全部，我正在使用以下续集查询...
select
u.full_name, u.id, sum(hit_count), tag_names
from MyTable as t
inner join MyUsers as u on u.id=t.user_id
where t.created_at >= '2015-1-01 04:00:00' AND t.created_at < '2015-2-01 04:00:00' AND tag_names like '%Tree%'
group by u.id
...使用大约1,000行呈现一个看起来像这样的数据表：
| full_name | id | sum(hit_count) | tag_names |
| Fred Johnson | 001 | 1500 | Tree, Apple, Blue, Banana |
| Joe Smith | 002 | 663 | Tree, Red, Pear, Yellow |
| Tina Cook | 001 | 1500 | Tree, Berry, Red, Grape |
| Kyle Cross | 001 | 1500 | Tree, Berry, Red, Grape |
但是，因为我知道有限数量的原色标签（例如红色，蓝色，黄色）可用，有没有办法调整我的续集查询也按主色标签分组（并添加一个主要颜色标签列到我的表）来渲染......
| full_name | id | sum(hit_count) | tag_names | Primary_Color_Tag |
| Fred Johnson | 001 | 1500 | Tree, Apple, Blue, Banana | Blue |
| Joe Smith | 002 | 500 | Tree, Red, Pear, Yellow | Red |
| Tina Cook | 001 | 1500 | Tree, Berry, Red, Grape | Red |
| Kyle Cross | 001 | 1500 | Tree, Berry, Red, Grape | Red |
| Joe Smith | 001 | 163 | Tree, Red, Pear, Yellow | Yellow |
我在similar questions上找到了this forum看似相关的解决方案，但我无法拼凑他们的逻辑来重现我追求的表格。任何见解将不胜感激！