pyshine-translator 1.3.6

Desktop translator with AI support - translate selected text with a global hotkey

PyShine Translator

🏠 PyShine.com • 📦 PyPI • 📖 Documentation

A cross-platform desktop translator application that translates selected text using a global hotkey. Supports offline translation (CPU-compatible, no internet required), Google Translate, and AI providers (OpenAI, Ollama, and custom APIs).

Features

• System Tray Application: Runs in the background with a system tray icon
• Global Hotkey: Press Ctrl+Shift+T (default) to translate selected text
• Multiple Translation Backends:
  • Offline (Default): CPU-compatible local translation using Argos Translate - works without internet!
  • Google Translate (free via googletrans)
  • AI Providers: OpenAI, Ollama, and custom OpenAI-compatible APIs
• Offline First: Default mode supports translations even without internet connection
• Ollama Integration: Automatically discovers and lists local Ollama models
• Bidirectional Translation: Auto-detects source language and translates accordingly
• Multi-language Support: English, Chinese, Japanese, Korean, Spanish, and more
• Customizable Settings: Configure hotkeys, languages, and AI providers via GUI
• Cross-Platform: Works on Windows, macOS, and Linux

Installation

From PyPI (Recommended)

pip install pyshine-translator

Platform-specific notes:

• Windows: All dependencies install automatically
• macOS/Linux: keyboard and pyautogui are Windows-only and will be skipped automatically
• All platforms: pynput provides cross-platform keyboard/mouse support

From Source

git clone https://github.com/pyshine-labs/pyshine-translator.git
cd pyshine-translator
pip install -e .

Quick Start

Command Line

After installation, run the application:

pyshine-translator

From Python

from src import ConfigManager, TranslationService

config = ConfigManager()
service = TranslationService(config)
result = service.translate("Hello, world!")

if result.success():
    print(f"Translation: {result.text}")
    print(f"Backend: {result.backend}")
else:
    print(f"Error: {result.error}")

Usage

1. Launch the Application: Run pyshine-translator from terminal/command prompt
2. System Tray Icon: Look for the translator icon in your system tray
3. Right-click Menu:
   • Enable Translation: Toggle the hotkey on/off
   • Settings: Open configuration window
   • Exit: Quit the application
4. Translate Text:
   • Select text in any application
   • Press Ctrl+Shift+T (or your custom hotkey)
   • The selected text will be replaced with the translation

Configuration

General Settings

Setting	Description	Default
Backend	Offline (Local), Google Translate, or AI Provider	Offline (Local)
Source Language	Source language for bidirectional translation	English
Target Language	Target language for translation	Chinese Simplified
Bidirectional	Auto-detect and swap translation direction	Enabled
Hotkey	Global hotkey combination	Ctrl+Shift+T

Offline Translation

The default backend uses Argos Translate for offline translation:

• Works without internet connection
• CPU-compatible (no GPU required)
• First run will download language models automatically
• Supports many language pairs

To install additional language models:

from argostranslate import package
package.update_package_index()
package.install_from_path(package.get_available_packages()[0].download())

AI Provider Settings

Navigate to the AI Providers tab in Settings to configure AI backends:

1. Ollama Models:

   • Automatically lists available local models
   • Double-click a model to add it as a provider
   • Requires Ollama to be running (ollama serve)

2. Custom Providers:

   • Click "Add Custom" to add OpenAI or custom API providers
   • Configure: Name, Type, API URL, API Key, Model Name
   • Use "Test Connection" to verify settings

Supported AI Providers

Provider	Type	API URL	API Key Required
OpenAI GPT-4	openai	https://api.openai.com/v1	Yes
OpenAI GPT-3.5	openai	https://api.openai.com/v1	Yes
Ollama	ollama	http://localhost:11434	No
Custom	custom	Your URL	Depends

Project Structure

pyshine-translator/
├── src/
│   ├── __init__.py          # Package initialization
│   ├── config_manager.py    # Configuration management
│   ├── translator.py        # Translation service
│   ├── ai_translator.py     # AI translation backends
│   ├── hotkey_manager.py    # Global hotkey handling
│   ├── logger.py            # Logging setup
│   └── tray_app.py          # System tray GUI
├── icons/
│   └── translate.png        # Application icon
├── main.py                  # Entry point
├── pyproject.toml           # Package configuration
├── requirements.txt         # Dependencies
└── README.md                # This file

Requirements

• Python 3.8 or higher
• Platform: Windows, macOS, or Linux

Dependencies

• PySide6 >= 6.5.0 (GUI)
• googletrans == 4.0.0rc1 (Google Translate)
• langdetect >= 1.0.9 (Language detection)
• keyboard >= 0.13.5 (Global hotkeys)
• pyautogui >= 0.9.54 (Clipboard simulation)
• pynput >= 1.7.6 (Input handling)
• requests >= 2.31.0 (HTTP requests)

Configuration Files

Configuration is stored in:

