Answer = sane == 'fill buffer, don't overfill buffer' ? 'yes' : 'no';
:-)
Seriously, I will add comments when I submit the revised patch to
demonstrate the loop invariants etc. so nobody need get worried when
reading the code.
As for session end logging, I'm leaning in favor of a new
log_session_end or log_connection_end boolean, so that there is minimal
backwards compatibility impact. (Interestingly, that is actually the
only part of this stuff that is of any use to me right now).
cheers
andrew
Tom Lane wrote:
>"Andrew Dunstan" <andrew@dunslane.net> writes:
>
>
>>Of course, I could parse the string when setting up the buffer looking for
>>the number of expansions required. Instead I made the (I think reasonable)
>>assumption that each possible expansion would be wanted at most once.
>>
>>
>
>I think that's a reasonable assumption; the question was just about
>whether your code acts sanely if the assumption isn't true.
>
> regards, tom lane
>
>
>