diff --git a/CHANGELOG.md b/CHANGELOG.md index 9e7e327..88f701e 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -9,6 +9,20 @@ Follows [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) · versioned wi --- +## [1.12.2] — 2026-06-27 + +### Added +- **PDF Search**: search input in the Read Aloud toolbar — press Enter to jump to the first page containing the term, Shift+Enter to go back. +- **Noise gate slider**: adjustable minimum-amplitude threshold in the Conversation input bar (the blue marker on the level meter shows the current gate). Short noise spikes below the gate or bursts shorter than 300 ms are silenced before STT. Persists across sessions. +- **Conversation stats panel** now has a collapse button (→) to hide the latency sidebar and give more chat space; a floating icon button restores it. + +### Changed +- **Sidebar tooltip** now extracts text correctly for all item types (sub-items that had no `.nav-label` span showed nothing before). +- **Active section** in the sidebar now gets a blue background highlight + right-border accent (`.nav-tree-item.active` was previously unstyled, making it impossible to tell which section you were in). +- **Casting audiobook**: LLM prompt now explicitly handles `?«` and `!«` as valid German quote endings, and instructs the model to treat an unclosed `»` at end-of-passage as dialogue. Deterministic fallback also handles the unclosed-quote edge case. + +--- + ## [1.12.1] — 2026-06-27 ### Changed diff --git a/static/index.html b/static/index.html index 73afba6..f4a987e 100644 --- a/static/index.html +++ b/static/index.html @@ -26,7 +26,7 @@ - + @@ -336,7 +336,7 @@ - + diff --git a/static/js/audiobook.js b/static/js/audiobook.js index dcda0f8..956c477 100644 --- a/static/js/audiobook.js +++ b/static/js/audiobook.js @@ -70,8 +70,19 @@ function audiobookSplitByQuotes(text) { } last = sp.end; } - const tail = text.slice(last); - if (tail.trim()) out.push({ speaker: 'Narrator', type: 'narration', text: tail.trim(), emotion: '' }); + const tail = text.slice(last).trimEnd(); + if (tail) { + // If the tail starts with an unclosed opening quote (PDF line-break cut it off), + // treat the quoted portion as dialogue rather than narrator. + const unclosed = tail.match(/^([\s\S]*?)(»)([\s\S]+)$/); + if (unclosed && !tail.includes('«')) { + if (unclosed[1].trim()) out.push({ speaker: 'Narrator', type: 'narration', text: unclosed[1].trim(), emotion: '' }); + const speaker = audiobookGuessSpeaker('', unclosed[1]) || 'Unknown'; + out.push({ speaker, type: 'dialogue', text: unclosed[3].trim(), emotion: '' }); + } else if (tail.trim()) { + out.push({ speaker: 'Narrator', type: 'narration', text: tail.trim(), emotion: '' }); + } + } return audiobookTurnTaking(out); } @@ -269,6 +280,8 @@ FÜR JEDES SEGMENT GIBST DU FOLGENDES AUS: STRIKTE FORMAT- UND TEXTREGELN: - Mische NIEMALS Narration und Dialog im selben Segment! Trenne sie strikt. Wenn ein Zitat durch eine Handlungsanweisung unterbrochen wird (»Nein«, sagte sie, »halt.«), erstelle 3 Segmente: dialogue ("Nein"), narration (", sagte sie, "), dialogue ("halt."). +- »Text?« und »Text!« sind vollständige Dialoge — das ?« bzw. !« schließt das Zitat ab, auch wenn es ungewohnt aussieht. +- Wenn ein Auszug mit einem offenen »-Zitat endet (kein schließendes «), behandle den Text ab » bis Textende als 'dialogue'. - Lasse NIEMALS Wörter aus, fasse nicht zusammen, dupliziere nichts und erfinde keinen Text. Der kombinierte Text all deiner Segmente MUSS den Originaltext exakt und lückenlos Wort für Wort rekonstruieren!`; const globalPrompt = (typeof _appSettings !