Keep audiobook casting moving on slow LLMs

CHANGELOG.md

@@ -9,6 +9,14 @@ Follows [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) · versioned wi
 
 ---
 
+## [1.12.58] — 2026-06-30
+
+### Fixed
+- **Audiobook cast stalls** — per-passage LLM attribution now uses bounded UI timeouts and falls back to deterministic quote detection when a passage or retry half takes too long.
+- **LLM server responsiveness** — audiobook attribution and character-sheet extraction now run blocking LLM HTTP calls in worker threads and honor a clamped `timeout_seconds` request value, so slow local LLM calls no longer block the whole app server event loop.
+
+---
+
 ## [1.12.57] — 2026-06-30
 
 ### Fixed

VERSION

@@ -1 +1 @@
-1.12.57
+1.12.58

routes/conversation.py

@@ -95,6 +95,15 @@ def _rewrite_with_persona_sync(text: str, persona: str, llm_url: str, model: str
     return result
 
 
+def _request_timeout_seconds(value, default: float = 600.0, minimum: float = 5.0, maximum: float = 600.0) -> float:
+    """Clamp caller-provided LLM timeouts so UI recovery cannot hang indefinitely."""
+    try:
+        timeout = float(value)
+    except Exception:
+        timeout = default
+    return max(minimum, min(maximum, timeout))
+
+
 def _resolve_speak_voice(settings: dict, client_id: str, explicit_voice: str) -> str:
     if explicit_voice:
         return explicit_voice
@@ -462,6 +471,7 @@ async def character_sheets(request: Request):
     _settings = _load_settings()
     llm_url: str = (data.get("llm_url") or _settings.get("llm_url") or "http://localhost:11434/v1").rstrip("/")
     model: str = (data.get("model") or _settings.get("llm_model") or "").strip()
+    timeout_seconds = _request_timeout_seconds(data.get("timeout_seconds"), 600.0)
     if not text:
         raise HTTPException(400, "No text provided")
 
@@ -541,9 +551,10 @@ async def character_sheets(request: Request):
     if model:
         payload["model"] = model
     try:
-        resp = requests.post(
+        resp = await asyncio.to_thread(
+            requests.post,
             f"{llm_url}/chat/completions", json=payload,
-            headers={"Authorization": f"Bearer {_settings.get('llm_api_key') or 'sk-dummy-key'}"}, timeout=600,
+            headers={"Authorization": f"Bearer {_settings.get('llm_api_key') or 'sk-dummy-key'}"}, timeout=timeout_seconds,
         )
         resp.raise_for_status()
         _msg = resp.json()["choices"][0]["message"]
@@ -697,6 +708,7 @@ async def attribute_dialogue(request: Request):
     _settings = _load_settings()
     llm_url: str = (data.get("llm_url") or _settings.get("llm_url") or "http://localhost:11434/v1").rstrip("/")
     model: str = (data.get("model") or _settings.get("llm_model") or "").strip()
+    timeout_seconds = _request_timeout_seconds(data.get("timeout_seconds"), 600.0)
     if not text:
         raise HTTPException(400, "No text provided")
 
@@ -762,9 +774,10 @@ async def attribute_dialogue(request: Request):
     if model:
         payload["model"] = model
     try:
-        resp = requests.post(
+        resp = await asyncio.to_thread(
+            requests.post,
             f"{llm_url}/chat/completions", json=payload,
-            headers={"Authorization": f"Bearer {_settings.get('llm_api_key') or 'sk-dummy-key'}"}, timeout=600,
+            headers={"Authorization": f"Bearer {_settings.get('llm_api_key') or 'sk-dummy-key'}"}, timeout=timeout_seconds,
         )
         resp.raise_for_status()
         msg = resp.json()["choices"][0]["message"]

static/index.html

@@ -10,7 +10,7 @@
   <meta name="format-detection" content="telephone=no">
   <meta name="color-scheme" content="light dark">
   <meta name="theme-color" content="#2563EB">
-  <meta name="app-version" content="1.12.57">
+  <meta name="app-version" content="1.12.58">
   <link rel="manifest" href="/manifest.webmanifest">
   <link rel="icon" href="/static/icon.svg" type="image/svg+xml">
   <link rel="apple-touch-icon" href="/static/icon.svg">
@@ -27,7 +27,7 @@
 
