ig
is a tool to interactively visualize include graphs for C++ projects
Point ig
at any directory containing C++ source or header files and it will
construct a full graph of all includes, serve you a local website and visualize
the graph interactively with sigma.js, for you to admire.
Usage is very easy:
$ ig -o include
will inspect the folder include
, serve a website on localhost:8080
and even
open your browser for you. The full set of options currently include:
usage: ig [-h] [--pattern PATTERNS] [-i PREFIXES] [-v] [-p PORT] [-o] [-j]
[-d DIRECTORY] [--relation {includes,included-by}]
[--min-degree MIN_DEGREE] [--group-granularity GROUP_GRANULARITY]
[--full-path] [--colors COLORS] [--color-variation COLOR_VARIATION]
[--color-alpha-min COLOR_ALPHA_MIN]
directories [directories ...]
Visualize C++ include graphs
positional arguments:
directories The directories to inspect
optional arguments:
-h, --help show this help message and exit
--pattern PATTERNS The file (glob) patterns to look for
-i PREFIXES, -I PREFIXES, --prefix PREFIXES
An include path for headers to recognize
-v, --verbose Turn on verbose output
-p PORT, --port PORT The port to serve the visualization on
-o, --open Open the webpage immediately
-j, --json Print the graph JSON and instead of serving it
-d DIRECTORY, --dir DIRECTORY
The directory to store the served files in. If not
supplied, a temporary directory is created.
--relation {includes,included-by}
The relation of edges in the graph
--min-degree MIN_DEGREE
The initial minimum degree nodes should have to be
displayed
--group-granularity GROUP_GRANULARITY
How coarse to group nodes (by folder)
--full-path If set, shows the full path for nodes
--colors COLORS The base RGB colors separated by commas
--color-variation COLOR_VARIATION
The variation in RGB around the base colors
--color-alpha-min COLOR_ALPHA_MIN
The minimum alpha value for colors
But does it scale? It scales quite well. The graph you see above is the include graph for the entire LLVM and clang codebase, which spans more than 5,000 files and 1.5M LOC. Note that the visualization also includes sliders to group nodes by folder and filter out low-degree nodes.
Get it with pip:
$ pip install ig-cpp
Works with Python 2 and 3.
Who ever said C++ was an ugly language?
Peter Goldsborough + cat ❤️