Babcore: Emacs Customizations everyone should have

Update (2017-05): babcore is at 0.2, but I cannot currently update the marmalade package.

C-g: abort

;; Copyright (C) 2013 Arne Babenhauserheide

;; Author: Arne Babenhauserheide (and various others in Emacswiki and elsewhere).
;; Maintainer: Arne Babenhauserheide
;; Created 03 April 2013
;; Version: 0.1.0
;; Version Keywords: core configuration

;; This program is free software; you can redistribute it and/or
;; modify it under the terms of the GNU General Public License
;; as published by the Free Software Foundation; either version 3
;; of the License, or (at your option) any later version.

;; This program is distributed in the hope that it will be useful,
;; but WITHOUT ANY WARRANTY; without even the implied warranty of
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
;; GNU General Public License for more details.

;; You should have received a copy of the GNU General Public License
;; along with this program. If not, see <http://www.gnu.org/licenses/ [19]>.

;;; Commentary:
;; Quick Start / installation:
;; 1. Download this file and put it next to other files Emacs includes
;; 2. Add this to you .emacs file and restart emacs:
;; 
;;      (require 'babcore)
;;
;; Alternatively install via package.el:
;;
;;      (require 'package)
;;      (add-to-list 'package-archives '("marmalade" . "http://marmalade-repo.org/packages/"))
;;      (package-refresh-contents)
;;      (package-install 'babcore)
;; 
;; Use Case: Use a common core configuration so you can avoid the
;;   tedious act of gathering all the basic stuff over the years and
;;   can instead concentrate on the really cool new stuff Emacs offers
;;   you.
;;

;;; Change Log:

;; 2016-06-05 - 0.1.0: replace desktop with better savehist config and
;;                     cleanup babcore. Replace flymake with flycheck.
;;                     Remove the eval-region key-chord. Simplify
;;                     x-urgent. Fix switching back from full-screen
;;                     mode. Remove babcore-shell-execute, since
;;                     async-shell-command (M-&) is a built-in which
;;                     does the job better. Add C-M-. as third alias
;;                     for goto-last-change. Add find-file-as-root and
;;                     a few fixes for encumbering behavior.
;; 2013-11-02 - Disable clipboard sync while exporting with org-mode
;;              org-export-dispatch
;; 2013-10-22 - More useful frame titles
;; 2013-04-03 - Minor adjustments
;; 2013-02-29 - Initial release

;;; Code:

;; Convenient package handling in emacs

(require 'package)
;; use packages from marmalade
(add-to-list 'package-archives '("marmalade" . "http://marmalade-repo.org/packages/"))
;; and the old elpa repo
(add-to-list 'package-archives '("elpa-old" . "http://tromey.com/elpa/"))
;; and automatically parsed versiontracking repositories.
(add-to-list 'package-archives '("melpa" . "http://melpa.milkbox.net/packages/"))

;; Make sure a package is installed
(defun package-require (package)
  "Install a PACKAGE unless it is already installed 
or a feature with the same name is already active.

Usage: (package-require 'package)"
  ; try to activate the package with at least version 0.
  (package-activate package '(0))
  ; try to just require the package. Maybe the user has it in his local config

  (condition-case nil
      (require package)
    ; if we cannot require it, it does not exist, yet. So install it.
    (error (progn
             (package-install package)
             (require package)))))

;; Initialize installed packages
(package-initialize)  
;; package init not needed, since it is done anyway in emacs 24 after reading the init
;; but we have to load the list of available packages, if it is not available, yet.
(when (not package-archive-contents)
  (with-timeout (15 (message "updating package lists failed due to timeout"))
    (package-refresh-contents)))

;; Flycheck: On the fly syntax checking
(package-require 'flycheck)
(add-hook 'after-init-hook #'global-flycheck-mode)
; stronger error display
(defface flycheck-error
  '((t (:foreground "red" :underline (:color "Red1" :style wave) :weight bold)))
  "Flycheck face for errors"
  :group "flycheck")

;; Inline auto completion and suggestions
(package-require 'auto-complete)
;; avoid competing with org-mode templates.
(add-hook 'org-mode-hook
          (lambda ()
            (make-local-variable 'ac-stop-words)
            (loop for template in org-structure-template-alist do
                  (add-to-list 'ac-stop-words 
                               (concat "<" (car template))))))

; use ido mode for file and buffer Completion when switching buffers
(require 'ido)
(ido-mode t)

;; Convenient printing
(require 'printing)
(pr-update-menus t)
; make sure we use localhost as cups server
(setenv "CUPS_SERVER" "localhost")
(package-require 'cups)

; use allout minor mode to have outlining everywhere.
(allout-mode)

; syntax highlighting everywhere
(global-font-lock-mode 1)

; Activate org-mode
(require 'org)
; and some more org stuff

; http://orgmode.org/guide/Activation.html#Activation

; The following lines are always needed.  Choose your own keys.
(add-to-list 'auto-mode-alist '("\\.org\\'" . org-mode))

; And add babel inline code execution
; babel, for executing code in org-mode.
(org-babel-do-load-languages
 'org-babel-load-languages
 ; load all language marked with (lang . t).
 '((C . t)
   (R . t)
   (asymptote)
   (awk)
   (calc)
   (clojure)
   (comint)
   (css)
   (ditaa . t)
   (dot . t)
   (emacs-lisp . t)
   (fortran)
   (gnuplot . t)
   (haskell)
   (io)
   (java)
   (js)
   (latex)
   (ledger)
   (lilypond)
   (lisp)
   (matlab)
   (maxima)
   (mscgen)
   (ocaml)
   (octave)
   (org . t)
   (perl)
   (picolisp)
   (plantuml)
   (python . t)
   (ref)
   (ruby)
   (sass)
   (scala)
   (scheme)
   (screen)
   (sh . t)
   (shen)
   (sql)
   (sqlite)))

; Add proper word wrapping
(global-visual-line-mode t)

; go to the last change
(package-require 'goto-chg)
(global-set-key [(control .)] 'goto-last-change)
; M-. can conflict with etags tag search. But C-. can get overwritten
; by flyspell-auto-correct-word. And goto-last-change needs a really
; fast key.
(global-set-key [(meta .)] 'goto-last-change)
; ensure that even in worst case some goto-last-change is available
(global-set-key [(control meta .)] 'goto-last-change)

(require 'flyspell)
; Make german umlauts work.
(setq locale-coding-system 'utf-8)
(set-terminal-coding-system 'utf-8)
(set-keyboard-coding-system 'utf-8)
(set-selection-coding-system 'utf-8)
(prefer-coding-system 'utf-8)

;aspell und flyspell
(setq-default ispell-program-name "aspell")

;make aspell faster but less correctly
(setq ispell-extra-args '("--sug-mode=ultra" "-w" "äöüÄÖÜßñ"))
(setq ispell-list-command "list")

; control-lock-mode, so we can enter a vi style command-mode with standard emacs keys.
(package-require 'control-lock)
; also bind M-ü and M-ä to toggling control lock.
(global-set-key (kbd "M-ü") 'control-lock-toggle)
(global-set-key (kbd "C-ü") 'control-lock-toggle)
(global-set-key (kbd "M-ä") 'control-lock-toggle)
(global-set-key (kbd "C-ä") 'control-lock-toggle)
(global-set-key (kbd "C-z") 'control-lock-toggle)

; use key chords invoke commands
(package-require 'key-chord)
(key-chord-mode 1)
; buffer actions
(key-chord-define-global "vb"     'eval-buffer)
(key-chord-define-global "cy"     'yank-pop)
(key-chord-define-global "cg"     "\C-c\C-c")
; frame actions
(key-chord-define-global "xo"     'other-window);
(key-chord-define-global "x1"     'delete-other-windows)
(key-chord-define-global "x0"     'delete-window)
(defun kill-this-buffer-if-not-modified ()
  (interactive)
  ; taken from menu-bar.el
  (if (menu-bar-non-minibuffer-window-p)
      (kill-buffer-if-not-modified (current-buffer))
    (abort-recursive-edit)))
(key-chord-define-global "xk"     'kill-this-buffer-if-not-modified)
; file actions
(key-chord-define-global "bf"     'ido-switch-buffer)
(key-chord-define-global "cf"     'ido-find-file)
(key-chord-define-global "vc"     'vc-next-action)

emacsclient -e "(show-frame)"

(defun show-frame (&optional frame)
  "Show the current Emacs frame or the FRAME given as argument.

And make sure that it really shows up!"
  (raise-frame)
  ; yes, you have to call this twice. Don’t ask me why…
  ; select-frame-set-input-focus calls x-focus-frame and does a bit of
  ; additional magic.
  (select-frame-set-input-focus (selected-frame))
  (select-frame-set-input-focus (selected-frame)))

;; let emacs blink when something interesting happens.
;; in KDE this marks the active Emacs icon in the tray.
(defun x-urgency-hint (frame arg &optional source)
  "Set the x-urgency hint for the frame to arg: 

- If arg is nil, unset the urgency.
- If arg is any other value, set the urgency.

If you unset the urgency, you still have to visit the frame to make the urgency setting disappear (at least in KDE)."
    (let* ((wm-hints (append (x-window-property 
                "WM_HINTS" frame "WM_HINTS" source nil t) nil))
     (flags (car wm-hints)))
    (setcar wm-hints
        (if arg
        (logior flags #x100)
          (logand flags (lognot #x100))))
    (x-change-window-property "WM_HINTS" wm-hints frame "WM_HINTS" 32 t)))

(defun x-urgent (&optional arg)
  "Mark the current emacs frame as requiring urgent attention. 

