函数:http://my.oschina.net/Kenyon/blog/108303
紧接上述,补充一下:
输入/输出参数的函数demo(输入作为变量影响sql结果,输出作为结果返回)
create or replace function f_dept_salary_out2(int, out o_dept text,out o_salary text) returns setof record as $$ declare v_rec record; begin for v_rec in EXECUTE 'select departmentid as dept_id, sum(salary) as total_salary from f_get_employee() group by departmentid having departmentid='||$1 loop o_dept:=v_rec.dept_id; o_salary:=v_rec.total_salary; return next; end loop; end; $$ language plpgsql; select * from f_dept_salary_out2(2);
递归查询:http://my.oschina.net/kenshiro/blog/160129
由于资源页的数据源没有,因此在这里重新做了个demo,如下:
1. 准备数据
create table Tag( id int, name text, parent_id int ); insert into Tag values(1,'all products',-1),(2,'plastic',1),(3,'metal',1),(4,'toy',2),(5,'furniture',2),(6,'knife',3);
这是一个产品的tag表。如下:
树形描述如下:
all products --------plastic--------toy
| |--furniture
|---metal --------knife
2. 找子孙
即:根据某个节点寻找子树。
例如:找到plastic及下面所有的tag,sql如下:
with recursive r as( select * from Tag where id = 2 union all select Tag.* from Tag, r where Tag.parent_id = r.id ) select * from r order by id;
输出会如下:
id name parent_id
2 plastic 1
4 toy 2
4 furniture 2
3. 找祖先
即:根据某个节点寻找其所有父系节点。
例如:找到knife的所有父辈们,sql如下:
with recursive r as ( select * from Tag where id = 6 union all select Tag.* from Tag, r where Tag.id = r.parent_id ) select * from r order by id desc;
输出会如下:
id name parent_id
6 knife 3
3 metal 1
1 all products -1
后记:把r理解成下"当前"记录就好了。