From: Thomas Jost <schnouki@schnouki.net>
To: David Bremner <david@tethera.net>, notmuch@notmuchmail.org
Subject: Re: [PATCH v4 1/3] show: don't use hex literals in JSON output
In-Reply-To: <87mx9id1wo.fsf@zancas.localnet>
References: <87fwfaemuc.fsf@thor.loria.fr>
	<1327052365-20012-1-git-send-email-schnouki@schnouki.net>
	<87mx9id1wo.fsf@zancas.localnet>
User-Agent: Notmuch/0.11+96~g5e187d3 (http://notmuchmail.org) Emacs/24.0.92.1
	(x86_64-unknown-linux-gnu)
Date: Fri, 20 Jan 2012 13:33:58 +0100
Message-ID: <87d3aeeeo9.fsf@thor.loria.fr>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
	micalg=pgp-sha1; protocol="application/pgp-signature"
Precedence: list

--=-=-=
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On Fri, 20 Jan 2012 07:55:03 -0400, David Bremner <david@tethera.net> wrote:
> On Fri, 20 Jan 2012 10:39:23 +0100, Thomas Jost <schnouki@schnouki.net> w=
rote:
> > JSON does not support hex literals (0x..) so numbers must be formatted =
as %d
> > instead of %x.
> > ---
> >  notmuch-show.c |    2 +-
> >  1 files changed, 1 insertions(+), 1 deletions(-)
>=20=20
> Probably I'm just being lazy here, but can you explain why this change
> does not require a corresponding change on the emacs side?

Because Emacs already does the right thing. JSON numbers are supposed to
be decimal only (see http://json.org/: digits are 0-9 only), but the
current code could result in displaying a hexadecimal number instead
("c" instead of "12"). This would then trigger an error in Emacs, or in
any other correct JSON parser.

However we are quite lucky: because of the possible values of the gmime
error codes, such an error cannot happen.

The most common gmime error codes are 1 (expired signature), 2 (no
public key), 4 (expired key) and 8 (revoked key). The other possible
value is 16 (unsupported algorithm) but obviously it is much more rare.
If this happens, the current code will add '"errors": 10' (hex for
16...). This is valid JSON (it looks like a decimal number) but it is
incorrect (should be 16, not 10).

With this patch, notmuch will correctly display '"errors": 16' if such a
case happens.

(By the way, this issue was spotted by Austin Clements in
id:"20120117034714.GG16740@mit.edu", so he deserves the credits :))

Regards,

=2D-=20
Thomas/Schnouki

--=-=-=
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iQEcBAEBAgAGBQJPGV83AAoJEMPdciX+bh5IjeoIAJScqo3bRI0pJumtxrLA1OqB
eb3iYDptZ4a8i4oK15mj+PLERvUKPZEW6aggSfrb3Y3FJXNQZ0ilPgvnIIzSvXrF
i7VYk7xSjZhsudRgcGYPV1AQckhQ0DAQvOWmcgiCG15vGbHddKNfnBuFOfigSAq7
qhO0AN0j/hYwMGH0srF29rbLrAfU5Sw1H8GGCqBe8aqsvILlLUv1L0wOj7y6UcgJ
MtFjXwfhI39fWZMgRDnVekcNhCWwLeJPbzRLl/VqinhZvGOBn7J5jDZR9lznosTb
UIZgGEV8vZ2YjmuROWT9sqbmBbNBH6FRghlwC5ftELNBiRrsW5nqp+PSalRTdSs=
=Y8PS
-----END PGP SIGNATURE-----
--=-=-=--