From 68cbe3c062294d8abe6cfa0dc146ee06ff28913f Mon Sep 17 00:00:00 2001
From: alemi <me@alemi.dev>
Date: Mon, 2 Oct 2023 07:23:44 +0200
Subject: [PATCH] feat: initial simple implementation

---
 Cargo.toml  |  18 ++++++
 src/main.rs | 170 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 188 insertions(+)
 create mode 100644 Cargo.toml
 create mode 100644 src/main.rs

diff --git a/Cargo.toml b/Cargo.toml
new file mode 100644
index 0000000..63046b8
--- /dev/null
+++ b/Cargo.toml
@@ -0,0 +1,18 @@
+[package]
+name = "akkoma-mapper"
+version = "0.1.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+async-recursion = "1.0.5"
+axum = "0.6.20"
+clap = { version = "4.4.6", features = ["derive"] }
+reqwest = { version = "0.11.20", features = ["json"] }
+serde = { version = "1.0.188", features = ["derive"] }
+serde_json = "1.0.107"
+tokio = { version = "1.32.0", features = ["full"] }
+tracing = "0.1.37"
+tracing-subscriber = "0.3.17"
+
diff --git a/src/main.rs b/src/main.rs
new file mode 100644
index 0000000..958d7fe
--- /dev/null
+++ b/src/main.rs
@@ -0,0 +1,170 @@
+use std::{sync::Arc, collections::{HashMap, HashSet}, time::Duration, net::SocketAddr};
+
+use serde::{Deserialize, Serialize};
+
+use tokio::sync::Mutex;
+
+use clap::Parser;
+
+use axum::{routing::get, extract::Query, Json, Router};
+
+#[derive(Debug, Parser)]
+/// an API crawling akkoma bubble instances network and creating a map
+struct CliArgs {
+	/// start domain for crawl, without proto base
+	domain: String,
+}
+
+#[derive(Debug, Deserialize)]
+struct Params {
+	domain: String
+}
+
+
+#[tokio::main]
+async fn main() {
+	// initialize tracing
+	tracing_subscriber::fmt::init();
+
+	// build our application with a route
+	let app = Router::new()
+		.route("/crawl", get(route_scan_domain));
+
+	// run our app with hyper, listening globally on port 3000
+	let addr = SocketAddr::from(([127, 0, 0, 1], 18811));
+	tracing::debug!("listening on {}", addr);
+	axum::Server::bind(&addr)
+		.serve(app.into_make_service())
+		.await
+		.unwrap();
+}
+
+async fn route_scan_domain(Query(params): Query<Params>) -> Json<MapResult> {
+	tracing::info!("starting new crawl from {}", params.domain);
+
+	let map = Arc::new(Mutex::new(Map {
+		scanned: HashSet::new(),
+		name_to_id: HashMap::new(),
+		counter: 0,
+		nodes: Vec::new(),
+		vertices: Vec::new(),
+	}));
+
+	scan_instance(&params.domain, map.clone()).await;
+
+	let _map = map.lock().await;
+
+	axum::Json(MapResult {
+		nodes: _map.nodes.clone(),
+		vertices: _map.vertices.clone(),
+	})
+}
+
+struct Map {
+	name_to_id: HashMap<String, usize>,
+	scanned: HashSet<String>,
+	nodes: Vec<Node>,
+	vertices: Vec<Vertex>,
+	counter: usize,
+}
+
+impl Map {
+	fn scanned(&mut self, name: &str) -> bool {
+		let out = self.scanned.contains(name);
+		self.scanned.insert(name.to_string());
+		out
+	}
+
+	fn node(&mut self, name: String) -> usize {
+		match self.name_to_id.get(&name) {
+			Some(id) => *id,
+			None => {
+				let id = self.counter;
+				self.name_to_id.insert(name.clone(), id);
+				self.nodes.push(Node { label: name, id });
+				self.counter += 1;
+				id
+			}
+		}
+	}
+
+	fn vertex(&mut self, from_name: String, to_name: String) {
+		let from = self.node(from_name);
+		let to = self.node(to_name);
+		self.vertices.push(Vertex { from, to });
+	}
+}
+
+#[derive(Serialize, Clone, Debug)]
+struct Node {
+	id: usize,
+	label: String,
+}
+
+#[derive(Serialize, Clone, Debug)]
+struct Vertex {
+	from: usize,
+	to: usize,
+}
+
+#[derive(Serialize, Clone, Debug)]
+struct MapResult {
+	nodes: Vec<Node>,
+	vertices: Vec<Vertex>,
+}
+
+#[async_recursion::async_recursion]
+async fn scan_instance(domain: &str, map: Arc<Mutex<Map>>) -> Option<()> {
+	if map.lock().await.scanned(domain) { return None };
+
+	tracing::debug!("scanning instance {}", domain);
+	let response = match instance_metadata(domain).await {
+		Ok(r) => r,
+		Err(e) => {
+			tracing::warn!("could not fetch metadata for {}: {}", domain, e);
+			return None
+		}
+	};
+
+	let node_name : String = response
+		.get("metadata")?
+		.get("nodeName")?
+		.as_str()?
+		.to_string();
+
+	tracing::info!("adding instance {}", node_name);
+
+	map.lock().await.node(domain.to_string());
+
+	let mut tasks = Vec::new();
+
+	for bubble_instance in response
+		.get("metadata")?
+		.get("localBubbleInstances")?
+		.as_array()?
+		.iter()
+		.filter_map(|x| x.as_str().map(|x| x.to_string()))
+	{
+		let _map = map.clone();
+		map.lock().await.vertex(domain.to_string(), bubble_instance.clone());
+		tasks.push(tokio::spawn(async move { scan_instance(&bubble_instance, _map).await; }));
+	}
+
+	for t in tasks {
+		t.await.expect("could not join task");
+	}
+
+	Some(())
+}
+
+async fn instance_metadata(domain: &str) -> reqwest::Result<serde_json::Value> {
+	reqwest::Client::builder()
+		.timeout(Duration::from_secs(5))
+		.build()?
+		.get(format!("https://{}/nodeinfo/2.0.json", domain))
+		.send()
+		.await?
+		.json::<serde_json::Value>()
+		.await
+}
+