Question

我有一张看起来像这样的表

ID   SchoolName                   SchoolURL     updflag  orderflag
1    Brighton School          Brighton.edu   0      2
2    Andover school               Andover.edu    0       2
3    Fremont school               Fremont.edu    0       2
4    Tulsa Public school          Tulsa.edu      0       1
5    Crescent school          Crescent.edu    0      1
55   NH school                    NH.edu          0      2
61   PennState Univ               pennstate.edu   0      1
2    john Adams public School     Andover.edu     0      2
2    Estabrook school         Andover.edu      0     2

我想只打印没有重复的记录，即不应打印ID = 2。我想首先选择orderflag = 2的记录，然后显示orderflag = 1。我还希望一次只显示10条记录。如果我更新记录，updflag会更新为1。

SELECT   TOP 10 r.orderflag,r.[id],
                     r.schoolurl,
                     updatedflag
                    FROM     [dbo].[schools] AS r

                    INNER JOIN
                    (SELECT   id
                    FROM     [dbo].[schools]
                    GROUP BY id
                    HAVING   COUNT(*) = 1) AS k
                    ON k.id = r.id
                    WHERE   updflag = 0 
                    GROUP BY r.orderflag,r.id, School URL, updflag order by r.orderflag desc;

我有数百万条记录，我希望先处理orderflag = 2，然后再处理orderflag = 1。我怎么能这样做？

由于 [R

Answer 1

一种可能的方法是简单地排除COUNT > 1存在ID的那些：

SELECT ... FROM schools
WHERE ID NOT IN (
    SELECT ID FROM schools
    GROUP BY ID
    HAVING COUNT(ID) > 1
)

接下来，要处理订单标记，您可以执行以下操作：

SELECT ... FROM schools
WHERE orderflag = 2 AND
    ID NOT IN (
        SELECT ID FROM schools
        WHERE orderflag = 2
        GROUP BY ID
        HAVING COUNT(ID) > 1
    )
UNION
SELECT ... FROM schools
WHERE orderflag = 1 AND
    ID NOT IN (
        SELECT ID FROM schools
        WHERE orderflag = 1
        GROUP BY ID
        HAVING COUNT(ID) > 1
    )

sqlserver处理记录的顺序

1 个答案: