Skip to content

P
physical
Commit 87920e61 authored by lixiaofang's avatar lixiaofang
Browse files
Options

add

parent 779336c0
Show whitespace changes
Inline Side-by-side
Showing with 111 additions and 55 deletions
search/utils/topic.py
View file @ 87920e61
......@@ -120,9 +120,11 @@ class TopicUtils(object):
return {}
@classmethod
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=[],index_type="topic",routing=None,attention_tag_list=[],linucb_user_id_list = [],disable_collpase=False):
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=[], index_type="topic", routing=None, attention_tag_list=[],
current_topic_id=-1, topic_tag_list=[], topic_user_id=-1):
"""
:remark:获取首页推荐帖子列表
:param user_id:
......@@ -158,14 +160,14 @@ class TopicUtils(object):
q["query"] = dict()
functions_list = [
# {
# "filter": {
# "term": {
# "language_type": 1
# }
# },
# "weight": 60
# },
{
"filter": {
"term": {
"language_type": 1
}
},
"weight": 60
},
{
"gauss": {
"create_time": {
......@@ -175,18 +177,18 @@ class TopicUtils(object):
},
"weight": 60
},
# {
# "filter": {
# "constant_score":{
# "filter":{
# "term": {
# "content_level": 6
# }
# }
# }
# },
# "weight": 600
# }
{
"filter": {
"constant_score": {
"filter": {
"term": {
"content_level": 6
}
}
}
},
"weight": 600
}
]
# if len(user_similar_score_list) > 0:
......@@ -216,7 +218,25 @@ class TopicUtils(object):
"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": {
"bool": {
......@@ -242,10 +262,7 @@ class TopicUtils(object):
}
}
],
"minimum_should_match": 1,
"must_not": [
{"term": {"is_history":True}}
]
"minimum_should_match": 1
}
},
"score_mode": "sum",
......@@ -260,18 +277,14 @@ class TopicUtils(object):
}
if len(filter_topic_id_list) > 0:
query_function_score["query"]["bool"]["must_not"] = [
{"terms":{"id":filter_topic_id_list}}
]
if len(linucb_user_id_list)>0:
if "must_not" in query_function_score["query"]["bool"]:
query_function_score["query"]["bool"]["must_not"] += [
{"terms": {"user_id": linucb_user_id_list}}
]
else:
query_function_score["query"]["bool"]["must_not"] = [
{"terms": {"user_id": linucb_user_id_list}}
{"terms": {"id": filter_topic_id_list}}
]
if current_topic_id != -1:
query_function_score["query"]["bool"]["must_not"] = [{
"term": {
"id": current_topic_id
}
}]
if query is not None: # 搜索帖子
multi_fields = {
'description': 200,
......@@ -289,8 +302,7 @@ class TopicUtils(object):
query_function_score["query"]["bool"]["should"] = [
{'multi_match': multi_match},
{"term": {"tag_list": tag_id}},
{"term": {"user_nick_name_pre": query.lower()}}
{"term": {"tag_list": tag_id}}
]
query_function_score["query"]["bool"]["minimum_should_match"] = 1
else:
......@@ -303,7 +315,6 @@ class TopicUtils(object):
{"term": {"is_operation_home_recommend": True}}
]
q["query"]["function_score"] = query_function_score
if not disable_collpase:
q["collapse"] = {
"field": "user_id"
}
......@@ -325,18 +336,18 @@ class TopicUtils(object):
# }
# },
{
"offline_score": {
"_score": {
"order": "desc"
}
},
{
"_score": {
"offline_score": {
"order": "desc"
}
}
]
result_dict = ESPerform.get_search_results(ESPerform.get_cli(), sub_index_name=index_type, query_body=q,
offset=offset, size=size,routing=routing)
offset=offset, size=size, routing=routing)
topic_id_list = list()
......@@ -349,7 +360,8 @@ class TopicUtils(object):
@classmethod
def get_topic_detail_recommend_list(cls, user_id, topic_id, topic_tag_list, topic_pictorial_id, topic_user_id,
filter_topic_user_id, have_read_topic_list, offset, size, es_cli_obj=None,index_type="topic",routing=None):
filter_topic_user_id, have_read_topic_list, offset, size, es_cli_obj=None,
