Question

我有一个查询，我试图找出在数据库的某个时间范围内查看某些内容的次数。

到目前为止，这是我的查询：

USE DB1
SELECT DISTINCT NAME,
COUNT(*) AS COUNT
FROM LOGS
WHERE EVENT = 'Viewed'
AND TIME BETWEEN '2015-04-22 00:00:00.000' AND '2015-07-24 00:00:00.000'
GROUP BY NAME
ORDER BY COUNT DESC;

此查询成功告诉我在该3个月范围内查看了多少次，但是当我将时间范围缩短为三个不同的查询时，该特定行更改为：

TIME BETWEEN '2015-04-22 00:00:00.000' AND '2015-05-22 00:00:00.000';
TIME BETWEEN '2015-05-22 00:00:00.000' AND '2015-06-22 00:00:00.000';
TIME BETWEEN '2015-06-22 00:00:00.000' AND '2015-04-24 00:00:00.000';

然后，当我加上不同的值时，它们并不相等。实际上，它们比我在三个月范围内的第一个查询获得的行数要高得多。我假设某些值重叠。我怎么能绕过这个？我已经检查了我的工作，我很确定第一个查询在时间范围内给出了正确的行数，但我需要更具体的数据

Answer 1

col BETWEEN a AND b计算为a <= col <= b，因此您的时间范围内存在重叠间隔

将其从午夜改为午夜之前的一秒：

TIME BETWEEN '2015-04-22 00:00:00.000' AND '2015-04-22 23:59:59.999';
TIME BETWEEN '2015-05-22 00:00:00.000' AND '2015-05-22 23:59:59.999';
TIME BETWEEN '2015-06-22 00:00:00.000' AND '2015-04-24 23:59:59.999';

<强>更新

试试这个：

TIME BETWEEN '2015-04-22 00:00:00' AND '2015-04-22 23:59:59';
TIME BETWEEN '2015-05-22 00:00:00' AND '2015-05-22 23:59:59';
TIME BETWEEN '2015-06-22 00:00:00' AND '2015-04-24 23:59:59';

Answer 2

X BETWEEN Y AND Z

转换为

X >= Y AND X <= Z

因此，您应该对整个范围内的句点使用不同的查询，例如：

TIME >= '2015-04-22 00:00:00.000' AND TIME < '2015-05-22 00:00:00.000';
TIME >= '2015-05-22 00:00:00.000' AND TIME < '2015-06-22 00:00:00.000';
TIME >= '2015-06-22 00:00:00.000' AND TIME <= '2015-07-24 00:00:00.000';

上一段时间的逻辑取决于你...你可以将其重新排列为：

TIME >= '2015-06-22 00:00:00.000' AND TIME < '2015-07-25 00:00:01.000';

重叠计数特定时间范围

2 个答案: