Speakr is a personal, self-hosted web application designed for transcribing audio recordings
Python
Make readable share links opt-in via READABLE_PUBLIC_LINKS env var
Murtaza Nasir9faef1e7da57
diff --git a/src/api/shares.py b/src/api/shares.pyindex e0ae6dac..50f88af5 100644--- a/src/api/shares.py+++ b/src/api/shares.py@@ -22,6 +22,7 @@ ENABLE_PUBLIC_SHARING = os.environ.get('ENABLE_PUBLIC_SHARING', 'true').lower()ENABLE_INTERNAL_SHARING = os.environ.get('ENABLE_INTERNAL_SHARING', 'false').lower() == 'true'SHOW_USERNAMES_IN_UI = os.environ.get('SHOW_USERNAMES_IN_UI', 'false').lower() == 'true'ENABLE_INQUIRE_MODE = os.environ.get('ENABLE_INQUIRE_MODE', 'false').lower() == 'true'+READABLE_PUBLIC_LINKS = os.environ.get('READABLE_PUBLIC_LINKS', 'false').lower() == 'true'# Create blueprintshares_bp = Blueprint('shares', __name__)@@ -122,8 +123,10 @@ def view_shared_recording(public_id):share = Share.query.filter_by(public_id=public_id).first_or_404()recording = share.recording- # Process transcription for server-side rendering- processed_transcript = process_transcription_for_template(recording.transcription)+ # Process transcription for server-side rendering (only if READABLE_PUBLIC_LINKS is enabled)+ processed_transcript = None+ if READABLE_PUBLIC_LINKS:+ processed_transcript = process_transcription_for_template(recording.transcription)# Create a limited dictionary for the public viewrecording_data = {@@ -142,7 +145,7 @@ def view_shared_recording(public_id):'audio_duration': recording.get_audio_duration()}- return render_template('share.html', recording=recording_data, transcript=processed_transcript)+ return render_template('share.html', recording=recording_data, transcript=processed_transcript, readable_mode=READABLE_PUBLIC_LINKS)@shares_bp.route('/share/audio/<string:public_id>')diff --git a/templates/share.html b/templates/share.htmlindex a9730598..787e0e6b 100644--- a/templates/share.html+++ b/templates/share.html@@ -216,7 +216,7 @@<div v-show="activeTab === 'transcription'" class="w-full flex-1 flex flex-col overflow-hidden"><!-- Transcription Controls --><div class="flex items-center justify-between mb-4 flex-shrink-0">- <div v-if="hasDialogue" class="view-mode-toggle">+ <div v-if="readableMode ? hasDialogue : processedTranscription.hasDialogue" class="view-mode-toggle"><button @click="transcriptView = 'simple'":class="['toggle-button', transcriptView === 'simple' ? 'active' : '']"><i class="fas fa-list mr-1"></i>Simple@@ -231,6 +231,8 @@</button></div>+ {% if readable_mode and transcript %}+ <!-- SERVER-RENDERED VERSION (for READABLE_PUBLIC_LINKS mode) --><!-- Speaker Legend (only for bubble view) -->{% if transcript.has_speakers and transcript.speakers %}<div v-show="transcriptView === 'bubble'"@@ -295,6 +297,72 @@<div class="whitespace-pre-wrap cursor-pointer hover:bg-[var(--bg-accent-hover)] p-2 rounded transition-colors">{{ transcript.plain_text }}</div>{% endif %}</div>+ {% else %}+ <!-- VUE-RENDERED VERSION (default) -->+ <!-- Speaker Legend (only for bubble view) -->+ <div v-if="processedTranscription.hasDialogue && processedTranscription.speakers.length > 0 && transcriptView === 'bubble'"+ :class="['speaker-legend', legendExpanded ? 'expanded' : '', 'flex-shrink-0', 'mb-4']">+ <div class="speaker-legend-header" @click="legendExpanded = !legendExpanded">+ <div class="speaker-legend-title">+ <i class="fas fa-users"></i>+ Speakers+ <span class="speaker-count-indicator">(${processedTranscription.speakers.length})</span>+ </div>+ <i class="fas fa-chevron-down speaker-legend-toggle"></i>+ </div>+ <div class="speaker-legend-content">+ <div v-for="(speaker, index) in processedTranscription.speakers"+ :key="index"+ :class="['speaker-legend-item', speaker.color]">+ <span class="speaker-name">${speaker.name}</span>+ </div>+ </div>+ </div>++ <!