FOSK

fosk is a lightweight embedded SQL engine for Rust applications. It allows you to define in-memory collections, seed them with JSON objects, and query using a SQL-like syntax.

✨ Features

• In-memory database with collections (tables)
• Configurable ID strategies: integer, UUID, or none
• Simple JSON storage (serde_json::Value)
• SQL parser with support for:
• SELECT, WHERE, GROUP BY, HAVING
• JOIN (inner, left, right, full)
• ORDER BY, LIMIT, OFFSET
• Parameterized queries (? placeholders, including arrays)
• Test-friendly: create databases on the fly and seed them

Installation

In your Cargo.toml:

[dependencies]
fosk = "0.1.1"
serde_json = "1"

Quick example

use fosk::{Db, Config, IdType};
use serde_json::json;

fn main() {
    let db = Db::new_db_with_config(Config {
        id_type: IdType::Int,       // auto-increment IDs
        id_key: "id".into(),
    });

    let people = db.create_collection("People");

    // Add JSON
    let added = people.add(json!({"name": "Alice", "age": 30})).unwrap();

    // Retrieve auto-generated ID (if configured)
    let id = added.get("id").unwrap().to_string();
    println!("Added person with ID: {}", id);

    // Add multiple documents
    let added_many = people.add_batch(json!([
        {"name": "Bob", "age": 25},
        {"name": "Carol", "age": 28}
    ]));
    println!("Added many: {:?}", added_many);

    // Query with parameters
    let rows = db.query_with_args(
        "SELECT id, name, age FROM People WHERE id = ?",
        json!(id),
    ).unwrap();

    println!("Query result: {rows:?}");
}

📚 Main Types & API

Db

Represents a database.

• new_db_with_config(config: Config) -> Db
• create_collection(name: &str) -> CollectionHandle
• query(sql: &str) -> Result<Vec, AnalyzerError>
• query_with_args(sql: &str, args: Value) -> Result<Vec, AnalyzerError>

Config

Defines collection behavior.

• Config::int("id") → auto-increment integer IDs
• Config::uuid("id") → UUID v4 IDs
• Config::none("id") → no automatic ID field

CollectionHandle

• add(item: Value) -> Option
• get_all() -> Vec
• get(id: &str) -> Option
• update(id: &str, item: Value) -> Option
• delete(id: &str) -> Option

extras

• get_paginated(, offset: usize, limit: usize) -> Vec
• exists(id: &str) -> bool
• count() -> usize
• add_batch(items: Value) -> Vec
• update_partial(id: &str, partial_item: Value) -> Option
• clear() -> usize

Load from existing data

fn load_from_file(path: &std::path::Path) -> Result<String, String>;

let db = Db::new();
let people = db.create_collection("People");
people.load_from_file("./people.json");

fn load_from_json(json_value: &serde_json::Value) -> Result<String, String>

let db = Db::new();
let people = db.create_collection("People");
people.load_from_json(json!([
    { "id": 1,  "full_name": "Alice Johnson",    "age": 29, "city": "Porto",    "vip": true  },
    { "id": 2,  "full_name": "Bruno Martins",    "age": 34, "city": "Lisboa",   "vip": false },
    { "id": 3,  "full_name": "Carla Sousa",      "age": 41, "city": "Braga",    "vip": false },
    { "id": 4,  "full_name": "David Pereira",    "age": 25, "city": "Coimbra",  "vip": true  }
]));

🧪 Testing & Seeding

Example test seed (see fixtures::seed_db):

pub fn seed_db() -> Db {
    let db = Db::new_db_with_config(Config {
        id_type: IdType::None,
        id_key: "id".into(),
    });

    create_people(&db);
    create_products(&db);
    create_orders(&db);
    create_order_items(&db);

    db
}

⚠️ Notes

• Collection names are case-sensitive: "People" is different from "people".
• Projections normally output unqualified field names (id, name), unless duplicates exist. In case of conflicts, names are disambiguated with their collection prefix (id, o.id).

📄 License

Licensed under the MIT License. See LICENSE for details.