On Thu, 2006-07-06 at 15:03, Jim Cser wrote:
> >The problem isn't permission for copy, it permission to access files.
>
> >only the super user can access the file system through pgsql, because
> of
> >security concerns.
>
> >copy can be run by anybody. copy from/to a FILE can only be done by
> the
> >super user.
>
> >Have you tried queueing up copy from stdin and a stream of data and
> \.
> >in ODBC? I'm pretty sure that'll work, but I've never tried it.
>
> I imagine that your solution would work, but the larger issue is that I
> am sending the SQL from an R program, and not using the psql console. I
> could launch it the console programmatically and send copy commands to
> it, but that's just adding more moving parts. Having everyone be a
> superuser is good enough for now, thanks anyway.
You don't need the psql console, that's what I've been trying to say.
I do this stuff routinely in php and other languages.
Note that if R is generating this data from postgresql to begin with,
you might be better off with pl/R
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
