SQL table join based on aggregated/equal values - sql

Question

I have a very horrible two table dataset to hand that I need to create a join query for. Its best to show an example:
+------+---------+-----------+--+
| Time | Sent | Received | |
+------+---------+-----------+--+
| 1 | 100 | NULL | |
| 2 | NULL | 100 | |
| 3 | 50 | NULL | |
| 4 | NULL | 40 | |
| 5 | NULL | 10 | |
| 6 | 400 | 200 | |
| 7 | 100 | 200 | |
| 8 | NULL | 100 | |
| 9 | 500 | 500 | |
+------+---------+-----------+--+
Assuming 'time' above is in hours - 'Sent' shows the number of items sent in that hour, and 'Received' shows the number received. The problem being that they likely will not arrive in the same hour they were sent (though they can).
I need to match the received against the appropriate sent to find the time the received item was Sent.
Using the above:
Received 100 at time 2 is obviously the items sent from hour 1, so
that would be assigned to hour 1.
50 Sent in time 3 arrived in two batches (40 and 10 in time 4/5 respectively). So received 40/10 should be lumped into the time 3 category
Received in 6/7 (each for 200) correspond to the 400 order in time 6 (note that half the order was received in the same hour, this can happen)
Also in time 7 a new order was sent which corresponds to received for time 8
Also in time 9 an order of 500 was sent and received in the same hour.
Below is an example of what the output would look like (Note that there are other values associated with each 'Received' row but they are orthogonal to the task and will just be summed to provide meaning)
+------+----------+
| Time | Received |
+------+----------+
| 1 | 100 |
| 3 | 50 |
| 6 | 400 |
| 7 | 100 |
| 8 | 100 |
| 9 | 500 |
+------+----------+
I have been trying to rack my head around this for a while. If I could do this outside of sql I would have some function that loops through the value for each 'Sent' incrementally through time and loop that through 'Received' until the values match then assign those Received values to the Time index, then delete both the sent and received from the array (or note where the loop got to and continue from there)
Unfortunately the project doesnt allow the that scope - This must be done as much in SQL as possible. I am really at a loss and hoping there is some SQL functionality I have overlooked. Any help is much appreciated

Answer 1

If this is in SQL Server, you can use a WHILE loop. Look at the documentation. So, your project might look something like this:
CREATE TABLE #temp ([Time] int, [Received] int)
DECLARE #i int = 1
DECLARE #value int = 0
WHILE #i <= 9
BEGIN
#value = SELECT [Received] FROM [table] WHERE [Time] = #i
--Your logic here
INSERT INTO #temp ...
END
SELECT * FROM #temp
DROP TABLE #temp