mysql – 如何使用IN子句检索不同日期的行
作者:互联网
我的表结构:
+-----------------+--------------+------+-----+-------------------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-----------------+--------------+------+-----+-------------------+----------------+
| id | int(11) | NO | PRI | NULL | auto_increment |
| user | varchar(100) | NO | | NULL | |
| uniq_name | varchar(100) | NO | | NULL | |
| info_name | varchar(100) | NO | | NULL | |
| delay_time | int(11) | NO | | NULL | |
| track_time | timestamp | NO | | CURRENT_TIMESTAMP | |
+-----------------+--------------+------+-----+-------------------+----------------+
我必须在过去七天内从表中检索数据,而不是总结一次.特定日期的数据,为此现在我必须使用like运算符七次触发相同的查询:
SELECT COUNT(DISTINCT `user`) FROM `tracks` WHERE `track_time` LIKE '2012-11-12%';
这不是一个好方法.我正在尝试做自己,也在论坛上讨论过,可能是用IN运算符或使用date1< date2< ....过去七天的日期.为此,我在PHP中使用一个函数,我在其中提供日期参数,但查询将花费大量时间在Web上显示结果.请看看这种情况.
解决方法:
如果您正在进行一系列连续日期,则无需使用IN运算符.
如果你用它来计算七天前的午夜
DATE(NOW()) - INTERVAL 7 DAY + INTERVAL 0 SECOND
这是你得到的回报:
mysql> SELECT DATE(NOW()) - INTERVAL 7 DAY + INTERVAL 0 SECOND;
+--------------------------------------------------+
| DATE(NOW()) - INTERVAL 7 DAY + INTERVAL 0 SECOND |
+--------------------------------------------------+
| 2012-11-21 00:00:00 |
+--------------------------------------------------+
1 row in set (0.01 sec)
mysql>
您的查询将如下所示:
SELECT COUNT(DISTINCT `user`) FROM `tracks`
WHERE `track_time` >= ( DATE(NOW()) - INTERVAL 7 DAY + INTERVAL 0 SECOND );
您还需要一个索引来支持WHERE和COUNT
ALTER TABLE tracks ADD INDEX track_time_user_ndx (track_time,user);
然后,查询应该更快.
如果您正在进行一些早期的连续日期,只需提供从午夜到午夜的第一个和最后一个日期.例如,对于2012-11-12到2012-11-19的范围:
SELECT COUNT(DISTINCT `user`) FROM `tracks`
WHERE `track_time` >= '2012-11-12 00:00:00'
AND `track_time` < '2012-11-19 00:00:00';
如果您正在执行一系列非连续日期,只需使用UNION构造查询.例如,对于日期2012-11-12,2012-11-19,2012-11-26:
SELECT COUNT(DISTINCT `user`) FROM
(
SELECT user,track_time FROM `tracks`
WHERE `track_time` >= '2012-11-12 00:00:00'
AND `track_time` <= '2012-11-12 23:59:59'
UNION
SELECT user,track_time FROM `tracks`
WHERE `track_time` >= '2012-11-19 00:00:00'
AND `track_time` <= '2012-11-19 23:59:59'
UNION
SELECT user,track_time FROM `tracks`
WHERE `track_time` >= '2012-11-26 00:00:00'
AND `track_time` <= '2012-11-26 23:59:59'
) A;
试试看 !!!
更新2012-11-27 13:52
如果您想在2012-11-12只有一个日期计算不同的用户数
SELECT COUNT(DISTINCT user) usercount FROM `tracks`
WHERE `track_time` >= '2012-11-12 00:00:00'
AND `track_time` <= '2012-11-12 23:59:59';
如果你想要连续7个日期的不同用户数2012-11-12-2012-11-18(你不能包括2012-11-19)
SELECT
DATE(track_time) track_date,
COUNT(DISTINCT user) usercount
FROM
(
SELECT user,track_time FROM `tracks`
WHERE `track_time` >= '2012-11-12 00:00:00'
AND `track_time` < '2012-11-19 00:00:00'
) A GROUP BY DATE(track_time);
更新2012-11-27 14:41美国东部时间
如果您希望从NOW()开始动态完成此操作,请尝试以下操作:
SELECT
DATE(track_time) track_date,
COUNT(DISTINCT user) usercount
FROM
(
SELECT user,track_time FROM `tracks`
WHERE `track_time` >= ( DATE(NOW()) - INTERVAL 7 DAY + INTERVAL 0 SECOND )
) A GROUP BY DATE(track_time);
标签:php,mysql,phpmyadmin,date-format 来源: https://codeday.me/bug/20190806/1596661.html