Can u help me on this query

Question

Hi all,
I am using 10G. With the following tables and data ( hope this help you to udnerstand my question)
create table case
( case_no number,
case_name varchar2(100),
pid number);
insert into case (case_no,case_name,pid)
values
(900,'Driving with loud music',1);
insert into case (case_no,case_name,pid)
values
(910,' Disturb public meeting',2);
insert into case (case_no,case_name,pid)
values
(920,'indecent conduct',3);
create table person
(pid number,
fname varchar2(25),
lname varchar2(25)
insert into person
(pid,fname,lname)
values
(1,'JAY','STINKY');
insert into person
(pid,fname,lname)
values
(2,'Jerry','Jasonx');
insert into person
(pid,fname,lname)
values
(3,'Wild','Manning');
create table penalty
(case_no number,
pid number,
fine number,
court_cost number
insert into penalty
(case_no,pid,fine,court_cost)
values
(900,1,2000,100);
insert into penalty
(case_no,pid,fine,court_cost)
values
(910,2,0,100);
create table community_Service
(case_no number,
pid number,
seq number,
service varchar2(250)
insert into community_service
(case_no,pid,seq,service)
values
(900,2,1,'Street Cleaning');
insert into community_service
(case_no,pid,seq,service)
values
(900,2,2,'Beach Cleaning');
insert into community_service
(case_no,pid,seq,service)
values
(920,3,1,'Volunteer work');
insert into community_service
(case_no,pid,seq,service)
values
(920,3,2,'Street Cleaning');I am looking for this answercase_no     case_name          fname     Lanme     fine     court_cost     seq     service
900     Driving with loud music JAY     STINKY     2000     100          1     street Cleaning
900     Driving with loud music                           2     Beach Cleaning
910     Disturb public meeting     Jerry     Jasonx     0     100
920     Indecent conduct     Wild     Manning                    1     Volunteer Work
920     Indecent conduct                              2     Street CleaningThe following is my unsuccessful codes( it has repetition on name and penalty when no entries involved)
select
A.case_no, A.case_name
,B.fname,B.lname
,C.fine,c.court_cost
,D.seq,D.service
from
case A,
Person B,
Penalty C,
community_Service D
where
A.pid=B.pid(+)
and A.case_no=C.case_no(+)
and A.pid = B.pid(+)
and A.case_no=D.case_no(+)
order by case_no,case_name,fname,lname,seq
Thanx
Zeng

Answer 1

Hi, Zeng,
zeng wrote:
Frank Kulash wrote:
Hi, Zeng,
So, you only want to display vertain columns (fname, lname, fine and court_cost) on the 1st row for a case; on the 2nd (and later) rows, you want to display NULL in those columns. Is that right?
Frank,
Always good to learn from your comments.
Nope. I am looking for a report that will display all persons involved in each case plus penalty and community services. If no penalty or community service then show the person's name ( revised desired output was in my reply to Nikolay)Are you saying that you do want the fname, lname and fine repeated on all lines after all? I assume you mean that, in addition to hiding those items, there's another issue, concenrning multuiple persons associated with the same case.
How do we know which fine and which service are related to which person? Is it the pid column in the penalty and community_service tables? If so, then we need to include pid in the join conditions. To display the fname and lname on the first row for every distinct case/person combination (instead of just for every cae), then we need to add pid to the analytic PARTITION BY clause, also.
WITH       got_r_num   AS
     SELECT       c.case_no
     ,        c.case_name
     ,       pr.fname
     ,       pr.lname
     ,       pn.fine
     ,       pn.court_cost
     ,       cs.seq
     ,       cs.service
     ,       ROW_NUMBER () OVER ( PARTITION BY  c.case_no
                                  ,          c.pid     -- ***** NEW *****
                                  ORDER BY          cs.seq
                         )           AS r_num
     FROM        cases               c
     ,       person          pr
     ,       penalty           pn
     ,       community_service     cs
     WHERE       c.pid          = pr.pid (+)
     AND       c.case_no     = pn.case_no (+)
     AND       c.pid          = pn.pid (+)               -- *****  NEW  *****
     AND       c.pid          = pr.pid (+)
     AND       c.case_no     = cs.case_no (+)
     AND       c.pid          = cs.pid (+)               -- *****  NEW  *****
SELECT       case_no
,        case_name
,       CASE WHEN r_num = 1 THEN fname      END     AS fname
,       CASE WHEN r_num = 1 THEN lname      END     AS lname
,       CASE WHEN r_num = 1 THEN fine           END     AS fine
,       CASE WHEN r_num = 1 THEN court_cost END     AS court_cost
,       seq
,       service
FROM        got_r_num
ORDER BY  case_no
,       fname
,       lname
,       seq
;The output I get from your sample data is:
CASE                                               COURT
_NO CASE_NAME                 FNAME LNAME    FINE _COST  SEQ SERVICE
900 Driving with loud music   JAY   STINKY   2000   100
910  Disturb public meeting   Jerry Jasonx      0   100
920 indecent conduct          Merry Lamb
920 indecent conduct          Wild  Manning                1 Volunteer work
920 indecent conduct                                       2 Street CleaningNote that there is no community_service associated with JAY STINKY, pid=1. In your sample data:
insert into community_service
(case_no,pid,seq,service)
values
(900,2,1,'Street Cleaning');
insert into community_service
(case_no,pid,seq,service)
values
(900,2,2,'Beach Cleaning');all the community_service for case_no=900 was assigned to pid=2, Jerry Jasonx, who is not related to case_no=900. I assume there's a mistake in the sample data. You can add foreign key constraints to the tables to prevent mistakes like that.