Expand description
MBTA-RS
A simple Rust client for interacting with the Massachusetts Bay Transport Authority's V3 API*
*This project is not affiliated with any official development work from the MBTAAbout
The MBTA V3 API is described as:
A fast, flexible, standards-based API for schedules, arrival predictions, vehicle locations, and service alerts.
This project provides a simple synchronous client and data models to easily consume data from the API within your Rust code.
Built With
- ureq as the underlying HTTP client
- Serde and Serde JSON for data deserialization/serialization
- Chrono for handling datetime data
Why provide a synchronous client rather than an asynchronous one?
- I didn’t want this crate to be tied down to a specific
async
runtime - I wanted to use the
ureq
crate for its simple API and small size, and it only provides a synchronous client
Why not auto-generate a client, given that the OpenAPI/Swagger client code-generators exists?
- I’m not very familiar with any of the code generation tools available
- I’d personally prefer to have a handcrafted client with some sharper data definitions than one that is auto-generated
- There aren’t too many API endpoints as of now, so manual maintenance shouldn’t be too much of an issue once kicked off
- I like subjecting myself to unnecessary and Sisyphean tasks
Usage
It is highly recommended to have the API Swagger docs handy, as it generally contains more detailed and thorough documentation for model field than what is provided here.
In your Cargo.toml
file:
[dependencies]
mbta-rs = "*"
chrono = "*"
serde_json = "*"
Simple example usage:
use std::env;
use mbta_rs::Client;
let client = match env::var("MBTA_TOKEN") {
Ok(token) => Client::with_key(token),
Err(_) => Client::without_key()
};
let query_params = [
("page[limit]", "3")
];
let alerts_response = client.alerts(&query_params);
if let Ok(response) = alerts_response {
for alert in response.data {
println!("MBTA alert: {}", alert.attributes.header);
}
}
Map Feature
This library comes with an optional module for plotting location-related data models (stops, vehicles, shapes, etc.) onto a simple tile map.
In your Cargo.toml
file:
[dependencies]
mbta-rs = { version = "*", features = ["map"] }
staticmap = "*"
Simple example usage:
use std::{collections::HashMap, env};
use staticmap::StaticMapBuilder;
use mbta_rs::{Client, map::{Plottable, PlotStyle}};
let client = match env::var("MBTA_TOKEN") {
Ok(token) => Client::with_key(token),
Err(_) => Client::without_key()
};
let routes = client.routes(&[("filter[type]", "0,1")]).expect("failed to get routes");
let mut map = StaticMapBuilder::new()
.width(1000)
.height(1000)
.zoom(12)
.lat_center(42.326768)
.lon_center(-71.100099)
.build()
.expect("failed to build map");
for route in routes.data {
let query_params = [("filter[route]", &route.id)];
let shapes = client
.shapes(&query_params)
.expect("failed to get shapes");
for shape in shapes.data {
shape
.plot(&mut map, true, PlotStyle::new((route.attributes.color.clone(), 3.0), Some(("#FFFFFF".into(), 1.0))))
.expect("failed to plot shape");
}
let stops = client
.stops(&query_params)
.expect("failed to get stops");
for stop in stops.data {
stop.plot(
&mut map,
true,
PlotStyle::new((route.attributes.color.clone(), 3.0), Some(("#FFFFFF".into(), 1.0))),
)
.expect("failed to plot stop");
}
}
// save to file...
Contribute
See CONTRIBUTE.md to get started!
Other Acknowledgements
- Matt Vertescher’s MBTA V3 API for already existing!
- Matt Boulanger’s Mattermost API client for some of the client design inspiration
- Othneil Drew’s Best README Template for templating the
README
andCONTRIBUTE
layouts