Date: Sun, 27 Jun 2021 21:35:35 +0800
From: Julien Rouhaud <rjuju123@gmail.com>
To: Ranier Vilela <ranier.vf@gmail.com>
Cc: Pg Hackers <pgsql-hackers@postgresql.org>,
	Pavel Stehule <pavel.stehule@gmail.com>
Subject: Re: Deparsing rewritten query
Message-ID: <20210627133535.hsuf6jwmacmktpnj@nol>
References: 
 <CAEudQAp1kX+G_uG3ZfyCDsdff4ue-vmwbWQzRwWb4oPYOFpunA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CAEudQAp1kX+G_uG3ZfyCDsdff4ue-vmwbWQzRwWb4oPYOFpunA@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/20210627133535.hsuf6jwmacmktpnj%40nol>
Precedence: bulk

On Sun, Jun 27, 2021 at 10:06:00AM -0300, Ranier Vilela wrote:
> 
> 1. strcmp(sql, "") could not be replaced by sql[0] == '\0'?

It's a debugging leftover that I forgot to remove.  There's no point trying
to catch an empty string as e.g. a single space would be exactly the same, and
both would be caught by the next (and correct) test.

> 3. initStringInfo(&buf) inside a loop, wouldn't it be exaggerated? each
> time call palloc0.

initStringInfo calls palloc, not palloc0.

It's unlikely to make any difference.  Rules have been strongly discouraged for
many years, and if you have enough to make a noticeable difference here, you
probably have bigger problems.  But no objection to reset the StringInfo
instead.