Elasticsearch client for Rust: overview of the elasticsearch-rs library

Elasticsearch client for Rust: elasticsearch-rs

Elasticsearch is a powerful distributed search and analytics system built on Apache Lucene. To interact with it from Rust applications, there is an official client — elasticsearch-rs. This library provides a complete set of tools for working with the Elasticsearch REST API, including indexing, search, aggregations, and cluster management. In this article, we will thoroughly cover installation, configuration, and the main features of elasticsearch-rs with code examples.

What is elasticsearch-rs?

elasticsearch-rs is the official asynchronous Elasticsearch client for the Rust language, developed by Elastic. It is based on the popular asynchronous HTTP client library reqwest and provides a type-safe API for all Elasticsearch endpoints. Key advantages:

• Full support for Elasticsearch 7.x and 8.x — all core operations: CRUD, search, aggregations, index management.
• Asynchronous — uses async/await, allowing efficient handling of many requests.
• Type safety — request and response structures are strictly typed, reducing runtime errors.
• Automatic serialization/deserialization — via serde and serde_json.
• Authentication support — Basic Auth, API keys, SSL/TLS.

Installation

Add the dependency to Cargo.toml:

[dependencies]elasticsearch = "8.14.0"serde = { version = "1.0", features = ["derive"] }serde_json = "1.0"tokio = { version = "1", features = ["full"] }

Basic connection

Let's create a client connecting to a local Elasticsearch instance:

use elasticsearch::Elasticsearch;use elasticsearch::http::transport::Transport;use serde_json::json;

#[tokio::main]async fn main() -> Result<(), Box<dyn std::error::Error>> {    let transport = Transport::single_node("http://localhost:9200")?;    let client = Elasticsearch::new(transport);    // Check connection    let response = client.ping().send().await?;    println!("Elasticsearch is available: {}", response.status_code());    Ok(())}

Basic operations: CRUD

1. Indexing a document

Add a document with book information to the books index:

#[derive(Serialize, Deserialize, Debug)]struct Book {    title: String,    author: String,    year: u16,    isbn: String,}

let book = Book {    title: "Rust in Action".into(),    author: "Tim McNamara".into(),    year: 2023,    isbn: "978-5-4461-2345-6".into(),};

let response = client    .index(IndexParts::IndexId("books", "1"))    .body(json!(book))    .send()    .await?;

println!("Document indexed: {:?}", response.body());

2. Getting a document

let response = client    .get(GetParts::IndexId("books", "1"))    .send()    .await?;

if response.status_code().is_success() {    let book: Book = response.json().await?;    println!("Document found: {:?}", book);}

3. Updating a document

let response = client    .update(UpdateParts::IndexId("books", "1"))    .body(json!({        "doc": { "year": 2024 }    }))    .send()    .await?;

println!("Updated: {:?}", response.body());

4. Deleting a document

let response = client    .delete(DeleteParts::IndexId("books", "1"))    .send()    .await?;

println!("Deleted: {:?}", response.body());

Search using Query DSL

Perform a full-text search on the title field:

use elasticsearch::SearchParts;

l