There is a limitation: the number of the 'columns' of the matrix is fixed at declaration time (in DECLARE section) and you need to manually initialize all the elements in the first 'row' of the matrix.
The number of rows is unlimited and can be sized at runtime.
Here is the code that can help you.
If someone know how to manage an NxN array without limitations, please replay to this thread.
I'm using PostgreSQL 7.4.1 on RH7.3
Cheers Sergio
CREATE OR REPLACE FUNCTION "testarray" (integer) RETURNS SETOF integer AS' DECLARE n alias for $1; -- number of rows is passed as argument i INTEGER; j integer; k INTEGER := 3; -- matrix columns number b integer[] := array[0,0,0]; -- need it to initialize the matrix!! a integer[][] := array[[0,0,0]]; -- need it to initialize the matrix!! begin
for i in 1..n loop -- the i loop can start obviously from 2 (the first row is already present...) but for our purpose here we use 1
a := array_cat(a,b);
for j in 1..k loop
a[i][j] := i*j;
end loop;
end loop;
for i in 1..n loop
return next null; return next i; -- need it to format in some way the output :) return next null;
for j in 1..k loop return next a[i][j]; end loop;
end loop; return; end 'LANGUAGE 'plpgsql';
here is the output:
=> select * from testarray(8); testarray -----------
1
1 2 3
2
2 4 6
3
3 6 9
4
4 8 12
5
5 10 15
6
6 12 18
7
7 14 21
8
8 16 24 (48 rows)
-- --------------------------------------------------------------------- Sergio Fantinel EGEE Project --------------------------------------------------------------------- INFN - Lab. Naz. di Legnaro phone: +39 049 8068 489 viale dell'Università n. 2, 35020 Legnaro (PD) ITALY [EMAIL PROTECTED] ---------------------------------------------------------------------
---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match