From: Michal Sojka <sojkam1@fel.cvut.cz>
To: David Bremner <david@tethera.net>, notmuch@notmuchmail.org
Subject: Re: [PATCH RFC] Emacs: Add address completion mechanism implemented
	in elisp
In-Reply-To: <87r411f604.fsf@maritornes.cs.unb.ca>
References: <87bns8f6hi.fsf@steelpick.2x.cz>
	<1406653070-13174-1-git-send-email-sojkam1@fel.cvut.cz>
	<87r411f604.fsf@maritornes.cs.unb.ca>
User-Agent: Notmuch/0.18.1+57~gbabacd5 (http://notmuchmail.org) Emacs/24.3.1
	(x86_64-pc-linux-gnu)
Date: Mon, 11 Aug 2014 16:31:06 +0200
Message-ID: <87r40np07p.fsf@steelpick.2x.cz>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: list

Hi David,

On Fri, Aug 01 2014, David Bremner wrote:
> Michal Sojka <sojkam1@fel.cvut.cz> writes:
>
>> notmuch-company.el hooks itself into message-mode and uses
>> company-mode to offer the completion to the user. The file is put into
>> the contrib directory which means that the use has to install it
>> himself. This is because company-mode is not a part of Emacs and
>> bytecompiling notmuch-company.el fails due to used --quick option that
>> causes user installed packages to be ignored. 
>
> what about for now just putting 
>
> ;; -*-no-byte-compile: t; -*-
>
> at the top of the file?  It seems usable uncompiled to me.
>
> Then somebody who really wants it compiled can figure out how to do the
> conditional compilation.
>
>> It would probably make sense to implement another completion frontend
>> based only on Emacs built-in functionality and integrate it with
>> notmuch-addresses.el.
>
> Agreed.
>
>> +  "`company-mode' completion back-end for `nevermore (nm)'."
>
> missed "nevermore" ;)
>
>
>>  
>> +(defun notmuch-flatten-thread-set (thread-set)
>> +  "Convert the result of 'notmuch show' to the plain list of messages."
>> +(defun notmuch-flatten-thread (thread)
>> +
>> +(defun notmuch-flatten-thread-node (thread-node)
>
>
> what about putting those functions in notmuch-query.el?
>
>
>> +(defun notmuch-async-harvest ()
>> +  "Collect possible addresses for completion. It queries the
>
>
> logically it seems like this might belong in notmuch-address.el,
> even if none of the functions in  in it currently use

Thanks for the comments. I incorporated them in a new version of the
patch, which I'll send in a while.

-Michal