Django 模型层多表操作
一、Django多表ORM设计规则
'''1. 关联的表之间建议建立外键，但可以取消关联关系(db_constraint=False)2. 关联表之间的外键字段建议采用对应类名的全小写3. 采用关联表的主键或对象均能进行操作'''
设计Models
# 一对多：出版社（一） 书籍（多，外键在多的一方，依赖于出版社）# 一对一：作者详情（一） 作者（一，外键在任意一方均可，一旦外键放在作者中，作者依赖于作者详情）# 多对多：作者（多）书籍（多）建立关系表（存放两个表的外键信息 => 将建表转化为关系对应字段）# Book书籍：id name price publish_date publish(publish_id)class Book(models.Model): id = models.AutoField(primary_key=True) name = models.CharField(max_length=20) price = models.DecimalField(max_digits=5, decimal_places=2) publish_date = models.DateField() publish = models.ForeignKey(to='Publish', to_field='id') # 多对多关系字段，该字段不会再book表中形成字段，是用来创建关系表的 author = models.ManyToManyField(to='Author')# Author作者：id nameclass Author(models.Model): id = models.AutoField(primary_key=True) name = models.CharField(max_length=20) # author_detail = models.ForeignKey(to='AuthorDetail', to_field='id', unique=True) author_detail = models.OneToOneField(to='AuthorDetail', to_field='id')# AuthorDetail作者详情: id age telephone infoclass AuthorDetail(models.Model): id = models.AutoField(primary_key=True) age = models.IntegerField() telephone = models.IntegerField() # 存大文本 info = models.TextField()# Publish出版社：id name addressclass Publish(models.Model): id = models.AutoField(primary_key=True) name = models.CharField(max_length=20) address = models.CharField(max_length=64)
二、一对多关系 Book 依赖于 Publish
规则：
'''1. 关系中 『多』 依赖于 『一』2. Django 1.x外键关联默认有级联删除，2.x需要手动明确外键的级联删除(on_delete=models.CASCADE)'''
1、增# 主键操作，外键字段为关联对象主键publish = Publish.objects.create(name='小女孩出版社', address='山东')publish_id = publish.idBook.objects.create(name='死亡Python', price=66.66, publish_date='2018-8-8', publish_id=publish_id)# 对象操作，外键对象为关联对象publish = Publish.objects.first()Book.objects.create(name='inPingMei', price=888.88, publish_date='1970-1-1', publish=publish)2、删# 默认有级联删除，关联数据贵全被删除book = Book.objects.last() # type: Bookbook.delete()Publish.objects.first().delete()3、改publish = Publish.objects.last()book = Book.objects.last() # type: Bookbook.publish = publishbook.save()publish_id = Publish.objects.last().idBook.objects.filter(name='死亡Python').update(publish_id=publish_id)
三、一对一
规则：通过外键所在表决定依赖关系
Author作者：id nameAuthorDetail作者详情: id age telephone info一对一关系 Author 依赖于 AuthorDetail
1、增detail = AuthorDetail.objects.create(age=8, telephone=13766668888, info='一个字，帅')Author.objects.create(name="Owen", author_detail=detail)Author.objects.create(name="Liuxx", author_detail_id=detail.id) # 三句一起执行，两个作者使用一条详情，不满足唯一约束detail = AuthorDetail.objects.create(age=88, telephone=13722222222, info='一个字，二')Author.objects.create(name="Liuxx", author_detail_id=detail.id)2、删：拥有级联删除AuthorDetail.objects.last().delete()3、改：一对一关系不考虑连表更新
四、多对多
规则：
'''1. 多对多关系存在关系表，关系表建议采用ManyToManyField字段处理2. 需要手动创建关系表时，在字段中明确through与through_field值'''
