Re: ...

Re: ...

[Carsten Dominik]

Hi Christopher,

I mant to copy the list, I am doing this again now.

Wow, I was not aware that Emacs caches by content, this is an important piece of information.  I guess this removed the main concern I had.  Thanks for looking it up in the code and showing it to me.  I am not sure if I understand that code completely, but i trust your judgment.

-Carsten

On 31 jan. 2013, at 10:29, Christopher Schmidt <christopher@ch.ristopher.com> wrote:

> Christopher Schmidt <christopher@ch.ristopher.com> writes:
>> I was not aware of Emacs caching the regexp on a per-object rather
>> than per-content basis.  If that is true, the regexp should be made a
>> buffer local variable, of course!
> 
> I do not think that's true.
> 
> In compile_pattern:
> 
>      for (cpp = &searchbuf_head; ; cpp = &cp->next)
>        {
>          cp = *cpp;
>          ...
>          if (SCHARS (cp->regexp) == SCHARS (pattern)
>              && STRING_MULTIBYTE (cp->regexp) == STRING_MULTIBYTE (pattern)
>              && !NILP (Fstring_equal (cp->regexp, pattern))
>                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>              && EQ (cp->buf.translate, (! NILP (translate) ? translate : make_number (0)))
>              && cp->posix == posix
>              && (EQ (cp->syntax_table, Qt)
>                  || EQ (cp->syntax_table, BVAR (current_buffer, syntax_table)))
>              && !NILP (Fequal (cp->whitespace_regexp, Vsearch_spaces_regexp))
>              && cp->buf.charset_unibyte == charset_unibyte)
>            break;
> 
>          /* If we're at the end of the cache, compile into the nil cell
>             we found, or the last (least recently used) cell with a
>             string value.  */
>          if (cp->next == 0)
>            {
>            compile_it:
>              compile_pattern_1 (cp, pattern, translate, posix);
>              break;
>            }
>        }
>      ...
> 
>        Christopher


-- 
They can't eat you. -- Merlin Mann in his "Scared Shitless" talk

Re: ...

[Christopher Schmidt]

Carsten Dominik <carsten.dominik@gmail.com> writes:
> Hi Christopher,
>
> I mant to copy the list, I am doing this again now.
>
> Wow, I was not aware that Emacs caches by content, this is an important piece of
> information.  I guess this removed the main concern I had.  Thanks for looking
> it up in the code and showing it to me.  I am not sure if I understand that code
> completely, but i trust your judgment.

Please, do not trust me.  This should be brought up to emacs-devel and
the definite answer should be documented in (info "(elisp)Regular
Expressions").

Thanks a lot for your input.  Greetings,

        Christopher

Re: ...

[Bastien]

Hi Carsten and Christopher,

Carsten Dominik <carsten.dominik@gmail.com> writes:

> I mant to copy the list, I am doing this again now.
>
> Wow, I was not aware that Emacs caches by content, this is an important
> piece of information.  I guess this removed the main concern I had.  Thanks
> for looking it up in the code and showing it to me.  I am not sure if I
> understand that code completely, but i trust your judgment.

I'm not sure I have all the background to understand the issue at
stake... can anyone educate me?  Thanks!

-- 
 Bastien

roundtrip integration with Asana?

[Meng Weng Wong]

I use org for personal task management, but I use asana.com with my workgroup.

I just learned that Asana offers an API that exposes all the usual interfaces.
http://developers.asana.com/documentation/

Roundtrip integration between org and Asana is therefore possible.

I'm sure I'm not the first person to have thought about this... which approach would be the preferred architecture for achieving such integration?

I can imagine writing, say, a bunch of Perl scripts using WWW::Asana, to act as a gateway between text file and web API. As far as org and emacs are concerned, the text files magically update themselves, and no additional elisp need be written.

