简体   繁体   English

SQL Server 2008中的拆分功能

[英]Split function in SQL Server 2008

I have Table1 with columns like this: 我有Table1的列:

+--+------+
|ID|Name  |
+--+------+
|1 |MSSQL |
+--+------+
|2 |MySQl |
+--+------+
|3 |Oracle|
+--+------+

In Table2 , I have a column like 在表Table2 ,我有一个列

+------------+
|Databasename|
+------------+
|1,3         |
+------------+
|2           |
+------------+
|1,2         |
+------------+

My output should be: 我的输出应该是:

+------------+
|Databasename|
+------------+
|MSSQL,Oracle|
+------------+
|MySQL       |
+------------+
|MSSQL,MYSQL |
+------------+

How do I get this, I need query for this.. 我怎么得到这个,我需要查询这个..

You are asking for a split function but you do not have to split your values to get the result you want. 您要求分割功能,但您不必分割您的值以获得所需的结果。

This query builds the comma separated names list in a correlated subquery using the for xml trick to concatenate values. 此查询使用for xml技巧在相关子查询中构建逗号分隔名称列表以连接值。 It uses like to figure out what values to use from Table1 for each row in Table2 . 它使用like弄清楚什么值从使用Table1中的每一行Table2

select (
       select ', '+T1.Name
       from Table1 as T1
       where ','+T2.Databasename+',' like '%,'+cast(T1.ID as varchar(10))+',%'
       for xml path(''), type
       ).value('substring(text()[1], 3)', 'varchar(max)') as Databasenames
from Table2 as T2

SQL Fiddle SQL小提琴

First, your best solution is to not store data in a comma-separated list in your database. 首先,您最好的解决方案是不将数据存储在数据库中以逗号分隔的列表中。 You should consider fixing the table structure. 您应该考虑修复表结构。

If you cannot alter the table structure, then you will need to split the data in the list to rows to assign the correct name. 如果无法更改表结构,则需要将列表中的数据拆分为行以指定正确的名称。 Once the data is split then you can concatenate the data back into the list. 分割数据后,您可以将数据连接回列表。

There are many different split function that you can find online but here is a version that I typically use: 您可以在线找到许多不同的split功能,但这是我通常使用的版本:

CREATE FUNCTION [dbo].[Split](@String varchar(MAX), @Delimiter char(1))       
returns @temptable TABLE (items varchar(MAX))       
as       
begin      
    declare @idx int       
    declare @slice varchar(8000)       

    select @idx = 1       
        if len(@String)<1 or @String is null  return       

    while @idx!= 0       
    begin       
        set @idx = charindex(@Delimiter,@String)       
        if @idx!=0       
            set @slice = left(@String,@idx - 1)       
        else       
            set @slice = @String       

        if(len(@slice)>0)  
            insert into @temptable(Items) values(@slice)       

        set @String = right(@String,len(@String) - @idx)       
        if len(@String) = 0 break       
    end   
return 
end;

To get your result, I would start by applying the split function and a row_number() since I do not see a unique key associated with each row. 为了得到你的结果,我将首先应用split函数和row_number()因为我没有看到与每一行相关联的唯一键。 If you have a unique key on each row then you will not need the row_number() : 如果每行都有唯一键,则不需要row_number()

;with cte as
(
  select rn, name, id
  from
  (
    select row_number() over(order by (select 1)) rn,
      databasename
    from table2
  ) t2
  cross apply dbo.split(t2.databasename, ',') i
  inner join table1 t1
    on i.items = t1.id
) 
select *
from cte

This query breaks your comma-separated list into the following: 此查询将以逗号分隔的列表分为以下内容:

| RN |   NAME | ID |
--------------------
|  1 |  MSSQL |  1 |
|  1 | Oracle |  3 |
|  2 |  MySQl |  2 |
|  3 |  MSSQL |  1 |
|  3 |  MySQl |  2 |

Once you have the data in multiple rows with the correct name , then you can use STUFF() and FOR XML PATH to concatenate it into the list. 一旦使用正确的name将数据存储在多个行中,就可以使用STUFF()FOR XML PATH将其连接到列表中。 You full query would be similar to this: 您完全查询将类似于:

;with cte as
(
  select rn, name, id
  from
  (
    select row_number() over(order by (select 1)) rn,
      databasename
    from table2
  ) t2
  cross apply dbo.split(t2.databasename, ',') i
  inner join table1 t1
    on i.items = t1.id
) 
select  
  STUFF(
         (SELECT ', ' + c2.name
          FROM cte c2
          where c1.rn = c2.rn
          order by c2.id
          FOR XML PATH (''))
          , 1, 1, '') Databasename
from cte c1
group by c1.rn
order by c1.rn;

