Added Wayback Machine Search scraper

Cargo.toml

@@ -4,3 +4,8 @@ version = "0.1.0"
 edition = "2021"
 
 [dependencies]
+rayon = "1.10.0"
+reqwest = "0.12.7"
+serde = { version = "1.0.210", features = ["derive"] }
+serde_json = "1.0.128"
+tokio = { version = "1.40.0", features = ["macros", "rt-multi-thread"] }

flake.nix

@@ -13,6 +13,8 @@
       devShells.${system}.default = pkgs.mkShell {
 
         buildInputs = with pkgs; [
+          openssl
+
           clippy
           rustfmt
 

src/main.rs

@@ -1,3 +1,14 @@
-fn main() {
-    println!("Hello, world!");
+mod wayback_machine;
+use wayback_machine::Search;
+
+#[tokio::main(flavor = "multi_thread", worker_threads = 10)]
+async fn main() -> Result<(), Box<dyn std::error::Error>> {
+    let records = Search::search("fermimn.gov.it").await;
+
+    for row in &records {
+        println!("{:#?}", row.get_wayback_machine_url());
+    }
+    println!("Length: {:#?}", records.len());
+
+    Ok(())
 }

src/wayback_machine/mod.rs

@@ -0,0 +1,53 @@
+use rayon::prelude::*;
+use reqwest::Client;
+use std::collections::HashMap;
+
+#[derive(Debug, serde::Deserialize, serde::Serialize)]
+pub struct Search {
+    urlkey: String,
+    timestamp: String,
+    original: String,
+    mimetype: String,
+    statuscode: String,
+    digest: String,
+    length: String,
+}
+
+impl Search {
+    pub async fn search(url: &str) -> Vec<Self> {
+        let url = format!("https://web.archive.org/cdx/search/cdx?output=json&url={url}");
+
+        let response = Client::new()
+            .get(url)
+            .send()
+            .await
+            .unwrap()
+            .text()
+            .await
+            .unwrap();
+
+        let mut records: Vec<Vec<&str>> = serde_json::from_str(&response).unwrap();
+        let headers = records.remove(0);
+
+        records
+            .par_iter()
+            .map(|x| headers.iter().zip(x).collect::<HashMap<_, _>>())
+            .map(|x| Self {
+                urlkey: (**x.get(&"urlkey").unwrap()).to_string(),
+                timestamp: (**x.get(&"timestamp").unwrap()).to_string(),
+                original: (**x.get(&"original").unwrap()).to_string(),
+                mimetype: (**x.get(&"mimetype").unwrap()).to_string(),
+                statuscode: (**x.get(&"statuscode").unwrap()).to_string(),
+                digest: (**x.get(&"digest").unwrap()).to_string(),
+                length: (**x.get(&"length").unwrap()).to_string(),
+            })
+            .collect()
+    }
+
+    pub fn get_wayback_machine_url(&self) -> String {
+        format!(
+            "https://web.archive.org/web/{}/{}",
+            self.timestamp, self.original
+        )
+    }
+}