在 SQL 中拆分字符串文本答案

【问题标题】：Split string text in TSQL在 SQL 中拆分字符串文本
【发布时间】：2021-08-28 02:18:39
【问题描述】：

我有下面的字符串需要拆分以允许插入到表中。

数据来自：

http://smartgriddashboard.eirgrid.com/DashboardService.svc/csv?area=co2Intensity&region=NI&datefrom=30-Aug-2021%2000:00&dateto=30-Aug-2021%2023:59

我正在开发一个在代理作业上运行的函数以读取数据，然后将其解析到 MSSQL DB 中。例如

DECLARE @fromTime NVARCHAR(MAX)
DECLARE @toTime NVARCHAR(MAX)
DECLARE @URL2 NVARCHAR(MAX)

Set @fromTime = REPLACE(FORMAT(DATEADD(DAY, -1, GETDATE()), 'dd-MMM-yyyy 00:00' ), ' ','%20')
Set @toTime = REPLACE(FORMAT(DATEADD(DAY, -1,GETDATE()), 'dd-MMM-yyyy 23:59' ), ' ','%20')

select @URL2 = 
concat('http://smartgriddashboard.eirgrid.com/DashboardService.svc/csv?area=co2Intensity&region=ALL&datefrom=',@fromTime,'&dateto=',@toTime)

DECLARE @URL NVARCHAR(MAX) = @URL2
Declare @Object as Int;
Declare @ResponseText as Varchar(8000);
Declare @currenttime as datetime;
set @currenttime = GETDATE()

Exec sp_OACreate 'MSXML2.XMLHTTP', @Object OUT;
Exec sp_OAMethod @Object, 'open', NULL, 'get',
   @URL,
   'False'
Exec sp_OAMethod @Object, 'send'
Exec sp_OAMethod @Object, 'responseText', @ResponseText OUTPUT

IF((Select @ResponseText) <> '')
BEGIN
   DECLARE @json NVARCHAR(MAX) = (Select @ResponseText)
  
  --do something here to parce the return data
 ;with CTE(raw) as 
 (

 select replace(replace(replace(replace(@json,' ','-'),' ','-'),' ','-'),' ','-')
 )
 select * from CTE
     
 select @json

--write to SQL table.

END
ELSE
BEGIN
 DECLARE @ErroMsg NVARCHAR(30) = 'No data found.';
 Print @ErroMsg;
END
Exec sp_OADestroy @Object

