我正在尝试将哈希转换为活动记录关系,但无法这样做。我打算使用活动记录关系对类别表进行排序,然后过滤。最终目标是创建一个实例方法来过滤访问的前 5 个类别,然后我可以在控制器中使用/调用它们。这就是我的想法:
品类型号:
def top_5_visited
Article.joins(:categories).group('categories.name').sum(:impressions_count)
// Add sort
// Limit to top 5
end
类别控制器:
@categories = Category.top_5 visited
哈希{"Simula"=>7, "Haskell"=>5, "JavaScript"=>10, "C#"=>112}
将通过以下查询创建:
total_count = Article.joins(:categories).group('categories.name').sum(:impressions_count)
我还尝试使用sort_by方法将其转换为数组:
total_count_sorted = total_count.sort_by {|_key, value| value}
我已经用谷歌搜索了"将数组转换为活动记录关系"并引用了这篇文章,但对此进行了测试:
Category.where(id: total_count_sort.map(&:id))
在 Rails 控制台中,显示此错误:
NoMethodError: ["Simula", 7]:Array 的未定义方法 ID
您要执行的操作从反向结束(类别(开始,并在 ORDER 子句中使用聚合。
Category.joins(:articles)
.order('SUM(articles.impressions_count) DESC')
.group(:id)
.limit(5)
irb(main):005:0> Category.joins(:articles).order("SUM(articles.impressions_count) DESC").group('categories.id').limit(5)
Category Load (1.5ms) SELECT "categories".* FROM "categories" INNER JOIN "articles" ON "articles"."category_id" = "categories"."id" GROUP BY categories.id ORDER BY SUM(articles.impressions_count) DESC LIMIT $1 [["LIMIT", 5]]
=> #<ActiveRecord::Relation [#<Category id: 4, name: "C#", created_at: "2017-11-15 15:06:32", updated_at: "2017-11-15 15:06:32">, #<Category id: 3, name: "JavaScript", created_at: "2017-11-15 15:06:32", updated_at: "2017-11-15 15:06:32">, #<Category id: 1, name: "Simula", created_at: "2017-11-15 15:03:37", updated_at: "2017-11-15 15:03:37">, #<Category id: 2, name: "Haskell", created_at: "2017-11-15 15:06:32", updated_at: "2017-11-15 15:06:32">]>
并且您应该创建一个类方法 - 而不是实例方法,因为这基本上只是一个范围,调用实例没有意义。
class Category < ApplicationRecord
has_many :articles
def self.order_by_article_impressions
self.joins(:articles)
.order('SUM(articles.impressions_count)')
.group(:id)
end
def self.top_5_visited
order_by_article_impressions.limit(5)
end
# Or use `scope` which is just syntactic sugar
scope(:top_5_visited) -> { order_by_article_impressions.limit(5) }
end
将代码更改为:
Category.where(id: total_count_sort.map(&:last))