import os
import sys
import re
import json

def apply_error_correction(text):
    # Load the JSON file that contains your translations
    with open('error_correction.json', 'r', encoding='utf-8') as file:
        correction_dict = json.load(file)

    # Combine keys into a single regex pattern
    pattern = r'\b(' + '|'.join(re.escape(key) for key in correction_dict.keys()) + r')\b'

    def replacement_func(match):
        key = match.group(0)
        return correction_dict.get(key, key)

    # re.subn returns a tuple (new_string, number_of_subs_made)
    corrected_text, count = re.subn(pattern, replacement_func, text)
    return corrected_text, count

def process_folder(root_folder):
    """
    Walk through root_folder and process .md files.
    """
    for dirpath, _, filenames in os.walk(root_folder):
        for filename in filenames:
            if filename.lower().endswith(".md"):
                file_path = os.path.join(dirpath, filename)
                
                with open(file_path, 'r', encoding='utf-8') as file:
                    text = file.read()

                corrected_text, error_count = apply_error_correction(text)
                
                # Only write to file if at least one error was corrected.
                if error_count > 0:
                    with open(file_path, 'w', encoding='utf-8') as file:
                        file.write(corrected_text)
                    print(f"{file_path} - {error_count} errors corrected.")

if __name__ == "__main__":
    if len(sys.argv) != 2:
        print("Usage: python apply_correction.py <root_folder>")
        sys.exit(1)
    
    root_folder = sys.argv[1]
    process_folder(root_folder)