Hello all .. i just wrote this script in plpgsql. Basically it performs a crosstab query. The difference between this and the one you already have in contrib is that this one has no limitations on the arguments passed to it.
I don't understand what you think is different. Your example can be produced using contrib/tablefunc/crosstab like so:
regression=# select art, coalesce(s,0) as "S", coalesce(m,0) as "M", coalesce(l,0) as "L", coalesce(xl,0) as "XL", coalesce(xxl,0) as "XXL" from crosstab('select art, tgl, sum(qty) from art group by art, tgl order by 1', 'select ''L'' union all select ''M'' union all select ''S'' union all select ''XL'' union all select ''XXL''') as (art varchar, L int, M int, S int, XL int, XXL int);
art | S | M | L | XL | XXL
----------+----+----+----+----+-----
508301 | 0 | 0 | 10 | 38 | 0
5001001 | 0 | 12 | 0 | 0 | 25
45370104 | 10 | 0 | 0 | 0 | 0
(3 rows)
Or with nulls in place of zeros, like this:
regression=# select * from crosstab('select art, tgl, sum(qty) from art group by art, tgl order by 1', 'select ''S'' union all select ''M'' union all select ''L'' union all select ''XL'' union all select ''XXL''') as (art varchar, L int, M int, S int, XL int, XXL int);
art | l | m | s | xl | xxl ----------+----+----+----+----+----- 508301 | | | 10 | 38 | 5001001 | | 12 | | | 25 45370104 | 10 | | | | (3 rows)
Joe
---------------------------(end of broadcast)--------------------------- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