表格中每一行的分隔符是一个双空格。每列的分隔符是一个','。

第一个双空格分隔符之前的第一组数据是列/表标题并且是固定的。例如

DATE & TIME  |  CO2 INTENSITY (gCO2/kWh)  |  REGION

来自上述 URL 的示例输出

DATE & TIME, CO2 INTENSITY (gCO2/kWh), REGION  28 August 2021 00:00,470,All Island  28 August 2021 00:15,471,All Island  28 August 2021 00:30,471,All Island  28 August 2021 00:45,472,All Island  28 August 2021 01:00,470,All Island  28 August 2021 01:15,471,All Island  28 August 2021 01:30,468,All Island  28 August 2021 01:45,471,All Island  28 August 2021 02:00,468,All Island  28 August 2021 02:15,460,All Island  28 August 2021 02:30,452,All Island  28 August 2021 02:45,450,All Island  28 August 2021 03:00,464,All Island  28 August 2021 03:15,469,All Island  28 August 2021 03:30,472,All Island  28 August 2021 03:45,476,All Island  28 August 2021 04:00,472,All Island  28 August 2021 04:15,477,All Island  28 August 2021 04:30,488,All Island  28 August 2021 04:45,485,All Island  28 August 2021 05:00,488,All Island  28 August 2021 05:15,489,All Island  28 August 2021 05:30,487,All Island  28 August 2021 05:45,493,All Island  28 August 2021 06:00,492,All Island  28 August 2021 06:15,472,All Island  28 August 2021 06:30,469,All Island  28 August 2021 06:45,468,All Island  28 August 2021 07:00,472,All Island  28 August 2021 07:15,476,All Island  28 August 2021 07:30,482,All Island  28 August 2021 07:45,473,All Island  28 August 2021 08:00,465,All Island  28 August 2021 08:15,463,All Island  28 August 2021 08:30,467,All Island  28 August 2021 08:45,470,All Island  28 August 2021 09:00,460,All Island  28 August 2021 09:15,461,All Island  28 August 2021 09:30,463,All Island  28 August 2021 09:45,472,All Island  28 August 2021 10:00,478,All Island  28 August 2021 10:15,475,All Island  28 August 2021 10:30,475,All Island  28 August 2021 10:45,473,All Island  28 August 2021 11:00,459,All Island  28 August 2021 11:15,462,All Island  28 August 2021 11:30,445,All Island  28 August 2021 11:45,438,All Island  28 August 2021 12:00,433,All Island  28 August 2021 12:15,435,All Island  28 August 2021 12:30,429,All Island  28 August 2021 12:45,424,All Island  28 August 2021 13:00,430,All Island  28 August 2021 13:15,430,All Island  28 August 2021 13:30,431,All Island  28 August 2021 13:45,429,All Island  28 August 2021 14:00,419,All Island  28 August 2021 14:15,418,All Island  28 August 2021 14:30,416,All Island  28 August 2021 14:45,421,All Island  28 August 2021 15:00,411,All Island  28 August 2021 15:15,424,All Island  28 August 2021 15:30,430,All Island  28 August 2021 15:45,429,All Island  28 August 2021 16:00,424,All Island  28 August 2021 16:15,423,All Island  28 August 2021 16:30,416,All Island  28 August 2021 16:45,422,All Island  28 August 2021 17:00,425,All Island  28 August 2021 17:15,439,All Island  28 August 2021 17:30,439,All Island  28 August 2021 17:45,440,All Island  28 August 2021 18:00,440,All Island  28 August 2021 18:15,426,All Island  28 August 2021 18:30,439,All Island  28 August 2021 18:45,429,All Island  28 August 2021 19:00,437,All Island  28 August 2021 19:15,440,All Island  28 August 2021 19:30,436,All Island  28 August 2021 19:45,441,All Island  28 August 2021 20:00,434,All Island  28 August 2021 20:15,446,All Island  28 August 2021 20:30,434,All Island  28 August 2021 20:45,447,All Island  28 August 2021 21:00,443,All Island  28 August 2021 21:15,456,All Island  28 August 2021 21:30,447,All Island  28 August 2021 21:45,448,All Island  28 August 2021 22:00,447,All Island  28 August 2021 22:15,426,All Island  28 August 2021 22:30,418,All Island  28 August 2021 22:45,433,All Island  28 August 2021 23:00,438,All Island  28 August 2021 23:15,440,All Island  28 August 2021 23:30,434,All Island  28 August 2021 23:45,434,All Island

前 3 行和前 3 列的示例是：

28 August 2021 00:00  |  470  |  All Island
28 August 2021 00:15  |  471  |  All Island
28 August 2021 00:30  |  471  |  All Island

所以我想要实现的输出是这样的表格

      DATE & TIME     |  CO2 INTENSITY (gCO2/kWh)  |  REGION
-----------------------------------------------------------------
28 August 2021 00:00  |          470               |  All Island
28 August 2021 00:15  |          471               |  All Island
28 August 2021 00:30  |          471               |  All Island

我已经尝试了许多拆分字符串和批量导入，但没有成功。例如

 --;with CTE(raw) as 
 --(
 ----select @Data as raw
 --select replace(replace(replace(replace(@Data ,' ','-'),' ','-'),' 
 ','- 
 '),' ','-')
 --   )
 --select * from CTE

