Hello, org-world. I experience a regression in org-replace-disputed-keys, and I git-bisected it.
* Symptom: org-replace-disputed-keys doesn't work for me. Emacs version: GNU Emacs 24.3.1 (x86_64-pc-linux-gnu, GTK+ Version 3.0.12) * How to reproduce it: 1. Put the org to "/tmp/org-mode" 2. Save the following lisp to "foo.el" ----------------------------------------------------------------------- (progn (setq org-replace-disputed-keys t) (setq org-disputed-keys '(([(shift up)] . [(ctrl up)]) ([(shift down)] . [(ctrl down)]) ([(shift left)] . [(ctrl left)]) ([(shift right)] . [(ctrl right)]) ([(shift meta right)] . [(shift control right)]) ([(shift meta left)] . [(shift control left)]) ([(shift meta up)] . [(shift control up)]) ([(shift meta down)] . [(shift control down)]) ([(shift control right)] . [(shift meta right)]) ([(shift control left)] . [(shift meta left)])) ) (add-to-list 'load-path (expand-file-name "/tmp/org-mode/lisp")) (pop-to-buffer "*scratch*") (insert "* " (org-version) " (" (org-git-version) ")\n ") (require 'org) (org-mode)) ------------------------------------------------------------------------ 3. Run emacs with $ emacs -Q -l foo.el 4. M-x org-schedule, and press ctrl+<cursor>. It should replace shift+<cursor>, but it doesn't. `describe-key' for C-down says: "<C-down> runs the command forward-paragraph, which is an interactive compiled Lisp function in `paragraphs.el'" * The bad commit The exact bad commit can't be tracked (org fails to initialize), but I narrowed it down to 6 commits. It was done between 9a0e84fbd7b0e25cf49613e787d572f3c71723dc and 8ad6f534f9b24f273b7699199df34c0f44f9059f . I here show the diff for "org.el" in that period. The last 2 hunks looks relevant: ------------------------------------------------------------------------ diff --git a/lisp/org.el b/lisp/org.el index aebf58f..7f0717e 100644 --- a/lisp/org.el +++ b/lisp/org.el @@ -6897,7 +6897,7 @@ of the first headline in the buffer. This is important, because if the first headline is not level one, then (hide-sublevels 1) gives confusing results." (interactive) - (let ((l (org-current-line)) + (let ((pos (point)) (level (save-excursion (goto-char (point-min)) (if (re-search-forward (concat "^" outline-regexp) nil t) @@ -6906,7 +6906,7 @@ results." (funcall outline-level)))))) (and level (hide-sublevels level)) (recenter '(4)) - (org-goto-line l))) + (goto-char pos))) (defun org-content (&optional arg) "Show all headlines in the buffer, like a table of contents. @@ -14050,10 +14050,19 @@ See also `org-scan-tags'. minus tag mm tagsmatch todomatch tagsmatcher todomatcher kwd matcher orterms term orlist re-p str-p level-p level-op time-p - prop-p pn pv po gv rest) + prop-p pn pv po gv rest (start 0) (ss 0)) ;; Expand group tags (setq match (org-tags-expand match)) - (if (string-match "/+" match) + + ;; Check if there is a TODO part of this match, which would be the + ;; part after a "/". TO make sure that this slash is not part of + ;; a property value to be matched against, we also check that there + ;; is no " after that slash. + ;; First, find the last slash + (while (string-match "/+" match ss) + (setq start (match-beginning 0) ss (match-end 0))) + (if (and (string-match "/+" match start) + (not (save-match-data (string-match "\"" match start)))) ;; match contains also a todo-matching request (progn (setq tagsmatch (substring match 0 (match-beginning 0)) @@ -16126,7 +16135,8 @@ So these are more for recording a certain time/date." (defvar org-read-date-inactive) (defvar org-read-date-minibuffer-local-map - (let ((map (make-sparse-keymap))) + (let* ((org-replace-disputed-keys nil) + (map (make-sparse-keymap))) (set-keymap-parent map minibuffer-local-map) (org-defkey map (kbd ".") (lambda () (interactive) @@ -16286,7 +16296,6 @@ user." (calendar-current-date)))) (org-eval-in-calendar nil t) (let* ((old-map (current-local-map)) - (org-replace-disputed-keys nil) (map (copy-keymap calendar-mode-map)) (minibuffer-local-map (copy-keymap org-read-date-minibuffer-local-map))) ------------------------------------------------------------------------ Thank you very much for developing org-mode. Kind regards, Teika (Teika kazura) # Hey, it was almost exacly a year ago. Only few use this feature? ;]