Building an alternative view of the open source applications written in Python

./bin/pydigger

This will fetch the RSS feed from PyPi, create a subdirectory called www, and download the zip files of the most recently uploaded packages.

Once the zip file is downloaded it is unzipped.

./bin/app.py

export PYTHONPATH=~/python easy_install -d ~/python Genshi

Create a JSON file in the conf/ directory similar to the already exising ones and run $ python bin/parse.py conf/yourfile.json

• Have a json file describe the source that needs to be analyzed Where to download the files, how to update them (e.g. a packaged file having a new version or a git or SVN repository)
• process and analyze the files and create a static display with a few static json files with meta-data

) Syntax highlight the python files) Each project has its own directory ) Each projects has a main page listing all the files in it) The main page has a listing of all the projects (for now) with links to the main page of each project ) Parse other, non-python files as well and show them in some nice way) Allow listing directories that need to be excluded from the display ) Has simple about page) Basic logging in the parser ) Navigation link from each file back to the project page) Timeout for processing files

In order to find out which token has which token number the mapping can be found here: /usr/lib/python2.7/token.py Apparently if I type pydoc tokenize on the command line, I get the list

) Change the page names so they will have the same name as the original file, but still serve correctly) Guessing the lexer better ) Each directory in the project tree has an index file listing all the files and subdirectories in it) Create a configuration file to process several projects, including the source code of Python itself. ) List the dependencies and link to the relevant project pages) Show the documentation of each file/project as a separate HTML page *) Handle the file we cannot parse and highlight (e.g. .png files)