blaise (/blɛz/) named after Blaise Pascal, a French mathematician and physicist who created the first public transit system in Paris in 1662 called the Carrosses à cinq sols.

blaise is a high-performance, fully local transit engine. It removes the "bottleneck" of developing transit applications and simulations by handling the complex infrastructure of schedule processing and routing entirely on your own hardware.

Why blaise?

Developing apps that use public transit data usually means dealing with high-latency external APIs or the overwhelming complexity of raw GTFS files. blaise provides a better way:

• Total Control: No API keys, no rate limits, and no "black-box" routing. You own the code and the data.

• Built for Scale: Designed for simulations and high-traffic tools, allowing you to run thousands of queries per second with zero network latency.

• Privacy First: Your users' location data and search queries never leave your environment.

• Transparent Logic: Built in Rust for speed and predictability. The relationship between stops, trips, and routes is clear and easy to query.

Core Features

• Fast Routing: Efficient schedule-based pathfinding using a highly optimized version of the RAPTOR algorithm.

• Smart Search: A built-in fuzzy search engine that handles partial names, typos, and abbreviations for stops and areas.

• Geospatial Lookups: Instant discovery of transit options based on geographic coordinates using spatial indexing.

Note

This project is early in development, if you like the idea and want to help improve it, please open an issue.

Server

While blaise is a library, there is also a standalone server for projects that cannot directly integrate with the Rust crate. The blaise server wraps this functionality in a ready-to-use REST API, supporting search, proximity queries, and routing out of the box.

Read more

Installation

cargo add blaise

Quick Start

use blaise::gtfs::{GtfsReader};
use blaise::repository::Repository;
use blaise::raptor::{Raptor, Location};
use blaise::shared::{time::Time, geo::Coordinate};

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let reader = GtfsReader::new().from_zip("gtfs_data.zip")?;
    let repo = Repository::new().load_gtfs(reader)?;

    let from = Location::Stop("STOP_ID_1".into());
    let to = Location::Coordinate(Coordinate { latitude: 59.3, longitude: 18.0 });
    let departure = Time::from_hms("16:00:00")?;

    let raptor = Raptor::new(&repo, from, to).departure_at(departure);
    let itinerary = raptor.solve()?;
    println!("Found a path with {} legs!", itinerary.legs.len());
    Ok(())
}

Roadmap

• Web server crate with docker image
• Multi-threaded routing (Switching to RAPTOR)
• Arrival and departure time constraints for routing
• Real-time data updates (GTFS-RT)
• Advanced date and holiday filtering

Refrences

• GTFS Specification
• Development Data (Sweden)
• RAPTOR

License

This project is licensed under the MIT License.