Horton

Re: Any objections to implementing LogicalDecodeMessageCB for pgoutput?

Euler Taveira <euler@eulerto.com>

View thread View on postgr.es
From: "Euler Taveira" <euler@eulerto.com>
To: "Amit Kapila" <amit.kapila16@gmail.com>
Cc: "David Pirotte" <dpirotte@gmail.com>, "Ashutosh Bapat" <ashutosh.bapat.oss@gmail.com>, "Dave Cramer" <davecramer@gmail.com>, "Michael Paquier" <michael@paquier.xyz>, "Andres Freund" <andres@anarazel.de>, pgsql-hackers <pgsql-hackers@postgresql.org>, "Ashutosh Bapat" <ashutosh.bapat@enterprisedb.com>
Date: 2021-04-02T23:55:52Z
Lists: pgsql-hackers

Attachments

On Thu, Apr 1, 2021, at 7:19 AM, Amit Kapila wrote:
> Your ideas/suggestions look good to me. Don't we need to provide a
> read function corresponding to logicalrep_write_message? We have it
> for other write functions. Can you please combine all of your changes
> into one patch?
Thanks for taking a look at this patch. I didn't consider a
logicalrep_read_message function because the protocol doesn't support it yet.

/*
* Logical replication does not use generic logical messages yet.
* Although, it could be used by other applications that use this
* output plugin.
*/

Someone that is inspecting the code in the future could possibly check this
discussion to understand why this function isn't available.

This new patch set version has 2 patches that is because there are 2 separate
changes: parse_output_parameters() refactor and logical decoding message
support.


--
Euler Taveira
EDB   https://www.enterprisedb.com/

Commits

  1. Allow pgoutput to send logical decoding messages.

  2. Refactor function parse_output_parameters.