that sounds like an interesting approach. xml seems like what you really want, since looking at the parsetree there is a lot of information (e.g. attributes, properties, etc...) that would be tricky to generate a fully representative json scheme.

This page suggests at the bottom you could export to texinfo, and convert that to docbook:
http://orgmode.org/worg/exporters/ox-overview.html

(1) DocBook export, available in previous Org-mode versions, has not currently been ported to the new exporter, however the new ox-texinfo backend can generate DocBook format. Once file.texi is created via ox-texinfo, simply execute:

makeinfo --docbook file.texi

John

-----------------------------------
John Kitchin
Associate Professor
Doherty Hall A207F
Department of Chemical Engineering
Carnegie Mellon University
Pittsburgh, PA 15213
412-268-7803
http://kitchingroup.cheme.cmu.edu

On Thu, Dec 5, 2013 at 3:30 PM, Brett Viren <bv@bnl.gov> wrote:

Has anyone written any new-style exporter which will produce a common
markup/data language format like JSON or YAML? 狢'm looking for
something that fully preserves the original org document structure and
does no semantic interpretation along the way.

What I really want is to parse arbitrary org files in Python. 狢've
looked at the entries at worg's "org-tool" node which do this but they
seem out of date or make assumptions about what org elements exist or
their URLs are not loading (NEO). 狢f any of that's a misrepresentation
please correct me.

In any case, using org's own exporter to produce JSON or YAML and then
relying on these format's Python modules for parsing seems like the best
way to go to let me author in org and process in Python.

I'm not very good with elisp (which is why I want to get org data into
Python) but I guess I can have a go at making such a "shunt" exporter.
Before I try, I just wanted to check if someone had this wheel already
spinning.

Thanks,
-Brett.