from dataclasses import dataclass, asdict
from .util import time_conversion
import logging

logger = logging.getLogger(__file__)


@dataclass
class DataPoint:
    """Represents a single Datapoint from the InfluxDB perspective."""
    measurement: str
    time: int
    fields: dict
    tags: dict

    def asdict(self):
        return asdict(self)


def data_point_factory(run, *,
                       time_key,
                       measurement_name,
                       field_keys: set(),
                       tag_keys: set() = None,
                       no_tag_keys: set() = None,
                       ) -> DataPoint:
    """
    Returns a data point from a dict.
        Parameters:
            run(dict): dict from that the data are extracted
            time_key(str): key to extract timestamp
            tag_keys(set): explicitly define tags
            no_tag_key(set): define what are not tags, the tag_keys are inferred automatically
    """
    if tag_keys is None and no_tag_keys is None:
        raise ValueError("You need either specify tag_keys or no_tag_keys")

    fields = {key: run[key] for key in field_keys}

    if tag_keys is None:
        tag_keys = run.keys() - no_tag_keys - field_keys - {time_key}

    tags = {key: run[key]for key in tag_keys}
    time = run[time_key]
    try:
        time = time_conversion(time)
    except ValueError:
        logger.warn(f"{time} could not be transformed with the current pattern")
        pass
    return DataPoint(measurement_name, time, fields=fields, tags=tags)