find_each 風味

find_each.rb

class Batches
  BATCH_SIZE = 2
  
  def initialize
    @data = (1..10).to_a
  end
  
  def find_each
    if block_given?
      find_in_batches do |records|
        records.each { |record| yield record }
      end
    else
      enum_for(:find_each)
    end
  end
  
  def find_in_batches
    unless block_given?
      return to_enum(:find_in_batches)
    end

    @start = 0
    loop do
      batch = @data[@start, BATCH_SIZE]
      yield batch
      @start += BATCH_SIZE
      break if batch.size < BATCH_SIZE
    end
  end
end

# 直接イテレート
Batches.new.find_each do |record|
  puts record
end

# Enumerator オブジェクト受け取ってからイテレート
result = Batches.new.find_each
result.each do |record|
  puts record
end

repository.rb

# Elasticsearch のページングを find_each 風に扱えるようにする

class CustomerRepository
  include Elasticsearch::Persistence::Repository
  include Elasticsearch::Persistence::Repository::DSL
  
  def find_each(query:)
    find_in_batches(query: query) do |records|
      records.each { |record| yield record }
    end
  end

  def find_in_batches(query:)
    search_after = {}

    loop do
      batch = search({
                       query: query,
                       size: BATCH_SIZE,
                       sort: [
                         { email: 'asc' }
                       ]
                     }.merge(search_after))

      yield batch

      break if batch.size < BATCH_SIZE

      search_after = { search_after: [batch.last.email] }
    end
  end
end

query = { bool: { must: { match: { foo: 'bar' } } } }
CustomerRepository.new.find_each(query: query) do |record|
  puts record
end