feat: voice message recording and send (Phase 8)#551
Draft
torlando-tech wants to merge 12 commits intomainfrom
Draft
feat: voice message recording and send (Phase 8)#551torlando-tech wants to merge 12 commits intomainfrom
torlando-tech wants to merge 12 commits intomainfrom
Conversation
- Add hasAudioAttachment field for audio bubble rendering - Add audioDurationMs field for playback duration display - Add audioCodecId field for decoder selection (e.g., "opus_vm") - Add audioWaveform field for pre-computed amplitude peaks - All fields have defaults so existing callers are unaffected
Wire audioData/audioCodecId through the full IPC stack: - ReticulumProtocol interface: add audioData/audioCodecId params with defaults - MockReticulumProtocol: match updated interface signature - IReticulumService.aidl: add audioData, audioCodecId, audioDataPath params - ServiceReticulumProtocol: handle large audio via temp file (same as image pattern) - ReticulumServiceBinder: pass audio params through to Python callAttr Large audio (>FILE_TRANSFER_THRESHOLD) is written to a temp file and passed as audioDataPath to bypass Android Binder IPC size limits, mirroring the existing imageDataPath pattern.
- Add hasAudioField() to detect LXMF field 7 audio (array format) distinguishing from legacy location data (string format) - Add AudioMetadata data class with codecId, durationMs, waveform - Add extractAudioBytes() for raw audio data extraction from field 7 - Add extractAudioMetadata() to read codec_id and optional waveform peaks - Update toMessageUi() to populate hasAudioAttachment and audio metadata - Preserve fieldsJson when hasAudio is true for lazy byte loading
Send path: - Add audio_data, audio_codec_id, audio_data_path params to send_lxmf_message_with_method() - Large audio files (via audio_data_path) are read from disk and temp file deleted after reading (same pattern as image_data_path) - Audio bytes packed into fields[FIELD_AUDIO] = [codec_id, audio_bytes] Receive path: - Disambiguate field 7 before attempting legacy location parse - Audio data arrives as [codec_id, audio_bytes] (list with string first) - Legacy location data arrives as bytes/string containing JSON - Type check prevents audio from being misinterpreted as location JSON and vice versa, ensuring both features coexist without breakage
- Add extractAudioBytes/extractAudioMetadata imports from MessageMapper - Add audioData, audioCodecId, audioWaveform params to buildFieldsJson() - Pack Field 7 as JSONArray ["codec_id", "hex"] with optional [waveform] - Large audio (>512KB) uses _file_ref pattern to avoid OOM - Declare audio vars in sendMessage() (null placeholders, Phase 8 wires them) - Pass audioData/audioCodecId to sendLxmfMessageWithMethod() in sendMessage() - Add audioData/audioCodecId/audioWaveform params to handleSendSuccess() - handleSendSuccess() stores audio in local fieldsJson via buildFieldsJson() - retryFailedMessage() extracts audio via extractAudioBytes/extractAudioMetadata - retryFailedMessage() passes extracted audio to sendLxmfMessageWithMethod() - All audio params nullable with null defaults, no breaking changes to callers Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- 3/3 plans executed across 2 waves - PROTO-01 through PROTO-05 requirements marked Complete - Phase 7 verified: 5/5 must-haves confirmed against codebase Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…heck - Mirrors MediaPermissionManager pattern with single hasPermission(context) method - Checks ContextCompat.checkSelfPermission for RECORD_AUDIO - Permission launcher to be added in UI layer (Plan 02)
- AudioRecord captures PCM at 24kHz mono, encodes to Opus VOICE_MEDIUM - 2-byte big-endian length-prefixed Opus frames accumulated in ByteArrayOutputStream - 300ms minimum duration: shorter recordings return null from stop() (silent discard) - 30s auto-stop via coroutine delay(MAX_DURATION_MS) - Audio focus AUDIOFOCUS_GAIN_TRANSIENT_EXCLUSIVE acquired before recording, released in stop() - Audio focus released BEFORE 300ms discard check (covers both send and discard paths) - stop() is suspend fun dispatching AudioRecord.stop()/release() to Dispatchers.IO (ANR safety) - Waveform peak amplitudes captured per frame for visualization - StateFlow<RecordingUiState> for observable recording state - VoiceRecording data class with audioBytes, codecId, durationMs, waveformPeaks
…params - Add VoiceMessageViewModel (HiltViewModel) with startRecording/stopRecording delegating to VoiceMessageRecorder - Add optional voiceRecording parameter to MessagingViewModel.sendMessage() - Wire audioBytes/codecId/waveformPeaks from VoiceRecording to send pipeline - Add voice-only message guard: bypass validateAndSanitizeContent with " " when audio present but no text/image/files (matches image-only pattern)
…n handling - Add VoiceMessageViewModel + recording state to MessagingScreen - Add RECORD_AUDIO permission launcher (requests on first mic press) - Add mic/send button toggle in MessageInputBar: mic when empty, send when text/attachments present - Mic button uses detectTapGestures onPress/tryAwaitRelease for hold gesture - onMicRelease launches coroutine to call suspend stopRecording() and sends VoiceRecording through sendMessage pipeline - Wire hasTextOrAttachments, isRecording, onMicPress, onMicRelease params
Extract helpers from sendMessage/buildFieldsJson/sendLxmfMessageWithMethod to satisfy LongMethod, LongParameterList, and CyclomaticComplexMethod thresholds. Fix ktlint parse error in MessageUi KDoc comment. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…nd bubble rendering - Fix mic button gesture: replace IconButton with Box so pointerInput receives press/release events (IconButton's internal clickable was swallowing them) - Fix recomposition killing gesture: keep mic button in composition during recording so tryAwaitRelease() completes - Add recording indicator: pulsing red dot + duration timer replaces text field during active recording - Add preview bar after recording: play/pause, waveform with progress, discard (trash), and send buttons - Add VoiceMessagePlayer: decodes length-prefixed Opus frames via LXST-kt Opus decoder, plays through AudioTrack MODE_STATIC - Add VoiceMessageBubble: play/pause, waveform, duration in message bubbles for sent/received voice messages - Extract shared WaveformBar composable for preview and bubble reuse - Hide text content for voice-only messages (content is " " for Sideband compatibility) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
torlando-tech
force-pushed
the
feat/phase-8-voice-recording
branch
from
c9fe3ef to
b462e15
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
VoiceMessageRecorderwith Opus encoding (24kHz mono, VOICE_MEDIUM profile) via LXST-ktAudioPermissionManagerfor RECORD_AUDIO permission checksVoiceMessageViewModel(Hilt) wrapping the recorderMessagingScreensendMessage()to acceptVoiceRecordingwith audio data passthroughFeatures
Test plan
🤖 Generated with Claude Code