Django REST框架：何时创建超链接资源和嵌套资源？如何发布嵌套资源？ [英] Django REST Framework: when to create a hyperlinked resource and when nested resource? How to POST a nested resource?

问题描述

我正在使用Django REST框架构建REST Web API。事情进展顺利，但是我却遇到了嵌套资源的问题。起初，REST API中的所有关系都被超链接。例如，一个帖子如下所示：

  {
path：http：// api。 myproject.com/posts/1.json，
id：1，
author：http://api.myproject.com/users/broak.json，
image：/images/posts/cxyUzlPo.jpg，
header：我是谁？，
footer：我是冠军！，
date：2014-11-09 15:16，
喜欢：http://api.myproject.com/posts/1/likes.json，
评论：http://api.myproject.com/posts/1/comments.json，
likes_count：0，
comments_count：0 
} 
  

文章与作者（用户）之间的关系是超链接的。当你想创建一个新的帖子时，你需要指定一个正确的超链接到一个特定的用户 - 这很正常。

当调用帖子列表时，事情变得无效，因为您必须为每个帖子的每个作者添加一个额外的API调用。我通过使用NESTED资源而不是HYPERLINKED资源解决了这个问题，所以每个帖子现在都包含有关作者的所有信息。

  {
path：http://api.myproject.com/posts/1.json，
id：1，
author：{
email ：[e-mail]
username：broak，
first_name：John，
last_name：Broak，
is_staff ：False，
is_active：True，
last_login：02-26-2016
}，
image：/ images / posts / cxyUzlPo 
标题：我是谁？，
footer：我是冠军！，
date：2014-11-09 15 ：16，
喜欢：http://api.myproject.com/posts/1/likes.json，
评论：http://api.myproject.com /posts/1/comments.json，
likes_count：0，
comments_count：0 
} 
  

我的第一个问题是：你有没有指导，是否应该创建一个嵌套数据st结构或具有超链接的单独端点。

我的第二个问题是：当我使用作者作为嵌套资源并且想要创建一个新的帖子时，我不想要指定所有关于作者的信息（用户名，电子邮件，...）。有没有办法只使用CREATE / UPDATE操作的用户链接？或修改一些东西，以便用户ID足够填写这个字段？

解决方案

如果我正确理解你的问题，在您检索数据时扩展作者，只需要在更新和创建时发送ID或URL。

1＃它不是关于任何指导，而是完全取决于你对 api 将如何使用的要求。

2＃所以你需要扩展你的 UserSerializer 并覆盖 to_internal_value 。示例代码可能看起来像

  class MyCustomSerializer（UserSerializer）：
 def to_internal_value（self，data）：
＃数据必须是有效的user-detail url 
 return serializers.HyperLinkedRelatedField（queryset = User.objects.all（），view_name ='user-detail'）to_internal_value（data）
  

通知，您必须拥有能够使用HyperLinkedRelatedField的用户详细信息的端点。

所以如果你想要发送 ID ，那么示例代码可能看起来像

$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ PrimaryKeyRelatedField（queryset = User.objects.all（））to_internal_value（data）

不过我想在 POST / PUT / PATCH 中发送ForeignKey字段的一致性。 （始终是URL或ID）。

然后在您的代码中使用它，如

  class PostSerializer（serializers.HyperlinkedModelSerializer）：
 author = MyCustomSerializer（）
 
 class Meta：
 model = Post 
  

请参阅可写嵌套序列化程序到嵌套资源的 POST 。

I'm building a REST web API using the Django REST Framework. Things are going great, but I have, however stumbled upon a problem with nested resources. At first, all relationships in the REST API were hyperlinked. A post, for example, looked like this:

{
    "path": "http://api.myproject.com/posts/1.json",
    "id": 1,
    "author": "http://api.myproject.com/users/broak.json",
    "image": "/images/posts/cxyUzlPo.jpg",
    "header": "Who am I?",
    "footer": "I am a champion!",
    "date": "2014-11-09 15:16",
    "likes": "http://api.myproject.com/posts/1/likes.json",
    "comments": "http://api.myproject.com/posts/1/comments.json",
    "likes_count": 0,
    "comments_count": 0
}

The relationship between the post and the author (user) is hyperlinked. When you want to create a new post, you need to specify a correct hyperlink to a specific user - this works fine.

When calling a list of posts, things become inefficient, because you have to make an extra API call for every author for every post. I solved this by using NESTED resources instead of HYPERLINKED resources, so every post now contains all the information about the author.

{
    "path": "http://api.myproject.com/posts/1.json",
    "id": 1,
    "author": {
        "email": "[e-mail]"
        "username": "broak",
        "first_name: "John",
        "last_name": "Broak",
        "is_staff": False,
        "is_active": True,
        "last_login": "02-26-2016"
    },
    "image": "/images/posts/cxyUzlPo.jpg",
    "header": "Who am I?",
    "footer": "I am a champion!",
    "date": "2014-11-09 15:16",
    "likes": "http://api.myproject.com/posts/1/likes.json",
    "comments": "http://api.myproject.com/posts/1/comments.json",
    "likes_count": 0,
    "comments_count": 0
}

My first question is: do you have a guideline, whether I should create a nested data structure or a separate endpoint with hyperlink to it.

My second question is: when I use author as a nested resource and want to create a new post, I don't want to specify all the information about the author (username, e-mail, ...). Is there any way to just use a link to a user for the CREATE/UPDATE operation? Or modify something so that the user ID is enough to fill in this field?

解决方案

If i understood your question correctly, You want to have expanded author while you are retrieving the data and just want to send ID or URL in case of update and create.

1# It is not about any guideline and it totally depends on your requirement of how your api is going to be used.

2# So you need to extend your UserSerializer and override to_internal_value. Sample code might look like

class MyCustomSerializer(UserSerializer):
    def to_internal_value(self, data):
        # data must be valid user-detail url
        return serializers.HyperLinkedRelatedField(queryset=User.objects.all(), view_name='user-detail').to_internal_value(data)

Notice that you must have a Endpoint for user-detail in able to work with HyperLinkedRelatedField.

So If you want to be able to send ID then sample code might look like

class MyCustomSerializer(UserSerializer):
    # data must be valid user id
    def to_internal_value(self, data):
        return serializers.PrimaryKeyRelatedField(queryset=User.objects.all()).to_internal_value(data)

However i would like to keep consistency in sending ForeignKey field in POST/PUT/PATCH. (Always Either URL or ID).

then use it in your code like

class PostSerializer(serializers.HyperlinkedModelSerializer):
    author = MyCustomSerializer()

    class Meta:
        model = Post

Please see the documentation on Writable nested serializers to POST on a Nested resource.

这篇关于Django REST框架：何时创建超链接资源和嵌套资源？如何发布嵌套资源？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！