On Tue, Sep 27, 2022 at 03:38:49PM -0700, Jacob Champion wrote:
> On 9/26/22 06:29, Drouvot, Bertrand wrote:
> Since there are only internal clients to the API, I'd argue this makes
> more sense as an Assert(authn_id != NULL), but I don't think it's a
> dealbreaker.
Using an assert() looks like a good idea from here. If this is called
with a NULL authn, this could reflect a problem in the authentication
logic.
>> As far the assertion failure mentioned by Michael when moving the
>> SVFOP_SYSTEM_USER from NAMEOID to TEXTOID: V4 is assuming that it is
>> safe to force the collation to C_COLLATION_OID for SQLValueFunction
>> having a TEXT type, but I would be happy to also hear your thoughts
>> about it.
>
> Unfortunately I don't have much to add here; I don't know enough about
> the underlying problems.
I have been looking at that, and after putting my hands on that this
comes down to the facility introduced in 40c24bf. So, I think that
we'd better use COERCE_SQL_SYNTAX so as there is no need to worry
about the shortcuts this patch is trying to use with the collation
setup. And there are a few tests for get_func_sql_syntax() in
create_view.sql. Note that this makes the patch slightly shorter, and
simpler.
The docs still mentioned "name", and not "text".
This brings in a second point. 40c24bf has refrained from removing
SQLValueFunction, but based the experience on this thread I see a
pretty good argument in doing the jump once and for all. This
deserves a separate discussion, though. I'll do that and create a new
thread.
--
Michael
