有什么方法可以始终仅使用Rails API的应用程序与子对象或子对象一起检索父对象?
例如我有一个@students数组。 @students数组中的每个学生对象都有两个外键,分别是standard_id和school_id。现在,所有对象默认都具有standard_id和school_id。相反,我希望在@students数组中的每个学生对象中使用标准对象和学校对象。
我得到的回应
[
{
"id": 1,
"standard_id": 1,
"school_id": 1,
"created_at": "2019-04-14T11:36:03.000Z",
"updated_at": "2019-04-14T11:36:03.000Z"
},
{
"id": 2,
"standard_id": 1,
"school_id": 1,
"created_at": "2019-04-14T11:41:38.000Z",
"updated_at": "2019-04-14T11:41:45.000Z"
}
]
我想要的答复
[
{
"id": 1,
"standard_id": 1,
"school_id": 1,
"created_at": "2019-04-14T11:36:03.000Z",
"updated_at": "2019-04-14T11:36:03.000Z",
"standard": {
"id": 1,
"name": "1",
"created_at": "2019-04-14T11:32:15.000Z",
"updated_at": "2019-04-14T11:32:15.000Z"
},
"school": {
"id": 1,
"name": "SACS",
"created_at": "2019-04-14T11:35:24.000Z",
"updated_at": "2019-04-14T11:35:24.000Z"
}
},
{
"id": 2,
"standard_id": 1,
"school_id": 1,
"created_at": "2019-04-14T11:41:38.000Z",
"updated_at": "2019-04-14T11:41:45.000Z",
"standard": {
"id": 1,
"name": "1",
"created_at": "2019-04-14T11:32:15.000Z",
"updated_at": "2019-04-14T11:32:15.000Z"
},
"school": {
"id": 1,
"name": "SACS",
"created_at": "2019-04-14T11:35:24.000Z",
"updated_at": "2019-04-14T11:35:24.000Z"
}
}
]
所有控制器都有共同的解决方案吗?因为该应用程序已经构建。现在非常忙于在每个控制器中手动格式化数据。预先感谢。
答案 0 :(得分:2)
如果您在控制器中使用Rails默认的as_json
序列化程序,即参见下文:
render json: @students
# ^ above will default call `.to_json` to `@students`, which will also call `.as_json`
# thereby, equivalently calling:
# render json: @students.as_json
...然后,您可以稍微修改as_json
(see docs),以便JSON将包含第一级嵌套关联;见下文
app / models / student.rb
class Student < ApplicationRecord
belongs_to :standard
belongs_to :school
def as_json(**options)
unless options.has_key? :include
options.merge!(include: [:standard, :school])
end
super(options)
end
# or if you don't want to manually include "each" association, and just dynamically include per association
# def as_json(**options)
# unless options.has_key? :include
# options.merge!(
# include: self.class.reflect_on_all_associations.map(&:name)
# )
# end
# super(options)
# end
end
与上述解决方案相同,但如果您希望此方法适用于 ALL 模型,而不仅适用于Student
模型,请执行以下操作:
app / models / application_record.rb
class ApplicationRecord < ActiveRecord::Base
def as_json(**options)
unless options.has_key? :include
options.merge!(
include: self.class.reflect_on_all_associations.map(&:name)
)
end
super(options)
end
end
# rails console
students = Student.all
puts students.as_json
# => [{"id"=>1, "standard_id"=>1, "school_id"=>1, "created_at"=>"2019-04-14T11:36:03.000Z", "updated_at"=>"2019-04-14T11:36:03.000Z", "standard"=>{"id"=>1, "name"=>"1", "created_at"=>"2019-04-14T11:32:15.000Z", "updated_at"=>"2019-04-14T11:32:15.000Z"}, "school"=>{"id"=>1, "name"=>"SACS", "created_at"=>"2019-04-14T11:35:24.000Z", "updated_at"=>"2019-04-14T11:35:24.000Z"}}, {"id"=>2, "standard_id"=>1, "school_id"=>1, "created_at"=>"2019-04-14T11:41:38.000Z", "updated_at"=>"2019-04-14T11:41:45.000Z", "standard"=>{"id"=>1, "name"=>"1", "created_at"=>"2019-04-14T11:32:15.000Z", "updated_at"=>"2019-04-14T11:32:15.000Z"}, "school"=>{"id"=>1, "name"=>"SACS", "created_at"=>"2019-04-14T11:35:24.000Z", "updated_at"=>"2019-04-14T11:35:24.000Z"}}]
以上解决方案仅将第一级关联呈现为JSON响应的一部分,而不会“深度”呈现第二级或第三级等关联。 >
我很好奇您提出的整合分页的请求;因此,下面是我可能的解决方案(经过测试,尽管尚不确定是否有副作用):
您将需要一个“分页”宝石,即在下面的示例中,我正在使用kaminari
app / models / application_record.rb
class ApplicationRecord < ActiveRecord::Base
self.abstract_class = true
def as_json(**options)
unless options.has_key? :include
options.merge!(
include: self.class.reflect_on_all_associations.map(&:name).inject({}) do |hash, name|
paginate = options.dig(:association_paginations, name.to_sym, :paginate)
paginate = true if paginate.nil?
