如果结束日期比下一行开始日期早 1 天，如何添加日期

Question

仅当第一行的结束日期是下一行开始日期的前一天时，我才尝试合计两行的天数。如果第一行的结束日期不是第二行开始日期的前一天，那么我想从查询中排除这两行。因此，对于下面的示例，我的总和应该是 365。我的活动表有数千行具有不同的名称和 orderid，我需要在执行此任务的同时保持每个人的 orderID 的完整性。

name	orderID	Start date	end date
Joe Smith	1	2020-01-01	2020-09-30
Joe Smith	2	2020-10-01	2020-12-30

Answer 1

基于@Gordon-Linoff 的回答和他关于间隙和孤岛的线索，但我在添加测试数据时遇到了错误，正如评论中提到的那样。我也用过这个帖子 https://bertwagner.com/posts/gaps-and-islands/

-- test data
DECLARE @t TABLE (name varchar(50), orderID int, StartDate DateTime, EndDate DateTime);

INSERT INTO @t
SELECT 'Joe Smith', 1, '2020-01-01', '2020-09-30' UNION
SELECT 'Joe Smith', 2, '2020-10-01', '2020-12-30' UNION
SELECT 'Joe Smith', 3, '2021-01-01', '2021-09-30' UNION
SELECT 'Joe Smith', 4, '2021-10-01', '2021-12-31' UNION
SELECT 'Joe Smith', 5, '2022-01-01', '2022-09-30' UNION
SELECT 'Jane Doe', 6, '2020-01-01', '2020-09-30' UNION
SELECT 'Jane Doe',  7, '2020-11-01', '2020-12-30';

-- caculate the difference add 1 because EndDate is inclusive (ends on the start of next day)
SELECT t.*, d.IslandStartDate, d.IslandEndDate, DATEDIFF(DAY, IslandStartDate, IslandEndDate) + 1 AS Days FROM (
-- return the minimum and maximum start and end dates 
SELECT
    name, 
    MIN(StartDate) AS IslandStartDate,
    MAX(EndDate) AS IslandEndDate
FROM (
    SELECT
        *,
        -- indicates when a new island begins by looking if the current row's StartDate occurs after the previous row's EndDate
        CASE WHEN Groups.PreviousEndDate >= DATEADD(DAY, -1, StartDate) THEN 0 ELSE 1 END AS IslandStartInd,
        -- indicates which island number the current row belongs to
        SUM(CASE WHEN Groups.PreviousEndDate >= DATEADD(DAY, -1, StartDate) THEN 0 ELSE 1 END) OVER (PARTITION BY name ORDER BY Groups.RN) AS IslandId
    FROM
    (
        -- create a row number column based on the sequence of start and end dates, as well as bring the previous row's EndDate to the current row
        SELECT
            name,
            orderID,
            ROW_NUMBER() OVER(PARTITION BY name ORDER BY StartDate,EndDate) AS RN,
            StartDate,
            EndDate,
            LAG(EndDate,1) OVER (PARTITION BY name ORDER BY StartDate, EndDate) AS PreviousEndDate
        FROM
            @t
    ) Groups
) Islands
GROUP BY
    name, IslandId
) d
-- join to get the orderID back
INNER JOIN @t t ON d.name = t.name AND t.StartDate >= d.IslandStartDate AND t.EndDate <= d.IslandEndDate
ORDER BY IslandStartDate, name

Answer 2

如果您想要满足您的条件的多行组，那么这是一种间隙和孤岛问题。

您要做的是为行分配一个“孤岛”编号。您可以通过在上一行达到峰值来查看它是否符合您的条件。如果没有，则开始一个岛屿。岛屿起点的累积总和会为这些组分配一个岛屿编号。

剩下的就是聚合

select name, min(startdate), max(enddate),
       datediff(day, min(startdate), max(enddate)) as num_days
from (select t.*,
             sum(case when prev_end_date <> dateadd(day, -1, startdate) then 0 else 1 end) over 
                 (partition by name order by startdate) as island
      from (select t.*,
                   lag(end_date) over (partition by name order by start_date) as prev_end_date
            from t
           ) t
     ) t
group by name, island
having count(*) > 1;