I can also imagine native org-mode extensions that talk directly to asana, in the same way that org2blog/wp talks directly to Wordpress. Until emacs gets threading support, though, blocking HTTP IO operations might seriously inconvenience one's editing.

Which approach do you think is preferable?

Re: ...

[Carsten Dominik]

Hi Bastien,

as you know, regular expressions are a language to do a programmed search for text.  The pattern string has to be compiled before it can be used.  That compilation is a costly process, so most languages that have pattern matching use some kind of cache to store compiled patterns, so that frequently used patterns can be reused without compilation.

I am aware of this very much from studying perl.  In perl, a compiled pattern is associated with a particular instance of a string.  Often you build the pattern by constructing it through concatenation of other parts etc.  In Perl this means that the pattern is recompiled each time a match.  You can work around this issue in Perl by telling it explicitly and on programmers authority that, "yes, this pattern is dynamically constructed, but only once, I guarantee that it will not change, so compile it only once".  So in Perl the difference is

/pattern/      will match against pattern
/$pattern/     will match agains the pattern contained in the
               variable $pattern, and recompilation will occur
               each time
/$pattern/o    will compile only once and trust the programmer.

So I am very aware of this speedup issue.  And I thought that in Emacs, the caching would work by associating a specific string object with the compiled pattern.  But the code Christopher pointed out seems to suggest that the pattern cache works also for strings that are `equal', not only for string that are `eq'.

If this is the case, this means that there is only a very small difference between

(defconst my-pattern (concat "^" "xyz"))
(re-search-forward my-pattern ....)      ; many times in different functions

and

(defconst my-partial-pattern "xyz")
(re-search-forward (concat "^" my-partial-pattern) ....)  ; many times

The difference is only the repeated concatenation operation, and not the recompilation.  I always thought that this would work differently, and that is why a lot of regexps get constructed and then stored in variables or constants.  Of course this is also a good practice for readable and maintainable code, but the impact on efficiency is not as big as I used to think.  So when I saw Christoher's initial patch, I thought a function to create
org-ooutline-regexp-bol would be a large burden in speed - but it now seems that it would only be a minor impact.

Still, I think making a local variable in buffers with org-struct-mode is also a good way to get the functionality Christopher wants.

Clearer?

- Carsten


On 31 jan. 2013, at 12:22, Bastien <bzg@altern.org> wrote:

> Hi Carsten and Christopher,
> 
> Carsten Dominik <carsten.dominik@gmail.com> writes:
> 
>> I mant to copy the list, I am doing this again now.
>> 
>> Wow, I was not aware that Emacs caches by content, this is an important
>> piece of information.  I guess this removed the main concern I had.  Thanks
>> for looking it up in the code and showing it to me.  I am not sure if I
>> understand that code completely, but i trust your judgment.
> 
> I'm not sure I have all the background to understand the issue at
> stake... can anyone educate me?  Thanks!
> 
> -- 
> Bastien


-- 
There is no unscripted life.  Only a badly scripted one. -- Brothers Bloom

Re: ...

[Bastien]

Hi Carsten,

thanks a lot for the very clear explanations.

Carsten Dominik <carsten.dominik@gmail.com> writes:

> The difference is only the repeated concatenation operation, and not the
> recompilation.  I always thought that this would work differently, and that
> is why a lot of regexps get constructed and then stored in variables or
> constants.  Of course this is also a good practice for readable and
> maintainable code, but the impact on efficiency is not as big as I used to
> think.  So when I saw Christoher's initial patch, I thought a function to
> create
> org-ooutline-regexp-bol would be a large burden in speed - but it now seems that it would only be a minor impact.

That was my assumption too... hence my reaction to Christopher's
patch.

It now makes sense -- I'll ask for confirmation on the emacs-devel as
suggested by Christopher.

> Still, I think making a local variable in buffers with
> org-struct-mode is also a good way to get the functionality
> Christopher wants.

