Translation Caching

Goal

• A web server that exposes an API to translate a text.

Task

• Create a web server with a RESTful API to translate a text from one language to another.
• For the actual translation, an external service like Google Translate can be used.
• The source and target language should be definable via the API.
• In addition, cache (store in Database) translations, in order to avoid repeated hits to the translation API. The cache must be persistent!
• The server should have an extensible architecture. E.g. We may want to change our caching strategy or switch out our translation service.

Tech stack used

• NodeJS & ExpressJS (a flexible Node.js web application framework) as beckend.
• Google translate as an external service for actual translation.
• node-cache package for caching.
• Jest to test the APIs.

Install Dependencies

npm install

Run the app

# Run in development mode
npm run devStart

# Run in production mode
npm start

Usage

To get the translation, we can Postman or any web browser and hit this API as follow:

{URL}/translate?sourceText=&targetLanguage=''

# URL: (http://localhost:5000) or Deployed link

We have to pass two Query Params with

 _____________________________________
|_______KEY_______|_______VALUE_______|
|                 |                   |
| sourceText      |        _          |
|-----------------|-------------------|
| targetLanguage  |        _          |
|_________________|___________________|

• sourceText: the text you want to translate

• targetLanguage: the langauge to be it has to be translated
  NOTE:

  • for the targetLangauge, only ISO Language code should be passed as value.
    For example:
    • "hi" for Hindi
    • "ja" for Japanese
    • "fa" for Persian, etc.
      A list of all ISO Language Codes is attached here.

• A sample example is shown here when we hit the route on Postman.

-Same thing can be done on any web browser by hitting the GET request.

Caching

In order to avoid repeated hits to the translation API, caching of translation is done.

If the HTTP method is GET then we cache the translation. first time when a unique request is made it will take actual time to fetch the response. After that it get stored in cached for some time(here in this project I use 10 seconds but can be modified as need.). Now after each time if same request is made it will get fetched from cached translation.

For this I use node-cache npm package.

Testing

Testing on our REST API is done using jest along with supertest(which provide less complex structure to handle request.)

Test cases where written to provide whether the the request return statusCode as 200.

• To test the APIs.

  # Test during development mode
  npm run jest:watch
  
  # Test once on hitting this command
  npm jest
  

• Version: 1.0.0

• Lisence: MIT