Convert files from JIS (Japanese ASCII) to UTF-8 encoding

jis2utf

#!/usr/bin/env python3

import chardet
import argparse
import mimetypes
import sys

def is_text_file(file_path):
    mime_type, _ = mimetypes.guess_type(file_path)
    return mime_type is not None and mime_type.startswith('text')

def process_file(file_path):
    with open(file_path, 'rb') as file:
        raw_data = file.read()
        result = chardet.detect(raw_data)
        encoding = result['encoding']
        confidence = result['confidence']

        if encoding:
            if 'JIS' in encoding:
                print(f"{file_path} has JIS encoding, {confidence*100:.2f}% confidence")
                convert_jis_to_utf8(file_path, raw_data, encoding)
            else:
                print(f"{file_path} has {encoding} encoding, {confidence*100:.2f}% confidence")

def convert_jis_to_utf8(file_path, raw_data, encoding):
    try:
        decoded_data = raw_data.decode(encoding)
        utf8_data = decoded_data.encode('utf-8')
        with open(file_path, 'wb') as new_file:
            new_file.write(utf8_data)
        print(f"Converted {file_path} to UTF-8")
    except Exception as e:
        print(f"Error converting {file_path} to UTF-8: {e}")

def main():
    parser = argparse.ArgumentParser(description="Convert files with JIS encoding to UTF-8 encoding.")
    parser.add_argument('file_path', help="Path to the file to be analyzed.")
    args = parser.parse_args()

    if not is_text_file(args.file_path):
        print(f"{args.file_path} is not a text file, skipping")
        sys.exit(1)

    process_file(args.file_path)

if __name__ == "__main__":
    main()