Feedback-ID: ic6394509:Fastmail
MIME-Version: 1.0
Date: Wed, 14 Aug 2024 14:34:06 +0200
From: "Joel Jacobson" <joel@compiler.org>
To: "Aleksander Alekseev" <aleksander@timescale.com>,
 pgsql-hackers@lists.postgresql.org
Message-Id: <a3531c01-f6a9-4c85-be86-8df932b78400@app.fastmail.com>
In-Reply-To: 
 <CAJ7c6TNVbeNsxbbj1QJTcjdkHRmJUBfF=iv_dOMDAGxoKY9CyQ@mail.gmail.com>
References: 
 <CAJ7c6TPtOp6+kFX5QX3fH1SVr7v65uHr-7yEJ=GMGQi5uhGtcA@mail.gmail.com>
 <7882ebcf-4f1f-4f57-afeb-e1301dd6c9d3@app.fastmail.com>
 <CAJ7c6TNVbeNsxbbj1QJTcjdkHRmJUBfF=iv_dOMDAGxoKY9CyQ@mail.gmail.com>
Subject: Re: [PATCH] Add get_bytes() and set_bytes() functions
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Archived-At: <https://www.postgresql.org/message-id/a3531c01-f6a9-4c85-be86-8df932b78400%40app.fastmail.com>
Precedence: bulk

On Wed, Aug 14, 2024, at 13:31, Aleksander Alekseev wrote:
>> I wonder if get_bytes() and set_bytes() will behave differently
>> on little-endian vs big-endian systems?
> No, the returned value will not depend on the CPU endiness. Current
> implementation uses big-endian / network order which in my humble
> opinion is what most users would expect.

Nice.

I've reviewed and tested the patch.
It looks straight-forward to me.
I don't see any potential problems.
I've marked it Ready for Committer.

> I believe we also need reverse(bytea) and repeat(bytea, integer)
> functions e.g. for those who want little-endian. However I want to
> propose them separately when we are done with this patch.

I agree those functions would be nice too.

I also think it would be nice to provide these convenience functions:
to_bytes(bigint) -> bytea
from_bytes(bytea) -> bigint

Since if not having a current bytea value,
and just wanting to convert a bigint to bytea,
then one would need to construct an zeroed bytea
of the proper size first, to then use set_bytes().

And if just wanting to convert the entire bytea to a bigint,
then one would need to pass 0 as offset and the length
of the bytea as size.

Regards,

Joel