html2json

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# Filename: html2json.py
# Version: 1.0.0
# Author: Jeoi Reqi

"""
Description:
This script converts an HTML file (.html) to a JSON file (.json).
It uses BeautifulSoup to parse the HTML content and saves the prettified HTML as a JSON file.

Requirements:
- Python 3.x
- BeautifulSoup library (install using: pip install beautifulsoup4)

Usage:
1. Save this script as 'html2json.py'.
2. Ensure your HTML file ('example.html') is in the same directory as the script.
3. Install the BeautifulSoup library using the command: 'pip install beautifulsoup4'
4. Run the script.
5. The converted JSON file ('html2json.json') will be generated in the same directory.

Note: Adjust the 'html_filename' and 'json_filename' variables in the script as needed.
"""
import json
from bs4 import BeautifulSoup

def html_to_json(html_filename, json_filename):
    with open(html_filename, 'r') as htmlfile, open(json_filename, 'w') as jsonfile:
        soup = BeautifulSoup(htmlfile, 'html.parser')
        data = {"html_content": soup.prettify()}
        json.dump(data, jsonfile, indent=2)

if __name__ == "__main__":
    html_filename = 'example.html'
    json_filename = 'html2json.json'
    html_to_json(html_filename, json_filename)
    print(f"Converted '{html_filename}' to '{json_filename}'.")