Hello,
I'm writing a paper with some colleagues for a conference, and we've
just stumbled on a strange bug. My colleague is using org from elpa, and
is having the problem, whereas I'm compiling using a very current org
version and everything works for me.
The problem is as follows.
Given this org file (exp.org):
--8<---------------cut here---------------start------------->8---
#+options: toc:nil
#+property: results drawer
#+name: fetch
#+BEGIN_SRC emacs-lisp :results raw :var f="foo" :exports none
f
#+END_SRC
#+name: wrap
#+BEGIN_SRC emacs-lisp :var text="" :results raw :exports none
(concat "#+BEGIN_EXAMPLE\n" text "\n#+END_EXAMPLE")
#+END_SRC
* Test
:PROPERTIES:
:post: wrap(text=*this*)
:END:
Hello world
#+call: fetch("testing")
--8<---------------cut here---------------end--------------->8---
and this initialization file (exp_init.el):
--8<---------------cut here---------------start------------->8---
(setq emacsd-dir "~/.emacs.d/")
(package-initialize)
(require 'org-loaddefs)
(require 'ox-html)
(setq org-confirm-babel-evaluate nil)
--8<---------------cut here---------------end--------------->8---
running this command
emacs --batch -Q -l exp_init.el exp.org -f org-html-export-to-html
yields in my colleague's case the following:
> Loading vc-git...
> executing Emacs-Lisp code block (fetch)...
>
> (f (quote "testing"))
>
> "testing"
> executing Emacs-Lisp code block...
>
> (results (quote "testing"))
> Reference 'wrap' not found in this buffer
The same command works for me (for completeness's sake, here is my
configuration file):
--8<---------------cut here---------------start------------->8---
(setq emacsd-dir "~/.emacs.d/")
(add-to-list 'load-path (concat emacsd-dir "org/emacs/site-lisp/org"))
(require 'ox-html)
(setq org-confirm-babel-evaluate nil)
--8<---------------cut here---------------end--------------->8---
He's using this org version:
Org-mode version 8.2.5h (8.2.5h-94-g91175a-elpa @
/Users/pmaksimo/.emacs.d/elpa/org-20140331/)
updated yesterday.
My questions are:
- is there a bug in the ELPA version of org mode?
- is there a workaround? We tried ":exports results" instead of
":exports none" but the blocks are not exported in that case.
Thanks,
Alan
