mdbq 1.1.4__py3-none-any.whl → 1.1.6__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
mdbq/aggregation/aggregation.py
CHANGED
@@ -849,8 +849,11 @@ class DatabaseUpdate:
|
|
849
849
|
return df
|
850
850
|
|
851
851
|
|
852
|
-
def upload(path, db_name, collection_name
|
852
|
+
def upload(path, db_name, collection_name):
|
853
853
|
""" 上传一个文件夹到数据库 """
|
854
|
+
if not os.path.isdir(path):
|
855
|
+
print(f'{os.path.splitext(os.path.basename(__file__))[0]}.upload: 函数只接受文件夹路径,不是一个文件夹: {path}')
|
856
|
+
return
|
854
857
|
username, password, host, port = get_myconf.select_config_values(
|
855
858
|
target_service='home_lx',
|
856
859
|
database='mongodb',
|
@@ -888,7 +891,7 @@ def upload(path, db_name, collection_name, one_file=False):
|
|
888
891
|
db_name=db_name,
|
889
892
|
collection_name=collection_name,
|
890
893
|
)
|
891
|
-
|
894
|
+
|
892
895
|
count = 0
|
893
896
|
for root, dirs, files in os.walk(path, topdown=False):
|
894
897
|
for name in files:
|
@@ -920,12 +923,18 @@ def upload(path, db_name, collection_name, one_file=False):
|
|
920
923
|
except Exception as e:
|
921
924
|
print(name, e)
|
922
925
|
i += 1
|
923
|
-
if one_file: # 给 file_dir 函数调用
|
924
|
-
break # 每个文件夹只上传一个文件
|
925
926
|
if d.client:
|
926
927
|
d.client.close() # 必须手动关闭数据库连接
|
927
928
|
|
928
929
|
|
930
|
+
def one_file_to_mysql(file, db_name, table_name, target_service, database):
|
931
|
+
username, password, host, port = get_myconf.select_config_values(target_service=target_service, database=database)
|
932
|
+
filename = os.path.basename(file)
|
933
|
+
df = pd.read_csv(file, encoding='utf-8_sig', header=0, na_filter=False, float_precision='high')
|
934
|
+
m = mysql.MysqlUpload(username=username, password=password, host=host, port=port)
|
935
|
+
m.df_to_mysql(df=df, db_name=db_name, table_name=table_name, filename=filename)
|
936
|
+
|
937
|
+
|
929
938
|
def file_dir():
|
930
939
|
"""
|
931
940
|
按照文件记录对照表
|
@@ -941,16 +950,26 @@ def file_dir():
|
|
941
950
|
for data in datas:
|
942
951
|
# print(data)
|
943
952
|
if data['入库进度'] == 0:
|
944
|
-
sub_path, db_name,
|
945
|
-
# print(os.path.join(path, sub_path), db_name,
|
946
|
-
|
947
|
-
|
948
|
-
|
949
|
-
|
950
|
-
|
951
|
-
|
953
|
+
sub_path, db_name, table_name = data['子文件夹'], data['数据库名'], data['数据表']
|
954
|
+
# print(os.path.join(path, sub_path), db_name, table_name)
|
955
|
+
# 从每个文件夹中取出一个文件
|
956
|
+
real_path_list = []
|
957
|
+
for root, dirs, files in os.walk(os.path.join(path, sub_path), topdown=False):
|
958
|
+
for name in files:
|
959
|
+
if name.endswith('.csv') and 'baidu' not in name and '~' not in name:
|
960
|
+
real_path_list.append(os.path.join(root, name))
|
961
|
+
break
|
962
|
+
for real_path in real_path_list:
|
963
|
+
one_file_to_mysql(
|
964
|
+
file=real_path,
|
965
|
+
db_name=db_name,
|
966
|
+
table_name=table_name,
|
967
|
+
target_service='home_lx',
|
968
|
+
database='mysql'
|
969
|
+
)
|
952
970
|
data.update({'入库进度': 1}) # 更新进度
|
953
971
|
df = pd.DataFrame.from_dict(datas, orient='columns')
|
972
|
+
# print(df)
|
954
973
|
df.to_csv(os.path.join(path, filename), encoding='utf-8_sig', index=False, header=True)
|
955
974
|
|
956
975
|
|
@@ -1,7 +1,7 @@
|
|
1
1
|
mdbq/__init__.py,sha256=Il5Q9ATdX8yXqVxtP_nYqUhExzxPC_qk_WXQ_4h0exg,16
|
2
2
|
mdbq/__version__.py,sha256=y9Mp_8x0BCZSHsdLT_q5tX9wZwd5QgqrSIENLrb6vXA,62
|
3
3
|
mdbq/aggregation/__init__.py,sha256=EeDqX2Aml6SPx8363J-v1lz0EcZtgwIBYyCJV6CcEDU,40
|
4
|
-
mdbq/aggregation/aggregation.py,sha256=
|
4
|
+
mdbq/aggregation/aggregation.py,sha256=3dCahUs562As08tK9kCyOQVQquYv5aNQNzwXd7l3GDI,57041
|
5
5
|
mdbq/aggregation/df_types.py,sha256=rHLIgv82PJSFmDvXkZyOJAffXkFyyMyFO23w9tUt8EQ,7525
|
6
6
|
mdbq/aggregation/mysql_types.py,sha256=kzUAGM4FYp77tA_dvHjZNGyPoUAOU1WY5QD0uanh9I4,10418
|
7
7
|
mdbq/aggregation/optimize_data.py,sha256=jLAWtxPUuhpo4XTVrhKtT4xK3grs7r73ePQfLhxlu1I,779
|
@@ -36,7 +36,7 @@ mdbq/pbix/__init__.py,sha256=Trtfaynu9RjoTyLLYBN2xdRxTvm_zhCniUkVTAYwcjo,24
|
|
36
36
|
mdbq/pbix/pbix_refresh.py,sha256=JUjKW3bNEyoMVfVfo77UhguvS5AWkixvVhDbw4_MHco,2396
|
37
37
|
mdbq/pbix/refresh_all.py,sha256=tgy762608HMaXWynbOURIf2UVMuSPybzrDXQnOOcnZU,6102
|
38
38
|
mdbq/spider/__init__.py,sha256=RBMFXGy_jd1HXZhngB2T2XTvJqki8P_Fr-pBcwijnew,18
|
39
|
-
mdbq-1.1.
|
40
|
-
mdbq-1.1.
|
41
|
-
mdbq-1.1.
|
42
|
-
mdbq-1.1.
|
39
|
+
mdbq-1.1.6.dist-info/METADATA,sha256=V1FyrzllnS0bwYxliNCbDaqc8_MaGJNSuGqvdeNA6AE,245
|
40
|
+
mdbq-1.1.6.dist-info/WHEEL,sha256=cpQTJ5IWu9CdaPViMhC9YzF8gZuS5-vlfoFihTBC86A,91
|
41
|
+
mdbq-1.1.6.dist-info/top_level.txt,sha256=2FQ-uLnCSB-OwFiWntzmwosW3X2Xqsg0ewh1axsaylA,5
|
42
|
+
mdbq-1.1.6.dist-info/RECORD,,
|
File without changes
|
File without changes
|