序列化组件之MoelSerializer

from rest_framework import serializers

serializers.py文件中有  Serializer   ModelSerializer  ListSerializer

今天我们研究ModelSerializer(Serializer)   它是继承Serializer的，在其之上重写了create update等方法

ModelSerializer的用法也有区别，下面开始研究  

课程准备

配置:settings.py

INSTALLED_APPS = [
    # ...
    'rest_framework',
]

DATABASES = {
    'default': {
        'ENGINE': 'django.db.backends.mysql',
        'NAME': 'dg_proj',
        'USER': 'root',
        'PASSWORD': '123',
    }
}
"""
任何__init__文件
import pymysql
pymysql.install_as_MySQLdb()
"""

LANGUAGE_CODE = 'zh-hans'
TIME_ZONE = 'Asia/Shanghai'
USE_I18N = True
USE_L10N = True
USE_TZ = False

MEDIA_URL = '/media/'
MEDIA_ROOT = os.path.join(BASE_DIR, 'media')

 路由

# 主
from django.conf.urls import url, include
from django.contrib import admin
from django.views.static import serve
from django.conf import settings
urlpatterns = [
    url(r'^admin/', admin.site.urls),
    url(r'^api/', include('api.urls')),
    url(r'^media/(?P<path>.*)', serve, {'document_root': settings.MEDIA_ROOT}),
]

# 子
from django.conf.urls import url
from . import views
urlpatterns = [
    
]

多表设计

"""
Book表：name、price、img、authors、publish、is_delete、create_time

Publish表：name、address、is_delete、create_time
    
Author表：name、age、is_delete、create_time

AuthorDetail表：mobile, author、is_delete、create_time
    
BaseModel基表
    is_delete、create_time
上面四表继承基表，可以继承两个字段
"""

基表

class BaseModel(models.Model):
    is_delete = models.BooleanField(default=False)
    create_time = models.DateTimeField(auto_now_add=True)

    # 设置 abstract = True 来声明基表，作为基表的Model不能在数据库中形成对应的表
    class Meta:
        abstract = True

段关联多表关系 (知识点  重点   之后我会从中分开  单独去整理ORM的集合知识点)

"""
1、外键位置：
    一对多 - 外键放多的一方
    一对一 - 从逻辑正反向考虑，如作者表与作者详情表，作者删除级联删除详情，详情删除作者依旧存在，所以建议外键在详情表中
    多对多 - 外键在关系表中
    
2、ORM正向方向连表查找：
    正向：通过外键字段 eg: author_detial_obj.author 
    反向：通过related_name的值 eg：author_obj.detail
    注：依赖代码见下方
    
3、连表操作关系：
    1）作者删除，详情级联 - on_delete=models.CASCADE
    2）作者删除，详情置空 - null=True, on_delete=models.SET_NULL
    3）作者删除，详情重置 - default=0, on_delete=models.SET_DEFAULT
    4）作者删除，详情不动 - on_delete=models.DO_NOTHING
    注：拿作者与作者详情表举例
    
4、外键关联字段的参数 - 如何实现 断关联、目前表间操作关系、方向查询字段
    i）作者详情表中的
    author = models.OneToOneField(
        to='Author',
        related_name='detail',
        db_constraint=False,
        on_delete=models.CASCADE
    )
    
    ii）图书表中的
    publish = models.ForeignKey(
        to='Publish',
        related_name='books',
        db_constraint=False,
        on_delete=models.DO_NOTHING,
    )
    authors = models.ManyToManyField(
        to='Author'
        related_name='books',
        db_constraint=False,
    )
    注：ManyToManyField不能设置on_delete，OneToOneField、ForeignKey必须设置on_delete（django1.x系统默认级联，但是django2.x必须手动明确）
"""

模型类:model.py

from django.db import models

# 图书管理系统：Book、Author、AuthorDetail、Publish
"""
Book表: name、price、img、authors、publish、is_delete、create_time
Publish表: name、address、is_delete、create_time
Author表: name、age、is_delete、create_time
AuthorDetail表: mobile, author、is_delete、create_time
"""

# 1) 基表
class BaseModel(models.Model):
    is_delete = models.BooleanField(default=False)
    create_time = models.DateTimeField(auto_now_add=True)

