试试这个

with t as ( select  unix_timestamp('10-01-2019' , 'dd-MM-yyyy') as dt )
select from_unixtime(dt,'dd')  as Date,
       from_unixtime(dt,'MMMM')  as Month,
       from_unixtime(dt,'YYYY')  as Year
 from t;

结果

Total MapReduce CPU Time Spent: 2 seconds 720 msec
OK
10  January 2019
Time taken: 23.206 seconds, Fetched: 1 row(s)

使用split（）：

with your_data as(
select stack(3,'10-01-2019',   
               '20-02-2019',
               '22-03-2019'
        ) as dt
) --use your table instead of this

select dt[0] as day,
       dt[1] as month,
       dt[2] as year
from ( select split(dt,'-') as dt from your_data )s;

结果：

OK
day     month   year
10      01      2019
20      02      2019
22      03      2019
Time taken: 0.081 seconds, Fetched: 3 row(s)

我们需要使用 from_unixtime and unix_timestamp 函数来解析日期。
那么 split the field 在 subquery 提取日期，月份，年份。。 Example:hive> select dt[0] day,dt[1] month,dt[2] year from( select split(from_unixtime(unix_timestamp("10-01-2019",'dd-MM-yyyy'),'dd-MMMM-yyyy'),'-')dt )e;Result: ```
day month year
10 January 2019