omg-activerecord 8.0.0.alpha1

data/lib/active_record/insert_all.rb

@@ -0,0 +1,328 @@
+# frozen_string_literal: true
+
+require "active_support/core_ext/enumerable"
+
+module ActiveRecord
+  class InsertAll # :nodoc:
+    attr_reader :model, :connection, :inserts, :keys
+    attr_reader :on_duplicate, :update_only, :returning, :unique_by, :update_sql
+
+    class << self
+      def execute(relation, ...)
+        relation.model.with_connection do |c|
+          new(relation, c, ...).execute
+        end
+      end
+    end
+
+    def initialize(relation, connection, inserts, on_duplicate:, update_only: nil, returning: nil, unique_by: nil, record_timestamps: nil)
+      @relation = relation
+      @model, @connection, @inserts = relation.model, connection, inserts.map(&:stringify_keys)
+      @on_duplicate, @update_only, @returning, @unique_by = on_duplicate, update_only, returning, unique_by
+      @record_timestamps = record_timestamps.nil? ? model.record_timestamps : record_timestamps
+
+      disallow_raw_sql!(on_duplicate)
+      disallow_raw_sql!(returning)
+
+      if @inserts.empty?
+        @keys = []
+      else
+        resolve_sti
+        resolve_attribute_aliases
+        @keys = @inserts.first.keys
+      end
+
+      @scope_attributes = relation.scope_for_create.except(@model.inheritance_column)
+      @keys |= @scope_attributes.keys
+      @keys = @keys.to_set
+
+      @returning = (connection.supports_insert_returning? ? primary_keys : false) if @returning.nil?
+      @returning = false if @returning == []
+
+      @unique_by = find_unique_index_for(@unique_by)
+
+      configure_on_duplicate_update_logic
+      ensure_valid_options_for_connection!
+    end
+
+    def execute
+      return ActiveRecord::Result.empty if inserts.empty?
+
+      message = +"#{model} "
+      message << "Bulk " if inserts.many?
+      message << (on_duplicate == :update ? "Upsert" : "Insert")
+      connection.exec_insert_all to_sql, message
+    end
+
+    def updatable_columns
+      @updatable_columns ||= keys - readonly_columns - unique_by_columns
+    end
+
+    def primary_keys
+      Array(@model.schema_cache.primary_keys(model.table_name))
+    end
+
+    def skip_duplicates?
+      on_duplicate == :skip
+    end
+
+    def update_duplicates?
+      on_duplicate == :update
+    end
+
+    def map_key_with_value
+      inserts.map do |attributes|
+        attributes = attributes.stringify_keys
+        attributes.merge!(@scope_attributes)
+        attributes.reverse_merge!(timestamps_for_create) if record_timestamps?
+
+        verify_attributes(attributes)
+
+        keys_including_timestamps.map do |key|
+          yield key, attributes[key]
+        end
+      end
+    end
+
+    def record_timestamps?
+      @record_timestamps
+    end
+
+    # TODO: Consider renaming this method, as it only conditionally extends keys, not always
+    def keys_including_timestamps
+      @keys_including_timestamps ||= if record_timestamps?
+        keys + model.all_timestamp_attributes_in_model
+      else
+        keys
+      end
+    end
+
+    private
+      def has_attribute_aliases?(attributes)
+        attributes.keys.any? { |attribute| model.attribute_alias?(attribute) }
+      end
+
+      def resolve_sti
+        return if model.descends_from_active_record?
+
+        sti_type = model.sti_name
+        @inserts = @inserts.map do |insert|
+          insert.reverse_merge(model.inheritance_column.to_s => sti_type)
+        end
+      end
+
+      def resolve_attribute_aliases
+        return unless has_attribute_aliases?(@inserts.first)
+
+        @inserts = @inserts.map do |insert|
+          insert.transform_keys { |attribute| resolve_attribute_alias(attribute) }
+        end
+
+        @update_only = Array(@update_only).map { |attribute| resolve_attribute_alias(attribute) } if @update_only
+        @unique_by = Array(@unique_by).map { |attribute| resolve_attribute_alias(attribute) } if @unique_by
+      end
+
+      def resolve_attribute_alias(attribute)
+        model.attribute_alias(attribute) || attribute
+      end
+
+      def configure_on_duplicate_update_logic
+        if custom_update_sql_provided? && update_only.present?
+          raise ArgumentError, "You can't set :update_only and provide custom update SQL via :on_duplicate at the same time"
+        end
+
+        if update_only.present?
