proper url decoding, better hop resolution

.vscode/settings.json

@@ -0,0 +1,10 @@
+{
+  "cSpell.words": [
+    "Hotwatch",
+    "actix",
+    "bunbun",
+    "bunbunsearch",
+    "itertools",
+    "opensearchdescription"
+  ]
+}

Cargo.lock

@@ -363,6 +363,8 @@ dependencies = [
  "actix-web 1.0.9 (registry+https://github.com/rust-lang/crates.io-index)",
  "handlebars 2.0.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "hotwatch 0.4.3 (registry+https://github.com/rust-lang/crates.io-index)",
+ "itertools 0.8.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "percent-encoding 2.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "serde 1.0.104 (registry+https://github.com/rust-lang/crates.io-index)",
  "serde_yaml 0.8.11 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
@@ -778,6 +780,14 @@ dependencies = [
  "winreg 0.6.2 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
+[[package]]
+name = "itertools"
+version = "0.8.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "either 1.5.3 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "itoa"
 version = "0.4.4"
@@ -1894,6 +1904,7 @@ dependencies = [
 "checksum inotify-sys 0.1.3 (registry+https://github.com/rust-lang/crates.io-index)" = "e74a1aa87c59aeff6ef2cc2fa62d41bc43f54952f55652656b18a02fd5e356c0"
 "checksum iovec 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)" = "b2b3ea6ff95e175473f8ffe6a7eb7c00d054240321b84c57051175fe3c1e075e"
 "checksum ipconfig 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)" = "aa79fa216fbe60834a9c0737d7fcd30425b32d1c58854663e24d4c4b328ed83f"
+"checksum itertools 0.8.2 (registry+https://github.com/rust-lang/crates.io-index)" = "f56a2d0bc861f9165be4eb3442afd3c236d8a98afd426f65d92324ae1091a484"
 "checksum itoa 0.4.4 (registry+https://github.com/rust-lang/crates.io-index)" = "501266b7edd0174f8530248f87f99c88fbe60ca4ef3dd486835b8d8d53136f7f"
 "checksum kernel32-sys 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)" = "7507624b29483431c0ba2d82aece8ca6cdba9382bff4ddd0f7490560c056098d"
 "checksum language-tags 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)" = "a91d884b6667cd606bb5a69aa0c99ba811a115fc68915e7056ec08a46e93199a"

Cargo.toml

@@ -13,7 +13,9 @@ actix-web = "1.0"
 serde = "1.0"
 serde_yaml = "0.8"
 handlebars = "2.0"
-hotwatch = "0.4.3"
+hotwatch = "0.4"
+percent-encoding = "2.1"
+itertools = "0.8"
 
 [profile.release]
 lto = true

rustfmt.toml

@@ -1,2 +1,3 @@
 tab_spaces = 2
 use_field_init_shorthand = true
+max_width = 80

src/main.rs

@@ -6,19 +6,24 @@ use actix_web::{
 };
 use handlebars::Handlebars;
 use hotwatch::{Event, Hotwatch};
+use itertools::Itertools;
+use percent_encoding::{utf8_percent_encode, AsciiSet, CONTROLS};
 use serde::Deserialize;
 use std::collections::{BTreeMap, HashMap};
-use std::error::Error;
 use std::fmt;
 use std::fs::{File, OpenOptions};
 use std::io::Write;
 use std::sync::{Arc, RwLock};
 use std::time::Duration;
 
+/// https://url.spec.whatwg.org/#fragment-percent-encode-set
+static FRAGMENT_ENCODE_SET: &AsciiSet =
+  &CONTROLS.add(b' ').add(b'"').add(b'<').add(b'>').add(b'`');
 static DEFAULT_CONFIG: &[u8] = include_bytes!("../bunbun.default.toml");
 static CONFIG_FILE: &str = "bunbun.toml";
 
 #[derive(Debug)]
+#[allow(clippy::enum_variant_names)]
 enum BunBunError {
   IoError(std::io::Error),
   ParseError(serde_yaml::Error),
@@ -35,8 +40,8 @@ impl fmt::Display for BunBunError {
   }
 }
 
-impl Error for BunBunError {}
-
+/// Generates a from implementation from the specified type to the provided
+/// bunbun error.
 macro_rules! from_error {
   ($from:ty, $to:ident) => {
     impl From<$from> for BunBunError {
@@ -63,60 +68,65 @@ struct SearchQuery {
 }
 
 #[get("/hop")]
-fn hop(data: Data<Arc<RwLock<State>>>, query: Query<SearchQuery>) -> impl Responder {
+fn hop(
+  data: Data<Arc<RwLock<State>>>,
+  query: Query<SearchQuery>,
+) -> impl Responder {
   let data = data.read().unwrap();
-  let mut raw_args = query.to.split_ascii_whitespace();
-  let command = raw_args.next();
-
-  if command.is_none() {
-    return HttpResponse::NotFound().body("not found");
-  }
-
-  // Reform args into url-safe string (probably want to go thru an actual parser)
-  let mut args = String::new();
-  if let Some(first_arg) = raw_args.next() {
-    args.push_str(first_arg);
-    for arg in raw_args {
-      args.push_str("+");
-      args.push_str(arg);
-    }
-  }
 
