今天做个后台服务,有个需求是批量生成一批表的数据,如果用BulkInsert会提升很大一截提交效率,但是如果用循环构造提交的Datable,则算法开销太高,所以用这种查询批量查出符合格式的DataTable结果集。
*思想就是将查询列做CASE判断值,然后放在 SELECT段中作为别名列显示字段值,然后再根据ID分组。
下面是内层的原始查询效果:
下面是行转列后查询效果如图:
下面是上面两张图的最终查询语句:
1 SELECT f_gameID,SUM(score) AS totNum, 2 SUM(CASE f_sub WHEN '01' THEN score ELSE 0 END) AS f_subject01,SUM(CASE f_sub WHEN '02' THEN score ELSE 0 END) AS f_subject02, 3 SUM(CASE f_sub WHEN '03' THEN score ELSE 0 END) AS f_subject03,SUM(CASE f_sub WHEN '04' THEN score ELSE 0 END) AS f_subject04, 4 SUM(CASE f_sub WHEN '05' THEN score ELSE 0 END) AS f_subject05,SUM(CASE f_sub WHEN '06' THEN score ELSE 0 END) AS f_subject06, 5 SUM(CASE f_sub WHEN '07' THEN score ELSE 0 END) AS f_subject07,SUM(CASE f_sub WHEN '08' THEN score ELSE 0 END) AS f_subject08, 6 SUM(CASE f_sub WHEN '09' THEN score ELSE 0 END) AS f_subject09 7 FROM 8 ( 9 SELECT u.f_gameID,u.f_sub,COUNT(u.f_sub) AS score 10 FROM t_mfg_GameUserAnswers AS u 11 LEFT JOIN t_mfg_Class AS c ON u.f_gameId = c.f_gameId 12 WHERE u.f_startTime > '2015-07-20' 13 AND u.f_endTime < '2015-07-24' 14 -- AND c.f_schoolId = 193034 15 GROUP BY u.f_gameID ,u.f_sub 16 ) AS a 17 GROUP BY f_gameID;