SQL Server 5 中lead() 函数的替代方法

Question

我想要与窗口函数的前导函数相同的输出。我想为 SQL Server 5.0 版做这件事

表名：表

EventID  CID  Date
123      1    01-12-2020
123      2    01-12-2020
123      3    01-12-2020
345      2    05-12-2020
345      4    05-12-2020
456      1    07-12-2020
456      4    07-12-2020
567      1    08-12-2020

输出

ID      CID     Date       ColumnA
123      1    01-12-2020    07-12-2020
456      1    07-12-2020    08-12-2020
567      1    08-12-2020    Null
123      2    01-12-2020    05-12-2020
345      2    05-12-2020    Null
123      3    01-12-2020    Null
345      4    05-12-2020    07-12-2020
456      4    07-12-2020    Null

Answer 1

您可以使用相关子查询来代替无法访问LEAD() 函数：

SELECT
    EventID AS ID,
    CID,
    Date,
    (SELECT TOP 1 t2.Date FROM yourTable t2
     WHERE t2.CID = t1.CID AND t2.Date > t1.Date
     ORDER BY t2.Date) AS ColumnA
FROM yourTable t1
ORDER BY
    CID,
    Date;

Demo

编辑：

您为 SQL Server 标记了您的问题，并解释了您收到此答案和其他答案的原因。如果你的 RDBMS 真的是 MySQL，那么使用这个版本：

SELECT
    EventID AS ID,
    CID,
    Date,
    (SELECT t2.Date FROM yourTable t2
     WHERE t2.CID = t1.CID AND t2.Date > t1.Date
     ORDER BY t2.Date LIMIT 1) AS ColumnA
FROM yourTable t1
ORDER BY
    CID,
    Date;

Answer 2

试试这个：

WITH DataSource AS
(
    SELECT *
          ,ROW_NUMBER() OVER (PARTITION BY [CID] ORDER BY [Date] ASC) AS [RowID]
    FROM [Table] 
)
SELECT DS1.EventID  
      ,DS1.[CID]
      ,DS1.[Date]
      ,DS2.[Date] AS ColumnA
FROM DataSource DS1
LEFT JOIN DataSource DS2
    ON DS1.[RowID] = DS2.[RowID] - 1