Re: support for MERGE

From: Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org>
To: Pg Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Cc: Simon Riggs <simon(dot)riggs(at)enterprisedb(dot)com>, Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com>, Zhihong Yu <zyu(at)yugabyte(dot)com>, Daniel Westermann <dwe(at)dbi-services(dot)com>, Amit Langote <amitlangote09(at)gmail(dot)com>, Justin Pryzby <pryzby(at)telsasoft(dot)com>, Japin Li <japinli(at)hotmail(dot)com>, Erik Rijkers <er(at)xs4all(dot)nl>, Jaime Casanova <jcasanov(at)systemguards(dot)com(dot)ec>
Subject: Re: support for MERGE
Date: 2022-01-28 20:27:37
Message-ID: 202201282027.jt555atu6hlh@alvherre.pgsql
Views: Raw Message | Whole Thread | Download mbox | Resend email
Lists: pgsql-hackers

MERGE, v10. I am much more comfortable with this version; I have
removed a bunch of temporary hacks and cleaned up the interactions with
table AM and executor, which is something that had been bothering me for
a while. The complete set of changes can be seen in github,

The most important one is probably
where I introduced a new struct used at executor time to pass to
ExecUpdate et al where they can install the various bits of status info
on its way out; this allowed cleanup of the function signatures, as well
as TM_FailureData which was being modified in a somewhat strange way.

I am not aware of anything of significance in terms of remaining work
for this project. The one thing I'm a bit bothered about is the fact
that we expose a lot of executor functions previously static. I am now
wondering if it would be better to move the MERGE executor support
functions into nodeModifyTable.c, which I think would mean we would not
have to expose those function prototypes.

Álvaro Herrera PostgreSQL Developer —

Attachment Content-Type Size
v10-0001-MERGE-SQL-Command-following-SQL-2016.patch text/x-diff 389.0 KB

In response to


Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2022-01-28 21:25:57 Re: Support tab completion for upper character inputs in psql
Previous Message Robert Haas 2022-01-28 20:20:41 Re: archive modules