Re: about help message for new pg_dump's --filter option - Mailing list pgsql-hackers

From Daniel Gustafsson
Subject Re: about help message for new pg_dump's --filter option
Date
Msg-id 48A06620-7B90-4B9A-9376-0F82BE40A230@yesql.se
Whole thread Raw
In response to Re: about help message for new pg_dump's --filter option  (Alvaro Herrera <alvherre@alvh.no-ip.org>)
List pgsql-hackers
> On 2 Dec 2023, at 17:02, Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
>
> On 2023-Nov-30, Kyotaro Horiguchi wrote:
>
>> Hello.
>>
>> Recently, a new --filter option was added to pg_dump. I might be
>> wrong, but the syntax of the help message for this feels off. Is the
>> word 'on' not necessary after 'based'?
>>
>>> --filter=FILENAME            include or exclude objects and data from dump
>>>                              based expressions in FILENAME
>
> Isn't this a bit too long?

I was trying to come up with a shorter description but didn't come up with one
that clearly enough described what it does.

>  Maybe we can do something shorter, like
>
>  --filter=FILENAME            determine objects in dump based on FILENAME

I don't think that's an improvement really, it's not obvious what "determine
objects" means.  How about a variation along these lines?:

--filter=FILENAME            include/exclude objects based on rules in FILENAME

If we want to use less horizontal space we can replace FILENAME with FILE,
though I'd prefer not to since FILENAME is already used in the help output for
--file setting a precedent.

--
Daniel Gustafsson




pgsql-hackers by date:

Previous
From: Nathan Bossart
Date:
Subject: Re: Emitting JSON to file using COPY TO
Next
From: Nathan Bossart
Date:
Subject: Re: Is WAL_DEBUG related code still relevant today?