Mh... here I'm not sure to understand what you suggest. 
Isn't the whole point of Christopher's patch to rely on
a buffer local value of org-outline-regexp?  I surely miss
something.

Thanks!

-- 
 Bastien

Re: ...

[Bastien]

Christopher Schmidt <christopher@ch.ristopher.com> writes:

> Please, do not trust me.  This should be brought up to emacs-devel and
> the definite answer should be documented in (info "(elisp)Regular
> Expressions").

I asked on emacs-devel.  Thanks,

-- 
 Bastien

Re: roundtrip integration with Asana?

[Bastien]

Hi Meng Weng,

Meng Weng Wong <mengwong@gmail.com> writes:

> I use org for personal task management, but I use asana.com with my workgroup.
>
> I just learned that Asana offers an API that exposes all the usual interfaces.
> http://developers.asana.com/documentation/
>
> Roundtrip integration between org and Asana is therefore possible.
>
> I'm sure I'm not the first person to have thought about
> this... which approach would be the preferred architecture for
> achieving such integration?

I don't know Asana but I'd suggest to have a look at org-sync

  http://orgmode.org/cgit.cgi/org-sync.git/
  ~$ git clone git://orgmode.org/org-sync.git

It allows to connect a local .org file with various APIs.

-- 
 Bastien

Re: ...

[Carsten Dominik]

On 31 jan. 2013, at 14:32, Bastien <bzg@altern.org> wrote:

> Hi Carsten,
> 
> thanks a lot for the very clear explanations.
> 
> Carsten Dominik <carsten.dominik@gmail.com> writes:
> 
>> The difference is only the repeated concatenation operation, and not the
>> recompilation.  I always thought that this would work differently, and that
>> is why a lot of regexps get constructed and then stored in variables or
>> constants.  Of course this is also a good practice for readable and
>> maintainable code, but the impact on efficiency is not as big as I used to
>> think.  So when I saw Christoher's initial patch, I thought a function to
>> create
>> org-ooutline-regexp-bol would be a large burden in speed - but it now seems that it would only be a minor impact.
> 
> That was my assumption too... hence my reaction to Christopher's
> patch.
> 
> It now makes sense -- I'll ask for confirmation on the emacs-devel as
> suggested by Christopher.
> 
>> Still, I think making a local variable in buffers with
>> org-struct-mode is also a good way to get the functionality
>> Christopher wants.
> 
> Mh... here I'm not sure to understand what you suggest. 
> Isn't the whole point of Christopher's patch to rely on
> a buffer local value of org-outline-regexp?  I surely miss
> something.

Not really, I am not particulrly clear here.

Christopher proposes to do this:

   ;; Local Variables:
   ;; eval: (orgstruct-mode 1)
   ;; org-outline-regexp: ";;; \\*+"
   ;; org-heading-regexp: "^;;; \\(\\*+\\)\\(.+\\)$"
   ;; End:

I was just thinking, if you are setting up two local variables, you could set up three and include org-outline-regexp-bol as well.  Or, you could just set up `org-outline-regexp', and make the function org-struct-mode set up the other two as additional local variables.  This would simplify things.  In fact, I think this would make for a more compact and better patch, Christopher, what do you think?

- Carsten

Re: orgstruct-mode with custom headline prefix (was: ...)

[Christopher Schmidt]

[-- Attachment #1: Type: text/plain, Size: 813 bytes --]

Carsten Dominik <carsten.dominik@gmail.com> writes:
> I was just thinking, if you are setting up two local variables, you
> could set up three and include org-outline-regexp-bol as well.  Or,
> you could just set up org-outline-regexp', and make the function
> org-struct-mode set up the other two as additional local variables.
> This would simplify things.  In fact, I think this would make for a
> more compact and better patch, Christopher, what do you think?

I agree.  I am not sure how to deduce org-heading-regexp from
org-outline-regexp, though.  Maybe we can use a buffer-local
orgstruct-heading-prefix and setup
org-heading-regexp/org-outline-regexp(-bol) from this one?

