incorporate base utility into all classes

model_data/epc_attributes/FloorAttributes.py

@@ -1,20 +1,27 @@
-from typing import Dict, Union, Optional
+from typing import Dict, Union
+from model_data.BaseUtility import BaseUtility
 from model_data.epc_attributes.attribute_utils import extract_thermal_transmittance, extract_component_types
 
 
-class FloorAttributes:
+class FloorAttributes(BaseUtility):
     DWELLING_BELOW = ["another dwelling below", "other premises below"]
     FLOOR_TYPES = ["assumed", "to unheated space", "to external air", "suspended", "solid"]
 
     def __init__(self, description: str):
         self.description: str = description.lower()
 
+        self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
+
         if not description or not any(
             rt in self.description for rt in self.FLOOR_TYPES + self.DWELLING_BELOW + ["average thermal transmittance"]
         ):
             raise ValueError('Invalid description')
 
     def process(self) -> Dict[str, Union[str, bool, int, None]]:
+
+        if self.nodata:
+            return {}
+
         result: Dict[str, Union[float, str, bool, None]] = {}
         description = self.description
 

model_data/epc_attributes/HotWaterAttributes.py

@@ -1,8 +1,9 @@
 from typing import Dict, Union
-from model_data.epc_attributes.attribute_utils import clean_description, remove_punctuation, find_keyword
+from model_data.BaseUtility import BaseUtility
+from model_data.epc_attributes.attribute_utils import clean_description, find_keyword
 
 
-class HotWaterAttributes:
+class HotWaterAttributes(BaseUtility):
     # HEATER_TYPES refer to the main devices used for heating water. These devices can be powered by different energy
     # sources.
     HEATER_TYPES = [
@@ -78,6 +79,8 @@ class HotWaterAttributes:
     def __init__(self, description: str):
         self.description: str = clean_description(description.lower())
 
+        self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
+
         if not any(
             self._keyword_in_description(keywords)
             for keywords in [
@@ -98,6 +101,22 @@ class HotWaterAttributes:
         return any(keyword in self.description for keyword in keywords)
 
     def process(self) -> Dict[str, Union[str, bool]]:
+
+        if self.nodata:
+            return {
+                "heater_type": None,
+                "system_type": None,
+                "thermostat_characteristics": None,
+                "heating_scope": None,
+                "energy_recovery": None,
+                "tariff_type": None,
+                "extra_features": None,
+                "chp_systems": None,
+                "distribution_system": None,
+                "no_system_present": None,
+                "assumed": None,
+            }
+
         result: Dict[str, Union[str, bool]] = {
             "heater_type": find_keyword(self.description, self.HEATER_TYPES),
             "system_type": find_keyword(self.description, self.SYSTEM_TYPES),

model_data/epc_attributes/MainheatAttributes.py

@@ -1,8 +1,9 @@
+from model_data.BaseUtility import BaseUtility
 from model_data.epc_attributes.attribute_utils import clean_description, process_part
 from typing import Dict, Union
 
 
-class MainHeatAttributes:
+class MainHeatAttributes(BaseUtility):
     HEAT_SYSTEMS = ["boiler", "air source heat pump", "room heaters", "electric storage heaters", "warm air",
                     "electric underfloor heating", "electric ceiling heating", "community scheme",
                     "ground source heat pump", "no system present", "portable electric heaters",
@@ -16,6 +17,7 @@ class MainHeatAttributes:
     def __init__(self, description: str):
         self.description: str = clean_description(description.lower())
         # Remove special characters
+        self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
 
         if not description or not any(
             rt in self.description for rt in
@@ -31,6 +33,9 @@ class MainHeatAttributes:
         result.update({f'has_{ot.replace(" ", "_")}': False for ot in self.OTHERS})
         result['has_underfloor_heating'] = False
 
+        if self.nodata:
+            return result
+
         description = self.description.split(',')
 
         # Process each part separately

model_data/epc_attributes/MainheatControlAttributes.py

@@ -1,8 +1,9 @@
 from typing import Dict, Union
-from model_data.epc_attributes.attribute_utils import clean_description, remove_punctuation, find_keyword
+from model_data.BaseUtility import BaseUtility
+from model_data.epc_attributes.attribute_utils import clean_description, find_keyword
 
 
-class MainheatControlAttributes:
+class MainheatControlAttributes(BaseUtility):
     # These systems allow for the automatic regulation of temperature
     THERMOSTATIC_CONTROL_KEYWORDS = [
         'room thermostats',
@@ -67,7 +68,7 @@ class MainheatControlAttributes:
 
     def __init__(self, description: str):
         self.description: str = clean_description(description.lower())
-        self.nodata = not description
+        self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
 
         if not self.nodata:
             if not any(

model_data/epc_attributes/RoofAttributes.py

@@ -1,9 +1,10 @@
 import re
 from typing import Dict, Union
+from model_data.BaseUtility import BaseUtility
 from model_data.epc_attributes.attribute_utils import extract_component_types, extract_thermal_transmittance
 
 
-class RoofAttributes:
+class RoofAttributes(BaseUtility):
     ROOF_TYPES = ['pitched', 'roof room', 'loft', 'flat', 'thatched', 'at rafters', 'assumed']
     DWELLING_ABOVE = ["another dwelling above", "other premises above"]
 
@@ -13,6 +14,7 @@ class RoofAttributes:
         """
 
         self.description: str = description.lower()
+        self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
 
         if not description or not any(
             rt in self.description for rt in self.ROOF_TYPES + self.DWELLING_ABOVE + ["average thermal transmittance"]
@@ -22,6 +24,10 @@ class RoofAttributes:
     def process(self) -> Dict[str, Union[float, str, bool, None]]:
 
         result: Dict[str, Union[float, str, bool, None]] = {}
+
+        if self.nodata:
+            return result
+
         description = self.description
 
         # thermal transmittance

model_data/epc_attributes/WallAttributes.py

@@ -1,8 +1,9 @@
 from typing import Dict, Union
+from model_data.BaseUtility import BaseUtility
 from model_data.epc_attributes.attribute_utils import extract_component_types, extract_thermal_transmittance
 
 
-class WallAttributes:
+class WallAttributes(BaseUtility):
     WALL_TYPES = ['cavity wall', 'filled cavity', 'solid brick', 'system built', 'timber frame', 'granite or whinstone',
                   'as built', 'cob', 'assumed', 'sandstone or limestone']
 
@@ -12,8 +13,13 @@ class WallAttributes:
         """
         self.description: str = description
 
+        self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
+
     def process(self) -> Dict[str, Union[float, str, bool, None]]:
         result: Dict[str, Union[float, str, bool, None]] = {}
+        if self.nodata:
+            return result
+
         description = self.description.lower()
 
         # thermal transmittance - it can be negative which is errneous however we'll still pull it out

model_data/epc_attributes/WindowAttributes.py

@@ -1,8 +1,9 @@
 from typing import Dict, Union
+from model_data.BaseUtility import BaseUtility
 from model_data.epc_attributes.attribute_utils import clean_description
 
 
-class WindowAttributes:
+class WindowAttributes(BaseUtility):
     GLAZING_KEYWORDS = ["glazing", "glazed", "glaze"]
     GLAZING_COVERAGE = ["fully", "mostly", "partial", "some", "full", "thoughout"]
     GLAZING_TYPES = ["double", "triple", "secondary", "multiple", "high performance", "single"]
@@ -21,7 +22,7 @@ class WindowAttributes:
 
         # In the case of an empty description, we want to return a dictionary with all values set to False
         # and indicate there was no data
-        self.nodata = not description
+        self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
 
         if not self.nodata:
             if not any(