+          @updatable_columns = Array(update_only)
+          @on_duplicate = :update
+        elsif custom_update_sql_provided?
+          @update_sql = on_duplicate
+          @on_duplicate = :update
+        elsif @on_duplicate == :update && updatable_columns.empty?
+          @on_duplicate = :skip
+        end
+      end
+
+      def custom_update_sql_provided?
+        @custom_update_sql_provided ||= Arel.arel_node?(on_duplicate)
+      end
+
+      def find_unique_index_for(unique_by)
+        if !connection.supports_insert_conflict_target?
+          return if unique_by.nil?
+
+          raise ArgumentError, "#{connection.class} does not support :unique_by"
+        end
+
+        name_or_columns = unique_by || model.primary_key
+        match = Array(name_or_columns).map(&:to_s)
+        sorted_match = match.sort
+
+        if index = unique_indexes.find { |i| match.include?(i.name) || Array(i.columns).sort == sorted_match }
+          index
+        elsif match == primary_keys
+          unique_by.nil? ? nil : ActiveRecord::ConnectionAdapters::IndexDefinition.new(model.table_name, "#{model.table_name}_primary_key", true, match)
+        else
+          raise ArgumentError, "No unique index found for #{name_or_columns}"
+        end
+      end
+
+      def unique_indexes
+        @model.schema_cache.indexes(model.table_name).select(&:unique)
+      end
+
+      def ensure_valid_options_for_connection!
+        if returning && !connection.supports_insert_returning?
+          raise ArgumentError, "#{connection.class} does not support :returning"
+        end
+
+        if skip_duplicates? && !connection.supports_insert_on_duplicate_skip?
+          raise ArgumentError, "#{connection.class} does not support skipping duplicates"
+        end
+
+        if update_duplicates? && !connection.supports_insert_on_duplicate_update?
+          raise ArgumentError, "#{connection.class} does not support upsert"
+        end
+
+        if unique_by && !connection.supports_insert_conflict_target?
+          raise ArgumentError, "#{connection.class} does not support :unique_by"
+        end
+      end
+
+
+      def to_sql
+        connection.build_insert_sql(ActiveRecord::InsertAll::Builder.new(self))
+      end
+
+
+      def readonly_columns
+        primary_keys + model.readonly_attributes
+      end
+
+      def unique_by_columns
+        Array(unique_by&.columns)
+      end
+
+
+      def verify_attributes(attributes)
+        if keys_including_timestamps != attributes.keys.to_set
+          raise ArgumentError, "All objects being inserted must have the same keys"
+        end
+      end
+
+      def disallow_raw_sql!(value)
+        return if !value.is_a?(String) || Arel.arel_node?(value)
+
+        raise ArgumentError, "Dangerous query method (method whose arguments are used as raw " \
+                             "SQL) called: #{value}. " \
+                             "Known-safe values can be passed " \
+                             "by wrapping them in Arel.sql()."
+      end
+
+      def timestamps_for_create
+        model.all_timestamp_attributes_in_model.index_with(connection.high_precision_current_timestamp)
+      end
+
+      class Builder # :nodoc:
+        attr_reader :model
+
+        delegate :skip_duplicates?, :update_duplicates?, :keys, :keys_including_timestamps, :record_timestamps?, to: :insert_all
+
+        def initialize(insert_all)
+          @insert_all, @model, @connection = insert_all, insert_all.model, insert_all.connection
+        end
+
+        def into
+          "INTO #{model.quoted_table_name} (#{columns_list})"
+        end
+
+        def values_list
+          types = extract_types_from_columns_on(model.table_name, keys: keys_including_timestamps)
+
+          values_list = insert_all.map_key_with_value do |key, value|
+            next value if Arel::Nodes::SqlLiteral === value
+            connection.with_yaml_fallback(types[key].serialize(value))
+          end
+
+          connection.visitor.compile(Arel::Nodes::ValuesList.new(values_list))
+        end
+
+        def returning
+          return unless insert_all.returning
+
+          if insert_all.returning.is_a?(String)
+            insert_all.returning
+          else
+            Array(insert_all.returning).map do |attribute|
+              if model.attribute_alias?(attribute)
+                "#{quote_column(model.attribute_alias(attribute))} AS #{quote_column(attribute)}"
+              else
+                quote_column(attribute)
+              end
+            end.join(",")
+          end
+        end
+
+        def conflict_target
+          if index = insert_all.unique_by
+            sql = +"(#{format_columns(index.columns)})"
+            sql << " WHERE #{index.where}" if index.where
+            sql
+          elsif update_duplicates?