Here is my patch with org-outline-regexp-bol being a variable and being
set up by the hijacker commands using org-outline-regexp.

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: Type: text/x-diff, Size: 16459 bytes --]

--- a/lisp/org.el
+++ b/lisp/org.el
@@ -92,6 +92,7 @@
 ;; job when `orgstruct-mode' is active.
 (defvar org-outline-regexp "\\*+ "
   "Regexp to match Org headlines.")
+;;;###autoload(put 'org-outline-regexp 'safe-local-variable 'stringp)
 
 (defvar org-outline-regexp-bol "^\\*+ "
   "Regexp to match Org headlines.
@@ -101,6 +102,7 @@ sure that we are at the beginning of the line.")
 (defvar org-heading-regexp "^\\(\\*+\\)\\(?: +\\(.*?\\)\\)?[ \t]*$"
   "Matches an headline, putting stars and text into groups.
 Stars are put in group 1 and the trimmed body in group 2.")
+;;;###autoload(put 'org-heading-regexp 'safe-local-variable 'stringp)
 
 ;; Emacs 22 calendar compatibility:  Make sure the new variables are available
 (when (fboundp 'defvaralias)
@@ -6223,8 +6225,10 @@ and subscripts."
 
 (defvar org-cycle-global-status nil)
 (make-variable-buffer-local 'org-cycle-global-status)
+(put 'org-cycle-global-status 'org-state t)
 (defvar org-cycle-subtree-status nil)
 (make-variable-buffer-local 'org-cycle-subtree-status)
+(put 'org-cycle-subtree-status 'org-state t)
 
 (defvar org-inlinetask-min-level)
 
@@ -7403,13 +7407,24 @@ This is a list with the following elements:
 - the tags string, or nil."
   (save-excursion
     (org-back-to-heading t)
-    (if (let (case-fold-search) (looking-at org-complex-heading-regexp))
-	(list (length (match-string 1))
-	      (org-reduced-level (length (match-string 1)))
-	      (org-match-string-no-properties 2)
-	      (and (match-end 3) (aref (match-string 3) 2))
-	      (org-match-string-no-properties 4)
-	      (org-match-string-no-properties 5)))))
+    (if (let (case-fold-search)
+          (looking-at
+           (if orgstruct-mode
+               org-heading-regexp
+             org-complex-heading-regexp)))
+        (if orgstruct-mode
+            (list (length (match-string 1))
+                  (org-reduced-level (length (match-string 1)))
+                  nil
+                  nil
+                  (match-string 2)
+                  nil)
+          (list (length (match-string 1))
+                (org-reduced-level (length (match-string 1)))
+                (org-match-string-no-properties 2)
+                (and (match-end 3) (aref (match-string 3) 2))
+                (org-match-string-no-properties 4)
+                (org-match-string-no-properties 5))))))
 
 (defun org-get-entry ()
   "Get the entry text, after heading, entire subtree."
@@ -8482,12 +8497,12 @@ If WITH-CASE is non-nil, the sorting will be case-sensitive."
 ;; command.  There might be problems if any of the keys is otherwise
 ;; used as a prefix key.
 
-;; Another challenge is that the key binding for TAB can be tab or \C-i,
-;; likewise the binding for RET can be return or \C-m.  Orgtbl-mode
-;; addresses this by checking explicitly for both bindings.
+(defcustom orgstruct-setup-hook nil
+  "Hook run after orgstruct-mode-map is filled."
+  :group 'org
+  :type 'hook)
 
-(defvar orgstruct-mode-map (make-sparse-keymap)
-  "Keymap for the minor `orgstruct-mode'.")
+(defvar orgstruct-initialized nil)
 
