Complete Phase 4: Comprehensive preferences, localization, and UX polish

- Rename application from MenuWhisper to Tell me with new domain com.fmartingr.tellme
- Implement comprehensive preferences window with 6 tabs (General, Models, Text Insertion, Interface, Advanced, Permissions)
- Add full English/Spanish localization for all UI elements
- Create functional onboarding flow with model download capability
- Implement preview dialog for transcription editing
- Add settings export/import functionality
- Fix HUD content display issues and add comprehensive permission checking
- Enhance build scripts and app bundle creation for proper localization support
This commit is contained in:
Felipe M 2025-09-19 13:55:46 +02:00
parent 7ba5895406
commit 54c3b65d4a
Signed by: fmartingr
GPG key ID: CCFBC5637D4000A8
25 changed files with 3086 additions and 235 deletions

View file

@ -1,10 +1,11 @@
import SwiftUI
import CoreUtils
import MenuWhisperAudio
import TellMeAudio
import CorePermissions
import CoreSTT
import CoreModels
import CoreInjection
import CoreSettings
import AVFoundation
public class AppController: ObservableObject {
@ -14,16 +15,21 @@ public class AppController: ObservableObject {
private let hotkeyManager = HotkeyManager()
private let audioEngine = AudioEngine()
private let permissionManager = PermissionManager()
private let soundManager = SoundManager()
private let soundManager: SoundManager
private let textInjector: TextInjector
// Settings
public let settings = Settings()
// STT components
public let whisperEngine = WhisperCPPEngine(numThreads: 4, useGPU: true)
public let whisperEngine: WhisperCPPEngine
public var modelManager: ModelManager!
// UI components
private var hudWindow: HUDWindow?
private var preferencesWindow: PreferencesWindowController?
private var onboardingWindow: OnboardingWindowController?
private var helpWindow: HelpWindowController?
private var statusItem: NSStatusItem?
// State management
@ -32,9 +38,13 @@ public class AppController: ObservableObject {
// Dictation timer
private var dictationTimer: Timer?
private let maxDictationDuration: TimeInterval = 600 // 10 minutes default
public init() {
whisperEngine = WhisperCPPEngine(
numThreads: settings.processingThreads,
useGPU: true
)
soundManager = SoundManager(settings: settings)
textInjector = TextInjector(permissionManager: permissionManager)
setupDelegates()
setupNotifications()
@ -89,6 +99,10 @@ public class AppController: ObservableObject {
public func start() {
logger.info("Starting app controller")
// Debug: Check localization
let testString = NSLocalizedString("preferences.title", comment: "Test")
logger.info("Localization test - preferences.title: '\(testString)'")
// Setup status item menu on main actor
Task { @MainActor in
setupStatusItemMenu()
@ -110,14 +124,14 @@ public class AppController: ObservableObject {
@MainActor
private func setupStatusItemMenu() {
statusItem = NSStatusBar.system.statusItem(withLength: NSStatusItem.squareLength)
statusItem?.button?.image = NSImage(systemSymbolName: "mic", accessibilityDescription: "MenuWhisper")
statusItem?.button?.image = NSImage(systemSymbolName: "mic", accessibilityDescription: "Tell me")
statusItem?.button?.imagePosition = .imageOnly
let menu = NSMenu()
// Status item
let statusMenuItem = NSMenuItem()
statusMenuItem.title = "MenuWhisper"
statusMenuItem.title = NSLocalizedString("app.name", comment: "App name")
statusMenuItem.isEnabled = false
menu.addItem(statusMenuItem)
@ -125,19 +139,33 @@ public class AppController: ObservableObject {
// Model status
let modelMenuItem = NSMenuItem()
modelMenuItem.title = "Loading model..."
modelMenuItem.title = NSLocalizedString("menubar.loading_model", comment: "Loading model...")
