Skip to content

P
physical
Commit 87138c5c authored by 段英荣's avatar 段英荣
Browse files
Options

Merge branch 'similar_sort' into 'master' 

modify

See merge request !175
parents e03214dd 0b8524b3
Hide whitespace changes
Inline Side-by-side
Showing with 75 additions and 57 deletions
search/utils/topic.py
View file @ 87138c5c
......@@ -121,7 +121,9 @@ class TopicUtils(object):
return {}
@classmethod
def get_recommend_topic_ids(cls,user_id,tag_id,offset,size,query=None,query_type=TopicPageType.FIND_PAGE,filter_topic_id_list=[],test_score=False,must_topic_id_list=[],recommend_tag_list=[],user_similar_score_list=[]):
def get_recommend_topic_ids(cls,user_id,tag_id,offset,size,single_size,query=None,query_type=TopicPageType.FIND_PAGE,
filter_topic_id_list=[],test_score=False,must_topic_id_list=[],recommend_tag_list=[],
user_similar_score_list=[]):
"""
:需增加打散逻辑
:remark:获取首页推荐帖子列表
......@@ -324,22 +326,40 @@ class TopicUtils(object):
result_dict = ESPerform.get_search_results(ESPerform.get_cli(), sub_index_name="topic", query_body=q,
offset=offset, size=size)
if not test_score:
topic_id_dict = dict()
for item in result_dict["hits"]:
topic_id_dict[item["_source"]["id"]] = [item["_source"]["group_id"],item["_source"]["user_id"]]
topic_id_list = list()
same_group_id_set = set()
same_user_id_set = set()
return topic_id_dict
else:
topic_id_dict = dict()
for item in result_dict["hits"]:
topic_id_dict[item["_source"]["id"]] = [item["_source"]["group_id"],item["_source"]["user_id"],item["_score"]]
return topic_id_dict
for item in result_dict["hits"]:
if item["_source"]["group_id"]>0 and item["_source"]["group_id"] not in same_group_id_set:
same_group_id_set.add(item["_source"]["id"])
topic_id_list.append(item["_source"]["id"])
else:
same_group_id_set.add(item["_source"]["id"])
if item["_source"]["user_id"] not in same_user_id_set:
same_user_id_set.add(item["_source"]["id"])
topic_id_list.append(item["_source"]["id"])
else:
same_user_id_set.add(item["_source"]["id"])
if len(topic_id_list) >= single_size:
break
if len(topic_id_list) < single_size:
for topic_id in same_group_id_set:
topic_id_list.append(topic_id)
if len(topic_id_list)>=single_size:
break
for topic_id in same_user_id_set:
topic_id_list.append(topic_id)
if len(topic_id_list)>=single_size:
break
return topic_id_list
except:
logging.error("catch exception,err_msg:%s" % traceback.format_exc())
return dict()
return list()
@classmethod
def get_topic_detail_recommend_list(cls,user_id,topic_id,topic_tag_list,topic_group_id,topic_user_id,filter_topic_user_id,have_read_topic_list,offset,size,es_cli_obj=None):
......
search/views/topic.py
View file @ 87138c5c
......@@ -25,13 +25,10 @@ def get_discover_page_topic_ids(user_id, device_id, size, query_type=TopicPageTy
have_read_topic_id_list = json.loads(redis_field_val_list[0]) if redis_field_val_list[0] else []
recommend_topic_ids = []
topic_id_dict = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=0, offset=0, size=size,
recommend_topic_ids = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=0, offset=0, size=size,
query_type=query_type,
filter_topic_id_list=have_read_topic_id_list)
for topic_id in topic_id_dict:
recommend_topic_ids.append(topic_id)
have_read_topic_id_list.extend(recommend_topic_ids)
redis_dict = {
......@@ -83,7 +80,8 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
user_similar_score_redis_list = json.loads(
redis_user_similar_score_redis_val) if redis_user_similar_score_redis_val else []
topic_id_dict = TopicUtils.get_recommend_topic_ids(user_id, tag_id, offset, size * size, query, query_type,
topic_id_list = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=tag_id, offset=offset, size=size * size,
