mdbq 1.4.3__tar.gz → 1.4.5__tar.gz

{mdbq-1.4.3 → mdbq-1.4.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: mdbq
-Version: 1.4.3
+Version: 1.4.5
 Home-page: https://pypi.org/project/mdbsql
 Author: xigua, 
 Author-email: 2587125111@qq.com

{mdbq-1.4.3 → mdbq-1.4.5}/mdbq/aggregation/aggregation.py

@@ -366,6 +366,8 @@ class DatabaseUpdate:
                     df = pd.read_csv(os.path.join(root, name), encoding='utf-8_sig', header=0, na_filter=False)
                 elif name.endswith('.csv') and '类目洞察_价格分析_分析明细_汇总' in name:
                     df = pd.read_csv(os.path.join(root, name), encoding='utf-8_sig', header=0, na_filter=False)
+                elif name.endswith('.csv') and '搜索排行_搜索' in name:
+                    df = pd.read_csv(os.path.join(root, name), encoding='utf-8_sig', header=0, na_filter=False)
                 elif name.endswith('.csv') and '竞店分析-销售分析-关键指标对比' in name:
                     df = pd.read_csv(os.path.join(root, name), encoding='utf-8_sig', header=0, na_filter=False)
                 elif name.endswith('.csv') and '竞店分析-销售分析-top商品榜' in name:
@@ -592,15 +594,15 @@ class DatabaseUpdate:
                     if is_move:
                         os.remove(os.path.join(root, name))
 
-        df = self.date_table()  # 创建一个日期表
-        self.datas.append(
-            {
-                '数据库名': '聚合数据',
-                '集合名称': '日期表',
-                '数据主体': df,
-                '文件名': '日期表文件名',
-            }
-        )
+        # df = self.date_table()  # 创建一个日期表
+        # self.datas.append(
+        #     {
+        #         '数据库名': '聚合数据',
+        #         '集合名称': '日期表',
+        #         '数据主体': df,
+        #         '文件名': '日期表文件名',
+        #     }
+        # )
 
     def upload_df(self, service_databases=[{}], path=None, system_name=None):
         """
@@ -631,6 +633,8 @@ class DatabaseUpdate:
                             is_file_dtype=True,  # 默认本地文件优先: True
                         )
                         d.df_to_mongo(df=df, db_name=db_name, collection_name=collection_name)
+                    if d.client:
+                        d.client.close()
 
                 elif database == 'mysql':
                     username, password, host, port = get_myconf.select_config_values(
@@ -780,7 +784,7 @@ class DatabaseUpdate:
             encod = chardet.detect(f1).get('encoding')
         return encod
 
-    def date_table(self):
+    def date_table(self, service_databases=[{}]):
         """
         生成 pbix使用的日期表
         """
@@ -805,7 +809,29 @@ class DatabaseUpdate:
         df['索引'] = p
         df['月索引'] = mon
         df.sort_values('日期', ascending=False, ignore_index=True, inplace=True)
-        return df
+
+        for service_database in service_databases:
+            for service_name, database in service_database.items():
+                username, password, host, port = get_myconf.select_config_values(
+                    target_service=service_name,
+                    database=database,
+                )
+                m = mysql.MysqlUpload(
+                    username=username,
+                    password=password,
+                    host=host,
+                    port=port,
+                )
+                m.df_to_mysql(
+                    df=df,
+                    db_name=db_name,
+                    table_name=collection_name,
+                    df_sql=False,  # 值为 True 时使用 df.to_sql 函数上传整个表, 不会排重
+                    drop_dup=True,  # 值为 True 时检查重复数据再插入，反之直接上传，会比较慢
+                    filename=None,  # 用来追踪处理进度
+                    system_name=service_name,  # 用来追踪处理进度
+                )
+        # return df
 
 
 def upload_dir(path, db_name, collection_name, dbs={'mysql': True, 'mongodb': True}, json_path=None, system_name=None):

{mdbq-1.4.3 → mdbq-1.4.5}/mdbq/aggregation/optimize_data.py

@@ -109,4 +109,4 @@ def op_data(db_name_lists, service_databases=None, days: int = 63):
 
 
 if __name__ == '__main__':
-    op_data(service_databases=[{'home_lx': 'mysql'}], days=3650)
+    op_data(db_name_lists=['聚合数据'], service_databases=[{'company': 'mysql'}], days=3650)

{mdbq-1.4.3 → mdbq-1.4.5}/mdbq/clean/data_clean.py

@@ -1046,6 +1046,9 @@ class DataClean:
                 elif name.endswith('.csv') and '类目洞察' in name and '价格分析_分析明细_商品发现' in name:
                     t_path = str(pathlib.Path(self.source_path, '市场数据/类目洞察/价格分析/商品发现'))
                     bib(t_path, _as_month=True)
+                elif name.endswith('.csv') and '搜索排行_搜索' in name:
+                    t_path = str(pathlib.Path(self.source_path, '市场数据/搜索排行'))
+                    bib(t_path, _as_month=True)
                 elif name.endswith('.csv') and '市场排行_店铺排行' in name:
                     t_path = str(pathlib.Path(self.source_path, '市场数据/市场二级类目店铺'))
                     bib(t_path, _as_month=True)

{mdbq-1.4.3 → mdbq-1.4.5}/mdbq/company/copysh.py

@@ -347,6 +347,7 @@ def main():
             dp.new_unzip(is_move=True)
             dp.cleaning(is_move=True)  # 公司台式机需要移除自身下载的文件
             dp.upload_df(service_databases=[{'company': 'mysql'}], system_name='company')
+            dp.date_table(service_databases=[{'company': 'mysql'}])  # 因为日期表不受 days 参数控制，因此单独更新日期表
 
             # 此操作用于修改 .copysh_conf 文件，将 ch_record 改为 false (更新完成)
             w = update_conf.UpdateConf()

{mdbq-1.4.3 → mdbq-1.4.5}/mdbq.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: mdbq
-Version: 1.4.3
+Version: 1.4.5
 Home-page: https://pypi.org/project/mdbsql
 Author: xigua, 
 Author-email: 2587125111@qq.com

{mdbq-1.4.3 → mdbq-1.4.5}/setup.py

@@ -3,7 +3,7 @@
 from setuptools import setup, find_packages
 
 setup(name='mdbq',
-      version='1.4.3',
+      version='1.4.5',
       author='xigua, ',
       author_email="2587125111@qq.com",
       url='https://pypi.org/project/mdbsql',