洞察探索open banking如何通过小程序容器技术助力金融企业实现数据安全和数字化转型
822
2022-10-22
在Mybatis中association标签多层嵌套的问题
目录association标签多层嵌套问题排查从代码上看没有什么问题正常代码如下association集合嵌套这个返回集合有什么用呢
association标签多层嵌套问题
mybatis里查询使用嵌套association标签时,发现内层的association查询的结果一直为null
排查
检查sql执行情况,发现有数据返回,排除检查property的值是否和pojo中的对应,值一致,排除检查column的值是否和数据库的相对应,相对应,排除
那么应该是mybatis没有把数据映射到位了,经过排查是association中columnPrefix被不对应
SELECT
ii.Id,
ii.model,
ii.status,
ii.work_time,
ui.id AS ui_id,
ui.interface_name AS ui_interface_name,
ui.interface_type AS ui_interface_type,
ui.frequency AS ui_frequency,
ui.address AS ui_address,
ui.template_or_sql AS ui_template_or_sql,
ui.status AS ui_status,
sys.id AS sys_id,
sys.system_name AS sys_system_name,
sys.system_name_en AS sys_system_name_en,
sys.belong AS sys_belong,
sys.status AS sys_status,
ser.id AS ser_id,
ser.ftp_ip AS ser_ftp_ip,
ser.ftp_port AS ser_ftp_port,
ser.ftp_account AS ser_ftp_account,
ser.ftp_password AS ser_ftp_password
从代码上看没有什么问题
原因是association在进行多层嵌套时,mybatis会将外层association的columnPrefix值与内层的进行并合,
如外层columnPrefix值位ui_, 内层为sys_, 那么在SQL中就不能这样 sys.id AS sys_id 了,需要将ui_前缀加上,变成 sys.id AS ui_sys_id ,这样mybatis在匹配的时候才会将数据映射到对应association上
正常代码如下
SELECT
ii.Id,
ii.model,
ii.status,
ii.work_time,
ui.id AS ui_id,
ui.interface_name AS ui_interface_name,
ui.interface_type AS ui_interface_type,
ui.frequency AS ui_frequency,
ui.address AS ui_address,
ui.template_or_sql AS ui_template_or_sql,
ui.status AS ui_status,
sys.id AS ui_sys_id,
sys.system_name AS ui_sys_system_name,
sys.system_name_en AS ui_sys_system_name_en,
sys.belong AS ui_sys_belong,
sys.status AS ui_sys_status,
ser.id AS ui_ser_id,
ser.ftp_ip AS ui_ser_ftp_ip,
ser.ftp_port AS ui_ser_ftp_port,
ser.ftp_account AS ui_ser_ftp_account,
ser.ftp_password AS ui_ser_ftp_password
问题解决!
association集合嵌套
学了一下mybatis的查询返回值的集合嵌套,先查了查官网:
这个返回集合有什么用呢
举个例子三张表
hr_job_department
hr_job_position
第三张表里在表示部门和职位的时候只用了上面两张表的主键
但是查询的时候,希望表示下面这样的结果
所以返回值是不止一个对象,这样就用到http://了集合嵌套
javaType="com.advancedc.hrsys.entity.JobDepartment">
javaType="com.advancedc.hrsys.entity.JobDepartment">
javaType="com.advancedc.hrsys.entity.JobPosition">
javaType="com.advancedc.hrsys.entity.JobPosition">
只需要知道:
(1)column表示数据库字段
(2)property表示Java里的值
而且我这里的主键都是id所以会出现重名的情况,在SQL语句里,查询时就要赋予别名才能加以区分,返回结果resultMap就如上图所示
SELECT
ui.id,
ui.name,
ui.gender,
ui.id_card,
ui.is_married,
ui.department_id,
ui.position_id,
ui.phone,
ui.priority,
ui.entry_time,
ui.full_time,
ui.created_time,
ui.edited_time,
jd.id jdid,
jd.name jdname,
jp.id jpid,
jp.name jpname
FROM
hr_user_info ui
INNER JOIN
hr_job_department jd
ON
ui.department_id=jd.id
INNER JOIN
hr_job_position jp
ON
ui.position_id=jp.id
and ui.id = #{someone.id}
and ui.gender = #{someone.gender}
and ui.name = #{someone.name}
and ui.id_card = #{someone.idCard}
and ui.is_married = #{someone.isMarried}
and ui.department_id = #{someone.jobDepartment.id}
and ui.position_id = #{someone.jonPosition.id}
and ui.phone = #{someone.phone}
and ui.entry_time = #{someone.entryTime}
and ui.full_time = #{someone.fullTime}
上图用了INNER JOIN来查询看上去挺简洁的,有一种不简洁的写法如下,虽然也能得到结果,但是不知道性能对比如何
SELECT
ui.id,
ui.name,
ui.gender,
ui.id_card,
ui.is_married,
ui.department_id,
ui.position_id,
ui.phone,
ui.priority,
ui.entry_time,
ui.full_time,
ui.created_time,
ui.edited_time,
(select id jdid from hr_job_department jd where jd.id=ui.department_id) jdid,
(select name jdname from hr_job_department jd where jd.id=ui.department_id) jdname,
(select id jpid from hr_job_position jp where jp.id=ui.position_id) jpid,
(select name jpname from hr_job_position jp where jp.id=ui.position_id) jpname
FROM
hr_user_info ui;
版权声明:本文内容由网络用户投稿,版权归原作者所有,本站不拥有其著作权,亦不承担相应法律责任。如果您发现本站中有涉嫌抄袭或描述失实的内容,请联系我们jiasou666@gmail.com 处理,核实后本网站将在24小时内删除侵权内容。
发表评论
暂时没有评论,来抢沙发吧~