t-sql looping through rows in a table

Use cursors

>Note: I'm not to worried about efficiency at this point.
note: you should ALWAYS worry about efficiency...

>I have a need to loop through the rows of a temp table, compare certain columns, and perform an update on same columns.
can you explain a bit in details?
usually, a simple UPDATE with CASE syntax can do the job.

declare @col varchar(10),
 @co2 varchar(10)
declare c cursor for select column1, column2 from yourtable

open c
fetch next from c into into @co1 , @co2

while @@fetch_status = 0
begin
if @co1= 1
update yourtable set @col=3 where @co2= @co2
fetch next from c into into @co1 , @co2
end
close c
deallocate c
go

Hi rutledgj,

do you realy need to loop through all the records this is very slow and totaly contary to set based processing

rather tell us your table structure and what you are tryong to update and we will try help you avoid cursors

rutledgj,
> Note: I'm not to worried about efficiency at this point
but soon you will be :)


HTH

R.

Suppose you have all the conditions linked
Update #temp set col1 = val1, col2 = val2, col3 = val3
     where col1 in (select col1 from #temp where col1 = someval1) and
               col2 in (select col1 from #temp where col2 = someval2) and
               col2 in (select col1 from #temp where col3 = someval3)

If the columns you have are independant then break up the above query as
Update #temp set col1 = val1
     where col1 in (select col1 from #temp where col1 = someval)

and repeat it for the number of columns

OK.
My temp table is like this:

act_id, prop_id, act_date, prop_date, act_name, prop_name, etc

In the above table, act_id WILL always = prop_id but prop_date may not be the same as act_date. So I need to update the prop_date with the act_date.


Rut