Merge branch 'set_tag_list' into 'master'

Set tag list

See merge request alpha/physical!455

linucb/views/collect_data.py

@@ -17,6 +17,8 @@ import libs.tools as Tools
 from trans2es.models.pictorial import CommunityPictorialHomeFeed
 from libs.error import logging_exception
 import os
+from search.views.tag import get_same_tagset_ids
+
 
 class KafkaManager(object):
     consumser_obj = None
@@ -91,7 +93,8 @@ class CollectData(object):
                 recommend_lin_pictorial_id_list = list()
 
                 if click_topic_tag_list and len(click_topic_tag_list)>0:
-                    recommend_topic_id_list_click,recommend_topic_id_list_click_dict = ESPerform.get_tag_topic_list_dict(click_topic_tag_list,
+                    click_topic_tag_list_same_tagset_ids = get_same_tagset_ids(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_same_tagset_ids,
                                                                                  have_read_topic_id_list,size=2)
                     if len(recommend_topic_id_list_click) > 0:
                         recommend_topic_id_list.extend(recommend_topic_id_list_click)
@@ -120,12 +123,14 @@ class CollectData(object):
                 # 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])
                 if len(new_user_click_tag_list)>0:
-                     tag_topic_id_list,tag_topic_dict = ESPerform.get_tag_topic_list_dict(new_user_click_tag_list, have_read_topic_id_list)
-                     recommend_lin_pictorial_id_list = ESPerform.get_tag_pictorial_id_list(new_user_click_tag_list,
+                     new_user_click_tag_list_same_tagset_ids = get_same_tagset_ids(new_user_click_tag_list)
+                     tag_topic_id_list,tag_topic_dict = ESPerform.get_tag_topic_list_dict(new_user_click_tag_list_same_tagset_ids, have_read_topic_id_list)
+                     recommend_lin_pictorial_id_list = ESPerform.get_tag_pictorial_id_list(new_user_click_tag_list_same_tagset_ids,
                                                                                       have_read_lin_pictorial_id_list)
                 else:
-                     tag_topic_id_list,tag_topic_dict = ESPerform.get_tag_topic_list_dict(tag_id_list,have_read_topic_id_list)
-                     recommend_lin_pictorial_id_list = ESPerform.get_tag_pictorial_id_list(tag_id_list,
+                     tag_id_list_same_tagset_ids = get_same_tagset_ids(tag_id_list)
+                     tag_topic_id_list,tag_topic_dict = ESPerform.get_tag_topic_list_dict(tag_id_list_same_tagset_ids,have_read_topic_id_list)
+                     recommend_lin_pictorial_id_list = ESPerform.get_tag_pictorial_id_list(tag_id_list_same_tagset_ids,
                                                                                       have_read_lin_pictorial_id_list)
 
                 if len(recommend_topic_id_list)>0 or len(tag_topic_id_list)>0 or len(new_user_click_tag_list) > 0:

search/utils/topic.py

@@ -4,16 +4,14 @@
 import logging
 import traceback
 import json
-
 from alpha_types.venus import TOPIC_SEARCH_SORT
-
+from search.views.tag import get_same_tagset_ids
 from libs.es import ESPerform
 from .common import TopicDocumentField
 from search.utils.common import *
 from trans2es.models.pictorial import PictorialTopics
 from libs.cache import redis_client
 
-
 class TopicUtils(object):
 
     @classmethod
@@ -239,9 +237,10 @@ class TopicUtils(object):
                 }
                 query_function_score["boost_mode"] = "replace"
 
+                tag_ids = get_same_tagset_ids(tag_id)
                 query_function_score["query"]["bool"]["should"] = [
                     {'multi_match': multi_match},
-                    {"term": {"tag_list": tag_id}},
+                    {"terms": {"tag_list": tag_ids}},
                     {"term": {"user_nick_name_pre": query.lower()}}
                 ]
                 query_function_score["query"]["bool"]["minimum_should_match"] = 1

search/views/group.py

@@ -17,6 +17,7 @@ from trans2es.models.topic import CommunityCategoryTagRelation
 from trans2es.models.pictorial import PictorialTopics
 from trans2es.models.pictorial import CommunityPictorialActivity
 from alpha_types.venus import PICTORIAL_ACTIVITY_SORT
+from search.views.tag import get_same_tagset_ids
 
