Download basic statistics for all repositories in list of GitHub organizations

download_github_organization_stats.py

#!/usr/bin/env python 
# Python 3

import datetime
import json
import os

from github import Github # pip install PyGithub

ORGANIZATIONS = ['my_organization', 'some_other_organization']
TARGET = '/path/to/data'
TOKEN = 'githubtoken'

def _get_clones_traffic(repo):
    data = repo.get_clones_traffic()
    data['clones'] = [clone.raw_data for clone in data['clones']]
    return data

def _get_views_traffic(repo):
    data = repo.get_views_traffic()
    data['views'] = [view.raw_data for view in data['views']]
    return data

def _get_top_paths(repo):
    return [path.raw_data for path in repo.get_top_paths()]

def _get_top_referrers(repo):
    return [ref.raw_data for ref in repo.get_top_referrers()]

def _get_repository_stats(repo):
    return {
        'clones_traffic': _get_clones_traffic(repo),
        'views_traffic': _get_views_traffic(repo),
        'top_paths': _get_top_paths(repo),
        'top_referrers': _get_top_referrers(repo),
        'stargazers_count': repo.stargazers_count,
        'watchers_count': repo.watchers_count,
        'forks_count': repo.forks_count,
        'network_count': repo.network_count,
        'subscribers_count': repo.subscribers_count,
    }

def _get_organization_stats(handle, orga_name):
    return {repo.name: _get_repository_stats(repo) for repo in handle.get_organization(orga_name).get_repos()}

def main():
    handle = Github(TOKEN)
    data = {orga: _get_organization_stats(handle, orga) for orga in ORGANIZATIONS}
    with open(
        os.path.join(TARGET, '%s.json' % datetime.datetime.now().strftime('%Y-%m-%d-%H-%M-%S')),
        'w', encoding = 'utf-8'
        ) as f:
        f.write(json.dumps(data, sort_keys = True, indent = '\t'))

if __name__ == '__main__':
    main()