GitHub Metrics

Notice: urllib already migrated to requests library.

Step 1 - Intro

I created this repo to gather some metrics for the company I used to work. Its purpose is to collect metrics around an organization repositories, such as commits, forks, stars etc.

I used GitHub’s REST API v3 and PyGitHub to develop the Python scripts.

In order to be able to run this project you have to

be a member of the organization you want to collect the metrics from
have a Github personal token (generate here: https://github.com/settings/tokens)

Step 0 - Installing requirements

Clone this repo
Create a virtual environment (I use venv)
Activate your environment: $ source [ENVIRONMENT_NAME]/bin/activate
Install dependencies: $ pip install -r requirements.txt

Using the repo

There are 2 Python scripts. Their usage and purposes are described below.

FIRST SCRIPT

You execute the file on command line (terminal, bash, whatever you name it) by typing:

$ python github_metrics.py -t [GITHUB-TOKEN] -o [ORGANIZATION NAME]

And github_metrics.py script runs the functions:

1) list_orgs(token)

function that returns a list of the organizations the user belongs to.

2) list_org_members(org, token)

function that builds a list with the members of an organization inside GitHub. Only current members are listed. It returns one list with members names and other list with members logins (usernames).

3) export_code_frequency(directory, org, token)

function that gather statistics for all repositories in an organization, by week/user/repo, and returns a csv file with: name of the repository, week in question, number of additions, number of deletions, author of such commit and if the author is a member or not of the organization.

4) list_unique_collaborators(directory, org, token)

function that returns the number of unique collaborators of an organization (for all its repos), with the information if the collaborator is a member of the organization or not.

5) export_community_engagement(directory, org, token)

function that gather basic metrics for all repos of an organization, such as number of forks, stars, commits and collaborators. The function returns a csv file with these informations for each repo.

SECOND SCRIPT

You execute the file on command line by typing:

$ python export_traffic.py -t [GITHUB-TOKEN] -o [ORGANIZATION NAME]

And export_traffic.py runs the functions:

1) test_push_access(org, token)

function that will check which repositories your token has push access to, inside the specified organization. It returns two lists: one with the repositories that the token doesn't have push access to, and other with the repos the token has access to.

2) export_traffic(directory, org, repos_ok, token)

function that exports the traffic from GitHub repos that your token has push access to, for the last 14 days. It will create a csv file for the following metrics: traffic (views), clones, paths and referrers. Please check further documentation from GitHub.

3) relevantrepos_noaccess(numstars, repos_noaccess, organization, authToken)

function that checks which repositories over numstars your token doesn't have access to, and returns a dictionary with the name of the repo and its number of stars.