opensim-core-code-analysis

To analyze cmake add --graphviz=deps.dot and make sure you have graphviz installed and pathed

For large graphs it may take a very long time to render. You can break out of the script (ctrl+c) as soon as the dot file is generated. For this repo

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

python3 dependency_graph.py -c --cluster-labels ~/opensim-workspace/opensim-core-source/ results/whole-opensim.dot
python3 dependency_graph.py -c --cluster-labels ~/opensim-workspace/opensim-core-source/OpenSim/ results/opensim.dot
python3 dependency_graph.py -c --cluster-labels ~/opensim-workspace/opensim-core-source/OpenSim/Common/ results/opensim-common.dot

f=opensim-common.dot
F=${f%.*}
T=png
out=j-opensim-common
sccmap $f -S > $out
gvpr -f ../stronglyColored.gvpr  $out  $f  | dot -T$T >$F.S.$T

find cycles:

python3 dot_find_cycles.py --only-shortest --print-labels results/whole-opensim.dot
python3 dot_find_cycles.py --only-shortest results/opensim-common.dot

References

• Cmake GraphViz
• Original Transative Reduction Script - dot_find_cycles.py
• Color strongly connected components GraphViz script
• scc.g script original
• Original dependency-graph repo

dependency-graph

A python script to show the "include" dependency of C++ classes.

It is useful to check the presence of circular dependencies.

Installation

The script depends on Graphviz to draw the graph.

On Ubuntu, you can install the dependencies with these two commands:

sudo apt install graphviz
pip3 install -r requirements.txt

Manual

usage: dependency_graph.py [-h] [-f {bmp,gif,jpg,png,pdf,svg}] [-v] [-c]
                           folder output

positional arguments:
  folder                Path to the folder to scan
  output                Path of the output file without the extension

optional arguments:
  -h, --help            show this help message and exit
  -f {bmp,gif,jpg,png,pdf,svg}, --format {bmp,gif,jpg,png,pdf,svg}
                        Format of the output
  -v, --view            View the graph
  -c, --cluster         Create a cluster for each subfolder

Examples

Example of a graph produced by the script:

Graph produced for the same project with clusters (-c):