From mboxrd@z Thu Jan 1 00:00:00 1970 From: Carsten Dominik Subject: Re: org-map-entries doesn't understand deletions Date: Sun, 5 Apr 2009 13:48:53 +0200 Message-ID: <1A2419E0-F631-4991-A62E-078F8602356A@gmail.com> References: <20524da70904041606t6e1dcb51ma568894b171e465b@mail.gmail.com> Mime-Version: 1.0 (Apple Message framework v930.3) Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes Content-Transfer-Encoding: 7bit Return-path: Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1LqQqX-00007N-7u for emacs-orgmode@gnu.org; Sun, 05 Apr 2009 07:49:05 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1LqQqS-00007B-Ms for emacs-orgmode@gnu.org; Sun, 05 Apr 2009 07:49:04 -0400 Received: from [199.232.76.173] (port=35902 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1LqQqS-000074-EV for emacs-orgmode@gnu.org; Sun, 05 Apr 2009 07:49:00 -0400 Received: from mx20.gnu.org ([199.232.41.8]:13936) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1LqQqR-0000zK-Vu for emacs-orgmode@gnu.org; Sun, 05 Apr 2009 07:49:00 -0400 Received: from mail-ew0-f160.google.com ([209.85.219.160]) by mx20.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1LqQqR-0000fI-41 for emacs-orgmode@gnu.org; Sun, 05 Apr 2009 07:48:59 -0400 Received: by ewy4 with SMTP id 4so1768035ewy.42 for ; Sun, 05 Apr 2009 04:48:57 -0700 (PDT) In-Reply-To: <20524da70904041606t6e1dcb51ma568894b171e465b@mail.gmail.com> List-Id: "General discussions about Org-mode." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org Errors-To: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org To: Samuel Wales Cc: emacs-orgmode@gnu.org Hi Samuel, you are quite right. The mapping function does wrap the call to your action function into a save-excursion form that assumes to return to the beginning of the headline. It then jumps to the end of that line and continues search from there. This is by design, because it allows "quick and dirty" uses of the mapper without having to take care where the cursor is after the function call. And is does the right thing in all cases where meta data of the entry at point i changed. But you are also right, that for your application this does not work out right. To remidy this situation, get the latest git version. Your action function may now set the variable `org-map-continue-from-here' to the buffer position from where you wish to continue the search. In your case you would now remove you attempt to fix this (outline-previous-visible-heading 1))) and replace it with (setq org-map-continue-from (point)))) This should fix the problems. On Apr 5, 2009, at 1:06 AM, Samuel Wales wrote: > A while back, I wanted to archive all tasks that were closed > more than n days ago (or, better, archive tasks merked with > a done todo keyword, since some tasks have vestigial closed > timestamps without having a done todo keyword). > > I tried using mapping. However, it took me months to get > this far, was extremely difficult for me, and it is still > not working. I cannot proceed further, but I also now think > that the remaining issue is probably not user error. > > The problem seems to be that archiving deletes entries, > which places point after the deleted entry, which causes > mapping to skip an entry. See the comments for more details, > including > why this is apparently not fixable by the user. > > If it is not user error, then there might be a fundamental > problem with how org-map-entries places point. I have a suggestion > for a possible solution in the comments. > > Thanks. > > > ;;i don't want to change the todo kw > (setf org-archive-mark-done nil) > > (defvar alpha-org-archive-expired-ndays 31 > "when done (or closed -- see code) entries are expired, expire > only older todo entries.") > (defun alpha-org-days-since-closed () > "age in days since closed" > (interactive) > (let ((ts (org-entry-get nil "CLOSED"))) > (when ts (org-days-to-time ts)))) > (defun alpha-org-expired-p () > (let ((n (alpha-org-days-since-closed))) > (and n (< n (- alpha-org-archive-expired-ndays))))) > (defun alpha-org-archive-heading () > (interactive) > ;;archiving moves point to the next entry, which is congruent to > ;;/forward/ killing. however, mapping moves point /after/ > ;;running this. save-excursion would do nothing useful here. > ;; > ;;i think it's undocumented whether mapping operates 1. > ;;according to point after execution, 2. goes to a prior > ;;concept of point (e.g. going to a marker placed at the next > ;;entry), or 3. does some kind of save-excursion thing that > ;;assumes no deletion. this question is important. > ;; > ;;it seems to be 3. try running this on 2 sequential entries > ;;that need archiving. it will skip over the second even if you > ;;manually move to the previous entry. thus, perhaps, any > ;;kludging around it in this function is reversed. the code is > ;;incomprehensible to me. > ;; > ;;i think that 3. is probably incorrect behavior because > ;;sometimes, as here, people want to delete > ;;entries. 1. would at least leave everything up to user > ;;control and I would recommend it, despite its drawbacks. if > save-excursion, or whatever, was placed there to fix some other > problem, then doing so might require care in implementing. perhaps > even with a parameter. > (when (alpha-org-expired-p) > ;;; ;;(insert (prin1-to-string (alpha-org-days-since-closed))) > (org-archive-subtree) > ;;the above moves point to the next entry, and exiting this > ;;form moves it again. therefore we have to move back. dunno > ;;if visibility matters, so playing it safe by saying visible > ;;here and show-all in the caller. this will instantiate a > ;;loop if archiving does not do as expected. that would be a > ;;bug. > (outline-previous-visible-heading 1))) > (defun alpha-org-archive-expired-closed () > (interactive) > ;;forestall issues with folding and visibility > (show-all) > (org-map-entries 'alpha-org-archive-heading > ;;this expression is hardcoded instead of > ;;specifying doneness. otoh, there are > ;;sometimes mistakenly closed items that are > ;;not done in the kw, and we don't want to > ;;archive those. > ;; > ;;"/+DONE|+MOOT" > ;;TODO={^DONE$\\|^MOOT$} > ;;"+CLOSED<\"<-1m>\"" > t > 'file)) > > -- > Myalgic encephalomyelitis denialism is causing death (decades early; > Jason et al. 2006) and severe suffering (worse than nearly all other > diseases studied; e.g. Schweitzer et al. 1995) and grossly corrupting > science. http://www.meactionuk.org.uk/What_Is_ME_What_Is_CFS.htm > > > _______________________________________________ > Emacs-orgmode mailing list > Remember: use `Reply All' to send replies to the list. > Emacs-orgmode@gnu.org > http://lists.gnu.org/mailman/listinfo/emacs-orgmode