drf序列化高级自定义只读只写序列化覆盖字段二次封装Response数据库查询优化(断关联)十大接口视图家族自动补全图片链接

Posted guyouyin123

tags:

篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了drf序列化高级自定义只读只写序列化覆盖字段二次封装Response数据库查询优化(断关联)十大接口视图家族自动补全图片链接相关的知识,希望对你有一定的参考价值。

自定义只读

自定义只读:Model类中
@property
def 字段(self): return "字段值"

models.py

class Book(BaseModel):
    name = models.CharField(max_length=64)
    price = models.DecimalField(max_digits=5, decimal_places=2, null=True)
    image = models.ImageField(upload_to='img', default='img/default.png')

    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)

    @property  # @property字段默认就是read_only,且不允许修改
    def publish_name(self):
        return self.publish.name

自定义只写

自定义只写:Serializer类中
字段 = serializers.字段类型(write_only=True)

serializers.py

class BookModelSerializer(serializers.ModelSerializer):
    class Meta:
        list_serializer_class = BookListSerializer
        model = models.Book
        fields = ['name', 'price', 'image', 'publish', 'authors', 'publish_name', 'author_list']
        extra_kwargs = {
            'publish': {
                'write_only': True
            }

序列化覆盖字段

覆盖字段:Serializer类中,规则在()中规定,可以是只读、只写、可读可写任何形式
系统同名字段 = serializers.字段类型()

# 如何覆盖外键字段
    publish = serializers.PrimaryKeyRelatedField(read_only=True)  # 只读
    # publish = serializers.PrimaryKeyRelatedField(queryset=models.Publish.objects.all(), write_only=True)  # 只写
    # publish = serializers.PrimaryKeyRelatedField(queryset=models.Publish.objects.all())  # 可读可写

    # publish = serializers.PrimaryKeyRelatedField(queryset=models.Publish.objects.all())
    # authors = serializers.PrimaryKeyRelatedField(queryset=models.Author.objects.all(), many=True)
    
class Meta:
        model = models.Book
        fields = ('name',  'publish', )
        extra_kwargs = {
            'publish': {  # 系统原有的外键字段,要留给反序列化过程使用,序列化外键内容,用@property自定义
                'write_only': True,
            },

二次封装Response

from rest_framework.response import Response


class APIResponse(Response):
    def __init__(self,  status=0, msg='ok', http_status=None, headers=None, exception=False, **kwargs):
        data = {
            'status': status,
            'msg': msg
        }
        # 在外界可以存值
        if kwargs:
            data.update(kwargs)
        super().__init__(data=data, status=http_status, headers=headers, exception=exception)

数据库关系分析

"""
1)之间有关系的两个表,增删改操作会相互影响(效率低),查询操作就是正常的连表操作
2)之间有关系的两个表,断开关联,但所有数据保持与原来一致
    每个表都可以单独操作,增删改操作效率极高,但是容易出现脏数据(开发中完全可以避免)
    由于数据没有任何变化,所以查询的连表操作不会受到任何影响
    
3)Django的ORM支持断关联操作关系表,且所有的操作方式和没有断关联操作一致
"""

断外键关联关系

有关联的表,如果直接存在关联关系(外键),增删改操作效率低(但是安全);删除外键(断关联),可以提高增删改操作效率,但是要手动处理数据安全问题(避免脏数据的参数)
db_constraint=False

eg:
# 关系字段
    brand = models.ForeignKey(to='Brand', related_name='Cars', db_constraint=False)
    sponsors = models.ManyToManyField(to="Sponsor", related_name="Cars", db_constraint=False)

ORM操作外键关系

外键位置

"""
外键位置:
1)一对多:ForeignKey必须放在多的一方,书与出版社,外键应该放在书表
2)多对多:ManyToManyField放在任何一方都可以,因为会创建关系表,在关系表中用两个外键分别关联两个表
3)一对一:OneToOneField放在依赖的表,作者与作者详情,放在详情表,OneToOneField会被转换为 外键 + 唯一约束
"""

外键操作

ForeignKey可以设置:related_name, db_constraint, on_delete
OneToOneField可以设置:related_name, db_constraint, on_delete
ManyToManyField只能设置:related_name, db_constraint
        不能设置on_delete的原因:不管是关联的A表,还是B表,数据修改,都会影响到关系表(默认级联),
    如果想控制,只能自定义关系表,在关系表的两个外键分别设置on_delete

参数含义:

related_name:表之间反向访问的名字,默认是 表名小写|表名小写_set
db_constraint:表之间的关联关系,默认为True,代表关联,设置False,可以提高增删改的效率,且不影响查等
on_delete:表之间级联更新,在django 1.x下默认是CASCADE(级联更新),在django 2.x下必须手动明确
                                        DO_NOTHING(级联不更新)

