How to fetch records from complex database? - sql

Question

I need to fetch multiple records and single records at a time with multiple tables from MSSQL. Tables from where I need to collect data is like below :
1. tblExpenseType :
--> ExpenseTypeId | ExpenseTypeName
--> 1 | OVERTIME EXP.
--> 2 | REPAIRING EXP.
--> 3 | LU EXP.
2. tblMemoDetails :
--> MemoID | MemoNumber | FkTruckId
--> 1011 | 1 | 5
--> 1012 | 2 | 6
--> 1013 | 3 | 5
--> 1014 | 4 | 7
3. tblMemoExpense :
--> FkMemoNumber | FkExpenseTypeId | Amount
--> 1 | 1 | 150
--> 1 | 3 | 225
--> 2 | 1 | 50
--> 2 | 2 | 100
--> 2 | 3 | 150
4. tblMemoTrips :
--> FkTripId | FkMemoNumber | TripAmount
--> 11 | 1 | 1000
--> 9 | 2 | 500
--> 3 | 2 | 100
--> 4 | 2 | 2000
I'm trying to fetch data using below logic but it's making me confusing:
with MemoList
As
(select
_jpMemo.MemoId as memoId
from TMS_JPMemo _jpMemo
where _jpMemo.FkTruckId = 5)
select
ML.memoId
--,ME.FkExpenseTypeId
,ME.Amount
,ET.ExpenseTypeName
from TMS_MemoExpense ME
join MemoList ML on ML.memoId = ME.FkMemoId
join TMS_ExpenseTypes ET on ET.ExpenseTypeId = ME.FkExpenseTypeId
I need result like if I select FkTruckId 5 then it will show me below result.
I need results like below for FkTruckId 5 :
MemoNumber | TripDetails | TripAmount | OVERTIME | REPAIRING | LU
1 | 11 | 1000 | 150 | -- | 225
| Total | 1000 | 150 | -- | 225
And If I select FkTruckId 6 then it would show me a result like :
MemoNumber | TripDetails | TripAmount | OVERTIME | REPAIRING | LU
2 | 9 | 500 | -- | -- | --
| 3 | 100 | -- | -- | --
| 4 | 2000 | 50 | 100 | 150
| Total | 2600 | 50 | 100 | 150
So it's making me confusing how to solve this and how to achieve this type of complex data from tables.

Answer 1

You are displaying two different things per memo:
the memo's trips with their trip amount
the memo's expenses
Apart from belonging to the same memo, the two are not related, i.e. there are no expenses per trip. So it's actually two different things you want to show. Make this two separate queries. You can glue them together with UNION ALL, as you obviously want them both in one query result. Something like:
with memo as
(
select memoid
from tblmemodetails
where fktruckid = 5
)
select memonumber, tripid, tripamount, overtime, repairing, lu
from
(
-- -------------------------------------------------------------------------------
-- The memos' trips. They have a trip amount but no further expenses of their own.
select
fkmemonumber as memonumber,
fktripid as tripid,
tripamount,
null as overtime,
null as repairing,
null as lu,
1 sortkey
from tblmemotrips
where fkmemonumber in (select memonumber from memo)
group by fkmemonumber, fktripid -- one row per memo and trip
-- -------------------------------------------------------------------------------
union all
-- -------------------------------------------------------------------------------
-- The memos' expenses. They are not reated to particular trips.
select
fkmemonumber as memonumber,
null as tripid,
null as tripamount,
sum(case when fkexpensetypeid = 1 then amount end) as overtime,
sum(case when fkexpensetypeid = 2 then amount end) as repairing,
sum(case when fkexpensetypeid = 3 then amount end) as lu,
2 as sortkey
from tblmemoexpense
where fkmemonumber in (select memonumber from memo)
group by fkmemonumber -- one row per memo
-- -------------------------------------------------------------------------------
) data
order by memonumber, sortkey;

Answer 2

You could use a series of JOINs to put the tables in relation, and then conditional aggregation to pivot the data. This implies that the number of columns to pivot is fixed (if more values are added to table tblExpenseType, the query would need to be adapted) :
SELECT
md.MemoNumber,
mt.FkTripId AS TripDetails,
mt.TripAmount,
MAX(CASE WHEN et.ExpenseTypeName = 'OVERTIME EXP.' THEN me.amount END) AS OVERTIME,
MAX(CASE WHEN et.ExpenseTypeName = 'REPARINING EXP.' THEN me.amount END) AS REPARING,
MAX(CASE WHEN et.ExpenseTypeName = 'LU EXP.' THEN me.amount END) AS LU
FROM
tblMemoDetails AS md
LEFT JOIN tblMemoExpense AS me ON me.FkMemoNumber = md.MemoNumber
LEFT JOIN tblMemoTrips AS mt ON mt.FkMemoNumber = md.MemoNumber
LEFT JOIN tblExpenseType AS et ON et.ExpenseTypeId = me.FkExpenseTypeId
GROUP BY
md.MemoNumber,
mt.FkTripId AS TripDetails,
mt.TripAmount

Answer 3

Conditional aggregation is the way to go. I think the query looks a little more like this:
SELECT md.TruckId, md.MemoNumber,
COALESCE(mt.FkTripId, 'Total') AS TripDetails, mt.TripAmount,
SUM(CASE WHEN et.ExpenseTypeName = 'OVERTIME EXP.' THEN me.amount ELSE 0 END) AS OVERTIME,
SUM(CASE WHEN et.ExpenseTypeName = 'REPARINING EXP.' THEN me.amount ELSE 0 END) AS REPARING,
SUM(CASE WHEN et.ExpenseTypeName = 'LU EXP.' THEN me.amount ELSE 0 END) AS LU
FROM tblMemoDetails md LEFT JOIN
tblMemoExpense me
ON me.FkMemoNumber = md.MemoNumber LEFT JOIN
tblMemoTrips mt
ON mt.FkMemoNumber = md.MemoNumber LEFT JOIN
tblExpenseType et
ON et.ExpenseTypeId = me.FkExpenseTypeId
WHERE md.TruckId = #TruckId
GROUP BY GROUPING SETS ( (md.TruckId, md.MemoNumber, mt.FkTripId, mt.TripAmount), () );
The GROUPING SETS generates the total.