Question

我有这个SQL工作正常

SELECT COUNT(*) AS total, e.FirstName, e.LastName
FROM [Messages] AS m 
INNER JOIN Employees AS e ON m.SenderId = e.UserId
GROUP BY e.EmployeeId, e.FirstName, e.LastName

但我希望还有完整的未读消息作为列。

如何在SQL中实现这一目标？

SELECT 
    COUNT(*) AS total,
    COUNT(where m.isRead = false) AS totalUnreadMessages,
    e.FirstName, e.LastName
FROM 
    [Messages] AS m 
INNER JOIN
    Employees AS e ON m.SenderId = e.UserId
GROUP BY
    e.EmployeeId, e.FirstName, e.LastName

Answer 1

你几乎就在那里，使用case表达式进行条件计数：

SELECT Count(*) as total,Count(case when m.isRead = false then 1 end) as totalUnreadMessage,
    e.FirstName,e.LastName
FROM [Messages] as m 
inner join Employees as e on m.SenderId=e.UserId
group by e.EmployeeId, e.FirstName,e.LastName

Answer 2

我会把它写成：

AttributeError: 'unicode' object has no attribute 'p'

Answer 3

您可以计算它们，然后使用原始选择加入它们：

SELECT Count(*) as total,tbNotRead.NotReadNumber AS totalUnreadMessage, e.FirstName,e.LastName
FROM [Messages] as m 
INNER JOIN Employees as e ON m.SenderId=e.UserId
LEFT OUTER JOIN
(
  SELECT m.Sender, COUNT(*) NotReadNumber
  FROM [Messages] msgs
  WHERE m.isRead = FALSE
  GROUP BY m.Sender
)tbNotRead
ON tbNotRear.Sender = m.sender
GROUP BY e.EmployeeId, e.FirstName,e.LastName, tbNotRead.NotReadNumber

在count sql中使用where

3 个答案: