mdbq 4.0.44__py3-none-any.whl → 4.0.46__py3-none-any.whl

mdbq/__version__.py

@@ -1 +1 @@
-VERSION = '4.0.44'
+VERSION = '4.0.46'

mdbq/aggregation/query_data.py

@@ -1514,25 +1514,29 @@ class MysqlDatasQuery:
             end_date=end_date,
             projection=projection,
         )
+        df.pop('data_sku')
         df['日期'] = pd.to_datetime(df['日期'], format='%Y-%m-%d', errors='ignore')  # 转换日期列
         df_set['商品id'] = df_set['商品id'].astype('int64')
         df['商品id'] = df['商品id'].astype('int64')
         df_set.sort_values('商品id', ascending=False, ignore_index=True, inplace=True)
 
+        # 仅保留最新日期的数据
+        idx = df.groupby(['商品id'])['更新时间'].idxmax()
+        df = df.loc[idx]
+
         def check_year(item_id):
             for item in df_set.to_dict(orient='records'):
                 if item_id > item['商品id']:
                     return item['上市年份']
 
         df['上市年份'] = df['商品id'].apply(lambda x: check_year(x))
-        p = df.pop('上市年份')
-        df.insert(loc=5, column='上市年份', value=p)
         set_typ = {
             '日期': 'date',
             '店铺id': 'bigint',
             '店铺名称': 'varchar(255)',
             '商家id': 'bigint',
             '商品id': 'bigint',
+            '上市年份': 'varchar(50)',
             '商品标题': 'varchar(255)',
             '商品链接': 'varchar(255)',
             '商品图片': 'varchar(255)',
@@ -1540,7 +1544,6 @@ class MysqlDatasQuery:
             '页面价': 'int',
             'data_sku': 'varchar(1000)',
             '更新时间': 'timestamp',
-            '上市年份': 'varchar(50)',
         }
         return df, {
             'db_name': db_name,
@@ -2267,7 +2270,7 @@ class MysqlDatasQuery:
             '日期': 'date',
             '店铺名称': 'varchar(100)',
             '词类型': 'varchar(100)',
-            '搜索词': 'varchar(100)',
+            '搜索词': 'varchar(255)',
             '访客数': 'int',
             '加购人数': 'int',
             '支付金额': 'decimal(10,2)',

mdbq/mysql/uploader.py

@@ -625,11 +625,45 @@ class MySQLUploader:
         根据列类型验证并转换数据值
         """
         column_type_lower = column_type.lower() if column_type else ''
-        # 统一判断None/NaN
+        
+        # JSON类型验证和转换（优先处理，避免pd.isna的问题）
+        if 'json' in column_type_lower:
+            if isinstance(value, (dict, list)):
+                try:
+                    import json
+                    return json.dumps(value, ensure_ascii=False)
+                except (TypeError, ValueError) as e:
+                    logger.error(f"JSON序列化失败: {e}", {"库": db_name, "表": table_name, "列": col_name, "值": value})
+                    raise ValueError(f"JSON序列化失败: {e}")
+            elif isinstance(value, str):
+                # 验证字符串是否为有效的JSON
+                try:
+                    import json
+                    json.loads(value)
+                    return value
+                except (TypeError, ValueError) as e:
+                    logger.error(f"无效的JSON字符串: {e}", {"库": db_name, "表": table_name, "列": col_name, "值": value})
+                    raise ValueError(f"无效的JSON字符串: {e}")
+            else:
+                # 其他类型转换为字符串
+                return str(value)
+        
+        # 统一判断None/NaN（排除列表和字典类型）
         if value == '':
             if any(t in column_type_lower for t in ['varchar', 'text', 'char', 'mediumtext', 'longtext']):
                 return ""
-        if value == '' or pd.isna(value) or (isinstance(value, (float, Decimal)) and math.isinf(value)):
+        
+        # 安全地检查NaN值，避免对列表和字典使用pd.isna
+        is_nan = False
+        if isinstance(value, (list, dict)):
+            is_nan = False  # 列表和字典不是NaN
+        else:
+            try:
+                is_nan = pd.isna(value) or (isinstance(value, (float, Decimal)) and math.isinf(value))
+            except (ValueError, TypeError):
+                is_nan = False
+        
+        if value == '' or is_nan:
             # 兜底填充值映射
             fallback_map = {
                 'int': 0,

{mdbq-4.0.44.dist-info → mdbq-4.0.46.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: mdbq
-Version: 4.0.44
+Version: 4.0.46
 Home-page: https://pypi.org/project/mdbq
 Author: xigua, 
 Author-email: 2587125111@qq.com

