Question

鉴于此代码：

class Mammal
  @@class_name = "mammal"
  def self.class_name
    @@class_name
  end
end

class Dog < Mammal
  @@class_name = "dog"
  def self.class_name
    @@class_name
  end
end

Dog.class_name
# => dog
Mammal.class_name
# => dog

为什么这是红宝石的特征？似乎类变量遵循与实例变量不同的范围。类变量的范围是否在其整个继承链中共享？如果是这样，为什么呢？它们是单独的类，应该有作用域类变量，不是吗？我错过了什么？

Answer 1

类变量在整个类层次结构中共享，它们不会被继承，因此赋值顺序很重要：

class Mammal
  @@class_name = "mammal"
  def self.class_name
    @@class_name
  end
end

class Dog < Mammal
  @@class_name = "dog"
  def self.class_name
    @@class_name
  end
end

class Mammal
  @@class_name = "fish"
end

Dog.class_name # fish
Mammal.class_name # fish

<强>更新

使用类实例变量，类变量很棘手：

class Mammal
  @class_name = "mammal"
  def self.class_name
    @class_name
  end
end

class Dog < Mammal
  @class_name = "dog"
  def self.class_name
    @class_name
  end
end

Dog.class_name # dog
Mammal.class_name # mammal

为什么ruby父类假定其子类的类变量？

1 个答案: