datapipelab 0.2.2__py3-none-any.whl → 0.2.4__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.
@@ -0,0 +1,27 @@
1
+ from datapipelab.app.node.tnode import TNode
2
+ from datapipelab.logger import logger
3
+
4
+ class SparkSinkNode(TNode):
5
+ def __init__(self, spark, tnode_config, df):
6
+ from pyspark.sql import DataFrame
7
+ super().__init__(spark=spark)
8
+ self.mode = tnode_config.get('mode', None)
9
+ self.stream = tnode_config.get('stream', None)
10
+ self.database_name = tnode_config['options']['database']
11
+ self.table_name = tnode_config['options']['table']
12
+ self.partition_by = tnode_config['options'].get('partition_by', None)
13
+ self.partition_count = tnode_config['options'].get('partition_count', None)
14
+ self.overwrite = tnode_config['options']['overwrite']
15
+ self.df = df
16
+
17
+ def __write_dynamic_partition(self):
18
+ if self.partition_count:
19
+ if self.partition_by:
20
+ self.df = self.df.repartition(int(self.partition_count))
21
+ else:
22
+ self.df = self.df.repartition(int(self.partition_count), self.partition_by)
23
+ logger.info("Start writing to Hive")
24
+ self.df.write.insertInto(f'{self.database_name}.{self.table_name}', overwrite=self.overwrite)
25
+
26
+ def _process(self):
27
+ self.__write_dynamic_partition()
@@ -88,6 +88,10 @@ class PipelineHandler:
88
88
  if tnode_format == 'hive':
89
89
  from datapipelab.app.node.sink import hive_node
90
90
  processor_df = hive_node.HiveSinkNode(self.spark, tnode_config, t_df[tnode_name_df]).run()
91
+ if tnode_type == 'spark':
92
+ if tnode_format == 'spark':
93
+ from datapipelab.app.node.sink import spark_node
94
+ processor_df = spark_node.SparkSinkNode(self.spark, tnode_config, t_df[tnode_name_df]).run()
91
95
 
92
96
 
93
97
 
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: datapipelab
3
- Version: 0.2.2
3
+ Version: 0.2.4
4
4
  Summary: A data pipeline library with connectors, sources, processors, and sinks.
5
5
  Requires-Dist: json5
6
6
  Requires-Dist: loguru
@@ -3,7 +3,7 @@ datapipelab/engine.py,sha256=3QRsedRYNov6xIDOZ1tukinFE-SKv39Fn3sNCnD3L6g,442
3
3
  datapipelab/logger.py,sha256=Ugv0A4TfD3JWCWXNWu0lURcnfAEyuVrK3IrvVVgcHBo,864
4
4
  datapipelab/pipeline.py,sha256=dw9D9KM_hztt9g_YzqoNgQBRyCYR92cRZwrU5duP_Pg,1464
5
5
  datapipelab/pipeline_config.py,sha256=2bFAJepViE7rT7CaRANZU07aeQpOYcZ954ISujm9pXA,3816
6
- datapipelab/pipeline_handler.py,sha256=dMFGpp8WgdTHlD4LuWkNzrmh3RqO2nQ1t_RyTCqIPIg,4515
6
+ datapipelab/pipeline_handler.py,sha256=1kH5jwT_YJ7LY8VfyLiRv2miLbH0TWd_6H3pKjqTABk,4763
7
7
  datapipelab/app/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
8
8
  datapipelab/app/connector_node/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
9
9
  datapipelab/app/node/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -20,12 +20,13 @@ datapipelab/app/node/sink/csv_node.py,sha256=d2hyufP5_Nmql0pfD0KeC4rFu1wXTnBxVso
20
20
  datapipelab/app/node/sink/delta_node.py,sha256=iKEdiTjJ7SHJZMrbm0jR5tms5JZ5iCFfQklZbI-Yr2o,2044
21
21
  datapipelab/app/node/sink/hive_node.py,sha256=ycknOPBBwZGH3oHram_6LjHy-ygFjhuFNvVoPaNGaCU,1220
22
22
  datapipelab/app/node/sink/pandas_csv_node.py,sha256=JsJFt2XRpwxGeJyt_PDUgqZafiQROf1Sk5TUhQPxh4c,870
23
+ datapipelab/app/node/sink/spark_node.py,sha256=tP3tZae2jzQtAtfIm8C-166WWSLdZs54mqoIyZOSy58,1221
23
24
  datapipelab/app/node/sink/teams_notification_node.py,sha256=6ZufdbhVvRXi3QTQafLo5uKl9kLyDnkYIE_VZFT0QNw,3581
24
25
  datapipelab/app/node/source/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
25
26
  datapipelab/app/node/source/delta_node.py,sha256=gg7SfuKBAAfjk6OX2jNrot9XX61HoBe3us3D8O-dscE,529
26
27
  datapipelab/app/node/source/hive_node.py,sha256=h_AMCnnmon7uLRIGsaHAPWEReD3VaWZXnz9r0TpLGNM,478
27
28
  datapipelab/app/node/source/spark_node.py,sha256=S_x2atRFPDnXmhCUtcmaLc4BDFd2H4uQq6wnEJb7Uug,480
28
- datapipelab-0.2.2.dist-info/METADATA,sha256=aRJaLx-19yUY47qYHcWyJ6pm8WCGr7YwPkeL6AH-FXg,220
29
- datapipelab-0.2.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
30
- datapipelab-0.2.2.dist-info/top_level.txt,sha256=HgeBjHvXorKzvNqU5BNPutoI771HtiqVit9_-0Zyrb4,12
31
- datapipelab-0.2.2.dist-info/RECORD,,
29
+ datapipelab-0.2.4.dist-info/METADATA,sha256=5QcU6LwZKP7_kkDtrzWFdKc3ttA9vAbRd6PurVsefYE,220
30
+ datapipelab-0.2.4.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
31
+ datapipelab-0.2.4.dist-info/top_level.txt,sha256=HgeBjHvXorKzvNqU5BNPutoI771HtiqVit9_-0Zyrb4,12
32
+ datapipelab-0.2.4.dist-info/RECORD,,