如何從SQL Server表中選擇不同的行

Question

我正在將現有的工作查詢從MySQL移動到MS SQL Server（版本11），我無法轉換在MySQL上正常工作的查詢之一。

它基本上只考慮4列值從表中選擇不同的行。

表是（有數據的最小的子集，同一類型的數據存在於不同的名稱，sub_name和用戶）：

+-------+-------+-------+-------+-----------------+----------+--------+ 
| col_1 | col_2 | col_3 | col_4 | name   | sub_name | user | 
+-------+-------+-------+-------+-----------------+----------+--------+ 
| 50 | 40 | 20 | 30 | one_four_eleven | One  | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Two  | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Three | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Four  | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Five  | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Six  | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Seven | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Eight | user_1 | 
| 50 | 40 | 20 | 30 | one_four_eleven | Nine  | user_1 | 
| 50 | 40 | 15 | 30 | one_four_eleven | Ten  | user_1 | 
+-------+-------+-------+-------+-----------------+----------+--------+ 

而我需要的是，如果所有COL_1，COL_2，col_3和col_4與其他行相比，在一行中具有相同的值，那麼應該只選擇其中一個相同的行。

由於在給定的數據中，只有最後一行在col_3中有不同的數據，所以結果應該只有兩行。一行在col_1，col_2，col_3，col_4和名稱中顯示相同的值。第二行應該是最後一行。

我能夠做到這一點使用MySQL查詢如下圖所示，它的結果是：

mysql> SELECT DISTINCT col_1, col_2, col_3, col_4, name FROM table_name 
WHERE (user = 'user_1' OR user = 'GLOBAL') AND name = 'one_four_eleven' AND sub_name != 'ALL' GROUP BY sub_name; 
+-------+-------+-------+-------+-----------------+ 
| col_1 | col_2 | col_3 | col_4 | name   | 
+-------+-------+-------+-------+-----------------+ 
| 50 | 40 | 20 | 30 | one_four_eleven | 
| 50 | 40 | 15 | 30 | one_four_eleven | 
+-------+-------+-------+-------+-----------------+ 

但是，當我試圖此查詢轉換爲MS SQL Server上運行，它給錯誤：

Column 'table_name.col_1' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause. 

如何爲SQL Server編寫查詢，該查詢會得到與上述MySQL查詢相同的結果。

Answer 1

隨着給出的示例數據，這將工作：

SELECT col_1, col_2, col_3, col_4, name 
FROM table_name 
WHERE (user = 'user_1' OR user = 'GLOBAL') 
    AND name = 'one_four_eleven' 
    AND sub_name != 'ALL' 
GROUP BY col_1, col_2, col_3, col_4, name 

或本

SELECT distinct col_1, col_2, col_3, col_4, name 
FROM table_name 
WHERE (user = 'user_1' OR user = 'GLOBAL') 
    AND name = 'one_four_eleven' 
    AND sub_name != 'ALL' 

哦，如果你是想知道mysql基本上是這樣的：

SELECT max(col_1), max(col_2), max(col_3), max(col_4), max(name) 
FROM table_name 
WHERE (user = 'user_1' OR user = 'GLOBAL') 
    AND name = 'one_four_eleven' 
    AND sub_name != 'ALL' 
GROUP BY sub_name 

但這並不如前兩個好。

Answer 2

，我認爲這可以幫助你：

;WITH cte AS (
SELECT * 
, ROW_NUMBER() OVER(PARTITION BY col_1, col_2, col_3, col_4, name ORDER BY (SELECT NULL)) AS rn 
FROM yourTable 
WHERE (user = 'user_1' OR user = 'GLOBAL') AND name = 'one_four_eleven' AND sub_name != 'ALL' 
) 

SELECT * FROM cte 
WHERE rn = 1