Transliteration API

This project is a Flask-based API for transliterating text from English to Indian languages (currently supporting Hindi and Marathi) using the ai4bharat-transliteration library. The API allows users to input a word and get the corresponding transliteration for the requested language.

Features

• Transliteration Support: Converts English text into Hindi (hi) or Marathi (mr).
• Customizable Top-K Results: Users can specify the number of transliterations (top-k) they want in the response.
• Dockerized: Easy to build and deploy using Docker.

Requirements

• Docker installed on your machine

Input and Output

API Endpoint

Format:

GET /tl/{lang}/{word}?k={optional_topk_value}

• lang: The language code for transliteration (hi for Hindi, mr for Marathi).
• word: The English word you want to transliterate.
• k: (Optional) The number of top-k transliterations to return (default is 5).

Example Request

GET /tl/hi/namaste?k=3

Example Response

{
    "at": "2024-09-18T13:15:31.682491845+05:30",
    "error": "",
    "input": "namaste",
    "result": ["नमस्ते", "नमास्थे", "नमस्थे"],
    "success": true
}

Error Response

If an unsupported language code is provided or any other error occurs:

{
    "at": "2024-09-18T13:15:31.682491845+05:30",
    "error": "Unsupported language code: fr. Supported languages are 'hi' and 'mr'.",
    "input": "bonjour",
    "result": [],
    "success": false
}

How to Build and Run the Docker Container

Step 1: Build the Docker Image

First, clone the repository and navigate to the project directory. Then run the following command to build the Docker image:

docker build -t transliteration-api .

Step 2: Run the Docker Container

Once the image is built, run the container using:

docker run -p 6000:6000 transliteration-api

The API will be available at http://localhost:6000.

Step 3: Test the API

You can test the API using curl, Postman, or directly from your browser.

Example using curl:

curl "http://localhost:6000/tl/hi/namaste?k=5"

Development Notes

• Model Loading: On the first API call, the models for Hindi and Marathi are loaded into memory. This may take some time initially but will be faster on subsequent calls.

• Docker Configuration: The Dockerfile ensures that the necessary models are downloaded and cached during the image build process, reducing the time needed during the first request after deployment.

Files in this Project

• Dockerfile: Docker configuration for building the API image.
• app.py: The main Flask application that handles API requests.
• download_model.py: Script to download and preload the transliteration models.
• requirements.txt: Dependencies for the project, including Flask and ai4bharat-transliteration.