HTML to Markdown Test Files for lightfeed-extract

This repository contains test files for validating HTML to LLM-extractor-ready Markdown conversion functionality. It specifically tests three conversion variants:

1. Basic Conversion - Converting all HTML content to Markdown (without images)
2. Main Content Extraction - Extracting and converting only the main content from HTML files (without images)
3. Conversion with Images - Converting all HTML content to Markdown including images

Repository Structure

├── html/             # Source HTML files
│   ├── forum/        # Forum HTML samples
│   │   ├── tech-0.html
│   │   └── ...
│   └── ...
│
└── groundtruth/      # Expected Markdown output files
    ├── forum/        # Expected forum conversion results
    │   ├── tech-0.md         # Basic conversion expected output
    │   ├── tech-0.main.md    # Main-content-only expected output
    │   ├── tech-0.images.md  # Conversion with images expected output
    │   └── ...
    └── ...

Naming Convention

Files follow a specific naming pattern to clearly indicate their purpose:

• html/[category]/[file-name].html - Original HTML source files
• groundtruth/[category]/[file-name].md - Expected output for basic HTML conversion
• groundtruth/[category]/[file-name].main.md - Expected output for main content extraction
• groundtruth/[category]/[file-name].images.md - Expected output for conversion with images

For example:

• html/forum/tech-0.html - Original forum HTML file
• groundtruth/forum/tech-0.md - Expected Markdown after basic conversion (no images)
• groundtruth/forum/tech-0.main.md - Expected Markdown when only extracting main content (no images)
• groundtruth/forum/tech-0.images.md - Expected Markdown with images included