1.springfox-swagger实体类无限递归
https://hacpai.com/article/1525674135818 里面有不完美的解决方案
不用动源码的解决方案也有,在swagger项目里的函数传入参数没有类互相包含就可以了
这也要求某些多对多关系的实体类的属性定义要仔细走关系图
比如User和SysClass,用在swagger传入参数的实体类不能互相包含List<SysClass>和List<User>
只能继承VO(这个VO不能作为swagger函数的传入参数),把对应的包含属性写进去,然后需要查询对应列表时返回VO就行了
2.有中间表的mybatis一对多查询
查看某些有明确一对多关系的实体类的mybatis xml,可以看到查询方式是主表join子表,然后在mybatis这端把主表的数据group起来作为某个对象的元素
由此可见没有直接关联的2个表对应的类也可以用这种方式查询
<?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE mapper PUBLIC "-//mybatis.org//DTD Mapper 3.0//EN" "http://mybatis.org/dtd/mybatis-3-mapper.dtd"> <mapper namespace="com.ass.core.mapper.SysClassMapper"> <!-- 可根据自己的需求,是否要使用 --> <resultMap type="com.ass.core.entity.SysClass" id="classMap"> <result property="id" column="cid"/> <result property="code" column="code"/> <result property="name" column="name"/> <result property="num" column="num"/> <result property="description" column="description"/> </resultMap> <resultMap type="com.ass.core.VO.ClassWithUser" id="classUser"> <result property="id" column="cid"/> <result property="code" column="code"/> <result property="name" column="name"/> <result property="num" column="num"/> <result property="description" column="description"/> <collection property="teachers" ofType="User"> <id column="id" jdbcType="INTEGER" property="id" /> <result column="uid" jdbcType="VARCHAR" property="uid" /> <result column="user_account" jdbcType="VARCHAR" property="userAccount" /> <result column="password" jdbcType="VARCHAR" property="password" /> <result column="username" jdbcType="VARCHAR" property="username" /> <result column="user_identity" jdbcType="INTEGER" property="userIdentity" /> <result column="user_status" jdbcType="INTEGER" property="userStatus" /> <result column="user_type" jdbcType="INTEGER" property="userType" /> <result column="user_email" jdbcType="VARCHAR" property="userEmail" /> <result column="user_mobile" jdbcType="VARCHAR" property="userMobile" /> <result column="user_remind" jdbcType="VARCHAR" property="userRemind" /> <result column="user_create_time" jdbcType="TIMESTAMP" property="userCreateTime" /> <result column="user_update_time" jdbcType="TIMESTAMP" property="userUpdateTime" /> <result column="class_id" jdbcType="VARCHAR" property="classId" /> <result column="user_col2" jdbcType="VARCHAR" property="userCol2" /> <result column="user_col3" jdbcType="VARCHAR" property="userCol3" /> <result column="user_token" jdbcType="VARCHAR" property="userToken" /> <result column="role_id" jdbcType="VARCHAR" property="roleId" /> </collection> </resultMap> <select id="selectClassListByName" resultMap="classUser"> SELECT c.id as cid, c.code, c.name, c.num, c.description, u.* FROM sys_class c LEFT JOIN user_class uc ON c.id = uc.class_id LEFT JOIN user u ON u.uid = uc.user_id WHERE c.name like '%${name}%' </select> </mapper>
3.swagger传入参数含有对象数组时的坑点
swagger前端在碰到对象数组(无论是直接传的还是实体类里面的)时,会显示xxx[0].id和xxx[0].name之类的参数,但是直接通过前端发送请求,会因为url含有非法字符被拒绝请求
如果是用其他程序提交,例如Python(先安装requests扩展包),可以提交这种带对象数组的请求,并让服务器接收
import requests import json def login(): res = requests.post('http://localhost:20005/api/user/login?userAccount=admin&userPassword=admin', verify=False) myj = json.loads(res.text) #print(myj) return myj['data']['token'] # def addContent(): token = login() params = { 'resourceSteps[0].procedureName': 'ytmym', 'resourceSteps[0].procedureContent': 'mytm', 'resourceSteps[2].procedureName': 'loveZJJ', 'resourceSteps[2].procedureContent': 'asswecan', 'id': '45', 'resourceContentId': '45j45j', 'experimentAdd': '5j5j', 'experimentAdd1': 'grh', 'stepsjson': '[{"procedureName":"a","procedureContent":"b"}]', 'token': token, 'resourceManagementId': 'j54j45j', 'experimentalObjective': '45j45j4', 'experimentalContext': '45j45j', 'experimentalRequirements': '5j45j', 'prepareKnowledge': '54j45j', 'experimentSummary': 'j54j', 'experimentTestId': 'h4h54' } url = 'http://localhost:20005/api/resource/add/content' res = requests.post(url, params=params) print(res.text) # if __name__ == '__main__': addContent()
但是这种实现方式不太推荐,太不直观了。建议使用@ResponseBody注解对象数组或包含此类型的实体类,然后让前端传入json,处理会简单一些