def marshall_table_full(table_dict: Dict) -> Dict:
"""
Forms the full version of a table Dict, with additional and sanitized fields
:param table_dict: Table Dict from metadata service
:return: Table Dict with sanitized fields
"""
schema = TableSchema(strict=True)
# TODO: consider migrating to validate() instead of roundtripping
table: Table = schema.load(table_dict).data
results: Dict[str, Any] = schema.dump(table).data
# Check if schema is uneditable
is_editable_schema = results['schema'] not in app.config[
'UNEDITABLE_SCHEMAS']
# Check if Table Description is uneditable
is_editable_table = True
uneditable_table_desc_match_rules = app.config[
'UNEDITABLE_TABLE_DESCRIPTION_MATCH_RULES']
for rule in uneditable_table_desc_match_rules:
is_editable_table = is_editable_table and _parse_editable_rule(
rule, results['schema'], results['name'])
is_editable = is_editable_schema and is_editable_table
results['is_editable'] = is_editable
# TODO - Cleanup https://github.com/lyft/amundsen/issues/296
# This code will try to supplement some missing data since the data here is incomplete.
# Once the metadata service response provides complete user objects we can remove this.
results['owners'] = [
_map_user_object_to_schema(owner) for owner in results['owners']
]
readers = results['table_readers']
for reader_object in readers:
reader_object['user'] = _map_user_object_to_schema(
reader_object['user'])
# If order is provided, we sort the column based on the pre-defined order
if app.config['COLUMN_STAT_ORDER']:
columns = results['columns']
for col in columns:
# the stat_type isn't defined in COLUMN_STAT_ORDER, we just use the max index for sorting
col['stats'].sort(
key=lambda x: app.config['COLUMN_STAT_ORDER'].get(
x['stat_type'], len(app.config['COLUMN_STAT_ORDER'])))
col['is_editable'] = is_editable
# TODO: Add the 'key' or 'id' to the base TableSchema
results[
'key'] = f'{table.database}://{table.cluster}.{table.schema}/{table.name}'
# Temp code to make 'partition_key' and 'partition_value' part of the table
results['partition'] = _get_partition_data(results['watermarks'])
# We follow same style as column stat order for arranging the programmatic descriptions
prog_descriptions = results['programmatic_descriptions']
if prog_descriptions:
_update_prog_descriptions(prog_descriptions)
return results
def get(self, table_uri: str) -> Iterable[Union[Mapping, int, None]]:
try:
table = self.client.get_table(table_uri=table_uri)
schema = TableSchema(strict=True)
return schema.dump(table).data, HTTPStatus.OK
except NotFoundException:
return {'message': 'table_uri {} does not exist'.format(table_uri)}, HTTPStatus.NOT_FOUND
def get(self, table_uri: str) -> Iterable[Union[Mapping, int, None]]:
try:
table, upstream, downstream = self.client.get_table(table_uri=table_uri)
schema = TableSchema(strict=True)
data = schema.dump(table).data
data['upstream'] = upstream
data['downstream'] = downstream
return data, HTTPStatus.OK
except NotFoundException:
return {'message': 'table_uri {} does not exist'.format(table_uri)}, HTTPStatus.NOT_FOUND
except Exception as e:
LOGGER.error("Error", e)
raise e