Re: libpq patch for pqtypes hook api and PGresult creation

>
> Should the hook only be called when the conn or result was
> successfull or should the hooks be called for failed
> connections/commands as well?
>

ISTM that the hooks should be called on success and error (doesn't
include cases where a NULL conn or result is returned). I think this
makes things more useful. If the hooker (pun intended) isn't interested
in error results or conns, it can easily ignore them.

connCreate: The best solution I found was to hook into PQconnectPoll.
This required making the current PQconnectPoll a static named
connectPoll and making PQconnectPoll a wrapper to it. The wrapper just
calls connectPoll and checks the status for hook points.

resultCreate: PQgetResult seemed like the best place. I only notify the
hooks if the resultStatus is NOT copyin or copyout.

I diff'd fe-connect.c and fe-exec.c against cvs which shows these changes.

--
Andrew Chernow
eSilo, LLC
every bit counts
http://www.esilo.com/