How I Support Over 100 Languages in My React app...and You Can Too!

I'm Richard Kerrigan, and I'm thrilled to be here today to show you how you can add multilingual support into your React apps. The problem we're trying to solve is that most websites, web apps, only publish content in one language, which is not convenient for users and is not inclusive for people for whom that language is not their primary language. We need a way to support other languages and dialects on our websites and web apps without adversely affecting our team's productivity. Now there are various ways we could approach solving the problem, depending on the resources we have available to us. But for the sake of this talk, let's assume that we don't have much budget to dedicate to this feature, nor do we have a team that can translate content for us. This is a fantastic use case for AI and automation. When my team started tackling this problem, we found the Azure AI translator service. It offers ad hoc text translation and document translation, with key features such as language support, cost-effectiveness, and accuracy. It also allows the use of custom glossaries to tweak the translation process. However, there are some limitations, such as compatibility with certain file types and the need for result validation and possible tweaking. Let's now focus on designing the workflow for translating content stored as JSON. Here's an example of how you could achieve this. We're taking the pieces of data that need to be translated and sending each piece through the translator. Then, we send the translated content as a response. For JSON content, the translation can be done at runtime in a React server component or an API route. For HTML or Markdown files, the translation can be incorporated into the CI-CD workflow, involving Azure Blob Storage, Azure Function App, and the app's source code files. Instead, we've moved the translation process to our GitHub Actions workflow. We provision the necessary resources in Azure, build and deploy the function app, and then perform the translation process. We configure the translation endpoint, specify the glossary for tweaking translations, and execute the translation. Once the translation is complete, we create a new blob in the output location. Our GitHub workflow ensures that all translations are done before proceeding to the next step, where we upload and download the artifact for deployment. This ensures that all content is translated and avoids errors for missing files. When setting up the translation process, it felt like magic. The pipeline for translating Markdown content is the same as for HTML content. The difference lies in how the metadata is handled. In HTML, the metadata is stored in the head of the file, while in Markdown, it is within the front matter section. When translating HTML files, the head needs to be translated as part of the process. For Markdown files, the metadata needs to be translated separately due to issues with Azure AI Translator. The translation function parses and translates the original files' metadata, then combines it with the translated files' content. The GitHub actions workflow provisions Azure resources, builds and deploys the function app, and runs the translation process for both HTML and Markdown files. The process is now deploying the ARM template that includes the necessary resources. We discussed the problem of limited language support on websites and web apps, the benefits of Azure AI Translator, and alternative solutions. We also explored different versions of the app for translating JSON, HTML, and markdown content without learning a new language. If interested, you can find more information and resources on GitHub and Azure AI Translator. The workflow is now deploying and translating the markdown files. It ensures all files are translated before deployment. Once the translation is complete, the files are put back into the repository. The deployment process guarantees the presence of all necessary resources. After a brief wait, the web app is ready to be viewed. The blog contains multiple posts. The metadata provides the title and excerpt for each post. The markdown is parsed into HTML for rendering. Formatting is preserved in different languages. Feel free to connect with me on LinkedIn or X.