Merge branch 'wordres' into 'master'

Wordres

See merge request !24

search/utils/auto_tips.py

@@ -38,8 +38,6 @@ def get_suggest_tips(query, lat, lng, offset=0, size=50):
         #
         # logging.info("get whole_query:%s" % whole_query)
 
-
-
         # hits = whole_query["hits"]["hits"]
         #
         # for item in hits:
@@ -169,21 +167,37 @@ def get_suggest_tips(query, lat, lng, offset=0, size=50):
             return ret_list
 
         else:
-            QUERY_KEY = "query:{}:search_tip"
+            query_ret_list = []
+            wordresemble_ret_list = []
+            value_data = []
+            QUERY_KEY = "query:search_tip"
             query_base64 = base64.b64encode(query.encode('utf8')).decode('utf8')
-            key = QUERY_KEY.format(query_base64)
-            labels = list(map(lambda x: x.decode("utf8"), list(redis_client.smembers(key))))
-            for i in labels:
-                ori = i.split(":")[0]
-                ori_name = base64.b64decode(ori.encode('utf8')).decode('utf8')
-                if ori_name not in have_read_tips_set:
-                    have_read_tips_set.add(ori_name)
-                    result_num = i.split(":")[1]
-                    ret_list.append(
-                        {"results_num": result_num, "ori_name": ori_name, "id": None, "is_online": True,
-                         "offline_score": 0,
-                         "type_flag": get_tips_word_type(ori_name), "highlight_name": ori_name, "describe": ""})
-
+            if redis_client.hget(QUERY_KEY, query_base64) is not None:
+                value_data = json.loads(redis_client.hget(QUERY_KEY, query_base64))
+
+            if len(value_data) > 0:
+                for i in value_data:
+                    key = list(i.keys())[0]
+                    ori_name = str(base64.b64decode(key), "utf-8")
+                    if ori_name not in have_read_tips_set:
+                        have_read_tips_set.add(ori_name)
+                        result_num = i.get(key, 0)
+                        highlight_marks = u'<ems>%s</ems>' % query
+                        highlight_name = ori_name.replace(query, highlight_marks)
+                        if ori_name == query:
+                            query_ret_list.append(
+                                {"results_num": result_num, "ori_name": ori_name, "id": None, "is_online": True,
+                                 "offline_score": 0,
+                                 "type_flag": get_tips_word_type(ori_name), "highlight_name": highlight_name,
+                                 "describe": ""})
+                        else:
+                            wordresemble_ret_list.append(
+                                {"results_num": result_num, "ori_name": ori_name, "id": None, "is_online": True,
+                                 "offline_score": 0,
+                                 "type_flag": get_tips_word_type(ori_name), "highlight_name": highlight_name,
+                                 "describe": ""})
+                ret_list.extend(query_ret_list)
+                ret_list.extend(wordresemble_ret_list)
             if len(ret_list) >= 50:
                 return ret_list[0:50]
 

trans2es/type_info.py

@@ -538,8 +538,8 @@ def tips_attr_sync_to_redis_type_info_map():
         ),
         TypeInfo(
             name='suggest',
-            type='wordrel_results_num',  # api_wordrelresemble
-            model=doctor.Hospital,
+            type='wordrel_results_num',  # api_wordrel
+            model=wordresemble.WordRel,
             query_deferred=lambda: wordresemble.WordRel.objects.filter(
                 category__in=[13, 12, 11, 9, 1]).query,
             get_data_func=WordResemble.get_all_data_name_mapping_results_to_redis,
@@ -550,7 +550,7 @@ def tips_attr_sync_to_redis_type_info_map():
         TypeInfo(
             name='suggest',
             type='wordresemble_results_num',  # api_wordrelresemble
-            model=doctor.Hospital,
+            model=wordresemble.WordRel,
             query_deferred=lambda: wordresemble.WordRel.objects.filter(
                 category__in=[13, 12, 11, 9, 1]).query,
             get_data_func=WordResemble.set_data_to_redis,

trans2es/utils/wordresemble.py

@@ -88,6 +88,23 @@ class WordResemble(object):
             # 获取日记的
             WordResemble.get_diary_data_name_mapping_results_to_redis(instance)
 