With a prefix argument which does not equal a boolean value of nil, remove the urgency flag (which might or might not change display, depending on the window manager)."
  (interactive "P")
  (let (frame (selected-frame))
  (x-urgency-hint frame (not arg))))

; fullscreen, taken from http://www.emacswiki.org/emacs/FullScreen#toc26
; should work for X und OSX with emacs 23.x (TODO find minimum version).
; for windows it uses (w32-send-sys-command #xf030) (#xf030 == 61488)
(defvar babcore-fullscreen-p nil
  "Check if fullscreen is on or off")
(defvar babcore-stored-frame-width nil 
  "width of the frame before going fullscreen")
(defvar babcore-stored-frame-height nil
  "width of the frame before going fullscreen")

(defun babcore-non-fullscreen ()
  (interactive)
  (if (fboundp 'w32-send-sys-command)
      ;; WM_SYSCOMMAND restore #xf120
      (w32-send-sys-command 61728)
    (progn (set-frame-parameter nil 'fullscreen nil)
           (set-frame-parameter nil 'width 
                                (if babcore-stored-frame-width
                                    babcore-stored-frame-width 82))
           (sleep-for 0 1) ; 1ms sleep: workaround to avoid unsetting the width in the next command
           (set-frame-parameter nil 'height
                                (if babcore-stored-frame-height 
                                    babcore-stored-frame-height 42)))))

(defun babcore-fullscreen ()
  (interactive)
  (setq babcore-stored-frame-width (frame-width))
  (setq babcore-stored-frame-height (frame-height))
  (if (fboundp 'w32-send-sys-command)
      ;; WM_SYSCOMMAND maximaze #xf030
      (w32-send-sys-command 61488)
    (set-frame-parameter nil 'fullscreen 'fullboth)))

(defun toggle-fullscreen ()
  (interactive)
  (setq babcore-fullscreen-p (not babcore-fullscreen-p))
  (if babcore-fullscreen-p
      (babcore-non-fullscreen)
    (babcore-fullscreen)))

(global-set-key [f11] 'toggle-fullscreen)

; Default KDE keybindings to make emacs nicer integrated into KDE. 

; can treat C-m as its own mapping.
; (define-key input-decode-map "\C-m" [?\C-1])

(defun revert-buffer-preserve-modes ()
  (interactive)
  (revert-buffer t nil t))

; C-m shows/hides the menu bar - thanks to http://stackoverflow.com/questions/2298811/how-to-turn-off-alternative-enter-with-ctrlm-in-linux
; f5 reloads
(defconst kde-default-keys-minor-mode-map
  (let ((map (make-sparse-keymap)))
    (set-keymap-parent map text-mode-map)
    (define-key map [f5] 'revert-buffer-preserve-modes)
    (define-key map [?\C-1] 'menu-bar-mode)
    (define-key map [?\C-+] 'text-scale-increase)
    (define-key map [?\C--] 'text-scale-decrease) ; shadows 'negative-argument which is also available via M-- and C-M--, though.
    (define-key map [C-kp-add] 'text-scale-increase)
    (define-key map [C-kp-subtract] 'text-scale-decrease)
    map)
  "Keymap for `kde-default-keys-minor-mode'.")

;; Minor mode for keypad control
(define-minor-mode kde-default-keys-minor-mode
  "Adds some default KDE keybindings"
  :global t
  :init-value t
  :lighter ""
  :keymap 'kde-default-keys-minor-mode-map
  )

;; Set the frame title as by http://www.emacswiki.org/emacs/FrameTitle
(setq frame-title-format (list "%b ☺ " (user-login-name) "@" (system-name) "%[ - GNU %F " emacs-version)
      icon-title-format (list "%b ☻ " (user-login-name) "@" (system-name) " - GNU %F " emacs-version))

M-x ucs-insert

;; Highlight TODO and FIXME in comments 
(package-require 'fic-ext-mode)
(defun add-something-to-mode-hooks (mode-list something)
  "helper function to add a callback to multiple hooks"
  (dolist (mode mode-list)
    (add-hook (intern (concat (symbol-name mode) "-mode-hook")) something)))

(add-something-to-mode-hooks '(c++ tcl emacs-lisp python text markdown latex) 'fic-ext-mode)

; save the current macro as reusable function.
(defun save-current-kbd-macro-to-dot-emacs (name)
  "Save the current macro as named function definition inside
your initialization file so you can reuse it anytime in the
future."
  (interactive "SSave Macro as: ")
  (name-last-kbd-macro name)
  (save-excursion 
    (find-file-literally user-init-file)
    (goto-char (point-max))
    (insert "\n\n;; Saved macro\n")
    (insert-kbd-macro name)
    (insert "\n")))

; Activate transparent GnuPG encryption.
(require 'epa-file)
(epa-file-enable)

(setq backup-by-copying t      ; don't clobber symlinks
      backup-directory-alist
      '(("." . "~/.local/share/emacs-saves"))    ; don't litter my fs tree
      delete-old-versions t
      kept-new-versions 6
      kept-old-versions 2
      version-control t)       ; use versioned backups

; save the place in files
(require 'saveplace)
(setq-default save-place t)

; save minibuffer history
(require 'savehist)
;; increase the default history cutoff
(setq history-length 500)
(savehist-mode t)
(setq savehist-additional-variables
      '(regexp-search-ring
        register-alist))

(setq x-select-enable-clipboard t)

; When I have x-select-enable-clipboard enabled, compiling an org-mode file to PDF locks
; KDE - I think it does so by filling up the clipboard.
(defadvice org-export-dispatch-no-clipboard-advice (around org-export-dispatch)
  "Do not clobber the system-clipboard while compiling an org-mode file with `org-export`."
  (let ((select-active-regions nil)
        (x-select-enable-clipboard nil)
        (x-select-enable-primary nil)
        (interprogram-cut-function nil)
        (interprogram-paste-function nil))
    ad-do-it))
(ad-activate 'org-export-dispatch-no-clipboard-advice t)

(package-require 'legalese)

;;; Open files as root - quickly
(defcustom find-file-root-prefix "/sudo:root@localhost:"
"Tramp root prefix to use.")

(defun find-file-as-root ()
  "Like `ido-find-file, but automatically edit the file with
root-privileges (using tramp/sudo), if the file is not writable by
user."
  (interactive)
  (let ((file (ido-read-file-name "Edit as root: ")))
    (unless (file-writable-p file)
      (setq file (concat find-file-root-prefix file)))
    (find-file file)))
;; or some other keybinding...
;; (global-set-key (kbd "C-x F") 'djcb-find-file-as-root)

(defun find-current-as-root ()
  "Reopen current file as root"
  (interactive)
  (set-visited-file-name (concat find-file-root-prefix (buffer-file-name)))
  (setq buffer-read-only nil))

;;;;;;;;;;;;;
;;; Fixes ;;;
;;;;;;;;;;;;;

;; Make comint recognize passwords in virtually all languages.
(defcustom comint-password-prompt-regexp
  (concat
   "\\(^ *\\|"
   (regexp-opt
    '("Enter" "enter" "Enter same" "enter same" "Enter the" "enter the"
      "Old" "old" "New" "new" "'s" "login"
      "Kerberos" "CVS" "UNIX" " SMB" "LDAP" "[sudo]" "Repeat" "Bad") t)
   " +\\)"
   (regexp-opt
    '("Adgangskode" "adgangskode" "Contrasenya" "contrasenya" "Contraseña" "contraseña" "Geslo" "geslo" "Hasło" "hasło" "Heslo" "heslo" "Iphasiwedi" "iphasiwedi" "Jelszó" "jelszó" "Lozinka" "lozinka" "Lösenord" "lösenord" "Mot de passe " "Mot de Passe " "mot de Passe " "mot de passe " "Mật khẩu " "mật khẩu" "Parola" "parola" "Pasahitza" "pasahitza" "Pass phrase" "pass Phrase" "pass phrase" "Passord" "passord" "Passphrase" "passphrase" "Password" "password" "Passwort" "passwort" "Pasvorto" "pasvorto" "Response" "response" "Salasana" "salasana" "Senha" "senha" "Wachtwoord" "wachtwoord" "slaptažodis" "slaptažodis" "Лозинка" "лозинка" "Пароль" "пароль" "ססמה" "كلمة السر" "गुप्तशब्द" "शब्दकूट" "গুপ্তশব্দ" "পাসওয়ার্ড" "ਪਾਸਵਰਡ" "પાસવર્ડ" "ପ୍ରବେଶ ସଙ୍କେତ" "கடவுச்சொல்" "సంకేతపదము" "ಗುಪ್ತಪದ" "അടയാളവാക്ക്" "රහස්පදය" "ពាក្យសម្ងាត់ ៖ " "パスワード" "密码" "密碼" "암호"))
   "\\(?:\\(?:, try\\)? *again\\| (empty for no passphrase)\\| (again)\\)?\
\\(?: for [^:]+\\)?:\\s *\\'")
  "Regexp matching prompts for passwords in the inferior process.
This is used by `comint-watch-for-password-prompt'."
  :version "24.3"
  :type 'regexp
  :group 'comint)

;; Mark all AC_* and AS_* functions as builtin.
(add-hook 'autoconf-mode-hook 
          (lambda () 
            (add-to-list 'autoconf-font-lock-keywords '("\\(\\(AC\\|AS\\|AM\\)_.+?\\)\\((\\|\n\\)" (1 font-lock-builtin-face)))))

; tell emacs to ignore alt-gr clicks needed for M4 in the Neo Layout.
(define-key special-event-map (kbd "<key-17>") 'ignore)
(define-key special-event-map (kbd "<M-key-17>") 'ignore)

; yank-pop should yank if the last command was no yank.
(defun yank-pop (&optional arg)
  "Replace just-yanked stretch of killed text with a different stretch.
At such a time, the region contains a stretch of reinserted
previously-killed text.  `yank-pop' deletes that text and inserts in its
place a different stretch of killed text.

With no argument, the previous kill is inserted.
With argument N, insert the Nth previous kill.
If N is negative, this is a more recent kill.

The sequence of kills wraps around, so that after the oldest one
comes the newest one.

When this command inserts killed text into the buffer, it honors
`yank-excluded-properties' and `yank-handler' as described in the
doc string for `insert-for-yank-1', which see."
  (interactive "*p")
  (if (not (eq last-command 'yank))
      (yank)
    (setq this-command 'yank)
    (unless arg (setq arg 1))
    (let ((inhibit-read-only t)
          (before (< (point) (mark t))))
      (if before
          (funcall (or yank-undo-function 'delete-region) (point) (mark t))
        (funcall (or yank-undo-function 'delete-region) (mark t) (point)))
      (setq yank-undo-function nil)
      (set-marker (mark-marker) (point) (current-buffer))
      (insert-for-yank (current-kill arg))
      ;; Set the window start back where it was in the yank command,
      ;; if possible.
      (set-window-start (selected-window) yank-window-start t)
      (if before
          ;; This is like exchange-point-and-mark, but doesn't activate the mark.
          ;; It is cleaner to avoid activation, even though the command
          ;; loop would deactivate the mark because we inserted text.
          (goto-char (prog1 (mark t)
                       (set-marker (mark-marker) (point) (current-buffer))))))
    nil))

(setq visible-bell t)

(provide 'babcore)
;;; babcore.el ends here

Note: As almost everything on this page, this text and code is available under the GPLv3 or later.