modelMenuItem.isEnabled = false
menu.addItem(modelMenuItem)
menu.addItem(NSMenuItem.separator())
// Preferences
let preferencesMenuItem = NSMenuItem(title: "Preferences...", action: #selector(openPreferences), keyEquivalent: ",")
let preferencesMenuItem = NSMenuItem(title: NSLocalizedString("menubar.preferences", comment: "Preferences..."), action: #selector(openPreferences), keyEquivalent: ",")
preferencesMenuItem.target = self
menu.addItem(preferencesMenuItem)
// Help
let helpMenuItem = NSMenuItem(title: NSLocalizedString("menubar.help", comment: "Help"), action: #selector(openHelp), keyEquivalent: "?")
helpMenuItem.target = self
menu.addItem(helpMenuItem)
// Debug: Reset Onboarding (only in development)
#if DEBUG
let resetOnboardingMenuItem = NSMenuItem(title: NSLocalizedString("menubar.reset_onboarding", comment: "Reset Onboarding"), action: #selector(resetOnboarding), keyEquivalent: "")
resetOnboardingMenuItem.target = self
menu.addItem(resetOnboardingMenuItem)
#endif
menu.addItem(NSMenuItem.separator())
// Quit
let quitMenuItem = NSMenuItem(title: "Quit MenuWhisper", action: #selector(quitApp), keyEquivalent: "q")
let quitMenuItem = NSMenuItem(title: NSLocalizedString("menubar.quit", comment: "Quit Tell me"), action: #selector(quitApp), keyEquivalent: "q")
quitMenuItem.target = self
menu.addItem(quitMenuItem)
@ -153,6 +181,24 @@ public class AppController: ObservableObject {
}
}
@objc private func openHelp() {
Task { @MainActor in
showHelp()
}
}
#if DEBUG
@objc private func resetOnboarding() {
UserDefaults.standard.removeObject(forKey: "hasShownPermissionOnboarding")
logger.info("Onboarding status reset - will show on next startup or permission check")
// Optionally show onboarding immediately
Task { @MainActor in
showPermissionOnboarding()
}
}
#endif
@objc private func quitApp() {
NSApplication.shared.terminate(nil)
}
@ -165,11 +211,11 @@ public class AppController: ObservableObject {
let modelMenuItem = menu.items[2] // Model status item
if let activeModel = modelManager?.activeModel, whisperEngine.isModelLoaded() {
modelMenuItem.title = "Model: \(activeModel.name)"
modelMenuItem.title = String(format: NSLocalizedString("menubar.model_status", comment: "Model status"), activeModel.name)
} else if modelManager?.activeModel != nil {
modelMenuItem.title = "Model: Loading..."
modelMenuItem.title = NSLocalizedString("menubar.model_loading", comment: "Model loading")
} else {
modelMenuItem.title = "No model - click Preferences"
modelMenuItem.title = NSLocalizedString("menubar.no_model", comment: "No model")
}
}
@ -222,21 +268,28 @@ public class AppController: ObservableObject {
@MainActor
private func showPermissionOnboarding() {
let alert = NSAlert()
alert.messageText = "Welcome to MenuWhisper"
alert.informativeText = "MenuWhisper needs some permissions to work properly:\n\n• Microphone: To capture your speech\n• Accessibility: To insert transcribed text\n• Input Monitoring: To send keyboard events\n\nWould you like to set up permissions now?"
