C/C++ dependency graphing using snakefood
Project description
Creates dependency files from C/C++ code for use with snakefood. This allows you to easily create dependency graphs on file/module/folder/etc. level from your C/C++ codebase.
Installation
You must have Python installed on your system.
If you don’t have snakefood installed yet, install it with pip install snakefood first.
Now install crowfood with pip install crowfood.
How to create dependency graphs
Getting started
Let’s assume /libab is the root path of your C/C++ project which contains:
A.c
B.c
ab/
A.h
B.h
We’ll create a simple file-based dependency graph by piping crowfood’s dependency information to snakefood’s sfood-graph tool to create a graph in DOT format which is then visualized with dot itself and saved to a pdf file:
cfood /libab | sfood-graph | dot -Tpdf > simple.pdf
crowfood can handle many folder layouts and will output warnings if it can’t find #include’s. Run cfood --help to see all options regarding folder layout and include paths.
TIP: As a quick fix or to get started with bigger projects, use --fuzzy to let crowfood search for #include files purely by their filename and ignore folder structure.
Grouping/Clustering
Depending on how big your project is it may make sense to group files together in some way.
One way is to group matching source and header files as “modules”:
cfood /libab --merge module | sfood-graph | dot -Tpdf > modules.pdf
Another way is to define clusters in terms of path prefixes. First, create a file clusters which contains the prefixes:
componenta
componentb/partc
And then run:
cfood /bigproject | sfood-cluster -f clusters | sfood-graph | dot -Tpdf > clustered.pdf
Have a look at the snakefood docs to get some more inspiration on how to transform the raw dependency output from crowfood into something that makes sense for your project. Always remember that the dependency output is line-based and very easy to handle with standard unix tools, e.g. grep for filtering.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.