Route Overview • Architecture • How To Run • Local Environment • Technologies
Managing multilingual content can be challenging, especially when new languages are introduced. Traditional approaches often leave gaps where new languages lack translations for existing content, creating an inconsistent user experience and significant manual translation overhead.
I have created this application to solve the above issue by leveraging Azure AI Translation to automatically generate translations for new content and retroactively fill gaps when new languages are added. When a user creates or updates an entity, the application translates each field into all supported languages, ensuring every piece of content is immediately accessible in all available languages. Additionally, when a new language is added, the application translates all fields from the application’s set default language, removing the need for manual back-filling. Users see content in their browser’s language if supported or default to English (en-US), ensuring seamless multilingual accessibility.
For a extensive list of available API routes, please refer to the wiki at: https://github.com/Dominicod/AutoFieldTranslationExperiment/wiki/API-Documentation.
This application also does support Swagger, which can be accessed at the root of the application when in development.
(e.g. https://localhost:7101/swagger)
When adding languages, the "Code" relates to a browser language code. There will be a route to help with this in the future, see issue: #3
This project is based on a semi-traditional "N-Layer" architecture. The project is divided into the following layers:
- Web: This layer contains all application/client facing logic (DTOs, Services, Controllers, Ect...).
- Domain: This layer contains all domain logic. (Entities, Value Objects, Ect...).
- Infrastructure: This layer contains all infrastructure logic. (Data Access, External Services, Ect...).
To clone and run this application, you'll need Git and .NET 8 installed on your computer. From your command line:
# Clone this repository
$ git clone https://github.com/Dominicod/AutoFieldTranslationExperiment.git
# Go into the repository
$ cd AutoFieldTranslationExperiment/src/Web
# Run the app
$ dotnet runA local environment can be set up by creating a .secrets.json file in the Web project.
The file should contain the following:
{
"ConnectionStrings": {
"DefaultConnection": "Server=tcp:{YOUR_SERVER_NAME}.database.windows.net,1433;Initial Catalog={YOUR_CATALOG};Persist Security Info=False;User ID={YOUR_USER_ID};Password={YOUR_PASSWORD};MultipleActiveResultSets=False;Encrypt=True;TrustServerCertificate=False;Connection Timeout=30;"
},
"Keys": {
"Azure": {
"AIService": "{YOUR_AZURE_TRANSLATION_SERVICE_KEY}"
}
},
"AzureRegions": {
"AIService": "{YOUR_AZURE_TRANSLATION_SERVICE_REGION}"
}
}These secrets are all required for the project to function, and if they are missing an exception will be thrown.
To setup Azure AI you will need to follow the documentation located here: Azure Translator Documentation
To setup the database you will need to create a new database in Azure SQL and replace the connection string with the one provided.
Running database migrations is easy. Ensure you add the following flags to your command (values assume you are executing from repository root)
--project src/Infrastructure--startup-project src/Web--output-dir Migrations
For example, to add a new migration from the root folder:
dotnet ef migrations add "SampleMigration" --project src\Infrastructure --startup-project src\Web --output-dir Migrations