Re: Cache lookup errors with functions manipulation object addresses

> On Tue, Jul 2, 2019 at 9:28 AM Michael Paquier <michael(at)paquier(dot)xyz> wrote:
>
> On Thu, Feb 21, 2019 at 04:40:13PM +0900, Michael Paquier wrote:
> > Rebased version fixing some conflicts with HEAD.
>
> And rebased version for this stuff on HEAD (66c5bd3), giving visibly
> v16.

Thanks for the patch! I couldn't check it in action, since looks like it
doesn't apply anymore [1] (although after a quick check I'm not entirely sure
why). Nevertheless I have a few short commentaries:

v16-0001-Add-flag-to-format_type_extended-to-enforce-NULL.patch

- if (type_oid == InvalidOid && (flags & FORMAT_TYPE_ALLOW_INVALID) != 0)
- return pstrdup("-");
+ if (type_oid == InvalidOid)
+ {
+ if ((flags & FORMAT_TYPE_FORCE_NULL) != 0)
+ return NULL;
+ else if ((flags & FORMAT_TYPE_ALLOW_INVALID) != 0)
+ return pstrdup("-");
+ }

Here and in format_operator_extened commentary says

* Returns a palloc'd string.

but now it's possible to return NULL, so I guess comments need to be adjusted,
right?

v16-0003-Eliminate-user-visible-cache-lookup-errors-for-o.patch

- appendStringInfo(&buffer, _("operator %s"),
- format_operator(object->objectId));
- break;
+ {
+ char *oprname = format_operator_extended(object->objectId,
+ FORMAT_PROC_FORCE_NULL);

Shouldn't it be FORMAT_OPERATOR_FORCE_NULL here?

I'll continue reviewing the last patch in more details.

[1]: http://cfbot.cputube.org/patch_24_1947.log