 
 @bind("physical/search/query_pictorial")
@@ -148,6 +149,10 @@ def pictorial_topic(topic_id=-1, offset=0, size=10):
             for item in result_dict["hits"]:
                 pict_pictorial_ids_list = item["_source"]["pictorial_id"]
                 topic_tag_list = item["_source"]["tag_list"]
+
+        ##拿到这些标签的同义词
+
+        topic_tag_list = get_same_tagset_ids(topic_tag_list)
         ##先拿到当前帖子对应的标签的分类ID
         tag_type_ids = CommunityCategoryTagRelation.objects.filter(tag_id__in=topic_tag_list,
                                                                    is_online=True, is_deleted=False).values_list(

search/views/topic.py

@@ -16,6 +16,7 @@ from django.conf import settings
 from libs.tools import get_have_read_lin_pictorial_id_list
 import datetime
 from trans2es.models.tag import Tag
+from search.views.tag import get_same_tagset_ids
 
 
 def get_discover_page_topic_ids(user_id, device_id, size, query_type=TopicPageType.FIND_PAGE):
@@ -32,7 +33,8 @@ def get_discover_page_topic_ids(user_id, device_id, size, query_type=TopicPageTy
         if have_read_topic_id_list == None:
             have_read_topic_id_list = list()
 
-        recommend_topic_ids,ret_data_list = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=0, offset=0, size=size,
+        recommend_topic_ids, ret_data_list = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=0, offset=0,
+                                                                                size=size,
                                                                                 single_size=size,
                                                                                 query_type=query_type,
                                                                                 filter_topic_id_list=have_read_topic_id_list,
@@ -165,11 +167,13 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
         rank_topic_id_list = list()
         ret_data_list = list()
         if size > 0:
-            rank_topic_id_list,ret_data_list = TopicUtils.get_recommend_topic_ids(user_id=user_id, tag_id=tag_id, offset=0, size=size,
+            rank_topic_id_list, ret_data_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,
                                                                                    filter_topic_id_list=have_read_topic_id_list,
-                                                                    index_type=index_type, routing=topic_star_routing,
+                                                                                   index_type=index_type,
+                                                                                   routing=topic_star_routing,
                                                                                    attention_tag_list=attention_tag_list,
                                                                                    linucb_user_id_list=recommend_topic_user_list,
                                                                                    disable_collpase=disable_collpase)
@@ -202,7 +206,7 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
         today = datetime.datetime.strptime(str(datetime.date.today()), "%Y-%m-%d")
         tomorrow = today + datetime.timedelta(days=1)
         nowTime = datetime.datetime.now()
-        expire_time = (tomorrow - nowTime).seconds + 3*60*60
+        expire_time = (tomorrow - nowTime).seconds + 3 * 60 * 60
         redis_client.expire(redis_key, expire_time)
 
         ret_list = rank_topic_id_list if query is None else ret_data_list
@@ -217,10 +221,12 @@ def get_home_recommend_topic_ids(user_id, device_id, tag_id, offset, size, query
         else:
             return [], []
 
