From b3053f8518727db1b00c2165b286988dbf5796e9 Mon Sep 17 00:00:00 2001
From: Khalim Conn-Kowlessar <kconnkowlessar@gmail.com>
Date: Wed, 25 Sep 2024 16:10:34 +0100
Subject: [PATCH] handing edge cases for epc cleaning

---
 etl/epc_clean/epc_attributes/HotWaterAttributes.py | 9 ++++++++-
 etl/epc_clean/epc_attributes/MainheatAttributes.py | 1 +
 etl/epc_clean/epc_attributes/RoofAttributes.py     | 3 +++
 etl/epc_clean/epc_attributes/WindowAttributes.py   | 1 +
 4 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/etl/epc_clean/epc_attributes/HotWaterAttributes.py b/etl/epc_clean/epc_attributes/HotWaterAttributes.py
index 78ee5f7d..b0105b10 100644
--- a/etl/epc_clean/epc_attributes/HotWaterAttributes.py
+++ b/etl/epc_clean/epc_attributes/HotWaterAttributes.py
@@ -127,13 +127,20 @@ class HotWaterAttributes(Definitions):
                                                                                    "thermostat, flue gas heat recovery",
         "ogçör brif system, gydag ynnigçör haul, adfer gwres nwyon ffliw": "from main system, plus solar, flue gas "
                                                                            "heat recovery",
+        "o r brif system, gydag ynni r haul, dim thermostat ar y silindr": "from main system, plus solar, no cylinder "
+                                                                           "thermostat",
     }
 
+    NODATA_DESCRIPTIONS = [
+        "sap05 hot-water",
+        "sap hot-water"
+    ]
+
     def __init__(self, description: str):
         self.description: str = clean_description(description.lower()).strip()
 
         self.nodata = not self.description or description in self.DATA_ANOMALY_MATCHES or (
-            self.description == "sap05 hot-water"
+            self.description in self.NODATA_DESCRIPTIONS
         )
 
         translation = self.WELSH_TEXT.get(self.description)
diff --git a/etl/epc_clean/epc_attributes/MainheatAttributes.py b/etl/epc_clean/epc_attributes/MainheatAttributes.py
index 430b418d..46cbf52b 100644
--- a/etl/epc_clean/epc_attributes/MainheatAttributes.py
+++ b/etl/epc_clean/epc_attributes/MainheatAttributes.py
@@ -61,6 +61,7 @@ class MainHeatAttributes(Definitions):
         "bwyler a rheiddiaduron, olew, st+¦r wresogyddion trydan": "boiler and radiators, oil, electric storage "
                                                                    "heaters",
         "pwmp gwres sygçön tarddu yn yr awyr, awyr gynnes, trydan": "air source heat pump, warm air, electric",
+        "stor wresogyddion trydan": "electric storage heaters",
         # Should be handled by edge cases
         ", trydan": ", electric",
     }
diff --git a/etl/epc_clean/epc_attributes/RoofAttributes.py b/etl/epc_clean/epc_attributes/RoofAttributes.py
index f36d445f..75cb8af1 100644
--- a/etl/epc_clean/epc_attributes/RoofAttributes.py
+++ b/etl/epc_clean/epc_attributes/RoofAttributes.py
@@ -19,6 +19,8 @@ class RoofAttributes(Definitions):
         "ar oleddf, inswleiddio cyfyngedig": "pitched, limited insulation",
         "ar oleddf, wedigçöi inswleiddio wrth y trawstiau": 'pitched, insulated at rafters',
         "ar oleddf, wedi?i inswleiddio wrth y trawstiau": 'pitched, insulated at rafters',
+        "ar oleddf, wedi?i inswleiddio wrth y trawstia": 'pitched, insulated at rafters',
+        "ar oleddf, wedigçöi inswleiddio wrth y trawstia": 'pitched, insulated at rafters',
         "yn wastad, inswleiddio cyfyngedig (rhagdybiaeth)": "flat, limited insulation (assumed)",
         "yn wastad, inswleiddio cyfyngedig": "flat, limited insulation",
         "yn wastad, dim inswleiddio (rhagdybiaeth)": "flat, no insulation (assumed)",
@@ -35,6 +37,7 @@ class RoofAttributes(Definitions):
         "ystafell(oedd) to, nenfwd wedigçöi inswleiddio": "roof room(s), ceiling insulated",
         "ystafell(oedd) to, dim inswleiddio (rhagdybiaeth)": "roof room(s), no insulation (assumed)",
         "ystafell(oedd) to, dim inswleiddio": "roof room(s), no insulation",
+        "to gwellt, gydag inswleiddio ychwanegol": "thatched, with additional insulation",
     }
 
     DEFAULT_KEYS = [
diff --git a/etl/epc_clean/epc_attributes/WindowAttributes.py b/etl/epc_clean/epc_attributes/WindowAttributes.py
index a52977e6..8c4d0c45 100644
--- a/etl/epc_clean/epc_attributes/WindowAttributes.py
+++ b/etl/epc_clean/epc_attributes/WindowAttributes.py
@@ -37,6 +37,7 @@ class WindowAttributes(Definitions):
     NO_DATA_CASES = [
         "SAP05:Windows",
         "Solid, no insulation (assumed)",  # A description typically associated with floors, not windows
+        "Suspended, no insulation (assumed)",  # A description typically associated with floors, not windows
     ]
 
     def __init__(self, description: str):