如何将方法添加到activerecord集合?

Leo*_*eon 17 activerecord ruby-on-rails

我想为特定模型的所有集合添加一个方法.假设我想将该方法添加my_complicated_averaging_method到WeatherData集合中:

WeatherData.all.limit(3).my_complicated_averaging_method()
Station.first.weatherdata.my_complicated_averaging_method()
Run Code Online (Sandbox Code Playgroud)

做这个的最好方式是什么?目前,我找到的唯一方法是这样的:

class WeatherData < ActiveRecord::Base
  def self.my_complicated_averaging_method
    weighted_average = 0
    @relation.each do |post|
      # do something complicated
      # weighted_average = 
    end
    return weighted_average
  end
end
Run Code Online (Sandbox Code Playgroud)

这是将方法添加到集合的好方法吗?有没有更好/支持的方式来做到这一点?

Edg*_*ons 12

有很多方法如何做到这一点,而你的是完全有效的(虽然我个人更喜欢包装类的方法为单独的块检查的出),但随着人们正在添加更多的业务逻辑,以他们的模型和盲从"瘦控制器,胖模型"概念,模型变成完全混乱.

为了避免这种混乱,引入服务对象是一个好主意,在你的情况下,它将是这样的:

class AverageWeatherData
  class << self
    def data(collection)
      new(collection).data
    end
  end

  def initialize(collection)
    @collection = collection
  end

  def data
    @collection.reduce do |avg, post|
      # reduce goes through every post, each next iteration receives in avg a value of the last line of iteration
      # do something with avg and post 
    end
    # no need for explicit return, every line of Ruby code returns it's value
    # so this method would return result of the reduce
    # more on reduce: http://ruby-doc.org/core-2.0.0/Enumerable.html#method-i-reduce
  end
end
Run Code Online (Sandbox Code Playgroud)

现在,您可以通过将您的集合传递给它来直接调用此类.但您也可以像这样代理呼叫:

def self.my_complicated_averaging_method
  AverageWeatherData.data(@relation)
end
Run Code Online (Sandbox Code Playgroud)

我鼓励您通过阅读此博客了解更多此方法:http: //blog.codeclimate.com/blog/2012/10/17/7-ways-to-decompose-fat-activerecord-models/

UPD

你是正确的使用实例变量是一种可能的方式搞乱对象内部(加上它不是一个公共接口,它可能会在未来发生变化).我的建议是使用方法scoped.基本上替换@relationscoped.

检查此示例.我使用了自己项目中的模型来证明它确实有效

2.0.0p247 :001 > Tracking # just asking console to load this class before modifying it
# => Tracking(id: integer, action: string, cookie_id: string, ext_object_id: integer, created_at: datetime, updated_at: datetime)
2.0.0p247 :002 > class Tracking
2.0.0p247 :003?>     def self.fetch_ids
2.0.0p247 :004?>         scoped.map(&:id)
2.0.0p247 :005?>       end
2.0.0p247 :006?>   end
# => nil
2.0.0p247 :007 >
2.0.0p247 :008 >   Tracking.where(id: (1..100)).fetch_ids
#  Tracking Load (2.0ms)  SELECT "trackings".* FROM "trackings" WHERE ("trackings"."id" BETWEEN 1 AND 100)
# => [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]
Run Code Online (Sandbox Code Playgroud)

UPD

在Rails 4 scoped中已弃用,因此使用正确all.

all.map(&:id)
Run Code Online (Sandbox Code Playgroud)

  • __UPD2__ 不要在集合上使用`.map(&amp;:id)`,而是使用`.pluck(:id)`。 (2认同)

el_*_*ick 5

在 Rails >= 4 上,您可以使用where(nil)代替scoped

class Foo < ActiveRecord::Base  
  def self.bar
    where(nil).pluck(:id)
  end
end

Foo.where(id: [1, 2, 3]).order(:id).bar
Run Code Online (Sandbox Code Playgroud)

此外,您可以使用#scope,例如:

class Foo < ActiveRecord::Base
  scope :bar, -> {where(nil).pluck(:id)}
end
Run Code Online (Sandbox Code Playgroud)

最后,您可以编写如下代码 Foo.all.bar