單向N-1
情景: 合租，假設你要租房子，多個人住同一個地方，對于你來說，你只需要關心我今天晚上要住到哪里，而不需要關心。某一個地址住了哪些人，（其實還是應該關心一下的）。
表結構如下
| Field | Type | Null | Key | Default | Extra |
+----------+--------------+------+-----+---------+-------+
| id | varchar(50) | NO | PRI | NULL | |
| location | varchar(100) | NO | | NULL | |
+----------+--------------+------+-----+---------+-------+
+------------+-------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+------------+-------------+------+-----+---------+-------+
| id | varchar(50) | NO | PRI | NULL | |
| name | varchar(50) | NO | | NULL | |
| age | int(5) | YES | | NULL | |
| address_id | varchar(50) | YES | MUL | NULL | |
+------------+-------------+------+-----+---------+-------+
無鏈接表的單向N-1
- 在N一端的實體增加@ManyToOne 修飾的 關聯實體
- 在N一端增加外鍵列，實體類中用@JoinColumn修飾關聯實體。
@Entity
@Table(name="tb_person")
public class Person extends AbstractEntity{
// id 我在AbstractEntity中聲明了
// 需要將AbstractEntity 添加@MappedSuperclass
@Column(name="name")
private String name;
@Column(name="age")
private Integer age;
//用CasecadeType=All 主要方便的是在 Person 更新，刪除，添加的時候自動把address持久化
@ManyToOne(cascade=CascadeType.ALL)
@JoinColumn(name="address_id")
private Address address;
//省略set.get方法
}
因為不需要從address獲取person 所以 address 只需要自己本身的屬性就行。
@Entity
@Table(name="tb_address")
public class Address extends AbstractEntity{
@Column(name="location")
private String addressDetail;
public String getAddressDetail() {
return addressDetail;
}
public void setAddressDetail(String addressDetail) {
this.addressDetail = addressDetail;
}
}
測試用例
@Test
@Transactional
@Rollback(value=false)
public void testEntity(){
Session session = sessionFactory.getCurrentSession();
Person p = new Person();
Address address = new Address();
address.setAddressDetail("ZheJiang");
p.setName("WhosYourdaddy");
p.setAge(89);
p.setAddress(address);
session.save(p);
}
測試結果
+--------------------+----------+
| id | location |
+--------------------+----------+
| 170413181716030431 | ZheJiang |
+--------------------+----------+
+--------------------+---------------+------+--------------------+
| id | name | age | address_id |
+--------------------+---------------+------+--------------------+
| 170413181716014566 | WhosYourdaddy | 89 | 170413181716030431 |
+--------------------+---------------+------+--------------------+
我們再來看一下sql輸出，
第一條
/* get current state com.statestr.verify.entity.Address */
select address_.id, address_.location as location2_0_ from tb_address address_ where address_.id=?
HibernateLog --> 18:17:16 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170413181716030431]
第二條
/* insert com.statestr.verify.entity.Address*/
insert into tb_address (location, id) values (?, ?)
HibernateLog --> 18:17:16 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [ZheJiang]
HibernateLog --> 18:17:16 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [VARCHAR] - [170413181716030431]
第三條
insert com.statestr.verify.entity.Person*/ insert into tb_person (address_id, age, name, id) values (?, ?, ?, ?)
HibernateLog --> 18:17:16 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170413181716030431]
HibernateLog --> 18:17:16 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [INTEGER] - [89]
HibernateLog --> 18:17:16 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [3] as [VARCHAR] - [WhosYourdaddy]
HibernateLog --> 18:17:16 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [4] as [VARCHAR] - [170413181716014566]
總共執行了3條語句。第一條是查詢是否數據庫已經有了identity 相等的address。如果沒有那就需要insert address.
那么如果數據庫已經存在該條數據庫呢？
Address address = (Address)session.get(Address.class, "170413181716030431");
// 替換
Address address = new Address();
address.setAddressDetail("ZheJiang");
第一條
select address0_.id as id1_0_0_, address0_.location as location2_0_0_ from tb_address address0_ where address0_.id=?
HibernateLog --> 19:08:43 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170413181716030431]
第二條
/* insert com.statestr.verify.entity.Person*/
insert into tb_person (address_id, age, name, id) values (?, ?, ?, ?)
HibernateLog --> 19:08:43 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170413181716030431]
HibernateLog --> 19:08:43 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [INTEGER] - [89]
HibernateLog --> 19:08:43 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [3] as [VARCHAR] - [WhosYourdaddy]
HibernateLog --> 19:08:43 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [4] as [VARCHAR] - [170413190843156966]
只執行了2條語句，因為它判斷addres已經在數據庫里面，而且沒有改變其內容。hibernate 會跟蹤每個持久化狀態對象的改動。（可以參閱 http://www.bbsmax.com/A/amd03O6X5g/ ，我發現在我這個4.3.11.Final，其實所有持久化對象都存儲在Session的 persistenceContext，變量中。持久化對象發生改變時，他下面的確改變了，但是沒有找到保存之前狀態的地方。）這里它發現對象并沒有被改變。那如果對象發生改變了呢？
Address address = (Address)session.get(Address.class, "170413181716030431");
address.setAddressDetail("QingDao");
// 替換
Address address = (Address)session.get(Address.class, "170413181716030431");
// 第一條 查找數據庫
select address0_.id as id1_0_0_, address0_.location as location2_0_0_ from tb_address address0_ where address0_.id=?
