[Docutils-users] Producing skeleton documents with data from RST - workflow advice?

[Matthew L. <lei...@ci...>]

Hi,

I've got a restructured text file lessons.txt with metadata on every  
class I want to teach in a course this term.  I'd like to create some  
LaTeX files to use for class notes that have that metadata in them  
already.  The actual math of the class has to be put in later, but I'm  
trying to automate some of what I do every day.  I have another file  
hw.txt with homework assignments in it; I'd like to merge those into  
the class notes as well.  There's some tedious data munging I want,  
too, like having the lesson files in a directory based on the lesson  
title, naming the files based on the lesson title (in a different  
way), etc.

I'm trying to figure out the best way to do this.  I use the  
rst2xxx.py tools a lot to produce latex or html files.  One possible  
solution would be to convert lessons.txt and hw.txt to docutils XML  
and use XSLT to make the LaTeX files.  But I'm not so excited about  
doing some of the directory-creating and text manipulation in XSLT.

Another solution is to start with the docutils parser and select what  
I need from the document's internal data structure, much like  
selecting nodes from an XML tree in XSLT.  This way I can use whatever  
python code I need to manipulate text or create/open files.  But I  
don't know too much about this internal structure, and I couldn't find  
it documented.

So I thought I'd throw it out to the list.  How would you go about  
this task?

Thanks for any advice,
Matt

--
Matthew Leingang
Clinical Associate Professor of Mathematics
lei...@ci...