def open_child(self, name: str):
"""Open the file <name> that is in the directory path for reading.
If files_as_path is True, return the path of the child."""
if not self.__path.is_dir():
raise InvalidDataError(f'"{self.__path}" is not a directory')
path = self.__path / name
if not path.exists():
raise InvalidDataError(f"'{path}' does not exist")
if self.__files_as_path:
return path
return path.open(mode=self.__mode)
def get_data(self, selector: SeriesSelector, start_date: datetime,
end_date: datetime) -> pa.Table:
"""Return data for the given time series in the given time period."""
if selector.name is None:
raise InvalidDataError("No series name")
measurement, tags, field_key = _parse_influx_series(selector.name)
query = f"""SELECT time, "{_escape(field_key)}"
FROM "{_escape(measurement)}"
WHERE time >= $start_date and time <= $end_date"""
bind_params = {
"start_date": start_date.strftime("%Y-%m-%dT%H:%M:%SZ"),
"end_date": end_date.strftime("%Y-%m-%dT%H:%M:%SZ"),
}
for i, (tag_key, tag_value) in enumerate(tags):
bind_params[str(i)] = tag_value
query = query + f' and "{_escape(tag_key)}" = ${str(i)}'
timestamps = []
values = []
for item in self.__client.query(query=query,
bind_params=bind_params).get_points():
timestamps.append(dateutil.parser.parse(item["time"]))
values.append(item[field_key])
return pa.Table.from_pydict({"ts": timestamps, "value": values})
def _read_pivot_data(loader: Loader, selector: SeriesSelector) -> pa.Table:
all_data = pyarrow.csv.read_csv(loader.open())
if selector.name not in all_data.column_names:
raise InvalidDataError(f'column "{selector.name}" not found')
schema = pa.schema([("ts", pa.timestamp("us", "utc")),
("value", pa.float64())])
return (all_data.select([0, selector.name
]).rename_columns(["ts", "value"]).cast(schema))
def get_metadata(self, selector: SeriesSelector) -> Metadata:
"""Get metadata from the Flight service."""
if selector.name is None:
raise InvalidDataError("No series name")
remote_selector = SeriesSelector(self.__source_name, selector.name)
metadata = self.__client.get_metadata(remote_selector)
metadata.series = selector
return metadata
def _read_pivot_data(self, selector: SeriesSelector) -> pa.Table:
all_data = self.read_file(self.__loader.open())
if selector.name not in all_data.column_names:
raise InvalidDataError(f'column "{selector.name}" not found')
data = all_data.select([0,
selector.name]).rename_columns(["ts", "value"])
schema = pa.schema([("ts", pa.timestamp("us", "utc")),
("value", _get_value_schema_type(data))])
return data.cast(schema)
def open(self):
"""Open the file at path for reading.
If files_as_path is True, return the path."""
if not self.__path.exists():
raise InvalidDataError(f"'{self.__path}' does not exist")
if self.__files_as_path:
return self.__path
return self.__path.open(mode=self.__mode)
def has_child(self, name: str) -> bool:
"""Test if the file <name> is in the directory pointed to by <path>."""
if not self.__path.is_dir():
raise InvalidDataError(f'"{self.__path}" is not a directory')
path = self.__path / name
return path.exists()