-- write to file first and then import e.g.
--BULK INSERT [AutoData].[dbo].[Eirgrid_Co2Data]
--FROM 'C:\TempData\FileName.csv'
--WITH ( FORMAT='CSV');

 -- Tried to read it as JSON and XML!
 --select * from OPENJSON(@Data)
 select @Data

 -- splitting!
 --  SELECT value  
 --FROM STRING_SPLIT(@Data, ',')
 ----FROM STRING_SPLIT(@Data, '  ')  'split of double space!
 --WHERE RTRIM(value) <> '';

【问题讨论】：

您的“数据”在我看来很混乱，这令人困惑。如果你解开它，那将鼓励其他人提供帮助。引用字符串然后引用“行”同样令人困惑。上下文会有所帮助。包含 DDL、示例数据和基于该示例数据的所需输出的脚本将大有帮助。
根据问题指南，请展示您的尝试并告诉我们您发现了什么（在本网站或其他地方）以及为什么它不能满足您的需求。
这确实需要更多细节。您是否提前知道列，即可以从数据中删除它们？还是需要根据数据动态创建表？如果是这样，确定列结束和数据开始的标准是什么？第一个“数据”行之前没有换行符，也没有换行符来指示新行。任何人都应该如何自信地处理这个问题？
提问时，您需要提供minimal reproducible example： (1) DDL 和样本数据填充，即 CREATE 表和 INSERT T-SQL 语句。 (2) 你需要做什么，即逻辑和你的代码尝试在 T-SQL 中实现它。 (3) 期望的输出，基于上面#1 中的样本数据。 (4) 您的 SQL Server 版本 (SELECT @@version;)。
感谢您的反馈我已经更新了问题，还有什么请告诉我。

标签： sql sql-server tsql

【解决方案1】：

看来行分隔符是 2 个空格。不幸的是，SQL Server 不能轻松地拆分为 2 个空格。似乎还有一些异常数据，即 ItemNumber=7 ，其中有一个实际的回车符。为了蛮力完成，这会将 2 个空格（以及 1 个带 1 个回车的空格）替换为波浪号 '~'，然后使用序号拆分器在波浪号上拆分。

序数分离器described here

CREATE FUNCTION [dbo].[DelimitedSplit8K_LEAD]
--===== Define I/O parameters
        (@pString VARCHAR(8000), @pDelimiter CHAR(1))
RETURNS TABLE WITH SCHEMABINDING AS
 RETURN
  WITH E1(N) AS (
                 SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL 
                 SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL 
                 SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1
                ),                          --10E+1 or 10 rows
       E2(N) AS (SELECT 1 FROM E1 a, E1 b), --10E+2 or 100 rows
       E4(N) AS (SELECT 1 FROM E2 a, E2 b), --10E+4 or 10,000 rows max
 cteTally(N) AS (--==== This provides the "zero base" and limits the number of rows right up front
                     -- for both a performance gain and prevention of accidental "overruns"
                 SELECT 0 UNION ALL
                 SELECT TOP (DATALENGTH(ISNULL(@pString,1))) ROW_NUMBER() OVER (ORDER BY (SELECT NULL)) FROM E4
                ),
cteStart(N1) AS (--==== This returns N+1 (starting position of each "element" just once for each delimiter)
                 SELECT t.N+1
                   FROM cteTally t
                  WHERE (SUBSTRING(@pString,t.N,1) = @pDelimiter OR t.N = 0) 
                )
--===== Do the actual split. The ISNULL/NULLIF combo handles the length for the final element when no delimiter is found.
 SELECT ItemNumber = ROW_NUMBER() OVER(ORDER BY s.N1),
        Item = SUBSTRING(@pString,s.N1,ISNULL(NULLIF((LEAD(s.N1,1,1) OVER (ORDER BY s.N1) - 1),0)-s.N1,8000))
   FROM cteStart s
;

