diff options
Diffstat (limited to 'activerecord/lib/active_record')
| -rw-r--r-- | activerecord/lib/active_record/querying.rb | 2 | ||||
| -rw-r--r-- | activerecord/lib/active_record/railtie.rb | 10 | ||||
| -rw-r--r-- | activerecord/lib/active_record/relation.rb | 7 | ||||
| -rw-r--r-- | activerecord/lib/active_record/relation/batches.rb | 98 | ||||
| -rw-r--r-- | activerecord/lib/active_record/relation/batches/batch_enumerator.rb | 67 | ||||
| -rw-r--r-- | activerecord/lib/active_record/serialization.rb | 2 | ||||
| -rw-r--r-- | activerecord/lib/active_record/serializers/xml_serializer.rb | 193 |
7 files changed, 169 insertions, 210 deletions
diff --git a/activerecord/lib/active_record/querying.rb b/activerecord/lib/active_record/querying.rb index 4e597590e9..87a1988f2f 100644 --- a/activerecord/lib/active_record/querying.rb +++ b/activerecord/lib/active_record/querying.rb @@ -6,7 +6,7 @@ module ActiveRecord delegate :find_or_create_by, :find_or_create_by!, :find_or_initialize_by, to: :all delegate :find_by, :find_by!, to: :all delegate :destroy, :destroy_all, :delete, :delete_all, :update, :update_all, to: :all - delegate :find_each, :find_in_batches, to: :all + delegate :find_each, :find_in_batches, :in_batches, to: :all delegate :select, :group, :order, :except, :reorder, :limit, :offset, :joins, :or, :where, :rewhere, :preload, :eager_load, :includes, :from, :lock, :readonly, :having, :create_with, :uniq, :distinct, :references, :none, :unscope, to: :all diff --git a/activerecord/lib/active_record/railtie.rb b/activerecord/lib/active_record/railtie.rb index da6b8447d3..459d6256fa 100644 --- a/activerecord/lib/active_record/railtie.rb +++ b/activerecord/lib/active_record/railtie.rb @@ -16,11 +16,11 @@ module ActiveRecord config.app_generators.orm :active_record, :migration => true, :timestamps => true - config.app_middleware.insert_after "::ActionDispatch::Callbacks", - "ActiveRecord::QueryCache" + config.app_middleware.insert_after ::ActionDispatch::Callbacks, + ActiveRecord::QueryCache - config.app_middleware.insert_after "::ActionDispatch::Callbacks", - "ActiveRecord::ConnectionAdapters::ConnectionManagement" + config.app_middleware.insert_after ::ActionDispatch::Callbacks, + ActiveRecord::ConnectionAdapters::ConnectionManagement config.action_dispatch.rescue_responses.merge!( 'ActiveRecord::RecordNotFound' => :not_found, @@ -78,7 +78,7 @@ module ActiveRecord initializer "active_record.migration_error" do if config.active_record.delete(:migration_error) == :page_load - config.app_middleware.insert_after "::ActionDispatch::Callbacks", + config.app_middleware.insert_after ::ActionDispatch::Callbacks, "ActiveRecord::Migration::CheckPending" end end diff --git a/activerecord/lib/active_record/relation.rb b/activerecord/lib/active_record/relation.rb index 3ed04dee3b..e47b7b1ed9 100644 --- a/activerecord/lib/active_record/relation.rb +++ b/activerecord/lib/active_record/relation.rb @@ -667,6 +667,13 @@ module ActiveRecord "#<#{self.class.name} [#{entries.join(', ')}]>" end + protected + + def load_records(records) + @records = records + @loaded = true + end + private def exec_queries diff --git a/activerecord/lib/active_record/relation/batches.rb b/activerecord/lib/active_record/relation/batches.rb index e07580a563..beb8fa511c 100644 --- a/activerecord/lib/active_record/relation/batches.rb +++ b/activerecord/lib/active_record/relation/batches.rb @@ -1,3 +1,5 @@ +require "active_record/relation/batches/batch_enumerator" + module ActiveRecord module Batches # Looping through a collection of records from the database @@ -122,24 +124,102 @@ module ActiveRecord end end + in_batches(of: batch_size, begin_at: begin_at, end_at: end_at, load: true) do |batch| + yield batch.to_a + end + end + + # Yields ActiveRecord::Relation objects to work with a batch of records. + # + # Person.where("age > 21").in_batches do |relation| + # relation.delete_all + # sleep(10) # Throttle the delete queries + # end + # + # If you do not provide a block to #in_batches, it will return a + # BatchEnumerator which is enumerable. + # + # Person.in_batches.with_index do |relation, batch_index| + # puts "Processing relation ##{batch_index}" + # relation.each { |relation| relation.delete_all } + # end + # + # Examples of calling methods on the returned BatchEnumerator object: + # + # Person.in_batches.delete_all + # Person.in_batches.update_all(awesome: true) + # Person.in_batches.each_record(&:party_all_night!) + # + # ==== Options + # * <tt>:of</tt> - Specifies the size of the batch. Default to 1000. + # * <tt>:load</tt> - Specifies if the relation should be loaded. Default to false. + # * <tt>:begin_at</tt> - Specifies the primary key value to start from, inclusive of the value. + # * <tt>:end_at</tt> - Specifies the primary key value to end at, inclusive of the value. + # + # This is especially useful if you want to work with the + # ActiveRecord::Relation object instead of the array of records, or if + # you want multiple workers dealing with the same processing queue. You can + # make worker 1 handle all the records between id 0 and 10,000 and worker 2 + # handle from 10,000 and beyond (by setting the +:begin_at+ and +:end_at+ + # option on each worker). + # + # # Let's process the next 2000 records + # Person.in_batches(of: 2000, begin_at: 2000).update_all(awesome: true) + # + # An example of calling where query method on the relation: + # + # Person.in_batches.each do |relation| + # relation.update_all('age = age + 1') + # relation.where('age > 21').update_all(should_party: true) + # relation.where('age <= 21').delete_all + # end + # + # NOTE: If you are going to iterate through each record, you should call + # #each_record on the yielded BatchEnumerator: + # + # Person.in_batches.each_record(&:party_all_night!) + # + # NOTE: It's not possible to set the order. That is automatically set to + # ascending on the primary key ("id ASC") to make the batch ordering + # consistent. Therefore the primary key must be orderable, e.g an integer + # or a string. + # + # NOTE: You can't set the limit either, that's used to control the batch + # sizes. + def in_batches(of: 1000, begin_at: nil, end_at: nil, load: false) + relation = self + unless block_given? + return BatchEnumerator.new(of: of, begin_at: begin_at, end_at: end_at, relation: self) + end + if logger && (arel.orders.present? || arel.taken.present?) logger.warn("Scoped order and limit are ignored, it's forced to be batch order and batch size") end - relation = relation.reorder(batch_order).limit(batch_size) + relation = relation.reorder(batch_order).limit(of) relation = apply_limits(relation, begin_at, end_at) - records = relation.to_a + batch_relation = relation + + loop do + if load + records = batch_relation.to_a + ids = records.map(&:id) + yielded_relation = self.where(primary_key => ids) + yielded_relation.load_records(records) + else + ids = batch_relation.pluck(primary_key) + yielded_relation = self.where(primary_key => ids) + end - while records.any? - records_size = records.size - primary_key_offset = records.last.id - raise "Primary key not included in the custom select clause" unless primary_key_offset + break if ids.empty? - yield records + primary_key_offset = ids.last + raise ArgumentError.new("Primary key not included in the custom select clause") unless primary_key_offset - break if records_size < batch_size + yield yielded_relation - records = relation.where(table[primary_key].gt(primary_key_offset)).to_a + break if ids.length < of + batch_relation = relation.where(table[primary_key].gt(primary_key_offset)) end end diff --git a/activerecord/lib/active_record/relation/batches/batch_enumerator.rb b/activerecord/lib/active_record/relation/batches/batch_enumerator.rb new file mode 100644 index 0000000000..153aae9584 --- /dev/null +++ b/activerecord/lib/active_record/relation/batches/batch_enumerator.rb @@ -0,0 +1,67 @@ +module ActiveRecord + module Batches + class BatchEnumerator + include Enumerable + + def initialize(of: 1000, begin_at: nil, end_at: nil, relation:) #:nodoc: + @of = of + @relation = relation + @begin_at = begin_at + @end_at = end_at + end + + # Looping through a collection of records from the database (using the + # +all+ method, for example) is very inefficient since it will try to + # instantiate all the objects at once. + # + # In that case, batch processing methods allow you to work with the + # records in batches, thereby greatly reducing memory