Pate - AI FB2/EPUB Reader & Translator 📚

Overview

Pate is an intelligent reader application that allows you to read FB2 and EPUB books while providing real-time AI-powered translation capabilities. Perfect for language learners and readers who want to enjoy books in their original language while having instant access to translations.

Features

• 📖 Support for FB2 and EPUB formats.
• 🗜️ Handles zipped FB2 files.
• 🤖 AI-powered translation using Ollama.
• 📱 Responsive side-by-side view of original and translated text.
• 🌓 Dark/Light theme support.
• 📑 Chapter navigation.

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/pate.git
   cd pate

2. Install the required dependencies:

   pip install -r requirements.txt

3. Install Ollama from ollama.ai and pull the required model:

   ollama pull aya-expanse:32b-q4_K_M

Usage

1. Make sure Ollama is served:

   ollama serve

2. Start the application:

   make run # or you can use this from project root folder: PYTHONPATH=$(PWD) python -m streamlit run src/app.py

3. Open your browser and navigate to http://localhost:8501

How to Use

Loading a Book

1. Click the "Upload an FB2, EPUB, or ZIP" button in the sidebar.
2. Select your book file (supported formats: .fb2, .epub, .zip).
3. Wait for the file to process.

You can also use 'drag&drop' feature to load your book.

Navigating Chapters

1. Use the chapter dropdown in the sidebar to select your desired chapter.
2. The content will automatically update with the original text and translation.

Reading with Translation

• Original text appears on the left.
• AI-translated text appears on the right.
• Translations are generated in real-time as you navigate through the content.

Configuration

You can modify the translation settings in src/app_config.py:

OLLAMA_MODEL = "aya-expanse:32b-q4_K_M" # Change the AI model
TRANSLATION_LANGUAGE = "Russian" # Change target language
TRANSLATION_PROMPT = "Translate this text to {language}, don't say anything else: {text} \n{language}:" #Change prompt

The main translation logic is stored in src/utils/llm_translator. You can modify this file if necessary (e.g. use OpenAI API for translation).

Requirements

• Python 3.8+
• Streamlit 1.39.0
• ebooklib 0.18
• beautifulsoup4 4.12.3
• Ollama 0.4.2+

Also, you need a machine that can handle Ollama running your selected LLM. In my case, it was a PC with an RTX 3090.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

🗓️ Backlog

External LLM Integration

• Support for OpenAI API integration
• Add Claude/Anthropic API support
• Configurable provider selection
• API key management interface
• Fallback options between providers

Translation Management

• Save translations back to FB2/EPUB files
• Export bilingual versions of books
• Cache translations for faster loading
• Batch translation of multiple chapters
• Translation memory system

Translator's Toolkit

• Interactive translation editing
• Translation quality verification
• Glossary management for consistent translations
• Comments and annotations support
• Translation style guidelines enforcement
• Version control for edited translations

User Experience

• Customizable side-by-side view
• Reading progress tracking
• Bookmarks and notes
• Translation quality rating system
• User preferences persistence

---