拆分和匹配行

Question

我使用SQL Server 2008

我的表看起來像這樣：

ID  Column 
---------------------------------- 
1  This is a Sample Text 
2  Sample Text is typed here 
3  Here the sample text is 
4  Typing a sample 

，我需要的輸出是這樣的：

ID Column      MostCommon Common1 Common2 NonCommon 
--------------------------------------------------------------------------------------- 
1 This is a Sample Text  Sample  Text  is  This a 
2 Sample Text is typed here Sample  Text  is  typed here 
3 Here the sample text is Sample  Text  is  Here the 
4 Typing a sample   sample  NULL  NULL  Typing A 

任何人都可以幫助我在SQL Server 2008中編寫sp /函數/查詢

'sample'出現在所有行中。所以我可以保留它作爲最常見的單詞，'文本'是'，下一個最常見的，可以在＃1,2,3行中找到。所有其他字與其他行不匹配，並將移動到非常見類別

Answer 1

以下是您可以如何操作的方法，首先必須創建一個函數，用於分割字符串，然後計算出現次數並按照需要進行顯示。這使得它更復雜，因爲你想顯示可變數量的列：

/* 
CREATE FUNCTION dbo.SplitStrings_XML 
(
    @List  NVARCHAR(MAX), 
    @Delimiter NVARCHAR(255) 
) 
RETURNS TABLE 
WITH SCHEMABINDING 
AS 
    RETURN 
    ( 
     SELECT Item = y.i.value('(./text())[1]', 'nvarchar(4000)') 
     FROM 
     ( 
     SELECT x = CONVERT(XML, '<i>' + 
        REPLACE(@List, @Delimiter, '</i><i>') + '</i>').query('.') 
    ) AS a CROSS APPLY x.nodes('i') AS y(i) 
    ); 
GO 
*/ 

CREATE TABLE #t(ID INT, Col VARCHAR(1000)) 
INSERT #t 
VALUES 
(1,  'This is a Sample Text'), 
(2,  'Sample Text is typed here'), 
(3,  'Here the sample text is'), 
(4,  'Typing a sample') 

DECLARE @MinimumNumberOfOccurances INT = 2 

SELECT a.ID, 
     a.Col, 
     b.Item 
INTO #SplitedStrings 
FROM #t a 
CROSS APPLY dbo.SplitStrings_XML(a.Col, N' ') b 


SELECT b.Item, 
     COUNT(*) cnt 
INTO #SplitedStringsGrouped 
FROM #t a 
CROSS APPLY dbo.SplitStrings_XML(a.Col, N' ') b 
GROUP BY b.Item 

SELECT  b.*, 
      a.cnt 
INTO  #ResultTable 
FROM  #SplitedStringsGrouped a 
RIGHT JOIN #SplitedStrings b ON 
      b.Item = a.Item 
      AND a.cnt > @MinimumNumberOfOccurances 
ORDER BY b.ID, a.cnt DESC, LEN(a.Item) DESC 

DECLARE @ColumnNames VARCHAR(1000) = STUFF(
(
    SELECT ',[' + Item + ']' 
    FROM #SplitedStringsGrouped 
    WHERE cnt > @MinimumNumberOfOccurances 
    FOR  XML PATH('') 
) 
, 1, 1, '') 

DECLARE @TableHeader VARCHAR(1000) = STUFF(
(
    SELECT ',MAX([' + Item + ']) AS [Common' + 
      CAST((ROW_NUMBER() OVER 
       (ORDER BY cnt DESC, LEN(Item) DESC) - 1) 
        AS VARCHAR(5)) 
      + ']' 
    FROM #SplitedStringsGrouped 
    WHERE cnt > @MinimumNumberOfOccurances 
    FOR  XML PATH('') 
) 
, 1, 1, '') 

SELECT ID, 
     Item, 
     ROW_NUMBER() OVER 
      (PARTITION BY ID ORDER BY ID) Num 
INTO #NonCommon 
FROM #ResultTable 
WHERE cnt IS NULL 

DECLARE @sql VARCHAR(1000) = 
' 
SELECT MAX(pvt.ID) ID, MAX(pvt.Col) [Column], 
     '+@TableHeader+', 
     RTRIM((
      SELECT a.Item + '' '' 
      FROM #NonCommon a 
      WHERE a.ID = pvt.ID 
      FOR  XML PATH('''') 
     )) NonCommon   
FROM #ResultTable a 
PIVOT (
    MAX(Item) FOR Item IN ('+@ColumnNames+') 
) pvt 
GROUP BY pvt.ID 
' 

EXEC(@sql) 

DROP TABLE #t 
DROP TABLE #SplitedStringsGrouped 
DROP TABLE #SplitedStrings 
DROP TABLE #ResultTable 
DROP TABLE #NonCommon