I've just solved a puzzle finding out why Python src blocks that exported
happily last year don't now. I'm using them for class tutorials and a of about
fifteen Python fragments, a couple of the blocks would expand in-situ (C-c C-c)
but when exported were treated as normal text.
I narrowed it down using the following file
==
#+options: num:nil toc:nil
** Try 1
#+BEGIN_SRC python :results output :exports both
print "hi"
#+END_SRC
#+RESULTS:
: hi
** Try 2
#+BEGIN_SRC python :results output :exports both
print "Fred"
#+END_SRC python
#+RESULTS:
: Fred
Both of the src blocks can be isolated and edited in its own buffer using
org-babel-expand-src-block - C-c C-v C-v) and both get run correctly if you C-c
C-c inside them, as shown by the #+RESULTS blocks, however on export, the second
one is treated as normal text .
Puzzlingly, a hex-dump showed the #+BEGIN_SRC lines were identical - then I
noticed that the #+END line was wrong. I'd written "#+END_SRC python" instead of
just "#+END_SRC".
It seems like the parser for the src blocks is now stricter that it was last
year.
Could I ask that an error message be generated for this? While not strictly a
bug, it's an insidious trap given that the in-buffer expansion (C-c C-c) runs
and inserts the results as expected.
I've also had similar problems with
#+BEGIN_SRC python :output results :exports both
print "hi"
#+END_SRC
which looks ok, and the ":output results" makes sense when you read it, but
should be ":results output". From memory, this also worked in-buffer but would
silently fail on export.
Orgmode + Babel is excellent for writing Python tutorial worksheets - so thank
you - your work is much appreciated.
Cheers
Giovanni
--
==
Dr Giovanni Moretti | School of Engineering and Advanced Technology
Senior Lecturer | Massey University, Palmerston North, New Zealand
Computer Science| Ph +64-6-3505799x84134 Fax +64-6-3502259 - ZL2GX
==
http://seat.massey.ac.nz/morettig.more...@massey.ac.nz