案例:

"""
表关系:
作者没,作者详情一定没:CASCADE  *****
作者没,书还是该作者出的:DO_NOTHING
部门们,部门内的员工全部进入未分组部门:SET_DEFAULT (需要配合default属性使用)  
部门们,部门内的员工部门外键字段设置为空:SET_NULL (需要配合null=True属性使用)  *****
"""
class Author(models.Model):
    name = models.CharField(max_length=64)


class AuthorDetail(models.Model):
    phone = models.CharField(max_length=11)
    author = models.OneToOneField(
        to=Author,
        related_name='detail',
        db_constraint=False,
        on_delete=models.SET_NULL,
        null=True
    )

ORM四种关联关系

级联、不处理、置为默认值(配合默认值)、置空(配合可以为空)

基表

# 基类:是抽象的(不会完成数据库迁移),目的是提供共有字段的
# 抽出来的基类
class BaseModel(models.Model):
    name = models.CharField(max_length=64)
    is_delete = models.BooleanField(default=0)
    updated_time = models.DateTimeField(auto_now=True)

    class Meta:
        abstract = True  # 必须完成该配置
        
class Car(BaseModel):
    pass

# 其他的表继承了基类就有基类的属性了

序列化类其他配置(了解)

class AuthorModelSerializer(serializers.ModelSerializer):
    class Meta:
        model = models.Author
        # 不常用,将全部字段提供给外界
        fields = '__all__' 
        
# ------------------------------------------------------------------

class AuthorModelSerializer(serializers.ModelSerializer):
    class Meta:
        model = models.Author
        # 不常用,非,排除指定字段的其他所有字段,不能自动包含 外键反向 字段
        exclude = ['is_delete', 'updated_time']  
        
# ------------------------------------------------------------------

class AuthorModelSerializer(serializers.ModelSerializer):
    class Meta:
        model = models.Author
        # 'detail', 'books' 是 外键(正向|反向) 字段
        fields = ['name', 'detail', 'books']
        # 不常用,自动深度,自动深度会显示外键关联表的所有字段
        depth = 2  
# 正向外键字段:就是外键的属性名
# 反向外键字段:就是外键属性设置的related_name

十大接口

BaseSerializer初始化方法

from rest_framework.serializers import BaseSerializer
    def __init__(self, instance=None, data=empty, **kwargs):
        self.instance = instance
        if data is not empty:
            self.initial_data = data
        self.partial = kwargs.pop('partial', False)
        self._context = kwargs.pop('context', {})
        kwargs.pop('many', None)
        super().__init__(**kwargs)

'''
instance:赋值对象的
data:赋值数据的
many:数据是单个还是多个
partial:在修改需求时使用,可以将所有校验字段required校验规则设置为False
context:用于视图类和序列化类直接传参使用
'''

十大接口序列化总结

六个必备:单查、群查、单增、单删、单整体改(了解),单局部该
四个额外接口:群增、群删、群整体改、群局部改

# 常见使用
# 单查接口  get请求
UserModelSerializer(instance=user_obj)

# 群查接口  get请求
UserModelSerializer(instance=user_query, many=True)

# 单增接口,request.data是字典   post请求
UserModelSerializer(data=request.data) 

# 群增接口,request.data是列表   post请求
UserModelSerializer(data=request.data, many=True)

# 单整体改接口,request.data是字典    put请求
UserModelSerializer(instance=user_obj, data=request.data)

# 群整体改接口,request.data是列表,且可以分离出pks,转换成user_queryset put请求
UserModelSerializer(instance=user_queryset, data=request.data, many=True)

# 单局部改接口,request.data是字典    patch请求
UserModelSerializer(instance=user_obj, data=request.data, partial=True)

# 群局部改接口,request.data是列表,且可以分离出pks,转换成user_queryset patch请求
UserModelSerializer(instance=user_queryset, data=request.data, partial=True, many=True)

# 删接口,用不到序列化类  delete请求
models.Car.objects.filter(is_delete=False, pk__in=pks).update(is_delete=True)

单查群查

判断有无pk

class CarAPIView(APIView):
    def get(self, request, *args, **kwargs):
        pk = kwargs.get('pk')
        if pk:  # 单查
            try:
                obj = models.Car.objects.get(pk=pk)
                serializer = serializers.CarModelSerializer(obj)
                return APIResponse(result=serializer.data)
            except:
                return APIResponse(status=1, msg='pk error', http_status=status.HTTP_400_BAD_REQUEST, exception=True)

        else:  # 群查
            queryset = models.Car.objects.all()
            serializer = serializers.CarModelSerializer(queryset, many=True)
            return APIResponse(results=serializer.data)