HibernateLog --> 19:22:02 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170413181716030431]
// 第二條 添加person到數據庫
/* insert com.statestr.verify.entity.Person*/ insert into tb_person (address_id, age, name, id) values (?, ?, ?, ?)
HibernateLog --> 19:30:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170413181716030431]
HibernateLog --> 19:30:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [INTEGER] - [89]
HibernateLog --> 19:30:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [3] as [VARCHAR] - [WhosYourdaddy]
HibernateLog --> 19:30:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [4] as [VARCHAR] - [170413192101018079]
//第三條 更新address，hibernate session 檢測到持久類已經發生改變
update com.statestr.verify.entity.Address */ update tb_address set location=? where id=?
HibernateLog --> 19:30:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [QingDao]
HibernateLog --> 19:30:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [VARCHAR] - [170413181716030431]
還要注意一下調用的順序，稍微有點偏題了...
有鏈接表的單向N-1
絕大多數單向N-1，都是用基于外鍵的。但hibernate 也支持了有連接表的N-1關系。（不常用，留個印象就行）
有連接表的單向N-1就需要用到@JoinTable，他主要用來映射底層數據庫中間表。主要屬性有
| 屬性 | 是否必須 | 說明 |
|---|---|---|
| name | 否 | （重要）中間表的表名 |
| catalog | 否 | 設置中間表放入制定的catalog |
| schema | 否 | 設置中間表放入制定的schema |
| targetEntity | 否 | 設定關聯實體的類名。一般不需要設置，hibernate自動會獲取 |
| indexes | 否 | 為連接表設置多個索引 |
| joinColumns | 否 | （重要）接收JoinColumn，這邊設置的是JoinTable表對應當前實體的外鍵列 |
| inverseJoinColumns | 否 | （重要）接收JoinColumn，這邊設置的是JoinTable表對應關聯實體的外鍵列 |
| uniqueConstraints | 否 | 增加唯一性約束 |
（安利一個好的網站 table Generate,可以生成markdow的表格）
好接下來我們來看例子:
表結構
mysql> describe tb_person_addr_map;
+------------+-------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+------------+-------------+------+-----+---------+-------+
| person_id | varchar(50) | NO | MUL | NULL | |
| address_id | varchar(50) | NO | MUL | NULL | |
+------------+-------------+------+-----+---------+-------+
2 rows in set (0.00 sec)
mysql> describe tb_person;
+-------+-------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-------+-------------+------+-----+---------+-------+
| id | varchar(50) | NO | PRI | NULL | |
| name | varchar(50) | NO | | NULL | |
| age | int(5) | YES | | NULL | |
+-------+-------------+------+-----+---------+-------+
3 rows in set (0.02 sec)
mysql> describe tb_address;
+----------+--------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+----------+--------------+------+-----+---------+-------+
| id | varchar(50) | NO | PRI | NULL | |
| location | varchar(100) | NO | | NULL | |
+----------+--------------+------+-----+---------+-------+
2 rows in set (0.00 sec)
這邊只需要該Person實體
@Entity
@Table(name="tb_person")
public class Person extends AbstractEntity{
@Column(name="name")
private String name;
@Column(name="age")
private Integer age;
@ManyToOne(cascade=CascadeType.ALL)
@JoinTable(name="tb_person_addr_map",joinColumns=@JoinColumn(name="person_id",unique=true),
inverseJoinColumns=@JoinColumn(name="address_id"))
private Address address;
// 省略set、get 方法
}
因為是N-1關聯，那么Person只能有一個Address，所以添加了unique=true
我們來看sql輸出:
/* get current state com.statestr.verify.entity.Address */
select address_.id, address_.location as location2_0_ from tb_address address_ where address_.id=?
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170414165050149527]
/* insert com.statestr.verify.entity.Address*/
insert into tb_address (location, id) values (?, ?)
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [ZheJiang]
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [VARCHAR] - [170414165050149527]
/* insert com.statestr.verify.entity.Person*/
insert into tb_person (age, name, id) values (?, ?, ?)
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [INTEGER] - [89]
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [VARCHAR] - [WhosYourdaddy]
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [3] as [VARCHAR] - [170414165050149349]
/* insert com.statestr.verify.entity.Person*/
insert into tb_person_addr_map (address_id, person_id) values (?, ?)
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [1] as [VARCHAR] - [170414165050149527]
HibernateLog --> 16:50:50 TRACE org.hibernate.type.descriptor.sql.BasicBinder - binding parameter [2] as [VARCHAR] - [170414165050149349]
一共執行了4條語句?？梢钥吹竭@邊最后才插入到中間表中。
