要查询的SUM查询,直到SUM(users_count)达到1000

Question

我需要一个SQL查询来从我的消息队列中选择行,直到SUM(users_count)达到最多1000. 但是如果只返回一行并且该行的users_count大于1000,则没有问题.

我需要类似的东西:(我添加了自己的关键字)

SELECT * FROM `messages_queue` UNTIL SUM(users_count) < 1000 AT LEAST 1 ROW

这是我的表结构:

messages_queue
- msg_id
- msg_body - users_count
(消息接收者的数量)
- 时间(插入时间)

Answer 1

此解决方案将执行累积总和,当总和超过1000时停止:

SELECT NULL AS users_count, NULL AS total
  FROM dual
 WHERE (@total := 0)
 UNION
SELECT users_count, @total := @total + users_count AS total
  FROM messages_queue
 WHERE @total < 1000;

这意味着如果你有两个值,比如800,总和将是1600.第一个SELECT只是初始化@total变量.

如果你想防止总和超过1000,除了单行的值大于1000的情况,那么我认为这是有效的,尽管你需要对它进行一些严格的测试:

SELECT NULL AS users_count, NULL AS total, NULL AS found
  FROM dual
 WHERE (@total := 0 OR @found := 0)
 UNION
SELECT users_count, @total AS total, @found := 1 AS found
  FROM messages_queue
 WHERE (@total := @total + users_count)
   AND @total < 1000
 UNION
SELECT users_count, users_count AS total, 0 AS found
  FROM messages_queue
 WHERE IF(@found = 0, @found := 1, 0);

Answer 2

我试图将此作为评论添加到 Mike 的答案，但是，变量的 @ 符号存在问题。

为了借鉴 Mike 的答案，实际上可以通过初始化 FROM 子句中的变量来缩短查询，例如：

SELECT users_count, @total := @total + users_count AS total
    FROM (messages_queue, (select @total := 0) t)
WHERE @total < 1000;