supervertaler 1.9.153__py3-none-any.whl

modules/voice_dictation.py

@@ -0,0 +1,477 @@
+"""
+Voice Dictation Module for Supervertaler
+Uses OpenAI Whisper for multilingual speech recognition
+Supports English, Dutch, and 90+ other languages
+"""
+
+import sounddevice as sd
+import numpy as np
+import tempfile
+import wave
+from pathlib import Path
+from PyQt6.QtWidgets import (
+    QWidget, QVBoxLayout, QHBoxLayout, QPushButton, QLabel,
+    QComboBox, QTextEdit, QGroupBox, QMessageBox, QProgressBar
+)
+from PyQt6.QtCore import Qt, QThread, pyqtSignal
+from PyQt6.QtGui import QFont, QShortcut, QKeySequence
+
+
+class RecordingThread(QThread):
+    """Background thread for audio recording"""
+    finished = pyqtSignal(str)  # Emits path to recorded file
+    error = pyqtSignal(str)
+
+    def __init__(self, duration=30, sample_rate=16000):
+        super().__init__()
+        self.duration = duration
+        self.sample_rate = sample_rate
+        self.is_recording = False
+
+    def run(self):
+        """Record audio in background"""
+        try:
+            self.is_recording = True
+
+            # Record audio
+            recording = sd.rec(
+                int(self.duration * self.sample_rate),
+                samplerate=self.sample_rate,
+                channels=1,
+                dtype='float32'
+            )
+
+            # Wait for recording to complete
+            sd.wait()
+
+            if not self.is_recording:
+                # Recording was stopped early
+                sd.stop()
+
+            # Convert to int16
+            audio_data = np.int16(recording * 32767)
+
+            # Save to temporary WAV file with explicit directory
+            import os
+            temp_dir = tempfile.gettempdir()
+            temp_path = os.path.join(temp_dir, f"voice_recording_{os.getpid()}.wav")
+
+            with wave.open(temp_path, 'wb') as wf:
+                wf.setnchannels(1)
+                wf.setsampwidth(2)  # 16-bit
+                wf.setframerate(self.sample_rate)
+                wf.writeframes(audio_data.tobytes())
+
+            self.finished.emit(temp_path)
+
+        except Exception as e:
+            import traceback
+            self.error.emit(f"{str(e)}\n{traceback.format_exc()}")
+
+    def stop(self):
+        """Stop recording"""
+        self.is_recording = False
+        sd.stop()
+
+
+class TranscriptionThread(QThread):
+    """Background thread for transcription"""
+    finished = pyqtSignal(str)  # Emits transcribed text
+    error = pyqtSignal(str)
+    progress = pyqtSignal(str)  # Progress updates
+
+    def __init__(self, audio_path, model_name="base", language=None):
+        super().__init__()
+        self.audio_path = audio_path
+        self.model_name = model_name
+        self.language = language
+
+    def run(self):
+        """Transcribe audio in background"""
+        try:
+            import os
+
+            try:
+                import whisper  # Local Whisper (optional)
+            except ImportError:
+                self.error.emit(
+                    "Local Whisper is not installed.\n\n"
+                    "To use offline speech recognition, install:\n"
+                    "  pip install supervertaler[local-whisper]\n\n"
+                    "Or switch to 'OpenAI Whisper API' in Supervoice settings."
+                )
+                return
+
+            # Verify file exists
+            if not os.path.exists(self.audio_path):
+                self.error.emit(f"Audio file not found: {self.audio_path}")
+                return
+
+            self.progress.emit("Loading Whisper model...")
+
+            # Load model
+            model = whisper.load_model(self.model_name)
+
+            self.progress.emit("Transcribing audio...")