• Windows: %APPDATA%\TranslateContextMenu\config.json
• macOS: ~/Library/Application Support/TranslateContextMenu/config.json
• Linux: ~/.config/TranslateContextMenu/config.json

Logs are written to:

• Windows: %APPDATA%\TranslateContextMenu\translation.log
• macOS: ~/Library/Application Support/TranslateContextMenu/translation.log
• Linux: ~/.config/TranslateContextMenu/translation.log

Acknowledgments

• argostranslate - Offline translation library
• googletrans - Free Google Translate API
• PySide6 - Qt for Python GUI
• keyboard - Global hotkey support
• pyautogui - GUI automation
• Ollama - Local LLM runtime

Changelog

v1.3.5 (2026)

• Fixed OpenMP conflict properly: Created wrapper script to set environment variable before imports
• Improved entry point: Changed from direct import to wrapper script
• No more OpenMP warnings: Environment variable set before any library imports

v1.3.4 (2026)

• Fixed OpenMP conflict on Windows: Resolved libiomp5md.dll initialization error
• Added environment variable fix: Automatically sets KMP_DUPLICATE_LIB_OK=TRUE
• Improved startup stability: No more OpenMP warnings or crashes

v1.3.3 (2026)

• Fixed model selection bug: Now correctly saves model name instead of display text
• Fixed "model not found" error: Resolved issue where selected model couldn't be found
• Improved model handling: Properly extracts model name from dropdown data

v1.3.2 (2026)

• Fixed TypeError: Resolved "setEnabled called with wrong argument types" error
• Fixed button enable logic: Changed from string to boolean check for URL field
• Improved stability: Better handling of empty URL strings

v1.3.1 (2026)

• Fixed bug: Resolved "NoneType has no attribute strip" error when saving provider without model
• Improved error handling: Better handling of None values in model selection

v1.3.0 (2026)

• Auto-fetch Ollama models: Added "Fetch Models" button to automatically retrieve available models from any Ollama URL
• Improved model selection: Changed model field to dropdown combo box for easier selection
• Better Ollama integration: Automatically enables "Fetch Models" button when Ollama URL is entered
• Enhanced UI: Better feedback when fetching models and connecting to Ollama servers
• Cross-platform Ollama support: Easily connect to Ollama running on any device (Windows, Mac, Linux) over WiFi

v1.2.9 (2026)

• Improved hotkey detection: Fixed hotkey combination detection to properly track modifier keys
• Added cooldown mechanism: Prevents multiple rapid triggers of the same hotkey
• Better key tracking: Properly tracks pressed/released keys for reliable hotkey detection
• Enhanced logging: Added detailed debug logging for key press/release events

v1.2.8 (2026)

• Fixed macOS tray icon visibility: Added macOS-specific fixes to ensure tray icon appears in menu bar
• Prevented garbage collection: Keep reference to tray app to prevent icon from disappearing
• Added visibility logging: Better debugging for tray icon display issues

v1.2.7 (2026)

• Fixed hotkey support for macOS/Linux: Replaced Windows-only keyboard library with cross-platform pynput
• Added pyperclip dependency: For reliable clipboard access across all platforms
• macOS hotkey improvements: Uses Cmd key instead of Ctrl for better macOS UX
• Hotkey now works on all platforms: Windows, macOS, and Linux

v1.2.6 (2026)

• Fixed cross-platform config directory: Now uses platform-specific paths
  • Windows: %APPDATA%/TranslateContextMenu
  • macOS: ~/Library/Application Support/PyShineTranslator
  • Linux: ~/.config/pyshine-translator
• App now works on macOS and Linux!

v1.2.5 (2026)

• Added upper bounds to all dependencies to speed up pip dependency resolution
• This significantly reduces installation time on macOS/Linux

v1.2.4 (2026)

• Fixed dependency resolution: Made keyboard and pyautogui Windows-only to speed up macOS/Linux installation
• Added OS-specific classifiers (Windows, Linux, macOS) for better PyPI filtering
• Relaxed googletrans version constraint from ==4.0.0rc1 to >=4.0.0rc1 for better compatibility

v1.2.3 (2026)

• Added prominent PyShine.com branding links in README

v1.2.2 (2026)

• Fixed notification message to show actual configured hotkey instead of hardcoded value

v1.2.1 (2026)

• Fixed bidirectional translation for offline mode - now correctly uses detected language as source

v1.2.0 (2026)

• Auto-install language models: Offline translation now auto-downloads required language pairs on first use
• Improved error handling for missing language models
• Fixed googletrans compatibility issues with newer httpcore versions

v1.1.0 (2026)

• Added offline translation backend using Argos Translate
• CPU-compatible local translation - works without internet!
• Set offline as default backend for immediate use
• Supports multiple language pairs

v1.0.0 (2024)

• Initial release
• Google Translate support
• AI provider support (OpenAI, Ollama, custom)
• Automatic Ollama model discovery
• Bidirectional translation
• Cross-platform support
• System tray integration
• Configurable hotkeys
• GUI settings panel

---

🏠 PyShine.com
_{Python tutorials, AI projects, and more}