consumption. + # + # Person.in_batches.each_record do |person| + # person.do_awesome_stuff + # end + # + # Person.where("age > 21").in_batches(of: 10).each_record do |person| + # person.party_all_night! + # end + # + # If you do not provide a block to #each_record, it will return an Enumerator + # for chaining with other methods: + # + # Person.in_batches.each_record.with_index do |person, index| + # person.award_trophy(index + 1) + # end + def each_record + return to_enum(:each_record) unless block_given? + + @relation.to_enum(:in_batches, of: @of, begin_at: @begin_at, end_at: @end_at, load: true).each do |relation| + relation.to_a.each { |record| yield record } + end + end + + # Delegates #delete_all, #update_all, #destroy_all methods to each batch. + # + # People.in_batches.delete_all + # People.in_batches.destroy_all('age < 10') + # People.in_batches.update_all('age = age + 1') + [:delete_all, :update_all, :destroy_all].each do |method| + define_method(method) do |*args, &block| + @relation.to_enum(:in_batches, of: @of, begin_at: @begin_at, end_at: @end_at, load: false).each do |relation| + relation.send(method, *args, &block) + end + end + end + + # Yields an ActiveRecord::Relation object for each batch of records. + # + # Person.in_batches.each do |relation| + # relation.update_all(awesome: true) + # end + def each + enum = @relation.to_enum(:in_batches, of: @of, begin_at: @begin_at, end_at: @end_at, load: false) + return enum.each { |relation| yield relation } if block_given? + enum + end + end + end +end diff --git a/activerecord/lib/active_record/serialization.rb b/activerecord/lib/active_record/serialization.rb index 48c12dcf9f..23dc6465af 100644 --- a/activerecord/lib/active_record/serialization.rb +++ b/activerecord/lib/active_record/serialization.rb @@ -18,5 +18,3 @@ module ActiveRecord #:nodoc: end end end - -require 'active_record/serializers/xml_serializer' diff --git a/activerecord/lib/active_record/serializers/xml_serializer.rb b/activerecord/lib/active_record/serializers/xml_serializer.rb deleted file mode 100644 index 89b7e0be82..0000000000 --- a/activerecord/lib/active_record/serializers/xml_serializer.rb +++ /dev/null @@ -1,193 +0,0 @@ -require 'active_support/core_ext/hash/conversions' - -module ActiveRecord #:nodoc: - module Serialization - include ActiveModel::Serializers::Xml - - # Builds an XML document to represent the model. Some configuration is - # available through +options+. However more complicated cases should - # override ActiveRecord::Base#to_xml. - # - # By default the generated XML document will include the processing - # instruction and all the object's attributes. For example: - # - # <?xml version="1.0" encoding="UTF-8"?> - # <topic> - # <title>The First Topic</title> - # <author-name>David</author-name> - # <id type="integer">1</id> - # <approved type="boolean">false</approved> - # <replies-count type="integer">0</replies-count> - # <bonus-time type="dateTime">2000-01-01T08:28:00+12:00</bonus-time> - # <written-on type="dateTime">2003-07-16T09:28:00+1200</written-on> - # <content>Have a nice day</content> - # <author-email-address>david@loudthinking.com</author-email-address> - # <parent-id></parent-id> - # <last-read type="date">2004-04-15</last-read> - # </topic> - # - # This behavior can be controlled with <tt>:only</tt>, <tt>:except</tt>, - # <tt>:skip_instruct</tt>, <tt>:skip_types</tt>, <tt>:dasherize</tt> and <tt>:camelize</tt> . - # The <tt>:only</tt> and <tt>:except</tt> options are the same as for the - # +attributes+ method. The default is to dasherize all column names, but you - # can disable this setting <tt>:dasherize</tt> to +false+. Setting <tt>:camelize</tt> - # to +true+ will camelize all column names - this also overrides <tt>:dasherize</tt>. - # To not have the column type included in the XML output set <tt>:skip_types</tt> to +true+. - # - # For instance: - # - # topic.to_xml(skip_instruct: true, except: [ :id, :bonus_time, :written_on, :replies_count ]) - # - # <topic> - # <title>The First Topic</title> - # <author-name>David</author-name> - # <approved type="boolean">false</approved> - # <content>Have a nice day</content> - # <author-email-address>david@loudthinking.com</author-email-address> - # <parent-id></parent-id> - # <last-read type="date">2004-04-15</last-read> - # </topic> - # - # To include first level associations use <tt>:include</tt>: - # - # firm.to_xml include: [ :account, :clients ] - # - # <?xml version="1.0" encoding="UTF-8"?