See SQL Fiddle with Demo . 请参阅SQL Fiddle with Demo

The result of the full query is: 完整查询的结果是:

|   DATABASENAME |
------------------
|  MSSQL, Oracle |
|          MySQl |
|   MSSQL, MySQl |

No splitting, also no XML Path, but achieves the right result. 没有拆分,也没有XML路径,但实现了正确的结果。

;with cte as (
    select *, cast(null as varchar(1024)) as str, cast(0 as int) as ID
    from Table2

    union all

    select DatabaseName, (case when DatabaseName like cast(t.ID as varchar(32)) + ',%' 
                                    or DatabaseName like '%,' + cast(t.ID as varchar(32)) + ',%'
                                    or DatabaseName like '%,' + cast(t.ID as varchar(32)) 
                                    or DatabaseName = cast(t.ID as varchar(32)) then cast(isnull(str, '') + ',' + t.Name as varchar(1024)) else str end), cte.ID + 1 as ID
    from cte
    inner join Table1 t on cte.ID + 1 = t.ID
)
select DatabaseName, (case when str like ',%' then substring(str, 2, len(str)) else null end) as str
from cte c
where ID = (select max(ID) from cte where DatabaseName = c.DatabaseName)
--Here it goes:

----------------
-- FieldCount --
----------------
CREATE FUNCTION [dbo].[FieldCount](@S VARCHAR(8000), @Separator VARCHAR(10))
  RETURNS INT
AS

BEGIN

  /*
  @Author: Leonardo Augusto Rezende Santos
  @Contact: http://www.linkedin.com/pub/leonardo-santos/0/2b1/890
  */

  DECLARE @Ptr INT, @p INT, @LenS INT, @LenSep INT, @Result INT

  IF @Separator = ' ' 
    BEGIN
      SET @S = REPLACE(@S, ' ', '|-|')
      SET @Separator = '|-|'
    END

  WHILE CHARINDEX(@Separator + @Separator, @S) > 0
    SET @S = Replace(@S, @Separator + @Separator, @Separator + '_-_' + @Separator)
  IF @S <> ''
    SET @Result = 1
  ELSE
    BEGIN
      SET @Result = 0
      RETURN(@Result)
    END
  SET @Ptr = 0
  SET @LenS = LEN(@S)
  SET @LenSep = LEN(@Separator)
  SET @p = CHARINDEX(@Separator, @S)
  WHILE @p > 0
    BEGIN
      SET @Result = @Result + 1
      SET @Ptr = @Ptr + @p + @LenSep
      SET @p = CHARINDEX(@Separator, SUBSTRING(@S, @Ptr, @LenS - @Ptr + 1))
    END

  RETURN(@Result)

END

--------------
-- GetField --
--------------
CREATE FUNCTION [dbo].[GetField](@S VARCHAR(8000), @Separator VARCHAR(10), @Field INT)
  RETURNS VARCHAR(8000)
AS

BEGIN

  /*
  @Author: Leonardo Augusto Rezende Santos
  @Contact: http://www.linkedin.com/pub/leonardo-santos/0/2b1/890
  */

  DECLARE @Ptr INT, @p INT, @LenS INT, @LenSep INT, @Fld INT, @Result VARCHAR(8000)

  IF @Separator = ' ' 
    BEGIN
      SET @S = REPLACE(@S, ' ', '|-|')
      SET @Separator = '|-|'
    END

  IF @Field > dbo.FieldCount(@S, @Separator)
    BEGIN
      SET @Result = ''
      RETURN(@Result)
    END
  SET @Fld = 1
  SET @Ptr = 1
  SET @LenS = LEN(@S)
  SET @LenSep = LEN(@Separator)
  SET @p = CHARINDEX(@Separator, @S)
  WHILE (@p > 0) and (@Fld < @Field)
    BEGIN
      SET @Fld = @Fld + 1
      SET @Ptr = @Ptr + @p + @LenSep - 1
      SET @p = CHARINDEX(@Separator, SUBSTRING(@S, @Ptr, @LenS - @Ptr + 1))
    END
  IF (@p = 0) and (@Fld = @Field)
    SET @p = @LenS - @Ptr + 2
  SET @Result = SUBSTRING(@S, @Ptr, @p - 1)

  RETURN(@Result)

END

/* USAGE*/

select dbo.FieldCount('A1 A2 A3 A4 A5', ' ')

--It will return 5

select dbo.GetField('A1 A2 A3 A4 A5', ' ', 3)

--It will return 'A3'

select dbo.GetField('A1/A2/A3/A4/A5', '/', 3)

--It will return 'A3'

--Hope it works for you.

--Leonardo Augusto

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM