Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.chataigne.ai/llms.txt

Use this file to discover all available pages before exploring further.

​
Translation System Implementation Guide

​
Overview

This guide explains how to use the translation system to replace hardcoded French text in agent.py.

​
Translation System Components

  1. translations.json - Contains all translations organized by context
  2. translations.py - Translation manager and helper functions
  3. t_customer() - Helper function to get translations with customer language preference

​
Usage Examples

​
1. Simple Translation (with access to payload and store_info)

# Get customer language
lang = self._get_customer_language(payload, store_info)
default_lang = store_info.get('main_language', 'fr')

# Use translation
message = t_customer(
    "errors.technical_difficulty",
    customer_language=lang,
    default_language=default_lang
)

​
2. Translation with Parameters

# Translation with placeholders
message = t_customer(
    "payment.verify_order_payment_disclaimer",
    customer_language=lang,
    default_language=default_lang,
    brand="Visa",
    last4="1234"
)

​
3. When payload/store_info not available

For methods that don’t have direct access to payload or store_info, you’ll need to:
  1. Pass language as a parameter to the method:
async def _generate_order_summary(
    self,
    order_draft: OrderDraftDTO,
    menu: CatalogDTO,
    customer_language: str = "fr",  # Add this parameter
    default_language: str = "fr",    # Add this parameter
    # ... other parameters
) -> str:
  1. Use the translation with passed language:
summary_lines = [t_customer(
    "order.order_summary",
    customer_language=customer_language,
    default_language=default_language
)]

​
Remaining Hardcoded Text to Replace

​
Payment Messages

  • Line 1751: “Votre commande est prête pour le paiement.”
  • Line 1767: “Vous allez être redirigé vers Twint…”
  • Line 1770: “Vous allez être redirigé vers Revolut…”
  • Line 1773: “Cliquez sur le lien ci-dessous pour procéder au paiement sécurisé par carte.”

​
Order Messages

  • Line 1822: “⚠️ Certains détails de votre commande ont changé :\n”
  • Line 1824: “\n\nVoici le résumé mis à jour de votre commande :”
  • Line 2572: “Résumé de commande :\n ```”
  • Line 2591: “Aucun article dans la commande, veuillez faire un choix.”

​
Delivery Messages

  • Line 2704: “Livraison prévue ”
  • Line 2711: “À récupérer en magasin ”
  • Line 2720: “Commande au nom de ”
  • Line 2723: “Notes de livraison : ”
  • Line 2727: “Notes générale : ”
  • Line 2783: “Frais de livraison : “

​
Button Labels

  • Line 1842: “Je confirme ✅”
  • Line 1846: “Payer autrement 💰”
  • Line 3946: “Je confirme ✅”
  • Line 3980: “Payer ✅”
  • Line 3984: “Changer de carte 🔄”
  • Line 3988: “Payer autrement 💰“

​
Delivery Status Messages

  • Line 5247: “Votre commande est en attente d’un livreur.”
  • Line 5250: “Un livreur a été assigné à votre commande! 🛵”
  • Line 5254: “Code de livraison: ”
  • Line 5257: “Votre livreur arrive au restaurant…”
  • Line 5262: “Votre commande est en route! 🛵”
  • Line 5276: “Votre commande a été livrée! Bon appétit! 😋“

​
Step-by-Step Process

  1. Identify the context - Determine where in the code the hardcoded text appears
  2. Check available variables - See if payload and store_info are available
  3. Get language preference - Use _get_customer_language() if possible
  4. Replace with translation - Use t_customer() with appropriate key
  5. Test - Ensure the translation works correctly

​
Adding New Translations

  1. Add to translations.json:
"new_category": {
  "new_key": {
    "fr": "French text",
    "en": "English text",
    "de": "German text",
    "it": "Italian text",
    "es": "Spanish text",
    "description": "Where this is used"
  }
}
  1. Use in code:
message = t_customer(
    "new_category.new_key",
    customer_language=lang,
    default_language=default_lang
)

​
Best Practices

  1. Always provide fallback - Use default_language parameter
  2. Group related translations - Keep translations organized by context
  3. Include descriptions - Document where each translation is used
  4. Test all languages - Ensure translations work for all supported languages
  5. Handle missing translations gracefully - The system returns the key if translation not found