page = options.dig(:association_paginations, name.to_sym, :page) || 1
per = options.dig(:association_paginations, name.to_sym, :per) || Kaminari.config.default_per_page
hash[name.to_sym] = {
paginate: paginate,
page: page,
per: per
}
hash
end
)
end
super(options)
end
end
app / config / initializers / active_model_serialization_patch.rb
module ActiveModel::Serialization
private def serializable_add_includes(options = {})
if Gem.loaded_specs['activemodel'].version.to_s != '6.0.0.beta3' # '5.2.3'
raise "Version mismatch! \
Not guaranteed to work properly without side effects! \
You'll have to copy and paste (and modify) to below correct code (and Gem version!) from \
https://github.com/rails/rails/blob/5-2-stable/activemodel/lib/active_model/serialization.rb#L178"
else
# copied code: start
return unless includes = options[:include]
unless includes.is_a?(Hash)
includes = Hash[Array(includes).flat_map { |n| n.is_a?(Hash) ? n.to_a : [[n, {}]] }]
end
includes.each do |association, opts|
if opts[:paginate]
opts[:page] ||= 1
opts[:per] ||= Kaminari.config.default_per_page
records = send(association).page(opts[:page]).per(opts[:per])
else
records = send(association)
end
if records
yield association, records, opts
end
end
# copied code: end
end
end
end
app / config / initializers / active_record_relation_patch.rb
class ActiveRecord::Relation
def as_json(**options)
if options[:paginate]
options[:page] ||= 1
options[:per] ||= Kaminari.config.default_per_page
options[:paginate] = false
page(options[:page]).per(options[:per]).as_json(options)
else
super(options)
end
end
end
your_controller.rb
def some_action
@students = Student.all
render json: @students.as_json(
paginate: true,
page: params[:page],
per: params[:per],
association_paginations: params[:association_paginations]
)
end
http://localhost:3000/your_controller/some_action?per=2&page=1
因为per
= 2
[
{
id: 1,
school_id: 101,
school: { id: 101, ... },
standard_id: 201,
standard: { id: 201, ... },
subjects: [
{ id: 301, ... },
{ id: 302, ... },
{ id: 303, ... },
{ id: 304, ... },
{ id: 305, ... },
...
],
attendances: [
{ id: 123, ... },
{ id: 124, ... },
{ id: 125, ... },
{ id: 126, ... },
{ id: 127, ... },
{ id: 128, ... },
...
]
},
{
id: 2,
...
}
]
http://localhost:3000/your_controller/some_action?per=2&page=1&association_paginations[subjects][per]=2
因为per
= 2
由于association_paginations[subjects][per]
= 2
[
{
id: 1,
school_id: 101,
school: { id: 101, ... },
standard_id: 201,
standard: { id: 201, ... },
subjects: [
{ id: 301, ... },
{ id: 302, ... },
],
attendances: [
{ id: 123, ... },
{ id: 124, ... },
{ id: 125, ... },
{ id: 126, ... },
{ id: 127, ... },
{ id: 128, ... },
...
]
},
{
id: 2,
...
}
]
P.S。由于上述解决方案涉及猴子补丁,因此我建议您改用ActiveModelSerializers::Model,因为您请求的功能现在变得越来越复杂。