 (defvar org-local-vars nil
   "List of local variables, for use by `orgstruct-mode'.")
@@ -8498,26 +8513,13 @@ If WITH-CASE is non-nil, the sorting will be case-sensitive."
 This mode is for using Org-mode structure commands in other
 modes.  The following keys behave as if Org-mode were active, if
 the cursor is on a headline, or on a plain list item (both as
-defined by Org-mode).
-
-M-up        Move entry/item up
-M-down	    Move entry/item down
-M-left	    Promote
-M-right	    Demote
-M-S-up	    Move entry/item up
-M-S-down    Move entry/item down
-M-S-left    Promote subtree
-M-S-right   Demote subtree
-M-q	    Fill paragraph and items like in Org-mode
-C-c ^	    Sort entries
-C-c -	    Cycle list bullet
-TAB         Cycle item visibility
-M-RET       Insert new heading/item
-S-M-RET     Insert new TODO heading / Checkbox item
-C-c C-c     Set tags / toggle checkbox"
-  nil " OrgStruct" nil
-  (org-load-modules-maybe)
-  (and (orgstruct-setup) (defun orgstruct-setup () nil)))
+defined by Org-mode)."
+  nil " OrgStruct" (make-sparse-keymap)
+  (when orgstruct-mode
+    (org-load-modules-maybe)
+    (unless orgstruct-initialized
+      (orgstruct-setup)
+      (setq orgstruct-initialized t))))
 
 ;;;###autoload
 (defun turn-on-orgstruct ()
@@ -8568,104 +8570,90 @@ buffer.  It will also recognize item context in multiline items."
   (error "This key has no function outside structure elements"))
 
