and updated_at/on
product.touch(:started_at, :ended_at) # updates started_at, ended_at and updated_at/on attributes

If used along with belongs_to then touch will invoke touch method on associated object.

class Brake < ActiveRecord::Base
  belongs_to :car, touch: true
end

class Car < ActiveRecord::Base
  belongs_to :corporation, touch: true
end

# triggers @brake.car.touch and @brake.car.corporation.touch
@brake.touch

Note that touch must be used on a persisted object, or else an ActiveRecordError will be thrown. For example:

ball = Ball.new
ball.touch(:updated_at)   # => raises ActiveRecordError
[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 793

def touch(*names, time: nil)
  _raise_record_not_touched_error unless persisted?
  _raise_readonly_record_error if readonly?

  attribute_names = timestamp_attributes_for_update_in_model
  attribute_names = (attribute_names | names).map! do |name|
    name = name.to_s
    name = self.class.attribute_aliases[name] || name
    verify_readonly_attribute(name)
    name
  end

  unless attribute_names.empty?
    affected_rows = _touch_row(attribute_names, time)
    @_trigger_update_callback = affected_rows == 1
  else
    true
  end
end

#update(attributes)

Updates the attributes of the model from the passed-in hash and saves the record, all wrapped in a transaction. If the object is invalid, the saving will fail and false will be returned.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 563

def update(attributes)
  # The following transaction covers any possible database side-effects of the
  # attributes assignment. For example, setting the IDs of a child collection.
  with_transaction_returning_status do
    assign_attributes(attributes)
    save
  end
end

#update!(attributes)

Updates its receiver just like #update but calls #save! instead of #save, so an exception is raised if the record is invalid and saving will fail.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 574

def update!(attributes)
  # The following transaction covers any possible database side-effects of the
  # attributes assignment. For example, setting the IDs of a child collection.
  with_transaction_returning_status do
    assign_attributes(attributes)
    save!
  end
end

#update_attribute(name, value)

Updates a single attribute and saves the record. This is especially useful for boolean flags on existing records. Also note that

  • Validation is skipped.

  • Callbacks are invoked.

  • updated_at/updated_on column is updated if that column is available.

  • Updates all the attributes that are dirty in this object.

This method raises an ActiveRecordError if the attribute is marked as readonly.

Also see #update_column.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 530

def update_attribute(name, value)
  name = name.to_s
  verify_readonly_attribute(name)
  public_send("#{name}=", value)

  save(validate: false)
end

#update_attribute!(name, value)

Updates a single attribute and saves the record. This is especially useful for boolean flags on existing records. Also note that

  • Validation is skipped.

  • Callbacks are invoked.

  • updated_at/updated_on column is updated if that column is available.

  • Updates all the attributes that are dirty in this object.

This method raises an ActiveRecordError if the attribute is marked as readonly.

If any of the before_* callbacks throws :abort the action is cancelled and #update_attribute! raises RecordNotSaved. See Callbacks for further details.

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 552

def update_attribute!(name, value)
  name = name.to_s
  verify_readonly_attribute(name)
  public_send("#{name}=", value)

  save!(validate: false)
end

#update_column(name, value)

Equivalent to update_columns(name => value).

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 584

def update_column(name, value)
  update_columns(name => value)
end

#update_columns(attributes)

Updates the attributes directly in the database issuing an UPDATE SQL statement and sets them in the receiver:

user.update_columns(last_request_at: Time.current)

This is the fastest way to update attributes because it goes straight to the database, but take into account that in consequence the regular update procedures are totally bypassed. In particular:

  • Validations are skipped.

  • Callbacks are skipped.

  • updated_at+/+updated_on are not updated.

  • However, attributes are serialized with the same rules as Relation#update_all

This method raises an ActiveRecordError when called on new objects, or when at least one of the attributes is marked as readonly.

Raises:

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 604

def update_columns(attributes)
  raise ActiveRecordError, "cannot update a new record" if new_record?
  raise ActiveRecordError, "cannot update a destroyed record" if destroyed?
  _raise_readonly_record_error if readonly?

  attributes = attributes.transform_keys do |key|
    name = key.to_s
    name = self.class.attribute_aliases[name] || name
    verify_readonly_attribute(name) || name
  end

  update_constraints = _query_constraints_hash
  attributes = attributes.each_with_object({}) do |(k, v), h|
    h[k] = @attributes.write_cast_value(k, v)
    clear_attribute_change(k)
  end

  affected_rows = self.class._update_record(
    attributes,
    update_constraints
  )

  affected_rows == 1
end

#verify_readonly_attribute(name) (private)

Raises:

[ GitHub ]

  
# File 'activerecord/lib/active_record/persistence.rb', line 945

def verify_readonly_attribute(name)
  raise ActiveRecordError, "#{name} is marked as readonly" if self.class.readonly_attribute?(name)
end