🈹 JSON Translator – OpenAI / Amazon

A desktop application to translate structured JSON content into multiple languages using OpenAI or Amazon Translate.
Now includes a separate Engine Selector window for credential management and a refined GUI workflow optimized for
Windows, macOS, and Linux.

---

🚀 Features

🌍 Multi-Language Translation

• Translate JSON content into multiple target languages in one go.
• All target translations (e.g., ar, fr, es) are merged into one output JSON (Non-Blog mode)
  or exported as separate files (Blog mode).
• Manage target languages dynamically through a popup interface.
  • Default languages: Arabic (ar), French (fr), Spanish (es).

---

🧠 Dual-Window Workflow

🪟 1. Engine Selector (engine_selector_gui.py)

• Choose your translation engine: OpenAI or Amazon Translate.
• Enter and save credentials to a local file (api_credentials.json) in the same folder as the app.
• Credentials never leave your device and are not stored in user profile folders.
• Once validated, click Continue → to open the translator interface.

🪟 2. Translator Main GUI (translator_main_gui.py)

• Clean, modern Tkinter interface for JSON translation.
• Handles file selection, mode selection, and progress display.
• Detailed logs go to the Messages tab; the main window shows only final success/failure status.

---

⚙️ Smart Engine Credential Handling

• Local storage only → api_credentials.json resides in the same directory as your .py or .exe.
• File format:

  {
    "openai": {
      "openai_key": "sk-xxxxxxx"
    },
    "amazon": {
      "aws_access_key": "AKIAXXXXXXX",
      "aws_secret_key": "xxxxxxxxxxxxxxxx"
    }
  }
  
  

🚀 Features

🌍 Multi-Language Translation

• Translate JSON content into multiple target languages in a single run.
• All target translations (e.g., ar, fr, es) are merged into one output JSON file.
• Manage target languages dynamically through a popup window.

Default languages: Arabic (ar), French (fr), Spanish (es)
Additional languages can be added anytime.

---

🧩 Deep JSON Handling

• Handles nested structures such as:
  • title, description, and localized text blocks.
  • additionalContent arrays and multilingual nodes.
• Maintains proper language separation and prevents overwriting of existing data.
• Removes empty or invalid text fields automatically before translation.

---

⚙️ Dual Translation Engines

• OpenAI Engine: Uses GPT-based translation for natural, human-like phrasing.
• Amazon Translate: Uses AWS neural translation for structured and enterprise-safe output.
• Built-in key/credential verification for both engines.

---

🧠 Two Operating Modes

• Blog Mode:
  Creates a separate translated file for each language (ideal for localized blog/article exports).

• Non-Blog Mode:
  Produces a single JSON file containing all translated languages (e.g., en, ar, fr, es).

Each mode automatically organizes outputs:

/YourFolder/
├── Blog/
│   ├── file_ar.json
│   ├── file_fr.json
│   └── file_es.json
└── Non-Blog/
    └── file_translated.json

---

🪟 Clean, Intuitive GUI

• Built with Tkinter for a lightweight desktop experience.
• Key components:
  • File selector for input JSON.
  • Language management popup.
  • Translation engine selector (OpenAI / Amazon).
  • Progress bar and live status messages.
  • Start / Cancel translation control.

---

🔒 Safe Data Handling

• The original JSON is never modified.
• Translations are applied to a deep copy of the structure.
• Results are saved in clearly organized folders (Blog / Non-Blog).
• Supports UTF-8 output with proper handling of non-Latin characters (e.g., Arabic, Hindi, Chinese).

---

🧰 Installation

1️⃣ Clone the repository

git clone <your-repo-url>
cd <your-repo-folder>

2️⃣ Install dependencies

pip install openai boto3 cryptography requests

💡 Tkinter comes pre-installed with most Python distributions.
If not available:

sudo apt install python3-tk  # (Linux)

---

▶️ Usage

Run the main GUI script:

python translator_main_gui.py

1. Select a JSON file containing your source language content (e.g., English).
2. Choose Source Language (default: en).
3. Add or remove Target Languages from the “Manage Languages” popup.
4. Pick your Translation Engine — OpenAI or Amazon.
5. Choose mode:
   • Blog Mode ✅ (separate per-language files)
   • Non-Blog Mode 🌐 (all languages in one file)
6. Click Start Translation — progress and messages appear in real time.

---

📁 Output Examples

✅ Non-Blog Mode

Output file:
/Non-Blog/data_translated.json

{
  "title": {
    "en": "Welcome",
    "ar": "مرحبا",
    "fr": "Bienvenue",
    "es": "Bienvenido"
  }
}

✅ Blog Mode

Folder /Blog/ contains:

data_ar.json
data_fr.json
data_es.json

---

🧩 File Structure Overview

engine_selector_gui.py        # Engine + credential setup window
translator_main_gui.py        # Main translation GUI
translator_gui_functions.py   # Threading, progress, button control
translator_logic.py           # Non-Blog translation engine
translator_blog_logic.py      # Blog translation engine
translator_lang.py            # Language management popup
translate_openai.py           # OpenAI translation utilities
translate_aws.py              # AWS Translate utilities
api_credentials.json          # Locally stored keys

---

🔑 Credentials

OpenAI

Create a .env file or provide your API key via popup:

OPENAI_API_KEY=sk-xxxxxxxxxxxx

Amazon Translate

Requires AWS access credentials:

AWS_ACCESS_KEY_ID=your-key
AWS_SECRET_ACCESS_KEY=your-secret

---

🧠 Notes

• Supports translation of large, nested JSON structures (like CMS exports, blogs, or apps).
• Automatically skips empty or invalid text nodes for Amazon Translate.
• Cleanly handles right-to-left scripts (Arabic, Hebrew) in JSON outputs.
• Translation progress and completion messages are shown in the GUI.
• Works seamlessly on Windows, macOS, and Linux.
• Background threads keep the GUI responsive at all times.
• Progress and error messages logged under “Messages”.
• Start/Stop button color feedback adds clarity to long translations.

---

🪪 License

© 2025 JSON Translator – Human-like Output
Licensed for personal and professional use under the MIT License.