mysql去重留一

首先先分析一下 我们现在的目的 是 查询到这俩张表的所有数据 然后进行删除重复记录 每条数据只保留一条

新北ssl适用于网站、小程序/APP、API接口等需要进行数据传输应用场景，ssl证书未来市场广阔！成为成都创新互联的ssl证书销售渠道，可以享受市场价格4-6折优惠！如果有意向欢迎电话联系或者加微信：18980820575（备注：SSL证书合作）期待与您的合作！

第一步：
查询以下俩张表的重复记录 (关键字段重复>1)
ks_examcity 、 ks_examdistrict

select * from ks_examcity group by examSubjectID,city,province having count(examSubjectID)>1;
select * from ks_examdistrict group by examSubjectID,district,city having count(examSubjectID)>1;

第二步：
   查询这两张表中 每条记录的第一条记录 (每条记录重复中的第一条 id最小)
select min(id)
  from ks_examcity
 group by examSubjectID, city, province
having count(examSubjectID)> 1

SELECT min(id)
  FROM `ks_examdistrict`
 GROUP BY `examSubjectID`, `district`, `city`
HAVING COUNT(`examSubjectID`)> 1

第三步：
    联查： 查询所有的重复数据以及重复记录中第一条以外的数据
select `examSubjectID`
  from ks_examcity
 group by examSubjectID, city, province
having count(examSubjectID)> 1 )
   and id not in(
select min(id)
  from ks_examcity
 group by examSubjectID, city, province
having count(examSubjectID)> 1

SELECT `examSubjectID`
  from `ks_examdistrict`
 group by `examSubjectID`, `district`, `city`
HAVING COUNT(`examSubjectID`)> 1)
   and id not in(
SELECT min(id)
  FROM `ks_examdistrict`
 GROUP BY `examSubjectID`, `district`, `city`
HAVING COUNT(`examSubjectID`)> 1

第四步：  查询以上查询到的数据的所有id  并以查询到的id作为条件 进行删除

delete from `ks_examcity`  where id IN(  select id from (
select  id
  from ks_examcity
 where `examSubjectID` in(
select `examSubjectID`
  from ks_examcity
 group by examSubjectID, city, province
having count(examSubjectID)> 1 )
   and id not in(
select min(id)
  from ks_examcity
 group by examSubjectID, city, province
having count(examSubjectID)> 1)) as  tmpresult)

======================================================

DELETE
  FROM `ks_examdistrict`
 where id IN(
SELECT id
  from(
select id
  from `ks_examdistrict`
 where `examSubjectID` in(
SELECT `examSubjectID`
  from `ks_examdistrict`
 group by `examSubjectID`, `district`, `city`
HAVING COUNT(`examSubjectID`)> 1)
   and id not in(
SELECT min(id)
  FROM `ks_examdistrict`
 GROUP BY `examSubjectID`, `district`, `city`
HAVING COUNT(`examSubjectID`)> 1)) as tmpresult)

参考文章 ：https://www.cnblogs.com/jdbeyond/p/8157224.html