创建表：
create table hive_wordcount(context string); load data local inpath ‘/home/hadoop/files/helloworld.txt‘ into table hive_wordcount;
执行查询SQL：
select word, count(*) from hive_wordcount lateral view explode(split(context,‘\t‘)) wc as word group by word;
原文：http://www.cnblogs.com/one--way/p/7543751.html