Hi all,
o We have a table t1 which has a primary key id and other fields.
(say, 10 rows with ids 1 to 10).
o We need to group rows (unordered, and arbritrary size) in t1 and
assign a unique id to each group. (e.g.: {1, 3, 4} is group 1, {1, 5}
is group 2 etc.)
o A group shouldn't change
I understand that an UPDATE is done using "old" values of the variables
involved so that we can write
T(n + 1) = U(T(n))
where T(n) and T(n + 1) are the values of the table before and after
update U.
Thanks for all who pointed this out.
Just as a matter of interest, I want to point out that t
uld have
> been because it alone met the before March 31 and update flag = H criteria.
> But it is a start for me thanks.
>
> And your second sql statement just didn't work. It said 'ERROR: parser:
> parse error at or near "select"'
>
> Steve
>
>
First, posting_date in journal can _NOT_ be of type char(4)! I guess it is
a "date".
Try
update journal set gl_update_flag='Y' from distrib where
journal.gl_update_flag = 'H' and journal.posting_date <= '2001-03-31' and
(journal.objectid = distrib.distrib_objectid or journal.objectid =
distrib.
If I have a query
UPDATE tablename SET c1 = 10 - c2, c2 = 4 where ...
will the two assignments be evaluated from left to right?
Thanks in advance.
Anuradha
---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an app
