Hi all
This is a follow-up to
http://thread.gmane.org/gmane.emacs.orgmode/17581
On 09-10-14 14:24 , I wrote:
(global-set-key (kbd "C->") 'my-orgstruct-dive)
(global-set-key (kbd "C-<") 'my-orgstruct-dive-out)
(defvar my-orgstruct-dive-level 0)
Now I did a rewrite for some improvements, mainly to support dive in
and out of headings also in a variant that leaves the visibility of
siblings. The code is at the end.
Example 1: C-> C-> C-< C-> [...]
This changes only the visibility of of the current heading and leaves
the visibility of siblings, similar to Tab.
Example 2: C-u C-> C-> C-< C-< C-< C-> [...]
This changes the visibility of all headings, similar to S-Tab.
Michael
-----------------------------------------------------------
(global-set-key (kbd "C->") 'my-orgstruct-dive-in)
(global-set-key (kbd "C-<") 'my-orgstruct-dive-out)
(defun my-orgstruct-dive-in (arg)
"This func is very similar to the func my-orgstruct-dive-out. They are
wrappers to simplify comparison with last-command in my-orgstruct-dive.
Without prefix C-u: Dive in and out the current heading (relative) and leave
the visibility of the siblings, similar to Tab. Having the point before the
first heading in a buffer does not make sense.
Example: `C-> C-> C-< C-> [...]'
With prefix C-u: Dive in and out all headings in the buffer (absolute),
similar to S-Tab. If the point is before the first heading then the behavior
is like if it is on a top level heading.
Example: `C-u C-> C-> C-< C-< C-< C-> [...]'"
(interactive "P")
(my-orgstruct-dive arg))
(defun my-orgstruct-dive-out (arg)
"See the docstring of the very similar func my-orgstruct-dive-in."
(interactive "P")
(my-orgstruct-dive arg t))
(defvar my-orgstruct-dive-start-arg nil
"State variable of func my-orgstruct-dive: The prefix state of the first
invoke of my-orgstruct-dive-in or my-orgstruct-dive-out, see their
docstrings.")
(defvar my-orgstruct-dive-level 1
"State variable of func my-orgstruct-dive: The reduced level which is the
level of the heading and not necessarily the number of stars if oddeven.
The type of level is relative if my-orgstruct-dive-start-arg is nil, else
absolute. For both types of level the smallest possible value is 1.")
(defun my-orgstruct-dive (arg &optional dive-out)
"Make visible one heading level more or less.
return value: Undefined.
arg: See the docstring about prefix C-u in func my-orgstruct-dive-in
dive-out: If nil dive in, else dive out."
;; determine new level
(if (or arg ; let any intermediate `C-u' break the invoke repetition
(and (not (eq last-command 'my-orgstruct-dive-in))
(not (eq last-command 'my-orgstruct-dive-out))))
;; this is the first invoke of dive
(progn
;; move back to beginning of line with heading:
;; - to prepare the point location for possibly hitting some Tabs
;; afterwards
;; - to skip list elements, they are not supported do dive in
;; (this alone could also be covered within a save-excursion)
;; - for the func org-outline-level that requires beginning of line
;; (this alone could also be covered within a save-excursion)
(unless (org-before-first-heading-p) (outline-back-to-heading))
;; reset start arg and level
(setq my-orgstruct-dive-start-arg arg)
(setq my-orgstruct-dive-level
(if my-orgstruct-dive-start-arg
;; absolute level
(if (org-before-first-heading-p)
'1
(org-reduced-level (org-outline-level)))
;; relative level
'1)))
;; this is a repeated invoke of dive: increase/decrease level
(setq my-orgstruct-dive-level
(+ my-orgstruct-dive-level (if dive-out '-1 '1)))
(when (< my-orgstruct-dive-level 1) (setq my-orgstruct-dive-level 1)))
;; update heading visibility according to new level
(if my-orgstruct-dive-start-arg
;; absolute level
(org-shifttab my-orgstruct-dive-level) ; logs the level itself
;; relative level
(hide-subtree)
(show-children (* (org-level-increment) (1- my-orgstruct-dive-level)))
(message "Content view to relative level: %d" my-orgstruct-dive-level))
;; move back to beginning of line with heading to prepare
;; the point location for possibly hitting some Tabs afterwards
(unless (org-before-first-heading-p) (outline-back-to-heading)))
-----------------------------------------------------------
_______________________________________________
Emacs-orgmode mailing list
Please use `Reply All' to send replies to the list.
Emacs-orgmode@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-orgmode
