def import_bulk(data_source, book_keeper):
"""
Imports bulk data from the given data source.
It can perform both 'full import' as well as 'incremental update'.
:param data_source: Data source to read input from
:param book_keeper: Book keeper to get info about recently ingested data
:return: None
"""
try:
# Now, get the last incremental update timestamp from the graph.
graph_meta = GraphPopulator.get_metadata()
# If the timestamp is unknown then it means graph is not populated yet and we need to do full import.
list_keys = []
if graph_meta is None:
# Collect all the files from data-source and group them by package-version.
logger.debug("Performing full import. Fetching all objects from : " + data_source.get_source_name())
list_keys = data_source.list_files()
# else if the timestamp is available then we need to perform incremental update.
else:
if book_keeper is None:
raise RuntimeError("Cannot perform incremental update without book keeper!")
# Collect all the package-version from RDS table that were updated recently.
# Note: If RDS table is unreachable then we should still live with S3 data.
min_finished_at = graph_meta.last_incr_update_ts
list_epv = book_keeper.get_recent_epv(min_finished_at)
# Collect relevant files from data-source and group them by package-version.
logger.debug("Performing incremental update. Fetching some objects from : " + data_source.get_source_name())
for epv in list_epv:
key_prefix = epv.get('ecosystem') + "/" + epv.get('name') + "/" + epv.get('version')
list_keys.extend(data_source.list_files(prefix=key_prefix))
# end of if graph_meta is None:
# Import the S3 data
dict_grouped_keys = _group_keys_by_epv(list_keys, data_source)
report = _import_grouped_keys(data_source, dict_grouped_keys)
# In the end, update the meta-data in the graph.
if report.get('max_finished_at') is not None:
dict_graph_meta = {
'last_incremental_update_timestamp': report.get('max_finished_at'),
'last_imported_epv': report.get('last_imported_EPV')
}
GraphPopulator.update_metadata(dict_graph_meta)
_log_report_msg("import_bulk()", report)
except Exception as e:
msg = _get_exception_msg("import_bulk() failed with error", e)
raise RuntimeError(msg)
return report
def update_graph_metadata(input_json):
GraphPopulator.update_metadata(input_json)
