我有两张桌子:
financials_standalone ('fin_id', 'attr_id', 'year', 'value');
financials_consolidated ('fin_id', 'attr_id', 'year', 'value');
('fin_id', 'attr_id', 'year') is the unique key
financials_consolidated表除了financials_standalone之外还会有数据。
例如:
financials_standalone
| fin_id | attr_id | year | value |
---------------------------------------
| fin01 | pe | 2016 | 33.23 |
| fin02 | pe | 2016 | 12.52 |
financials_consolidated
| fin_id | attr_id | year | value |
---------------------------------------
| fin02 | pe | 2016 | 20.41 |
现在我想将两个表组合成一个视图: - 如果行存在于合并中,则选择该行,否则从financials_standalone表中选择行。
所以最终的视图输出应该是这样的
financials_data_view
| fin_id | attr_id | year | value |
---------------------------------------
| fin01 | pe | 2016 | 33.23 |
| fin02 | pe | 2016 | 20.41 |
我无法通过case-when或left outer join找到解决方案。如何获得此视图输出?
答案 0 :(得分:1)
在financials_standalone
上左连接financials_consolidated
以在所有情况下从financials_consolidated
获取值,并使用coalesce()函数从2个表中返回第一个非空值。然后执行联合以从financials_consolidated
获取这些记录以从该表中获取另一个中不存在的记录。如果情况不是这样,那么你就不需要工会了。
select fs.fin_id, fs.attr_id, fs.year, coalesce(fc.value, fs.value) as val
from `financials_standalone` fs
left join `financials_consolidated` fc
on fs.fin_id=fc.fin_id
and fs.attr_id=fc.attr_id
and fs.year=fc.year
union
select fc.fin_id, fc.attr_id, fc.year, fc.value
from `financials_consolidated` fc
left join `financials_standalone` fs
on fs.fin_id=fc.fin_id
and fs.attr_id=fc.attr_id
and fs.year=fc.year
where fs.fin_id is null