Joining 2 related records using PL SQL in Apex - Problems when there are more than 2 related records?

Question

Hi
I am combining 2 related records of legacy data together that make up a marriage record.  I am doing this in APEX using a before header process using the following code below which works well when there are only 2 related records which joins the bride and groom record together on screen in apex.  I have appended a field called principle which is set to 'Y' for the groom and 'N' for the bride to this legacy data
However there are lots of records where in some instances there are 3, 4 , 5, 6 or even 1 record which causes the PL/SQL in APEX to not return the correct data.  The difference in these related columns is that the name of the bride or groom could be different but it is the same person, its just that from the old system if a person had another name or was formally known as they would create another duplicate record for the marriage with the different name, but the book and entry number is the same as this is unique for each couple who get married.
How can I adapt the script below so that if there are more than 2 records that match the entry and book values then it will display a message or is there a better possible work around?  Cleaning the data would be not an option as there are thousands of rows of where these occurrences occur
declare 
     cursor c_mar_principle(b_entry in number, b_book in varchar2) 
     is 
          select DISTINCT  id, forename, surname, marriagedate, entry, book,  formername, principle
          from   MARRIAGES mar 
          where  mar.entry   = b_entry
          and    mar.book = b_book
          order by principle desc, id asc; 
     rec c_mar_principle%rowtype;
begin 
open c_mar_principle(:p16_entry,:p16_book)  ;
fetch c_mar_principle into rec;
:P16_SURNAME_GROOM   := rec.surname; 
:P16_FORNAME_GROOM   := rec.forename;
               :P16_ENTRY := rec.entry; 
               :P16_BOOK :=rec.book;
:P16_FORMERNAME :=rec.formername;
:P16_MARRIAGEDATE :=rec.marriagedate;
:P16_GROOMID  := rec.id;
fetch c_mar_principle into rec;
:P16_SURNAME_BRIDE   := rec.surname; 
:P16_FORNAME_BRIDE   := rec.forename;
               :P16_ENTRY := rec.entry; 
               :P16_BOOK :=rec.book;
:P16_FORMERNAME :=rec.formername;
:P16_MARRIAGEDATE :=rec.marriagedate;
:P16_BRIDEID  := rec.id;
close c_mar_principle;
end;

Answer 1

rambo81 wrote:
True but that answer is not really helping this situation either?
It's indisputably true, which is more than can be said for the results of querying this data.
The data is from an old legacy flat file database that has been exported into a relational database.
It should have been normalized at the time it was imported.
Without having to redesign the data model what options do I have in changing the PL/SQL to cater for multiple occurances
In my professional opinion, none. The actual problem is the data model, so that's what should be changed.