    # 作为基表的Model不能在数据库中形成对应的表，设置 abstract = True
    class Meta:
        abstract = True


class Book(BaseModel):
    """name、price、img、authors、publish、is_delete、create_time"""
    name = models.CharField(max_length=64)
    price = models.DecimalField(max_digits=5, decimal_places=2)
    img = models.ImageField(upload_to='img', default='img/default.jpg')
    publish = models.ForeignKey(
        to='Publish',
        db_constraint=False,  # 断关联
        related_name='books',  # 反向查询字段：publish_obj.books 就能访问所有出版的书
        on_delete=models.DO_NOTHING,  # 设置连表操作关系
    )
    authors = models.ManyToManyField(
        to='Author',
        db_constraint=False,
        related_name='books'
    )

    # 序列化插拔式属性 - 完成自定义字段名完成连表查询
    @property
    def publish_name(self):
        return self.publish.name

    @property
    def author_list(self):
        return self.authors.values('name', 'age', 'detail__mobile').all()

    class Meta:
        db_table = 'book'
        verbose_name = '书籍'
        verbose_name_plural = verbose_name
    def __str__(self):
        return self.name

class Publish(BaseModel):
    """name、address、is_delete、create_time"""
    name = models.CharField(max_length=64)
    address = models.CharField(max_length=64)

    class Meta:
        db_table = 'publish'
        verbose_name = '出版社'
        verbose_name_plural = verbose_name
    def __str__(self):
        return self.name

class Author(BaseModel):
    """name、age、is_delete、create_time"""
    name = models.CharField(max_length=64)
    age = models.IntegerField()

    class Meta:
        db_table = 'author'
        verbose_name = '作者'
        verbose_name_plural = verbose_name
    def __str__(self):
        return self.name

class AuthorDetail(BaseModel):
    """mobile, author、is_delete、create_time"""
    mobile = models.CharField(max_length=11)
    author = models.OneToOneField(
        to='Author',
        db_constraint=False,
        related_name='detail',
        on_delete=models.CASCADE,
    )

    class Meta:
        db_table = 'author_detail'
        verbose_name = '作者详情'
        verbose_name_plural = verbose_name
    def __str__(self):
        return '%s的详情' % self.author.name

下面开始正题  ModelSerializer的序列化与反序列化整合   之十大接口    （重点）

路由层

urlpatterns = [
    url(r'^v2/books/$', views.V2Book.as_view()),
    url(r'^v2/books/(?P<pk>.*)/$', views.V2Book.as_view()),
]

序列化层 : serializers.py

在当前项目app01文件夹下创建一个 serializers.py文件

"""
1) fields中设置所有序列化与反序列化字段
2) extra_kwargs划分只序列化或只反序列化字段
    write_only：只反序列化
    read_only：只序列化
    自定义字段默认只序列化(read_only)
3) 设置反序列化所需的 系统、局部钩子、全局钩子 等校验规则
"""
class V2BookModelSerializer(ModelSerializer):
    class Meta:
        model = models.Book
        fields = ('name', 'price', 'img', 'author_list', 'publish_name', 'publish', 'authors')
        extra_kwargs = {
            'name': {
                'required': True,
                'min_length': 1,
                'error_messages': {
                    'required': '必填项',
                    'min_length': '太短',
                }
            },
            'publish': {
                'write_only': True
            },
            'authors': {
                'write_only': True
            },
            'img': {
                'read_only': True,
            },
            'author_list': {
                'read_only': True,
            },
            'publish_name': {
                'read_only': True,
            }
        }

    def validate_name(self, value):
        # 书名不能包含 g 字符
        if 'g' in value.lower():
            raise ValidationError('该g书不能出版')
        return value

    def validate(self, attrs):
        publish = attrs.get('publish')
        name = attrs.get('name')
        if models.Book.objects.filter(name=name, publish=publish):
            raise ValidationError({'book': '该书已存在'})
        return attrs

视图层: views.py

查询两大接口：

1.单查  http://127.0.0.1:8000/app01/v2/books/2/   2.群查  http://127.0.0.1:8000/app01/v2/books/       判断条件：有无pk值

class V2Book(APIView):
    # 单查：有pk
    # 群查：无pk
    def get(self, request, *args, **kwargs):
        pk = kwargs.get('pk')
        if pk:
            try:
                book_obj = models.Book.objects.get(pk=pk, is_delete=False)
                book_data = serializers.V2BookModelSerializer(book_obj).data
            except:
                return Response({
                    'status': 1,
                    'msg': '书籍不存在'
                })
        else:
            book_query = models.Book.objects.filter(is_delete=False).all()
            book_data = serializers.V2BookModelSerializer(book_query, many=True).data
        return Response({
            'status': 0,
            'msg': 'ok',
            'results': book_data
        })