{mdbq-4.0.44.dist-info → mdbq-4.0.46.dist-info}/RECORD

@@ -1,7 +1,7 @@
 mdbq/__init__.py,sha256=Il5Q9ATdX8yXqVxtP_nYqUhExzxPC_qk_WXQ_4h0exg,16
-mdbq/__version__.py,sha256=Yq_JgKwKONwVexEyE66trDbripXgbesAkvt1eQ___20,18
+mdbq/__version__.py,sha256=_sPwKl25zozCOVLAfNKDLrJOiNvhHbBB2Oh22FfZsGk,18
 mdbq/aggregation/__init__.py,sha256=EeDqX2Aml6SPx8363J-v1lz0EcZtgwIBYyCJV6CcEDU,40
-mdbq/aggregation/query_data.py,sha256=ZWLJghNiEtyA4rvgUqMCLorY0R4-Likd6i4mVMuOni0,170025
+mdbq/aggregation/query_data.py,sha256=gQUcdrK0QCA0nEBkRA9zBieRWqSnkvpV5Eb-hXKw9Y8,170094
 mdbq/log/__init__.py,sha256=Mpbrav0s0ifLL7lVDAuePEi1hJKiSHhxcv1byBKDl5E,15
 mdbq/log/mylogger.py,sha256=kPe3wsQNaB1slfX-Z7VMqzZoMoqPfc7ylYXZDBeFzzI,24945
 mdbq/myconf/__init__.py,sha256=jso1oHcy6cJEfa7udS_9uO5X6kZLoPBF8l3wCYmr5dM,18
@@ -11,7 +11,7 @@ mdbq/mysql/deduplicator.py,sha256=AB3gL7ZwhcmzGHSu4UY4M6YZVPFZ2wlAN3BCcwAhegQ,73
 mdbq/mysql/mysql.py,sha256=pDg771xBugCMSTWeskIFTi3pFLgaqgyG3smzf-86Wn8,56772
 mdbq/mysql/s_query.py,sha256=1wJ3HVjHEF6FA-bVeeesRlsf73CZSvVTEQ51CF1OsE4,46786
 mdbq/mysql/unique_.py,sha256=MaztT-WIyEQUs-OOYY4pFulgHVcXR1BfCy3QUz0XM_U,21127
-mdbq/mysql/uploader.py,sha256=LPfYEj7ywoAynY2Nl9gg0IurgIWd_bTwkda2ifD1TeE,86443
+mdbq/mysql/uploader.py,sha256=NKwuBZcm4597_3gnY9M5m-jnD5joaVjf28twfQfP_jE,87955
 mdbq/other/__init__.py,sha256=jso1oHcy6cJEfa7udS_9uO5X6kZLoPBF8l3wCYmr5dM,18
 mdbq/other/download_sku_picture.py,sha256=X66sVdvVgzoNzmgVJyPtd7bjEvctEKtLPblEPF65EWc,46940
 mdbq/other/error_handler.py,sha256=4p5haAXSY-P78stp4Xwo_MwAngWYqyKj5ogWIuYXMeY,12631
@@ -25,7 +25,7 @@ mdbq/redis/__init__.py,sha256=YtgBlVSMDphtpwYX248wGge1x-Ex_mMufz4-8W0XRmA,12
 mdbq/redis/getredis.py,sha256=vpBuNc22uj9Vr-_Dh25_wpwWM1e-072EAAIBdB_IpL0,23494
 mdbq/spider/__init__.py,sha256=RBMFXGy_jd1HXZhngB2T2XTvJqki8P_Fr-pBcwijnew,18
 mdbq/spider/aikucun.py,sha256=XptHjGzbout9IYzWAOQUpMMV5qEgLTU8pL1ZGt8oNEA,21868
-mdbq-4.0.44.dist-info/METADATA,sha256=6BCrO5mef08KVeODcA3rgk-gOsNtI_8_CTh0mghCKeE,364
-mdbq-4.0.44.dist-info/WHEEL,sha256=jB7zZ3N9hIM9adW7qlTAyycLYW9npaWKLRzaoVcLKcM,91
-mdbq-4.0.44.dist-info/top_level.txt,sha256=2FQ-uLnCSB-OwFiWntzmwosW3X2Xqsg0ewh1axsaylA,5
-mdbq-4.0.44.dist-info/RECORD,,
+mdbq-4.0.46.dist-info/METADATA,sha256=gwdYseVCsqG4wlrM-Upti194tEpmc1PB7TKshTGKI30,364
+mdbq-4.0.46.dist-info/WHEEL,sha256=jB7zZ3N9hIM9adW7qlTAyycLYW9npaWKLRzaoVcLKcM,91
+mdbq-4.0.46.dist-info/top_level.txt,sha256=2FQ-uLnCSB-OwFiWntzmwosW3X2Xqsg0ewh1axsaylA,5
+mdbq-4.0.46.dist-info/RECORD,,