让我们说我有表联盟(城市,团队)和排名(城市,评级)(所有都是字符串,除了评级是真实类型)。排名中有额外的城市不一定在联赛中。我正在尝试创建一个名为nearestRating的函数,该函数将接收一个城市并返回其城市具有最小评级差距(意味着不是自身)的团队的名称与函数参数中的城市或对应于该城市的团队的名称城市参数如果我们正在寻找的城市也存在于联盟中。 (城市在联赛和排名中被定义为varchar(20))。
我对它的尝试有一些非常奇怪的语法错误
CREATE FUNCTION closestRating (city varchar(20)) RETURNS VARCHAR(20) AS $$
DECLARE minimum real;
DECLARE rat real;
BEGIN
SET minimum = -1;
FOR r.city, rating IN (SELECT r.city, r.rating FROM rankings r)
LOOP
IF city = r.city
BEGIN
SET rat = rating;
END IF;
END LOOP;
FOR rating IN (SELECT r.city, r.rating FROM rankings r, league l WHERE r.city = l.city)
LOOP
IF minimum = -1 OR minimum < ABS(rat – rating)
BEGIN
SET minimum_dist = ABS(r – rating);
SET rat = rating;
END IF;
END LOOP;
RETURN (SELECT team from rankings, league WHERE rating = rat));
END;
$func$
LANGUAGE plpgsql;
答案 0 :(得分:0)
鉴于此数据:
create table league (
city varchar(20),
team varchar(20)
);
create table rankings (
city varchar(20),
rating real
);
insert into rankings (city, rating) values
('citya', 1),('cityb', 1.1),('cityc', 1.01),('cityd', 1.102),('citye', 1.2)
;
insert into league (city, team) values
('citya', 'teama'),('cityc', 'teamc'),('citye', 'teame')
;
此查询将返回预期输出:
with given_city as (
select rating, city
from rankings
where city = 'citya'
)
select l.team
from
league l
inner join
rankings r using (city)
cross join
given_city gc
order by abs(r.rating - gc.rating)
limit 1
;
team
-------
teama