+            "(#{format_columns(insert_all.primary_keys)})"
+          end
+        end
+
+        def updatable_columns
+          quote_columns(insert_all.updatable_columns)
+        end
+
+        def touch_model_timestamps_unless(&block)
+          return "" unless update_duplicates? && record_timestamps?
+
+          model.timestamp_attributes_for_update_in_model.filter_map do |column_name|
+            if touch_timestamp_attribute?(column_name)
+              "#{column_name}=(CASE WHEN (#{updatable_columns.map(&block).join(" AND ")}) THEN #{model.quoted_table_name}.#{column_name} ELSE #{connection.high_precision_current_timestamp} END),"
+            end
+          end.join
+        end
+
+        def raw_update_sql
+          insert_all.update_sql
+        end
+
+        alias raw_update_sql? raw_update_sql
+
+        private
+          attr_reader :connection, :insert_all
+
+          def touch_timestamp_attribute?(column_name)
+            insert_all.updatable_columns.exclude?(column_name)
+          end
+
+          def columns_list
+            format_columns(insert_all.keys_including_timestamps)
+          end
+
+          def extract_types_from_columns_on(table_name, keys:)
+            columns = @model.schema_cache.columns_hash(table_name)
+
+            unknown_column = (keys - columns.keys).first
+            raise UnknownAttributeError.new(model.new, unknown_column) if unknown_column
+
+            keys.index_with { |key| model.type_for_attribute(key) }
+          end
+
+          def format_columns(columns)
+            columns.respond_to?(:map) ? quote_columns(columns).join(",") : columns
+          end
+
+          def quote_columns(columns)
+            columns.map { |column| quote_column(column) }
+          end
+
+          def quote_column(column)
+            connection.quote_column_name(column)
+          end
+      end
+  end
+end

data/lib/active_record/integration.rb

@@ -0,0 +1,209 @@
+# frozen_string_literal: true
+
+require "active_support/core_ext/string/filters"
+
+module ActiveRecord
+  module Integration
+    extend ActiveSupport::Concern
+
+    included do
+      ##
+      # :singleton-method:
+      # Indicates the format used to generate the timestamp in the cache key, if
+      # versioning is off. Accepts any of the symbols in +Time::DATE_FORMATS+.
+      #
+      # This is +:usec+, by default.
+      class_attribute :cache_timestamp_format, instance_writer: false, default: :usec
+
+      ##
+      # :singleton-method:
+      # Indicates whether to use a stable #cache_key method that is accompanied
+      # by a changing version in the #cache_version method.
+      #
+      # This is +true+, by default on \Rails 5.2 and above.
+      class_attribute :cache_versioning, instance_writer: false, default: false
+
+      ##
+      # :singleton-method:
+      # Indicates whether to use a stable #cache_key method that is accompanied
+      # by a changing version in the #cache_version method on collections.
+      #
+      # This is +false+, by default until \Rails 6.1.
+      class_attribute :collection_cache_versioning, instance_writer: false, default: false
+    end
+
+    # Returns a +String+, which Action Pack uses for constructing a URL to this
+    # object. The default implementation returns this record's id as a +String+,
+    # or +nil+ if this record's unsaved.
+    #
+    # For example, suppose that you have a User model, and that you have a
+    # <tt>resources :users</tt> route. Normally, +user_path+ will
+    # construct a path with the user object's 'id' in it:
+    #
+    #   user = User.find_by(name: 'Phusion')
+    #   user_path(user)  # => "/users/1"
+    #
+    # You can override +to_param+ in your model to make +user_path+ construct
+    # a path using the user's name instead of the user's id:
+    #
+    #   class User < ActiveRecord::Base
+    #     def to_param  # overridden
+    #       name
+    #     end
+    #   end
+    #
+    #   user = User.find_by(name: 'Phusion')
+    #   user_path(user)  # => "/users/Phusion"
+    def to_param
+      return unless id
+      Array(id).join(self.class.param_delimiter)
+    end
+
+    # Returns a stable cache key that can be used to identify this record.
+    #
+    #   Product.new.cache_key     # => "products/new"
+    #   Product.find(5).cache_key # => "products/5"
+    #
+    # If ActiveRecord::Base.cache_versioning is turned off, as it was in \Rails 5.1 and earlier,
+    # the cache key will also include a version.
+    #
+    #   Product.cache_versioning = false
+    #   Product.find(5).cache_key  # => "products/5-20071224150000" (updated_at available)
+    def cache_key
+      if new_record?
