Script to get all the gravatars in a git repo

get_gravatars.py

#!/usr/bin/env python

import os
import hashlib
import argparse
import subprocess
from urllib import request, error

start_dir = os.path.abspath('.')


parser = argparse.ArgumentParser()
parser.add_argument('-w', '--where', default=start_dir)
parser.add_argument('-o', '--output', default='avatars')

args = parser.parse_args()

if os.path.exists(args.output):
    raise ValueError('output dir already exists')

res = subprocess.check_output('git shortlog -sne', shell=True, cwd=args.where)
nameaddr = [line.split('\t')[-1] for line in res.decode().strip().split('\n')]

nametoaddr = {}
for na in nameaddr:
    n, a = na.split('<')
    n = n.strip()
    a = a.strip()
    if a.endswith('>'):
        a = a[:-1]
    nametoaddr[n] = a

os.mkdir(args.output)
for n, a in nametoaddr.items():
    hsh = hashlib.md5(a.encode()).hexdigest()
    gravatar_url = 'http://www.gravatar.com/avatar/' + hsh + '?d=404'

    ofn = os.path.join(args.output, n+'.jpg')
    print('Writing', ofn, 'from', gravatar_url)
    try:
        request.urlretrieve(gravatar_url, ofn)
    except error.HTTPError as e:
        print('Failed due to:', e)