alert.alertStyle = .informational
alert.addButton(withTitle: "Set Up Permissions")
alert.addButton(withTitle: "Later")
let response = alert.runModal()
// Mark that we've shown the onboarding
UserDefaults.standard.set(true, forKey: "hasShownPermissionOnboarding")
if response == .alertFirstButtonReturn {
showPreferences(initialTab: 1) // Open Permissions tab
guard let modelManager = modelManager else {
logger.error("ModelManager not initialized for onboarding")
return
}
onboardingWindow = OnboardingWindowController(
permissionManager: permissionManager,
modelManager: modelManager,
whisperEngine: whisperEngine,
onComplete: { [weak self] in
// Mark that we've shown the onboarding
UserDefaults.standard.set(true, forKey: "hasShownPermissionOnboarding")
// Clear the reference after a brief delay to allow window to close
DispatchQueue.main.asyncAfter(deadline: .now() + 0.1) {
self?.onboardingWindow = nil
}
}
)
onboardingWindow?.showWindow(nil)
onboardingWindow?.window?.makeKeyAndOrderFront(nil)
NSApp.activate(ignoringOtherApps: true)
}
private func checkMicrophonePermission(completion: @escaping (Bool) -> Void) {
@ -252,12 +305,53 @@ public class AppController: ObservableObject {
cancelDictation()
}
private func checkAllRequiredPermissions() -> Bool {
// Refresh permission status first (they might have been granted since startup)
permissionManager.checkAllPermissions()
// Check microphone permission (critical for audio capture)
if permissionManager.microphoneStatus != .granted {
logger.warning("Microphone permission not granted")
Task { @MainActor in
showMicrophonePermissionAlert()
}
return false
}
// Check accessibility permission (required for text insertion)
if permissionManager.accessibilityStatus != .granted {
logger.warning("Accessibility permission not granted")
Task { @MainActor in
showAccessibilityPermissionAlert()
}
return false
}
// Input monitoring (required for hotkeys and text insertion)
if permissionManager.inputMonitoringStatus != .granted {
logger.warning("Input monitoring permission not granted")
Task { @MainActor in
showInputMonitoringPermissionAlert()
}
return false
}
logger.info("All required permissions granted")
return true
}
private func startListening() {
guard currentState == .idle else {
logger.warning("Cannot start listening from state: \(currentState)")
return
}
// Check all required permissions before starting
let allPermissionsGranted = checkAllRequiredPermissions()
guard allPermissionsGranted else {
return // Permission alerts will be shown by checkAllRequiredPermissions
}
// Check if a model is loaded before starting
guard whisperEngine.isModelLoaded() else {
logger.warning("No model loaded - showing setup alert")
@ -324,7 +418,7 @@ public class AppController: ObservableObject {
}
let startTime = Date()
let transcription = try await whisperEngine.transcribe(audioData: audioData, language: "auto")
let transcription = try await whisperEngine.transcribe(audioData: audioData, language: settings.forcedLanguage ?? "auto")
let duration = Date().timeIntervalSince(startTime)
logger.info("Transcription completed in \(String(format: "%.2f", duration))s: \"\(transcription)\"")
@ -345,10 +439,37 @@ public class AppController: ObservableObject {
private func injectTranscriptionResult(_ text: String) {
logger.info("Attempting to inject transcription result: \(text)")
// Check if preview is enabled
if settings.showPreview {
showPreviewDialog(text: text)
} else {
performTextInjection(text)
}
}
@MainActor
private func showPreviewDialog(text: String) {
let previewDialog = PreviewDialogController(
text: text,
onInsert: { [weak self] editedText in
self?.performTextInjection(editedText)
},
onCancel: { [weak self] in
self?.finishProcessing()
}
)
previewDialog.showWindow(nil)
}
@MainActor
private func performTextInjection(_ text: String) {
do {
// Attempt to inject the text using paste method with fallback enabled
try textInjector.injectText(text, method: .paste, enableFallback: true)
logger.info("Text injection successful")
// Determine injection method from settings
let injectionMethod: InjectionMethod = settings.insertionMethod == .paste ? .paste : .typing
// Attempt to inject the text using the configured method
try textInjector.injectText(text, method: injectionMethod, enableFallback: true)
logger.info("Text injection successful using \(injectionMethod) method")
// Show success and finish processing
finishProcessing()
@ -403,7 +524,7 @@ public class AppController: ObservableObject {
private func startDictationTimer() {
stopDictationTimer() // Clean up any existing timer
dictationTimer = Timer.scheduledTimer(withTimeInterval: maxDictationDuration, repeats: false) { [weak self] _ in
dictationTimer = Timer.scheduledTimer(withTimeInterval: settings.dictationTimeLimit, repeats: false) { [weak self] _ in
self?.logger.info("Dictation timeout reached")
self?.stopListening()
}
@ -416,7 +537,7 @@ public class AppController: ObservableObject {
private func showHUD(state: HUDState) {
if hudWindow == nil {
hudWindow = HUDWindow()
hudWindow = HUDWindow(settings: settings)
}
hudWindow?.show(state: state)
}
@ -445,7 +566,7 @@ public class AppController: ObservableObject {
private func showPermissionRequiredNotice() {
let alert = NSAlert()
alert.messageText = "Permission Required"
alert.informativeText = "MenuWhisper needs Accessibility and Input Monitoring permissions to insert text into other applications.\n\nWould you like to open System Settings to grant these permissions?"
