ENSAM Traduction — Technical Document Translator

A modular, production-ready translation tool for technical documents, powered exclusively by the Groq API with a modern PyQt6 dark-themed GUI.

---

✨ Features

• 🔄 Groq API Translation — All translations powered by Groq's ultra-fast LLM inference
• 🧹 Text Preprocessing — Regex-based cleaning with technical token preservation
• 🌍 Language Detection — Automatic source language identification
• 📚 Custom Glossary — Manage technical terminology with priority-based replacement
• 🧠 Translation Memory — SQLite-backed cache to avoid redundant API calls
• 📊 Quality Evaluation — Length consistency, glossary coverage, and diff highlighting
• 📂 Multi-Format I/O — Support for TXT, DOCX, and PDF documents
• 🖥️ Modern GUI — Dark-themed PyQt6 interface with drag-and-drop support

---

📁 Project Structure

ENSAM_Traduction/
├── main.py                           # Application entry point
├── config.py                         # Configuration & settings
├── requirements.txt                  # Python dependencies
├── LOGO.png                          # Application icon
│
├── src/                              # Source code (MVC architecture)
│   ├── models/                       # Business logic modules
│   │   ├── text_cleaner.py           # Regex preprocessing pipeline
│   │   ├── language_detector.py      # Language detection
│   │   ├── glossary_model.py         # Technical glossary management
│   │   ├── translation_memory.py     # SQLite TM cache
│   │   ├── groq_translator.py        # Groq API integration
│   │   ├── document_parser.py        # TXT/DOCX/PDF parsing
│   │   ├── document_writer.py        # Multi-format output
│   │   ├── quality_evaluator.py      # Translation quality metrics
│   │   ├── pipeline.py               # Orchestration pipeline
│   │   ├── term_detector.py          # Technical term detection
│   │   ├── term_explainer.py         # AI-powered term definitions
│   │   ├── term_dictionary.py        # Personal term dictionary
│   │   └── ocr_engine.py            # Tesseract OCR integration
│   │
│   ├── views/                        # PyQt6 UI layer (pure UI, no logic)
│   │   ├── main_view.py              # Main application window
│   │   ├── glossary_editor_view.py   # Glossary editor dialog
│   │   ├── tm_viewer_view.py         # Translation memory viewer
│   │   ├── diff_viewer_view.py       # Diff comparison viewer
│   │   ├── term_dictionary_panel.py  # Term dictionary panel
│   │   ├── term_highlighter.py       # Syntax highlighter for terms
│   │   ├── drop_zone.py              # Drag-and-drop file zone
│   │   ├── styles_somber.qss         # Somber (dark) theme
│   │   └── styles_light.qss          # Light theme
│   │
│   └── controllers/                  # Application logic
│       ├── app_controller.py         # Main controller
│       └── workers.py                # QThread background workers
│
├── data/                             # Persistent data
│   ├── glossary.json                 # Default technical glossary
│   ├── term_dictionary.json          # Personal term dictionary
│   └── translation_memory.db         # Auto-created SQLite database
│
└── samples/                          # Example input files
    └── sample_technical.txt          # Sample technical document

---

🚀 Installation

Prerequisites

• Python 3.10 or higher
• A valid Groq API key

Steps

1. Clone the repository:

git clone <repo-url>
cd ENSAM_Traduction

2. Install dependencies:

pip install -r requirements.txt

3. Run the application:

python main.py

---

🎮 Usage

Text Translation

1. Type or paste text into the Source Text area
2. Select source language (or leave on "Auto-detect")
3. Select target language
4. Click ▶ Translate (or press Ctrl+Enter)

Document Translation

1. Drag & drop a .txt, .docx, or .pdf file onto the drop zone
2. Alternatively, use File → Open File (Ctrl+O)
3. Click ▶ Translate
4. Export via File → Export as TXT/DOCX/PDF

Glossary Management

Open the glossary editor via Tools → Glossary Editor (Ctrl+G):

• Add, edit, or delete technical terms
• Import/export glossaries as JSON
• Terms are automatically applied during translation

Translation Memory

View cached translations via Tools → Translation Memory (Ctrl+M):

• Browse all cached translations
• View hit statistics
• Search and manage entries

Quality Reports

After translation, check the 📊 Quality tab for:

• Overall quality score (A-F grade)
• Length ratio analysis
• Glossary term coverage
• Warnings for potential issues

---

⌨️ Keyboard Shortcuts

Shortcut	Action
Ctrl+Enter	Start translation
Ctrl+O	Open file
Ctrl+G	Open glossary editor
Ctrl+M	Open translation memory
Ctrl+D	View translation diff
Ctrl+Q	Quit

---

📦 Tech Stack

Component	Technology
Core	Python 3.10+
GUI	PyQt6
Translation	Groq API (llama-3.3-70b-versatile)
Language Detection	langdetect
Document Parsing	python-docx, PyMuPDF
PDF Generation	reportlab
Caching	SQLite3

---

🏗️ Architecture

The application follows a clean pipeline architecture:

Input → Parse → Clean → Detect Language → Glossary Pre-process
  → TM Cache Lookup → Groq API Translation
  → Glossary Post-process → Quality Evaluation → Output

All modules are loosely coupled and can be used independently.

---

📄 License

MIT License — see LICENSE for details."# ENSAMTransLAB"