+        "#{model_name.cache_key}/new"
+      else
+        if cache_version
+          "#{model_name.cache_key}/#{id}"
+        else
+          timestamp = max_updated_column_timestamp
+
+          if timestamp
+            timestamp = timestamp.utc.to_fs(cache_timestamp_format)
+            "#{model_name.cache_key}/#{id}-#{timestamp}"
+          else
+            "#{model_name.cache_key}/#{id}"
+          end
+        end
+      end
+    end
+
+    # Returns a cache version that can be used together with the cache key to form
+    # a recyclable caching scheme. By default, the #updated_at column is used for the
+    # cache_version, but this method can be overwritten to return something else.
+    #
+    # Note, this method will return nil if ActiveRecord::Base.cache_versioning is set to
+    # +false+.
+    def cache_version
+      return unless cache_versioning
+
+      if has_attribute?("updated_at")
+        timestamp = updated_at_before_type_cast
+        if can_use_fast_cache_version?(timestamp)
+          raw_timestamp_to_cache_version(timestamp)
+
+        elsif timestamp = updated_at
+          timestamp.utc.to_fs(cache_timestamp_format)
+        end
+      elsif self.class.has_attribute?("updated_at")
+        raise ActiveModel::MissingAttributeError, "missing attribute 'updated_at' for #{self.class}"
+      end
+    end
+
+    # Returns a cache key along with the version.
+    def cache_key_with_version
+      if version = cache_version
+        "#{cache_key}-#{version}"
+      else
+        cache_key
+      end
+    end
+
+    module ClassMethods
+      # Defines your model's +to_param+ method to generate "pretty" URLs
+      # using +method_name+, which can be any attribute or method that
+      # responds to +to_s+.
+      #
+      #   class User < ActiveRecord::Base
+      #     to_param :name
+      #   end
+      #
+      #   user = User.find_by(name: 'Fancy Pants')
+      #   user.id         # => 123
+      #   user_path(user) # => "/users/123-fancy-pants"
+      #
+      # Values longer than 20 characters will be truncated. The value
+      # is truncated word by word.
+      #
+      #   user = User.find_by(name: 'David Heinemeier Hansson')
+      #   user.id         # => 125
+      #   user_path(user) # => "/users/125-david-heinemeier"
+      #
+      # Because the generated param begins with the record's +id+, it is
+      # suitable for passing to +find+. In a controller, for example:
+      #
+      #   params[:id]               # => "123-fancy-pants"
+      #   User.find(params[:id]).id # => 123
+      def to_param(method_name = nil)
+        if method_name.nil?
+          super()
+        else
+          define_method :to_param do
+            if (default = super()) &&
+                 (result = send(method_name).to_s).present? &&
+                   (param = result.squish.parameterize.truncate(20, separator: /-/, omission: "")).present?
+              "#{default}-#{param}"
+            else
+              default
+            end
+          end
+        end
+      end
+
+      def collection_cache_key(collection = all, timestamp_column = :updated_at) # :nodoc:
+        collection.send(:compute_cache_key, timestamp_column)
+      end
+    end
+
+    private
+      # Detects if the value before type cast
+      # can be used to generate a cache_version.
+      #
+      # The fast cache version only works with a
+      # string value directly from the database.
+      #
+      # We also must check if the timestamp format has been changed
+      # or if the timezone is not set to UTC then
+      # we cannot apply our transformations correctly.
+      def can_use_fast_cache_version?(timestamp)
+        timestamp.is_a?(String) &&
+          cache_timestamp_format == :usec &&
+          # FIXME: checking out a connection for this is wasteful
+          # we should store/cache this information in the schema cache
+          # or similar.
+          self.class.with_connection(&:default_timezone) == :utc &&
+          !updated_at_came_from_user?
+      end
+
+      # Converts a raw database string to `:usec`
+      # format.
+      #
+      # Example:
+      #
+      #   timestamp = "2018-10-15 20:02:15.266505"
+      #   raw_timestamp_to_cache_version(timestamp)
+      #   # => "20181015200215266505"
+      #
+      # PostgreSQL truncates trailing zeros,
+      # https://github.com/postgres/postgres/commit/3e1beda2cde3495f41290e1ece5d544525810214
+      # to account for this we pad the output with zeros
+      def raw_timestamp_to_cache_version(timestamp)
+        key = timestamp.delete("- :.")