   <!-- ── Core styles (local — no CDN dependency for first paint) ────────── -->
   <link rel="stylesheet" href="/static/vendor/mdi/materialdesignicons.min.css">
-  <link rel="stylesheet" href="/static/style.css?v=1.12.57">
+  <link rel="stylesheet" href="/static/style.css?v=1.12.58">
 
 
   <!-- ── Flag icons — non-blocking (loaded async, icons appear after JS) ── -->
@@ -362,7 +362,7 @@ window.toggleNavTree = function(treeId, chevronId) {
 </script>
 
 <!-- loader.js: fetches sections → loads JS modules → removes skeleton -->
-<script src="/static/loader.js?v=1.12.57"></script>
+<script src="/static/loader.js?v=1.12.58"></script>
 
 </body>
 </html>

static/js/audiobook.js

@@ -12,8 +12,44 @@
 // (rehearser.js), splitTextIntoChunks (generation.js), $ / toast (utils.js).
 
 const AUDIOBOOK_CHUNK_CHARS = 3000;   // passage size per LLM attribution call
+const AUDIOBOOK_WARMUP_TIMEOUT_MS = 180000;
+const AUDIOBOOK_ATTRIBUTION_TIMEOUT_MS = 90000;
+const AUDIOBOOK_ATTRIBUTION_RETRY_TIMEOUT_MS = 60000;
+const AUDIOBOOK_RECAST_TIMEOUT_MS = 75000;
 const _audiobook = { running: false, cancel: false };
 
+async function audiobookFetchWithTimeout(url, options = {}, timeoutMs = AUDIOBOOK_ATTRIBUTION_TIMEOUT_MS) {
+  const parentSignal = options.signal;
+  const ac = new AbortController();
+  let timedOut = false;
+  const timer = setTimeout(() => {
+    timedOut = true;
+    ac.abort();
+  }, timeoutMs);
+  const onParentAbort = () => ac.abort(parentSignal?.reason);
+  if (parentSignal) {
+    if (parentSignal.aborted) onParentAbort();
+    else parentSignal.addEventListener('abort', onParentAbort, { once: true });
+  }
+  try {
+    return await fetch(url, { ...options, signal: ac.signal });
+  } catch (err) {
+    if (timedOut) {
+      const timeoutErr = new Error(`Timed out after ${Math.ceil(timeoutMs / 1000)}s`);
+      timeoutErr.name = 'TimeoutError';
+      throw timeoutErr;
+    }
+    throw err;
+  } finally {
+    clearTimeout(timer);
+    if (parentSignal) parentSignal.removeEventListener('abort', onParentAbort);
+  }
+}
+
+function audiobookTimeoutSeconds(timeoutMs) {
+  return Math.max(5, Math.round(timeoutMs / 1000));
+}
+
 // Same hue algorithm as library.js _charHue so avatar colours match across views
 function _abCharHue(name) {
   return Math.abs((name || '?').split('').reduce(function (h, c) { return (h * 31 + c.charCodeAt(0)) % 360; }, 0));
@@ -1582,14 +1618,15 @@ async function audiobookRecastUnknown(overrideUrl, overrideModel) {
   
   view.processing('Waking up LLM model (this may take a few minutes if cold-booting)…');
   try {
-    await fetch('/api/attribute-dialogue', {
+    await audiobookFetchWithTimeout('/api/attribute-dialogue', {
       method: 'POST', headers: { 'Content-Type': 'application/json' }, signal: ac.signal,
       body: JSON.stringify({
         text: 'Wake up.', known_characters: [], recent: '', language,
         llm_url: document.getElementById('ab-cv-llm-url')?.value.trim() || llm_url,
-        model: document.getElementById('ab-cv-llm-select')?.value || model
+        model: document.getElementById('ab-cv-llm-select')?.value || model,
+        timeout_seconds: audiobookTimeoutSeconds(AUDIOBOOK_WARMUP_TIMEOUT_MS)
       })
-    });
+    }, AUDIOBOOK_WARMUP_TIMEOUT_MS + 5000);
   } catch (err) {
     if (err.name === 'AbortError') {
       _audiobook.cancel = true;
@@ -1627,7 +1664,7 @@ async function audiobookRecastUnknown(overrideUrl, overrideModel) {
       
