Re: PostgreSQL Gotchas

From: Chris Travers <chris(at)travelamericas(dot)com>
To: Martijn van Oosterhout <kleptog(at)svana(dot)org>
Cc: Chris Travers <chris(at)travelamericas(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Scott Marlowe <smarlowe(at)g2switchworks(dot)com>, Stephen Frost <sfrost(at)snowman(dot)net>, "Aly S(dot)P Dharshi" <aly(dot)dharshi(at)telus(dot)net>, "Gavin M(dot) Roy" <gmr(at)ehpg(dot)net>, pgsql-general(at)postgresql(dot)org
Subject: Re: PostgreSQL Gotchas
Date: 2005-10-13 21:16:00
Message-ID: 434ECE90.5040005@travelamericas.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Martijn van Oosterhout wrote:

>On Thu, Oct 13, 2005 at 12:21:14PM -0700, Chris Travers wrote:
>
>
>>>Since the end reward for all this work would be having to read CATALOGS
>>>WRITTEN IN ALL UPPER CASE, none of the key developers seem very
>>>interested ...
>>>
>>>
>>>
>>>
>
>I think you're slightly missing the point. Most people won't need to
>change any queries because unquoted identifiers are case-folded anyway.
>
>
Right. That is part of the case for at least putting this on the TODO list.

>Which way doesn't matter, as long as it's consistant. No the main
>objection would be this (see below). Does anyone really want this? I'll
>take lowercase anyday... (Note, it's not exact, some things aren't SQL
>identifiers so would be left alone. But idea is the same.)
>
>
Ok. If someone wanted it, we would see a patch already. I understand
this. Probably making AS optional would be a bigger priority because
such a patch exists already :-)

However.....

This is not only non-standard behavior but can also cause headaches in
certain porting circumstances. Adding it to the TODO list would at least
invite others to take this up *if* they (or their customers) required
it. The work I see required would minimally be:

1) Add a GUC variable to allow for changing the case folding behavior.
2) Add duplicate entries for builtin aggregates and functions in the
catalogs but with upper-case names
3) Adding a duplicate INFORMATION_SCHEMA with views off information_schema

Additionally if people depend on case folding for catalogs, you would
want to create a PG_CATALOG schema with views off pg_catalog. However,
this seems to be peripheral and not central to the issues of managing
the data given that identifiers are treated as already double-quoted.

I *could* see a case where a company that does portability work (like
EnterpriseDB, which specializes in Oracle portability) might find such a
feature as important. Adding the item to the TODO list would invite
patches where they are needed.

Best Wishes,
Chris Travers
Metatron Technology Consulting

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Tatsuo Ishii 2005-10-13 22:48:00 Re: On "multi-master"
Previous Message Chris Travers 2005-10-13 21:04:39 Re: PostgreSQL Gotchas