根据字符串拆分由多列分隔的行中的单列 - SQL Server

Question

我有 2 个需要字符串拆分的表。结果表将拆分字符串以创建 new year、make、model 列。这是示例 2 表。

CREATE TABLE Table1
(
    sku varchar(50), 
    Conca varchar(max)
);

INSERT INTO Table1 (sku, Conca)
VALUES ('AVS1234','Game;Consoles/Year;2001/ Year;2002/ Year;2006/ Year;2007/ Year;2009/ Year;2011/Make;XBox/Model;100/'),
       ('AVS1234','Game;Consoles/Year;2016/Make;Nintendo/Model;DX/'),
       ('AVS1234','Game;Consoles/Year;2001/ Year;2002/ Year;2006/ Year;2007/ Year;2009/ Year;2011/Make;PS/Model;300/Model;500/');

CREATE TABLE Table2
(
    sku2 varchar(50), 
    Conca2 varchar(max)
);

INSERT INTO Table2 (sku2, Conca2)
VALUES ('AVS1234','<tr><td>2011 </td><td>Xbox </td><td>100; Notes : 2 pc.; Stainless</td></tr>'), 
       ('AVS1234','<tr><td>2005 </td><td>Xbox </td><td>100; Notes : 2 pc.; Stainless</td></tr>'), 
       ('AVS1234','<tr><td>2016 </td><td>Xbox </td><td>300</td></tr>'); 

预期结果表1

        Sku     year    make   model
        --------------------------
        avs1234 2001    XBox    100
        avs1234 2002    XBox    100
        avs1234 2006    XBox    100
        avs1234 2007    XBox    100
        avs1234 2009    XBox    100
        avs1234 2011    XBox    100
        AVS1234 2016    Nintendo    DX
        AVS1234 2001    PS      300
        AVS1234 2002    PS      300
        AVS1234 2006    PS      300
        AVS1234 2007    PS      300
        AVS1234 2009    PS      300
        AVS1234 2011    PS      300
        AVS1234 2001    PS      500
        AVS1234 2002    PS      500
        AVS1234 2006    PS      500
        AVS1234 2007    PS      500
        AVS1234 2009    PS      500
        AVS1234 2011    PS      500

预期结果表2

        Sku       year  make   model  Notes
        ------------------------------------
        avs1234   2001  XBox    100   2 pc.; Stainless
        avs1234   2002  XBox    100   2 pc.; Stainless
        AVS1234   2005  Xbox    100   2 pc.; Stainless
        AVS1234   2016  Xbox    300

首先我想尝试以下位置/修剪方法：

Cross Apply 
    (Select Pos1 = ltrim(rtrim(xDim.value('/x[1]','varchar(max)'))) 

但 # 年的范围可能从 1950 年代到 2018 年及以后。所以这会产生太多的“pos”。

然后我尽我所能：

with firstpass as
(
    SELECT A.sku,  
           Split.a.value('.', 'VARCHAR(100)') AS Data  
    FROM  
        (SELECT sku,  
                Cast('<M>' + replace((Select replace(Conca,'year','§§Split§§') as [*] For XML Path('')),'§§Split§§','</M><M>')+'</M>' as xml) as Data
         FROM  table1) AS A 
    CROSS APPLY 
        Data.nodes ('/M') AS Split(a)
)
SELECT
    *, Year = SUBSTRING([data], 1 ,
            case when  CHARINDEX('/', [data] ) = 0 then LEN([data]) 
            else CHARINDEX('/', [data]) -1 end)
                   from firstpass

我不知道下一步该去哪里……请帮忙。

Answer 1

如果打开表值辅助函数。我修改了一个拆分/解析函数来接受两个不同的分隔符，因为我厌倦了提取字符串（左、右、charindex、patindex 等）。

示例或dbFiddle

Select SKU
      ,Year  = B.RetVal
      ,Make  = C.RetVal
      ,Model = D.RetVal
 From  Table1 A
 Cross Apply  [dbo].[tvf-Str-Extract]('/'+A.Conca+'/','Year;','/')  B
 Cross Apply  [dbo].[tvf-Str-Extract]('/'+A.Conca+'/','Make;' ,'/') C
 Cross Apply  [dbo].[tvf-Str-Extract]('/'+A.Conca+'/','Model;','/') D


Select SKU2
      ,Year  = XMLData.value('tr[1]/td[1]','varchar(max)')
      ,Make  = XMLData.value('tr[1]/td[2]','varchar(max)')
      ,Model = ltrim(rtrim(left(XMLData.value('tr[1]/td[3]','varchar(max)'),charindex(';',XMLData.value('tr[1]/td[3]','varchar(max)')+';')-1)))
      ,Notes = ltrim(rtrim(substring(XMLData.value('tr[1]/td[3]','varchar(max)'),charindex(';',XMLData.value('tr[1]/td[3]','varchar(max)')+';')+1,100)))
 From  Table2
 Cross Apply (values (cast(Conca2 as xml)) )B(XMLdata)

退货

有兴趣的 TVF

CREATE FUNCTION [dbo].[tvf-Str-Extract] (@String varchar(max),@Delimiter1 varchar(100),@Delimiter2 varchar(100))
Returns Table 
As
Return (  

with   cte1(N)   As (Select 1 From (Values(1),(1),(1),(1),(1),(1),(1),(1),(1),(1)) N(N)),
       cte2(N)   As (Select Top (IsNull(DataLength(@String),0)) Row_Number() over (Order By (Select NULL)) From (Select N=1 From cte1 N1,cte1 N2,cte1 N3,cte1 N4,cte1 N5,cte1 N6) A ),
       cte3(N)   As (Select 1 Union All Select t.N+DataLength(@Delimiter1) From cte2 t Where Substring(@String,t.N,DataLength(@Delimiter1)) = @Delimiter1),
       cte4(N,L) As (Select S.N,IsNull(NullIf(CharIndex(@Delimiter1,@String,s.N),0)-S.N,8000) From cte3 S)

Select RetSeq = Row_Number() over (Order By N)
      ,RetPos = N
      ,RetVal = left(RetVal,charindex(@Delimiter2,RetVal)-1) 
 From  (
        Select *,RetVal = Substring(@String, N, L) 
         From  cte4
       ) A
 Where charindex(@Delimiter2,RetVal)>1

)
/*
Max Length of String 1MM characters

Declare @String varchar(max) = 'Dear [[FirstName]] [[LastName]], ...'
Select * From [dbo].[tvf-Str-Extract] (@String,'[[',']]')
*/

编辑 - 您也可以将 TVF 用于 table2

Select SKU2
      ,B.Year
      ,B.Make
      ,Model = ltrim(rtrim(left(B.Model,charindex(';',B.Model+';')-1)))
      ,Notes = ltrim(rtrim(substring(B.Model,charindex(';',B.Model+';')+1,100)))
 From  Table2 A
 Cross Apply (
                Select Year  = max(case when RetSeq=1 then RetVal end)
                      ,Make  = max(case when RetSeq=2 then RetVal end)
                      ,Model = max(case when RetSeq=3 then RetVal end)
                 From [dbo].[tvf-Str-Extract](A.Conca2,'<td>','</td>')  
             )B