== 'undefined' && _appSettings.audiobook_prompt) ? _appSettings.audiobook_prompt : AB_DEFAULT_PROMPT; diff --git a/static/js/conversation.js b/static/js/conversation.js index 819f746..8d44266 100644 --- a/static/js/conversation.js +++ b/static/js/conversation.js @@ -194,11 +194,14 @@ $('s-import-voices-file')?.addEventListener('change', async function () { let vadRafId = null; let liveInterimText = ''; let speechRec = null; - const VAD_THRESHOLD = 0.02; // raised to ignore background noise const VAD_MIN_REC_MS = 650; // wait before VAD starts checking (avoids click/noise at start) const VAD_SILENCE_MS = 1000; const INTERRUPT_THRESHOLD = 0.04; // higher than VAD to avoid echo triggering interruption const INTERRUPT_HOLD_MS = 350; // speech must persist this long to interrupt + const MIN_SPEECH_FRAMES_MS = 300; // speech must exceed threshold for at least this long + const gateSlider = () => document.getElementById('conv-gate-slider'); + const vadThreshold = () => { const s = gateSlider(); return s ? (parseInt(s.value, 10) / 1000) : 0.02; }; + let vadSpeechStartMs = 0; // time when speech level first crossed threshold in this recording const liveAgentOn = () => liveAgentToggle ? liveAgentToggle.checked : true; const vadSilenceMs = () => liveAgentOn() ? 650 : VAD_SILENCE_MS; const interruptHoldMs = () => liveAgentOn() ? 120 : INTERRUPT_HOLD_MS; @@ -225,6 +228,23 @@ $('s-import-voices-file')?.addEventListener('change', async function () { let vadHadSpeech = false; // true once RMS crossed threshold during this recording let vadLastVoiceMs = 0; // last timestamp speech was detected (for preview gate) let cancelNextBlob = false; // set by VAD when no speech was detected → skip STT + + // Update gate marker position and persist on slider change + function updateGateMarker() { + const s = gateSlider(); if (!s) return; + const gate = document.getElementById('conv-level-gate'); + if (gate) gate.style.left = s.value + '%'; + try { localStorage.setItem('ttsvc_conv_gate', s.value); } catch (_) {} + } + document.addEventListener('change', function (e) { + if (e.target.id === 'conv-gate-slider') updateGateMarker(); + }); + // Restore persisted gate + (function () { + const s = gateSlider(); if (!s) return; + try { const v = localStorage.getItem('ttsvc_conv_gate'); if (v) s.value = v; } catch (_) {} + updateGateMarker(); + })(); let convCurrentSentenceBubble = null; // assistant bubble to update with current TTS sentence // ── Populate STT backends ──────────────────────────────────────────────── @@ -650,6 +670,7 @@ $('s-import-voices-file')?.addEventListener('change', async function () { previewTranscribing = false; vadHadSpeech = false; vadLastVoiceMs = 0; + vadSpeechStartMs = 0; cancelNextBlob = false; recChunks = []; mediaRecorder = new MediaRecorder(stream); @@ -738,11 +759,17 @@ $('s-import-voices-file')?.addEventListener('change', async function () { let rms = 0; for (const s of vadBuf) rms += s * s; rms = Math.sqrt(rms / vadBuf.length); - if (levelFill) levelFill.style.width = Math.min(100, rms * 5000) + '%'; - if (rms > VAD_THRESHOLD) { - vadLastVoice = Date.now(); - vadLastVoiceMs = Date.now(); - vadHadSpeech = true; + const thr = vadThreshold(); + if (levelFill) levelFill.style.width = Math.min(100, rms / thr * 40) + '%'; + if (rms > thr) { + const now = Date.now(); + if (!vadSpeechStartMs) vadSpeechStartMs = now; + vadLastVoice = now; + vadLastVoiceMs = now; + // Only mark as valid speech once it persists long enough + if (!vadHadSpeech && (now - vadSpeechStartMs) >= MIN_SPEECH_FRAMES_MS) vadHadSpeech = true; + } else { + vadSpeechStartMs = 0; // reset burst timer on silence } const elapsed = Date.now() - recStart; const silence = Date.now() - vadLastVoice; @@ -1178,4 +1205,21 @@ $('s-import-voices-file')?.addEventListener('change', async function () { // Keep TTS backend select in sync after global backend refresh window._ttsRefreshHooks = window._ttsRefreshHooks || []; window._ttsRefreshHooks.push(populateConvTtsBackends); + + // ── Stats panel collapse ───────────────────────────────────────────────── + (function () { + const panel = document.getElementById('conv-stats-panel'); + const colBtn = document.getElementById('conv-stats-collapse-btn'); + const expBtn = document.getElementById('conv-stats-expand-btn'); + const KEY = 'ttsvc_conv_stats_collapsed'; + function setCollapsed(on) { + if (!panel) return; + panel.classList.toggle('collapsed', on); + if (expBtn) expBtn.style.display = on ? 