Get public Google profile information from email addresses

get_public_google_profile_info_from_emails.py

# Some scratch code to get public profile information from Google
# based on a list of email addresses. Look at the People Search API
# for more details

import httplib2
import logging

from urllib.parse import urlparse, parse_qs
# via https://github.com/google/google-api-python-client
from apiclient.discovery import build
from django.conf import settings

# assume a contact model
from .models import Contact

logger = logging.getLogger('default')


def chunks(l, n):
    """
    http://stackoverflow.com/questions/312443/how-do-you-split-a-list-into-evenly-sized-chunks
    Yield successive n-sized chunks from l.
    """
    for i in range(0, len(l), n):
        yield l[i:i + n]


def on_contact_information(self, request_id, response, exception):
    """
    Called on receiving Google profile information. Attempts to save a few fields
    in the database via a Model named "Contact"
    """
    if exception is not None:
        # woops bai
        print(exception)
        return

    url = response.get('selfLink')
    parsed = urlparse(url)
    email = parse_qs(parsed.query)['query']
    contacts = Contact.objects.filter(email=email[0])
    if not contacts.exists():
        Contact.objects.create(email=email[0])
        # refresh the queryset
        # https://github.com/django/django/commit/6d5daa30cf29d0bada0586213ecb2959152566b4
        contacts.all()

    updates = {}
    for item in response['items']:
        if item.get('image') and item['image'].get('url'):
            updates['avatar_url'] = item['image'].get('url')

        if item.get('gender'):
            updates['gender'] = item.get('gender')

        if item.get('ageRange'):
            age_range = item['ageRange']
            updates['age_min'] = age_range.get('min')
            updates['age_max'] = age_range.get('max')

        if item.get('placesLived'):
            places_lived = item['placesLived']
            for place in places_lived:
                updates['location'] = place.get('value')
                # if we've found the primary one, get outta here
                if place.get('primary'):
                    break

    if updates:
        logger.debug(updates)

    contacts.update(**updates)


def discover_contact_information(self, email_addresses):
    """
    Discover public Google profile information for the given email
    addresses and call on_contact_information with the results
    """
    limit = 100

    http = httplib2.Http()
    service = build('plus', 'v1', developerKey=settings.GOOGLE_API_KEY)
    people_resource = service.people()

    for chunk in chunks(email_addresses, limit):
        batch = service.new_batch_http_request()
        for email in chunk:
            logger.debug('Queueing {}'.format(email))
            batch.add(people_resource.search(query=email), callback=self.on_contact_information)
        batch.execute(http=http)