Question

假设我在1NF中有一个带有以下表格的MySQL数据库：

sandwich_id  |  sandwich_name  | sandwich_price
_______________________________________________
0            |  BLT            |  5.5
1            |  Reuben         |  7.0
3            |  Grilled Cheese |  3.75
...

和一个存储所有成分值的单独表格：

sandwich_id |  ingredient
__________________________
0           |  bacon
0           |  lettuce
0           |  tomato
1           |  corned beef
1           |  swiss cheese
...

如何比较所有三明治的成分，以确定哪些最相似？

（另外，我缺少第二个表的技术术语吗？我想把它称为地图表，但我知道这不太对，因为地图表存储了两个表的外键并且这个更多的是第一个分支......）

Answer 1

select sw1.sandwich_name, sw2.sandwich_name, COUNT(i1.ingredient) as [Ingredients in common]
from sandwiches sw1
join sandwiches sw2 on sw2.sandwich_id < sw1.sandwich_id
join ingredients i1 on i1.sandwich_id = sw1.sandwich_id
where exists (select 1
                from sandwiches sw
                join ingredients i on i.sandwich_id = sw.sandwich_id
               where i.ingredient = i1.ingredient
                 and sw.sandwich_id = sw2.sandwich_id)
group by sw1.sandwich_name, sw2.sandwich_name
order by [Ingredients in common] desc, sw1.sandwich_name, sw2.sandwich_name

如何比较两个外键值之间存储在array-table中的所有值？

1 个答案: