Updated documentation for get_bcast_params() method

Signed-off-by: Constantin M Adam <[email protected]>

data-processing-lib/doc/spark-runtime.md

@@ -41,9 +41,11 @@ of this parameter:
 
 ## Transforms
 
-* [SparkTransformRuntimeConfiguration](../spark/src/data_processing_spark/transform/runtime_configuration.py) allows
-    to configure transform to use PySpark
-
+* [SparkTransformRuntimeConfiguration](../spark/src/data_processing_spark/runtime/spark/runtime_configuration.py)
+  allows to configure transform to use PySpark. In addition to its base class
+  [TransformRuntimeConfiguration](../python//src/data_processing/runtime/runtime_configuration.py) features,
+  this class includes `get_bcast_params()` method to get very large configuration settings. Before starting the
+  transform execution, the Spark runtime will broadcast these settings to all the workers.
 
 ## Runtime
 

data-processing-lib/spark/src/data_processing_spark/runtime/spark/runtime_configuration.py

@@ -36,7 +36,8 @@ def get_bcast_params(self, data_access_factory: DataAccessFactoryBase) -> dict[s
         """Allows retrieving and broadcasting to all the workers very large
         configuration parameters, like the list of document IDs to remove for
         fuzzy dedup, or the list of blocked web domains for block listing.  This
-        function is called after spark initialization, and before spark_context.parallelize()
+        function is called by the spark runtime after spark initialization, and
+        before spark_context.parallelize()
         :param data_access_factory - creates data_access object to download the large config parameter
         """
         return {}

data-processing-lib/spark/src/data_processing_spark/runtime/spark/transform_runtime.py

@@ -48,7 +48,8 @@ def get_bcast_params(self, data_access_factory: DataAccessFactoryBase) -> dict[s
         """Allows retrieving and broadcasting to all the workers very large
         configuration parameters, like the list of document IDs to remove for
         fuzzy dedup, or the list of blocked web domains for block listing. This
-        function is called after spark initialization, and before spark_context.parallelize().
+        function is called by the spark runtime after spark initialization, and
+        before spark_context.parallelize()
         :param data_access_factory - creates data_access object to download the large config parameter
         """
         return {}