1、group by 后想把列合并起来,可以用WMSYS.WM_CONCAT(FIELD_NAME),这在实际写sql中很常用。
2、我们一般在写sql里用group by来分组,有时候要求得到group by的数目时,常常会加上count。这时候如果除了分组的字段,我还要想select其它的字段的话,会报不是分组的字段的错误。而且,我们常常有通过select 出来的内容是根据count(*)排序后的内容,这时候可以用:
OVER(partition by FIELD_NAME1 order by FIELD_NAME2)来进行,且可以根据分组的count的数量进行排序,如:
select *
from (select tunnel_name 隧道名称,CONTENT 事件内容,STYLE 事件类型,AFF_LEVEL 事件等级,location 位置,PLACE_MILE 地点桩号,BEGIN_TIME 发生时间,CENTER_MILE 隧道中心桩号,TUNNEL_LEN 隧道长度,ROAD_LINE_NAME 所属路线,
count(*) OVER(partition by tunnel_name order by tunnel_name) CNT
from epub.tunnel_rank_temp)
order by CNTdesc
3、网上摘录的资料:(原文:http://hi.baidu.com/jinliyixi/blog/item/41f180c75cc0fca48326ace4.html)
wmsys.wm_concat、sys_connect_by_path、自定义函数实现行列转换
构建测试表:
Sql代码
create table TABLE1
(
ID INTEGER,
NAME VARCHAR2(10)
)
create table TABLE2
(
ID INTEGER,
ROLE VARCHAR2(10)
)
insert into TABLE1 (ID, NAME) values (1, '张三');
insert into TABLE1 (ID, NAME) values (2, '李四');
commit;
insert into TABLE2 (ID, ROLE) values (1, '查询');
insert into TABLE2 (ID, ROLE) values (1, '分析');
insert into TABLE2 (ID, ROLE) values (1, '决策');
insert into TABLE2 (ID, ROLE) values (2, '查询');
commit;
create table TABLE1
(
ID INTEGER,
NAME VARCHAR2(10)
)
create table TABLE2
(
ID INTEGER,
ROLE VARCHAR2(10)
)
insert into TABLE1 (ID, NAME) values (1, '张三');
insert into TABLE1 (ID, NAME) values (2, '李四');
commit;
insert into TABLE2 (ID, ROLE) values (1, '查询');
insert into TABLE2 (ID, ROLE) values (1, '分析');
insert into TABLE2 (ID, ROLE) values (1, '决策');
insert into TABLE2 (ID, ROLE) values (2, '查询');
commit;
要求输出结果:
Sql代码
ID NAME ROLE
1 张三 查询,分析,决策
2 李四 查询
ID NAME ROLE
1 张三 查询,分析,决策
2 李四 查询
方法一、使用wmsys.wm_concat
Sql代码
select table1.*,wmsys.wm_concat(role) from table1,table2 where table1.id=table2.id
group by table1.id,table1.name
select table1.*,wmsys.wm_concat(role) from table1,table2 where table1.id=table2.id
group by table1.id,table1.name
方法二、使用sys_connect_by_path
Sql代码
select id, name, ltrim(max(sys_connect_by_path(role, ',')), ',') from
(select row_number() over(partition by table1.id order by name) rn,table1.*, role from table1, table2 where table1.id =
table2.id)
start with rn = 1
connect by prior rn = rn - 1 and prior id = id
group by id, name
order by id
select id, name, ltrim(max(sys_connect_by_path(role, ',')), ',') from
(select row_number() over(partition by table1.id order by name) rn,table1.*, role from table1, table2 where table1.id =
table2.id)
start with rn = 1
connect by prior rn = rn - 1 and prior id = id
group by id, name
order by id
方法三、使用自定义函数
Sql代码
create or replace function my_concat(mid in integer) return varchar2 --记住:参数和返回值里的数据类型都不用定义长度
is
result varchar2(4000); --定义变量,记住Oracle中定义变量不需要
begin
for temp_cursor in (select role from table2 where id=mid) loop --此处在游标FOR循环中使用查询
result :=result || temp_cursor.role || ','; --Oracle中字符连接使用||,而sql server中用+
end loop;
result := rtrim(result,','); --去掉最后一个空格,还有Oracle中的赋值前面没有set
return result;
end;
select table1.*,my_concat(table1.id) from table1,table2 where table1.id=table2.id
group by table1.id,table1.name
order by table1.id