harvest_email_from_github.py

#!/usr/bin/env python3
"""Usage: harvest_email_from_github.py <username>

E.g.:

    $ ./harvest_email_from_github.py chadwhitacre
      7  chad@zetaweb.com  Chad Whitacre
    $

Set GITHUB_TOKEN in the environment to increase your rate limit. Get a token here:

    https://github.com/settings/tokens/new

Or use as a library:

    >>> from harvest_email_from_github import harvest_one
    >>> harvest_one('chadwhitacre')
    ({'chad@zetaweb.com': [7, {'Chad Whitacre'}]}, 1, 16)
    >>>

Requires the Requests library:

    https://pypi.org/project/requests/

"""
import os
import sys
from collections import defaultdict

import requests


log = lambda *a: print(*a, file=sys.stderr, flush=True)


# Configure
# =========

TOKEN = os.environ.get('GITHUB_TOKEN', '')


def harvest_one(username, token=TOKEN):
    """Given a username on GitHub, return a 3-tuple:

    - a dict mapping email addresses to [count, {author name, ...}] 2-lists
    - a column width for displaying counts
    - a column width for displaying email addresses

    """

    # Hit GitHub API
    # ==============

    response = requests.get(
        f'https://api.github.com/users/{username}/events/public',
        headers={'Authorization': f'token {token}'} if token else None
    )
    if response.status_code == 403:
        for k, v in response.headers.items():
            if k.startswith('X-RateLimit'):
                log(f'{k:<16}: {v}')
        if not TOKEN:
            log()
            log('Set the GITHUB_TOKEN environment variable to increase your rate limit.')
            log('Get a token here: https://github.com/settings/tokens/new.')
        exit()
    elif response.status_code != 200:
        exit(f'request failed with {response.status_code}')


    # Parse results
    # =============

    w1, w2, counts = 0, 0, defaultdict(lambda: [0, set()])
    for event in response.json():
        for commit in event['payload'].get('commits', []):
            author = commit['author']
            name, email = author['name'], author['email'].lower()
            counts[email][0] += 1
            counts[email][1].add(name)
            w1 = max(len(str(counts[email][0])), w1)
            w2 = max(len(email), w2)

    return dict(counts), w1, w2


# Display
# =======

if __name__ == '__main__':
    username = sys.argv[1] if len(sys.argv) > 1 else exit(f'usage: {sys.argv[0]} <username>')
    counts, w1, w2 = harvest_one(username)
    for email, (n, names) in sorted(counts.items(), key=lambda x: x[1], reverse=True):
        print(f"  {n:>{w1}}  {email:<{w2}}  {'; '.join(names)}")