My scenario is I have to populate data into a table by performing join on two tables A and B on the basis of column quoteid and compid
Table A
------------------------------
quoteid compid ................
10004 1
10004 1
10004 1
10004 22
10004 22
10004 22
Table B
------------------------------------
quoteid compid quartercode cost
10004 1 1q10 14
10004 1 2q09 10
10004 1 3q10 12
10004 22 4q12 32
10004 22 3q11 30
10004 22 2q11 43
Now, the result of select query should be like
quoteid compid quarter1cost quarter2cost quarter3cost
10004 1 10 14 12
10004 22 43 30 32
The concept to select cost for quarter is with quartercode which is combination of quarter in year(1st,2nd...) and last two digits of year. So, oldest quarter will with quarter1 , second oldest will be quarter2 and most recent will be quarter3. Here, there cost will be available for only 3 recent quarters due to join conditions. For example, here for quoteid 10004 and compid 1, quarter1 will be 2q09, quarter2 will be 1q10 and quarter3 will be 3q10 and hence the cost.
I am trying to do it with cursor. But since I am new so unable to get the desired result.
Table A seems to have nothing to do with your result.
The basic idea is to use row_number() and conditional aggregation. This is complicated because you the quarter identifier is stored backwards, so it doesn't sort correctly. But you can still do it:
select quoteid, compid,
max(case when seqnum = 1 then cost end) as cost_q1,
max(case when seqnum = 2 then cost end) as cost_q2,
max(case when seqnum = 3 then cost end) as cost_q3
from (select b.*,
row_number() over (partition by quoteid, compid
order by substr(quartercode, 1, 1), substr(quartercode, 3, 2)
) as seqnum
from b
) b
group by quoteid, compid