       let data = null;
       try {
-        const r = await fetch('/api/attribute-dialogue', {
+        const r = await audiobookFetchWithTimeout('/api/attribute-dialogue', {
           method: 'POST', headers: { 'Content-Type': 'application/json' },
           signal: ac.signal,
           body: JSON.stringify({
@@ -1636,9 +1673,10 @@ async function audiobookRecastUnknown(overrideUrl, overrideModel) {
             recent: '',
             language,
             llm_url: document.getElementById('ab-cv-llm-url')?.value.trim() || llm_url,
-            model: document.getElementById('ab-cv-llm-select')?.value || model
+            model: document.getElementById('ab-cv-llm-select')?.value || model,
+            timeout_seconds: audiobookTimeoutSeconds(AUDIOBOOK_RECAST_TIMEOUT_MS)
           })
-        });
+        }, AUDIOBOOK_RECAST_TIMEOUT_MS + 5000);
         if (!r.ok) { const e = await r.json().catch(() => ({})); throw new Error(e.detail || e.error || r.statusText); }
         data = await r.json();
       } catch (err) {
@@ -1790,14 +1828,15 @@ async function audiobookCast(overrideUrl, overrideModel) {
 
   view.processing('Waking up LLM model (this may take a few minutes if cold-booting)…');
   try {
-    await fetch('/api/attribute-dialogue', {
+    await audiobookFetchWithTimeout('/api/attribute-dialogue', {
       method: 'POST', headers: { 'Content-Type': 'application/json' }, signal: ac.signal,
       body: JSON.stringify({
         text: 'Wake up.', known_characters: [], recent: '', language,
         llm_url: document.getElementById('ab-cv-llm-url')?.value.trim() || llm_url,
-        model: document.getElementById('ab-cv-llm-select')?.value || model
+        model: document.getElementById('ab-cv-llm-select')?.value || model,
+        timeout_seconds: audiobookTimeoutSeconds(AUDIOBOOK_WARMUP_TIMEOUT_MS)
       })
-    });
+    }, AUDIOBOOK_WARMUP_TIMEOUT_MS + 5000);
   } catch (err) {
     if (err.name === 'AbortError') {
       _audiobook.cancel = true;
@@ -1843,13 +1882,13 @@ async function audiobookCast(overrideUrl, overrideModel) {
       
       view.processing(chunks[i]);
       // ── Helper: run one attribution call and return parsed segments (or null on error) ──
-      const attributeChunk = async (chunkText, recentCtx) => {
+      const attributeChunk = async (chunkText, recentCtx, timeoutMs = AUDIOBOOK_ATTRIBUTION_TIMEOUT_MS) => {
         try {
-          const r = await fetch('/api/attribute-dialogue', {
+          const r = await audiobookFetchWithTimeout('/api/attribute-dialogue', {
             method: 'POST', headers: { 'Content-Type': 'application/json' },
             signal: ac.signal,
-            body: JSON.stringify({ text: chunkText, known_characters: roster.slice(-40), recent: recentCtx, language, llm_url: document.getElementById('ab-cv-llm-url')?.value.trim() || llm_url, model: document.getElementById('ab-cv-llm-select')?.value || model }),
+            body: JSON.stringify({ text: chunkText, known_characters: roster.slice(-40), recent: recentCtx, language, llm_url: document.getElementById('ab-cv-llm-url')?.value.trim() || llm_url, model: document.getElementById('ab-cv-llm-select')?.value || model, timeout_seconds: audiobookTimeoutSeconds(timeoutMs) }),
-          });
+          }, timeoutMs + 5000);
           if (!r.ok) { const e = await r.json().catch(() => ({})); throw new Error(e.detail || e.error || r.statusText); }
           const d = await r.json();
           return Array.isArray(d.segments) ? d.segments : null;
@@ -1872,8 +1911,8 @@ async function audiobookCast(overrideUrl, overrideModel) {
         const splitAt = chunks[i].lastIndexOf(' ', half) || half;
         const chunkA = chunks[i].slice(0, splitAt).trim();
         const chunkB = chunks[i].slice(splitAt).trim();
-        const resA = await attributeChunk(chunkA, recent);
+        const resA = await attributeChunk(chunkA, recent, AUDIOBOOK_ATTRIBUTION_RETRY_TIMEOUT_MS);
-        const resB = await attributeChunk(chunkB, recent);
+        const resB = await attributeChunk(chunkB, recent, AUDIOBOOK_ATTRIBUTION_RETRY_TIMEOUT_MS);
 
         const segsA = (resA && !resA.error) ? resA : null;
         const segsB = (resB && !resB.error) ? resB : null;