Skip to content

Visual code dependency graph creation for C/C++ projects

License

Notifications You must be signed in to change notification settings

uohmak4fvpqe/codeviz

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

95 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CodeViz

Visualize code dependencies in your C/C++ projects.

codeviz is a super simple, cross-platform, python script that only uses built-in libraries to create code dependency graphs of your source files.

It works like this:

  1. Find the source files and parse them to get their headers.
  2. Create a dot file that describes the code graph.
  3. Pass the dot file into graphviz, which will create a visual graphic of your code dependencies.

Installation

1. Install Graphviz

codeviz uses graphviz to create the visual map.

Follow the graphviz installation instructions for your platform:

Once graphviz is installed, ensure that it is available in your environment.

2. Install CodeViz

There are a several ways you can install codeviz.

Option A: Just copying the script to your executable path

For example, if you have ~/bin mapped to your environment's path:

$ git clone https://github.com/jmarkowski/codeviz.git
$ cp codeviz/codeviz.py ~/bin/codeviz

Option B: Using setup

To install codeviz into your system path:

$ git clone https://github.com/jmarkowski/codeviz.git
$ cd codeviz
$ sudo python3 setup.py install

The generated install-record.txt file will list the paths of files that were installed on your system.

Option C: In a virtual environment with pip

First, build the distribution package:

$ git clone https://github.com/jmarkowski/codeviz.git
$ python setup.py sdist

Next, create and start a virtual environment:

$ virtualenv pyenv
$ source pyenv/bin/activate

Finally, install the package:

$ pip install dist/codeviz*

Now, while in the virtual environment, you'll have the codeviz command in your path.

Example

The image below shows the result of generating the visual code map for the original git source at hash e83c516.

git e83c516

Usage

General

Run codeviz with a path to your source directory to search for source & header files in the given path:

$ codeviz path/to/src

Specify multiple code paths, mixing directories and files:

$ codeviz path/to/src/subdir1 path/to/src/subdir2 path/to/src/subdir3/*.h

Search directories recursively:

$ codeviz -r path/to/src

Ignore certain files and directories from being used:

# Ignore tests
$ codeviz -r src/ --ignore=src/unit-tests/* --ignore=src/test

Limit the source files to those with certain headers:

# Any source files that do not include the a.h and b.h headers are ignored.
$ codeviz *.cpp a.h b.h --must-include

Style Options

If you want a black and white version of the generated dependency graph:

$ codeviz --no-color path/to/src

If you wish to highlight particular files in the graph, you may do so:

$ codeviz src/ --highlight src/foo*.cpp

Format Options

You can specify the name of an output file with a variety of output formats.

$ codeviz path/to/src -o jpeg-file.jpg
$ codeviz path/to/src -o postscript-file.ps
$ codeviz path/to/src -o png-file.png

See here for a complete list of supported file formats.

Limitations

  1. System header files (e.g. stdio.h) are ignored and not referenced. This is intentional as it's presumed that you only care about dependencies of your source code.

  2. Because codeviz does not know anything about your build system and it's include paths, there may be "dependency collisions" if it finds multiple header files in separate directories that share the same filename. However, codeviz will warn you if such collisions occur and still show the dependencies in dotted lines labeled with a ?. You can resolve these collisions yourself by modifying the contents of the outputted dot file.

  3. For very, very large codebases, the graphviz tool may fail to create the graphic. In such cases, you're better off specifying a smaller subset of paths within the codebase that you're really interested in.

About

Visual code dependency graph creation for C/C++ projects

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%