← Back to team overview

drizzle-discuss team mailing list archive

Re: about replacing loadfile with filesystem storage engine


On Sat, Jul 31, 2010 at 2:34 AM, ZQ <ziminq@xxxxxxxxx> wrote:
> Thanks for all the advice. I'm not familiar with PG's COPY, so I would take
> a look.
> anyway, I will start small, and make it more functional bit by bit :-) .

I have been thinking about this a bit. I have never liked LOAD DATA
INFILE. My big gripe is that as far as I know there is no way to get
it to work like a normal *nix program in terms out stdin. The same
sort of thought applies to SELECT INTO OUTFILE, I want to use stdout.
(can one emulate stdin/stdout usage by using a FIFO?).

I wonder if there would be some way for two servers to be using the
same file as a table through the filesystem storage engine in order to
achieve something like what I would like with stdin/stdout.

I wonder if something like the pgdump "custom" format might be
desirable in the future. PG uses their version of SELECT INTO OUTFILE
(COPY) to create a compressed single file of table(s) (or entire
schemas) archive including all the DDL. Individual schems/tables can
be restored through the use of the pgrestore utility. Are any
decisions being made now that would make it easier have something

Rob Wultsch