T-SQL - Merge two tables without create temp table - sql

Question

The prefer result i want I have those 2 T-SQL queries:
SELECT
CONVERT(NVARCHAR, AL.FECHA, 103) AS [DATE],
CONVERT(NVARCHAR, AL.HORA, 108) AS [HOUR],
SUM(CAST(AL.TOTALNETO AS money)) AS [AMOUNT]
FROM
ALBVENTACAB AL
WHERE
CONVERT(NVARCHAR, AL.FECHA, 103) = '10/09/2020'
-- AND CONVERT(nvarchar, GETDATE(), 103) AND
AND CONVERT(nvarchar, AL.HORA, 108) BETWEEN (SELECT CONVERT(nvarchar, AL.HORA, 108)) AND CONVERT(nvarchar, GETDATE(), 108)
GROUP BY
AL.FECHA, AL.HORA
ORDER BY
AL.FECHA
SELECT
CONVERT(NVARCHAR, AL.FECHA, 103) AS [DATE],
CONVERT(NVARCHAR, AL.HORA, 108) AS [HOUR],
SUM(CAST(AL.TOTALNETO AS money)) AS [AMOUNT]
FROM
ALBVENTACAB AL
WHERE
FECHA = CAST(DATEADD(DD, -7, GETDATE()) AS date)
AND CONVERT(nvarchar, AL.HORA, 108) BETWEEN (SELECT CONVERT(nvarchar, AL.HORA, 108)) AND CONVERT(nvarchar, GETDATE(), 108)
GROUP BY
AL.FECHA, AL.HORA
ORDER BY
AL.FECHA
Individually, their results are correct. I want to create a result as shown in the screenshot.
I tried to create a CTE, but that didn't work. What is the proper way?
Thank you!

Answer 1

First, the query I think you're looking for is something like this. The Friday date (#fecha) was removed to a variable of type DATE. Instead of 2 query, 1 for the week, and 1 for the particular Friday, it's now 1 query. The friday sales are broken out using conditional aggregation into a new column called [FECHA_AMOUNT]. Note: this only works if the variable #fecha falls within the last 7 days.
declare #fecha date='20201009';
SELECT
CONVERT(NVARCHAR, AL.FECHA, 103) AS [DATE],
CONVERT(NVARCHAR, AL.HORA, 108) AS [HOUR],
SUM(CAST(AL.TOTALNETO AS money)) AS [AMOUNT],
SUM(case when cast(AL.FECHA as date)=#fecha then CAST(AL.TOTALNETO AS money)
else cast(0 as money) end) AS [FECHA_AMOUNT]
FROM
ALBVENTACAB AL
WHERE
FECHA = CAST(DATEADD(DD, -7, GETDATE()) AS date)
AND CONVERT(nvarchar, AL.HORA, 108) BETWEEN (SELECT CONVERT(nvarchar, AL.HORA, 108))
AND CONVERT(nvarchar, GETDATE(), 108)
GROUP BY
AL.FECHA, AL.HORA
ORDER BY
AL.FECHA;
Finally, as note in the comments. When converting it's always good to use lengths. Also, the money data type is not considered ideal. Fwiw