How to synthesize too many SQL calls into one - sql

Question

A vb.net desktop application must make hundreds of these calls on a SQL Server (2012). The declaration is composed in the application and requires hundreds (or sometimes thousands) of times from the database.
I was looking for a way to synthesize the call into a single SQL statement or stored procedure if possible.
The difference between one statement and another is the "day" variable (in the example code is '20190724') , which is incremented by one day for each call.
SELECT SUM(table1.qta)
FROM table1
WHERE (table1.id = 35)
AND ('20190724' BETWEEN table1.date1 AND table1.date2)
The expected result would be 2 columns:
day1 SumQta1
day2 SumQta2
day3 SumQta3
....

Answer 1

What you should do is to change your approach.
Instead of querying the sum for each day individually, create a user defined table type to hold the dates, and use it to pass a table valued parameter to the database.
Then query the table joined to the variable using sum and group by the date.
Your SQL should look something like this:
CREATE TYPE dbo.DatesList
(
Date Date NOT NULL PRIMARY KEY
);
CREATE PROCEDURE GetSums
(
#Dates dbo.DatesList readonly
)
AS
SELECT Dates.Date, sum(table1.qta)
FROM table1
JOIN #Dates As Dates
ON Dates.Date BETWEEN table1.date1 AND table1.date2
WHERE table1.id = 35
GROUP BY Dates.Date
As for how to use a table valued parameter from Vb.Net - Well, that depends on how you connect to the database in the first place. I'm sure there are plenty of resources here on stackoverflow and also out on the web to help with that part.

Answer 2

Well, you could build a VALUES clause in your program with the days of interest. Then left join the table and aggregate to get the sum. Back in your application you can then iterate the result day for day.
SELECT d.day,
coalesce(sum(t.qta)) qta
FROM (VALUES ('20190101'),
('20190102'),
...) d (day)
LEFT JOIN table1 t
ON t.id = 35
AND d.day BETWEEN t.date1
AND t.date2
GROUP BY d.day
ORDER BY d.day;
Another option would be the use of a recursive CTE to generate the days from a minimum and maximum and then do the join on that, again aggregating to get the sums.
WITH
d (day)
AS
(
SELECT convert(date, '20190101') day
UNION ALL
SELECT dateadd(day, 1, day)
FROM d
WHERE dateadd(day, 1, day) <= convert(date, '20190131')
)
SELECT d.day,
coalesce(sum(t.qta)) qta
FROM d
LEFT JOIN table1 t
ON t.id = 35
AND d.day BETWEEN t.date1
AND t.date2
GROUP BY d.day
ORDER BY d.day;