Mysql 查询以从SQL中删除重复项

我在桌子上有一个名字叫“距离”。它有4列id、起点、终点和距离

我有一些重复的记录。在某种意义上，重复记录

start_from   |   end_to    | distance
Chennai        Bangalore     350
Bangalore      Chennai       350
Chennai        Hyderabad     500
Hyderabad      Chennai       510

在上表中，钦奈到班加罗尔和班加罗尔到钦奈都有相同的距离。因此，我需要查询以删除选择上的记录

我想要一个像

start_from   |   end_to    | distance
Chennai        Bangalore     350
Chennai        Hyderabad     500
Hyderabad      Chennai       510

在查询中设置字段顺序（使用值）有助于获得唯一的行：

select distinct
    case when start_from  > end_to then end_to     else  start_from end as _start,
    case when start_from  > end_to then start_from else  end_to     end as _end,
    distance
from distance;

经过测试，我得到：

+-----------+-----------+----------+
| _start    | _end      | distance |
+-----------+-----------+----------+
| Bangalore | Chennai   |      350 |
| Chennai   | Hyderabad |      500 |
| Chennai   | Hyderabad |      510 |
+-----------+-----------+----------+

---

您可以使用以下查询查找重复项：

SELECT LEAST(start_from, end_to) AS start_from, 
       GREATEST(start_from, end_to) AS end_to, 
       distance
FROM mytable 
GROUP BY LEAST(start_from, end_to), GREATEST(start_from, end_to), distance
HAVING COUNT(*) > 1

SELECT t1.*
FROM mytable AS t1
LEFT JOIN (
    SELECT LEAST(start_from, end_to) AS start_from, 
           GREATEST(start_from, end_to) AS end_to, 
           distance
    FROM mytable 
    GROUP BY LEAST(start_from, end_to), GREATEST(start_from, end_to), distance
    HAVING COUNT(*) > 1
) AS t2 ON t1.start_from = t2.start_from AND 
           t1.end_to = t2.end_to AND 
           t1.distance = t2.distance    
WHERE t2.start_from IS NULL

输出：

start_from,   end_to,  distance
--------------------------------
Bangalore,    Chennai, 350

start_from  end_to     distance
--------------------------------
Chennai     Bangalore  350
Chennai     Hyderabad  500
Hyderabad   Chennai    510

现在，您可以使用上述查询作为派生表来筛选重复项：

SELECT LEAST(start_from, end_to) AS start_from, 
       GREATEST(start_from, end_to) AS end_to, 
       distance
FROM mytable 
GROUP BY LEAST(start_from, end_to), GREATEST(start_from, end_to), distance
HAVING COUNT(*) > 1

SELECT t1.*
FROM mytable AS t1
LEFT JOIN (
    SELECT LEAST(start_from, end_to) AS start_from, 
           GREATEST(start_from, end_to) AS end_to, 
           distance
    FROM mytable 
    GROUP BY LEAST(start_from, end_to), GREATEST(start_from, end_to), distance
    HAVING COUNT(*) > 1
) AS t2 ON t1.start_from = t2.start_from AND 
           t1.end_to = t2.end_to AND 
           t1.distance = t2.distance    
WHERE t2.start_from IS NULL

WHERE

子句谓词

t2.start\u from为NULL

，过滤掉重复记录

输出：

start_from,   end_to,  distance
--------------------------------
Bangalore,    Chennai, 350

start_from  end_to     distance
--------------------------------
Chennai     Bangalore  350
Chennai     Hyderabad  500
Hyderabad   Chennai    510

摆出你的桌子

id  start_from              end_to                  distance
0   Chennai                 Bangalore               350
1   Bangalore               Chennai                 350
2   Chennai                 Hyderabad               500
3   Hyderabad               Chennai                 510

然后可以使用查询compare with id

Select 
    O.start_from,
    O.end_to,
    O.distance 
From 
    distance O
Left Join
    distance P
On 
    1 = 1
    and O.start_from = P.end_to 
    and O.end_to = P.start_from
Where 
    1 = 1
    and O.distance <> P.distance 
    or(O.distance = P.distance and O.id < P.id)

选择
从……开始，
O.结束，
O.距离
从…起
距离O
左连接
距离P
在…上
1 = 1
和O.start\u from=P.end\u to
O.end_to=P.start_from
哪里
1 = 1
和O.distance P.distance
或（O.distance=P.distance，O.id

如果

钦奈到班加罗尔

或

班加罗尔到钦奈

之间没有区别，您可以尝试以下方法：

select
    max(`start_from`) as `start_from`,
    min(`end_to`) as `end_to`,
    `distance`
from yourtable
group by
    case when `start_from` > `end_to` then `end_to` else `start_from` end,
    case when `start_from` > `end_to` then `start_from` else `end_to` end,
    `distance`

这是一个在线测试仪

即使是从钦奈到海得拉巴也可以

如果您想保留从班加罗尔到钦奈的

max

和

min

的位置：

select
    min(`start_from`) as `start_from`,
    max(`end_to`) as `end_to`,
    `distance`
from yourtable
group by
    case when `start_from` > `end_to` then `end_to` else `start_from` end,
    case when `start_from` > `end_to` then `start_from` else `end_to` end,
    `distance`

也是一个

---

当将与大多数数据库兼容时，

将与大多数数据库兼容。
请共享确切的所需输出。字段值可能重复，但根据需要，我们需要重写查询或重新设计表。@Saurabhjhunjhunla添加了所需的输出。我不能换桌子。你为什么要金奈去班加罗尔，而不是班加罗尔去金奈？如果钦奈到海得拉巴也是350，你想要什么？钦奈到海得拉巴是不同的起点和终点。所以我不想删除这一行。那你为什么选择留在金奈到班加罗尔，而不是班加罗尔到金奈？但最后一行应该是
海得拉巴，金奈，510
是的，很好，使用值对字段排序，然后过滤唯一的记录。但正如@Ullas在changedYep中提到的start_，你是对的。使用case和
case
vs
JOIN
时效果更好。情况更好。因此最好使用
CASE