查询

declare @string         nvarchar(max)=N'DATE & TIME, CO2 INTENSITY (gCO2/kWh), REGION  27 August 2021 00:00,465,All Island  27 August 2021 00:15,459,All Island  27 August
2021 00:30,463,All Island  27 August 2021 00:45,464,All Island  27
August 2021 01:00,461,All Island  27 August 2021 01:15,465,All Island 
27 August 2021 01:30,466,All Island  27 August 2021 01:45,470,All
Island  27 August 2021 02:00,466,All Island';


select *
from dbo.DelimitedSPlit8K_LEAD(replace(replace(@string, ' 
','~'), '  ','~'), '~') dl;

ItemNumber  Item
1           DATE & TIME, CO2 INTENSITY (gCO2/kWh), REGION
2           27 August 2021 00:00,465,All Island
3           27 August 2021 00:15,459,All Island
4           27 August  2021 00:30,463,All Island
5           27 August 2021 00:45,464,All Island
6           27  August 2021 01:00,461,All Island
7           27 August 2021 01:15,465,All Island
8           27 August 2021 01:30,466,All Island
9           27 August 2021 01:45,470,All  Island
10          27 August 2021 02:00,466,All Island

【讨论】：

感谢@SteveC 非常感谢，我已经更新了我的问题，以显示有关我正在寻找的内容的更多详细信息。

【解决方案2】：

我多年来使用的方法在这里找到tally table solution。

最后是 CTE varchar 版本的函数。如果需要，上面的链接也有一个 nvarchar 版本。

请注意，您的数据似乎没有一致的分隔符，因此您可能需要多次使用以下内容。

例如，使用上面的第一行数据：

DECLARE @Rowdata Varchar(100) = '27 August 2021 02:25,465,All Island' ;

SELECT * FROM [dbo].[DelimitedSplit8K](@Rowdata,',')

产生以下结果集：

ItemNumber Item

1 2021 年 8 月 27 日 02:25

2 465

3 全岛

    CREATE FUNCTION [dbo].[DelimitedSplit8K]
--===== Define I/O parameters
        (@pString VARCHAR(8000), @pDelimiter CHAR(1))
--WARNING!!! DO NOT USE MAX DATA-TYPES HERE!  IT WILL KILL PERFORMANCE!
RETURNS TABLE WITH SCHEMABINDING AS
 RETURN
--===== "Inline" CTE Driven "Tally Table" produces values from 1 up to 10,000...
     -- enough to cover VARCHAR(8000)
  WITH E1(N) AS (
                 SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL
                 SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL
                 SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1
                ),                          --10E+1 or 10 rows
       E2(N) AS (SELECT 1 FROM E1 a, E1 b), --10E+2 or 100 rows
       E4(N) AS (SELECT 1 FROM E2 a, E2 b), --10E+4 or 10,000 rows max
 cteTally(N) AS (--==== This provides the "base" CTE and limits the number of rows right up front
                     -- for both a performance gain and prevention of accidental "overruns"
                 SELECT TOP (ISNULL(DATALENGTH(@pString),0)) ROW_NUMBER() OVER (ORDER BY (SELECT NULL)) FROM E4
                ),
cteStart(N1) AS (--==== This returns N+1 (starting position of each "element" just once for each delimiter)
                 SELECT 1 UNION ALL
                 SELECT t.N+1 FROM cteTally t WHERE SUBSTRING(@pString,t.N,1) = @pDelimiter
                ),
cteLen(N1,L1) AS(--==== Return start and length (for use in substring)
                 SELECT s.N1,
                        ISNULL(NULLIF(CHARINDEX(@pDelimiter,@pString,s.N1),0)-s.N1,8000)
                   FROM cteStart s
                )
--===== Do the actual split. The ISNULL/NULLIF combo handles the length for the final element when no delimiter is found.
 SELECT ItemNumber = ROW_NUMBER() OVER(ORDER BY l.N1),
        Item       = SUBSTRING(@pString, l.N1, l.L1)
   FROM cteLen l
;

【讨论】：

答案需要包含问题的完整答案。不是指向外部资源的链接。