Eric,
You might check out this thread:
http://archives.postgresql.org/pgsql-general/2005-06/msg00527.php
Someone can almost certainly point to a better thread, but it is a
start....
Sean
On Jun 13, 2005, at 12:41 PM, Eric E wrote:
Hi David,
I ended up doing what you suggest, and it seems to work OK right now.
I'll definitely look for return_next in 8.1. On a related subject,
is there any interest in eliminating the requirement that types be
specified at runtime, either by the function returning the type, or in
the select statement? It would be helpful to me, but it's not a
dealkiller.
Anyway, thanks a lot for your help, it's much appreciated.
Eric
On Fri, Jun 10, 2005 at 07:24:49PM -0400, Eric E wrote:
Hi all,
I'm working on implementing a function in PL/PERL that will ready
many rows. As such I'd like to use return_next to keep memory usage
down. When I call return next, I get the following error message:
ERROR: error from Perl function: Can't call method "return_next" on
unblessed reference at (eval 18) line 25.
likewise if I use spi_return_next: ERROR: error from Perl function:
Can't call method "spi_return_next" on unblessed reference at (eval
16) line 25.
After some Google and list searching, I inferred that
spi_return_next may not be implemented? Is this correct? If so, is
there a workaround?
For 8.0, not exactly, although you can batch the entire rowset into an
array ref of hash references, then return that.
In the forthcoming 8.1 (CVS TIP), there is a return_next() method.
Cheers,
D
--
David Fetter david ( at ) fetter ( dot ) org http://fetter.org/
phone: +1 510 893 6100 mobile: +1 415 235 3778
Remember to vote!
---------------------------(end of
broadcast)---------------------------
TIP 1: subscribe and unsubscribe commands go to
majordomo@xxxxxxxxxxxxxx
---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster
