Kenneth McDonald wrote: > More seriously, there is a major problem with docstrings in that they > can only document something that has a docstring; classes, functions, > methods, and modules. But what if I have constants that are > important? The only place to document them is in the module > docstring, and everything else--examples, concepts, and so on--must > be thrown in there as well. But there are no agreed on formats and > processing pipelines that then allow such a large module docstring, > plus other docstrings, to produce a good final document.
fwiw, that's one of reason why I developed PythonDoc (which supports JavaDoc-style documentation for all the usual suspects, but also for con- stants, attributes, and variables) > It's too bad that there is no equivalent of d'oxygen for Python. That > is a _nice_ program. doesn't doxygen support Python? </F> -- http://mail.python.org/mailman/listinfo/python-list
