Обсуждение: RE: [HACKERS] ecpg stuff

Or we should change the declare/open behaviour in that it ignores the
embedded SQL declare command and just sends the declare command when it
finds an open. But then standard may say that the actual way is correct.

Michael

--
Dr. Michael Meskes, Project-Manager    | topsystem Systemhaus GmbH
meskes@topsystem.de                    | Europark A2, Adenauerstr. 20
meskes@debian.org                      | 52146 Wuerselen
Go SF49ers! Go Rhein Fire!             | Tel: (+49) 2405/4670-44
Use Debian GNU/Linux!                  | Fax: (+49) 2405/4670-10

> -----Original Message-----
> From:    Vadim B. Mikheev [SMTP:vadim@sable.krasnoyarsk.su]
> Sent:    Wednesday, February 18, 1998 9:56 AM
> To:    Michael Meskes
> Cc:    Thomas G. Lockhart; PostgreSQL Hacker
> Subject:    Re: [HACKERS] ecpg stuff
>
> Michael Meskes wrote:
> >
> > - and check resp. tell me whether the cursor behaviour is correct.
> Currently
> > the declare statement is send to the backend via PQexec. The open
> statement
> > is ignored and the fetch is executed as fetch via PQexec. I think
> the data
> > shouldn't be processed before the cursor is opened. But I do not
> know
> > what PostgreSQL does with the declare command.
>
> DECLARE: parser + optimizer + ExecutorStart (initializes plan nodes:
> checks permissions, opens tables & indices).
> Is OPEN statement in standard ?
> If yes then we could call ExecutorStart() for the OPEN someday.
>
> Vadim

Meskes, Michael wrote:
>
> Or we should change the declare/open behaviour in that it ignores the
> embedded SQL declare command and just sends the declare command when it
> finds an open. But then standard may say that the actual way is correct.

It seems that OPEN is in standard and this is really bad that we havn't
OPEN statement: using DECLARE we could plan query once and then
re-use this plan many times - OPEN, FETCH-es, CLOSE; OPEN, ...
Also, we could use queries with parameters (like in SQL-funcs: $1, ...)
in DECLARE and then assign parameter' values in OPEN and execute
cursor for different values.
Isn't this what Oracle allow ?

Vadim

>
> Meskes, Michael wrote:
> >
> > Or we should change the declare/open behaviour in that it ignores the
> > embedded SQL declare command and just sends the declare command when it
> > finds an open. But then standard may say that the actual way is correct.
>
> It seems that OPEN is in standard and this is really bad that we havn't
> OPEN statement: using DECLARE we could plan query once and then
> re-use this plan many times - OPEN, FETCH-es, CLOSE; OPEN, ...
> Also, we could use queries with parameters (like in SQL-funcs: $1, ...)
> in DECLARE and then assign parameter' values in OPEN and execute
> cursor for different values.
> Isn't this what Oracle allow ?

Yes, I agree we should have it, and allow cursors outside transactions
with OPEN/CLOSE behavior.

Added to TODO.

--
Bruce Momjian
maillist@candle.pha.pa.us