public-mirrors/NewsBlur
1
0
Fork 0
mirror of https://github.com/samuelclay/NewsBlur.git synced 2025-09-18 21:50:56 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Major optimization to both single feed loads and river loads. Both are now checking training only for trained feeds, and river makes only a single subscirption call instead of N subscription calls.

Browse source
This commit is contained in:
Samuel Clay 2013-06-25 15:41:42 -07:00
parent b8e6ff4a70
commit e62b058638
2 changed files with 24 additions and 14 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
apps/reader/models.py
View file

@ -100,7 +100,7 @@ class UserSubscription(models.Model):
else:
r.delete(unread_ranked_stories_key)
if not r.exists(stories_key):
print " ---> No stories on feed: %s" % self
# print " ---> No stories on feed: %s" % self
return []
elif read_filter != 'unread' or not r.exists(read_stories_key):
ignore_user_stories = True
@ -128,7 +128,7 @@ class UserSubscription(models.Model):
max_score = int(time.mktime(self.mark_read_date.timetuple())) + 1
else:
max_score = 0
if settings.DEBUG and False:
debug_stories = r.zrevrange(unread_ranked_stories_key, 0, -1, withscores=True)
print " ---> Unread all stories (%s - %s) %s stories: %s" % (
@ -157,7 +157,8 @@ class UserSubscription(models.Model):
return []
@classmethod
def feed_stories(cls, user_id, feed_ids, offset=0, limit=6, order='newest', read_filter='all'):
def feed_stories(cls, user_id, feed_ids, offset=0, limit=6, order='newest', read_filter='all',
usersubs=None):
r = redis.Redis(connection_pool=settings.REDIS_STORY_HASH_POOL)
if order == 'oldest':
@ -178,11 +179,15 @@ class UserSubscription(models.Model):
r.delete(ranked_stories_keys)
cache.delete(unread_ranked_stories_keys)
if not usersubs:
usersubs = cls.objects.get(user=user_id, feed__in=feed_ids)
usersubs = dict((sub.feed_id, sub) for sub in usersubs)
unread_feed_story_hashes = {}
for feed_id in feed_ids:
try:
us = cls.objects.get(user=user_id, feed=feed_id)
except cls.DoesNotExist:
if feed_id in usersubs:
us = usersubs[feed_id]
else:
continue
story_hashes = us.get_stories(offset=0, limit=200,
order=order, read_filter=read_filter,

21
apps/reader/views.py
View file

@ -786,9 +786,11 @@ def load_river_stories__redis(request):
now = localtime_for_timezone(datetime.datetime.now(), user.profile.timezone)
if not feed_ids and not story_hashes:
usersubs = UserSubscription.objects.filter(user=user, active=True).only('feed')
usersubs = UserSubscription.objects.filter(user=user, active=True)
feed_ids = [sub.feed_id for sub in usersubs]
else:
usersubs = UserSubscription.objects.filter(user=user, active=True, feed__in=feed_ids)
offset = (page-1) * limit
limit = page * limit - 1
story_date_order = "%sstory_date" % ('' if order == 'oldest' else '-')
@ -800,11 +802,14 @@ def load_river_stories__redis(request):
story_hashes, unread_feed_story_hashes = UserSubscription.feed_stories(user.pk, feed_ids,
offset=offset, limit=limit,
order=order,
read_filter=read_filter)
read_filter=read_filter,
usersubs=usersubs)
mstories = MStory.objects(story_hash__in=story_hashes).order_by(story_date_order)
stories = Feed.format_stories(mstories)
found_feed_ids = list(set([story['story_feed_id'] for story in stories]))
stories, user_profiles = MSharedStory.stories_with_comments_and_profiles(stories, user.pk)
trained_feed_ids = [sub.feed_id for sub in usersubs if sub.is_trained]
found_trained_feed_ids = list(set(trained_feed_ids) & set(found_feed_ids))
# Find starred stories
if found_feed_ids:
@ -818,15 +823,15 @@ def load_river_stories__redis(request):
starred_stories = {}
# Intelligence classifiers for all feeds involved
if found_feed_ids:
if found_trained_feed_ids:
classifier_feeds = list(MClassifierFeed.objects(user_id=user.pk,
feed_id__in=found_feed_ids))
feed_id__in=found_trained_feed_ids))
classifier_authors = list(MClassifierAuthor.objects(user_id=user.pk,
feed_id__in=found_feed_ids))
feed_id__in=found_trained_feed_ids))
classifier_titles = list(MClassifierTitle.objects(user_id=user.pk,
feed_id__in=found_feed_ids))
feed_id__in=found_trained_feed_ids))
classifier_tags = list(MClassifierTag.objects(user_id=user.pk,
feed_id__in=found_feed_ids))
feed_id__in=found_trained_feed_ids))
else:
classifier_feeds = []
classifier_authors = []