Query to get data for current week+13

Question

Hi Friends,
I have two tables
BACKLOG_WEEK_AFTER_ATP (LE)
BACKLOG_ATP_GT_CW (RE)
** First I have to query whats the current week and year and it should come in this format ---- 2011-WK30
columns in table BACKLOG_WEEK_AFTER_ATP are:
ITEM_NUMBER      QUANTITY
1N5418                 20
1N5614                 30
1N5806SM               10
1N5811                  0
2PFF6                  60columns in table BACKLOG_ATP_GT_CW are:
ITEM_NUMBER     QUANTITY        YEAR_WEEK
1N5418                30        2011-WK30
1N5418                 5        2011-WK31
1N5614                30        2011-WK32
1N5806SM              30        2011-WK33
1N5811                20        2011-WK32
3EX473K1              20        2011-WK30My report should look like
ITEM_NUMBER    2011-WK30  2011-WK31  2011-WK32  2011-WK33  ...............till 13th week
1N5418                10         -5         -5         -5  ...............till 13t week
1N5614                30         30          0          0  ................till 13th week
1N5806SM              10         10         10         20  ................till 13th week
1N5811                 0          0         20         20  ................till 13th week
2PFF6                 60         60         60         60  ................till 13th week
3EX473K1              20         20         20         20  ................till 13th weekTo get this report i have these conditions to keep in mind.
1) If item_number not present in LE table and present in RE table then repeat what it is in RE table till 13th week
2) If item_number not present in RE table and present in LE table then repeat what it is in LE table till 13th week
3) If item_number present in LE and also present in RE table then do subtraction for RE - LE for that particular item_number till 13th week.
4) If item_number is there in LE table but not present in RE table for current_week+1(today week comes as 29th week) then repeat the same which is there in LE table. If item is found in RE table for (example 32th week) then subtract RE -LE for that particular item_number
Thanks in advance.
Regards

Answer 1

Hello,
If you don't need the PIVOT display, then this may help you :
with le as
(select '1N5418' item_number, 20 quantity from dual union all
select '1N5614' item_number, 30 quantity from dual union all
select '1N5806SM' item_number, 10 quantity from dual union all
select '1N5811' item_number, 0 quantity from dual union all
select '2PFF6' item_number, 60 quantity from dual ),
re as
(select '1N5418' item_number, 30 quantity, '2011-WK30' year_week from dual union all
select '1N5418' item_number, 5 quantity, '2011-WK31' year_week from dual union all
select '1N5614' item_number, 30 quantity, '2011-WK32' year_week from dual union all
select '1N5806SM' item_number, 30 quantity, '2011-WK33' year_week from dual union all
select '1N5811' item_number, 20 quantity, '2011-WK32' year_week from dual union all
select '3EX473K1' item_number, 20 quantity, '2011-WK30' year_week from dual
row_gen as (
select item_number, calc_year_week,
       row_number() over(partition by item_number order by calc_year_week) rn
from
    (select le.item_number from le union select item_number from re) item,
    (select to_char(level*7+sysdate,'YYYY-"WK"WW') calc_year_week from dual connect by level<=13) week)
select item_number, calc_year_week, calc_qty
from row_gen, le, re
where row_gen.item_number=le.item_number(+)
and row_gen.item_number=re.item_number(+)
and row_gen.calc_year_week=re.year_week(+)
model
partition by (row_gen.item_number)
dimension by (rn)
measures (calc_year_week, year_week, le.quantity le_qty,re.quantity re_qty,0 calc_qty )
rules  (
calc_qty[1]  =
    case when re_qty[cv()] is null then le_qty[cv()]
    when le_qty[cv()] is null then re_qty[cv()]
    else  re_qty[cv()]-le_qty[cv()]
end,     
calc_qty[rn>1] order by rn =
    case when re_qty[cv()] is null then calc_qty[cv()-1]
    else re_qty[cv()] - calc_qty[cv()-1]
    end           )
order by 1,2;I am not sure this is the simplest way to do it, but the results seem to match your example.
Regards,
Sylvie
Edited by: Troll35 on Jul 19, 2011 3:08 PM