Re: Patch: add --if-exists to pg_recvlogical

On 9/20/17 02:26, Rosser Schwarz wrote:
> The more I think about it, I don't think it's nonsensical, though.
> --create-slot --if-exists or --drop-slot --if-not-exists, obviously. I
> mean, do you even logic?

Those pieces make sense. We have many CREATE IF NOT EXISTS and DROP IF
EXISTS commands. The use is clear.

> Those aside, --if-exists just means a non-existent slot isn't an error
> condition, doesn't it? --start --if-exists will start, if the slot
> exists. Otherwise it won't, in neither case raising an error. Exactly
> what it says on the tin. Perhaps the docs could make clear that
> combination implies --no-loop (or at least means we'll exit immediately)
> if the slot does not, in fact, exist?

A non-terrible definition could perhaps be arrived at here, but it's not
clear why one would need this. We don't have SELECT FROM IF EXISTS, either.

I suggest you create a patch that focuses on the --create part.

You can create a separate follow-on patch for the --start part if you
wish, but I think that that patch would be rejected.

--
Peter Eisentraut http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services