-def get_home_recommend_pictorial_ids(user_id=-1,device_id="",size=4):
+
+def get_home_recommend_pictorial_ids(user_id=-1, device_id="", size=4):
     try:
         pictorial_recommend_redis_key = "physical:linucb:pictorial_recommend:device_id:" + str(device_id)
-        have_read_lin_pictorial_id_list = get_have_read_lin_pictorial_id_list(device_id,user_id,TopicPageType.HOME_RECOMMEND)
+        have_read_lin_pictorial_id_list = get_have_read_lin_pictorial_id_list(device_id, user_id,
+                                                                              TopicPageType.HOME_RECOMMEND)
         pictorial_recommend_dict = redis_client.hgetall(pictorial_recommend_redis_key)
         pictorial_recommend_list = list()
         if b"data" in pictorial_recommend_dict:
@@ -238,7 +244,7 @@ def get_home_recommend_pictorial_ids(user_id=-1,device_id="",size=4):
                 redis_key = "physical:home_pictorial_recommend" + ":device_id:" + str(device_id) + ":query_type:" + str(
                     TopicPageType.HOME_RECOMMEND)
             redis_field_list = 'have_read_pictorial_list'
-            redis_client.hset(redis_key, redis_field_list,have_read_lin_pictorial_id_list)
+            redis_client.hset(redis_key, redis_field_list, have_read_lin_pictorial_id_list)
             today = datetime.datetime.strptime(str(datetime.date.today()), "%Y-%m-%d")
             tomorrow = today + datetime.timedelta(days=1)
             nowTime = datetime.datetime.now()
@@ -313,7 +319,7 @@ def home_recommend(device_id="", user_id=-1, offset=0, size=10, query_type=Topic
                                                                                                      usefulrecall=usefulrecall,
                                                                                                      useful_tag_list=useful_tag_list)
                 return {"linucb_topic_ids": recommend_topic_ids, "rank_topic_ids": rank_topic_ids,
-                        "useful_topic_ids": useful_topic_ids,"linucb_pictorial_ids":recommend_pictorial_ids}
+                        "useful_topic_ids": useful_topic_ids, "linucb_pictorial_ids": recommend_pictorial_ids}
             else:
                 recommend_pictorial_ids = get_home_recommend_pictorial_ids(user_id, device_id, 4)
                 size = size - len(recommend_pictorial_ids)
@@ -322,13 +328,14 @@ def home_recommend(device_id="", user_id=-1, offset=0, size=10, query_type=Topic
                                                                                    query_type=query_type,
                                                                                    promote_topic_list=promote_topic_list)
 
-                return {"linucb_topic_ids": recommend_topic_ids, "rank_topic_ids": rank_topic_ids,"linucb_pictorial_ids":recommend_pictorial_ids}
+                return {"linucb_topic_ids": recommend_topic_ids, "rank_topic_ids": rank_topic_ids,
+                        "linucb_pictorial_ids": recommend_pictorial_ids}
     except:
         logging.error("catch exception,err_msg:%s" % traceback.format_exc())
         if usefulrecall != -1:
-            return {"linucb_topic_ids": [], "rank_topic_ids": [], "useful_topic_ids": [],"linucb_pictorial_ids":[]}
+            return {"linucb_topic_ids": [], "rank_topic_ids": [], "useful_topic_ids": [], "linucb_pictorial_ids": []}
         else:
-            return {"linucb_topic_ids": [], "rank_topic_ids": [],"linucb_pictorial_ids":[]}
+            return {"linucb_topic_ids": [], "rank_topic_ids": [], "linucb_pictorial_ids": []}
 
 
 @bind("physical/search/discover_page")
@@ -434,6 +441,7 @@ def topic_detail_page_recommend(device_id="", user_id=-1, topic_id=-1, topic_pic
         normal_tag_result = list()
         category_tag_reuslt = list()
         result = list()
+        topic_tag_list = get_same_tagset_ids(topic_tag_list)
         tag_query_results = Tag.objects.using(settings.SLAVE_DB_NAME).filter(
             id__in=topic_tag_list, is_online=True, is_deleted=False).values_list("id", "is_category")
 
@@ -589,9 +597,11 @@ 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_list,ret_data_list = TopicUtils.get_recommend_topic_ids(tag_id=0, user_id=-1, offset=offset, size=size,
+        topic_id_list, ret_data_list = TopicUtils.get_recommend_topic_ids(tag_id=0, user_id=-1, offset=offset,
+                                                                          size=size,
                                                                           single_size=size,
-                                                           must_topic_id_list=must_topic_id_list, index_type="topic",
+                                                                          must_topic_id_list=must_topic_id_list,
+                                                                          index_type="topic",
                                                                           routing="4,5,6")
 
         return {"recommend_topic_ids": topic_id_list}

trans2es/models/tag.py

@@ -107,4 +107,15 @@ class SettingsConfig(models.Model):
     key = models.IntegerField(verbose_name=u"值")
     val = models.IntegerField(verbose_name="标签内容")
 
+class CommunityTagSetRelation(models.Model):
+    class Meta:
+        verbose_name = "搜索热词"
+        db_table = "community_tag_set_relation"
+
+    id = models.IntegerField(primary_key=True, verbose_name=u"主键ID")
+    tag_id = models.IntegerField(verbose_name=u"标签ID")
+    tag_set_id = models.IntegerField(verbose_name=u"标签ID")
+    is_deleted = models.BooleanField(verbose_name=u"是否删除")
+
+