From d8e8b997a46bf278154cea08444f9b8add3386c5 Mon Sep 17 00:00:00 2001 From: Khalim Conn-Kowlessar Date: Tue, 29 Oct 2024 17:31:23 +0000 Subject: [PATCH] extend to get dimensions from 2nd floor --- etl/customers/stonewater/Wave 3 Preparation.py | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/etl/customers/stonewater/Wave 3 Preparation.py b/etl/customers/stonewater/Wave 3 Preparation.py index 84d67f56..ad35e2d5 100644 --- a/etl/customers/stonewater/Wave 3 Preparation.py +++ b/etl/customers/stonewater/Wave 3 Preparation.py @@ -149,7 +149,8 @@ def extract_summary_report(pdf_path): if data["Existing Secondary Heating System"] == "": data["Secondary Heating Code"] = "" else: - data["Secondary Heating Code"] = secondary_heating_code_match.group(1).strip() + data["Secondary Heating Code"] = secondary_heating_code_match.group( + 1).strip() if secondary_heating_code_match else "" data["Water Heating Code"] = water_heating_code_match.group(1).strip() @@ -236,7 +237,7 @@ def extract_building_parts_epr(text): # Pattern to match each floor's measurements floor_pattern = re.compile( - r"(Lowest floor|First floor)\s+([\d.]+)\s+([\d.]+)\s+([\d.]+)\s+([\d.]+)" + r"(Lowest floor|First floor|Second floor)\s+([\d.]+)\s+([\d.]+)\s+([\d.]+)\s+([\d.]+)" ) # Extract floor details for each building part @@ -305,7 +306,7 @@ def extract_building_parts_summary(text): # Pattern to extract floor details: Floor Level, Floor Area, Room Height, Perimeter, Party Wall Length floor_pattern = re.compile( - r"(1st Floor|Lowest Floor):\s*([\d.]+)\s+([\d.]+)\s+([\d.]+)\s+([\d.]+)" + r"(1st Floor|Lowest Floor|Second floor):\s*([\d.]+)\s+([\d.]+)\s+([\d.]+)\s+([\d.]+)" ) # Extract data for each floor within the building part @@ -634,6 +635,7 @@ def main(): extracted_data.append(summary_data) extracted_data = pd.DataFrame(extracted_data) + # Save this as a csv # extracted_data.to_csv("Wave 3 Summary Data - first 200 files.csv", index=False)