commonFunc.py 6.7 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193
  1. from sqlalchemy import create_engine
  2. from urllib import parse
  3. import pandas as pd
  4. import yaml
  5. def update_dict(x, y):
  6. # 用字典y,来更新x
  7. # key 存在则更新,不存在则新增
  8. x.update(y)
  9. return x
  10. def get_db_engine(db_info):
  11. db_con_str = 'mysql+pymysql://%s:%s@%s:%d/%s' % \
  12. (db_info['username'],
  13. parse.quote_plus(db_info['password']),
  14. db_info['host'],
  15. db_info['port'],
  16. db_info['database'])
  17. engine = create_engine(db_con_str, connect_args={'charset': 'utf8'})
  18. return engine
  19. def is_contains_region(val1, val2):
  20. """
  21. level1
  22. region long[] 可选 地域 传值为[]表示不限;传递上一级ID时,childrenID可以不传;不允许同时传parentID和childrenID;
  23. 仅计划的campaign_type为5时,支持设置三级地域(例:山西-大同-左云,左云是三级地域)
  24. :param val1: 推荐的城市等级组合,如 "一线城市|新一线城市|二线城市|三线城市" , "不限"
  25. :param val2: 从账户配置信息表里读取的地区信息, 如 '[]', None, '[14,15,21,22,23,31,32]'
  26. :return:
  27. """
  28. # 如果val1为空,表示该字段没有参与定向组合运算,就直接取客户投放策略里面的值
  29. if val1 is None:
  30. return True, {'region': val2}
  31. sql = """
  32. select t1.city_name, t2.level, t2.region_id, t2.parent from ctop_kuaishou_city_level t1
  33. left join ctop_kuaishou_region_list_parent t2
  34. on t1.city_name = t2.name
  35. where t1.city_level in %s
  36. """ % (tuple(val1.split('|')),)
  37. with open('config/config.yaml', mode='r', encoding='utf-8') as f:
  38. config = yaml.load(f.read(), Loader=yaml.FullLoader)
  39. city_df = pd.read_sql(sql, get_db_engine(config['productDB']))
  40. # 如果【吉林】在city_level里面的话,在ctop_kuaishou_region_list_parent会同时把 吉林省和吉林市过滤出来
  41. # 按level进行降序排列后,按city_name进行去重,保留低等级的值
  42. city_df = city_df.sort_values(by='level', ascending=False)
  43. city_df.drop_duplicates(subset=['city_name'], keep='first', inplace=True)
  44. if not val2 or val2 == '不限' or val2 == '[]':
  45. return True, {'region': list(city_df['region_id'].values)}
  46. else:
  47. if val1 == '不限':
  48. return False, -1
  49. else:
  50. val = []
  51. # level1
  52. if set(city_df[city_df.level == 1].region_id.values).issubset(set(eval(val2))):
  53. val.extend(city_df[city_df.level == 1].region_id.values)
  54. else:
  55. return False, -1
  56. # level2
  57. # 元素的父级是否存在于 val2,如果存在则为包含于的关系
  58. # 元素本身是否存在于 val2 ,如果存在则为包含于的关系
  59. # 以上都不满足,val1 不包含于 val2
  60. for item in city_df[city_df.level == 2].region_id.values:
  61. parent = city_df[city_df.region_id == item].parent.values[0]
  62. if parent in eval(val2):
  63. val.append(item)
  64. elif item in eval(val2):
  65. val.append(item)
  66. else:
  67. return False, -1
  68. # level3
  69. for item in city_df[city_df.level == 3].region_id.values:
  70. parent = city_df[city_df.region_id == item].parent.values[0]
  71. if parent in eval(val2):
  72. val.append(item)
  73. elif item in eval(val2):
  74. val.append(item)
  75. else:
  76. return False, -1
  77. return True, {'region': val}
  78. def is_contains_gender(val1, val2):
  79. """
  80. :param val1: ctop_ai_kuaishou_signature_recommended_target_combine 中的 gender 字段: '男'、'女'、'不限'、None
  81. :param val2: ctop_ai_kuaishou_advertiser_strategy 中的 gender 字段: 1:女性, 2:男性,0表示不限
  82. :return: val1 是否为 val2的子集,以及传递给快手后台的值
  83. """
  84. is_contains = False
  85. val = None
  86. if val1 == '男' and val2 in (2, 0):
  87. is_contains = True
  88. val = 2
  89. if val1 == '女' and val2 in (1, 0):
  90. is_contains = True
  91. val = 1
  92. if val1 == '不限' and val2 == 0:
  93. is_contains = True
  94. val = val2
  95. # 如果val1为空,表示该字段没有参与定向组合运算,就直接取客户投放策略里面的值
  96. if val1 is None:
  97. is_contains = True
  98. val = val2
  99. return is_contains, {'gender': val}
  100. def is_contains_platform_os(val1, val2):
  101. """
  102. :param val1: ctop_ai_kuaishou_signature_recommended_target_combine 中的 client 字段: '男'、'女'、'不限'、None
  103. :param val2:
  104. :return:
  105. """
  106. pass
  107. def is_contains_age(val1, val2_min, val2_max, val2_range):
  108. """
  109. :param val1: 推荐的年龄等级组合,如 "31-40岁|41-49岁|50+岁" , "不限"
  110. age struct 可选 自定义年龄段 不传值表示不限,传值具体见下方表格;与ages_range不能同时传
  111. min int 必填 年龄最小限制 年龄区间最小为18岁
  112. max int 必填 年龄最大限制 年龄区间最大为55岁,且年龄最大限制须大于等于年龄最小限制
  113. :param val2_min: age_min
  114. :param val2_max: age_max
  115. :param val2_range: ctop_ai_kuaishou_advertiser_strategy 中的 ages_range [] varchar
  116. 与age不能同时传;【18:表示18-23岁】;【24:表示24-30岁】;【31:表示31-40岁】;【41:表示41-49岁】;【50:表示50-100岁】
  117. :return:
  118. """
  119. # 如果val1为空,表示该字段没有参与定向组合运算,就直接取客户投放策略里面的值
  120. if val1 is None:
  121. return True, {'age_min': val2_min, 'age_max': val2_max, 'ages_range': val2_range}
  122. age_dict = {'18-23岁': 18, '24-30岁': 24, '31-40岁': 31, '41-49岁': 41, '50+岁': 50, '50-100岁': 50}
  123. if (not val2_min) and (not val2_max) and (not val2_range or val2_range == '不限'):
  124. return True, {'ages_range': val1}
  125. else:
  126. if val1 == '不限':
  127. return False, -1
  128. else:
  129. age1 = [age_dict[item] for item in val1.split('|')]
  130. if (val2_min is None and val2_max is None) and val2_range:
  131. if set(age1).issubset(set(eval(val2_range))):
  132. return True, {'ages_range': age1}
  133. else:
  134. return False, -1
  135. if (val2_min and val2_max) and (not val2_range):
  136. if age1[0] >= val2_min and age1[-1] <= val2_max:
  137. return True, {'age_min': age1[0], 'age_max': age1[-1]}
  138. else:
  139. return False, -1
  140. def is_contains_business_interest(val1, val2):
  141. pass