+
+            # Transcribe
+            if self.language:
+                result = model.transcribe(self.audio_path, language=self.language)
+            else:
+                result = model.transcribe(self.audio_path)
+
+            # Clean up temp file
+            try:
+                Path(self.audio_path).unlink()
+            except:
+                pass
+
+            self.finished.emit(result["text"].strip())
+
+        except Exception as e:
+            import traceback
+            self.error.emit(f"{str(e)}\n\nFull error:\n{traceback.format_exc()}")
+
+
+class VoiceDictationWidget(QWidget):
+    """
+    Voice Dictation Widget using Whisper
+
+    Features:
+    - Push-to-record button
+    - Multilingual support (100+ languages)
+    - Multiple model sizes (tiny, base, small, medium, large)
+    - Copy to clipboard functionality
+    """
+
+    MODELS = {
+        "tiny": "Tiny (fastest, ~1GB RAM)",
+        "base": "Base (good balance, ~1GB RAM)",
+        "small": "Small (better quality, ~2GB RAM)",
+        "medium": "Medium (high quality, ~5GB RAM)",
+        "large": "Large (best quality, ~10GB RAM)"
+    }
+
+    LANGUAGES = {
+        "auto": "Auto-detect",
+        "en": "English",
+        "nl": "Dutch",
+        "de": "German",
+        "fr": "French",
+        "es": "Spanish",
+        "it": "Italian",
+        "pt": "Portuguese",
+        "pl": "Polish",
+        "ru": "Russian",
+        "zh": "Chinese",
+        "ja": "Japanese",
+        "ko": "Korean",
+        # Add more as needed
+    }
+
+    def __init__(self, parent=None):
+        super().__init__(parent)
+        self.recording_thread = None
+        self.transcription_thread = None
+        self.init_ui()
+        self.setup_shortcuts()
+
+    def init_ui(self):
+        """Initialize the user interface"""
+        layout = QVBoxLayout(self)
+        layout.setContentsMargins(10, 10, 10, 10)
+
+        # Title
+        title = QLabel("🎤 Voice Dictation")
+        title.setFont(QFont("Segoe UI", 14, QFont.Weight.Bold))
+        title.setStyleSheet("color: #1976D2;")
+        layout.addWidget(title)
+
+        # Description
+        desc = QLabel(
+            "Record your voice and get instant transcription in 100+ languages.\n"
+            "Powered by OpenAI Whisper."
+        )
+        desc.setStyleSheet("color: #666; padding: 5px; background-color: #E3F2FD; border-radius: 3px;")
+        desc.setWordWrap(True)
+        layout.addWidget(desc)
+
+        # Settings group
+        settings_group = QGroupBox("Settings")
+        settings_layout = QHBoxLayout()
+
+        # Model selection
+        settings_layout.addWidget(QLabel("Model:"))
+        self.model_combo = QComboBox()
+        for key, label in self.MODELS.items():
+            self.model_combo.addItem(label, key)
+        self.model_combo.setCurrentIndex(1)  # Default to "base"
+        settings_layout.addWidget(self.model_combo)
+
+        # Language selection
+        settings_layout.addWidget(QLabel("Language:"))
+        self.language_combo = QComboBox()
+        for key, label in self.LANGUAGES.items():
+            self.language_combo.addItem(label, key)
+        settings_layout.addWidget(self.language_combo)
+
+        settings_group.setLayout(settings_layout)
+        layout.addWidget(settings_group)
+
+        # Recording controls
+        controls_layout = QHBoxLayout()
+
+        self.record_btn = QPushButton("🎙️ Start Recording")
+        self.record_btn.setStyleSheet("""
+            QPushButton {
+                background-color: #2196F3;
+                color: white;
+                border: none;
+                padding: 10px 20px;
+                font-size: 14px;
+                border-radius: 5px;
+            }
+            QPushButton:hover {
+                background-color: #1976D2;
+            }
+            QPushButton:pressed {
+                background-color: #0D47A1;
+            }
+            QPushButton:disabled {
+                background-color: #BDBDBD;
+            }
+        """)
+        self.record_btn.clicked.connect(self.toggle_recording)
+        controls_layout.addWidget(self.record_btn)
+
+        self.copy_btn = QPushButton("📋 Copy to Clipboard")
+        self.copy_btn.setEnabled(False)
+        self.copy_btn.clicked.connect(self.copy_to_clipboard)
+        controls_layout.addWidget(self.copy_btn)
+
+        self.clear_btn = QPushButton("🗑️ Clear")
+        self.clear_btn.clicked.connect(self.clear_text)
+        controls_layout.addWidget(self.clear_btn)
+
+        layout.addLayout(controls_layout)
+
+        # Progress bar
+        self.progress_bar = QProgressBar()
+        self.progress_bar.setVisible(False)
+        self.progress_bar.setTextVisible(False)
+        layout.addWidget(self.progress_bar)
+
+        # Status label
+        self.status_label = QLabel("Ready")
+        self.status_label.setStyleSheet("color: #666; font-style: italic;")
+        layout.addWidget(self.status_label)
+
+        # Transcription output
+        output_group = QGroupBox("Transcription")
+        output_layout = QVBoxLayout()
+
+        self.transcription_text = QTextEdit()
+        self.transcription_text.setPlaceholderText("Transcribed text will appear here...")