新增两大接口：

1.单增     2.群增          http://127.0.0.1:8000/app01/v2/books/ 

通过JSON数据包方式提交post

单增的

{
        "name":"红楼3梦",
        "price":8.88,
        "publish":2,
        "authors":[1,2]
}

群增的

[
    {
            "name":"红楼3梦",
            "price":8.88,
            "publish":2,
            "authors":[1,2]
    },
    {
            "name":"红楼1梦",
            "price":8.88,
            "publish":2,
            "authors":[1,2]
    }
]

具体代码：

    # 单增：传的数据是与model对应的字典
    # 群增：传的数据是 装多个 model对应字典 的列表
    def post(self, request, *args, **kwargs):
        request_data = request.data
        if isinstance(request_data, dict):
            many = False
        elif isinstance(request_data, list):
            many = True
        else:
            return Response({
                'status': 1,
                'msg': '数据有误',
            })
        book_ser = serializers.V2BookModelSerializer(data=request_data, many=many)
        # 当校验失败，马上终止当前视图方法，抛异常返回给前台
        book_ser.is_valid(raise_exception=True)
        book_result = book_ser.save()
        return Response({
            'status': 0,
            'msg': 'ok',
            'results': serializers.V2BookModelSerializer(book_result, many=many).data
        })

删除两大接口：

1.单删     2.群删

 # 单删：有pk
    # 群删：有pks   |  {"pks": [1, 2, 3]}
    def delete(self, request, *args, **kwargs):
        pk = kwargs.get('pk')
        if pk:
            pks = [pk]
        else:
            pks = request.data.get('pks')
        if models.Book.objects.filter(pk__in=pks, is_delete=False).update(is_delete=True):
            return Response({
                'status': 0,
                'msg': '删除成功',
            })
        return Response({
            'status': 1,
            'msg': '删除失败',
        })

整体单改

注意点：

"""
1) 单整体改，说明前台要提供修改的数据，那么数据就需要校验，校验的数据应该在实例化“序列化类对象”时，赋值给data
2）修改，就必须明确被修改的模型类对象，并在实例化“序列化类对象”时，赋值给instance
3）整体修改，所有校验规则有required=True的字段，都必须提供，因为在实例化“序列化类对象”时，参数partial默认为False


注：如果partial值设置为True，就是可以局部改
1）单整体修改，一般用put请求：
V2BookModelSerializer(
    instance=要被更新的对象, 
    data=用来更新的数据,
    partial=默认False，必须的字段全部参与校验
)
2）单局部修改，一般用patch请求：
V2BookModelSerializer(
    instance=要被更新的对象, 
    data=用来更新的数据,
    partial=设置True，必须的字段都变为选填字段
)
    注：partial设置True的本质就是使字段 required=True 校验规则失效
"""

具体代码：

class V2Book(APIView):
    # 单整体改: 对 v2/books/(pk)/ 传的数据是与model对应的字典{name|price|publish|authors}
    def put(self, request, *args, **kwargs):
        request_data = request.data
        pk = kwargs.get('pk')
        old_book_obj = models.Book.objects.filter(pk=pk).first()
        # 目的：将众多数据的校验交给序列化类来处理 - 让序列化类扮演反序列化角色，校验成功后，序列化类来帮你入库
        book_ser = serializers.V2BookModelSerializer(instance=old_book_obj, data=request_data, partial=False)
        book_ser.is_valid(raise_exception=True)
        # 校验通过，完成数据的更新：要更新的目标，用来更新的新数据
        book_obj = book_ser.save()

        return Response({
            'status': 0,
            'msg': 'ok',
            'results': serializers.V2BookModelSerializer(book_obj).data
        })

单与整体局部修改

序列化层:serializers.py

# 重点：ListSerializer与ModelSerializer建立关联的是：
# ModelSerializer的Meta类的 - list_serializer_class
class V2BookListSerializer(ListSerializer):
    def update(self, instance, validated_data):
        # print(instance)  # 要更新的对象们
        # print(validated_data)  # 更新的对象对应的数据们
        # print(self.child)  # 服务的模型序列化类 - V2BookModelSerializer
        for index, obj in enumerate(instance):
            self.child.update(obj, validated_data[index])
        return instance
    
# 原模型序列化类变化
class V2BookModelSerializer(ModelSerializer):
    class Meta:
        # ...
        # 群改，需要设置 自定义ListSerializer，重写群改的 update 方法
        list_serializer_class = V2BookListSerializer
    # ...

