Model/epc_data/attributes/MainFuelAttributes.py

from typing import Dict, Union
from epc_data.attributes.attribute_utils import clean_description, remove_punctuation


class MainFuelAttributes:
    FUEL_KEYWORDS = [
        'heat network',
        'mains gas',
        'electricity',
        'oil',
        'biomass',
        'biodiesel',
        # Note: there is als a category called 'bottled LPG', but only 2/50k entries had this
        'LPG',
        'waste combustion',
        'biogas',
        'wood logs',
        'dual fuel - mineral + wood',
        'gas',
        'anthracite',
        'smokeless coal',
        'house coal'
    ]

    NO_FUEL_KEYWORDS = [
        'To be used only when there is no heating/hot-water system or data is from a community network'
    ]

    def __init__(self, description: str):
        self.description: str = remove_punctuation(clean_description(description.lower()))

        self.nodata = not description or any(keyword in self.description for keyword in self.NO_FUEL_KEYWORDS)

        self.is_community = False if 'not community' in self.description else 'community' in self.description

        if not self.nodata and not any(
            self._keyword_in_description(keywords)
            for keywords in [
                self.FUEL_KEYWORDS
            ]
        ):
            raise ValueError('Invalid description')

    def _keyword_in_description(self, keywords):
        return any(keyword in self.description for keyword in keywords)

    def process(self) -> Dict[str, Union[str, bool]]:
        result: Dict[str, Union[str, bool]] = {
            "fuel_type": self._find_keyword(self.FUEL_KEYWORDS),
            "no_fuel": self._find_keyword(self.NO_FUEL_KEYWORDS),
            "is_community": self.is_community
        }

        return result

    def _find_keyword(self, keywords):
        for keyword in keywords:
            if keyword in self.description:
                return keyword

        return None