.net - LINQ 到 SQL。级联删除模拟

Question

我在人和地址之间有多对多的关系

Person { Id, ... }
Address { Id, ... }
Address2Person { Id, PersonId, AddressId }

DBML 生成的类看起来像(图中有 PK 和 FK):

class Person {
  Id, Address2Person 
}

class Address2Person {
  Id, Person, PersonId, Address, AddressId
}

class Address {
  Id, Address2Person
}

当我删除 person 时，我必须删除 AddressPerson 行(这很容易)以及适当的 Address 行，但前提是没有对 address 的外部引用，除非删除一个。

在数据库级别实现级联删除可能更聪明，但我如何在纯 LINQ 2 SQL 中模拟这一点？

最后我编写了这样的代码:

// finding out addresses to delete
var addressesToDelete = 
  (from ap in DataBase.Address2Person
    group ap by ap.Address
    into g
      where g.All(x => x.PersonId == personId2Delete)
        select g.Key).ToList();

// killing references
DataBase.Address2Person.DeleteAllOnSubmit(
    DataBase.Address2Person.Where(x => x.PersonId == personId2Delete));

// killing addresses
DataBase.Address.DeleteAllOnSubmit(addressesToDelete);

有没有更好的办法？

先感谢您!

Answer 1

如果您的规则是不能有孤立地址，您可能会从以下顺序获得更好的性能和更容易的代码维护:

DataBase.Address2Person.DeleteAllOnSubmit(
    DataBase.Address2Person.Where(x => x.PersonId == personId2Delete));
DataBase.Address.DeleteAllOnSubmit(
    DataBase.Address.Where(x => !x.Address2Person.Any());