[大数据]Hive str to array / regexp_extract

with releases as (select '["us","ca","fr"]' as country)

select split(regexp_extract(country,'^\\["(.*)\\"]$',1),'","')

from releases;

["us","ca","fr"]

--------

with releases as (select '["us:[abc]","ca:[abc]","fr:[abc]"]' as country)

select split(regexp_extract(country,'^\\["(.*)\\"]$',1),'","') -- 提取开头的[和结尾的]之间的内容

from releases;

["us:[abc]","ca:[abc]","fr:[abc]"]

中间的[]依然存在。

-------

with releases as (select '["us:[abc]","ca:[abc]","fr:[abc]"]' as country)
select  split(regexp_extract(country,'^\\["(.*)\\"]$',1),'","')  as  c1 , -- ["us:[abc]","ca:[abc]","fr:[abc]"]
split(regexp_extract(country,'^\\["(.*)\\"]$',0),'","')  as  c2  -- ["[\"us:[abc]","ca:[abc]","fr:[abc]\"]"]
     from    releases;

0表示把整个正则表达式对应的结果全部返回
1表示返回正则表达式中第一个() 对应的结果 以此类推

Hive,regexp_extract用法--注意圆括号！！

end