def main(user, google_cloud_credentials_file_path, pipeline_configuration_file_path, raw_data_dir):
# Read the settings from the configuration file
log.info("Loading Pipeline Configuration File...")
with open(pipeline_configuration_file_path) as f:
pipeline_configuration = PipelineConfiguration.from_configuration_file(f)
log.info("Downloading Firestore UUID Table credentials...")
firestore_uuid_table_credentials = json.loads(google_cloud_utils.download_blob_to_string(
google_cloud_credentials_file_path,
pipeline_configuration.phone_number_uuid_table.firebase_credentials_file_url
))
phone_number_uuid_table = FirestoreUuidTable(
pipeline_configuration.phone_number_uuid_table.table_name,
firestore_uuid_table_credentials,
"avf-phone-uuid-"
)
log.info("Initialised the Firestore UUID table")
log.info(f"Fetching data from {len(pipeline_configuration.raw_data_sources)} sources...")
for i, raw_data_source in enumerate(pipeline_configuration.raw_data_sources):
log.info(f"Fetching from source {i + 1}/{len(pipeline_configuration.raw_data_sources)}...")
if isinstance(raw_data_source, RapidProSource):
fetch_from_rapid_pro(user, google_cloud_credentials_file_path, raw_data_dir, phone_number_uuid_table,
raw_data_source)
elif isinstance(raw_data_source, GCloudBucketSource):
fetch_from_gcloud_bucket(google_cloud_credentials_file_path, raw_data_dir, raw_data_source)
elif isinstance(raw_data_source, ShaqadoonCSVSource):
fetch_from_shaqadoon_csv(user, google_cloud_credentials_file_path, raw_data_dir, phone_number_uuid_table,
raw_data_source)
else:
assert False, f"Unknown raw_data_source type {type(raw_data_source)}"
def main(user, google_cloud_credentials_file_path,
pipeline_configuration_file_path, raw_data_dir):
# Read the settings from the configuration file
log.info("Loading Pipeline Configuration File...")
with open(pipeline_configuration_file_path) as f:
pipeline_configuration = PipelineConfiguration.from_configuration_file(
f)
Logger.set_project_name(pipeline_configuration.pipeline_name)
log.debug(f"Pipeline name is {pipeline_configuration.pipeline_name}")
log.info("Downloading Firestore UUID Table credentials...")
firestore_uuid_table_credentials = json.loads(
google_cloud_utils.download_blob_to_string(
google_cloud_credentials_file_path,
pipeline_configuration.uuid_table.firebase_credentials_file_url))
uuid_table = FirestoreUuidTable(
pipeline_configuration.uuid_table.table_name,
firestore_uuid_table_credentials,
pipeline_configuration.uuid_table.uuid_prefix)
log.info("Initialised the Firestore UUID table")
log.info(
f"Fetching data from {len(pipeline_configuration.raw_data_sources)} sources..."
)
for i, raw_data_source in enumerate(
pipeline_configuration.raw_data_sources):
log.info(
f"Fetching from source {i + 1}/{len(pipeline_configuration.raw_data_sources)}..."
)
if isinstance(raw_data_source, RapidProSource):
fetch_from_rapid_pro(user, google_cloud_credentials_file_path,
raw_data_dir, uuid_table, raw_data_source)
elif isinstance(raw_data_source, GCloudBucketSource):
fetch_from_gcloud_bucket(google_cloud_credentials_file_path,
raw_data_dir, raw_data_source)
elif isinstance(raw_data_source, RecoveryCSVSource):
fetch_from_recovery_csv(user, google_cloud_credentials_file_path,
raw_data_dir, uuid_table, raw_data_source)
elif isinstance(raw_data_source, FacebookSource):
fetch_from_facebook(user, google_cloud_credentials_file_path,
raw_data_dir, uuid_table, raw_data_source)
else:
assert False, f"Unknown raw_data_source type {type(raw_data_source)}"
metavar="csv-output-file-path",
help=
"Path to a CSV file to write the contacts from the locations of interest to. "
"Exported file is in a format suitable for direct upload to Rapid Pro")
args = parser.parse_args()
exclusion_list_file_path = args.exclusion_list_file_path
google_cloud_credentials_file_path = args.google_cloud_credentials_file_path
pipeline_configuration_file_path = args.pipeline_configuration_file_path
traced_data_paths = args.traced_data_paths
csv_output_file_path = args.csv_output_file_path
log.info("Loading Pipeline Configuration File...")
with open(pipeline_configuration_file_path) as f:
pipeline_configuration = PipelineConfiguration.from_configuration_file(
f)
Logger.set_project_name(pipeline_configuration.pipeline_name)
log.debug(f"Pipeline name is {pipeline_configuration.pipeline_name}")
log.info("Downloading Firestore UUID Table credentials...")
firestore_uuid_table_credentials = json.loads(
google_cloud_utils.download_blob_to_string(
google_cloud_credentials_file_path, pipeline_configuration.
phone_number_uuid_table.firebase_credentials_file_url))
phone_number_uuid_table = FirestoreUuidTable(
pipeline_configuration.phone_number_uuid_table.table_name,
firestore_uuid_table_credentials, "avf-phone-uuid-")
log.info("Initialised the Firestore UUID table")
uuids = set()
