simon wrote:
> "Gerard H. Pille" <ghp_at_skynet.be> ¦b¶l¥ó
>
>>Not if id can be 0 or 99, else it is OK. It is abnormal to select nothing
>
> or
>
>>multiple values in to a single, simple variable.
>
> Ok, assuming id cannot be 0 or 99, or I can just use another variable as the
> flag.
>
> What i am trying clarify is that if this way of using exception for testing
> row count returned by SELECT INTO a good practice? Because I am thinking
> exception should be used to handle abnormal cases, not an IF test of
> expected possibilities. However, given the fact that SELECT INTO will raise
> exception, it seems this technique have to be used.
>
>
I wouldn't do it that way, but I don't think your way is worse.
declare
v_number number;
cursor c_test (p_param varchar2) is
select id from table where varchar_col = p_param;
begin
v_number := 0;
for r_test in c_test( something_else) loop
v_number := v_number + 1;
if v_number > 1 then
v_number := 99;
exit;
end if;
end loop;
end;
/
Received on Fri Jun 10 2005 - 14:11:52 CDT
