Skip to content

P
physical
Commit 8a589d3c authored by lixiaofang's avatar lixiaofang
Browse files
Options

add

parent efc03161
Hide whitespace changes
Inline Side-by-side
Showing with 113 additions and 54 deletions
.idea/misc.xml
View file @ 8a589d3c
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="ProjectRootManager" version="2" project-jdk-name="Python 3.6 (physical1)" project-jdk-type="Python SDK" />
<component name="ProjectRootManager" version="2" project-jdk-name="Python 3.6 (LiXiaoFang)" project-jdk-type="Python SDK" />
<component name="PyCharmProfessionalAdvertiser">
<option name="shown" value="true" />
</component>
......
.idea/physical.iml
View file @ 8a589d3c
......@@ -2,7 +2,7 @@
<module type="PYTHON_MODULE" version="4">
<component name="NewModuleRootManager">
<content url="file://$MODULE_DIR$" />
<orderEntry type="jdk" jdkName="Python 3.6 (physical1)" jdkType="Python SDK" />
<orderEntry type="jdk" jdkName="Python 3.6 (LiXiaoFang)" jdkType="Python SDK" />
<orderEntry type="sourceFolder" forTests="false" />
</component>
<component name="TestRunnerService">
......
search/utils/topic.py
View file @ 8a589d3c
......@@ -14,7 +14,6 @@ from trans2es.models.pictorial import PictorialTopics
from libs.cache import redis_client
class TopicUtils(object):
@classmethod
......@@ -35,7 +34,7 @@ class TopicUtils(object):
}
q["_source"] = {
"include": ["tag_list","attention_user_id_list", "pick_user_id_list", "same_pictorial_user_id_list"]
"include": ["tag_list", "attention_user_id_list", "pick_user_id_list", "same_pictorial_user_id_list"]
}
result_dict = ESPerform.get_search_results(ESPerform.get_cli(), "user", q, offset, size)
......@@ -126,7 +125,8 @@ class TopicUtils(object):
def get_recommend_topic_ids(cls, user_id, tag_id, offset, size, single_size, query=None,
query_type=TopicPageType.HOME_RECOMMEND,
filter_topic_id_list=[], test_score=False, must_topic_id_list=[], recommend_tag_list=[],
user_similar_score_list=[], index_type="topic-test", routing=None, attention_tag_list=[],
user_similar_score_list=[], index_type="topic-test", routing=None,
attention_tag_list=[],
linucb_user_id_list=[], disable_collpase=False):
"""
:remark:获取首页推荐帖子列表
......@@ -174,18 +174,18 @@ class TopicUtils(object):
}
]
if user_id and user_id>0:
if user_id and user_id > 0:
redis_key_prefix = "physical:user_similar:participant_user_id:"
similar_redis_key = redis_key_prefix + str(user_id)
redis_user_similar_data = redis_client.get(similar_redis_key)
user_similar_list = json.loads(redis_user_similar_data) if redis_user_similar_data else []
if len(user_similar_list)>0:
if len(user_similar_list) > 0:
functions_list.extend(user_similar_list)
if len(attention_user_id_list) > 0:
functions_list.append(
{
"filter": {"constant_score":{"filter":{"terms": {"user_id": attention_user_id_list}}}},
"filter": {"constant_score": {"filter": {"terms": {"user_id": attention_user_id_list}}}},
"weight": 100,
}
)
......@@ -272,7 +272,7 @@ class TopicUtils(object):
]
query_function_score["query"]["bool"]["minimum_should_match"] = 1
query_function_score["query"]["bool"]["filter"].append(
{"range": {"content_level": {"gte":3,"lte":6}}}
{"range": {"content_level": {"gte": 3, "lte": 6}}}
)
else:
if "must_not" in query_function_score["query"]["bool"]:
......@@ -337,7 +337,7 @@ class TopicUtils(object):
@classmethod
def userful_tag_topic_list(cls, user_id, have_read_topic_list, size,
index_type="topic-high-star",routing=None,useful_tag_list =[]):
index_type="topic-high-star", routing=None, useful_tag_list=[]):
"""
:remark 帖子详情页推荐列表,缺少按时间衰减
:param user_id:
......@@ -371,20 +371,20 @@ class TopicUtils(object):
