为什么JDBI 3 @ColumnName注释不起作用?

时间:2018-11-02 14:11:14

标签: java orm jdbi

我想使用jdbi 3 @ColumnName批注来映射表和Java bean字段的列名,但是在运行时jdbi会引发异常。 我的问题是为什么会引发异常?我的用法不正确吗?如何解决这个问题呢? 谢谢

数据库表:

CREATE TABLE `department` (
  `id` bigint(20)  NOT NULL AUTO_INCREMENT,
  `name` varchar(32) NOT NULL DEFAULT '' ,
  `business_system` tinyint(4)  NOT NULL ,
  `functional` tinyint(4)  NOT NULL ,
  `remark` varchar(100) NOT NULL DEFAULT ''
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

Java Bean:

@Data
public class Department {
    private Long id;

    @NotNull
    @Size(min = 3, max = 20)
    private String name;

    @NotNull
    @ColumnName("business_system")
    private Integer businessSystem;

    @NotNull
    @ColumnName("functional")
    private Integer functional;

    @Max(value = 200)
    private String remark;
  }

JDBI更新:

@Override
public Long create(Department department) {
    String sql = "INSERT INTO department (name, business_system, functional, remark)"
        + "VALUES (:name, :business_system, :functional, :remark)";

return jdbi.withHandle(handle -> handle.createUpdate(sql)
    .bindBean(department)
    .executeAndReturnGeneratedKeys()
    .mapTo(Long.class)
    .findOnly());

}

单元测试运行结果:

org.jdbi.v3.core.statement.UnableToExecuteStatementException: Unable to 
execute, no named parameter matches 'business_system'. [statement:"INSERT INTO 
department (name, business_system, functional, remark)VALUES (:name, 
:business_system, :functional, :remark)", rewritten:"INSERT INTO department 
(name, business_system, functional, remark)VALUES (:name, :business_system, 
:functional, :remark)", parsed:"ParsedSql{sql='INSERT INTO department (name, 
business_system, functional, remark)VALUES (?, ?, ?, ?)', 
parameters=ParsedParameters{positional=false, parameterNames=[name, 
business_system, functional, remark]}}", 
arguments:{ positional:{}, named:{}, finder:[{lazy bean property arguments 
"Department(id=null, name=test, businessSystem=2, functional=2, 
remark=test)"]}]

1 个答案:

答案 0 :(得分:1)

Jdbi维护者在这里。四个误解共同导致您的代码无法正常工作:

首先,@ColumnName定义了一种单向映射,仅用于将列名映射到对象属性/字段/构造函数参数。因此,BeanMapper将使用@ColumnName注释,但是bindBean()会忽略它。

第二,BeanMapper仅在获取器和设置器上检测到@ColumnName,而不是字段。 由于您使用的是Lombok的@Data批注,因此@ColumnName批注实际上放在字段上。龙目岛(Lombok)有一种非常难看的方法来为getter或setter指定注释:

@NotNull
@Getter(onMethod=@__(@ColumnName("business_system")))
private Integer businessSystem;

@NotNull
@Getter(onMethod=@__(@ColumnName("functional")))
private Integer functional;

第三,bindBean()绑定基于bean属性名称而不是列名称的命名参数。因此,您的命名参数应位于:camelCase中,而不是:snake_case中:

INSERT INTO department (name, business_system, functional, remark)
VALUES (:name, :businessSystem, :functional, :remark)
--                      ^ camel case property name

最后,Jdbi 3已经将蛇形案例的列名与驼峰式属性名进行了匹配。例如:一列foo_bar将自动映射到fooBar属性,而无需@ColumnName注释。