Back in May, there was quite an extensive discussion of whether or not Python should support Unicode identifiers (with the final result being that this would be supported in Python 3). In my periodic googling for pyparsing users, I stumbled upon Zhpy, a preprocessor that renders on the fly Chinese Python keywords and indentifiers written using traditional Chinese characters into ASCII counterparts, and then executes the resulting script. I was very intrigued at the idea, and I thought back to Martin v. Löwis's PEP. The Zhpy approach will work with current versions of Python (going back to 2.3, since pyparsing requires that version).
You can view examples at the pyparsing "Who's Using Pyparsing" wiki page - http://pyparsing.wikispaces.com/WhosUsingPyparsing#Zhpy. Given the breadth of response to Martin v. Löwis's post, I'm curious as to what the reaction might be to something like Zhpy. -- Paul -- http://mail.python.org/mailman/listinfo/python-list