else:
q = dict()
q["query"] = {
"bool": {
"must": [
{"term": {"is_online": True}},
{"term": {"is_deleted": False}},
{"terms": {"useful_tag_list": useful_tag_list}},
{"term":{"content_level":6}}
],
"must_not": {
"terms": {
"id": have_read_topic_list
}
}
"bool": {
"must": [
{"term": {"is_online": True}},
{"term": {"is_deleted": False}},
{"terms": {"useful_tag_list": useful_tag_list}},
{"term": {"content_level": 6}}
],
"must_not": {
"terms": {
"id": have_read_topic_list
}
}
}
}
# logging.warning("topic_tag_list:%s"%str(topic_tag_list))
# query_function_score = {
# "query": {
......@@ -406,7 +406,8 @@ class TopicUtils(object):
"includes": ["id"]
}
result_dict = ESPerform.get_search_results(es_cli_obj, sub_index_name=index_type, query_body=q, size=size,
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()
......@@ -928,7 +929,35 @@ class TopicUtils(object):
"order": "desc"
},
})
elif sort_by == TOPIC_SEARCH_SORT.REAL_VOTE_AEC:
sort_rule.append({
"related_billboard.real_vote_cnt": {
"order": "asc",
"nested_path": "related_billboard"
},
})
elif sort_by == TOPIC_SEARCH_SORT.REAL_VOTE_DESC:
sort_rule.append({
"related_billboard.real_vote_cnt": {
"order": "desc",
"nested_path": "related_billboard"
},
})
elif sort_by == TOPIC_SEARCH_SORT.VIRT_VOTE_AEC:
sort_rule.append({
"related_billboard.virt_vote_cnt": {
"order": "asc",
"nested_path": "related_billboard"
},
})
elif sort_by == TOPIC_SEARCH_SORT.VIRT_VOTE_DESC:
sort_rule.append({
"related_billboard.virt_vote_cnt": {
"order": "desc",
"nested_path": "related_billboard"
},
})
return sort_rule
@classmethod
......@@ -976,7 +1005,6 @@ class TopicUtils(object):
def business_topic_ids(cls, filters, nfilters, sorts_by, offset=0, size=10, index_name="topic", filter_online=True):
must = cls.business_filters(filters, filter_online=filter_online)
logging.info("get must:%s" % must)
q = {
"query": {
"bool": {
......@@ -1072,6 +1100,24 @@ class TopicUtils(object):
}
}
})
# elif k == "pictorial_id":
# f.append({
# "nested": {
# "path": "related_billboard",
# "query": {
# "bool": {
# "must": [
# {
# "term": {
# "related_billboard.pictorial_id": v
# }
# }
# ]
# }
# }
# }
# })
else:
if isinstance(v, list):
......
search/views/group.py
View file @ 8a589d3c
......@@ -59,7 +59,8 @@ def pictorial_sort(user_id=-1, sort_type=GroupSortTypes.HOT_RECOMMEND, offset=0,
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 []
pictorial_ids_list = GroupUtils.get_hot_pictorial_recommend_result_list(offset, size, es_cli_obj, attention_tag_list)
pictorial_ids_list = GroupUtils.get_hot_pictorial_recommend_result_list(offset, size, es_cli_obj,
attention_tag_list)
return {"pictorial_recommend_ids": pictorial_ids_list}
elif sort_type == GroupSortTypes.ATTENTION_RECOMMEND:
......@@ -251,34 +252,39 @@ def pictorial_topic_sort(pictorial_id=-1, offset=0, size=10):
"query": {
"bool": {
"must": [
{"term": {"pictorial_id": pictorial_id}},
{"term": {"is_online": True}},
{"term": {"is_deleted": False}},
# {
# "nested": {
# "path": "related_billboard",
# "query": {
# "bool": {
# "must": [
# {
# "term": {
# "pictorial_id": pictorial_id
# }
# }
# ]
# }
# }
# }
# }
{
"nested": {
"path": "related_billboard",
"query": {
"bool": {
"must": [
{
"term": {
"related_billboard.pictorial_id": pictorial_id
}
}
]
}
}
}
}
]
}
},
"sort": [
{"total_vote_num": {"order": "desc"}},
{"related_billboard.total_vote_cnt": {
"order": "desc",
# "mode": "min",
"nested_path": "related_billboard"
}},
{"create_time": {"order": "desc"}}
]
}
logging.info("get qqqqqq:%s" % q)
pict_pictorial_ids_list = []
# 获取es链接对象
es_cli_obj = ESPerform.get_cli()
......