index_type="topic", routing=None):
"""
:remark 帖子详情页推荐列表,缺少按时间衰减
:param user_id:
......@@ -426,7 +438,7 @@ class TopicUtils(object):
}
result_dict = ESPerform.get_search_results(es_cli_obj, sub_index_name=index_type, query_body=q,
offset=offset, size=size,routing=routing)
offset=offset, size=size, routing=routing)
return result_dict["hits"]
except:
......@@ -434,9 +446,10 @@ class TopicUtils(object):
return []
@classmethod
def top_get_topic_detail_recommend_list(cls, user_id, topic_id,have_read_topic_list, size, es_cli_obj=None,
index_type="topic", routing=None,collection_topic_tag_list = [],topic_tag_list = [],
topic_user_id =-1):
def top_get_topic_detail_recommend_list(cls, user_id, topic_id, have_read_topic_list, size, es_cli_obj=None,
index_type="topic", routing=None, collection_topic_tag_list=[],
topic_tag_list=[],
topic_user_id=-1):
"""
:remark 帖子详情页推荐列表,缺少按时间衰减
:param user_id:
......@@ -480,7 +493,7 @@ class TopicUtils(object):
# "weight": 5000
# }
# )
if len(topic_tag_list) != 0 or topic_user_id!= -1:
if len(topic_tag_list) != 0 or topic_user_id != -1:
query_function_score = {
"query": {
"bool": {
......@@ -548,7 +561,8 @@ class TopicUtils(object):
# {"create_time": {"order": "desc"}}
# ]
result_dict = ESPerform.get_search_results(es_cli_obj, sub_index_name=index_type, query_body=q, size=size, routing=routing)
result_dict = ESPerform.get_search_results(es_cli_obj, sub_index_name=index_type, query_body=q, size=size,
routing=routing)
topic_id_list = list()
for item in result_dict["hits"]:
......@@ -691,7 +705,7 @@ class TopicUtils(object):
{"term": {"is_deleted": False}},
]
logging.info("get filters:%s"%filters)
logging.info("get filters:%s" % filters)
if not filters:
return f
......@@ -797,7 +811,10 @@ class TopicUtils(object):
return nf
for k, v in nfilters.items():
pass
if k == "is_his":
nf.append({
"term": {k: v},
})
return nf
......@@ -894,13 +911,14 @@ class TopicUtils(object):
}
@classmethod
def business_topic_ids(cls, filters, nfilters, sorts_by, offset=0, size=10, index_name="topic", filter_online=True):
def business_topic_id(cls, filters, nfilters, sorts_by, offset=0, size=10, index_name="topic", filter_online=True):
must = cls.process_filters(filters, filter_online=filter_online)
query = ''
for k, v in filters.items():
if k == "content":
query = filters[k]
q = {}
q["query"] = {
"function_score": {
......@@ -915,7 +933,7 @@ class TopicUtils(object):
}],
"query": {
"multi_match": {
"fields":["content"],
"fields": ["content"],
"type": "cross_fields",
"operator": "and",
"query": query
......@@ -923,6 +941,7 @@ class TopicUtils(object):
}
}
}
}
if query == '':
q["query"] = {
......@@ -930,8 +949,45 @@ class TopicUtils(object):
"must": must,
"must_not": cls.process_nfilters(nfilters),
}
}
if sorts_by:
sorts = cls.process_sort(sorts_by)
if sorts:
q["sort"] = sorts
try:
result_dict = ESPerform.get_search_results(
ESPerform.get_cli(), sub_index_name=index_name,
query_body=q, offset=offset, size=size
)
return {
"hits": result_dict["hits"],
"total_count": result_dict["total_count"]
}
except:
logging.error("catch exception,err_msg:%s" % traceback.format_exc())
return {
"hits": [],
"total_count": 0
}
@classmethod
def business_topic_ids(cls, filters, nfilters, sorts_by, offset=0, size=10, index_name="topic", filter_online=True):
must = cls.process_filters(filters, filter_online=filter_online)
q = {
"query": {
"bool": {
"must": must,
"must_not": cls.process_nfilters(nfilters),
}
}
}
if sorts_by:
sorts = cls.process_sort(sorts_by)
if sorts:
......
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