Description

Generate dependency graphs from Python code. This dependency tracker package has a few distinguishing characteristics:

• It uses the AST to parse the Python files. This is very reliable, it always runs.
• No module is loaded. Loading modules to figure out dependencies is almost always problem, because a lot of codebases run initialization code in the global namespace, which often requires additional setup. Snakefood is guaranteed not to have this problem (it just runs, no matter what).
• It works on a set of files, i.e. you do not have to specify a single script, you can select a directory (package or else) or a set of files. It finds all the Python files recursively automatically.
• Automatic/no configuration: your PYTHONPATH is automatically adjusted to include the required package roots. It figures out the paths that are required from the files/directories given as input. You should not have to setup ANYTHING.
• It does not have to automatically 'follow' dependencies between modules, i.e. by default it only considers the files and directories you specify on the command-line and their immediate dependencies. It also has an option to automatically include only the dependencies within the packages of the files you specify.
• It follows the UNIX philosophy of small programs that do one thing well: it consists of a few simple programs whose outputs you combine via pipes. Graphing dependencies always requires the user to filter and cluster the filenames, so this is appropriate. You can combine it with your favourite tools, grep, sed, etc.

A problem with dependency trackers that run code is that they are unreliable, due to the dynamic nature of Python (the presence of imports within function calls and __import__ hooks makes it almost impossible to always do the right thing). This script aims at being right 99% of the time, and we think that given the trade-offs, 99% is good enough for 99% of the uses.

I fully intend that this program work on all codebases. It has been tested on a number of popular open source codes (see the test directory).

Dependencies

• Python 2.5 or higher. That's it.

Documentation

• CHANGES
• TODO
• User's Manual
• Example Outputs

Download

A Mercurial repository can be found at:

http://github.com/blais/snakefood

Links

• dottoxml, a tool by Dirk Bächle, that converts dot files into yEd inputs, useful for large graphs.

Copyright and License

Copyright (C) 2001-2007 Martin Blais. All Rights Reserved.

This code is distributed under the GNU General Public License;

Author

Martin Blais <blais@furius.ca>