def handle(self, *args, **options):
#items = Item.objects.filter(deleted=False, is_tagged=False)
#Some items were not tagged the last time, so should not be tagged this time unless the amount
#of topics has grown a significant amount.
if len(args) > 0 and args[0] == 'video': #extend the mmseg dict by adding features in train_video_channel
extend_dict_by_video_features()
return
recent_item_set = Item.objects.filter(deleted=False, is_tagged=True).order_by('-id')
if recent_item_set:
recent_item_id = recent_item_set[0].pk
items = Item.objects.filter(pk__gt=recent_item_id, deleted=False, is_tagged=False)
else:
items = Item.objects.filter(deleted=False, is_tagged=False)
tag_count = 0
for item in items:
file = settings.MEDIA_ROOT + item.file
sorted_terms = tag_html(file)
if len(sorted_terms) > 10:
range_len = 10
else:
range_len = len(sorted_terms)
is_tagged = False
for i in xrange(range_len):
topic_name = sorted_terms[i][0]
if SensitiveWord.objects.filter(name=topic_name, disabled=False):
continue
topic_set = Topic.objects.filter(name=topic_name)
if topic_set:
try:
is_tagged = True
item_topic = Item_Topic(item=item, topic=topic_set[0]) #might duplicate with user generated one
item_topic.save()
tag_count += 1
except:
continue
else: #place in candidate topic
candidate_topic_set = CandidateTopic.objects.filter(name=topic_name)
if candidate_topic_set:
candidate_topic = candidate_topic_set[0]
else:
language = 'en'
for char in topic_name:
if ord(char) >= 128:
language = 'zh'
break
candidate_topic = CandidateTopic(name=topic_name, language=language)
candidate_topic.count = candidate_topic.count + 1
candidate_topic.save()
if is_tagged:
item.is_tagged = True
item.save()
self.stdout.write('Tagged ' + str(len(items)) + ' items with ' + str(tag_count) + ' tag(s)\n')
def add_item_topic_action(item_id, user_id, topic_id=0, topic_name=''):
"""
Args:
#. item_id (long type)
#. user_id (long type)
#. topic_id (long type, if it's null, then should use topic_name instead)
#. topic_name (string type, if it's null, then should use topic_id instead)
Return:
#. Success: topic_id (long type) in case that topic id is required
#. Failed: error_message (string type)s
"""
#Topic name is only used if there is no topic found by query suggestion(or deleted)
#A rare case could also occur if query suggestion is not quick enough
if not topic_id and not topic_name:
return '话题名字不能为空'
if topic_name:
#Ajax might not be able to respond in time, so user may enter a topic_name that already exists
topic_set = Topic.objects.filter(name=topic_name)
if topic_set:
topic_id = topic_set[0].pk
else:
return_value = add_topic_action(user_id, topic_name=topic_name)
if return_value.get('topic_id', None):
topic_id = return_value.get('topic_id', None)
else:
return return_value['error_message']
topic = Topic.objects.get(pk=topic_id)
if topic.deleted or topic.merged_to_id:
return '此话题已被删除或合并'
item_topic_set = Item_Topic.objects.filter(item__id=item_id, topic__id=topic_id)
previous_user_id = None
if item_topic_set:
item_topic = item_topic_set[0]
if item_topic.deleted == False:
return '此话题已经添加这个网址'
item_topic.deleted = False
previous_user_id = item_topic.recent_user_id
else:
item_topic = Item_Topic()
item_topic.item_id = item_id
item_topic.topic_id = topic_id
item_topic.recent_user_id = user_id
item_topic.save()
if previous_user_id:
utf_previous = User_Topic_Fame.objects.get(user__pk=previous_user_id, topic__pk=topic_id)
utf_previous.reverted_in_item_count = utf_previous.reverted_in_item_count + 1
utf_previous.save()
utf_current_set = User_Topic_Fame.objects.filter(user__pk=user_id, topic__pk=topic_id)
if utf_current_set:
utf_current = utf_current_set[0]
utf_current.action_in_item_count = utf_current.action_in_item_count + 1
utf_current.save()
else:
utf_current = User_Topic_Fame()
utf_current.topic_id = topic_id
utf_current.user_id = user_id
utf_current.action_in_item_count = 1
utf_current.save()
return topic_id #success