'flex' : 'none'; + try { localStorage.setItem(KEY, on ? '1' : '0'); } catch (_) {} + } + if (colBtn) colBtn.addEventListener('click', function () { setCollapsed(true); }); + if (expBtn) expBtn.addEventListener('click', function () { setCollapsed(false); }); + try { if (localStorage.getItem(KEY) === '1') setCollapsed(true); } catch (_) {} + })(); })(); diff --git a/static/js/reader.js b/static/js/reader.js index 41a964d..e584b42 100644 --- a/static/js/reader.js +++ b/static/js/reader.js @@ -1398,6 +1398,43 @@ $('reader-zoom-two') ?.addEventListener('click', () => readerApplyZoom('two')); $('reader-zoom-in') ?.addEventListener('click', () => { readerState.scale = Math.min(4, readerState.scale * 1.2); readerApplyZoom('custom'); }); $('reader-zoom-out') ?.addEventListener('click', () => { readerState.scale = Math.max(0.2, readerState.scale / 1.2); readerApplyZoom('custom'); }); +// ── Document search ────────────────────────────────────────────────────────── +let _readerSearchIdx = -1; // last hit sentence index +function readerSearchGo(term, direction) { + if (!term) { const st = $('reader-search-status'); if (st) st.textContent = ''; return; } + const lc = term.toLowerCase(); + const sents = readerState.sentences; + if (!sents || !sents.length) { const st = $('reader-search-status'); if (st) st.textContent = 'No text loaded'; return; } + const dir = direction === -1 ? -1 : 1; + const start = _readerSearchIdx < 0 ? 0 : (_readerSearchIdx + dir + sents.length) % sents.length; + for (let i = 0; i < sents.length; i++) { + const idx = (start + i * dir + sents.length * sents.length) % sents.length; + const s = sents[idx]; + const txt = (s.words || []).map(w => w.text).join(' ').toLowerCase(); + if (txt.includes(lc)) { + _readerSearchIdx = idx; + const pageIdx = s.words?.[0]?.page ?? 0; + const pgDiv = readerState.pages[pageIdx]?.pageDiv; + if (pgDiv) pgDiv.scrollIntoView({ behavior: 'smooth', block: 'start' }); + const st = $('reader-search-status'); + const total = sents.filter(s2 => (s2.words||[]).map(w=>w.text).join(' ').toLowerCase().includes(lc)).length; + if (st) st.textContent = total ? `p.${pageIdx + 1}` : ''; + return; + } + } + const st = $('reader-search-status'); + if (st) st.textContent = 'Not found'; +} + +$('reader-search')?.addEventListener('keydown', e => { + if (e.key === 'Enter') { e.preventDefault(); readerSearchGo(e.target.value.trim(), e.shiftKey ? -1 : 1); } + if (e.key === 'Escape') { e.target.value = ''; _readerSearchIdx = -1; const st = $('reader-search-status'); if (st) st.textContent = ''; } +}); +$('reader-search')?.addEventListener('input', e => { + _readerSearchIdx = -1; + if (!e.target.value) { const st = $('reader-search-status'); if (st) st.textContent = ''; } +}); + // Library + export $('reader-save-lib')?.addEventListener('click', readerSaveLibrary); $('reader-export-btn')?.addEventListener('click', () => readerExport($('reader-export-mode')?.value || 'page')); diff --git a/static/nav.js b/static/nav.js index 2de13e6..d47158b 100644 --- a/static/nav.js +++ b/static/nav.js @@ -393,8 +393,14 @@ var item = e.target.closest('.nav-item, .nav-tree-item, .nav-tree-head'); if (!item || item === _railTipActive) return; var label = item.querySelector('.nav-label'); - if (!label) return; - var text = label.textContent.trim(); + var text = ''; + if (label) { + text = label.textContent.trim(); + } else { + var clone = item.cloneNode(true); + clone.querySelectorAll('.mdi, .ntc, .nav-badge, .nav-chevron').forEach(function (el) { el.remove(); }); + text = clone.textContent.replace(/\s+/g, ' ').trim(); + } if (!text) return; var r = item.getBoundingClientRect(); _railTip.textContent = text; diff --git a/static/sections/s-conversation.html b/static/sections/s-conversation.html index 7772de5..884af80 100644 --- a/static/sections/s-conversation.html +++ b/static/sections/s-conversation.html @@ -59,7 +59,14 @@
Ready
-
+
+
+
+
+
+ + +
- -
-
Latency
+ +
+
+ Latency + +
@@ -125,4 +135,8 @@
No turns yet.
+ +
diff --git a/static/sections/s-reader.html b/static/sections/s-reader.html index ce0eceb..6fdcf12 100644 --- a/static/sections/s-reader.html +++ b/static/sections/s-reader.html @@ -69,7 +69,7 @@
- +