技术图片

技术图片

单增群增

判断数据类型是否为列表

    def post(self, request, *args, **kwargs):
        # 单增
        if not isinstance(request.data, list):
            # 单增
            serializer = serializers.CarModelSerializer(data=request.data)
            serializer.is_valid(raise_exception=True)  # 如果校验失败,直接抛异常,返回给前台
            car_obj = serializer.save()
            return APIResponse(
                result=serializers.CarModelSerializer(car_obj).data,
                status=status.HTTP_201_CREATED
            )
        else:
            # 群增
            serializer = serializers.CarModelSerializer(data=request.data,many=True)
            serializer.is_valid(raise_exception=True)  # 如果校验失败,直接抛异常,返回给前台
            objs = serializer.save()

            return APIResponse(result=serializers.CarModelSerializer(objs, many=True).data, http_status=201)

技术图片

技术图片

单删群删

判断前端传过来的数据:是url参数就是单删,是data数据就是群删

只是更改了表示是否删除的字段

# 单删群删
"""
单删:接口:/books/(pk)/
群删:接口:/books/   数据:[pk1, ..., pkn]
"""
    def delete(self,request, *args,**kwargs):
        pk = kwargs.get('pk')
        if pk:
            pks = [pk]  # [1]  单删
        else:
            pks = request.data  # [2,3,4] 群删
        try:
            rows = models.Car.objects.filter(is_delete=False, pk__in=pks).update(is_delete=True)
        except:
            return APIResponse(1, '数据有误')
        if rows:
            return APIResponse(0, '删除成功')
        return APIResponse(2, '删除失败')

技术图片

技术图片

技术图片

单整体改/群整体改

# 单整体改群整体改
"""
单整体改:接口:/books/(pk)/ 数据:dict
群整体改:接口:/books/   数据:[{pk1, ...}, ..., {pkn, ...}] | {pks: [pk1, ..., pkn], data: [{}, ..., {}]}
"""
    def put(self,request, *args, **kwargs):
        pk = kwargs.get('pk')
        if pk:  # 单
            try:
                instance = models.Car.objects.get(is_delete=False, pk=pk)
            except:
                return APIResponse(1,'pk error', http_status=400)
            serializer = serializers.CarModelSerializer(instance=instance, data=request.data)
            serializer.is_valid(raise_exception=True)
            obj = serializer.save()  # 返回的是受影响的对象
            return APIResponse(result=serializers.CarModelSerializer(obj).data)
        else:  # 群
            pks = []
            try:
                for dic in request.data:
                    pks.append(dic.pop('pk'))
                objs = models.Car.objects.filter(is_delete=False, pk__in=pks) # 筛选  __in(或),双下线方法
                assert len(objs) == len(request.data)  # 两个列表长度必须一致  断言方法
            except:
                return APIResponse(1, '数据有误', http_status=400)

            serializer = serializers.CarModelSerializer(instance=objs, data=request.data, many=True)
            serializer.is_valid(raise_exception=True)
            objs = serializer.save()
            return APIResponse(result=serializers.CarModelSerializer(objs, many=True).data)

技术图片

技术图片

单局部改/群局部改

# 单局部改群局部改
"""
单局部改:接口:/books/(pk)/ 数据:dict
群局部改:接口:/books/   数据:[{pk1, ...}, ..., {pkn, ...}] | {pks: [pk1, ..., pkn], data: [{}, ..., {}]}
"""
    def patch(self,request, *args, **kwargs):
        pk = kwargs.get('pk')
        if pk:  # 单
            try:
                instance = models.Car.objects.get(is_delete=False, pk=pk)
            except:
                return APIResponse(1, 'pk error', http_status=400)
            # partial=True就是将所有反序列化字段的 required 设置为 False(提供就校验,不提供不校验)
            serializer = serializers.CarModelSerializer(
                instance=instance,
                data=request.data,
                partial=True
            )
            serializer.is_valid(raise_exception=True)
            obj = serializer.save()
            return APIResponse(result=serializers.CarModelSerializer(obj).data)
        else:
            pks = []
            try:
                for dic in request.data:
                    pks.append(dic.get('pk'))
                objs = models.Car.objects.filter(is_delete=False, pk__in=pks)
                assert len(objs) == len(request.data)
            except:
                return APIResponse(1, '数据有误', http_status=400)
            serializer = serializers.CarModelSerializer(
                instance=objs,
                data=request.data,
                many=True,
                partial=True,
            )
            serializer.is_valid(raise_exception=True)
            objs = serializer.save()
            return APIResponse(result=serializers.CarModelSerializer(objs, many=True).data)

