Sean Lange (8/6/2013)
Lowell (8/6/2013)
no, you should say "Because @@rowcount is affected by any commands in the session, it's my practice to re-select from the table in case there was a trigger that might returna different @@rowcount i might not expect."just blame force of habit to resolve rare edge cases like that.
I read this post and thought the same thing. So I decided I should test my theory. It seems that @@rowcount returns the correct number at least in this simple test.
create table RowCountTest
(
id int
)
go
create trigger RowCountTestTrigger on RowCountTest after insert as
select top 1000 *
into #Something
from Tally
go
insert RowCountTest
select 1
select @@ROWCOUNT
If this was a QOTD I would have answered that the value of @@rowcount would be 1000. Interesting. Maybe I will have to do a little more digging and submit as a qotd soon.
An INSTEAD OF trigger can definitely result in @@rowcount being different to SELECT count(x) FROM ...
MM
select geometry::STGeomFromWKB(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