转置和识别带有标志的单元格

Question

我有一个包含三列的 table。值列始终为 NULL。我知道如何将行转置为列，但我还想用 'X'.

标识单元格

Table

EditorTypeSymbolicName ClaimFieldSymbolicName Value
01                     AA                     NULL
01                     BB                     NULL
01                     CC                     NULL
02                     BB                     NULL
02                     CC                     NULL
03                     AA                     NULL

这是我的代码

DECLARE @cols NVARCHAR(MAX), @sql NVARCHAR(MAX)

SET @cols = STUFF((SELECT DISTINCT ',' + QUOTENAME(ClaimFieldSymbolicName)
            FROM #TempEditElements
            ORDER BY 1
            FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)'),1,1,'')

SET @sql = 'SELECT EditorTypeSymbolicName, ' + @cols + '
              FROM
            (
              select EditorTypeSymbolicName, ClaimFieldSymbolicName, Value 
              from #TempEditElements
            ) s
            PIVOT
            (
              MAX(Value) FOR ClaimFieldSymbolicName IN (' + @cols + ')
            ) p
            order by EditorTypeSymbolicName'

EXECUTE(@sql)

我的代码转置了它，但我不知道如何用 'X' 标记它们。我希望输出如下：

EditorTypeSymbolicName  AA   BB   CC
01                      X    X    X
02                      NULL X    X
03                      X    NULL NULL

当前输出如下所示

EditorTypeSymbolicName  AA   BB   CC
01                      NULL NULL NULL
02                      NULL NULL NULL
03                      NULL NULL NULL

Answer 1

静态解决方案

select      *

from        (select 'X' as x,EditorTypeSymbolicName,ClaimFieldSymbolicName from t) t 
                pivot (max(x) for ClaimFieldSymbolicName in ([AA],[BB],[CC]))   t

order by    EditorTypeSymbolicName
;

动态解决方案

declare @columns    nvarchar(max)   = ''
       ,@stmt       nvarchar(max)
;

----------

select      @columns += case @columns when '' then '' else ',' end + QUOTENAME (ClaimFieldSymbolicName,'[')

from        t

group by    ClaimFieldSymbolicName

order by    ClaimFieldSymbolicName
;

----------

set @stmt =

    '
    select      *

    from        (select ''X'' as x,EditorTypeSymbolicName,ClaimFieldSymbolicName from t) t 
                    pivot (max(x) for ClaimFieldSymbolicName in ('+ @columns + '))   t

    order by    EditorTypeSymbolicName
    '
;

----------

exec (@stmt)

半静态解决方案

select      (select top 0 EditorTypeSymbolicName from t) as EditorTypeSymbolicName,t.*

from        (select distinct ClaimFieldSymbolicName,dense_rank () over (order by ClaimFieldSymbolicName) as dr from t) t 
                pivot (max(ClaimFieldSymbolicName) for dr in ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10])) t


union all

select      *

from        (select 'X' as x,EditorTypeSymbolicName,dense_rank () over (order by ClaimFieldSymbolicName) as dr from t) t 
                pivot (max(x) for dr in ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10])) t

order by    1
;

╔════════════════════════╦══════╦══════╦══════╦══════╦══════╦══════╦══════╦══════╦══════╦══════╗
║ EditorTypeSymbolicName ║  1   ║  2   ║  3   ║  4   ║  5   ║  6   ║  7   ║  8   ║  9   ║  10  ║
╠════════════════════════╬══════╬══════╬══════╬══════╬══════╬══════╬══════╬══════╬══════╬══════╣
║ NULL                   ║ AA   ║ BB   ║ CC   ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║
║ 01                     ║ X    ║ X    ║ X    ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║
║ 02                     ║ NULL ║ X    ║ X    ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║
║ 03                     ║ X    ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║ NULL ║
╚════════════════════════╩══════╩══════╩══════╩══════╩══════╩══════╩══════╩══════╩══════╩══════╝

Answer 2

这是一种方法。在动态查询中，

为 EditorTypeSymbolicName X ClaimFieldSymbolicName（CROSS JOIN 部分）创建了一个 "matrix"
然后链接到原始 table 以查找每个组合是否存在一行（LEFT JOIN 部分）
如果存在一行，Value 将替换为您需要的 X（Value=CASE ... 部分）

CREATE TABLE #TempEditElements(EditorTypeSymbolicName VARCHAR(128),ClaimFieldSymbolicName VARCHAR(128),Value VARCHAR(128));
INSERT INTO #TempEditElements(EditorTypeSymbolicName,ClaimFieldSymbolicName,Value)
VALUES ('01','AA',NULL),('01','BB',NULL),('01','CC',NULL),('02','BB',NULL),('02','CC',NULL),('03','AA',NULL);

DECLARE @cols NVARCHAR(MAX), @sql NVARCHAR(MAX);

SET @cols = STUFF((SELECT DISTINCT ',' + QUOTENAME(ClaimFieldSymbolicName)
            FROM #TempEditElements
            ORDER BY 1
            FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)'),1,1,'')

SET @sql = 'SELECT EditorTypeSymbolicName, ' + @cols + '
              FROM
            (
                SELECT 
                    ets.EditorTypeSymbolicName,
                    cfs.ClaimFieldSymbolicName,
                    Value=CASE WHEN tee.EditorTypeSymbolicName IS NULL AND tee.ClaimFieldSymbolicName IS NULL THEN NULL ELSE ''X'' END
                FROM
                    (SELECT DISTINCT EditorTypeSymbolicName FROM #TempEditElements) AS ets
                    CROSS JOIN (SELECT DISTINCT ClaimFieldSymbolicName FROM #TempEditElements) AS cfs
                    LEFT JOIN #TempEditElements AS tee ON 
                        tee.EditorTypeSymbolicName=ets.EditorTypeSymbolicName AND 
                        tee.ClaimFieldSymbolicName=cfs.ClaimFieldSymbolicName
            ) s
            PIVOT
            (
              MAX(Value) FOR ClaimFieldSymbolicName IN (' + @cols + ')
            ) p
            order by EditorTypeSymbolicName'

EXECUTE (@sql);

DROP TABLE #TempEditElements;

结果是：

╔════════════════════════╦══════╦══════╦══════╗
║ EditorTypeSymbolicName ║  AA  ║  BB  ║  CC  ║
╠════════════════════════╬══════╬══════╬══════╣
║                     01 ║ X    ║ X    ║ X    ║
║                     02 ║ NULL ║ X    ║ X    ║
║                     03 ║ X    ║ NULL ║ NULL ║
╚════════════════════════╩══════╩══════╩══════╝

转置和识别带有标志的单元格

Transposing and identifying the cell with a flag

tsql

sql-server

pivot

sql-server-2008

静态解决方案

动态解决方案

半静态解决方案