Skip to content

P
physical
Commit 4e6d1d70 authored by 黄凯's avatar 黄凯
Browse files
Options

Merge branch 'hk' into 'test' 

revise read history

See merge request alpha/physical!308
parents f4fea10d 5f58738a
Hide whitespace changes
Inline Side-by-side
Showing with 106 additions and 34 deletions
.idea/misc.xml
View file @ 4e6d1d70
<?xml version="1.0" encoding="UTF-8"?> <?xml version="1.0" encoding="UTF-8"?>
<project version="4"> <project version="4">
<component name="ProjectRootManager" version="2" project-jdk-name="Python 3.6 (venv)" project-jdk-type="Python SDK" /> <component name="ProjectRootManager" version="2" project-jdk-name="Python 3.6 (physical1)" project-jdk-type="Python SDK" />
<component name="PyCharmProfessionalAdvertiser"> <component name="PyCharmProfessionalAdvertiser">
<option name="shown" value="true" /> <option name="shown" value="true" />
</component> </component>
......
.idea/physical.iml
View file @ 4e6d1d70
...@@ -2,7 +2,7 @@ ...@@ -2,7 +2,7 @@
<module type="PYTHON_MODULE" version="4"> <module type="PYTHON_MODULE" version="4">
<component name="NewModuleRootManager"> <component name="NewModuleRootManager">
<content url="file://$MODULE_DIR$" /> <content url="file://$MODULE_DIR$" />
<orderEntry type="jdk" jdkName="Python 3.6 (venv)" jdkType="Python SDK" /> <orderEntry type="jdk" jdkName="Python 3.6 (physical1)" jdkType="Python SDK" />
<orderEntry type="sourceFolder" forTests="false" /> <orderEntry type="sourceFolder" forTests="false" />
</component> </component>
<component name="TestRunnerService"> <component name="TestRunnerService">
......
libs/es.py
View file @ 4e6d1d70
...@@ -366,3 +366,74 @@ class ESPerform(object): ...@@ -366,3 +366,74 @@ class ESPerform(object):
except: except:
logging.error("catch exception,err_msg:%s" % traceback.format_exc()) logging.error("catch exception,err_msg:%s" % traceback.format_exc())
return list() return list()
@classmethod
def get_tag_topic_list_dict(cls, tag_id, have_read_topic_id_list, size=100):
try:
functions_list = list()
for id in tag_id:
functions_list.append(
{
"filter": {"term": {"tag_list": id}},
"weight": 1
}
)
functions_list += [
{
"filter": {"term": {"content_level": 6}},
"weight": 6000
},
{
"filter": {"term": {"content_level": 5}},
"weight": 5000
},
{
"filter": {"term": {"content_level": 4}},
"weight": 4000
}
]
q = {
"query": {
"function_score": {
"query": {
"bool": {
"must": [
{"range": {"content_level": {"gte": 4, "lte": 6}}},
{"term": {"is_online": True}},
{"term": {"is_deleted": False}},
{"terms": {"tag_list": tag_id}}
]
}
},
"boost_mode": "sum",
"score_mode": "sum",
"functions": functions_list
}
},
"_source": {
"include": ["id"]
},
"sort": [
{"_score": {"order": "desc"}},
{"create_time_val": {"order": "desc"}},
{"language_type": {"order": "asc"}},
]
}
if len(have_read_topic_id_list) > 0:
q["query"]["function_score"]["query"]["bool"]["must_not"] = {
"terms": {
"id": have_read_topic_id_list
}
}
result_dict = ESPerform.get_search_results(ESPerform.get_cli(), sub_index_name="topic", query_body=q,
offset=0, size=size, routing="4,5,6")
topic_id_list = [item["_source"]["id"] for item in result_dict["hits"]]
# logging.info("topic_id_list:%s" % str(topic_id_list))
topic_id_dict = [{str(item["_source"]["id"]):item["_source"]["user_id"]} for item in result_dict["hits"]]
# logging.info("topic_id_list:%s" % str(topic_id_list))
return topic_id_list,topic_id_dict
except:
logging.error("catch exception,err_msg:%s" % traceback.format_exc())
return list()
linucb/views/collect_data.py
View file @ 4e6d1d70
...@@ -74,18 +74,23 @@ class CollectData(object): ...@@ -74,18 +74,23 @@ class CollectData(object):
