unmhtml 0.3.0

MHTML to HTML converter library using Python stdlib

unmhtml

Convert MHTML files to standalone HTML with embedded CSS and resources.

Installation

pip install unmhtml

Usage

from unmhtml import MHTMLConverter

# Convert MHTML file to HTML (secure by default)
converter = MHTMLConverter()
html_content = converter.convert_file('saved_page.mhtml')

# Save as standalone HTML
with open('output.html', 'w') as f:
    f.write(html_content)

# Unsafe conversion preserving original content
unsafe_converter = MHTMLConverter(
    remove_javascript=False,
    sanitize_css=False,
    remove_forms=False,
    remove_meta_redirects=False
)
html_content = unsafe_converter.convert_file('trusted_page.mhtml')

Features

• Pure Python - No external dependencies, uses only standard library
• Standalone HTML - Embeds CSS and converts resources to data URIs
• Secure by Default - All security sanitization enabled by default for safe processing
• Graceful degradation - Handles malformed MHTML files
• Memory efficient - Processes large files without excessive memory usage

Security

The library is secure by default - all security features are enabled automatically to safely display untrusted content:

Available Security Options (All Enabled by Default)

• remove_javascript=True - Removes <script> tags, event handlers (onclick, onload, etc.), and converts javascript: URLs to safe # anchors
• sanitize_css=True - Removes external CSS URLs (http://, https://, //, /absolute) while preserving relative URLs (image.png, ../fonts/font.woff) and data URIs
• remove_forms=True - Removes form elements (<form>, <input>, <textarea>, <select>) that could submit data externally
• remove_meta_redirects=True - Removes dangerous meta tags (refresh redirects, set-cookie, dns-prefetch) that could be used maliciously

Requirements

• Python 3.8+

License

MIT