SELECT DISTINCT Campaign_id
FROM Impressions
WHERE Date BETWEEN '2015-03-01' AND '2015-03-31' ;
以上查询为我提供了2015-03-01和2015-03-31之间任何日期活跃的Campaign_id
的结果。
如果所有在2015-03-01和2015-03-31之间的日期一直有效,我希望结果集包含campaign_id
。
我该怎么做?
答案 0 :(得分:6)
假设DATE
是DATE
数据类型并且没有时间组件。
DECLARE @Start DATE = '2015-03-01',
@End DATE = '2015-03-31'
SELECT Campaign_id
FROM Impressions
WHERE Date BETWEEN @Start AND @End
GROUP BY Campaign_id
HAVING COUNT(DISTINCT Date) = 1 + DATEDIFF(DAY, @Start, @End);
或没有变量的版本
SELECT Campaign_id
FROM Impressions
CROSS APPLY (VALUES ({ d '2015-03-01' },
{ d '2015-03-31' })) V([Start], [End])
WHERE [Date] BETWEEN [Start] AND [End]
GROUP BY Campaign_id, [Start], [End]
HAVING COUNT(DISTINCT Date) = 1 + DATEDIFF(DAY, [Start], [End]);
答案 1 :(得分:4)
将HAVING
子句与COUNT(DISTINCT)
一起使用:
SELECT Campaign_id
FROM Impressions
WHERE Date between '2015-03-01' and '2015-03-31'
GROUP BY Campaign_id
HAVING COUNT(DISTINCT Date) = 31;
您还应该查看Aaron Betrand的blog post,了解为什么在日期中使用BETWEEN
是一个坏主意。
您可以通过执行以下操作来安排查询,只提及日期:
WITH params as (
SELECT CAST('2015-03-01' as DATE) as date1, CAST('2015-03-31' as DATE) date2
)
SELECT i.Campaign_id
FROM params CROSS JOIN
Impressions i
WHERE i.Date >= params.Date1 and i.Date < DATEADD(day, 1, params.Date2)
GROUP BY i.Campaign_id, params.date1, params.date2
HAVING COUNT(DISTINCT i.Date) = 1 + DATEDIFF(day, params.date1, params.date2);
注意:在这种情况下,有些人更喜欢JOIN
到CROSS JOIN
。根据习惯,我总是使用CROSS JOIN
在查询中放置参数CTE。