posts - 431,  comments - 344,  trackbacks - 0

So, here is my version, using the cache backend:

from datetime import timedelta, datetime
from django.core.cache import cache
from django.contrib.sites.models import Site

CACHE_KEY = '%s_online_user_ids' % Site.objects.get_current().domain

_last_purged =

def get_online_user_ids():
    user_dict = cache.get(CACHE_KEY)
    return hasattr(user_dict, 'keys') and user_dict.keys() or []

class OnlineUsers(object):
    def process_request(self, request):
        if request.user.is_anonymous():

        user_dict = cache.get(CACHE_KEY)
        if not user_dict:
            # initialization
            user_dict = {}

        now =
        user_dict[] = now

        # purge
        global _last_purged
        if _last_purged + timedelta(minutes=ONLINE_MINUTES) < now:
            purge_older_than = now - timedelta(minutes=ONLINE_MINUTES)
            for user_id, last_seen in user_dict.items():
                if last_seen < purge_older_than:
            _last_purged = now

        cache.set(CACHE_KEY, user_dict, 60*60*24)

This stores a dictionary in the form: {user_id: last_seen_time, ...}
in the cache and updates the cache once for every request by an
authenticated user.

An alternative would be to store a structure like Jeremy's,
{minute_seen: set(user_id, ...), ...} which I think will result in
nearly the same amount of cache hits on average.

I would like to hear your comments.

posted on 2009-07-08 08:41 周锐 阅读(728) 评论(1)  编辑  收藏 所属分类: Python