8

我创建了一个与我原来的场景相同的示例表。数据库“test”中的表名“records”

数据库时区设置为 UTC (SET time_zone = "+00:00";)

`records` (`id`, `name`, `time_created`)

(1, '议案', '2017-09-13 16:20:41'),
(2, '议案', '2017-09-13 16:20:57'),
(3, '议案', '2017-09-13 16:21:24'),
(4, '议案', '2017-09-13 16:21:40'),
(5, '议案', '2017-09-13 16:26:38'),
(6, '议案', '2017-09-13 17:09:00'),
(7, '议案', '2017-09-13 17:09:16'),
(8, '议案', '2017-09-13 22:14:37'),
(9, '议案', '2017-09-13 22:23:53'),
(10, '议案', '2017-09-13 22:24:08'),
(11, '议案', '2017-09-13 22:24:24'),
(12, '议案', '2017-09-13 23:45:17'),
(13, '议案', '2017-09-13 23:45:36'),
(14, '议案', '2017-09-13 23:45:54'),
(15, '议案', '2017-09-14 00:07:09'),
(16, '议案', '2017-09-14 00:07:24'),
(17, '议案', '2017-09-14 00:07:42'),
(18, '议案', '2017-09-19 09:42:11'),
(19, '议案', '2017-09-19 09:42:27'),
(20, '议案', '2017-09-19 09:42:44'),
(21, '议案', '2017-09-19 11:21:08'),
(22, '议案', '2017-09-19 11:21:23'),
(23, '议案', '2017-09-19 11:21:38'),
(24, '议案', '2017-09-19 11:21:54'),
(25, '议案', '2017-09-19 11:48:13'),
(26, '议案', '2017-09-13 16:20:41'),
(27, '议案', '2017-09-13 16:20:57'),
(28, '议案', '2017-09-13 16:21:24'),
(29, '议案', '2017-09-13 16:21:40'),
(30, '议案', '2017-09-13 16:26:38'),
(31, '议案', '2017-09-13 17:09:00'),
(32, '议案', '2017-09-13 17:09:16'),
(33, '议案', '2017-09-13 22:14:37'),
(34, '议案', '2017-09-13 22:23:53'),
(35, '议案', '2017-09-13 22:24:08'),
(36, '议案', '2017-09-13 22:24:24'),
(37, '议案', '2017-09-13 23:45:17'),
(38, '议案', '2017-09-13 23:45:36'),
(39, '议案', '2017-09-13 23:45:54'),
(40, '议案', '2017-09-14 00:07:09'),
(41, '议案', '2017-09-14 00:07:24'),
(42, '议案', '2017-09-14 00:07:42'),
(43, '议案', '2017-09-19 09:42:11'),
(44, '议案', '2017-09-19 09:42:27'),
(45, '议案', '2017-09-19 09:42:44'),
(46, '议案', '2017-09-19 11:21:08'),
(47, '议案', '2017-09-19 11:21:23'),
(48, '议案', '2017-09-19 11:21:38'),
(49, '议案', '2017-09-19 11:21:54'),
(50, '议案', '2017-09-19 11:48:13');

我必须使用 CONVERT_TZ（ mysql 转换函数）使用时间转换进行两个查询

我需要两个查询：1. 获取日期“今天”和“今天 - 30 天前”之间的记录 2. 获取给定日期的记录，例如“2017-09-14”

我尝试过以下查询

  1. 日期之间的记录：
SELECT * FROM test.records WHERE name LIKE '%motion%'
AND CONVERT_TZ(time_created ,'+00:00','-7:0') BETWEEN DATE_SUB(CONVERT_TZ('2017-09-20 11:48:13' ,'+00:00','-7:0' ), 间隔 30 天) 和 CONVERT_TZ('2017-09-20 11:48:13','+00:00','-7:0')
GROUP BY DATE(time_created) ORDER BY ID DESC;

结果 ：
18议案 2017-09-19 09:42:11
15 议案 2017-09-14 00:07:09
1 议案 2017-09-13 16:20:41
  1. 给定日期的记录：
SELECT name,id, CONVERT_TZ(time_created ,'+00:00','-7:0') as time_created,
DATE_FORMAT( CONVERT_TZ(time_created ,'+00:00','-7:0') , '%h:%i:%s %p') as new_format_time
FROM test.records WHERE name LIKE '%motion%'
AND DATE(CONVERT_TZ(time_created,'+00:00','-7:0')) = '2017-09-14' ORDER BY ID DESC
结果：
0 条记录
4

1 回答 1

2

首先，假设您的系统在 UTC 运行

   SELECT * FROM test.records
   WHERE name LIKE '%motion%' AND
      time_created >= DATE_SUB(CONVERT_TZ(NOW(),'+00:00','-7:0') , INTERVAL 30 DAY)
   GROUP BY DATE(time_created) ORDER BY ID DESC;

如果 NOW() 是当前时区

   SELECT * FROM test.records
   WHERE name LIKE '%motion%' AND
      CONVERT_TZ(time_created,'+00:00','-7:0') >= DATE_SUB(NOW(), INTERVAL 30 DAY)
   GROUP BY DATE(time_created) ORDER BY ID DESC;

--- 201/09/25 编辑 ---

其次，知道要与哪个时区进行比较非常重要。

select DATE(CONVERT_TZ('2017-09-14 00:07:25','+00:00', '-07:00'));

将日期放在前一天或2017-09-13（我们已将比较移至等式的另一边，因此我们必须更改顺序

更改>=为=并仅使用DATE()

   SELECT *
   FROM test.records
   WHERE name LIKE '%motion%' AND
     DATE(CONVERT_TZ(time_created,'-7:0', '+00:00')) = '2017-09-14'
   GROUP BY DATE(time_created) ORDER BY ID DESC;

最简单的比较是针对 UTC，公式为

   SELECT *
   FROM test.records
   WHERE name LIKE '%motion%' AND
     DATE(time_created) = '2017-09-14'
   GROUP BY DATE(time_created) ORDER BY ID DESC;
于 2017-09-22T23:29:06.167 回答