Question

我在MSSQL中有两个表;

视频： ID，艺术家，标题

runlog：视频ID，日期时间

我试图整理的结果显示：

ID，艺术家，标题，在过去1小时内播放的时间ID，在过去12小时内播放的时间ID，过去24小时内播放的时间ID，时代艺术家

我找到了除最后一栏之外的所有内容。

SELECT ID,Artist,Title,
(select COUNT(*) from runlog where runlog.videoID = videos.id ) as plays,
(select COUNT(*) from runlog where datetime > DATEADD(hh,-1,GETDATE()) and runlog.videoID = videos.id) as hr1plays ,
(select COUNT(*) from runlog where datetime > DATEADD(hh,-12,GETDATE()) and runlog.videoID = videos.id) as hr12plays ,
(select COUNT(*) from runlog where datetime > DATEADD(hh,-24,GETDATE()) and runlog.videoID = videos.id) as day1plays ,

(select COUNT(*) from runlog left join videos on videos.ID = runlog.videoID 
where videos.artist = Artist and runlog.videoID = videos.id) as artistplays 


FROM videos

显然，最后一件事只是显示了runlog表中的记录总数，但是我想把它放在我想要做的事情上。

我首先尝试寻找解决方案，但我找不到任何可以帮助的方法。

谢谢！

Answer 1

我认为你可以通过聚合和窗口函数来实现这一点，结合起来：

SELECT v.ID, v.Artist, v.Title,
       count(rl.videoid) as plays,
       sum(case when datetime > DATEADD(hh, -1 ,GETDATE()) then 1 else 0
           end) as hr1plays,
       sum(case when datetime > DATEADD(hh, -12, GETDATE()) then 1 else 0
           end) as hr12plays,
       sum(case when datetime > DATEADD(hh, -24, GETDATE()) then 1 else 0
           end) as hr24plays,
       sum(count(rl.videoid)) over (partition by v.Artist) as artistplays
FROM videos v left outer join
     runlog rl
     on v.id = rl.videoId
group by v.ID, v.Artist, v.Title;

尝试计算连接表中不同字段的数量

1 个答案: