sws-spark-dissemination-helper 0.0.98__py3-none-any.whl → 0.0.100__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.
- sws_spark_dissemination_helper/SWSBronzeIcebergSparkHelper.py +19 -18
- {sws_spark_dissemination_helper-0.0.98.dist-info → sws_spark_dissemination_helper-0.0.100.dist-info}/METADATA +1 -1
- {sws_spark_dissemination_helper-0.0.98.dist-info → sws_spark_dissemination_helper-0.0.100.dist-info}/RECORD +5 -5
- {sws_spark_dissemination_helper-0.0.98.dist-info → sws_spark_dissemination_helper-0.0.100.dist-info}/WHEEL +0 -0
- {sws_spark_dissemination_helper-0.0.98.dist-info → sws_spark_dissemination_helper-0.0.100.dist-info}/licenses/LICENSE +0 -0
|
@@ -464,6 +464,7 @@ class SWSBronzeIcebergSparkHelper:
|
|
|
464
464
|
logging.info(f"bronze refs: {refs}")
|
|
465
465
|
|
|
466
466
|
create_branch_query = f"ALTER TABLE {self.iceberg_tables.BRONZE.iceberg_id}.`tag_{self.tag_name}` CREATE OR REPLACE BRANCH `diss_tag_{self.tag_name}`" # AS OF VERSION `{tag_name}`
|
|
467
|
+
create_branch_query = f"ALTER TABLE {self.iceberg_tables.BRONZE.iceberg_id} CREATE OR REPLACE BRANCH `diss_tag_{self.tag_name}`"
|
|
467
468
|
logging.info(f"create_branch_query: {create_branch_query}")
|
|
468
469
|
create_branch_query_result = self.spark.sql(create_branch_query).collect()
|
|
469
470
|
|
|
@@ -493,27 +494,27 @@ class SWSBronzeIcebergSparkHelper:
|
|
|
493
494
|
logging.info(f"dimension_name: {dimension_name}")
|
|
494
495
|
logging.info(f"codes: {codes}")
|
|
495
496
|
if len(codes) != 0:
|
|
496
|
-
|
|
497
|
-
|
|
498
|
-
|
|
499
|
-
|
|
500
|
-
|
|
501
|
-
|
|
502
|
-
|
|
503
|
-
|
|
504
|
-
|
|
505
|
-
# )
|
|
506
|
-
self.disseminated_tag_df = self.disseminated_tag_df.filter(
|
|
507
|
-
col(dimension_name).isin(codes)
|
|
497
|
+
not_in_codes = ",".join([f"'{code}'" for code in codes])
|
|
498
|
+
delete_from_branch_query = f"DELETE FROM {self.iceberg_tables.BRONZE.iceberg_id}.`branch_diss_tag_{self.tag_name}` WHERE {dimension_name} NOT IN ({not_in_codes})"
|
|
499
|
+
logging.info(f"delete_from_branch_query: {delete_from_branch_query}")
|
|
500
|
+
delete_from_branch_query_result = self.spark.sql(
|
|
501
|
+
delete_from_branch_query
|
|
502
|
+
).collect()
|
|
503
|
+
|
|
504
|
+
logging.info(
|
|
505
|
+
f"result of delete_from_branch_query: {delete_from_branch_query_result}"
|
|
508
506
|
)
|
|
507
|
+
# self.disseminated_tag_df = self.disseminated_tag_df.filter(
|
|
508
|
+
# col(dimension_name).isin(codes)
|
|
509
|
+
# )
|
|
509
510
|
|
|
510
|
-
|
|
511
|
-
|
|
512
|
-
|
|
511
|
+
self.disseminated_tag_df = self.spark.read.option(
|
|
512
|
+
"branch", f"`diss_tag_{self.tag_name}`"
|
|
513
|
+
).table(self.iceberg_tables.BRONZE.iceberg_id)
|
|
513
514
|
|
|
514
|
-
self.disseminated_tag_df.writeTo(
|
|
515
|
-
|
|
516
|
-
).overwritePartitions()
|
|
515
|
+
# self.disseminated_tag_df.writeTo(
|
|
516
|
+
# f"{self.iceberg_tables.BRONZE.iceberg_id}.`branch_diss_tag_{self.tag_name}`"
|
|
517
|
+
# ).overwritePartitions()
|
|
517
518
|
|
|
518
519
|
disseminated_tag_df = self.disseminated_tag_df.withColumn(
|
|
519
520
|
"metadata", F.to_json(col("metadata"))
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.4
|
|
2
2
|
Name: sws-spark-dissemination-helper
|
|
3
|
-
Version: 0.0.
|
|
3
|
+
Version: 0.0.100
|
|
4
4
|
Summary: A Python helper package providing streamlined Spark functions for efficient data dissemination processes
|
|
5
5
|
Project-URL: Repository, https://bitbucket.org/cioapps/sws-it-python-spark-dissemination-helper
|
|
6
6
|
Author-email: Daniele Mansillo <danielemansillo@gmail.com>
|
|
@@ -1,11 +1,11 @@
|
|
|
1
|
-
sws_spark_dissemination_helper/SWSBronzeIcebergSparkHelper.py,sha256=
|
|
1
|
+
sws_spark_dissemination_helper/SWSBronzeIcebergSparkHelper.py,sha256=vCnw4PVJ41Fiqwl0nuC-bXbJ4nqelx_EsFuraP2OAuk,22079
|
|
2
2
|
sws_spark_dissemination_helper/SWSGoldIcebergSparkHelper.py,sha256=ZC7hxkppo6qmfCc2z5vm2Y2iH1901F-rx9Er9cxuzP4,16037
|
|
3
3
|
sws_spark_dissemination_helper/SWSPostgresSparkReader.py,sha256=ja7AbOfbmC_EXHCJk7UMDzzbA-LRxzPkaaUmuvcihJ8,17449
|
|
4
4
|
sws_spark_dissemination_helper/SWSSilverIcebergSparkHelper.py,sha256=zEppNq5shiHZH2yt5faWGsb5QEmpAQS0ToIrG6fmv6o,22231
|
|
5
5
|
sws_spark_dissemination_helper/__init__.py,sha256=Efjoe9V4vGXWVp-DY5P6NbRwIUr_zkZJkDmMi-lf5Bc,262
|
|
6
6
|
sws_spark_dissemination_helper/constants.py,sha256=hpHHlbojShMWRfyIelXz6c5BqFzO48Oap1zmztlMMrs,11349
|
|
7
7
|
sws_spark_dissemination_helper/utils.py,sha256=6SzrXX0xhvynRyv-vRFDbc6V4UNe_RzKKETZAtefnhg,21341
|
|
8
|
-
sws_spark_dissemination_helper-0.0.
|
|
9
|
-
sws_spark_dissemination_helper-0.0.
|
|
10
|
-
sws_spark_dissemination_helper-0.0.
|
|
11
|
-
sws_spark_dissemination_helper-0.0.
|
|
8
|
+
sws_spark_dissemination_helper-0.0.100.dist-info/METADATA,sha256=8z7Q9un-K_28KTvKWQC9jhK42XaHzepACSOFycGxFRU,2824
|
|
9
|
+
sws_spark_dissemination_helper-0.0.100.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
|
|
10
|
+
sws_spark_dissemination_helper-0.0.100.dist-info/licenses/LICENSE,sha256=zFzeb_j_6pXEHwH8Z0OpIkKFJk7vmhZjdem-K0d4zU4,1073
|
|
11
|
+
sws_spark_dissemination_helper-0.0.100.dist-info/RECORD,,
|
|
File without changes
|