Question

After originally composing several reports by hand as part of an MVC 3 (EF4) project for a customer, I have decided it is far easier to compose reports in Microsoft reporting services (SSRS 2008 on SQL server 2008). For most of my reports this has proven to be a breeze, not even having to write SQL procs for most of my reports (semantic query in BIDS based on model).

However one report remains daunting due to the nature of its input parameters and originally being computed in a function returning a list, not a LINQ query. The goal is to return all people in the people table who are less than x percent busy on any day in the range between date a and date b (variables input by report requester). This busy level information only exists in the Job table which has information on which person it is assigned to, the start and end dates of the job, and its task level. As a result, my original C# function does pretty much the following (psuedo-code):

given startDateRange a, stopDateRange b, busyLevel x
{
for each person in people:
   List returnPeople = new List()
   create array consisting of days from a to b
   for each job in jobs:
      for each intersecting date of DateRange and job dateRange:
         add task level of job to array
   for each day in array:
      if array[day] < x
         returnPeople.Add(person)
         break;
return returnPeople
}

Table structure for relevant pieces:

People Table:
-PersonID (PK)
-other stuff...

Jobs Table:
-JobID (PK)
-StartDate
-StopDate
-Allocation
-AssignedPersonID (FK to People Table)

As you can see, a job is on someones schedule if that job’s AssignedPersonID is = that someone

So… since I’m now working in SSRS 2008, I am looking for guidance on how to get these same results in SSRS from the same input parameters. Some trick to semantic queries that I am currently unaware of? SQL stored proc with temp table or table variable? Some sort of data processing extension? Any information on what the best approach to this type of problem is would be much appreciated.

Let me know if additional information/code is required.

EDIT: Currently working on a solution using cursors and temp tables that pretty much copies the structure of my original function. This is DEFINITELY not ideal however due to the complexity and inefficient nature of cursors, so advice/alternatives would be helpful.

EDIT: Found a solution, definitely not the most efficient. Will change accepted answer if someone has a significantly faster method.

Answer 1

Sorry that it took a while to get back to you. I don’t think there’s a way to get around having to loop through the days with a WHILE loop, since that gives you individual days needed to compare to your users’ input range. I think this query should help somewhat for speed though, since it doesn’t use any cursors. I added the logic within the WHILE loop, so you only loop one time through, and you have your temp table as a result. The variable names might be a little different, and you might switch a <= to < or something, but the main logic is there.

The query includes the start date and end date, and is set to include anyone with an allocation less than or equal to the users’ input. I hope it helps you:

DECLARE @startDate datetime, @endDate datetime, @maxAllocation float, @dayCounter datetime

DECLARE @PeopleList TABLE (PersonId int, PersonName varchar(50))

SET @dayCounter = @startDate
WHILE @dayCounter <= @endDate
BEGIN
    INSERT INTO @PeopleList SELECT P.PersonId, P.PersonName FROM People P INNER JOIN Jobs J ON P.PersonId = J.AssignedPersonId
       WHERE J.StartDate <= @dayCounter AND J.StopDate >= @dayCounter
       GROUP BY P.PersonId, P.PersonName
       HAVING SUM(J.Allocation) <= @maxAllocation
    SET @dayCounter = DATEADD(DAY, 1, @dayCounter)
END

SELECT DISTINCT PersonId, PersonName FROM @PeopleList