Spring Data JPA自动添加查询前缀致列名错误，如何禁用该行为？

这个问题我之前也碰到过！当多对多关联的连接表需要加额外字段时，直接改原来的@ManyToMany配置肯定会踩坑——JPA默认的命名规则会自作主张生成你没定义的列名，就是你看到的employee_EMPLOYEE_ID_PK。解决的核心是把多对多拆成两个一对多，用中间实体来映射带额外字段的连接表，并且明确指定所有关联列的名称，不让JPA瞎自动生成。

具体配置步骤 #

1. 定义复合主键类

首先需要为连接表创建一个复合主键类（因为连接表的主键是员工ID+项目ID的组合），实现Serializable接口：

@Embeddable
public class EmployeeProjectId implements Serializable {
    @Column(name = "EMPLOYEE_ID_PK")
    private Long employeeId;

    @Column(name = "PROJECT_ID_PK")
    private Long projectId;

    // 必须实现无参构造器、equals()和hashCode()方法
    public EmployeeProjectId() {}

    public EmployeeProjectId(Long employeeId, Long projectId) {
        this.employeeId = employeeId;
        this.projectId = projectId;
    }

    @Override
    public boolean equals(Object o) {
        if (this == o) return true;
        if (o == null || getClass() != o.getClass()) return false;
        EmployeeProjectId that = (EmployeeProjectId) o;
        return Objects.equals(employeeId, that.employeeId) && Objects.equals(projectId, that.projectId);
    }

    @Override
    public int hashCode() {
        return Objects.hash(employeeId, projectId);
    }

    // getter和setter方法
}

2. 创建中间实体映射连接表

用这个实体来对应带额外字段的连接表，同时关联Employee和Project实体：

@Entity
@Table(name = "EMPLOYEE_PROJECT") // 你的连接表名称
public class EmployeeProject {
    @EmbeddedId
    private EmployeeProjectId id;

    @ManyToOne
    @MapsId("employeeId") // 关联复合主键中的employeeId字段
    @JoinColumn(name = "EMPLOYEE_ID_PK") // 明确指定连接表中的外键列名
    private Employee employee;

    @ManyToOne
    @MapsId("projectId") // 关联复合主键中的projectId字段
    @JoinColumn(name = "PROJECT_ID_PK") // 明确指定连接表中的外键列名
    private Project project;

    // 你的额外字段，比如
    private LocalDate joinDate;
    private String role;

    // 构造器、getter和setter方法
    public EmployeeProject() {}

    public EmployeeProject(Employee employee, Project project, LocalDate joinDate, String role) {
        this.id = new EmployeeProjectId(employee.getId(), project.getId());
        this.employee = employee;
        this.project = project;
        this.joinDate = joinDate;
        this.role = role;
    }
}

3. 修改Employee和Project实体的关联

把原来的@ManyToMany替换成@OneToMany关联中间实体：

@Entity
@Table(name = "EMPLOYEE")
public class Employee {
    @Id
    @Column(name = "EMPLOYEE_ID_PK")
    private Long id;

    // 其他原有字段...

    @OneToMany(mappedBy = "employee", cascade = CascadeType.ALL, orphanRemoval = true)
    private List<EmployeeProject> employeeProjects = new ArrayList<>();

    // 方便操作的工具方法，比如添加项目
    public void addProject(Project project, LocalDate joinDate, String role) {
        EmployeeProject ep = new EmployeeProject(this, project, joinDate, role);
        employeeProjects.add(ep);
        project.getEmployeeProjects().add(ep);
    }

    // getter和setter方法
}

Project实体的修改类似：

@Entity
@Table(name = "PROJECT")
public class Project {
    @Id
    @Column(name = "PROJECT_ID_PK")
    private Long id;

    // 其他原有字段...

    @OneToMany(mappedBy = "project", cascade = CascadeType.ALL, orphanRemoval = true)
    private List<EmployeeProject> employeeProjects = new ArrayList<>();

    // getter和setter方法
}

关键说明 #

你遇到的employee_EMPLOYEE_ID_PK错误，是因为JPA在没有明确指定@JoinColumn的情况下，会默认按照实体名_主键列名的规则生成关联列名。通过上面的配置：

• 用@JoinColumn(name = "你的列名")明确指定了连接表中的外键列，完全禁用了JPA的自动命名
• 用@MapsId把中间实体的关联字段和复合主键的字段绑定，保证了关联关系的正确性

这样改造后，查询员工的关联项目时，就不会再出现这个异常了。

内容的提问来源于stack exchange，提问作者Sabin Antohe