Kavita-mokuro

This is a fork of Kavita with enhanced mokuro support and Live OCR capabilities.

⚠️ Important Notice: This fork is NOT keeping up with the original Kavita. There will be bugs and it will not be updated. Do not under ANY conditions bother the original Kavita team with support questions or complaints if you are using this fork.

Features

• Mokuro Support: Read Japanese manga with selectable text overlays
• Live OCR: Real-time OCR processing with dual engine support:
  • 🔥 manga-ocr (default) - Fast, offline, specialized for Japanese manga
  • 🌐 Google Lens OCR - Multilingual, excellent for mixed content
• Advanced Reader Controls: OCR type selection, status monitoring, export capabilities
• Preloading: Automatic processing of upcoming pages for smooth reading

Setup

1. Clone with Submodules

git clone --recurse-submodules https://github.com/your-username/Kavita-mokuro.git
cd Kavita-mokuro

2. Mokuro OCR Service (Optional)

Mokuro OCR functionality runs as a separate service and is not included in the Kavita build.

If you need OCR capabilities, you should run mokuro in a separate container or process. The mokuro submodule in this repository (from xrishox/mokuro) provides:

• Enhanced dual OCR engine support (manga-ocr + Google Lens)
• Improved text detection via comic-text-detector
• Separate file naming for different OCR engines

3. Build and Run Kavita

See BUILD.md for detailed instructions on building Kavita-mokuro as a Docker container or standalone application.

Configuration

1. Enable Mokuro Support: Go to Settings → Server → Mokuro and enable mokuro support
2. Configure Live OCR: Set up Live OCR server URL and port if using real-time OCR
3. Add Libraries: Add your manga libraries containing mokuro files or raw images

Usage

Pre-processed Mokuro Files

• Works with .mokuro files produced by the enhanced mokuro fork
• Supports both manga-ocr (.mo.mokuro) and Google Lens (.gl.mokuro) outputs

Live OCR Processing

• Real-time OCR processing of manga pages
• Automatic preloading of upcoming pages
• Status monitoring and progress tracking
• Export completed OCR data as mokuro files

Reader Controls

• OCR Toggle: Eye icon at the top to enable/disable overlays
• OCR Type Selection: Choose between Mokuro tracks and Live OCR engines
• Status Window: View processing progress for all pages
• Export: Convert Live OCR results to mokuro format

Dependencies

This project includes xrishox/mokuro as a git submodule, which provides:

• Enhanced dual OCR engine support (manga-ocr + Google Lens)
• Improved text detection via comic-text-detector
• Separate file naming for different OCR engines
• Advanced caching and processing capabilities

Troubleshooting

If you encounter issues:

1. For build issues, see BUILD.md
2. Ensure your mokuro service is running if using Live OCR
3. Check that manga files are in supported formats
4. Verify that mokuro files match the expected naming convention (.mo.mokuro or .gl.mokuro)