Conveniently convert CamelCase to words_with_underscores using a small emacs hack

I currently cope with refactoring in an upstream project to which I maintain some changes which upstream does not merge. One nasty part is that the project converted from CamelCase for function names to words_with_underscores. And that created lots of merge errors.

Today I finally decided to speed up my work.

The first thing I needed was a function to convert a string in CamelCase to words_with_underscores. Since I’m lazy, I used google, and that turned up the CamelCase page of Emacswiki [28] - and with it the following string functions:

(defun split-name (s)
  (split-string
   (let ((case-fold-search nil))
     (downcase
      (replace-regexp-in-string "\\([a-z]\\)\\([A-Z]\\)" "\\1 \\2" s)))
   "[^A-Za-z0-9]+"))
(defun underscore-string (s) (mapconcat 'downcase   (split-name s) "_"))

Quite handy - and elegantly executed. Now I just need to make this available for interactive use. For this, Emacs Lisp offers many useful ways to turn Editor information into program information, called interactive codes [29] - in my case the region-code: "r". This gives the function the beginning and the end of the currently selected region as arguments.

With this, I created an interactive function which de-camelcases and underscores the selected region:

(defun underscore-region (begin end) (interactive "r")
  (let* ((word (buffer-substring begin end))
         (underscored (underscore-string word)))
    (save-excursion
      (widen) ; break out of the subregion so we can fix every usage of the function
      (replace-string word underscored nil (point-min) (point-max)))))

And now we’re almost there. Just create a macro which searches for a function, selects its name, de-camelcaeses and underscores it and then replaces every usage of the CamelCase name by the underscored name. This isn’t perfect refactoring (can lead to errors), but it’s fast and I see every change it does.

C-x C-(
C-s def 
M-x mark-word
M-x underscore-region
C-x C-)

That’s it, now just call the macro repeatedly.

C-x eeeeee…

Now check the diff to fix where this 13 lines hack got something wrong ( like changing __init__ into _init_ - I won’t debug this, you’ve been warned ☺).

Happy Hacking!

Custom link completion for org-mode in 25 lines (emacs)

Update (2013-01-23): The new org-mode removed (org-make-link), so I replaced it with (concat) and uploaded a new example-file: org-custom-link-completion.el [31].
Happy Hacking!

1 Intro

I recently set up custom completion for two of my custom link types in Emacs [32] org-mode [5]. When I wrote on identi.ca about that [33], Greg Tucker-Kellog said that he’d like to see that [34]. So I decided, I’d publish my code.

The link types I regularly need are papers (PDFs of research papers I take notes about) and bib (the bibtex entries for the papers). The following are my custom link definitions :

(setq org-link-abbrev-alist
      '(("bib" . "~/Dokumente/Uni/Doktorarbeit-inverse-co2-ch4/aufschriebe/ref.bib::%s")
       ("notes" . "~/Dokumente/Uni/Doktorarbeit-inverse-co2-ch4/aufschriebe/papers.org::#%s")
       ("papers" . "~/Dokumente/Uni/Doktorarbeit-inverse-co2-ch4/aufschriebe/papers/%s.pdf")))

For some weeks I had copied the info into the links by hand. Thus an entry about a paper looks like the following.

* Title [[bib:identifier]] [[papers:name_without_suffix]]

This already suffices to be able to click the links for opening the PDF or showing the bibtex entry. Entering the links was quite inconvenient, though.

(defun string-replace (this withthat in)
  "replace THIS with WITHTHAT' in the string IN"
  (with-temp-buffer
    (insert in)
    (goto-char (point-min))
    (replace-string this withthat)
    (buffer-substring (point-min) (point-max))))

<<string-replace>>
(defun org-papers-complete-link (&optional arg)
  "Create a papers link using completion."
  (let (file link)
       (setq file (read-file-name "papers: " "papers/"))
       <<cleanup-link>>
    link))

(let ((pwd (file-name-as-directory (expand-file-name ".")))
  (pwd1 (file-name-as-directory (abbreviate-file-name
                 (expand-file-name ".")))))
  (setq file (string-replace "papers/" "" file))
  (setq file (string-replace pwd "" (string-replace pwd1 "" file)))
  (setq file (string-replace ".pdf" "" file))
  (setq link (concat "papers:" file)))

<<reftex-setup>>
(defun org-bib-complete-link (&optional arg)
  "Create a bibtex link using reftex autocompletion."
  (concat "bib:" (reftex-do-citation nil t nil)))

(defun org-mode-reftex-setup ()
  (interactive)
  (and (buffer-file-name) (file-exists-p (buffer-file-name))
       (progn
        ; Reftex should use the org file as master file. See C-h v TeX-master for infos.
        (setq TeX-master t)
        (turn-on-reftex)
        ; don’t ask for the tex master on every start.
        (reftex-parse-all)
        ;add a custom reftex cite format to insert links
        (reftex-set-cite-format
         '((?b . "[[bib:%l][%l-bib]]")
           (?n . "[[notes:%l][%l-notes]]")
           (?p . "[[papers:%l][%l-paper]]")
           (?t . "%t")
           (?h . "** %t\n:PROPERTIES:\n:Custom_ID: %l\n:END:\n[[papers:%l][%l-paper]]")))))
  (define-key org-mode-map (kbd "C-c )") 'reftex-citation)
  (define-key org-mode-map (kbd "C-c (") 'org-mode-reftex-search))

(add-hook 'org-mode-hook 'org-mode-reftex-setup)

<<org-papers-complete-link>>
<<org-bib-complete-link>>

Easily converting ris-citations to bibtex with emacs and bibutils

The problem

Nature [40] only gives me ris-formatted [41] citations, but I use bibtex [42].

Also ris is far from human readable.

The background

ris can be reformatted to bibtext, but doing that manually disturbs my workflow when getting references while taking note about a paper in emacs.

I tend to search online for references, often just using google scholar [43], so when I find a ris reference, the first data I get for the ris-citation is a link.

The solution

Making it possible

bibutils [44]¹ can convert ris to an intermediate xml format and then convert that to bibtex.

wget -O reference.ris RIS_URL
cat reference.ris | ris2xml | xml2bib >> ref.bib

This solves the problem, but it is not convenient, because I have to switch to the terminal, download the file, convert it and append the result to my bibtex file.

Making it convenient

With the first step, getting the ris-citation is quite inconvenient. I need 3 steps just for getting a citation.

But those steps are always the same, and since I use Emacs [45], I can automate and integrate them very easily. So I created a simple function in emacs, which takes the url of a ris citation, converts it to bibtex and appends the result to my local bibtex file. Now I get a ris citation with a simple call to

M-x ris-citation-to-bib

Then I enter the url and the function appends the citation to my bibtex file.²

Feel free to integrate it into your own emacs setup (additionally to the GPLv3 [46] you can use any license used by emacswiki [3] or worg [47]).

(defun ris-citation-to-bib (&optional ris-url) 
  "get a ris citation as bibtex in one step. Just call M-x
ris-citation-to-bib and enter the ris url. 
Requires bibutils: http://sourceforge.net/p/bibutils/home/Bibutils/ 
"
  (interactive "Mris-url: ")
  (save-excursion
    (let ((bib-file "/home/arne/aufschriebe/ref.bib")
          (bib-buffer (get-buffer "ref.bib"))
          (ris-buffer (url-retrieve-synchronously ris-url)))
      ; firstoff check if we have the bib buffer. If yes, move point to the last line.
      (if (not (member bib-buffer (buffer-list)))
          (setq bib-buffer (find-file-noselect bib-file)))
      (progn 
        (set-buffer bib-buffer)
        (goto-char (point-max)))
      (if ris-buffer
          (set-buffer ris-buffer))
      (shell-command-on-region (point-min) (point-max) "ris2xml | xml2bib" ris-buffer)
      (let ((pmin (- (search-forward "@") 1))
            (pmax (search-forward "}

")))
        (if (member bib-buffer (buffer-list))
            (progn 
              (append-to-buffer bib-buffer pmin pmax)
              (kill-buffer ris-buffer)
              (set-buffer bib-buffer)
              (save-buffer)
          ))))))

Happy Hacking!

PS: When I don’t have the URL (many thanks to journals giving me only a download button), I open the file, select the content and hit M-| (shell-command-on-region) with ris2xml | xml2bib (searching backwards via C-r ris so I to avoid typing the exact command) and get the bibtex version in the results buffer.

El Kanban Org: parse org-mode todo-states to use org-tables as Kanban tables

Kanban [48] for emacs [32] org-mode [49].

Update (2020): Kanban moved to sourcehut: https://hg.sr.ht/~arnebab/kanban.el [50]

Update (2013-04-13): Kanban.el now lives in its own repository: on bitbucket [51] and on a statically served http-repo [52] (to be independent from unfree software).

Update (2013-04-10): Thanks to Han Duply, kanban links now work for entries from other files. And I uploaded kanban.el on marmalade [53].

Some time ago I learned about kanban, and the obvious next step was: “I want to have a kanban board from org-mode”. I searched for it, but did not find any. Not wanting to give up on the idea, I implemented my own :)

The result are two functions: kanban-todo and kanban-zero.

“Screenshot” :)

TODO	DOING	DONE
Refactor in such a way that the
	let Presentation manage dumb sprites
return all actions on every command:
Make the UiState adhere the list of
Turn the model into a pure state

kanban-todo

kanban-todo provides your TODO items as kanban-fields. You can move them in the table without having duplicates, so all the state maintenance is done in the kanban table. Once you are finished, you mark them as done and delete them from the table.

To set it up, put kanban.el [54] somewhere in your load path and (require 'kanban) (more recent but potentially unstable version [55]). Then just add a table like the following:

|   |   |   |
|---+---+---|
|   |   |   |
|   |   |   |
|   |   |   |
|   |   |   |
#+TBLFM: $1='(kanban-todo @# @2$2..@>$>)::@1='(kanban-headers $#)

Click C-c C-c with the point on the TBLFMT line to update the table.

