module ActiveRecord::Batches

def act_on_ignored_order(error_on_ignore)

def act_on_ignored_order(error_on_ignore)
  raise_error = (error_on_ignore.nil? ? ActiveRecord.error_on_ignored_order : error_on_ignore)
  if raise_error
    raise ArgumentError.new(ORDER_IGNORE_MESSAGE)
  elsif logger
    logger.warn(ORDER_IGNORE_MESSAGE)
  end
end

def apply_finish_limit(relation, finish, order)

def apply_finish_limit(relation, finish, order)
  relation.where(predicate_builder[primary_key, finish, order == :desc ? :gteq : :lteq])
end

def apply_limits(relation, start, finish, order)

def apply_limits(relation, start, finish, order)
  relation = apply_start_limit(relation, start, order) if start
  relation = apply_finish_limit(relation, finish, order) if finish
  relation
end

def apply_start_limit(relation, start, order)

def apply_start_limit(relation, start, order)
  relation.where(predicate_builder[primary_key, start, order == :desc ? :lteq : :gteq])
end

def batch_order(order)

def batch_order(order)
  table[primary_key].public_send(order)
end

def find_each(start: nil, finish: nil, batch_size: 1000, error_on_ignore: nil, order: :asc, &block)

other processes are modifying the database.
NOTE: By its nature, batch processing is subject to race conditions if

orderable (e.g. an integer or string).
This also means that this method only works when the primary key is
ascending on the primary key ("id ASC").
NOTE: Order can be ascending (:asc) or descending (:desc). It is automatically set to

end
person.party_all_night!
Person.find_each(start: 10_000) do |person|
# In worker 2, let's process from record 10_000 and onwards.

end
person.party_all_night!
Person.find_each(finish: 9_999) do |person|
# In worker 1, let's process until 9999 records.

option on each worker.
handle from 10000 and beyond by setting the +:start+ and +:finish+
worker 1 handle all the records between id 1 and 9999 and worker 2
multiple workers dealing with the same processing queue. You can make
The options +start+ and +finish+ are especially useful if you want

size: it can be less than, equal to, or greater than the limit.
Limits are honored, and if present there is no requirement for the batch

* :order - Specifies the primary key order (can be +:asc+ or +:desc+). Defaults to +:asc+.
an order is present in the relation.
* :error_on_ignore - Overrides the application config to specify if an error should be raised when
* :finish - Specifies the primary key value to end at, inclusive of the value.
* :start - Specifies the primary key value to start from, inclusive of the value.
* :batch_size - Specifies the size of the batch. Defaults to 1000.
==== Options

end
person.award_trophy(index + 1)
Person.find_each.with_index do |person, index|

for chaining with other methods:
If you do not provide a block to #find_each, it will return an Enumerator

end
person.party_all_night!
Person.where("age > 21").find_each do |person|

end
person.do_awesome_stuff
Person.find_each do |person|

specified by the +:batch_size+ option).
The #find_each method uses #find_in_batches with a batch size of 1000 (or as

with the records in batches, thereby greatly reducing memory consumption.
In that case, batch processing methods allow you to work

is very inefficient since it will try to instantiate all the objects at once.
(using the Scoping::Named::ClassMethods.all method, for example)
Looping through a collection of records from the database
def find_each(start: nil, finish: nil, batch_size: 1000, error_on_ignore: nil, order: :asc, &block)
  if block_given?
    find_in_batches(start: start, finish: finish, batch_size: batch_size, error_on_ignore: error_on_ignore, order: order) do |records|
      records.each(&block)
    end
  else
    enum_for(:find_each, start: start, finish: finish, batch_size: batch_size, error_on_ignore: error_on_ignore, order: order) do
      relation = self
      apply_limits(relation, start, finish, order).size
    end
  end
end

def find_in_batches(start: nil, finish: nil, batch_size: 1000, error_on_ignore: nil, order: :asc)

other processes are modifying the database.
NOTE: By its nature, batch processing is subject to race conditions if

orderable (e.g. an integer or string).
This also means that this method only works when the primary key is
ascending on the primary key ("id ASC").
NOTE: Order can be ascending (:asc) or descending (:desc). It is automatically set to

end
group.each { |person| person.party_all_night! }
Person.find_in_batches(start: 10_000) do |group|
# Let's process from record 10_000 on.

option on each worker.
handle from 10000 and beyond by setting the +:start+ and +:finish+
worker 1 handle all the records between id 1 and 9999 and worker 2
multiple workers dealing with the same processing queue. You can make
The options +start+ and +finish+ are especially useful if you want

size: it can be less than, equal to, or greater than the limit.
Limits are honored, and if present there is no requirement for the batch

* :order - Specifies the primary key order (can be +:asc+ or +:desc+). Defaults to +:asc+.
an order is present in the relation.
* :error_on_ignore - Overrides the application config to specify if an error should be raised when
* :finish - Specifies the primary key value to end at, inclusive of the value.
* :start - Specifies the primary key value to start from, inclusive of the value.
* :batch_size - Specifies the size of the batch. Defaults to 1000.
==== Options