技术图片

技术图片

群增群改配置

serializers.py:

# 群增群改辅助类(了解)
class BookListSerializer(serializers.ListSerializer):
    """
    1)create群增方法不需要重新
    2)update群改方法需要重写,且需要和views中处理request.data的逻辑配套使用
    3)self.child就代表该ListSerializer类绑定的ModelSerializer类
        BookListSerializer的self.child就是BookModelSerializer
    """
    # 重写update方法
    def update(self, queryset, validated_data_list):
        return [
            self.child.update(queryset[index], validated_data) for index, validated_data in enumerate(validated_data_list)
        ]
    
class CarModelSerializer(serializers.ModelSerializer):
        class Meta:
        # 配置自定义群增群改序列化类
        list_serializer_class = BookListSerializer

十大接口小结

"""
1)初始化序列化类,设置partial=True可以将所有反序列化字段的 required 设置为 False(提供就校验,不提供不校验),可以运用在局部修改接口

2)初始化序列化类,设置context={...},在序列化类操作self.context,实现视图类和序列化类数据互通

3)只有要完成资源的群改这种特殊需求时,才需要自定义ListSerializer绑定给自定义的ModelSerializer,重写update方法,来完成需求
"""

字段提供就校验,不提供拉到

实现单局部该/群局部改

技术图片

'required':False
# partial=True就是将所有反序列化字段的 required 设置为 False(提供就校验,不提供不校验)

DjangoORM内置优化机制:一次最多查21条数据

queryset = models.Book.objects.filter(is_delete=False).all()

models.py

from django.db import models


# 抽出来的基类
class BaseModel(models.Model):
    name = models.CharField(max_length=64)
    is_delete = models.BooleanField(default=0)
    updated_time = models.DateTimeField(auto_now=True)

    class Meta:
        abstract = True


class Car(BaseModel):
    price = models.DecimalField(max_digits=9, decimal_places=2)

    # 关系字段
    brand = models.ForeignKey(to='Brand', related_name='Cars', db_constraint=False)
    sponsors = models.ManyToManyField(to="Sponsor", related_name="Cars", db_constraint=False)
    class Meta:
        verbose_name_plural = '汽车表'

    def __str__(self):
        return self.name

    def brand_(self):
        return self.brand.name

    # 多对多跨表查询
    def sponsors_(self):
        temp_sponsor_list = []
        for sponsor in self.sponsors.all():
            sponsor_dic = {"name": sponsor.name}
            try:
                sponsor_dic['电话'] = sponsor.detail.phone
            except:
                sponsor_dic['电话'] = ''
            temp_sponsor_list.append(sponsor_dic)
        return temp_sponsor_list



class Brand(BaseModel):

    class Meta:
        verbose_name_plural = '品牌表'
    def __str__(self):
        return self.name

    # 一对多跨表查询
    def car_list(self):
        temp_car_list = []
        for car_obj in self.Cars.all():
            car_dic = {
                "name": car_obj.name
            }
            try:
                car_dic['price'] = car_obj.price
            except:
                car_dic['price'] = ''
            temp_car_list.append(car_dic)
        return temp_car_list


class Sponsor(BaseModel):

    class Meta:
        verbose_name_plural = '赞助商表'
    def __str__(self):
        return self.name



class SponsorDetail(models.Model):

    class Meta:
        verbose_name_plural = '赞助商详情表'

    phone = models.CharField(max_length=11)
    is_delete = models.BooleanField(default=1)
    updated_time = models.DateTimeField(auto_now=True)
    sponsor = models.OneToOneField(to="Sponsor", related_name="detail", db_constraint=False)

response.py

from rest_framework.response import Response


class APIResponse(Response):
    def __init__(self,  status=0, msg='ok', http_status=None, headers=None, exception=False, **kwargs):
        data = {
            'status': status,
            'msg': msg
        }
        # 在外界可以存值
        if kwargs:
            data.update(kwargs)
        super().__init__(data=data, status=http_status, headers=headers, exception=exception)

serializers.py

from rest_framework import serializers
from rest_framework import exceptions


from . import models

# 群增群改辅助类(了解)
class BookListSerializer(serializers.ListSerializer):
    """
    1)create群增方法不需要重新
    2)update群改方法需要重写,且需要和views中处理request.data的逻辑配套使用
    3)self.child就代表该ListSerializer类绑定的ModelSerializer类
        BookListSerializer的self.child就是BookModelSerializer
    """
    # 重写update方法
    def update(self, queryset, validated_data_list):
        return [
            self.child.update(queryset[index], validated_data) for index, validated_data in enumerate(validated_data_list)
        ]

