This repository is collection of scripts, that import every function in C code as a node in a graph database, in this case Neo4j. Every function call is represented as an edge. Once the compilation process is finished, there should be a represantation of the call graph in the database.

The general purpose is to be able to query the database and possibly discover patterns in the codebase and attain a greater understanding of it.

Dependencies:

gcc-python-plugin (select the branch that is correct for the gcc version you have installed).

pyzmq

Make sure that you build the gcc python plugin correctly and export LD_LIBRARY_PATH=(yourpath)/gcc-python-plugin/gcc-c-api/

Also check out the usage of the gcc-python-plugin.

You'll also need a Neo4j database instance and the abillity to connect to it via a bolt connection.

The first step is to create a directory named feeds in /tmp (e.g. mkdir /tmp/feeds). Then execute receiver.py. The receiver will standby until there is a gcc compilation. Use gcc as you would normally, just add the parameters indicated in the gcc-python-plugin basic usage documentation.

This project works only with gcc.