如果您有一个数字表，您可以：

select s.product, s.category, 1 as qty
from sales s join
     numbers n
     on n.n <= s.qty and n.n > 0;

在mysql中，可以使用变量生成这样的表。例如，如果 sales 足够大：

select s.product, s.category, 1 as qty
from sales s join
     (select (@rn := @rn + 1) as n
      from sales s cross join
           (select @rn := 0) params
     ) n
     on n.n <= s.qty and n.n > 0;

Hive

select  t.product
       ,t.category
       ,1           as qty

from    mytable t
        lateral view explode(split(space(t.qty - 1),' ')) e
;

+---------+----------+-----+
| product | category | qty |
+---------+----------+-----+
| 123     | A        | 1   |
+---------+----------+-----+
| 123     | A        | 1   |
+---------+----------+-----+
| 123     | A        | 1   |
+---------+----------+-----+
| 345     | B        | 1   |
+---------+----------+-----+
| 345     | B        | 1   |
+---------+----------+-----+

我自己想办法回答（这是在Hive里）

select category,main1.product,main2.qty from (select category,product,
split(substring(repeat(concat(1,','),int(qty)),0,int(qty*2)-1),",") as qty from 
(select collect_set(product) product,category,
sum(qty) qty
from sales) main
lateral view explode(product) main1 as product
lateral view explode(qty) main2 AS qty;