trans2es/mapping/topic.json
View file @ 8a589d3c
......@@ -8,18 +8,18 @@
"vote_num":{"type":"long"},
"total_vote_num":{"type":"long"},
"reply_num":{"type":"long"},
"name":{"type":"text","analyzer":"gm_default_index","search_analyzer":"gm_default_index"},
"description":{"type":"text","analyzer":"gm_default_index","search_analyzer":"gm_default_index"},
"content":{"type":"text","analyzer":"gm_default_index","search_analyzer":"gm_default_index"},
"name":{"type":"text","analyzer":"keyword","search_analyzer":"keyword"},
"description":{"type":"text","analyzer":"keyword","search_analyzer":"keyword"},
"content":{"type":"text","analyzer":"keyword","search_analyzer":"keyword"},
"content_level":{"type":"text"},
"user_id":{"type":"long"},
"user_nick_name":{"type":"text","analyzer":"gm_default_index","search_analyzer":"gm_default_index"},//帖子用户名
"user_nick_name":{"type":"text","analyzer":"keyword","search_analyzer":"keyword"},//帖子用户名
"user_nick_name_pre": {"type":"text","analyzer":"keyword"}, //不切词的用户名
"group_id":{"type":"long"}, //所在组ID
"tag_list":{"type":"long"},//标签属性
"useful_tag_list":{"type":"long"},//有用标签属性
"edit_tag_list":{"type":"long"},//编辑标签
"tag_name_list":{"type":"text","analyzer":"gm_default_index","search_analyzer":"gm_default_index"},
"tag_name_list":{"type":"text","analyzer":"keyword","search_analyzer":"keyword"},
"share_num":{"type":"long"},
"pick_id_list":{"type":"long"},
"offline_score":{"type":"double"},//离线算分
......@@ -50,8 +50,8 @@
"pictorial_id":{"type": "long"}, //所在组ID
"pictorial_name":{ // 所在组名称
"type": "text",
"analyzer": "gm_default_index",
"search_analyzer": "gm_default_index"
"analyzer": "keyword",
"search_analyzer": "keyword"
},
"is_excellent":{"type": "long"},
"is_operation_home_recommend": {"type": "boolean"}, //是否首页运营推荐
......@@ -61,7 +61,8 @@
"properties":{
"pictorial_id":{"type":"long"},
"real_vote_cnt":{"type":"long"},
"virt_vote_cnt":{"type":"long"}
"virt_vote_cnt":{"type":"long"},
"total_vote_cnt":{"type":"long"}
}
}
}
......
trans2es/models/topic.py
View file @ 8a589d3c
......@@ -113,9 +113,15 @@ class Topic(models.Model):
def get_pictorial_id(self):
try:
pictorial_id_list = []
logging.info("get pictorial_id:%s" % self.id)
pictorial_id = PictorialTopic.objects.filter(topic_id=self.id).values_list("pictorial_id", flat=True)
logging.info("get pictorial_id:%s" % pictorial_id_list)
for i in pictorial_id:
pictorial_id_list.append(i)
logging.info("get pictorial_id:%s" % pictorial_id_list)
return pictorial_id_list
except:
logging.error("catch exception,err_msg:%s" % traceback.format_exc())
......@@ -270,16 +276,16 @@ class Topic(models.Model):
logging.error("catch exception,err_msg:%s" % traceback.format_exc())
return 0.0
def get_related_billboard(self):
try:
related_billboard_list = list()
query_results = TopicBillBoard.objects.filter(id=self.id)
query_results = TopicBillBoard.objects.filter(topic_id=self.id)
for items in query_results:
total_vote_cnt = int(items.virt_vote_cnt) + int(items.real_vote_cnt)
related_billboard_list.append({"pictorial_id": items.pictorial_id, "real_vote_cnt": items.real_vote_cnt,
"virt_vote_cnt": items.virt_vote_cnt})
"virt_vote_cnt": items.virt_vote_cnt, "total_vote_cnt": total_vote_cnt})
logging.info("product_brand_info" % related_billboard_list)
......
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