Horton

Re: Cache lookup errors with functions manipulation object addresses

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Alvaro Herrera <alvherre@2ndquadrant.com>
Cc: Daniel Gustafsson <daniel@yesql.se>, Dmitry Dolgov <9erthalion6@gmail.com>, Andrew Dunstan <andrew.dunstan@2ndquadrant.com>, PostgreSQL Developers <pgsql-hackers@lists.postgresql.org>
Date: 2020-07-06T07:45:43Z
Lists: pgsql-hackers

Attachments

On Fri, Jul 03, 2020 at 11:04:17AM -0400, Alvaro Herrera wrote:
> 0001 and 0002 look good to me.

Thanks for the review.

> I think 0003 could be a little more careful about indentation; some long
> lines are going to result after pgindent that might be better to handle
> in a different way before commit, e.g., here
> 
>> +    {
>> +        char *proname = format_procedure_extended(object->objectId,
>> +            FORMAT_PROC_FORCE_QUALIFY | FORMAT_PROC_INVALID_AS_NULL);

Yes, I was looking at that for a couple of hours and pgindent made
that a bit weird.  So I have changed the code to just use a separate
variable.  That makes things a bit cleaner.

While refreshing my mind with this code, I got surprised with the
choice of "routine" in getProcedureTypeDescription() when we need a
default object type name for an object not found, so I have switched
that to "procedure" to be more consistent.

I have also spent some time analyzing the coverage of the patch, and
did not find any obvious holes or any remaining missing_ok paths not
covered.  Some comments were also a bit weird after re-reading them,
so I tweaked a couple of places.

Attached is for now a rebased patch.  If there are any comments,
please feel free.  Daniel, Alvaro, does that look fine for you?  I am
letting this stuff aside for a couple of days for the moment.
--
Michael

Commits

  1. Eliminate cache lookup errors in SQL functions for object addresses

  2. Refactor routines for name lookups of procedures and operators

  3. Add new flag to format_type_extended() to get NULL for undefined type

  4. Introduce new extended routines for FDW and foreign server lookups

  5. Refactor routines for subscription and publication lookups

  6. get_relid_attribute_name is dead, long live get_attname

  7. Tweak parser so that there is a defined representation for datatypes