The important line is the #+TBLFM. That says “use my TODO items in the TODO column, except if they are in another column” and “add kanban headers for my TODO states”

The kanban-todo function takes an optional parameter match, which you can use to restrict the kanban table to given tags. The syntax is the same as for org-mode matchers [56]. The third argument allows you to provide a scope [57], for example a list of files.

To only set the scope, use nil for the matcher.

See C-h f org-map-entries and C-h v org-agenda-files for details.

kanban-zero

kanban-zero is a zero-state Kanban: All state is managed in org-mode and the table only displays the kanban items.

To set it up, put kanban.el [54] somwhere in your load path and (require 'kanban). Then just add a table like the following:

|   |   |   |
|---+---+---|
|   |   |   |
|   |   |   |
|   |   |   |
|   |   |   |
#+TBLFM: @2$1..@>$>='(kanban-zero @# $#)::@1='(kanban-headers $#)

The important line is the #+TBLFM. That says “show my org items in the appropriate column” and “add kanban headers for my TODO states”.

Click C-c C-c with the point on the TBLFMT line to update the table.

The kanban-zero function takes an optional parameter match, which you can use to restrict the kanban table to given tags. The syntax is the same as for org-mode matchers [56]. The third argument allows you to provide a scope [57], for example a list of files.

To only set the scope, use nil for the matcher.

An example for matcher and scope would be:

#+TBLFM: @2$1..@>$>='(kanban-zero @# $# "1w6" '("/home/arne/.emacs.d/private/org/emacs-plan.org"))::@1='(kanban-headers $#)

See C-h f org-map-entries and C-h v org-agenda-files for details.

Contribute

To contribute to kanban.el, just change the file and write a comment about your changes. Maybe I’ll setup a repo on Bitbucket at some point…

Example

In the Hexbattle [58] game-draft, I use kanban to track my progress:

Table of Contents

• 1 Kanban
• 2 refactor Hexbattle
  • 2.1 Structure plan:
  • 2.2 First all logic from UiState into the model. Model gets all state which is independent of the UI. Any necessary interaction in UiState calls model methods. √
  • 2.3 Refactor in such a way that the UiState and (even more importantly) the input do not need to call the model directly anymore.
    • 2.3.1 let Presentation manage dumb sprites for all chars which are active. Then I can remove a lot of UiState dependent functionality from the chars.
    • 2.3.2 return all actions on every command: command(…) → commands (instance of Commands).
  • 2.4 Make the UiState adhere the list of possible commands.
  • 2.5 Turn the model into a pure state machine.

1 Kanban

❢	STARTED	✔
Refactor in such a way that the
	let Presentation manage dumb sprites
return all actions on every command:
Make the UiState adhere the list of
Turn the model into a pure state

2 refactor Hexbattle    1w6

… and so on …

Advanced usage

“Graphical” TODO states

To make the todo states easier to grok directly you can use unicode symbols for them. Example:

#+SEQ_TODO: ❢ ☯ ⧖ | ☺ ✔ DEFERRED ✘

| ❢ | ☯ | ⧖ | ☺ |
|---+---+---+---|
|   |   |   |   |
#+TBLFM: @1='(kanban-headers $#)::@2$1..@>$>='(kanban-zero @# $#)

In my setup they are ❢ (todo) ☯ (doing) ⧖ (waiting) and ☺ (to report). Not shown in the kanban Table are ✔ (finished), ✘ (dropped) and deferred (later), because they don’t require any action from me, so I don’t need to see them all the time.

Collecting kanban entries via SSH

If you want to create a shared kanban table, you can use the excellent transparent network access options from Emacs tramp to collect kanban entries directly via SSH.

To use that, simply pass an explicit list of files to kanban-zero as 4th argument (if you don’t use tag matching just use nil as 3rd argument). "/ssh:host:path/to/file.org" retrieves the file ~/path/to/file.org from the host.

| ❢ | ☯ |
|---+---|
|   |   |
#+TBLFM: @1='(kanban-headers $#)::@2$1..@>$>='(kanban-zero @# $# nil (list (buffer-file-name) "/ssh:localhost:plan.org"))

Caveeat: all included kanban files have to use at least some of the same todo states: kanban.el only retrieves TODO states which are used in the current buffer.

How to show the abstract before the table of contents in org-mode

#+OPTIONS: toc:nil

#+BEGIN_ABSTRACT
Abstract
#+END_ABSTRACT
#+TOC: headlines 2

# Local Variables:
# org-export-with-toc: nil
# End:

#+BEGIN_ABSTRACT
Abstract
#+END_ABSTRACT
#+LATEX: \tableofcontents

IRC-chat via Tor with Emacs on Gentoo

As example: Connecting to #youbroketheinternet.

emerge privoxy torsocks net-vpn/tor
# rc-config start privoxy tor
# rc-update add privoxy default
# rc-update add tor default
mkdir -p ~/.local/EMACS_TOR_HOME/.emacs.d
echo "(require 'socks)" >> ~/.local/EMACS_TOR_HOME/.emacs.d/init.el
HOME=~/.local/EMACS_TOR_HOME torify emacs --title "Emacs-torified"
# M-x customize-variable RET socks-server RET
#   host: localhost
#   port: 9050
#   type: Socks v5
#   (C-x C-s to save and set)
# M-x erc-select
#   server loupsycedyglgamf.onion
#   port 67
# the welcome channel is good to go.

See https://www.emacswiki.org/emacs/ErcProxy#toc2 [63]

and http://youbroketheinternet.org/#overlay [64]

Insert a scaled screenshot in emacs org-mode

@marjoleink asked on identi.ca [65]¹, if it is possible to use emacs org-mode for showing scaled screenshots inline while writing. Since I thought I’d enjoy some hacking, I decided to take the challenge.

It does not do auto-scaling of embedded images, as far as I know, but the use case of screenshots can be done with a simple function (add this to your ~/.emacs or ~/.emacs.d/init.el):

(defun org-insert-scaled-screenshot ()
  "Insert a scaled screenshot 
for inline display 
into your org-mode buffer."
  (interactive)
  (let ((filename 
         (concat "screenshot-" 
                 (substring 
                  (shell-command-to-string 
                   "date +%Y%m%d%H%M%S")
                  0 -1 )
                 ".png")))
    (let ((scaledname 
           (concat filename "-width300.png")))

      (shell-command 
       (concat "import -window root " 
               filename))
      (shell-command 
       (concat "convert -adaptive-resize 300 " 
               filename " " scaledname))
      (insert (concat "[[./" scaledname "]]")))))

Now just call M-x org-redisplay-inline-images to see the screenshot (or add it to the function).

In action:

Have fun with Emacs - and happy hacking!

PS: In case it’s not obvious: The screenshot shows emacs just as the screenshot is being shot - with the method shown here ☺)

Minimal example for literate programming with noweb in emacs org-mode

If you want to use the literate programming features in emacs org-mode, you can try this minimal example to get started: Activate org-babel-tangle [70], then put this into the file noweb-test.org:

Minimal example for noweb in org-mode

* Assign 

First we assign abc:

#+begin_src python :noweb-ref assign_abc
abc = "abc"
#+end_src

* Use

Then we use it in a function:

#+begin_src python :noweb tangle :tangle noweb-test.py
def x():
  <<assign_abc>>
  return abc

print(x())
#+end_src

noweb-test.org [71]

Hit C-c C-c to evaluate the source block. Hit C-c C-v C-t to put the expanded code into the file noweb-test.py.

The exported code looks like this:

def x():
  abc = "abc"
  return abc

print(x())

noweb-test.py [72]

(html generated with org-export-as-html-to-buffer and slightly reniced to escape the additional parsing I have on my site)

And with org-export-as-pdf we get this:

[73]

noweb-test.pdf [73]

Add :results output to the #+begin_src line of the second block to see the print results under that block when you hit C-c C-c in the block.

You can also use properties of headlines for giving the noweb-ref. Org-mode can then even concatenate several source blocks into one noweb reference. Just hit C-c C-x p to set a property (or use M-x org-set-property), then set noweb-ref to the name you want to use to embed all blocks under this heading together.

Note: org-babel prefixes each line of an included code-block with the prefix used for the reference (here <<assign_abc>>). This way you can easily include blocks inside python [74] functions.

Note: To keep noweb-references literally in the output or similar, have a look at the different options to :noweb [75].

Note: To do this with shell-code, it’s useful to change the noweb markers to {{{ and }}}, because << and >> are valid shell-syntax, so they disturb the highlighting in sh-mode. Also confirming the evaluation every time makes plain exporting problematic. To fix this, just add the following somewhere in the file (to keep this simple, just add it to the end):

# Local Variables:
# org-babel-noweb-wrap-start: "{{{"
# org-babel-noweb-wrap-end: "}}}"
# org-confirm-babel-evaluate: nil
# org-export-allow-bind-keywords: t
# End:

Have fun with Emacs [32] and org-mode [5]!