 (defun orgstruct-setup ()
-  "Setup orgstruct keymaps."
-  (let ((nfunc 0)
-	(bindings
-	 (list
-	  '([(meta up)]           org-metaup)
-	  '([(meta down)]         org-metadown)
-	  '([(meta left)]         org-metaleft)
-	  '([(meta right)]        org-metaright)
-	  '([(meta shift up)]     org-shiftmetaup)
-	  '([(meta shift down)]   org-shiftmetadown)
-	  '([(meta shift left)]   org-shiftmetaleft)
-	  '([(meta shift right)]  org-shiftmetaright)
-	  '([?\e (up)]            org-metaup)
-	  '([?\e (down)]          org-metadown)
-	  '([?\e (left)]          org-metaleft)
-	  '([?\e (right)]         org-metaright)
-	  '([?\e (shift up)]      org-shiftmetaup)
-	  '([?\e (shift down)]    org-shiftmetadown)
-	  '([?\e (shift left)]    org-shiftmetaleft)
-	  '([?\e (shift right)]   org-shiftmetaright)
-	  '([(shift up)]          org-shiftup)
-	  '([(shift down)]        org-shiftdown)
-	  '([(shift left)]        org-shiftleft)
-	  '([(shift right)]       org-shiftright)
-	  '("\C-c\C-c"            org-ctrl-c-ctrl-c)
-	  '("\M-q"                fill-paragraph)
-	  '("\C-c^"               org-sort)
-	  '("\C-c-"               org-cycle-list-bullet)))
-	elt key fun cmd)
-    (while (setq elt (pop bindings))
-      (setq nfunc (1+ nfunc))
-      (setq key (org-key (car elt))
-	    fun (nth 1 elt)
-	    cmd (orgstruct-make-binding fun nfunc key))
-      (org-defkey orgstruct-mode-map key cmd))
-
-    ;; Prevent an error for users who forgot to make autoloads
-    (require 'org-element)
-
-    ;; Special treatment needed for TAB and RET
-    (org-defkey orgstruct-mode-map [(tab)]
-		(orgstruct-make-binding 'org-cycle 102 [(tab)] "\C-i"))
-    (org-defkey orgstruct-mode-map "\C-i"
-		(orgstruct-make-binding 'org-cycle 103 "\C-i" [(tab)]))
-
-    (org-defkey orgstruct-mode-map "\M-\C-m"
-		(orgstruct-make-binding 'org-insert-heading 105
-					"\M-\C-m" [(meta return)]))
-    (org-defkey orgstruct-mode-map [(meta return)]
-		(orgstruct-make-binding 'org-insert-heading 106
-					[(meta return)] "\M-\C-m"))
-
-    (org-defkey orgstruct-mode-map [(shift meta return)]
-		(orgstruct-make-binding 'org-insert-todo-heading 107
-					[(meta return)] "\M-\C-m"))
-
-    (org-defkey orgstruct-mode-map "\e\C-m"
-		(orgstruct-make-binding 'org-insert-heading 108
-					"\e\C-m" [?\e (return)]))
-    (org-defkey orgstruct-mode-map [?\e (return)]
-		(orgstruct-make-binding 'org-insert-heading 109
-					[?\e (return)] "\e\C-m"))
-    (org-defkey orgstruct-mode-map [?\e (shift return)]
-		(orgstruct-make-binding 'org-insert-todo-heading 110
-					[?\e (return)] "\e\C-m"))
-
-    (unless org-local-vars
-      (setq org-local-vars (org-get-local-variables)))
-
-    t))
-
-(defun orgstruct-make-binding (fun n &rest keys)
+  "Setup orgstruct keymap."
+  (dolist (f
+           '("org-meta"
+             "org-shiftmeta"
+             org-shifttab
+             org-backward-element
+             org-backward-heading-same-level
+             org-ctrl-c-ret
+             org-cycle
+             org-forward-heading-same-level
+             org-insert-heading
+             org-insert-heading-respect-content
+             org-kill-note-or-show-branches
+             org-mark-subtree
+             org-narrow-to-subtree
+             org-promote-subtree
+             org-reveal
+             org-show-subtree
+             org-sort
+             org-up-element
+             outline-demote
+             outline-next-visible-heading
+             outline-previous-visible-heading
+             outline-promote
+             outline-up-heading
+             show-children))
+    (dolist (f (if (stringp f)
+                   (let ((flist))
+                     (dolist (postfix
+                              '("-return" "tab" "left" "right" "up" "down")
+                              flist)
+                       (let ((f (intern (concat f postfix))))
+                         (when (fboundp f)
+                           (push f flist)))))
+                 (list f)))
+      (dolist (binding (nconc (where-is-internal f org-mode-map)
+                              (where-is-internal f outline-mode-map)))
+        (dolist (rep '(("<tab>" . "TAB")
+                       ("<ret>" . "RET")
+                       ("<esc>" . "ESC")
+                       ("<del>" . "DEL")))
+          (setq binding (kbd (replace-regexp-in-string
+                              (regexp-quote (car rep))
+                              (cdr rep)
+                              (key-description binding)))))
+        (let ((key (lookup-key orgstruct-mode-map binding)))
+          (when (or (not key) (numberp key))
+            (org-defkey orgstruct-mode-map
+                        binding
+                        (orgstruct-make-binding f binding)))))))
+  (run-hooks 'orgstruct-setup-hook))
+
+(defun orgstruct-make-binding (fun key)
   "Create a function for binding in the structure minor mode.
-FUN is the command to call inside a table.  N is used to create a unique
-command name.  KEYS are keys that should be checked in for a command
-to execute outside of tables."
-  (eval
-   (list 'defun
-	 (intern (concat "orgstruct-hijacker-command-" (int-to-string n)))
-	 '(arg)
-	 (concat "In Structure, run `" (symbol-name fun) "'.\n"
-		 "Outside of structure, run the binding of `"
-		 (mapconcat (lambda (x) (format "%s" x)) keys "' or `")
-		 "'.")
-	 '(interactive "p")
-	 (list 'if
-	       `(org-context-p 'headline 'item
-			       (and orgstruct-is-++
-				    ,(and (memq fun '(org-insert-heading org-insert-todo-heading)) t)
-				    'item-body))
-	       (list 'org-run-like-in-org-mode (list 'quote fun))
-	       (list 'let '(orgstruct-mode)
-		     (list 'call-interactively
-			   (append '(or)
-				   (mapcar (lambda (k)
-					     (list 'key-binding k))
-					   keys)
-				   '('orgstruct-error))))))))
+FUN is the command to call inside a table.  KEY is the key that
+should be checked in for a command to execute outside of tables."
+  (let ((name (concat "orgstruct-hijacker-" (symbol-name fun))))
+    (let ((nname name)
+          (i 0))
+      (while (fboundp (intern nname))
+        (setq nname (format "%s-%d" name (setq i (1+ i)))))
+      (setq name (intern nname)))
+    (eval
+     `(defun ,name (arg)
+        ,(concat "In Structure, run `" (symbol-name fun) "'.\n"
+                 "Outside of structure, run the binding of `"
+                 (key-description key) "'.")
+        (interactive "p")
+        (unless
+            (let ((outline-regexp org-outline-regexp)
+                  (org-outline-regexp-bol
+                   (concat "^" org-outline-regexp)))
+              (when (org-context-p 'headline 'item
+                                   ,(when (memq fun '(org-insert-heading))
+                                      '(when orgstruct-is-++
+                                         'item-body)))
+                (org-run-like-in-org-mode ',fun)
+                t))
+          (let ((binding (let ((orgstruct-mode)) (key-binding ,key))))
+            (if (keymapp binding)
+                (set-temporary-overlay-map binding)
+              (call-interactively
+               (or binding 'orgstruct-error)))))))
+    name))
 