+        self.transcription_text.setMinimumHeight(200)
+        output_layout.addWidget(self.transcription_text)
+
+        output_group.setLayout(output_layout)
+        layout.addWidget(output_group)
+
+        # Help text
+        help_text = QLabel(
+            "💡 Tip: Use 'base' model for quick transcription, 'medium' or 'large' for best quality. "
+            "First use downloads the model (~1-10GB depending on size)."
+        )
+        help_text.setWordWrap(True)
+        help_text.setStyleSheet("color: #666; font-size: 10px; padding: 5px;")
+        layout.addWidget(help_text)
+
+        # Keyboard shortcuts info
+        shortcuts_text = QLabel(
+            "⌨️ Shortcuts: F9 = Start/Stop Recording | Esc = Cancel | Ctrl+C = Copy"
+        )
+        shortcuts_text.setWordWrap(True)
+        shortcuts_text.setStyleSheet(
+            "color: #1976D2; font-size: 10px; font-weight: bold; "
+            "padding: 8px; background-color: #E3F2FD; border-radius: 3px; margin-top: 5px;"
+        )
+        layout.addWidget(shortcuts_text)
+
+    def setup_shortcuts(self):
+        """Setup keyboard shortcuts"""
+        # F9 - Start/Stop recording
+        self.shortcut_record = QShortcut(QKeySequence("F9"), self)
+        self.shortcut_record.activated.connect(self.toggle_recording)
+
+        # Esc - Cancel recording
+        self.shortcut_cancel = QShortcut(QKeySequence(Qt.Key.Key_Escape), self)
+        self.shortcut_cancel.activated.connect(self.cancel_recording)
+
+        # Ctrl+C - Copy to clipboard
+        self.shortcut_copy = QShortcut(QKeySequence("Ctrl+C"), self)
+        self.shortcut_copy.activated.connect(self.copy_to_clipboard)
+
+    def cancel_recording(self):
+        """Cancel ongoing recording"""
+        if self.recording_thread and self.recording_thread.is_recording:
+            self.stop_recording()
+            self.status_label.setText("⚠️ Recording cancelled")
+            self.status_label.setStyleSheet("color: #FF9800;")
+
+    def toggle_recording(self):
+        """Start or stop recording"""
+        if self.recording_thread and self.recording_thread.is_recording:
+            # Stop recording
+            self.stop_recording()
+        else:
+            # Start recording
+            self.start_recording()
+
+    def start_recording(self):
+        """Start recording audio"""
+        self.status_label.setText("🔴 Recording... (max 30 seconds)")
+        self.status_label.setStyleSheet("color: #D32F2F; font-weight: bold;")
+        self.record_btn.setText("⏹️ Stop Recording")
+        self.record_btn.setStyleSheet("""
+            QPushButton {
+                background-color: #D32F2F;
+                color: white;
+                border: none;
+                padding: 10px 20px;
+                font-size: 14px;
+                border-radius: 5px;
+            }
+            QPushButton:hover {
+                background-color: #C62828;
+            }
+        """)
+
+        # Disable controls
+        self.model_combo.setEnabled(False)
+        self.language_combo.setEnabled(False)
+
+        # Start recording thread
+        self.recording_thread = RecordingThread()
+        self.recording_thread.finished.connect(self.on_recording_finished)
+        self.recording_thread.error.connect(self.on_recording_error)
+        self.recording_thread.start()
+
+    def stop_recording(self):
+        """Stop recording audio"""
+        if self.recording_thread:
+            self.recording_thread.stop()
+            self.status_label.setText("Processing...")
