From: Jameson Rollins <jrollins@finestructure.net>
To: Notmuch Mail <notmuch@notmuchmail.org>
Subject: Re: problem with notmuch-search-authors-width in JSON emacs
	implementation
In-Reply-To: <87633ko11s.fsf@servo.finestructure.net>
References: <87633ko11s.fsf@servo.finestructure.net>
Date: Thu, 22 Apr 2010 00:37:25 -0400
Message-ID: <878w8gxfui.fsf@servo.finestructure.net>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
	micalg=pgp-sha256; protocol="application/pgp-signature"
Precedence: list

--=-=-=
Content-Transfer-Encoding: quoted-printable

On Wed, 21 Apr 2010 19:09:35 -0400, Jameson Rollins <jrollins@finestructure=
.net> wrote:
> Hey, folks.  I just build from Carl's HEAD which includes the new emacs
> JSON interface and I'm noticing a problem with the
> notmuch-search-authors-width variable.  If I set the variable as
> follows:
>=20
> (setq notmuch-search-authors-width 40)
>=20
> author lists that overrun the specified width start spitting out into
> the subject field.  I think bremner just confirmed this on irc.  I'll
> try to look into this when I get a chance, but I don't have time right
> now.  Maybe someone with more intimate knowledge of this piece of code
> could figure it out more quickly.

So I think I see what might be happening here, but I don't know elisp
well enough to fix it.  Maybe one of the elisp experts out there can
help me figure it out.

notmuch-search-authors-width is actually called three times in
notmuch.el:

0 servo:~/src/notmuch/git [master] $ grep -nH notmuch-search-authors-width =
emacs/*
emacs/notmuch.el:57:(defcustom notmuch-search-authors-width 20
emacs/notmuch.el:65:    ("authors" . ,(format "%%-%ds " notmuch-search-auth=
ors-width))
emacs/notmuch.el:626:		      (if (> authors-length notmuch-search-authors-w=
idth)
emacs/notmuch.el:627:			  (set 'authors (concat (substring authors 0 (- not=
much-search-authors-width 3)) "...")))
0 servo:~/src/notmuch/git [master] $=20

The first call is inside the defcustom definition for
notmuch-search-result-format.  The second two are in the
notmuch-search-process-filter function.  It's pretty clear from the
behavior that setting notmuch-search-authors-width, either with setq or
with "Customize Options", affects what is returned in
notmuch-search-process-filter, but *not* in the defcustom for
notmuch-search-result-format.  I guess what's happening is that the
value used in the defcustom definition of notmuch-search-result-format
is set when the defcustom is defined, and isn't changed later with the
new user-set value.

Do any elisp experts out there know any way out of this?  Is there a way
to escape the expansion of the variable used in
notmuch-search-result-format until it is actually called?

I guess we're using notmuch-search-authors-width to "nicely" truncate
the author list and add the "..." at the end.  We could just straight
truncate the authors field with something like '%-.40s', I think, which
would make the code a little simpler, but might not be as nice.

jamie.

--=-=-=
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
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=5Y5J
-----END PGP SIGNATURE-----
--=-=-=--