vesuvius-phalanx 0.1.12

No project description provided

phalanx

A specialized tool for downloading scrolls and fragments from the Vesuvius Challenge dataset. Phalanx allows users to efficiently retrieve specific slices from volumes and fragments with parallel downloading capability.

The name "phalanx" draws inspiration from the mythical phoenix, symbolizing rebirth and revival. Just as the phoenix rises from its ashes, phalanx helps resurrect and breathe new life into ancient scrolls buried under the ashes of Mount Vesuvius.

Key Features

• Selective Downloads: Download specific slice ranges instead of entire volumes
• Multithreaded Downloading: Parallel download operations for improved performance
• Progress Tracking: Visual progress bars showing completion status and data transfer
• Smart Defaults: Automatically selects appropriate volume/fragment IDs when possible
• Error Handling: Robust retry mechanism and helpful error messages
• Clean CLI: Intuitive command-line interface with comprehensive help documentation

Installation

pip install vesuvius-phalanx

Dependencies

Package	Purpose
requests	HTTP requests and content downloading
beautifulsoup4	HTML parsing and web page data extraction
tqdm	Progress bar visualization
click	User-friendly command-line interface

Command-line Usage

Download Volume Slices

phalanx download-volume SCROLL_ID --output-path OUTPUT_PATH [OPTIONS]

Parameters:

• SCROLL_ID: Numeric ID (e.g., 5) - will be prepended with "Scroll" automatically
• --output-path: Path to store downloaded data (full_scrolls directory)
• --volpkg-name: (Optional) Specific volpkg name if multiple are available
• --volume-id: (Optional) Specific volume identifier
• --slices: (Optional) Slice ranges to download (default: all)
  • Format: "start-end" or "start-end,another-range"
  • Example: "1-5,10,15-20"

Download Fragment Slices

phalanx download-fragment SCROLL_ID FRAG_ID [OPTIONS]

• SCROLL_ID: Numeric scroll ID (e.g., 5). This value will be prepended with "Scroll" (e.g., Scroll5).
• FRAG_ID: Fragment identifier (e.g., 20241024131838).
• --output-dir: Output data root directory (default: data).
• --volpkg-name: Name of the volpkg (optional).
• --slices: Slice ranges to download (default is all).
• --mask: Download mask (default is true).

Parameters:

• SCROLL_ID: Numeric ID (e.g., 5) - will be prepended with "Scroll" automatically
• FRAG_ID: Fragment identifier (e.g., 20241024131838)
• --output-dir: (Optional) Output data root directory (default: data)
• --volpkg-name: (Optional) Specific volpkg name if multiple are available
• --slices: (Optional) Slice ranges to download (default: all)
• --mask: (Optional) Whether to download the mask (default: true)

Examples

Basic Usage

Download all volume slices for Scroll 1 with auto-detected defaults:

phalanx download-volume 1 --output-path ./data

Download specific slices from Scroll 1:

phalanx download-volume 1 --output-path ./data --slices 1-5

Advanced Usage

Download with explicit volume identifier:

phalanx download-volume 1 --output-path ./data --volume-id 20230205180739 --slices 1-5

Download a fragment with its mask:

phalanx download-fragment 1 20230503225234 --output-dir ./vesuvius_data --slices all

Download a fragment without its mask:

phalanx download-fragment 1 20230503225234 --mask false

How It Works

Phalanx operates by:

1. Discovery: Identifying available volpkgs and volumes for the requested scroll
2. Selection: Choosing appropriate defaults or using user-specified identifiers
3. Metadata Retrieval: Fetching slice information and other metadata
4. Parallel Downloads: Creating multiple worker threads for faster downloading
5. Error Handling: Implementing retries and cleanup for failed or interrupted downloads

Technical Details

• Intelligent Defaults: The tool maintains a defaults.json file with preferred volpkg and volume IDs for common scrolls
• Session Management: Creates optimized HTTP sessions with retry capabilities
• Resource Management: Cleans up partial downloads on startup and handles interruptions
• Progress Reporting: Shows real-time download progress with data transfer rates

Command Aliases

For convenience, you can also use the shorter phnx command instead of phalanx:

phnx download-volume 1 --output-path ./data

Contributing

Contributions are welcome! Please submit a pull request or open an issue on GitHub.

Before contributing:

• Fork the repository
• Create a feature branch
• Add your changes
• Submit a pull request

License

This project is licensed under the MIT License - see the LICENSE file for details.