Debugging roof and lighting welsh translations

model_data/epc_attributes/LightingAttributes.py

@@ -5,7 +5,8 @@ from model_data.utils import correct_spelling
 
 class LightingAttributes:
     WELSH_TEXT = {
-        "goleuadau ynni-isel ym mhob un ogçör mannau gosod": "low energy lighting in all fixed outlets"
+        "goleuadau ynni-isel ym mhob un ogçör mannau gosod": "low energy lighting in all fixed outlets",
+        "dim goleuadau ynni-isel": "no low energy lighting"
     }
 
     def __init__(self, description, averages):

model_data/epc_attributes/RoofAttributes.py

@@ -22,16 +22,34 @@ class RoofAttributes(Definitions):
         self.description: str = description.lower()
         self.nodata = not description or description in self.DATA_ANOMALY_MATCHES
 
-        translation = self.WELSH_TEXT.get(self.description)
+        self.welsh_translation_search()
-        if translation:
-            self.nodata = False
-            self.description = translation
 
         if not self.nodata and not any(
             rt in self.description for rt in self.ROOF_TYPES + self.DWELLING_ABOVE + ["average thermal transmittance"]
         ):
             raise ValueError('Invalid description')
 
+    def welsh_translation_search(self):
+        """
+        For some descriptions,
+        we want to translate, however they have a consistent structure, where the only change
+        is the thickness of insulation. Instead of manually adding a record for each translation, we
+        search for regular expressions and translate
+        """
+
+        insulation_thickness_match = re.search(r"(\d+ mm) o inswleiddio yn y llofft", self.description)
+
+        # Step 2: Generalized translation with placeholder
+        if insulation_thickness_match:
+            insulation_thickness = insulation_thickness_match.group(1)
+            self.description = self.description.replace(insulation_thickness_match.group(0), "")
+            self.description = f"pitched, {insulation_thickness} loft insulation"
+        else:
+            translation = self.WELSH_TEXT.get(self.description)
+            if translation:
+                self.nodata = False
+                self.description = translation
+
     def process(self) -> Dict[str, Union[float, str, bool, None]]:
 
         result: Dict[str, Union[float, str, bool, None]] = {}

model_data/tests/test_data/test_lighting_attributes_cases.py

@@ -32,4 +32,5 @@ test_cases = [
     {'original_description': 'No Low energy lighting', 'low_energy_proportion': 0},
     {'original_description': 'Goleuadau ynni-isel mewn 60% oGÇÖr mannau gosod', 'low_energy_proportion': 0.6},
     {'original_description': 'Goleuadau ynni-isel ym mhob un oGÇÖr mannau gosod', 'low_energy_proportion': 1},
+    {'original_description': 'Dim goleuadau ynni-isel', 'low_energy_proportion': 0},
 ]