Hrvoje Niksic <hniksic(a)iskon.hr> writes in xemacs-beta(a)xemacs.org:
SL Baur <steve(a)beopen.com> writes:
> (setq R (pq-exec P "This is a bad query;"))
> #<PGresult PGRES_FATAL_ERROR - >
This should signal a real error. We've got an error system;
let's
use it.
O.K.
The PostgreSQL error object should be encapsulated within the error
data, so that a handler can use it.
There is no real error object. There is an errno-like string area
set aside in the PGconn object for the most recent backend error
message.[1] It is not guaranteed to work, for various reasons.
Footnotes:
[1] Comments in the source indicate that the programmers are aware
that this is perhaps not the best mechanism in the world.