list_github_users.py

# -*- coding: utf-8 -*-

# sort list.txt | uniq | grep -v '#' | grep -v 'noreply' | grep -v 'local' | grep -e '\.' | grep -v 'internal' | grep -v 'contact'

import os
import sys
import requests
import time
from github3 import login
from tqdm import tqdm

URL = "https://api.github.com/users/{name}/events/public?access_token={token}"
token = os.environ['STARGAZER_ACCESS_TOKEN']

gh = login(token=token)
repo = gh.repository('myui', 'hivemall')

def extract_email(name):
  url = URL.format(name=name, token=token)
  li = requests.get(url).json()
  time.sleep(1) # sleep 1 sec to avoid API rate limit
  try:
    for d in li:
      if d['type'] == 'PushEvent':
        return d['payload']['commits'][0]['author']['email']
  except:
    return None
  return None

def print_user(u):
  if u.email is None:
    email = extract_email(u.login)
    if email is not None:
      print email
      # print "{}|{}|{}|{}|{}".format(u.login,u.name.encode('utf-8'),email,u.company.encode('utf-8'),u.location.encode('utf-8'))
return 1
  else:
    print u.email
    return 1
  return 0

def gh_rate_limit_remaining():
  return gh.ratelimit_remaining

# print "login|name|email|company|location"
  
print '# Stargazers'
print >> sys.stderr, "API limit remaining: " + str(gh_rate_limit_remaining())
print >> sys.stderr, '# Stargazers'
cnt = 0
for x in tqdm(repo.iter_stargazers()):
  u = gh.user(x)
  cnt += print_user(u)
print >> sys.stderr, "Found: " + str(cnt)

print '# Forked Users'
print >> sys.stderr, "API limit remaining: " + str(gh_rate_limit_remaining())
print >> sys.stderr, '# Forked users'
cnt = 0
for fork in tqdm(repo.iter_forks()):
  u = fork.owner
  cnt += print_user(u)
print >> sys.stderr, "Found: " + str(cnt)

print '# Watchers'
print >> sys.stderr, "API limit remaining: " + str(gh_rate_limit_remaining())
print >> sys.stderr, '# Watchers'
cnt = 0
for u in tqdm(repo.iter_subscribers()):
  cnt += print_user(u)
print >> sys.stderr, "Found: " + str(cnt)