From e207dda2fed450807bb838d7b394b99ca59e9719 Mon Sep 17 00:00:00 2001
From: Astro <astro@spaceboyz.net>
Date: Fri, 3 Mar 2023 02:19:48 +0100
Subject: [PATCH] actor: normalize unicode away for TagRelay

---
 Cargo.lock   |  7 +++++++
 Cargo.toml   |  1 +
 src/actor.rs | 10 ++++++++++
 src/main.rs  |  4 ++--
 src/relay.rs |  4 ++--
 5 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index fa0414e..8946a27 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -237,6 +237,7 @@ dependencies = [
  "axum-extra",
  "axum-macros",
  "chrono",
+ "deunicode",
  "eventsource-stream",
  "futures",
  "http",
@@ -417,6 +418,12 @@ dependencies = [
  "syn",
 ]
 
+[[package]]
+name = "deunicode"
+version = "1.3.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8c1bba4f227a4a53d12b653f50ca7bf10c9119ae2aba56aff9e0338b5c98f36a"
+
 [[package]]
 name = "digest"
 version = "0.10.6"
diff --git a/Cargo.toml b/Cargo.toml
index 98893a2..e4527e5 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -27,3 +27,4 @@ systemd = "0.10"
 metrics = "0.20"
 metrics-util = "0.14"
 metrics-exporter-prometheus = "0.11"
+deunicode = "1.3"
diff --git a/src/actor.rs b/src/actor.rs
index b2e6c63..9a759f2 100644
--- a/src/actor.rs
+++ b/src/actor.rs
@@ -1,4 +1,5 @@
 use std::sync::Arc;
+use deunicode::deunicode;
 use sigh::{PublicKey, Key};
 
 use crate::activitypub;
@@ -9,6 +10,15 @@ pub enum ActorKind {
     InstanceRelay(String),
 }
 
+impl ActorKind {
+    pub fn from_tag(tag: &str) -> Self {
+        let tag = deunicode(tag)
+            .to_lowercase()
+            .replace(char::is_whitespace, "");
+        ActorKind::TagRelay(tag)
+    }
+}
+
 #[derive(Debug, Clone, Hash, PartialEq, Eq, PartialOrd, Ord)]
 pub struct Actor {
     pub host: Arc<String>,
diff --git a/src/main.rs b/src/main.rs
index 43e035e..59b7856 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -97,7 +97,7 @@ async fn get_tag_actor(
     track_request("GET", "actor", "tag");
     let target = actor::Actor {
         host: state.hostname.clone(),
-        kind: actor::ActorKind::TagRelay(tag.to_lowercase()),
+        kind: actor::ActorKind::from_tag(&tag),
     };
     target.as_activitypub(&state.pub_key)
         .into_response()
@@ -123,7 +123,7 @@ async fn post_tag_relay(
 ) -> Response {
     let target = actor::Actor {
         host: state.hostname.clone(),
-        kind: actor::ActorKind::TagRelay(tag.to_lowercase()),
+        kind: actor::ActorKind::from_tag(&tag),
     };
     post_relay(state, endpoint, target).await
 }
diff --git a/src/relay.rs b/src/relay.rs
index d402059..351b209 100644
--- a/src/relay.rs
+++ b/src/relay.rs
@@ -30,7 +30,7 @@ impl Post<'_> {
                 vec![],
             Some(tags) =>
                 tags.iter()
-                .map(|tag| tag.name.to_lowercase())
+                .map(|tag| tag.name.to_string())
                 .collect()
         }
     }
@@ -42,7 +42,7 @@ impl Post<'_> {
             .chain(
                 self.tags()
                     .into_iter()
-                    .map(actor::ActorKind::TagRelay)
+                    .map(|ref s| actor::ActorKind::from_tag(s))
             )
     }