+            # 获取该关键词的同义词的数据
+            query_sql_item = wordresemble.WordRelResemble.objects.filter(wordrel_id=instance.id)
+            if query_sql_item:
+                for item in query_sql_item:
+                    instance.name = item.word
+                    # 获取百科的
+                    ItemWikiTransfer.get_wiki_data_name_mapping_results_to_redis(instance)
+
+                    # 获取美购的
+                    TagTransfer.get_tag_data_name_mapping_results_to_redis(instance)
+
+                    # 获取医生医院的
+                    DoctorTransfer.get_doctor_data_name_mapping_results_to_redis(instance)
+
+                    # 获取日记的
+                    WordResemble.get_diary_data_name_mapping_results_to_redis(instance)
+
         except:
             logging.error("catch exception,err_msg:%s" % traceback.format_exc())
 
@@ -153,41 +170,37 @@ class WordResemble(object):
     def set_data_to_redis(cls, instance):
         try:
 
-            QUERY_KEY = "query:{}:search_tip"
+            keyword_value = []
+            QUERY_KEY = "query:search_tip"
             tips_num_redis_key_prefix = "search_tips:tips_mapping_num"
             query_sql_item = wordresemble.WordRel.objects.filter(keyword=instance.keyword)
             for sql_obj in query_sql_item:
+                count = 0
                 words = list(sql_obj.all_resembles.all().values_list('word', flat=True))
+                query_base64 = base64.b64encode(instance.keyword.encode('utf8')).decode('utf8')
                 for items in words:
+                    count += 1
+                    wordresemble_value = []
                     # 先存储本体词
-                    query_base64 = base64.b64encode(instance.keyword.encode('utf8')).decode('utf8')
-                    key = QUERY_KEY.format(query_base64)
-                    value_name = base64.b64encode(items.encode('utf8')).decode('utf8')
-                    # 获取本体词的结果
+                    item_name = base64.b64encode(items.encode('utf8')).decode('utf8')
+                    # 获取该同义词的结果
                     redis_data = redis_client.hget(tips_num_redis_key_prefix, items)
                     logging.info("get redis_data:%s" % redis_data)
                     redis_val_dict = json.loads(str(redis_data, encoding="utf-8")) if redis_data else {}
                     total_count = 0
                     if 't' in redis_val_dict:
                         total_count = int(redis_val_dict['t'])
-                    # 存储到Redis
-                    value = value_name + str(":" + str(total_count))
-                    redis_client.sadd(key, value)
-
-                    # 存储近义词
-                    # query_base64 = base64.b64encode(items.encode('utf8')).decode('utf8')
-                    key = QUERY_KEY.format(value_name)
-                    # value_name = base64.b64encode(wordrm["keyword"].encode('utf8')).decode('utf8')
-                    value_names = query_base64
-                    # 获取结果
+                    value = {item_name: total_count}
+                    keyword_value.append(value)
                     redis_data = redis_client.hget(tips_num_redis_key_prefix, instance.keyword)
                     redis_val_dict = json.loads(str(redis_data, encoding="utf-8")) if redis_data else {}
                     total_count = 0
                     if 't' in redis_val_dict:
                         total_count = int(redis_val_dict['t'])
-                    # 存储结果
-                    value = value_names + str(":" + str(total_count))
-                    redis_client.sadd(key, value)
+                    value = {query_base64: total_count}
+                    wordresemble_value.append(value)
+                    if count == 1:
+                        keyword_value.append(value)
                     if words:
                         for w in words:
                             value_name_w = base64.b64encode(w.encode('utf8')).decode('utf8')
@@ -196,8 +209,12 @@ class WordResemble(object):
                             total_count = 0
                             if 't' in redis_val_dict:
                                 total_count = int(redis_val_dict['t'])
-                            value = value_name_w + str(":" + str(total_count))
-                            redis_client.sadd(key, value)
+                            value = {value_name_w: total_count}
+                            wordresemble_value.append(value)
+                    if len(wordresemble_value) > 0:
+                        redis_client.hset(QUERY_KEY, item_name, json.dumps(wordresemble_value))
+                if len(keyword_value) > 0:
+                    redis_client.hset(QUERY_KEY, query_base64, json.dumps(keyword_value))
 
         except:
             logging.error("catch exception,err_msg:%s" % traceback.format_exc())