DA Morgan wrote:
> EscVector wrote:
>> Anurag Varma wrote:
>>> EscVector wrote:
>>>> As demonstrated below, a table can be created with data type
>>>> "timestamp" as a column name.
>>>>
>>>> This causes a subsequent issue when creating triggers.
>>>>
>>>> Shouldn't the table create throw a "ORA-00904: : invalid identifier"
>>>> error?
>>>>
>>>> create sequence seq;
>>>>
>>>> create table test (
>>>> col1 number,
>>>> timestamp timestamp
>>>> )
>>>> /
>>>>
>>>> create or replace trigger trg
>>>> before insert on test
>>>> for each row
>>>> begin
>>>> select seq.nextval into :new.col1 from dual;
>>>> end;
>>>> /
>>>>
>>>> alter table test rename column timestamp to ts
>>>> /
>>>>
>>>> create or replace trigger trg
>>>> before insert on test
>>>> for each row
>>>> begin
>>>> select seq.nextval into :new.col1 from dual;
>>>> end;
>>>> /
>>> You are right .. it should be a bug. However, probably not easy
>>> to fix ...
>>>
>>> You might want to read Bug 4703738 where this is discussed.
>>>
>>> Anurag
>>
>> Quote from doc:
>> This issue is fixed in
>> * 9.2.0.8 (Server Patch Set)
>>
>> Well they didn't kill this bug. It is back again in 10.2.0.2 :)
>
> I'm not quite sure I can agree that something violating Oracle's advice,
> and generating an error, is a bug. At least not as long as the word
> TIMESTAMP is present in gv$reserved_words.
I wish my customers were as tolerant as you are....
--
Serge Rielau
DB2 Solutions Development
IBM Toronto Lab
WAIUG Conference
http://www.iiug.org/waiug/present/Forum2006/Forum2006.html
Received on Wed Jan 03 2007 - 19:28:35 CST