class CarModelSerializer(serializers.ModelSerializer):

    def validate(self, attrs):
        print(self.context)  # 可以获得视图类在初始化序列化对象时传入的context
        # self.context.update({'a': 10})  # 序列化类内部更新context,传递给视图类
        return attrs

    class Meta:
        # 配置自定义群增群改序列化类
        list_serializer_class = BookListSerializer

        model = models.Car
        fields = ['name', 'is_delete', 'price', 'brand', 'brand_', 'updated_time', 'sponsors_']

        extra_kwargs = {
            'brand': {
                'write_only': True,
            }
            # 'name':{
            #     'required':False
            # },
            # 'price': {
            #     'required': False
            # },
        }


class BrandModelSerializer(serializers.ModelSerializer):
    class Meta:
        model = models.Brand
        fields = ['name', 'is_delete', 'updated_time', 'car_list']

urls.py

from django.conf.urls import url

from . import views
from django.views.static import serve
from d_car import settings
urlpatterns = [
    url(r'v1/car/$', views.CarAPIView.as_view()),
    url(r'v1/car/(?P<pk>d+)/$', views.CarAPIView.as_view()),

    url(r'brand/$', views.BrandAPIView.as_view()),
    url(r'brand/(?P<pk>d+)/$', views.BrandAPIView.as_view()),

    url(r'v2/car/$', views.CarV2APIView.as_view()),
    url(r'v2/car/(?P<pk>d+)/$', views.CarV2APIView.as_view()),

    # url(r'v3/car/$', views.CarV3APIView.as_view()),
    # url(r'v3/car/(?P<pk>d+)/$', views.CarV3APIView.as_view()),

    # 暴露文件夹
    url(r'^media/(?P<path>.*)', serve, {'document_root': settings.MEDIA_ROOT}),

]

视图家族

"""
视图基类:APIView、GenericAPIView
视图工具类:mixins包下的五个类(六个方法)
工具视图类:generics包下的所有GenericAPIView的子类
视图集:viewsets包下的类
"""
""" GenericAPIView基类(基本不会单独使用,了解即可,但是是高级视图类的依赖基础)
1)GenericAPIView继承APIView,所有APIView子类写法在继承GenericAPIView时可以保持一致
2)GenericAPIView给我们提供了三个属性 queryset、serializer_class、lookup_field
3)GenericAPIView给我们提供了三个方法 get_queryset、get_serializer、get_obj
"""
""" mixins包存放了视图工具类(不能单独使用,必须配合GenericAPIView使用)
CreateModelMixin:单增工具类
    create方法
    
ListModelMixin:群查工具类
    list方法

RetrieveModelMixin:单查工具类
    retrieve方法

UpdateModelMixin:单整体局部改工具类
    update方法

DestroyModelMixin:单删工具类
    destory方法
"""
""" generics包下的所有GenericAPIView的子类(就是继承GenericAPIView和不同mixins下的工具类的组合)
1)定义的视图类,继承generics包下已有的特点的GenericAPIView子类,可以在只初始化queryset和serializer_class两个类属性后,就获得特定的功能

2)定义的视图类,自己手动继承GenericAPIView基类,再任意组合mixins包下的一个或多个工具类,可以实现自定义的工具视图类,获得特定的功能或功能们

注:
i)在这些模式下,不能实现单查群查共存(可以加逻辑区分,也可以用视图集知识)
ii)DestroyModelMixin工具类提供的destory方法默认是从数据库中删除数据,所以一般删除数据的需求需要自定义逻辑
"""

视图基类 APIView、GenericAPIView

视图基类:APIView、GenericAPIView

i)继承APIView的,所以拥有APIView的全部
ii)三个类属性:queryset、serializer_class、lookup_field
iii)三个方法:get_queryset、get_serializer、get_object
过渡:单独继承GenericAPIView类的视图类,需要自己定义请求方法如get,还需要属性方法体完成请求

定义的视图类,自己手动继承GenericAPIView基类,再任意组合mixins包下的一个或多个工具类,可以实现自定义的工具视图类,获得特定的功能或功能们

# ----------------------------- 过渡写法:了解 -----------------------------

from rest_framework.generics import GenericAPIView
class BookV1APIView(GenericAPIView):
    # 将数据和序列化提示为类属性,所有的请求方法都可以复用
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer
    lookup_field = 'pk'  # 可以省略,默认是pk,与url有名分组对应的