# 多对多：Book 与 Author表均可以单独操作，关联关系存在于 关系表# 问题：如果获取关系表，通过Book的对象获得author属性，代表关系表的对象# book = Book.objects.first() # type: Book# book.author 代表关系表的对象
detail = AuthorDetail.objects.create(age=88, telephone=13744444444, info='一个字，衰')Author.objects.create(name="Egon", author_detail_id=detail.id)b1 = Book.objects.first() # type: Bookb2 = Book.objects.all()[1] # type: Booka1 = Author.objects.first() # type: Authora2 = Author.objects.all()[1] # type: Author1、增 add()book.author.add(*authors)b1.author.add(a1.id, a2.id)b2.author.add(a1, a2)2、删将第二本书的关系记录全部清除b2.author.clear()通过 主键 或 对象 操作指定某条记录b1.author.remove(a1)b1.author.remove(a2.id)3、改清除该本书的之前所有关系记录，添加新的关系b1.author.clear()b1.author.add(a1)去除新数据中不存在的值，添加新数据中新有的值，与之前重复的值保留b2.author.set([a2.id])a3 = Author.objects.last()b2.author.set([a2.id, a3.id])
连表查询规则
- 正向逆向概念：从存放外键的表到关系表称之为正向跨表查询，反之称之为逆向查询
- 正向查询通过外键属性名进行跨表查询
- 逆向查询通过关联表对应类名小写进行跨表查询
五、基于对象的多表查询
在连表查询规则规则之上，逆向查询遇到多条结果，在类名后再添加_set
1、一对多表查询
需求：打印第一本书的出版社名book = Book.objects.first() # 查询得到书对象print(book.publish.name)需求：第一个出版社出版过的书们的名字， publish = Publish.objects.first() # 查询得到出版社对象print(publish.book_set.all())for book in publish.book_set.all():# 得到书的类名 print(book.name)需求：打印第二本书的出版社地址address = Book.objects.all()[1].publish.addressprint(address)需求：第二个出版社出版过的书们的价格publish = Publish.objects.all()[1] # type: Publishfor book in publish.book_set.all(): print(book.price)
2、一对一
author = Author.objects.first() # type: Authorprint(author.author_detail.telephone)author_detail = AuthorDetail.objects.last() # type: AuthorDetailprint(author_detail.author.name)
3、多对多
需求：第一本书的作者们的姓名book = Book.objects.filter(name="inPingMei").first() # type: Book# print(book.author)for author in book.author.filter(name__endswith="xx"): print(author.name)需求：第二个作者写过的书们的书名author = Author.objects.all()[1] # type: Authorfor book in author.book_set.all(): print(book.name)需求：第一个出版社出版过的书们的名字与价格publish = Publish.objects.first()books = publish.book_set.all() # 数据所有字段数据都被保留存放在对象(们)中for book in books: print(book.name, book.price)books_infos = publish.book_set.values('name', 'price') # 数据指定字段数据被保留存放在列表中print(books_infos) 5、多级跨表查询
需求：第一个出版社出版过书们的作者们的电话号码publish = Publish.objects.first() # type: Publishbooks = publish.book_set.all()for book in books: # type: Book authors = book.author.all() # type: Author for author in authors: # type: Author print(author.author_detail.telephone)
六、基于双下划线的跨表查询
注：不能出现对象
1、一对多需求：第一本书的出版社名publish_name = Book.objects.filter(id=5).values('publish__name')print(publish_name)需求：第一个出版社出版过的书们，books_infos = Publish.objects.filter(id=2).values('book__name', 'book__price')print(books_infos)2、一对一需求: 查询所有大于80岁作者的名字与实际年龄authors_infos = Author.objects.filter(author_detail__age__gt=80).values('name', 'author_detail__age')print(authors_infos)多级连表查询多表关联：查询出版社在上海的出版过的所有书的 作者姓名、作者电话、具体出版社名 的相关信息infos = Publish.objects.filter(address__contains='上海').values('book__author__name', 'book__author__author_detail__telephone', 'name')print(infos)