Search for possible hire candidates from github profiles

find_candidates.py

"""
Welcome to github candidate finder!

Note: Current github api hardcode limit for searches is 1000 results

Valid arguments are either:
a. set the CREDS tuple in the code or,
b. pass one of the following:
    * A github OAUTH_TOKEN
    * A valid set of credentials (<Username>, <Password>) with which we'll
      obtain a token
"""
import sys

from requests.auth import HTTPBasicAuth
import requests
import ujson as json

PROVIDER_PREFIX = 'https://api.github.com'
SEARCH_ONE_URL = PROVIDER_PREFIX + '/users/'
AUTH_URL = PROVIDER_PREFIX + '/authorizations'
SEARCH_URL = PROVIDER_PREFIX + '/search/users'
GENERIC_HEADER = {"User-Agent": 'Find Some Candidates'}
PREVIEW_HEADER = {"Accept": 'application/vnd.github.preview'}
PREVIEW_HEADER.update(GENERIC_HEADER)
SEARCH_QUERY = 'repos:>=2 location:israel language:python'

# enter github user/pass here for hardcoded basic auth
# CREDS = ('user', 'password')


def get_access_token(username, password):
    auth_payload = {
        'client_id': '<add client id>',
        'client_secret': '<add client secret>',
        'scopes': 'user'
    }

    auth_res = requests.post(
        AUTH_URL, data=json.dumps(auth_payload),
        auth=HTTPBasicAuth(username, password),
        headers=GENERIC_HEADER)

    if auth_res.status_code in [200, 201]:
        return auth_res.json()['token']
    return None


def search_for_users(query, token):
    payload = {
        'q': query,
        'access_token': token,
        'sort': 'followers',
        'per_page': 100,
    }

    first_res = requests.get(SEARCH_URL, params=payload,
                             headers=PREVIEW_HEADER)

    results = first_res.json()
    users = results.get('items', [])
    total_count = results.get('total_count', 0)
    print "found total of {results} results matching the query {query}".format(
        results=total_count, query=query)
    if total_count > 100:
        pages = results['total_count'] // 100
        for i in range(2, int(pages + 2)):
            payload['page'] = i
            res = requests.get(SEARCH_URL, params=payload,
                               headers=PREVIEW_HEADER)
            results = res.json()
            users.extend(results.get('items', []))
    return users


def get_user_details(user, token):
    payload = {
        'access_token': token,
    }
    res = requests.get(SEARCH_ONE_URL + user['login'], params=payload)

    if res.status_code == 200:
        user['details'] = res.json()


if __name__ == '__main__':
    if 'CREDS' not in globals():
        if len(sys.argv) > 1:
            if len(sys.argv) == 2:
                # use token for authentication
                access_token = sys.argv[1]
            if len(sys.argv) > 2:
                access_token = get_access_token(*sys.argv[1:3])
        else:
            sys.exit(__doc__)
    else:
        access_token = get_access_token(*globals()['CREDS'])

    users = search_for_users(SEARCH_QUERY, access_token)
    for user in users:
        get_user_details(user, access_token)
    f = open('candidates_dump.json', 'w')
    f.write(json.dumps(users))
    f.close()

    hirable = [u for u in users if u.get('details', {}).get('hireable', False)]
    f = open('hirable_dump.json', 'w')
    f.write(json.dumps(hirable))
    f.close()

    print "Results were written to candidates_dump.json and hirable_dump.json"