    # 群查
    def get(self, request, *args, **kwargs):
        # queryset = models.Book.objects.filter(is_delete=False).all()  # => 方法+属性两行代码
        queryset = self.get_queryset()
        # serializer = serializers.BookModelSerializer(instance=queryset, many=True)  # => 方法+属性两行代码
        serializer = self.get_serializer(instance=queryset, many=True)
        return APIResponse(results=serializer.data)

    # 单查
    # def get(self, request, *args, **kwargs):
    #     obj = self.get_object()
    #     serializer = self.get_serializer(obj)
    #     return APIResponse(results=serializer.data)

    # 单增
    def post(self, request, *args, **kwargs):
        # serializer = serializers.BookModelSerializer(data=request.data)
        serializer = self.get_serializer(data=request.data)  # 同样的步骤多了,好处就来了
        serializer.is_valid(raise_exception=True)
        obj = serializer.save()
        return APIResponse(result=self.get_serializer(obj).data, http_status=201)

基于GenericAPIView的十大接口

views.py:

# 十大接口:
# 1)单查、群查、单增、单整体改、单局部改都可以直接使用
# 2)单删不能直接使用,因为默认提供的功能是删除数据库数据,不是我们自定义is_delete字段值修改
# 3)除了群查以为的接口,都要自己来实现

# 注:给序列化类context赋值{'request': request},序列化类就可以自动补全后台图片链接
from rest_framework.generics import GenericAPIView
from rest_framework import mixins
from . import models, serializers
from rest_framework.response import Response

class BookV1APIView(GenericAPIView,
                    mixins.RetrieveModelMixin,
                    mixins.ListModelMixin,
                    mixins.CreateModelMixin,
                    mixins.UpdateModelMixin):

    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer

    def get(self, request, *args, **kwargs):
        if 'pk' in kwargs:
            return self.retrieve(request, *args, **kwargs)  # 单查
        
        # queryset = models.Book.objects.filter(is_delete=False).all()
        # 注:给序列化类context赋值{'request': request},序列化类就可以自动补全后台图片链接
        # serializer = serializers.BookModelSerializer(queryset, many=True, context={'request': request})
        # return Response(serializer.data)
        return self.list(request, *args, **kwargs)  # 群查

    def post(self, request, *args, **kwargs):
        if not isinstance(request.data, list):
            return self.create(request, *args, **kwargs)

        serializer = self.get_serializer(data=request.data, many=True)
        serializer.is_valid(raise_exception=True)
        self.perform_create(serializer)
        headers = self.get_success_headers(serializer.data)
        return Response(serializer.data, status=201, headers=headers)

    def delete(self, request, *args, **kwargs):
        pk = kwargs.get('pk')
        if pk:
            pks = [pk]
        else:
            pks = request.data
        try:
            rows = models.Book.objects.filter(is_delete=False, pk__in=pks).update(is_delete=True)
        except:
            return Response(status=400)
        if rows:
            return Response(status=204)
        return Response(status=400)

    def put(self, request, *args, **kwargs):
        if 'pk' in kwargs:
            return self.update(request, *args, **kwargs)

        pks = []
        try:
            for dic in request.data:
                pks.append(dic.pop('pk'))
            objs = models.Book.objects.filter(is_delete=False, pk__in=pks)
            assert len(objs) == len(request.data)
        except:
            return Response(status=400)
        serializer = serializers.BookModelSerializer(instance=objs, data=request.data, many=True)
        serializer.is_valid(raise_exception=True)
        objs = serializer.save()
        return Response(serializers.BookModelSerializer(objs, many=True).data)

    def patch(self, request, *args, **kwargs):
        if 'pk' in kwargs:
            return self.partial_update(request, *args, **kwargs)

        pks = []
        try:
            for dic in request.data:
                pks.append(dic.pop('pk'))
            objs = models.Book.objects.filter(is_delete=False, pk__in=pks)
            assert len(objs) == len(request.data)
        except:
            return Response(status=400)
        serializer = serializers.BookModelSerializer(instance=objs, data=request.data, many=True, partial=True)
        serializer.is_valid(raise_exception=True)
        objs = serializer.save()
        return Response(serializers.BookModelSerializer(objs, many=True).data)

视图工具类:mixins包

from rest_framework import mixins

""" mixins包存放了视图工具类(不能单独使用,必须配合GenericAPIView使用)
CreateModelMixin:单增工具类
    create方法
    
ListModelMixin:群查工具类
    list方法

RetrieveModelMixin:单查工具类
    retrieve方法

UpdateModelMixin:单整体局部改工具类
    update方法

DestroyModelMixin:单删工具类
    destory方法
"""
# ----------------------------- 过渡写法:了解 -----------------------------

from rest_framework.generics import GenericAPIView
from rest_framework import mixins
class BookV2APIView(GenericAPIView, mixins.RetrieveModelMixin, mixins.CreateModelMixin):
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer

