How ramjke/translumo works
Translumo is a Windows real-time screen translator. The principal workflow is: a user invokes a hotkey, selects or uses a capture region, OCR engines recognize screen text, machine-learning validation selects the best result, a configured translator produces the target language, and the chat overlay displays it. The tree also shows multiple OCR and translation providers, optional speech output, encrypted settings, and gamepad input. Some orchestration links are inferred from documented behavior and naming; sampled source directly confirms OCR creation, Python integration, and hotkey dispatch.
Main entry points: TranslatorFactory.cs Translator Factory · TranslationProcessingService.cs Translation Processing · TextDetectionProvider.cs Text Detection · OcrEnginesFactory.cs OCR Factory
Components 28
Input & Capture
- Hotkey Manager
src/Translumo/HotKeys/HotKeysServiceManager.cs - Selection Window
src/Translumo/SelectionAreaWindow.xaml.cs - Screen Capturer
src/Translumo/Services/ScreenCapturerFactory.cs - Translation Processing
src/Translumo.Processing/TranslationProcessingService.cs
Text Recognition
- Text Detection
src/Translumo.Processing/TextProcessing/TextDetectionProvider.cs - OCR Factory
src/Translumo.OCR/OcrEnginesFactory.cs - Windows OCR
src/Translumo.OCR/WindowsOCR/WindowsOCREngine.cs - Tesseract OCR
src/Translumo.OCR/Tesseract/TesseractOCREngine.cs - EasyOCR
src/Translumo.OCR/EasyOCR/EasyOCREngine.cs - Python Runtime
src/Translumo.Infrastructure/Python/PythonEngineWrapper.cs - Text Validity Model
src/Translumo.Processing/TextProcessing/TextValidityPredictor.cs - Text Result Cache
src/Translumo.Processing/TextProcessing/TextResultCacheService.cs
Translation
- Translator Factory
src/Translumo.Translation/TranslatorFactory.cs - DeepL Translator
src/Translumo.Translation/Deepl/DeeplTranslator.cs - Google Translator
src/Translumo.Translation/Google/GoogleTranslator.cs - Yandex Translator
src/Translumo.Translation/Yandex/YandexTranslator.cs - Papago Translator
src/Translumo.Translation/Papago/PapagoTranslator.cs
Output & Speech
- Chat Mediator
src/Translumo/Services/ChatUITextMediator.cs - Chat Window Model
src/Translumo/MVVM/Models/ChatWindowModel.cs - TTS Factory
src/Translumo.TTS/TtsFactory.cs - Windows Speech
src/Translumo.TTS/Engines/WindowsTTSEngine.cs
Configuration
- Encrypted Settings
src/Translumo/Configuration/ConfigurationStorage.cs
Other components
- User
- Desktop Screen
- DeepL Service
- Google Service
- Yandex Service
- Papago Service
Connections 27
- User → Hotkey Manager: presses hotkey
- Desktop Screen → Screen Capturer: provides pixels
- Hotkey Manager → Selection Window: opens selection
- Selection Window → Screen Capturer: sets region
- Hotkey Manager → Translation Processing: starts translation
- Screen Capturer → Translation Processing: provides image
- Translation Processing → Text Detection: detects text
- Text Detection → OCR Factory: requests engines
- OCR Factory → Windows OCR: creates engine
- OCR Factory → Tesseract OCR: creates engine
- OCR Factory → EasyOCR: creates engine
- EasyOCR → Python Runtime: executes Python
- Text Detection → Text Validity Model: scores results
- Text Detection → Text Result Cache: caches results
- Translation Processing → Translator Factory: requests translation
- Translator Factory → DeepL Translator: selects provider
- Translator Factory → Google Translator: selects provider
- Translator Factory → Yandex Translator: selects provider
- Translator Factory → Papago Translator: selects provider
- DeepL Translator → DeepL Service: sends request
- Google Translator → Google Service: sends request
- Yandex Translator → Yandex Service: sends request
- Papago Translator → Papago Service: sends request
- Translation Processing → Chat Mediator: publishes result
- Chat Mediator → Chat Window Model: updates overlay
- Chat Mediator → TTS Factory: speaks text
- TTS Factory → Windows Speech: creates engine