详讲jpa

一：JPA（Java Persistent API） 是Sun 公式官方提出的Java持久操作形式 ，它为Java开发人员提供了一种对象/关系映射工具，

但是提供的不如hibernate功能那么强大，可以很方便的切换数据库(本人觉得比hibernate还要方便)

核心提示：JPA 注解的几个要点 1.设置Pojo为实体 @Entity //标识这个pojo是一个jpa实体 public class Users implements Serializable{ } 2.设置表名 @Entity @Table (name= users ) //指定表名为users public class Users implements Serializable{ } 3.设置主键 public JPA 注解的几个要点1.设置Pojo为实体@Entity //标识这个pojo是一个jpa实体     public class Users implements Serializable {     }    2.设置表名@Entity     @Table(name = "users") //指定表名为users     public class Users implements Serializable {     }    3.设置主键(其中主键生成策略很多种，本人以为不愿意切换数据库所以喜欢自己写一个生成策略)public class Users implements Serializable {     @Id     private String userCode;    4. 设置字段类型通过@Column注解设置，包含的设置如下.name：字段名.unique：是否唯一.nullable：是否可以为空.inserttable：是否可以插入.updateable：是否可以更新.columnDefinition: 定义建表时创建此列的DDL.secondaryTable: 从表名。如果此列不建在主表上（默认建在主表），该属性定义该列所在从表的名字。@Column(name = "user_code", nullable = false, length=32)//设置属性userCode对应的字段为user_code，长度为32，非空     private String userCode;     @Column(name = "user_wages", nullable = true, precision=12, scale=2)//设置属性wages对应的字段为user_wages，12位数字可保留两位小数，可以为空     private double wages;     @Temporal(TemporalType.DATE)//设置为时间类型     private Date joinDate;    5.字段排序在加载数据的时候可以为其指定顺序，使用@OrderBy注解实现@Table(name = "USERS")     public class User {     @OrderBy(name = "group_name ASC, name DESC")     private List books = new ArrayList();     }    6.主键生成策略public class Users implements Serializable {     @Id     @GeneratedValue(strategy=GenerationType.IDENTITY)//主键自增，注意，这种方式依赖于具体的数据库，如果数据库不支持自增主键，那么这个类型是没法用的     @Column(name = "user_id", nullable = false)     private int userId;           public class Users implements Serializable {     @Id     @GeneratedValue(strategy=GenerationType.TABLE)//通过一个表来实现主键id的自增，这种方式不依赖于具体的数据库，可以解决数据迁移的问题     @Column(name = "user_code", nullable = false)     private String userCode;           public class Users implements Serializable {     @Id     @GeneratedValue(strategy=GenerationType.SEQUENCE)//通过Sequence来实现表主键自增，这种方式依赖于数据库是否有SEQUENCE，如果没有就不能用     @SequenceGenerator(name="seq_user")     @Column(name = "user_id", nullable = false)     private int userId;    7.一对多映射关系有T_One和T_Many两个表，他们是一对多的关系，注解范例如下主Pojo@Entity     @Table(name = "T_ONE")     public class One implements Serializable {     private static final long serialVersionUID = 1L;     @Id     @Column(name = "ONE_ID", nullable = false)     private String oneId;     @Column(name = "DESCRIPTION")     private String description;     @OneToMany(cascade = CascadeType.ALL, mappedBy = "oneId")//指向多的那方的pojo的关联外键字段     private Collection<Many> manyCollection;     子Pojo@Entity     @Table(name = "T_MANY")     public class Many implements Serializable {     private static final long serialVersionUID = 1L;     @Id     @Column(name = "MANY_ID", nullable = false)     private String manyId;     @Column(name = "DESCRIPTION")     private String description;        @JoinColumn(name = "ONE_ID", referencedColumnName = "ONE_ID")//设置对应数据表的列名和引用的数据表的列名     @ManyToOne//设置在“一方”pojo的外键字段上     private One oneId;     8.多对多映射关系貌似多对多关系不需要设置级联，以前用hibernate的时候着实为多对多的级联头疼了一阵子，JPA的多对多还需要实际的尝试一下才能有所体会。估计JPA的多对多也是可以转换成两个一对多的。第一个Pojo@Entity     @Table(name = "T_MANYA")     public class ManyA implements Serializable {     private static final long serialVersionUID = 1L;     @Id     @Column(name = "MANYA_ID", nullable = false)     private String manyaId;     @Column(name = "DESCRIPTION")     private String description;     @ManyToMany     @JoinTable(name = "TMANY1_TMANY2", joinColumns = {@JoinColumn(name = "MANYA_ID", referencedColumnName = "MANYA_ID")}, inverseJoinColumns = {@JoinColumn(name = "MANYB_ID", referencedColumnName = "MANYB_ID")})     private Collection<ManyB> manybIdCollection;     第二个Pojo@Entity     @Table(name = "T_MANYB")     public class ManyB implements Serializable {     private static final long serialVersionUID = 1L;     @Id     @Column(name = "MANYB_ID", nullable = false)     private String manybId;     @Column(name = "DESCRIPTION")     private String description;     @ManyToMany(mappedBy = "manybIdCollection")     private Collection<ManyA> manyaIdCollection;     9.一对一映射关系主Pojo@Entity     @Table(name = "T_ONEA")     public class OneA implements Serializable {     private static final long serialVersionUID = 1L;     @Id     @Column(name = "ONEA_ID", nullable = false)     private String oneaId;     @Column(name = "DESCRIPTION")     private String description;     @OneToOne(cascade = CascadeType.ALL, mappedBy = "oneA")//主Pojo这方的设置比较简单，只要设置好级联和映射到从Pojo的外键就可以了。     private OneB oneB;    从Pojo@Entity     @Table(name = "T_ONEB")     public class OneB implements Serializable {     private static final long serialVersionUID = 1L;     @Id     @Column(name = "ONEA_ID", nullable = false)     private String oneaId;     @Column(name = "DESCRIPTION")     private String description;     @JoinColumn(name = "ONEA_ID", referencedColumnName = "ONEA_ID", insertable = false, updatable = false)//设置从方指向主方的关联外键，这个ONEA_ID其实是表T_ONEA的主键     @OneToOne     private OneA oneA;     10 大字段@Lob //对应Blob字段类型     @Column(name = "PHOTO")     private Serializable photo;     @Lob //对应Clob字段类型     @Column(name = "DESCRIPTION")     private String description;    11.瞬时字段不需要与数据库映射的字段，在保存的时候不需要保存倒数据库@Transient     private int tempValue;        public int getTempValue(){     get tempValue;     }        public void setTempValue(int value){     this.tempValue = value;     }