def load_registry() -> List[StationSchema]:
"""
Loads the facility registry into a list of Station schema's
"""
stations = load_data("facility_registry.json")
records = []
for station_id, station_record in stations.items():
facilities = []
for duid, facility_record in station_record["duid_data"].items():
status = map_compat_facility_state(
station_record.get("status", {}).get("state", "")
)
fuel_tech = map_compat_fueltech(
facility_record.get("fuel_tech", "")
)
registered_capacity = clean_capacity(
facility_record.get("registered_capacity")
)
facility = FacilitySchema(
name=normalize_whitespace(station_record.get("display_name")),
network_region=map_compat_network_region(
station_record["region_id"]
),
status=status,
duid=duid,
fueltech=fuel_tech,
capacity=registered_capacity,
)
facilities.append(facility)
record = StationSchema(
name=normalize_whitespace(station_record.get("display_name")),
code=station_id,
state=station_record.get("location", {}).get("state", None),
facilities=_sort_facilities(facilities),
)
records.append(record)
records = _sort_stations(records)
return records
def optimized_data_parser(date_str: str) -> Optional[datetime]:
"""
Turns out that dateutil's date parser is slow since
it does a lot of string parsing. Here we try matching
the date using known string formats.
"""
dt_return = None
date_str = normalize_whitespace(date_str).strip()
for date_format_str in DATE_FORMATS:
try:
dt_return = datetime.strptime(date_str, date_format_str)
except (KeyError, OverflowError, ValueError, AttributeError):
continue
if dt_return:
return dt_return
return dt_return
def name_clean(cls, v):
name = normalize_whitespace(v)
return name