Re: [HACKERS] libpq: why we need to fetch all rows?

Поиск
Список
Период
Сортировка
От Igor
Тема Re: [HACKERS] libpq: why we need to fetch all rows?
Дата
Msg-id 036befc3b9949a30e6a2f6ec22074780
обсуждение исходный текст
Ответ на [HACKERS] libpq: why we need to fetch all rows?  (Alexander Demenshin <aldem@techie.com>)
Список pgsql-hackers
Does the portal store all the rows at the backend and feed them one by one
to the client, or does it go out to disk to get new rows?

=+=------------------------/\---------------------------------=+=
       Igor Natanzon      |**|   E-mail: igor@sba.miami.edu
=+=------------------------\/---------------------------------=+=

On Sun, 29 Jun 1997, Bruce Momjian wrote:

> >
> > On Sun, Jun 29, 1997 at 01:44:36PM -0400, Bruce Momjian wrote:
> >
> > > Because it is not standard SQL.  No database I know of returns multiple
> > > rows with out a 'declare' statement and a cursor.
>
> Actually, standard SQL requires a cursor so you can name the cursor to
> fetch from, but PostgreSQL requires it to be in a transaction, which is
> not standard.
>
> >
> >   Not so good... But then, why not to modify backend and libpq
> >   in a way, which will allow to fetch rows one-by-one?
>
> OK, can you run a test?  If you start a transaction, declare a cursor,
> and do a fetch over and over again, does it return one line at a time
> from the server.  I think it does.
>
> Now, it will not return from the first fetch until the query has
> finished executing, but I don't think it will send all the results to
> the client, only one at a time.
>
> --
> Bruce Momjian
> maillist@candle.pha.pa.us
>

------------------------------

End of hackers-digest V1 #404
*****************************

В списке pgsql-hackers по дате отправления: