经典hive面试题
输入
1,2,4,5,6,8,9
输出 就是能连续上的搞在一起 用hive解决
1,2
4,5,6
8,9
直接进入主题：
建表CREATE TABLE people(
number varchar(256),
id int,
pt_d int)
ROW FORMAT SERDE
‘org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe’
STORED AS INPUTFORMAT
‘org.apache.hadoop.mapred.TextInputFormat’
OUTPUTFORMAT
‘org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat’
LOCATION
‘hdfs://centos01:9000/user/hive/warehouse/people’
插入数据：insert into people values(‘1,2,4,5,6,8,9’,1,20200607)；
查看是否插入成功：
执行如下语句
SELECT
collect_set(CONCAT_WS(’,’,ranknum)) AS ranknum
FROM
(SELECT
id
,ranknum
,ranknum - next_ranknum AS dif
FROM
(
SELECT
id
,ranknum
,ROW_NUMBER() OVER(PARTITION BY id ORDER BY ranknum) AS next_ranknum
FROM
(
SELECT
id
,ranknum
FROM people LATERAL VIEW explode(split(number,’,’))rank AS ranknum
) t
) tt
) ttt
group by dif
同样的在mysql中如何实现呢？
SELECT
group_concat(val) as val
FROM
(
select
val
,rank-val AS dif
FROM
(select
val
,@num:= @num + 1 AS rank
from
(select substring_index(substring_index(a.address,’,’,b.help_topic_id),’,’,-1) AS val
from
people a
join
mysql.help_topic b
on b.help_topic_id <= (length(a.address) - length(replace(a.address,’,’,’’))+1)
) c
,(select @num:=0)d
order by rank
) e WHERE val is NOT NULL AND val!=’’
) f
group by dif
order by dif desc
结果如图：
