Horton

Re: Implement targetlist SRFs using ROWS FROM() (was Changed SRF in targetlist handling)

Kevin Grittner <kgrittn@gmail.com>

View thread View on postgr.es
From: Kevin Grittner <kgrittn@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Andres Freund <andres@anarazel.de>, Robert Haas <robertmhaas@gmail.com>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2016-09-02T16:04:20Z
Lists: pgsql-hackers 
On Fri, Sep 2, 2016 at 10:10 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:

> Also, before getting too high and mighty with users who expect
> "select * from table" to produce rows in a predictable order,
> you should reflect on the number of places in our regression
> tests that assume exactly that ...

An assumption that not infrequently breaks.  AFAIK, we generally
adjust the tests when that happens, rather than considering it a
bug in the code.  I never thought we did that because there was a
secret, undocumented guarantee of order, but to allow different
code paths to be tested than we would test if we always specified
an order.

--
Kevin Grittner
EDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Commits

  1. Remove obsoleted code relating to targetlist SRF evaluation.

  2. Doc: improve documentation of new SRF-in-tlist behavior.

  3. Move targetlist SRF handling from expression evaluation to new executor node.

  4. Don't split up SRFs when choosing to postpone SELECT output expressions.