+            self.status_label.setStyleSheet("color: #FF9800; font-weight: bold;")
+
+    def on_recording_finished(self, audio_path):
+        """Handle recording completion"""
+        # Reset button
+        self.record_btn.setText("🎙️ Start Recording")
+        self.record_btn.setStyleSheet("""
+            QPushButton {
+                background-color: #2196F3;
+                color: white;
+                border: none;
+                padding: 10px 20px;
+                font-size: 14px;
+                border-radius: 5px;
+            }
+            QPushButton:hover {
+                background-color: #1976D2;
+            }
+        """)
+
+        # Start transcription
+        self.transcribe_audio(audio_path)
+
+    def on_recording_error(self, error_msg):
+        """Handle recording error"""
+        self.status_label.setText(f"❌ Recording error: {error_msg}")
+        self.status_label.setStyleSheet("color: #D32F2F;")
+        self.record_btn.setEnabled(True)
+        self.model_combo.setEnabled(True)
+        self.language_combo.setEnabled(True)
+
+    def transcribe_audio(self, audio_path):
+        """Transcribe recorded audio"""
+        model_name = self.model_combo.currentData()
+        language = self.language_combo.currentData()
+        if language == "auto":
+            language = None
+
+        # Show progress
+        self.progress_bar.setVisible(True)
+        self.progress_bar.setRange(0, 0)  # Indeterminate
+        self.record_btn.setEnabled(False)
+
+        # Start transcription thread
+        self.transcription_thread = TranscriptionThread(audio_path, model_name, language)
+        self.transcription_thread.finished.connect(self.on_transcription_finished)
+        self.transcription_thread.error.connect(self.on_transcription_error)
+        self.transcription_thread.progress.connect(self.on_transcription_progress)
+        self.transcription_thread.start()
+
+    def on_transcription_progress(self, message):
+        """Update progress message"""
+        self.status_label.setText(message)
+
+    def on_transcription_finished(self, text):
+        """Handle transcription completion"""
+        self.progress_bar.setVisible(False)
+        self.status_label.setText("✅ Transcription complete")
+        self.status_label.setStyleSheet("color: #388E3C; font-weight: bold;")
+
+        # Add text to output
+        current_text = self.transcription_text.toPlainText()
+        if current_text:
+            self.transcription_text.setPlainText(current_text + "\n\n" + text)
+        else:
+            self.transcription_text.setPlainText(text)
+
+        # Re-enable controls
+        self.record_btn.setEnabled(True)
+        self.model_combo.setEnabled(True)
+        self.language_combo.setEnabled(True)
+        self.copy_btn.setEnabled(True)
+
+    def on_transcription_error(self, error_msg):
+        """Handle transcription error"""
+        self.progress_bar.setVisible(False)
+        self.status_label.setText(f"❌ Transcription error: {error_msg}")
+        self.status_label.setStyleSheet("color: #D32F2F;")
+
+        QMessageBox.critical(self, "Error", f"Transcription failed:\n{error_msg}")
+
+        self.record_btn.setEnabled(True)
+        self.model_combo.setEnabled(True)
+        self.language_combo.setEnabled(True)
+
+    def copy_to_clipboard(self):
+        """Copy transcription to clipboard"""
+        from PyQt6.QtWidgets import QApplication
+        text = self.transcription_text.toPlainText()
+        if text:
+            QApplication.clipboard().setText(text)
+            self.status_label.setText("✅ Copied to clipboard")
+            self.status_label.setStyleSheet("color: #388E3C;")
+
+    def clear_text(self):
+        """Clear transcription text"""
+        self.transcription_text.clear()
+        self.copy_btn.setEnabled(False)
+        self.status_label.setText("Ready")
+        self.status_label.setStyleSheet("color: #666; font-style: italic;")
+
+
+# Standalone test application
+if __name__ == "__main__":
+    import sys
+    from PyQt6.QtWidgets import QApplication
+
+    app = QApplication(sys.argv)
+    window = VoiceDictationWidget()
+    window.setWindowTitle("Voice Dictation - Supervertaler")
+    window.resize(600, 700)
+    window.show()
+    sys.exit(app.exec())