sws-spark-dissemination-helper 0.0.161__py3-none-any.whl → 0.0.163__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/SWSEasyIcebergSparkHelper.py +4 -0
- sws_spark_dissemination_helper/SWSGoldIcebergSparkHelper.py +25 -0
- {sws_spark_dissemination_helper-0.0.161.dist-info → sws_spark_dissemination_helper-0.0.163.dist-info}/METADATA +1 -1
- {sws_spark_dissemination_helper-0.0.161.dist-info → sws_spark_dissemination_helper-0.0.163.dist-info}/RECORD +6 -6
- {sws_spark_dissemination_helper-0.0.161.dist-info → sws_spark_dissemination_helper-0.0.163.dist-info}/WHEEL +0 -0
- {sws_spark_dissemination_helper-0.0.161.dist-info → sws_spark_dissemination_helper-0.0.163.dist-info}/licenses/LICENSE +0 -0
|
@@ -510,6 +510,10 @@ class SWSEasyIcebergSparkHelper:
|
|
|
510
510
|
)
|
|
511
511
|
|
|
512
512
|
def _gen_denormalied_data_sql(self) -> DataFrame:
|
|
513
|
+
self._gen_denormalized_observation_sql()
|
|
514
|
+
self._gen_denormalized_metadata_sql()
|
|
515
|
+
self._gen_grouped_metadata_sql()
|
|
516
|
+
|
|
513
517
|
return self.spark.sql(
|
|
514
518
|
f"""
|
|
515
519
|
SELECT
|
|
@@ -271,6 +271,31 @@ class SWSGoldIcebergSparkHelper:
|
|
|
271
271
|
|
|
272
272
|
return df
|
|
273
273
|
|
|
274
|
+
def write_gold_faostat_data_to_iceberg_and_csv(self, df: DataFrame) -> DataFrame:
|
|
275
|
+
"""The expected input to this function is the output of the sws disseminated function"""
|
|
276
|
+
df.writeTo(self.iceberg_tables.GOLD_FAOSTAT.iceberg_id).createOrReplace()
|
|
277
|
+
|
|
278
|
+
logging.info(
|
|
279
|
+
f"Gold FAOSTAT table written to {self.iceberg_tables.GOLD_FAOSTAT.iceberg_id}"
|
|
280
|
+
)
|
|
281
|
+
|
|
282
|
+
self.spark.sql(
|
|
283
|
+
f"ALTER TABLE {self.iceberg_tables.GOLD_FAOSTAT.iceberg_id} CREATE OR REPLACE TAG `{self.tag_name}`"
|
|
284
|
+
)
|
|
285
|
+
|
|
286
|
+
logging.info(f"gold FAOSTAT tag '{self.tag_name}' created")
|
|
287
|
+
|
|
288
|
+
df_1 = df.coalesce(1)
|
|
289
|
+
|
|
290
|
+
save_cache_csv(
|
|
291
|
+
df=df_1,
|
|
292
|
+
bucket=self.bucket,
|
|
293
|
+
prefix=self.iceberg_tables.GOLD_FAOSTAT.csv_prefix,
|
|
294
|
+
tag_name=self.tag_name,
|
|
295
|
+
)
|
|
296
|
+
|
|
297
|
+
return df
|
|
298
|
+
|
|
274
299
|
def write_gold_sws_validated_sws_dissemination_tag(
|
|
275
300
|
self, df: DataFrame, tags: Tags
|
|
276
301
|
) -> DataFrame:
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.4
|
|
2
2
|
Name: sws-spark-dissemination-helper
|
|
3
|
-
Version: 0.0.
|
|
3
|
+
Version: 0.0.163
|
|
4
4
|
Summary: A Python helper package providing streamlined Spark functions for efficient data dissemination processes
|
|
5
5
|
Project-URL: Repository, https://github.com/un-fao/fao-sws-it-python-spark-dissemination-helper
|
|
6
6
|
Author-email: Daniele Mansillo <danielemansillo@gmail.com>
|
|
@@ -1,13 +1,13 @@
|
|
|
1
1
|
sws_spark_dissemination_helper/SWSBronzeIcebergSparkHelper.py,sha256=ocuau0WtpyRwui0qwdQ_Rxh4nYPOyZoHpGKaWRa6B3Q,28868
|
|
2
2
|
sws_spark_dissemination_helper/SWSDatatablesExportHelper.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
|
3
|
-
sws_spark_dissemination_helper/SWSEasyIcebergSparkHelper.py,sha256=
|
|
4
|
-
sws_spark_dissemination_helper/SWSGoldIcebergSparkHelper.py,sha256=
|
|
3
|
+
sws_spark_dissemination_helper/SWSEasyIcebergSparkHelper.py,sha256=abNk2rsBvDcHESczU07Ouc731HH69ypFEXA0PuvxOYM,26372
|
|
4
|
+
sws_spark_dissemination_helper/SWSGoldIcebergSparkHelper.py,sha256=cHEKA4whB4tVzZrsRoGkbdSDRgdfZA77xifYvA2XZMU,22091
|
|
5
5
|
sws_spark_dissemination_helper/SWSPostgresSparkReader.py,sha256=KpG8gp8Ai9pHDiKhUOTcXWxxmFGeKEE3XKlI_Y-SveU,18453
|
|
6
6
|
sws_spark_dissemination_helper/SWSSilverIcebergSparkHelper.py,sha256=qioLv3SlJEfk0LzTiwfXRtZXVImPOJUeh9k1XwHC-pA,26225
|
|
7
7
|
sws_spark_dissemination_helper/__init__.py,sha256=42TPbk7KxAud_qY3Sr_F4F7VjyofUlxEJkUXAFQsjRo,327
|
|
8
8
|
sws_spark_dissemination_helper/constants.py,sha256=zviO6huxWTWonHv4v2M8zKr7HXCDMBGqjHx-eTfGT2A,13487
|
|
9
9
|
sws_spark_dissemination_helper/utils.py,sha256=G7lQqNRrvqZpgm9WmddD7fWsI8IVn09x1p3cV3458EA,21963
|
|
10
|
-
sws_spark_dissemination_helper-0.0.
|
|
11
|
-
sws_spark_dissemination_helper-0.0.
|
|
12
|
-
sws_spark_dissemination_helper-0.0.
|
|
13
|
-
sws_spark_dissemination_helper-0.0.
|
|
10
|
+
sws_spark_dissemination_helper-0.0.163.dist-info/METADATA,sha256=j3TwgkmurtYUIOg6LQ83EelFI88FyUTK1g4hOTsareU,2824
|
|
11
|
+
sws_spark_dissemination_helper-0.0.163.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
|
|
12
|
+
sws_spark_dissemination_helper-0.0.163.dist-info/licenses/LICENSE,sha256=zFzeb_j_6pXEHwH8Z0OpIkKFJk7vmhZjdem-K0d4zU4,1073
|
|
13
|
+
sws_spark_dissemination_helper-0.0.163.dist-info/RECORD,,
|
|
File without changes
|