perf(proofing): run proof_text on a blocking thread

proof_text loaded the dictionary and ran spell/grammar checks
synchronously inside an async command, blocking an async-runtime worker
and holding the engine mutex for the whole call. Move the work to
spawn_blocking so IPC stays responsive under load.

src-tauri/src/proofing.rs

@@ -176,12 +176,13 @@ fn collect_grammar(harper: &mut Harper, text: &str, chars: &[char], issues: &mut
 #[tauri::command]
 pub async fn proof_text(
     app: tauri::AppHandle,
-    state: tauri::State<'_, ProofState>,
     text: String,
     spelling: bool,
     grammar: bool,
 ) -> Result<Vec<Issue>, String> {
-    let mut guard = state.inner().lock().map_err(|e| e.to_string())?;
+    tauri::async_runtime::spawn_blocking(move || -> Result<Vec<Issue>, String> {
+        let state = app.state::<ProofState>();
+        let mut guard = state.lock().map_err(|e| e.to_string())?;
         if guard.is_none() {
             *guard = Some(Engine {
                 speller: None,
@@ -206,6 +207,9 @@ pub async fn proof_text(
             collect_grammar(engine.harper.as_mut().unwrap(), &text, &chars, &mut issues);
         }
         Ok(issues)
+    })
+    .await
+    .map_err(|e| e.to_string())?
 }
 
 #[tauri::command]