Procedure to take inpot data set and return result set

Question

Hi all,
I have a situation where there will be one "standard" set of data (source_data below) and I will need to get information on results for certain groups of clients (client_data being an example). As straight SQL this would be very easy (see below) -- real world problem is a little more complicated. However, what I would like to do is set up a procedure so that I can pass it my variable client data and it will spit back out a data set the same as the output from the given SQL.
A pointer in the right direction would be appreciated. If I could "pass the client data" as a string containing a SQL query, that would be even better, e.g.
GetResults('select client_id,min(whatever_date) from some_client_data where .....',MyOutputRefCursor?)
create table source_data
(client_id integer,
tdate date,
amount number(6,2));
create table client_data
(client_id integer,
critical_date date);
insert into source_data values(1,to_date('20090104','yyyymmdd'),1000);
insert into source_data values(1,to_date('20100104','yyyymmdd'),2000);
insert into source_data values(1,to_date('20110104','yyyymmdd'),3000);
insert into source_data values(1,to_date('20120104','yyyymmdd'),4000);
insert into source_data values(2,to_date('20090104','yyyymmdd'),5000);
insert into source_data values(2,to_date('20090604','yyyymmdd'),1000);
insert into source_data values(2,to_date('20100104','yyyymmdd'),2000);
insert into source_data values(3,to_date('20091004','yyyymmdd'),3000);
insert into source_data values(3,to_date('20091104','yyyymmdd'),4000);
insert into source_data values(4,to_date('20090104','yyyymmdd'),5000);
insert into source_data values(4,to_date('20090604','yyyymmdd'),2000);
insert into client_data values(1,to_date('20110104','yyyymmdd'));
insert into client_data values(2,to_date('20090604','yyyymmdd'));
select c.client_id,
  sum(CASE WHEN tdate < critical_date then amount else null end) used_before,
  sum(CASE WHEN tdate >= critical_date then amount else null end) used_after
from source_data s
inner join client_data c on s.client_id = c.client_id
GROUP BY c.client_id;Thanks,
Jon

Answer 1

JonWat wrote:
I have a situation where there will be one "standard" set of data (source_data below) and I will need to get information on results for certain groups of clients > (client_data being an example). As straight SQL this would be very easy (see below) -- real world problem is a little more complicated. Can you give a more accurate representation of the real world problem?
A pointer in the right direction would be appreciated. If I could "pass the client data" as a string containing a SQL query, that would be even better, e.g.Passing a query as as string for execution is rarely the correct approach. This approach won't scale and most likely will result in maintenance and security problems.
If you can update your post to have a more accurate representation of the problem you are trying to solve you'll probably get an acceptable answer. Other than that all I can suggest is that you need to rethink the approach a bit. For example, if you know you need to filter out a set of data to join to your SOURCE_DATA table, why couldn't you do that inside the procedure?
SELECT c.client_id
     , SUM
       ( CASE
         WHEN tdate < critical_date
         THEN amount
         END
       ) AS used_before
    ,  SUM
       ( CASE
         WHEN tdate >= critical_date
         THEN amount
         END
       ) AS used_after
FROM   source_data s
JOIN   ( SELECT client_id
              , MIN(<date column>) AS critical_date
         FROM   client_data
         WHERE  <where clause>
         GROUP BY client_id
       ) c on s.client_id = c.client_id
GROUP BY c.client_id;