bulk update of records in rails with UPDATE SET FROM (VALUES ...) AS WHERE

rails_bulk_update.rb

    # NOTE assumes the relation is very very simple
    def bulk_update(relation, values, where_columns, set_columns)
      return if values.empty?

      connection = relation.connection
      base_class = relation.base_class

      if relation.arel.where_sql.present?
        relation_ids = relation.ids
      end

      quote_column_name_fn = connection.method(:quote_column_name)
      quote_column_value_fn = ->(value, col_name) { base_class.quote_value(value, base_class.column_for_attribute(col_name)) }

      destination_table_name_quoted = relation.quoted_table_name
      temporary_table_name_quoted = quote_column_name_fn.('tmp_data')

      temporary_table_columns = [*where_columns, *set_columns]
      temporary_table_def = "#{temporary_table_name_quoted}(#{temporary_table_columns.map(&quote_column_name_fn).join(', ')})"
      temporary_table_values_list = values.map do |record|
        record.to_h.values_at(*temporary_table_columns).map.with_index do |val, i|
          # cast is to make sure temporary table is correctly typed (e.g. timestamp without explicit type coercion is text)
          "CAST(#{quote_column_value_fn.(val, temporary_table_columns[i])} AS #{base_class.column_for_attribute(temporary_table_columns[i]).sql_type})"
        end
      end

      set_column_names_quoted = set_columns.map(&quote_column_name_fn)
      set_fields = set_column_names_quoted.map { |col| "#{col} = #{temporary_table_name_quoted}.#{col}" }.join(', ')

      where_column_names_quoted = where_columns.map(&quote_column_name_fn)
      where_cond = where_column_names_quoted.map { |col| "#{destination_table_name_quoted}.#{col} = #{temporary_table_name_quoted}.#{col}" }.join(' AND ')

      if relation_ids
        relation_cond = "#{destination_table_name_quoted}.#{quote_column_name_fn.(relation.primary_key)} IN (#{relation_ids.map {|val| quote_column_value_fn.(val, relation.primary_key) }.join(', ')})"
      end

      connection.update(<<-SQL)
        UPDATE #{destination_table_name_quoted}
         SET #{set_fields}
         FROM
          (VALUES
            #{temporary_table_values_list.map { |vals| "(#{vals.join(', ')})" }.join(', ')}
          ) AS #{temporary_table_def}
         WHERE
          #{[where_cond, relation_cond].compact.join(' AND ')}
      SQL
    end

bulk_update(Users.all, [{id: 1, name: 'User 1'}, {id: 2, name: 'User 2'}], [:id], [:name])

Note:
This is primarily for PostgreSQL. Does not work with MySQL because it does not support (VALUES ...) AS.

I'd love to know how to do this with MySQL. INSERT with ON DUPLICATE KEY UPDATE? Maybe CASE-WHEN is most universal...

maybe this http://stackoverflow.com/a/19033152/4212945