前两天产品有个需求，相同的商品因为价格不同而分开展示，但是明细还是算一条明细，具体区分展示出商品的价格和数量信息，其他重复的商品信息要置空。
需求并不难，用程序代码循环处理就可以了。但是后面涉及到打印报表，只能用纯sql语句生成。开始自己用了union 写的复杂了一些，后来百度了下文章看到一种实现思路，在此记录一下。下面的内容是根据其思路分析自己修改实现的一个demo，也简单扩展了一点实现功能。
直接上代码好了：
-- ---------------------------- -- Table structure for act -- ---------------------------- DROP TABLE IF EXISTS `act`; CREATE TABLE `act` ( `id` int(10) NOT NULL AUTO_INCREMENT COMMENT 'id', `name` varchar(10) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NULL DEFAULT NULL COMMENT 'name', `type` varchar(12) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NULL DEFAULT NULL COMMENT 'type', `peple` varchar(10) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NULL DEFAULT NULL COMMENT 'peple', `age` int(3) NULL DEFAULT NULL COMMENT 'age', `sex` int(1) NULL DEFAULT NULL COMMENT 'sex', PRIMARY KEY (`id`) USING BTREE ) ENGINE = InnoDB AUTO_INCREMENT = 10 CHARACTER SET = utf8mb4 COLLATE = utf8mb4_general_ci ROW_FORMAT = Dynamic;
-- ---------------------------- -- Records of act -- ---------------------------- INSERT INTO `act` VALUES (1, '张三', 'A', 'a', 10, 1); INSERT INTO `act` VALUES (2, '张三', 'A', 'a', 10, 0); INSERT INTO `act` VALUES (3, '王五', 'B', 'b', 10, 1); INSERT INTO `act` VALUES (4, '赵六', 'B', 'b', 20, 0); INSERT INTO `act` VALUES (5, '刘七', 'C', 'c', 30, 1); INSERT INTO `act` VALUES (6, '阿三', 'D', 'd', 10, 0); INSERT INTO `act` VALUES (7, '阿四', 'D', 'd', 20, 0); INSERT INTO `act` VALUES (8, '阿五', 'A', 'a', 30, 1); INSERT INTO `act` VALUES (9, '张三', 'A', 'a', 30, 1);
生成表数据：
下面案列是以type相同与否作为判断条件分组，peple就是type的小写，方便对照观察。后面的Bk结尾的别名字段也是方便对照观察加上的。
实现sql:
SELECT id, CASE WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.name = a.name GROUP BY type HAVING COUNT(*) > 1 ) THEN name WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.name = a.name GROUP BY type HAVING COUNT(*) = 1 ) THEN name ELSE '' END AS 'name' , CASE WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) >= 1 ) THEN type WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) = 1 ) THEN type ELSE '' END AS type, CASE WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.age = a.age GROUP BY type HAVING COUNT(*) >= 1 ) THEN age WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.age = a.age GROUP BY type HAVING COUNT(*) = 1 ) THEN age ELSE '' END AS age, CASE WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.sex = a.sex GROUP BY type HAVING COUNT(*) >= 1 ) THEN sex WHEN id in ( SELECT MIN(id) FROM (select * from act) as a WHERE b.sex = a.sex GROUP BY type HAVING COUNT(*) = 1 ) THEN sex ELSE '' END AS sex, peple,name as nameBk,age as ageBk,sex as sexBk FROM (select * from act order by type,id asc) as b
执行结果：
如果干脆要把任意字段置空（不管是不是重复的）也可以：
SELECT id, CASE WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) > 1 ) THEN name WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) = 1 ) THEN name ELSE '' END AS name, CASE WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) >= 1 ) THEN type WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) = 1 ) THEN type ELSE '' END AS type, CASE WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) >= 1 ) THEN age WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) = 1 ) THEN age ELSE '' END AS age, CASE WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) >= 1 ) THEN sex WHEN id = ( SELECT MIN(id) FROM (select * from act) as a WHERE b.type = a.type GROUP BY type HAVING COUNT(*) = 1 ) THEN sex ELSE '' END AS sex, peple,name as nameBk,age as ageBk,sex as sexBk FROM (select * from act order by type,id asc) as b
执行结果：
上面案例都是以type作为判断分组条件，如果是有多个字段，直接在后面接着增加就可以了。