Org-mode with Parallel Babel

Update 2017: a block with sem -j ... seems to block in recent versions of Emacs until all subtasks are done. It would be great if someone could figure out why (though it likely is the right thing to do). To circumvent that, you can daemonize the job in sem, but that might have unwanted side-effects: sem "[job] &"

; Execute babel source blocks asynchronously by just opening a new emacs.
(defun bab/org-babel-execute-src-block-new-emacs ()
  "Execute the current source block in a separate emacs,
so we do not block the current emacs."
  (interactive)
  (let ((line (line-number-at-pos))
        (file (buffer-file-name)))
    (async-shell-command (concat 
                          "TERM=vt200 emacs -nw --find-file " 
                          file 
                          " --eval '(goto-line "
                          (number-to-string line) 
                          ")' --eval "
     "'(let ((org-confirm-babel-evaluate nil))(org-babel-execute-src-block t))' "
                          "--eval '(kill-emacs 0)'"))))

; Export as pdf asynchronously by just opening a new emacs.
(defun bab/org-beamer-export-new-emacs ()
  "Export the current file in a separate emacs,
so we do not block the current emacs."
  (interactive)
  (let ((line (line-number-at-pos))
        (file (buffer-file-name)))
    (async-shell-command (concat 
                          "TERM=vt200 emacs -nw --find-file " 
                          file 
                          " --eval '(goto-line " 
                          (number-to-string line) 
                          ")' --eval "
     "'(let ((org-confirm-babel-evaluate nil))(org-beamer-export-to-pdf))' "
                          "--eval '(kill-emacs 0)'"))))

#+BEGIN_SRC sh :exports none
  oldPWD=$(pwd)
  cd ~/tm5tools/plotting
  filename="./obsheat-increasing.png" >/dev/null 2>/dev/null
  sem -j -1 ./plotstation.py -c ~/sun-work/ct-production-out-5x7e300m1.0 -C "aircraft" -c ~/sun-work/ct-production-out-5x7e300m1.0no-aircraft -C "continuous"  --obsheat --station allnoaa --title "\"Reducing observation coverage\"" -o ${oldPWD}/${filename}
  cd -
#+END_SRC

for i in {1..100}; do 
    sem -j -1 [code] # run N-1 processes with N as the number of
                     # pocessors in my computer
done

sem --wait

sem -j -1 ./plotstation.py --title "\"Reducing observation coverage\""

(setq org-src-fontify-natively t)

sem -j -1 [escaped code]

Publish a single file with emacs org-mode

(setq private-publish-ftp-proj (concat "/ftp:" USER "@" HOST ":arnebab/proj/"))

(setq org-publish-project-alist
      '(("orgmode-single-file"
         :base-directory "~/.emacs.d/private/journal"
         :publishing-directory (concat private-publish-ftp-proj "orgmode-single-file/")
         :base-extension "org"
         :publishing-function org-html-publish-to-html
         :completion-function (lambda () (rename-file 
                                          (concat private-publish-ftp-proj 
                                                  "orgmode-single-file/2013-11-25-Mo-publish-single-file-org-mode.html") 
                                          (concat private-publish-ftp-proj 
                                                  "orgmode-single-file/index.html") t))
         :section-numbers nil
         :with-toc t
         :html-preamble t
         :exclude ".*"
         :include ["2013-11-25-Mo-publish-single-file-org-mode.org"])))

Read your python module documentation from emacs

Update 2021: Fixed links that died with Bitbuckets hosting.

I just found the excellent pydoc-info mode [82] for emacs [32] from Jon Waltman. It allows me to hit C-h S in a python file and enter a module name to see the documentation right away. If the point is on a symbol (=module or class or function), I can just hit enter to see its docs.

In its default configuration (see the Readme [83]) it “only” reads the python documentation. This alone is really cool when writing new python code, but it s not enough, since I often use third party modules.

And now comes the treat: If those modules use sphinx [84] for documentation (≥1.1), I can integrate them just like the standard python documentation!

It took me some time to get it right, but now I have all the documentation for the inverse modelling framework I contribute to directly at my fingertips: Just hit C-h S ENTER when I’m on some symbol and a window shows me the docs:

The text in this image is from Wouter Peters. Used here as short citation which should be legal almost everywhere under citation rules.

I want to save you the work of figuring out how to do that yourself, so here’s a short guide for integrating the documentation for your python program into emacs.

Integrating your own documentation into emacs

The prerequisite for integrating your own documentation is to use sphinx [85] for documenting your code. See their tutorial [86] for info how to set it up. As soon as sphinx works for you, follow this guide to integrate your docs in your emacs.

Install pydoc-info

First get pydoc-info and the python infofile (adapt this to your local setup):

# get the mode
cd ~/.emacs.d/libs
hg clone https://hg.sr.ht/~arnebab/pydoc-info
# and the pregenerated info-file for python
wget http://www.draketo.de/dateien/python.info.gz
gunzip python.info.gz
sudo cp python.info /usr/share/info
sudo install-info --info-dir=/usr/share/info python.info

To build the info file for python yourself, have a look at the Readme [83].

Turn your documentation into info

Now turn your own documentation into an info document and install it.

Sphinx uses a core configuration file named conf.py. Add the following to that file, replacing all values but index and False by the appropriate names for you project:

# One entry per manual page. 
# list of tuples (startdocname, 
# targetname, title, author, dir_entry, 
# description, category, toctree_only).
texinfo_documents = [
  ('index', # startdocname, keep this!
   'TARGETNAME', # targetname
   u'Long Title', # title
   u'Author Name', # author
   'Name in the Directory Index of Info', # dir_entry
   u'Long Description', # description
   'Software Development', # cathegory
   False), # better keep this, too, i think.
]

Then call sphinx and install the info files like this (maybe adapted to your local setup):

sphinx-build -b texinfo source/ texinfo/ 
cd texinfo
sudo install-info --info-dir=/usr/share/info TARGETNAME.info
sudo cp TARGETNAME.info /usr/share/info/

Activate pydoc-info, including your documentation

Finally add the following to your .emacs (or wherever you store your personal adaptions):

; Show python-documentation as info-pages via C-h S
(setq load-path (cons "~/.emacs.d/libs/pydoc-info" load-path))
(require 'pydoc-info)
(info-lookup-add-help
   :mode 'python-mode
   :parse-rule 'pydoc-info-python-symbol-at-point
   :doc-spec
   '(("(python)Index" pydoc-info-lookup-transform-entry)
     ("(TARGETNAME)Index" pydoc-info-lookup-transform-entry)))

Recipes for presentations with beamer latex using emacs org-mode

I wrote some recipes for creating the kinds of slides I need with emacs org-mode export to beamer latex.

Update: Read ox-beamer [89] to see how to adapt this to work with the new export engine in org-mode 0.8.

[90] The recipes as PDF (21 slides, 247 KiB) [90]

[91] The org-mode sources (12.2 KiB) [91]

Below is an html export of the org-mode file. Naturally it does not look as impressive as the real slides, but it captures all the sources, so I think it has some value.

Note: To be able to use the simple block-creation commands, you need to add #+startup: beamer to the header of your file or explicitely activate org-beamer with M-x org-beamer-mode.

PS: I hereby allow use of these slides under any of the licenses used by worg [47] and/or the emacs wiki [3].

Hello World

#+LaTeX_CLASS: beamer
#+BEAMER_FRAME_LEVEL: 2

* Hello
** Hello GNU
Nice to see you!

Recipes
#+LaTeX_CLASS: beamer
#+BEAMER_FRAME_LEVEL: 2
* Introduction
** org-mode + beamer =  love
*** Code :BMCOL:
    :PROPERTIES:
    :BEAMER_col: 0.7
    :END:
<example block>
*** Simple block  :BMCOL:B_block:
    :PROPERTIES:
    :BEAMER_col: 0.3
    :BEAMER_env: block
    :END:
it's that easy!

** Two columns - in commands
*** Commands
C-c C-b | 0.7
C-c C-b b
C-n
<eTAB (write example) C-n C-n
*** Result
C-c C-b | 0.3
C-c C-b b
even easier - and faster!

*** Column 1 :B_ignoreheading:BMCOL:
    :PROPERTIES:
    :BEAMER_env: ignoreheading
    :BEAMER_col: 0.5
    :END:

*** One
*** Three                                                           

*** Column 2 :BMCOL:B_ignoreheading:
    :PROPERTIES:
    :BEAMER_col: 0.5
    :BEAMER_env: ignoreheading
    :END:

*** Two
*** Four

*** 
C-c C-b | 0.5 # column
C-c C-b i # ignore heading
*** One 
C-c C-b b # block
*** Three 
C-c C-b b
*** 
C-c C-b | 0.5
C-c C-b i
*** Two 
C-c C-b b
*** Four 
C-c C-b b

*** Code                                                      :B_block:BMCOL:
    :PROPERTIES:
    :BEAMER_env: block
    :BEAMER_col: 0.5
    :BEAMER_envargs: C[t]
    :END:

*** Result                                                    :B_block:BMCOL:
    :PROPERTIES:
    :BEAMER_env: block
    :BEAMER_col: 0.5
    :END:
pretty nice!

#+begin_quote
Emacs org-mode is a 
great presentation tool - 
Fast to beautiful slides.
- Arne Babenhauserheide
#+end_quote

\( 1 + 2 = 3 \) is clear

\[ 1 + 2 \cdot 3 = 7 \]

\( \LaTeX \) gives a space 
after math mode.

\LaTeX{} does it, too.

\LaTeX does not.

At the end of a sentence 
both work.
Try \LaTeX. Or try \LaTeX{}.

Only \( \LaTeX \) and \( \LaTeX{} \) 
also work with HTML export.

: #+bla: foo
: * Example Header

#+begin_example
content
#+end_example

#+bla: foo
* Example Header

content

<Title>

#+startup: beamer
#+LaTeX_CLASS: beamer
#+LaTeX_CLASS_OPTIONS: [bigger]
#+AUTHOR: <empty for none, if missing: inferred>
#+DATE: <empty for none, if missing: today>
#+BEAMER_FRAME_LEVEL: 2
#+TITLE: <causes <Title> to be regular content!>

(org-babel-do-load-languages ; babel, for executing 
 'org-babel-load-languages   ; code in org-mode.
 '((sh . t)
   (emacs-lisp . t)))

(require 'org-latex) ; latex export 
(add-to-list         ; with highlighting
  'org-export-latex-packages-alist '("" "minted"))
(add-to-list 
  'org-export-latex-packages-alist '("" "color"))
(setq org-export-latex-listings 'minted)

(custom-set-variables ; in ~/.emacs, only one instance 
 '(org-export-latex-classes (quote ; in the init file!
    (("beamer" "\\documentclass{beamer}" 
        org-beamer-sectioning))))
 '(org-latex-to-pdf-process (quote 
    ((concat "pdflatex -interaction nonstopmode" 
             "-shell-escape -output-directory %o %f") 
     "bibtex $(basename %b)" 
     (concat "pdflatex -interaction nonstopmode" 
             "-shell-escape -output-directory %o %f")
     (concat "pdflatex -interaction nonstopmode" 
             "-shell-escape -output-directory %o %f")))))

emerge emacs

emerge dev-tex/latex-beamer app-text/texlive

emerge dev-python/pygments

Sending email to many people with Emacs Wanderlust

I recently needed to send an email to many people¹.

Putting all of them into the BCC field did not work (mail rejected by provider) and when I split it into 2 emails, many did not see my mail because it was flagged as potential spam (they were not in the To-Field)².