have_read_topic_id_list.extend(promote_recommend_topic_id_list) have_read_topic_id_list.extend(promote_recommend_topic_id_list)
recommend_topic_id_list = list() recommend_topic_id_list = list()
recommend_topic_id_list_dict = list()
recommend_topic_id_list_click = list() recommend_topic_id_list_click = list()
recommend_topic_id_list_click_dict = list()
if click_topic_tag_list: if click_topic_tag_list:
if len(click_topic_tag_list)>0: if len(click_topic_tag_list)>0:
recommend_topic_id_list_click = ESPerform.get_tag_topic_list(click_topic_tag_list, recommend_topic_id_list_click,recommend_topic_id_list_click_dict = ESPerform.get_tag_topic_list_dict(click_topic_tag_list,
have_read_topic_id_list,size=2) have_read_topic_id_list,size=2)
if len(recommend_topic_id_list_click) > 0: if len(recommend_topic_id_list_click) > 0:
recommend_topic_id_list.extend(recommend_topic_id_list_click) recommend_topic_id_list.extend(recommend_topic_id_list_click)
have_read_topic_id_list.extend(recommend_topic_id_list) recommend_topic_id_list_dict.extend(recommend_topic_id_list_click_dict)
have_read_topic_id_list.extend(recommend_topic_id_list_click)
click_recommend_redis_key = self.click_recommend_redis_key_prefix + str(device_id) click_recommend_redis_key = self.click_recommend_redis_key_prefix + str(device_id)
click_redis_data_dict = { click_redis_data_dict = {
"data": json.dumps(recommend_topic_id_list), "data": json.dumps(recommend_topic_id_list),
"datadict":json.dumps(recommend_topic_id_list_dict),
"cursor": 0 "cursor": 0
} }
redis_client.hmset(click_recommend_redis_key, click_redis_data_dict) redis_client.hmset(click_recommend_redis_key, click_redis_data_dict)
...@@ -101,25 +106,30 @@ class CollectData(object): ...@@ -101,25 +106,30 @@ class CollectData(object):
b"data"] else [] b"data"] else []
cursor = int(str(redis_topic_data_dict[b"cursor"], encoding="utf-8")) cursor = int(str(redis_topic_data_dict[b"cursor"], encoding="utf-8"))
if len(recommend_topic_id_list)==0 and cursor==0 and len(redis_topic_list)>0: if len(recommend_topic_id_list)==0 and cursor==0 and len(redis_topic_list)>0:
have_read_topic_id_list.extend(redis_topic_list[:2]) have_read_topic_id_list.extend(redis_topic_list[:2])
tag_topic_dict = list()
if len(new_user_click_tag_list)>0: if len(new_user_click_tag_list)>0:
tag_topic_id_list = ESPerform.get_tag_topic_list(new_user_click_tag_list, have_read_topic_id_list) tag_topic_id_list,tag_topic_dict = ESPerform.get_tag_topic_list_dict(new_user_click_tag_list, have_read_topic_id_list)
else: else:
tag_topic_id_list = ESPerform.get_tag_topic_list(tag_id_list,have_read_topic_id_list) tag_topic_id_list,tag_topic_dict = ESPerform.get_tag_topic_list_dict(tag_id_list,have_read_topic_id_list)
if len(recommend_topic_id_list)>0 or len(new_user_click_tag_list) > 0: if len(recommend_topic_id_list)>0 or len(new_user_click_tag_list) > 0:
tag_topic_id_list = recommend_topic_id_list + tag_topic_id_list tag_topic_id_list = recommend_topic_id_list + tag_topic_id_list
tag_topic_dict = recommend_topic_id_list_dict + tag_topic_dict
redis_data_dict = { redis_data_dict = {
"data": json.dumps(tag_topic_id_list), "data": json.dumps(tag_topic_id_list),
"datadict":json.dumps(tag_topic_dict),
"cursor":0 "cursor":0
} }
redis_client.hmset(topic_recommend_redis_key,redis_data_dict) redis_client.hmset(topic_recommend_redis_key,redis_data_dict)
else: else:
if cursor<=0 and len(redis_topic_list)>0: if cursor<=0 and len(redis_topic_list)>0:
tag_topic_id_list = redis_topic_list[:2] + tag_topic_id_list tag_topic_dict = redis_topic_list[:2] + tag_topic_dict
tag_topic_id_list = list(set(tag_topic_id_list)) tag_topic_dict = list(set(tag_topic_dict))
return True return True
except: except:
......
