import streamlit as st
import csv
import os
import re

def delete_existing_state_csv_files():
    """Delete all CSV files with 6-character names in the current directory."""
    for file in os.listdir('.'):
        if re.match(r'^[A-Z]{2}\.csv$', file, re.I):
            os.remove(file)

def clean_column_name(column_name):
    """Remove spaces and punctuation from column names, keeping only alphabets."""
    return re.sub(r'[^a-zA-Z]', '', column_name)

def process_file(input_file_path):
    # Delete existing state CSV files before processing
    delete_existing_state_csv_files()
    
    # Open the input file for reading
    with open(input_file_path, mode='r', encoding='utf-8') as input_file:
        reader = csv.DictReader(input_file)
        # Prepare headers with cleaned column names and truncate to the first 107 fields
        headers = [clean_column_name(header) for header in reader.fieldnames[:107]]
        
        state_files = {}  # Dictionary to keep track of open file handles for each state
        
        for row in reader:
            state = row['Provider Business Mailing Address State Name']
            
            # Skip if state is not exactly two letters
            if not re.match(r'^[A-Z]{2}$', state, re.I):
                continue
            
            # Generate the file name based on the state
            file_name = f'{state}.csv'
            
            # Check if we've already opened this state file
            if state not in state_files:
                # Open a new file for the state and write the header
                state_file = open(file_name, mode='w', newline='', encoding='utf-8')
                writer = csv.DictWriter(state_file, fieldnames=headers)
                writer.writeheader()
                state_files[state] = state_file
            else:
                # Get the writer for the already opened file
                writer = csv.DictWriter(state_files[state], fieldnames=headers)
            
            # Write the current row to the state's file, cleaning and truncating each row to the first 50 fields
            cleaned_row = {clean_column_name(k): v for k, v in list(row.items())[:50]}
            writer.writerow(cleaned_row)
    
    # Close all open state files
    for file in state_files.values():
        file.close()

if __name__ == "__main__":
    input_file_path = 'npidata_pfile_20050523-20240107.csv'  # Replace with the path to your large file
    process_file(input_file_path)