add merge list

eda/smart_rank/new_tag3_cold_start.py

 import json
+from itertools import chain, zip_longest
 
 import redis
 from es_tool import es_query
 
+
+def merge_lists(*lol):
+    a = zip_longest(*lol)
+    return [i for i in chain(*a) if i is not None]
+
+
 keywords = [
     "瘦脸", "双眼皮", "补水", "美白嫩肤", "瘦小腿", "除皱", "隆鼻", "牙齿矫正", "脱毛", "祛斑", "控油", "面部祛脂", "隆胸", "垫鼻基底", "缩鼻翼", "生头发", "开眼角", "瘦肩", "祛痘",
     "缩毛孔", "轮廓改善", "除眼底细纹", "眼部修复", "垫下巴", "颧骨内推", "除法令纹", "缩咬肌", "瘦臀部", "丰眼窝", "瘦手臂", "丰唇", "瘦腰腹", "瘦大腿", "填充泪沟", "洁面", "除面部细纹",
@@ -124,15 +131,22 @@ def get_es_diary(keywords, city_tag_id=-1, version=False):
     }
     q["sort"] = sort_list
     q["_source"] = {"includes": ["id", "second_demands"]}
-    es_res = es_query("diary", q, offset=0, size=500)
-    diary_pairs = []
+    es_res = es_query("diary", q, offset=0, size=5000)
+    diary_dict = {}
     for diary_info in es_res["hits"]["hits"]:
-        sd = ""
         second_demands = diary_info["_source"].get("second_demands", [])
         if second_demands and len(second_demands) > 0:
+            id = diary_info["_source"]["id"]
             sd = second_demands[0]
-        diary_pairs.append((diary_info["_source"]["id"], sd))
-    return diary_pairs
+            if sd not in diary_dict:
+                diary_dict[sd] = [id]
+            else:
+                diary_dict[sd].append(id)
+    res_pair = sorted(diary_dict.items(), key=lambda x: len(x[1]), reverse=True)
+    res_list = []
+    for _, v in res_pair:
+        res_list.append(v)
+    return merge_lists(res_list)
 
 
 if __name__ == "__main__":