Rule system

From: jwieck(at)debis(dot)com (Jan Wieck)
To: pgsql-hackers(at)postgreSQL(dot)org (PostgreSQL HACKERS)
Subject: Rule system
Date: 1998-08-12 08:55:11
Message-ID: m0z6Wga-000EBPC@orion.SAPserv.Hamburg.dsh.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

It's time,

there are a few small things and one big left in PL/pgSQL.
The small things are two new statements (easy) and the big
one is to be able to return tuples and sets.

I decided to implement the statements now and leave the
tuples for later (cause they require changes in the main
executor).

So next on my TODO will be the rule system.

But before starting on it I want to have a clear view onto
the target. What is the minimum of capabilities, the rule
system must have at least?

Here's a start of the list:

Retrieve-instead-retrieve rules on the relation level.
This is what builds a view (and works already - but order
by and distinct would be nice).

All other instead rules (insert, update, delete) on the
relation level, so views can behave like real tables
(yeah - these can be tricky).

What else must be there? I think everything on the instance
level is better done by triggers. And if we add
row-/statement-level triggers on SELECT, there would be no
reason left to have non-instead rules. Or am I missing
something?

Jan

--

#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me. #
#======================================== jwieck(at)debis(dot)com (Jan Wieck) #

Browse pgsql-hackers by date

  From Date Subject
Next Message Andreas Zeugswetter 1998-08-12 08:57:32 AW: [HACKERS] partial index
Previous Message Serj 1998-08-12 08:01:25 Re: [HACKERS] Indexes bug