hello, > As a stream processor, M4 would seem to be a natural candidate for > implementation with Python generators.
i'm really curious and trying to find just one good reason to do that. i use both python and m4 but i can't see the point. * from the python user perspective, i would say that there is a lot of template engine outhere (like jinja) and if i miss something, it would be the port of the perl template toolkit (http://template-toolkit.org/). * from the m4 user perspective, i would say that the GNU m4 implementation is mature and written in C so rewrite it in a dynamic language (whatever it is) seems useless. a guile extension to m4 could be much valuable. regards -- Marc Chantreux (eiro on github and freenode) http://eiro.github.com/ http://eiro.github.com/atom.xml "Don't believe everything you read on the Internet" -- Abraham Lincoln
