#csv #nosql #db #embedded-database

csv_db

A simple embedded NoSQL database using CSV files for storage

5 releases

0.2.0 Sep 12, 2023
0.1.3 Sep 12, 2023
0.1.2 Sep 12, 2023
0.1.1 Sep 11, 2023
0.1.0 Sep 11, 2023

#1459 in Database interfaces

MIT/Apache

17KB
118 lines

CSV DB

A simple embedded NoSQL database using CSV files for storage.

It allows using CSV files to perform these operations on collections of documents:

  • find
  • insert
  • delete
  • update

Design

csv_db wraps a thin layer around the csv crate, providing a Database struct with the usual methods expected from a database library. Its main purpose is to abstract the user further away from the low level details of CSV files, namely dealing with opening files and working with records. Additionally the crate tries to be as generic as possible, allowing the user to seamlessly use any custom type as document. In order to achieve this easily, csv_db's methods expect generic documents to implement Serialize/Deserialize traits using serde's derive macros.

The methods are all asynchronous, so that you can easily integrate this crate into asynchronous code. To achieve this, and since the csv crate isn't asynchronous, each method of csv_db wraps any potentially blocking code (like opening files or dealing with records) inside a tokio::task::spawn_blocking function.

Starting with version 0.2.0, all method invocations that write data, namely insert, delete and update, will automatically try to create the collection file, if it doesn't already exist, including all parent folders. For instance, if you try to insert into the users collection, the file {path}/users.csv, will be created if it doesn't exist, including all folders in {path}.

Examples

use csv_db::Database;
use serde::{Deserialize, Serialize};

#[derive(Debug, Deserialize, PartialEq, Serialize)]
struct User {
    id: usize,
    first_name: String,
    last_name: String,
    age: u32,
}

#[tokio::main]
async fn main() {
    // Create a new Database with a path for the base folder.
    // Optionally provide an extension for the files (csv by default).
    let db = Database::new("data", None);

    let user = User {
        id: 1,
        first_name: String::from("First"),
        last_name: String::from("Last"),
        age: 20,
    };

    // Insert a new user into the users collection.
    db.insert("users", user)
        .await
        .expect("Problem inserting user.");

    // Find users by filtering with a predicate on the users collection.
    let adults = db
        .find("users", |u: &User| u.age >= 18)
        .await
        .expect("Problem searching user.");

    let user = User {
        id: 1,
        first_name: String::from("First"),
        last_name: String::from("Last"),
        age: 21,
    };

    // Update a user by filtering with a predicate on the users collection.
    db.update("users", user, |u: &&User| u.id == 1)
        .await
        .expect("Problem updating user.");

    // Delete a user by filtering with a predicate from the users collection.
    db.delete("users", |u: &&User| u.id == 1)
        .await
        .expect("Problem deleting user.");
}

Dependencies

~4–11MB
~91K SLoC