Avoid duplicate inserts using SQL statement?

Question

I have primary key on three columns and I want to write a sql statement to insert a row only if the matching three columns doesn't exist in the database.
This is the condition I wrote but doesn't work. Could someone suggest me a query which meets this criteria.
INSERT INTO JAVA_SERVICES (SERVICE_PAGE, SERVICE_TYPE, SERVICE_NAME,
  ROOT_PATH)
  VALUES ('AAAA', 'BBBB', 'CCCC', 'njdex') WHERE SERVICE_PAGE != 'AAAA' AND SERVICE_TYPE != 'BBBB' AND SERVICE_NAME != 'CCCC'Edited by: user626618 on Jun 12, 2009 2:39 PM

Answer 1

you need to provide us more details such as what table name and column names are you using. using an insert into statement an example would be like:
insert into java_services
(service_page, service_type, service_name, root_path)
select [column name], [column name], [column name]
  from [table name]
where ([column name], [column name], [column name]) not in (select SERVICE_PAGE, SERVICE_TYPE, SERVICE_NAME
                                                               from JAVA_SERVICES);
begin
  for i in (select [column name] col1, [column name] col2, [column name] col3
              from [table name]
             where ([column name], [column name], [column name]) not in (select SERVICE_PAGE, SERVICE_TYPE, SERVICE_NAME
                                                                from JAVA_SERVICES)) loop
    insert into java_services
      (service_page, service_type, service_name, root_path)
    values
      (i.col1, i.col2, i.col3);
  end loop;
end;
/