如何选择相同`userid`其他字段具有特定值的行？

Question

我有这种表（简化）：

640

并希望选择所有订单，其中每个orders sample data below --------------------------------------------- id INT: 1 2 3 4 5 userid INT 10 10 10 20 20 status CHAR(1) A A B A C 状态为userid，但根本没有订单IN ('A','B')。

因此，上述数据的输出会产生ID = 1,2和3 的订单。用户ID = 10有订单A和B，但没有C或D.

换句话说：为订单状态为A或B但没有状态C或D的客户选择订单。

我从这开始：

IN ('C','D')

但是认为SELECT xcart_orders.orderid, xcart_orders.* FROM xcart_orders JOIN ( select count(*) as bad_statuses, userid from xcart_orders where status in ('C','D') group by userid ) bo ON bo.userid=xcart_orders.userid JOIN ( select count(*) as good_statuses, userid from xcart_orders where status in ('A','B') group by userid ) bo2 ON bo2.userid=xcart_orders.userid WHERE bo2.good_statuses>0 and bo.bad_statuses=0 对于'坏'状态不会返回零，所以我没有得到任何结果。

Answer 1

您的汇总没有GROUP BY，要检查结果，您需要我们HAVING而不是WHERE

SELECT 
    xcart_orders.orderid, 
    xcart_orders.*, 
    SUM(CASE WHEN xcart_orders.status in ('C','D') THEN 1 ELSE 0 END) AS bad_statuses, 
    SUM(CASE WHEN xcart_orders.status in ('A','B') THEN 1 ELSE 0 END) AS good_statuses 
FROM xcart_orders 
GROUP BY orderid
HAVING bad_statuses = 0 
   AND good_statuses > 0 

请注意，如果您需要首先订购的特定字段，那么您从xcart_orders.*获得的字段将是随机的（或非确定性字段）。

• 首先，您要对GROUP BY user_id进行检查，以确定其状态是否与'A', 'B'不同

• 然后您从user_id：

中选择订单

SQL DEMO

SELECT `user_id`
FROM orders1
GROUP BY `user_id`
HAVING COUNT(*) = COUNT(CASE WHEN `status` IN ('A', 'B') THEN 1 END);


SELECT *
FROM orders1
WHERE `user_id` IN (SELECT `user_id`
                    FROM orders1
                    GROUP BY `user_id`
                    HAVING COUNT(*) = COUNT(CASE WHEN `status` IN ('A', 'B') THEN 1 END)
                   );

输出