single_size=size,query=query, query_type=query_type,
filter_topic_id_list=have_read_topic_id_list,
recommend_tag_list=recommend_tag_list,
user_similar_score_list=user_similar_score_redis_list)
......@@ -91,42 +89,43 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
have_read_user_id_set = set()
unread_topic_id_dict = dict()
# 当前页小组数量
cur_page_group_num = 0
# 当前页用户数量
cur_page_user_num = 0
for topic_id in topic_id_dict:
if topic_id_dict[topic_id][0] in have_read_group_id_set or topic_id_dict[topic_id][
1] in have_read_user_id_set:
unread_topic_id_dict[topic_id] = topic_id_dict[topic_id]
else:
if isinstance(topic_id_dict[topic_id][0], int) and topic_id_dict[topic_id][
0] > 0 and cur_page_group_num < (size * 0.9):
have_read_group_id_set.add(topic_id_dict[topic_id][0])
have_read_user_id_set.add(topic_id_dict[topic_id][1])
have_read_topic_id_list.append(topic_id)
cur_page_group_num += 1
recommend_topic_ids.append(topic_id)
elif topic_id_dict[topic_id] and cur_page_user_num < (size * 0.1):
have_read_user_id_set.add(topic_id_dict[topic_id][1])
cur_page_user_num += 1
recommend_topic_ids.append(topic_id)
have_read_topic_id_list.append(topic_id)
else:
unread_topic_id_dict[topic_id] = topic_id_dict[topic_id]
if len(recommend_topic_ids) >= size:
break
if len(recommend_topic_ids) < size and len(unread_topic_id_dict) > 0:
for unread_topic_id in unread_topic_id_dict:
if len(recommend_topic_ids) < size:
recommend_topic_ids.append(unread_topic_id)
have_read_topic_id_list.append(unread_topic_id)
else:
break
# # 当前页小组数量
# cur_page_group_num = 0
# # 当前页用户数量
# cur_page_user_num = 0
#
# for topic_id in topic_id_dict:
# if topic_id_dict[topic_id][0] in have_read_group_id_set or topic_id_dict[topic_id][
# 1] in have_read_user_id_set:
# unread_topic_id_dict[topic_id] = topic_id_dict[topic_id]
# else:
# if isinstance(topic_id_dict[topic_id][0], int) and topic_id_dict[topic_id][
# 0] > 0 and cur_page_group_num < (size * 0.9):
# have_read_group_id_set.add(topic_id_dict[topic_id][0])
# have_read_user_id_set.add(topic_id_dict[topic_id][1])
# have_read_topic_id_list.append(topic_id)
# cur_page_group_num += 1
# recommend_topic_ids.append(topic_id)
# elif topic_id_dict[topic_id] and cur_page_user_num < (size * 0.1):
# have_read_user_id_set.add(topic_id_dict[topic_id][1])
# cur_page_user_num += 1
# recommend_topic_ids.append(topic_id)
# have_read_topic_id_list.append(topic_id)
# else:
# unread_topic_id_dict[topic_id] = topic_id_dict[topic_id]
#
# if len(recommend_topic_ids) >= size:
# break
# if len(recommend_topic_ids) < size and len(unread_topic_id_dict) > 0:
# for unread_topic_id in unread_topic_id_dict:
# if len(recommend_topic_ids) < size:
# recommend_topic_ids.append(unread_topic_id)
# have_read_topic_id_list.append(unread_topic_id)
# else:
# break
have_read_topic_id_list.extend(topic_id_list)
if len(have_read_topic_id_list) > 5000:
cut_len = len(have_read_topic_id_list)-5000
have_read_topic_id_list = have_read_topic_id_list[cut_len:]
......@@ -321,11 +320,10 @@ def query_topic_by_user_similarity(topic_similarity_score_dict, offset=0, size=1
"""
try:
must_topic_id_list = list(topic_similarity_score_dict.keys())
topic_id_dict = TopicUtils.get_recommend_topic_ids(tag_id=0, user_id=-1, offset=offset, size=size,
topic_id_list = TopicUtils.get_recommend_topic_ids(tag_id=0, user_id=-1, offset=offset, size=size,
must_topic_id_list=must_topic_id_list)
recommend_topic_ids = list(topic_id_dict.keys())
return {"recommend_topic_ids": recommend_topic_ids}
return {"recommend_topic_ids": topic_id_list}
except:
logging.error("catch exception,err_msg:%s" % traceback.format_exc())
return {"recommend_topic_id": []}
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment