@tinytars/frame 0.1.30 → 0.1.31

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@tinytars/frame",
-  "version": "0.1.30",
+  "version": "0.1.31",
   "description": "Domain-neutral Svelte app-shell: session/auth controllers, account chrome, and menu/card/modal primitives built on @tinytars/vault.",
   "license": "MIT",
   "type": "module",

package/speech-registry.svelte.ts

@@ -1,11 +1,12 @@
 // The one shared read-aloud player. It mirrors menu-registry.svelte.ts's singleton shape: playing one
 // thing stops whatever else was playing.
 //
-// Text is spoken one chunk (roughly one sentence) at a time, and each chunk's `end` starts the next.
-// Chromium silently cuts off a single long utterance after ~15s, and chunking is also what makes
-// progress and resume possible. Pause is cancel() plus a remembered index, not
-// speechSynthesis.pause(), which is a no-op or broken on Android Chrome and some Linux voices. Resume
-// restarts the current chunk.
+// Text is spoken one sentence at a time, and each chunk's `end` starts the next. Chromium silently
+// cuts off a single long utterance after ~15s, and chunking is also what makes progress possible.
+// Browser-voice pause is cancel() plus a remembered position, not speechSynthesis.pause(), which is a
+// no-op or broken on Android Chrome and some Linux voices. The position comes from the voice's word
+// `boundary` events, and resume rewinds one word before it for context. A voice that sends no
+// boundaries resumes at the start of its current sentence instead.
 //
 // The engine finishes or fails utterances on its own. cancel()'s `end` also arrives as a separate
 // task, after the call that caused it. So every utterance captures the `generation` it was spoken
@@ -14,7 +15,8 @@
 //
 // An app may configure a SpeechEngine: each chunk is then synthesized to audio (a neural voice) and
 // played through an HTMLAudioElement, the next chunk prefetched while the current one plays. If the
-// engine or playback fails, the rest of that playback falls back to the browser voice.
+// engine or playback fails, the rest of that playback falls back to the browser voice. A playing clip
+// pauses for real and resumes REWIND_SECONDS back.
 export type SpeechStatus = "idle" | "playing" | "paused";
 
 export interface SpeechEngine {
@@ -23,6 +25,7 @@ export interface SpeechEngine {
 }
 
 const MAX_CHUNK = 200;
+const REWIND_SECONDS = 1;
 
 const state = $state({
   id: null as string | null,
@@ -32,6 +35,8 @@ const state = $state({
   index: 0,
 });
 let generation = 0;
+// Character offset within the current chunk of the last word the browser voice reached.
+let reached = 0;
 let engine: SpeechEngine | null = null;
 let voice: string | undefined;
 let neural = false;
@@ -69,7 +74,13 @@ function pack(parts: string[]): string[] {
 export function speechChunks(text: string): string[] {
   const sentences = Array.from(new Intl.Segmenter(undefined, { granularity: "sentence" }).segment(text), (s) => s.segment.trim())
     .filter(Boolean);
-  return pack(sentences.flatMap((s) => (s.length > MAX_CHUNK ? pack(s.split(/\s+/)) : [s])));
+  return sentences.flatMap((s) => (s.length > MAX_CHUNK ? pack(s.split(/\s+/)) : [s]));
+}
+
+function oneWordBefore(text: string, at: number): number {
+  const starts = Array.from(text.matchAll(/\S+/g), (m) => m.index);
+  const current = starts.findLastIndex((i) => i <= at);
+  return starts[Math.max(current - 1, 0)] ?? 0;
 }
 
 function halt() {
@@ -104,14 +115,16 @@ function advance(gen: number) {
   }
 }
 
-function speakCurrent() {
+function speakCurrent(from = 0) {
   const gen = ++generation;
   if (neural) speakNeural(gen);
-  else speakBrowser(gen);
+  else speakBrowser(gen, from);
 }
 
-function speakBrowser(gen: number) {
-  const utterance = new SpeechSynthesisUtterance(state.chunks[state.index]);
+function speakBrowser(gen: number, from = 0) {
+  reached = from;
+  const utterance = new SpeechSynthesisUtterance(state.chunks[state.index].slice(from));
+  utterance.onboundary = (e) => { if (gen === generation) reached = from + e.charIndex; };
   utterance.onend = () => advance(gen);
   utterance.onerror = () => { if (gen === generation) reset(); };
   synth()!.speak(utterance);
@@ -166,13 +179,23 @@ export const speechRegistry = {
   },
   pause(): void {
     if (state.status !== "playing") return;
-    halt();
+    if (neural && audio) {
+      audio.pause();
+      audio.currentTime = Math.max(0, audio.currentTime - REWIND_SECONDS);
+    } else {
+      halt();
+    }
     state.status = "paused";
   },
   resume(): void {
     if (state.status !== "paused" || !isSpeechSupported()) return;
     state.status = "playing";
-    speakCurrent();
+    if (neural && audio) {
+      const gen = generation;
+      audio.play().catch(() => fallBack(gen));
+    } else {
+      speakCurrent(oneWordBefore(state.chunks[state.index], reached));
+    }
   },
   toggle(id: string, text: string, label: string, withVoice?: string): void {
     if (state.id !== id) this.play(id, text, label, withVoice);