(T-SQL): How do I use result set A to calculate result set B? - sql

Question

I have 2 queries, A and B, which I currently use independently from each other. They both return an employee ID and some metric values. I now want to use the employee IDs returned from result set A in query B.
A's query is structured like this:
select employee_id from employee where employee_team = 1 and employee_role = 1
B's query is structured like this:
declare #tester int = 123450 --plug in employee ID
select employee_id
,employee_name
,sum(case
when notes.note_author!=employee_id and logs.log_date<#today
then 1 else 0 end) as metric
from notes
inner join note_to_log_bridge as br on notes.note_id=br.note_id
inner join logs on br.log_id=logs.log_id
inner join employee on employee_id=#Tester
If I want to get B's metrics for 5 employees, I have to run query B 5 times, changing the #Tester variable each time. I'd like to instead find some way of automating that, so that I get the metrics for query B for every employee_id in result set A.
I tried stored result set A as a CTE and using a while loop to run through query B:
declare #line=1
with cte (employee_id) as <query_a>
while (#line<=count(cte.employee_id))
begin <query b>...
I never finished this query because I discovered that while cannot follow the creation of a CTE.
I tried using a table variable:
declare #set_a (employee_id int)
insert into #set_a <query a>
but when I try to use #set_a in query B, I get a message saying that I need to declare the scalar variable #set_a.
I tried using a temp table and got a "could not be bound" error.
I am out of ideas. Am I approaching this problem in anything resembling the right direction? Is this even possible?
Thank you!

Answer 1

Yes, you can use cursor and it will work perfectly fine.
But, if you have significantly more than 5 rows you may consider using CROSS APPLY to make it all in one query. It may work faster than cursor.
select
employee.employee_id
,CA.*
from
employee
CROSS APPLY
(
<put your query B here
and replace all references to parameter #Tester
with employee.employee_id>
) AS CA
where employee.employee_team = 1 and employee.employee_role = 1
You can think of this operator like this: for each row in the main outer query A CROSS APPLY runs the inner query B with the possibility to reference values of the row from the outer query A (in this case employee.employee_id).

Answer 2

Use Cursor ?
If I want to get B's metrics for 5 employees, I have to run query B 5
times, changing the #Tester variable each time.
DECLARE #empid int;
DECLARE vend_cursor CURSOR
FOR select employee_id from employee where employee_team = 1 and employee_role = 1
OPEN vend_cursor
FETCH NEXT FROM vend_cursor into #empid;
WHILE ##FETCH_STATUS = 0
BEGIN
// your query with #mpid
FETCH NEXT FROM db_cursor INTO #name
END
CLOSE vend_cursor
DEALLOCATE vend_cursor