David Fraser wrote:
Hi
I spoke to Andi on IRC about doing docs for pylucene.
I hacked together a system this afternoon that strips out the Java function declarations and javadoc comments, and tries to add them to the generated PyLucene.py as docstrings.
There are three python modules involved:
- sparse.py which is a general tokenizer I use for other projects.
- java2doc.py which parses a Java file and finds the class and function declarations, matching them to comments
- redoc.py which takes a generated PyLucene.py file as input and uses the given java files (parsed with java2doc) to add docstrings to it, outputing the new file to stdout
To run it, place the three files in a PyLucene source directory, and go: ./redoc.py . < PyLucene.py > PyLucene_d.py The generated PyLucene_d.py will have the docstrings in it. It takes about 25 seconds to run on my machine.
I haven't yet done anything to strip off the /* * * */ that you get in the comments.
You can now generate an HTML doc for PyLucene using pydoc, it ends up being a single 1.8MB html file though.
It may be nice to split up the generated PyLucene_d.py into different files to make doc browsing easier.
The only negative impact observed is that Python complains about having a docstring that has a non-ASCII character in it, since the encoding isn't specified. I haven't done extensive tests to make sure I haven't messed anything up but it looks OK to me.
David
Has anybody tried this out? Any comments appreciated :-)
David _______________________________________________ pylucene-dev mailing list [email protected] http://lists.osafoundation.org/mailman/listinfo/pylucene-dev