 (defun org-contextualize-keys (alist contexts)
   "Return valid elements in ALIST depending on CONTEXTS.
@@ -8762,17 +8750,18 @@ Possible values in the list of contexts are `table', `headline', and `item'."
       (setq varlist (buffer-local-variables)))
     (kill-buffer "*Org tmp*")
     (delq nil
-	  (mapcar
-	   (lambda (x)
-	     (setq x
-		   (if (symbolp x)
-		       (list x)
-		     (list (car x) (list 'quote (cdr x)))))
-	     (if (string-match
-		  "^\\(org-\\|orgtbl-\\|outline-\\|comment-\\|paragraph-\\|auto-fill\\|normal-auto-fill\\|fill-paragraph\\|indent-\\)"
-		  (symbol-name (car x)))
-		 x nil))
-	   varlist))))
+          (mapcar
+           (lambda (x)
+             (setq x
+                   (if (symbolp x)
+                       (list x)
+                     (list (car x) (cdr x))))
+             (if (and (not (get (car x) 'org-state))
+                      (string-match
+                       "^\\(org-\\|orgtbl-\\|outline-\\|comment-\\|paragraph-\\|auto-fill\\|normal-auto-fill\\|fill-paragraph\\|indent-\\)"
+                       (symbol-name (car x))))
+                 x nil))
+           varlist))))
 
 (defun org-clone-local-variables (from-buffer &optional regexp)
   "Clone local variables from FROM-BUFFER.
@@ -8795,8 +8784,14 @@ call CMD."
   (org-load-modules-maybe)
   (unless org-local-vars
     (setq org-local-vars (org-get-local-variables)))
-  (eval (list 'let org-local-vars
-	      (list 'call-interactively (list 'quote cmd)))))
+  (let (symbols values)
+    (dolist (var org-local-vars)
+      (when (eq (symbol-value (car var))
+                (default-value (car var)))
+        (push (car var) symbols)
+        (push (cadr var) values)))
+    (progv symbols values
+      (call-interactively cmd))))
 
 ;;;; Archiving
 
@@ -22598,46 +22593,43 @@ clocking lines, and drawers."
     (point)))
 