I did not want to put them all into the To-Field, because that would have spread their email-addresses around, which many would not want³.

So I needed a different solution. Which I found in the extensibility of emacs [45] and wanderlust [98]⁴. It now carries the name wl-draft-send-to-multiple-receivers-from-buffer.

You simply write the email as usual via wl-draft, then put all email addresses you want write to into a buffer and call M-x wl-draft-send-to-multiple-receivers-from-buffer. It asks you about the buffer with email addresses, then shows you all addresses and asks for confirmation.

Then it sends one email after the other, with a randomized wait of 0-10 seconds between messages to avoid flagging as spam.

If you want to use it, just add the following to your .emacs:

(defun wl-draft-clean-mail-address (address)
  (replace-regexp-in-string "," "" address))

(defun wl-draft-send-to-multiple-receivers (addresses)
  (loop for address in addresses do
        (progn 
          (wl-user-agent-insert-header "To" 
            (wl-draft-clean-mail-address address))
          (let ((wl-interactive-send nil))
            (wl-draft-send))
          (sleep-for (random 10)))))

(defun wl-draft-send-to-multiple-receivers-from-buffer 
    (&optional addresses-buffer-name)
  "Send a mail to multiple recipients - one recipient at a time"
  (interactive "BBuffer with one address per line")
  (let ((addresses nil))
    (with-current-buffer addresses-buffer-name
      (setq addresses (split-string (buffer-string) "\n")))
    (if (y-or-n-p (concat "Send this mail to " 
            (mapconcat 'identity addresses ", ")))
        (wl-draft-send-to-multiple-receivers addresses))))

Happy Hacking!

Simple Emacs DarkRoom

I just realized that I let myself be distracted by all kinds of not-so-useful stuff instead of finally getting to type the text I already wanted to transcribe from stenografic at the beginning of … last week.

Screenshot!

Let’s take a break for a screenshot of the final version, because that’s what we really want to gain from this article: a distraction-free screenshot as distraction from the text :)

[100]

As you can see, the distractions are removed — the screenshot is completely full screen and only the text is left. If you switch to the minibuffer (i.e. via M-x), the status bar (modeline) is shown.

Background

To remove the distractions I looked again at WriteRoom [101] and DarkRoom and similar which show just the text I want to write. More exactly: I thought about looking at them again, but at second thought I decided to see if I could not just customize emacs to do the same, backed with all the power you get from several decades of being THE editor for many great hackers.

It took some googling and reading emacs wiki, and then some Lisp-hacking, but finally it’s 4 o’clock in the morning and I’m writing this in my own darkroom mode¹, toggled on and off by just hitting F11.

Implementation

I build on hide-mode-line (livejournal post [102] or webonastick [103]) as well as the full-screen info in the emacs wiki [104].

The whole code just takes 76 lines of code plus 26 lines comments and whitespace:

;;;; Activate distraction free editing with F11

; hide mode line, from http://dse.livejournal.com/66834.html / http://webonastick.com
(autoload 'hide-mode-line "hide-mode-line" nil t)
; word counting
(require 'wc)

(defun count-words-and-characters-buffer ()
  "Display the number of words and characters in the current buffer."
  (interactive)
  (message (concat "The current buffer contains "
           (number-to-string
            (wc-non-interactive (point-min) (point-max)))
           " words and "
           (number-to-string 
            (- (point-max) (point-min)))
           " letters.")))

; fullscreen, taken from http://www.emacswiki.org/emacs/FullScreen#toc26
; should work for X und OSX with emacs 23.x (TODO find minimum version).
; for windows it uses (w32-send-sys-command #xf030) (#xf030 == 61488)
(defvar babcore-fullscreen-p t "Check if fullscreen is on or off")
(setq babcore-stored-frame-width nil)
(setq babcore-stored-frame-height nil)

(defun babcore-non-fullscreen ()
  (interactive)
  (if (fboundp 'w32-send-sys-command)
      ;; WM_SYSCOMMAND restore #xf120
      (w32-send-sys-command 61728)
    (progn (set-frame-parameter nil 'width 
                                (if babcore-stored-frame-width
                                    babcore-stored-frame-width 82))
           (set-frame-parameter nil 'height
                                (if babcore-stored-frame-height 
                                    babcore-stored-frame-height 42))
           (set-frame-parameter nil 'fullscreen nil))))

(defun babcore-fullscreen ()
  (interactive)
  (setq babcore-stored-frame-width (frame-width))
  (setq babcore-stored-frame-height (frame-height))
  (if (fboundp 'w32-send-sys-command)
      ;; WM_SYSCOMMAND maximaze #xf030
      (w32-send-sys-command 61488)
    (set-frame-parameter nil 'fullscreen 'fullboth)))

(defun toggle-fullscreen ()
  (interactive)
  (setq babcore-fullscreen-p (not babcore-fullscreen-p))
  (if babcore-fullscreen-p
      (babcore-non-fullscreen)
    (babcore-fullscreen)))

(global-set-key [f11] 'toggle-fullscreen)

; simple darkroom with fullscreen, fringe, mode-line, menu-bar and scroll-bar hiding.
(defvar darkroom-enabled nil)
; TODO: Find out if menu bar is enabled when entering darkroom. If yes: reenable.
(defvar darkroom-menu-bar-enabled nil)

(defun toggle-darkroom ()
  (interactive)
  (if (not darkroom-enabled)
      (setq darkroom-enabled t)
    (setq darkroom-enabled nil))
  (hide-mode-line)
  (if darkroom-enabled
      (progn
        (toggle-fullscreen)
        ; if the menu bar was enabled, reenable it when disabling darkroom
        (if menu-bar-mode
            (setq darkroom-menu-bar-enabled t)
          (setq darkroom-menu-bar-enabled nil))
        ; save the frame configuration to be able to restore to the exact previous state.
        (if darkroom-menu-bar-enabled
            (menu-bar-mode -1))
        (scroll-bar-mode -1)
        (let ((fringe-width 
               (* (window-width (get-largest-window)) 
                  (/ (- 1 0.61803) (1+ (count-windows)))))
              (char-width-pixels 6))
        ; 8 pixels is the default, 6 is the average char width in pixels
        ; for some fonts:
        ; http://www.gnu.org/software/emacs/manual/html_node/emacs/Fonts.html
           (set-fringe-mode (truncate (* fringe-width char-width-pixels))))
    
        (add-hook 'after-save-hook 'count-words-and-characters-buffer))
    
    (progn 
      (if darkroom-menu-bar-enabled
          (menu-bar-mode))
      (scroll-bar-mode t)
      (set-fringe-mode nil)
      (remove-hook 'after-save-hook 'count-words-and-characters-buffer)
      (toggle-fullscreen))))

; Activate with M-F11 -> enhanced fullscreen :)
(global-set-key [M-f11] 'toggle-darkroom)

(provide 'activate-darkroom)

Also I now activated cua-mode to make it easier to interact with other programs: C-c and C-x now copy/cut when the mark is active. Otherwise they are the usual prefix keys. To force them to be the prefix keys, I can use control-shift-c/-x. I thought this would disturb me, but it does not.

To make it faster, I also told cua-mode to have a maximum delay of 50ms, so I don’t feel the delay. Essentially I just put this in my ~/.emacs:

(cua-mode t)
(setq cua-prefix-override-inhibit-delay 0.005)

Epilog

Well, did this get me to transcribe the text? Not really, since I spent the time building my own DarkRoom/WriteRoom, but I enjoyed the little hacking and it might help me get it done tomorrow - and get far more other stuff done.

And it is really fun to write in DarkRoom mode ;)

PS: If you like the simple darkroom, please leave a comment [105]!

I hereby declare that anyone is allowed to use this post and the screenshot under the same licensing as if it had been written in emacswiki [3].

Staying sane with Emacs (when facing drudge work)

I have to sift through 6 really boring config files. To stay sane, I call in Emacs [107] for support.

My task looks like this:

[108]
(click for full size [108])

In the lower left window I check the identifier in the table I have to complete (left column), then I search for all instances of that identifier in the right window and insert the instrument type, the SIGMA (uncertainty due to representation error defined for the type of the instrument and the location of the site) and note whether the site is marked as assimilated in the config file.

Then I also check all the other config files and note whether the site is assimilated there.

Drudge work. There are people who can do this kind of work. My wife would likely be able to do it faster without tool support than I can do it with tool support. But I’m really bad at that: When the task gets too boring I tend to get distracted - for example by writing this article.

To get the task done anyway, I create tools which make it more enjoyable. And with Emacs that’s actually quite easy, because Emacs provides most required tools out of the box.

Firstoff: My workflow before adding tools was like this:

• hit xo to switch from the lower left window to the config file at the right.
• Use M-x occur then type the station identifier. This displays all occurances of the station identifier within the config file in the upper left window.
• Hit xo twice to switch to the lower left window again.
• Type the information into the file.
• Switch to the next line and repeat the process.

I now want to simplify this to a single command per line. I’ll use F9 as key, because that isn’t yet used for other things in my Emacs setup and because it is easy to reach and my default keybinding as “useful shortcut for this file”. Other single-keystroke options would be F7 and F8. All other F-keys are already used :)

To make this easy, I define a macro:

• Move to the line above the line I want to edit.
• Start Macro-recording with C-x C-(.
• Go to the beginning of the next line with C-n and C-a.
• Activate the mark with C-SPACE and select the whole identifier with M-f.
• Make the identifier lowercase with M-x downcase-region, copy it with M-w and undo the downcasing with C-x u (or use the undo key; I defined one in my xmodmap).
• Switch to the config file with C-x o
• Search the buffer with M-x occur, inserting the identifier with C-y.
• Hit C-x o C-x o (yes, twice) to get back into the list of sites.
• Move to the end of the instrument column with M-f and kill the word with C-BACKSPACE.
• Save the macro with C-x C-).
• Bind kmacro-call-macro to F9 with M-x local-set-key F9 kmacro-call-macro.

Done.

My workflow is now reduced to this:

• Hit F9
• Type the information.
• Repeat.

I’m pretty sure that this will save me more time today than I spent writing this text ☺

Happy hacking!

