pydoll-substack2md

A CAPTCHA-safe Substack scraper with automatic Cloudflare bypass and human-like behavior

pydoll-substack2md is a Python tool for downloading free and premium Substack posts that handles modern web challenges:

🛡️ Automatic Cloudflare bypass - No manual intervention needed 🤖 CAPTCHA handling - Built-in solver for common challenges 🕰️ Human-like scraping - Random delays and respectful rate limiting 🔒 Premium content support - Login capability for paid subscriptions 📁 Organized output - Numbered posts by date, Markdown + HTML formats

Built on Pydoll, a powerful browser automation library that handles anti-bot measures automatically.

Key Features

🛡️ Anti-Bot Protection Handling

• Automatic Cloudflare bypass - No manual solving needed
• CAPTCHA support - Built-in handling for common challenges
• Stealth mode - Mimics real browser behavior
• Smart retries - Automatic retry with backoff strategies

🤖 Human-Like Scraping

• Random delays - Configurable delay ranges between requests
• Respectful rate limiting - Default 1-3 second delays
• Browser fingerprinting - Realistic browser profiles
• Session persistence - Maintains cookies and state

📥 Content Management

• Markdown conversion - Clean, readable Markdown files
• Image downloading - Local storage with smart naming
• Post numbering - Chronological ordering (01-oldest to newest)
• Continuous updates - Fetch only new posts on subsequent runs
• Premium content - Login support for paid subscriptions

⚡ Performance & Reliability

• Concurrent scraping - Multiple posts at once
• Async architecture - Non-blocking I/O operations
• Resource optimization - Blocks unnecessary assets
• Error resilience - Continues on individual post failures

How It Handles Protected Sites

Cloudflare Protection

When encountering Cloudflare's "Checking your browser" page, pydoll-substack2md:

1. Automatically detects the challenge
2. Waits for JavaScript execution
3. Solves challenges without user intervention
4. Proceeds with scraping once verified

CAPTCHA Handling

The tool uses Pydoll's built-in CAPTCHA solving capabilities:

# Automatic handling in the code
async with tab.expect_and_bypass_cloudflare_captcha():
    await tab.go_to(url)

Human-Like Behavior

To avoid detection and respect servers:

• Random delays between 1-3 seconds (configurable)
• Realistic mouse movements and clicks
• Maintains browser session and cookies
• Uses real Chrome/Edge browser (not headless by default)

Requirements

• Python 3.10 or higher, Python 3.11 recommended
• Chrome or Edge browser installed

Installation

Install from PyPI (Recommended)

pip install substack2md

View on PyPI: https://pypi.org/project/substack2md/

Install from Source

Clone the repository:

git clone https://github.com/cognitive-glitch/pydoll-substack2md.git
cd pydoll-substack2md

Usage

After installing with pip install substack2md, you can use the command directly:

# Use the short command
substack2md https://example.substack.com

# Or the full command
substack2markdown https://example.substack.com

# With login for premium content
substack2md https://example.substack.com --login

# Manual login mode (works with any login method)
substack2md https://example.substack.com --manual-login

# Run with custom options
substack2md https://example.substack.com -n 10 --headless

Running from Source with uv

If you cloned the repository and want to run without installing:

# Run directly with uv - it handles all dependencies automatically
uv run substack2md https://example.substack.com

# With login for premium content
uv run substack2md https://example.substack.com --login

# Run with custom options
uv run substack2md https://example.substack.com -n 10 --headless

Configuration

For premium content access, create a .env file:

# Copy the example file
cp .env.example .env

# Edit .env with your credentials
SUBSTACK_EMAIL=your-email@domain.com
SUBSTACK_PASSWORD=your-password

Advanced Options

# Scrape only 10 posts
substack2md https://example.substack.com -n 10

# Run in headless mode (default is non-headless for user intervention)
substack2md https://example.substack.com --headless

# Use concurrent scraping for better performance
substack2md https://example.substack.com --concurrent --max-concurrent 5

# Specify custom directories
substack2md https://example.substack.com -d ./posts --html-directory ./html

# Custom browser path
substack2md https://example.substack.com --browser-path "/path/to/chrome"

# Custom delay between requests (respectful rate limiting)
substack2md https://example.substack.com --delay-min 2 --delay-max 5

# Continuous/incremental mode - only fetch new posts since last run
substack2md https://example.substack.com --continuous

Continuous Fetching & Post Numbering

Automatic Post Numbering

Posts are automatically numbered based on their publication date (oldest first):

• 01-first-post-title.md
• 02-second-post-title.md
• 03-latest-post-title.md

This makes it easy to read posts in chronological order.

Continuous/Incremental Mode

Use the --continuous or -c flag to only fetch new posts since your last run:

# First run - fetches all posts
substack2md https://example.substack.com

# Later runs - only fetches new posts
substack2md https://example.substack.com --continuous

The tool maintains a .scraping_state.json file in the output directory to track:

• The latest post date and URL
• The highest number used
• Previously scraped URLs

This allows you to run the scraper periodically to keep your collection up-to-date without re-downloading existing posts.

Output Structure

After running the tool, you'll find:

├── substack_md_files/      # Markdown versions of posts
│   └── {author_name}/      # Organized by Substack author
│       ├── images/         # Downloaded images for posts
│       │   ├── image1.jpg
│       │   └── image2.png
│       ├── post1.md
│       ├── post2.md
│       └── ...
├── substack_html_pages/    # HTML versions for browsing
│   └── {author_name}.html  # Single HTML file per author
├── data/                   # JSON metadata files
└── assets/                 # CSS/JS for HTML interface

pydoll-substack2md/ ├── substack_md_files/ # Markdown files organized by author │ └── author-name/ │ ├── post-title-1.md │ ├── post-title-2.md │ ├── ... │ └── images/ # Downloaded images from posts │ ├── image1.jpg │ └── image2.png ├── substack_html_pages/ # HTML interface for browsing │ └── author-name.html └── data/ # JSON metadata for the HTML interface └── author-name_data.json

## Development

```bash
# Install development dependencies
uv pip install -e ".[dev]"

# Run tests
uv run pytest

# Format code
uv run black .

# Lint
uv run ruff check . --fix

# Type check
uv run pyright

# Run pre-commit hooks
pre-commit run --all-files

Migration to Pydoll

This project has been migrated from Selenium to Pydoll for improved performance and reliability. Key benefits include:

• Faster execution: Direct Chrome DevTools Protocol connection
• Better reliability: Event-driven architecture for dynamic content
• Async support: Concurrent post scraping capabilities
• Cloudflare handling: Built-in bypass for protected sites
• Resource optimization: Block images/fonts for faster loading

Environment Variables

Configure the tool using a .env file (see .env.example for template):

• SUBSTACK_EMAIL: Your Substack account email
• SUBSTACK_PASSWORD: Your Substack account password
• HEADLESS: Set to true for headless browser mode (default: false)
• BROWSER_PATH: Custom path to Chrome/Edge binary (optional)
• USER_AGENT: Custom user agent string (optional)

Viewing Output

The tool generates both Markdown files and an HTML interface for easy viewing. To view the raw Markdown files in your browser, you can install the Markdown Viewer browser extension.

Alternatively, you can use the Substack Reader online tool built by @Firevvork, which allows you to read and export free Substack articles directly in your browser without any installation. Note that premium content export is only available in the local version.

Contributing

Contributions are welcome! Please ensure all tests pass and code is formatted before submitting a PR.

License

MIT License - see LICENSE file for details.

Acknowledgments

• Original project by timf34
• Web version by @Firevvork
• Built with Pydoll and html-to-markdown