    # 单查
    def get(self, request, *args, **kwargs):
        # obj = self.get_object()
        # serializer = self.get_serializer(obj)
        # return APIResponse(results=serializer.data)

        # return self.retrieve(request, *args, **kwargs)

        response = self.retrieve(request, *args, **kwargs)
        return APIResponse(result=response.data)

    # 单增
    def post(self, request, *args, **kwargs):
        return self.create(request, *args, **kwargs)

    
# ----------------------------- 开发写法:常用 -----------------------------

from rest_framework.generics import RetrieveAPIView
class BookV3APIView(RetrieveAPIView):
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer

    # 单查
    pass

工具视图类 generics工具包:继承指定类

generics包:工具视图类
i)系统完成的GenericAPIView与mixins包下工具类的不同组合
只需要配置三个类属性

ii)自己完成GenericAPIView与mixins包下工具类的不同组合
需要配置三个类属性,需要书写请求方法,如post,内部直接调用self.create方法

iii)自己继承一堆generics包下的工具视图,完成组合
只需要配置三个类属性,但是单查、群查不能共存

工具

技术图片

继承指定的类,完成相应的功能

from rest_framework import generics  点进去看看又那些类

RetrieveAPIView   : 单查             内部:get   方法写好了
ListAPIView       :群查              内部:get  方法写好了
CreateAPIView     :单增              内部:post  方法
DestroyAPIView    : 单删             内部:delete    方法写好了
UpdateAPIView    : 单整体改,单局部改   内部:put、patch 方法写好了

ListCreateAPIView   : 群查,和单增  内部: get、post 方法写好了
ListCreateAPIView   : 群增            内部:post方法

。。。

单查:

from rest_framework.generics import RetrieveAPIView
class BookV3APIView(RetrieveAPIView):
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer

    # 单查  RetrieveAPIView内部已经写好了get方法
    pass

基于generics包下工具视图类的六大基础接口

六个必备:单查、群查、单增、单删、单整体改(了解),单局部该
四个额外接口:群增、群删、群整体改、群局部改

# 六大基础接口  
# 1)直接继承generics包下的工具视图类,可以完成六大基础接口
# 2)单查群查不能共存,因为都是get方法
# 3)单删一般会重写


views.py:

from . import models, serializers
from rest_framework.response import Response
from rest_framework import generics
class BookV2APIView(generics.ListAPIView,
                    generics.RetrieveAPIView,
                    generics.CreateAPIView,
                    generics.UpdateAPIView,
                    generics.DestroyAPIView):
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer

    def get(self, request, *args, **kwargs):  # 重写单查群查的get
        if 'pk' in kwargs:
            return self.retrieve(request, *args, **kwargs)
        return self.list(request, *args, **kwargs)

    def delete(self, request, *args, **kwargs):  # 重写删除(不是真的删除数据,是修改字段)
        pk = kwargs.get('pk')
        models.Book.objects.filter(is_delete=False, pk=pk).update(is_delete=True)
        return Response(status=204)

视图集(开发常用,最高级)

路由组件(繁琐,可以自定义映射关系)

urls.py

url(r'^v3/books/$', views.BookV3APIView.as_view(
    {'get': 'list', 'post': 'create', 'delete': 'multiple_destroy'}
    )),

url(r'^v3/books/(?P<pk>d+)/$', views.BookV3APIView.as_view(
        {'get': 'my_list'}
    )),

views.py

from rest_framework.viewsets import ModelViewSet
from rest_framework.response import Response
class BookV3APIView(ModelViewSet):
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer

    # 可以在urls.py中as_view({'get': 'my_list'})自定义请求映射
    def my_list(self, request, *args, **kwargs):
        return Response('ok')

路由组件(简化)

urls.py

from django.conf.urls import url, include
from . import views
# 路由组件,必须配合视图集使用
from rest_framework.routers import SimpleRouter
router = SimpleRouter()

# 以后就写视图集的注册即可:BookV3APIView和BookV4APIView都是视图集
router.register('v3/books', views.BookV3APIView, 'book')
router.register('v4/books', views.BookV4APIView, 'book')

urlpatterns = [
    url('', include(router.urls))
]

视图组件 views.py

from rest_framework.viewsets import ReadOnlyModelViewSet
class BookV4APIView(ReadOnlyModelViewSet):
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookModelSerializer

重写删除方法:destroy

因为继承的删除方法:是真的删除了数据

我们修改的:修改is_delete字段

# 需要完成字段删除,不是重写delete方法,而是重写destroy方法
    def destroy(self, request, *args, **kwargs):
        pk = kwargs.get('pk')
        models.Book.objects.filter(is_delete=False, pk=pk).update(is_delete=True)
        return Response(status=204)

