Re: [PATCH] clarify palloc comment on quote_literal_cstr - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: [PATCH] clarify palloc comment on quote_literal_cstr
Date
Msg-id Z_L2zipq-qAYClxk@paquier.xyz
Whole thread Raw
In response to [PATCH] clarify palloc comment on quote_literal_cstr  (Steve Chavez <steve@supabase.io>)
Responses Re: [PATCH] clarify palloc comment on quote_literal_cstr
List pgsql-hackers
On Sun, Apr 06, 2025 at 12:37:24PM -0500, Steve Chavez wrote:
> I found the numbers in `quote_literal_cstr` palloc quite magical. So I've
> added a comment clarifying what they mean. The change is small:
>
>         /* We make a worst-case result area; wasting a little space is OK */
> -       result = palloc(len * 2 + 3 + 1);
> +       result = palloc(
> +               (len * 2)    /* worst-case doubling for every character if
> each one is a quote */
> +               + 3          /* two outer quotes + possibly 'E' if needed */
> +               + 1          /* null terminator */
> +       );

Agreed that this is a good idea to have a better documentation here if
someone decides to tweak this area in the future, rather than have
them guess the numbers.

Looking at what quote_literal_internal() does, it seems to me that you
are right based on ESCAPE_STRING_SYNTAX, SQL_STR_DOUBLE(), and the
extra backslashes added to the destination buffer.
--
Michael

Attachment

pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: Typo in comment for pgstat_database_flush_cb()
Next
From: Daniel Gustafsson
Date:
Subject: Re: Possibly hard-to-read message