def run()

in python/dataproc_templates/elasticsearch/elasticsearch_to_gcs.py [0:0]

• 30 lines of code
• 6 McCabe index (conditional complexity)

    def run(self, spark: SparkSession, args: Dict[str, Any]) -> None:

        logger: Logger = self.get_logger(spark=spark)

        # Arguments
        es_node: str = args[constants.ES_GCS_INPUT_NODE]
        es_index: str = args[constants.ES_GCS_INPUT_INDEX]
        es_user: str = args[constants.ES_GCS_NODE_USER]
        es_password: str = args[constants.ES_GCS_NODE_PASSWORD]
        es_api_key: str = args[constants.ES_GCS_NODE_API_KEY]
        flatten_struct = args[constants.ES_GCS_FLATTEN_STRUCT]
        flatten_array = args[constants.ES_GCS_FLATTEN_ARRAY]
        output_format: str = args[constants.ES_GCS_OUTPUT_FORMAT]
        output_mode: str = args[constants.ES_GCS_OUTPUT_MODE]
        output_location: str = args[constants.ES_GCS_OUTPUT_LOCATION]

        ignore_keys = {constants.ES_GCS_NODE_PASSWORD, constants.ES_GCS_NODE_API_KEY}
        filtered_args = {key:val for key,val in args.items() if key not in ignore_keys}
        logger.info(
            "Starting ElasticSearch to Cloud Storage Spark job with parameters:\n"
            f"{pprint.pformat(filtered_args)}"
        )

        # Read
        input_data = ingest_dataframe_from_elasticsearch(
            spark, es_node, es_index, es_user, es_password, es_api_key, args, "es.gcs.input."
        )

        if flatten_struct:
            # Flatten the Struct Fields
            input_data = flatten_struct_fields(input_data)

            if flatten_array:
                # Flatten the n-D array fields to 1-D array fields
                input_data = flatten_array_fields(input_data)

        if not input_data.head(1):
            logger.info("No records in dataframe, Skipping the GCS Load")
            return

        # Write
        writer: DataFrameWriter = input_data.write.mode(output_mode)
        persist_dataframe_to_cloud_storage(writer, args, output_location, output_format, "es.gcs.output.")