+        if key.length < 20
+          key.ljust(20, "0")
+        else
+          key
+        end
+      end
+  end
+end

data/lib/active_record/internal_metadata.rb

@@ -0,0 +1,164 @@
+# frozen_string_literal: true
+
+require "active_record/scoping/default"
+require "active_record/scoping/named"
+
+module ActiveRecord
+  # This class is used to create a table that keeps track of values and keys such
+  # as which environment migrations were run in.
+  #
+  # This is enabled by default. To disable this functionality set
+  # `use_metadata_table` to false in your database configuration.
+  class InternalMetadata # :nodoc:
+    class NullInternalMetadata # :nodoc:
+    end
+
+    attr_reader :arel_table
+
+    def initialize(pool)
+      @pool = pool
+      @arel_table = Arel::Table.new(table_name)
+    end
+
+    def primary_key
+      "key"
+    end
+
+    def value_key
+      "value"
+    end
+
+    def table_name
+      "#{ActiveRecord::Base.table_name_prefix}#{ActiveRecord::Base.internal_metadata_table_name}#{ActiveRecord::Base.table_name_suffix}"
+    end
+
+    def enabled?
+      @pool.db_config.use_metadata_table?
+    end
+
+    def []=(key, value)
+      return unless enabled?
+
+      @pool.with_connection do |connection|
+        update_or_create_entry(connection, key, value)
+      end
+    end
+
+    def [](key)
+      return unless enabled?
+
+      @pool.with_connection do |connection|
+        if entry = select_entry(connection, key)
+          entry[value_key]
+        end
+      end
+    end
+
+    def delete_all_entries
+      dm = Arel::DeleteManager.new(arel_table)
+
+      @pool.with_connection do |connection|
+        connection.delete(dm, "#{self.class} Destroy")
+      end
+    end
+
+    def count
+      sm = Arel::SelectManager.new(arel_table)
+      sm.project(*Arel::Nodes::Count.new([Arel.star]))
+
+      @pool.with_connection do |connection|
+        connection.select_values(sm, "#{self.class} Count").first
+      end
+    end
+
+    def create_table_and_set_flags(environment, schema_sha1 = nil)
+      return unless enabled?
+
+      @pool.with_connection do |connection|
+        create_table
+        update_or_create_entry(connection, :environment, environment)
+        update_or_create_entry(connection, :schema_sha1, schema_sha1) if schema_sha1
+      end
+    end
+
+    # Creates an internal metadata table with columns +key+ and +value+
+    def create_table
+      return unless enabled?
+
+      @pool.with_connection do |connection|
+        unless connection.table_exists?(table_name)
+          connection.create_table(table_name, id: false) do |t|
+            t.string :key, **connection.internal_string_options_for_primary_key
+            t.string :value
+            t.timestamps
+          end
+        end
+      end
+    end
+
+    def drop_table
+      return unless enabled?
+
+      @pool.with_connection do |connection|
+        connection.drop_table table_name, if_exists: true
+      end
+    end
+
+    def table_exists?
+      @pool.schema_cache.data_source_exists?(table_name)
+    end
+
+    private
+      def update_or_create_entry(connection, key, value)
+        entry = select_entry(connection, key)
+
+        if entry
+          if entry[value_key] != value
+            update_entry(connection, key, value)
+          else
+            entry[value_key]
+          end
+        else
+          create_entry(connection, key, value)
+        end
+      end
+
+      def current_time(connection)
+        connection.default_timezone == :utc ? Time.now.utc : Time.now
+      end
+
+      def create_entry(connection, key, value)
+        im = Arel::InsertManager.new(arel_table)
+        im.insert [
+          [arel_table[primary_key], key],
+          [arel_table[value_key], value],
+          [arel_table[:created_at], current_time(connection)],
+          [arel_table[:updated_at], current_time(connection)]
+        ]
+
+        connection.insert(im, "#{self.class} Create", primary_key, key)
+      end
+
+      def update_entry(connection, key, new_value)
+        um = Arel::UpdateManager.new(arel_table)
+        um.set [
+          [arel_table[value_key], new_value],
+          [arel_table[:updated_at], current_time(connection)]
+        ]
+
+        um.where(arel_table[primary_key].eq(key))
+
+        connection.update(um, "#{self.class} Update")
+      end
+
+      def select_entry(connection, key)
+        sm = Arel::SelectManager.new(arel_table)
+        sm.project(Arel::Nodes::SqlLiteral.new("*", retryable: true))
+        sm.where(arel_table[primary_key].eq(Arel::Nodes::BindParam.new(key)))
+        sm.order(arel_table[primary_key].asc)
+        sm.limit = 1
+
+        connection.select_all(sm, "#{self.class} Load").first
+      end
+  end
+end