From: Carl Worth <cworth@cworth.org>
To: notmuch@notmuchmail.org
Subject: Fixed recent regression (corruption of binary parts)
User-Agent: Notmuch/0.5 (http://notmuchmail.org) Emacs/23.3.1
	(i486-pc-linux-gnu)
Date: Tue, 31 May 2011 16:31:43 -0700
Message-ID: <874o4a1m74.fsf@yoom.home.cworth.org>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
	micalg=pgp-sha1; protocol="application/pgp-signature"
Precedence: list

--=-=-=
Content-Transfer-Encoding: quoted-printable

I just fixed a bug that I should at least mention to the list.

On the notmuch master branch, since May 23, there has been a bug which
could lead to corruption when saving binary attachments from
messages.

A recent change to share more code internally (to provide "notmuch show
=2D-part" rather than a separate top-level command of "notmuch part
=2D-part") inadvertently passed the part content through a function that
would convert CRLF character pairs to LF and would also potentially
convert to UTF-8, (though only if the part being saved had a non-UTF-8
charset explicitly declared).

Fortunately, I happened to be sent a large attachment hitting the
CRLF->LF corruption so I was able to find and fix this bug.

I've just pushed a fix for this (along with a test case). So if you're
running notmuch newer than 0.5, you'll probably want to update to the
latest to avoid the chance of hitting this bug.

Thanks,

=2DCarl

=2D-=20
carl.d.worth@intel.com

--=-=-=
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iEYEARECAAYFAk3lel8ACgkQ6JDdNq8qSWgjpgCfRX9fJ/m4neW6YgM88AEGBnWo
16EAn2aOmc5CYjHm1LpoBEcaMQGjgYeo
=jVZk
-----END PGP SIGNATURE-----
--=-=-=--