群删接口

# 群删接口
    def multiple_destroy(self, request, *args, **kwargs):
        try:
            models.Book.objects.filter(is_delete=False, pk__in=request.data).update(is_delete=True)
        except:
            return Response(status=400)
        return Response(status=204)

视图级使用总结

六个必备:单查、群查、单增、单删、单整体改(了解),单局部该
四个额外接口:群增、群删、群整体改、群局部改
# 注:给序列化类context赋值{'request': request},序列化类就可以自动补全后台图片链接
"""
视图集的使用总结
1)可以直接继承ModelViewSet,实现六大继承接口(是否重写destroy方法,或其他方法,根据需求决定)
2)可以直接继承ReadOnlyModelViewSet,实现只读需求(只有单查群查)
3)继承ViewSet类,与Model类关系不是很密切的接口:登录的post请求,是查询操作;短信验证码发生接口,借助第三方平台
4)继承GenericViewSet类,就代表要配合mixins包,自己完成任意组合
5)继承以上4个视图集任何一个,都可以与路由as_view({映射})配合,完成自定义请求响应方法
"""

自定义路由组件(了解)

from rest_framework.routers import SimpleRouter as DrfSimpleRouter
from rest_framework.routers import Route, DynamicRoute

class SimpleRouter(DrfSimpleRouter):
    routes = [
        # List route.
        Route(
            url=r'^{prefix}{trailing_slash}$',
            mapping={
                'get': 'list',
                'post': 'create',  # 注:群增只能自己在视图类中重写create方法,完成区分
                'delete': 'multiple_destroy',  # 新增:群删
                'put': 'multiple_update',  # 新增:群整体改
                'patch': 'multiple_partial_update'  # 新增:群局部改
            },
            name='{basename}-list',
            detail=False,
            initkwargs={'suffix': 'List'}
        ),
        # Dynamically generated list routes. Generated using
        # @action(detail=False) decorator on methods of the viewset.
        DynamicRoute(
            url=r'^{prefix}/{url_path}{trailing_slash}$',
            name='{basename}-{url_name}',
            detail=False,
            initkwargs={}
        ),
        # Detail route.
        Route(
            url=r'^{prefix}/{lookup}{trailing_slash}$',
            mapping={
                'get': 'retrieve',
                'put': 'update',
                'patch': 'partial_update',
                'delete': 'destroy'
            },
            name='{basename}-detail',
            detail=True,
            initkwargs={'suffix': 'Instance'}
        ),
        # Dynamically generated detail routes. Generated using
        # @action(detail=True) decorator on methods of the viewset.
        DynamicRoute(
            url=r'^{prefix}/{lookup}/{url_path}{trailing_slash}$',
            name='{basename}-{url_name}',
            detail=True,
            initkwargs={}
        ),
    ]

上传文件接口

urls.py

from django.conf.urls import url, include
from . import views
# 路由组件,必须配合视图集使用
from rest_framework.routers import SimpleRouter
router = SimpleRouter()

# /books/image/(pk) 提交 form-data:用image携带图片
router.register('books/image', views.BookUpdateImageAPIView, 'book')

urlpatterns = [
    url('', include(router.urls))
]

serializers.py

class BookUpdateImageModelSerializer(serializers.ModelSerializer):
    class Meta:
        model = models.Book
        fields = ['image']

views.py

# 上次文件 - 修改头像 - 修改海报
from rest_framework.viewsets import GenericViewSet
from rest_framework import mixins
class BookUpdateImageAPIView(GenericViewSet, mixins.UpdateModelMixin):
    queryset = models.Book.objects.filter(is_delete=False).all()
    serializer_class = serializers.BookUpdateImageModelSerializer

图片自动补全后台图片链接

queryset = models.Book.objects.filter(is_delete=False).all()
        # 注:给序列化类context赋值{'request': request},序列化类就可以自动补全后台图片链接
serializer = serializers.BookModelSerializer(queryset, many=True, context={'request': request})
return Response(serializer.data)

以上是关于drf序列化高级自定义只读只写序列化覆盖字段二次封装Response数据库查询优化(断关联)十大接口视图家族自动补全图片链接的主要内容,如果未能解决你的问题,请参考以下文章

序列化组件三,视图函数家族

序列化类外键字段的覆盖,十大接口序列化总结,视图家族

DRF:Djoser 覆盖自定义用户序列化程序

序列化 DRF 中的自定义相关字段

drf框架4-views视图家族操作

DRF--验证