Tutorial: Writing scientific papers for ACP using emacs org-mode

Update 2023: I no longer work at the University [113]. Nowadays I would use this setup as starting point, but with more focus on using org for reproducibility [114] (I have German instructions [115] for that) when the first version is ready for submission, export it as LaTeX, and work directly on that, because you’ll need detailed changes in LaTeX.

  (require 'org)
  (require 'org-latex)
  (require 'ox-latex)
  (setq org-latex-packages-alist 
        (quote (("" "color" t) ("" "minted" t) ("" "parskip" t)))
        org-latex-pdf-process 
        (quote (
"pdflatex -interaction nonstopmode -shell-escape -output-directory %o %f" 
"bibtex $(basename %b)" 
"pdflatex -interaction nonstopmode -shell-escape -output-directory %o %f" 
"pdflatex -interaction nonstopmode -shell-escape -output-directory %o %f")))
  (add-to-list 'org-latex-classes
               `("copernicus_discussions"
                 "\\documentclass{copernicus_discussions}
               [NO-DEFAULT-PACKAGES]
               [PACKAGES]
               [EXTRA]"
                 ("\\section{%s}" . "\\section*{%s}")
                 ("\\subsection{%s}" "\\newpage" "\\subsection*{%s}" "\\newpage")
                 ("\\subsubsection{%s}" . "\\subsubsection*{%s}")
                 ("\\paragraph{%s}" . "\\paragraph*{%s}")
                 ("\\subparagraph{%s}" . "\\subparagraph*{%s}"))
               )

(require 'reftex-cite)
(defun org-mode-reftex-setup ()
  (interactive)
  (and (buffer-file-name) (file-exists-p (buffer-file-name))
       (progn
        ; Reftex should use the org file as master file. See C-h v TeX-master for infos.
        (setq TeX-master t)
        (turn-on-reftex)
        ; enable auto-revert-mode to update reftex when bibtex file changes on disk
        (global-auto-revert-mode t) ; careful: this can kill the undo
                                    ; history when you change the file
                                    ; on-disk.
        (reftex-parse-all)
        ; add a custom reftex cite format to insert links
        ; This also changes any call to org-citation!
        (reftex-set-cite-format
         '((?c . "\\citet{%l}") ; natbib inline text
           (?i . "\\citep{%l}") ; natbib with parens
           ))))
  (define-key org-mode-map (kbd "C-c )") 'reftex-citation)
  (define-key org-mode-map (kbd "C-c (") 'org-mode-reftex-search))

(add-hook 'org-mode-hook 'org-mode-reftex-setup)

(add-to-list 'org-latex-packages-alist '("" "minted"))
(add-to-list 'org-latex-packages-alist '("" "color"))
(setq org-latex-listings 'minted)

; add emacs lisp support for minted
(setq org-latex-custom-lang-environments
      '((emacs-lisp "common-lispcode")))

#+title: YOUR TITLE
#+Options: toc:nil ^:nil
#+BIND: org-latex-title-command ""
#+Latex_Class: copernicus_discussions
#+LaTeX_CLASS_OPTIONS: [acpd, hvmath, online]

# Nice code-blocks
#+BEGIN_SRC elisp :noweb no-export :exports results
  (setq org-latex-minted-options
    '(("bgcolor" "mintedbg") ("frame" "single") ("framesep" "6pt") 
      ("mathescape" "true") ("fontsize" "\\footnotesize")))
  nil
#+END_SRC

#+BEGIN_ABSTRACT
Abstract
#+END_ABSTRACT
#+TOC: headlines 2

#+Latex: \runningtitle{SHORT TITLE}
#+Latex: \runningauthor{SHORT AUTHOR}
#+Latex: \correspondence{AUTHOR NAME\\ EMAIL}
#+Latex: \affil{YOUR UNIVERSITY}
#+Latex: \author[2,*]{SECOND AUTHOR}
#+Latex: \author[1]{THIRD AUTHOR SAME INSTITUTE}
#+Latex: \affil[2]{SECOND UNIVERSITY}
#+Latex: \affil[*]{now at: THIRD UNIVERSITY}

#+Latex: \received{}
#+Latex: \pubdiscuss{}
#+Latex: \revised{}
#+Latex: \accepted{}
#+Latex: \published{}
#+Latex: %% These dates will be inserted by ACPD
#+Latex: \firstpage{1}

#+Latex: \maketitle

#+Latex: \introduction
# * Introduction

* Second section

* Discussion

#+Latex: \conclusions
# * Conclusions

#+Latex: \appendix

# use acknowledgements for multiple
#+BEGIN_acknowledgement
Foo Bar Baz.
#+END_acknowledgement

#+Latex: \bibliographystyle{copernicus}
#+Latex: \bibliography{ABSOLUTE_PATH_TO_YOUR_BIBTEX_FILE_WITHOUT_.bib_SUFFIX}{}

# Local Variables:
# org-confirm-babel-evaluate: nil
# org-export-allow-bind-keywords: t
# End:

#+Latex_Class: copernicus_discussions
#+LaTeX_CLASS_OPTIONS: [acpd, hvmath, online]

#+BEGIN_SRC elisp :noweb no-export :exports results
  (setq org-latex-minted-options
    '(("bgcolor" "mintedbg") ("frame" "single") ("framesep" "6pt") 
      ("mathescape" "true") ("fontsize" "\\footnotesize")))
  nil
#+END_SRC

#+Latex: \appendix

# Local Variables:
# org-confirm-babel-evaluate: nil
# org-export-allow-bind-keywords: t
# End:

Unicode char \u8:χ not set up for use with LaTeX: Solution (made easy with Emacs)

For years I regularly stumbled over LaTeX-Errors in the form of Unicode char \u8:χ not set up for use with LaTeX. I always took the chickens path and replaced the unicode characters with the tex-escapes in the file. That was easy, but it made my files needlessly unreadable. Today I decided to FIX the problem once and for all. And it worked. Easily.

Firstoff: The problem I’m facing is that my keyboard layout [128] makes it effortless for me to input characters like ℂ Σ and χ. But LaTeX cannot cope with them out-of-the-box. Org-mode [5] already catches most of these problems, so I can write things like x² instead of x^2, but occasionally it stumbles.

The solution to that is actually pretty simple: I only need to declare the escapes-sequences LaTeX should use when it sees one of the characters (to be used before \begin{document}!):

\DeclareUnicodeCharacter{03C7}{\chi}

Or in org-mode:

#+LaTeX_HEADER: \DeclareUnicodeCharacter{03C7}{\chi}

To do this more easily, you can use the uniinput.ins [129] and uniinput.dtx [130] from the neo-layout project [128]. Run latex uniinput.ins to generate uniinput.sty which you can put next to your latex files and use with \usepackage{uniinput} (instructions in German [131]).

Thanks go to Wikibooks:LaTeX [132] for this. Their solution suggests then to read several Unicode definition documents for tracking down the codepoint of the character. But we can make that easier with Emacs [4] (almost everything is easier with Emacs ☺).

Instead of browsing huge documents manually, we simply rely on the unicode-definitions in Emacs: Move the cursor over the char and execute M-x describe-char.

When used with χ, this shows the following output:

             position: 672 of 35513 (2%), column: 0
            character: χ (displayed as χ) (codepoint 967, #o1707, #x3c7)
    preferred charset: unicode-bmp (Unicode Basic Multilingual Plane (U+0000..U+FFFF))
code point in charset: 0x03C7
… (and a bit more) …

What we need is code point in charset: Just leave out the 0x and you have the codepoint.

For the document I currently write, I now use the following definitions:

#+LaTeX_HEADER: \DeclareUnicodeCharacter{03C7}{\chi}
#+LaTeX_HEADER: \DeclareUnicodeCharacter{B2}{^{2}}

And that makes χ² work.

Happy Hacking - and have fun with Emacs Org-Mode!

Unicode-Characters for TODO-States in Emacs Orgmode

By default Emacs [107] Orgmode [133] uses uppercase words for todo keywords. But having tens of entries marked with TODO and DONE in my file looked horribly cluttered to me. So I searched for alternatives. After a few months of experimentation, I decided on the following scheme. It served me well ever since:

• ❢ To do
• ☯ In progress
  • ⚙ A program is running (optional detail)
  • ✍ I’m writing (optional detail)
• ⧖ Waiting
• ☺ To report
• ✔ Done
• ⌚ Maybe do this at some later time
• ✘ Won’t do

To set this in org-mode, just add the following to the header (and reopen the document, for example with C-x C-v):

#+SEQ_TODO: ❢ ☯ ⧖ | ☺ ✔ ⌚ ✘

or for the complex case (with details on what I do)

#+SEQ_TODO: ❢ ☯ ⚙ ✍ ⧖ | ☺ ✔ ⌚ ✘

Then use C-c C-t or SHIFT-→ (shift + right arrow) to switch to the next state or SHIFT-← (shift + left arrow) to switch to the previous state.

Anything before the | in the SEQ_TODO is shown in red (not yet done), anything after the | is show in green (done). Things which get triggered when something is done (like storing the time of a scheduled entry [134]) happen when the state crosses the |.

And with that, my orgmode documents are not only very useful but also look pretty lean. Just as good as having a GUI with images, but additionally I can access them over SSH and edit the todo state with any tool - because it’s just text.

Use the source, Luke! — Emacs org-mode beamer export with images in figure

I just needed to tweak my Emacs [45] org-mode [5] to beamer-latex [135] export to embed images into a figure environment (not wrapfigure!). After lots of googling and documentation reading I decided to bite the bullet and just read the source. Which proved to be much easier than I had expected.

This tutorial requires at least org-mode 8.0 (before that you had to use hacks to get figure without a caption). It is only tested for org-mode 8.0.2: The code you see when you read the source might look different in other versions.

#+BEGIN_SRC sh :exports results :results output raw
echo '[[./image.png]]'
#+END_SRC

\includegraphics[width=.9\linewidth]{./image.png}

\begin{figure}[htb]
\centering
\includegraphics[width=.9\linewidth]{./image.png}
\end{figure}

C-h f org-latex-export-as-latex
C-x C-o
C-s .el C-b ENTER
C-s figure C-s C-s C-s ...

