Skip to content
Layout C call graphs from cflow using GraphViz dot
Branch: master
Clone or download
Pull request Compare This branch is 8 commits ahead of johnyf:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
examples/simple
.gitignore
LICENSE
README.md
pycflow2dot.py
setup.py

README.md

pycflow2dot

Summary

Draw call graphs for C source codes using dot and cflow. Typeset PDF with a page per source file and clickable cross-file function references.

cflow2dot -i hello_simple.c -f png produces:

from hello_simple.c.

Description

Draw the call graph of C source code using cflow and dot. Output to LaTeX, .dot, .PDF, .SVG, .PNG and from dot to all formats supported from it. The LaTeX output is obtained by including the SVG via Inkscape's LaTeX export functionality.

Multi-file sources are converted to multiple SVG files, one for each source. These contain links using the LaTeX package hyperref, so that after compilation one can click on the name of a function call and be taken to its definition, even if that definition is in another page of the PDF, because the function is defined in another source file than the one corresponding to the current PDF page.

Note that if a file containing the definition is missing, then the hyperref link is omitted, so that no dead links result after compiling with LaTeX. This might be the case of for example the file with the definitions is available, but is not passed to pycflow2dot, e.g., for the purpose of focusing on a subset of the sources.

For now the LaTeX result has to be manually compiled, though this extra step will be automated in the future. Multi-SVG export will still be available, so that the results can be included in a larger document, e.g., a report.

PyCflow2dot is a Python port of the Perl script cflow2dot. Tested with Python 3.2 (NetworkX not yet available in 3.3.).

Installation

Use pip with Python 3 or 2:

pip install pycflow2dot

You also need to install the following non-Python dependencies:

  • GNU cflow:
    • apt-get install cflow on Debian GNU/Linux
    • port install cflow with MacPorts
  • dot:
    • apt-get install graphviz
    • port install graphviz

Optionally, cpp too.

License

PyCflow2dot is licensed under the GNU GPL v3.

You can’t perform that action at this time.