From 2ee9ba9dddb114a5565e4f1e891695c4e29e674d Mon Sep 17 00:00:00 2001 From: Khalim Conn-Kowlessar Date: Mon, 4 Sep 2023 10:36:39 +0100 Subject: [PATCH] Implemented area data extraction for first 6 files --- model_data/simulation_system/area_data.py | 179 ++++++++++++++++++++-- 1 file changed, 162 insertions(+), 17 deletions(-) diff --git a/model_data/simulation_system/area_data.py b/model_data/simulation_system/area_data.py index ee74012b..e381b0e5 100644 --- a/model_data/simulation_system/area_data.py +++ b/model_data/simulation_system/area_data.py @@ -4,7 +4,9 @@ of insulation measures within homes """ import boto3 import PyPDF2 -import tempfile +import re +import json +from io import BytesIO bucket = "retrofit-datalake-dev" @@ -43,29 +45,132 @@ def list_files_in_s3_folder(bucket_name, folder_name): return files -def fetch_pdf_from_s3(bucket_name, pdf_key, local_path): +def fetch_and_parse_pdf_from_s3(bucket_name, filename): """ - Fetch a PDF from an S3 bucket and save it locally. + Fetch a PDF from an S3 bucket and parse its content. Parameters: - bucket_name: Name of the S3 bucket. - pdf_key: Path (key) of the PDF file within the bucket. - - local_path: Local path where the PDF should be saved. + + Returns: + - text: Extracted text from the PDF. """ s3_client = boto3.client('s3') - response = s3_client.get_object(Bucket=bucket_name, Key=pdf_key) + response = s3_client.get_object(Bucket=bucket_name, Key=filename) - # Read the PDF bytes and save locally - with open(local_path, 'wb') as f: - f.write(response['Body'].read()) + # Create a BytesIO object from the PDF bytes + pdf_content = BytesIO(response['Body'].read()) + + # Use PyPDF2 to read the PDF content + reader = PyPDF2.PdfReader(pdf_content) + + # Extract text from each page + pages = [] + for page_num in range(len(reader.pages)): + page = reader.pages[page_num] + + text = page.extract_text() + text = remove_excess_newlines(text) + pages.append(text.split("\n")) + + return pages -# Usage -bucket_name = 'YOUR_BUCKET_NAME' -pdf_key = 'path/to/your/pdf_file.pdf' -local_path = 'local_file_name.pdf' -fetch_pdf_from_s3(bucket_name, pdf_key, local_path) +def fetch_json_from_s3(bucket_name, file_name): + # Create an S3 client + s3 = boto3.client('s3') + + # Fetch the file from S3 + response = s3.get_object(Bucket=bucket_name, Key=file_name) + + # Parse and return the JSON data + return json.loads(response['Body'].read().decode('utf-8')) + + +def write_json_to_s3(bucket_name, file_name, json_data): + """ + Write JSON data to a file in an S3 bucket. + + Parameters: + - bucket_name: Name of the S3 bucket. + - file_name: Path (key) of the file within the bucket. + - json_data: JSON data to be saved. + """ + + s3_client = boto3.client('s3') + + # Convert the JSON data to a string + json_string = json.dumps(json_data) + + # Upload the JSON string to S3 + s3_client.put_object(Bucket=bucket_name, Key=file_name, Body=json_string) + + +def check_s3_file_exists(bucket_name, file_name): + """ + Check if a file exists in an S3 bucket. + + Parameters: + - bucket_name: Name of the S3 bucket. + - file_name: Path (key) of the file within the bucket. + + Returns: + - bool: True if the file exists, False otherwise. + """ + + s3_client = boto3.client('s3') + + try: + # Check if the object exists by attempting to retrieve its metadata + s3_client.head_object(Bucket=bucket_name, Key=file_name) + return True + except s3_client.exceptions.ClientError as e: + # If the error code is 404 (Not Found), then the file doesn't exist + if e.response['Error']['Code'] == '404': + return False + # If there's any other exception, raise it + raise + + +def remove_excess_newlines(text): + return re.sub('\n+', '\n', text).strip() + + +def search_pages(pages, search_term) -> ( + str | None, int | None, int | None +): + """ + This method looks for a search term in the EPR and returns the first instance of it + :param pages: list of pages to search through + :param search_term: The term to search for + :return: The text, page number and page index of the first instance of the search term + """ + + to_page = len(pages) + from_page = 0 + from_index = 0 + + for page_num in range(from_page, to_page + 1): + + page_to_index = len(pages[page_num]) + + for page_index in range(from_index, page_to_index): + if search_term in pages[page_num][page_index]: + return pages[page_num][page_index], page_num, page_index + + return None, None, None + + +def check_page(pages, page_num, page_index): + if page_num > len(pages): + return False + + if page_index > len(pages[page_num]): + return False + + return True def handler(): @@ -75,10 +180,50 @@ def handler(): sap_calulation_pdfs = [file for file in files if file.endswith(".pdf")] # For each pdf, we pull out the net & gross wall areas + if check_s3_file_exists(bucket_name=bucket, file_name="wall-area-data/wall-area.json"): + data = fetch_json_from_s3(bucket_name=bucket, file_name="wall-area-data/wall-area.json") + data = json.loads(data) + else: + data = [] + + used_files = [x["filename"] for x in data] + + sap_calulation_pdfs = [filename for filename in sap_calulation_pdfs if filename.split("/")[-1] not in used_files] - data = [] for sap_calculation_file in sap_calulation_pdfs: - # Create a temp file to store the PDF - temp_filename = tempfile.NamedTemporaryFile(suffix=".pdf").name - pdf_file = fetch_pdf_from_s3(bucket, sap_calculation_file, temp_filename) + # Download pdf + pdf_pages = fetch_and_parse_pdf_from_s3(bucket, sap_calculation_file) + + # We search for net and gross wall areas + result = search_pages(pdf_pages, "External walls Main")[0] + # This is a row in a table where the columns are: + # Element, Gross, Openings, NetArea, U-value, A x U, K-value, A x K + # The values we're interested in are Gross and NetArea + values = result.split("External walls Main")[1].strip().split(" ") + # Remove the empty white space - we should now have the fields we want + values = [v for v in values if v] + gross_area = float(values[0]) + net_area = float(values[2]) + + # Search for property identifiers + _, pagenum, page_idx = search_pages(pdf_pages, 'Prop Type Ref') + if pagenum != 0: + raise ValueError("Property reference not found on the first page") + # the reference will be on the next line + property_reference = pdf_pages[pagenum][page_idx + 1] + property_reference_number = pdf_pages[pagenum][page_idx + 2] + address = pdf_pages[pagenum][page_idx + 4] + + data.append( + { + "property_reference": property_reference, + "reference_number": property_reference_number, + "address": address, + "gross_area": gross_area, + "net_area": net_area, + "filename": sap_calculation_file + } + ) + + write_json_to_s3(bucket_name=bucket, file_name="wall-area-data/wall-area.json", json_data=json.dumps(data))