Horton

Re: COPY FROM WHEN condition

Surafel Temsgen <surafel3000@gmail.com>

View thread View on postgr.es
From: Surafel Temesgen <surafel3000@gmail.com>
To: tomas.vondra@2ndquadrant.com
Cc: Adam Berlin <berlin.ab@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2018-11-30T13:00:12Z
Lists: pgsql-hackers

Attachments

On Thu, Nov 29, 2018 at 2:17 AM Tomas Vondra <tomas.vondra@2ndquadrant.com>
wrote:

(c) allow VOLATILE functions in the FILTER clause, but change the
> behavior to make the behavior sane
>

 Did changing the behavior means getting new snapshot before evaluating a
tuple to ensure the function sees results of any previously executed
queries or there are   other mechanism that can make the behavior sane?

Which leaves us with (b) and (c). Clearly, (b) is simpler to implement,
> because it (c) needs to do the detection too, and then some additional
> stuff. I'm not sure how much more complex (c) is, compared to (b).
>
> The attache patch implement option b prohibit VOLATILE functions but i am
open to change
regards
Surafel

Commits

  1. Remove unused struct member, enforce multi_insert callback presence.

  2. Separate per-batch and per-tuple memory contexts in COPY

  3. Fix handling of volatile expressions in COPY FROM ... WHERE

  4. Allow COPY FROM to filter data using WHERE conditions

  5. Remove obsolete netbsd dynloader code