有这样一个操作表：
+----+------------+-----------+--------+---------------------------------+---------------------------------+---------------------------------+
| id | oper_date | parent_id | amount | p1 | p2 | p3 |
+----+------------+-----------+--------+---------------------------------+---------------------------------+---------------------------------+
| 1 | 01.09.2018 | 1 | 5 | p1 value at the date 01.09.2018 | p2 value at the date 01.09.2018 | p3 value at the date 01.09.2018 |
| 2 | 01.09.2018 | | 2 | p1 value at the date 01.09.2018 | p2 value at the date 01.09.2018 | p3 value at the date 01.09.2018 |
| 3 | 02.09.2018 | 1 | 7 | p1 value at the date 02.09.2018 | p2 value at the date 02.09.2018 | p3 value at the date 01.09.2018 |
| 4 | 02.09.2018 | | 4 | p1 value at the date 02.09.2018 | p2 value at the date 02.09.2018 | p3 value at the date 01.09.2018 |
| 5 | 02.09.2018 | | 6 | p1 value at the date 02.09.2018 | p2 value at the date 02.09.2018 | p3 value at the date 01.09.2018 |
| 6 | 02.09.2018 | | 10 | p1 value at the date 02.09.2018 | p2 value at the date 02.09.2018 | p3 value at the date 01.09.2018 |
| 7 | 03.09.2018 | | 4 | p1 value at the date 03.09.2018 | p2 value at the date 03.09.2018 | p3 value at the date 01.09.2018 |
| 8 | 03.09.2018 | 1 | 3 | p1 value at the date 03.09.2018 | p2 value at the date 03.09.2018 | p3 value at the date 01.09.2018 |
| 9 | 03.09.2018 | | 2 | p1 value at the date 03.09.2018 | p2 value at the date 03.09.2018 | p3 value at the date 01.09.2018 |
| 10 | 04.09.2018 | | 6 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
| 11 | 04.09.2018 | 1 | 7 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
| 12 | 04.09.2018 | | 11 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
| 13 | 04.09.2018 | | 4 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
| 14 | 05.09.2018 | | 8 | p1 value at the date 05.09.2018 | p2 value at the date 05.09.2018 | p3 value at the date 05.09.2018 |
| 15 | 05.09.2018 | 1 | 2 | p1 value at the date 05.09.2018 | p2 value at the date 05.09.2018 | p3 value at the date 05.09.2018 |
| 16 | 05.09.2018 | | 6 | p1 value at the date 05.09.2018 | p2 value at the date 05.09.2018 | p3 value at the date 05.09.2018 |
+----+------------+-----------+--------+---------------------------------+---------------------------------+---------------------------------+
获取oper_date为<=特定日期的行时，
parent_id字段具有相同非空值的几行
以便在amount字段中显示该日期的所有操作的总和。
在字段p1/p2/p3/id/oper_date中，我需要显示此日期的这些字段的值。
例如，如果我选择oper_date <= 04.09.2018的行，则需要显示以下结果：
+----+------------+-----------+--------+---------------------------------+---------------------------------+---------------------------------+
| id | oper_date | parent_id | amount | p1 | p2 | p3 |
+----+------------+-----------+--------+---------------------------------+---------------------------------+---------------------------------+
| 2 | 01.09.2018 | | 2 | p1 value at the date 01.09.2018 | p2 value at the date 01.09.2018 | p3 value at the date 01.09.2018 |
| 4 | 02.09.2018 | | 4 | p1 value at the date 02.09.2018 | p2 value at the date 02.09.2018 | p3 value at the date 01.09.2018 |
| 5 | 02.09.2018 | | 6 | p1 value at the date 02.09.2018 | p2 value at the date 02.09.2018 | p3 value at the date 01.09.2018 |
| 6 | 02.09.2018 | | 10 | p1 value at the date 02.09.2018 | p2 value at the date 02.09.2018 | p3 value at the date 01.09.2018 |
| 7 | 03.09.2018 | | 4 | p1 value at the date 03.09.2018 | p2 value at the date 03.09.2018 | p3 value at the date 01.09.2018 |
| 9 | 03.09.2018 | | 2 | p1 value at the date 03.09.2018 | p2 value at the date 03.09.2018 | p3 value at the date 01.09.2018 |
| 10 | 04.09.2018 | | 6 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
| 11 | 04.09.2018 | 1 | 22 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
| 12 | 04.09.2018 | | 11 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
| 13 | 04.09.2018 | | 4 | p1 value at the date 04.09.2018 | p2 value at the date 04.09.2018 | p3 value at the date 04.09.2018 |
+----+------------+-----------+--------+---------------------------------+---------------------------------+---------------------------------+
对于id = 11的行，字段AMOUNT显示值的总和（22 = 5 + 7 + 3 + 7）
parent_id = 1并且oper_date为<= 04/09/2018的所有行中的所有行。
如何使用sql-query执行此操作？
P.S。设计表时，对于parent_id字段中的父行，我可以保留null或其id。
答案 0 :(得分：0)
假设您已将感兴趣的日期（在本例中为04/09/2018）另存为@SetDate。根据您的预期输出，我想您希望所有在parent_id或之前带有NULL @SetDate的记录，尤其是要返回@SetDate的记录，而您想要@SetDate上和之前的所有金额之和，其中parent_id作为单独记录不为NULL。
SELECT id, oper_date, parent_id, amount, p1, p2, p3
FROM your_table
WHERE parent_id IS NULL AND oper_date <= @SetDate
UNION
SELECT
(SELECT id FROM your_table WHERE oper_date = @SetDate AND parent_id = 1 LIMIT 1) AS id,
@SetDate AS oper_date,
1 AS parent_id,
(SELECT SUM(Amount) FROM your_table WHERE oper_date <= @SetDate AND parent_id = 1) AS amount,
(SELECT p1 FROM your_table WHERE oper_date = @SetDate AND parent_id = 1 LIMIT 1) AS p1,
(SELECT p2 FROM your_table WHERE oper_date = @SetDate AND parent_id = 1 LIMIT 1) AS p2,
(SELECT p3 FROM your_table WHERE oper_date = @SetDate AND parent_id = 1 LIMIT 1) AS p3
FROM your_table
答案 1 :(得分：0)
在此，“解析函数划分依据”将很有帮助。使用按ID排序将使总和递增，最大ID将具有总和。按照您的解释
使用MY_VIEW AS
（SELECT id，oper_date，parent_id，SUM（amount）超过（partition by parent_id order by id）AS数量，p1，p2，p3
来自my_table
如果parent_id不为NULL并且oper_date <=
）
SELECT * FROM MY_VIEW WHERE id =
（从MY_VIEW中选择MAX（id）个
）
联合
SELECT id，oper_date，parent_id，amount，p1，p2，p3
来自my_table
其中parent_id为NULL并且oper_date <=
按ID排序