Rust engine
Search STAC catalogs from a Rust application running on Tokio.
Use SuperSTACEngine to search catalogs from a Tokio application. It checks catalog health, discovers collections, and combines search results.
Dependencies
[dependencies]
superstac-core = "0.3"
superstac-config = "0.3"
superstac-search = "0.3"
superstac-engine = "0.3"
tokio = { version = "1", features = ["macros", "rt-multi-thread"] }You need Rust 1.88 or newer. These examples use SuperSTAC 0.3.
A complete search
Save the quickstart configuration as superstac.yml next to where you run the binary.
use superstac_config::init_from_yaml;
use superstac_core::models::storage::Storage;
use superstac_engine::SuperSTACEngine;
use superstac_search::query::SearchQuery;
#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
let storage = init_from_yaml(Storage::Memory, "superstac.yml")?;
let engine = SuperSTACEngine::new(storage);
let result = engine.search(SearchQuery {
collections: vec!["sentinel-2-l2a".into()],
ids: None,
intersects: None,
bbox: None,
datetime: Some("2024-01-01T00:00:00Z/2024-01-31T23:59:59Z".into()),
limit: Some(10),
sortby: None,
}).await;
engine.shutdown().await;
let response = result?;
for entry in response.items {
println!("{} from {:?}", entry.item.id, entry.seen_in);
}
for failure in response.metadata.failures {
eprintln!("{}: {}", failure.catalog_id, failure.reason);
}
Ok(())
}Search starts the engine if needed. Explicit engine.start().await? is useful when you want to finish startup before serving application requests. shutdown().await cancels background health tasks.
Discovery
| Method | Purpose |
|---|---|
list_collections().await | Aggregated collection availability. |
collections_by_catalog().await | Collections grouped by catalog. |
catalogs_supporting(id).await | Catalogs advertising a canonical collection. |
describe_collection(catalog_id, collection_id).await | Full collection document from one catalog. |
Shared storage
Use SuperSTACEngine::new when the engine can own the boxed storage backend. Use SuperSTACEngine::from_shared with a SharedStorage handle when your application also needs access to the catalog registry. Add catalogs before startup; if you change them later, restart the engine to repeat discovery.
Which crates to use
| Crate | Responsibility |
|---|---|
superstac-core | Catalog/provider models, settings, errors, and storage traits. |
superstac-config | YAML loading and backend population. |
superstac-search | Query translation, concurrent execution, normalization, aggregation. |
superstac-engine | Lifecycle, health, discovery, source selection. |
superstac-cli | Terminal interface over the engine. |
Generate API documentation for the exact checkout with:
cargo doc --no-deps -p superstac-core -p superstac-config -p superstac-search -p superstac-engine --openOnly memory storage is implemented. The Sqlite and Postgres variants currently construct memory backends; they do not provide persistence.
