hive j简单邮件过滤

select min(call_log),
a.mail_subject from
(select mail_to,mail_subject from dw_user_deviler_201408 where
log_date>=20140825 and mail_subject not like ‘%?%‘ and mail_subject not like ‘%时%‘ and mail_to!=‘‘ and
mail_subject!=‘‘ and mail_subject rlike ‘.*[\u4e00-\u9fa5]+.*‘) a
lateral view explode(split(mail_to,"\;")) b as call_log group by a.mail_subject;

原文：http://www.cnblogs.com/charlie-badegg/p/3944986.html