+  match resolve_hop(&query.to, &data.routes, &data.default_route) {
+    (Some(path), args) => {
       let mut template_args = HashMap::new();
-  template_args.insert("query", args);
-
-  match data.routes.get(command.unwrap()) {
-    Some(template) => HttpResponse::Found()
-      .header(
-        header::LOCATION,
-        data
-          .renderer
-          .render_template(template, &template_args)
-          .unwrap(),
-      )
-      .finish(),
-    None => match &data.default_route {
-      Some(route) => {
       template_args.insert(
         "query",
-          format!(
-            "{}+{}",
-            command.unwrap(),
-            template_args.get("query").unwrap()
-          ),
+        utf8_percent_encode(&args, FRAGMENT_ENCODE_SET).to_string(),
       );
+
       HttpResponse::Found()
         .header(
           header::LOCATION,
           data
             .renderer
-              .render_template(data.routes.get(route).unwrap(), &template_args)
+            .render_template(&path, &template_args)
             .unwrap(),
         )
         .finish()
     }
-      None => HttpResponse::NotFound().body("not found"),
+    (None, _) => HttpResponse::NotFound().body("not found"),
+  }
+}
+
+/// Attempts to resolve the provided string into its route and its arguments.
+/// If a default route was provided, then this will consider that route before
+/// failing to resolve a route.
+///
+/// The first element in the tuple describes the route, while the second element
+/// returns the remaining arguments. If none remain, an empty string is given.
+fn resolve_hop(
+  query: &str,
+  routes: &BTreeMap<String, String>,
+  default_route: &Option<String>,
+) -> (Option<String>, String) {
+  let mut split_args = query.split_ascii_whitespace().peekable();
+  let command = match split_args.peek() {
+    Some(command) => command,
+    None => return (None, String::new()),
+  };
+
+  match (routes.get(*command), default_route) {
+    // Found a route
+    (Some(resolved), _) => (
+      Some(resolved.clone()),
+      match split_args.next() {
+        // Discard the first result, we found the route using the first arg
+        Some(_) => split_args.join(" "),
+        None => String::new(),
       },
+    ),
+    // Unable to find route, but had a default route
+    (None, Some(route)) => (routes.get(route).cloned(), split_args.join(" ")),
+    // No default route and no match
+    (None, None) => (None, String::new()),
   }
 }
 
@@ -125,7 +135,8 @@ fn index(data: Data<Arc<RwLock<State>>>) -> impl Responder {
   let data = data.read().unwrap();
   let mut template_args = HashMap::new();
   template_args.insert("hostname", &data.public_address);
-  HttpResponse::Ok().body(data.renderer.render("index", &template_args).unwrap())
+  HttpResponse::Ok()
+    .body(data.renderer.render("index", &template_args).unwrap())
 }
 
 #[get("/bunbunsearch.xml")]
@@ -141,14 +152,8 @@ fn opensearch(data: Data<Arc<RwLock<State>>>) -> impl Responder {
     .body(data.renderer.render("opensearch", &template_args).unwrap())
 }
 
-#[derive(Deserialize)]
-struct Config {
-  bind_address: String,
-  public_address: String,
-  default_route: Option<String>,
-  routes: BTreeMap<String, String>,
-}
-
+/// Dynamic variables that either need to be present at runtime, or can be
+/// changed during runtime.
 struct State {
   public_address: String,
   default_route: Option<String>,
@@ -197,6 +202,16 @@ fn main() -> Result<(), BunBunError> {
   Ok(())
 }
 
+#[derive(Deserialize)]
+struct Config {
+  bind_address: String,
+  public_address: String,
+  default_route: Option<String>,
+  routes: BTreeMap<String, String>,
+}
+
+/// Attempts to read the config file. If it doesn't exist, generate one a
+/// default config file before attempting to parse it.
 fn read_config(config_file_path: &str) -> Result<Config, BunBunError> {
   let config_file = match File::open(config_file_path) {
     Ok(file) => file,
@@ -217,11 +232,13 @@ fn read_config(config_file_path: &str) -> Result<Config, BunBunError> {
   Ok(serde_yaml::from_reader(config_file)?)
 }
 
+/// Returns an instance with all pre-generated templates included into the
+/// binary. This allows for users to have a portable binary without needed the
+/// templates at runtime.
 fn compile_templates() -> Handlebars {
   let mut handlebars = Handlebars::new();
-
   macro_rules! register_template {
-    ( $( $template:expr ),* ) => {
+    [ $( $template:expr ),* ] => {
       $(
         handlebars
           .register_template_string(
@@ -233,8 +250,6 @@ fn compile_templates() -> Handlebars {
       )*
     };
   }
-
-  register_template!("index", "list", "opensearch");
-
+  register_template!["index", "list", "opensearch"];
   handlebars
 }