From a9ce5b68bb6b506b62179c7abac5f43da2498ad1 Mon Sep 17 00:00:00 2001 From: Khalim Conn-Kowlessar Date: Tue, 29 Oct 2024 15:11:16 +0000 Subject: [PATCH] debug extract of main heating code --- etl/customers/stonewater/Wave 3 Preparation.py | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/etl/customers/stonewater/Wave 3 Preparation.py b/etl/customers/stonewater/Wave 3 Preparation.py index 0af43310..bb100ae1 100644 --- a/etl/customers/stonewater/Wave 3 Preparation.py +++ b/etl/customers/stonewater/Wave 3 Preparation.py @@ -123,8 +123,8 @@ def extract_summary_report(pdf_path): secondary_heating_section = re.search(r"Main\s*Heating2\s*(.*?)\s*Water\s*Heating", text, re.DOTALL) secondary_text = secondary_heating_section.group(1) - data["Existing Secondary Heating System"] = re.search(r"Main Heating Code\s*(.*?)\n", secondary_text).group( - 1).strip() + main_heating_code_match = re.search(r"Main Heating Code\s*(.*?)(?=\n|Percentage of Heat)", secondary_text) + data["Existing Secondary Heating System"] = main_heating_code_match.group(1).strip() data["Existing Secondary Heating PCDF Reference"] = re.search(r"PCDF boiler Reference\s*(\d+)", secondary_text).group(1) second_heating_controls_match = re.search(r"Main Heating Controls\s*(.*?)\n", secondary_text) @@ -139,7 +139,11 @@ def extract_summary_report(pdf_path): secondary_heating_code_match = re.search(r"Secondary Heating Code\s*(.*?)\n", text) water_heating_code_match = re.search(r"Water Heating Code\s*(.*?)\n", text) - data["Secondary Heating Code"] = secondary_heating_code_match.group(1).strip() + if data["Existing Secondary Heating System"] == "": + data["Secondary Heating Code"] = "" + else: + data["Secondary Heating Code"] = secondary_heating_code_match.group(1).strip() + data["Water Heating Code"] = water_heating_code_match.group(1).strip() return data