Multilingual Elasticsearch Search

A Spring Boot REST API for storing and searching multilingual documents using Elasticsearch. This project enables efficient full-text search across multiple languages, making it ideal for applications requiring internationalization and multilingual content management.

Table of Contents

• Features
• Architecture
• Requirements
• Getting Started
  • Running with Docker
  • Manual Setup
• API Documentation
  • Insert Document
  • Search Documents
• Data Model
• Configuration
• Testing
• Troubleshooting
• License
• Author

---

Features

• Store documents with content in multiple languages.
• Full-text search across all languages in a single query.
• RESTful API for document management.
• Optimized Elasticsearch index and mapping.
• Dockerized for easy deployment.
• Secure Elasticsearch setup with authentication.

---

Architecture

graph TD
  Client["Client (e.g., Postman, Frontend App)"]
  API["Spring Boot REST API"]
  ES["Elasticsearch 9.0.3"]

  Client -->|HTTP/JSON| API
  API -->|Elasticsearch Java API| ES

Loading

---

Requirements

• Java 17+
• Maven 3.6+
• Docker & Docker Compose (for containerized setup)

---

Getting Started

Running with Docker

1. Clone the repository:

   git clone https://github.com/ArianKheir/Multilingual-ElasticSearch.git
   cd Multilingual-ElasticSearch

2. Start the services:

   docker compose up --build

   This will start:

   • Elasticsearch (secured, port 9200)
   • Spring Boot API (port 8080)

3. Access the API:

   • API base URL: http://localhost:8080/api/documents
   • Elasticsearch: http://localhost:9200 (user: elastic, password: see docker-compose.yaml)

Manual Setup

1. Start Elasticsearch:

   • Install and run Elasticsearch 9.0.3 locally, or use Docker as above.

2. Configure environment variables:

   • Set SPRING_ELASTICSEARCH_URIS, SPRING_ELASTICSEARCH_USERNAME, and SPRING_ELASTICSEARCH_PASSWORD as needed.

3. Build and run the Spring Boot app:

   cd multilingual-search
   mvn clean package
   java -jar target/*.jar

---

API Documentation

Insert Document

• Endpoint: POST /api/documents
• Description: Insert a new multilingual document.
• Request Body:

  {
    "identifier": "doc1",
    "body": {
      "en": "Hello",
      "fa": "سلام"
    }
  }

• Response: 200 OK

  {
    "identifier": "doc1",
    "body": {
      "en": "Hello",
      "fa": "سلام"
    },
    "bodyAll": "Hello سلام"
  }

• Errors:
  • 409 Conflict if the identifier already exists.

Search Documents

• Endpoint: GET /api/documents/search?text={query}&size={N}
• Description: Search for documents containing the given text in any language.
• Parameters:
  • text (required): The search query.
  • size (optional): Max number of results (default: 10).
• Response: 200 OK

  [
    {
      "identifier": "doc1",
      "body": {
        "en": "Hello",
        "fa": "سلام"
      },
      "bodyAll": "Hello سلام"
    }
    // ... more documents
  ]

---

Data Model

A document is represented as:

Field	Type	Description
identifier	string	Unique document ID
body	map[string]string	Language code to text (e.g., "en", "fa")
bodyAll	string	Concatenated text for optimized search

---

Configuration

• Elasticsearch index: multilingual_documents
• Index mapping: Ensures bodyAll is mapped as text for full-text search.
• Environment variables:
  • SPRING_ELASTICSEARCH_URIS
  • SPRING_ELASTICSEARCH_USERNAME
  • SPRING_ELASTICSEARCH_PASSWORD

See docker-compose.yaml for example values.

---

Testing

• Unit and integration tests are provided using JUnit and Mockito.
• To run tests:

  cd multilingual-search
  mvn test

---

Troubleshooting

• Elasticsearch connection errors: Ensure the service is running and credentials match.
• Port conflicts: Change the exposed ports in docker-compose.yaml if needed.
• Duplicate document error: Identifiers must be unique.

---

License

This project is licensed under the MIT License.

---

Author

Arian Kheirandish

---

Contributions are welcome! For issues or feature requests, please open an issue or submit a pull request.