视图层：views.py

class V2Book(APIView):
    # 单局部改：对 v2/books/(pk)/ 传的数据，数据字段key都是选填
    # 群局部改：对 v2/books/ 
    # 请求数据 - [{pk:1, name:123}, {pk:3, price:7}, {pk:7, publish:2}]
    def patch(self, request, *args, **kwargs):
        request_data = request.data
        pk = kwargs.get('pk')

        # 将单改，群改的数据都格式化成 pks=[要需要的对象主键标识] | request_data=[每个要修改的对象对应的修改数据]
        if pk and isinstance(request_data, dict):  # 单改
            pks = [pk, ]
            request_data = [request_data, ]
        elif not pk and isinstance(request_data, list): # 群改
            pks = []
            for dic in request_data:  # 遍历前台数据[{pk:1, name:123}, {pk:3, price:7}, {pk:7, publish:2}]，拿一个个字典
                pk = dic.pop('pk', None)
                if pk:
                    pks.append(pk)
                else:
                    return Response({
                        'status': 1,
                        'msg': '数据有误',
                    })
        else:
            return Response({
                'status': 1,
                'msg': '数据有误',
            })

        # pks与request_data数据筛选，
        # 1）将pks中的没有对应数据的pk与数据已删除的pk移除，request_data对应索引位上的数据也移除
        # 2）将合理的pks转换为 objs
        objs = []
        new_request_data = []
        for index, pk in enumerate(pks):
            try:
                # pk对应的数据合理，将合理的对象存储
                obj = models.Book.objects.get(pk=pk)
                objs.append(obj)
                # 对应索引的数据就需要保存下来
                new_request_data.append(request_data[index])
            except:
                # 重点：反面教程 - pk对应的数据有误，将对应索引的data中request_data中移除
                # index = pks.index(pk)
                # request_data.pop(index)
                continue

        book_ser = serializers.V2BookModelSerializer(instance=objs, data=new_request_data, partial=True, many=True)
        book_ser.is_valid(raise_exception=True)
        book_objs = book_ser.save()

        return Response({
            'status': 0,
            'msg': 'ok',
            'results': serializers.V2BookModelSerializer(book_objs, many=True).data
        })

 总结 :不管是群增 还是 群改  都需要设置属性 many=True  是通过ModelSerializer的父类的父类BaseSerializer中many_init实现的

具体代码如下:

    def __new__(cls, *args, **kwargs):
        # We override this method in order to automagically create
        # `ListSerializer` classes instead when `many=True` is set.
        if kwargs.pop('many', False):
            return cls.many_init(*args, **kwargs)
        return super().__new__(cls, *args, **kwargs)

    @classmethod
    def many_init(cls, *args, **kwargs):
        """
        This method implements the creation of a `ListSerializer` parent
        class when `many=True` is used. You can customize it if you need to
        control which keyword arguments are passed to the parent, and
        which are passed to the child.

        Note that we're over-cautious in passing most arguments to both parent
        and child classes in order to try to cover the general case. If you're
        overriding this method you'll probably want something much simpler, eg:

        @classmethod
        def many_init(cls, *args, **kwargs):
            kwargs['child'] = cls()
            return CustomListSerializer(*args, **kwargs)
        """
        allow_empty = kwargs.pop('allow_empty', None)
        child_serializer = cls(*args, **kwargs)
        list_kwargs = {
            'child': child_serializer,
        }
        if allow_empty is not None:
            list_kwargs['allow_empty'] = allow_empty
        list_kwargs.update({
            key: value for key, value in kwargs.items()
            if key in LIST_SERIALIZER_KWARGS
        })
        meta = getattr(cls, 'Meta', None)
        list_serializer_class = getattr(meta, 'list_serializer_class', ListSerializer)
        return list_serializer_class(*args, **list_kwargs)

其中重要的是下面两句代码：

list_serializer_class = getattr(meta, 'list_serializer_class', ListSerializer)
     return list_serializer_class(*args, **list_kwargs)

从中映射到ListSerializer类  而ListSerializer类中有update,和create方法 这就是实现群增和群删的关键点

    def update(self, instance, validated_data):
        raise NotImplementedError(
            "Serializers with many=True do not support multiple update by "
            "default, only multiple create. For updates it is unclear how to "
            "deal with insertions and deletions. If you need to support "
            "multiple update, use a `ListSerializer` class and override "
            "`.update()` so you can specify the behavior exactly."
        )

    def create(self, validated_data):
        return [
            self.child.create(attrs) for attrs in validated_data
        ]

群增create方法是通过遍历之后再去一个个调用单增， 而群改update方法需要我们手动去重写 因此我们就可以模仿create方法去这样写.