personnal-accounting/scripts/process_bourso.py

#!/usr/bin/env python3

import re
import csv
import os
from collections import defaultdict

def categorize_bourso_transaction(description):
    description = description.lower()
    
    if 'ech pret' in description:
        return 'Loan Repayment'
    if 'american express' in description:
        return 'Credit Card Payment (Amex)'
    if 'orange sa' in description or 'sfr' in description or 'ste reunionnaise du radiotelep' in description:
        return 'Utilities'
    if 'be rock' in description:
        return 'Subscription (BE ROCK)'
    if 'paypal' in description:
        return 'Online Purchases (Paypal)'
    if 'vir virement interne' in description:
        return 'Internal Transfer'
    if 'retrait dab' in description:
        return 'Cash Withdrawal'
    if description.startswith('carte'):
        return 'Card Payment'
    return 'Other'

def process_bourso_statement(file_path, output_csv=False, output_dir='../../output/csv'):
    with open(file_path, 'r', encoding='utf-8') as f:
        content = f.read()

    expense_summary = defaultdict(float)
    total_expenses = 0
    transactions_data = []  # Store all transaction data for CSV output
    
    # A more robust regex to handle slight variations
    transaction_regex = re.compile(r"^ (\d{2}/\d{2}/\d{4})\s+(.*?)\s+(\d{2}/\d{2}/\d{4})\s+([\d,.]+\s*)?([\d,.]+\s*)?$", re.MULTILINE)
    
    transactions = transaction_regex.findall(content)
    
    print("--- Matched Transactions ---")
    for op_date, description, val_date, debit_str, credit_str in transactions:
        description = description.strip()
        debit = 0
        if debit_str:
            try:
                debit = float(debit_str.strip().replace(',', '.'))
            except ValueError:
                continue # Skip if debit is not a valid number

        category = categorize_bourso_transaction(description)
        print(f"Found: {description} -> {category} -> {debit}") # DEBUG
        
        # Store transaction data for potential CSV output
        transactions_data.append({
            'Date': op_date,
            'Description': description,
            'Category': category,
            'Debit': debit,
            'Credit': 0,
            'Value Date': val_date
        })
        
        if debit > 0 and category != 'Internal Transfer':
            expense_summary[category] += debit
            total_expenses += debit
    
    # Output CSV if requested
    if output_csv:
        csv_file = os.path.join(output_dir, 'boursobank_all_transactions.csv')
        os.makedirs(output_dir, exist_ok=True)
        with open(csv_file, 'w', newline='', encoding='utf-8') as csvfile:
            fieldnames = ['Date', 'Description', 'Category', 'Debit', 'Credit', 'Value Date']
            writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
            writer.writeheader()
            writer.writerows(transactions_data)
        print(f"\nTransaction data saved to {csv_file}")

    print("\n--- Boursobank Expense Summary (Dec 2025) - Final ---")
    print(f"Total Expenses Analyzed: €{total_expenses:,.2f}")
    print("\n--- Spending by Category ---")
    
    sorted_expenses = sorted(expense_summary.items(), key=lambda item: item[1], reverse=True)
    
    if total_expenses > 0:
        for category, total in sorted_expenses:
            percentage = (total / total_expenses) * 100
            print(f"{category:<25} €{total:9,.2f} ({percentage:5.2f}%)")
    else:
        print("No expenses found.")
    
    return transactions_data

def process_bourso_pdf_files(directory, output_csv=False, output_dir='../../output/csv'):
    import subprocess
    import glob
    
    # Get all PDF files in the directory
    pdf_files = glob.glob(os.path.join(directory, "*.pdf"))
    all_transactions = []
    
    for pdf_file in pdf_files:
        try:
            # Convert PDF to text
            result = subprocess.run(['pdftotext', '-layout', pdf_file, '-'], 
                                  capture_output=True, text=True, check=True)
            content = result.stdout
            
            # Save text to temporary file
            temp_file = os.path.splitext(pdf_file)[0] + '.txt'
            with open(temp_file, 'w', encoding='utf-8') as f:
                f.write(content)
            
            # Process the text file
            transactions = process_bourso_statement(temp_file, output_csv, output_dir)
            all_transactions.extend(transactions)
            
            # Clean up temporary file
            os.remove(temp_file)
            
        except (subprocess.CalledProcessError, FileNotFoundError) as e:
            print(f"Error processing {pdf_file}: {e}")
            continue
    
    # Output consolidated CSV if requested
    if output_csv and all_transactions:
        csv_file = os.path.join(output_dir, 'boursobank_all_transactions.csv')
        os.makedirs(output_dir, exist_ok=True)
        with open(csv_file, 'w', newline='', encoding='utf-8') as csvfile:
            fieldnames = ['Date', 'Description', 'Category', 'Debit', 'Credit', 'Value Date']
            writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
            writer.writeheader()
            writer.writerows(all_transactions)
        print(f"\nAll transaction data saved to {csv_file}")
    
    return all_transactions

if __name__ == "__main__":
    import argparse
    parser = argparse.ArgumentParser(description='Process Boursobank statements')
    parser.add_argument('--pdf-dir', default='../data/pdf/boursobank', 
                       help='Directory containing Boursobank PDF files')
    parser.add_argument('--output-dir', default='../../output/csv', 
                       help='Directory to save CSV output files')
    parser.add_argument('--csv', action='store_true', 
                       help='Output transaction data to CSV files')
    args = parser.parse_args()
    
    # Process all PDF files in the directory
    process_bourso_pdf_files(args.pdf_dir, args.csv, args.output_dir)