-- Transcription Content - Scrollable Box -->+ <div class="w-full flex-1 overflow-y-auto transcription-box" @click="handleTranscriptClick">+ <!-- Simple View -->+ <div v-if="transcriptView === 'simple'" class="transcription-simple-view">+ <div v-for="(segment, index) in processedTranscription.simpleSegments"+ :key="segment.startTime || Math.random()"+ :class="['speaker-segment cursor-pointer hover:bg-[var(--bg-accent-hover)] p-2 rounded transition-colors mb-2', { 'active-playing-segment': currentPlayingSegmentIndex === index }]"+ :data-start-time="segment.startTime"+ :data-end-time="segment.endTime"+ :data-segment-index="index"+ style="margin-bottom: 0.5rem !important;">+ <div v-if="segment.showSpeaker" :class="['speaker-tablet', segment.color]">+ ${segment.speaker}+ </div>+ <div class="speaker-text">+ ${segment.sentence}+ </div>+ </div>+ </div>++ <!-- Bubble View -->+ <div v-else-if="transcriptView === 'bubble'" class="transcription-with-speakers">+ <div v-for="(row, rowIndex) in processedTranscription.bubbleRows"+ :key="rowIndex"+ :class="['bubble-row', row.isMe ? 'speaker-me' : '']">+ <div v-for="(bubble, bubbleIndex) in row.bubbles"+ :key="bubble.startTime || Math.random()"+ :class="['speaker-bubble', bubble.color, row.isMe ? 'speaker-me' : '', 'cursor-pointer', { 'active-playing-segment': currentPlayingSegmentIndex === getBubbleGlobalIndex(rowIndex, bubbleIndex) }]"+ :data-start-time="bubble.startTime"+ :data-end-time="bubble.endTime"+ :data-segment-index="getBubbleGlobalIndex(rowIndex, bubbleIndex)">+ <div class="speaker-bubble-content">+ ${bubble.sentence}+ </div>+ </div>+ </div>+ </div>++ <!-- Plain Text View (for non-JSON transcriptions) -->+ <div v-if="!processedTranscription.isJson" class="whitespace-pre-wrap cursor-pointer hover:bg-[var(--bg-accent-hover)] p-2 rounded transition-colors">+ ${processedTranscription.content}+ </div>+ </div>+ {% endif %}</div><!-- Summary View -->@@ -368,11 +436,14 @@const legendExpanded = ref(false);const isDarkMode = ref(document.documentElement.classList.contains('dark'));- // Server-rendered transcript info (avoids duplicate JSON parsing)- const hasDialogue = ref({{ 'true' if transcript.has_speakers else 'false' }});- const plainTextTranscript = {{ transcript.plain_text|tojson|safe }};+ // Readable mode flag from server+ const readableMode = {{ 'true' if readable_mode else 'false' }};const currentPlayingSegmentIndex = ref(-1);+ // Server-rendered transcript info (only used in readable mode)+ const hasDialogue = ref({{ 'true' if transcript and transcript.has_speakers else 'false' }});+ const plainTextTranscript = {{ (transcript.plain_text if transcript else '')|tojson|safe }};+// Custom audio player stateconst audioIsPlaying = ref(false);const audioCurrentTime = ref(0);@@ -522,8 +593,113 @@}}- // Note: Transcript content is now server-rendered for accessibility (Claude, screen readers, etc.)- // The processedTranscription computed property has been removed - content comes from Jinja2 templates+ // processedTranscription computed property (used in Vue mode, not readable mode)+ const processedTranscription = computed(() => {+ if (readableMode) {+ // In readable mode, content is server-rendered, so just return minimal info+ return { hasDialogue: hasDialogue.value, isJson: true, speakers: [], simpleSegments: [], bubbleRows: [] };+ }++ if (!recording.value?.transcription) {+ return { hasDialogue: false, content: '', speakers: [], simpleSegments: [], bubbleRows: [] };+ }++ const transcription = recording.value.transcription;+ let transcriptionData;++ try {+ transcriptionData = JSON.parse(transcription);+ } catch (e) {+ transcriptionData = null;+ }++ if (transcriptionData && Array.isArray(transcriptionData)) {+ const wasDiarized = transcriptionData.some(segment => segment.speaker);++ if (!wasDiarized) {+ const segments = transcriptionData.map(segment => ({+ sentence: segment.sentence,+ startTime: segment.start_time,+ endTime: segment.end_time+ }));+ return {+ hasDialogue: false,+ isJson: true,+ content: segments.map(s => s.sentence).join('\n'),+ simpleSegments: segments,+ speakers: [],+ bubbleRows: []+ };+ }++ const speakers = [...new Set(transcriptionData.map(segment => segment.speaker).filter(Boolean))];+ const speakerColors = {};+ speakers.forEach((speaker, index) => {+ speakerColors[speaker] = `speaker-color-${(index % 8) + 1}`;+ });++ const simpleSegments = transcriptionData.map(segment => ({+ speakerId: segment.speaker,+ speaker: segment.speaker,+ sentence: segment.sentence,+ startTime: segment.start_time || segment.startTime,+ endTime: segment.end_time || segment.endTime,+ color: speakerColors[segment.speaker] || 'speaker-color-1'+ }));++ const processedSimpleSegments = [];+ let lastSpeaker = null;+ simpleSegments.forEach(segment => {+ processedSimpleSegments.push({+ ...segment,+ showSpeaker: segment.speaker !== lastSpeaker+ });+ lastSpeaker = segment.speaker;+ });++ const bubbleRows = [];+ let lastBubbleSpeaker = null;+ simpleSegments.forEach(segment => {+ if (bubbleRows.length === 0 || segment.speaker !== lastBubbleSpeaker) {+ bubbleRows.push({+ speaker: segment.speaker,+ color: segment.color,+ isMe: segment.speaker && (typeof segment.speaker === 'string') && segment.speaker.toLowerCase().includes('me'),+ bubbles: []+ });+ lastBubbleSpeaker = segment.speaker;+ }+ bubbleRows[bubbleRows.length - 1].bubbles.push({+ sentence: segment.sentence,+ startTime: segment.startTime || segment.start_time,+ endTime: segment.endTime || segment.end_time,+ color: segment.color+ });+ });++ return {+ hasDialogue: true,+ isJson: true,+ segments: simpleSegments,+ simpleSegments: processedSimpleSegments,+ bubbleRows: bubbleRows,+ speakers: speakers.map(speaker => ({+ name: speaker,+ color: speakerColors[speaker]+ }))+ };+ }+ return { hasDialogue: false, content: transcription, speakers: [], simpleSegments: [], bubbleRows: [] };+ });++ // Helper to get global index for bubble view highlighting+ function getBubbleGlobalIndex(rowIndex, bubbleIndex) {+ let globalIndex = 0;+ for (let i = 0; i < rowIndex; i++) {+ globalIndex += processedTranscription.value.bubbleRows[i]?.bubbles?.length || 0;+ }+ return globalIndex + bubbleIndex;+ }function seekAudio(startTime) {if (recording.value.audio_deleted_at) return;@@ -546,8 +722,20 @@function copyTranscript(event) {const button = event?.currentTarget;- // Use server-provided plain text (already formatted with speaker labels if diarized)- const textToCopy = plainTextTranscript || recording.value.transcription;+ let textToCopy = '';+ if (readableMode) {+ // Use server-provided plain text (already formatted with speaker labels if diarized)+ textToCopy = plainTextTranscript || recording.value.transcription;+ } else {+ // Use Vue-computed data+ if (processedTranscription.value.isJson && processedTranscription.value.simpleSegments) {+ textToCopy = processedTranscription.value.simpleSegments.map(s =>+ s.speaker ? `[${s.speaker}]: ${s.sentence}` : s.sentence+ ).join('\n');+ } else {+ textToCopy = recording.value.transcription;+ }+ }navigator.clipboard.writeText(textToCopy).then(() => {animateCopyButton(button);showToast('Transcription copied to clipboard!');@@ -634,7 +822,11 @@activeTab,transcriptView,audioPlayer,+ readableMode,hasDialogue,+ processedTranscription,+ currentPlayingSegmentIndex,+ getBubbleGlobalIndex,legendExpanded,isDarkMode,seekAudio,