What is this?

gitwalker is a tool for collecting data from git repositories. It automates the process of checking out each revision, running some command and logging the output to a JSON file. Additional commands can be added by writing Python classes.

What can it do?

Currently gitwalker supports two built in commands:

• A LaTeX word count
• du disk usage command
• Arbitrary shell commands

Its straightforward to add additional commands - see the file tools.py

The included script gitwalk_plot uses the matplotlib framework to produce time-series graphs overlaying multiple data files.

Getting It

If you have pip installed, simply pip install gitwalker

Usage

To word count a git-tracked LaTeX project across all commits:

gitwalk --wordcount myfile.tex --out wordcount.json /path/to/project

This will clone the repository at /path/to/project to a temporary directory before checking out each revision and running a word count on the file myfile.tex in the repository. The results will be output to the file wordcount.json

gitwalker also supports incremental update of a previously produced log file. To add newly committed revisions,

gitwalk --update wordcount.json --wordcount myfile.tex /path/to/project

There is an attached script to plot a number of such output files on the same axes using matplotlib. e.g.

gitwalk_plot --plot file1.json me red --plot you.json you blue wordcount/wordcount

Will plot the files file1.json and file2.json on the same axes using the specified labels and colours. The value will be dug out from the JSON file via the path format at the end of the command line - in this case wordcount/wordcount. One could also run

gitwalk_plot --plot file1.json me red --plot you.json you blue wordcount/nfigures

to plot the number of LaTeX figures present in each commit.

To run a shell command on each commit,

gitwalk --shell="ls | wc -l" /path/to/project

Todo

• Add git-notes option