Question

这是我的问题;

我运行此查询并获得13469个具有多行的用户，其中'FirstTimeBooking'= 1：

Select CustUserId, COUNT(CustUserId)
FROM Bookings with (nolock)
WHERE FirstTimeBooking = 1
GROUP BY CustUserId
HAVING COUNT(CustUserId) > 1
ORDER BY COUNT(CustUserId) DESC

我需要更新所有这些记录，为每个行设置FirstTimeBooking = 0除了每个用户的Min（BookingDate）（这是实际的FirstTimeBooking）。任何帮助将不胜感激。

Answer 1

我试图保持简单并保持语法：

CREATE TABLE Bookings 
(
 CustUserID int
,FirstTimeBooking int
,BookingDate datetime
)

INSERT INTO Bookings VALUES 
 (1, 1, '20170101') 
,(1, 1, '20170101') -- Duplicate on same date
,(2, 1, '20170201')
,(2, 1, '20170301') -- On a second date
,(3, 1, '20170401')

UPDATE a
SET FirstTimeBooking = 0
FROM Bookings as a
LEFT OUTER JOIN (
    -- Only CustUserID 1 and 2
    SELECT CustUserID, min(BookingDate) as BookingDate
    FROM Bookings
    WHERE FirstTimeBooking = 1
    GROUP BY CustUserID
    HAVING COUNT(*) > 1
) as b
on a.CustUserID = b.CustUserID
and a.BookingDate = b.BookingDate
WHERE b.BookingDate IS NULL -- only where the first booking date has not been found in subquery

-- Still got the Problem with CustUserID
SELECT * 
FROM Bookings
WHERE CustUserID = 1

UPDATE b
set FirstTimeBooking = 0
FROM (
SELECT *, ROW_NUMBER() OVER (PARTITION BY CustUserID ORDER BY CustUserID) As Rown
FROM Bookings as b
WHERE FirstTimeBooking = 1
) as b
WHERE Rown = 2

-- This is now also solved
SELECT * 
FROM Bookings
WHERE CustUserID = 1

也许你没有两个相同的预订日期，但我加上这个以防万一。

此致 TGR

更新重复记录，保持记录与最小（日期）相同

1 个答案: