我怎么做'分组'部分匹配

Question

我在SQL-server中有一个表，其中包含相同字段中的项目代码和子项目代码。

结构是这样的

+----+------+-------------+-------+--------+--------+
| id | date | projectcode | debit | credit | budget |
+----+------+-------------+-------+--------+--------+
| 1  | bla  | A100        | bla
| 2  | bla  | A100.01     |
| 3  | bla  | A112        |
| 4  | bla  | A112.02

如何选择这样的

SELECT projectcode
  , sum(debit) as debit
  , sum(credit) as credit
  , sum(budget) as budget
FROM table1
GROUP BY -insert-answer-here-

我希望输出到group by A100和A100.01和A100.x以及A112 + A112.x

我该怎么做？

我无法控制桌子的结构。

Answer 1

GROUP BY LEFT(projectcode ,CHARINDEX('.',projectcode  + '.')-1)

Answer 2

如果项目代码始终遵循相同的模式（cnnn / cnnn.nn），您可以获得前四个字符：

group by substring(projectcode, 1, 4)

Answer 3

也许这会奏效：

SELECT Substring(projectcode, 1, 4) as Project
  , sum(debit) as debit
  , sum(credit) as credit
  , sum(budget) as budget
FROM table1
GROUP BY Substring(projectcode, 1, 4)