search/utils/topic.py
View file @ 4e6d1d70
...@@ -122,7 +122,7 @@ class TopicUtils(object): ...@@ -122,7 +122,7 @@ class TopicUtils(object):
@classmethod @classmethod
def get_recommend_topic_ids(cls,user_id,tag_id,offset,size,single_size,query=None,query_type=TopicPageType.FIND_PAGE, 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=[], filter_topic_id_list=[],test_score=False,must_topic_id_list=[],recommend_tag_list=[],
user_similar_score_list=[],index_type="topic",routing=None,attention_tag_list=[],current_topic_id =-1,topic_tag_list =[],topic_user_id =-1): user_similar_score_list=[],index_type="topic",routing=None,attention_tag_list=[],linucb_user_id_list = []):
""" """
:remark:获取首页推荐帖子列表 :remark:获取首页推荐帖子列表
:param user_id: :param user_id:
...@@ -216,25 +216,7 @@ class TopicUtils(object): ...@@ -216,25 +216,7 @@ class TopicUtils(object):
"weight": 100 "weight": 100
} }
) )
if current_topic_id != -1:
if len(topic_tag_list) > 0:
functions_list.append(
{
"filter": {"bool": {
"should": {"terms": {"tag_list": topic_tag_list}}}},
"weight": 2000
}
)
if topic_user_id != -1:
functions_list.append(
{
"filter": {"bool": {
"should": {"term": {"user_id": topic_user_id}}}},
"weight": 1500
}
)
query_function_score = { query_function_score = {
"query": { "query": {
"bool": { "bool": {
...@@ -277,12 +259,14 @@ class TopicUtils(object): ...@@ -277,12 +259,14 @@ class TopicUtils(object):
query_function_score["query"]["bool"]["must_not"] = [ query_function_score["query"]["bool"]["must_not"] = [
{"terms":{"id":filter_topic_id_list}} {"terms":{"id":filter_topic_id_list}}
] ]
if current_topic_id != -1: if "must_not" in query_function_score["query"]["bool"]:
query_function_score["query"]["bool"]["must_not"] = [{ query_function_score["query"]["bool"]["must_not"] += [
"term": { {"terms": {"user_id": linucb_user_id_list}}
"id": current_topic_id ]
} else:
}] query_function_score["query"]["bool"]["must_not"] = [
{"term": {"user_id": linucb_user_id_list}}
]
if query is not None: # 搜索帖子 if query is not None: # 搜索帖子
multi_fields = { multi_fields = {
'description': 200, 'description': 200,
......
search/views/topic.py
View file @ 4e6d1d70
...@@ -87,6 +87,8 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query ...@@ -87,6 +87,8 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
recommend_topic_dict = redis_client.hgetall(topic_recommend_redis_key) recommend_topic_dict = redis_client.hgetall(topic_recommend_redis_key)
if b"data" in recommend_topic_dict: if b"data" in recommend_topic_dict:
recommend_topic_id_list = json.loads(recommend_topic_dict[b"data"]) recommend_topic_id_list = json.loads(recommend_topic_dict[b"data"])
recommend_topic_id_dict = json.loads(recommend_topic_dict[b"datadict"])
# 推荐帖子是强插的，要保证推荐帖子不在已读里 # 推荐帖子是强插的，要保证推荐帖子不在已读里
recommend_topic_id_list = list(set(recommend_topic_id_list) - set(have_read_topic_id_list)) recommend_topic_id_list = list(set(recommend_topic_id_list) - set(have_read_topic_id_list))
cursor = int(str(recommend_topic_dict[b"cursor"], encoding="utf-8")) cursor = int(str(recommend_topic_dict[b"cursor"], encoding="utf-8"))
...@@ -95,6 +97,10 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query ...@@ -95,6 +97,10 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
recommend_topic_list = recommend_topic_id_list[cursor:newcursor] recommend_topic_list = recommend_topic_id_list[cursor:newcursor]
redis_client.hset(topic_recommend_redis_key, "cursor", newcursor) redis_client.hset(topic_recommend_redis_key, "cursor", newcursor)
recommend_topic_user_list = list()
if len(recommend_topic_list) == 6:
for i in recommend_topic_list:
recommend_topic_user_list.append(i.values())
# 用户关注标签 # 用户关注标签
redis_tag_data = redis_client.hget("physical:linucb:register_user_tag_info", user_id) redis_tag_data = redis_client.hget("physical:linucb:register_user_tag_info", user_id)
attention_tag_list = json.loads(redis_tag_data) if redis_tag_data else [] attention_tag_list = json.loads(redis_tag_data) if redis_tag_data else []
...@@ -113,7 +119,8 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query ...@@ -113,7 +119,8 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
topic_id_list = list() topic_id_list = list()
rank_topic_id_list = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=tag_id, offset=0, size=size, rank_topic_id_list = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=tag_id, offset=0, size=size,
single_size=size,query=query, query_type=query_type, single_size=size,query=query, query_type=query_type,
filter_topic_id_list=have_read_topic_id_list,index_type="topic-high-star",routing="4,5,6",attention_tag_list=attention_tag_list) filter_topic_id_list=have_read_topic_id_list,
user_similar_score_list=user_similar_score_redis_list,index_type="topic-high-star",routing="4,5,6",attention_tag_list=attention_tag_list,linucb_user_id_list=recommend_topic_user_list)
if len(recommend_topic_list) == 6 and query is None: if len(recommend_topic_list) == 6 and query is None:
if (size < 11): if (size < 11):
......
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