Difference in Triggers for populating Primary Keys

Question

version: Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production
Hello,
I have a question regarding using Triggers to populate Primary Keys.
One Trigger I've seen written like this:
CREATE OR REPLACE TRIGGER "BI_MY_TABLE"
    BEFORE INSERT
    ON my_table
    FOR EACH ROW
BEGIN
    IF :new."MY_TABLE_ID" IS NULL THEN
        SELECT "MY_TABLE_SEQ".NEXTVAL INTO :new."MY_TABLE_ID" FROM DUAL;
    END IF;
END;
And another Trigger written like this:
CREATE OR REPLACE TRIGGER "BI_MY_TABLE"
    BEFORE INSERT
    ON my_table
    FOR EACH ROW
BEGIN
   FOR c1 IN (SELECT my_table_seq.NEXTVAL next_val FROM dual)
   LOOP
      :new.my_table_id := c1.next_val;
   END LOOP;
END;
The first version is using an IF statement and the second version is using a LOOP.
My questions are:
Is there really any difference in the technique used to create the Trigger?
Is there any "benefit" in using one technique over the other?
Is there a more preferred method/technique to populating Primary Key columns?
Thanks,
Joe

Answer 1

Hi, Joe,
The second way (using a loop) is just silly.  Don't use a loop.
Starting in Oracle 11, you don't need to use dual at all; you can do it like this:
CREATE OR REPLACE TRIGGER BI_MY_TABLE 
    BEFORE INSERT 
    ON my_table  
    FOR EACH ROW  
BEGIN 
    IF  :new.MY_TABLE_ID  IS NULL  THEN 
        :new.MY_TABLE_ID  := MY_TABLE_SEQ.NEXTVAL;  
    END IF;  
END  BI_MY_TABLE;