sparebanken-sor-ynab/src/convert.py

import glob
import logging
import pandas as pd
from pathlib import Path

from bank_parsers.sparebank1 import parse_sparebank1
from bank_parsers.bank_norwegian import parse_bank_norwegian
from bank_parsers.sparebanken_norge import parse_sparebanken_norge


# Dictionary of banks, filename patterns, and parsing functions
BANKS = {
    "SparebankenNorge": {
        "patterns": ["Transaksjoner*.csv"],
        "encoding": "latin1",
        "output_filename": "YNAB-{bank}-FROM-{first_date}-TO-{last_date}.csv",
        "parse_function": parse_sparebanken_norge,
        "delimiter": ";"
    },
    "Sparebank1": {
        "patterns": ["OversiktKonti*.csv"],
        "output_filename": "YNAB-{bank}-FROM-{first_date}-TO-{last_date}.csv",
        "parse_function": parse_sparebank1,
        "delimiter": ";"
    },
    "Norwegian": {
        "patterns": ["BankNorwegian*.xlsx", "Statement*.xlsx"], 
        "output_filename": "YNAB-{bank}-FROM-{first_date}-TO-{last_date}.csv",
        "parse_function": parse_bank_norwegian
    }
    # Add more banks and patterns as needed
}


def find_bank_config(filename):
    """
    Find the appropriate bank configuration for a given filename
    
    Args:
        filename (str): Name of the file to match
    
    Returns:
        tuple: (bank_name, bank_config) or (None, None) if no match
    """
    import fnmatch
    
    for bank_name, bank_config in BANKS.items():
        for pattern in bank_config["patterns"]:
            if fnmatch.fnmatch(filename, pattern):
                return bank_name, bank_config
    
    return None, None


def process_bank_statement(file_path, parse_function, delimiter, encoding):
    """
    Process a single bank statement file
    
    Args:
        file_path (str): Path to the bank statement file
        parse_function (callable): Function to parse the specific bank format
        delimiter (Optional<str>): Field delimiter
    
    Returns:
        pd.DataFrame: Processed YNAB-compatible data
    """
    file_extension = Path(file_path).suffix.lower()
    
    try:
        # Handle CSV files
        if file_extension == ".csv":
            data = pd.read_csv(file_path, delimiter=delimiter, encoding=encoding)
        # Handle Excel files
        elif file_extension in [".xlsx", ".xls"]:
            data = pd.read_excel(file_path)
        else:
            logging.warning(f"Skipping unsupported file type: {file_path}")
            return pd.DataFrame()
        
        # Call the appropriate bank-specific parsing function
        ynab_data = parse_function(data)
        return ynab_data
        
    except Exception as e:
        logging.error(f"Error processing file {file_path}: {e}")
        raise e
        return pd.DataFrame()


def get_unique_filename(original_path):
    file_retry_count = 0

    while True:
        result = Path(original_path)

        if file_retry_count > 0:
            result = result.with_stem(result.stem + f"({file_retry_count})")
        
        if result.exists():
            logging.debug(f"File {result} exists. Looking for available alternative")
            file_retry_count += 1
            continue
        else:
            return result


def convert_bank_statements_to_ynab(input_paths, output_directory, archive_directory, on_success, overwrite):
    """
    Convert bank statements to YNAB format
    
    Args:
        input_paths (list): List of specific files or directories to process
    """
    # Create output directory if it doesn't exist
    output_directory.mkdir(exist_ok=True, parents=True)
    
    # Get list of files to process
    files_to_process = []
    for path in input_paths:
        if not path.exists():
            logging.warning(f"Path does not exist: {file_path}")
        elif path.is_file():
            files_to_process.append(path)
        elif path.is_dir():
            logging.debug(f"Looking for matching files in {path}")
            for bank_config in BANKS.values():
                for pattern in bank_config["patterns"]:
                    matching_files = glob.glob(str(path / pattern))
                    files_to_process.extend([Path(f) for f in matching_files])

    files_processed = False
    
    # Process each file
    logging.info(f"Processing {len(files_to_process)} file(s)...")
    for file_path in files_to_process:
        logging.debug(f"Processing {file_path}")
        if not file_path.exists():
            logging.warning(f"File not found: {file_path}")
            continue
            
        # Find matching bank configuration
        bank_name, bank_config = find_bank_config(file_path.name)
        
        if not bank_config:
            logging.warning(f"No bank configuration found for file: {file_path.name}")
            continue
            
        logging.info(f"Processing file: {file_path} for {bank_name}")
        
        parse_function = bank_config["parse_function"]
        delimiter = bank_config.get("delimiter", ",")
        encoding = bank_config.get("encoding", "utf-8")
        
        # Process the file
        ynab_data = process_bank_statement(str(file_path), parse_function, delimiter, encoding)
        
        if ynab_data.empty:
            logging.warning(f"No data processed for {file_path}")
            continue
        
        filename_placeholders = {
            'bank': bank_name,
            'first_date': ynab_data['Date'].min().date(),
            'last_date': ynab_data['Date'].max().date(),
        }

        output_file = output_directory / Path(bank_config["output_filename"].format(**filename_placeholders))
        if not overwrite:
            output_file = get_unique_filename(output_file)
        
        # Export to CSV for YNAB import
        ynab_data.to_csv(output_file, index=False)
        logging.info(f"Data saved to {output_file}")
        files_processed = True

        if on_success == 'delete':
            logging.info(f"Deleting {file_path}")
            file_path.unlink()
        elif on_success == 'archive':
            archive_directory.mkdir(exist_ok=True, parents=True)
            file_archive_path = archive_directory / file_path.name
            if not overwrite:
                file_archive_path = get_unique_filename(file_archive_path)
            logging.debug(f"Archiving {file_path} to {file_archive_path}")
            file_path.rename(file_archive_path)
        elif on_success and on_success != 'nothing':
            logger.warning(f"Invalid operation after conversion: {on_success}")
    
    if not files_processed:
        logging.warning("No files were processed. Make sure your files match the expected patterns.")