Re: bbdbv3 and wanderlust

Yes, I'm certain that this is the problem.

When I use the normal setup, and hit the : key, an entry gets entered
into my bbdb.  However, when I set the update to happen on every
redisplay, it does not.

When we are in bbdb-wl.el, inside of the function bbdb/wl-header,
(wl-summary-message-number) returns nil, if we get there via the
redisplay hook.  It think that the problem is that we have already
moved into the mime-view buffer below, and then the

    (if (or (re-search-forward "\r\\(-?[0-9]+\\)" (point-at-eol) t)
            (re-search-forward "^ *\\(-?[0-9]+\\)" (point-at-eol) t))
        (string-to-number (wl-match-buffer 1))
      nil))

code doesn't work.

Is it possible to make this function make sure it is in the right
window somehow?  Also, what happens when the user has configured their
wanderlust to not show the message number?  Maybe we could get that
information from some magical elmo-* calls?

Thanks,
Rob


On Sun, 10 Sep 2017 00:06:16 -0700,
Rob Walker wrote:
> 
> 
> I can't be certain, but I *think* that when we are inside of the
> bbdb/wl-header function, the call to wl-summary-message-number returns
> nil.  I have tried to fake it with hardcoded values, but gotten no
> success, so I don't know.
> 
> Thanks,
> Rob
> 
> 
> On Sat, 09 Sep 2017 16:23:48 -0700,
> Rob Walker wrote:
> > 
> > Hi everyone,
> > 
> > tl;dr I tried a bunch of things, and I get stuck debugging in the
> > function bbdb/wl-header in the file bbdb-wl.el.
> > 
> > I have some questions about using bbdbv3 with Wanderlust.  I have the
> > most simple setup I could implement.
> > 
> > I have only added melpa, installed wl and bbdb from melpa.  ("adduser
> > g" and "ssh -X g@localhost" are great to create pristine environments)
> > 
> > My bbdb initialization code consists of the following lines:
> > 
> > (require 'bbdb) 
> > (bbdb-initialize 'wl)
> > (bbdb-mua-auto-update-init 'wl)
> > 
> > The call to bbdb-initialize works, because
> > 
> > wl-init-hook is a variable defined in ‘wl-vars.el’.
> > Its value is (bbdb-insinuate-wl wl-plugged-init-icons wl-biff-init-icons)
> > 
> > Also, while I am in the Summary buffer, C-h k : returns
> > 
> > : runs the command bbdb-mua-display-sender (found in
> > wl-summary-mode-map), which is an interactive autoloaded compiled Lisp
> > function in ‘bbdb-mua.el’.
> > 
> > It is bound to :.
> > 
> > (bbdb-mua-display-sender &optional UPDATE-P)
> > 
> > The first time I do this, I get the message:
> > 
> > (New file)
> > 
> > At that moment, I am able to bring up a buffer called "bbdb", which
> > has the following in it:
> > 
> > ;; -*- mode: Emacs-Lisp; coding: utf-8; -*-
> > ;;; file-format: 7
> > 
> > So far, so good, right?  Yes, except that file has no data in it.  I
> > then hit : on a few more emails, and still nothing is added to the
> > bbdb file.
> > 
> > The second line seems to work as well, since:
> > 
> > wl-message-redisplay-hook is a variable defined in ‘wl-vars.el’.
> > Its value is (bbdb-mua-auto-update)
> > 
> > However, nothing happens when I'm reading emails.  We enter
> > (bbdb-mua-auto-update), but we never exit the function.
> > 
> >   (let* ((bbdb-silent-internal t)
> >          (records (bbdb-mua-update-records header-class
> >                                            (or update-p
> >                                                bbdb-mua-auto-update-p)))
> >          (bbdb-pop-up-window-size bbdb-mua-pop-up-window-size))
> >     (if bbdb-mua-pop-up
> >         (if records
> >               (bbdb-display-records records bbdb-pop-up-layout
> >                                     nil nil (bbdb-mua-window-p))
> >           ;; If there are no records, empty the BBDB window.
> >           (bbdb-undisplay-records)))
> >     records)
> > 
> > is the code that is failing.
> > 
> > We enter (bbdb-mua-update-records), but never exit it.
> > 
> > We enter (bbdb-mua), but never exit it.  At the time of entering
> > (bbdb-mua), 
> > 
> > bbdb-mua-mode-alist - ((vm vm-mode vm-virtual-mode vm-summary-mode vm-presentation-mode) (gnus gnus-summary-mode gnus-article-mode gnus-tree-mode) (rmail rmail-mode rmail-summary-mode) (mh mhe-mode mhe-summary-mode mh-folder-mode) (mu4e mu4e-view-mode) (wl wl-summary-mode wl-draft-mode) (message message-mode mu4e-compose-mode notmuch-message-mode) (mail mail-mode))
> > 
> > I believe the problem is that our display buffer is in
> > 
> > major-mode is a variable defined in ‘C source code’.
> > Its value is ‘mime-view-mode’
> > Original value was fundamental-mode
> > Local in buffer  *WL:Message*-608594; global value is fundamental-mode
> > 
> >   Automatically becomes buffer-local when set.
> > 
> > Documentation:
> >  
> > I added mime-view-mode to the (defconst bbdb-mua-mode-alist) and now
> > we exit (bbdb-mua) and (bbdb-mua-update-records).  So with that one
> > change, I think we are able to get to the code at line 615 in
> > bbdb-mua.el which looks like this:
> > 
> >        ;; Wanderlust
> >        ((eq mua 'wl)
> >         (bbdb-update-records (bbdb-get-address-components header-class)
> >                              update-p sort))
> > 
> > 
> > This is where I get lost.  I am currently looking at
> > bbdb-get-adress-components, but I don't know how it is getting the
> > addresses.  I suspect it isn't able to jump into the mime-view file
> > somehow.  My thoughts moving forward are to figure out how the code at
> > https://bitbucket.org/cnngimenez/bbdb-wl grabs the headers, and use
> > that instead of what bbdb.el is doing.  But that's just a guess.
> > 
> > Thanks,
> > Rob
> > 
>