SQL Server：从分隔符中提取字符串

Question

我正在查询一个名为 Description 的列，我需要从每个“-”分隔符中提取字符串；

例如

Description            
---------------------------------
abc@abc.com - Invoice - A12222203
FGH@fgh.com - Credit -  C12222333

因此理想情况下需要将每个段提取到三个单独的列中；

例如

Email       | Doc Type | Ref       
------------+----------+----------
abc@abc.com | Invoice  | A12222203 
FGH@fgh.com | Credit   | C12222333

我已设法使用

提取电子邮件地址

Substring(SL_Reference,0,charindex('-',SL_Reference))Email

有什么想法可以将剩余的两个部分分成单独的列（即文档类型和参考）？

非常感谢

Answer 1

必须有数百种方法来进行这种字符串操作，这里有几种。

这使用apply 来获取每个分隔符的位置，然后通过简单的字符串操作来获取每个部分。

with myTable as (
    select * from (values('abc@abc.com - Invoice - A12222203'),('FGH@fgh.com - Credit - C12222333'))v(Description)
)
select
    Trim(Left(description,h1-1)) Email, 
    Trim(Substring(description,h1+1,Len(description)-h2-h1-1)) DocType,
    Trim(Right(description,h2-1)) Ref
from mytable
cross apply(values(CharIndex('-',description)))v1(h1)
cross apply(values(CharIndex('-',Reverse(description))))v2(h2)

这会将字符串拆分为行，然后有条件地聚合回一行。

with myTable as (
    select * from (values('abc@abc.com - Invoice - A12222203'),('FGH@fgh.com - Credit -  C12222333'))v(Description)
)
select 
    max(Iif(rn=1,v,null)) Email,
    max(Iif(rn=2,v,null)) Doctype,
    max(Iif(rn=3,v,null)) Ref
from mytable
cross apply (
    select Trim(value)v,row_number() over(order by (select null)) rn
    from String_Split(Description,'-') 
)s
group by Description