To be yielded each record one by one, use #find_each instead.

end
group.each(&:recover_from_last_night!)
puts "Processing group ##{batch}"
Person.find_in_batches.with_index do |group, batch|

for chaining with other methods:
If you do not provide a block to #find_in_batches, it will return an Enumerator

end
group.each { |person| person.party_all_night! }
sleep(50) # Make sure it doesn't get too crowded in there!
Person.where("age > 21").find_in_batches do |group|

an array.
Yields each batch of records that was found by the find options as
def find_in_batches(start: nil, finish: nil, batch_size: 1000, error_on_ignore: nil, order: :asc)
  relation = self
  unless block_given?
    return to_enum(:find_in_batches, start: start, finish: finish, batch_size: batch_size, error_on_ignore: error_on_ignore, order: order) do
      total = apply_limits(relation, start, finish, order).size
      (total - 1).div(batch_size) + 1
    end
  end
  in_batches(of: batch_size, start: start, finish: finish, load: true, error_on_ignore: error_on_ignore, order: order) do |batch|
    yield batch.to_a
  end
end

def in_batches(of: 1000, start: nil, finish: nil, load: false, error_on_ignore: nil, order: :asc)

other processes are modifying the database.
NOTE: By its nature, batch processing is subject to race conditions if

orderable (e.g. an integer or string).
This also means that this method only works when the primary key is
ascending on the primary key ("id ASC").
NOTE: Order can be ascending (:asc) or descending (:desc). It is automatically set to

Person.in_batches.each_record(&:party_all_night!)

#each_record on the yielded BatchEnumerator:
NOTE: If you are going to iterate through each record, you should call

end
relation.where('age <= 21').delete_all
relation.where('age > 21').update_all(should_party: true)
relation.update_all('age = age + 1')
Person.in_batches.each do |relation|

An example of calling where query method on the relation:

Person.in_batches(start: 10_000).update_all(awesome: true)
# Let's process from record 10_000 on.

option on each worker.
handle from 10000 and beyond by setting the +:start+ and +:finish+
worker 1 handle all the records between id 1 and 9999 and worker 2
multiple workers dealing with the same processing queue. You can make
The options +start+ and +finish+ are especially useful if you want

size, it can be less than, equal, or greater than the limit.
Limits are honored, and if present there is no requirement for the batch

* :order - Specifies the primary key order (can be +:asc+ or +:desc+). Defaults to +:asc+.
an order is present in the relation.
* :error_on_ignore - Overrides the application config to specify if an error should be raised when
* :finish - Specifies the primary key value to end at, inclusive of the value.
* :start - Specifies the primary key value to start from, inclusive of the value.
* :load - Specifies if the relation should be loaded. Defaults to false.
* :of - Specifies the size of the batch. Defaults to 1000.
==== Options

Person.in_batches.each_record(&:party_all_night!)
Person.in_batches.update_all(awesome: true)
Person.in_batches.delete_all

Examples of calling methods on the returned BatchEnumerator object:

end
relation.delete_all
puts "Processing relation ##{batch_index}"
Person.in_batches.each_with_index do |relation, batch_index|

BatchEnumerator which is enumerable.
If you do not provide a block to #in_batches, it will return a

end
sleep(10) # Throttle the delete queries
relation.delete_all
Person.where("age > 21").in_batches do |relation|

Yields ActiveRecord::Relation objects to work with a batch of records.
def in_batches(of: 1000, start: nil, finish: nil, load: false, error_on_ignore: nil, order: :asc)
  relation = self
  unless block_given?
    return BatchEnumerator.new(of: of, start: start, finish: finish, relation: self)
  end
  unless [:asc, :desc].include?(order)
    raise ArgumentError, ":order must be :asc or :desc, got #{order.inspect}"
  end
  if arel.orders.present?
    act_on_ignored_order(error_on_ignore)
  end
  batch_limit = of
  if limit_value
    remaining   = limit_value
    batch_limit = remaining if remaining < batch_limit
  end
  relation = relation.reorder(batch_order(order)).limit(batch_limit)
  relation = apply_limits(relation, start, finish, order)
  relation.skip_query_cache! # Retaining the results in the query cache would undermine the point of batching
  batch_relation = relation
  loop do
    if load
      records = batch_relation.records
      ids = records.map(&:id)
      yielded_relation = where(primary_key => ids)
      yielded_relation.load_records(records)
    else
      ids = batch_relation.pluck(primary_key)
      yielded_relation = where(primary_key => ids)
    end
    break if ids.empty?
    primary_key_offset = ids.last
    raise ArgumentError.new("Primary key not included in the custom select clause") unless primary_key_offset
    yield yielded_relation
    break if ids.length < batch_limit
    if limit_value
      remaining -= ids.length
      if remaining == 0
        # Saves a useless iteration when the limit is a multiple of the
        # batch size.
        break
      elsif remaining < batch_limit
        relation = relation.limit(remaining)
      end
    end
    batch_relation = relation.where(
      predicate_builder[primary_key, primary_key_offset, order == :desc ? :lt : :gt]
    )
  end
end