Question

有没有办法简化下面的查询？我得到了我想要的但我想学习如何通过使用更短的查询来实现结果。至于结果，我可以将值NULL替换为0吗？

SELECT ad.Staff, 
       ad.TYear, 
       ad.TWeek, 
       b.TMonth 
  FROM (SELECT a.Staff, 
               a.TYear, 
               d.TWeek
          FROM (SELECT t2.creatorfullname AS Staff, 
                      count(distinct(t1.ticketmaskid)) AS TYear 
                 FROM swtickets t1 
                      JOIN swticketauditlogs t2 
                           ON t1.ticketid = t2.ticketid 
                WHERE (YEAR(from_unixtime(t2.dateline)) = YEAR(CURRENT_TIMESTAMP) 
                       AND FIND_IN_SET (t2.creatorfullname,'Andy Murray,Rafael Nadal,Maria Sharapova') > 0 
                       AND t2.actionmsg LIKE '%Ticket status changed from: % to: Closed%') 
                       AND totalreplies > 0 
              GROUP BY t2.creatorfullname) AS a 
                LEFT OUTER JOIN (SELECT t2.creatorfullname AS Staff, 
                                        count(distinct(t1.ticketmaskid)) AS TWeek 
                                   FROM swtickets t1 
                                        JOIN swticketauditlogs t2 
                                             ON t1.ticketid = t2.ticketid 
                                  WHERE (WEEK(from_unixtime(t2.dateline)) = WEEK(CURRENT_TIMESTAMP) 
                                         AND FIND_IN_SET (t2.creatorfullname,'Andy Murray,Rafael Nadal,Maria Sharapova') > 0 
                                         AND t2.actionmsg LIKE '%Ticket status changed from: % to: Closed%') 
                                         AND totalreplies > 0  
                                GROUP BY t2.creatorfullname) AS d 
                                 ON (a.Staff = d.Staff)) AS ad 
                LEFT OUTER JOIN (SELECT t2.creatorfullname AS Staff, 
                                        count(distinct(t1.ticketmaskid)) AS TMonth 
                                   FROM swtickets t1 
                                        JOIN swticketauditlogs t2 
                                             ON t1.ticketid = t2.ticketid 
                                  WHERE (MONTH(from_unixtime(t2.dateline)) = MONTH(CURRENT_TIMESTAMP) 
                                         AND FIND_IN_SET (t2.creatorfullname,'Andy Murray,Rafael Nadal,Maria Sharapova') > 0 
                                         AND t2.actionmsg LIKE '%Ticket status changed from: % to: Closed%') 
                                         AND totalreplies > 0  
                                GROUP BY t2.creatorfullname) AS b 
                                 ON (ad.Staff = b.Staff);

+----------------+-------+-------+--------+
| Staff          | TYear | TWeek | TMonth |
+----------------+-------+-------+--------+
| Andy Murray    |   337 |    37 |    142 |
| Rafael Nadal   |    49 |  NULL |     13 |
| Maria Sharapova|    49 |     1 |      4 |
+----------------+-------+-------+--------+

Answer 1

尝试：

 SELECT t2.creatorfullname AS Staff, 
        count(distinct case when YEAR(from_unixtime(t2.dateline)) = YEAR(CURRENT_TIMESTAMP) 
                       then t1.ticketmaskid end) AS TYear,
        count(distinct case when WEEK(from_unixtime(t2.dateline)) = WEEK(CURRENT_TIMESTAMP) 
                       then t1.ticketmaskid end) AS TWeek,
        count(distinct case when MONTH(from_unixtime(t2.dateline)) = MONTH(CURRENT_TIMESTAMP)
                       then t1.ticketmaskid end) AS TMonth  
 FROM swtickets t1 
 JOIN swticketauditlogs t2 
      ON t1.ticketid = t2.ticketid 
 WHERE FIND_IN_SET (t2.creatorfullname,'Andy Murray,Rafael Nadal,Maria Sharapova') > 0 
   AND t2.actionmsg LIKE '%Ticket status changed from: % to: Closed%' 
   AND totalreplies > 0 
 GROUP BY t2.creatorfullname

- 编辑 -

当我们需要计算满足许多条件的记录时，可以使用这种方法在这种方法中，我们使用最简单的案例表达式形式：

CASE WHEN condition THEN expression1 [ ELSE expression2 ] END

case表达式的作用类似于IF语句 - 它计算条件，当条件为真时，则返回第一个表达式的结果，否则返回第二个表达式的结果，或者 - 当没有给出第二个表达式时（{ {1}}部分被跳过） - 它返回else MySql有一个IF函数：NULL，其作用方式与case表达式相同，但case表达式是ANSII SQL Standard的一部分，由大多数数据库实现，而IF(expr1,expr2,expr3)函数是MySql的专有扩展。登记/> 请查看此链接了解详情：https://dev.mysql.com/doc/refman/5.6/en/control-flow-functions.html

计数记录的查询可能看起来像：

IF

SELECT count(*), count( some_column ), count( CASE WHEN condition1 THEN 1 END ), count( IF( condition2, 1, null ) ), sum( CASE WHEN condition3 THEN 1 ELSE 0 END ), count( DISTINCT CASE WHEN condition4 THEN some_column END ), count( DISTINCT IF( condition2, 1, some_column ) ) FROM sometable - 计算总行数
COUNT(*) - 当COUNT( expression )不为空（它跳过空值）时计算多个行 expression - 与上述相同，但仅考虑不同的值

COUNT( DISTINCT expression )

- 对于每一行，它评估count( DISTINCT CASE WHEN condition THEN some_column END )并在条件为真时返回值condition，否则返回null。由于跳过了空值，因此它只计算满足条件的行的不同值。

简化MYSQL查询

1 个答案: