SQL remove and filter similar results - sql

Question

I have the below query which gives me a list of orders where an address comes up twice so we can double pack same orders to the same address
select * from Orders
where
address in (select address from orders group by address having count(*) = 2)
CustID StockID Address Company
-----------------------------------------------------------------
1217 23185 1 Some Road Stockton
58458 23185 1 Some Road
58459 23185 4 John St
58457 23185 4 John St
299576 23185 9 Roadway PDE Graceland
59470 23185 9 Roadway PDE Cahill Tow
97656 23185 24 Kent St
97677 23185 24 Kent St
212732 23185 23 Best Rd
226583 23185 23 Best Rd c/o John
191718 23185 98 King St
156363 23185 98 King St
121106 23185 19 Broadway
156362 23185 19 Broadway
I want the result to look like this which excludes any addresses which come up which have a company name in either of the 2 results that come up for it. Some addresses have nothing in the Company name however i want to exclude them as well if the other result for the same address contains a company name.
CustID StockID Address Company
-----------------------------------------------------------------
58459 23185 4 John St
58457 23185 4 John St
97656 23185 24 Kent St
97677 23185 24 Kent St
191718 23185 98 King St
156363 23185 98 King St
121106 23185 19 Broadway
156362 23185 19 Broadway
Hope this all makes sense and appreciate any help
Thank you!

Answer 1

select * from Orders o1
where
0 = (select count(*) from orders o2
where o1.address = o2.address
and company is not null)
This is a correlated sub-query which ties the main query and subquery together over address. I have assumed "empty" company is NULL, replace with <> "" if you don't use nulls.

Answer 2

AS #ClockWork-Muse mentions in the comments, there's a lot of things you need to consider in this scenario, pertaining to data cleanliness as well as business logic. Having said that, you can try this for your issue:
;with filtered as
(
select * from Orders
where
address in (select address from orders group by address having count(*) = 2)
)
,cte as
(select address, max(company) as max
from filtered
group by address
having max(company) = '')
select f.* from filtered f
inner join cte c on f.address = c.address
Basically, you create a CTE to identify those companies which have only blank values for Company, and then join back to the results of your query