> - # <firm> - # <id type="integer">1</id> - # <rating type="integer">1</rating> - # <name>37signals</name> - # <clients type="array"> - # <client> - # <rating type="integer">1</rating> - # <name>Summit</name> - # </client> - # <client> - # <rating type="integer">1</rating> - # <name>Microsoft</name> - # </client> - # </clients> - # <account> - # <id type="integer">1</id> - # <credit-limit type="integer">50</credit-limit> - # </account> - # </firm> - # - # Additionally, the record being serialized will be passed to a Proc's second - # parameter. This allows for ad hoc additions to the resultant document that - # incorporate the context of the record being serialized. And by leveraging the - # closure created by a Proc, to_xml can be used to add elements that normally fall - # outside of the scope of the model -- for example, generating and appending URLs - # associated with models. - # - # proc = Proc.new { |options, record| options[:builder].tag!('name-reverse', record.name.reverse) } - # firm.to_xml procs: [ proc ] - # - # <firm> - # # ... normal attributes as shown above ... - # <name-reverse>slangis73</name-reverse> - # </firm> - # - # To include deeper levels of associations pass a hash like this: - # - # firm.to_xml include: {account: {}, clients: {include: :address}} - # <?xml version="1.0" encoding="UTF-8"?> - # <firm> - # <id type="integer">1</id> - # <rating type="integer">1</rating> - # <name>37signals</name> - # <clients type="array"> - # <client> - # <rating type="integer">1</rating> - # <name>Summit</name> - # <address> - # ... - # </address> - # </client> - # <client> - # <rating type="integer">1</rating> - # <name>Microsoft</name> - # <address> - # ... - # </address> - # </client> - # </clients> - # <account> - # <id type="integer">1</id> - # <credit-limit type="integer">50</credit-limit> - # </account> - # </firm> - # - # To include any methods on the model being called use <tt>:methods</tt>: - # - # firm.to_xml methods: [ :calculated_earnings, :real_earnings ] - # - # <firm> - # # ... normal attributes as shown above ... - # <calculated-earnings>100000000000000000</calculated-earnings> - # <real-earnings>5</real-earnings> - # </firm> - # - # To call any additional Procs use <tt>:procs</tt>. The Procs are passed a - # modified version of the options hash that was given to +to_xml+: - # - # proc = Proc.new { |options| options[:builder].tag!('abc', 'def') } - # firm.to_xml procs: [ proc ] - # - # <firm> - # # ... normal attributes as shown above ... - # <abc>def</abc> - # </firm> - # - # Alternatively, you can yield the builder object as part of the +to_xml+ call: - # - # firm.to_xml do |xml| - # xml.creator do - # xml.first_name "David" - # xml.last_name "Heinemeier Hansson" - # end - # end - # - # <firm> - # # ... normal attributes as shown above ... - # <creator> - # <first_name>David</first_name> - # <last_name>Heinemeier Hansson</last_name> - # </creator> - # </firm> - # - # As noted above, you may override +to_xml+ in your ActiveRecord::Base - # subclasses to have complete control about what's generated. The general - # form of doing this is: - # - # class IHaveMyOwnXML < ActiveRecord::Base - # def to_xml(options = {}) - # require 'builder' - # options[:indent] ||= 2 - # xml = options[:builder] ||= ::Builder::XmlMarkup.new(indent: options[:indent]) - # xml.instruct! unless options[:skip_instruct] - # xml.level_one do - # xml.tag!(:second_level, 'content') - # end - # end - # end - def to_xml(options = {}, &block) - XmlSerializer.new(self, options).serialize(&block) - end - end - - class XmlSerializer < ActiveModel::Serializers::Xml::Serializer #:nodoc: - class Attribute < ActiveModel::Serializers::Xml::Serializer::Attribute #:nodoc: - def compute_type - klass = @serializable.class - cast_type = klass.type_for_attribute(name) - - type = ActiveSupport::XmlMini::TYPE_NAMES[value.class.name] || cast_type.type - - { :text => :string, - :time => :datetime }[type] || type - end - protected :compute_type - end - end -end |