aegis-geo

Geospatial engine for Aegis-DB — named collections of {id, lat, lon, metadata} features on a uniform grid spatial index, with radius / bounding-box / nearest-k queries over great-circle (Haversine) distance. The ninth data paradigm in the Aegis engine.

Features

• Grid spatial index — points bucketed into fixed-size lat/lon cells, so radius and bbox queries only scan the cells overlapping the query region.
• Haversine distance — great-circle distance in metres (spherical Earth).
• Radius query — every feature within radius_m of a point, nearest first.
• Bounding-box query — every feature inside a lat/lon rectangle.
• Nearest-k — exact k-nearest; the search radius expands geometrically until ≥ k points are enclosed, so the true k nearest are guaranteed.
• Metadata filtering — exact-match JSON filter on results.
• Snapshot persistence — a serializable snapshot the server stores and rebuilds the index from on load.

Example

use aegis_geo::GeoEngine;

let geo = GeoEngine::new();
geo.create_collection("cities")?;
geo.upsert("cities", "nyc", 40.7128, -74.0060, serde_json::json!({"country": "us"}))?;
geo.upsert("cities", "chicago", 41.8781, -87.6298, serde_json::json!({"country": "us"}))?;

// Everything within 2000 km of New York, nearest first.
let hits = geo.within_radius("cities", 40.7128, -74.0060, 2_000_000.0, &serde_json::Value::Null)?;
for hit in hits {
    println!("{}  {:.0} km", hit.id, hit.distance_m / 1000.0);
}

// The 3 nearest cities to a point near Washington DC.
let near = geo.nearest("cities", 39.0, -77.0, 3, &serde_json::Value::Null)?;

HTTP API (via aegis-server)

Tests

Workspace total includes Haversine accuracy vs a known city-pair distance, nearest-k vs brute force, radius + bbox membership, metadata filter, CRUD with in-place moves, and snapshot round-trip.