 (defun org-forward-heading-same-level (arg &optional invisible-ok)
-  "Move forward to the arg'th subheading at same level as this one.
+  "Move forward to the ARG'th subheading at same level as this one.
 Stop at the first and last subheadings of a superior heading.
 Normally this only looks at visible headings, but when INVISIBLE-OK is
 non-nil it will also look at invisible ones."
   (interactive "p")
   (if (not (ignore-errors (org-back-to-heading invisible-ok)))
-      (outline-next-heading)
+      (if (and arg (< arg 0))
+          (goto-char (point-min))
+        (outline-next-heading))
     (org-at-heading-p)
-    (let* ((level (- (match-end 0) (match-beginning 0) 1))
-	   (re (format "^\\*\\{1,%d\\} " level))
-	   l)
-      (forward-char 1)
-      (while (> arg 0)
-	(while (and (re-search-forward re nil 'move)
-		    (setq l (- (match-end 0) (match-beginning 0) 1))
-		    (= l level)
-		    (not invisible-ok)
-		    (progn (backward-char 1) (outline-invisible-p)))
-	  (if (< l level) (setq arg 1)))
-	(setq arg (1- arg)))
-      (beginning-of-line 1))))
+    (let ((level (- (match-end 0) (match-beginning 0) 1))
+          (f (if (and arg (< arg 0))
+                 're-search-backward
+               're-search-forward))
+          (count (if arg (abs arg) 1))
+          (result (point)))
+      (forward-char (if (and arg (< arg 0)) -1 1))
+      (while (and (> count 0)
+                  (funcall f org-outline-regexp-bol nil 'move))
+        (let ((l (- (match-end 0) (match-beginning 0) 1)))
+          (cond ((< l level) (setq count 0))
+                ((and (= l level)
+                      (or invisible-ok
+                          (progn
+                            (goto-char (line-beginning-position))
+                            (not (outline-invisible-p)))))
+                 (setq count (1- count))
+                 (when (eq l level)
+                   (setq result (point)))))))
+      (goto-char result))
+    (beginning-of-line 1)))
 
 (defun org-backward-heading-same-level (arg &optional invisible-ok)
-  "Move backward to the arg'th subheading at same level as this one.
+  "Move backward to the ARG'th subheading at same level as this one.
 Stop at the first and last subheadings of a superior heading."
   (interactive "p")
-  (if (not (ignore-errors (org-back-to-heading)))
-      (goto-char (point-min))
-    (org-at-heading-p)
-    (let* ((level (- (match-end 0) (match-beginning 0) 1))
-	   (re (format "^\\*\\{1,%d\\} " level))
-	   l)
-      (while (> arg 0)
-	(while (and (re-search-backward re nil 'move)
-		    (setq l (- (match-end 0) (match-beginning 0) 1))
-		    (= l level)
-		    (not invisible-ok)
-		    (outline-invisible-p))
-	  (if (< l level) (setq arg 1)))
-	(setq arg (1- arg))))))
+  (org-forward-heading-same-level (if arg (- arg) -1) invisible-ok))
 
 (defun org-forward-element ()
   "Move forward by one element.

[-- Attachment #3: Type: text/plain, Size: 21 bytes --]


        Christopher

Re: roundtrip integration with Asana?

[Meng Weng Wong]

On 31 Jan, 2013, at 9:49 PM, Bastien <bzg@altern.org> wrote:
> 
> Meng Weng Wong <mengwong@gmail.com> writes:
>> http://developers.asana.com/documentation/
>> Roundtrip integration between org and Asana is therefore possible.
> 
> I don't know Asana but I'd suggest to have a look at org-sync
>  http://orgmode.org/cgit.cgi/org-sync.git/ 

Thank you. I have also learned of
https://github.com/christopherjwhite/org-toodledo

I will review the sync architecture of both packages if I get the chance.

For now I am taking the Perl route as we have over 400 tasks across the enterprise, so it's probably better to sync them in a separate background process. When the code is done I will publish it to CPAN and drop an announcement here.

thanks all