import pandas as pd
from shapely.geometry import Polygon, Point
import ast
import logging
import sys

# --- Configure Logging ---
log_formatter = logging.Formatter('%(asctime)s - %(levelname)s - %(message)s')

# File Handler
file_handler = logging.FileHandler('run.log')
file_handler.setFormatter(log_formatter)

# Console Handler
console_handler = logging.StreamHandler(sys.stdout)
console_handler.setFormatter(log_formatter)

# Root Logger
root_logger = logging.getLogger()
root_logger.setLevel(logging.INFO)
root_logger.addHandler(file_handler)
root_logger.addHandler(console_handler)


# --- Load and Prepare Territory Data ---
logging.info("Loading and preparing territory data...")
territory_file = 'TerritoryExport.csv'
territories_df = pd.read_csv(territory_file)

def parse_boundary_to_polygon(boundary_str):
    try:
        coords = ast.literal_eval(boundary_str)
        return Polygon(coords)
    except (ValueError, SyntaxError, TypeError):
        return None

territories_df['Polygon'] = territories_df['Boundary'].apply(parse_boundary_to_polygon)
territories_df.dropna(subset=['Polygon'], inplace=True)
logging.info(f"Loaded {len(territories_df)} territories.")

# --- Load and Prepare Address Data ---
logging.info("Loading address data...")
addresses_file = 'Addrsses.csv'
addresses_df = pd.read_csv(addresses_file)
logging.info(f"Found {len(addresses_df)} addresses to process.")

# --- Process Each Address ---
results = []
logging.info("Processing addresses...")
for index, address_row in addresses_df.iterrows():
    user_lat = address_row['Latitude']
    user_lon = address_row['Longitude']
    user_point = Point(user_lon, user_lat)

    found_territory_id = None
    for _, territory_row in territories_df.iterrows():
        if territory_row['Polygon'].contains(user_point):
            found_territory_id = territory_row['TerritoryID']
            break
    
    # If no territory was found, assign 'OUTSIDE_TERRITORY'
    if found_territory_id is None:
        found_territory_id = 'OUTSIDE_TERRITORY'

    # Replace the first column with the found TerritoryID
    address_row.iloc[0] = found_territory_id
    results.append(address_row)
    logging.info(f"  Processed address {index + 1}/{len(addresses_df)}")

logging.info("Processing complete.")

# --- Save Results to a New CSV ---
results_df = pd.DataFrame(results)
output_file = 'Addresses_with_Territory.csv'
results_df.to_csv(output_file, index=False)

logging.info(f"Results saved to {output_file}")