alert.informativeText = "Tell me needs Accessibility and Input Monitoring permissions to insert text into other applications.\n\nWould you like to open System Settings to grant these permissions?"
alert.alertStyle = .warning
alert.addButton(withTitle: "Open System Settings")
alert.addButton(withTitle: "Cancel")
@ -478,6 +599,7 @@ public class AppController: ObservableObject {
modelManager: modelManager,
whisperEngine: whisperEngine,
permissionManager: permissionManager,
settings: settings,
initialTab: initialTab
)
} else {
@ -490,11 +612,67 @@ public class AppController: ObservableObject {
NSApp.activate(ignoringOtherApps: true)
}
@MainActor
public func showHelp() {
if helpWindow == nil {
helpWindow = HelpWindowController()
}
helpWindow?.showWindow(nil)
helpWindow?.window?.makeKeyAndOrderFront(nil)
NSApp.activate(ignoringOtherApps: true)
}
@MainActor
private func showMicrophonePermissionAlert() {
let alert = NSAlert()
alert.messageText = NSLocalizedString("permissions.microphone.title", comment: "Microphone Access Required")
alert.informativeText = NSLocalizedString("permissions.microphone.message", comment: "Microphone permission message")
alert.alertStyle = .warning
alert.addButton(withTitle: NSLocalizedString("permissions.open_settings", comment: "Open System Settings"))
alert.addButton(withTitle: NSLocalizedString("general.cancel", comment: "Cancel"))
let response = alert.runModal()
if response == .alertFirstButtonReturn {
permissionManager.openSystemSettings(for: .microphone)
}
}
@MainActor
private func showAccessibilityPermissionAlert() {
let alert = NSAlert()
alert.messageText = NSLocalizedString("permissions.accessibility.title", comment: "Accessibility Access Required")
alert.informativeText = NSLocalizedString("permissions.accessibility.message", comment: "Accessibility permission message")
alert.alertStyle = .warning
alert.addButton(withTitle: NSLocalizedString("permissions.open_settings", comment: "Open System Settings"))
alert.addButton(withTitle: NSLocalizedString("general.cancel", comment: "Cancel"))
let response = alert.runModal()
if response == .alertFirstButtonReturn {
permissionManager.openSystemSettings(for: .accessibility)
}
}
@MainActor
private func showInputMonitoringPermissionAlert() {
let alert = NSAlert()
alert.messageText = NSLocalizedString("permissions.input_monitoring.title", comment: "Input Monitoring Required")
alert.informativeText = NSLocalizedString("permissions.input_monitoring.message", comment: "Input monitoring permission message")
alert.alertStyle = .warning
alert.addButton(withTitle: NSLocalizedString("permissions.open_settings", comment: "Open System Settings"))
alert.addButton(withTitle: NSLocalizedString("general.cancel", comment: "Cancel"))
let response = alert.runModal()
if response == .alertFirstButtonReturn {
permissionManager.openSystemSettings(for: .inputMonitoring)
}
}
@MainActor
private func showModelSetupAlert() {
let alert = NSAlert()
alert.messageText = "No Speech Recognition Model"
alert.informativeText = "You need to download and select a speech recognition model before using MenuWhisper.\n\nWould you like to open Preferences to download a model?"
alert.informativeText = "You need to download and select a speech recognition model before using Tell me.\n\nWould you like to open Preferences to download a model?"
alert.alertStyle = .informational
alert.addButton(withTitle: "Open Preferences")
alert.addButton(withTitle: "Cancel")
@ -511,6 +689,8 @@ public class AppController: ObservableObject {
audioEngine.stopCapture()
hotkeyManager.disableHotkey()
preferencesWindow?.close()
onboardingWindow?.close()
helpWindow?.close()
NotificationCenter.default.removeObserver(self)
}
}