isu-admin/hearbit-ai-app
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 2 Wiki Activity

Release 1.1.0: Add Import Audio Files feature

Browse Source 
- New Import tab with drag-and-drop support for audio files
- Support for 8 formats: MP3, MP4, WAV, M4A, FLAC, OGG, AAC, WMA
- File metadata display (duration, size, format)
- Editable meeting titles
- Progress tracking with visual indicators
- Smart template selection
- Auto-navigation to Transcription view
- Updated README with BlackHole requirement and Teams config
- Added get_audio_metadata Rust command
- Version bump to 1.1.0
This commit is contained in:
michael.borak
2026-01-21 09:08:56 +01:00
parent 79f509951c
commit a06e473e85
12 changed files with 1041 additions and 171 deletions
Download Patch File Download Diff File Expand all files Collapse all files

137
src-tauri/src/lib.rs
View File

@@ -1,4 +1,9 @@
use tauri::{AppHandle, Manager, State, Emitter};
use tauri::{
AppHandle, Manager, State, Emitter,
menu::{Menu, MenuItem},
tray::{TrayIconBuilder, TrayIconEvent},
WindowEvent
};
use std::sync::{Arc, Mutex};
use std::process::Command;
use cpal::traits::{DeviceTrait, HostTrait, StreamTrait};
@@ -65,7 +70,7 @@ fn get_input_devices() -> Result<Vec<AudioDevice>, String> {
#[tauri::command]
fn start_recording(app: AppHandle, state: State<'_, AppState>, device_id: String, save_path: Option<String>, custom_filename: Option<String>) -> Result<(), String> {
fn start_recording(app: AppHandle, state: State<'_, AppState>, device_id: String, save_path: Option<String>, custom_filename: Option<String>, wait_for_speech: Option<bool>) -> Result<(), String> {
emit_log(&app, "INFO", &format!("Starting recording on device: {}", device_id));
let host = cpal::default_host();
@@ -77,16 +82,17 @@ fn start_recording(app: AppHandle, state: State<'_, AppState>, device_id: String
.or_else(|| host.default_input_device())
.ok_or("No input device found")?;
let config = device.default_input_config().map_err(|e| e.to_string())?;
// VAD requires 16Hz or 8kHz, typically. Silero likes 16k.
// We might need to resample or just check if the device supports it.
// For MVP VAD, we will try to stick to standard rates.
// Actually, simple energy VAD is easier to start with if Silero is too heavy or requires ONNX runtime.
// Let's check the crate docs or usage first.
// Wait, the user wants to IGNORE music. Energy VAD will fail on music.
// voice_activity_detector crate usually uses Silero or similar.
// Select the configuration with the MAXIMUM number of channels
// This is crucial for "Hearbit Audio" (Aggregate) which lists 3 channels but might default to 2.
// We want the raw 3 channels to separate Mic (Ch0) from System (Ch1+2).
let supported_configs = device.supported_input_configs().map_err(|e| e.to_string())?;
let config = supported_configs
.max_by_key(|c| c.channels())
.map(|c| c.with_max_sample_rate())
.ok_or("No supported input configurations found")?;
emit_log(&app, "INFO", &format!("Selected Audio Config: {} Channels, {} Hz", config.channels(), config.sample_rate()));
let spec = hound::WavSpec {
channels: config.channels(),
sample_rate: config.sample_rate(),
@@ -122,7 +128,12 @@ fn start_recording(app: AppHandle, state: State<'_, AppState>, device_id: String
// Initialize AudioProcessor (VAD)
// We pass the writer to it.
let processor = AudioProcessor::new(config.sample_rate(), writer.clone(), app.clone())
let should_wait = wait_for_speech.unwrap_or(false);
if should_wait {
emit_log(&app, "INFO", "Recording started in WAITING mode (buffer-only until speech).");
}
let processor = AudioProcessor::new(config.sample_rate(), config.channels(), writer.clone(), app.clone(), should_wait)
.map_err(|e| format!("Failed to create AudioProcessor: {}", e))?;
// Wrap processor in Arc<Mutex> so we can share/move it into callback
@@ -560,6 +571,62 @@ async fn summarize_text(app: AppHandle, text: String, api_key: String, product_i
}
}
#[derive(serde::Serialize)]
struct AudioMetadata {
duration: f64,
size: u64,
format: String,
}
#[tauri::command]
fn get_audio_metadata(app: AppHandle, file_path: String) -> Result<AudioMetadata, String> {
emit_log(&app, "INFO", &format!("Getting metadata for: {}", file_path));
// Get file size
let metadata = std::fs::metadata(&file_path).map_err(|e| e.to_string())?;
let size = metadata.len();
// Extract format from extension
let path = std::path::Path::new(&file_path);
let format = path.extension()
.and_then(|e| e.to_str())
.unwrap_or("unknown")
.to_string();
// Get duration using ffprobe (requires ffmpeg to be installed)
let duration = match Command::new("ffprobe")
.args([
"-v", "error",
"-show_entries", "format=duration",
"-of", "default=noprint_wrappers=1:nokey=1",
&file_path
])
.output()
{
Ok(output) => {
if output.status.success() {
let duration_str = String::from_utf8_lossy(&output.stdout);
duration_str.trim().parse::<f64>().unwrap_or(0.0)
} else {
emit_log(&app, "WARN", "ffprobe failed, duration = 0");
0.0
}
},
Err(_) => {
emit_log(&app, "WARN", "ffprobe not found, duration = 0");
0.0
}
};
emit_log(&app, "SUCCESS", &format!("Metadata: {}s, {} bytes", duration, size));
Ok(AudioMetadata {
duration,
size,
format,
})
}
#[tauri::command]
fn open_audio_midi_setup() -> Result<(), String> {
Command::new("open")
@@ -640,6 +707,49 @@ async fn read_log_file(app: AppHandle) -> Result<String, String> {
#[cfg_attr(mobile, tauri::mobile_entry_point)]
pub fn run() {
tauri::Builder::default()
.setup(|app| {
// Setup Tray Icon
let quit_i = MenuItem::with_id(app, "quit", "Quit Hearbit AI", true, None::<&str>).unwrap();
let show_i = MenuItem::with_id(app, "show", "Show Window", true, None::<&str>).unwrap();
let menu = Menu::with_items(app, &[&show_i, &quit_i]).unwrap();
let _tray = TrayIconBuilder::new()
.icon(app.default_window_icon().unwrap().clone())
.menu(&menu)
.show_menu_on_left_click(true)
.on_menu_event(|app, event| {
match event.id.as_ref() {
"quit" => app.exit(0),
"show" => {
if let Some(window) = app.get_webview_window("main") {
let _ = window.show();
let _ = window.set_focus();
}
}
_ => {}
}
})
.on_tray_icon_event(|tray, event| {
if let TrayIconEvent::Click { .. } = event {
let app = tray.app_handle();
if let Some(window) = app.get_webview_window("main") {
let _ = window.show();
let _ = window.set_focus();
}
}
})
.build(app)?;
Ok(())
})
.on_window_event(|window, event| {
if let WindowEvent::CloseRequested { api, .. } = event {
// Prevent window from closing, just hide it
window.hide().unwrap();
api.prevent_close();
}
})
.plugin(tauri_plugin_shell::init())
.plugin(tauri_plugin_log::Builder::default()
.targets([
tauri_plugin_log::Target::new(tauri_plugin_log::TargetKind::Stdout),
@@ -670,6 +780,7 @@ pub fn run() {
auth::get_calendar_events,
save_text_file,
read_log_file,
get_audio_metadata,
email::send_smtp_email
])
.run(tauri::generate_context!())