(float (let ((float (plist-get attr :float)))
     (cond ((string= float "wrap") 'wrap)
       ((string= float "multicolumn") 'multicolumn)
       ((or (string= float "figure")
            (org-element-property :caption parent))
        'figure))))

#+attr_latex: :float figure

#+BEGIN_SRC sh :exports results :results output raw
echo '#+attr_latex: :float figure'
echo '[[./image.png]]'
#+END_SRC

\begin{figure}[htb]
\centering
\includegraphics[width=.9\linewidth]{./image.png}
\end{figure}

C-h f org-latex-export-as-latex

C-x C-o

C-s .el C-b ENTER

C-s figure C-s C-s C-s ...

Using Macros to avoid tedious tasks (screencast)

Because I am lazy,¹ and that makes me fast.

Screencast

(download (ogg theora video) [137])

Using Macros to avoid tedious tasks

Plan

• [X] Show the task
• [X] Record Macro
• [X] Use Macro

Explanation

I record a macro to find ~, then activate the mark and find a space.

C-s ~, C-SPACE, C-s SPACE

Then kill the region and type ${}

C-w ${}

That’s it.

Why??

• It is resilient: I check each change I do.
• I avoid repeating unnerving stuff.

Thank you

☺

recorded with recordmydesktop: recordmydesktop --delay 10 --width 800 --height 600 --on-the-fly-encoding

Wish: KDE with Emacs-style keyboard shortcuts

I would love to be able to use KDE with emacs-style keyboard shortcuts, because Emacs offers a huge set of already clearly defined shortcuts for many different situations. Since its users tend to do very much with the keyboard alone, even more obscure tasks are available via shortcuts.

I think that this would be useful, because Emacs is like a kind of nongraphical desktop environment itself (just look at emacspeak [139]!). For all those who use Emacs in a KDE environment, it could be a nice timesaver to be able to just use their accustomed bindings.

It also has a mostly clean structure for the bindings:

• "C-x anything" does changes which affect things outside the content of the current buffer.
• "C-c anything" is kept for specific actions of programs. For example "C-c C-c" in an email sends the email, while "C-c C-c" in a version tracking commit message finishes the message and starts the actual commit.
• "C-anything but x or c" acts on the content you're currently editing.
• "M-x" opens a 'command-selection-dialog' (just like alt-f2). You can run commands by name.
• "M-anything but x" is a different flavor of "C-anything but x or c". For example "C-f" moves the cursor one character forward, while "M-f" moves one word forward. "C-v" moves one page forward, while "M-v" moves one page backwards.

On the backend side, this would require being able to define multistep shortcuts. Everything else is just porting the emacs shortcuts to KDE actions.

The actual porting of shortcuts would then require mapping of the Emacs commands to KDE actions.

Some examples:

• "C-s" searches in a file. Replaces C-f.
• "C-r" searches backwards.
• "C-x C-s" saves a file -> close. Replaces C-w.
• "C-x C-f" opens a file -> Open. Replaces C-o.
• "C-x C-c" closes the program -> quit. Replaces C-q.
• "C-x C-b" switches between buffers/files/tabs -> switch the open file. Replaces alt-right_arrow and a few other (to my knowledge) inconsistent bindings.
• "C-x C-2" splits a window (or part of a window) vertically. "C-x C-o" switches between the parts. "C-x C-1" undoes the split and keeps the currently selected part. "C-x C-0" undoes the split and hides the currently selected part.

Write multiple images on a single page in org-mode.

• PDF-File [140]
• Org-File [141]

#+latex_header: \usepackage{subfig}

import pylab as pl
import numpy as np
x = np.random.random(size=(2,1000))
pl.scatter(x[0,:], x[1,:], marker=".")
pl.savefig("test.png")
print "\label{fig:image}"
print "[[./test.png]]"

#+BEGIN_LaTeX
\begin{figure}\centering
\subfloat[A gull]{\label{fig:latex-gull} 
\includegraphics[width=0.3\textwidth]{test}
} 
\subfloat[A tiger]{\label{fig:latex-tiger} 
\includegraphics[width=0.3\textwidth]{test}
} 
\subfloat[A mouse]{\label{fig:latex-mouse} 
\includegraphics[width=0.3\textwidth]{test}
}
\caption{Multiple pictures}\label{fig:latex-animals}
\end{figure}
#+END_LaTeX

#+latex: \begin{figure}\centering
#+latex: \subfloat[A gull]{\label{fig:org-gull} 
#+attr_latex: :width 0.3\textwidth
[[./test.png]]
#+latex: }\subfloat[A tiger]{\label{fig:org-tiger} 
#+attr_latex: :width 0.3\textwidth
[[./test.png]]
#+latex: }\subfloat[A mouse]{\label{fig:org-mouse} 
#+attr_latex: :width 0.3\textwidth
[[./test.png]]
#+latex: }\caption{Multiple pictures}\label{fig:org-animals}
#+latex: \end{figure}

emacs wanderlust.el setup for reading kmail maildir

This is my wanderlust.el file to read kmail maildirs. You need to define every folder you want to read.

;; mode:-*-emacs-lisp-*-
;; wanderlust 
(setq 
  elmo-maildir-folder-path "~/.kde/share/apps/kmail/mail"
          ;; where i store my mail

  wl-stay-folder-window t                       ;; show the folder pane (left)
  wl-folder-window-width 25                     ;; toggle on/off with 'i'
  
  wl-smtp-posting-server "smtp.web.de"            ;; put the smtp server here
  wl-local-domain "draketo.de"          ;; put something here...
  wl-message-id-domain "web.de"     ;; ...

file continued:

  wl-from "Arne Babenhauserheide "                  ;; my From:

  ;; note: all below are dirs (Maildirs) under elmo-maildir-folder-path 
  ;; the '.'-prefix is for marking them as maildirs
  wl-fcc ".sent-mail"                       ;; sent msgs go to the "sent"-folder
  wl-fcc-force-as-read t               ;; mark sent messages as read 
  wl-default-folder ".inbox"           ;; my main inbox 
  wl-draft-folder ".drafts"            ;; store drafts in 'postponed'
  wl-trash-folder ".trash"             ;; put trash in 'trash'
  wl-spam-folder ".gruppiert/Spam"              ;; ...spam as well
  wl-queue-folder ".queue"             ;; we don't use this

  ;; check this folder periodically, and update modeline
  wl-biff-check-folder-list '(".todo") ;; check every 180 seconds
                                       ;; (default: wl-biff-check-interval)

  ;; hide many fields from message buffers
  wl-message-ignored-field-list '("^.*:")
  wl-message-visible-field-list
  '("^\\(To\\|Cc\\):"
    "^Subject:"
    "^\\(From\\|Reply-To\\):"
    "^Organization:"
    "^Message-Id:"
    "^\\(Posted\\|Date\\):"
    )
  wl-message-sort-field-list
  '("^From"
    "^Organization:"
    "^X-Attribution:"
     "^Subject"
     "^Date"
     "^To"
     "^Cc"))


; Encryption via GnuPG

(require 'mailcrypt)
 (load-library "mailcrypt") ; provides "mc-setversion"
(mc-setversion "gpg")    ; for PGP 2.6 (default); also "5.0" and "gpg"

(autoload 'mc-install-write-mode "mailcrypt" nil t)
(autoload 'mc-install-read-mode "mailcrypt" nil t)
(add-hook 'mail-mode-hook 'mc-install-write-mode)

(add-hook 'wl-summary-mode-hook 'mc-install-read-mode)
(add-hook 'wl-mail-setup-hook 'mc-install-write-mode)

;(setq mc-pgp-keydir "~/.gnupg")
;(setq mc-pgp-path "gpg")
(setq mc-encrypt-for-me t)
(setq mc-pgp-user-id "FE96C404")

(defun mc-wl-verify-signature ()
  (interactive)
  (save-window-excursion
    (wl-summary-jump-to-current-message)
    (mc-verify)))

(defun mc-wl-decrypt-message ()
  (interactive)
  (save-window-excursion
    (wl-summary-jump-to-current-message)
    (let ((inhibit-read-only t))
      (mc-decrypt))))

(eval-after-load "mailcrypt"
  '(setq mc-modes-alist
       (append
        (quote
         ((wl-draft-mode (encrypt . mc-encrypt-message)
            (sign . mc-sign-message))
          (wl-summary-mode (decrypt . mc-wl-decrypt-message)
            (verify . mc-wl-verify-signature))))
        mc-modes-alist)))


; flowed text

 ;; Reading f=f
 (autoload 'fill-flowed "flow-fill")
 (add-hook 'mime-display-text/plain-hook
          (lambda ()
            (when (string= "flowed"
                           (cdr (assoc "format"
                                       (mime-content-type-parameters
                                        (mime-entity-content-type entity)))))
              (fill-flowed))))
; writing f=f
;(mime-edit-insert-tag "text" "plain" "; format=flowed")


(provide 'private-wanderlust)

UPDATE (2012-05-07): ~/.folders

I now use a ~/.folders file, to manage my non-kmail maildir subscriptions, too. It looks like this:

.sent-mail
.~/.local/share/mail/mgl_spam   "mgl spam" 
.~/.local/share/mail/to.arne_bab    "to arne_bab"
.inbox  "inbox" 
.trash  "Trash"
..gruppiert.directory/.inbox.directory/Freunde  "Freunde"
.drafts "Drafts"
..gruppiert.directory/.alt.directory/Posteingang-2011-09-18 "2011-09-18"
.outbox

The mail in ~/.local/share/mail is fetched via fetchmail and procmail to have a really reliable mail fetching system which does not rely on a non-broken database or free space on the disk to keep working…

keep auto-complete from competing with org-mode structure-templates

For a long time it bothered me that auto-complete made it necessary for me to abort completion before being able to use org-mode templates [143].

I typed <s and auto-complete showed stuff like <string, forcing me to hit C-g before I could use TAB to complete the template with org-mode.

I fixed this for me by adding all the org-mode structure templates as stop-words:

;; avoid competing with org-mode templates.
(add-hook 'org-mode-hook
          (lambda ()
            (make-local-variable 'ac-stop-words)
            (loop for template in org-structure-template-alist do
                  (add-to-list 'ac-stop-words 
                               (concat "<" (car template))))))

Note, that with this snippet you will have to reopen a file if you add an org-mode template and want it recognized as stop-word in that file.

PS: I added this as bug-report to auto-complete [144], so with some luck you might not have to bother with this, if you’re willing to simply wait for the next release ☺